Merge branch 'linus' into perf/core, to pick up fixes

Ingo Molnar committed 9 years ago

Commit 2b95bd72 parents ffa86c2 69eea5a586 files · +7737 −3452Browse files

Unified · Split

586 files

Documentation/admin-guide/kernel-parameters.txt +9 −0
Documentation/arm64/silicon-errata.txt +1 −0
Documentation/cgroup-v2.txt +6 −5
Documentation/dev-tools/kcov.rst +1 −1
Documentation/devicetree/bindings/powerpc/4xx/emac.txt +60 −2
Documentation/devicetree/bindings/regulator/ti-abb-regulator.txt +1 −1
Documentation/devicetree/bindings/usb/usb251xb.txt +18 −35
Documentation/networking/ip-sysctl.txt +2 −1
Documentation/virtual/kvm/api.txt +4 −0
Documentation/vm/userfaultfd.txt +0 −4
MAINTAINERS +0 −1
Makefile +1 −1
arch/arc/include/asm/hugepage.h +1 −0
arch/arc/include/asm/pgtable.h +1 −0
arch/arm/include/asm/kvm_arm.h +1 −0
arch/arm/include/asm/kvm_host.h +0 −1
arch/arm/include/asm/pgtable.h +1 −0
arch/arm/kvm/arm.c +3 −0
arch/arm/kvm/handle_exit.c +12 −7
arch/arm/xen/mm.c +2 −0
arch/arm64/Kconfig +10 −0
arch/arm64/include/asm/kvm_host.h +1 −2
arch/arm64/include/asm/pgtable-types.h +4 −0
arch/arm64/kvm/handle_exit.c +12 −7
arch/arm64/kvm/hyp/tlb.c +55 −9
arch/avr32/include/asm/pgtable-2level.h +1 −0
arch/avr32/oprofile/backtrace.c +1 −1
arch/cris/arch-v32/drivers/cryptocop.c +1 −1
arch/cris/include/asm/pgtable.h +1 −0
arch/frv/include/asm/pgtable.h +1 −0
arch/h8300/include/asm/pgtable.h +1 −0
arch/h8300/kernel/ptrace_h.c +1 −1
arch/hexagon/include/asm/pgtable.h +1 −0
arch/ia64/include/asm/pgtable.h +2 −0
arch/metag/include/asm/pgtable.h +1 −0
arch/microblaze/include/asm/page.h +2 −1
arch/mips/cavium-octeon/cpu.c +2 −0
arch/mips/cavium-octeon/crypto/octeon-crypto.c +1 −0
arch/mips/cavium-octeon/smp.c +1 −0
arch/mips/include/asm/fpu.h +1 −0
arch/mips/include/asm/pgtable-32.h +1 −0
arch/mips/include/asm/pgtable-64.h +1 −0
arch/mips/kernel/smp-bmips.c +1 −0
arch/mips/kernel/smp-mt.c +1 −0
arch/mips/loongson64/loongson-3/cop2-ex.c +1 −0
arch/mips/netlogic/common/smp.c +1 −0
arch/mips/netlogic/xlp/cop2-ex.c +3 −0
arch/mips/sgi-ip22/ip28-berr.c +1 −0
arch/mips/sgi-ip27/ip27-berr.c +2 −0
arch/mips/sgi-ip27/ip27-smp.c +3 −0
arch/mips/sgi-ip32/ip32-berr.c +1 −0
arch/mips/sgi-ip32/ip32-reset.c +1 −0
arch/mn10300/include/asm/page.h +1 −0
arch/nios2/include/asm/pgtable.h +1 −0
arch/openrisc/include/asm/pgtable.h +1 −0
arch/powerpc/Kconfig +72 −66
arch/powerpc/Makefile +10 −1
arch/powerpc/boot/zImage.lds.S +1 −0
arch/powerpc/crypto/crc32c-vpmsum_glue.c +1 −1
arch/powerpc/include/asm/bitops.h +4 −0
arch/powerpc/include/asm/book3s/32/pgtable.h +1 −0
arch/powerpc/include/asm/book3s/64/pgtable.h +58 −30
arch/powerpc/include/asm/checksum.h +1 −1
arch/powerpc/include/asm/cpuidle.h +2 −2
arch/powerpc/include/asm/elf.h +2 −2
arch/powerpc/include/asm/mce.h +107 −1
arch/powerpc/include/asm/nohash/32/pgtable.h +1 −0
arch/powerpc/include/asm/nohash/64/pgtable-4k.h +3 −0
arch/powerpc/include/asm/nohash/64/pgtable-64k.h +1 −0
arch/powerpc/include/asm/nohash/pgtable.h +1 −1
arch/powerpc/include/asm/ppc-opcode.h +7 −0
arch/powerpc/include/asm/prom.h +12 −6
arch/powerpc/kernel/cputable.c +3 −0
arch/powerpc/kernel/idle_book3s.S +6 −4
arch/powerpc/kernel/mce.c +86 −2
arch/powerpc/kernel/mce_power.c +237 −0
arch/powerpc/kernel/prom_init.c +113 −7
arch/powerpc/kernel/setup_64.c +4 −1
arch/powerpc/kvm/book3s_64_mmu_hv.c +1 −1
arch/powerpc/kvm/book3s_hv_rm_mmu.c +1 −1
arch/powerpc/lib/Makefile +1 −0
arch/powerpc/lib/sstep.c +0 −20
arch/powerpc/lib/test_emulate_step.c +434 −0
arch/powerpc/mm/init_64.c +30 −6
arch/powerpc/mm/pgtable-radix.c +4 −0
arch/powerpc/perf/core-book3s.c +2 −0
arch/powerpc/perf/isa207-common.c +36 −7
arch/powerpc/perf/isa207-common.h +1 −0
arch/powerpc/platforms/powernv/opal-wrappers.S +2 −2
arch/powerpc/platforms/powernv/opal.c +6 −15
arch/powerpc/platforms/powernv/pci-ioda.c +15 −5
arch/powerpc/purgatory/trampoline.S +6 −6
arch/powerpc/sysdev/axonram.c +4 −1
arch/powerpc/sysdev/xics/icp-opal.c +10 −0
arch/powerpc/sysdev/xics/xics-common.c +14 −3
arch/s390/crypto/paes_s390.c +4 −1
arch/s390/include/asm/cputime.h +8 −12
arch/s390/include/asm/pgtable.h +1 −0
arch/s390/include/asm/timex.h +4 −8
arch/s390/include/uapi/asm/unistd.h +3 −1
arch/s390/kernel/compat_wrapper.c +1 −0
arch/s390/kernel/entry.S +5 −5
arch/s390/kernel/ipl.c +2 −0
arch/s390/kernel/process.c +3 −0
arch/s390/kernel/syscalls.S +2 −0
arch/s390/kernel/vtime.c +1 −1
arch/s390/mm/pgtable.c +18 −1
arch/score/include/asm/pgtable.h +1 −0
arch/score/kernel/traps.c +1 −0
arch/score/mm/extable.c +2 −0
arch/sh/boards/mach-cayman/setup.c +0 −2
arch/sh/include/asm/pgtable-2level.h +1 −0
arch/sh/include/asm/pgtable-3level.h +1 −0
arch/sparc/include/asm/pgtable_64.h +1 −0
arch/tile/include/asm/pgtable_32.h +1 −0
arch/tile/include/asm/pgtable_64.h +1 −0
arch/um/include/asm/pgtable-2level.h +1 −0
arch/um/include/asm/pgtable-3level.h +1 −0
arch/unicore32/include/asm/pgtable.h +1 −0
arch/x86/configs/x86_64_defconfig +1 −0
arch/x86/events/amd/core.c +1 −1
arch/x86/events/intel/cstate.c +1 −1
arch/x86/events/intel/rapl.c +1 −1
arch/x86/events/intel/uncore.h +3 −3
arch/x86/hyperv/hv_init.c +1 −1
arch/x86/include/asm/cpufeatures.h +2 −1
arch/x86/include/asm/pgtable_types.h +4 −0
arch/x86/include/asm/pkeys.h +9 −6
arch/x86/include/asm/purgatory.h +20 −0
arch/x86/include/asm/tlbflush.h +1 −1
arch/x86/include/uapi/asm/bootparam.h +1 −1
arch/x86/kernel/apic/apic.c +7 −16
arch/x86/kernel/cpu/amd.c +0 −4
arch/x86/kernel/cpu/centaur.c +0 −2
arch/x86/kernel/cpu/common.c +0 −3
arch/x86/kernel/cpu/cyrix.c +0 −1
arch/x86/kernel/cpu/intel.c +0 −4
arch/x86/kernel/cpu/intel_rdt_rdtgroup.c +0 −1
arch/x86/kernel/cpu/transmeta.c +0 −2
arch/x86/kernel/cpu/vmware.c +0 −1
arch/x86/kernel/ftrace.c +1 −1
arch/x86/kernel/hpet.c +1 −1
arch/x86/kernel/kdebugfs.c +1 −1
arch/x86/kernel/machine_kexec_64.c +6 −3
arch/x86/kernel/reboot.c +16 −0
arch/x86/kernel/tsc.c +23 −12
arch/x86/kvm/vmx.c +8 −22
arch/x86/mm/gup.c +21 −16
arch/x86/pci/common.c +9 −0
arch/x86/pci/xen.c +7 −16
arch/x86/platform/uv/tlb_uv.c +0 −1
arch/x86/purgatory/purgatory.c +18 −18
arch/x86/purgatory/setup-x86_64.S +1 −0
arch/x86/purgatory/sha256.h +0 −1
arch/xtensa/include/asm/pgtable.h +1 −0
block/bio.c +9 −3
block/blk-core.c +31 −10
block/blk-mq-sysfs.c +25 −15
block/blk-mq-tag.c +3 −0
block/blk-mq.c +15 −22
block/blk-mq.h +2 −0
block/genhd.c +10 −27
block/sed-opal.c +2 −8
crypto/af_alg.c +5 −4
crypto/algif_hash.c +5 −4
drivers/acpi/internal.h +2 −0
drivers/acpi/ioapic.c +16 −6
drivers/acpi/pci_root.c +2 −2
drivers/ata/ahci_qoriq.c +4 −2
drivers/ata/libata-sff.c +0 −1
drivers/ata/libata-transport.c +3 −6
drivers/block/paride/pcd.c +1 −1
drivers/block/paride/pd.c +1 −1
drivers/block/paride/pf.c +1 −1
drivers/block/paride/pg.c +1 −1
drivers/block/paride/pt.c +1 −1
drivers/block/rbd.c +12 −4
drivers/block/zram/zram_drv.c +2 −0
drivers/char/hw_random/omap-rng.c +14 −2
drivers/char/nwbutton.c +1 −1
drivers/char/random.c +50 −79
drivers/cpufreq/cpufreq.c +1 −0
drivers/cpufreq/intel_pstate.c +31 −36
drivers/crypto/s5p-sss.c +85 −47
drivers/crypto/ux500/cryp/cryp.c +1 −1
drivers/firmware/efi/arm-runtime.c +1 −0
drivers/firmware/efi/libstub/secureboot.c +2 −2
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +4 −1
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +2 −1
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +21 −0
drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +1 −1
drivers/gpu/drm/drm_edid.c +3 −0
drivers/gpu/drm/i915/gvt/cfg_space.c +54 −3
drivers/gpu/drm/i915/gvt/cmd_parser.c +5 −5
drivers/gpu/drm/i915/gvt/display.c +94 −45
drivers/gpu/drm/i915/gvt/display.h +19 −1
drivers/gpu/drm/i915/gvt/firmware.c +1 −1
drivers/gpu/drm/i915/gvt/gtt.c +26 −14
drivers/gpu/drm/i915/gvt/gvt.h +10 −2
drivers/gpu/drm/i915/gvt/handlers.c +307 −132
drivers/gpu/drm/i915/gvt/kvmgt.c +6 −6
drivers/gpu/drm/i915/gvt/mmio.c +65 −1
drivers/gpu/drm/i915/gvt/opregion.c +2 −3
drivers/gpu/drm/i915/gvt/render.c +16 −0
drivers/gpu/drm/i915/gvt/scheduler.c +36 −16
drivers/gpu/drm/i915/gvt/vgpu.c +45 −27
drivers/gpu/drm/mxsfb/mxsfb_crtc.c +43 −6
drivers/gpu/drm/mxsfb/mxsfb_drv.c +4 −0
drivers/gpu/drm/mxsfb/mxsfb_out.c +4 −0
drivers/gpu/drm/mxsfb/mxsfb_regs.h +1 −0
drivers/gpu/drm/rcar-du/rcar_du_vsp.c +6 −2
drivers/hv/channel.c +1 −1
drivers/i2c/busses/i2c-brcmstb.c +21 −6
drivers/i2c/busses/i2c-designware-core.h +1 −0
drivers/i2c/busses/i2c-designware-platdrv.c +24 −4
drivers/i2c/busses/i2c-exynos5.c +2 −1
drivers/i2c/busses/i2c-meson.c +1 −1
drivers/i2c/busses/i2c-mt65xx.c +0 −9
drivers/i2c/busses/i2c-riic.c +5 −1
drivers/i2c/i2c-mux.c +2 −0
drivers/irqchip/irq-crossbar.c +5 −4
drivers/irqchip/irq-gic-v3-its.c +16 −0
drivers/isdn/gigaset/bas-gigaset.c +3 −0
drivers/isdn/hisax/st5481_b.c +1 −1
drivers/macintosh/macio_asic.c +1 −0
drivers/md/bcache/util.h +0 −1
drivers/md/dm.c +16 −13
drivers/md/raid10.c +2 −1
drivers/media/dvb-frontends/drx39xyj/drx_driver.h +3 −5
drivers/media/platform/vsp1/vsp1_drm.c +16 −17
drivers/media/rc/lirc_dev.c +2 −2
drivers/media/rc/nuvoton-cir.c +3 −2
drivers/media/rc/rc-main.c +15 −11
drivers/media/rc/serial_ir.c +62 −61
drivers/media/usb/dvb-usb/dw2102.c +145 −99
drivers/misc/sgi-gru/grufault.c +7 −2
drivers/mtd/spi-nor/spi-nor.c +1 −1
drivers/net/ethernet/amd/xgbe/xgbe-drv.c +2 −8
drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c +1 −1
drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +30 −6
drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.c +16 −8
drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h +1 −0
drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c +28 −12
drivers/net/ethernet/broadcom/bnxt/bnxt.c +18 −7
drivers/net/ethernet/broadcom/bnxt/bnxt.h +1 −0
drivers/net/ethernet/broadcom/bnxt/bnxt_dcb.c +1 −1
drivers/net/ethernet/broadcom/genet/bcmgenet.c +147 −59
drivers/net/ethernet/broadcom/genet/bcmgenet.h +11 −5
drivers/net/ethernet/cavium/liquidio/lio_main.c +55 −55
drivers/net/ethernet/cavium/liquidio/lio_vf_main.c +55 −49
drivers/net/ethernet/cavium/liquidio/octeon_config.h +3 −3
drivers/net/ethernet/cavium/liquidio/octeon_droq.c +2 −15
drivers/net/ethernet/cavium/liquidio/octeon_droq.h +2 −2
drivers/net/ethernet/cavium/liquidio/octeon_main.h +0 −42
drivers/net/ethernet/cavium/liquidio/octeon_network.h +27 −16
drivers/net/ethernet/cavium/thunder/nic.h +1 −0
drivers/net/ethernet/cavium/thunder/nicvf_main.c +11 −1
drivers/net/ethernet/cavium/thunder/nicvf_queues.c +145 −39
drivers/net/ethernet/cavium/thunder/nicvf_queues.h +3 −1
drivers/net/ethernet/cavium/thunder/thunder_bgx.c +45 −19
drivers/net/ethernet/cavium/thunder/thunder_bgx.h +0 −1
drivers/net/ethernet/ibm/emac/core.c +17 −8
drivers/net/ethernet/ibm/ibmvnic.c +34 −9
drivers/net/ethernet/ibm/ibmvnic.h +1 −0
drivers/net/ethernet/mellanox/mlx5/core/Kconfig +1 −0
drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c +5 −5
drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c +1 −4
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +18 −15
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +1 −1
drivers/net/ethernet/mellanox/mlx5/core/main.c +3 −2
drivers/net/ethernet/mellanox/mlxsw/reg.h +2 −2
drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c +2 −2
drivers/net/ethernet/qlogic/qed/qed_cxt.c +2 −1
drivers/net/ethernet/qlogic/qed/qed_dev.c +2 −3
drivers/net/ethernet/qlogic/qed/qed_iscsi.c +31 −0
drivers/net/ethernet/qlogic/qed/qed_ll2.c +8 −5
drivers/net/ethernet/qlogic/qed/qed_ooo.c +2 −0
drivers/net/ethernet/qlogic/qlge/qlge.h +2 −2
drivers/net/ethernet/smsc/smc91x.c +25 −22
drivers/net/hyperv/hyperv_net.h +2 −1
drivers/net/hyperv/netvsc.c +2 −6
drivers/net/hyperv/netvsc_drv.c +3 −8
drivers/net/phy/marvell.c +0 −15
drivers/net/phy/phy_device.c +1 −1
drivers/net/phy/spi_ks8995.c +2 −1
drivers/net/team/team.c +1 −0
drivers/net/tun.c +15 −4
drivers/net/vrf.c +2 −1
drivers/net/vxlan.c +40 −33
drivers/net/wan/fsl_ucc_hdlc.c +2 −2
drivers/net/wimax/i2400m/usb.c +3 −0
drivers/net/xen-netback/interface.c +17 −9
drivers/net/xen-netback/netback.c +1 −1
drivers/net/xen-netback/xenbus.c +10 −10
drivers/pci/dwc/pci-exynos.c +4 −4
drivers/pci/pcie/aspm.c +2 −3
drivers/pci/quirks.c +1 −0
drivers/pinctrl/qcom/pinctrl-msm.c +15 −0
drivers/pinctrl/uniphier/pinctrl-uniphier-ld11.c +6 −6
drivers/platform/x86/asus-nb-wmi.c +2 −47
drivers/platform/x86/asus-wmi.c +17 −5
drivers/platform/x86/asus-wmi.h +0 −1
drivers/platform/x86/fujitsu-laptop.c +222 −229
drivers/scsi/Kconfig +16 −3
drivers/scsi/aacraid/src.c +1 −1
drivers/scsi/aic7xxx/aic79xx_core.c +1 −1
drivers/scsi/libiscsi.c +25 −1
drivers/scsi/lpfc/lpfc.h +3 −1
drivers/scsi/lpfc/lpfc_attr.c +8 −1
drivers/scsi/lpfc/lpfc_crtn.h +3 −1
drivers/scsi/lpfc/lpfc_ct.c +1 −1
drivers/scsi/lpfc/lpfc_debugfs.c +0 −22
drivers/scsi/lpfc/lpfc_els.c +14 −8
drivers/scsi/lpfc/lpfc_hbadisc.c +13 −11
drivers/scsi/lpfc/lpfc_hw4.h +2 −2
drivers/scsi/lpfc/lpfc_init.c +85 −43
drivers/scsi/lpfc/lpfc_mem.c +1 −1
drivers/scsi/lpfc/lpfc_nvme.c +82 −25
drivers/scsi/lpfc/lpfc_nvme.h +1 −0
drivers/scsi/lpfc/lpfc_nvmet.c +35 −8
drivers/scsi/lpfc/lpfc_scsi.c +2 −2
drivers/scsi/lpfc/lpfc_sli.c +59 −9
drivers/scsi/lpfc/lpfc_sli4.h +6 −0
drivers/scsi/lpfc/lpfc_version.h +1 −1
drivers/scsi/mpt3sas/mpt3sas_base.h +0 −3
drivers/scsi/mpt3sas/mpt3sas_scsih.c +2 −17
drivers/scsi/qedf/qedf_dbg.h +8 −5
drivers/scsi/qedf/qedf_fip.c +1 −1
drivers/scsi/qedf/qedf_io.c +2 −2
drivers/scsi/qedf/qedf_main.c +2 −2
drivers/scsi/qedi/qedi_debugfs.c +8 −8
drivers/scsi/qedi/qedi_fw.c +2 −2
drivers/scsi/qedi/qedi_gbl.h +7 −1
drivers/scsi/qedi/qedi_iscsi.c +4 −4
drivers/scsi/qedi/qedi_main.c +1 −1
drivers/scsi/qla2xxx/qla_dbg.c +4 −8
drivers/scsi/scsi_lib.c +10 −4
drivers/scsi/scsi_priv.h +0 −3
drivers/scsi/sd.c +25 −33
drivers/scsi/storvsc_drv.c +17 −10
drivers/scsi/ufs/ufs.h +8 −14
drivers/scsi/ufs/ufshcd.c +173 −58
drivers/scsi/ufs/ufshcd.h +15 −0
drivers/scsi/vmw_pvscsi.c +1 −1
drivers/staging/lustre/lnet/lnet/lib-socket.c +2 −2
drivers/staging/octeon/ethernet-rx.c +0 −1
drivers/staging/vc04_services/Kconfig +1 −0
drivers/tty/n_hdlc.c +69 −63
drivers/tty/serial/samsung.c +4 −2
drivers/usb/dwc3/dwc3-omap.c +2 −1
drivers/usb/dwc3/gadget.c +67 −9
drivers/usb/dwc3/gadget.h +7 −7
drivers/usb/gadget/configfs.c +1 −0
drivers/usb/gadget/function/f_fs.c +14 −3
drivers/usb/gadget/function/f_uvc.c +0 −7
drivers/usb/gadget/legacy/inode.c +4 −3
drivers/usb/gadget/udc/atmel_usba_udc.c +2 −2
drivers/usb/gadget/udc/dummy_hcd.c +2 −0
drivers/usb/gadget/udc/net2280.c +13 −12
drivers/usb/gadget/udc/pxa27x_udc.c +3 −2
drivers/usb/host/ohci-at91.c +2 −2
drivers/usb/host/xhci-dbg.c +1 −1
drivers/usb/host/xhci-mtk.c +3 −4
drivers/usb/host/xhci-plat.c +2 −0
drivers/usb/host/xhci-tegra.c +0 −1
drivers/usb/host/xhci.c +2 −2
drivers/usb/misc/iowarrior.c +15 −6
drivers/usb/misc/usb251xb.c +24 −35
drivers/usb/phy/phy-isp1301.c +7 −0
drivers/usb/serial/digi_acceleport.c +1 −1
drivers/usb/serial/io_ti.c +7 −1
drivers/usb/serial/omninet.c +0 −13
drivers/usb/serial/safe_serial.c +5 −0
drivers/usb/storage/unusual_devs.h +14 −0
drivers/xen/swiotlb-xen.c +47 −0
drivers/xen/xenbus/xenbus_dev_frontend.c +0 −1
fs/dlm/lowcomms.c +1 −1
fs/fat/inode.c +12 −1
fs/fs-writeback.c +21 −14
fs/gfs2/incore.h +1 −1
fs/iomap.c +10 −7
fs/ocfs2/cluster/tcp.c +1 −1
fs/overlayfs/util.c +0 −1
fs/timerfd.c +4 −4
fs/userfaultfd.c +26 −49
fs/xfs/kmem.c +0 −18
fs/xfs/kmem.h +0 −2
fs/xfs/libxfs/xfs_bmap.c +21 −13
fs/xfs/libxfs/xfs_bmap_btree.c +3 −3
fs/xfs/xfs_aops.c +27 −32
fs/xfs/xfs_icache.c +1 −1
fs/xfs/xfs_inode.c +1 −1
fs/xfs/xfs_iomap.c +18 −7
fs/xfs/xfs_itable.c +2 −4
fs/xfs/xfs_mount.c +1 −2
fs/xfs/xfs_reflink.c +16 −7
fs/xfs/xfs_reflink.h +2 −2
fs/xfs/xfs_super.c +1 −1
include/asm-generic/4level-fixup.h +2 −1
include/asm-generic/5level-fixup.h +41 −0
include/asm-generic/pgtable-nop4d-hack.h +62 −0
include/asm-generic/pgtable-nop4d.h +56 −0
include/asm-generic/pgtable-nopud.h +27 −21
include/asm-generic/pgtable.h +45 −3
include/asm-generic/tlb.h +11 −3
include/crypto/if_alg.h +1 −1
include/dt-bindings/sound/cs42l42.h +1 −1
include/linux/blkdev.h +0 −1
include/linux/ceph/libceph.h +2 −0
include/linux/ceph/osd_client.h +1 −0
include/linux/dccp.h +1 −0
include/linux/filter.h +12 −4
include/linux/fs.h +1 −1
include/linux/genhd.h +0 −8
include/linux/hugetlb.h +4 −1
include/linux/irqchip/arm-gic-v3.h +2 −0
include/linux/irqdomain.h +4 −0
include/linux/jump_label.h +9 −2
include/linux/kasan.h +1 −0
include/linux/list_nulls.h +5 −0
include/linux/mm.h +28 −6
include/linux/net.h +1 −1
include/linux/phy.h +4 −0
include/linux/purgatory.h +23 −0
include/linux/random.h +15 −3
include/linux/rculist_nulls.h +14 −0
include/linux/regulator/machine.h +1 −1
include/linux/user_namespace.h +1 −1
include/linux/userfaultfd_k.h +3 −10
include/linux/vm_event_item.h +3 −0
include/linux/wait.h +10 −21
include/media/vsp1.h +11 −2
include/net/inet_common.h +2 −1
include/net/inet_connection_sock.h +1 −1
include/net/irda/timer.h +1 −1
include/net/sctp/structs.h +2 −1
include/net/sock.h +6 −3
include/scsi/libiscsi.h +1 −0
include/scsi/scsi_device.h +4 −0
include/trace/events/syscalls.h +1 −0
include/uapi/linux/packet_diag.h +1 −1
include/uapi/linux/userfaultfd.h +1 −4
include/xen/swiotlb-xen.h +11 −0
init/main.c +0 −1
kernel/bpf/hashtab.c +80 −39
kernel/bpf/lpm_trie.c +6 −0
kernel/cgroup/cgroup-v1.c +1 −1
kernel/cgroup/cgroup.c +1 −1
kernel/cgroup/pids.c +1 −1
kernel/events/core.c +1 −1
kernel/exit.c +0 −1
kernel/kexec_file.c +4 −4
kernel/kexec_internal.h +1 −5
kernel/locking/lockdep.c +9 −2
kernel/locking/test-ww_mutex.c +3 −3
kernel/sched/core.c +8 −3
kernel/sched/cpufreq_schedutil.c +9 −10
kernel/sched/fair.c +1 −1
kernel/sched/features.h +5 −0
kernel/sched/wait.c +39 −0
kernel/time/jiffies.c +1 −1
kernel/trace/ftrace.c +18 −5
kernel/trace/trace_stack.c +1 −1
kernel/ucount.c +11 −7
kernel/workqueue.c +1 −0
lib/ioremap.c +36 −3
lib/radix-tree.c +2 −2
lib/refcount.c +7 −7
mm/backing-dev.c +28 −15
mm/gup.c +40 −6
mm/huge_memory.c +7 −2
mm/hugetlb.c +18 −11
mm/kasan/kasan_init.c +39 −5
mm/kasan/quarantine.c +44 −7
mm/madvise.c +41 −3
mm/memblock.c +4 −1
mm/memcontrol.c +13 −5
mm/memory.c +196 −34
mm/mlock.c +5 −5
mm/mprotect.c +23 −3
mm/mremap.c +11 −2
mm/page_alloc.c +2 −1
mm/page_vma_mapped.c +5 −1
mm/pagewalk.c +29 −3
mm/percpu-vm.c +3 −4
mm/percpu.c +4 −1
mm/pgtable-generic.c +6 −0
mm/rmap.c +13 −7
mm/sparse-vmemmap.c +19 −3
mm/swapfile.c +23 −3
mm/userfaultfd.c +13 −10
mm/vmalloc.c +60 −21
mm/vmstat.c +3 −0
net/atm/svc.c +3 −2
net/ax25/af_ax25.c +2 −1
net/bluetooth/l2cap_sock.c +1 −1
net/bluetooth/rfcomm/sock.c +2 −1
net/bluetooth/sco.c +1 −1
net/bridge/br_input.c +1 −0
net/bridge/br_netfilter_hooks.c +0 −21
net/ceph/ceph_common.c +15 −0
net/ceph/osd_client.c +35 −1
net/ceph/osdmap.c +1 −3
net/core/dev.c +1 −0
net/core/net-sysfs.c +3 −3
net/core/skbuff.c +16 −14
net/core/sock.c +57 −49
net/dccp/ccids/ccid2.c +1 −0
net/dccp/ipv4.c +2 −1
net/dccp/ipv6.c +5 −3
net/dccp/minisocks.c +16 −8
net/decnet/af_decnet.c +3 −2
net/ipv4/af_inet.c +6 −3
net/ipv4/inet_connection_sock.c +1 −1
net/ipv4/ip_output.c +1 −1
net/ipv4/tcp_ipv4.c +7 −3
net/ipv4/tcp_timer.c +4 −2
net/ipv6/af_inet6.c +5 −5
net/ipv6/ip6_fib.c +2 −0
net/ipv6/ip6_offload.c +3 −1
net/ipv6/ip6_output.c +7 −2
net/ipv6/ip6_vti.c +6 −2
net/ipv6/route.c +6 −5
net/ipv6/tcp_ipv6.c +5 −3
net/irda/af_irda.c +3 −2
net/iucv/af_iucv.c +1 −1
net/llc/af_llc.c +3 −1
net/mpls/af_mpls.c +3 −1
net/netrom/af_netrom.c +2 −1
net/nfc/llcp_sock.c +1 −1
net/phonet/pep.c +4 −2
net/phonet/socket.c +2 −2
net/rds/connection.c +1 −0
net/rds/ib_cm.c +36 −11
net/rds/rds.h +3 −3
net/rds/tcp.c +21 −17
net/rds/tcp.h +1 −1
net/rds/tcp_listen.c +8 −3
net/rose/af_rose.c +2 −1
net/rxrpc/input.c +19 −8
net/rxrpc/recvmsg.c +2 −2
net/rxrpc/sendmsg.c +31 −18
net/sched/act_connmark.c +3 −0
net/sched/act_skbmod.c +0 −1
net/sctp/ipv6.c +3 −2
net/sctp/protocol.c +3 −2
net/sctp/socket.c +2 −2
net/smc/af_smc.c +1 −1
net/socket.c +3 −2
net/tipc/socket.c +5 −3
net/unix/af_unix.c +3 −2
net/vmw_vsock/af_vsock.c +2 −1
net/x25/af_x25.c +2 −1
net/xfrm/xfrm_policy.c +9 −10
scripts/gcc-plugins/sancov_plugin.c +1 −1
scripts/module-common.lds +2 −0
scripts/spelling.txt +3 −0
sound/soc/amd/acp-pcm-dma.c +1 −1
tools/include/uapi/linux/bpf_perf_event.h +18 −0
tools/lguest/lguest.c +1 −1
tools/lib/bpf/Makefile +1 −1
tools/lib/traceevent/Makefile +1 −1
tools/lib/traceevent/event-parse.h +1 −1
tools/objtool/builtin-check.c +12 −3
tools/objtool/elf.c +12 −0
tools/objtool/elf.h +1 −0
tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c +1 −1
tools/testing/ktest/ktest.pl +13 −8
tools/testing/radix-tree/Makefile +9 −6
tools/testing/radix-tree/benchmark.c +166 −7
tools/testing/radix-tree/idr-test.c +75 −3
tools/testing/radix-tree/main.c +1 −0
tools/testing/radix-tree/tag_check.c +29 −0
tools/testing/radix-tree/test.h +1 −0
tools/testing/selftests/bpf/Makefile +3 −1
tools/testing/selftests/bpf/test_verifier.c +4 −0
tools/testing/selftests/powerpc/harness.c +4 −2
tools/testing/selftests/powerpc/include/vsx_asm.h +24 −24
tools/testing/selftests/vm/Makefile +4 −0
tools/testing/selftests/x86/fsgsbase.c +1 −1
tools/testing/selftests/x86/ldt_gdt.c +11 −5
tools/testing/selftests/x86/ptrace_syscall.c +2 −1
tools/testing/selftests/x86/single_step_syscall.c +4 −1
virt/kvm/arm/vgic/vgic-its.c +65 −44
virt/kvm/arm/vgic/vgic-mmio.c +24 −8
virt/kvm/arm/vgic/vgic-v3.c +4 −1

Documentation/admin-guide/kernel-parameters.txt+9−0

@@ -653,6 +653,9 @@ · 652 unmodified lines
653	653	cpuidle.off=1 [CPU_IDLE]
654	654	disable the cpuidle sub-system
655	655
	656	+ cpufreq.off=1 [CPU_FREQ]
	657	+ disable the cpufreq sub-system
	658	+
656	659	cpu_init_udelay=N
657	660	[X86] Delay for N microsec between assert and de-assert
658	661	of APIC INIT to start processors. This delay occurs
@@ -1183,6 +1186,12 @@ · 524 unmodified lines
1183	1186	functions that can be changed at run time by the
1184	1187	set_graph_notrace file in the debugfs tracing directory.
1185	1188
	1189	+ ftrace_graph_max_depth=<uint>
	1190	+ [FTRACE] Used with the function graph tracer. This is
	1191	+ the max depth it will trace into a function. This value
	1192	+ can be changed at run time by the max_graph_depth file
	1193	+ in the tracefs tracing directory. default: 0 (no limit)
	1194	+
1186	1195	gamecon.map[2\|3]=
1187	1196	[HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1188	1197	support via parallel port (up to 5 devices per port)

Documentation/arm64/silicon-errata.txt+1−0

@@ -68,3 +68,4 @@ stable kernels. · 67 unmodified lines
68	68	\| \| \| \| \|
69	69	\| Qualcomm Tech. \| Falkor v1 \| E1003 \| QCOM_FALKOR_ERRATUM_1003 \|
70	70	\| Qualcomm Tech. \| Falkor v1 \| E1009 \| QCOM_FALKOR_ERRATUM_1009 \|
	71	+\| Qualcomm Tech. \| QDF2400 ITS \| E0065 \| QCOM_QDF2400_ERRATUM_0065 \|

Documentation/cgroup-v2.txt+6−5

@@ -1142,16 +1142,17 @@ used by the kernel. · 1141 unmodified lines
1142	1142
1143	1143	pids.max
1144	1144
1145		− A read-write single value file which exists on non-root cgroups. The
1146		− default is "max".
	1145	+ A read-write single value file which exists on non-root
	1146	+ cgroups. The default is "max".
1147	1147
1148		− Hard limit of number of processes.
	1148	+ Hard limit of number of processes.
1149	1149
1150	1150	pids.current
1151	1151
1152		− A read-only single value file which exists on all cgroups.
	1152	+ A read-only single value file which exists on all cgroups.
1153	1153
1154		− The number of processes currently in the cgroup and its descendants.
	1154	+ The number of processes currently in the cgroup and its
	1155	+ descendants.
1155	1156
1156	1157	Organisational operations are not blocked by cgroup policies, so it is
1157	1158	possible to have pids.current > pids.max. This can be done by either

Documentation/dev-tools/kcov.rst+1−1

@@ -10,7 +10,7 @@ Note that kcov does not aim to collect as much coverage as possible. It aims · 9 unmodified lines
10	10	to collect more or less stable coverage that is function of syscall inputs.
11	11	To achieve this goal it does not collect coverage in soft/hard interrupts
12	12	and instrumentation of some inherently non-deterministic parts of kernel is
13		−disbled (e.g. scheduler, locking).
	13	+disabled (e.g. scheduler, locking).
14	14
15	15	Usage
16	16	-----

Documentation/devicetree/bindings/powerpc/4xx/emac.txt+60−2

@@ -71,6 +71,9 @@ · 70 unmodified lines
71	71	For Axon it can be absent, though my current driver
72	72	doesn't handle phy-address yet so for now, keep
73	73	0x00ffffff in it.
	74	+ - phy-handle : Used to describe configurations where a external PHY
	75	+ is used. Please refer to:
	76	+ Documentation/devicetree/bindings/net/ethernet.txt
74	77	- rx-fifo-size-gige : 1 cell, Rx fifo size in bytes for 1000 Mb/sec
75	78	operations (if absent the value is the same as
76	79	rx-fifo-size). For Axon, either absent or 2048.
@@ -81,8 +84,22 @@ · 4 unmodified lines
81	84	offload, phandle of the TAH device node.
82	85	- tah-channel : 1 cell, optional. If appropriate, channel used on the
83	86	TAH engine.
	87	+ - fixed-link : Fixed-link subnode describing a link to a non-MDIO
	88	+ managed entity. See
	89	+ Documentation/devicetree/bindings/net/fixed-link.txt
	90	+ for details.
	91	+ - mdio subnode : When the EMAC has a phy connected to its local
	92	+ mdio, which us supported by the kernel's network
	93	+ PHY library in drivers/net/phy, there must be device
	94	+ tree subnode with the following required properties:
	95	+ - #address-cells: Must be <1>.
	96	+ - #size-cells: Must be <0>.
84	97
85		− Example:
	98	+ For PHY definitions: Please refer to
	99	+ Documentation/devicetree/bindings/net/phy.txt and
	100	+ Documentation/devicetree/bindings/net/ethernet.txt
	101	+
	102	+ Examples:
86	103
87	104	EMAC0: ethernet@40000800 {
88	105	device_type = "network";
@@ -104,6 +121,48 @@ · 15 unmodified lines
104	121	zmii-channel = <0>;
105	122	};
106	123
	124	+ EMAC1: ethernet@ef600c00 {
	125	+ device_type = "network";
	126	+ compatible = "ibm,emac-apm821xx", "ibm,emac4sync";
	127	+ interrupt-parent = <&EMAC1>;
	128	+ interrupts = <0 1>;
	129	+ #interrupt-cells = <1>;
	130	+ #address-cells = <0>;
	131	+ #size-cells = <0>;
	132	+ interrupt-map = <0 &UIC2 0x10 IRQ_TYPE_LEVEL_HIGH /* Status */
	133	+ 1 &UIC2 0x14 IRQ_TYPE_LEVEL_HIGH /* Wake */>;
	134	+ reg = <0xef600c00 0x000000c4>;
	135	+ local-mac-address = [000000000000]; /* Filled in by U-Boot */
	136	+ mal-device = <&MAL0>;
	137	+ mal-tx-channel = <0>;
	138	+ mal-rx-channel = <0>;
	139	+ cell-index = <0>;
	140	+ max-frame-size = <9000>;
	141	+ rx-fifo-size = <16384>;
	142	+ tx-fifo-size = <2048>;
	143	+ fifo-entry-size = <10>;
	144	+ phy-mode = "rgmii";
	145	+ phy-handle = <&phy0>;
	146	+ phy-map = <0x00000000>;
	147	+ rgmii-device = <&RGMII0>;
	148	+ rgmii-channel = <0>;
	149	+ tah-device = <&TAH0>;
	150	+ tah-channel = <0>;
	151	+ has-inverted-stacr-oc;
	152	+ has-new-stacr-staopc;
	153	+
	154	+ mdio {
	155	+ #address-cells = <1>;
	156	+ #size-cells = <0>;
	157	+
	158	+ phy0: ethernet-phy@0 {
	159	+ compatible = "ethernet-phy-ieee802.3-c22";
	160	+ reg = <0>;
	161	+ };
	162	+ };
	163	+ };
	164	+
	165	+
107	166	ii) McMAL node
108	167
109	168	Required properties:
@@ -145,4 +204,3 @@ · 35 unmodified lines
145	204	- revision : as provided by the RGMII new version register if
146	205	available.
147	206	For Axon: 0x0000012a
148		−

Documentation/devicetree/bindings/regulator/ti-abb-regulator.txt+1−1

@@ -45,7 +45,7 @@ Required Properties: · 44 unmodified lines
45	45	Optional Properties:
46	46	- reg-names: In addition to the required properties, the following are optional
47	47	- "efuse-address" - Contains efuse base address used to pick up ABB info.
48		− - "ldo-address" - Contains address of ABB LDO overide register address.
	48	+ - "ldo-address" - Contains address of ABB LDO override register.
49	49	"efuse-address" is required for this.
50	50	- ti,ldovbb-vset-mask - Required if ldo-address is set, mask for LDO override
51	51	register to provide override vset value.

Documentation/devicetree/bindings/usb/usb251xb.txt+18−35

@@ -7,18 +7,18 @@ Required properties : · 6 unmodified lines
7	7	- compatible : Should be "microchip,usb251xb" or one of the specific types:
8	8	"microchip,usb2512b", "microchip,usb2512bi", "microchip,usb2513b",
9	9	"microchip,usb2513bi", "microchip,usb2514b", "microchip,usb2514bi"
10		− - hub-reset-gpios : Should specify the gpio for hub reset
	10	+ - reset-gpios : Should specify the gpio for hub reset
	11	+ - reg : I2C address on the selected bus (default is <0x2C>)
11	12
12	13	Optional properties :
13		− - reg : I2C address on the selected bus (default is <0x2C>)
14	14	- skip-config : Skip Hub configuration, but only send the USB-Attach command
15		− - vendor-id : USB Vendor ID of the hub (16 bit, default is 0x0424)
16		− - product-id : USB Product ID of the hub (16 bit, default depends on type)
17		− - device-id : USB Device ID of the hub (16 bit, default is 0x0bb3)
18		− - language-id : USB Language ID (16 bit, default is 0x0000)
19		− - manufacturer : USB Manufacturer string (max 31 characters long)
20		− - product : USB Product string (max 31 characters long)
21		− - serial : USB Serial string (max 31 characters long)
	15	+ - vendor-id : Set USB Vendor ID of the hub (16 bit, default is 0x0424)
	16	+ - product-id : Set USB Product ID of the hub (16 bit, default depends on type)
	17	+ - device-id : Set USB Device ID of the hub (16 bit, default is 0x0bb3)
	18	+ - language-id : Set USB Language ID (16 bit, default is 0x0000)
	19	+ - manufacturer : Set USB Manufacturer string (max 31 characters long)
	20	+ - product : Set USB Product string (max 31 characters long)
	21	+ - serial : Set USB Serial string (max 31 characters long)
22	22	- {bus,self}-powered : selects between self- and bus-powered operation (default
23	23	is self-powered)
24	24	- disable-hi-speed : disable USB Hi-Speed support
@@ -31,8 +31,10 @@ Optional properties : · 6 unmodified lines
31	31	(default is individual)
32	32	- dynamic-power-switching : enable auto-switching from self- to bus-powered
33	33	operation if the local power source is removed or unavailable
34		− - oc-delay-{100us,4ms,8ms,16ms} : set over current timer delay (default is 8ms)
35		− - compound-device : indicated the hub is part of a compound device
	34	+ - oc-delay-us : Delay time (in microseconds) for filtering the over-current
	35	+ sense inputs. Valid values are 100, 4000, 8000 (default) and 16000. If
	36	+ an invalid value is given, the default is used instead.
	37	+ - compound-device : indicate the hub is part of a compound device
36	38	- port-mapping-mode : enable port mapping mode
37	39	- string-support : enable string descriptor support (required for manufacturer,
38	40	product and serial string configuration)
@@ -40,34 +42,15 @@ Optional properties : · 1 unmodified line
40	42	device connected.
41	43	- sp-disabled-ports : Specifies the ports which will be self-power disabled
42	44	- bp-disabled-ports : Specifies the ports which will be bus-power disabled
43		− - max-sp-power : Specifies the maximum current the hub consumes from an
44		− upstream port when operating as self-powered hub including the power
45		− consumption of a permanently attached peripheral if the hub is
46		− configured as a compound device. The value is given in mA in a 0 - 500
47		− range (default is 2).
48		− - max-bp-power : Specifies the maximum current the hub consumes from an
49		− upstream port when operating as bus-powered hub including the power
50		− consumption of a permanently attached peripheral if the hub is
51		− configured as a compound device. The value is given in mA in a 0 - 500
52		− range (default is 100).
53		− - max-sp-current : Specifies the maximum current the hub consumes from an
54		− upstream port when operating as self-powered hub EXCLUDING the power
55		− consumption of a permanently attached peripheral if the hub is
56		− configured as a compound device. The value is given in mA in a 0 - 500
57		− range (default is 2).
58		− - max-bp-current : Specifies the maximum current the hub consumes from an
59		− upstream port when operating as bus-powered hub EXCLUDING the power
60		− consumption of a permanently attached peripheral if the hub is
61		− configured as a compound device. The value is given in mA in a 0 - 500
62		− range (default is 100).
63		− - power-on-time : Specifies the time it takes from the time the host initiates
64		− the power-on sequence to a port until the port has adequate power. The
65		− value is given in ms in a 0 - 510 range (default is 100ms).
	45	+ - power-on-time-ms : Specifies the time it takes from the time the host
	46	+ initiates the power-on sequence to a port until the port has adequate
	47	+ power. The value is given in ms in a 0 - 510 range (default is 100ms).
66	48
67	49	Examples:
68	50	usb2512b@2c {
69	51	compatible = "microchip,usb2512b";
70		− hub-reset-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>;
	52	+ reg = <0x2c>;
	53	+ reset-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>;
71	54	};
72	55
73	56	usb2514b@2c {

Documentation/networking/ip-sysctl.txt+2−1

@@ -1006,7 +1006,8 @@ accept_redirects - BOOLEAN · 1005 unmodified lines
1006	1006	FALSE (router)
1007	1007
1008	1008	forwarding - BOOLEAN
1009		− Enable IP forwarding on this interface.
	1009	+ Enable IP forwarding on this interface. This controls whether packets
	1010	+ received _on_ this interface can be forwarded.
1010	1011
1011	1012	mc_forwarding - BOOLEAN
1012	1013	Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE

Documentation/virtual/kvm/api.txt+4−0

@@ -951,6 +951,10 @@ This ioctl allows the user to create or modify a guest physical memory · 950 unmodified lines
951	951	slot. When changing an existing slot, it may be moved in the guest
952	952	physical memory space, or its flags may be modified. It may not be
953	953	resized. Slots may not overlap in guest physical address space.
	954	+Bits 0-15 of "slot" specifies the slot id and this value should be
	955	+less than the maximum number of user memory slots supported per VM.
	956	+The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS,
	957	+if this capability is supported by the architecture.
954	958
955	959	If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot"
956	960	specifies the address space which is being modified. They must be

Documentation/vm/userfaultfd.txt+0−4

@@ -172,10 +172,6 @@ the same read(2) protocol as for the page fault notifications. The · 171 unmodified lines
172	172	manager has to explicitly enable these events by setting appropriate
173	173	bits in uffdio_api.features passed to UFFDIO_API ioctl:
174	174
175		−UFFD_FEATURE_EVENT_EXIT - enable notification about exit() of the
176		−non-cooperative process. When the monitored process exits, the uffd
177		−manager will get UFFD_EVENT_EXIT.
178		−
179	175	UFFD_FEATURE_EVENT_FORK - enable userfaultfd hooks for fork(). When
180	176	this feature is enabled, the userfaultfd context of the parent process
181	177	is duplicated into the newly created process. The manager receives

MAINTAINERS+0−1

@@ -8307,7 +8307,6 @@ M: Richard Leitner <richard.leitner@skidata.com> · 8306 unmodified lines
8307	8307	L: linux-usb@vger.kernel.org
8308	8308	S: Maintained
8309	8309	F: drivers/usb/misc/usb251xb.c
8310		−F: include/linux/platform_data/usb251xb.h
8311	8310	F: Documentation/devicetree/bindings/usb/usb251xb.txt
8312	8311
8313	8312	MICROSOFT SURFACE PRO 3 BUTTON DRIVER

Makefile+1−1

@@ -1,7 +1,7 @@
1	1	VERSION = 4
2	2	PATCHLEVEL = 11
3	3	SUBLEVEL = 0
4		−EXTRAVERSION = -rc1
	4	+EXTRAVERSION = -rc2
5	5	NAME = Fearless Coyote
6	6
7	7	# DOCUMENTATION

arch/arc/include/asm/hugepage.h+1−0

@@ -11,6 +11,7 @@ · 10 unmodified lines
11	11	#define _ASM_ARC_HUGEPAGE_H
12	12
13	13	#include <linux/types.h>
	14	+#define __ARCH_USE_5LEVEL_HACK
14	15	#include <asm-generic/pgtable-nopmd.h>
15	16
16	17	static inline pte_t pmd_pte(pmd_t pmd)

arch/arc/include/asm/pgtable.h+1−0

@@ -37,6 +37,7 @@ · 36 unmodified lines
37	37
38	38	#include <asm/page.h>
39	39	#include <asm/mmu.h>
	40	+#define __ARCH_USE_5LEVEL_HACK
40	41	#include <asm-generic/pgtable-nopmd.h>
41	42	#include <linux/const.h>
42	43

arch/arm/include/asm/kvm_arm.h+1−0

@@ -209,6 +209,7 @@ · 208 unmodified lines
209	209	#define HSR_EC_IABT_HYP (0x21)
210	210	#define HSR_EC_DABT (0x24)
211	211	#define HSR_EC_DABT_HYP (0x25)
	212	+#define HSR_EC_MAX (0x3f)
212	213
213	214	#define HSR_WFI_IS_WFE (_AC(1, UL) << 0)
214	215

arch/arm/include/asm/kvm_host.h+0−1

@@ -30,7 +30,6 @@ · 29 unmodified lines
30	30	#define __KVM_HAVE_ARCH_INTC_INITIALIZED
31	31
32	32	#define KVM_USER_MEM_SLOTS 32
33		−#define KVM_PRIVATE_MEM_SLOTS 4
34	33	#define KVM_COALESCED_MMIO_PAGE_OFFSET 1
35	34	#define KVM_HAVE_ONE_REG
36	35	#define KVM_HALT_POLL_NS_DEFAULT 500000

arch/arm/include/asm/pgtable.h+1−0

@@ -20,6 +20,7 @@ · 19 unmodified lines
20	20
21	21	#else
22	22
	23	+#define __ARCH_USE_5LEVEL_HACK
23	24	#include <asm-generic/pgtable-nopud.h>
24	25	#include <asm/memory.h>
25	26	#include <asm/pgtable-hwdef.h>

arch/arm/kvm/arm.c+3−0

@@ -221,6 +221,9 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) · 220 unmodified lines
221	221	case KVM_CAP_MAX_VCPUS:
222	222	r = KVM_MAX_VCPUS;
223	223	break;
	224	+ case KVM_CAP_NR_MEMSLOTS:
	225	+ r = KVM_USER_MEM_SLOTS;
	226	+ break;
224	227	case KVM_CAP_MSI_DEVID:
225	228	if (!kvm)
226	229	r = -EINVAL;

arch/arm/kvm/handle_exit.c+12−7

@@ -79,7 +79,19 @@ static int kvm_handle_wfx(struct kvm_vcpu vcpu, struct kvm_run run) · 78 unmodified lines
79	79	return 1;
80	80	}
81	81
	82	+static int kvm_handle_unknown_ec(struct kvm_vcpu vcpu, struct kvm_run run)
	83	+{
	84	+ u32 hsr = kvm_vcpu_get_hsr(vcpu);
	85	+
	86	+ kvm_pr_unimpl("Unknown exception class: hsr: %#08x\n",
	87	+ hsr);
	88	+
	89	+ kvm_inject_undefined(vcpu);
	90	+ return 1;
	91	+}
	92	+
82	93	static exit_handle_fn arm_exit_handlers[] = {
	94	+ [0 ... HSR_EC_MAX] = kvm_handle_unknown_ec,
83	95	[HSR_EC_WFI] = kvm_handle_wfx,
84	96	[HSR_EC_CP15_32] = kvm_handle_cp15_32,
85	97	[HSR_EC_CP15_64] = kvm_handle_cp15_64,
@@ -98,13 +110,6 @@ static exit_handle_fn kvm_get_exit_handler(struct kvm_vcpu *vcpu) · 12 unmodified lines
98	110	{
99	111	u8 hsr_ec = kvm_vcpu_trap_get_class(vcpu);
100	112
101		− if (hsr_ec >= ARRAY_SIZE(arm_exit_handlers) \|\|
102		− !arm_exit_handlers[hsr_ec]) {
103		− kvm_err("Unknown exception class: hsr: %#08x\n",
104		− (unsigned int)kvm_vcpu_get_hsr(vcpu));
105		− BUG();
106		− }
107		−
108	113	return arm_exit_handlers[hsr_ec];
109	114	}
110	115

arch/arm/xen/mm.c+2−0

@@ -198,6 +198,8 @@ static const struct dma_map_ops xen_swiotlb_dma_ops = { · 197 unmodified lines
198	198	.unmap_page = xen_swiotlb_unmap_page,
199	199	.dma_supported = xen_swiotlb_dma_supported,
200	200	.set_dma_mask = xen_swiotlb_set_dma_mask,
	201	+ .mmap = xen_swiotlb_dma_mmap,
	202	+ .get_sgtable = xen_swiotlb_get_sgtable,
201	203	};
202	204
203	205	int __init xen_mm_init(void)

arch/arm64/Kconfig+10−0

@@ -508,6 +508,16 @@ config QCOM_FALKOR_ERRATUM_1009 · 507 unmodified lines
508	508
509	509	If unsure, say Y.
510	510
	511	+config QCOM_QDF2400_ERRATUM_0065
	512	+ bool "QDF2400 E0065: Incorrect GITS_TYPER.ITT_Entry_size"
	513	+ default y
	514	+ help
	515	+ On Qualcomm Datacenter Technologies QDF2400 SoC, ITS hardware reports
	516	+ ITE size incorrectly. The GITS_TYPER.ITT_Entry_size field should have
	517	+ been indicated as 16Bytes (0xf), not 8Bytes (0x7).
	518	+
	519	+ If unsure, say Y.
	520	+
511	521	endmenu
512	522
513	523

arch/arm64/include/asm/kvm_host.h+1−2

@@ -30,8 +30,7 @@ · 29 unmodified lines
30	30
31	31	#define __KVM_HAVE_ARCH_INTC_INITIALIZED
32	32
33		−#define KVM_USER_MEM_SLOTS 32
34		−#define KVM_PRIVATE_MEM_SLOTS 4
	33	+#define KVM_USER_MEM_SLOTS 512
35	34	#define KVM_COALESCED_MMIO_PAGE_OFFSET 1
36	35	#define KVM_HALT_POLL_NS_DEFAULT 500000
37	36

arch/arm64/include/asm/pgtable-types.h+4−0

@@ -55,9 +55,13 @@ typedef struct { pteval_t pgprot; } pgprot_t; · 54 unmodified lines
55	55	#define __pgprot(x) ((pgprot_t) { (x) } )
56	56
57	57	#if CONFIG_PGTABLE_LEVELS == 2
	58	+#define __ARCH_USE_5LEVEL_HACK
58	59	#include <asm-generic/pgtable-nopmd.h>
59	60	#elif CONFIG_PGTABLE_LEVELS == 3
	61	+#define __ARCH_USE_5LEVEL_HACK
60	62	#include <asm-generic/pgtable-nopud.h>
	63	+#elif CONFIG_PGTABLE_LEVELS == 4
	64	+#include <asm-generic/5level-fixup.h>
61	65	#endif
62	66
63	67	#endif /* __ASM_PGTABLE_TYPES_H */

arch/arm64/kvm/handle_exit.c+12−7

@@ -135,7 +135,19 @@ static int kvm_handle_guest_debug(struct kvm_vcpu vcpu, struct kvm_run run) · 134 unmodified lines
135	135	return ret;
136	136	}
137	137
	138	+static int kvm_handle_unknown_ec(struct kvm_vcpu vcpu, struct kvm_run run)
	139	+{
	140	+ u32 hsr = kvm_vcpu_get_hsr(vcpu);
	141	+
	142	+ kvm_pr_unimpl("Unknown exception class: hsr: %#08x -- %s\n",
	143	+ hsr, esr_get_class_string(hsr));
	144	+
	145	+ kvm_inject_undefined(vcpu);
	146	+ return 1;
	147	+}
	148	+
138	149	static exit_handle_fn arm_exit_handlers[] = {
	150	+ [0 ... ESR_ELx_EC_MAX] = kvm_handle_unknown_ec,
139	151	[ESR_ELx_EC_WFx] = kvm_handle_wfx,
140	152	[ESR_ELx_EC_CP15_32] = kvm_handle_cp15_32,
141	153	[ESR_ELx_EC_CP15_64] = kvm_handle_cp15_64,
@@ -162,13 +174,6 @@ static exit_handle_fn kvm_get_exit_handler(struct kvm_vcpu *vcpu) · 20 unmodified lines
162	174	u32 hsr = kvm_vcpu_get_hsr(vcpu);
163	175	u8 hsr_ec = ESR_ELx_EC(hsr);
164	176
165		− if (hsr_ec >= ARRAY_SIZE(arm_exit_handlers) \|\|
166		− !arm_exit_handlers[hsr_ec]) {
167		− kvm_err("Unknown exception class: hsr: %#08x -- %s\n",
168		− hsr, esr_get_class_string(hsr));
169		− BUG();
170		− }
171		−
172	177	return arm_exit_handlers[hsr_ec];
173	178	}
174	179

arch/arm64/kvm/hyp/tlb.c+55−9

@@ -18,14 +18,62 @@ · 17 unmodified lines
18	18	#include <asm/kvm_hyp.h>
19	19	#include <asm/tlbflush.h>
20	20
	21	+static void __hyp_text __tlb_switch_to_guest_vhe(struct kvm *kvm)
	22	+{
	23	+ u64 val;
	24	+
	25	+ /*
	26	+ * With VHE enabled, we have HCR_EL2.{E2H,TGE} = {1,1}, and
	27	+ * most TLB operations target EL2/EL0. In order to affect the
	28	+ * guest TLBs (EL1/EL0), we need to change one of these two
	29	+ * bits. Changing E2H is impossible (goodbye TTBR1_EL2), so
	30	+ * let's flip TGE before executing the TLB operation.
	31	+ */
	32	+ write_sysreg(kvm->arch.vttbr, vttbr_el2);
	33	+ val = read_sysreg(hcr_el2);
	34	+ val &= ~HCR_TGE;
	35	+ write_sysreg(val, hcr_el2);
	36	+ isb();
	37	+}
	38	+
	39	+static void __hyp_text __tlb_switch_to_guest_nvhe(struct kvm *kvm)
	40	+{
	41	+ write_sysreg(kvm->arch.vttbr, vttbr_el2);
	42	+ isb();
	43	+}
	44	+
	45	+static hyp_alternate_select(__tlb_switch_to_guest,
	46	+ __tlb_switch_to_guest_nvhe,
	47	+ __tlb_switch_to_guest_vhe,
	48	+ ARM64_HAS_VIRT_HOST_EXTN);
	49	+
	50	+static void __hyp_text __tlb_switch_to_host_vhe(struct kvm *kvm)
	51	+{
	52	+ /*
	53	+ * We're done with the TLB operation, let's restore the host's
	54	+ * view of HCR_EL2.
	55	+ */
	56	+ write_sysreg(0, vttbr_el2);
	57	+ write_sysreg(HCR_HOST_VHE_FLAGS, hcr_el2);
	58	+}
	59	+
	60	+static void __hyp_text __tlb_switch_to_host_nvhe(struct kvm *kvm)
	61	+{
	62	+ write_sysreg(0, vttbr_el2);
	63	+}
	64	+
	65	+static hyp_alternate_select(__tlb_switch_to_host,
	66	+ __tlb_switch_to_host_nvhe,
	67	+ __tlb_switch_to_host_vhe,
	68	+ ARM64_HAS_VIRT_HOST_EXTN);
	69	+
21	70	void __hyp_text __kvm_tlb_flush_vmid_ipa(struct kvm *kvm, phys_addr_t ipa)
22	71	{
23	72	dsb(ishst);
24	73
25	74	/* Switch to requested VMID */
26	75	kvm = kern_hyp_va(kvm);
27		− write_sysreg(kvm->arch.vttbr, vttbr_el2);
28		− isb();
	76	+ __tlb_switch_to_guest()(kvm);
29	77
30	78	/*
31	79	* We could do so much better if we had the VA as well.
@@ -46,7 +94,7 @@ void __hyp_text __kvm_tlb_flush_vmid_ipa(struct kvm *kvm, phys_addr_t ipa) · 14 unmodified lines
46	94	dsb(ish);
47	95	isb();
48	96
49		− write_sysreg(0, vttbr_el2);
	97	+ __tlb_switch_to_host()(kvm);
50	98	}
51	99
52	100	void __hyp_text __kvm_tlb_flush_vmid(struct kvm *kvm)
@@ -55,14 +103,13 @@ void __hyp_text __kvm_tlb_flush_vmid(struct kvm *kvm) · 2 unmodified lines
55	103
56	104	/* Switch to requested VMID */
57	105	kvm = kern_hyp_va(kvm);
58		− write_sysreg(kvm->arch.vttbr, vttbr_el2);
59		− isb();
	106	+ __tlb_switch_to_guest()(kvm);
60	107
61	108	__tlbi(vmalls12e1is);
62	109	dsb(ish);
63	110	isb();
64	111
65		− write_sysreg(0, vttbr_el2);
	112	+ __tlb_switch_to_host()(kvm);
66	113	}
67	114
68	115	void __hyp_text __kvm_tlb_flush_local_vmid(struct kvm_vcpu *vcpu)
@@ -70,14 +117,13 @@ void __hyp_text __kvm_tlb_flush_local_vmid(struct kvm_vcpu *vcpu) · 1 unmodified line
70	117	struct kvm *kvm = kern_hyp_va(kern_hyp_va(vcpu)->kvm);
71	118
72	119	/* Switch to requested VMID */
73		− write_sysreg(kvm->arch.vttbr, vttbr_el2);
74		− isb();
	120	+ __tlb_switch_to_guest()(kvm);
75	121
76	122	__tlbi(vmalle1);
77	123	dsb(nsh);
78	124	isb();
79	125
80		− write_sysreg(0, vttbr_el2);
	126	+ __tlb_switch_to_host()(kvm);
81	127	}
82	128
83	129	void __hyp_text __kvm_flush_vm_context(void)

arch/avr32/include/asm/pgtable-2level.h+1−0

@@ -8,6 +8,7 @@ · 7 unmodified lines
8	8	#ifndef __ASM_AVR32_PGTABLE_2LEVEL_H
9	9	#define __ASM_AVR32_PGTABLE_2LEVEL_H
10	10
	11	+#define __ARCH_USE_5LEVEL_HACK
11	12	#include <asm-generic/pgtable-nopmd.h>
12	13
13	14	/*

arch/avr32/oprofile/backtrace.c+1−1

@@ -14,7 +14,7 @@ · 13 unmodified lines
14	14	*/
15	15
16	16	#include <linux/oprofile.h>
17		−#include <linux/sched.h>
	17	+#include <linux/ptrace.h>
18	18	#include <linux/uaccess.h>
19	19
20	20	/* The first two words of each frame on the stack look like this if we have

arch/cris/arch-v32/drivers/cryptocop.c+1−1

@@ -2086,7 +2086,7 @@ static void cryptocop_job_queue_close(void) · 2085 unmodified lines
2086	2086	dma_in_cfg.en = regk_dma_no;
2087	2087	REG_WR(dma, IN_DMA_INST, rw_cfg, dma_in_cfg);
2088	2088
2089		− /* Disble the cryptocop. */
	2089	+ /* Disable the cryptocop. */
2090	2090	rw_cfg = REG_RD(strcop, regi_strcop, rw_cfg);
2091	2091	rw_cfg.en = 0;
2092	2092	REG_WR(strcop, regi_strcop, rw_cfg, rw_cfg);

arch/cris/include/asm/pgtable.h+1−0

@@ -6,6 +6,7 @@ · 5 unmodified lines
6	6	#define _CRIS_PGTABLE_H
7	7
8	8	#include <asm/page.h>
	9	+#define __ARCH_USE_5LEVEL_HACK
9	10	#include <asm-generic/pgtable-nopmd.h>
10	11
11	12	#ifndef __ASSEMBLY__

arch/frv/include/asm/pgtable.h+1−0

@@ -16,6 +16,7 @@ · 15 unmodified lines
16	16	#ifndef _ASM_PGTABLE_H
17	17	#define _ASM_PGTABLE_H
18	18
	19	+#include <asm-generic/5level-fixup.h>
19	20	#include <asm/mem-layout.h>
20	21	#include <asm/setup.h>
21	22	#include <asm/processor.h>

arch/h8300/include/asm/pgtable.h+1−0

@@ -1,5 +1,6 @@
1	1	#ifndef _H8300_PGTABLE_H
2	2	#define _H8300_PGTABLE_H
	3	+#define __ARCH_USE_5LEVEL_HACK
3	4	#include <asm-generic/pgtable-nopud.h>
4	5	#include <asm-generic/pgtable.h>
5	6	#define pgtable_cache_init() do { } while (0)

arch/h8300/kernel/ptrace_h.c+1−1

@@ -9,7 +9,7 @@ · 8 unmodified lines
9	9	*/
10	10
11	11	#include <linux/linkage.h>
12		−#include <linux/sched.h>
	12	+#include <linux/sched/signal.h>
13	13	#include <asm/ptrace.h>
14	14
15	15	#define BREAKINST 0x5730 /* trapa #3 */

arch/hexagon/include/asm/pgtable.h+1−0

@@ -26,6 +26,7 @@ · 25 unmodified lines
26	26	*/
27	27	#include <linux/swap.h>
28	28	#include <asm/page.h>
	29	+#define __ARCH_USE_5LEVEL_HACK
29	30	#include <asm-generic/pgtable-nopmd.h>
30	31
31	32	/* A handy thing to have if one has the RAM. Declared in head.S */

arch/ia64/include/asm/pgtable.h+2−0

@@ -587,8 +587,10 @@ extern struct page *zero_page_memmap_ptr; · 586 unmodified lines
587	587
588	588
589	589	#if CONFIG_PGTABLE_LEVELS == 3
	590	+#define __ARCH_USE_5LEVEL_HACK
590	591	#include <asm-generic/pgtable-nopud.h>
591	592	#endif
	593	+#include <asm-generic/5level-fixup.h>
592	594	#include <asm-generic/pgtable.h>
593	595
594	596	#endif /* _ASM_IA64_PGTABLE_H */

arch/metag/include/asm/pgtable.h+1−0

@@ -6,6 +6,7 @@ · 5 unmodified lines
6	6	#define _METAG_PGTABLE_H
7	7
8	8	#include <asm/pgtable-bits.h>
	9	+#define __ARCH_USE_5LEVEL_HACK
9	10	#include <asm-generic/pgtable-nopmd.h>
10	11
11	12	/* Invalid regions on Meta: 0x00000000-0x001FFFFF and 0xFFFF0000-0xFFFFFFFF */

arch/microblaze/include/asm/page.h+2−1

@@ -95,7 +95,8 @@ typedef struct { unsigned long pgd; } pgd_t; · 94 unmodified lines
95	95	# else /* CONFIG_MMU */
96	96	typedef struct { unsigned long ste[64]; } pmd_t;
97	97	typedef struct { pmd_t pue[1]; } pud_t;
98		−typedef struct { pud_t pge[1]; } pgd_t;
	98	+typedef struct { pud_t p4e[1]; } p4d_t;
	99	+typedef struct { p4d_t pge[1]; } pgd_t;
99	100	# endif /* CONFIG_MMU */
100	101
101	102	# define pte_val(x) ((x).pte)

arch/mips/cavium-octeon/cpu.c+2−0

@@ -10,7 +10,9 @@ · 9 unmodified lines
10	10	#include <linux/irqflags.h>
11	11	#include <linux/notifier.h>
12	12	#include <linux/prefetch.h>
	13	+#include <linux/ptrace.h>
13	14	#include <linux/sched.h>
	15	+#include <linux/sched/task_stack.h>
14	16
15	17	#include <asm/cop2.h>
16	18	#include <asm/current.h>

arch/mips/cavium-octeon/crypto/octeon-crypto.c+1−0

@@ -9,6 +9,7 @@ · 8 unmodified lines
9	9	#include <asm/cop2.h>
10	10	#include <linux/export.h>
11	11	#include <linux/interrupt.h>
	12	+#include <linux/sched/task_stack.h>
12	13
13	14	#include "octeon-crypto.h"
14	15

arch/mips/cavium-octeon/smp.c+1−0

@@ -12,6 +12,7 @@ · 11 unmodified lines
12	12	#include <linux/kernel_stat.h>
13	13	#include <linux/sched.h>
14	14	#include <linux/sched/hotplug.h>
	15	+#include <linux/sched/task_stack.h>
15	16	#include <linux/init.h>
16	17	#include <linux/export.h>
17	18

arch/mips/include/asm/fpu.h+1−0

@@ -12,6 +12,7 @@ · 11 unmodified lines
12	12
13	13	#include <linux/sched.h>
14	14	#include <linux/sched/task_stack.h>
	15	+#include <linux/ptrace.h>
15	16	#include <linux/thread_info.h>
16	17	#include <linux/bitops.h>
17	18

arch/mips/include/asm/pgtable-32.h+1−0

@@ -16,6 +16,7 @@ · 15 unmodified lines
16	16	#include <asm/cachectl.h>
17	17	#include <asm/fixmap.h>
18	18
	19	+#define __ARCH_USE_5LEVEL_HACK
19	20	#include <asm-generic/pgtable-nopmd.h>
20	21
21	22	extern int temp_tlb_entry;

arch/mips/include/asm/pgtable-64.h+1−0

@@ -17,6 +17,7 @@ · 16 unmodified lines
17	17	#include <asm/cachectl.h>
18	18	#include <asm/fixmap.h>
19	19
	20	+#define __ARCH_USE_5LEVEL_HACK
20	21	#if defined(CONFIG_PAGE_SIZE_64KB) && !defined(CONFIG_MIPS_VA_BITS_48)
21	22	#include <asm-generic/pgtable-nopmd.h>
22	23	#else

arch/mips/kernel/smp-bmips.c+1−0

@@ -11,6 +11,7 @@ · 10 unmodified lines
11	11	#include <linux/init.h>
12	12	#include <linux/sched.h>
13	13	#include <linux/sched/hotplug.h>
	14	+#include <linux/sched/task_stack.h>
14	15	#include <linux/mm.h>
15	16	#include <linux/delay.h>
16	17	#include <linux/smp.h>

arch/mips/kernel/smp-mt.c+1−0

@@ -23,6 +23,7 @@ · 22 unmodified lines
23	23	#include <linux/interrupt.h>
24	24	#include <linux/irqchip/mips-gic.h>
25	25	#include <linux/compiler.h>
	26	+#include <linux/sched/task_stack.h>
26	27	#include <linux/smp.h>
27	28
28	29	#include <linux/atomic.h>

arch/mips/loongson64/loongson-3/cop2-ex.c+1−0

@@ -13,6 +13,7 @@ · 12 unmodified lines
13	13	#include <linux/init.h>
14	14	#include <linux/sched.h>
15	15	#include <linux/notifier.h>
	16	+#include <linux/ptrace.h>
16	17
17	18	#include <asm/fpu.h>
18	19	#include <asm/cop2.h>

arch/mips/netlogic/common/smp.c+1−0

@@ -35,6 +35,7 @@ · 34 unmodified lines
35	35	#include <linux/kernel.h>
36	36	#include <linux/delay.h>
37	37	#include <linux/init.h>
	38	+#include <linux/sched/task_stack.h>
38	39	#include <linux/smp.h>
39	40	#include <linux/irq.h>
40	41

arch/mips/netlogic/xlp/cop2-ex.c+3−0

@@ -9,11 +9,14 @@ · 8 unmodified lines
9	9	* Copyright (C) 2009 Wind River Systems,
10	10	* written by Ralf Baechle <ralf@linux-mips.org>
11	11	*/
	12	+#include <linux/capability.h>
12	13	#include <linux/init.h>
13	14	#include <linux/irqflags.h>
14	15	#include <linux/notifier.h>
15	16	#include <linux/prefetch.h>
	17	+#include <linux/ptrace.h>
16	18	#include <linux/sched.h>
	19	+#include <linux/sched/task_stack.h>
17	20
18	21	#include <asm/cop2.h>
19	22	#include <asm/current.h>

arch/mips/sgi-ip22/ip28-berr.c+1−0

@@ -9,6 +9,7 @@ · 8 unmodified lines
9	9	#include <linux/kernel.h>
10	10	#include <linux/sched.h>
11	11	#include <linux/sched/debug.h>
	12	+#include <linux/sched/signal.h>
12	13	#include <linux/seq_file.h>
13	14
14	15	#include <asm/addrspace.h>

arch/mips/sgi-ip27/ip27-berr.c+2−0

@@ -12,7 +12,9 @@ · 11 unmodified lines
12	12	#include <linux/signal.h> /* for SIGBUS */
13	13	#include <linux/sched.h> /* schow_regs(), force_sig() */
14	14	#include <linux/sched/debug.h>
	15	+#include <linux/sched/signal.h>
15	16
	17	+#include <asm/ptrace.h>
16	18	#include <asm/sn/addrs.h>
17	19	#include <asm/sn/arch.h>
18	20	#include <asm/sn/sn0/hub.h>

arch/mips/sgi-ip27/ip27-smp.c+3−0

@@ -8,10 +8,13 @@ · 7 unmodified lines
8	8	*/
9	9	#include <linux/init.h>
10	10	#include <linux/sched.h>
	11	+#include <linux/sched/task_stack.h>
11	12	#include <linux/topology.h>
12	13	#include <linux/nodemask.h>
	14	+
13	15	#include <asm/page.h>
14	16	#include <asm/processor.h>
	17	+#include <asm/ptrace.h>
15	18	#include <asm/sn/arch.h>
16	19	#include <asm/sn/gda.h>
17	20	#include <asm/sn/intr.h>

arch/mips/sgi-ip32/ip32-berr.c+1−0

@@ -11,6 +11,7 @@ · 10 unmodified lines
11	11	#include <linux/kernel.h>
12	12	#include <linux/sched.h>
13	13	#include <linux/sched/debug.h>
	14	+#include <linux/sched/signal.h>
14	15	#include <asm/traps.h>
15	16	#include <linux/uaccess.h>
16	17	#include <asm/addrspace.h>

arch/mips/sgi-ip32/ip32-reset.c+1−0

@@ -13,6 +13,7 @@ · 12 unmodified lines
13	13	#include <linux/kernel.h>
14	14	#include <linux/module.h>
15	15	#include <linux/sched.h>
	16	+#include <linux/sched/signal.h>
16	17	#include <linux/notifier.h>
17	18	#include <linux/delay.h>
18	19	#include <linux/rtc/ds1685.h>

arch/mn10300/include/asm/page.h+1−0

@@ -57,6 +57,7 @@ typedef struct page *pgtable_t; · 56 unmodified lines
57	57	#define __pgd(x) ((pgd_t) { (x) })
58	58	#define __pgprot(x) ((pgprot_t) { (x) })
59	59
	60	+#define __ARCH_USE_5LEVEL_HACK
60	61	#include <asm-generic/pgtable-nopmd.h>
61	62
62	63	#endif /* !__ASSEMBLY__ */

arch/nios2/include/asm/pgtable.h+1−0

@@ -22,6 +22,7 @@ · 21 unmodified lines
22	22	#include <asm/tlbflush.h>
23	23
24	24	#include <asm/pgtable-bits.h>
	25	+#define __ARCH_USE_5LEVEL_HACK
25	26	#include <asm-generic/pgtable-nopmd.h>
26	27
27	28	#define FIRST_USER_ADDRESS 0UL

arch/openrisc/include/asm/pgtable.h+1−0

@@ -25,6 +25,7 @@ · 24 unmodified lines
25	25	#ifndef __ASM_OPENRISC_PGTABLE_H
26	26	#define __ASM_OPENRISC_PGTABLE_H
27	27
	28	+#define __ARCH_USE_5LEVEL_HACK
28	29	#include <asm-generic/pgtable-nopmd.h>
29	30
30	31	#ifndef __ASSEMBLY__

arch/powerpc/Kconfig+72−66

@@ -80,93 +80,99 @@ config ARCH_HAS_DMA_SET_COHERENT_MASK · 79 unmodified lines
80	80	config PPC
81	81	bool
82	82	default y
83		− select BUILDTIME_EXTABLE_SORT
	83	+ #
	84	+ # Please keep this list sorted alphabetically.
	85	+ #
	86	+ select ARCH_HAS_DEVMEM_IS_ALLOWED
	87	+ select ARCH_HAS_DMA_SET_COHERENT_MASK
	88	+ select ARCH_HAS_ELF_RANDOMIZE
	89	+ select ARCH_HAS_GCOV_PROFILE_ALL
	90	+ select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE
	91	+ select ARCH_HAS_SG_CHAIN
	92	+ select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
	93	+ select ARCH_HAS_UBSAN_SANITIZE_ALL
	94	+ select ARCH_HAVE_NMI_SAFE_CMPXCHG
84	95	select ARCH_MIGHT_HAVE_PC_PARPORT
85	96	select ARCH_MIGHT_HAVE_PC_SERIO
	97	+ select ARCH_SUPPORTS_ATOMIC_RMW
	98	+ select ARCH_SUPPORTS_DEFERRED_STRUCT_PAGE_INIT
	99	+ select ARCH_USE_BUILTIN_BSWAP
	100	+ select ARCH_USE_CMPXCHG_LOCKREF if PPC64
	101	+ select ARCH_WANT_IPC_PARSE_VERSION
86	102	select BINFMT_ELF
87		− select ARCH_HAS_ELF_RANDOMIZE
88		− select OF
89		− select OF_EARLY_FLATTREE
90		− select OF_RESERVED_MEM
91		− select HAVE_FTRACE_MCOUNT_RECORD
	103	+ select BUILDTIME_EXTABLE_SORT
	104	+ select CLONE_BACKWARDS
	105	+ select DCACHE_WORD_ACCESS if PPC64 && CPU_LITTLE_ENDIAN
	106	+ select EDAC_ATOMIC_SCRUB
	107	+ select EDAC_SUPPORT
	108	+ select GENERIC_ATOMIC64 if PPC32
	109	+ select GENERIC_CLOCKEVENTS
	110	+ select GENERIC_CLOCKEVENTS_BROADCAST if SMP
	111	+ select GENERIC_CMOS_UPDATE
	112	+ select GENERIC_CPU_AUTOPROBE
	113	+ select GENERIC_IRQ_SHOW
	114	+ select GENERIC_IRQ_SHOW_LEVEL
	115	+ select GENERIC_SMP_IDLE_THREAD
	116	+ select GENERIC_STRNCPY_FROM_USER
	117	+ select GENERIC_STRNLEN_USER
	118	+ select GENERIC_TIME_VSYSCALL_OLD
	119	+ select HAVE_ARCH_AUDITSYSCALL
	120	+ select HAVE_ARCH_HARDENED_USERCOPY
	121	+ select HAVE_ARCH_JUMP_LABEL
	122	+ select HAVE_ARCH_KGDB
	123	+ select HAVE_ARCH_SECCOMP_FILTER
	124	+ select HAVE_ARCH_TRACEHOOK
	125	+ select HAVE_CBPF_JIT if !PPC64
	126	+ select HAVE_CONTEXT_TRACKING if PPC64
	127	+ select HAVE_DEBUG_KMEMLEAK
	128	+ select HAVE_DEBUG_STACKOVERFLOW
	129	+ select HAVE_DMA_API_DEBUG
92	130	select HAVE_DYNAMIC_FTRACE
93		− select HAVE_DYNAMIC_FTRACE_WITH_REGS if MPROFILE_KERNEL
94		− select HAVE_FUNCTION_TRACER
	131	+ select HAVE_DYNAMIC_FTRACE_WITH_REGS if MPROFILE_KERNEL
	132	+ select HAVE_EBPF_JIT if PPC64
	133	+ select HAVE_EFFICIENT_UNALIGNED_ACCESS if !(CPU_LITTLE_ENDIAN && POWER7_CPU)
	134	+ select HAVE_FTRACE_MCOUNT_RECORD
95	135	select HAVE_FUNCTION_GRAPH_TRACER
	136	+ select HAVE_FUNCTION_TRACER
96	137	select HAVE_GCC_PLUGINS
97		− select SYSCTL_EXCEPTION_TRACE
98		− select VIRT_TO_BUS if !PPC64
	138	+ select HAVE_GENERIC_RCU_GUP
	139	+ select HAVE_HW_BREAKPOINT if PERF_EVENTS && (PPC_BOOK3S \|\| PPC_8xx)
99	140	select HAVE_IDE
100	141	select HAVE_IOREMAP_PROT
101		− select HAVE_EFFICIENT_UNALIGNED_ACCESS if !(CPU_LITTLE_ENDIAN && POWER7_CPU)
	142	+ select HAVE_IRQ_EXIT_ON_IRQ_STACK
	143	+ select HAVE_KERNEL_GZIP
102	144	select HAVE_KPROBES
103		− select HAVE_OPTPROBES if PPC64
104		− select HAVE_ARCH_KGDB
105	145	select HAVE_KRETPROBES
106		− select HAVE_ARCH_TRACEHOOK
	146	+ select HAVE_LIVEPATCH if HAVE_DYNAMIC_FTRACE_WITH_REGS
107	147	select HAVE_MEMBLOCK
108	148	select HAVE_MEMBLOCK_NODE_MAP
109		− select HAVE_DMA_API_DEBUG
	149	+ select HAVE_MOD_ARCH_SPECIFIC
	150	+ select HAVE_NMI if PERF_EVENTS
110	151	select HAVE_OPROFILE
111		− select HAVE_DEBUG_KMEMLEAK
112		− select ARCH_HAS_SG_CHAIN
113		− select GENERIC_ATOMIC64 if PPC32
	152	+ select HAVE_OPTPROBES if PPC64
114	153	select HAVE_PERF_EVENTS
	154	+ select HAVE_PERF_EVENTS_NMI if PPC64
115	155	select HAVE_PERF_REGS
116	156	select HAVE_PERF_USER_STACK_DUMP
	157	+ select HAVE_RCU_TABLE_FREE if SMP
117	158	select HAVE_REGS_AND_STACK_ACCESS_API
118		− select HAVE_HW_BREAKPOINT if PERF_EVENTS && (PPC_BOOK3S \|\| PPC_8xx)
119		− select ARCH_WANT_IPC_PARSE_VERSION
120		− select SPARSE_IRQ
	159	+ select HAVE_SYSCALL_TRACEPOINTS
	160	+ select HAVE_VIRT_CPU_ACCOUNTING
121	161	select IRQ_DOMAIN
122		− select GENERIC_IRQ_SHOW
123		− select GENERIC_IRQ_SHOW_LEVEL
124	162	select IRQ_FORCED_THREADING
125		− select HAVE_RCU_TABLE_FREE if SMP
126		− select HAVE_SYSCALL_TRACEPOINTS
127		− select HAVE_CBPF_JIT if !PPC64
128		− select HAVE_EBPF_JIT if PPC64
129		− select HAVE_ARCH_JUMP_LABEL
130		− select ARCH_HAVE_NMI_SAFE_CMPXCHG
131		− select ARCH_HAS_GCOV_PROFILE_ALL
132		− select GENERIC_SMP_IDLE_THREAD
133		− select GENERIC_CMOS_UPDATE
134		− select GENERIC_TIME_VSYSCALL_OLD
135		− select GENERIC_CLOCKEVENTS
136		− select GENERIC_CLOCKEVENTS_BROADCAST if SMP
137		− select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
138		− select GENERIC_STRNCPY_FROM_USER
139		− select GENERIC_STRNLEN_USER
140		− select HAVE_MOD_ARCH_SPECIFIC
141	163	select MODULES_USE_ELF_RELA
142		− select CLONE_BACKWARDS
143		− select ARCH_USE_BUILTIN_BSWAP
144		− select OLD_SIGSUSPEND
145		− select OLD_SIGACTION if PPC32
146		− select HAVE_DEBUG_STACKOVERFLOW
147		− select HAVE_IRQ_EXIT_ON_IRQ_STACK
148		− select ARCH_USE_CMPXCHG_LOCKREF if PPC64
149		− select HAVE_ARCH_AUDITSYSCALL
150		− select ARCH_SUPPORTS_ATOMIC_RMW
151		− select DCACHE_WORD_ACCESS if PPC64 && CPU_LITTLE_ENDIAN
152	164	select NO_BOOTMEM
153		− select HAVE_GENERIC_RCU_GUP
154		− select HAVE_PERF_EVENTS_NMI if PPC64
155		− select HAVE_NMI if PERF_EVENTS
156		− select EDAC_SUPPORT
157		− select EDAC_ATOMIC_SCRUB
158		− select ARCH_HAS_DMA_SET_COHERENT_MASK
159		− select ARCH_HAS_DEVMEM_IS_ALLOWED
160		− select HAVE_ARCH_SECCOMP_FILTER
161		− select ARCH_HAS_UBSAN_SANITIZE_ALL
162		− select ARCH_SUPPORTS_DEFERRED_STRUCT_PAGE_INIT
163		− select HAVE_LIVEPATCH if HAVE_DYNAMIC_FTRACE_WITH_REGS
164		− select GENERIC_CPU_AUTOPROBE
165		− select HAVE_VIRT_CPU_ACCOUNTING
166		− select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE
167		− select HAVE_ARCH_HARDENED_USERCOPY
168		− select HAVE_KERNEL_GZIP
169		− select HAVE_CONTEXT_TRACKING if PPC64
	165	+ select OF
	166	+ select OF_EARLY_FLATTREE
	167	+ select OF_RESERVED_MEM
	168	+ select OLD_SIGACTION if PPC32
	169	+ select OLD_SIGSUSPEND
	170	+ select SPARSE_IRQ
	171	+ select SYSCTL_EXCEPTION_TRACE
	172	+ select VIRT_TO_BUS if !PPC64
	173	+ #
	174	+ # Please keep this list sorted alphabetically.
	175	+ #
170	176
171	177	config GENERIC_CSUM
172	178	def_bool n

arch/powerpc/Makefile+10−1

@@ -72,8 +72,15 @@ GNUTARGET := powerpc · 71 unmodified lines
72	72	MULTIPLEWORD := -mmultiple
73	73	endif
74	74
75		−cflags-$(CONFIG_CPU_BIG_ENDIAN) += $(call cc-option,-mbig-endian)
	75	+ifdef CONFIG_PPC64
	76	+cflags-$(CONFIG_CPU_BIG_ENDIAN) += $(call cc-option,-mabi=elfv1)
	77	+cflags-$(CONFIG_CPU_BIG_ENDIAN) += $(call cc-option,-mcall-aixdesc)
	78	+aflags-$(CONFIG_CPU_BIG_ENDIAN) += $(call cc-option,-mabi=elfv1)
	79	+aflags-$(CONFIG_CPU_LITTLE_ENDIAN) += -mabi=elfv2
	80	+endif
	81	+
76	82	cflags-$(CONFIG_CPU_LITTLE_ENDIAN) += -mlittle-endian
	83	+cflags-$(CONFIG_CPU_BIG_ENDIAN) += $(call cc-option,-mbig-endian)
77	84	ifneq ($(cc-name),clang)
78	85	cflags-$(CONFIG_CPU_LITTLE_ENDIAN) += -mno-strict-align
79	86	endif
@@ -113,7 +120,9 @@ ifeq ($(CONFIG_CPU_LITTLE_ENDIAN),y) · 33 unmodified lines
113	120	CFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mabi=elfv2,$(call cc-option,-mcall-aixdesc))
114	121	AFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mabi=elfv2)
115	122	else
	123	+CFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mabi=elfv1)
116	124	CFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mcall-aixdesc)
	125	+AFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mabi=elfv1)
117	126	endif
118	127	CFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mcmodel=medium,$(call cc-option,-mminimal-toc))
119	128	CFLAGS-$(CONFIG_PPC64) += $(call cc-option,-mno-pointers-to-nested-functions)

arch/powerpc/boot/zImage.lds.S+1−0

@@ -68,6 +68,7 @@ SECTIONS · 67 unmodified lines
68	68	}
69	69
70	70	#ifdef CONFIG_PPC64_BOOT_WRAPPER
	71	+ . = ALIGN(256);
71	72	.got :
72	73	{
73	74	__toc_start = .;

arch/powerpc/crypto/crc32c-vpmsum_glue.c+1−1

@@ -52,7 +52,7 @@ static int crc32c_vpmsum_cra_init(struct crypto_tfm *tfm) · 51 unmodified lines
52	52	{
53	53	u32 *key = crypto_tfm_ctx(tfm);
54	54
55		− *key = 0;
	55	+ *key = ~0;
56	56
57	57	return 0;
58	58	}

arch/powerpc/include/asm/bitops.h+4−0

@@ -51,6 +51,10 @@ · 50 unmodified lines
51	51	#define PPC_BIT(bit) (1UL << PPC_BITLSHIFT(bit))
52	52	#define PPC_BITMASK(bs, be) ((PPC_BIT(bs) - PPC_BIT(be)) \| PPC_BIT(bs))
53	53
	54	+/* Put a PPC bit into a "normal" bit position */
	55	+#define PPC_BITEXTRACT(bits, ppc_bit, dst_bit) \
	56	+ ((((bits) >> PPC_BITLSHIFT(ppc_bit)) & 1) << (dst_bit))
	57	+
54	58	#include <asm/barrier.h>
55	59
56	60	/* Macro for generating the **_bits() functions /

arch/powerpc/include/asm/book3s/32/pgtable.h+1−0

@@ -1,6 +1,7 @@
1	1	#ifndef _ASM_POWERPC_BOOK3S_32_PGTABLE_H
2	2	#define _ASM_POWERPC_BOOK3S_32_PGTABLE_H
3	3
	4	+#define __ARCH_USE_5LEVEL_HACK
4	5	#include <asm-generic/pgtable-nopmd.h>
5	6
6	7	#include <asm/book3s/32/hash.h>

arch/powerpc/include/asm/book3s/64/pgtable.h+58−30

@@ -1,9 +1,12 @@
1	1	#ifndef _ASM_POWERPC_BOOK3S_64_PGTABLE_H_
2	2	#define _ASM_POWERPC_BOOK3S_64_PGTABLE_H_
3	3
	4	+#include <asm-generic/5level-fixup.h>
	5	+
4	6	#ifndef __ASSEMBLY__
5	7	#include <linux/mmdebug.h>
6	8	#endif
	9	+
7	10	/*
8	11	* Common bits between hash and Radix page table
9	12	*/
@@ -347,23 +350,58 @@ static inline int __ptep_test_and_clear_young(struct mm_struct *mm, · 337 unmodified lines
347	350	__r; \
348	351	})
349	352
	353	+static inline int __pte_write(pte_t pte)
	354	+{
	355	+ return !!(pte_raw(pte) & cpu_to_be64(_PAGE_WRITE));
	356	+}
	357	+
	358	+#ifdef CONFIG_NUMA_BALANCING
	359	+#define pte_savedwrite pte_savedwrite
	360	+static inline bool pte_savedwrite(pte_t pte)
	361	+{
	362	+ /*
	363	+ * Saved write ptes are prot none ptes that doesn't have
	364	+ * privileged bit sit. We mark prot none as one which has
	365	+ * present and pviliged bit set and RWX cleared. To mark
	366	+ * protnone which used to have _PAGE_WRITE set we clear
	367	+ * the privileged bit.
	368	+ */
	369	+ return !(pte_raw(pte) & cpu_to_be64(_PAGE_RWX \| _PAGE_PRIVILEGED));
	370	+}
	371	+#else
	372	+#define pte_savedwrite pte_savedwrite
	373	+static inline bool pte_savedwrite(pte_t pte)
	374	+{
	375	+ return false;
	376	+}
	377	+#endif
	378	+
	379	+static inline int pte_write(pte_t pte)
	380	+{
	381	+ return __pte_write(pte) \|\| pte_savedwrite(pte);
	382	+}
	383	+
350	384	#define __HAVE_ARCH_PTEP_SET_WRPROTECT
351	385	static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr,
352	386	pte_t *ptep)
353	387	{
354		− if ((pte_raw(*ptep) & cpu_to_be64(_PAGE_WRITE)) == 0)
355		− return;
356		−
357		− pte_update(mm, addr, ptep, _PAGE_WRITE, 0, 0);
	388	+ if (__pte_write(*ptep))
	389	+ pte_update(mm, addr, ptep, _PAGE_WRITE, 0, 0);
	390	+ else if (unlikely(pte_savedwrite(*ptep)))
	391	+ pte_update(mm, addr, ptep, 0, _PAGE_PRIVILEGED, 0);
358	392	}
359	393
360	394	static inline void huge_ptep_set_wrprotect(struct mm_struct *mm,
361	395	unsigned long addr, pte_t *ptep)
362	396	{
363		− if ((pte_raw(*ptep) & cpu_to_be64(_PAGE_WRITE)) == 0)
364		− return;
365		−
366		− pte_update(mm, addr, ptep, _PAGE_WRITE, 0, 1);
	397	+ /*
	398	+ * We should not find protnone for hugetlb, but this complete the
	399	+ * interface.
	400	+ */
	401	+ if (__pte_write(*ptep))
	402	+ pte_update(mm, addr, ptep, _PAGE_WRITE, 0, 1);
	403	+ else if (unlikely(pte_savedwrite(*ptep)))
	404	+ pte_update(mm, addr, ptep, 0, _PAGE_PRIVILEGED, 1);
367	405	}
368	406
369	407	#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
@@ -397,11 +435,6 @@ static inline void pte_clear(struct mm_struct *mm, unsigned long addr, · 27 unmodified lines
397	435	pte_update(mm, addr, ptep, ~0UL, 0, 0);
398	436	}
399	437
400		−static inline int pte_write(pte_t pte)
401		−{
402		− return !!(pte_raw(pte) & cpu_to_be64(_PAGE_WRITE));
403		−}
404		−
405	438	static inline int pte_dirty(pte_t pte)
406	439	{
407	440	return !!(pte_raw(pte) & cpu_to_be64(_PAGE_DIRTY));
@@ -465,19 +498,12 @@ static inline pte_t pte_clear_savedwrite(pte_t pte) · 57 unmodified lines
465	498	VM_BUG_ON(!pte_protnone(pte));
466	499	return __pte(pte_val(pte) \| _PAGE_PRIVILEGED);
467	500	}
468		−
469		−#define pte_savedwrite pte_savedwrite
470		−static inline bool pte_savedwrite(pte_t pte)
	501	+#else
	502	+#define pte_clear_savedwrite pte_clear_savedwrite
	503	+static inline pte_t pte_clear_savedwrite(pte_t pte)
471	504	{
472		− /*
473		− * Saved write ptes are prot none ptes that doesn't have
474		− * privileged bit sit. We mark prot none as one which has
475		− * present and pviliged bit set and RWX cleared. To mark
476		− * protnone which used to have _PAGE_WRITE set we clear
477		− * the privileged bit.
478		− */
479		− VM_BUG_ON(!pte_protnone(pte));
480		− return !(pte_raw(pte) & cpu_to_be64(_PAGE_RWX \| _PAGE_PRIVILEGED));
	505	+ VM_WARN_ON(1);
	506	+ return __pte(pte_val(pte) & ~_PAGE_WRITE);
481	507	}
482	508	#endif /* CONFIG_NUMA_BALANCING */
483	509
@@ -506,6 +532,8 @@ static inline unsigned long pte_pfn(pte_t pte) · 22 unmodified lines
506	532	/* Generic modifiers for PTE bits */
507	533	static inline pte_t pte_wrprotect(pte_t pte)
508	534	{
	535	+ if (unlikely(pte_savedwrite(pte)))
	536	+ return pte_clear_savedwrite(pte);
509	537	return __pte(pte_val(pte) & ~_PAGE_WRITE);
510	538	}
511	539
@@ -926,6 +954,7 @@ static inline int pmd_protnone(pmd_t pmd) · 414 unmodified lines
926	954
927	955	#define __HAVE_ARCH_PMD_WRITE
928	956	#define pmd_write(pmd) pte_write(pmd_pte(pmd))
	957	+#define __pmd_write(pmd) __pte_write(pmd_pte(pmd))
929	958	#define pmd_savedwrite(pmd) pte_savedwrite(pmd_pte(pmd))
930	959
931	960	#ifdef CONFIG_TRANSPARENT_HUGEPAGE
@@ -982,11 +1011,10 @@ static inline int __pmdp_test_and_clear_young(struct mm_struct *mm, · 50 unmodified lines
982	1011	static inline void pmdp_set_wrprotect(struct mm_struct *mm, unsigned long addr,
983	1012	pmd_t *pmdp)
984	1013	{
985		−
986		− if ((pmd_raw(*pmdp) & cpu_to_be64(_PAGE_WRITE)) == 0)
987		− return;
988		−
989		− pmd_hugepage_update(mm, addr, pmdp, _PAGE_WRITE, 0);
	1014	+ if (__pmd_write((*pmdp)))
	1015	+ pmd_hugepage_update(mm, addr, pmdp, _PAGE_WRITE, 0);
	1016	+ else if (unlikely(pmd_savedwrite(*pmdp)))
	1017	+ pmd_hugepage_update(mm, addr, pmdp, 0, _PAGE_PRIVILEGED);
990	1018	}
991	1019
992	1020	static inline int pmd_trans_huge(pmd_t pmd)

arch/powerpc/include/asm/checksum.h+1−1

@@ -112,7 +112,7 @@ static inline __wsum csum_add(__wsum csum, __wsum addend) · 111 unmodified lines
112	112
113	113	#ifdef __powerpc64__
114	114	res += (__force u64)addend;
115		− return (__force __wsum)((u32)res + (res >> 32));
	115	+ return (__force __wsum) from64to32(res);
116	116	#else
117	117	asm("addc %0,%0,%1;"
118	118	"addze %0,%0;"

arch/powerpc/include/asm/cpuidle.h+2−2

@@ -70,8 +70,8 @@ static inline void report_invalid_psscr_val(u64 psscr_val, int err) · 69 unmodified lines
70	70	std r0,0(r1); \
71	71	ptesync; \
72	72	ld r0,0(r1); \
73		−1: cmpd cr0,r0,r0; \
74		− bne 1b; \
	73	+236: cmpd cr0,r0,r0; \
	74	+ bne 236b; \
75	75	IDLE_INST; \
76	76
77	77	#define IDLE_STATE_ENTER_SEQ_NORET(IDLE_INST) \

arch/powerpc/include/asm/elf.h+2−2

@@ -144,8 +144,8 @@ extern int arch_setup_additional_pages(struct linux_binprm *bprm, · 143 unmodified lines
144	144	#define ARCH_DLINFO_CACHE_GEOMETRY \
145	145	NEW_AUX_ENT(AT_L1I_CACHESIZE, ppc64_caches.l1i.size); \
146	146	NEW_AUX_ENT(AT_L1I_CACHEGEOMETRY, get_cache_geometry(l1i)); \
147		− NEW_AUX_ENT(AT_L1D_CACHESIZE, ppc64_caches.l1i.size); \
148		− NEW_AUX_ENT(AT_L1D_CACHEGEOMETRY, get_cache_geometry(l1i)); \
	147	+ NEW_AUX_ENT(AT_L1D_CACHESIZE, ppc64_caches.l1d.size); \
	148	+ NEW_AUX_ENT(AT_L1D_CACHEGEOMETRY, get_cache_geometry(l1d)); \
149	149	NEW_AUX_ENT(AT_L2_CACHESIZE, ppc64_caches.l2.size); \
150	150	NEW_AUX_ENT(AT_L2_CACHEGEOMETRY, get_cache_geometry(l2)); \
151	151	NEW_AUX_ENT(AT_L3_CACHESIZE, ppc64_caches.l3.size); \

arch/powerpc/include/asm/mce.h+107−1

@@ -66,6 +66,55 @@ · 65 unmodified lines
66	66
67	67	#define P8_DSISR_MC_SLB_ERRORS (P7_DSISR_MC_SLB_ERRORS \| \
68	68	P8_DSISR_MC_ERAT_MULTIHIT_SEC)
	69	+
	70	+/*
	71	+ * Machine Check bits on power9
	72	+ */
	73	+#define P9_SRR1_MC_LOADSTORE(srr1) (((srr1) >> PPC_BITLSHIFT(42)) & 1)
	74	+
	75	+#define P9_SRR1_MC_IFETCH(srr1) ( \
	76	+ PPC_BITEXTRACT(srr1, 45, 0) \| \
	77	+ PPC_BITEXTRACT(srr1, 44, 1) \| \
	78	+ PPC_BITEXTRACT(srr1, 43, 2) \| \
	79	+ PPC_BITEXTRACT(srr1, 36, 3) )
	80	+
	81	+/* 0 is reserved */
	82	+#define P9_SRR1_MC_IFETCH_UE 1
	83	+#define P9_SRR1_MC_IFETCH_SLB_PARITY 2
	84	+#define P9_SRR1_MC_IFETCH_SLB_MULTIHIT 3
	85	+#define P9_SRR1_MC_IFETCH_ERAT_MULTIHIT 4
	86	+#define P9_SRR1_MC_IFETCH_TLB_MULTIHIT 5
	87	+#define P9_SRR1_MC_IFETCH_UE_TLB_RELOAD 6
	88	+/* 7 is reserved */
	89	+#define P9_SRR1_MC_IFETCH_LINK_TIMEOUT 8
	90	+#define P9_SRR1_MC_IFETCH_LINK_TABLEWALK_TIMEOUT 9
	91	+/* 10 ? */
	92	+#define P9_SRR1_MC_IFETCH_RA 11
	93	+#define P9_SRR1_MC_IFETCH_RA_TABLEWALK 12
	94	+#define P9_SRR1_MC_IFETCH_RA_ASYNC_STORE 13
	95	+#define P9_SRR1_MC_IFETCH_LINK_ASYNC_STORE_TIMEOUT 14
	96	+#define P9_SRR1_MC_IFETCH_RA_TABLEWALK_FOREIGN 15
	97	+
	98	+/* DSISR bits for machine check (On Power9) */
	99	+#define P9_DSISR_MC_UE (PPC_BIT(48))
	100	+#define P9_DSISR_MC_UE_TABLEWALK (PPC_BIT(49))
	101	+#define P9_DSISR_MC_LINK_LOAD_TIMEOUT (PPC_BIT(50))
	102	+#define P9_DSISR_MC_LINK_TABLEWALK_TIMEOUT (PPC_BIT(51))
	103	+#define P9_DSISR_MC_ERAT_MULTIHIT (PPC_BIT(52))
	104	+#define P9_DSISR_MC_TLB_MULTIHIT_MFTLB (PPC_BIT(53))
	105	+#define P9_DSISR_MC_USER_TLBIE (PPC_BIT(54))
	106	+#define P9_DSISR_MC_SLB_PARITY_MFSLB (PPC_BIT(55))
	107	+#define P9_DSISR_MC_SLB_MULTIHIT_MFSLB (PPC_BIT(56))
	108	+#define P9_DSISR_MC_RA_LOAD (PPC_BIT(57))
	109	+#define P9_DSISR_MC_RA_TABLEWALK (PPC_BIT(58))
	110	+#define P9_DSISR_MC_RA_TABLEWALK_FOREIGN (PPC_BIT(59))
	111	+#define P9_DSISR_MC_RA_FOREIGN (PPC_BIT(60))
	112	+
	113	+/* SLB error bits */
	114	+#define P9_DSISR_MC_SLB_ERRORS (P9_DSISR_MC_ERAT_MULTIHIT \| \
	115	+ P9_DSISR_MC_SLB_PARITY_MFSLB \| \
	116	+ P9_DSISR_MC_SLB_MULTIHIT_MFSLB)
	117	+
69	118	enum MCE_Version {
70	119	MCE_V1 = 1,
71	120	};
@@ -93,6 +142,9 @@ enum MCE_ErrorType { · 21 unmodified lines
93	142	MCE_ERROR_TYPE_SLB = 2,
94	143	MCE_ERROR_TYPE_ERAT = 3,
95	144	MCE_ERROR_TYPE_TLB = 4,
	145	+ MCE_ERROR_TYPE_USER = 5,
	146	+ MCE_ERROR_TYPE_RA = 6,
	147	+ MCE_ERROR_TYPE_LINK = 7,
96	148	};
97	149
98	150	enum MCE_UeErrorType {
@@ -121,6 +173,32 @@ enum MCE_TlbErrorType { · 22 unmodified lines
121	173	MCE_TLB_ERROR_MULTIHIT = 2,
122	174	};
123	175
	176	+enum MCE_UserErrorType {
	177	+ MCE_USER_ERROR_INDETERMINATE = 0,
	178	+ MCE_USER_ERROR_TLBIE = 1,
	179	+};
	180	+
	181	+enum MCE_RaErrorType {
	182	+ MCE_RA_ERROR_INDETERMINATE = 0,
	183	+ MCE_RA_ERROR_IFETCH = 1,
	184	+ MCE_RA_ERROR_PAGE_TABLE_WALK_IFETCH = 2,
	185	+ MCE_RA_ERROR_PAGE_TABLE_WALK_IFETCH_FOREIGN = 3,
	186	+ MCE_RA_ERROR_LOAD = 4,
	187	+ MCE_RA_ERROR_STORE = 5,
	188	+ MCE_RA_ERROR_PAGE_TABLE_WALK_LOAD_STORE = 6,
	189	+ MCE_RA_ERROR_PAGE_TABLE_WALK_LOAD_STORE_FOREIGN = 7,
	190	+ MCE_RA_ERROR_LOAD_STORE_FOREIGN = 8,
	191	+};
	192	+
	193	+enum MCE_LinkErrorType {
	194	+ MCE_LINK_ERROR_INDETERMINATE = 0,
	195	+ MCE_LINK_ERROR_IFETCH_TIMEOUT = 1,
	196	+ MCE_LINK_ERROR_PAGE_TABLE_WALK_IFETCH_TIMEOUT = 2,
	197	+ MCE_LINK_ERROR_LOAD_TIMEOUT = 3,
	198	+ MCE_LINK_ERROR_STORE_TIMEOUT = 4,
	199	+ MCE_LINK_ERROR_PAGE_TABLE_WALK_LOAD_STORE_TIMEOUT = 5,
	200	+};
	201	+
124	202	struct machine_check_event {
125	203	enum MCE_Version version:8; /* 0x00 */
126	204	uint8_t in_use; /* 0x01 */
@@ -166,6 +244,30 @@ struct machine_check_event { · 39 unmodified lines
166	244	uint64_t effective_address;
167	245	uint8_t reserved_2[16];
168	246	} tlb_error;
	247	+
	248	+ struct {
	249	+ enum MCE_UserErrorType user_error_type:8;
	250	+ uint8_t effective_address_provided;
	251	+ uint8_t reserved_1[6];
	252	+ uint64_t effective_address;
	253	+ uint8_t reserved_2[16];
	254	+ } user_error;
	255	+
	256	+ struct {
	257	+ enum MCE_RaErrorType ra_error_type:8;
	258	+ uint8_t effective_address_provided;
	259	+ uint8_t reserved_1[6];
	260	+ uint64_t effective_address;
	261	+ uint8_t reserved_2[16];
	262	+ } ra_error;
	263	+
	264	+ struct {
	265	+ enum MCE_LinkErrorType link_error_type:8;
	266	+ uint8_t effective_address_provided;
	267	+ uint8_t reserved_1[6];
	268	+ uint64_t effective_address;
	269	+ uint8_t reserved_2[16];
	270	+ } link_error;
169	271	} u;
170	272	};
171	273
@@ -176,8 +278,12 @@ struct mce_error_info { · 4 unmodified lines
176	278	enum MCE_SlbErrorType slb_error_type:8;
177	279	enum MCE_EratErrorType erat_error_type:8;
178	280	enum MCE_TlbErrorType tlb_error_type:8;
	281	+ enum MCE_UserErrorType user_error_type:8;
	282	+ enum MCE_RaErrorType ra_error_type:8;
	283	+ enum MCE_LinkErrorType link_error_type:8;
179	284	} u;
180		− uint8_t reserved[2];
	285	+ enum MCE_Severity severity:8;
	286	+ enum MCE_Initiator initiator:8;
181	287	};
182	288
183	289	#define MAX_MC_EVT 100

arch/powerpc/include/asm/nohash/32/pgtable.h+1−0

@@ -1,6 +1,7 @@
1	1	#ifndef _ASM_POWERPC_NOHASH_32_PGTABLE_H
2	2	#define _ASM_POWERPC_NOHASH_32_PGTABLE_H
3	3
	4	+#define __ARCH_USE_5LEVEL_HACK
4	5	#include <asm-generic/pgtable-nopmd.h>
5	6
6	7	#ifndef __ASSEMBLY__

arch/powerpc/include/asm/nohash/64/pgtable-4k.h+3−0

@@ -1,5 +1,8 @@
1	1	#ifndef _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H
2	2	#define _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H
	3	+
	4	+#include <asm-generic/5level-fixup.h>
	5	+
3	6	/*
4	7	* Entries per page directory level. The PTE level must use a 64b record
5	8	* for each page table entry. The PMD and PGD level use a 32b record for

arch/powerpc/include/asm/nohash/64/pgtable-64k.h+1−0

@@ -1,6 +1,7 @@
1	1	#ifndef _ASM_POWERPC_NOHASH_64_PGTABLE_64K_H
2	2	#define _ASM_POWERPC_NOHASH_64_PGTABLE_64K_H
3	3
	4	+#define __ARCH_USE_5LEVEL_HACK
4	5	#include <asm-generic/pgtable-nopud.h>
5	6
6	7

arch/powerpc/include/asm/nohash/pgtable.h+1−1

@@ -230,7 +230,7 @@ static inline int hugepd_ok(hugepd_t hpd) · 229 unmodified lines
230	230	return ((hpd_val(hpd) & 0x4) != 0);
231	231	#else
232	232	/* We clear the top bit to indicate hugepd */
233		− return ((hpd_val(hpd) & PD_HUGE) == 0);
	233	+ return (hpd_val(hpd) && (hpd_val(hpd) & PD_HUGE) == 0);
234	234	#endif
235	235	}
236	236

arch/powerpc/include/asm/ppc-opcode.h+7−0

@@ -284,6 +284,13 @@ · 283 unmodified lines
284	284	#define PPC_INST_BRANCH_COND 0x40800000
285	285	#define PPC_INST_LBZCIX 0x7c0006aa
286	286	#define PPC_INST_STBCIX 0x7c0007aa
	287	+#define PPC_INST_LWZX 0x7c00002e
	288	+#define PPC_INST_LFSX 0x7c00042e
	289	+#define PPC_INST_STFSX 0x7c00052e
	290	+#define PPC_INST_LFDX 0x7c0004ae
	291	+#define PPC_INST_STFDX 0x7c0005ae
	292	+#define PPC_INST_LVX 0x7c0000ce
	293	+#define PPC_INST_STVX 0x7c0001ce
287	294
288	295	/* macros to insert fields into opcodes */
289	296	#define ___PPC_RA(a) (((a) & 0x1f) << 16)

arch/powerpc/include/asm/prom.h+12−6

@@ -160,12 +160,18 @@ struct of_drconf_cell { · 159 unmodified lines
160	160	#define OV5_PFO_HW_ENCR 0x1120 /* PFO Encryption Accelerator */
161	161	#define OV5_SUB_PROCESSORS 0x1501 /* 1,2,or 4 Sub-Processors supported */
162	162	#define OV5_XIVE_EXPLOIT 0x1701 /* XIVE exploitation supported */
163		−#define OV5_MMU_RADIX_300 0x1880 /* ISA v3.00 radix MMU supported */
164		−#define OV5_MMU_HASH_300 0x1840 /* ISA v3.00 hash MMU supported */
165		−#define OV5_MMU_SEGM_RADIX 0x1820 /* radix mode (no segmentation) */
166		−#define OV5_MMU_PROC_TBL 0x1810 /* hcall selects SLB or proc table */
167		−#define OV5_MMU_SLB 0x1800 /* always use SLB */
168		−#define OV5_MMU_GTSE 0x1808 /* Guest translation shootdown */
	163	+/* MMU Base Architecture */
	164	+#define OV5_MMU_SUPPORT 0x18C0 /* MMU Mode Support Mask */
	165	+#define OV5_MMU_HASH 0x1800 /* Hash MMU Only */
	166	+#define OV5_MMU_RADIX 0x1840 /* Radix MMU Only */
	167	+#define OV5_MMU_EITHER 0x1880 /* Hash or Radix Supported */
	168	+#define OV5_MMU_DYNAMIC 0x18C0 /* Hash or Radix Can Switch Later */
	169	+#define OV5_NMMU 0x1820 /* Nest MMU Available */
	170	+/* Hash Table Extensions */
	171	+#define OV5_HASH_SEG_TBL 0x1980 /* In Memory Segment Tables Available */
	172	+#define OV5_HASH_GTSE 0x1940 /* Guest Translation Shoot Down Avail */
	173	+/* Radix Table Extensions */
	174	+#define OV5_RADIX_GTSE 0x1A40 /* Guest Translation Shoot Down Avail */
169	175
170	176	/* Option Vector 6: IBM PAPR hints */
171	177	#define OV6_LINUX 0x02 /* Linux is our OS */

arch/powerpc/kernel/cputable.c+3−0

@@ -77,6 +77,7 @@ extern void __flush_tlb_power8(unsigned int action); · 76 unmodified lines
77	77	extern void __flush_tlb_power9(unsigned int action);
78	78	extern long __machine_check_early_realmode_p7(struct pt_regs *regs);
79	79	extern long __machine_check_early_realmode_p8(struct pt_regs *regs);
	80	+extern long __machine_check_early_realmode_p9(struct pt_regs *regs);
80	81	#endif /* CONFIG_PPC64 */
81	82	#if defined(CONFIG_E500)
82	83	extern void __setup_cpu_e5500(unsigned long offset, struct cpu_spec* spec);
@@ -540,6 +541,7 @@ static struct cpu_spec __initdata cpu_specs[] = { · 457 unmodified lines
540	541	.cpu_setup = __setup_cpu_power9,
541	542	.cpu_restore = __restore_cpu_power9,
542	543	.flush_tlb = __flush_tlb_power9,
	544	+ .machine_check_early = __machine_check_early_realmode_p9,
543	545	.platform = "power9",
544	546	},
545	547	{ /* Power9 */
@@ -559,6 +561,7 @@ static struct cpu_spec __initdata cpu_specs[] = { · 13 unmodified lines
559	561	.cpu_setup = __setup_cpu_power9,
560	562	.cpu_restore = __restore_cpu_power9,
561	563	.flush_tlb = __flush_tlb_power9,
	564	+ .machine_check_early = __machine_check_early_realmode_p9,
562	565	.platform = "power9",
563	566	},
564	567	{ /* Cell Broadband Engine */

arch/powerpc/kernel/idle_book3s.S+6−4

@@ -276,19 +276,21 @@ power_enter_stop: · 275 unmodified lines
276	276	*/
277	277	andis. r4,r3,PSSCR_EC_ESL_MASK_SHIFTED
278	278	clrldi r3,r3,60 /* r3 = Bits[60:63] = Requested Level (RL) */
279		− bne 1f
	279	+ bne .Lhandle_esl_ec_set
280	280	IDLE_STATE_ENTER_SEQ(PPC_STOP)
281	281	li r3,0 /* Since we didn't lose state, return 0 */
282	282	b pnv_wakeup_noloss
	283	+
	284	+.Lhandle_esl_ec_set:
283	285	/*
284	286	* Check if the requested state is a deep idle state.
285	287	*/
286		−1: LOAD_REG_ADDRBASE(r5,pnv_first_deep_stop_state)
	288	+ LOAD_REG_ADDRBASE(r5,pnv_first_deep_stop_state)
287	289	ld r4,ADDROFF(pnv_first_deep_stop_state)(r5)
288	290	cmpd r3,r4
289		− bge 2f
	291	+ bge .Lhandle_deep_stop
290	292	IDLE_STATE_ENTER_SEQ_NORET(PPC_STOP)
291		−2:
	293	+.Lhandle_deep_stop:
292	294	/*
293	295	* Entering deep idle state.
294	296	* Clear thread bit in PACA_CORE_IDLE_STATE, save SPRs to

arch/powerpc/kernel/mce.c+86−2

@@ -58,6 +58,15 @@ static void mce_set_error_info(struct machine_check_event *mce, · 57 unmodified lines
58	58	case MCE_ERROR_TYPE_TLB:
59	59	mce->u.tlb_error.tlb_error_type = mce_err->u.tlb_error_type;
60	60	break;
	61	+ case MCE_ERROR_TYPE_USER:
	62	+ mce->u.user_error.user_error_type = mce_err->u.user_error_type;
	63	+ break;
	64	+ case MCE_ERROR_TYPE_RA:
	65	+ mce->u.ra_error.ra_error_type = mce_err->u.ra_error_type;
	66	+ break;
	67	+ case MCE_ERROR_TYPE_LINK:
	68	+ mce->u.link_error.link_error_type = mce_err->u.link_error_type;
	69	+ break;
61	70	case MCE_ERROR_TYPE_UNKNOWN:
62	71	default:
63	72	break;
@@ -90,13 +99,14 @@ void save_mce_event(struct pt_regs *regs, long handled, · 26 unmodified lines
90	99	mce->gpr3 = regs->gpr[3];
91	100	mce->in_use = 1;
92	101
93		− mce->initiator = MCE_INITIATOR_CPU;
94	102	/* Mark it recovered if we have handled it and MSR(RI=1). */
95	103	if (handled && (regs->msr & MSR_RI))
96	104	mce->disposition = MCE_DISPOSITION_RECOVERED;
97	105	else
98	106	mce->disposition = MCE_DISPOSITION_NOT_RECOVERED;
99		− mce->severity = MCE_SEV_ERROR_SYNC;
	107	+
	108	+ mce->initiator = mce_err->initiator;
	109	+ mce->severity = mce_err->severity;
100	110
101	111	/*
102	112	* Populate the mce error_type and type-specific error_type.
@@ -115,6 +125,15 @@ void save_mce_event(struct pt_regs *regs, long handled, · 12 unmodified lines
115	125	} else if (mce->error_type == MCE_ERROR_TYPE_ERAT) {
116	126	mce->u.erat_error.effective_address_provided = true;
117	127	mce->u.erat_error.effective_address = addr;
	128	+ } else if (mce->error_type == MCE_ERROR_TYPE_USER) {
	129	+ mce->u.user_error.effective_address_provided = true;
	130	+ mce->u.user_error.effective_address = addr;
	131	+ } else if (mce->error_type == MCE_ERROR_TYPE_RA) {
	132	+ mce->u.ra_error.effective_address_provided = true;
	133	+ mce->u.ra_error.effective_address = addr;
	134	+ } else if (mce->error_type == MCE_ERROR_TYPE_LINK) {
	135	+ mce->u.link_error.effective_address_provided = true;
	136	+ mce->u.link_error.effective_address = addr;
118	137	} else if (mce->error_type == MCE_ERROR_TYPE_UE) {
119	138	mce->u.ue_error.effective_address_provided = true;
120	139	mce->u.ue_error.effective_address = addr;
@@ -239,6 +258,29 @@ void machine_check_print_event_info(struct machine_check_event *evt) · 118 unmodified lines
239	258	"Parity",
240	259	"Multihit",
241	260	};
	261	+ static const char *mc_user_types[] = {
	262	+ "Indeterminate",
	263	+ "tlbie(l) invalid",
	264	+ };
	265	+ static const char *mc_ra_types[] = {
	266	+ "Indeterminate",
	267	+ "Instruction fetch (bad)",
	268	+ "Page table walk ifetch (bad)",
	269	+ "Page table walk ifetch (foreign)",
	270	+ "Load (bad)",
	271	+ "Store (bad)",
	272	+ "Page table walk Load/Store (bad)",
	273	+ "Page table walk Load/Store (foreign)",
	274	+ "Load/Store (foreign)",
	275	+ };
	276	+ static const char *mc_link_types[] = {
	277	+ "Indeterminate",
	278	+ "Instruction fetch (timeout)",
	279	+ "Page table walk ifetch (timeout)",
	280	+ "Load (timeout)",
	281	+ "Store (timeout)",
	282	+ "Page table walk Load/Store (timeout)",
	283	+ };
242	284
243	285	/* Print things out */
244	286	if (evt->version != MCE_V1) {
@@ -315,6 +357,36 @@ void machine_check_print_event_info(struct machine_check_event *evt) · 70 unmodified lines
315	357	printk("%s Effective address: %016llx\n",
316	358	level, evt->u.tlb_error.effective_address);
317	359	break;
	360	+ case MCE_ERROR_TYPE_USER:
	361	+ subtype = evt->u.user_error.user_error_type <
	362	+ ARRAY_SIZE(mc_user_types) ?
	363	+ mc_user_types[evt->u.user_error.user_error_type]
	364	+ : "Unknown";
	365	+ printk("%s Error type: User [%s]\n", level, subtype);
	366	+ if (evt->u.user_error.effective_address_provided)
	367	+ printk("%s Effective address: %016llx\n",
	368	+ level, evt->u.user_error.effective_address);
	369	+ break;
	370	+ case MCE_ERROR_TYPE_RA:
	371	+ subtype = evt->u.ra_error.ra_error_type <
	372	+ ARRAY_SIZE(mc_ra_types) ?
	373	+ mc_ra_types[evt->u.ra_error.ra_error_type]
	374	+ : "Unknown";
	375	+ printk("%s Error type: Real address [%s]\n", level, subtype);
	376	+ if (evt->u.ra_error.effective_address_provided)
	377	+ printk("%s Effective address: %016llx\n",
	378	+ level, evt->u.ra_error.effective_address);
	379	+ break;
	380	+ case MCE_ERROR_TYPE_LINK:
	381	+ subtype = evt->u.link_error.link_error_type <
	382	+ ARRAY_SIZE(mc_link_types) ?
	383	+ mc_link_types[evt->u.link_error.link_error_type]
	384	+ : "Unknown";
	385	+ printk("%s Error type: Link [%s]\n", level, subtype);
	386	+ if (evt->u.link_error.effective_address_provided)
	387	+ printk("%s Effective address: %016llx\n",
	388	+ level, evt->u.link_error.effective_address);
	389	+ break;
318	390	default:
319	391	case MCE_ERROR_TYPE_UNKNOWN:
320	392	printk("%s Error type: Unknown\n", level);
@@ -341,6 +413,18 @@ uint64_t get_mce_fault_addr(struct machine_check_event *evt) · 20 unmodified lines
341	413	if (evt->u.tlb_error.effective_address_provided)
342	414	return evt->u.tlb_error.effective_address;
343	415	break;
	416	+ case MCE_ERROR_TYPE_USER:
	417	+ if (evt->u.user_error.effective_address_provided)
	418	+ return evt->u.user_error.effective_address;
	419	+ break;
	420	+ case MCE_ERROR_TYPE_RA:
	421	+ if (evt->u.ra_error.effective_address_provided)
	422	+ return evt->u.ra_error.effective_address;
	423	+ break;
	424	+ case MCE_ERROR_TYPE_LINK:
	425	+ if (evt->u.link_error.effective_address_provided)
	426	+ return evt->u.link_error.effective_address;
	427	+ break;
344	428	default:
345	429	case MCE_ERROR_TYPE_UNKNOWN:
346	430	break;

arch/powerpc/kernel/mce_power.c+237−0

@@ -116,6 +116,51 @@ static void flush_and_reload_slb(void) · 115 unmodified lines
116	116	}
117	117	#endif
118	118
	119	+static void flush_erat(void)
	120	+{
	121	+ asm volatile(PPC_INVALIDATE_ERAT : : :"memory");
	122	+}
	123	+
	124	+#define MCE_FLUSH_SLB 1
	125	+#define MCE_FLUSH_TLB 2
	126	+#define MCE_FLUSH_ERAT 3
	127	+
	128	+static int mce_flush(int what)
	129	+{
	130	+#ifdef CONFIG_PPC_STD_MMU_64
	131	+ if (what == MCE_FLUSH_SLB) {
	132	+ flush_and_reload_slb();
	133	+ return 1;
	134	+ }
	135	+#endif
	136	+ if (what == MCE_FLUSH_ERAT) {
	137	+ flush_erat();
	138	+ return 1;
	139	+ }
	140	+ if (what == MCE_FLUSH_TLB) {
	141	+ if (cur_cpu_spec && cur_cpu_spec->flush_tlb) {
	142	+ cur_cpu_spec->flush_tlb(TLB_INVAL_SCOPE_GLOBAL);
	143	+ return 1;
	144	+ }
	145	+ }
	146	+
	147	+ return 0;
	148	+}
	149	+
	150	+static int mce_handle_flush_derrors(uint64_t dsisr, uint64_t slb, uint64_t tlb, uint64_t erat)
	151	+{
	152	+ if ((dsisr & slb) && mce_flush(MCE_FLUSH_SLB))
	153	+ dsisr &= ~slb;
	154	+ if ((dsisr & erat) && mce_flush(MCE_FLUSH_ERAT))
	155	+ dsisr &= ~erat;
	156	+ if ((dsisr & tlb) && mce_flush(MCE_FLUSH_TLB))
	157	+ dsisr &= ~tlb;
	158	+ /* Any other errors we don't understand? */
	159	+ if (dsisr)
	160	+ return 0;
	161	+ return 1;
	162	+}
	163	+
119	164	static long mce_handle_derror(uint64_t dsisr, uint64_t slb_error_bits)
120	165	{
121	166	long handled = 1;
@@ -281,6 +326,9 @@ long __machine_check_early_realmode_p7(struct pt_regs *regs) · 159 unmodified lines
281	326	long handled = 1;
282	327	struct mce_error_info mce_error_info = { 0 };
283	328
	329	+ mce_error_info.severity = MCE_SEV_ERROR_SYNC;
	330	+ mce_error_info.initiator = MCE_INITIATOR_CPU;
	331	+
284	332	srr1 = regs->msr;
285	333	nip = regs->nip;
286	334
@@ -352,6 +400,9 @@ long __machine_check_early_realmode_p8(struct pt_regs *regs) · 65 unmodified lines
352	400	long handled = 1;
353	401	struct mce_error_info mce_error_info = { 0 };
354	402
	403	+ mce_error_info.severity = MCE_SEV_ERROR_SYNC;
	404	+ mce_error_info.initiator = MCE_INITIATOR_CPU;
	405	+
355	406	srr1 = regs->msr;
356	407	nip = regs->nip;
357	408
@@ -372,3 +423,189 @@ long __machine_check_early_realmode_p8(struct pt_regs *regs) · 14 unmodified lines
372	423	save_mce_event(regs, handled, &mce_error_info, nip, addr);
373	424	return handled;
374	425	}
	426	+
	427	+static int mce_handle_derror_p9(struct pt_regs *regs)
	428	+{
	429	+ uint64_t dsisr = regs->dsisr;
	430	+
	431	+ return mce_handle_flush_derrors(dsisr,
	432	+ P9_DSISR_MC_SLB_PARITY_MFSLB \|
	433	+ P9_DSISR_MC_SLB_MULTIHIT_MFSLB,
	434	+
	435	+ P9_DSISR_MC_TLB_MULTIHIT_MFTLB,
	436	+
	437	+ P9_DSISR_MC_ERAT_MULTIHIT);
	438	+}
	439	+
	440	+static int mce_handle_ierror_p9(struct pt_regs *regs)
	441	+{
	442	+ uint64_t srr1 = regs->msr;
	443	+
	444	+ switch (P9_SRR1_MC_IFETCH(srr1)) {
	445	+ case P9_SRR1_MC_IFETCH_SLB_PARITY:
	446	+ case P9_SRR1_MC_IFETCH_SLB_MULTIHIT:
	447	+ return mce_flush(MCE_FLUSH_SLB);
	448	+ case P9_SRR1_MC_IFETCH_TLB_MULTIHIT:
	449	+ return mce_flush(MCE_FLUSH_TLB);
	450	+ case P9_SRR1_MC_IFETCH_ERAT_MULTIHIT:
	451	+ return mce_flush(MCE_FLUSH_ERAT);
	452	+ default:
	453	+ return 0;
	454	+ }
	455	+}
	456	+
	457	+static void mce_get_derror_p9(struct pt_regs *regs,
	458	+ struct mce_error_info mce_err, uint64_t addr)
	459	+{
	460	+ uint64_t dsisr = regs->dsisr;
	461	+
	462	+ mce_err->severity = MCE_SEV_ERROR_SYNC;
	463	+ mce_err->initiator = MCE_INITIATOR_CPU;
	464	+
	465	+ if (dsisr & P9_DSISR_MC_USER_TLBIE)
	466	+ *addr = regs->nip;
	467	+ else
	468	+ *addr = regs->dar;
	469	+
	470	+ if (dsisr & P9_DSISR_MC_UE) {
	471	+ mce_err->error_type = MCE_ERROR_TYPE_UE;
	472	+ mce_err->u.ue_error_type = MCE_UE_ERROR_LOAD_STORE;
	473	+ } else if (dsisr & P9_DSISR_MC_UE_TABLEWALK) {
	474	+ mce_err->error_type = MCE_ERROR_TYPE_UE;
	475	+ mce_err->u.ue_error_type = MCE_UE_ERROR_PAGE_TABLE_WALK_LOAD_STORE;
	476	+ } else if (dsisr & P9_DSISR_MC_LINK_LOAD_TIMEOUT) {
	477	+ mce_err->error_type = MCE_ERROR_TYPE_LINK;
	478	+ mce_err->u.link_error_type = MCE_LINK_ERROR_LOAD_TIMEOUT;
	479	+ } else if (dsisr & P9_DSISR_MC_LINK_TABLEWALK_TIMEOUT) {
	480	+ mce_err->error_type = MCE_ERROR_TYPE_LINK;
	481	+ mce_err->u.link_error_type = MCE_LINK_ERROR_PAGE_TABLE_WALK_LOAD_STORE_TIMEOUT;
	482	+ } else if (dsisr & P9_DSISR_MC_ERAT_MULTIHIT) {
	483	+ mce_err->error_type = MCE_ERROR_TYPE_ERAT;
	484	+ mce_err->u.erat_error_type = MCE_ERAT_ERROR_MULTIHIT;
	485	+ } else if (dsisr & P9_DSISR_MC_TLB_MULTIHIT_MFTLB) {
	486	+ mce_err->error_type = MCE_ERROR_TYPE_TLB;
	487	+ mce_err->u.tlb_error_type = MCE_TLB_ERROR_MULTIHIT;
	488	+ } else if (dsisr & P9_DSISR_MC_USER_TLBIE) {
	489	+ mce_err->error_type = MCE_ERROR_TYPE_USER;
	490	+ mce_err->u.user_error_type = MCE_USER_ERROR_TLBIE;
	491	+ } else if (dsisr & P9_DSISR_MC_SLB_PARITY_MFSLB) {
	492	+ mce_err->error_type = MCE_ERROR_TYPE_SLB;
	493	+ mce_err->u.slb_error_type = MCE_SLB_ERROR_PARITY;
	494	+ } else if (dsisr & P9_DSISR_MC_SLB_MULTIHIT_MFSLB) {
	495	+ mce_err->error_type = MCE_ERROR_TYPE_SLB;
	496	+ mce_err->u.slb_error_type = MCE_SLB_ERROR_MULTIHIT;
	497	+ } else if (dsisr & P9_DSISR_MC_RA_LOAD) {
	498	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	499	+ mce_err->u.ra_error_type = MCE_RA_ERROR_LOAD;
	500	+ } else if (dsisr & P9_DSISR_MC_RA_TABLEWALK) {
	501	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	502	+ mce_err->u.ra_error_type = MCE_RA_ERROR_PAGE_TABLE_WALK_LOAD_STORE;
	503	+ } else if (dsisr & P9_DSISR_MC_RA_TABLEWALK_FOREIGN) {
	504	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	505	+ mce_err->u.ra_error_type = MCE_RA_ERROR_PAGE_TABLE_WALK_LOAD_STORE_FOREIGN;
	506	+ } else if (dsisr & P9_DSISR_MC_RA_FOREIGN) {
	507	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	508	+ mce_err->u.ra_error_type = MCE_RA_ERROR_LOAD_STORE_FOREIGN;
	509	+ }
	510	+}
	511	+
	512	+static void mce_get_ierror_p9(struct pt_regs *regs,
	513	+ struct mce_error_info mce_err, uint64_t addr)
	514	+{
	515	+ uint64_t srr1 = regs->msr;
	516	+
	517	+ switch (P9_SRR1_MC_IFETCH(srr1)) {
	518	+ case P9_SRR1_MC_IFETCH_RA_ASYNC_STORE:
	519	+ case P9_SRR1_MC_IFETCH_LINK_ASYNC_STORE_TIMEOUT:
	520	+ mce_err->severity = MCE_SEV_FATAL;
	521	+ break;
	522	+ default:
	523	+ mce_err->severity = MCE_SEV_ERROR_SYNC;
	524	+ break;
	525	+ }
	526	+
	527	+ mce_err->initiator = MCE_INITIATOR_CPU;
	528	+
	529	+ *addr = regs->nip;
	530	+
	531	+ switch (P9_SRR1_MC_IFETCH(srr1)) {
	532	+ case P9_SRR1_MC_IFETCH_UE:
	533	+ mce_err->error_type = MCE_ERROR_TYPE_UE;
	534	+ mce_err->u.ue_error_type = MCE_UE_ERROR_IFETCH;
	535	+ break;
	536	+ case P9_SRR1_MC_IFETCH_SLB_PARITY:
	537	+ mce_err->error_type = MCE_ERROR_TYPE_SLB;
	538	+ mce_err->u.slb_error_type = MCE_SLB_ERROR_PARITY;
	539	+ break;
	540	+ case P9_SRR1_MC_IFETCH_SLB_MULTIHIT:
	541	+ mce_err->error_type = MCE_ERROR_TYPE_SLB;
	542	+ mce_err->u.slb_error_type = MCE_SLB_ERROR_MULTIHIT;
	543	+ break;
	544	+ case P9_SRR1_MC_IFETCH_ERAT_MULTIHIT:
	545	+ mce_err->error_type = MCE_ERROR_TYPE_ERAT;
	546	+ mce_err->u.erat_error_type = MCE_ERAT_ERROR_MULTIHIT;
	547	+ break;
	548	+ case P9_SRR1_MC_IFETCH_TLB_MULTIHIT:
	549	+ mce_err->error_type = MCE_ERROR_TYPE_TLB;
	550	+ mce_err->u.tlb_error_type = MCE_TLB_ERROR_MULTIHIT;
	551	+ break;
	552	+ case P9_SRR1_MC_IFETCH_UE_TLB_RELOAD:
	553	+ mce_err->error_type = MCE_ERROR_TYPE_UE;
	554	+ mce_err->u.ue_error_type = MCE_UE_ERROR_PAGE_TABLE_WALK_IFETCH;
	555	+ break;
	556	+ case P9_SRR1_MC_IFETCH_LINK_TIMEOUT:
	557	+ mce_err->error_type = MCE_ERROR_TYPE_LINK;
	558	+ mce_err->u.link_error_type = MCE_LINK_ERROR_IFETCH_TIMEOUT;
	559	+ break;
	560	+ case P9_SRR1_MC_IFETCH_LINK_TABLEWALK_TIMEOUT:
	561	+ mce_err->error_type = MCE_ERROR_TYPE_LINK;
	562	+ mce_err->u.link_error_type = MCE_LINK_ERROR_PAGE_TABLE_WALK_IFETCH_TIMEOUT;
	563	+ break;
	564	+ case P9_SRR1_MC_IFETCH_RA:
	565	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	566	+ mce_err->u.ra_error_type = MCE_RA_ERROR_IFETCH;
	567	+ break;
	568	+ case P9_SRR1_MC_IFETCH_RA_TABLEWALK:
	569	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	570	+ mce_err->u.ra_error_type = MCE_RA_ERROR_PAGE_TABLE_WALK_IFETCH;
	571	+ break;
	572	+ case P9_SRR1_MC_IFETCH_RA_ASYNC_STORE:
	573	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	574	+ mce_err->u.ra_error_type = MCE_RA_ERROR_STORE;
	575	+ break;
	576	+ case P9_SRR1_MC_IFETCH_LINK_ASYNC_STORE_TIMEOUT:
	577	+ mce_err->error_type = MCE_ERROR_TYPE_LINK;
	578	+ mce_err->u.link_error_type = MCE_LINK_ERROR_STORE_TIMEOUT;
	579	+ break;
	580	+ case P9_SRR1_MC_IFETCH_RA_TABLEWALK_FOREIGN:
	581	+ mce_err->error_type = MCE_ERROR_TYPE_RA;
	582	+ mce_err->u.ra_error_type = MCE_RA_ERROR_PAGE_TABLE_WALK_IFETCH_FOREIGN;
	583	+ break;
	584	+ default:
	585	+ break;
	586	+ }
	587	+}
	588	+
	589	+long __machine_check_early_realmode_p9(struct pt_regs *regs)
	590	+{
	591	+ uint64_t nip, addr;
	592	+ long handled;
	593	+ struct mce_error_info mce_error_info = { 0 };
	594	+
	595	+ nip = regs->nip;
	596	+
	597	+ if (P9_SRR1_MC_LOADSTORE(regs->msr)) {
	598	+ handled = mce_handle_derror_p9(regs);
	599	+ mce_get_derror_p9(regs, &mce_error_info, &addr);
	600	+ } else {
	601	+ handled = mce_handle_ierror_p9(regs);
	602	+ mce_get_ierror_p9(regs, &mce_error_info, &addr);
	603	+ }
	604	+
	605	+ /* Handle UE error. */
	606	+ if (mce_error_info.error_type == MCE_ERROR_TYPE_UE)
	607	+ handled = mce_handle_ue_error(regs);
	608	+
	609	+ save_mce_event(regs, handled, &mce_error_info, nip, addr);
	610	+ return handled;
	611	+}

arch/powerpc/kernel/prom_init.c+113−7

@@ -168,6 +168,14 @@ static unsigned long __initdata prom_tce_alloc_start; · 167 unmodified lines
168	168	static unsigned long __initdata prom_tce_alloc_end;
169	169	#endif
170	170
	171	+static bool __initdata prom_radix_disable;
	172	+
	173	+struct platform_support {
	174	+ bool hash_mmu;
	175	+ bool radix_mmu;
	176	+ bool radix_gtse;
	177	+};
	178	+
171	179	/* Platforms codes are now obsolete in the kernel. Now only used within this
172	180	* file and ultimately gone too. Feel free to change them if you need, they
173	181	* are not shared with anything outside of this file anymore
@@ -626,6 +634,12 @@ static void __init early_cmdline_parse(void) · 452 unmodified lines
626	634	prom_memory_limit = ALIGN(prom_memory_limit, 0x1000000);
627	635	#endif
628	636	}
	637	+
	638	+ opt = strstr(prom_cmd_line, "disable_radix");
	639	+ if (opt) {
	640	+ prom_debug("Radix disabled from cmdline\n");
	641	+ prom_radix_disable = true;
	642	+ }
629	643	}
630	644
631	645	#if defined(CONFIG_PPC_PSERIES) \|\| defined(CONFIG_PPC_POWERNV)
@@ -695,6 +709,8 @@ struct option_vector5 { · 63 unmodified lines
695	709	u8 byte22;
696	710	u8 intarch;
697	711	u8 mmu;
	712	+ u8 hash_ext;
	713	+ u8 radix_ext;
698	714	} __packed;
699	715
700	716	struct option_vector6 {
@@ -850,8 +866,9 @@ struct ibm_arch_vec __cacheline_aligned ibm_architecture_vec = { · 149 unmodified lines
850	866	.reserved3 = 0,
851	867	.subprocessors = 1,
852	868	.intarch = 0,
853		− .mmu = OV5_FEAT(OV5_MMU_RADIX_300) \| OV5_FEAT(OV5_MMU_HASH_300) \|
854		− OV5_FEAT(OV5_MMU_PROC_TBL) \| OV5_FEAT(OV5_MMU_GTSE),
	869	+ .mmu = 0,
	870	+ .hash_ext = 0,
	871	+ .radix_ext = 0,
855	872	},
856	873
857	874	/* option vector 6: IBM PAPR hints */
@@ -990,6 +1007,92 @@ static int __init prom_count_smt_threads(void) · 132 unmodified lines
990	1007
991	1008	}
992	1009
	1010	+static void __init prom_parse_mmu_model(u8 val,
	1011	+ struct platform_support *support)
	1012	+{
	1013	+ switch (val) {
	1014	+ case OV5_FEAT(OV5_MMU_DYNAMIC):
	1015	+ case OV5_FEAT(OV5_MMU_EITHER): /* Either Available */
	1016	+ prom_debug("MMU - either supported\n");
	1017	+ support->radix_mmu = !prom_radix_disable;
	1018	+ support->hash_mmu = true;
	1019	+ break;
	1020	+ case OV5_FEAT(OV5_MMU_RADIX): /* Only Radix */
	1021	+ prom_debug("MMU - radix only\n");
	1022	+ if (prom_radix_disable) {
	1023	+ /*
	1024	+ * If we __have__ to do radix, we're better off ignoring
	1025	+ * the command line rather than not booting.
	1026	+ */
	1027	+ prom_printf("WARNING: Ignoring cmdline option disable_radix\n");
	1028	+ }
	1029	+ support->radix_mmu = true;
	1030	+ break;
	1031	+ case OV5_FEAT(OV5_MMU_HASH):
	1032	+ prom_debug("MMU - hash only\n");
	1033	+ support->hash_mmu = true;
	1034	+ break;
	1035	+ default:
	1036	+ prom_debug("Unknown mmu support option: 0x%x\n", val);
	1037	+ break;
	1038	+ }
	1039	+}
	1040	+
	1041	+static void __init prom_parse_platform_support(u8 index, u8 val,
	1042	+ struct platform_support *support)
	1043	+{
	1044	+ switch (index) {
	1045	+ case OV5_INDX(OV5_MMU_SUPPORT): /* MMU Model */
	1046	+ prom_parse_mmu_model(val & OV5_FEAT(OV5_MMU_SUPPORT), support);
	1047	+ break;
	1048	+ case OV5_INDX(OV5_RADIX_GTSE): /* Radix Extensions */
	1049	+ if (val & OV5_FEAT(OV5_RADIX_GTSE)) {
	1050	+ prom_debug("Radix - GTSE supported\n");
	1051	+ support->radix_gtse = true;
	1052	+ }
	1053	+ break;
	1054	+ }
	1055	+}
	1056	+
	1057	+static void __init prom_check_platform_support(void)
	1058	+{
	1059	+ struct platform_support supported = {
	1060	+ .hash_mmu = false,
	1061	+ .radix_mmu = false,
	1062	+ .radix_gtse = false
	1063	+ };
	1064	+ int prop_len = prom_getproplen(prom.chosen,
	1065	+ "ibm,arch-vec-5-platform-support");
	1066	+ if (prop_len > 1) {
	1067	+ int i;
	1068	+ u8 vec[prop_len];
	1069	+ prom_debug("Found ibm,arch-vec-5-platform-support, len: %d\n",
	1070	+ prop_len);
	1071	+ prom_getprop(prom.chosen, "ibm,arch-vec-5-platform-support",
	1072	+ &vec, sizeof(vec));
	1073	+ for (i = 0; i < prop_len; i += 2) {
	1074	+ prom_debug("%d: index = 0x%x val = 0x%x\n", i / 2
	1075	+ , vec[i]
	1076	+ , vec[i + 1]);
	1077	+ prom_parse_platform_support(vec[i], vec[i + 1],
	1078	+ &supported);
	1079	+ }
	1080	+ }
	1081	+
	1082	+ if (supported.radix_mmu && supported.radix_gtse) {
	1083	+ /* Radix preferred - but we require GTSE for now */
	1084	+ prom_debug("Asking for radix with GTSE\n");
	1085	+ ibm_architecture_vec.vec5.mmu = OV5_FEAT(OV5_MMU_RADIX);
	1086	+ ibm_architecture_vec.vec5.radix_ext = OV5_FEAT(OV5_RADIX_GTSE);
	1087	+ } else if (supported.hash_mmu) {
	1088	+ /* Default to hash mmu (if we can) */
	1089	+ prom_debug("Asking for hash\n");
	1090	+ ibm_architecture_vec.vec5.mmu = OV5_FEAT(OV5_MMU_HASH);
	1091	+ } else {
	1092	+ /* We're probably on a legacy hypervisor */
	1093	+ prom_debug("Assuming legacy hash support\n");
	1094	+ }
	1095	+}
993	1096
994	1097	static void __init prom_send_capabilities(void)
995	1098	{
@@ -997,6 +1100,9 @@ static void __init prom_send_capabilities(void) · 1 unmodified line
997	1100	prom_arg_t ret;
998	1101	u32 cores;
999	1102
	1103	+ /* Check ibm,arch-vec-5-platform-support and fixup vec5 if required */
	1104	+ prom_check_platform_support();
	1105	+
1000	1106	root = call_prom("open", 1, 1, ADDR("/"));
1001	1107	if (root != 0) {
1002	1108	/* We need to tell the FW about the number of cores we support.
@@ -2993,6 +3099,11 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, · 1990 unmodified lines
2993	3099	*/
2994	3100	prom_check_initrd(r3, r4);
2995	3101
	3102	+ /*
	3103	+ * Do early parsing of command line
	3104	+ */
	3105	+ early_cmdline_parse();
	3106	+
2996	3107	#if defined(CONFIG_PPC_PSERIES) \|\| defined(CONFIG_PPC_POWERNV)
2997	3108	/*
2998	3109	* On pSeries, inform the firmware about our capabilities
@@ -3008,11 +3119,6 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, · 9 unmodified lines
3008	3119	if (of_platform != PLATFORM_POWERMAC)
3009	3120	copy_and_flush(0, kbase, 0x100, 0);
3010	3121
3011		− /*
3012		− * Do early parsing of command line
3013		− */
3014		− early_cmdline_parse();
3015		−
3016	3122	/*
3017	3123	* Initialize memory management within prom_init
3018	3124	*/

arch/powerpc/kernel/setup_64.c+4−1

@@ -408,7 +408,10 @@ static void init_cache_info(struct ppc_cache_info *info, u32 size, u32 lsize, · 407 unmodified lines
408	408	info->line_size = lsize;
409	409	info->block_size = bsize;
410	410	info->log_block_size = __ilog2(bsize);
411		− info->blocks_per_page = PAGE_SIZE / bsize;
	411	+ if (bsize)
	412	+ info->blocks_per_page = PAGE_SIZE / bsize;
	413	+ else
	414	+ info->blocks_per_page = 0;
412	415
413	416	if (sets == 0)
414	417	info->assoc = 0xffff;

arch/powerpc/kvm/book3s_64_mmu_hv.c+1−1

@@ -601,7 +601,7 @@ int kvmppc_book3s_hv_page_fault(struct kvm_run run, struct kvm_vcpu vcpu, · 600 unmodified lines
601	601	hva, NULL, NULL);
602	602	if (ptep) {
603	603	pte = kvmppc_read_update_linux_pte(ptep, 1);
604		− if (pte_write(pte))
	604	+ if (__pte_write(pte))
605	605	write_ok = 1;
606	606	}
607	607	local_irq_restore(flags);

arch/powerpc/kvm/book3s_hv_rm_mmu.c+1−1

@@ -256,7 +256,7 @@ long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags, · 255 unmodified lines
256	256	}
257	257	pte = kvmppc_read_update_linux_pte(ptep, writing);
258	258	if (pte_present(pte) && !pte_protnone(pte)) {
259		− if (writing && !pte_write(pte))
	259	+ if (writing && !__pte_write(pte))
260	260	/* make the actual HPTE be read-only */
261	261	ptel = hpte_make_readonly(ptel);
262	262	is_ci = pte_ci(pte);

arch/powerpc/lib/Makefile+1−0

@@ -20,6 +20,7 @@ obj64-y += copypage_64.o copyuser_64.o usercopy_64.o mem_64.o hweight_64.o \ · 19 unmodified lines
20	20
21	21	obj64-$(CONFIG_SMP) += locks.o
22	22	obj64-$(CONFIG_ALTIVEC) += vmx-helper.o
	23	+obj64-$(CONFIG_KPROBES_SANITY_TEST) += test_emulate_step.o
23	24
24	25	obj-y += checksum_$(BITS).o checksum_wrappers.o
25	26

arch/powerpc/lib/sstep.c+0−20

@@ -1799,8 +1799,6 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 1798 unmodified lines
1799	1799	goto instr_done;
1800	1800
1801	1801	case LARX:
1802		− if (regs->msr & MSR_LE)
1803		− return 0;
1804	1802	if (op.ea & (size - 1))
1805	1803	break; /* can't handle misaligned */
1806	1804	if (!address_ok(regs, op.ea, size))
@@ -1823,8 +1821,6 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 16 unmodified lines
1823	1821	goto ldst_done;
1824	1822
1825	1823	case STCX:
1826		− if (regs->msr & MSR_LE)
1827		− return 0;
1828	1824	if (op.ea & (size - 1))
1829	1825	break; /* can't handle misaligned */
1830	1826	if (!address_ok(regs, op.ea, size))
@@ -1849,8 +1845,6 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 18 unmodified lines
1849	1845	goto ldst_done;
1850	1846
1851	1847	case LOAD:
1852		− if (regs->msr & MSR_LE)
1853		− return 0;
1854	1848	err = read_mem(&regs->gpr[op.reg], op.ea, size, regs);
1855	1849	if (!err) {
1856	1850	if (op.type & SIGNEXT)
@@ -1862,8 +1856,6 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 5 unmodified lines
1862	1856
1863	1857	#ifdef CONFIG_PPC_FPU
1864	1858	case LOAD_FP:
1865		− if (regs->msr & MSR_LE)
1866		− return 0;
1867	1859	if (size == 4)
1868	1860	err = do_fp_load(op.reg, do_lfs, op.ea, size, regs);
1869	1861	else
@@ -1872,15 +1864,11 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 2 unmodified lines
1872	1864	#endif
1873	1865	#ifdef CONFIG_ALTIVEC
1874	1866	case LOAD_VMX:
1875		− if (regs->msr & MSR_LE)
1876		− return 0;
1877	1867	err = do_vec_load(op.reg, do_lvx, op.ea & ~0xfUL, regs);
1878	1868	goto ldst_done;
1879	1869	#endif
1880	1870	#ifdef CONFIG_VSX
1881	1871	case LOAD_VSX:
1882		− if (regs->msr & MSR_LE)
1883		− return 0;
1884	1872	err = do_vsx_load(op.reg, do_lxvd2x, op.ea, regs);
1885	1873	goto ldst_done;
1886	1874	#endif
@@ -1903,8 +1891,6 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 16 unmodified lines
1903	1891	goto instr_done;
1904	1892
1905	1893	case STORE:
1906		− if (regs->msr & MSR_LE)
1907		− return 0;
1908	1894	if ((op.type & UPDATE) && size == sizeof(long) &&
1909	1895	op.reg == 1 && op.update_reg == 1 &&
1910	1896	!(regs->msr & MSR_PR) &&
@@ -1917,8 +1903,6 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 6 unmodified lines
1917	1903
1918	1904	#ifdef CONFIG_PPC_FPU
1919	1905	case STORE_FP:
1920		− if (regs->msr & MSR_LE)
1921		− return 0;
1922	1906	if (size == 4)
1923	1907	err = do_fp_store(op.reg, do_stfs, op.ea, size, regs);
1924	1908	else
@@ -1927,15 +1911,11 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr) · 2 unmodified lines
1927	1911	#endif
1928	1912	#ifdef CONFIG_ALTIVEC
1929	1913	case STORE_VMX:
1930		− if (regs->msr & MSR_LE)
1931		− return 0;
1932	1914	err = do_vec_store(op.reg, do_stvx, op.ea & ~0xfUL, regs);
1933	1915	goto ldst_done;
1934	1916	#endif
1935	1917	#ifdef CONFIG_VSX
1936	1918	case STORE_VSX:
1937		− if (regs->msr & MSR_LE)
1938		− return 0;
1939	1919	err = do_vsx_store(op.reg, do_stxvd2x, op.ea, regs);
1940	1920	goto ldst_done;
1941	1921	#endif

arch/powerpc/lib/test_emulate_step.c+434−0

@@ -0,0 +1,434 @@
	1	+/*
	2	+ * Simple sanity test for emulate_step load/store instructions.
	3	+ *
	4	+ * Copyright IBM Corp. 2016
	5	+ *
	6	+ * This program is free software; you can redistribute it and/or modify
	7	+ * it under the terms of the GNU General Public License as published by
	8	+ * the Free Software Foundation; either version 2 of the License, or
	9	+ * (at your option) any later version.
	10	+ */
	11	+
	12	+#define pr_fmt(fmt) "emulate_step_test: " fmt
	13	+
	14	+#include <linux/ptrace.h>
	15	+#include <asm/sstep.h>
	16	+#include <asm/ppc-opcode.h>
	17	+
	18	+#define IMM_L(i) ((uintptr_t)(i) & 0xffff)
	19	+
	20	+/*
	21	+ * Defined with TEST_ prefix so it does not conflict with other
	22	+ * definitions.
	23	+ */
	24	+#define TEST_LD(r, base, i) (PPC_INST_LD \| ___PPC_RT(r) \| \
	25	+ ___PPC_RA(base) \| IMM_L(i))
	26	+#define TEST_LWZ(r, base, i) (PPC_INST_LWZ \| ___PPC_RT(r) \| \
	27	+ ___PPC_RA(base) \| IMM_L(i))
	28	+#define TEST_LWZX(t, a, b) (PPC_INST_LWZX \| ___PPC_RT(t) \| \
	29	+ ___PPC_RA(a) \| ___PPC_RB(b))
	30	+#define TEST_STD(r, base, i) (PPC_INST_STD \| ___PPC_RS(r) \| \
	31	+ ___PPC_RA(base) \| ((i) & 0xfffc))
	32	+#define TEST_LDARX(t, a, b, eh) (PPC_INST_LDARX \| ___PPC_RT(t) \| \
	33	+ ___PPC_RA(a) \| ___PPC_RB(b) \| \
	34	+ __PPC_EH(eh))
	35	+#define TEST_STDCX(s, a, b) (PPC_INST_STDCX \| ___PPC_RS(s) \| \
	36	+ ___PPC_RA(a) \| ___PPC_RB(b))
	37	+#define TEST_LFSX(t, a, b) (PPC_INST_LFSX \| ___PPC_RT(t) \| \
	38	+ ___PPC_RA(a) \| ___PPC_RB(b))
	39	+#define TEST_STFSX(s, a, b) (PPC_INST_STFSX \| ___PPC_RS(s) \| \
	40	+ ___PPC_RA(a) \| ___PPC_RB(b))
	41	+#define TEST_LFDX(t, a, b) (PPC_INST_LFDX \| ___PPC_RT(t) \| \
	42	+ ___PPC_RA(a) \| ___PPC_RB(b))
	43	+#define TEST_STFDX(s, a, b) (PPC_INST_STFDX \| ___PPC_RS(s) \| \
	44	+ ___PPC_RA(a) \| ___PPC_RB(b))
	45	+#define TEST_LVX(t, a, b) (PPC_INST_LVX \| ___PPC_RT(t) \| \
	46	+ ___PPC_RA(a) \| ___PPC_RB(b))
	47	+#define TEST_STVX(s, a, b) (PPC_INST_STVX \| ___PPC_RS(s) \| \
	48	+ ___PPC_RA(a) \| ___PPC_RB(b))
	49	+#define TEST_LXVD2X(s, a, b) (PPC_INST_LXVD2X \| VSX_XX1((s), R##a, R##b))
	50	+#define TEST_STXVD2X(s, a, b) (PPC_INST_STXVD2X \| VSX_XX1((s), R##a, R##b))
	51	+
	52	+
	53	+static void __init init_pt_regs(struct pt_regs *regs)
	54	+{
	55	+ static unsigned long msr;
	56	+ static bool msr_cached;
	57	+
	58	+ memset(regs, 0, sizeof(struct pt_regs));
	59	+
	60	+ if (likely(msr_cached)) {
	61	+ regs->msr = msr;
	62	+ return;
	63	+ }
	64	+
	65	+ asm volatile("mfmsr %0" : "=r"(regs->msr));
	66	+
	67	+ regs->msr \|= MSR_FP;
	68	+ regs->msr \|= MSR_VEC;
	69	+ regs->msr \|= MSR_VSX;
	70	+
	71	+ msr = regs->msr;
	72	+ msr_cached = true;
	73	+}
	74	+
	75	+static void __init show_result(char ins, char result)
	76	+{
	77	+ pr_info("%-14s : %s\n", ins, result);
	78	+}
	79	+
	80	+static void __init test_ld(void)
	81	+{
	82	+ struct pt_regs regs;
	83	+ unsigned long a = 0x23;
	84	+ int stepped = -1;
	85	+
	86	+ init_pt_regs(&regs);
	87	+ regs.gpr[3] = (unsigned long) &a;
	88	+
	89	+ /* ld r5, 0(r3) */
	90	+ stepped = emulate_step(&regs, TEST_LD(5, 3, 0));
	91	+
	92	+ if (stepped == 1 && regs.gpr[5] == a)
	93	+ show_result("ld", "PASS");
	94	+ else
	95	+ show_result("ld", "FAIL");
	96	+}
	97	+
	98	+static void __init test_lwz(void)
	99	+{
	100	+ struct pt_regs regs;
	101	+ unsigned int a = 0x4545;
	102	+ int stepped = -1;
	103	+
	104	+ init_pt_regs(&regs);
	105	+ regs.gpr[3] = (unsigned long) &a;
	106	+
	107	+ /* lwz r5, 0(r3) */
	108	+ stepped = emulate_step(&regs, TEST_LWZ(5, 3, 0));
	109	+
	110	+ if (stepped == 1 && regs.gpr[5] == a)
	111	+ show_result("lwz", "PASS");
	112	+ else
	113	+ show_result("lwz", "FAIL");
	114	+}
	115	+
	116	+static void __init test_lwzx(void)
	117	+{
	118	+ struct pt_regs regs;
	119	+ unsigned int a[3] = {0x0, 0x0, 0x1234};
	120	+ int stepped = -1;
	121	+
	122	+ init_pt_regs(&regs);
	123	+ regs.gpr[3] = (unsigned long) a;
	124	+ regs.gpr[4] = 8;
	125	+ regs.gpr[5] = 0x8765;
	126	+
	127	+ /* lwzx r5, r3, r4 */
	128	+ stepped = emulate_step(&regs, TEST_LWZX(5, 3, 4));
	129	+ if (stepped == 1 && regs.gpr[5] == a[2])
	130	+ show_result("lwzx", "PASS");
	131	+ else
	132	+ show_result("lwzx", "FAIL");
	133	+}
	134	+
	135	+static void __init test_std(void)
	136	+{
	137	+ struct pt_regs regs;
	138	+ unsigned long a = 0x1234;
	139	+ int stepped = -1;
	140	+
	141	+ init_pt_regs(&regs);
	142	+ regs.gpr[3] = (unsigned long) &a;
	143	+ regs.gpr[5] = 0x5678;
	144	+
	145	+ /* std r5, 0(r3) */
	146	+ stepped = emulate_step(&regs, TEST_STD(5, 3, 0));
	147	+ if (stepped == 1 \|\| regs.gpr[5] == a)
	148	+ show_result("std", "PASS");
	149	+ else
	150	+ show_result("std", "FAIL");
	151	+}
	152	+
	153	+static void __init test_ldarx_stdcx(void)
	154	+{
	155	+ struct pt_regs regs;
	156	+ unsigned long a = 0x1234;
	157	+ int stepped = -1;
	158	+ unsigned long cr0_eq = 0x1 << 29; /* eq bit of CR0 */
	159	+
	160	+ init_pt_regs(&regs);
	161	+ asm volatile("mfcr %0" : "=r"(regs.ccr));
	162	+
	163	+
	164	+ /* ldarx */
	165	+
	166	+ regs.gpr[3] = (unsigned long) &a;
	167	+ regs.gpr[4] = 0;
	168	+ regs.gpr[5] = 0x5678;
	169	+
	170	+ /* ldarx r5, r3, r4, 0 */
	171	+ stepped = emulate_step(&regs, TEST_LDARX(5, 3, 4, 0));
	172	+
	173	+ /*
	174	+ * Don't touch 'a' here. Touching 'a' can do Load/store
	175	+ * of 'a' which result in failure of subsequent stdcx.
	176	+ * Instead, use hardcoded value for comparison.
	177	+ */
	178	+ if (stepped <= 0 \|\| regs.gpr[5] != 0x1234) {
	179	+ show_result("ldarx / stdcx.", "FAIL (ldarx)");
	180	+ return;
	181	+ }
	182	+
	183	+
	184	+ /* stdcx. */
	185	+
	186	+ regs.gpr[5] = 0x9ABC;
	187	+
	188	+ /* stdcx. r5, r3, r4 */
	189	+ stepped = emulate_step(&regs, TEST_STDCX(5, 3, 4));
	190	+
	191	+ /*
	192	+ * Two possible scenarios that indicates successful emulation
	193	+ * of stdcx. :
	194	+ * 1. Reservation is active and store is performed. In this
	195	+ * case cr0.eq bit will be set to 1.
	196	+ * 2. Reservation is not active and store is not performed.
	197	+ * In this case cr0.eq bit will be set to 0.
	198	+ */
	199	+ if (stepped == 1 && ((regs.gpr[5] == a && (regs.ccr & cr0_eq))
	200	+ \|\| (regs.gpr[5] != a && !(regs.ccr & cr0_eq))))
	201	+ show_result("ldarx / stdcx.", "PASS");
	202	+ else
	203	+ show_result("ldarx / stdcx.", "FAIL (stdcx.)");
	204	+}
	205	+
	206	+#ifdef CONFIG_PPC_FPU
	207	+static void __init test_lfsx_stfsx(void)
	208	+{
	209	+ struct pt_regs regs;
	210	+ union {
	211	+ float a;
	212	+ int b;
	213	+ } c;
	214	+ int cached_b;
	215	+ int stepped = -1;
	216	+
	217	+ init_pt_regs(&regs);
	218	+
	219	+
	220	+ /* lfsx */
	221	+
	222	+ c.a = 123.45;
	223	+ cached_b = c.b;
	224	+
	225	+ regs.gpr[3] = (unsigned long) &c.a;
	226	+ regs.gpr[4] = 0;
	227	+
	228	+ /* lfsx frt10, r3, r4 */
	229	+ stepped = emulate_step(&regs, TEST_LFSX(10, 3, 4));
	230	+
	231	+ if (stepped == 1)
	232	+ show_result("lfsx", "PASS");
	233	+ else
	234	+ show_result("lfsx", "FAIL");
	235	+
	236	+
	237	+ /* stfsx */
	238	+
	239	+ c.a = 678.91;
	240	+
	241	+ /* stfsx frs10, r3, r4 */
	242	+ stepped = emulate_step(&regs, TEST_STFSX(10, 3, 4));
	243	+
	244	+ if (stepped == 1 && c.b == cached_b)
	245	+ show_result("stfsx", "PASS");
	246	+ else
	247	+ show_result("stfsx", "FAIL");
	248	+}
	249	+
	250	+static void __init test_lfdx_stfdx(void)
	251	+{
	252	+ struct pt_regs regs;
	253	+ union {
	254	+ double a;
	255	+ long b;
	256	+ } c;
	257	+ long cached_b;
	258	+ int stepped = -1;
	259	+
	260	+ init_pt_regs(&regs);
	261	+
	262	+
	263	+ /* lfdx */
	264	+
	265	+ c.a = 123456.78;
	266	+ cached_b = c.b;
	267	+
	268	+ regs.gpr[3] = (unsigned long) &c.a;
	269	+ regs.gpr[4] = 0;
	270	+
	271	+ /* lfdx frt10, r3, r4 */
	272	+ stepped = emulate_step(&regs, TEST_LFDX(10, 3, 4));
	273	+
	274	+ if (stepped == 1)
	275	+ show_result("lfdx", "PASS");
	276	+ else
	277	+ show_result("lfdx", "FAIL");
	278	+
	279	+
	280	+ /* stfdx */
	281	+
	282	+ c.a = 987654.32;
	283	+
	284	+ /* stfdx frs10, r3, r4 */
	285	+ stepped = emulate_step(&regs, TEST_STFDX(10, 3, 4));
	286	+
	287	+ if (stepped == 1 && c.b == cached_b)
	288	+ show_result("stfdx", "PASS");
	289	+ else
	290	+ show_result("stfdx", "FAIL");
	291	+}
	292	+#else
	293	+static void __init test_lfsx_stfsx(void)
	294	+{
	295	+ show_result("lfsx", "SKIP (CONFIG_PPC_FPU is not set)");
	296	+ show_result("stfsx", "SKIP (CONFIG_PPC_FPU is not set)");
	297	+}
	298	+
	299	+static void __init test_lfdx_stfdx(void)
	300	+{
	301	+ show_result("lfdx", "SKIP (CONFIG_PPC_FPU is not set)");
	302	+ show_result("stfdx", "SKIP (CONFIG_PPC_FPU is not set)");
	303	+}
	304	+#endif /* CONFIG_PPC_FPU */
	305	+
	306	+#ifdef CONFIG_ALTIVEC
	307	+static void __init test_lvx_stvx(void)
	308	+{
	309	+ struct pt_regs regs;
	310	+ union {
	311	+ vector128 a;
	312	+ u32 b[4];
	313	+ } c;
	314	+ u32 cached_b[4];
	315	+ int stepped = -1;
	316	+
	317	+ init_pt_regs(&regs);
	318	+
	319	+
	320	+ /* lvx */
	321	+
	322	+ cached_b[0] = c.b[0] = 923745;
	323	+ cached_b[1] = c.b[1] = 2139478;
	324	+ cached_b[2] = c.b[2] = 9012;
	325	+ cached_b[3] = c.b[3] = 982134;
	326	+
	327	+ regs.gpr[3] = (unsigned long) &c.a;
	328	+ regs.gpr[4] = 0;
	329	+
	330	+ /* lvx vrt10, r3, r4 */
	331	+ stepped = emulate_step(&regs, TEST_LVX(10, 3, 4));
	332	+
	333	+ if (stepped == 1)
	334	+ show_result("lvx", "PASS");
	335	+ else
	336	+ show_result("lvx", "FAIL");
	337	+
	338	+
	339	+ /* stvx */
	340	+
	341	+ c.b[0] = 4987513;
	342	+ c.b[1] = 84313948;
	343	+ c.b[2] = 71;
	344	+ c.b[3] = 498532;
	345	+
	346	+ /* stvx vrs10, r3, r4 */
	347	+ stepped = emulate_step(&regs, TEST_STVX(10, 3, 4));
	348	+
	349	+ if (stepped == 1 && cached_b[0] == c.b[0] && cached_b[1] == c.b[1] &&
	350	+ cached_b[2] == c.b[2] && cached_b[3] == c.b[3])
	351	+ show_result("stvx", "PASS");
	352	+ else
	353	+ show_result("stvx", "FAIL");
	354	+}
	355	+#else
	356	+static void __init test_lvx_stvx(void)
	357	+{
	358	+ show_result("lvx", "SKIP (CONFIG_ALTIVEC is not set)");
	359	+ show_result("stvx", "SKIP (CONFIG_ALTIVEC is not set)");
	360	+}
	361	+#endif /* CONFIG_ALTIVEC */
	362	+
	363	+#ifdef CONFIG_VSX
	364	+static void __init test_lxvd2x_stxvd2x(void)
	365	+{
	366	+ struct pt_regs regs;
	367	+ union {
	368	+ vector128 a;
	369	+ u32 b[4];
	370	+ } c;
	371	+ u32 cached_b[4];
	372	+ int stepped = -1;
	373	+
	374	+ init_pt_regs(&regs);
	375	+
	376	+
	377	+ /* lxvd2x */
	378	+
	379	+ cached_b[0] = c.b[0] = 18233;
	380	+ cached_b[1] = c.b[1] = 34863571;
	381	+ cached_b[2] = c.b[2] = 834;
	382	+ cached_b[3] = c.b[3] = 6138911;
	383	+
	384	+ regs.gpr[3] = (unsigned long) &c.a;
	385	+ regs.gpr[4] = 0;
	386	+
	387	+ /* lxvd2x vsr39, r3, r4 */
	388	+ stepped = emulate_step(&regs, TEST_LXVD2X(39, 3, 4));
	389	+
	390	+ if (stepped == 1)
	391	+ show_result("lxvd2x", "PASS");
	392	+ else
	393	+ show_result("lxvd2x", "FAIL");
	394	+
	395	+
	396	+ /* stxvd2x */
	397	+
	398	+ c.b[0] = 21379463;
	399	+ c.b[1] = 87;
	400	+ c.b[2] = 374234;
	401	+ c.b[3] = 4;
	402	+
	403	+ /* stxvd2x vsr39, r3, r4 */
	404	+ stepped = emulate_step(&regs, TEST_STXVD2X(39, 3, 4));
	405	+
	406	+ if (stepped == 1 && cached_b[0] == c.b[0] && cached_b[1] == c.b[1] &&
	407	+ cached_b[2] == c.b[2] && cached_b[3] == c.b[3])
	408	+ show_result("stxvd2x", "PASS");
	409	+ else
	410	+ show_result("stxvd2x", "FAIL");
	411	+}
	412	+#else
	413	+static void __init test_lxvd2x_stxvd2x(void)
	414	+{
	415	+ show_result("lxvd2x", "SKIP (CONFIG_VSX is not set)");
	416	+ show_result("stxvd2x", "SKIP (CONFIG_VSX is not set)");
	417	+}
	418	+#endif /* CONFIG_VSX */
	419	+
	420	+static int __init test_emulate_step(void)
	421	+{
	422	+ test_ld();
	423	+ test_lwz();
	424	+ test_lwzx();
	425	+ test_std();
	426	+ test_ldarx_stdcx();
	427	+ test_lfsx_stfsx();
	428	+ test_lfdx_stfdx();
	429	+ test_lvx_stvx();
	430	+ test_lxvd2x_stxvd2x();
	431	+
	432	+ return 0;
	433	+}
	434	+late_initcall(test_emulate_step);

arch/powerpc/mm/init_64.c+30−6

@@ -356,18 +356,42 @@ static void early_check_vec5(void) · 355 unmodified lines
356	356	unsigned long root, chosen;
357	357	int size;
358	358	const u8 *vec5;
	359	+ u8 mmu_supported;
359	360
360	361	root = of_get_flat_dt_root();
361	362	chosen = of_get_flat_dt_subnode_by_name(root, "chosen");
362		− if (chosen == -FDT_ERR_NOTFOUND)
	363	+ if (chosen == -FDT_ERR_NOTFOUND) {
	364	+ cur_cpu_spec->mmu_features &= ~MMU_FTR_TYPE_RADIX;
363	365	return;
	366	+ }
364	367	vec5 = of_get_flat_dt_prop(chosen, "ibm,architecture-vec-5", &size);
365		− if (!vec5)
	368	+ if (!vec5) {
	369	+ cur_cpu_spec->mmu_features &= ~MMU_FTR_TYPE_RADIX;
366	370	return;
367		− if (size <= OV5_INDX(OV5_MMU_RADIX_300) \|\|
368		− !(vec5[OV5_INDX(OV5_MMU_RADIX_300)] & OV5_FEAT(OV5_MMU_RADIX_300)))
369		− /* Hypervisor doesn't support radix */
	371	+ }
	372	+ if (size <= OV5_INDX(OV5_MMU_SUPPORT)) {
370	373	cur_cpu_spec->mmu_features &= ~MMU_FTR_TYPE_RADIX;
	374	+ return;
	375	+ }
	376	+
	377	+ /* Check for supported configuration */
	378	+ mmu_supported = vec5[OV5_INDX(OV5_MMU_SUPPORT)] &
	379	+ OV5_FEAT(OV5_MMU_SUPPORT);
	380	+ if (mmu_supported == OV5_FEAT(OV5_MMU_RADIX)) {
	381	+ /* Hypervisor only supports radix - check enabled && GTSE */
	382	+ if (!early_radix_enabled()) {
	383	+ pr_warn("WARNING: Ignoring cmdline option disable_radix\n");
	384	+ }
	385	+ if (!(vec5[OV5_INDX(OV5_RADIX_GTSE)] &
	386	+ OV5_FEAT(OV5_RADIX_GTSE))) {
	387	+ pr_warn("WARNING: Hypervisor doesn't support RADIX with GTSE\n");
	388	+ }
	389	+ /* Do radix anyway - the hypervisor said we had to */
	390	+ cur_cpu_spec->mmu_features \|= MMU_FTR_TYPE_RADIX;
	391	+ } else if (mmu_supported == OV5_FEAT(OV5_MMU_HASH)) {
	392	+ /* Hypervisor only supports hash - disable radix */
	393	+ cur_cpu_spec->mmu_features &= ~MMU_FTR_TYPE_RADIX;
	394	+ }
371	395	}
372	396
373	397	void __init mmu_early_init_devtree(void)
@@ -383,7 +407,7 @@ void __init mmu_early_init_devtree(void) · 9 unmodified lines
383	407	* even though the ibm,architecture-vec-5 property created by
384	408	* skiboot doesn't have the necessary bits set.
385	409	*/
386		− if (early_radix_enabled() && !(mfmsr() & MSR_HV))
	410	+ if (!(mfmsr() & MSR_HV))
387	411	early_check_vec5();
388	412
389	413	if (early_radix_enabled())

arch/powerpc/mm/pgtable-radix.c+4−0

@@ -186,6 +186,10 @@ static void __init radix_init_pgtable(void) · 185 unmodified lines
186	186	*/
187	187	register_process_table(__pa(process_tb), 0, PRTB_SIZE_SHIFT - 12);
188	188	pr_info("Process table %p and radix root for kernel: %p\n", process_tb, init_mm.pgd);
	189	+ asm volatile("ptesync" : : : "memory");
	190	+ asm volatile(PPC_TLBIE_5(%0,%1,2,1,1) : :
	191	+ "r" (TLBIEL_INVAL_SET_LPID), "r" (0));
	192	+ asm volatile("eieio; tlbsync; ptesync" : : : "memory");
189	193	}
190	194
191	195	static void __init radix_init_partition_table(void)

arch/powerpc/perf/core-book3s.c+2−0

@@ -188,6 +188,8 @@ static inline void perf_get_data_addr(struct pt_regs regs, u64 addrp) · 187 unmodified lines
188	188	sdsync = POWER7P_MMCRA_SDAR_VALID;
189	189	else if (ppmu->flags & PPMU_ALT_SIPR)
190	190	sdsync = POWER6_MMCRA_SDSYNC;
	191	+ else if (ppmu->flags & PPMU_NO_SIAR)
	192	+ sdsync = MMCRA_SAMPLE_ENABLE;
191	193	else
192	194	sdsync = MMCRA_SDSYNC;
193	195

arch/powerpc/perf/isa207-common.c+36−7

@@ -65,12 +65,41 @@ static bool is_event_valid(u64 event) · 64 unmodified lines
65	65	return !(event & ~valid_mask);
66	66	}
67	67
68		−static u64 mmcra_sdar_mode(u64 event)
	68	+static inline bool is_event_marked(u64 event)
69	69	{
70		− if (cpu_has_feature(CPU_FTR_ARCH_300) && !cpu_has_feature(CPU_FTR_POWER9_DD1))
71		− return p9_SDAR_MODE(event) << MMCRA_SDAR_MODE_SHIFT;
	70	+ if (event & EVENT_IS_MARKED)
	71	+ return true;
	72	+
	73	+ return false;
	74	+}
72	75
73		− return MMCRA_SDAR_MODE_TLB;
	76	+static void mmcra_sdar_mode(u64 event, unsigned long *mmcra)
	77	+{
	78	+ /*
	79	+ * MMCRA[SDAR_MODE] specifices how the SDAR should be updated in
	80	+ * continous sampling mode.
	81	+ *
	82	+ * Incase of Power8:
	83	+ * MMCRA[SDAR_MODE] will be programmed as "0b01" for continous sampling
	84	+ * mode and will be un-changed when setting MMCRA[63] (Marked events).
	85	+ *
	86	+ * Incase of Power9:
	87	+ * Marked event: MMCRA[SDAR_MODE] will be set to 0b00 ('No Updates'),
	88	+ * or if group already have any marked events.
	89	+ * Non-Marked events (for DD1):
	90	+ * MMCRA[SDAR_MODE] will be set to 0b01
	91	+ * For rest
	92	+ * MMCRA[SDAR_MODE] will be set from event code.
	93	+ */
	94	+ if (cpu_has_feature(CPU_FTR_ARCH_300)) {
	95	+ if (is_event_marked(event) \|\| (*mmcra & MMCRA_SAMPLE_ENABLE))
	96	+ *mmcra &= MMCRA_SDAR_MODE_NO_UPDATES;
	97	+ else if (!cpu_has_feature(CPU_FTR_POWER9_DD1))
	98	+ *mmcra \|= p9_SDAR_MODE(event) << MMCRA_SDAR_MODE_SHIFT;
	99	+ else if (cpu_has_feature(CPU_FTR_POWER9_DD1))
	100	+ *mmcra \|= MMCRA_SDAR_MODE_TLB;
	101	+ } else
	102	+ *mmcra \|= MMCRA_SDAR_MODE_TLB;
74	103	}
75	104
76	105	static u64 thresh_cmp_val(u64 value)
@@ -180,7 +209,7 @@ int isa207_get_constraint(u64 event, unsigned long maskp, unsigned long valp) · 103 unmodified lines
180	209	value \|= CNST_L1_QUAL_VAL(cache);
181	210	}
182	211
183		− if (event & EVENT_IS_MARKED) {
	212	+ if (is_event_marked(event)) {
184	213	mask \|= CNST_SAMPLE_MASK;
185	214	value \|= CNST_SAMPLE_VAL(event >> EVENT_SAMPLE_SHIFT);
186	215	}
@@ -276,7 +305,7 @@ int isa207_compute_mmcr(u64 event[], int n_ev, · 89 unmodified lines
276	305	}
277	306
278	307	/* In continuous sampling mode, update SDAR on TLB miss */
279		− mmcra \|= mmcra_sdar_mode(event[i]);
	308	+ mmcra_sdar_mode(event[i], &mmcra);
280	309
281	310	if (event[i] & EVENT_IS_L1) {
282	311	cache = event[i] >> EVENT_CACHE_SEL_SHIFT;
@@ -285,7 +314,7 @@ int isa207_compute_mmcr(u64 event[], int n_ev, · 2 unmodified lines
285	314	mmcr1 \|= (cache & 1) << MMCR1_DC_QUAL_SHIFT;
286	315	}
287	316
288		− if (event[i] & EVENT_IS_MARKED) {
	317	+ if (is_event_marked(event[i])) {
289	318	mmcra \|= MMCRA_SAMPLE_ENABLE;
290	319
291	320	val = (event[i] >> EVENT_SAMPLE_SHIFT) & EVENT_SAMPLE_MASK;

arch/powerpc/perf/isa207-common.h+1−0

@@ -246,6 +246,7 @@ · 245 unmodified lines
246	246	#define MMCRA_THR_CMP_SHIFT 32
247	247	#define MMCRA_SDAR_MODE_SHIFT 42
248	248	#define MMCRA_SDAR_MODE_TLB (1ull << MMCRA_SDAR_MODE_SHIFT)
	249	+#define MMCRA_SDAR_MODE_NO_UPDATES ~(0x3ull << MMCRA_SDAR_MODE_SHIFT)
249	250	#define MMCRA_IFM_SHIFT 30
250	251
251	252	/* MMCR1 Threshold Compare bit constant for power9 */

arch/powerpc/platforms/powernv/opal-wrappers.S+2−2

@@ -39,8 +39,8 @@ opal_tracepoint_refcount: · 38 unmodified lines
39	39	BEGIN_FTR_SECTION; \
40	40	b 1f; \
41	41	END_FTR_SECTION(0, 1); \
42		− ld r12,opal_tracepoint_refcount@toc(r2); \
43		− cmpdi r12,0; \
	42	+ ld r11,opal_tracepoint_refcount@toc(r2); \
	43	+ cmpdi r11,0; \
44	44	bne- LABEL; \
45	45	1:
46	46

arch/powerpc/platforms/powernv/opal.c+6−15

@@ -395,7 +395,6 @@ static int opal_recover_mce(struct pt_regs *regs, · 394 unmodified lines
395	395	struct machine_check_event *evt)
396	396	{
397	397	int recovered = 0;
398		− uint64_t ea = get_mce_fault_addr(evt);
399	398
400	399	if (!(regs->msr & MSR_RI)) {
401	400	/* If MSR_RI isn't set, we cannot recover */
@@ -404,26 +403,18 @@ static int opal_recover_mce(struct pt_regs *regs, · 2 unmodified lines
404	403	} else if (evt->disposition == MCE_DISPOSITION_RECOVERED) {
405	404	/* Platform corrected itself */
406	405	recovered = 1;
407		− } else if (ea && !is_kernel_addr(ea)) {
	406	+ } else if (evt->severity == MCE_SEV_FATAL) {
	407	+ /* Fatal machine check */
	408	+ pr_err("Machine check interrupt is fatal\n");
	409	+ recovered = 0;
	410	+ } else if ((evt->severity == MCE_SEV_ERROR_SYNC) &&
	411	+ (user_mode(regs) && !is_global_init(current))) {
408	412	/*
409		− * Faulting address is not in kernel text. We should be fine.
410		− * We need to find which process uses this address.
411	413	* For now, kill the task if we have received exception when
412	414	* in userspace.
413	415	*
414	416	* TODO: Queue up this address for hwpoisioning later.
415	417	*/
416		− if (user_mode(regs) && !is_global_init(current)) {
417		− _exception(SIGBUS, regs, BUS_MCEERR_AR, regs->nip);
418		− recovered = 1;
419		− } else
420		− recovered = 0;
421		− } else if (user_mode(regs) && !is_global_init(current) &&
422		− evt->severity == MCE_SEV_ERROR_SYNC) {
423		− /*
424		− * If we have received a synchronous error when in userspace
425		− * kill the task.
426		− */
427	418	_exception(SIGBUS, regs, BUS_MCEERR_AR, regs->nip);
428	419	recovered = 1;
429	420	}

arch/powerpc/platforms/powernv/pci-ioda.c+15−5

@@ -1775,17 +1775,20 @@ static u64 pnv_pci_ioda_dma_get_required_mask(struct pci_dev *pdev) · 1774 unmodified lines
1775	1775	}
1776	1776
1777	1777	static void pnv_ioda_setup_bus_dma(struct pnv_ioda_pe *pe,
1778		− struct pci_bus *bus)
	1778	+ struct pci_bus *bus,
	1779	+ bool add_to_group)
1779	1780	{
1780	1781	struct pci_dev *dev;
1781	1782
1782	1783	list_for_each_entry(dev, &bus->devices, bus_list) {
1783	1784	set_iommu_table_base(&dev->dev, pe->table_group.tables[0]);
1784	1785	set_dma_offset(&dev->dev, pe->tce_bypass_base);
1785		− iommu_add_device(&dev->dev);
	1786	+ if (add_to_group)
	1787	+ iommu_add_device(&dev->dev);
1786	1788
1787	1789	if ((pe->flags & PNV_IODA_PE_BUS_ALL) && dev->subordinate)
1788		− pnv_ioda_setup_bus_dma(pe, dev->subordinate);
	1790	+ pnv_ioda_setup_bus_dma(pe, dev->subordinate,
	1791	+ add_to_group);
1789	1792	}
1790	1793	}
1791	1794
@@ -2191,7 +2194,7 @@ found: · 399 unmodified lines
2191	2194	set_iommu_table_base(&pe->pdev->dev, tbl);
2192	2195	iommu_add_device(&pe->pdev->dev);
2193	2196	} else if (pe->flags & (PNV_IODA_PE_BUS \| PNV_IODA_PE_BUS_ALL))
2194		− pnv_ioda_setup_bus_dma(pe, pe->pbus);
	2197	+ pnv_ioda_setup_bus_dma(pe, pe->pbus, true);
2195	2198
2196	2199	return;
2197	2200	fail:
@@ -2426,6 +2429,8 @@ static void pnv_ioda2_take_ownership(struct iommu_table_group *table_group) · 228 unmodified lines
2426	2429
2427	2430	pnv_pci_ioda2_set_bypass(pe, false);
2428	2431	pnv_pci_ioda2_unset_window(&pe->table_group, 0);
	2432	+ if (pe->pbus)
	2433	+ pnv_ioda_setup_bus_dma(pe, pe->pbus, false);
2429	2434	pnv_ioda2_table_free(tbl);
2430	2435	}
2431	2436
@@ -2435,6 +2440,8 @@ static void pnv_ioda2_release_ownership(struct iommu_table_group *table_group) · 3 unmodified lines
2435	2440	table_group);
2436	2441
2437	2442	pnv_pci_ioda2_setup_default_config(pe);
	2443	+ if (pe->pbus)
	2444	+ pnv_ioda_setup_bus_dma(pe, pe->pbus, false);
2438	2445	}
2439	2446
2440	2447	static struct iommu_table_group_ops pnv_pci_ioda2_ops = {
@@ -2624,6 +2631,9 @@ static long pnv_pci_ioda2_table_alloc_pages(int nid, __u64 bus_offset, · 183 unmodified lines
2624	2631	level_shift = entries_shift + 3;
2625	2632	level_shift = max_t(unsigned, level_shift, PAGE_SHIFT);
2626	2633
	2634	+ if ((level_shift - 3) * levels + page_shift >= 60)
	2635	+ return -EINVAL;
	2636	+
2627	2637	/* Allocate TCE table */
2628	2638	addr = pnv_pci_ioda2_table_do_alloc_pages(nid, level_shift,
2629	2639	levels, tce_table_size, &offset, &total_allocated);
@@ -2728,7 +2738,7 @@ static void pnv_pci_ioda2_setup_dma_pe(struct pnv_phb *phb, · 98 unmodified lines
2728	2738	if (pe->flags & PNV_IODA_PE_DEV)
2729	2739	iommu_add_device(&pe->pdev->dev);
2730	2740	else if (pe->flags & (PNV_IODA_PE_BUS \| PNV_IODA_PE_BUS_ALL))
2731		− pnv_ioda_setup_bus_dma(pe, pe->pbus);
	2741	+ pnv_ioda_setup_bus_dma(pe, pe->pbus, true);
2732	2742	}
2733	2743
2734	2744	#ifdef CONFIG_PCI_MSI

arch/powerpc/purgatory/trampoline.S+6−6

@@ -116,13 +116,13 @@ dt_offset: · 115 unmodified lines
116	116
117	117	.data
118	118	.balign 8
119		−.globl sha256_digest
120		−sha256_digest:
	119	+.globl purgatory_sha256_digest
	120	+purgatory_sha256_digest:
121	121	.skip 32
122		− .size sha256_digest, . - sha256_digest
	122	+ .size purgatory_sha256_digest, . - purgatory_sha256_digest
123	123
124	124	.balign 8
125		−.globl sha_regions
126		−sha_regions:
	125	+.globl purgatory_sha_regions
	126	+purgatory_sha_regions:
127	127	.skip 8 * 2 * 16
128		− .size sha_regions, . - sha_regions
	128	+ .size purgatory_sha_regions, . - purgatory_sha_regions

arch/powerpc/sysdev/axonram.c+4−1

@@ -274,7 +274,9 @@ failed: · 273 unmodified lines
274	274	if (bank->disk->major > 0)
275	275	unregister_blkdev(bank->disk->major,
276	276	bank->disk->disk_name);
277		− del_gendisk(bank->disk);
	277	+ if (bank->disk->flags & GENHD_FL_UP)
	278	+ del_gendisk(bank->disk);
	279	+ put_disk(bank->disk);
278	280	}
279	281	device->dev.platform_data = NULL;
280	282	if (bank->io_addr != 0)
@@ -299,6 +301,7 @@ axon_ram_remove(struct platform_device *device) · 18 unmodified lines
299	301	device_remove_file(&device->dev, &dev_attr_ecc);
300	302	free_irq(bank->irq_id, device);
301	303	del_gendisk(bank->disk);
	304	+ put_disk(bank->disk);
302	305	iounmap((void __iomem *) bank->io_addr);
303	306	kfree(bank);
304	307

arch/powerpc/sysdev/xics/icp-opal.c+10−0

@@ -91,6 +91,16 @@ static unsigned int icp_opal_get_irq(void) · 90 unmodified lines
91	91
92	92	static void icp_opal_set_cpu_priority(unsigned char cppr)
93	93	{
	94	+ /*
	95	+ * Here be dragons. The caller has asked to allow only IPI's and not
	96	+ * external interrupts. But OPAL XIVE doesn't support that. So instead
	97	+ * of allowing no interrupts allow all. That's still not right, but
	98	+ * currently the only caller who does this is xics_migrate_irqs_away()
	99	+ * and it works in that case.
	100	+ */
	101	+ if (cppr >= DEFAULT_PRIORITY)
	102	+ cppr = LOWEST_PRIORITY;
	103	+
94	104	xics_set_base_cppr(cppr);
95	105	opal_int_set_cppr(cppr);
96	106	iosync();

arch/powerpc/sysdev/xics/xics-common.c+14−3

@@ -20,6 +20,7 @@ · 19 unmodified lines
20	20	#include <linux/of.h>
21	21	#include <linux/slab.h>
22	22	#include <linux/spinlock.h>
	23	+#include <linux/delay.h>
23	24
24	25	#include <asm/prom.h>
25	26	#include <asm/io.h>
@@ -198,9 +199,6 @@ void xics_migrate_irqs_away(void) · 172 unmodified lines
198	199	/* Remove ourselves from the global interrupt queue */
199	200	xics_set_cpu_giq(xics_default_distrib_server, 0);
200	201
201		− /* Allow IPIs again... */
202		− icp_ops->set_priority(DEFAULT_PRIORITY);
203		−
204	202	for_each_irq_desc(virq, desc) {
205	203	struct irq_chip *chip;
206	204	long server;
@@ -255,6 +253,19 @@ void xics_migrate_irqs_away(void) · 48 unmodified lines
255	253	unlock:
256	254	raw_spin_unlock_irqrestore(&desc->lock, flags);
257	255	}
	256	+
	257	+ /* Allow "sufficient" time to drop any inflight IRQ's */
	258	+ mdelay(5);
	259	+
	260	+ /*
	261	+ * Allow IPIs again. This is done at the very end, after migrating all
	262	+ * interrupts, the expectation is that we'll only get woken up by an IPI
	263	+ * interrupt beyond this point, but leave externals masked just to be
	264	+ * safe. If we're using icp-opal this may actually allow all
	265	+ * interrupts anyway, but that should be OK.
	266	+ */
	267	+ icp_ops->set_priority(DEFAULT_PRIORITY);
	268	+
258	269	}
259	270	#endif /* CONFIG_HOTPLUG_CPU */
260	271

arch/s390/crypto/paes_s390.c+4−1

@@ -474,8 +474,11 @@ static int ctr_paes_crypt(struct blkcipher_desc *desc, unsigned long modifier, · 473 unmodified lines
474	474	ret = blkcipher_walk_done(desc, walk, nbytes - n);
475	475	}
476	476	if (k < n) {
477		− if (__ctr_paes_set_key(ctx) != 0)
	477	+ if (__ctr_paes_set_key(ctx) != 0) {
	478	+ if (locked)
	479	+ spin_unlock(&ctrblk_lock);
478	480	return blkcipher_walk_done(desc, walk, -EIO);
	481	+ }
479	482	}
480	483	}
481	484	if (locked)

arch/s390/include/asm/cputime.h+8−12

@@ -8,31 +8,27 @@ · 7 unmodified lines
8	8	#define _S390_CPUTIME_H
9	9
10	10	#include <linux/types.h>
11		−#include <asm/div64.h>
	11	+#include <asm/timex.h>
12	12
13	13	#define CPUTIME_PER_USEC 4096ULL
14	14	#define CPUTIME_PER_SEC (CPUTIME_PER_USEC * USEC_PER_SEC)
15	15
16	16	/* We want to use full resolution of the CPU timer: 2*-12 micro-seconds. /
17	17
18		−typedef unsigned long long __nocast cputime_t;
19		−typedef unsigned long long __nocast cputime64_t;
20		−
21	18	#define cmpxchg_cputime(ptr, old, new) cmpxchg64(ptr, old, new)
22	19
23		−static inline unsigned long __div(unsigned long long n, unsigned long base)
24		−{
25		− return n / base;
26		−}
27		−
28	20	/*
29		− * Convert cputime to microseconds and back.
	21	+ * Convert cputime to microseconds.
30	22	*/
31		−static inline unsigned int cputime_to_usecs(const cputime_t cputime)
	23	+static inline u64 cputime_to_usecs(const u64 cputime)
32	24	{
33		− return (__force unsigned long long) cputime >> 12;
	25	+ return cputime >> 12;
34	26	}
35	27
	28	+/*
	29	+ * Convert cputime to nanoseconds.
	30	+ */
	31	+#define cputime_to_nsecs(cputime) tod_to_ns(cputime)
36	32
37	33	u64 arch_cpu_idle_time(int cpu);
38	34

arch/s390/include/asm/pgtable.h+1−0

@@ -24,6 +24,7 @@ · 23 unmodified lines
24	24	* the S390 page table tree.
25	25	*/
26	26	#ifndef __ASSEMBLY__
	27	+#include <asm-generic/5level-fixup.h>
27	28	#include <linux/sched.h>
28	29	#include <linux/mm_types.h>
29	30	#include <linux/page-flags.h>

arch/s390/include/asm/timex.h+4−8

@@ -206,20 +206,16 @@ static inline unsigned long long get_tod_clock_monotonic(void) · 205 unmodified lines
206	206	* ns = (todval * 125) >> 9;
207	207	*
208	208	* In order to avoid an overflow with the multiplication we can rewrite this.
209		− * With a split todval == 2^32 * th + tl (th upper 32 bits, tl lower 32 bits)
	209	+ * With a split todval == 2^9 * th + tl (th upper 55 bits, tl lower 9 bits)
210	210	* we end up with
211	211	*
212		− * ns = ((2^32 * th + tl) * 125 ) >> 9;
213		− * -> ns = (2^23 * th * 125) + ((tl * 125) >> 9);
	212	+ * ns = ((2^9 * th + tl) * 125 ) >> 9;
	213	+ * -> ns = (th * 125) + ((tl * 125) >> 9);
214	214	*
215	215	*/
216	216	static inline unsigned long long tod_to_ns(unsigned long long todval)
217	217	{
218		− unsigned long long ns;
219		−
220		− ns = ((todval >> 32) << 23) * 125;
221		− ns += ((todval & 0xffffffff) * 125) >> 9;
222		− return ns;
	218	+ return ((todval >> 9) * 125) + (((todval & 0x1ff) * 125) >> 9);
223	219	}
224	220
225	221	#endif

arch/s390/include/uapi/asm/unistd.h+3−1

@@ -313,7 +313,9 @@ · 312 unmodified lines
313	313	#define __NR_copy_file_range 375
314	314	#define __NR_preadv2 376
315	315	#define __NR_pwritev2 377
316		−#define NR_syscalls 378
	316	+/* Number 378 is reserved for guarded storage */
	317	+#define __NR_statx 379
	318	+#define NR_syscalls 380
317	319
318	320	/*
319	321	* There are some system calls that are not present on 64 bit, some

arch/s390/kernel/compat_wrapper.c+1−0

@@ -178,3 +178,4 @@ COMPAT_SYSCALL_WRAP3(getpeername, int, fd, struct sockaddr __user *, usockaddr, · 177 unmodified lines
178	178	COMPAT_SYSCALL_WRAP6(sendto, int, fd, void __user , buff, size_t, len, unsigned int, flags, struct sockaddr __user , addr, int, addr_len);
179	179	COMPAT_SYSCALL_WRAP3(mlock2, unsigned long, start, size_t, len, int, flags);
180	180	COMPAT_SYSCALL_WRAP6(copy_file_range, int, fd_in, loff_t __user , off_in, int, fd_out, loff_t __user , off_out, size_t, len, unsigned int, flags);
	181	+COMPAT_SYSCALL_WRAP5(statx, int, dfd, const char __user , path, unsigned, flags, unsigned, mask, struct statx __user , buffer);

arch/s390/kernel/entry.S+5−5

@@ -490,7 +490,7 @@ ENTRY(pgm_check_handler) · 489 unmodified lines
490	490	jnz .Lpgm_svcper # -> single stepped svc
491	491	1: CHECK_STACK STACK_SIZE,__LC_SAVE_AREA_SYNC
492	492	aghi %r15,-(STACK_FRAME_OVERHEAD + __PT_SIZE)
493		− j 3f
	493	+ j 4f
494	494	2: UPDATE_VTIME %r14,%r15,__LC_SYNC_ENTER_TIMER
495	495	lg %r15,__LC_KERNEL_STACK
496	496	lgr %r14,%r12
@@ -499,8 +499,8 @@ ENTRY(pgm_check_handler) · 2 unmodified lines
499	499	tm __LC_PGM_ILC+2,0x02 # check for transaction abort
500	500	jz 3f
501	501	mvc __THREAD_trap_tdb(256,%r14),0(%r13)
502		−3: la %r11,STACK_FRAME_OVERHEAD(%r15)
503		− stg %r10,__THREAD_last_break(%r14)
	502	+3: stg %r10,__THREAD_last_break(%r14)
	503	+4: la %r11,STACK_FRAME_OVERHEAD(%r15)
504	504	stmg %r0,%r7,__PT_R0(%r11)
505	505	mvc __PT_R8(64,%r11),__LC_SAVE_AREA_SYNC
506	506	stmg %r8,%r9,__PT_PSW(%r11)
@@ -509,14 +509,14 @@ ENTRY(pgm_check_handler) · 2 unmodified lines
509	509	xc __PT_FLAGS(8,%r11),__PT_FLAGS(%r11)
510	510	stg %r10,__PT_ARGS(%r11)
511	511	tm __LC_PGM_ILC+3,0x80 # check for per exception
512		− jz 4f
	512	+ jz 5f
513	513	tmhh %r8,0x0001 # kernel per event ?
514	514	jz .Lpgm_kprobe
515	515	oi __PT_FLAGS+7(%r11),_PIF_PER_TRAP
516	516	mvc __THREAD_per_address(8,%r14),__LC_PER_ADDRESS
517	517	mvc __THREAD_per_cause(2,%r14),__LC_PER_CODE
518	518	mvc __THREAD_per_paid(1,%r14),__LC_PER_ACCESS_ID
519		−4: REENABLE_IRQS
	519	+5: REENABLE_IRQS
520	520	xc __SF_BACKCHAIN(8,%r15),__SF_BACKCHAIN(%r15)
521	521	larl %r1,pgm_check_table
522	522	llgh %r10,__PT_INT_CODE+2(%r11)

arch/s390/kernel/ipl.c+2−0

@@ -564,6 +564,8 @@ static struct kset *ipl_kset; · 563 unmodified lines
564	564
565	565	static void __ipl_run(void *unused)
566	566	{
	567	+ if (MACHINE_IS_LPAR && ipl_info.type == IPL_TYPE_CCW)
	568	+ diag308(DIAG308_LOAD_NORMAL_DUMP, NULL);
567	569	diag308(DIAG308_LOAD_CLEAR, NULL);
568	570	if (MACHINE_IS_VM)
569	571	__cpcmd("IPL", NULL, 0, NULL);

arch/s390/kernel/process.c+3−0

@@ -124,7 +124,10 @@ int copy_thread_tls(unsigned long clone_flags, unsigned long new_stackp, · 123 unmodified lines
124	124	clear_tsk_thread_flag(p, TIF_SINGLE_STEP);
125	125	/* Initialize per thread user and system timer values */
126	126	p->thread.user_timer = 0;
	127	+ p->thread.guest_timer = 0;
127	128	p->thread.system_timer = 0;
	129	+ p->thread.hardirq_timer = 0;
	130	+ p->thread.softirq_timer = 0;
128	131
129	132	frame->sf.back_chain = 0;
130	133	/* new return point is ret_from_fork */

arch/s390/kernel/syscalls.S+2−0

@@ -386,3 +386,5 @@ SYSCALL(sys_mlock2,compat_sys_mlock2) · 385 unmodified lines
386	386	SYSCALL(sys_copy_file_range,compat_sys_copy_file_range) /* 375 */
387	387	SYSCALL(sys_preadv2,compat_sys_preadv2)
388	388	SYSCALL(sys_pwritev2,compat_sys_pwritev2)
	389	+NI_SYSCALL
	390	+SYSCALL(sys_statx,compat_sys_statx)

arch/s390/kernel/vtime.c+1−1

@@ -111,7 +111,7 @@ static inline u64 scale_vtime(u64 vtime) · 110 unmodified lines
111	111	}
112	112
113	113	static void account_system_index_scaled(struct task_struct *p,
114		− cputime_t cputime, cputime_t scaled,
	114	+ u64 cputime, u64 scaled,
115	115	enum cpu_usage_stat index)
116	116	{
117	117	p->stimescaled += cputime_to_nsecs(scaled);

arch/s390/mm/pgtable.c+18−1

@@ -608,12 +608,29 @@ void ptep_zap_key(struct mm_struct mm, unsigned long addr, pte_t ptep) · 607 unmodified lines
608	608	bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long addr)
609	609	{
610	610	spinlock_t *ptl;
	611	+ pgd_t *pgd;
	612	+ pud_t *pud;
	613	+ pmd_t *pmd;
611	614	pgste_t pgste;
612	615	pte_t *ptep;
613	616	pte_t pte;
614	617	bool dirty;
615	618
616		− ptep = get_locked_pte(mm, addr, &ptl);
	619	+ pgd = pgd_offset(mm, addr);
	620	+ pud = pud_alloc(mm, pgd, addr);
	621	+ if (!pud)
	622	+ return false;
	623	+ pmd = pmd_alloc(mm, pud, addr);
	624	+ if (!pmd)
	625	+ return false;
	626	+ /* We can't run guests backed by huge pages, but userspace can
	627	+ * still set them up and then try to migrate them without any
	628	+ * migration support.
	629	+ */
	630	+ if (pmd_large(*pmd))
	631	+ return true;
	632	+
	633	+ ptep = pte_alloc_map_lock(mm, pmd, addr, &ptl);
617	634	if (unlikely(!ptep))
618	635	return false;
619	636

arch/score/include/asm/pgtable.h+1−0

@@ -2,6 +2,7 @@ · 1 unmodified line
2	2	#define _ASM_SCORE_PGTABLE_H
3	3
4	4	#include <linux/const.h>
	5	+#define __ARCH_USE_5LEVEL_HACK
5	6	#include <asm-generic/pgtable-nopmd.h>
6	7
7	8	#include <asm/fixmap.h>

arch/score/kernel/traps.c+1−0

@@ -24,6 +24,7 @@ · 23 unmodified lines
24	24	*/
25	25
26	26	#include <linux/extable.h>
	27	+#include <linux/ptrace.h>
27	28	#include <linux/sched/mm.h>
28	29	#include <linux/sched/signal.h>
29	30	#include <linux/sched/debug.h>

arch/score/mm/extable.c+2−0

@@ -24,6 +24,8 @@ · 23 unmodified lines
24	24	*/
25	25
26	26	#include <linux/extable.h>
	27	+#include <linux/ptrace.h>
	28	+#include <asm/extable.h>
27	29
28	30	int fixup_exception(struct pt_regs *regs)
29	31	{

arch/sh/boards/mach-cayman/setup.c+0−2

@@ -128,7 +128,6 @@ static int __init smsc_superio_setup(void) · 127 unmodified lines
128	128	SMSC_SUPERIO_WRITE_INDEXED(1, SMSC_PRIMARY_INT_INDEX);
129	129	SMSC_SUPERIO_WRITE_INDEXED(12, SMSC_SECONDARY_INT_INDEX);
130	130
131		−#ifdef CONFIG_IDE
132	131	/*
133	132	* Only IDE1 exists on the Cayman
134	133	*/
@@ -158,7 +157,6 @@ static int __init smsc_superio_setup(void) · 23 unmodified lines
158	157	SMSC_SUPERIO_WRITE_INDEXED(0x01, 0xc5); /* GP45 = IDE1_IRQ */
159	158	SMSC_SUPERIO_WRITE_INDEXED(0x00, 0xc6); /* GP46 = nIOROP */
160	159	SMSC_SUPERIO_WRITE_INDEXED(0x00, 0xc7); /* GP47 = nIOWOP */
161		−#endif
162	160
163	161	/* Exit the configuration state */
164	162	outb(SMSC_EXIT_CONFIG_KEY, SMSC_CONFIG_PORT_ADDR);

arch/sh/include/asm/pgtable-2level.h+1−0

@@ -1,6 +1,7 @@
1	1	#ifndef __ASM_SH_PGTABLE_2LEVEL_H
2	2	#define __ASM_SH_PGTABLE_2LEVEL_H
3	3
	4	+#define __ARCH_USE_5LEVEL_HACK
4	5	#include <asm-generic/pgtable-nopmd.h>
5	6
6	7	/*

arch/sh/include/asm/pgtable-3level.h+1−0

@@ -1,6 +1,7 @@
1	1	#ifndef __ASM_SH_PGTABLE_3LEVEL_H
2	2	#define __ASM_SH_PGTABLE_3LEVEL_H
3	3
	4	+#define __ARCH_USE_5LEVEL_HACK
4	5	#include <asm-generic/pgtable-nopud.h>
5	6
6	7	/*

arch/sparc/include/asm/pgtable_64.h+1−0

@@ -12,6 +12,7 @@ · 11 unmodified lines
12	12	* the SpitFire page tables.
13	13	*/
14	14
	15	+#include <asm-generic/5level-fixup.h>
15	16	#include <linux/compiler.h>
16	17	#include <linux/const.h>
17	18	#include <asm/types.h>

arch/tile/include/asm/pgtable_32.h+1−0

@@ -74,6 +74,7 @@ extern unsigned long VMALLOC_RESERVE /* = CONFIG_VMALLOC_RESERVE */; · 73 unmodified lines
74	74	#define MAXMEM (_VMALLOC_START - PAGE_OFFSET)
75	75
76	76	/* We have no pmd or pud since we are strictly a two-level page table */
	77	+#define __ARCH_USE_5LEVEL_HACK
77	78	#include <asm-generic/pgtable-nopmd.h>
78	79
79	80	static inline int pud_huge_page(pud_t pud) { return 0; }

arch/tile/include/asm/pgtable_64.h+1−0

@@ -59,6 +59,7 @@ · 58 unmodified lines
59	59	#ifndef __ASSEMBLY__
60	60
61	61	/* We have no pud since we are a three-level page table. */
	62	+#define __ARCH_USE_5LEVEL_HACK
62	63	#include <asm-generic/pgtable-nopud.h>
63	64
64	65	/*

arch/um/include/asm/pgtable-2level.h+1−0

@@ -8,6 +8,7 @@ · 7 unmodified lines
8	8	#ifndef __UM_PGTABLE_2LEVEL_H
9	9	#define __UM_PGTABLE_2LEVEL_H
10	10
	11	+#define __ARCH_USE_5LEVEL_HACK
11	12	#include <asm-generic/pgtable-nopmd.h>
12	13
13	14	/* PGDIR_SHIFT determines what a third-level page table entry can map */

arch/um/include/asm/pgtable-3level.h+1−0

@@ -7,6 +7,7 @@ · 6 unmodified lines
7	7	#ifndef __UM_PGTABLE_3LEVEL_H
8	8	#define __UM_PGTABLE_3LEVEL_H
9	9
	10	+#define __ARCH_USE_5LEVEL_HACK
10	11	#include <asm-generic/pgtable-nopud.h>
11	12
12	13	/* PGDIR_SHIFT determines what a third-level page table entry can map */

arch/unicore32/include/asm/pgtable.h+1−0

@@ -12,6 +12,7 @@ · 11 unmodified lines
12	12	#ifndef __UNICORE_PGTABLE_H__
13	13	#define __UNICORE_PGTABLE_H__
14	14
	15	+#define __ARCH_USE_5LEVEL_HACK
15	16	#include <asm-generic/pgtable-nopmd.h>
16	17	#include <asm/cpu-single.h>
17	18

arch/x86/configs/x86_64_defconfig+1−0

@@ -176,6 +176,7 @@ CONFIG_E1000E=y · 175 unmodified lines
176	176	CONFIG_SKY2=y
177	177	CONFIG_FORCEDETH=y
178	178	CONFIG_8139TOO=y
	179	+CONFIG_R8169=y
179	180	CONFIG_FDDI=y
180	181	CONFIG_INPUT_POLLDEV=y
181	182	# CONFIG_INPUT_MOUSEDEV_PSAUX is not set

arch/x86/events/amd/core.c+1−1

@@ -604,7 +604,7 @@ amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, int idx, · 603 unmodified lines
604	604	return &amd_f15_PMC20;
605	605	}
606	606	case AMD_EVENT_NB:
607		− /* moved to perf_event_amd_uncore.c */
	607	+ /* moved to uncore.c */
608	608	return &emptyconstraint;
609	609	default:
610	610	return &emptyconstraint;

arch/x86/events/intel/cstate.c+1−1

@@ -1,5 +1,5 @@
1	1	/*
2		− * perf_event_intel_cstate.c: support cstate residency counters
	2	+ * Support cstate residency counters
3	3	*
4	4	* Copyright (C) 2015, Intel Corp.
5	5	* Author: Kan Liang (kan.liang@intel.com)

arch/x86/events/intel/rapl.c+1−1

@@ -1,5 +1,5 @@
1	1	/*
2		− * perf_event_intel_rapl.c: support Intel RAPL energy consumption counters
	2	+ * Support Intel RAPL energy consumption counters
3	3	* Copyright (C) 2013 Google, Inc., Stephane Eranian
4	4	*
5	5	* Intel RAPL interface is specified in the IA-32 Manual Vol3b

arch/x86/events/intel/uncore.h+3−3

@@ -360,7 +360,7 @@ extern struct list_head pci2phy_map_head; · 359 unmodified lines
360	360	extern struct pci_extra_dev *uncore_extra_pci_dev;
361	361	extern struct event_constraint uncore_constraint_empty;
362	362
363		−/* perf_event_intel_uncore_snb.c */
	363	+/* uncore_snb.c */
364	364	int snb_uncore_pci_init(void);
365	365	int ivb_uncore_pci_init(void);
366	366	int hsw_uncore_pci_init(void);
@@ -371,7 +371,7 @@ void nhm_uncore_cpu_init(void); · 4 unmodified lines
371	371	void skl_uncore_cpu_init(void);
372	372	int snb_pci2phy_map_init(int devid);
373	373
374		−/* perf_event_intel_uncore_snbep.c */
	374	+/* uncore_snbep.c */
375	375	int snbep_uncore_pci_init(void);
376	376	void snbep_uncore_cpu_init(void);
377	377	int ivbep_uncore_pci_init(void);
@@ -385,5 +385,5 @@ void knl_uncore_cpu_init(void); · 7 unmodified lines
385	385	int skx_uncore_pci_init(void);
386	386	void skx_uncore_cpu_init(void);
387	387
388		−/* perf_event_intel_uncore_nhmex.c */
	388	+/* uncore_nhmex.c */
389	389	void nhmex_uncore_cpu_init(void);

arch/x86/hyperv/hv_init.c+1−1

@@ -158,13 +158,13 @@ void hyperv_init(void) · 157 unmodified lines
158	158	clocksource_register_hz(&hyperv_cs_tsc, NSEC_PER_SEC/100);
159	159	return;
160	160	}
	161	+register_msr_cs:
161	162	#endif
162	163	/*
163	164	* For 32 bit guests just use the MSR based mechanism for reading
164	165	* the partition counter.
165	166	*/
166	167
167		−register_msr_cs:
168	168	hyperv_cs = &hyperv_cs_msr;
169	169	if (ms_hyperv.features & HV_X64_MSR_TIME_REF_COUNT_AVAILABLE)
170	170	clocksource_register_hz(&hyperv_cs_msr, NSEC_PER_SEC/100);

arch/x86/include/asm/cpufeatures.h+2−1

@@ -289,7 +289,8 @@ · 288 unmodified lines
289	289	#define X86_FEATURE_PKU (1632+ 3) / Protection Keys for Userspace */
290	290	#define X86_FEATURE_OSPKE (1632+ 4) / OS Protection Keys Enable */
291	291	#define X86_FEATURE_AVX512_VPOPCNTDQ (1632+14) / POPCNT for vectors of DW/QW */
292		−#define X86_FEATURE_RDPID (1632+ 22) / RDPID instruction */
	292	+#define X86_FEATURE_LA57 (1632+16) / 5-level page tables */
	293	+#define X86_FEATURE_RDPID (1632+22) / RDPID instruction */
293	294
294	295	/* AMD-defined CPU features, CPUID level 0x80000007 (ebx), word 17 */
295	296	#define X86_FEATURE_OVERFLOW_RECOV (1732+0) / MCA overflow recovery support */

arch/x86/include/asm/pgtable_types.h+4−0

@@ -273,6 +273,8 @@ static inline pgdval_t pgd_flags(pgd_t pgd) · 272 unmodified lines
273	273	}
274	274
275	275	#if CONFIG_PGTABLE_LEVELS > 3
	276	+#include <asm-generic/5level-fixup.h>
	277	+
276	278	typedef struct { pudval_t pud; } pud_t;
277	279
278	280	static inline pud_t native_make_pud(pmdval_t val)
@@ -285,6 +287,7 @@ static inline pudval_t native_pud_val(pud_t pud) · 6 unmodified lines
285	287	return pud.pud;
286	288	}
287	289	#else
	290	+#define __ARCH_USE_5LEVEL_HACK
288	291	#include <asm-generic/pgtable-nopud.h>
289	292
290	293	static inline pudval_t native_pud_val(pud_t pud)
@@ -306,6 +309,7 @@ static inline pmdval_t native_pmd_val(pmd_t pmd) · 15 unmodified lines
306	309	return pmd.pmd;
307	310	}
308	311	#else
	312	+#define __ARCH_USE_5LEVEL_HACK
309	313	#include <asm-generic/pgtable-nopmd.h>
310	314
311	315	static inline pmdval_t native_pmd_val(pmd_t pmd)

arch/x86/include/asm/pkeys.h+9−6

@@ -46,6 +46,15 @@ extern int __arch_set_user_pkey_access(struct task_struct *tsk, int pkey, · 45 unmodified lines
46	46	static inline
47	47	bool mm_pkey_is_allocated(struct mm_struct *mm, int pkey)
48	48	{
	49	+ /*
	50	+ * "Allocated" pkeys are those that have been returned
	51	+ * from pkey_alloc(). pkey 0 is special, and never
	52	+ * returned from pkey_alloc().
	53	+ */
	54	+ if (pkey <= 0)
	55	+ return false;
	56	+ if (pkey >= arch_max_pkey())
	57	+ return false;
49	58	return mm_pkey_allocation_map(mm) & (1U << pkey);
50	59	}
51	60
@@ -82,12 +91,6 @@ int mm_pkey_alloc(struct mm_struct *mm) · 30 unmodified lines
82	91	static inline
83	92	int mm_pkey_free(struct mm_struct *mm, int pkey)
84	93	{
85		− /*
86		− * pkey 0 is special, always allocated and can never
87		− * be freed.
88		− */
89		− if (!pkey)
90		− return -EINVAL;
91	94	if (!mm_pkey_is_allocated(mm, pkey))
92	95	return -EINVAL;
93	96

arch/x86/include/asm/purgatory.h+20−0

@@ -0,0 +1,20 @@
	1	+#ifndef _ASM_X86_PURGATORY_H
	2	+#define _ASM_X86_PURGATORY_H
	3	+
	4	+#ifndef __ASSEMBLY__
	5	+#include <linux/purgatory.h>
	6	+
	7	+extern void purgatory(void);
	8	+/*
	9	+ * These forward declarations serve two purposes:
	10	+ *
	11	+ * 1) Make sparse happy when checking arch/purgatory
	12	+ * 2) Document that these are required to be global so the symbol
	13	+ * lookup in kexec works
	14	+ */
	15	+extern unsigned long purgatory_backup_dest;
	16	+extern unsigned long purgatory_backup_src;
	17	+extern unsigned long purgatory_backup_sz;
	18	+#endif /* __ASSEMBLY__ */
	19	+
	20	+#endif /* _ASM_PURGATORY_H */

arch/x86/include/asm/tlbflush.h+1−1

@@ -188,7 +188,7 @@ static inline void __native_flush_tlb_single(unsigned long addr) · 187 unmodified lines
188	188
189	189	static inline void __flush_tlb_all(void)
190	190	{
191		− if (static_cpu_has(X86_FEATURE_PGE))
	191	+ if (boot_cpu_has(X86_FEATURE_PGE))
192	192	__flush_tlb_global();
193	193	else
194	194	__flush_tlb();

arch/x86/include/uapi/asm/bootparam.h+1−1

@@ -58,7 +58,7 @@ struct setup_header { · 57 unmodified lines
58	58	__u32 header;
59	59	__u16 version;
60	60	__u32 realmode_swtch;
61		− __u16 start_sys;
	61	+ __u16 start_sys_seg;
62	62	__u16 kernel_version;
63	63	__u8 type_of_loader;
64	64	__u8 loadflags;

arch/x86/kernel/apic/apic.c+7−16

@@ -1610,24 +1610,15 @@ static inline void try_to_enable_x2apic(int remap_mode) { } · 1609 unmodified lines
1610	1610	static inline void __x2apic_enable(void) { }
1611	1611	#endif /* !CONFIG_X86_X2APIC */
1612	1612
1613		−static int __init try_to_enable_IR(void)
1614		−{
1615		−#ifdef CONFIG_X86_IO_APIC
1616		− if (!x2apic_enabled() && skip_ioapic_setup) {
1617		− pr_info("Not enabling interrupt remapping due to skipped IO-APIC setup\n");
1618		− return -1;
1619		− }
1620		−#endif
1621		− return irq_remapping_enable();
1622		−}
1623		−
1624	1613	void __init enable_IR_x2apic(void)
1625	1614	{
1626	1615	unsigned long flags;
1627	1616	int ret, ir_stat;
1628	1617
1629		− if (skip_ioapic_setup)
	1618	+ if (skip_ioapic_setup) {
	1619	+ pr_info("Not enabling interrupt remapping due to skipped IO-APIC setup\n");
1630	1620	return;
	1621	+ }
1631	1622
1632	1623	ir_stat = irq_remapping_prepare();
1633	1624	if (ir_stat < 0 && !x2apic_supported())
@@ -1645,7 +1636,7 @@ void __init enable_IR_x2apic(void) · 11 unmodified lines
1645	1636
1646	1637	/* If irq_remapping_prepare() succeeded, try to enable it */
1647	1638	if (ir_stat >= 0)
1648		− ir_stat = try_to_enable_IR();
	1639	+ ir_stat = irq_remapping_enable();
1649	1640	/* ir_stat contains the remap mode or an error code */
1650	1641	try_to_enable_x2apic(ir_stat);
1651	1642
@@ -2062,10 +2053,10 @@ static int allocate_logical_cpuid(int apicid) · 410 unmodified lines
2062	2053
2063	2054	/* Allocate a new cpuid. */
2064	2055	if (nr_logical_cpuids >= nr_cpu_ids) {
2065		− WARN_ONCE(1, "Only %d processors supported."
	2056	+ WARN_ONCE(1, "APIC: NR_CPUS/possible_cpus limit of %i reached. "
2066	2057	"Processor %d/0x%x and the rest are ignored.\n",
2067		− nr_cpu_ids - 1, nr_logical_cpuids, apicid);
2068		− return -1;
	2058	+ nr_cpu_ids, nr_logical_cpuids, apicid);
	2059	+ return -EINVAL;
2069	2060	}
2070	2061
2071	2062	cpuid_to_apicid[nr_logical_cpuids] = apicid;

arch/x86/kernel/cpu/amd.c+0−4

@@ -556,10 +556,6 @@ static void early_init_amd(struct cpuinfo_x86 *c) · 555 unmodified lines
556	556	if (c->x86_power & (1 << 8)) {
557	557	set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
558	558	set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
559		− if (check_tsc_unstable())
560		− clear_sched_clock_stable();
561		− } else {
562		− clear_sched_clock_stable();
563	559	}
564	560
565	561	/* Bit 12 of 8000_0007 edx is accumulated power mechanism. */

arch/x86/kernel/cpu/centaur.c+0−2

@@ -105,8 +105,6 @@ static void early_init_centaur(struct cpuinfo_x86 *c) · 104 unmodified lines
105	105	#ifdef CONFIG_X86_64
106	106	set_cpu_cap(c, X86_FEATURE_SYSENTER32);
107	107	#endif
108		−
109		− clear_sched_clock_stable();
110	108	}
111	109
112	110	static void init_centaur(struct cpuinfo_x86 *c)

arch/x86/kernel/cpu/common.c+0−3

@@ -88,7 +88,6 @@ static void default_init(struct cpuinfo_x86 *c) · 87 unmodified lines
88	88	strcpy(c->x86_model_id, "386");
89	89	}
90	90	#endif
91		− clear_sched_clock_stable();
92	91	}
93	92
94	93	static const struct cpu_dev default_cpu = {
@@ -1077,8 +1076,6 @@ static void identify_cpu(struct cpuinfo_x86 *c) · 982 unmodified lines
1077	1076	*/
1078	1077	if (this_cpu->c_init)
1079	1078	this_cpu->c_init(c);
1080		− else
1081		− clear_sched_clock_stable();
1082	1079
1083	1080	/* Disable the PN if appropriate */
1084	1081	squash_the_stupid_serial_number(c);

arch/x86/kernel/cpu/cyrix.c+0−1

@@ -185,7 +185,6 @@ static void early_init_cyrix(struct cpuinfo_x86 *c) · 184 unmodified lines
185	185	set_cpu_cap(c, X86_FEATURE_CYRIX_ARR);
186	186	break;
187	187	}
188		− clear_sched_clock_stable();
189	188	}
190	189
191	190	static void init_cyrix(struct cpuinfo_x86 *c)

arch/x86/kernel/cpu/intel.c+0−4

@@ -162,10 +162,6 @@ static void early_init_intel(struct cpuinfo_x86 *c) · 161 unmodified lines
162	162	if (c->x86_power & (1 << 8)) {
163	163	set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
164	164	set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
165		− if (check_tsc_unstable())
166		− clear_sched_clock_stable();
167		− } else {
168		− clear_sched_clock_stable();
169	165	}
170	166
171	167	/* Penwell and Cloverview have the TSC which doesn't sleep on S3 */

arch/x86/kernel/cpu/intel_rdt_rdtgroup.c+0−1

@@ -28,7 +28,6 @@ · 27 unmodified lines
28	28	#include <linux/sched/signal.h>
29	29	#include <linux/sched/task.h>
30	30	#include <linux/slab.h>
31		−#include <linux/cpu.h>
32	31	#include <linux/task_work.h>
33	32
34	33	#include <uapi/linux/magic.h>

arch/x86/kernel/cpu/transmeta.c+0−2

@@ -16,8 +16,6 @@ static void early_init_transmeta(struct cpuinfo_x86 *c) · 15 unmodified lines
16	16	if (xlvl >= 0x80860001)
17	17	c->x86_capability[CPUID_8086_0001_EDX] = cpuid_edx(0x80860001);
18	18	}
19		−
20		− clear_sched_clock_stable();
21	19	}
22	20
23	21	static void init_transmeta(struct cpuinfo_x86 *c)

arch/x86/kernel/cpu/vmware.c+0−1

@@ -30,7 +30,6 @@ · 29 unmodified lines
30	30	#include <asm/hypervisor.h>
31	31	#include <asm/timer.h>
32	32	#include <asm/apic.h>
33		−#include <asm/timer.h>
34	33
35	34	#undef pr_fmt
36	35	#define pr_fmt(fmt) "vmware: " fmt

arch/x86/kernel/ftrace.c+1−1

@@ -535,7 +535,7 @@ static void run_sync(void) · 534 unmodified lines
535	535	{
536	536	int enable_irqs = irqs_disabled();
537	537
538		− /* We may be called with interrupts disbled (on bootup). */
	538	+ /* We may be called with interrupts disabled (on bootup). */
539	539	if (enable_irqs)
540	540	local_irq_enable();
541	541	on_each_cpu(do_sync_core, NULL, 1);

arch/x86/kernel/hpet.c+1−1

@@ -354,7 +354,7 @@ static int hpet_resume(struct clock_event_device *evt, int timer) · 353 unmodified lines
354	354
355	355	irq_domain_deactivate_irq(irq_get_irq_data(hdev->irq));
356	356	irq_domain_activate_irq(irq_get_irq_data(hdev->irq));
357		− disable_irq(hdev->irq);
	357	+ disable_hardirq(hdev->irq);
358	358	irq_set_affinity(hdev->irq, cpumask_of(hdev->cpu));
359	359	enable_irq(hdev->irq);
360	360	}

arch/x86/kernel/kdebugfs.c+1−1

@@ -167,7 +167,7 @@ static int __init boot_params_kdebugfs_init(void) · 166 unmodified lines
167	167	struct dentry dbp, version, *data;
168	168	int error = -ENOMEM;
169	169
170		− dbp = debugfs_create_dir("boot_params", NULL);
	170	+ dbp = debugfs_create_dir("boot_params", arch_debugfs_dir);
171	171	if (!dbp)
172	172	return -ENOMEM;
173	173

arch/x86/kernel/machine_kexec_64.c+6−3

@@ -194,19 +194,22 @@ static int arch_update_purgatory(struct kimage *image) · 193 unmodified lines
194	194
195	195	/* Setup copying of backup region */
196	196	if (image->type == KEXEC_TYPE_CRASH) {
197		− ret = kexec_purgatory_get_set_symbol(image, "backup_dest",
	197	+ ret = kexec_purgatory_get_set_symbol(image,
	198	+ "purgatory_backup_dest",
198	199	&image->arch.backup_load_addr,
199	200	sizeof(image->arch.backup_load_addr), 0);
200	201	if (ret)
201	202	return ret;
202	203
203		− ret = kexec_purgatory_get_set_symbol(image, "backup_src",
	204	+ ret = kexec_purgatory_get_set_symbol(image,
	205	+ "purgatory_backup_src",
204	206	&image->arch.backup_src_start,
205	207	sizeof(image->arch.backup_src_start), 0);
206	208	if (ret)
207	209	return ret;
208	210
209		− ret = kexec_purgatory_get_set_symbol(image, "backup_sz",
	211	+ ret = kexec_purgatory_get_set_symbol(image,
	212	+ "purgatory_backup_sz",
210	213	&image->arch.backup_src_sz,
211	214	sizeof(image->arch.backup_src_sz), 0);
212	215	if (ret)

arch/x86/kernel/reboot.c+16−0

@@ -223,6 +223,22 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = { · 222 unmodified lines
223	223	DMI_MATCH(DMI_BOARD_NAME, "P4S800"),
224	224	},
225	225	},
	226	+ { /* Handle problems with rebooting on ASUS EeeBook X205TA */
	227	+ .callback = set_acpi_reboot,
	228	+ .ident = "ASUS EeeBook X205TA",
	229	+ .matches = {
	230	+ DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
	231	+ DMI_MATCH(DMI_PRODUCT_NAME, "X205TA"),
	232	+ },
	233	+ },
	234	+ { /* Handle problems with rebooting on ASUS EeeBook X205TAW */
	235	+ .callback = set_acpi_reboot,
	236	+ .ident = "ASUS EeeBook X205TAW",
	237	+ .matches = {
	238	+ DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
	239	+ DMI_MATCH(DMI_PRODUCT_NAME, "X205TAW"),
	240	+ },
	241	+ },
226	242
227	243	/* Certec */
228	244	{ /* Handle problems with rebooting on Certec BPC600 */

arch/x86/kernel/tsc.c+23−12

@@ -327,9 +327,16 @@ unsigned long long sched_clock(void) · 326 unmodified lines
327	327	{
328	328	return paravirt_sched_clock();
329	329	}
	330	+
	331	+static inline bool using_native_sched_clock(void)
	332	+{
	333	+ return pv_time_ops.sched_clock == native_sched_clock;
	334	+}
330	335	#else
331	336	unsigned long long
332	337	sched_clock(void) __attribute__((alias("native_sched_clock")));
	338	+
	339	+static inline bool using_native_sched_clock(void) { return true; }
333	340	#endif
334	341
335	342	int check_tsc_unstable(void)
@@ -1112,8 +1119,10 @@ static void tsc_cs_mark_unstable(struct clocksource *cs) · 776 unmodified lines
1112	1119	{
1113	1120	if (tsc_unstable)
1114	1121	return;
	1122	+
1115	1123	tsc_unstable = 1;
1116		− clear_sched_clock_stable();
	1124	+ if (using_native_sched_clock())
	1125	+ clear_sched_clock_stable();
1117	1126	disable_sched_clock_irqtime();
1118	1127	pr_info("Marking TSC unstable due to clocksource watchdog\n");
1119	1128	}
@@ -1135,18 +1144,20 @@ static struct clocksource clocksource_tsc = { · 15 unmodified lines
1135	1144
1136	1145	void mark_tsc_unstable(char *reason)
1137	1146	{
1138		− if (!tsc_unstable) {
1139		− tsc_unstable = 1;
	1147	+ if (tsc_unstable)
	1148	+ return;
	1149	+
	1150	+ tsc_unstable = 1;
	1151	+ if (using_native_sched_clock())
1140	1152	clear_sched_clock_stable();
1141		− disable_sched_clock_irqtime();
1142		− pr_info("Marking TSC unstable due to %s\n", reason);
1143		− /* Change only the rating, when not registered */
1144		− if (clocksource_tsc.mult)
1145		− clocksource_mark_unstable(&clocksource_tsc);
1146		− else {
1147		− clocksource_tsc.flags \|= CLOCK_SOURCE_UNSTABLE;
1148		− clocksource_tsc.rating = 0;
1149		− }
	1153	+ disable_sched_clock_irqtime();
	1154	+ pr_info("Marking TSC unstable due to %s\n", reason);
	1155	+ /* Change only the rating, when not registered */
	1156	+ if (clocksource_tsc.mult) {
	1157	+ clocksource_mark_unstable(&clocksource_tsc);
	1158	+ } else {
	1159	+ clocksource_tsc.flags \|= CLOCK_SOURCE_UNSTABLE;
	1160	+ clocksource_tsc.rating = 0;
1150	1161	}
1151	1162	}
1152	1163

arch/x86/kvm/vmx.c+8−22

@@ -7258,9 +7258,8 @@ static int handle_vmoff(struct kvm_vcpu *vcpu) · 7257 unmodified lines
7258	7258	static int handle_vmclear(struct kvm_vcpu *vcpu)
7259	7259	{
7260	7260	struct vcpu_vmx *vmx = to_vmx(vcpu);
	7261	+ u32 zero = 0;
7261	7262	gpa_t vmptr;
7262		− struct vmcs12 *vmcs12;
7263		− struct page *page;
7264	7263
7265	7264	if (!nested_vmx_check_permission(vcpu))
7266	7265	return 1;
@@ -7271,22 +7270,9 @@ static int handle_vmclear(struct kvm_vcpu *vcpu) · 4 unmodified lines
7271	7270	if (vmptr == vmx->nested.current_vmptr)
7272	7271	nested_release_vmcs12(vmx);
7273	7272
7274		− page = nested_get_page(vcpu, vmptr);
7275		− if (page == NULL) {
7276		− /*
7277		− * For accurate processor emulation, VMCLEAR beyond available
7278		− * physical memory should do nothing at all. However, it is
7279		− * possible that a nested vmx bug, not a guest hypervisor bug,
7280		− * resulted in this case, so let's shut down before doing any
7281		− * more damage:
7282		− */
7283		− kvm_make_request(KVM_REQ_TRIPLE_FAULT, vcpu);
7284		− return 1;
7285		− }
7286		− vmcs12 = kmap(page);
7287		− vmcs12->launch_state = 0;
7288		− kunmap(page);
7289		− nested_release_page(page);
	7273	+ kvm_vcpu_write_guest(vcpu,
	7274	+ vmptr + offsetof(struct vmcs12, launch_state),
	7275	+ &zero, sizeof(zero));
7290	7276
7291	7277	nested_free_vmcs02(vmx, vmptr);
7292	7278
@@ -9694,10 +9680,8 @@ static inline bool nested_vmx_merge_msr_bitmap(struct kvm_vcpu *vcpu, · 2401 unmodified lines
9694	9680	return false;
9695	9681
9696	9682	page = nested_get_page(vcpu, vmcs12->msr_bitmap);
9697		− if (!page) {
9698		− WARN_ON(1);
	9683	+ if (!page)
9699	9684	return false;
9700		− }
9701	9685	msr_bitmap_l1 = (unsigned long *)kmap(page);
9702	9686
9703	9687	memset(msr_bitmap_l0, 0xff, PAGE_SIZE);
@@ -11121,8 +11105,10 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason, · 1417 unmodified lines
11121	11105	*/
11122	11106	static void vmx_leave_nested(struct kvm_vcpu *vcpu)
11123	11107	{
11124		− if (is_guest_mode(vcpu))
	11108	+ if (is_guest_mode(vcpu)) {
	11109	+ to_vmx(vcpu)->nested.nested_run_pending = 0;
11125	11110	nested_vmx_vmexit(vcpu, -1, 0, 0);
	11111	+ }
11126	11112	free_nested(to_vmx(vcpu));
11127	11113	}
11128	11114

arch/x86/mm/gup.c+21−16

@@ -106,32 +106,35 @@ static noinline int gup_pte_range(pmd_t pmd, unsigned long addr, · 105 unmodified lines
106	106	unsigned long end, int write, struct page *pages, int nr)
107	107	{
108	108	struct dev_pagemap *pgmap = NULL;
109		− int nr_start = *nr;
110		− pte_t *ptep;
	109	+ int nr_start = *nr, ret = 0;
	110	+ pte_t ptep, ptem;
111	111
112		− ptep = pte_offset_map(&pmd, addr);
	112	+ /*
	113	+ * Keep the original mapped PTE value (ptem) around since we
	114	+ * might increment ptep off the end of the page when finishing
	115	+ * our loop iteration.
	116	+ */
	117	+ ptem = ptep = pte_offset_map(&pmd, addr);
113	118	do {
114	119	pte_t pte = gup_get_pte(ptep);
115	120	struct page *page;
116	121
117	122	/* Similar to the PMD case, NUMA hinting must take slow path */
118		− if (pte_protnone(pte)) {
119		− pte_unmap(ptep);
120		− return 0;
121		− }
	123	+ if (pte_protnone(pte))
	124	+ break;
	125	+
	126	+ if (!pte_allows_gup(pte_val(pte), write))
	127	+ break;
122	128
123	129	if (pte_devmap(pte)) {
124	130	pgmap = get_dev_pagemap(pte_pfn(pte), pgmap);
125	131	if (unlikely(!pgmap)) {
126	132	undo_dev_pagemap(nr, nr_start, pages);
127		− pte_unmap(ptep);
128		− return 0;
	133	+ break;
129	134	}
130		− } else if (!pte_allows_gup(pte_val(pte), write) \|\|
131		− pte_special(pte)) {
132		− pte_unmap(ptep);
133		− return 0;
134		− }
	135	+ } else if (pte_special(pte))
	136	+ break;
	137	+
135	138	VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
136	139	page = pte_page(pte);
137	140	get_page(page);
@@ -141,9 +144,11 @@ static noinline int gup_pte_range(pmd_t pmd, unsigned long addr, · 3 unmodified lines
141	144	(*nr)++;
142	145
143	146	} while (ptep++, addr += PAGE_SIZE, addr != end);
144		− pte_unmap(ptep - 1);
	147	+ if (addr == end)
	148	+ ret = 1;
	149	+ pte_unmap(ptem);
145	150
146		− return 1;
	151	+ return ret;
147	152	}
148	153
149	154	static inline void get_head_page_multiple(struct page *page, int nr)

arch/x86/pci/common.c+9−0

@@ -735,6 +735,15 @@ void pcibios_disable_device (struct pci_dev *dev) · 734 unmodified lines
735	735	pcibios_disable_irq(dev);
736	736	}
737	737
	738	+#ifdef CONFIG_ACPI_HOTPLUG_IOAPIC
	739	+void pcibios_release_device(struct pci_dev *dev)
	740	+{
	741	+ if (atomic_dec_return(&dev->enable_cnt) >= 0)
	742	+ pcibios_disable_device(dev);
	743	+
	744	+}
	745	+#endif
	746	+
738	747	int pci_ext_cfg_avail(void)
739	748	{
740	749	if (raw_pci_ext_ops)

arch/x86/pci/xen.c+7−16

@@ -234,23 +234,14 @@ static int xen_hvm_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) · 233 unmodified lines
234	234	return 1;
235	235
236	236	for_each_pci_msi_entry(msidesc, dev) {
237		− __pci_read_msi_msg(msidesc, &msg);
238		− pirq = MSI_ADDR_EXT_DEST_ID(msg.address_hi) \|
239		− ((msg.address_lo >> MSI_ADDR_DEST_ID_SHIFT) & 0xff);
240		− if (msg.data != XEN_PIRQ_MSI_DATA \|\|
241		− xen_irq_from_pirq(pirq) < 0) {
242		− pirq = xen_allocate_pirq_msi(dev, msidesc);
243		− if (pirq < 0) {
244		− irq = -ENODEV;
245		− goto error;
246		− }
247		− xen_msi_compose_msg(dev, pirq, &msg);
248		− __pci_write_msi_msg(msidesc, &msg);
249		− dev_dbg(&dev->dev, "xen: msi bound to pirq=%d\n", pirq);
250		− } else {
251		− dev_dbg(&dev->dev,
252		− "xen: msi already bound to pirq=%d\n", pirq);
	237	+ pirq = xen_allocate_pirq_msi(dev, msidesc);
	238	+ if (pirq < 0) {
	239	+ irq = -ENODEV;
	240	+ goto error;
253	241	}
	242	+ xen_msi_compose_msg(dev, pirq, &msg);
	243	+ __pci_write_msi_msg(msidesc, &msg);
	244	+ dev_dbg(&dev->dev, "xen: msi bound to pirq=%d\n", pirq);
254	245	irq = xen_bind_pirq_msi_to_irq(dev, msidesc, pirq,
255	246	(type == PCI_CAP_ID_MSI) ? nvec : 1,
256	247	(type == PCI_CAP_ID_MSIX) ?

arch/x86/platform/uv/tlb_uv.c+0−1

@@ -1847,7 +1847,6 @@ static void pq_init(int node, int pnode) · 1846 unmodified lines
1847	1847
1848	1848	ops.write_payload_first(pnode, first);
1849	1849	ops.write_payload_last(pnode, last);
1850		− ops.write_g_sw_ack(pnode, 0xffffUL);
1851	1850
1852	1851	/* in effect, all msg_type's are set to MSG_NOOP */
1853	1852	memset(pqp, 0, sizeof(struct bau_pq_entry) * DEST_Q_SIZE);

arch/x86/purgatory/purgatory.c+18−18

@@ -10,21 +10,19 @@ · 9 unmodified lines
10	10	* Version 2. See the file COPYING for more details.
11	11	*/
12	12
	13	+#include <linux/bug.h>
	14	+#include <asm/purgatory.h>
	15	+
13	16	#include "sha256.h"
14	17	#include "../boot/string.h"
15	18
16		−struct sha_region {
17		− unsigned long start;
18		− unsigned long len;
19		−};
20		−
21		−unsigned long backup_dest = 0;
22		−unsigned long backup_src = 0;
23		−unsigned long backup_sz = 0;
	19	+unsigned long purgatory_backup_dest __section(.kexec-purgatory);
	20	+unsigned long purgatory_backup_src __section(.kexec-purgatory);
	21	+unsigned long purgatory_backup_sz __section(.kexec-purgatory);
24	22
25		−u8 sha256_digest[SHA256_DIGEST_SIZE] = { 0 };
	23	+u8 purgatory_sha256_digest[SHA256_DIGEST_SIZE] __section(.kexec-purgatory);
26	24
27		−struct sha_region sha_regions[16] = {};
	25	+struct kexec_sha_region purgatory_sha_regions[KEXEC_SEGMENT_MAX] __section(.kexec-purgatory);
28	26
29	27	/*
30	28	* On x86, second kernel requries first 640K of memory to boot. Copy
@@ -33,26 +31,28 @@ struct sha_region sha_regions[16] = {}; · 2 unmodified lines
33	31	*/
34	32	static int copy_backup_region(void)
35	33	{
36		− if (backup_dest)
37		− memcpy((void )backup_dest, (void )backup_src, backup_sz);
38		−
	34	+ if (purgatory_backup_dest) {
	35	+ memcpy((void *)purgatory_backup_dest,
	36	+ (void *)purgatory_backup_src, purgatory_backup_sz);
	37	+ }
39	38	return 0;
40	39	}
41	40
42		−int verify_sha256_digest(void)
	41	+static int verify_sha256_digest(void)
43	42	{
44		− struct sha_region ptr, end;
	43	+ struct kexec_sha_region ptr, end;
45	44	u8 digest[SHA256_DIGEST_SIZE];
46	45	struct sha256_state sctx;
47	46
48	47	sha256_init(&sctx);
49		− end = &sha_regions[sizeof(sha_regions)/sizeof(sha_regions[0])];
50		− for (ptr = sha_regions; ptr < end; ptr++)
	48	+ end = purgatory_sha_regions + ARRAY_SIZE(purgatory_sha_regions);
	49	+
	50	+ for (ptr = purgatory_sha_regions; ptr < end; ptr++)
51	51	sha256_update(&sctx, (uint8_t *)(ptr->start), ptr->len);
52	52
53	53	sha256_final(&sctx, digest);
54	54
55		− if (memcmp(digest, sha256_digest, sizeof(digest)))
	55	+ if (memcmp(digest, purgatory_sha256_digest, sizeof(digest)))
56	56	return 1;
57	57
58	58	return 0;

arch/x86/purgatory/setup-x86_64.S+1−0

@@ -9,6 +9,7 @@ · 8 unmodified lines
9	9	* This source code is licensed under the GNU General Public License,
10	10	* Version 2. See the file COPYING for more details.
11	11	*/
	12	+#include <asm/purgatory.h>
12	13
13	14	.text
14	15	.globl purgatory_start

arch/x86/purgatory/sha256.h+0−1

@@ -10,7 +10,6 @@ · 9 unmodified lines
10	10	#ifndef SHA256_H
11	11	#define SHA256_H
12	12
13		−
14	13	#include <linux/types.h>
15	14	#include <crypto/sha.h>
16	15

arch/xtensa/include/asm/pgtable.h+1−0

@@ -11,6 +11,7 @@ · 10 unmodified lines
11	11	#ifndef _XTENSA_PGTABLE_H
12	12	#define _XTENSA_PGTABLE_H
13	13
	14	+#define __ARCH_USE_5LEVEL_HACK
14	15	#include <asm-generic/pgtable-nopmd.h>
15	16	#include <asm/page.h>
16	17	#include <asm/kmem_layout.h>

block/bio.c+9−3

@@ -376,10 +376,14 @@ static void punt_bios_to_rescuer(struct bio_set *bs) · 375 unmodified lines
376	376	bio_list_init(&punt);
377	377	bio_list_init(&nopunt);
378	378
379		− while ((bio = bio_list_pop(current->bio_list)))
	379	+ while ((bio = bio_list_pop(&current->bio_list[0])))
380	380	bio_list_add(bio->bi_pool == bs ? &punt : &nopunt, bio);
	381	+ current->bio_list[0] = nopunt;
381	382
382		− *current->bio_list = nopunt;
	383	+ bio_list_init(&nopunt);
	384	+ while ((bio = bio_list_pop(&current->bio_list[1])))
	385	+ bio_list_add(bio->bi_pool == bs ? &punt : &nopunt, bio);
	386	+ current->bio_list[1] = nopunt;
383	387
384	388	spin_lock(&bs->rescue_lock);
385	389	bio_list_merge(&bs->rescue_list, &punt);
@@ -466,7 +470,9 @@ struct bio bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set bs) · 80 unmodified lines
466	470	* we retry with the original gfp_flags.
467	471	*/
468	472
469		− if (current->bio_list && !bio_list_empty(current->bio_list))
	473	+ if (current->bio_list &&
	474	+ (!bio_list_empty(&current->bio_list[0]) \|\|
	475	+ !bio_list_empty(&current->bio_list[1])))
470	476	gfp_mask &= ~__GFP_DIRECT_RECLAIM;
471	477
472	478	p = mempool_alloc(bs->bio_pool, gfp_mask);

block/blk-core.c+31−10

@@ -578,8 +578,6 @@ void blk_cleanup_queue(struct request_queue *q) · 577 unmodified lines
578	578	q->queue_lock = &q->__queue_lock;
579	579	spin_unlock_irq(lock);
580	580
581		− put_disk_devt(q->disk_devt);
582		−
583	581	/* @q is and will stay empty, shutdown and put */
584	582	blk_put_queue(q);
585	583	}
@@ -1975,7 +1973,14 @@ end_io: · 1389 unmodified lines
1975	1973	*/
1976	1974	blk_qc_t generic_make_request(struct bio *bio)
1977	1975	{
1978		− struct bio_list bio_list_on_stack;
	1976	+ /*
	1977	+ * bio_list_on_stack[0] contains bios submitted by the current
	1978	+ * make_request_fn.
	1979	+ * bio_list_on_stack[1] contains bios that were submitted before
	1980	+ * the current make_request_fn, but that haven't been processed
	1981	+ * yet.
	1982	+ */
	1983	+ struct bio_list bio_list_on_stack[2];
1979	1984	blk_qc_t ret = BLK_QC_T_NONE;
1980	1985
1981	1986	if (!generic_make_request_checks(bio))
@@ -1992,7 +1997,7 @@ blk_qc_t generic_make_request(struct bio *bio) · 10 unmodified lines
1992	1997	* should be added at the tail
1993	1998	*/
1994	1999	if (current->bio_list) {
1995		− bio_list_add(current->bio_list, bio);
	2000	+ bio_list_add(&current->bio_list[0], bio);
1996	2001	goto out;
1997	2002	}
1998	2003
@@ -2011,23 +2016,39 @@ blk_qc_t generic_make_request(struct bio *bio) · 12 unmodified lines
2011	2016	* bio_list, and call into ->make_request() again.
2012	2017	*/
2013	2018	BUG_ON(bio->bi_next);
2014		− bio_list_init(&bio_list_on_stack);
2015		− current->bio_list = &bio_list_on_stack;
	2019	+ bio_list_init(&bio_list_on_stack[0]);
	2020	+ current->bio_list = bio_list_on_stack;
2016	2021	do {
2017	2022	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
2018	2023
2019	2024	if (likely(blk_queue_enter(q, false) == 0)) {
	2025	+ struct bio_list lower, same;
	2026	+
	2027	+ /* Create a fresh bio_list for all subordinate requests */
	2028	+ bio_list_on_stack[1] = bio_list_on_stack[0];
	2029	+ bio_list_init(&bio_list_on_stack[0]);
2020	2030	ret = q->make_request_fn(q, bio);
2021	2031
2022	2032	blk_queue_exit(q);
2023	2033
2024		− bio = bio_list_pop(current->bio_list);
	2034	+ /* sort new bios into those for a lower level
	2035	+ * and those for the same level
	2036	+ */
	2037	+ bio_list_init(&lower);
	2038	+ bio_list_init(&same);
	2039	+ while ((bio = bio_list_pop(&bio_list_on_stack[0])) != NULL)
	2040	+ if (q == bdev_get_queue(bio->bi_bdev))
	2041	+ bio_list_add(&same, bio);
	2042	+ else
	2043	+ bio_list_add(&lower, bio);
	2044	+ /* now assemble so we handle the lowest level first */
	2045	+ bio_list_merge(&bio_list_on_stack[0], &lower);
	2046	+ bio_list_merge(&bio_list_on_stack[0], &same);
	2047	+ bio_list_merge(&bio_list_on_stack[0], &bio_list_on_stack[1]);
2025	2048	} else {
2026		− struct bio *bio_next = bio_list_pop(current->bio_list);
2027		−
2028	2049	bio_io_error(bio);
2029		− bio = bio_next;
2030	2050	}
	2051	+ bio = bio_list_pop(&bio_list_on_stack[0]);
2031	2052	} while (bio);
2032	2053	current->bio_list = NULL; /* deactivate */
2033	2054

block/blk-mq-sysfs.c+25−15

@@ -17,6 +17,15 @@ static void blk_mq_sysfs_release(struct kobject *kobj) · 16 unmodified lines
17	17	{
18	18	}
19	19
	20	+static void blk_mq_hw_sysfs_release(struct kobject *kobj)
	21	+{
	22	+ struct blk_mq_hw_ctx *hctx = container_of(kobj, struct blk_mq_hw_ctx,
	23	+ kobj);
	24	+ free_cpumask_var(hctx->cpumask);
	25	+ kfree(hctx->ctxs);
	26	+ kfree(hctx);
	27	+}
	28	+
20	29	struct blk_mq_ctx_sysfs_entry {
21	30	struct attribute attr;
22	31	ssize_t (show)(struct blk_mq_ctx , char *);
@@ -200,7 +209,7 @@ static struct kobj_type blk_mq_ctx_ktype = { · 177 unmodified lines
200	209	static struct kobj_type blk_mq_hw_ktype = {
201	210	.sysfs_ops = &blk_mq_hw_sysfs_ops,
202	211	.default_attrs = default_hw_ctx_attrs,
203		− .release = blk_mq_sysfs_release,
	212	+ .release = blk_mq_hw_sysfs_release,
204	213	};
205	214
206	215	static void blk_mq_unregister_hctx(struct blk_mq_hw_ctx *hctx)
@@ -242,24 +251,15 @@ static int blk_mq_register_hctx(struct blk_mq_hw_ctx *hctx) · 35 unmodified lines
242	251	static void __blk_mq_unregister_dev(struct device dev, struct request_queue q)
243	252	{
244	253	struct blk_mq_hw_ctx *hctx;
245		− struct blk_mq_ctx *ctx;
246		− int i, j;
	254	+ int i;
247	255
248		− queue_for_each_hw_ctx(q, hctx, i) {
	256	+ queue_for_each_hw_ctx(q, hctx, i)
249	257	blk_mq_unregister_hctx(hctx);
250	258
251		− hctx_for_each_ctx(hctx, ctx, j)
252		− kobject_put(&ctx->kobj);
253		−
254		− kobject_put(&hctx->kobj);
255		− }
256		−
257	259	blk_mq_debugfs_unregister_hctxs(q);
258	260
259	261	kobject_uevent(&q->mq_kobj, KOBJ_REMOVE);
260	262	kobject_del(&q->mq_kobj);
261		− kobject_put(&q->mq_kobj);
262		−
263	263	kobject_put(&dev->kobj);
264	264
265	265	q->mq_sysfs_init_done = false;
@@ -277,7 +277,19 @@ void blk_mq_hctx_kobj_init(struct blk_mq_hw_ctx *hctx) · 11 unmodified lines
277	277	kobject_init(&hctx->kobj, &blk_mq_hw_ktype);
278	278	}
279	279
280		−static void blk_mq_sysfs_init(struct request_queue *q)
	280	+void blk_mq_sysfs_deinit(struct request_queue *q)
	281	+{
	282	+ struct blk_mq_ctx *ctx;
	283	+ int cpu;
	284	+
	285	+ for_each_possible_cpu(cpu) {
	286	+ ctx = per_cpu_ptr(q->queue_ctx, cpu);
	287	+ kobject_put(&ctx->kobj);
	288	+ }
	289	+ kobject_put(&q->mq_kobj);
	290	+}
	291	+
	292	+void blk_mq_sysfs_init(struct request_queue *q)
281	293	{
282	294	struct blk_mq_ctx *ctx;
283	295	int cpu;
@@ -297,8 +309,6 @@ int blk_mq_register_dev(struct device dev, struct request_queue q) · 13 unmodified lines
297	309
298	310	blk_mq_disable_hotplug();
299	311
300		− blk_mq_sysfs_init(q);
301		−
302	312	ret = kobject_add(&q->mq_kobj, kobject_get(&dev->kobj), "%s", "mq");
303	313	if (ret < 0)
304	314	goto out;

block/blk-mq-tag.c+3−0

@@ -295,6 +295,9 @@ int blk_mq_reinit_tagset(struct blk_mq_tag_set *set) · 294 unmodified lines
295	295	for (i = 0; i < set->nr_hw_queues; i++) {
296	296	struct blk_mq_tags *tags = set->tags[i];
297	297
	298	+ if (!tags)
	299	+ continue;
	300	+
298	301	for (j = 0; j < tags->nr_tags; j++) {
299	302	if (!tags->static_rqs[j])
300	303	continue;

block/blk-mq.c+15−22

@@ -1434,7 +1434,8 @@ static blk_qc_t request_to_qc_t(struct blk_mq_hw_ctx hctx, struct request rq) · 1433 unmodified lines
1434	1434	return blk_tag_to_qc_t(rq->internal_tag, hctx->queue_num, true);
1435	1435	}
1436	1436
1437		−static void blk_mq_try_issue_directly(struct request rq, blk_qc_t cookie)
	1437	+static void blk_mq_try_issue_directly(struct request rq, blk_qc_t cookie,
	1438	+ bool may_sleep)
1438	1439	{
1439	1440	struct request_queue *q = rq->q;
1440	1441	struct blk_mq_queue_data bd = {
@@ -1475,7 +1476,7 @@ static void blk_mq_try_issue_directly(struct request rq, blk_qc_t cookie) · 34 unmodified lines
1475	1476	}
1476	1477
1477	1478	insert:
1478		− blk_mq_sched_insert_request(rq, false, true, true, false);
	1479	+ blk_mq_sched_insert_request(rq, false, true, false, may_sleep);
1479	1480	}
1480	1481
1481	1482	/*
@@ -1569,11 +1570,11 @@ static blk_qc_t blk_mq_make_request(struct request_queue q, struct bio bio) · 87 unmodified lines
1569	1570
1570	1571	if (!(data.hctx->flags & BLK_MQ_F_BLOCKING)) {
1571	1572	rcu_read_lock();
1572		− blk_mq_try_issue_directly(old_rq, &cookie);
	1573	+ blk_mq_try_issue_directly(old_rq, &cookie, false);
1573	1574	rcu_read_unlock();
1574	1575	} else {
1575	1576	srcu_idx = srcu_read_lock(&data.hctx->queue_rq_srcu);
1576		− blk_mq_try_issue_directly(old_rq, &cookie);
	1577	+ blk_mq_try_issue_directly(old_rq, &cookie, true);
1577	1578	srcu_read_unlock(&data.hctx->queue_rq_srcu, srcu_idx);
1578	1579	}
1579	1580	goto done;
@@ -1955,16 +1956,6 @@ static void blk_mq_exit_hw_queues(struct request_queue *q, · 375 unmodified lines
1955	1956	}
1956	1957	}
1957	1958
1958		−static void blk_mq_free_hw_queues(struct request_queue *q,
1959		− struct blk_mq_tag_set *set)
1960		−{
1961		− struct blk_mq_hw_ctx *hctx;
1962		− unsigned int i;
1963		−
1964		− queue_for_each_hw_ctx(q, hctx, i)
1965		− free_cpumask_var(hctx->cpumask);
1966		−}
1967		−
1968	1959	static int blk_mq_init_hctx(struct request_queue *q,
1969	1960	struct blk_mq_tag_set *set,
1970	1961	struct blk_mq_hw_ctx *hctx, unsigned hctx_idx)
@@ -2045,7 +2036,6 @@ static void blk_mq_init_cpu_queues(struct request_queue *q, · 74 unmodified lines
2045	2036	struct blk_mq_ctx *__ctx = per_cpu_ptr(q->queue_ctx, i);
2046	2037	struct blk_mq_hw_ctx *hctx;
2047	2038
2048		− memset(__ctx, 0, sizeof(*__ctx));
2049	2039	__ctx->cpu = i;
2050	2040	spin_lock_init(&__ctx->lock);
2051	2041	INIT_LIST_HEAD(&__ctx->rq_list);
@@ -2257,15 +2247,19 @@ void blk_mq_release(struct request_queue *q) · 205 unmodified lines
2257	2247	queue_for_each_hw_ctx(q, hctx, i) {
2258	2248	if (!hctx)
2259	2249	continue;
2260		− kfree(hctx->ctxs);
2261		− kfree(hctx);
	2250	+ kobject_put(&hctx->kobj);
2262	2251	}
2263	2252
2264	2253	q->mq_map = NULL;
2265	2254
2266	2255	kfree(q->queue_hw_ctx);
2267	2256
2268		− /* ctx kobj stays in queue_ctx */
	2257	+ /*
	2258	+ * release .mq_kobj and sw queue's kobject now because
	2259	+ * both share lifetime with request queue.
	2260	+ */
	2261	+ blk_mq_sysfs_deinit(q);
	2262	+
2269	2263	free_percpu(q->queue_ctx);
2270	2264	}
2271	2265
@@ -2330,10 +2324,7 @@ static void blk_mq_realloc_hw_ctxs(struct blk_mq_tag_set *set, · 58 unmodified lines
2330	2324	if (hctx->tags)
2331	2325	blk_mq_free_map_and_requests(set, j);
2332	2326	blk_mq_exit_hctx(q, set, hctx, j);
2333		− free_cpumask_var(hctx->cpumask);
2334	2327	kobject_put(&hctx->kobj);
2335		− kfree(hctx->ctxs);
2336		− kfree(hctx);
2337	2328	hctxs[j] = NULL;
2338	2329
2339	2330	}
@@ -2352,6 +2343,9 @@ struct request_queue blk_mq_init_allocated_queue(struct blk_mq_tag_set set, · 12 unmodified lines
2352	2343	if (!q->queue_ctx)
2353	2344	goto err_exit;
2354	2345
	2346	+ /* init q->mq_kobj and sw queues' kobjects */
	2347	+ blk_mq_sysfs_init(q);
	2348	+
2355	2349	q->queue_hw_ctx = kzalloc_node(nr_cpu_ids * sizeof(*(q->queue_hw_ctx)),
2356	2350	GFP_KERNEL, set->numa_node);
2357	2351	if (!q->queue_hw_ctx)
@@ -2442,7 +2436,6 @@ void blk_mq_free_queue(struct request_queue *q) · 84 unmodified lines
2442	2436	blk_mq_del_queue_tag_set(q);
2443	2437
2444	2438	blk_mq_exit_hw_queues(q, set, set->nr_hw_queues);
2445		− blk_mq_free_hw_queues(q, set);
2446	2439	}
2447	2440
2448	2441	/* Basically redo blk_mq_init_queue with queue frozen */

block/blk-mq.h+2−0

@@ -77,6 +77,8 @@ static inline struct blk_mq_hw_ctx blk_mq_map_queue(struct request_queue q, · 76 unmodified lines
77	77	/*
78	78	* sysfs helpers
79	79	*/
	80	+extern void blk_mq_sysfs_init(struct request_queue *q);
	81	+extern void blk_mq_sysfs_deinit(struct request_queue *q);
80	82	extern int blk_mq_sysfs_register(struct request_queue *q);
81	83	extern void blk_mq_sysfs_unregister(struct request_queue *q);
82	84	extern void blk_mq_hctx_kobj_init(struct blk_mq_hw_ctx *hctx);

block/genhd.c+10−27

@@ -572,20 +572,6 @@ exit: · 571 unmodified lines
572	572	disk_part_iter_exit(&piter);
573	573	}
574	574
575		−void put_disk_devt(struct disk_devt *disk_devt)
576		−{
577		− if (disk_devt && atomic_dec_and_test(&disk_devt->count))
578		− disk_devt->release(disk_devt);
579		−}
580		−EXPORT_SYMBOL(put_disk_devt);
581		−
582		−void get_disk_devt(struct disk_devt *disk_devt)
583		−{
584		− if (disk_devt)
585		− atomic_inc(&disk_devt->count);
586		−}
587		−EXPORT_SYMBOL(get_disk_devt);
588		−
589	575	/**
590	576	* device_add_disk - add partitioning information to kernel list
591	577	* @parent: parent device for the disk
@@ -626,13 +612,6 @@ void device_add_disk(struct device parent, struct gendisk disk) · 34 unmodified lines
626	612
627	613	disk_alloc_events(disk);
628	614
629		− /*
630		− * Take a reference on the devt and assign it to queue since it
631		− * must not be reallocated while the bdi is registered
632		− */
633		− disk->queue->disk_devt = disk->disk_devt;
634		− get_disk_devt(disk->disk_devt);
635		−
636	615	/* Register BDI before referencing it from bdev */
637	616	bdi = disk->queue->backing_dev_info;
638	617	bdi_register_owner(bdi, disk_to_dev(disk));
@@ -681,12 +660,16 @@ void del_gendisk(struct gendisk *disk) · 42 unmodified lines
681	660	disk->flags &= ~GENHD_FL_UP;
682	661
683	662	sysfs_remove_link(&disk_to_dev(disk)->kobj, "bdi");
684		− /*
685		− * Unregister bdi before releasing device numbers (as they can get
686		− * reused and we'd get clashes in sysfs).
687		− */
688		− bdi_unregister(disk->queue->backing_dev_info);
689		− blk_unregister_queue(disk);
	663	+ if (disk->queue) {
	664	+ /*
	665	+ * Unregister bdi before releasing device numbers (as they can
	666	+ * get reused and we'd get clashes in sysfs).
	667	+ */
	668	+ bdi_unregister(disk->queue->backing_dev_info);
	669	+ blk_unregister_queue(disk);
	670	+ } else {
	671	+ WARN_ON(1);
	672	+ }
690	673	blk_unregister_region(disk_devt(disk), disk->minors);
691	674
692	675	part_stat_set_all(&disk->part0, 0);

block/sed-opal.c+2−8

@@ -1023,7 +1023,6 @@ static int finalize_and_send(struct opal_dev *dev, cont_fn cont) · 1022 unmodified lines
1023	1023
1024	1024	static int gen_key(struct opal_dev dev, void data)
1025	1025	{
1026		− const u8 *method;
1027	1026	u8 uid[OPAL_UID_LENGTH];
1028	1027	int err = 0;
1029	1028
@@ -1031,7 +1030,6 @@ static int gen_key(struct opal_dev dev, void data) · 1 unmodified line
1031	1030	set_comid(dev, dev->comid);
1032	1031
1033	1032	memcpy(uid, dev->prev_data, min(sizeof(uid), dev->prev_d_len));
1034		− method = opalmethod[OPAL_GENKEY];
1035	1033	kfree(dev->prev_data);
1036	1034	dev->prev_data = NULL;
1037	1035
@@ -1669,7 +1667,6 @@ static int add_user_to_lr(struct opal_dev dev, void data) · 631 unmodified lines
1669	1667	static int lock_unlock_locking_range(struct opal_dev dev, void data)
1670	1668	{
1671	1669	u8 lr_buffer[OPAL_UID_LENGTH];
1672		− const u8 *method;
1673	1670	struct opal_lock_unlock *lkul = data;
1674	1671	u8 read_locked = 1, write_locked = 1;
1675	1672	int err = 0;
@@ -1677,7 +1674,6 @@ static int lock_unlock_locking_range(struct opal_dev dev, void data) · 1 unmodified line
1677	1674	clear_opal_cmd(dev);
1678	1675	set_comid(dev, dev->comid);
1679	1676
1680		− method = opalmethod[OPAL_SET];
1681	1677	if (build_locking_range(lr_buffer, sizeof(lr_buffer),
1682	1678	lkul->session.opal_key.lr) < 0)
1683	1679	return -ERANGE;
@@ -1733,14 +1729,12 @@ static int lock_unlock_locking_range_sum(struct opal_dev dev, void data) · 49 unmodified lines
1733	1729	{
1734	1730	u8 lr_buffer[OPAL_UID_LENGTH];
1735	1731	u8 read_locked = 1, write_locked = 1;
1736		− const u8 *method;
1737	1732	struct opal_lock_unlock *lkul = data;
1738	1733	int ret;
1739	1734
1740	1735	clear_opal_cmd(dev);
1741	1736	set_comid(dev, dev->comid);
1742	1737
1743		− method = opalmethod[OPAL_SET];
1744	1738	if (build_locking_range(lr_buffer, sizeof(lr_buffer),
1745	1739	lkul->session.opal_key.lr) < 0)
1746	1740	return -ERANGE;
@@ -2133,7 +2127,7 @@ static int opal_add_user_to_lr(struct opal_dev *dev, · 386 unmodified lines
2133	2127	pr_err("Locking state was not RO or RW\n");
2134	2128	return -EINVAL;
2135	2129	}
2136		− if (lk_unlk->session.who < OPAL_USER1 &&
	2130	+ if (lk_unlk->session.who < OPAL_USER1 \|\|
2137	2131	lk_unlk->session.who > OPAL_USER9) {
2138	2132	pr_err("Authority was not within the range of users: %d\n",
2139	2133	lk_unlk->session.who);
@@ -2316,7 +2310,7 @@ static int opal_activate_user(struct opal_dev *dev, · 176 unmodified lines
2316	2310	int ret;
2317	2311
2318	2312	/* We can't activate Admin1 it's active as manufactured */
2319		− if (opal_session->who < OPAL_USER1 &&
	2313	+ if (opal_session->who < OPAL_USER1 \|\|
2320	2314	opal_session->who > OPAL_USER9) {
2321	2315	pr_err("Who was not a valid user: %d\n", opal_session->who);
2322	2316	return -EINVAL;

crypto/af_alg.c+5−4

@@ -266,7 +266,7 @@ unlock: · 265 unmodified lines
266	266	return err;
267	267	}
268	268
269		−int af_alg_accept(struct sock sk, struct socket newsock)
	269	+int af_alg_accept(struct sock sk, struct socket newsock, bool kern)
270	270	{
271	271	struct alg_sock *ask = alg_sk(sk);
272	272	const struct af_alg_type *type;
@@ -281,7 +281,7 @@ int af_alg_accept(struct sock sk, struct socket newsock) · 8 unmodified lines
281	281	if (!type)
282	282	goto unlock;
283	283
284		− sk2 = sk_alloc(sock_net(sk), PF_ALG, GFP_KERNEL, &alg_proto, 0);
	284	+ sk2 = sk_alloc(sock_net(sk), PF_ALG, GFP_KERNEL, &alg_proto, kern);
285	285	err = -ENOMEM;
286	286	if (!sk2)
287	287	goto unlock;
@@ -323,9 +323,10 @@ unlock: · 35 unmodified lines
323	323	}
324	324	EXPORT_SYMBOL_GPL(af_alg_accept);
325	325
326		−static int alg_accept(struct socket sock, struct socket newsock, int flags)
	326	+static int alg_accept(struct socket sock, struct socket newsock, int flags,
	327	+ bool kern)
327	328	{
328		− return af_alg_accept(sock->sk, newsock);
	329	+ return af_alg_accept(sock->sk, newsock, kern);
329	330	}
330	331
331	332	static const struct proto_ops alg_proto_ops = {

crypto/algif_hash.c+5−4

@@ -239,7 +239,8 @@ unlock: · 238 unmodified lines
239	239	return err ?: len;
240	240	}
241	241
242		−static int hash_accept(struct socket sock, struct socket newsock, int flags)
	242	+static int hash_accept(struct socket sock, struct socket newsock, int flags,
	243	+ bool kern)
243	244	{
244	245	struct sock *sk = sock->sk;
245	246	struct alg_sock *ask = alg_sk(sk);
@@ -260,7 +261,7 @@ static int hash_accept(struct socket sock, struct socket newsock, int flags) · 14 unmodified lines
260	261	if (err)
261	262	return err;
262	263
263		− err = af_alg_accept(ask->parent, newsock);
	264	+ err = af_alg_accept(ask->parent, newsock, kern);
264	265	if (err)
265	266	return err;
266	267
@@ -378,7 +379,7 @@ static int hash_recvmsg_nokey(struct socket sock, struct msghdr msg, · 111 unmodified lines
378	379	}
379	380
380	381	static int hash_accept_nokey(struct socket sock, struct socket newsock,
381		− int flags)
	382	+ int flags, bool kern)
382	383	{
383	384	int err;
384	385
@@ -386,7 +387,7 @@ static int hash_accept_nokey(struct socket sock, struct socket newsock, · 1 unmodified line
386	387	if (err)
387	388	return err;
388	389
389		− return hash_accept(sock, newsock, flags);
	390	+ return hash_accept(sock, newsock, flags, kern);
390	391	}
391	392
392	393	static struct proto_ops algif_hash_ops_nokey = {

drivers/acpi/internal.h+2−0

@@ -41,8 +41,10 @@ void acpi_gpe_apply_masked_gpes(void); · 40 unmodified lines
41	41	void acpi_container_init(void);
42	42	void acpi_memory_hotplug_init(void);
43	43	#ifdef CONFIG_ACPI_HOTPLUG_IOAPIC
	44	+void pci_ioapic_remove(struct acpi_pci_root *root);
44	45	int acpi_ioapic_remove(struct acpi_pci_root *root);
45	46	#else
	47	+static inline void pci_ioapic_remove(struct acpi_pci_root *root) { return; }
46	48	static inline int acpi_ioapic_remove(struct acpi_pci_root *root) { return 0; }
47	49	#endif
48	50	#ifdef CONFIG_ACPI_DOCK

drivers/acpi/ioapic.c+16−6

@@ -206,24 +206,34 @@ int acpi_ioapic_add(acpi_handle root_handle) · 205 unmodified lines
206	206	return ACPI_SUCCESS(status) && ACPI_SUCCESS(retval) ? 0 : -ENODEV;
207	207	}
208	208
209		−int acpi_ioapic_remove(struct acpi_pci_root *root)
	209	+void pci_ioapic_remove(struct acpi_pci_root *root)
210	210	{
211		− int retval = 0;
212	211	struct acpi_pci_ioapic ioapic, tmp;
213	212
214	213	mutex_lock(&ioapic_list_lock);
215	214	list_for_each_entry_safe(ioapic, tmp, &ioapic_list, list) {
216	215	if (root->device->handle != ioapic->root_handle)
217	216	continue;
218		−
219		− if (acpi_unregister_ioapic(ioapic->handle, ioapic->gsi_base))
220		− retval = -EBUSY;
221		−
222	217	if (ioapic->pdev) {
223	218	pci_release_region(ioapic->pdev, 0);
224	219	pci_disable_device(ioapic->pdev);
225	220	pci_dev_put(ioapic->pdev);
226	221	}
	222	+ }
	223	+ mutex_unlock(&ioapic_list_lock);
	224	+}
	225	+
	226	+int acpi_ioapic_remove(struct acpi_pci_root *root)
	227	+{
	228	+ int retval = 0;
	229	+ struct acpi_pci_ioapic ioapic, tmp;
	230	+
	231	+ mutex_lock(&ioapic_list_lock);
	232	+ list_for_each_entry_safe(ioapic, tmp, &ioapic_list, list) {
	233	+ if (root->device->handle != ioapic->root_handle)
	234	+ continue;
	235	+ if (acpi_unregister_ioapic(ioapic->handle, ioapic->gsi_base))
	236	+ retval = -EBUSY;
227	237	if (ioapic->res.flags && ioapic->res.parent)
228	238	release_resource(&ioapic->res);
229	239	list_del(&ioapic->list);

drivers/acpi/pci_root.c+2−2

@@ -648,12 +648,12 @@ static void acpi_pci_root_remove(struct acpi_device *device) · 647 unmodified lines
648	648
649	649	pci_stop_root_bus(root->bus);
650	650
651		− WARN_ON(acpi_ioapic_remove(root));
652		−
	651	+ pci_ioapic_remove(root);
653	652	device_set_run_wake(root->bus->bridge, false);
654	653	pci_acpi_remove_bus_pm_notifier(device);
655	654
656	655	pci_remove_root_bus(root->bus);
	656	+ WARN_ON(acpi_ioapic_remove(root));
657	657
658	658	dmar_device_remove(device->handle);
659	659

drivers/ata/ahci_qoriq.c+4−2

@@ -177,7 +177,8 @@ static int ahci_qoriq_phy_init(struct ahci_host_priv *hpriv) · 176 unmodified lines
177	177	case AHCI_LS1043A:
178	178	if (!qpriv->ecc_addr)
179	179	return -EINVAL;
180		− writel(ECC_DIS_ARMV8_CH2, qpriv->ecc_addr);
	180	+ writel(readl(qpriv->ecc_addr) \| ECC_DIS_ARMV8_CH2,
	181	+ qpriv->ecc_addr);
181	182	writel(AHCI_PORT_PHY_1_CFG, reg_base + PORT_PHY1);
182	183	writel(AHCI_PORT_TRANS_CFG, reg_base + PORT_TRANS);
183	184	if (qpriv->is_dmacoherent)
@@ -194,7 +195,8 @@ static int ahci_qoriq_phy_init(struct ahci_host_priv *hpriv) · 10 unmodified lines
194	195	case AHCI_LS1046A:
195	196	if (!qpriv->ecc_addr)
196	197	return -EINVAL;
197		− writel(ECC_DIS_ARMV8_CH2, qpriv->ecc_addr);
	198	+ writel(readl(qpriv->ecc_addr) \| ECC_DIS_ARMV8_CH2,
	199	+ qpriv->ecc_addr);
198	200	writel(AHCI_PORT_PHY_1_CFG, reg_base + PORT_PHY1);
199	201	writel(AHCI_PORT_TRANS_CFG, reg_base + PORT_TRANS);
200	202	if (qpriv->is_dmacoherent)

drivers/ata/libata-sff.c+0−1

@@ -1482,7 +1482,6 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc) · 1481 unmodified lines
1482	1482	break;
1483	1483
1484	1484	default:
1485		− WARN_ON_ONCE(1);
1486	1485	return AC_ERR_SYSTEM;
1487	1486	}
1488	1487

drivers/ata/libata-transport.c+3−6

@@ -224,7 +224,6 @@ static DECLARE_TRANSPORT_CLASS(ata_port_class, · 223 unmodified lines
224	224
225	225	static void ata_tport_release(struct device *dev)
226	226	{
227		− put_device(dev->parent);
228	227	}
229	228
230	229	/**
@@ -284,7 +283,7 @@ int ata_tport_add(struct device *parent, · 53 unmodified lines
284	283	device_initialize(dev);
285	284	dev->type = &ata_port_type;
286	285
287		− dev->parent = get_device(parent);
	286	+ dev->parent = parent;
288	287	dev->release = ata_tport_release;
289	288	dev_set_name(dev, "ata%d", ap->print_id);
290	289	transport_setup_device(dev);
@@ -348,7 +347,6 @@ static DECLARE_TRANSPORT_CLASS(ata_link_class, · 57 unmodified lines
348	347
349	348	static void ata_tlink_release(struct device *dev)
350	349	{
351		− put_device(dev->parent);
352	350	}
353	351
354	352	/**
@@ -410,7 +408,7 @@ int ata_tlink_add(struct ata_link *link) · 55 unmodified lines
410	408	int error;
411	409
412	410	device_initialize(dev);
413		− dev->parent = get_device(&ap->tdev);
	411	+ dev->parent = &ap->tdev;
414	412	dev->release = ata_tlink_release;
415	413	if (ata_is_host_link(link))
416	414	dev_set_name(dev, "link%d", ap->print_id);
@@ -589,7 +587,6 @@ static DECLARE_TRANSPORT_CLASS(ata_dev_class, · 172 unmodified lines
589	587
590	588	static void ata_tdev_release(struct device *dev)
591	589	{
592		− put_device(dev->parent);
593	590	}
594	591
595	592	/**
@@ -662,7 +659,7 @@ static int ata_tdev_add(struct ata_device *ata_dev) · 66 unmodified lines
662	659	int error;
663	660
664	661	device_initialize(dev);
665		− dev->parent = get_device(&link->tdev);
	662	+ dev->parent = &link->tdev;
666	663	dev->release = ata_tdev_release;
667	664	if (ata_is_host_link(link))
668	665	dev_set_name(dev, "dev%d.%d", ap->print_id,ata_dev->devno);

drivers/block/paride/pcd.c+1−1

@@ -50,7 +50,7 @@ · 49 unmodified lines
50	50	the slower the port i/o. In some cases, setting
51	51	this to zero will speed up the device. (default -1)
52	52
53		− major You may use this parameter to overide the
	53	+ major You may use this parameter to override the
54	54	default major number (46) that this driver
55	55	will use. Be sure to change the device
56	56	name as well.

drivers/block/paride/pd.c+1−1

@@ -61,7 +61,7 @@ · 60 unmodified lines
61	61	first drive found.
62	62
63	63
64		− major You may use this parameter to overide the
	64	+ major You may use this parameter to override the
65	65	default major number (45) that this driver
66	66	will use. Be sure to change the device
67	67	name as well.

drivers/block/paride/pf.c+1−1

@@ -59,7 +59,7 @@ · 58 unmodified lines
59	59	the slower the port i/o. In some cases, setting
60	60	this to zero will speed up the device. (default -1)
61	61
62		− major You may use this parameter to overide the
	62	+ major You may use this parameter to override the
63	63	default major number (47) that this driver
64	64	will use. Be sure to change the device
65	65	name as well.

drivers/block/paride/pg.c+1−1

@@ -84,7 +84,7 @@ · 83 unmodified lines
84	84	the slower the port i/o. In some cases, setting
85	85	this to zero will speed up the device. (default -1)
86	86
87		− major You may use this parameter to overide the
	87	+ major You may use this parameter to override the
88	88	default major number (97) that this driver
89	89	will use. Be sure to change the device
90	90	name as well.

drivers/block/paride/pt.c+1−1

@@ -61,7 +61,7 @@ · 60 unmodified lines
61	61	the slower the port i/o. In some cases, setting
62	62	this to zero will speed up the device. (default -1)
63	63
64		− major You may use this parameter to overide the
	64	+ major You may use this parameter to override the
65	65	default major number (96) that this driver
66	66	will use. Be sure to change the device
67	67	name as well.

drivers/block/rbd.c+12−4

@@ -120,10 +120,11 @@ static int atomic_dec_return_safe(atomic_t *v) · 119 unmodified lines
120	120
121	121	/* Feature bits */
122	122
123		−#define RBD_FEATURE_LAYERING (1<<0)
124		−#define RBD_FEATURE_STRIPINGV2 (1<<1)
125		−#define RBD_FEATURE_EXCLUSIVE_LOCK (1<<2)
126		−#define RBD_FEATURE_DATA_POOL (1<<7)
	123	+#define RBD_FEATURE_LAYERING (1ULL<<0)
	124	+#define RBD_FEATURE_STRIPINGV2 (1ULL<<1)
	125	+#define RBD_FEATURE_EXCLUSIVE_LOCK (1ULL<<2)
	126	+#define RBD_FEATURE_DATA_POOL (1ULL<<7)
	127	+
127	128	#define RBD_FEATURES_ALL (RBD_FEATURE_LAYERING \| \
128	129	RBD_FEATURE_STRIPINGV2 \| \
129	130	RBD_FEATURE_EXCLUSIVE_LOCK \| \
@@ -499,16 +500,23 @@ static bool rbd_is_lock_owner(struct rbd_device *rbd_dev) · 369 unmodified lines
499	500	return is_lock_owner;
500	501	}
501	502
	503	+static ssize_t rbd_supported_features_show(struct bus_type bus, char buf)
	504	+{
	505	+ return sprintf(buf, "0x%llx\n", RBD_FEATURES_SUPPORTED);
	506	+}
	507	+
502	508	static BUS_ATTR(add, S_IWUSR, NULL, rbd_add);
503	509	static BUS_ATTR(remove, S_IWUSR, NULL, rbd_remove);
504	510	static BUS_ATTR(add_single_major, S_IWUSR, NULL, rbd_add_single_major);
505	511	static BUS_ATTR(remove_single_major, S_IWUSR, NULL, rbd_remove_single_major);
	512	+static BUS_ATTR(supported_features, S_IRUGO, rbd_supported_features_show, NULL);
506	513
507	514	static struct attribute *rbd_bus_attrs[] = {
508	515	&bus_attr_add.attr,
509	516	&bus_attr_remove.attr,
510	517	&bus_attr_add_single_major.attr,
511	518	&bus_attr_remove_single_major.attr,
	519	+ &bus_attr_supported_features.attr,
512	520	NULL,
513	521	};
514	522

drivers/block/zram/zram_drv.c+2−0

@@ -1189,6 +1189,8 @@ static int zram_add(void) · 1188 unmodified lines
1189	1189	blk_queue_io_min(zram->disk->queue, PAGE_SIZE);
1190	1190	blk_queue_io_opt(zram->disk->queue, PAGE_SIZE);
1191	1191	zram->disk->queue->limits.discard_granularity = PAGE_SIZE;
	1192	+ zram->disk->queue->limits.max_sectors = SECTORS_PER_PAGE;
	1193	+ zram->disk->queue->limits.chunk_sectors = 0;
1192	1194	blk_queue_max_discard_sectors(zram->disk->queue, UINT_MAX);
1193	1195	/*
1194	1196	* zram_bio_discard() will clear all logical blocks if logical block

drivers/char/hw_random/omap-rng.c+14−2

@@ -397,9 +397,8 @@ static int of_get_omap_rng_device_details(struct omap_rng_dev *priv, · 396 unmodified lines
397	397	irq, err);
398	398	return err;
399	399	}
400		− omap_rng_write(priv, RNG_INTMASK_REG, RNG_SHUTDOWN_OFLO_MASK);
401	400
402		− priv->clk = of_clk_get(pdev->dev.of_node, 0);
	401	+ priv->clk = devm_clk_get(&pdev->dev, NULL);
403	402	if (IS_ERR(priv->clk) && PTR_ERR(priv->clk) == -EPROBE_DEFER)
404	403	return -EPROBE_DEFER;
405	404	if (!IS_ERR(priv->clk)) {
@@ -408,6 +407,19 @@ static int of_get_omap_rng_device_details(struct omap_rng_dev *priv, · 2 unmodified lines
408	407	dev_err(&pdev->dev, "unable to enable the clk, "
409	408	"err = %d\n", err);
410	409	}
	410	+
	411	+ /*
	412	+ * On OMAP4, enabling the shutdown_oflo interrupt is
	413	+ * done in the interrupt mask register. There is no
	414	+ * such register on EIP76, and it's enabled by the
	415	+ * same bit in the control register
	416	+ */
	417	+ if (priv->pdata->regs[RNG_INTMASK_REG])
	418	+ omap_rng_write(priv, RNG_INTMASK_REG,
	419	+ RNG_SHUTDOWN_OFLO_MASK);
	420	+ else
	421	+ omap_rng_write(priv, RNG_CONTROL_REG,
	422	+ RNG_SHUTDOWN_OFLO_MASK);
411	423	}
412	424	return 0;
413	425	}

drivers/char/nwbutton.c+1−1

@@ -6,7 +6,7 @@ · 5 unmodified lines
6	6
7	7	#include <linux/module.h>
8	8	#include <linux/kernel.h>
9		−#include <linux/sched.h>
	9	+#include <linux/sched/signal.h>
10	10	#include <linux/interrupt.h>
11	11	#include <linux/time.h>
12	12	#include <linux/timer.h>

drivers/char/random.c+50−79

@@ -312,13 +312,6 @@ static int random_read_wakeup_bits = 64; · 311 unmodified lines
312	312	*/
313	313	static int random_write_wakeup_bits = 28 * OUTPUT_POOL_WORDS;
314	314
315		−/*
316		− * The minimum number of seconds between urandom pool reseeding. We
317		− * do this to limit the amount of entropy that can be drained from the
318		− * input pool even if there are heavy demands on /dev/urandom.
319		− */
320		−static int random_min_urandom_seed = 60;
321		−
322	315	/*
323	316	* Originally, we used a primitive polynomial of degree .poolwords
324	317	* over GF(2). The taps for various sizes are defined below. They
@@ -409,7 +402,6 @@ static struct poolinfo { · 84 unmodified lines
409	402	*/
410	403	static DECLARE_WAIT_QUEUE_HEAD(random_read_wait);
411	404	static DECLARE_WAIT_QUEUE_HEAD(random_write_wait);
412		−static DECLARE_WAIT_QUEUE_HEAD(urandom_init_wait);
413	405	static struct fasync_struct *fasync;
414	406
415	407	static DEFINE_SPINLOCK(random_ready_list_lock);
@@ -467,7 +459,6 @@ struct entropy_store { · 51 unmodified lines
467	459	int entropy_count;
468	460	int entropy_total;
469	461	unsigned int initialized:1;
470		− unsigned int limit:1;
471	462	unsigned int last_data_init:1;
472	463	__u8 last_data[EXTRACT_SIZE];
473	464	};
@@ -485,7 +476,6 @@ static __u32 blocking_pool_data[OUTPUT_POOL_WORDS] __latent_entropy; · 11 unmodified lines
485	476	static struct entropy_store input_pool = {
486	477	.poolinfo = &poolinfo_table[0],
487	478	.name = "input",
488		− .limit = 1,
489	479	.lock = __SPIN_LOCK_UNLOCKED(input_pool.lock),
490	480	.pool = input_pool_data
491	481	};
@@ -493,7 +483,6 @@ static struct entropy_store input_pool = { · 1 unmodified line
493	483	static struct entropy_store blocking_pool = {
494	484	.poolinfo = &poolinfo_table[1],
495	485	.name = "blocking",
496		− .limit = 1,
497	486	.pull = &input_pool,
498	487	.lock = __SPIN_LOCK_UNLOCKED(blocking_pool.lock),
499	488	.pool = blocking_pool_data,
@@ -855,13 +844,6 @@ static void crng_reseed(struct crng_state crng, struct entropy_store r) · 355 unmodified lines
855	844	spin_unlock_irqrestore(&primary_crng.lock, flags);
856	845	}
857	846
858		−static inline void maybe_reseed_primary_crng(void)
859		−{
860		− if (crng_init > 2 &&
861		− time_after(jiffies, primary_crng.init_time + CRNG_RESEED_INTERVAL))
862		− crng_reseed(&primary_crng, &input_pool);
863		−}
864		−
865	847	static inline void crng_wait_ready(void)
866	848	{
867	849	wait_event_interruptible(crng_init_wait, crng_ready());
@@ -1220,15 +1202,6 @@ static void xfer_secondary_pool(struct entropy_store *r, size_t nbytes) · 352 unmodified lines
1220	1202	r->entropy_count > r->poolinfo->poolfracbits)
1221	1203	return;
1222	1204
1223		− if (r->limit == 0 && random_min_urandom_seed) {
1224		− unsigned long now = jiffies;
1225		−
1226		− if (time_before(now,
1227		− r->last_pulled + random_min_urandom_seed * HZ))
1228		− return;
1229		− r->last_pulled = now;
1230		− }
1231		−
1232	1205	_xfer_secondary_pool(r, nbytes);
1233	1206	}
1234	1207
@@ -1236,8 +1209,6 @@ static void _xfer_secondary_pool(struct entropy_store *r, size_t nbytes) · 1 unmodified line
1236	1209	{
1237	1210	__u32 tmp[OUTPUT_POOL_WORDS];
1238	1211
1239		− /* For /dev/random's pool, always leave two wakeups' worth */
1240		− int rsvd_bytes = r->limit ? 0 : random_read_wakeup_bits / 4;
1241	1212	int bytes = nbytes;
1242	1213
1243	1214	/* pull at least as much as a wakeup */
@@ -1248,7 +1219,7 @@ static void _xfer_secondary_pool(struct entropy_store *r, size_t nbytes) · 4 unmodified lines
1248	1219	trace_xfer_secondary_pool(r->name, bytes * 8, nbytes * 8,
1249	1220	ENTROPY_BITS(r), ENTROPY_BITS(r->pull));
1250	1221	bytes = extract_entropy(r->pull, tmp, bytes,
1251		− random_read_wakeup_bits / 8, rsvd_bytes);
	1222	+ random_read_wakeup_bits / 8, 0);
1252	1223	mix_pool_bytes(r, tmp, bytes);
1253	1224	credit_entropy_bits(r, bytes*8);
1254	1225	}
@@ -1276,7 +1247,7 @@ static void push_to_pool(struct work_struct *work) · 21 unmodified lines
1276	1247	static size_t account(struct entropy_store *r, size_t nbytes, int min,
1277	1248	int reserved)
1278	1249	{
1279		− int entropy_count, orig;
	1250	+ int entropy_count, orig, have_bytes;
1280	1251	size_t ibytes, nfrac;
1281	1252
1282	1253	BUG_ON(r->entropy_count > r->poolinfo->poolfracbits);
@@ -1285,14 +1256,12 @@ static size_t account(struct entropy_store *r, size_t nbytes, int min, · 2 unmodified lines
1285	1256	retry:
1286	1257	entropy_count = orig = ACCESS_ONCE(r->entropy_count);
1287	1258	ibytes = nbytes;
1288		− /* If limited, never pull more than available */
1289		− if (r->limit) {
1290		− int have_bytes = entropy_count >> (ENTROPY_SHIFT + 3);
	1259	+ /* never pull more than available */
	1260	+ have_bytes = entropy_count >> (ENTROPY_SHIFT + 3);
1291	1261
1292		− if ((have_bytes -= reserved) < 0)
1293		− have_bytes = 0;
1294		− ibytes = min_t(size_t, ibytes, have_bytes);
1295		− }
	1262	+ if ((have_bytes -= reserved) < 0)
	1263	+ have_bytes = 0;
	1264	+ ibytes = min_t(size_t, ibytes, have_bytes);
1296	1265	if (ibytes < min)
1297	1266	ibytes = 0;
1298	1267
@@ -1912,6 +1881,7 @@ SYSCALL_DEFINE3(getrandom, char __user *, buf, size_t, count, · 613 unmodified lines
1912	1881	static int min_read_thresh = 8, min_write_thresh;
1913	1882	static int max_read_thresh = OUTPUT_POOL_WORDS * 32;
1914	1883	static int max_write_thresh = INPUT_POOL_WORDS * 32;
	1884	+static int random_min_urandom_seed = 60;
1915	1885	static char sysctl_bootid[16];
1916	1886
1917	1887	/*
@@ -2042,63 +2012,64 @@ struct ctl_table random_table[] = { · 124 unmodified lines
2042	2012	};
2043	2013	#endif /* CONFIG_SYSCTL */
2044	2014
2045		−static u32 random_int_secret[MD5_MESSAGE_BYTES / 4] ____cacheline_aligned;
2046		−
2047		−int random_int_secret_init(void)
2048		−{
2049		− get_random_bytes(random_int_secret, sizeof(random_int_secret));
2050		− return 0;
2051		−}
2052		−
2053		−static DEFINE_PER_CPU(__u32 [MD5_DIGEST_WORDS], get_random_int_hash)
2054		− __aligned(sizeof(unsigned long));
	2015	+struct batched_entropy {
	2016	+ union {
	2017	+ u64 entropy_u64[CHACHA20_BLOCK_SIZE / sizeof(u64)];
	2018	+ u32 entropy_u32[CHACHA20_BLOCK_SIZE / sizeof(u32)];
	2019	+ };
	2020	+ unsigned int position;
	2021	+};
2055	2022
2056	2023	/*
2057		− * Get a random word for internal kernel use only. Similar to urandom but
2058		− * with the goal of minimal entropy pool depletion. As a result, the random
2059		− * value is not cryptographically secure but for several uses the cost of
2060		− * depleting entropy is too high
	2024	+ * Get a random word for internal kernel use only. The quality of the random
	2025	+ * number is either as good as RDRAND or as good as /dev/urandom, with the
	2026	+ * goal of being quite fast and not depleting entropy.
2061	2027	*/
2062		−unsigned int get_random_int(void)
	2028	+static DEFINE_PER_CPU(struct batched_entropy, batched_entropy_u64);
	2029	+u64 get_random_u64(void)
2063	2030	{
2064		− __u32 *hash;
2065		− unsigned int ret;
	2031	+ u64 ret;
	2032	+ struct batched_entropy *batch;
2066	2033
2067		− if (arch_get_random_int(&ret))
	2034	+#if BITS_PER_LONG == 64
	2035	+ if (arch_get_random_long((unsigned long *)&ret))
2068	2036	return ret;
	2037	+#else
	2038	+ if (arch_get_random_long((unsigned long *)&ret) &&
	2039	+ arch_get_random_long((unsigned long *)&ret + 1))
	2040	+ return ret;
	2041	+#endif
2069	2042
2070		− hash = get_cpu_var(get_random_int_hash);
2071		−
2072		− hash[0] += current->pid + jiffies + random_get_entropy();
2073		− md5_transform(hash, random_int_secret);
2074		− ret = hash[0];
2075		− put_cpu_var(get_random_int_hash);
2076		−
	2043	+ batch = &get_cpu_var(batched_entropy_u64);
	2044	+ if (batch->position % ARRAY_SIZE(batch->entropy_u64) == 0) {
	2045	+ extract_crng((u8 *)batch->entropy_u64);
	2046	+ batch->position = 0;
	2047	+ }
	2048	+ ret = batch->entropy_u64[batch->position++];
	2049	+ put_cpu_var(batched_entropy_u64);
2077	2050	return ret;
2078	2051	}
2079		−EXPORT_SYMBOL(get_random_int);
	2052	+EXPORT_SYMBOL(get_random_u64);
2080	2053
2081		−/*
2082		− * Same as get_random_int(), but returns unsigned long.
2083		− */
2084		−unsigned long get_random_long(void)
	2054	+static DEFINE_PER_CPU(struct batched_entropy, batched_entropy_u32);
	2055	+u32 get_random_u32(void)
2085	2056	{
2086		− __u32 *hash;
2087		− unsigned long ret;
	2057	+ u32 ret;
	2058	+ struct batched_entropy *batch;
2088	2059
2089		− if (arch_get_random_long(&ret))
	2060	+ if (arch_get_random_int(&ret))
2090	2061	return ret;
2091	2062
2092		− hash = get_cpu_var(get_random_int_hash);
2093		−
2094		− hash[0] += current->pid + jiffies + random_get_entropy();
2095		− md5_transform(hash, random_int_secret);
2096		− ret = (unsigned long )hash;
2097		− put_cpu_var(get_random_int_hash);
2098		−
	2063	+ batch = &get_cpu_var(batched_entropy_u32);
	2064	+ if (batch->position % ARRAY_SIZE(batch->entropy_u32) == 0) {
	2065	+ extract_crng((u8 *)batch->entropy_u32);
	2066	+ batch->position = 0;
	2067	+ }
	2068	+ ret = batch->entropy_u32[batch->position++];
	2069	+ put_cpu_var(batched_entropy_u32);
2099	2070	return ret;
2100	2071	}
2101		−EXPORT_SYMBOL(get_random_long);
	2072	+EXPORT_SYMBOL(get_random_u32);
2102	2073
2103	2074	/**
2104	2075	* randomize_page - Generate a random, page aligned address

drivers/cpufreq/cpufreq.c+1−0

@@ -2532,4 +2532,5 @@ static int __init cpufreq_core_init(void) · 2531 unmodified lines
2532	2532
2533	2533	return 0;
2534	2534	}
	2535	+module_param(off, int, 0444);
2535	2536	core_initcall(cpufreq_core_init);

drivers/cpufreq/intel_pstate.c+31−36

@@ -377,6 +377,7 @@ static void intel_pstate_set_performance_limits(struct perf_limits *limits) · 376 unmodified lines
377	377	intel_pstate_init_limits(limits);
378	378	limits->min_perf_pct = 100;
379	379	limits->min_perf = int_ext_tofp(1);
	380	+ limits->min_sysfs_pct = 100;
380	381	}
381	382
382	383	static DEFINE_MUTEX(intel_pstate_driver_lock);
@@ -968,11 +969,20 @@ static int intel_pstate_resume(struct cpufreq_policy *policy) · 585 unmodified lines
968	969	}
969	970
970	971	static void intel_pstate_update_policies(void)
	972	+ __releases(&intel_pstate_limits_lock)
	973	+ __acquires(&intel_pstate_limits_lock)
971	974	{
	975	+ struct perf_limits *saved_limits = limits;
972	976	int cpu;
973	977
	978	+ mutex_unlock(&intel_pstate_limits_lock);
	979	+
974	980	for_each_possible_cpu(cpu)
975	981	cpufreq_update_policy(cpu);
	982	+
	983	+ mutex_lock(&intel_pstate_limits_lock);
	984	+
	985	+ limits = saved_limits;
976	986	}
977	987
978	988	/************************ debugfs begin **********************/
@@ -1180,10 +1190,10 @@ static ssize_t store_no_turbo(struct kobject a, struct attribute b, · 201 unmodified lines
1180	1190
1181	1191	limits->no_turbo = clamp_t(int, input, 0, 1);
1182	1192
1183		− mutex_unlock(&intel_pstate_limits_lock);
1184		−
1185	1193	intel_pstate_update_policies();
1186	1194
	1195	+ mutex_unlock(&intel_pstate_limits_lock);
	1196	+
1187	1197	mutex_unlock(&intel_pstate_driver_lock);
1188	1198
1189	1199	return count;
@@ -1217,10 +1227,10 @@ static ssize_t store_max_perf_pct(struct kobject a, struct attribute b, · 27 unmodified lines
1217	1227	limits->max_perf_pct);
1218	1228	limits->max_perf = div_ext_fp(limits->max_perf_pct, 100);
1219	1229
1220		− mutex_unlock(&intel_pstate_limits_lock);
1221		−
1222	1230	intel_pstate_update_policies();
1223	1231
	1232	+ mutex_unlock(&intel_pstate_limits_lock);
	1233	+
1224	1234	mutex_unlock(&intel_pstate_driver_lock);
1225	1235
1226	1236	return count;
@@ -1254,10 +1264,10 @@ static ssize_t store_min_perf_pct(struct kobject a, struct attribute b, · 27 unmodified lines
1254	1264	limits->min_perf_pct);
1255	1265	limits->min_perf = div_ext_fp(limits->min_perf_pct, 100);
1256	1266
1257		− mutex_unlock(&intel_pstate_limits_lock);
1258		−
1259	1267	intel_pstate_update_policies();
1260	1268
	1269	+ mutex_unlock(&intel_pstate_limits_lock);
	1270	+
1261	1271	mutex_unlock(&intel_pstate_driver_lock);
1262	1272
1263	1273	return count;
@@ -1874,13 +1884,11 @@ static int intel_pstate_prepare_request(struct cpudata *cpu, int pstate) · 610 unmodified lines
1874	1884
1875	1885	intel_pstate_get_min_max(cpu, &min_perf, &max_perf);
1876	1886	pstate = clamp_t(int, pstate, min_perf, max_perf);
1877		− trace_cpu_frequency(pstate * cpu->pstate.scaling, cpu->cpu);
1878	1887	return pstate;
1879	1888	}
1880	1889
1881	1890	static void intel_pstate_update_pstate(struct cpudata *cpu, int pstate)
1882	1891	{
1883		− pstate = intel_pstate_prepare_request(cpu, pstate);
1884	1892	if (pstate == cpu->pstate.current_pstate)
1885	1893	return;
1886	1894
@@ -1900,6 +1908,8 @@ static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu) · 13 unmodified lines
1900	1908
1901	1909	update_turbo_state();
1902	1910
	1911	+ target_pstate = intel_pstate_prepare_request(cpu, target_pstate);
	1912	+ trace_cpu_frequency(target_pstate * cpu->pstate.scaling, cpu->cpu);
1903	1913	intel_pstate_update_pstate(cpu, target_pstate);
1904	1914
1905	1915	sample = &cpu->sample;
@@ -2132,16 +2142,11 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy) · 226 unmodified lines
2132	2142	mutex_lock(&intel_pstate_limits_lock);
2133	2143
2134	2144	if (policy->policy == CPUFREQ_POLICY_PERFORMANCE) {
	2145	+ pr_debug("set performance\n");
2135	2146	if (!perf_limits) {
2136	2147	limits = &performance_limits;
2137	2148	perf_limits = limits;
2138	2149	}
2139		− if (policy->max >= policy->cpuinfo.max_freq &&
2140		− !limits->no_turbo) {
2141		− pr_debug("set performance\n");
2142		− intel_pstate_set_performance_limits(perf_limits);
2143		− goto out;
2144		− }
2145	2150	} else {
2146	2151	pr_debug("set powersave\n");
2147	2152	if (!perf_limits) {
@@ -2152,7 +2157,7 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy) · 4 unmodified lines
2152	2157	}
2153	2158
2154	2159	intel_pstate_update_perf_limits(policy, perf_limits);
2155		− out:
	2160	+
2156	2161	if (cpu->policy == CPUFREQ_POLICY_PERFORMANCE) {
2157	2162	/*
2158	2163	* NOHZ_FULL CPUs need this as the governor callback may not
@@ -2198,9 +2203,9 @@ static int intel_pstate_verify_policy(struct cpufreq_policy *policy) · 39 unmodified lines
2198	2203	unsigned int max_freq, min_freq;
2199	2204
2200	2205	max_freq = policy->cpuinfo.max_freq *
2201		− limits->max_sysfs_pct / 100;
	2206	+ perf_limits->max_sysfs_pct / 100;
2202	2207	min_freq = policy->cpuinfo.max_freq *
2203		− limits->min_sysfs_pct / 100;
	2208	+ perf_limits->min_sysfs_pct / 100;
2204	2209	cpufreq_verify_within_limits(policy, min_freq, max_freq);
2205	2210	}
2206	2211
@@ -2243,13 +2248,8 @@ static int __intel_pstate_cpu_init(struct cpufreq_policy *policy) · 36 unmodified lines
2243	2248
2244	2249	cpu = all_cpu_data[policy->cpu];
2245	2250
2246		− /*
2247		− * We need sane value in the cpu->perf_limits, so inherit from global
2248		− * perf_limits limits, which are seeded with values based on the
2249		− * CONFIG_CPU_FREQ_DEFAULT_GOV_*, during boot up.
2250		− */
2251	2251	if (per_cpu_limits)
2252		− memcpy(cpu->perf_limits, limits, sizeof(struct perf_limits));
	2252	+ intel_pstate_init_limits(cpu->perf_limits);
2253	2253
2254	2254	policy->min = cpu->pstate.min_pstate * cpu->pstate.scaling;
2255	2255	policy->max = cpu->pstate.turbo_pstate * cpu->pstate.scaling;
@@ -2301,7 +2301,6 @@ static struct cpufreq_driver intel_pstate = { · 45 unmodified lines
2301	2301	static int intel_cpufreq_verify_policy(struct cpufreq_policy *policy)
2302	2302	{
2303	2303	struct cpudata *cpu = all_cpu_data[policy->cpu];
2304		− struct perf_limits *perf_limits = limits;
2305	2304
2306	2305	update_turbo_state();
2307	2306	policy->cpuinfo.max_freq = limits->turbo_disabled ?
@@ -2309,15 +2308,6 @@ static int intel_cpufreq_verify_policy(struct cpufreq_policy *policy) · 1 unmodified line
2309	2308
2310	2309	cpufreq_verify_within_cpu_limits(policy);
2311	2310
2312		− if (per_cpu_limits)
2313		− perf_limits = cpu->perf_limits;
2314		−
2315		− mutex_lock(&intel_pstate_limits_lock);
2316		−
2317		− intel_pstate_update_perf_limits(policy, perf_limits);
2318		−
2319		− mutex_unlock(&intel_pstate_limits_lock);
2320		−
2321	2311	return 0;
2322	2312	}
2323	2313
@@ -2370,6 +2360,7 @@ static int intel_cpufreq_target(struct cpufreq_policy *policy, · 46 unmodified lines
2370	2360	wrmsrl_on_cpu(policy->cpu, MSR_IA32_PERF_CTL,
2371	2361	pstate_funcs.get_val(cpu, target_pstate));
2372	2362	}
	2363	+ freqs.new = target_pstate * cpu->pstate.scaling;
2373	2364	cpufreq_freq_transition_end(policy, &freqs, false);
2374	2365
2375	2366	return 0;
@@ -2383,8 +2374,9 @@ static unsigned int intel_cpufreq_fast_switch(struct cpufreq_policy *policy, · 7 unmodified lines
2383	2374
2384	2375	target_freq = intel_cpufreq_turbo_update(cpu, policy, target_freq);
2385	2376	target_pstate = DIV_ROUND_UP(target_freq, cpu->pstate.scaling);
	2377	+ target_pstate = intel_pstate_prepare_request(cpu, target_pstate);
2386	2378	intel_pstate_update_pstate(cpu, target_pstate);
2387		− return target_freq;
	2379	+ return target_pstate * cpu->pstate.scaling;
2388	2380	}
2389	2381
2390	2382	static int intel_cpufreq_cpu_init(struct cpufreq_policy *policy)
@@ -2437,8 +2429,11 @@ static int intel_pstate_register_driver(void) · 46 unmodified lines
2437	2429
2438	2430	intel_pstate_init_limits(&powersave_limits);
2439	2431	intel_pstate_set_performance_limits(&performance_limits);
2440		− limits = IS_ENABLED(CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE) ?
2441		− &performance_limits : &powersave_limits;
	2432	+ if (IS_ENABLED(CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE) &&
	2433	+ intel_pstate_driver == &intel_pstate)
	2434	+ limits = &performance_limits;
	2435	+ else
	2436	+ limits = &powersave_limits;
2442	2437
2443	2438	ret = cpufreq_register_driver(intel_pstate_driver);
2444	2439	if (ret) {

drivers/crypto/s5p-sss.c+85−47

@@ -270,7 +270,7 @@ static void s5p_sg_copy_buf(void buf, struct scatterlist sg, · 269 unmodified lines
270	270	scatterwalk_done(&walk, out, 0);
271	271	}
272	272
273		−static void s5p_aes_complete(struct s5p_aes_dev *dev, int err)
	273	+static void s5p_sg_done(struct s5p_aes_dev *dev)
274	274	{
275	275	if (dev->sg_dst_cpy) {
276	276	dev_dbg(dev->dev,
@@ -281,8 +281,11 @@ static void s5p_aes_complete(struct s5p_aes_dev *dev, int err) · 4 unmodified lines
281	281	}
282	282	s5p_free_sg_cpy(dev, &dev->sg_src_cpy);
283	283	s5p_free_sg_cpy(dev, &dev->sg_dst_cpy);
	284	+}
284	285
285		− /* holding a lock outside */
	286	+/* Calls the completion. Cannot be called with dev->lock hold. */
	287	+static void s5p_aes_complete(struct s5p_aes_dev *dev, int err)
	288	+{
286	289	dev->req->base.complete(&dev->req->base, err);
287	290	dev->busy = false;
288	291	}
@@ -368,51 +371,44 @@ exit: · 79 unmodified lines
368	371	}
369	372
370	373	/*
371		− * Returns true if new transmitting (output) data is ready and its
372		− * address+length have to be written to device (by calling
373		− * s5p_set_dma_outdata()). False otherwise.
	374	+ * Returns -ERRNO on error (mapping of new data failed).
	375	+ * On success returns:
	376	+ * - 0 if there is no more data,
	377	+ * - 1 if new transmitting (output) data is ready and its address+length
	378	+ * have to be written to device (by calling s5p_set_dma_outdata()).
374	379	*/
375		−static bool s5p_aes_tx(struct s5p_aes_dev *dev)
	380	+static int s5p_aes_tx(struct s5p_aes_dev *dev)
376	381	{
377		− int err = 0;
378		− bool ret = false;
	382	+ int ret = 0;
379	383
380	384	s5p_unset_outdata(dev);
381	385
382	386	if (!sg_is_last(dev->sg_dst)) {
383		− err = s5p_set_outdata(dev, sg_next(dev->sg_dst));
384		− if (err)
385		− s5p_aes_complete(dev, err);
386		− else
387		− ret = true;
388		− } else {
389		− s5p_aes_complete(dev, err);
390		−
391		− dev->busy = true;
392		− tasklet_schedule(&dev->tasklet);
	387	+ ret = s5p_set_outdata(dev, sg_next(dev->sg_dst));
	388	+ if (!ret)
	389	+ ret = 1;
393	390	}
394	391
395	392	return ret;
396	393	}
397	394
398	395	/*
399		− * Returns true if new receiving (input) data is ready and its
400		− * address+length have to be written to device (by calling
401		− * s5p_set_dma_indata()). False otherwise.
	396	+ * Returns -ERRNO on error (mapping of new data failed).
	397	+ * On success returns:
	398	+ * - 0 if there is no more data,
	399	+ * - 1 if new receiving (input) data is ready and its address+length
	400	+ * have to be written to device (by calling s5p_set_dma_indata()).
402	401	*/
403		−static bool s5p_aes_rx(struct s5p_aes_dev *dev)
	402	+static int s5p_aes_rx(struct s5p_aes_dev dev/, bool set_dma/)
404	403	{
405		− int err;
406		− bool ret = false;
	404	+ int ret = 0;
407	405
408	406	s5p_unset_indata(dev);
409	407
410	408	if (!sg_is_last(dev->sg_src)) {
411		− err = s5p_set_indata(dev, sg_next(dev->sg_src));
412		− if (err)
413		− s5p_aes_complete(dev, err);
414		− else
415		− ret = true;
	409	+ ret = s5p_set_indata(dev, sg_next(dev->sg_src));
	410	+ if (!ret)
	411	+ ret = 1;
416	412	}
417	413
418	414	return ret;
@@ -422,33 +418,73 @@ static irqreturn_t s5p_aes_interrupt(int irq, void *dev_id) · 3 unmodified lines
422	418	{
423	419	struct platform_device *pdev = dev_id;
424	420	struct s5p_aes_dev *dev = platform_get_drvdata(pdev);
425		− bool set_dma_tx = false;
426		− bool set_dma_rx = false;
	421	+ int err_dma_tx = 0;
	422	+ int err_dma_rx = 0;
	423	+ bool tx_end = false;
427	424	unsigned long flags;
428	425	uint32_t status;
	426	+ int err;
429	427
430	428	spin_lock_irqsave(&dev->lock, flags);
431	429
	430	+ /*
	431	+ * Handle rx or tx interrupt. If there is still data (scatterlist did not
	432	+ * reach end), then map next scatterlist entry.
	433	+ * In case of such mapping error, s5p_aes_complete() should be called.
	434	+ *
	435	+ * If there is no more data in tx scatter list, call s5p_aes_complete()
	436	+ * and schedule new tasklet.
	437	+ */
432	438	status = SSS_READ(dev, FCINTSTAT);
433	439	if (status & SSS_FCINTSTAT_BRDMAINT)
434		− set_dma_rx = s5p_aes_rx(dev);
435		− if (status & SSS_FCINTSTAT_BTDMAINT)
436		− set_dma_tx = s5p_aes_tx(dev);
	440	+ err_dma_rx = s5p_aes_rx(dev);
	441	+
	442	+ if (status & SSS_FCINTSTAT_BTDMAINT) {
	443	+ if (sg_is_last(dev->sg_dst))
	444	+ tx_end = true;
	445	+ err_dma_tx = s5p_aes_tx(dev);
	446	+ }
437	447
438	448	SSS_WRITE(dev, FCINTPEND, status);
439	449
440		− /*
441		− * Writing length of DMA block (either receiving or transmitting)
442		− * will start the operation immediately, so this should be done
443		− * at the end (even after clearing pending interrupts to not miss the
444		− * interrupt).
445		− */
446		− if (set_dma_tx)
447		− s5p_set_dma_outdata(dev, dev->sg_dst);
448		− if (set_dma_rx)
449		− s5p_set_dma_indata(dev, dev->sg_src);
	450	+ if (err_dma_rx < 0) {
	451	+ err = err_dma_rx;
	452	+ goto error;
	453	+ }
	454	+ if (err_dma_tx < 0) {
	455	+ err = err_dma_tx;
	456	+ goto error;
	457	+ }
	458	+
	459	+ if (tx_end) {
	460	+ s5p_sg_done(dev);
	461	+
	462	+ spin_unlock_irqrestore(&dev->lock, flags);
	463	+
	464	+ s5p_aes_complete(dev, 0);
	465	+ dev->busy = true;
	466	+ tasklet_schedule(&dev->tasklet);
	467	+ } else {
	468	+ /*
	469	+ * Writing length of DMA block (either receiving or
	470	+ * transmitting) will start the operation immediately, so this
	471	+ * should be done at the end (even after clearing pending
	472	+ * interrupts to not miss the interrupt).
	473	+ */
	474	+ if (err_dma_tx == 1)
	475	+ s5p_set_dma_outdata(dev, dev->sg_dst);
	476	+ if (err_dma_rx == 1)
	477	+ s5p_set_dma_indata(dev, dev->sg_src);
450	478
	479	+ spin_unlock_irqrestore(&dev->lock, flags);
	480	+ }
	481	+
	482	+ return IRQ_HANDLED;
	483	+
	484	+error:
	485	+ s5p_sg_done(dev);
451	486	spin_unlock_irqrestore(&dev->lock, flags);
	487	+ s5p_aes_complete(dev, err);
452	488
453	489	return IRQ_HANDLED;
454	490	}
@@ -597,8 +633,9 @@ outdata_error: · 142 unmodified lines
597	633	s5p_unset_indata(dev);
598	634
599	635	indata_error:
600		− s5p_aes_complete(dev, err);
	636	+ s5p_sg_done(dev);
601	637	spin_unlock_irqrestore(&dev->lock, flags);
	638	+ s5p_aes_complete(dev, err);
602	639	}
603	640
604	641	static void s5p_tasklet_cb(unsigned long data)
@@ -805,8 +842,9 @@ static int s5p_aes_probe(struct platform_device *pdev) · 200 unmodified lines
805	842	dev_warn(dev, "feed control interrupt is not available.\n");
806	843	goto err_irq;
807	844	}
808		− err = devm_request_irq(dev, pdata->irq_fc, s5p_aes_interrupt,
809		− IRQF_SHARED, pdev->name, pdev);
	845	+ err = devm_request_threaded_irq(dev, pdata->irq_fc, NULL,
	846	+ s5p_aes_interrupt, IRQF_ONESHOT,
	847	+ pdev->name, pdev);
810	848	if (err < 0) {
811	849	dev_warn(dev, "feed control interrupt is not available.\n");
812	850	goto err_irq;

drivers/crypto/ux500/cryp/cryp.c+1−1

@@ -82,7 +82,7 @@ void cryp_activity(struct cryp_device_data *device_data, · 81 unmodified lines
82	82	void cryp_flush_inoutfifo(struct cryp_device_data *device_data)
83	83	{
84	84	/*
85		− * We always need to disble the hardware before trying to flush the
	85	+ * We always need to disable the hardware before trying to flush the
86	86	* FIFO. This is something that isn't written in the design
87	87	* specification, but we have been informed by the hardware designers
88	88	* that this must be done.

drivers/firmware/efi/arm-runtime.c+1−0

@@ -65,6 +65,7 @@ static bool __init efi_virtmap_init(void) · 64 unmodified lines
65	65	bool systab_found;
66	66
67	67	efi_mm.pgd = pgd_alloc(&efi_mm);
	68	+ mm_init_cpumask(&efi_mm);
68	69	init_new_context(NULL, &efi_mm);
69	70
70	71	systab_found = false;

drivers/firmware/efi/libstub/secureboot.c+2−2

@@ -45,6 +45,8 @@ enum efi_secureboot_mode efi_get_secureboot(efi_system_table_t *sys_table_arg) · 44 unmodified lines
45	45	size = sizeof(secboot);
46	46	status = get_efi_var(efi_SecureBoot_name, &efi_variable_guid,
47	47	NULL, &size, &secboot);
	48	+ if (status == EFI_NOT_FOUND)
	49	+ return efi_secureboot_mode_disabled;
48	50	if (status != EFI_SUCCESS)
49	51	goto out_efi_err;
50	52
@@ -78,7 +80,5 @@ secure_boot_enabled: · 27 unmodified lines
78	80
79	81	out_efi_err:
80	82	pr_efi_err(sys_table_arg, "Could not determine UEFI Secure Boot status.\n");
81		− if (status == EFI_NOT_FOUND)
82		− return efi_secureboot_mode_disabled;
83	83	return efi_secureboot_mode_unknown;
84	84	}

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c+4−1

@@ -2094,8 +2094,11 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) · 2093 unmodified lines
2094	2094	}
2095	2095
2096	2096	r = amdgpu_late_init(adev);
2097		− if (r)
	2097	+ if (r) {
	2098	+ if (fbcon)
	2099	+ console_unlock();
2098	2100	return r;
	2101	+ }
2099	2102
2100	2103	/* pin cursors */
2101	2104	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c+2−1

@@ -59,9 +59,10 @@ · 58 unmodified lines
59	59	* - 3.7.0 - Add support for VCE clock list packet
60	60	* - 3.8.0 - Add support raster config init in the kernel
61	61	* - 3.9.0 - Add support for memory query info about VRAM and GTT.
	62	+ * - 3.10.0 - Add support for new fences ioctl, new gem ioctl flags
62	63	*/
63	64	#define KMS_DRIVER_MAJOR 3
64		−#define KMS_DRIVER_MINOR 9
	65	+#define KMS_DRIVER_MINOR 10
65	66	#define KMS_DRIVER_PATCHLEVEL 0
66	67
67	68	int amdgpu_vram_limit = 0;

drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c+21−0

@@ -202,6 +202,27 @@ int amdgpu_gem_create_ioctl(struct drm_device dev, void data, · 201 unmodified lines
202	202	bool kernel = false;
203	203	int r;
204	204
	205	+ /* reject invalid gem flags */
	206	+ if (args->in.domain_flags & ~(AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED \|
	207	+ AMDGPU_GEM_CREATE_NO_CPU_ACCESS \|
	208	+ AMDGPU_GEM_CREATE_CPU_GTT_USWC \|
	209	+ AMDGPU_GEM_CREATE_VRAM_CLEARED\|
	210	+ AMDGPU_GEM_CREATE_SHADOW \|
	211	+ AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)) {
	212	+ r = -EINVAL;
	213	+ goto error_unlock;
	214	+ }
	215	+ /* reject invalid gem domains */
	216	+ if (args->in.domains & ~(AMDGPU_GEM_DOMAIN_CPU \|
	217	+ AMDGPU_GEM_DOMAIN_GTT \|
	218	+ AMDGPU_GEM_DOMAIN_VRAM \|
	219	+ AMDGPU_GEM_DOMAIN_GDS \|
	220	+ AMDGPU_GEM_DOMAIN_GWS \|
	221	+ AMDGPU_GEM_DOMAIN_OA)) {
	222	+ r = -EINVAL;
	223	+ goto error_unlock;
	224	+ }
	225	+
205	226	/* create a gem object to contain this object in */
206	227	if (args->in.domains & (AMDGPU_GEM_DOMAIN_GDS \|
207	228	AMDGPU_GEM_DOMAIN_GWS \| AMDGPU_GEM_DOMAIN_OA)) {

drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c+1−1

@@ -788,7 +788,7 @@ static int sdma_v3_0_start(struct amdgpu_device *adev) · 787 unmodified lines
788	788	}
789	789	}
790	790
791		− /* disble sdma engine before programing it */
	791	+ /* disable sdma engine before programing it */
792	792	sdma_v3_0_ctx_switch_enable(adev, false);
793	793	sdma_v3_0_enable(adev, false);
794	794

drivers/gpu/drm/drm_edid.c+3−0

@@ -148,6 +148,9 @@ static const struct edid_quirk { · 147 unmodified lines
148	148
149	149	/* Panel in Samsung NP700G7A-S01PL notebook reports 6bpc */
150	150	{ "SEC", 0xd033, EDID_QUIRK_FORCE_8BPC },
	151	+
	152	+ /* Rotel RSX-1058 forwards sink's EDID but only does HDMI 1.1*/
	153	+ { "ETR", 13896, EDID_QUIRK_FORCE_8BPC },
151	154	};
152	155
153	156	/*

drivers/gpu/drm/i915/gvt/cfg_space.c+54−3

@@ -41,6 +41,54 @@ enum { · 40 unmodified lines
41	41	INTEL_GVT_PCI_BAR_MAX,
42	42	};
43	43
	44	+/* bitmap for writable bits (RW or RW1C bits, but cannot co-exist in one
	45	+ * byte) byte by byte in standard pci configuration space. (not the full
	46	+ * 256 bytes.)
	47	+ */
	48	+static const u8 pci_cfg_space_rw_bmp[PCI_INTERRUPT_LINE + 4] = {
	49	+ [PCI_COMMAND] = 0xff, 0x07,
	50	+ [PCI_STATUS] = 0x00, 0xf9, /* the only one RW1C byte */
	51	+ [PCI_CACHE_LINE_SIZE] = 0xff,
	52	+ [PCI_BASE_ADDRESS_0 ... PCI_CARDBUS_CIS - 1] = 0xff,
	53	+ [PCI_ROM_ADDRESS] = 0x01, 0xf8, 0xff, 0xff,
	54	+ [PCI_INTERRUPT_LINE] = 0xff,
	55	+};
	56	+
	57	+/**
	58	+ * vgpu_pci_cfg_mem_write - write virtual cfg space memory
	59	+ *
	60	+ * Use this function to write virtual cfg space memory.
	61	+ * For standard cfg space, only RW bits can be changed,
	62	+ * and we emulates the RW1C behavior of PCI_STATUS register.
	63	+ */
	64	+static void vgpu_pci_cfg_mem_write(struct intel_vgpu *vgpu, unsigned int off,
	65	+ u8 *src, unsigned int bytes)
	66	+{
	67	+ u8 *cfg_base = vgpu_cfg_space(vgpu);
	68	+ u8 mask, new, old;
	69	+ int i = 0;
	70	+
	71	+ for (; i < bytes && (off + i < sizeof(pci_cfg_space_rw_bmp)); i++) {
	72	+ mask = pci_cfg_space_rw_bmp[off + i];
	73	+ old = cfg_base[off + i];
	74	+ new = src[i] & mask;
	75	+
	76	+ /**
	77	+ * The PCI_STATUS high byte has RW1C bits, here
	78	+ * emulates clear by writing 1 for these bits.
	79	+ * Writing a 0b to RW1C bits has no effect.
	80	+ */
	81	+ if (off + i == PCI_STATUS + 1)
	82	+ new = (~new & old) & mask;
	83	+
	84	+ cfg_base[off + i] = (old & ~mask) \| new;
	85	+ }
	86	+
	87	+ /* For other configuration space directly copy as it is. */
	88	+ if (i < bytes)
	89	+ memcpy(cfg_base + off + i, src + i, bytes - i);
	90	+}
	91	+
44	92	/**
45	93	* intel_vgpu_emulate_cfg_read - emulate vGPU configuration space read
46	94	*
@@ -123,7 +171,7 @@ static int emulate_pci_command_write(struct intel_vgpu *vgpu, · 76 unmodified lines
123	171	u8 changed = old ^ new;
124	172	int ret;
125	173
126		− memcpy(vgpu_cfg_space(vgpu) + offset, p_data, bytes);
	174	+ vgpu_pci_cfg_mem_write(vgpu, offset, p_data, bytes);
127	175	if (!(changed & PCI_COMMAND_MEMORY))
128	176	return 0;
129	177
@@ -237,6 +285,9 @@ int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset, · 107 unmodified lines
237	285	{
238	286	int ret;
239	287
	288	+ if (vgpu->failsafe)
	289	+ return 0;
	290	+
240	291	if (WARN_ON(bytes > 4))
241	292	return -EINVAL;
242	293
@@ -274,10 +325,10 @@ int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset, · 31 unmodified lines
274	325	if (ret)
275	326	return ret;
276	327
277		− memcpy(vgpu_cfg_space(vgpu) + offset, p_data, bytes);
	328	+ vgpu_pci_cfg_mem_write(vgpu, offset, p_data, bytes);
278	329	break;
279	330	default:
280		− memcpy(vgpu_cfg_space(vgpu) + offset, p_data, bytes);
	331	+ vgpu_pci_cfg_mem_write(vgpu, offset, p_data, bytes);
281	332	break;
282	333	}
283	334	return 0;

drivers/gpu/drm/i915/gvt/cmd_parser.c+5−5

@@ -668,7 +668,7 @@ static inline void print_opcode(u32 cmd, int ring_id) · 667 unmodified lines
668	668	if (d_info == NULL)
669	669	return;
670	670
671		− gvt_err("opcode=0x%x %s sub_ops:",
	671	+ gvt_dbg_cmd("opcode=0x%x %s sub_ops:",
672	672	cmd >> (32 - d_info->op_len), d_info->name);
673	673
674	674	for (i = 0; i < d_info->nr_sub_op; i++)
@@ -693,23 +693,23 @@ static void parser_exec_state_dump(struct parser_exec_state *s) · 18 unmodified lines
693	693	int cnt = 0;
694	694	int i;
695	695
696		− gvt_err(" vgpu%d RING%d: ring_start(%08lx) ring_end(%08lx)"
	696	+ gvt_dbg_cmd(" vgpu%d RING%d: ring_start(%08lx) ring_end(%08lx)"
697	697	" ring_head(%08lx) ring_tail(%08lx)\n", s->vgpu->id,
698	698	s->ring_id, s->ring_start, s->ring_start + s->ring_size,
699	699	s->ring_head, s->ring_tail);
700	700
701		− gvt_err(" %s %s ip_gma(%08lx) ",
	701	+ gvt_dbg_cmd(" %s %s ip_gma(%08lx) ",
702	702	s->buf_type == RING_BUFFER_INSTRUCTION ?
703	703	"RING_BUFFER" : "BATCH_BUFFER",
704	704	s->buf_addr_type == GTT_BUFFER ?
705	705	"GTT" : "PPGTT", s->ip_gma);
706	706
707	707	if (s->ip_va == NULL) {
708		− gvt_err(" ip_va(NULL)");
	708	+ gvt_dbg_cmd(" ip_va(NULL)");
709	709	return;
710	710	}
711	711
712		− gvt_err(" ip_va=%p: %08x %08x %08x %08x\n",
	712	+ gvt_dbg_cmd(" ip_va=%p: %08x %08x %08x %08x\n",
713	713	s->ip_va, cmd_val(s, 0), cmd_val(s, 1),
714	714	cmd_val(s, 2), cmd_val(s, 3));
715	715

drivers/gpu/drm/i915/gvt/display.c+94−45

@@ -83,44 +83,80 @@ static int pipe_is_enabled(struct intel_vgpu *vgpu, int pipe) · 82 unmodified lines
83	83	return 0;
84	84	}
85	85
	86	+static unsigned char virtual_dp_monitor_edid[GVT_EDID_NUM][EDID_SIZE] = {
	87	+ {
	88	+/* EDID with 1024x768 as its resolution */
	89	+ /Header/
	90	+ 0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00,
	91	+ /* Vendor & Product Identification */
	92	+ 0x22, 0xf0, 0x54, 0x29, 0x00, 0x00, 0x00, 0x00, 0x04, 0x17,
	93	+ /* Version & Revision */
	94	+ 0x01, 0x04,
	95	+ /* Basic Display Parameters & Features */
	96	+ 0xa5, 0x34, 0x20, 0x78, 0x23,
	97	+ /* Color Characteristics */
	98	+ 0xfc, 0x81, 0xa4, 0x55, 0x4d, 0x9d, 0x25, 0x12, 0x50, 0x54,
	99	+ /* Established Timings: maximum resolution is 1024x768 */
	100	+ 0x21, 0x08, 0x00,
	101	+ /* Standard Timings. All invalid */
	102	+ 0x00, 0xc0, 0x00, 0xc0, 0x00, 0x40, 0x00, 0x80, 0x00, 0x00,
	103	+ 0x00, 0x40, 0x00, 0x00, 0x00, 0x01,
	104	+ /* 18 Byte Data Blocks 1: invalid */
	105	+ 0x00, 0x00, 0x80, 0xa0, 0x70, 0xb0,
	106	+ 0x23, 0x40, 0x30, 0x20, 0x36, 0x00, 0x06, 0x44, 0x21, 0x00, 0x00, 0x1a,
	107	+ /* 18 Byte Data Blocks 2: invalid */
	108	+ 0x00, 0x00, 0x00, 0xfd, 0x00, 0x18, 0x3c, 0x18, 0x50, 0x11, 0x00, 0x0a,
	109	+ 0x20, 0x20, 0x20, 0x20, 0x20, 0x20,
	110	+ /* 18 Byte Data Blocks 3: invalid */
	111	+ 0x00, 0x00, 0x00, 0xfc, 0x00, 0x48,
	112	+ 0x50, 0x20, 0x5a, 0x52, 0x32, 0x34, 0x34, 0x30, 0x77, 0x0a, 0x20, 0x20,
	113	+ /* 18 Byte Data Blocks 4: invalid */
	114	+ 0x00, 0x00, 0x00, 0xff, 0x00, 0x43, 0x4e, 0x34, 0x33, 0x30, 0x34, 0x30,
	115	+ 0x44, 0x58, 0x51, 0x0a, 0x20, 0x20,
	116	+ /* Extension Block Count */
	117	+ 0x00,
	118	+ /* Checksum */
	119	+ 0xef,
	120	+ },
	121	+ {
86	122	/* EDID with 1920x1200 as its resolution */
87		−static unsigned char virtual_dp_monitor_edid[] = {
88		− /Header/
89		− 0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00,
90		− /* Vendor & Product Identification */
91		− 0x22, 0xf0, 0x54, 0x29, 0x00, 0x00, 0x00, 0x00, 0x04, 0x17,
92		− /* Version & Revision */
93		− 0x01, 0x04,
94		− /* Basic Display Parameters & Features */
95		− 0xa5, 0x34, 0x20, 0x78, 0x23,
96		− /* Color Characteristics */
97		− 0xfc, 0x81, 0xa4, 0x55, 0x4d, 0x9d, 0x25, 0x12, 0x50, 0x54,
98		− /* Established Timings: maximum resolution is 1024x768 */
99		− 0x21, 0x08, 0x00,
100		− /*
101		− * Standard Timings.
102		− * below new resolutions can be supported:
103		− * 1920x1080, 1280x720, 1280x960, 1280x1024,
104		− * 1440x900, 1600x1200, 1680x1050
105		− */
106		− 0xd1, 0xc0, 0x81, 0xc0, 0x81, 0x40, 0x81, 0x80, 0x95, 0x00,
107		− 0xa9, 0x40, 0xb3, 0x00, 0x01, 0x01,
108		− /* 18 Byte Data Blocks 1: max resolution is 1920x1200 */
109		− 0x28, 0x3c, 0x80, 0xa0, 0x70, 0xb0,
110		− 0x23, 0x40, 0x30, 0x20, 0x36, 0x00, 0x06, 0x44, 0x21, 0x00, 0x00, 0x1a,
111		− /* 18 Byte Data Blocks 2: invalid */
112		− 0x00, 0x00, 0x00, 0xfd, 0x00, 0x18, 0x3c, 0x18, 0x50, 0x11, 0x00, 0x0a,
113		− 0x20, 0x20, 0x20, 0x20, 0x20, 0x20,
114		− /* 18 Byte Data Blocks 3: invalid */
115		− 0x00, 0x00, 0x00, 0xfc, 0x00, 0x48,
116		− 0x50, 0x20, 0x5a, 0x52, 0x32, 0x34, 0x34, 0x30, 0x77, 0x0a, 0x20, 0x20,
117		− /* 18 Byte Data Blocks 4: invalid */
118		− 0x00, 0x00, 0x00, 0xff, 0x00, 0x43, 0x4e, 0x34, 0x33, 0x30, 0x34, 0x30,
119		− 0x44, 0x58, 0x51, 0x0a, 0x20, 0x20,
120		− /* Extension Block Count */
121		− 0x00,
122		− /* Checksum */
123		− 0x45,
	123	+ /Header/
	124	+ 0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00,
	125	+ /* Vendor & Product Identification */
	126	+ 0x22, 0xf0, 0x54, 0x29, 0x00, 0x00, 0x00, 0x00, 0x04, 0x17,
	127	+ /* Version & Revision */
	128	+ 0x01, 0x04,
	129	+ /* Basic Display Parameters & Features */
	130	+ 0xa5, 0x34, 0x20, 0x78, 0x23,
	131	+ /* Color Characteristics */
	132	+ 0xfc, 0x81, 0xa4, 0x55, 0x4d, 0x9d, 0x25, 0x12, 0x50, 0x54,
	133	+ /* Established Timings: maximum resolution is 1024x768 */
	134	+ 0x21, 0x08, 0x00,
	135	+ /*
	136	+ * Standard Timings.
	137	+ * below new resolutions can be supported:
	138	+ * 1920x1080, 1280x720, 1280x960, 1280x1024,
	139	+ * 1440x900, 1600x1200, 1680x1050
	140	+ */
	141	+ 0xd1, 0xc0, 0x81, 0xc0, 0x81, 0x40, 0x81, 0x80, 0x95, 0x00,
	142	+ 0xa9, 0x40, 0xb3, 0x00, 0x01, 0x01,
	143	+ /* 18 Byte Data Blocks 1: max resolution is 1920x1200 */
	144	+ 0x28, 0x3c, 0x80, 0xa0, 0x70, 0xb0,
	145	+ 0x23, 0x40, 0x30, 0x20, 0x36, 0x00, 0x06, 0x44, 0x21, 0x00, 0x00, 0x1a,
	146	+ /* 18 Byte Data Blocks 2: invalid */
	147	+ 0x00, 0x00, 0x00, 0xfd, 0x00, 0x18, 0x3c, 0x18, 0x50, 0x11, 0x00, 0x0a,
	148	+ 0x20, 0x20, 0x20, 0x20, 0x20, 0x20,
	149	+ /* 18 Byte Data Blocks 3: invalid */
	150	+ 0x00, 0x00, 0x00, 0xfc, 0x00, 0x48,
	151	+ 0x50, 0x20, 0x5a, 0x52, 0x32, 0x34, 0x34, 0x30, 0x77, 0x0a, 0x20, 0x20,
	152	+ /* 18 Byte Data Blocks 4: invalid */
	153	+ 0x00, 0x00, 0x00, 0xff, 0x00, 0x43, 0x4e, 0x34, 0x33, 0x30, 0x34, 0x30,
	154	+ 0x44, 0x58, 0x51, 0x0a, 0x20, 0x20,
	155	+ /* Extension Block Count */
	156	+ 0x00,
	157	+ /* Checksum */
	158	+ 0x45,
	159	+ },
124	160	};
125	161
126	162	#define DPCD_HEADER_SIZE 0xb
@@ -140,14 +176,20 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) · 13 unmodified lines
140	176	vgpu_vreg(vgpu, SDEISR) &= ~(SDE_PORTA_HOTPLUG_SPT \|
141	177	SDE_PORTE_HOTPLUG_SPT);
142	178
143		− if (intel_vgpu_has_monitor_on_port(vgpu, PORT_B))
	179	+ if (intel_vgpu_has_monitor_on_port(vgpu, PORT_B)) {
144	180	vgpu_vreg(vgpu, SDEISR) \|= SDE_PORTB_HOTPLUG_CPT;
	181	+ vgpu_vreg(vgpu, SFUSE_STRAP) \|= SFUSE_STRAP_DDIB_DETECTED;
	182	+ }
145	183
146		− if (intel_vgpu_has_monitor_on_port(vgpu, PORT_C))
	184	+ if (intel_vgpu_has_monitor_on_port(vgpu, PORT_C)) {
147	185	vgpu_vreg(vgpu, SDEISR) \|= SDE_PORTC_HOTPLUG_CPT;
	186	+ vgpu_vreg(vgpu, SFUSE_STRAP) \|= SFUSE_STRAP_DDIC_DETECTED;
	187	+ }
148	188
149		− if (intel_vgpu_has_monitor_on_port(vgpu, PORT_D))
	189	+ if (intel_vgpu_has_monitor_on_port(vgpu, PORT_D)) {
150	190	vgpu_vreg(vgpu, SDEISR) \|= SDE_PORTD_HOTPLUG_CPT;
	191	+ vgpu_vreg(vgpu, SFUSE_STRAP) \|= SFUSE_STRAP_DDID_DETECTED;
	192	+ }
151	193
152	194	if (IS_SKYLAKE(dev_priv) &&
153	195	intel_vgpu_has_monitor_on_port(vgpu, PORT_E)) {
@@ -160,6 +202,8 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) · 6 unmodified lines
160	202	GEN8_PORT_DP_A_HOTPLUG;
161	203	else
162	204	vgpu_vreg(vgpu, SDEISR) \|= SDE_PORTA_HOTPLUG_SPT;
	205	+
	206	+ vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_A)) \|= DDI_INIT_DISPLAY_DETECTED;
163	207	}
164	208	}
165	209
@@ -175,10 +219,13 @@ static void clean_virtual_dp_monitor(struct intel_vgpu *vgpu, int port_num) · 9 unmodified lines
175	219	}
176	220
177	221	static int setup_virtual_dp_monitor(struct intel_vgpu *vgpu, int port_num,
178		− int type)
	222	+ int type, unsigned int resolution)
179	223	{
180	224	struct intel_vgpu_port *port = intel_vgpu_port(vgpu, port_num);
181	225
	226	+ if (WARN_ON(resolution >= GVT_EDID_NUM))
	227	+ return -EINVAL;
	228	+
182	229	port->edid = kzalloc(sizeof(*(port->edid)), GFP_KERNEL);
183	230	if (!port->edid)
184	231	return -ENOMEM;
@@ -189,7 +236,7 @@ static int setup_virtual_dp_monitor(struct intel_vgpu *vgpu, int port_num, · 4 unmodified lines
189	236	return -ENOMEM;
190	237	}
191	238
192		− memcpy(port->edid->edid_block, virtual_dp_monitor_edid,
	239	+ memcpy(port->edid->edid_block, virtual_dp_monitor_edid[resolution],
193	240	EDID_SIZE);
194	241	port->edid->data_valid = true;
195	242
@@ -322,16 +369,18 @@ void intel_vgpu_clean_display(struct intel_vgpu *vgpu) · 126 unmodified lines
322	369	* Zero on success, negative error code if failed.
323	370	*
324	371	*/
325		−int intel_vgpu_init_display(struct intel_vgpu *vgpu)
	372	+int intel_vgpu_init_display(struct intel_vgpu *vgpu, u64 resolution)
326	373	{
327	374	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
328	375
329	376	intel_vgpu_init_i2c_edid(vgpu);
330	377
331	378	if (IS_SKYLAKE(dev_priv))
332		− return setup_virtual_dp_monitor(vgpu, PORT_D, GVT_DP_D);
	379	+ return setup_virtual_dp_monitor(vgpu, PORT_D, GVT_DP_D,
	380	+ resolution);
333	381	else
334		− return setup_virtual_dp_monitor(vgpu, PORT_B, GVT_DP_B);
	382	+ return setup_virtual_dp_monitor(vgpu, PORT_B, GVT_DP_B,
	383	+ resolution);
335	384	}
336	385
337	386	/**

drivers/gpu/drm/i915/gvt/display.h+19−1

@@ -154,10 +154,28 @@ struct intel_vgpu_port { · 153 unmodified lines
154	154	int type;
155	155	};
156	156
	157	+enum intel_vgpu_edid {
	158	+ GVT_EDID_1024_768,
	159	+ GVT_EDID_1920_1200,
	160	+ GVT_EDID_NUM,
	161	+};
	162	+
	163	+static inline char *vgpu_edid_str(enum intel_vgpu_edid id)
	164	+{
	165	+ switch (id) {
	166	+ case GVT_EDID_1024_768:
	167	+ return "1024x768";
	168	+ case GVT_EDID_1920_1200:
	169	+ return "1920x1200";
	170	+ default:
	171	+ return "";
	172	+ }
	173	+}
	174	+
157	175	void intel_gvt_emulate_vblank(struct intel_gvt *gvt);
158	176	void intel_gvt_check_vblank_emulation(struct intel_gvt *gvt);
159	177
160		−int intel_vgpu_init_display(struct intel_vgpu *vgpu);
	178	+int intel_vgpu_init_display(struct intel_vgpu *vgpu, u64 resolution);
161	179	void intel_vgpu_reset_display(struct intel_vgpu *vgpu);
162	180	void intel_vgpu_clean_display(struct intel_vgpu *vgpu);
163	181

drivers/gpu/drm/i915/gvt/firmware.c+1−1

@@ -80,7 +80,7 @@ static int expose_firmware_sysfs(struct intel_gvt *gvt) · 79 unmodified lines
80	80	int ret;
81	81
82	82	size = sizeof(*h) + info->mmio_size + info->cfg_space_size - 1;
83		− firmware = vmalloc(size);
	83	+ firmware = vzalloc(size);
84	84	if (!firmware)
85	85	return -ENOMEM;
86	86

drivers/gpu/drm/i915/gvt/gtt.c+26−14

@@ -1825,11 +1825,8 @@ static int emulate_gtt_mmio_write(struct intel_vgpu *vgpu, unsigned int off, · 1824 unmodified lines
1825	1825	gma = g_gtt_index << GTT_PAGE_SHIFT;
1826	1826
1827	1827	/* the VM may configure the whole GM space when ballooning is used */
1828		− if (WARN_ONCE(!vgpu_gmadr_is_valid(vgpu, gma),
1829		− "vgpu%d: found oob ggtt write, offset %x\n",
1830		− vgpu->id, off)) {
	1828	+ if (!vgpu_gmadr_is_valid(vgpu, gma))
1831	1829	return 0;
1832		− }
1833	1830
1834	1831	ggtt_get_guest_entry(ggtt_mm, &e, g_gtt_index);
1835	1832
@@ -2015,6 +2012,22 @@ int intel_vgpu_init_gtt(struct intel_vgpu *vgpu) · 179 unmodified lines
2015	2012	return create_scratch_page_tree(vgpu);
2016	2013	}
2017	2014
	2015	+static void intel_vgpu_free_mm(struct intel_vgpu *vgpu, int type)
	2016	+{
	2017	+ struct list_head pos, n;
	2018	+ struct intel_vgpu_mm *mm;
	2019	+
	2020	+ list_for_each_safe(pos, n, &vgpu->gtt.mm_list_head) {
	2021	+ mm = container_of(pos, struct intel_vgpu_mm, list);
	2022	+ if (mm->type == type) {
	2023	+ vgpu->gvt->gtt.mm_free_page_table(mm);
	2024	+ list_del(&mm->list);
	2025	+ list_del(&mm->lru_list);
	2026	+ kfree(mm);
	2027	+ }
	2028	+ }
	2029	+}
	2030	+
2018	2031	/**
2019	2032	* intel_vgpu_clean_gtt - clean up per-vGPU graphics memory virulization
2020	2033	* @vgpu: a vGPU
@@ -2027,19 +2040,11 @@ int intel_vgpu_init_gtt(struct intel_vgpu *vgpu) · 6 unmodified lines
2027	2040	*/
2028	2041	void intel_vgpu_clean_gtt(struct intel_vgpu *vgpu)
2029	2042	{
2030		− struct list_head pos, n;
2031		− struct intel_vgpu_mm *mm;
2032		−
2033	2043	ppgtt_free_all_shadow_page(vgpu);
2034	2044	release_scratch_page_tree(vgpu);
2035	2045
2036		− list_for_each_safe(pos, n, &vgpu->gtt.mm_list_head) {
2037		− mm = container_of(pos, struct intel_vgpu_mm, list);
2038		− vgpu->gvt->gtt.mm_free_page_table(mm);
2039		− list_del(&mm->list);
2040		− list_del(&mm->lru_list);
2041		− kfree(mm);
2042		− }
	2046	+ intel_vgpu_free_mm(vgpu, INTEL_GVT_MM_PPGTT);
	2047	+ intel_vgpu_free_mm(vgpu, INTEL_GVT_MM_GGTT);
2043	2048	}
2044	2049
2045	2050	static void clean_spt_oos(struct intel_gvt *gvt)
@@ -2322,6 +2327,13 @@ void intel_vgpu_reset_gtt(struct intel_vgpu *vgpu, bool dmlr) · 276 unmodified lines
2322	2327	int i;
2323	2328
2324	2329	ppgtt_free_all_shadow_page(vgpu);
	2330	+
	2331	+ /* Shadow pages are only created when there is no page
	2332	+ * table tracking data, so remove page tracking data after
	2333	+ * removing the shadow pages.
	2334	+ */
	2335	+ intel_vgpu_free_mm(vgpu, INTEL_GVT_MM_PPGTT);
	2336	+
2325	2337	if (!dmlr)
2326	2338	return;
2327	2339

drivers/gpu/drm/i915/gvt/gvt.h+10−2

@@ -143,6 +143,8 @@ struct intel_vgpu { · 142 unmodified lines
143	143	int id;
144	144	unsigned long handle; /* vGPU handle used by hypervisor MPT modules */
145	145	bool active;
	146	+ bool pv_notified;
	147	+ bool failsafe;
146	148	bool resetting;
147	149	void *sched_data;
148	150
@@ -203,18 +205,18 @@ struct intel_gvt_firmware { · 54 unmodified lines
203	205	};
204	206
205	207	struct intel_gvt_opregion {
206		− void __iomem *opregion_va;
	208	+ void *opregion_va;
207	209	u32 opregion_pa;
208	210	};
209	211
210	212	#define NR_MAX_INTEL_VGPU_TYPES 20
211	213	struct intel_vgpu_type {
212	214	char name[16];
213		− unsigned int max_instance;
214	215	unsigned int avail_instance;
215	216	unsigned int low_gm_size;
216	217	unsigned int high_gm_size;
217	218	unsigned int fence;
	219	+ enum intel_vgpu_edid resolution;
218	220	};
219	221
220	222	struct intel_gvt {
@@ -317,6 +319,7 @@ struct intel_vgpu_creation_params { · 96 unmodified lines
317	319	__u64 low_gm_sz; /* in MB */
318	320	__u64 high_gm_sz; /* in MB */
319	321	__u64 fence_sz;
	322	+ __u64 resolution;
320	323	__s32 primary;
321	324	__u64 vgpu_id;
322	325	};
@@ -449,6 +452,11 @@ struct intel_gvt_ops { · 126 unmodified lines
449	452	};
450	453
451	454
	455	+enum {
	456	+ GVT_FAILSAFE_UNSUPPORTED_GUEST,
	457	+ GVT_FAILSAFE_INSUFFICIENT_RESOURCE,
	458	+};
	459	+
452	460	#include "mpt.h"
453	461
454	462	#endif

drivers/gpu/drm/i915/gvt/handlers.c+307−132

@@ -121,6 +121,7 @@ static int new_mmio_info(struct intel_gvt *gvt, · 120 unmodified lines
121	121	info->size = size;
122	122	info->length = (i + 4) < end ? 4 : (end - i);
123	123	info->addr_mask = addr_mask;
	124	+ info->ro_mask = ro_mask;
124	125	info->device = device;
125	126	info->read = read ? read : intel_vgpu_default_mmio_read;
126	127	info->write = write ? write : intel_vgpu_default_mmio_write;
@@ -150,15 +151,44 @@ static int render_mmio_to_ring_id(struct intel_gvt *gvt, unsigned int reg) · 23 unmodified lines
150	151	#define fence_num_to_offset(num) \
151	152	(num * 8 + i915_mmio_reg_offset(FENCE_REG_GEN6_LO(0)))
152	153
	154	+
	155	+static void enter_failsafe_mode(struct intel_vgpu *vgpu, int reason)
	156	+{
	157	+ switch (reason) {
	158	+ case GVT_FAILSAFE_UNSUPPORTED_GUEST:
	159	+ pr_err("Detected your guest driver doesn't support GVT-g.\n");
	160	+ break;
	161	+ case GVT_FAILSAFE_INSUFFICIENT_RESOURCE:
	162	+ pr_err("Graphics resource is not enough for the guest\n");
	163	+ default:
	164	+ break;
	165	+ }
	166	+ pr_err("Now vgpu %d will enter failsafe mode.\n", vgpu->id);
	167	+ vgpu->failsafe = true;
	168	+}
	169	+
153	170	static int sanitize_fence_mmio_access(struct intel_vgpu *vgpu,
154	171	unsigned int fence_num, void *p_data, unsigned int bytes)
155	172	{
156	173	if (fence_num >= vgpu_fence_sz(vgpu)) {
157		− gvt_err("vgpu%d: found oob fence register access\n",
158		− vgpu->id);
159		− gvt_err("vgpu%d: total fence num %d access fence num %d\n",
160		− vgpu->id, vgpu_fence_sz(vgpu), fence_num);
	174	+
	175	+ /* When guest access oob fence regs without access
	176	+ * pv_info first, we treat guest not supporting GVT,
	177	+ * and we will let vgpu enter failsafe mode.
	178	+ */
	179	+ if (!vgpu->pv_notified)
	180	+ enter_failsafe_mode(vgpu,
	181	+ GVT_FAILSAFE_UNSUPPORTED_GUEST);
	182	+
	183	+ if (!vgpu->mmio.disable_warn_untrack) {
	184	+ gvt_err("vgpu%d: found oob fence register access\n",
	185	+ vgpu->id);
	186	+ gvt_err("vgpu%d: total fence %d, access fence %d\n",
	187	+ vgpu->id, vgpu_fence_sz(vgpu),
	188	+ fence_num);
	189	+ }
161	190	memset(p_data, 0, bytes);
	191	+ return -EINVAL;
162	192	}
163	193	return 0;
164	194	}
@@ -369,6 +399,74 @@ static int pipeconf_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, · 204 unmodified lines
369	399	return 0;
370	400	}
371	401
	402	+/* ascendingly sorted */
	403	+static i915_reg_t force_nonpriv_white_list[] = {
	404	+ GEN9_CS_DEBUG_MODE1, //_MMIO(0x20ec)
	405	+ GEN9_CTX_PREEMPT_REG,//_MMIO(0x2248)
	406	+ GEN8_CS_CHICKEN1,//_MMIO(0x2580)
	407	+ _MMIO(0x2690),
	408	+ _MMIO(0x2694),
	409	+ _MMIO(0x2698),
	410	+ _MMIO(0x4de0),
	411	+ _MMIO(0x4de4),
	412	+ _MMIO(0x4dfc),
	413	+ GEN7_COMMON_SLICE_CHICKEN1,//_MMIO(0x7010)
	414	+ _MMIO(0x7014),
	415	+ HDC_CHICKEN0,//_MMIO(0x7300)
	416	+ GEN8_HDC_CHICKEN1,//_MMIO(0x7304)
	417	+ _MMIO(0x7700),
	418	+ _MMIO(0x7704),
	419	+ _MMIO(0x7708),
	420	+ _MMIO(0x770c),
	421	+ _MMIO(0xb110),
	422	+ GEN8_L3SQCREG4,//_MMIO(0xb118)
	423	+ _MMIO(0xe100),
	424	+ _MMIO(0xe18c),
	425	+ _MMIO(0xe48c),
	426	+ _MMIO(0xe5f4),
	427	+};
	428	+
	429	+/* a simple bsearch */
	430	+static inline bool in_whitelist(unsigned int reg)
	431	+{
	432	+ int left = 0, right = ARRAY_SIZE(force_nonpriv_white_list);
	433	+ i915_reg_t *array = force_nonpriv_white_list;
	434	+
	435	+ while (left < right) {
	436	+ int mid = (left + right)/2;
	437	+
	438	+ if (reg > array[mid].reg)
	439	+ left = mid + 1;
	440	+ else if (reg < array[mid].reg)
	441	+ right = mid;
	442	+ else
	443	+ return true;
	444	+ }
	445	+ return false;
	446	+}
	447	+
	448	+static int force_nonpriv_write(struct intel_vgpu *vgpu,
	449	+ unsigned int offset, void *p_data, unsigned int bytes)
	450	+{
	451	+ u32 reg_nonpriv = (u32 )p_data;
	452	+ int ret = -EINVAL;
	453	+
	454	+ if ((bytes != 4) \|\| ((offset & (bytes - 1)) != 0)) {
	455	+ gvt_err("vgpu(%d) Invalid FORCE_NONPRIV offset %x(%dB)\n",
	456	+ vgpu->id, offset, bytes);
	457	+ return ret;
	458	+ }
	459	+
	460	+ if (in_whitelist(reg_nonpriv)) {
	461	+ ret = intel_vgpu_default_mmio_write(vgpu, offset, p_data,
	462	+ bytes);
	463	+ } else {
	464	+ gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x\n",
	465	+ vgpu->id, reg_nonpriv);
	466	+ }
	467	+ return ret;
	468	+}
	469	+
372	470	static int ddi_buf_ctl_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
373	471	void *p_data, unsigned int bytes)
374	472	{
@@ -1001,6 +1099,7 @@ static int pvinfo_mmio_read(struct intel_vgpu *vgpu, unsigned int offset, · 626 unmodified lines
1001	1099	if (invalid_read)
1002	1100	gvt_err("invalid pvinfo read: [%x:%x] = %x\n",
1003	1101	offset, bytes, (u32 )p_data);
	1102	+ vgpu->pv_notified = true;
1004	1103	return 0;
1005	1104	}
1006	1105
@@ -1039,7 +1138,7 @@ static int send_display_ready_uevent(struct intel_vgpu *vgpu, int ready) · 32 unmodified lines
1039	1138	char vmid_str[20];
1040	1139	char display_ready_str[20];
1041	1140
1042		− snprintf(display_ready_str, 20, "GVT_DISPLAY_READY=%d\n", ready);
	1141	+ snprintf(display_ready_str, 20, "GVT_DISPLAY_READY=%d", ready);
1043	1142	env[0] = display_ready_str;
1044	1143
1045	1144	snprintf(vmid_str, 20, "VMID=%d", vgpu->id);
@@ -1078,6 +1177,9 @@ static int pvinfo_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, · 32 unmodified lines
1078	1177	case _vgtif_reg(execlist_context_descriptor_lo):
1079	1178	case _vgtif_reg(execlist_context_descriptor_hi):
1080	1179	break;
	1180	+ case _vgtif_reg(rsv5[0])..._vgtif_reg(rsv5[3]):
	1181	+ enter_failsafe_mode(vgpu, GVT_FAILSAFE_INSUFFICIENT_RESOURCE);
	1182	+ break;
1081	1183	default:
1082	1184	gvt_err("invalid pvinfo write offset %x bytes %x data %x\n",
1083	1185	offset, bytes, data);
@@ -1203,26 +1305,37 @@ static int mailbox_write(struct intel_vgpu *vgpu, unsigned int offset, · 119 unmodified lines
1203	1305	u32 *data0 = &vgpu_vreg(vgpu, GEN6_PCODE_DATA);
1204	1306
1205	1307	switch (cmd) {
1206		− case 0x6:
1207		− /**
1208		− * "Read memory latency" command on gen9.
1209		− * Below memory latency values are read
1210		− * from skylake platform.
1211		− */
1212		− if (!*data0)
1213		− *data0 = 0x1e1a1100;
1214		− else
1215		− *data0 = 0x61514b3d;
	1308	+ case GEN9_PCODE_READ_MEM_LATENCY:
	1309	+ if (IS_SKYLAKE(vgpu->gvt->dev_priv)) {
	1310	+ /**
	1311	+ * "Read memory latency" command on gen9.
	1312	+ * Below memory latency values are read
	1313	+ * from skylake platform.
	1314	+ */
	1315	+ if (!*data0)
	1316	+ *data0 = 0x1e1a1100;
	1317	+ else
	1318	+ *data0 = 0x61514b3d;
	1319	+ }
	1320	+ break;
	1321	+ case SKL_PCODE_CDCLK_CONTROL:
	1322	+ if (IS_SKYLAKE(vgpu->gvt->dev_priv))
	1323	+ *data0 = SKL_CDCLK_READY_FOR_CHANGE;
1216	1324	break;
1217		− case 0x5:
	1325	+ case GEN6_PCODE_READ_RC6VIDS:
1218	1326	*data0 \|= 0x1;
1219	1327	break;
1220	1328	}
1221	1329
1222	1330	gvt_dbg_core("VM(%d) write %x to mailbox, return data0 %x\n",
1223	1331	vgpu->id, value, *data0);
1224		−
1225		− value &= ~(1 << 31);
	1332	+ /**
	1333	+ * PCODE_READY clear means ready for pcode read/write,
	1334	+ * PCODE_ERROR_MASK clear means no error happened. In GVT-g we
	1335	+ * always emulate as pcode read/write success and ready for access
	1336	+ * anytime, since we don't touch real physical registers here.
	1337	+ */
	1338	+ value &= ~(GEN6_PCODE_READY \| GEN6_PCODE_ERROR_MASK);
1226	1339	return intel_vgpu_default_mmio_write(vgpu, offset, &value, bytes);
1227	1340	}
1228	1341
@@ -1318,6 +1431,17 @@ static int ring_mode_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, · 89 unmodified lines
1318	1431	bool enable_execlist;
1319	1432
1320	1433	write_vreg(vgpu, offset, p_data, bytes);
	1434	+
	1435	+ /* when PPGTT mode enabled, we will check if guest has called
	1436	+ * pvinfo, if not, we will treat this guest as non-gvtg-aware
	1437	+ * guest, and stop emulating its cfg space, mmio, gtt, etc.
	1438	+ */
	1439	+ if (((data & _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE)) \|\|
	1440	+ (data & _MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE)))
	1441	+ && !vgpu->pv_notified) {
	1442	+ enter_failsafe_mode(vgpu, GVT_FAILSAFE_UNSUPPORTED_GUEST);
	1443	+ return 0;
	1444	+ }
1321	1445	if ((data & _MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE))
1322	1446	\|\| (data & _MASKED_BIT_DISABLE(GFX_RUN_LIST_ENABLE))) {
1323	1447	enable_execlist = !!(data & GFX_RUN_LIST_ENABLE);
@@ -1400,6 +1524,9 @@ static int ring_reset_ctl_write(struct intel_vgpu *vgpu, · 76 unmodified lines
1400	1524	#define MMIO_GM(reg, d, r, w) \
1401	1525	MMIO_F(reg, 4, F_GMADR, 0xFFFFF000, 0, d, r, w)
1402	1526
	1527	+#define MMIO_GM_RDR(reg, d, r, w) \
	1528	+ MMIO_F(reg, 4, F_GMADR \| F_CMD_ACCESS, 0xFFFFF000, 0, d, r, w)
	1529	+
1403	1530	#define MMIO_RO(reg, d, f, rm, r, w) \
1404	1531	MMIO_F(reg, 4, F_RO \| f, 0, rm, d, r, w)
1405	1532
@@ -1419,6 +1546,9 @@ static int ring_reset_ctl_write(struct intel_vgpu *vgpu, · 13 unmodified lines
1419	1546	#define MMIO_RING_GM(prefix, d, r, w) \
1420	1547	MMIO_RING_F(prefix, 4, F_GMADR, 0xFFFF0000, 0, d, r, w)
1421	1548
	1549	+#define MMIO_RING_GM_RDR(prefix, d, r, w) \
	1550	+ MMIO_RING_F(prefix, 4, F_GMADR \| F_CMD_ACCESS, 0xFFFF0000, 0, d, r, w)
	1551	+
1422	1552	#define MMIO_RING_RO(prefix, d, f, rm, r, w) \
1423	1553	MMIO_RING_F(prefix, 4, F_RO \| f, 0, rm, d, r, w)
1424	1554
@@ -1427,73 +1557,81 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) · 2 unmodified lines
1427	1557	struct drm_i915_private *dev_priv = gvt->dev_priv;
1428	1558	int ret;
1429	1559
1430		− MMIO_RING_DFH(RING_IMR, D_ALL, 0, NULL, intel_vgpu_reg_imr_handler);
	1560	+ MMIO_RING_DFH(RING_IMR, D_ALL, F_CMD_ACCESS, NULL,
	1561	+ intel_vgpu_reg_imr_handler);
1431	1562
1432	1563	MMIO_DFH(SDEIMR, D_ALL, 0, NULL, intel_vgpu_reg_imr_handler);
1433	1564	MMIO_DFH(SDEIER, D_ALL, 0, NULL, intel_vgpu_reg_ier_handler);
1434	1565	MMIO_DFH(SDEIIR, D_ALL, 0, NULL, intel_vgpu_reg_iir_handler);
1435	1566	MMIO_D(SDEISR, D_ALL);
1436	1567
1437		− MMIO_RING_D(RING_HWSTAM, D_ALL);
	1568	+ MMIO_RING_DFH(RING_HWSTAM, D_ALL, F_CMD_ACCESS, NULL, NULL);
1438	1569
1439		− MMIO_GM(RENDER_HWS_PGA_GEN7, D_ALL, NULL, NULL);
1440		− MMIO_GM(BSD_HWS_PGA_GEN7, D_ALL, NULL, NULL);
1441		− MMIO_GM(BLT_HWS_PGA_GEN7, D_ALL, NULL, NULL);
1442		− MMIO_GM(VEBOX_HWS_PGA_GEN7, D_ALL, NULL, NULL);
	1570	+ MMIO_GM_RDR(RENDER_HWS_PGA_GEN7, D_ALL, NULL, NULL);
	1571	+ MMIO_GM_RDR(BSD_HWS_PGA_GEN7, D_ALL, NULL, NULL);
	1572	+ MMIO_GM_RDR(BLT_HWS_PGA_GEN7, D_ALL, NULL, NULL);
	1573	+ MMIO_GM_RDR(VEBOX_HWS_PGA_GEN7, D_ALL, NULL, NULL);
1443	1574
1444	1575	#define RING_REG(base) (base + 0x28)
1445		− MMIO_RING_D(RING_REG, D_ALL);
	1576	+ MMIO_RING_DFH(RING_REG, D_ALL, F_CMD_ACCESS, NULL, NULL);
1446	1577	#undef RING_REG
1447	1578
1448	1579	#define RING_REG(base) (base + 0x134)
1449		− MMIO_RING_D(RING_REG, D_ALL);
	1580	+ MMIO_RING_DFH(RING_REG, D_ALL, F_CMD_ACCESS, NULL, NULL);
1450	1581	#undef RING_REG
1451	1582
1452		− MMIO_GM(0x2148, D_ALL, NULL, NULL);
1453		− MMIO_GM(CCID, D_ALL, NULL, NULL);
1454		− MMIO_GM(0x12198, D_ALL, NULL, NULL);
	1583	+ MMIO_GM_RDR(0x2148, D_ALL, NULL, NULL);
	1584	+ MMIO_GM_RDR(CCID, D_ALL, NULL, NULL);
	1585	+ MMIO_GM_RDR(0x12198, D_ALL, NULL, NULL);
1455	1586	MMIO_D(GEN7_CXT_SIZE, D_ALL);
1456	1587
1457		− MMIO_RING_D(RING_TAIL, D_ALL);
1458		− MMIO_RING_D(RING_HEAD, D_ALL);
1459		− MMIO_RING_D(RING_CTL, D_ALL);
1460		− MMIO_RING_D(RING_ACTHD, D_ALL);
1461		− MMIO_RING_GM(RING_START, D_ALL, NULL, NULL);
	1588	+ MMIO_RING_DFH(RING_TAIL, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1589	+ MMIO_RING_DFH(RING_HEAD, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1590	+ MMIO_RING_DFH(RING_CTL, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1591	+ MMIO_RING_DFH(RING_ACTHD, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1592	+ MMIO_RING_GM_RDR(RING_START, D_ALL, NULL, NULL);
1462	1593
1463	1594	/* RING MODE */
1464	1595	#define RING_REG(base) (base + 0x29c)
1465		− MMIO_RING_DFH(RING_REG, D_ALL, F_MODE_MASK, NULL, ring_mode_mmio_write);
	1596	+ MMIO_RING_DFH(RING_REG, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL,
	1597	+ ring_mode_mmio_write);
1466	1598	#undef RING_REG
1467	1599
1468		− MMIO_RING_DFH(RING_MI_MODE, D_ALL, F_MODE_MASK, NULL, NULL);
1469		− MMIO_RING_DFH(RING_INSTPM, D_ALL, F_MODE_MASK, NULL, NULL);
	1600	+ MMIO_RING_DFH(RING_MI_MODE, D_ALL, F_MODE_MASK \| F_CMD_ACCESS,
	1601	+ NULL, NULL);
	1602	+ MMIO_RING_DFH(RING_INSTPM, D_ALL, F_MODE_MASK \| F_CMD_ACCESS,
	1603	+ NULL, NULL);
1470	1604	MMIO_RING_DFH(RING_TIMESTAMP, D_ALL, F_CMD_ACCESS,
1471	1605	ring_timestamp_mmio_read, NULL);
1472	1606	MMIO_RING_DFH(RING_TIMESTAMP_UDW, D_ALL, F_CMD_ACCESS,
1473	1607	ring_timestamp_mmio_read, NULL);
1474	1608
1475		− MMIO_DFH(GEN7_GT_MODE, D_ALL, F_MODE_MASK, NULL, NULL);
1476		− MMIO_DFH(CACHE_MODE_0_GEN7, D_ALL, F_MODE_MASK, NULL, NULL);
	1609	+ MMIO_DFH(GEN7_GT_MODE, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1610	+ MMIO_DFH(CACHE_MODE_0_GEN7, D_ALL, F_MODE_MASK \| F_CMD_ACCESS,
	1611	+ NULL, NULL);
1477	1612	MMIO_DFH(CACHE_MODE_1, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
1478		−
1479		− MMIO_DFH(0x20dc, D_ALL, F_MODE_MASK, NULL, NULL);
1480		− MMIO_DFH(_3D_CHICKEN3, D_ALL, F_MODE_MASK, NULL, NULL);
1481		− MMIO_DFH(0x2088, D_ALL, F_MODE_MASK, NULL, NULL);
1482		− MMIO_DFH(0x20e4, D_ALL, F_MODE_MASK, NULL, NULL);
1483		− MMIO_DFH(0x2470, D_ALL, F_MODE_MASK, NULL, NULL);
1484		− MMIO_D(GAM_ECOCHK, D_ALL);
1485		− MMIO_DFH(GEN7_COMMON_SLICE_CHICKEN1, D_ALL, F_MODE_MASK, NULL, NULL);
	1613	+ MMIO_DFH(CACHE_MODE_0, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1614	+ MMIO_DFH(0x2124, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1615	+
	1616	+ MMIO_DFH(0x20dc, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1617	+ MMIO_DFH(_3D_CHICKEN3, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1618	+ MMIO_DFH(0x2088, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1619	+ MMIO_DFH(0x20e4, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1620	+ MMIO_DFH(0x2470, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	1621	+ MMIO_DFH(GAM_ECOCHK, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1622	+ MMIO_DFH(GEN7_COMMON_SLICE_CHICKEN1, D_ALL, F_MODE_MASK \| F_CMD_ACCESS,
	1623	+ NULL, NULL);
1486	1624	MMIO_DFH(COMMON_SLICE_CHICKEN2, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
1487		− MMIO_D(0x9030, D_ALL);
1488		− MMIO_D(0x20a0, D_ALL);
1489		− MMIO_D(0x2420, D_ALL);
1490		− MMIO_D(0x2430, D_ALL);
1491		− MMIO_D(0x2434, D_ALL);
1492		− MMIO_D(0x2438, D_ALL);
1493		− MMIO_D(0x243c, D_ALL);
1494		− MMIO_DFH(0x7018, D_ALL, F_MODE_MASK, NULL, NULL);
	1625	+ MMIO_DFH(0x9030, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1626	+ MMIO_DFH(0x20a0, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1627	+ MMIO_DFH(0x2420, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1628	+ MMIO_DFH(0x2430, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1629	+ MMIO_DFH(0x2434, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1630	+ MMIO_DFH(0x2438, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1631	+ MMIO_DFH(0x243c, D_ALL, F_CMD_ACCESS, NULL, NULL);
	1632	+ MMIO_DFH(0x7018, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
1495	1633	MMIO_DFH(HALF_SLICE_CHICKEN3, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
1496		− MMIO_DFH(0xe100, D_ALL, F_MODE_MASK, NULL, NULL);
	1634	+ MMIO_DFH(GEN7_HALF_SLICE_CHICKEN1, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
1497	1635
1498	1636	/* display */
1499	1637	MMIO_F(0x60220, 0x20, 0, 0, 0, D_ALL, NULL, NULL);
@@ -2022,8 +2160,8 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) · 522 unmodified lines
2022	2160	MMIO_D(FORCEWAKE_ACK, D_ALL);
2023	2161	MMIO_D(GEN6_GT_CORE_STATUS, D_ALL);
2024	2162	MMIO_D(GEN6_GT_THREAD_STATUS_REG, D_ALL);
2025		− MMIO_D(GTFIFODBG, D_ALL);
2026		− MMIO_D(GTFIFOCTL, D_ALL);
	2163	+ MMIO_DFH(GTFIFODBG, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2164	+ MMIO_DFH(GTFIFOCTL, D_ALL, F_CMD_ACCESS, NULL, NULL);
2027	2165	MMIO_DH(FORCEWAKE_MT, D_PRE_SKL, NULL, mul_force_wake_write);
2028	2166	MMIO_DH(FORCEWAKE_ACK_HSW, D_HSW \| D_BDW, NULL, NULL);
2029	2167	MMIO_D(ECOBUS, D_ALL);
@@ -2080,7 +2218,7 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) · 50 unmodified lines
2080	2218
2081	2219	MMIO_F(0x4f000, 0x90, 0, 0, 0, D_ALL, NULL, NULL);
2082	2220
2083		− MMIO_D(GEN6_PCODE_MAILBOX, D_PRE_SKL);
	2221	+ MMIO_D(GEN6_PCODE_MAILBOX, D_PRE_BDW);
2084	2222	MMIO_D(GEN6_PCODE_DATA, D_ALL);
2085	2223	MMIO_D(0x13812c, D_ALL);
2086	2224	MMIO_DH(GEN7_ERR_INT, D_ALL, NULL, NULL);
@@ -2159,36 +2297,35 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) · 72 unmodified lines
2159	2297	MMIO_D(0x1a054, D_ALL);
2160	2298
2161	2299	MMIO_D(0x44070, D_ALL);
2162		−
2163		− MMIO_D(0x215c, D_HSW_PLUS);
	2300	+ MMIO_DFH(0x215c, D_HSW_PLUS, F_CMD_ACCESS, NULL, NULL);
2164	2301	MMIO_DFH(0x2178, D_ALL, F_CMD_ACCESS, NULL, NULL);
2165	2302	MMIO_DFH(0x217c, D_ALL, F_CMD_ACCESS, NULL, NULL);
2166	2303	MMIO_DFH(0x12178, D_ALL, F_CMD_ACCESS, NULL, NULL);
2167	2304	MMIO_DFH(0x1217c, D_ALL, F_CMD_ACCESS, NULL, NULL);
2168	2305
2169		− MMIO_F(0x2290, 8, 0, 0, 0, D_HSW_PLUS, NULL, NULL);
2170		− MMIO_D(GEN7_OACONTROL, D_HSW);
	2306	+ MMIO_F(0x2290, 8, F_CMD_ACCESS, 0, 0, D_HSW_PLUS, NULL, NULL);
	2307	+ MMIO_DFH(GEN7_OACONTROL, D_HSW, F_CMD_ACCESS, NULL, NULL);
2171	2308	MMIO_D(0x2b00, D_BDW_PLUS);
2172	2309	MMIO_D(0x2360, D_BDW_PLUS);
2173		− MMIO_F(0x5200, 32, 0, 0, 0, D_ALL, NULL, NULL);
2174		− MMIO_F(0x5240, 32, 0, 0, 0, D_ALL, NULL, NULL);
2175		− MMIO_F(0x5280, 16, 0, 0, 0, D_ALL, NULL, NULL);
	2310	+ MMIO_F(0x5200, 32, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2311	+ MMIO_F(0x5240, 32, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2312	+ MMIO_F(0x5280, 16, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
2176	2313
2177	2314	MMIO_DFH(0x1c17c, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2178	2315	MMIO_DFH(0x1c178, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2179		− MMIO_D(BCS_SWCTRL, D_ALL);
2180		−
2181		− MMIO_F(HS_INVOCATION_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2182		− MMIO_F(DS_INVOCATION_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2183		− MMIO_F(IA_VERTICES_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2184		− MMIO_F(IA_PRIMITIVES_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2185		− MMIO_F(VS_INVOCATION_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2186		− MMIO_F(GS_INVOCATION_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2187		− MMIO_F(GS_PRIMITIVES_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2188		− MMIO_F(CL_INVOCATION_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2189		− MMIO_F(CL_PRIMITIVES_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2190		− MMIO_F(PS_INVOCATION_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
2191		− MMIO_F(PS_DEPTH_COUNT, 8, 0, 0, 0, D_ALL, NULL, NULL);
	2316	+ MMIO_DFH(BCS_SWCTRL, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2317	+
	2318	+ MMIO_F(HS_INVOCATION_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2319	+ MMIO_F(DS_INVOCATION_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2320	+ MMIO_F(IA_VERTICES_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2321	+ MMIO_F(IA_PRIMITIVES_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2322	+ MMIO_F(VS_INVOCATION_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2323	+ MMIO_F(GS_INVOCATION_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2324	+ MMIO_F(GS_PRIMITIVES_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2325	+ MMIO_F(CL_INVOCATION_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2326	+ MMIO_F(CL_PRIMITIVES_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2327	+ MMIO_F(PS_INVOCATION_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
	2328	+ MMIO_F(PS_DEPTH_COUNT, 8, F_CMD_ACCESS, 0, 0, D_ALL, NULL, NULL);
2192	2329	MMIO_DH(0x4260, D_BDW_PLUS, NULL, gvt_reg_tlb_control_handler);
2193	2330	MMIO_DH(0x4264, D_BDW_PLUS, NULL, gvt_reg_tlb_control_handler);
2194	2331	MMIO_DH(0x4268, D_BDW_PLUS, NULL, gvt_reg_tlb_control_handler);
@@ -2196,6 +2333,17 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) · 1 unmodified line
2196	2333	MMIO_DH(0x4270, D_BDW_PLUS, NULL, gvt_reg_tlb_control_handler);
2197	2334	MMIO_DFH(0x4094, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2198	2335
	2336	+ MMIO_DFH(ARB_MODE, D_ALL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	2337	+ MMIO_RING_GM_RDR(RING_BBADDR, D_ALL, NULL, NULL);
	2338	+ MMIO_DFH(0x2220, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2339	+ MMIO_DFH(0x12220, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2340	+ MMIO_DFH(0x22220, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2341	+ MMIO_RING_DFH(RING_SYNC_1, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2342	+ MMIO_RING_DFH(RING_SYNC_0, D_ALL, F_CMD_ACCESS, NULL, NULL);
	2343	+ MMIO_DFH(0x22178, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2344	+ MMIO_DFH(0x1a178, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2345	+ MMIO_DFH(0x1a17c, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2346	+ MMIO_DFH(0x2217c, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2199	2347	return 0;
2200	2348	}
2201	2349
@@ -2204,7 +2352,7 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 2 unmodified lines
2204	2352	struct drm_i915_private *dev_priv = gvt->dev_priv;
2205	2353	int ret;
2206	2354
2207		− MMIO_DH(RING_IMR(GEN8_BSD2_RING_BASE), D_BDW_PLUS, NULL,
	2355	+ MMIO_DFH(RING_IMR(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_CMD_ACCESS, NULL,
2208	2356	intel_vgpu_reg_imr_handler);
2209	2357
2210	2358	MMIO_DH(GEN8_GT_IMR(0), D_BDW_PLUS, NULL, intel_vgpu_reg_imr_handler);
@@ -2269,24 +2417,31 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 58 unmodified lines
2269	2417	MMIO_DH(GEN8_MASTER_IRQ, D_BDW_PLUS, NULL,
2270	2418	intel_vgpu_reg_master_irq_handler);
2271	2419
2272		− MMIO_D(RING_HWSTAM(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
2273		− MMIO_D(0x1c134, D_BDW_PLUS);
2274		−
2275		− MMIO_D(RING_TAIL(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
2276		− MMIO_D(RING_HEAD(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
2277		− MMIO_GM(RING_START(GEN8_BSD2_RING_BASE), D_BDW_PLUS, NULL, NULL);
2278		− MMIO_D(RING_CTL(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
2279		− MMIO_D(RING_ACTHD(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
2280		− MMIO_D(RING_ACTHD_UDW(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
2281		− MMIO_DFH(0x1c29c, D_BDW_PLUS, F_MODE_MASK, NULL, ring_mode_mmio_write);
2282		− MMIO_DFH(RING_MI_MODE(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_MODE_MASK,
2283		− NULL, NULL);
2284		− MMIO_DFH(RING_INSTPM(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_MODE_MASK,
2285		− NULL, NULL);
	2420	+ MMIO_DFH(RING_HWSTAM(GEN8_BSD2_RING_BASE), D_BDW_PLUS,
	2421	+ F_CMD_ACCESS, NULL, NULL);
	2422	+ MMIO_DFH(0x1c134, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2423	+
	2424	+ MMIO_DFH(RING_TAIL(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_CMD_ACCESS,
	2425	+ NULL, NULL);
	2426	+ MMIO_DFH(RING_HEAD(GEN8_BSD2_RING_BASE), D_BDW_PLUS,
	2427	+ F_CMD_ACCESS, NULL, NULL);
	2428	+ MMIO_GM_RDR(RING_START(GEN8_BSD2_RING_BASE), D_BDW_PLUS, NULL, NULL);
	2429	+ MMIO_DFH(RING_CTL(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_CMD_ACCESS,
	2430	+ NULL, NULL);
	2431	+ MMIO_DFH(RING_ACTHD(GEN8_BSD2_RING_BASE), D_BDW_PLUS,
	2432	+ F_CMD_ACCESS, NULL, NULL);
	2433	+ MMIO_DFH(RING_ACTHD_UDW(GEN8_BSD2_RING_BASE), D_BDW_PLUS,
	2434	+ F_CMD_ACCESS, NULL, NULL);
	2435	+ MMIO_DFH(0x1c29c, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS, NULL,
	2436	+ ring_mode_mmio_write);
	2437	+ MMIO_DFH(RING_MI_MODE(GEN8_BSD2_RING_BASE), D_BDW_PLUS,
	2438	+ F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	2439	+ MMIO_DFH(RING_INSTPM(GEN8_BSD2_RING_BASE), D_BDW_PLUS,
	2440	+ F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
2286	2441	MMIO_DFH(RING_TIMESTAMP(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_CMD_ACCESS,
2287	2442	ring_timestamp_mmio_read, NULL);
2288	2443
2289		− MMIO_RING_D(RING_ACTHD_UDW, D_BDW_PLUS);
	2444	+ MMIO_RING_DFH(RING_ACTHD_UDW, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2290	2445
2291	2446	#define RING_REG(base) (base + 0xd0)
2292	2447	MMIO_RING_F(RING_REG, 4, F_RO, 0,
@@ -2303,13 +2458,16 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 10 unmodified lines
2303	2458	#undef RING_REG
2304	2459
2305	2460	#define RING_REG(base) (base + 0x234)
2306		− MMIO_RING_F(RING_REG, 8, F_RO, 0, ~0, D_BDW_PLUS, NULL, NULL);
2307		− MMIO_F(RING_REG(GEN8_BSD2_RING_BASE), 4, F_RO, 0, ~0LL, D_BDW_PLUS, NULL, NULL);
	2461	+ MMIO_RING_F(RING_REG, 8, F_RO \| F_CMD_ACCESS, 0, ~0, D_BDW_PLUS,
	2462	+ NULL, NULL);
	2463	+ MMIO_F(RING_REG(GEN8_BSD2_RING_BASE), 4, F_RO \| F_CMD_ACCESS, 0,
	2464	+ ~0LL, D_BDW_PLUS, NULL, NULL);
2308	2465	#undef RING_REG
2309	2466
2310	2467	#define RING_REG(base) (base + 0x244)
2311		− MMIO_RING_D(RING_REG, D_BDW_PLUS);
2312		− MMIO_D(RING_REG(GEN8_BSD2_RING_BASE), D_BDW_PLUS);
	2468	+ MMIO_RING_DFH(RING_REG, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2469	+ MMIO_DFH(RING_REG(GEN8_BSD2_RING_BASE), D_BDW_PLUS, F_CMD_ACCESS,
	2470	+ NULL, NULL);
2313	2471	#undef RING_REG
2314	2472
2315	2473	#define RING_REG(base) (base + 0x370)
@@ -2331,6 +2489,8 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 15 unmodified lines
2331	2489	MMIO_D(GEN7_MISCCPCTL, D_BDW_PLUS);
2332	2490	MMIO_D(0x1c054, D_BDW_PLUS);
2333	2491
	2492	+ MMIO_DH(GEN6_PCODE_MAILBOX, D_BDW_PLUS, NULL, mailbox_write);
	2493	+
2334	2494	MMIO_D(GEN8_PRIVATE_PAT_LO, D_BDW_PLUS);
2335	2495	MMIO_D(GEN8_PRIVATE_PAT_HI, D_BDW_PLUS);
2336	2496
@@ -2341,14 +2501,14 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 4 unmodified lines
2341	2501	MMIO_F(RING_REG(GEN8_BSD2_RING_BASE), 32, 0, 0, 0, D_BDW_PLUS, NULL, NULL);
2342	2502	#undef RING_REG
2343	2503
2344		− MMIO_RING_GM(RING_HWS_PGA, D_BDW_PLUS, NULL, NULL);
2345		− MMIO_GM(0x1c080, D_BDW_PLUS, NULL, NULL);
	2504	+ MMIO_RING_GM_RDR(RING_HWS_PGA, D_BDW_PLUS, NULL, NULL);
	2505	+ MMIO_GM_RDR(RING_HWS_PGA(GEN8_BSD2_RING_BASE), D_BDW_PLUS, NULL, NULL);
2346	2506
2347	2507	MMIO_DFH(HDC_CHICKEN0, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
2348	2508
2349		− MMIO_D(CHICKEN_PIPESL_1(PIPE_A), D_BDW);
2350		− MMIO_D(CHICKEN_PIPESL_1(PIPE_B), D_BDW);
2351		− MMIO_D(CHICKEN_PIPESL_1(PIPE_C), D_BDW);
	2509	+ MMIO_D(CHICKEN_PIPESL_1(PIPE_A), D_BDW_PLUS);
	2510	+ MMIO_D(CHICKEN_PIPESL_1(PIPE_B), D_BDW_PLUS);
	2511	+ MMIO_D(CHICKEN_PIPESL_1(PIPE_C), D_BDW_PLUS);
2352	2512
2353	2513	MMIO_D(WM_MISC, D_BDW);
2354	2514	MMIO_D(BDW_EDP_PSR_BASE, D_BDW);
@@ -2362,27 +2522,31 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 7 unmodified lines
2362	2522	MMIO_D(GEN8_EU_DISABLE1, D_BDW_PLUS);
2363	2523	MMIO_D(GEN8_EU_DISABLE2, D_BDW_PLUS);
2364	2524
2365		− MMIO_D(0xfdc, D_BDW);
2366		− MMIO_DFH(GEN8_ROW_CHICKEN, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2367		− MMIO_D(GEN7_ROW_CHICKEN2, D_BDW_PLUS);
2368		− MMIO_D(GEN8_UCGCTL6, D_BDW_PLUS);
	2525	+ MMIO_D(0xfdc, D_BDW_PLUS);
	2526	+ MMIO_DFH(GEN8_ROW_CHICKEN, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS,
	2527	+ NULL, NULL);
	2528	+ MMIO_DFH(GEN7_ROW_CHICKEN2, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS,
	2529	+ NULL, NULL);
	2530	+ MMIO_DFH(GEN8_UCGCTL6, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2369	2531
2370		− MMIO_D(0xb1f0, D_BDW);
2371		− MMIO_D(0xb1c0, D_BDW);
	2532	+ MMIO_DFH(0xb1f0, D_BDW, F_CMD_ACCESS, NULL, NULL);
	2533	+ MMIO_DFH(0xb1c0, D_BDW, F_CMD_ACCESS, NULL, NULL);
2372	2534	MMIO_DFH(GEN8_L3SQCREG4, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2373		− MMIO_D(0xb100, D_BDW);
2374		− MMIO_D(0xb10c, D_BDW);
	2535	+ MMIO_DFH(0xb100, D_BDW, F_CMD_ACCESS, NULL, NULL);
	2536	+ MMIO_DFH(0xb10c, D_BDW, F_CMD_ACCESS, NULL, NULL);
2375	2537	MMIO_D(0xb110, D_BDW);
2376	2538
2377		− MMIO_DFH(0x24d0, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2378		− MMIO_DFH(0x24d4, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2379		− MMIO_DFH(0x24d8, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2380		− MMIO_DFH(0x24dc, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2539	+ MMIO_F(0x24d0, 48, F_CMD_ACCESS, 0, 0, D_BDW_PLUS,
	2540	+ NULL, force_nonpriv_write);
	2541	+
	2542	+ MMIO_D(0x22040, D_BDW_PLUS);
	2543	+ MMIO_D(0x44484, D_BDW_PLUS);
	2544	+ MMIO_D(0x4448c, D_BDW_PLUS);
2381	2545
2382		− MMIO_D(0x83a4, D_BDW);
	2546	+ MMIO_DFH(0x83a4, D_BDW, F_CMD_ACCESS, NULL, NULL);
2383	2547	MMIO_D(GEN8_L3_LRA_1_GPGPU, D_BDW_PLUS);
2384	2548
2385		− MMIO_D(0x8430, D_BDW);
	2549	+ MMIO_DFH(0x8430, D_BDW, F_CMD_ACCESS, NULL, NULL);
2386	2550
2387	2551	MMIO_D(0x110000, D_BDW_PLUS);
2388	2552
@@ -2394,10 +2558,19 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt) · 5 unmodified lines
2394	2558	MMIO_DFH(0xe194, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
2395	2559	MMIO_DFH(0xe188, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
2396	2560	MMIO_DFH(HALF_SLICE_CHICKEN2, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
2397		− MMIO_DFH(0x2580, D_BDW_PLUS, F_MODE_MASK, NULL, NULL);
2398		−
2399		− MMIO_D(0x2248, D_BDW);
2400		−
	2561	+ MMIO_DFH(0x2580, D_BDW_PLUS, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
	2562	+
	2563	+ MMIO_DFH(0x2248, D_BDW, F_CMD_ACCESS, NULL, NULL);
	2564	+
	2565	+ MMIO_DFH(0xe220, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2566	+ MMIO_DFH(0xe230, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2567	+ MMIO_DFH(0xe240, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2568	+ MMIO_DFH(0xe260, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2569	+ MMIO_DFH(0xe270, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2570	+ MMIO_DFH(0xe280, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2571	+ MMIO_DFH(0xe2a0, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2572	+ MMIO_DFH(0xe2b0, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
	2573	+ MMIO_DFH(0xe2c0, D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
2401	2574	return 0;
2402	2575	}
2403	2576
@@ -2420,7 +2593,6 @@ static int init_skl_mmio_info(struct intel_gvt *gvt) · 16 unmodified lines
2420	2593	MMIO_D(HSW_PWR_WELL_BIOS, D_SKL);
2421	2594	MMIO_DH(HSW_PWR_WELL_DRIVER, D_SKL, NULL, skl_power_well_ctl_write);
2422	2595
2423		− MMIO_DH(GEN6_PCODE_MAILBOX, D_SKL, NULL, mailbox_write);
2424	2596	MMIO_D(0xa210, D_SKL_PLUS);
2425	2597	MMIO_D(GEN9_MEDIA_PG_IDLE_HYSTERESIS, D_SKL_PLUS);
2426	2598	MMIO_D(GEN9_RENDER_PG_IDLE_HYSTERESIS, D_SKL_PLUS);
@@ -2578,16 +2750,16 @@ static int init_skl_mmio_info(struct intel_gvt *gvt) · 151 unmodified lines
2578	2750	MMIO_F(0xb020, 0x80, F_CMD_ACCESS, 0, 0, D_SKL, NULL, NULL);
2579	2751
2580	2752	MMIO_D(0xd08, D_SKL);
2581		− MMIO_D(0x20e0, D_SKL);
2582		− MMIO_D(0x20ec, D_SKL);
	2753	+ MMIO_DFH(0x20e0, D_SKL, F_MODE_MASK, NULL, NULL);
	2754	+ MMIO_DFH(0x20ec, D_SKL, F_MODE_MASK \| F_CMD_ACCESS, NULL, NULL);
2583	2755
2584	2756	/* TRTT */
2585		− MMIO_D(0x4de0, D_SKL);
2586		− MMIO_D(0x4de4, D_SKL);
2587		− MMIO_D(0x4de8, D_SKL);
2588		− MMIO_D(0x4dec, D_SKL);
2589		− MMIO_D(0x4df0, D_SKL);
2590		− MMIO_DH(0x4df4, D_SKL, NULL, gen9_trtte_write);
	2757	+ MMIO_DFH(0x4de0, D_SKL, F_CMD_ACCESS, NULL, NULL);
	2758	+ MMIO_DFH(0x4de4, D_SKL, F_CMD_ACCESS, NULL, NULL);
	2759	+ MMIO_DFH(0x4de8, D_SKL, F_CMD_ACCESS, NULL, NULL);
	2760	+ MMIO_DFH(0x4dec, D_SKL, F_CMD_ACCESS, NULL, NULL);
	2761	+ MMIO_DFH(0x4df0, D_SKL, F_CMD_ACCESS, NULL, NULL);
	2762	+ MMIO_DFH(0x4df4, D_SKL, F_CMD_ACCESS, NULL, gen9_trtte_write);
2591	2763	MMIO_DH(0x4dfc, D_SKL, NULL, gen9_trtt_chicken_write);
2592	2764
2593	2765	MMIO_D(0x45008, D_SKL);
@@ -2611,7 +2783,7 @@ static int init_skl_mmio_info(struct intel_gvt *gvt) · 17 unmodified lines
2611	2783	MMIO_D(0x65f08, D_SKL);
2612	2784	MMIO_D(0x320f0, D_SKL);
2613	2785
2614		− MMIO_D(_REG_VCS2_EXCC, D_SKL);
	2786	+ MMIO_DFH(_REG_VCS2_EXCC, D_SKL, F_CMD_ACCESS, NULL, NULL);
2615	2787	MMIO_D(0x70034, D_SKL);
2616	2788	MMIO_D(0x71034, D_SKL);
2617	2789	MMIO_D(0x72034, D_SKL);
@@ -2624,6 +2796,9 @@ static int init_skl_mmio_info(struct intel_gvt *gvt) · 6 unmodified lines
2624	2796	MMIO_D(_PLANE_KEYMSK_1(PIPE_C), D_SKL);
2625	2797
2626	2798	MMIO_D(0x44500, D_SKL);
	2799	+ MMIO_DFH(GEN9_CSFE_CHICKEN1_RCS, D_SKL_PLUS, F_CMD_ACCESS, NULL, NULL);
	2800	+ MMIO_DFH(GEN8_HDC_CHICKEN1, D_SKL, F_MODE_MASK \| F_CMD_ACCESS,
	2801	+ NULL, NULL);
2627	2802	return 0;
2628	2803	}
2629	2804

drivers/gpu/drm/i915/gvt/kvmgt.c+6−6

@@ -96,10 +96,10 @@ static int gvt_dma_map_iova(struct intel_vgpu *vgpu, kvm_pfn_t pfn, · 95 unmodified lines
96	96	struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev;
97	97	dma_addr_t daddr;
98	98
99		− page = pfn_to_page(pfn);
100		− if (is_error_page(page))
	99	+ if (unlikely(!pfn_valid(pfn)))
101	100	return -EFAULT;
102	101
	102	+ page = pfn_to_page(pfn);
103	103	daddr = dma_map_page(dev, page, 0, PAGE_SIZE,
104	104	PCI_DMA_BIDIRECTIONAL);
105	105	if (dma_mapping_error(dev, daddr))
@@ -295,10 +295,10 @@ static ssize_t description_show(struct kobject kobj, struct device dev, · 189 unmodified lines
295	295	return 0;
296	296
297	297	return sprintf(buf, "low_gm_size: %dMB\nhigh_gm_size: %dMB\n"
298		− "fence: %d\n",
299		− BYTES_TO_MB(type->low_gm_size),
300		− BYTES_TO_MB(type->high_gm_size),
301		− type->fence);
	298	+ "fence: %d\nresolution: %s\n",
	299	+ BYTES_TO_MB(type->low_gm_size),
	300	+ BYTES_TO_MB(type->high_gm_size),
	301	+ type->fence, vgpu_edid_str(type->resolution));
302	302	}
303	303
304	304	static MDEV_TYPE_ATTR_RO(available_instances);

drivers/gpu/drm/i915/gvt/mmio.c+65−1

@@ -57,6 +57,58 @@ int intel_vgpu_gpa_to_mmio_offset(struct intel_vgpu *vgpu, u64 gpa) · 56 unmodified lines
57	57	(reg >= gvt->device_info.gtt_start_offset \
58	58	&& reg < gvt->device_info.gtt_start_offset + gvt_ggtt_sz(gvt))
59	59
	60	+static void failsafe_emulate_mmio_rw(struct intel_vgpu *vgpu, uint64_t pa,
	61	+ void *p_data, unsigned int bytes, bool read)
	62	+{
	63	+ struct intel_gvt *gvt = NULL;
	64	+ void *pt = NULL;
	65	+ unsigned int offset = 0;
	66	+
	67	+ if (!vgpu \|\| !p_data)
	68	+ return;
	69	+
	70	+ gvt = vgpu->gvt;
	71	+ mutex_lock(&gvt->lock);
	72	+ offset = intel_vgpu_gpa_to_mmio_offset(vgpu, pa);
	73	+ if (reg_is_mmio(gvt, offset)) {
	74	+ if (read)
	75	+ intel_vgpu_default_mmio_read(vgpu, offset, p_data,
	76	+ bytes);
	77	+ else
	78	+ intel_vgpu_default_mmio_write(vgpu, offset, p_data,
	79	+ bytes);
	80	+ } else if (reg_is_gtt(gvt, offset) &&
	81	+ vgpu->gtt.ggtt_mm->virtual_page_table) {
	82	+ offset -= gvt->device_info.gtt_start_offset;
	83	+ pt = vgpu->gtt.ggtt_mm->virtual_page_table + offset;
	84	+ if (read)
	85	+ memcpy(p_data, pt, bytes);
	86	+ else
	87	+ memcpy(pt, p_data, bytes);
	88	+
	89	+ } else if (atomic_read(&vgpu->gtt.n_write_protected_guest_page)) {
	90	+ struct intel_vgpu_guest_page *gp;
	91	+
	92	+ /* Since we enter the failsafe mode early during guest boot,
	93	+ * guest may not have chance to set up its ppgtt table, so
	94	+ * there should not be any wp pages for guest. Keep the wp
	95	+ * related code here in case we need to handle it in furture.
	96	+ */
	97	+ gp = intel_vgpu_find_guest_page(vgpu, pa >> PAGE_SHIFT);
	98	+ if (gp) {
	99	+ /* remove write protection to prevent furture traps */
	100	+ intel_vgpu_clean_guest_page(vgpu, gp);
	101	+ if (read)
	102	+ intel_gvt_hypervisor_read_gpa(vgpu, pa,
	103	+ p_data, bytes);
	104	+ else
	105	+ intel_gvt_hypervisor_write_gpa(vgpu, pa,
	106	+ p_data, bytes);
	107	+ }
	108	+ }
	109	+ mutex_unlock(&gvt->lock);
	110	+}
	111	+
60	112	/**
61	113	* intel_vgpu_emulate_mmio_read - emulate MMIO read
62	114	* @vgpu: a vGPU
@@ -75,6 +127,11 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, uint64_t pa, · 12 unmodified lines
75	127	unsigned int offset = 0;
76	128	int ret = -EINVAL;
77	129
	130	+
	131	+ if (vgpu->failsafe) {
	132	+ failsafe_emulate_mmio_rw(vgpu, pa, p_data, bytes, true);
	133	+ return 0;
	134	+ }
78	135	mutex_lock(&gvt->lock);
79	136
80	137	if (atomic_read(&vgpu->gtt.n_write_protected_guest_page)) {
@@ -188,6 +245,11 @@ int intel_vgpu_emulate_mmio_write(struct intel_vgpu *vgpu, uint64_t pa, · 107 unmodified lines
188	245	u32 old_vreg = 0, old_sreg = 0;
189	246	int ret = -EINVAL;
190	247
	248	+ if (vgpu->failsafe) {
	249	+ failsafe_emulate_mmio_rw(vgpu, pa, p_data, bytes, false);
	250	+ return 0;
	251	+ }
	252	+
191	253	mutex_lock(&gvt->lock);
192	254
193	255	if (atomic_read(&vgpu->gtt.n_write_protected_guest_page)) {
@@ -236,7 +298,7 @@ int intel_vgpu_emulate_mmio_write(struct intel_vgpu *vgpu, uint64_t pa, · 42 unmodified lines
236	298
237	299	mmio = intel_gvt_find_mmio_info(gvt, rounddown(offset, 4));
238	300	if (!mmio && !vgpu->mmio.disable_warn_untrack)
239		− gvt_err("vgpu%d: write untracked MMIO %x len %d val %x\n",
	301	+ gvt_dbg_mmio("vgpu%d: write untracked MMIO %x len %d val %x\n",
240	302	vgpu->id, offset, bytes, (u32 )p_data);
241	303
242	304	if (!intel_gvt_mmio_is_unalign(gvt, offset)) {
@@ -322,6 +384,8 @@ void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu) · 79 unmodified lines
322	384
323	385	/* set the bit 0:2(Core C-State ) to C0 */
324	386	vgpu_vreg(vgpu, GEN6_GT_CORE_STATUS) = 0;
	387	+
	388	+ vgpu->mmio.disable_warn_untrack = false;
325	389	}
326	390
327	391	/**

drivers/gpu/drm/i915/gvt/opregion.c+2−3

@@ -27,7 +27,6 @@ · 26 unmodified lines
27	27
28	28	static int init_vgpu_opregion(struct intel_vgpu *vgpu, u32 gpa)
29	29	{
30		− void __iomem *host_va = vgpu->gvt->opregion.opregion_va;
31	30	u8 *buf;
32	31	int i;
33	32
@@ -43,8 +42,8 @@ static int init_vgpu_opregion(struct intel_vgpu *vgpu, u32 gpa) · 9 unmodified lines
43	42	if (!vgpu_opregion(vgpu)->va)
44	43	return -ENOMEM;
45	44
46		− memcpy_fromio(vgpu_opregion(vgpu)->va, host_va,
47		− INTEL_GVT_OPREGION_SIZE);
	45	+ memcpy(vgpu_opregion(vgpu)->va, vgpu->gvt->opregion.opregion_va,
	46	+ INTEL_GVT_OPREGION_SIZE);
48	47
49	48	for (i = 0; i < INTEL_GVT_OPREGION_PAGES; i++)
50	49	vgpu_opregion(vgpu)->gfn[i] = (gpa >> PAGE_SHIFT) + i;

drivers/gpu/drm/i915/gvt/render.c+16−0

@@ -53,6 +53,14 @@ static struct render_mmio gen8_render_mmio_list[] = { · 52 unmodified lines
53	53	{RCS, _MMIO(0x24d4), 0, false},
54	54	{RCS, _MMIO(0x24d8), 0, false},
55	55	{RCS, _MMIO(0x24dc), 0, false},
	56	+ {RCS, _MMIO(0x24e0), 0, false},
	57	+ {RCS, _MMIO(0x24e4), 0, false},
	58	+ {RCS, _MMIO(0x24e8), 0, false},
	59	+ {RCS, _MMIO(0x24ec), 0, false},
	60	+ {RCS, _MMIO(0x24f0), 0, false},
	61	+ {RCS, _MMIO(0x24f4), 0, false},
	62	+ {RCS, _MMIO(0x24f8), 0, false},
	63	+ {RCS, _MMIO(0x24fc), 0, false},
56	64	{RCS, _MMIO(0x7004), 0xffff, true},
57	65	{RCS, _MMIO(0x7008), 0xffff, true},
58	66	{RCS, _MMIO(0x7000), 0xffff, true},
@@ -76,6 +84,14 @@ static struct render_mmio gen9_render_mmio_list[] = { · 17 unmodified lines
76	84	{RCS, _MMIO(0x24d4), 0, false},
77	85	{RCS, _MMIO(0x24d8), 0, false},
78	86	{RCS, _MMIO(0x24dc), 0, false},
	87	+ {RCS, _MMIO(0x24e0), 0, false},
	88	+ {RCS, _MMIO(0x24e4), 0, false},
	89	+ {RCS, _MMIO(0x24e8), 0, false},
	90	+ {RCS, _MMIO(0x24ec), 0, false},
	91	+ {RCS, _MMIO(0x24f0), 0, false},
	92	+ {RCS, _MMIO(0x24f4), 0, false},
	93	+ {RCS, _MMIO(0x24f8), 0, false},
	94	+ {RCS, _MMIO(0x24fc), 0, false},
79	95	{RCS, _MMIO(0x7004), 0xffff, true},
80	96	{RCS, _MMIO(0x7008), 0xffff, true},
81	97	{RCS, _MMIO(0x7000), 0xffff, true},

drivers/gpu/drm/i915/gvt/scheduler.c+36−16

@@ -139,6 +139,9 @@ static int shadow_context_status_change(struct notifier_block *nb, · 138 unmodified lines
139	139	struct intel_vgpu_workload *workload =
140	140	scheduler->current_workload[req->engine->id];
141	141
	142	+ if (unlikely(!workload))
	143	+ return NOTIFY_OK;
	144	+
142	145	switch (action) {
143	146	case INTEL_CONTEXT_SCHEDULE_IN:
144	147	intel_gvt_load_render_mmio(workload->vgpu,
@@ -148,6 +151,15 @@ static int shadow_context_status_change(struct notifier_block *nb, · 3 unmodified lines
148	151	case INTEL_CONTEXT_SCHEDULE_OUT:
149	152	intel_gvt_restore_render_mmio(workload->vgpu,
150	153	workload->ring_id);
	154	+ /* If the status is -EINPROGRESS means this workload
	155	+ * doesn't meet any issue during dispatching so when
	156	+ * get the SCHEDULE_OUT set the status to be zero for
	157	+ * good. If the status is NOT -EINPROGRESS means there
	158	+ * is something wrong happened during dispatching and
	159	+ * the status should not be set to zero
	160	+ */
	161	+ if (workload->status == -EINPROGRESS)
	162	+ workload->status = 0;
151	163	atomic_set(&workload->shadow_ctx_active, 0);
152	164	break;
153	165	default:
@@ -359,15 +371,23 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id) · 205 unmodified lines
359	371	workload = scheduler->current_workload[ring_id];
360	372	vgpu = workload->vgpu;
361	373
362		− if (!workload->status && !vgpu->resetting) {
	374	+ /* For the workload w/ request, needs to wait for the context
	375	+ * switch to make sure request is completed.
	376	+ * For the workload w/o request, directly complete the workload.
	377	+ */
	378	+ if (workload->req) {
363	379	wait_event(workload->shadow_ctx_status_wq,
364	380	!atomic_read(&workload->shadow_ctx_active));
365	381
366		− update_guest_context(workload);
	382	+ i915_gem_request_put(fetch_and_zero(&workload->req));
	383	+
	384	+ if (!workload->status && !vgpu->resetting) {
	385	+ update_guest_context(workload);
367	386
368		− for_each_set_bit(event, workload->pending_events,
369		− INTEL_GVT_EVENT_MAX)
370		− intel_vgpu_trigger_virtual_event(vgpu, event);
	387	+ for_each_set_bit(event, workload->pending_events,
	388	+ INTEL_GVT_EVENT_MAX)
	389	+ intel_vgpu_trigger_virtual_event(vgpu, event);
	390	+ }
371	391	}
372	392
373	393	gvt_dbg_sched("ring id %d complete workload %p status %d\n",
@@ -397,7 +417,6 @@ static int workload_thread(void *priv) · 23 unmodified lines
397	417	int ring_id = p->ring_id;
398	418	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
399	419	struct intel_vgpu_workload *workload = NULL;
400		− long lret;
401	420	int ret;
402	421	bool need_force_wake = IS_SKYLAKE(gvt->dev_priv);
403	422	DEFINE_WAIT_FUNC(wait, woken_wake_function);
@@ -446,23 +465,24 @@ static int workload_thread(void *priv) · 42 unmodified lines
446	465
447	466	gvt_dbg_sched("ring id %d wait workload %p\n",
448	467	workload->ring_id, workload);
449		−
450		− lret = i915_wait_request(workload->req,
	468	+retry:
	469	+ i915_wait_request(workload->req,
451	470	0, MAX_SCHEDULE_TIMEOUT);
452		− if (lret < 0) {
453		− workload->status = lret;
454		− gvt_err("fail to wait workload, skip\n");
455		− } else {
456		− workload->status = 0;
	471	+ /* I915 has replay mechanism and a request will be replayed
	472	+ * if there is i915 reset. So the seqno will be updated anyway.
	473	+ * If the seqno is not updated yet after waiting, which means
	474	+ * the replay may still be in progress and we can wait again.
	475	+ */
	476	+ if (!i915_gem_request_completed(workload->req)) {
	477	+ gvt_dbg_sched("workload %p not completed, wait again\n",
	478	+ workload);
	479	+ goto retry;
457	480	}
458	481
459	482	complete:
460	483	gvt_dbg_sched("will complete workload %p, status: %d\n",
461	484	workload, workload->status);
462	485
463		− if (workload->req)
464		− i915_gem_request_put(fetch_and_zero(&workload->req));
465		−
466	486	complete_current_workload(gvt, ring_id);
467	487
468	488	if (need_force_wake)

drivers/gpu/drm/i915/gvt/vgpu.c+45−27

@@ -64,6 +64,20 @@ void populate_pvinfo_page(struct intel_vgpu *vgpu) · 63 unmodified lines
64	64	WARN_ON(sizeof(struct vgt_if) != VGT_PVINFO_SIZE);
65	65	}
66	66
	67	+static struct {
	68	+ unsigned int low_mm;
	69	+ unsigned int high_mm;
	70	+ unsigned int fence;
	71	+ enum intel_vgpu_edid edid;
	72	+ char *name;
	73	+} vgpu_types[] = {
	74	+/* Fixed vGPU type table */
	75	+ { MB_TO_BYTES(64), MB_TO_BYTES(512), 4, GVT_EDID_1024_768, "8" },
	76	+ { MB_TO_BYTES(128), MB_TO_BYTES(512), 4, GVT_EDID_1920_1200, "4" },
	77	+ { MB_TO_BYTES(256), MB_TO_BYTES(1024), 4, GVT_EDID_1920_1200, "2" },
	78	+ { MB_TO_BYTES(512), MB_TO_BYTES(2048), 4, GVT_EDID_1920_1200, "1" },
	79	+};
	80	+
67	81	/**
68	82	* intel_gvt_init_vgpu_types - initialize vGPU type list
69	83	* @gvt : GVT device
@@ -78,9 +92,8 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) · 8 unmodified lines
78	92	unsigned int min_low;
79	93
80	94	/* vGPU type name is defined as GVTg_Vx_y which contains
81		− * physical GPU generation type and 'y' means maximum vGPU
82		− * instances user can create on one physical GPU for this
83		− * type.
	95	+ * physical GPU generation type (e.g V4 as BDW server, V5 as
	96	+ * SKL server).
84	97	*
85	98	* Depend on physical SKU resource, might see vGPU types like
86	99	* GVTg_V4_8, GVTg_V4_4, GVTg_V4_2, etc. We can create
@@ -92,7 +105,7 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) · 5 unmodified lines
92	105	*/
93	106	low_avail = gvt_aperture_sz(gvt) - HOST_LOW_GM_SIZE;
94	107	high_avail = gvt_hidden_sz(gvt) - HOST_HIGH_GM_SIZE;
95		− num_types = 4;
	108	+ num_types = sizeof(vgpu_types) / sizeof(vgpu_types[0]);
96	109
97	110	gvt->types = kzalloc(num_types * sizeof(struct intel_vgpu_type),
98	111	GFP_KERNEL);
@@ -101,28 +114,29 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) · 2 unmodified lines
101	114
102	115	min_low = MB_TO_BYTES(32);
103	116	for (i = 0; i < num_types; ++i) {
104		− if (low_avail / min_low == 0)
	117	+ if (low_avail / vgpu_types[i].low_mm == 0)
105	118	break;
106		− gvt->types[i].low_gm_size = min_low;
107		− gvt->types[i].high_gm_size = max((min_low<<3), MB_TO_BYTES(384U));
108		− gvt->types[i].fence = 4;
109		− gvt->types[i].max_instance = min(low_avail / min_low,
110		− high_avail / gvt->types[i].high_gm_size);
111		− gvt->types[i].avail_instance = gvt->types[i].max_instance;
	119	+
	120	+ gvt->types[i].low_gm_size = vgpu_types[i].low_mm;
	121	+ gvt->types[i].high_gm_size = vgpu_types[i].high_mm;
	122	+ gvt->types[i].fence = vgpu_types[i].fence;
	123	+ gvt->types[i].resolution = vgpu_types[i].edid;
	124	+ gvt->types[i].avail_instance = min(low_avail / vgpu_types[i].low_mm,
	125	+ high_avail / vgpu_types[i].high_mm);
112	126
113	127	if (IS_GEN8(gvt->dev_priv))
114		− sprintf(gvt->types[i].name, "GVTg_V4_%u",
115		− gvt->types[i].max_instance);
	128	+ sprintf(gvt->types[i].name, "GVTg_V4_%s",
	129	+ vgpu_types[i].name);
116	130	else if (IS_GEN9(gvt->dev_priv))
117		− sprintf(gvt->types[i].name, "GVTg_V5_%u",
118		− gvt->types[i].max_instance);
	131	+ sprintf(gvt->types[i].name, "GVTg_V5_%s",
	132	+ vgpu_types[i].name);
119	133
120		− min_low <<= 1;
121		− gvt_dbg_core("type[%d]: %s max %u avail %u low %u high %u fence %u\n",
122		− i, gvt->types[i].name, gvt->types[i].max_instance,
	134	+ gvt_dbg_core("type[%d]: %s avail %u low %u high %u fence %u res %s\n",
	135	+ i, gvt->types[i].name,
123	136	gvt->types[i].avail_instance,
124	137	gvt->types[i].low_gm_size,
125		− gvt->types[i].high_gm_size, gvt->types[i].fence);
	138	+ gvt->types[i].high_gm_size, gvt->types[i].fence,
	139	+ vgpu_edid_str(gvt->types[i].resolution));
126	140	}
127	141
128	142	gvt->num_types = i;
@@ -138,7 +152,7 @@ static void intel_gvt_update_vgpu_types(struct intel_gvt *gvt) · 9 unmodified lines
138	152	{
139	153	int i;
140	154	unsigned int low_gm_avail, high_gm_avail, fence_avail;
141		− unsigned int low_gm_min, high_gm_min, fence_min, total_min;
	155	+ unsigned int low_gm_min, high_gm_min, fence_min;
142	156
143	157	/* Need to depend on maxium hw resource size but keep on
144	158	* static config for now.
@@ -154,12 +168,11 @@ static void intel_gvt_update_vgpu_types(struct intel_gvt *gvt) · 9 unmodified lines
154	168	low_gm_min = low_gm_avail / gvt->types[i].low_gm_size;
155	169	high_gm_min = high_gm_avail / gvt->types[i].high_gm_size;
156	170	fence_min = fence_avail / gvt->types[i].fence;
157		− total_min = min(min(low_gm_min, high_gm_min), fence_min);
158		− gvt->types[i].avail_instance = min(gvt->types[i].max_instance,
159		− total_min);
	171	+ gvt->types[i].avail_instance = min(min(low_gm_min, high_gm_min),
	172	+ fence_min);
160	173
161		− gvt_dbg_core("update type[%d]: %s max %u avail %u low %u high %u fence %u\n",
162		− i, gvt->types[i].name, gvt->types[i].max_instance,
	174	+ gvt_dbg_core("update type[%d]: %s avail %u low %u high %u fence %u\n",
	175	+ i, gvt->types[i].name,
163	176	gvt->types[i].avail_instance, gvt->types[i].low_gm_size,
164	177	gvt->types[i].high_gm_size, gvt->types[i].fence);
165	178	}
@@ -248,7 +261,7 @@ static struct intel_vgpu __intel_gvt_create_vgpu(struct intel_gvt gvt, · 82 unmodified lines
248	261	if (ret)
249	262	goto out_detach_hypervisor_vgpu;
250	263
251		− ret = intel_vgpu_init_display(vgpu);
	264	+ ret = intel_vgpu_init_display(vgpu, param->resolution);
252	265	if (ret)
253	266	goto out_clean_gtt;
254	267
@@ -312,6 +325,7 @@ struct intel_vgpu intel_gvt_create_vgpu(struct intel_gvt gvt, · 57 unmodified lines
312	325	param.low_gm_sz = type->low_gm_size;
313	326	param.high_gm_sz = type->high_gm_size;
314	327	param.fence_sz = type->fence;
	328	+ param.resolution = type->resolution;
315	329
316	330	/* XXX current param based on MB */
317	331	param.low_gm_sz = BYTES_TO_MB(param.low_gm_sz);
@@ -387,8 +401,12 @@ void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr, · 69 unmodified lines
387	401	populate_pvinfo_page(vgpu);
388	402	intel_vgpu_reset_display(vgpu);
389	403
390		− if (dmlr)
	404	+ if (dmlr) {
391	405	intel_vgpu_reset_cfg_space(vgpu);
	406	+ /* only reset the failsafe mode when dmlr reset */
	407	+ vgpu->failsafe = false;
	408	+ vgpu->pv_notified = false;
	409	+ }
392	410	}
393	411
394	412	vgpu->resetting = false;

drivers/gpu/drm/mxsfb/mxsfb_crtc.c+43−6

@@ -65,13 +65,11 @@ static int mxsfb_set_pixel_fmt(struct mxsfb_drm_private *mxsfb) · 64 unmodified lines
65	65	switch (format) {
66	66	case DRM_FORMAT_RGB565:
67	67	dev_dbg(drm->dev, "Setting up RGB565 mode\n");
68		− ctrl \|= CTRL_SET_BUS_WIDTH(STMLCDIF_16BIT);
69	68	ctrl \|= CTRL_SET_WORD_LENGTH(0);
70	69	ctrl1 \|= CTRL1_SET_BYTE_PACKAGING(0xf);
71	70	break;
72	71	case DRM_FORMAT_XRGB8888:
73	72	dev_dbg(drm->dev, "Setting up XRGB8888 mode\n");
74		− ctrl \|= CTRL_SET_BUS_WIDTH(STMLCDIF_24BIT);
75	73	ctrl \|= CTRL_SET_WORD_LENGTH(3);
76	74	/* Do not use packed pixels = one pixel per word instead. */
77	75	ctrl1 \|= CTRL1_SET_BYTE_PACKAGING(0x7);
@@ -87,6 +85,36 @@ static int mxsfb_set_pixel_fmt(struct mxsfb_drm_private *mxsfb) · 9 unmodified lines
87	85	return 0;
88	86	}
89	87
	88	+static void mxsfb_set_bus_fmt(struct mxsfb_drm_private *mxsfb)
	89	+{
	90	+ struct drm_crtc *crtc = &mxsfb->pipe.crtc;
	91	+ struct drm_device *drm = crtc->dev;
	92	+ u32 bus_format = MEDIA_BUS_FMT_RGB888_1X24;
	93	+ u32 reg;
	94	+
	95	+ reg = readl(mxsfb->base + LCDC_CTRL);
	96	+
	97	+ if (mxsfb->connector.display_info.num_bus_formats)
	98	+ bus_format = mxsfb->connector.display_info.bus_formats[0];
	99	+
	100	+ reg &= ~CTRL_BUS_WIDTH_MASK;
	101	+ switch (bus_format) {
	102	+ case MEDIA_BUS_FMT_RGB565_1X16:
	103	+ reg \|= CTRL_SET_BUS_WIDTH(STMLCDIF_16BIT);
	104	+ break;
	105	+ case MEDIA_BUS_FMT_RGB666_1X18:
	106	+ reg \|= CTRL_SET_BUS_WIDTH(STMLCDIF_18BIT);
	107	+ break;
	108	+ case MEDIA_BUS_FMT_RGB888_1X24:
	109	+ reg \|= CTRL_SET_BUS_WIDTH(STMLCDIF_24BIT);
	110	+ break;
	111	+ default:
	112	+ dev_err(drm->dev, "Unknown media bus format %d\n", bus_format);
	113	+ break;
	114	+ }
	115	+ writel(reg, mxsfb->base + LCDC_CTRL);
	116	+}
	117	+
90	118	static void mxsfb_enable_controller(struct mxsfb_drm_private *mxsfb)
91	119	{
92	120	u32 reg;
@@ -168,13 +196,22 @@ static void mxsfb_crtc_mode_set_nofb(struct mxsfb_drm_private *mxsfb) · 75 unmodified lines
168	196	vdctrl0 \|= VDCTRL0_HSYNC_ACT_HIGH;
169	197	if (m->flags & DRM_MODE_FLAG_PVSYNC)
170	198	vdctrl0 \|= VDCTRL0_VSYNC_ACT_HIGH;
171		− if (bus_flags & DRM_BUS_FLAG_DE_HIGH)
	199	+ /* Make sure Data Enable is high active by default */
	200	+ if (!(bus_flags & DRM_BUS_FLAG_DE_LOW))
172	201	vdctrl0 \|= VDCTRL0_ENABLE_ACT_HIGH;
173		− if (bus_flags & DRM_BUS_FLAG_PIXDATA_NEGEDGE)
	202	+ /*
	203	+ * DRM_BUS_FLAG_PIXDATA_ defines are controller centric,
	204	+ * controllers VDCTRL0_DOTCLK is display centric.
	205	+ * Drive on positive edge -> display samples on falling edge
	206	+ * DRM_BUS_FLAG_PIXDATA_POSEDGE -> VDCTRL0_DOTCLK_ACT_FALLING
	207	+ */
	208	+ if (bus_flags & DRM_BUS_FLAG_PIXDATA_POSEDGE)
174	209	vdctrl0 \|= VDCTRL0_DOTCLK_ACT_FALLING;
175	210
176	211	writel(vdctrl0, mxsfb->base + LCDC_VDCTRL0);
177	212
	213	+ mxsfb_set_bus_fmt(mxsfb);
	214	+
178	215	/* Frame length in lines. */
179	216	writel(m->crtc_vtotal, mxsfb->base + LCDC_VDCTRL1);
180	217
@@ -184,8 +221,8 @@ static void mxsfb_crtc_mode_set_nofb(struct mxsfb_drm_private *mxsfb) · 3 unmodified lines
184	221	VDCTRL2_SET_HSYNC_PERIOD(m->crtc_htotal),
185	222	mxsfb->base + LCDC_VDCTRL2);
186	223
187		− writel(SET_HOR_WAIT_CNT(m->crtc_hblank_end - m->crtc_hsync_end) \|
188		− SET_VERT_WAIT_CNT(m->crtc_vblank_end - m->crtc_vsync_end),
	224	+ writel(SET_HOR_WAIT_CNT(m->crtc_htotal - m->crtc_hsync_start) \|
	225	+ SET_VERT_WAIT_CNT(m->crtc_vtotal - m->crtc_vsync_start),
189	226	mxsfb->base + LCDC_VDCTRL3);
190	227
191	228	writel(SET_DOTCLK_H_VALID_DATA_CNT(m->hdisplay),

drivers/gpu/drm/mxsfb/mxsfb_drv.c+4−0

@@ -102,14 +102,18 @@ static void mxsfb_pipe_enable(struct drm_simple_display_pipe *pipe, · 101 unmodified lines
102	102	{
103	103	struct mxsfb_drm_private *mxsfb = drm_pipe_to_mxsfb_drm_private(pipe);
104	104
	105	+ drm_panel_prepare(mxsfb->panel);
105	106	mxsfb_crtc_enable(mxsfb);
	107	+ drm_panel_enable(mxsfb->panel);
106	108	}
107	109
108	110	static void mxsfb_pipe_disable(struct drm_simple_display_pipe *pipe)
109	111	{
110	112	struct mxsfb_drm_private *mxsfb = drm_pipe_to_mxsfb_drm_private(pipe);
111	113
	114	+ drm_panel_disable(mxsfb->panel);
112	115	mxsfb_crtc_disable(mxsfb);
	116	+ drm_panel_unprepare(mxsfb->panel);
113	117	}
114	118
115	119	static void mxsfb_pipe_update(struct drm_simple_display_pipe *pipe,

drivers/gpu/drm/mxsfb/mxsfb_out.c+4−0

@@ -112,6 +112,7 @@ static int mxsfb_attach_endpoint(struct drm_device *drm, · 111 unmodified lines
112	112
113	113	int mxsfb_create_output(struct drm_device *drm)
114	114	{
	115	+ struct mxsfb_drm_private *mxsfb = drm->dev_private;
115	116	struct device_node *ep_np = NULL;
116	117	struct of_endpoint ep;
117	118	int ret;
@@ -127,5 +128,8 @@ int mxsfb_create_output(struct drm_device *drm) · 9 unmodified lines
127	128	}
128	129	}
129	130
	131	+ if (!mxsfb->panel)
	132	+ return -EPROBE_DEFER;
	133	+
130	134	return 0;
131	135	}

drivers/gpu/drm/mxsfb/mxsfb_regs.h+1−0

@@ -44,6 +44,7 @@ · 43 unmodified lines
44	44	#define CTRL_DATA_SELECT (1 << 16)
45	45	#define CTRL_SET_BUS_WIDTH(x) (((x) & 0x3) << 10)
46	46	#define CTRL_GET_BUS_WIDTH(x) (((x) >> 10) & 0x3)
	47	+#define CTRL_BUS_WIDTH_MASK (0x3 << 10)
47	48	#define CTRL_SET_WORD_LENGTH(x) (((x) & 0x3) << 8)
48	49	#define CTRL_GET_WORD_LENGTH(x) (((x) >> 8) & 0x3)
49	50	#define CTRL_MASTER (1 << 5)

drivers/gpu/drm/rcar-du/rcar_du_vsp.c+6−2

@@ -32,6 +32,10 @@ void rcar_du_vsp_enable(struct rcar_du_crtc *crtc) · 31 unmodified lines
32	32	{
33	33	const struct drm_display_mode *mode = &crtc->crtc.state->adjusted_mode;
34	34	struct rcar_du_device *rcdu = crtc->group->dev;
	35	+ struct vsp1_du_lif_config cfg = {
	36	+ .width = mode->hdisplay,
	37	+ .height = mode->vdisplay,
	38	+ };
35	39	struct rcar_du_plane_state state = {
36	40	.state = {
37	41	.crtc = &crtc->crtc,
@@ -66,12 +70,12 @@ void rcar_du_vsp_enable(struct rcar_du_crtc *crtc) · 28 unmodified lines
66	70	*/
67	71	crtc->group->need_restart = true;
68	72
69		− vsp1_du_setup_lif(crtc->vsp->vsp, mode->hdisplay, mode->vdisplay);
	73	+ vsp1_du_setup_lif(crtc->vsp->vsp, &cfg);
70	74	}
71	75
72	76	void rcar_du_vsp_disable(struct rcar_du_crtc *crtc)
73	77	{
74		− vsp1_du_setup_lif(crtc->vsp->vsp, 0, 0);
	78	+ vsp1_du_setup_lif(crtc->vsp->vsp, NULL);
75	79	}
76	80
77	81	void rcar_du_vsp_atomic_begin(struct rcar_du_crtc *crtc)

drivers/hv/channel.c+1−1

@@ -543,7 +543,7 @@ static int vmbus_close_internal(struct vmbus_channel *channel) · 542 unmodified lines
543	543	/*
544	544	* In case a device driver's probe() fails (e.g.,
545	545	* util_probe() -> vmbus_open() returns -ENOMEM) and the device is
546		− * rescinded later (e.g., we dynamically disble an Integrated Service
	546	+ * rescinded later (e.g., we dynamically disable an Integrated Service
547	547	* in Hyper-V Manager), the driver's remove() invokes vmbus_close():
548	548	* here we should skip most of the below cleanup work.
549	549	*/

drivers/i2c/busses/i2c-brcmstb.c+21−6

@@ -465,6 +465,7 @@ static int brcmstb_i2c_xfer(struct i2c_adapter *adapter, · 464 unmodified lines
465	465	u8 *tmp_buf;
466	466	int len = 0;
467	467	int xfersz = brcmstb_i2c_get_xfersz(dev);
	468	+ u32 cond, cond_per_msg;
468	469
469	470	if (dev->is_suspended)
470	471	return -EBUSY;
@@ -481,10 +482,11 @@ static int brcmstb_i2c_xfer(struct i2c_adapter *adapter, · 10 unmodified lines
481	482	pmsg->buf ? pmsg->buf[0] : '0', pmsg->len);
482	483
483	484	if (i < (num - 1) && (msgs[i + 1].flags & I2C_M_NOSTART))
484		− brcmstb_set_i2c_start_stop(dev, ~(COND_START_STOP));
	485	+ cond = ~COND_START_STOP;
485	486	else
486		− brcmstb_set_i2c_start_stop(dev,
487		− COND_RESTART \| COND_NOSTOP);
	487	+ cond = COND_RESTART \| COND_NOSTOP;
	488	+
	489	+ brcmstb_set_i2c_start_stop(dev, cond);
488	490
489	491	/* Send slave address */
490	492	if (!(pmsg->flags & I2C_M_NOSTART)) {
@@ -497,13 +499,24 @@ static int brcmstb_i2c_xfer(struct i2c_adapter *adapter, · 6 unmodified lines
497	499	}
498	500	}
499	501
	502	+ cond_per_msg = cond;
	503	+
500	504	/* Perform data transfer */
501	505	while (len) {
502	506	bytes_to_xfer = min(len, xfersz);
503	507
504		− if (len <= xfersz && i == (num - 1))
505		− brcmstb_set_i2c_start_stop(dev,
506		− ~(COND_START_STOP));
	508	+ if (len <= xfersz) {
	509	+ if (i == (num - 1))
	510	+ cond_per_msg = cond_per_msg &
	511	+ ~(COND_RESTART \| COND_NOSTOP);
	512	+ else
	513	+ cond_per_msg = cond;
	514	+ } else {
	515	+ cond_per_msg = (cond_per_msg & ~COND_RESTART) \|
	516	+ COND_NOSTOP;
	517	+ }
	518	+
	519	+ brcmstb_set_i2c_start_stop(dev, cond_per_msg);
507	520
508	521	rc = brcmstb_i2c_xfer_bsc_data(dev, tmp_buf,
509	522	bytes_to_xfer, pmsg);
@@ -512,6 +525,8 @@ static int brcmstb_i2c_xfer(struct i2c_adapter *adapter, · 2 unmodified lines
512	525
513	526	len -= bytes_to_xfer;
514	527	tmp_buf += bytes_to_xfer;
	528	+
	529	+ cond_per_msg = COND_NOSTART \| COND_NOSTOP;
515	530	}
516	531	}
517	532

drivers/i2c/busses/i2c-designware-core.h+1−0

@@ -88,6 +88,7 @@ struct dw_i2c_dev { · 87 unmodified lines
88	88	void __iomem *base;
89	89	struct completion cmd_complete;
90	90	struct clk *clk;
	91	+ struct reset_control *rst;
91	92	u32 (get_clk_rate_khz) (struct dw_i2c_dev dev);
92	93	struct dw_pci_controller *controller;
93	94	int cmd_err;

drivers/i2c/busses/i2c-designware-platdrv.c+24−4

@@ -38,6 +38,7 @@ · 37 unmodified lines
38	38	#include <linux/pm_runtime.h>
39	39	#include <linux/property.h>
40	40	#include <linux/io.h>
	41	+#include <linux/reset.h>
41	42	#include <linux/slab.h>
42	43	#include <linux/acpi.h>
43	44	#include <linux/platform_data/i2c-designware.h>
@@ -199,6 +200,14 @@ static int dw_i2c_plat_probe(struct platform_device *pdev) · 155 unmodified lines
199	200	dev->irq = irq;
200	201	platform_set_drvdata(pdev, dev);
201	202
	203	+ dev->rst = devm_reset_control_get_optional_exclusive(&pdev->dev, NULL);
	204	+ if (IS_ERR(dev->rst)) {
	205	+ if (PTR_ERR(dev->rst) == -EPROBE_DEFER)
	206	+ return -EPROBE_DEFER;
	207	+ } else {
	208	+ reset_control_deassert(dev->rst);
	209	+ }
	210	+
202	211	if (pdata) {
203	212	dev->clk_freq = pdata->i2c_scl_freq;
204	213	} else {
@@ -235,12 +244,13 @@ static int dw_i2c_plat_probe(struct platform_device *pdev) · 30 unmodified lines
235	244	&& dev->clk_freq != 1000000 && dev->clk_freq != 3400000) {
236	245	dev_err(&pdev->dev,
237	246	"Only 100kHz, 400kHz, 1MHz and 3.4MHz supported");
238		− return -EINVAL;
	247	+ r = -EINVAL;
	248	+ goto exit_reset;
239	249	}
240	250
241	251	r = i2c_dw_eval_lock_support(dev);
242	252	if (r)
243		− return r;
	253	+ goto exit_reset;
244	254
245	255	dev->functionality = I2C_FUNC_10BIT_ADDR \| DW_IC_DEFAULT_FUNCTIONALITY;
246	256
@@ -286,10 +296,18 @@ static int dw_i2c_plat_probe(struct platform_device *pdev) · 39 unmodified lines
286	296	}
287	297
288	298	r = i2c_dw_probe(dev);
289		− if (r && !dev->pm_runtime_disabled)
290		− pm_runtime_disable(&pdev->dev);
	299	+ if (r)
	300	+ goto exit_probe;
291	301
292	302	return r;
	303	+
	304	+exit_probe:
	305	+ if (!dev->pm_runtime_disabled)
	306	+ pm_runtime_disable(&pdev->dev);
	307	+exit_reset:
	308	+ if (!IS_ERR_OR_NULL(dev->rst))
	309	+ reset_control_assert(dev->rst);
	310	+ return r;
293	311	}
294	312
295	313	static int dw_i2c_plat_remove(struct platform_device *pdev)
@@ -306,6 +324,8 @@ static int dw_i2c_plat_remove(struct platform_device *pdev) · 10 unmodified lines
306	324	pm_runtime_put_sync(&pdev->dev);
307	325	if (!dev->pm_runtime_disabled)
308	326	pm_runtime_disable(&pdev->dev);
	327	+ if (!IS_ERR_OR_NULL(dev->rst))
	328	+ reset_control_assert(dev->rst);
309	329
310	330	return 0;
311	331	}

drivers/i2c/busses/i2c-exynos5.c+2−1

@@ -457,7 +457,6 @@ static irqreturn_t exynos5_i2c_irq(int irqno, void *dev_id) · 456 unmodified lines
457	457
458	458	int_status = readl(i2c->regs + HSI2C_INT_STATUS);
459	459	writel(int_status, i2c->regs + HSI2C_INT_STATUS);
460		− trans_status = readl(i2c->regs + HSI2C_TRANS_STATUS);
461	460
462	461	/* handle interrupt related to the transfer status */
463	462	if (i2c->variant->hw == HSI2C_EXYNOS7) {
@@ -482,11 +481,13 @@ static irqreturn_t exynos5_i2c_irq(int irqno, void *dev_id) · 18 unmodified lines
482	481	goto stop;
483	482	}
484	483
	484	+ trans_status = readl(i2c->regs + HSI2C_TRANS_STATUS);
485	485	if ((trans_status & HSI2C_MASTER_ST_MASK) == HSI2C_MASTER_ST_LOSE) {
486	486	i2c->state = -EAGAIN;
487	487	goto stop;
488	488	}
489	489	} else if (int_status & HSI2C_INT_I2C) {
	490	+ trans_status = readl(i2c->regs + HSI2C_TRANS_STATUS);
490	491	if (trans_status & HSI2C_NO_DEV_ACK) {
491	492	dev_dbg(i2c->dev, "No ACK from device\n");
492	493	i2c->state = -ENXIO;

drivers/i2c/busses/i2c-meson.c+1−1

@@ -175,7 +175,7 @@ static void meson_i2c_put_data(struct meson_i2c i2c, char buf, int len) · 174 unmodified lines
175	175	wdata1 \|= buf++ << ((i - 4) 8);
176	176
177	177	writel(wdata0, i2c->regs + REG_TOK_WDATA0);
178		− writel(wdata0, i2c->regs + REG_TOK_WDATA1);
	178	+ writel(wdata1, i2c->regs + REG_TOK_WDATA1);
179	179
180	180	dev_dbg(i2c->dev, "%s: data %08x %08x len %d\n", __func__,
181	181	wdata0, wdata1, len);

drivers/i2c/busses/i2c-mt65xx.c+0−9

@@ -172,14 +172,6 @@ static const struct i2c_adapter_quirks mt6577_i2c_quirks = { · 171 unmodified lines
172	172	.max_comb_2nd_msg_len = 31,
173	173	};
174	174
175		−static const struct i2c_adapter_quirks mt8173_i2c_quirks = {
176		− .max_num_msgs = 65535,
177		− .max_write_len = 65535,
178		− .max_read_len = 65535,
179		− .max_comb_1st_msg_len = 65535,
180		− .max_comb_2nd_msg_len = 65535,
181		−};
182		−
183	175	static const struct mtk_i2c_compatible mt6577_compat = {
184	176	.quirks = &mt6577_i2c_quirks,
185	177	.pmic_i2c = 0,
@@ -199,7 +191,6 @@ static const struct mtk_i2c_compatible mt6589_compat = { · 13 unmodified lines
199	191	};
200	192
201	193	static const struct mtk_i2c_compatible mt8173_compat = {
202		− .quirks = &mt8173_i2c_quirks,
203	194	.pmic_i2c = 0,
204	195	.dcm = 1,
205	196	.auto_restart = 1,

drivers/i2c/busses/i2c-riic.c+5−1

@@ -218,8 +218,12 @@ static irqreturn_t riic_tend_isr(int irq, void *data) · 217 unmodified lines
218	218	}
219	219
220	220	if (riic->is_last \|\| riic->err) {
221		− riic_clear_set_bit(riic, 0, ICIER_SPIE, RIIC_ICIER);
	221	+ riic_clear_set_bit(riic, ICIER_TEIE, ICIER_SPIE, RIIC_ICIER);
222	222	writeb(ICCR2_SP, riic->base + RIIC_ICCR2);
	223	+ } else {
	224	+ /* Transfer is complete, but do not send STOP */
	225	+ riic_clear_set_bit(riic, ICIER_TEIE, 0, RIIC_ICIER);
	226	+ complete(&riic->msg_done);
223	227	}
224	228
225	229	return IRQ_HANDLED;

drivers/i2c/i2c-mux.c+2−0

@@ -429,6 +429,7 @@ void i2c_mux_del_adapters(struct i2c_mux_core *muxc) · 428 unmodified lines
429	429	while (muxc->num_adapters) {
430	430	struct i2c_adapter *adap = muxc->adapter[--muxc->num_adapters];
431	431	struct i2c_mux_priv *priv = adap->algo_data;
	432	+ struct device_node *np = adap->dev.of_node;
432	433
433	434	muxc->adapter[muxc->num_adapters] = NULL;
434	435
@@ -438,6 +439,7 @@ void i2c_mux_del_adapters(struct i2c_mux_core *muxc) · 3 unmodified lines
438	439
439	440	sysfs_remove_link(&priv->adap.dev.kobj, "mux_device");
440	441	i2c_del_adapter(adap);
	442	+ of_node_put(np);
441	443	kfree(priv);
442	444	}
443	445	}

drivers/irqchip/irq-crossbar.c+5−4

@@ -198,7 +198,8 @@ static const struct irq_domain_ops crossbar_domain_ops = { · 197 unmodified lines
198	198
199	199	static int __init crossbar_of_init(struct device_node *node)
200	200	{
201		− int i, size, max = 0, reserved = 0, entry;
	201	+ u32 max = 0, entry, reg_size;
	202	+ int i, size, reserved = 0;
202	203	const __be32 *irqsr;
203	204	int ret = -ENOMEM;
204	205
@@ -275,9 +276,9 @@ static int __init crossbar_of_init(struct device_node *node) · 70 unmodified lines
275	276	if (!cb->register_offsets)
276	277	goto err_irq_map;
277	278
278		− of_property_read_u32(node, "ti,reg-size", &size);
	279	+ of_property_read_u32(node, "ti,reg-size", &reg_size);
279	280
280		− switch (size) {
	281	+ switch (reg_size) {
281	282	case 1:
282	283	cb->write = crossbar_writeb;
283	284	break;
@@ -303,7 +304,7 @@ static int __init crossbar_of_init(struct device_node *node) · 19 unmodified lines
303	304	continue;
304	305
305	306	cb->register_offsets[i] = reserved;
306		− reserved += size;
	307	+ reserved += reg_size;
307	308	}
308	309
309	310	of_property_read_u32(node, "ti,irqs-safe-map", &cb->safe_map);

drivers/irqchip/irq-gic-v3-its.c+16−0

@@ -1601,6 +1601,14 @@ static void __maybe_unused its_enable_quirk_cavium_23144(void *data) · 1600 unmodified lines
1601	1601	its->flags \|= ITS_FLAGS_WORKAROUND_CAVIUM_23144;
1602	1602	}
1603	1603
	1604	+static void __maybe_unused its_enable_quirk_qdf2400_e0065(void *data)
	1605	+{
	1606	+ struct its_node *its = data;
	1607	+
	1608	+ /* On QDF2400, the size of the ITE is 16Bytes */
	1609	+ its->ite_size = 16;
	1610	+}
	1611	+
1604	1612	static const struct gic_quirk its_quirks[] = {
1605	1613	#ifdef CONFIG_CAVIUM_ERRATUM_22375
1606	1614	{
@@ -1617,6 +1625,14 @@ static const struct gic_quirk its_quirks[] = { · 10 unmodified lines
1617	1625	.mask = 0xffff0fff,
1618	1626	.init = its_enable_quirk_cavium_23144,
1619	1627	},
	1628	+#endif
	1629	+#ifdef CONFIG_QCOM_QDF2400_ERRATUM_0065
	1630	+ {
	1631	+ .desc = "ITS: QDF2400 erratum 0065",
	1632	+ .iidr = 0x00001070, /* QDF2400 ITS rev 1.x */
	1633	+ .mask = 0xffffffff,
	1634	+ .init = its_enable_quirk_qdf2400_e0065,
	1635	+ },
1620	1636	#endif
1621	1637	{
1622	1638	}

drivers/isdn/gigaset/bas-gigaset.c+3−0

@@ -2317,6 +2317,9 @@ static int gigaset_probe(struct usb_interface *interface, · 2316 unmodified lines
2317	2317	return -ENODEV;
2318	2318	}
2319	2319
	2320	+ if (hostif->desc.bNumEndpoints < 1)
	2321	+ return -ENODEV;
	2322	+
2320	2323	dev_info(&udev->dev,
2321	2324	"%s: Device matched (Vendor: 0x%x, Product: 0x%x)\n",
2322	2325	__func__, le16_to_cpu(udev->descriptor.idVendor),

drivers/isdn/hisax/st5481_b.c+1−1

@@ -239,7 +239,7 @@ static void st5481B_mode(struct st5481_bcs *bcs, int mode) · 238 unmodified lines
239	239	}
240	240	}
241	241	} else {
242		− // Disble B channel interrupts
	242	+ // Disable B channel interrupts
243	243	st5481_usb_device_ctrl_msg(adapter, FFMSK_B1+(bcs->channel * 2), 0, NULL, NULL);
244	244
245	245	// Disable B channel FIFOs

drivers/macintosh/macio_asic.c+1−0

@@ -392,6 +392,7 @@ static struct macio_dev * macio_add_one_device(struct macio_chip *chip, · 391 unmodified lines
392	392	* To get all the fields, copy all archdata
393	393	*/
394	394	dev->ofdev.dev.archdata = chip->lbus.pdev->dev.archdata;
	395	+ dev->ofdev.dev.dma_ops = chip->lbus.pdev->dev.dma_ops;
395	396	#endif /* CONFIG_PCI */
396	397
397	398	#ifdef DEBUG

drivers/md/bcache/util.h+0−1

@@ -4,7 +4,6 @@ · 3 unmodified lines
4	4
5	5	#include <linux/blkdev.h>
6	6	#include <linux/errno.h>
7		−#include <linux/blkdev.h>
8	7	#include <linux/kernel.h>
9	8	#include <linux/sched/clock.h>
10	9	#include <linux/llist.h>

drivers/md/dm.c+16−13

@@ -989,26 +989,29 @@ static void flush_current_bio_list(struct blk_plug_cb *cb, bool from_schedule) · 988 unmodified lines
989	989	struct dm_offload *o = container_of(cb, struct dm_offload, cb);
990	990	struct bio_list list;
991	991	struct bio *bio;
	992	+ int i;
992	993
993	994	INIT_LIST_HEAD(&o->cb.list);
994	995
995	996	if (unlikely(!current->bio_list))
996	997	return;
997	998
998		− list = *current->bio_list;
999		− bio_list_init(current->bio_list);
1000		−
1001		− while ((bio = bio_list_pop(&list))) {
1002		− struct bio_set *bs = bio->bi_pool;
1003		− if (unlikely(!bs) \|\| bs == fs_bio_set) {
1004		− bio_list_add(current->bio_list, bio);
1005		− continue;
	999	+ for (i = 0; i < 2; i++) {
	1000	+ list = current->bio_list[i];
	1001	+ bio_list_init(&current->bio_list[i]);
	1002	+
	1003	+ while ((bio = bio_list_pop(&list))) {
	1004	+ struct bio_set *bs = bio->bi_pool;
	1005	+ if (unlikely(!bs) \|\| bs == fs_bio_set) {
	1006	+ bio_list_add(&current->bio_list[i], bio);
	1007	+ continue;
	1008	+ }
	1009	+
	1010	+ spin_lock(&bs->rescue_lock);
	1011	+ bio_list_add(&bs->rescue_list, bio);
	1012	+ queue_work(bs->rescue_workqueue, &bs->rescue_work);
	1013	+ spin_unlock(&bs->rescue_lock);
1006	1014	}
1007		−
1008		− spin_lock(&bs->rescue_lock);
1009		− bio_list_add(&bs->rescue_list, bio);
1010		− queue_work(bs->rescue_workqueue, &bs->rescue_work);
1011		− spin_unlock(&bs->rescue_lock);
1012	1015	}
1013	1016	}
1014	1017

drivers/md/raid10.c+2−1

@@ -974,7 +974,8 @@ static void wait_barrier(struct r10conf *conf) · 973 unmodified lines
974	974	!conf->barrier \|\|
975	975	(atomic_read(&conf->nr_pending) &&
976	976	current->bio_list &&
977		− !bio_list_empty(current->bio_list)),
	977	+ (!bio_list_empty(&current->bio_list[0]) \|\|
	978	+ !bio_list_empty(&current->bio_list[1]))),
978	979	conf->resync_lock);
979	980	conf->nr_waiting--;
980	981	if (!conf->nr_waiting)

drivers/media/dvb-frontends/drx39xyj/drx_driver.h+3−5

@@ -256,8 +256,7 @@ int drxbsp_tuner_default_i2c_write_read(struct tuner_instance *tuner, · 255 unmodified lines
256	256	*
257	257	* The actual DAP implementation may be restricted to only one of the modes.
258	258	* A compiler warning or error will be generated if the DAP implementation
259		−* overides or cannot handle the mode defined below.
260		−*
	259	+* overrides or cannot handle the mode defined below.
261	260	*/
262	261	#ifndef DRXDAP_SINGLE_MASTER
263	262	#define DRXDAP_SINGLE_MASTER 1
@@ -272,7 +271,7 @@ int drxbsp_tuner_default_i2c_write_read(struct tuner_instance *tuner, · 8 unmodified lines
272	271	*
273	272	* This maximum size may be restricted by the actual DAP implementation.
274	273	* A compiler warning or error will be generated if the DAP implementation
275		−* overides or cannot handle the chunksize defined below.
	274	+* overrides or cannot handle the chunksize defined below.
276	275	*
277	276	* Beware that the DAP uses DRXDAP_MAX_WCHUNKSIZE to create a temporary data
278	277	* buffer. Do not undefine or choose too large, unless your system is able to
@@ -292,8 +291,7 @@ int drxbsp_tuner_default_i2c_write_read(struct tuner_instance *tuner, · 13 unmodified lines
292	291	*
293	292	* This maximum size may be restricted by the actual DAP implementation.
294	293	* A compiler warning or error will be generated if the DAP implementation
295		−* overides or cannot handle the chunksize defined below.
296		−*
	294	+* overrides or cannot handle the chunksize defined below.
297	295	*/
298	296	#ifndef DRXDAP_MAX_RCHUNKSIZE
299	297	#define DRXDAP_MAX_RCHUNKSIZE 60

drivers/media/platform/vsp1/vsp1_drm.c+16−17

@@ -54,12 +54,11 @@ EXPORT_SYMBOL_GPL(vsp1_du_init); · 53 unmodified lines
54	54	/**
55	55	* vsp1_du_setup_lif - Setup the output part of the VSP pipeline
56	56	* @dev: the VSP device
57		− * @width: output frame width in pixels
58		− * @height: output frame height in pixels
	57	+ * @cfg: the LIF configuration
59	58	*
60		− * Configure the output part of VSP DRM pipeline for the given frame @width and
61		− * @height. This sets up formats on the BRU source pad, the WPF0 sink and source
62		− * pads, and the LIF sink pad.
	59	+ * Configure the output part of VSP DRM pipeline for the given frame @cfg.width
	60	+ * and @cfg.height. This sets up formats on the BRU source pad, the WPF0 sink
	61	+ * and source pads, and the LIF sink pad.
63	62	*
64	63	* As the media bus code on the BRU source pad is conditioned by the
65	64	* configuration of the BRU sink 0 pad, we also set up the formats on all BRU
@@ -69,8 +68,7 @@ EXPORT_SYMBOL_GPL(vsp1_du_init); · 3 unmodified lines
69	68	*
70	69	* Return 0 on success or a negative error code on failure.
71	70	*/
72		−int vsp1_du_setup_lif(struct device *dev, unsigned int width,
73		− unsigned int height)
	71	+int vsp1_du_setup_lif(struct device dev, const struct vsp1_du_lif_config cfg)
74	72	{
75	73	struct vsp1_device *vsp1 = dev_get_drvdata(dev);
76	74	struct vsp1_pipeline *pipe = &vsp1->drm->pipe;
@@ -79,11 +77,8 @@ int vsp1_du_setup_lif(struct device *dev, unsigned int width, · 2 unmodified lines
79	77	unsigned int i;
80	78	int ret;
81	79
82		− dev_dbg(vsp1->dev, "%s: configuring LIF with format %ux%u\n",
83		− __func__, width, height);
84		−
85		− if (width == 0 \|\| height == 0) {
86		− /* Zero width or height means the CRTC is being disabled, stop
	80	+ if (!cfg) {
	81	+ /* NULL configuration means the CRTC is being disabled, stop
87	82	* the pipeline and turn the light off.
88	83	*/
89	84	ret = vsp1_pipeline_stop(pipe);
@@ -108,6 +103,9 @@ int vsp1_du_setup_lif(struct device *dev, unsigned int width, · 18 unmodified lines
108	103	return 0;
109	104	}
110	105
	106	+ dev_dbg(vsp1->dev, "%s: configuring LIF with format %ux%u\n",
	107	+ __func__, cfg->width, cfg->height);
	108	+
111	109	/* Configure the format at the BRU sinks and propagate it through the
112	110	* pipeline.
113	111	*/
@@ -117,8 +115,8 @@ int vsp1_du_setup_lif(struct device *dev, unsigned int width, · 3 unmodified lines
117	115	for (i = 0; i < bru->entity.source_pad; ++i) {
118	116	format.pad = i;
119	117
120		− format.format.width = width;
121		− format.format.height = height;
	118	+ format.format.width = cfg->width;
	119	+ format.format.height = cfg->height;
122	120	format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32;
123	121	format.format.field = V4L2_FIELD_NONE;
124	122
@@ -133,8 +131,8 @@ int vsp1_du_setup_lif(struct device *dev, unsigned int width, · 8 unmodified lines
133	131	}
134	132
135	133	format.pad = bru->entity.source_pad;
136		− format.format.width = width;
137		− format.format.height = height;
	134	+ format.format.width = cfg->width;
	135	+ format.format.height = cfg->height;
138	136	format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32;
139	137	format.format.field = V4L2_FIELD_NONE;
140	138
@@ -180,7 +178,8 @@ int vsp1_du_setup_lif(struct device *dev, unsigned int width, · 39 unmodified lines
180	178	/* Verify that the format at the output of the pipeline matches the
181	179	* requested frame size and media bus code.
182	180	*/
183		− if (format.format.width != width \|\| format.format.height != height \|\|
	181	+ if (format.format.width != cfg->width \|\|
	182	+ format.format.height != cfg->height \|\|
184	183	format.format.code != MEDIA_BUS_FMT_ARGB8888_1X32) {
185	184	dev_dbg(vsp1->dev, "%s: format mismatch\n", __func__);
186	185	return -EPIPE;

drivers/media/rc/lirc_dev.c+2−2

@@ -436,6 +436,8 @@ int lirc_dev_fop_open(struct inode inode, struct file file) · 435 unmodified lines
436	436	return -ERESTARTSYS;
437	437
438	438	ir = irctls[iminor(inode)];
	439	+ mutex_unlock(&lirc_dev_lock);
	440	+
439	441	if (!ir) {
440	442	retval = -ENODEV;
441	443	goto error;
@@ -476,8 +478,6 @@ int lirc_dev_fop_open(struct inode inode, struct file file) · 34 unmodified lines
476	478	}
477	479
478	480	error:
479		− mutex_unlock(&lirc_dev_lock);
480		−
481	481	nonseekable_open(inode, file);
482	482
483	483	return retval;

drivers/media/rc/nuvoton-cir.c+3−2

@@ -176,12 +176,13 @@ static void nvt_write_wakeup_codes(struct rc_dev *dev, · 175 unmodified lines
176	176	{
177	177	u8 tolerance, config;
178	178	struct nvt_dev *nvt = dev->priv;
	179	+ unsigned long flags;
179	180	int i;
180	181
181	182	/* hardcode the tolerance to 10% */
182	183	tolerance = DIV_ROUND_UP(count, 10);
183	184
184		− spin_lock(&nvt->lock);
	185	+ spin_lock_irqsave(&nvt->lock, flags);
185	186
186	187	nvt_clear_cir_wake_fifo(nvt);
187	188	nvt_cir_wake_reg_write(nvt, count, CIR_WAKE_FIFO_CMP_DEEP);
@@ -203,7 +204,7 @@ static void nvt_write_wakeup_codes(struct rc_dev *dev, · 15 unmodified lines
203	204
204	205	nvt_cir_wake_reg_write(nvt, config, CIR_WAKE_IRCON);
205	206
206		− spin_unlock(&nvt->lock);
	207	+ spin_unlock_irqrestore(&nvt->lock, flags);
207	208	}
208	209
209	210	static ssize_t wakeup_data_show(struct device *dev,

drivers/media/rc/rc-main.c+15−11

@@ -1663,6 +1663,7 @@ static int rc_setup_rx_device(struct rc_dev *dev) · 1662 unmodified lines
1663	1663	{
1664	1664	int rc;
1665	1665	struct rc_map *rc_map;
	1666	+ u64 rc_type;
1666	1667
1667	1668	if (!dev->map_name)
1668	1669	return -EINVAL;
@@ -1677,15 +1678,18 @@ static int rc_setup_rx_device(struct rc_dev *dev) · 8 unmodified lines
1677	1678	if (rc)
1678	1679	return rc;
1679	1680
1680		− if (dev->change_protocol) {
1681		− u64 rc_type = (1ll << rc_map->rc_type);
	1681	+ rc_type = BIT_ULL(rc_map->rc_type);
1682	1682
	1683	+ if (dev->change_protocol) {
1683	1684	rc = dev->change_protocol(dev, &rc_type);
1684	1685	if (rc < 0)
1685	1686	goto out_table;
1686	1687	dev->enabled_protocols = rc_type;
1687	1688	}
1688	1689
	1690	+ if (dev->driver_type == RC_DRIVER_IR_RAW)
	1691	+ ir_raw_load_modules(&rc_type);
	1692	+
1689	1693	set_bit(EV_KEY, dev->input_dev->evbit);
1690	1694	set_bit(EV_REP, dev->input_dev->evbit);
1691	1695	set_bit(EV_MSC, dev->input_dev->evbit);
@@ -1777,12 +1781,6 @@ int rc_register_device(struct rc_dev *dev) · 85 unmodified lines
1777	1781	dev->input_name ?: "Unspecified device", path ?: "N/A");
1778	1782	kfree(path);
1779	1783
1780		− if (dev->driver_type != RC_DRIVER_IR_RAW_TX) {
1781		− rc = rc_setup_rx_device(dev);
1782		− if (rc)
1783		− goto out_dev;
1784		− }
1785		−
1786	1784	if (dev->driver_type == RC_DRIVER_IR_RAW \|\|
1787	1785	dev->driver_type == RC_DRIVER_IR_RAW_TX) {
1788	1786	if (!raw_init) {
@@ -1791,7 +1789,13 @@ int rc_register_device(struct rc_dev *dev) · 2 unmodified lines
1791	1789	}
1792	1790	rc = ir_raw_event_register(dev);
1793	1791	if (rc < 0)
1794		− goto out_rx;
	1792	+ goto out_dev;
	1793	+ }
	1794	+
	1795	+ if (dev->driver_type != RC_DRIVER_IR_RAW_TX) {
	1796	+ rc = rc_setup_rx_device(dev);
	1797	+ if (rc)
	1798	+ goto out_raw;
1795	1799	}
1796	1800
1797	1801	/* Allow the RC sysfs nodes to be accessible */
@@ -1803,8 +1807,8 @@ int rc_register_device(struct rc_dev *dev) · 5 unmodified lines
1803	1807
1804	1808	return 0;
1805	1809
1806		−out_rx:
1807		− rc_free_rx_device(dev);
	1810	+out_raw:
	1811	+ ir_raw_event_unregister(dev);
1808	1812	out_dev:
1809	1813	device_del(&dev->dev);
1810	1814	out_unlock:

drivers/media/rc/serial_ir.c+62−61

@@ -487,10 +487,69 @@ static void serial_ir_timeout(unsigned long arg) · 486 unmodified lines
487	487	ir_raw_event_handle(serial_ir.rcdev);
488	488	}
489	489
	490	+/* Needed by serial_ir_probe() */
	491	+static int serial_ir_tx(struct rc_dev dev, unsigned int txbuf,
	492	+ unsigned int count);
	493	+static int serial_ir_tx_duty_cycle(struct rc_dev *dev, u32 cycle);
	494	+static int serial_ir_tx_carrier(struct rc_dev *dev, u32 carrier);
	495	+static int serial_ir_open(struct rc_dev *rcdev);
	496	+static void serial_ir_close(struct rc_dev *rcdev);
	497	+
490	498	static int serial_ir_probe(struct platform_device *dev)
491	499	{
	500	+ struct rc_dev *rcdev;
492	501	int i, nlow, nhigh, result;
493	502
	503	+ rcdev = devm_rc_allocate_device(&dev->dev, RC_DRIVER_IR_RAW);
	504	+ if (!rcdev)
	505	+ return -ENOMEM;
	506	+
	507	+ if (hardware[type].send_pulse && hardware[type].send_space)
	508	+ rcdev->tx_ir = serial_ir_tx;
	509	+ if (hardware[type].set_send_carrier)
	510	+ rcdev->s_tx_carrier = serial_ir_tx_carrier;
	511	+ if (hardware[type].set_duty_cycle)
	512	+ rcdev->s_tx_duty_cycle = serial_ir_tx_duty_cycle;
	513	+
	514	+ switch (type) {
	515	+ case IR_HOMEBREW:
	516	+ rcdev->input_name = "Serial IR type home-brew";
	517	+ break;
	518	+ case IR_IRDEO:
	519	+ rcdev->input_name = "Serial IR type IRdeo";
	520	+ break;
	521	+ case IR_IRDEO_REMOTE:
	522	+ rcdev->input_name = "Serial IR type IRdeo remote";
	523	+ break;
	524	+ case IR_ANIMAX:
	525	+ rcdev->input_name = "Serial IR type AnimaX";
	526	+ break;
	527	+ case IR_IGOR:
	528	+ rcdev->input_name = "Serial IR type IgorPlug";
	529	+ break;
	530	+ }
	531	+
	532	+ rcdev->input_phys = KBUILD_MODNAME "/input0";
	533	+ rcdev->input_id.bustype = BUS_HOST;
	534	+ rcdev->input_id.vendor = 0x0001;
	535	+ rcdev->input_id.product = 0x0001;
	536	+ rcdev->input_id.version = 0x0100;
	537	+ rcdev->open = serial_ir_open;
	538	+ rcdev->close = serial_ir_close;
	539	+ rcdev->dev.parent = &serial_ir.pdev->dev;
	540	+ rcdev->allowed_protocols = RC_BIT_ALL_IR_DECODER;
	541	+ rcdev->driver_name = KBUILD_MODNAME;
	542	+ rcdev->map_name = RC_MAP_RC6_MCE;
	543	+ rcdev->min_timeout = 1;
	544	+ rcdev->timeout = IR_DEFAULT_TIMEOUT;
	545	+ rcdev->max_timeout = 10 * IR_DEFAULT_TIMEOUT;
	546	+ rcdev->rx_resolution = 250000;
	547	+
	548	+ serial_ir.rcdev = rcdev;
	549	+
	550	+ setup_timer(&serial_ir.timeout_timer, serial_ir_timeout,
	551	+ (unsigned long)&serial_ir);
	552	+
494	553	result = devm_request_irq(&dev->dev, irq, serial_ir_irq_handler,
495	554	share_irq ? IRQF_SHARED : 0,
496	555	KBUILD_MODNAME, &hardware);
@@ -516,9 +575,6 @@ static int serial_ir_probe(struct platform_device *dev) · 19 unmodified lines
516	575	return -EBUSY;
517	576	}
518	577
519		− setup_timer(&serial_ir.timeout_timer, serial_ir_timeout,
520		− (unsigned long)&serial_ir);
521		−
522	578	result = hardware_init_port();
523	579	if (result < 0)
524	580	return result;
@@ -552,7 +608,8 @@ static int serial_ir_probe(struct platform_device *dev) · 27 unmodified lines
552	608	sense ? "low" : "high");
553	609
554	610	dev_dbg(&dev->dev, "Interrupt %d, port %04x obtained\n", irq, io);
555		− return 0;
	611	+
	612	+ return devm_rc_register_device(&dev->dev, rcdev);
556	613	}
557	614
558	615	static int serial_ir_open(struct rc_dev *rcdev)
@@ -723,7 +780,6 @@ static void serial_ir_exit(void) · 164 unmodified lines
723	780
724	781	static int __init serial_ir_init_module(void)
725	782	{
726		− struct rc_dev *rcdev;
727	783	int result;
728	784
729	785	switch (type) {
@@ -754,63 +810,9 @@ static int __init serial_ir_init_module(void) · 24 unmodified lines
754	810	sense = !!sense;
755	811
756	812	result = serial_ir_init();
757		− if (result)
758		− return result;
759		−
760		− rcdev = devm_rc_allocate_device(&serial_ir.pdev->dev, RC_DRIVER_IR_RAW);
761		− if (!rcdev) {
762		− result = -ENOMEM;
763		− goto serial_cleanup;
764		− }
765		−
766		− if (hardware[type].send_pulse && hardware[type].send_space)
767		− rcdev->tx_ir = serial_ir_tx;
768		− if (hardware[type].set_send_carrier)
769		− rcdev->s_tx_carrier = serial_ir_tx_carrier;
770		− if (hardware[type].set_duty_cycle)
771		− rcdev->s_tx_duty_cycle = serial_ir_tx_duty_cycle;
772		−
773		− switch (type) {
774		− case IR_HOMEBREW:
775		− rcdev->input_name = "Serial IR type home-brew";
776		− break;
777		− case IR_IRDEO:
778		− rcdev->input_name = "Serial IR type IRdeo";
779		− break;
780		− case IR_IRDEO_REMOTE:
781		− rcdev->input_name = "Serial IR type IRdeo remote";
782		− break;
783		− case IR_ANIMAX:
784		− rcdev->input_name = "Serial IR type AnimaX";
785		− break;
786		− case IR_IGOR:
787		− rcdev->input_name = "Serial IR type IgorPlug";
788		− break;
789		− }
790		−
791		− rcdev->input_phys = KBUILD_MODNAME "/input0";
792		− rcdev->input_id.bustype = BUS_HOST;
793		− rcdev->input_id.vendor = 0x0001;
794		− rcdev->input_id.product = 0x0001;
795		− rcdev->input_id.version = 0x0100;
796		− rcdev->open = serial_ir_open;
797		− rcdev->close = serial_ir_close;
798		− rcdev->dev.parent = &serial_ir.pdev->dev;
799		− rcdev->allowed_protocols = RC_BIT_ALL_IR_DECODER;
800		− rcdev->driver_name = KBUILD_MODNAME;
801		− rcdev->map_name = RC_MAP_RC6_MCE;
802		− rcdev->min_timeout = 1;
803		− rcdev->timeout = IR_DEFAULT_TIMEOUT;
804		− rcdev->max_timeout = 10 * IR_DEFAULT_TIMEOUT;
805		− rcdev->rx_resolution = 250000;
806		−
807		− serial_ir.rcdev = rcdev;
808		−
809		− result = rc_register_device(rcdev);
810		−
811	813	if (!result)
812	814	return 0;
813		−serial_cleanup:
	815	+
814	816	serial_ir_exit();
815	817	return result;
816	818	}
@@ -818,7 +820,6 @@ serial_cleanup: · 1 unmodified line
818	820	static void __exit serial_ir_exit_module(void)
819	821	{
820	822	del_timer_sync(&serial_ir.timeout_timer);
821		− rc_unregister_device(serial_ir.rcdev);
822	823	serial_ir_exit();
823	824	}
824	825

drivers/media/usb/dvb-usb/dw2102.c+145−99

@@ -68,6 +68,7 @@ · 67 unmodified lines
68	68	struct dw2102_state {
69	69	u8 initialized;
70	70	u8 last_lock;
	71	+ u8 data[MAX_XFER_SIZE + 4];
71	72	struct i2c_client *i2c_client_demod;
72	73	struct i2c_client *i2c_client_tuner;
73	74
@@ -661,62 +662,72 @@ static int su3000_i2c_transfer(struct i2c_adapter *adap, struct i2c_msg msg[], · 587 unmodified lines
661	662	int num)
662	663	{
663	664	struct dvb_usb_device *d = i2c_get_adapdata(adap);
664		− u8 obuf[0x40], ibuf[0x40];
	665	+ struct dw2102_state *state;
665	666
666	667	if (!d)
667	668	return -ENODEV;
	669	+
	670	+ state = d->priv;
	671	+
668	672	if (mutex_lock_interruptible(&d->i2c_mutex) < 0)
669	673	return -EAGAIN;
	674	+ if (mutex_lock_interruptible(&d->data_mutex) < 0) {
	675	+ mutex_unlock(&d->i2c_mutex);
	676	+ return -EAGAIN;
	677	+ }
670	678
671	679	switch (num) {
672	680	case 1:
673	681	switch (msg[0].addr) {
674	682	case SU3000_STREAM_CTRL:
675		− obuf[0] = msg[0].buf[0] + 0x36;
676		− obuf[1] = 3;
677		− obuf[2] = 0;
678		− if (dvb_usb_generic_rw(d, obuf, 3, ibuf, 0, 0) < 0)
	683	+ state->data[0] = msg[0].buf[0] + 0x36;
	684	+ state->data[1] = 3;
	685	+ state->data[2] = 0;
	686	+ if (dvb_usb_generic_rw(d, state->data, 3,
	687	+ state->data, 0, 0) < 0)
679	688	err("i2c transfer failed.");
680	689	break;
681	690	case DW2102_RC_QUERY:
682		− obuf[0] = 0x10;
683		− if (dvb_usb_generic_rw(d, obuf, 1, ibuf, 2, 0) < 0)
	691	+ state->data[0] = 0x10;
	692	+ if (dvb_usb_generic_rw(d, state->data, 1,
	693	+ state->data, 2, 0) < 0)
684	694	err("i2c transfer failed.");
685		− msg[0].buf[1] = ibuf[0];
686		− msg[0].buf[0] = ibuf[1];
	695	+ msg[0].buf[1] = state->data[0];
	696	+ msg[0].buf[0] = state->data[1];
687	697	break;
688	698	default:
689	699	/* always i2c write*/
690		− obuf[0] = 0x08;
691		− obuf[1] = msg[0].addr;
692		− obuf[2] = msg[0].len;
	700	+ state->data[0] = 0x08;
	701	+ state->data[1] = msg[0].addr;
	702	+ state->data[2] = msg[0].len;
693	703
694		− memcpy(&obuf[3], msg[0].buf, msg[0].len);
	704	+ memcpy(&state->data[3], msg[0].buf, msg[0].len);
695	705
696		− if (dvb_usb_generic_rw(d, obuf, msg[0].len + 3,
697		− ibuf, 1, 0) < 0)
	706	+ if (dvb_usb_generic_rw(d, state->data, msg[0].len + 3,
	707	+ state->data, 1, 0) < 0)
698	708	err("i2c transfer failed.");
699	709
700	710	}
701	711	break;
702	712	case 2:
703	713	/* always i2c read */
704		− obuf[0] = 0x09;
705		− obuf[1] = msg[0].len;
706		− obuf[2] = msg[1].len;
707		− obuf[3] = msg[0].addr;
708		− memcpy(&obuf[4], msg[0].buf, msg[0].len);
709		−
710		− if (dvb_usb_generic_rw(d, obuf, msg[0].len + 4,
711		− ibuf, msg[1].len + 1, 0) < 0)
	714	+ state->data[0] = 0x09;
	715	+ state->data[1] = msg[0].len;
	716	+ state->data[2] = msg[1].len;
	717	+ state->data[3] = msg[0].addr;
	718	+ memcpy(&state->data[4], msg[0].buf, msg[0].len);
	719	+
	720	+ if (dvb_usb_generic_rw(d, state->data, msg[0].len + 4,
	721	+ state->data, msg[1].len + 1, 0) < 0)
712	722	err("i2c transfer failed.");
713	723
714		− memcpy(msg[1].buf, &ibuf[1], msg[1].len);
	724	+ memcpy(msg[1].buf, &state->data[1], msg[1].len);
715	725	break;
716	726	default:
717	727	warn("more than 2 i2c messages at a time is not handled yet.");
718	728	break;
719	729	}
	730	+ mutex_unlock(&d->data_mutex);
720	731	mutex_unlock(&d->i2c_mutex);
721	732	return num;
722	733	}
@@ -844,17 +855,23 @@ static int su3000_streaming_ctrl(struct dvb_usb_adapter *adap, int onoff) · 121 unmodified lines
844	855	static int su3000_power_ctrl(struct dvb_usb_device *d, int i)
845	856	{
846	857	struct dw2102_state state = (struct dw2102_state )d->priv;
847		− u8 obuf[] = {0xde, 0};
	858	+ int ret = 0;
848	859
849	860	info("%s: %d, initialized %d", __func__, i, state->initialized);
850	861
851	862	if (i && !state->initialized) {
	863	+ mutex_lock(&d->data_mutex);
	864	+
	865	+ state->data[0] = 0xde;
	866	+ state->data[1] = 0;
	867	+
852	868	state->initialized = 1;
853	869	/* reset board */
854		− return dvb_usb_generic_rw(d, obuf, 2, NULL, 0, 0);
	870	+ ret = dvb_usb_generic_rw(d, state->data, 2, NULL, 0, 0);
	871	+ mutex_unlock(&d->data_mutex);
855	872	}
856	873
857		− return 0;
	874	+ return ret;
858	875	}
859	876
860	877	static int su3000_read_mac_address(struct dvb_usb_device *d, u8 mac[6])
@@ -1309,49 +1326,57 @@ static int prof_7500_frontend_attach(struct dvb_usb_adapter *d) · 448 unmodified lines
1309	1326	return 0;
1310	1327	}
1311	1328
1312		−static int su3000_frontend_attach(struct dvb_usb_adapter *d)
	1329	+static int su3000_frontend_attach(struct dvb_usb_adapter *adap)
1313	1330	{
1314		− u8 obuf[3] = { 0xe, 0x80, 0 };
1315		− u8 ibuf[] = { 0 };
	1331	+ struct dvb_usb_device *d = adap->dev;
	1332	+ struct dw2102_state *state = d->priv;
	1333	+
	1334	+ mutex_lock(&d->data_mutex);
	1335	+
	1336	+ state->data[0] = 0xe;
	1337	+ state->data[1] = 0x80;
	1338	+ state->data[2] = 0;
1316	1339
1317		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1340	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1318	1341	err("command 0x0e transfer failed.");
1319	1342
1320		− obuf[0] = 0xe;
1321		− obuf[1] = 0x02;
1322		− obuf[2] = 1;
	1343	+ state->data[0] = 0xe;
	1344	+ state->data[1] = 0x02;
	1345	+ state->data[2] = 1;
1323	1346
1324		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1347	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1325	1348	err("command 0x0e transfer failed.");
1326	1349	msleep(300);
1327	1350
1328		− obuf[0] = 0xe;
1329		− obuf[1] = 0x83;
1330		− obuf[2] = 0;
	1351	+ state->data[0] = 0xe;
	1352	+ state->data[1] = 0x83;
	1353	+ state->data[2] = 0;
1331	1354
1332		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1355	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1333	1356	err("command 0x0e transfer failed.");
1334	1357
1335		− obuf[0] = 0xe;
1336		− obuf[1] = 0x83;
1337		− obuf[2] = 1;
	1358	+ state->data[0] = 0xe;
	1359	+ state->data[1] = 0x83;
	1360	+ state->data[2] = 1;
1338	1361
1339		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1362	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1340	1363	err("command 0x0e transfer failed.");
1341	1364
1342		− obuf[0] = 0x51;
	1365	+ state->data[0] = 0x51;
1343	1366
1344		− if (dvb_usb_generic_rw(d->dev, obuf, 1, ibuf, 1, 0) < 0)
	1367	+ if (dvb_usb_generic_rw(d, state->data, 1, state->data, 1, 0) < 0)
1345	1368	err("command 0x51 transfer failed.");
1346	1369
1347		− d->fe_adap[0].fe = dvb_attach(ds3000_attach, &su3000_ds3000_config,
1348		− &d->dev->i2c_adap);
1349		− if (d->fe_adap[0].fe == NULL)
	1370	+ mutex_unlock(&d->data_mutex);
	1371	+
	1372	+ adap->fe_adap[0].fe = dvb_attach(ds3000_attach, &su3000_ds3000_config,
	1373	+ &d->i2c_adap);
	1374	+ if (adap->fe_adap[0].fe == NULL)
1350	1375	return -EIO;
1351	1376
1352		− if (dvb_attach(ts2020_attach, d->fe_adap[0].fe,
	1377	+ if (dvb_attach(ts2020_attach, adap->fe_adap[0].fe,
1353	1378	&dw2104_ts2020_config,
1354		− &d->dev->i2c_adap)) {
	1379	+ &d->i2c_adap)) {
1355	1380	info("Attached DS3000/TS2020!");
1356	1381	return 0;
1357	1382	}
@@ -1360,47 +1385,55 @@ static int su3000_frontend_attach(struct dvb_usb_adapter *d) · 2 unmodified lines
1360	1385	return -EIO;
1361	1386	}
1362	1387
1363		−static int t220_frontend_attach(struct dvb_usb_adapter *d)
	1388	+static int t220_frontend_attach(struct dvb_usb_adapter *adap)
1364	1389	{
1365		− u8 obuf[3] = { 0xe, 0x87, 0 };
1366		− u8 ibuf[] = { 0 };
	1390	+ struct dvb_usb_device *d = adap->dev;
	1391	+ struct dw2102_state *state = d->priv;
	1392	+
	1393	+ mutex_lock(&d->data_mutex);
1367	1394
1368		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1395	+ state->data[0] = 0xe;
	1396	+ state->data[1] = 0x87;
	1397	+ state->data[2] = 0x0;
	1398	+
	1399	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1369	1400	err("command 0x0e transfer failed.");
1370	1401
1371		− obuf[0] = 0xe;
1372		− obuf[1] = 0x86;
1373		− obuf[2] = 1;
	1402	+ state->data[0] = 0xe;
	1403	+ state->data[1] = 0x86;
	1404	+ state->data[2] = 1;
1374	1405
1375		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1406	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1376	1407	err("command 0x0e transfer failed.");
1377	1408
1378		− obuf[0] = 0xe;
1379		− obuf[1] = 0x80;
1380		− obuf[2] = 0;
	1409	+ state->data[0] = 0xe;
	1410	+ state->data[1] = 0x80;
	1411	+ state->data[2] = 0;
1381	1412
1382		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1413	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1383	1414	err("command 0x0e transfer failed.");
1384	1415
1385	1416	msleep(50);
1386	1417
1387		− obuf[0] = 0xe;
1388		− obuf[1] = 0x80;
1389		− obuf[2] = 1;
	1418	+ state->data[0] = 0xe;
	1419	+ state->data[1] = 0x80;
	1420	+ state->data[2] = 1;
1390	1421
1391		− if (dvb_usb_generic_rw(d->dev, obuf, 3, ibuf, 1, 0) < 0)
	1422	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1392	1423	err("command 0x0e transfer failed.");
1393	1424
1394		− obuf[0] = 0x51;
	1425	+ state->data[0] = 0x51;
1395	1426
1396		− if (dvb_usb_generic_rw(d->dev, obuf, 1, ibuf, 1, 0) < 0)
	1427	+ if (dvb_usb_generic_rw(d, state->data, 1, state->data, 1, 0) < 0)
1397	1428	err("command 0x51 transfer failed.");
1398	1429
1399		− d->fe_adap[0].fe = dvb_attach(cxd2820r_attach, &cxd2820r_config,
1400		− &d->dev->i2c_adap, NULL);
1401		− if (d->fe_adap[0].fe != NULL) {
1402		− if (dvb_attach(tda18271_attach, d->fe_adap[0].fe, 0x60,
1403		− &d->dev->i2c_adap, &tda18271_config)) {
	1430	+ mutex_unlock(&d->data_mutex);
	1431	+
	1432	+ adap->fe_adap[0].fe = dvb_attach(cxd2820r_attach, &cxd2820r_config,
	1433	+ &d->i2c_adap, NULL);
	1434	+ if (adap->fe_adap[0].fe != NULL) {
	1435	+ if (dvb_attach(tda18271_attach, adap->fe_adap[0].fe, 0x60,
	1436	+ &d->i2c_adap, &tda18271_config)) {
1404	1437	info("Attached TDA18271HD/CXD2820R!");
1405	1438	return 0;
1406	1439	}
@@ -1410,23 +1443,30 @@ static int t220_frontend_attach(struct dvb_usb_adapter *d) · 3 unmodified lines
1410	1443	return -EIO;
1411	1444	}
1412	1445
1413		−static int m88rs2000_frontend_attach(struct dvb_usb_adapter *d)
	1446	+static int m88rs2000_frontend_attach(struct dvb_usb_adapter *adap)
1414	1447	{
1415		− u8 obuf[] = { 0x51 };
1416		− u8 ibuf[] = { 0 };
	1448	+ struct dvb_usb_device *d = adap->dev;
	1449	+ struct dw2102_state *state = d->priv;
	1450	+
	1451	+ mutex_lock(&d->data_mutex);
1417	1452
1418		− if (dvb_usb_generic_rw(d->dev, obuf, 1, ibuf, 1, 0) < 0)
	1453	+ state->data[0] = 0x51;
	1454	+
	1455	+ if (dvb_usb_generic_rw(d, state->data, 1, state->data, 1, 0) < 0)
1419	1456	err("command 0x51 transfer failed.");
1420	1457
1421		− d->fe_adap[0].fe = dvb_attach(m88rs2000_attach, &s421_m88rs2000_config,
1422		− &d->dev->i2c_adap);
	1458	+ mutex_unlock(&d->data_mutex);
1423	1459
1424		− if (d->fe_adap[0].fe == NULL)
	1460	+ adap->fe_adap[0].fe = dvb_attach(m88rs2000_attach,
	1461	+ &s421_m88rs2000_config,
	1462	+ &d->i2c_adap);
	1463	+
	1464	+ if (adap->fe_adap[0].fe == NULL)
1425	1465	return -EIO;
1426	1466
1427		− if (dvb_attach(ts2020_attach, d->fe_adap[0].fe,
	1467	+ if (dvb_attach(ts2020_attach, adap->fe_adap[0].fe,
1428	1468	&dw2104_ts2020_config,
1429		− &d->dev->i2c_adap)) {
	1469	+ &d->i2c_adap)) {
1430	1470	info("Attached RS2000/TS2020!");
1431	1471	return 0;
1432	1472	}
@@ -1439,44 +1479,50 @@ static int tt_s2_4600_frontend_attach(struct dvb_usb_adapter *adap) · 6 unmodified lines
1439	1479	{
1440	1480	struct dvb_usb_device *d = adap->dev;
1441	1481	struct dw2102_state *state = d->priv;
1442		− u8 obuf[3] = { 0xe, 0x80, 0 };
1443		− u8 ibuf[] = { 0 };
1444	1482	struct i2c_adapter *i2c_adapter;
1445	1483	struct i2c_client *client;
1446	1484	struct i2c_board_info board_info;
1447	1485	struct m88ds3103_platform_data m88ds3103_pdata = {};
1448	1486	struct ts2020_config ts2020_config = {};
1449	1487
1450		− if (dvb_usb_generic_rw(d, obuf, 3, ibuf, 1, 0) < 0)
	1488	+ mutex_lock(&d->data_mutex);
	1489	+
	1490	+ state->data[0] = 0xe;
	1491	+ state->data[1] = 0x80;
	1492	+ state->data[2] = 0x0;
	1493	+
	1494	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1451	1495	err("command 0x0e transfer failed.");
1452	1496
1453		− obuf[0] = 0xe;
1454		− obuf[1] = 0x02;
1455		− obuf[2] = 1;
	1497	+ state->data[0] = 0xe;
	1498	+ state->data[1] = 0x02;
	1499	+ state->data[2] = 1;
1456	1500
1457		− if (dvb_usb_generic_rw(d, obuf, 3, ibuf, 1, 0) < 0)
	1501	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1458	1502	err("command 0x0e transfer failed.");
1459	1503	msleep(300);
1460	1504
1461		− obuf[0] = 0xe;
1462		− obuf[1] = 0x83;
1463		− obuf[2] = 0;
	1505	+ state->data[0] = 0xe;
	1506	+ state->data[1] = 0x83;
	1507	+ state->data[2] = 0;
1464	1508
1465		− if (dvb_usb_generic_rw(d, obuf, 3, ibuf, 1, 0) < 0)
	1509	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1466	1510	err("command 0x0e transfer failed.");
1467	1511
1468		− obuf[0] = 0xe;
1469		− obuf[1] = 0x83;
1470		− obuf[2] = 1;
	1512	+ state->data[0] = 0xe;
	1513	+ state->data[1] = 0x83;
	1514	+ state->data[2] = 1;
1471	1515
1472		− if (dvb_usb_generic_rw(d, obuf, 3, ibuf, 1, 0) < 0)
	1516	+ if (dvb_usb_generic_rw(d, state->data, 3, state->data, 1, 0) < 0)
1473	1517	err("command 0x0e transfer failed.");
1474	1518
1475		− obuf[0] = 0x51;
	1519	+ state->data[0] = 0x51;
1476	1520
1477		− if (dvb_usb_generic_rw(d, obuf, 1, ibuf, 1, 0) < 0)
	1521	+ if (dvb_usb_generic_rw(d, state->data, 1, state->data, 1, 0) < 0)
1478	1522	err("command 0x51 transfer failed.");
1479	1523
	1524	+ mutex_unlock(&d->data_mutex);
	1525	+
1480	1526	/* attach demod */
1481	1527	m88ds3103_pdata.clk = 27000000;
1482	1528	m88ds3103_pdata.i2c_wr_max = 33;

drivers/misc/sgi-gru/grufault.c+7−2

@@ -219,15 +219,20 @@ static int atomic_pte_lookup(struct vm_area_struct *vma, unsigned long vaddr, · 218 unmodified lines
219	219	int write, unsigned long paddr, int pageshift)
220	220	{
221	221	pgd_t *pgdp;
222		− pmd_t *pmdp;
	222	+ p4d_t *p4dp;
223	223	pud_t *pudp;
	224	+ pmd_t *pmdp;
224	225	pte_t pte;
225	226
226	227	pgdp = pgd_offset(vma->vm_mm, vaddr);
227	228	if (unlikely(pgd_none(*pgdp)))
228	229	goto err;
229	230
230		− pudp = pud_offset(pgdp, vaddr);
	231	+ p4dp = p4d_offset(pgdp, vaddr);
	232	+ if (unlikely(p4d_none(*p4dp)))
	233	+ goto err;
	234	+
	235	+ pudp = pud_offset(p4dp, vaddr);
231	236	if (unlikely(pud_none(*pudp)))
232	237	goto err;
233	238

drivers/mtd/spi-nor/spi-nor.c+1−1

@@ -186,7 +186,7 @@ static inline int write_enable(struct spi_nor *nor) · 185 unmodified lines
186	186	}
187	187
188	188	/*
189		− * Send write disble instruction to the chip.
	189	+ * Send write disable instruction to the chip.
190	190	*/
191	191	static inline int write_disable(struct spi_nor *nor)
192	192	{

drivers/net/ethernet/amd/xgbe/xgbe-drv.c+2−8

@@ -2272,10 +2272,7 @@ static int xgbe_one_poll(struct napi_struct *napi, int budget) · 2271 unmodified lines
2272	2272	processed = xgbe_rx_poll(channel, budget);
2273	2273
2274	2274	/* If we processed everything, we are done */
2275		− if (processed < budget) {
2276		− /* Turn off polling */
2277		− napi_complete_done(napi, processed);
2278		−
	2275	+ if ((processed < budget) && napi_complete_done(napi, processed)) {
2279	2276	/* Enable Tx and Rx interrupts */
2280	2277	if (pdata->channel_irq_mode)
2281	2278	xgbe_enable_rx_tx_int(pdata, channel);
@@ -2317,10 +2314,7 @@ static int xgbe_all_poll(struct napi_struct *napi, int budget) · 35 unmodified lines
2317	2314	} while ((processed < budget) && (processed != last_processed));
2318	2315
2319	2316	/* If we processed everything, we are done */
2320		− if (processed < budget) {
2321		− /* Turn off polling */
2322		− napi_complete_done(napi, processed);
2323		−
	2317	+ if ((processed < budget) && napi_complete_done(napi, processed)) {
2324	2318	/* Enable Tx and Rx interrupts */
2325	2319	xgbe_enable_rx_tx_ints(pdata);
2326	2320	}

drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c+1−1

@@ -213,9 +213,9 @@ void aq_pci_func_free_irqs(struct aq_pci_func_s *self) · 212 unmodified lines
213	213	if (!((1U << i) & self->msix_entry_mask))
214	214	continue;
215	215
216		− free_irq(pci_irq_vector(pdev, i), self->aq_vec[i]);
217	216	if (pdev->msix_enabled)
218	217	irq_set_affinity_hint(pci_irq_vector(pdev, i), NULL);
	218	+ free_irq(pci_irq_vector(pdev, i), self->aq_vec[i]);
219	219	self->msix_entry_mask &= ~(1U << i);
220	220	}
221	221	}

drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c+30−6

@@ -13292,17 +13292,15 @@ static int bnx2x_init_dev(struct bnx2x bp, struct pci_dev pdev, · 13291 unmodified lines
13292	13292	dev->vlan_features = NETIF_F_SG \| NETIF_F_IP_CSUM \| NETIF_F_IPV6_CSUM \|
13293	13293	NETIF_F_TSO \| NETIF_F_TSO_ECN \| NETIF_F_TSO6 \| NETIF_F_HIGHDMA;
13294	13294
13295		− /* VF with OLD Hypervisor or old PF do not support filtering */
13296	13295	if (IS_PF(bp)) {
13297	13296	if (chip_is_e1x)
13298	13297	bp->accept_any_vlan = true;
13299	13298	else
13300	13299	dev->hw_features \|= NETIF_F_HW_VLAN_CTAG_FILTER;
13301		−#ifdef CONFIG_BNX2X_SRIOV
13302		− } else if (bp->acquire_resp.pfdev_info.pf_cap & PFVF_CAP_VLAN_FILTER) {
13303		− dev->hw_features \|= NETIF_F_HW_VLAN_CTAG_FILTER;
13304		−#endif
13305	13300	}
	13301	+ /* For VF we'll know whether to enable VLAN filtering after
	13302	+ * getting a response to CHANNEL_TLV_ACQUIRE from PF.
	13303	+ */
13306	13304
13307	13305	dev->features \|= dev->hw_features \| NETIF_F_HW_VLAN_CTAG_RX;
13308	13306	dev->features \|= NETIF_F_HIGHDMA;
@@ -13738,7 +13736,7 @@ static int bnx2x_ptp_adjfreq(struct ptp_clock_info *ptp, s32 ppb) · 429 unmodified lines
13738	13736	if (!netif_running(bp->dev)) {
13739	13737	DP(BNX2X_MSG_PTP,
13740	13738	"PTP adjfreq called while the interface is down\n");
13741		− return -EFAULT;
	13739	+ return -ENETDOWN;
13742	13740	}
13743	13741
13744	13742	if (ppb < 0) {
@@ -13797,6 +13795,12 @@ static int bnx2x_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta) · 52 unmodified lines
13797	13795	{
13798	13796	struct bnx2x *bp = container_of(ptp, struct bnx2x, ptp_clock_info);
13799	13797
	13798	+ if (!netif_running(bp->dev)) {
	13799	+ DP(BNX2X_MSG_PTP,
	13800	+ "PTP adjtime called while the interface is down\n");
	13801	+ return -ENETDOWN;
	13802	+ }
	13803	+
13800	13804	DP(BNX2X_MSG_PTP, "PTP adjtime called, delta = %llx\n", delta);
13801	13805
13802	13806	timecounter_adjtime(&bp->timecounter, delta);
@@ -13809,6 +13813,12 @@ static int bnx2x_ptp_gettime(struct ptp_clock_info ptp, struct timespec64 ts) · 6 unmodified lines
13809	13813	struct bnx2x *bp = container_of(ptp, struct bnx2x, ptp_clock_info);
13810	13814	u64 ns;
13811	13815
	13816	+ if (!netif_running(bp->dev)) {
	13817	+ DP(BNX2X_MSG_PTP,
	13818	+ "PTP gettime called while the interface is down\n");
	13819	+ return -ENETDOWN;
	13820	+ }
	13821	+
13812	13822	ns = timecounter_read(&bp->timecounter);
13813	13823
13814	13824	DP(BNX2X_MSG_PTP, "PTP gettime called, ns = %llu\n", ns);
@@ -13824,6 +13834,12 @@ static int bnx2x_ptp_settime(struct ptp_clock_info *ptp, · 9 unmodified lines
13824	13834	struct bnx2x *bp = container_of(ptp, struct bnx2x, ptp_clock_info);
13825	13835	u64 ns;
13826	13836
	13837	+ if (!netif_running(bp->dev)) {
	13838	+ DP(BNX2X_MSG_PTP,
	13839	+ "PTP settime called while the interface is down\n");
	13840	+ return -ENETDOWN;
	13841	+ }
	13842	+
13827	13843	ns = timespec64_to_ns(ts);
13828	13844
13829	13845	DP(BNX2X_MSG_PTP, "PTP settime called, ns = %llu\n", ns);
@@ -13991,6 +14007,14 @@ static int bnx2x_init_one(struct pci_dev *pdev, · 161 unmodified lines
13991	14007	rc = bnx2x_vfpf_acquire(bp, tx_count, rx_count);
13992	14008	if (rc)
13993	14009	goto init_one_freemem;
	14010	+
	14011	+#ifdef CONFIG_BNX2X_SRIOV
	14012	+ /* VF with OLD Hypervisor or old PF do not support filtering */
	14013	+ if (bp->acquire_resp.pfdev_info.pf_cap & PFVF_CAP_VLAN_FILTER) {
	14014	+ dev->hw_features \|= NETIF_F_HW_VLAN_CTAG_FILTER;
	14015	+ dev->features \|= NETIF_F_HW_VLAN_CTAG_FILTER;
	14016	+ }
	14017	+#endif
13994	14018	}
13995	14019
13996	14020	/* Enable SRIOV if capability found in configuration space */

drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.c+16−8

@@ -434,7 +434,9 @@ static int bnx2x_vf_mac_vlan_config(struct bnx2x *bp, · 433 unmodified lines
434	434
435	435	/* Add/Remove the filter */
436	436	rc = bnx2x_config_vlan_mac(bp, &ramrod);
437		− if (rc && rc != -EEXIST) {
	437	+ if (rc == -EEXIST)
	438	+ return 0;
	439	+ if (rc) {
438	440	BNX2X_ERR("Failed to %s %s\n",
439	441	filter->add ? "add" : "delete",
440	442	(filter->type == BNX2X_VF_FILTER_VLAN_MAC) ?
@@ -444,6 +446,8 @@ static int bnx2x_vf_mac_vlan_config(struct bnx2x *bp, · 3 unmodified lines
444	446	return rc;
445	447	}
446	448
	449	+ filter->applied = true;
	450	+
447	451	return 0;
448	452	}
449	453
@@ -469,8 +473,10 @@ int bnx2x_vf_mac_vlan_config_list(struct bnx2x bp, struct bnx2x_virtf vf, · 19 unmodified lines
469	473	/* Rollback if needed */
470	474	if (i != filters->count) {
471	475	BNX2X_ERR("Managed only %d/%d filters - rolling back\n",
472		− i, filters->count + 1);
	476	+ i, filters->count);
473	477	while (--i >= 0) {
	478	+ if (!filters->filters[i].applied)
	479	+ continue;
474	480	filters->filters[i].add = !filters->filters[i].add;
475	481	bnx2x_vf_mac_vlan_config(bp, vf, qid,
476	482	&filters->filters[i],
@@ -1899,7 +1905,8 @@ void bnx2x_iov_adjust_stats_req(struct bnx2x *bp) · 1422 unmodified lines
1899	1905	continue;
1900	1906	}
1901	1907
1902		− DP(BNX2X_MSG_IOV, "add addresses for vf %d\n", vf->abs_vfid);
	1908	+ DP_AND((BNX2X_MSG_IOV \| BNX2X_MSG_STATS),
	1909	+ "add addresses for vf %d\n", vf->abs_vfid);
1903	1910	for_each_vfq(vf, j) {
1904	1911	struct bnx2x_vf_queue *rxq = vfq_get(vf, j);
1905	1912
@@ -1920,11 +1927,12 @@ void bnx2x_iov_adjust_stats_req(struct bnx2x *bp) · 14 unmodified lines
1920	1927	cpu_to_le32(U64_HI(q_stats_addr));
1921	1928	cur_query_entry->address.lo =
1922	1929	cpu_to_le32(U64_LO(q_stats_addr));
1923		− DP(BNX2X_MSG_IOV,
1924		− "added address %x %x for vf %d queue %d client %d\n",
1925		− cur_query_entry->address.hi,
1926		− cur_query_entry->address.lo, cur_query_entry->funcID,
1927		− j, cur_query_entry->index);
	1930	+ DP_AND((BNX2X_MSG_IOV \| BNX2X_MSG_STATS),
	1931	+ "added address %x %x for vf %d queue %d client %d\n",
	1932	+ cur_query_entry->address.hi,
	1933	+ cur_query_entry->address.lo,
	1934	+ cur_query_entry->funcID,
	1935	+ j, cur_query_entry->index);
1928	1936	cur_query_entry++;
1929	1937	cur_data_offset += sizeof(struct per_queue_stats);
1930	1938	stats_count++;

drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h+1−0

@@ -114,6 +114,7 @@ struct bnx2x_vf_mac_vlan_filter { · 113 unmodified lines
114	114	(BNX2X_VF_FILTER_MAC \| BNX2X_VF_FILTER_VLAN) /shortcut/
115	115
116	116	bool add;
	117	+ bool applied;
117	118	u8 *mac;
118	119	u16 vid;
119	120	};

drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c+28−12

@@ -868,7 +868,7 @@ int bnx2x_vfpf_set_mcast(struct net_device *dev) · 867 unmodified lines
868	868	struct bnx2x *bp = netdev_priv(dev);
869	869	struct vfpf_set_q_filters_tlv *req = &bp->vf2pf_mbox->req.set_q_filters;
870	870	struct pfvf_general_resp_tlv *resp = &bp->vf2pf_mbox->resp.general_resp;
871		− int rc, i = 0;
	871	+ int rc = 0, i = 0;
872	872	struct netdev_hw_addr *ha;
873	873
874	874	if (bp->state != BNX2X_STATE_OPEN) {
@@ -883,6 +883,15 @@ int bnx2x_vfpf_set_mcast(struct net_device *dev) · 8 unmodified lines
883	883	/* Get Rx mode requested */
884	884	DP(NETIF_MSG_IFUP, "dev->flags = %x\n", dev->flags);
885	885
	886	+ /* We support PFVF_MAX_MULTICAST_PER_VF mcast addresses tops */
	887	+ if (netdev_mc_count(dev) > PFVF_MAX_MULTICAST_PER_VF) {
	888	+ DP(NETIF_MSG_IFUP,
	889	+ "VF supports not more than %d multicast MAC addresses\n",
	890	+ PFVF_MAX_MULTICAST_PER_VF);
	891	+ rc = -EINVAL;
	892	+ goto out;
	893	+ }
	894	+
886	895	netdev_for_each_mc_addr(ha, dev) {
887	896	DP(NETIF_MSG_IFUP, "Adding mcast MAC: %pM\n",
888	897	bnx2x_mc_addr(ha));
@@ -890,16 +899,6 @@ int bnx2x_vfpf_set_mcast(struct net_device *dev) · 1 unmodified line
890	899	i++;
891	900	}
892	901
893		− /* We support four PFVF_MAX_MULTICAST_PER_VF mcast
894		− * addresses tops
895		− */
896		− if (i >= PFVF_MAX_MULTICAST_PER_VF) {
897		− DP(NETIF_MSG_IFUP,
898		− "VF supports not more than %d multicast MAC addresses\n",
899		− PFVF_MAX_MULTICAST_PER_VF);
900		− return -EINVAL;
901		− }
902		−
903	902	req->n_multicast = i;
904	903	req->flags \|= VFPF_SET_Q_FILTERS_MULTICAST_CHANGED;
905	904	req->vf_qid = 0;
@@ -924,7 +923,7 @@ int bnx2x_vfpf_set_mcast(struct net_device *dev) · 18 unmodified lines
924	923	out:
925	924	bnx2x_vfpf_finalize(bp, &req->first_tlv);
926	925
927		− return 0;
	926	+ return rc;
928	927	}
929	928
930	929	/* request pf to add a vlan for the vf */
@@ -1778,6 +1777,23 @@ static int bnx2x_vf_mbx_qfilters(struct bnx2x bp, struct bnx2x_virtf vf) · 847 unmodified lines
1778	1777	goto op_err;
1779	1778	}
1780	1779
	1780	+ /* build vlan list */
	1781	+ fl = NULL;
	1782	+
	1783	+ rc = bnx2x_vf_mbx_macvlan_list(bp, vf, msg, &fl,
	1784	+ VFPF_VLAN_FILTER);
	1785	+ if (rc)
	1786	+ goto op_err;
	1787	+
	1788	+ if (fl) {
	1789	+ /* set vlan list */
	1790	+ rc = bnx2x_vf_mac_vlan_config_list(bp, vf, fl,
	1791	+ msg->vf_qid,
	1792	+ false);
	1793	+ if (rc)
	1794	+ goto op_err;
	1795	+ }
	1796	+
1781	1797	}
1782	1798
1783	1799	if (msg->flags & VFPF_SET_Q_FILTERS_RX_MASK_CHANGED) {

drivers/net/ethernet/broadcom/bnxt/bnxt.c+18−7

@@ -4465,6 +4465,10 @@ static int bnxt_hwrm_func_qcfg(struct bnxt *bp) · 4464 unmodified lines
4465	4465	vf->vlan = le16_to_cpu(resp->vlan) & VLAN_VID_MASK;
4466	4466	}
4467	4467	#endif
	4468	+ if (BNXT_PF(bp) && (le16_to_cpu(resp->flags) &
	4469	+ FUNC_QCFG_RESP_FLAGS_FW_DCBX_AGENT_ENABLED))
	4470	+ bp->flags \|= BNXT_FLAG_FW_LLDP_AGENT;
	4471	+
4468	4472	switch (resp->port_partition_type) {
4469	4473	case FUNC_QCFG_RESP_PORT_PARTITION_TYPE_NPAR1_0:
4470	4474	case FUNC_QCFG_RESP_PORT_PARTITION_TYPE_NPAR1_5:
@@ -5507,8 +5511,9 @@ static int bnxt_hwrm_phy_qcaps(struct bnxt *bp) · 1036 unmodified lines
5507	5511	bp->lpi_tmr_hi = le32_to_cpu(resp->valid_tx_lpi_timer_high) &
5508	5512	PORT_PHY_QCAPS_RESP_TX_LPI_TIMER_HIGH_MASK;
5509	5513	}
5510		− link_info->support_auto_speeds =
5511		− le16_to_cpu(resp->supported_speeds_auto_mode);
	5514	+ if (resp->supported_speeds_auto_mode)
	5515	+ link_info->support_auto_speeds =
	5516	+ le16_to_cpu(resp->supported_speeds_auto_mode);
5512	5517
5513	5518	hwrm_phy_qcaps_exit:
5514	5519	mutex_unlock(&bp->hwrm_cmd_lock);
@@ -6495,8 +6500,14 @@ static void bnxt_reset_task(struct bnxt *bp, bool silent) · 980 unmodified lines
6495	6500	if (!silent)
6496	6501	bnxt_dbg_dump_states(bp);
6497	6502	if (netif_running(bp->dev)) {
	6503	+ int rc;
	6504	+
	6505	+ if (!silent)
	6506	+ bnxt_ulp_stop(bp);
6498	6507	bnxt_close_nic(bp, false, false);
6499		− bnxt_open_nic(bp, false, false);
	6508	+ rc = bnxt_open_nic(bp, false, false);
	6509	+ if (!silent && !rc)
	6510	+ bnxt_ulp_start(bp);
6500	6511	}
6501	6512	}
6502	6513
@@ -7444,6 +7455,10 @@ static int bnxt_init_one(struct pci_dev pdev, const struct pci_device_id ent) · 941 unmodified lines
7444	7455	if (rc)
7445	7456	goto init_err_pci_clean;
7446	7457
	7458	+ rc = bnxt_hwrm_func_reset(bp);
	7459	+ if (rc)
	7460	+ goto init_err_pci_clean;
	7461	+
7447	7462	bnxt_hwrm_fw_set_time(bp);
7448	7463
7449	7464	dev->hw_features = NETIF_F_IP_CSUM \| NETIF_F_IPV6_CSUM \| NETIF_F_SG \|
@@ -7554,10 +7569,6 @@ static int bnxt_init_one(struct pci_dev pdev, const struct pci_device_id ent) · 104 unmodified lines
7554	7569	if (rc)
7555	7570	goto init_err_pci_clean;
7556	7571
7557		− rc = bnxt_hwrm_func_reset(bp);
7558		− if (rc)
7559		− goto init_err_pci_clean;
7560		−
7561	7572	rc = bnxt_init_int_mode(bp);
7562	7573	if (rc)
7563	7574	goto init_err_pci_clean;

drivers/net/ethernet/broadcom/bnxt/bnxt.h+1−0

@@ -993,6 +993,7 @@ struct bnxt { · 992 unmodified lines
993	993	BNXT_FLAG_ROCEV2_CAP)
994	994	#define BNXT_FLAG_NO_AGG_RINGS 0x20000
995	995	#define BNXT_FLAG_RX_PAGE_MODE 0x40000
	996	+ #define BNXT_FLAG_FW_LLDP_AGENT 0x80000
996	997	#define BNXT_FLAG_CHIP_NITRO_A0 0x1000000
997	998
998	999	#define BNXT_FLAG_ALL_CONFIG_FEATS (BNXT_FLAG_TPA \| \

drivers/net/ethernet/broadcom/bnxt/bnxt_dcb.c+1−1

@@ -474,7 +474,7 @@ void bnxt_dcb_init(struct bnxt *bp) · 473 unmodified lines
474	474	return;
475	475
476	476	bp->dcbx_cap = DCB_CAP_DCBX_VER_IEEE;
477		− if (BNXT_PF(bp))
	477	+ if (BNXT_PF(bp) && !(bp->flags & BNXT_FLAG_FW_LLDP_AGENT))
478	478	bp->dcbx_cap \|= DCB_CAP_DCBX_HOST;
479	479	else
480	480	bp->dcbx_cap \|= DCB_CAP_DCBX_LLD_MANAGED;

drivers/net/ethernet/broadcom/genet/bcmgenet.c+147−59

@@ -1,7 +1,7 @@
1	1	/*
2	2	* Broadcom GENET (Gigabit Ethernet) controller driver
3	3	*
4		− * Copyright (c) 2014 Broadcom Corporation
	4	+ * Copyright (c) 2014-2017 Broadcom
5	5	*
6	6	* This program is free software; you can redistribute it and/or modify
7	7	* it under the terms of the GNU General Public License version 2 as
@@ -450,6 +450,22 @@ static inline void bcmgenet_rdma_ring_writel(struct bcmgenet_priv *priv, · 442 unmodified lines
450	450	genet_dma_ring_regs[r]);
451	451	}
452	452
	453	+static int bcmgenet_begin(struct net_device *dev)
	454	+{
	455	+ struct bcmgenet_priv *priv = netdev_priv(dev);
	456	+
	457	+ /* Turn on the clock */
	458	+ return clk_prepare_enable(priv->clk);
	459	+}
	460	+
	461	+static void bcmgenet_complete(struct net_device *dev)
	462	+{
	463	+ struct bcmgenet_priv *priv = netdev_priv(dev);
	464	+
	465	+ /* Turn off the clock */
	466	+ clk_disable_unprepare(priv->clk);
	467	+}
	468	+
453	469	static int bcmgenet_get_link_ksettings(struct net_device *dev,
454	470	struct ethtool_link_ksettings *cmd)
455	471	{
@@ -778,8 +794,9 @@ static const struct bcmgenet_stats bcmgenet_gstrings_stats[] = { · 322 unmodified lines
778	794	STAT_GENET_RUNT("rx_runt_bytes", mib.rx_runt_bytes),
779	795	/* Misc UniMAC counters */
780	796	STAT_GENET_MISC("rbuf_ovflow_cnt", mib.rbuf_ovflow_cnt,
781		− UMAC_RBUF_OVFL_CNT),
782		− STAT_GENET_MISC("rbuf_err_cnt", mib.rbuf_err_cnt, UMAC_RBUF_ERR_CNT),
	797	+ UMAC_RBUF_OVFL_CNT_V1),
	798	+ STAT_GENET_MISC("rbuf_err_cnt", mib.rbuf_err_cnt,
	799	+ UMAC_RBUF_ERR_CNT_V1),
783	800	STAT_GENET_MISC("mdf_err_cnt", mib.mdf_err_cnt, UMAC_MDF_ERR_CNT),
784	801	STAT_GENET_SOFT_MIB("alloc_rx_buff_failed", mib.alloc_rx_buff_failed),
785	802	STAT_GENET_SOFT_MIB("rx_dma_failed", mib.rx_dma_failed),
@@ -821,6 +838,45 @@ static void bcmgenet_get_strings(struct net_device *dev, u32 stringset, · 35 unmodified lines
821	838	}
822	839	}
823	840
	841	+static u32 bcmgenet_update_stat_misc(struct bcmgenet_priv *priv, u16 offset)
	842	+{
	843	+ u16 new_offset;
	844	+ u32 val;
	845	+
	846	+ switch (offset) {
	847	+ case UMAC_RBUF_OVFL_CNT_V1:
	848	+ if (GENET_IS_V2(priv))
	849	+ new_offset = RBUF_OVFL_CNT_V2;
	850	+ else
	851	+ new_offset = RBUF_OVFL_CNT_V3PLUS;
	852	+
	853	+ val = bcmgenet_rbuf_readl(priv, new_offset);
	854	+ /* clear if overflowed */
	855	+ if (val == ~0)
	856	+ bcmgenet_rbuf_writel(priv, 0, new_offset);
	857	+ break;
	858	+ case UMAC_RBUF_ERR_CNT_V1:
	859	+ if (GENET_IS_V2(priv))
	860	+ new_offset = RBUF_ERR_CNT_V2;
	861	+ else
	862	+ new_offset = RBUF_ERR_CNT_V3PLUS;
	863	+
	864	+ val = bcmgenet_rbuf_readl(priv, new_offset);
	865	+ /* clear if overflowed */
	866	+ if (val == ~0)
	867	+ bcmgenet_rbuf_writel(priv, 0, new_offset);
	868	+ break;
	869	+ default:
	870	+ val = bcmgenet_umac_readl(priv, offset);
	871	+ /* clear if overflowed */
	872	+ if (val == ~0)
	873	+ bcmgenet_umac_writel(priv, 0, offset);
	874	+ break;
	875	+ }
	876	+
	877	+ return val;
	878	+}
	879	+
824	880	static void bcmgenet_update_mib_counters(struct bcmgenet_priv *priv)
825	881	{
826	882	int i, j = 0;
@@ -836,19 +892,28 @@ static void bcmgenet_update_mib_counters(struct bcmgenet_priv *priv) · 9 unmodified lines
836	892	case BCMGENET_STAT_NETDEV:
837	893	case BCMGENET_STAT_SOFT:
838	894	continue;
839		− case BCMGENET_STAT_MIB_RX:
840		− case BCMGENET_STAT_MIB_TX:
841	895	case BCMGENET_STAT_RUNT:
842		− if (s->type != BCMGENET_STAT_MIB_RX)
843		− offset = BCMGENET_STAT_OFFSET;
	896	+ offset += BCMGENET_STAT_OFFSET;
	897	+ /* fall through */
	898	+ case BCMGENET_STAT_MIB_TX:
	899	+ offset += BCMGENET_STAT_OFFSET;
	900	+ /* fall through */
	901	+ case BCMGENET_STAT_MIB_RX:
844	902	val = bcmgenet_umac_readl(priv,
845	903	UMAC_MIB_START + j + offset);
	904	+ offset = 0; /* Reset Offset */
846	905	break;
847	906	case BCMGENET_STAT_MISC:
848		− val = bcmgenet_umac_readl(priv, s->reg_offset);
849		− /* clear if overflowed */
850		− if (val == ~0)
851		− bcmgenet_umac_writel(priv, 0, s->reg_offset);
	907	+ if (GENET_IS_V1(priv)) {
	908	+ val = bcmgenet_umac_readl(priv, s->reg_offset);
	909	+ /* clear if overflowed */
	910	+ if (val == ~0)
	911	+ bcmgenet_umac_writel(priv, 0,
	912	+ s->reg_offset);
	913	+ } else {
	914	+ val = bcmgenet_update_stat_misc(priv,
	915	+ s->reg_offset);
	916	+ }
852	917	break;
853	918	}
854	919
@@ -973,6 +1038,8 @@ static int bcmgenet_set_eee(struct net_device dev, struct ethtool_eee e) · 118 unmodified lines
973	1038
974	1039	/* standard ethtool support functions. */
975	1040	static const struct ethtool_ops bcmgenet_ethtool_ops = {
	1041	+ .begin = bcmgenet_begin,
	1042	+ .complete = bcmgenet_complete,
976	1043	.get_strings = bcmgenet_get_strings,
977	1044	.get_sset_count = bcmgenet_get_sset_count,
978	1045	.get_ethtool_stats = bcmgenet_get_ethtool_stats,
@@ -1167,7 +1234,6 @@ static unsigned int __bcmgenet_tx_reclaim(struct net_device *dev, · 188 unmodified lines
1167	1234	struct bcmgenet_priv *priv = netdev_priv(dev);
1168	1235	struct device *kdev = &priv->pdev->dev;
1169	1236	struct enet_cb *tx_cb_ptr;
1170		− struct netdev_queue *txq;
1171	1237	unsigned int pkts_compl = 0;
1172	1238	unsigned int bytes_compl = 0;
1173	1239	unsigned int c_index;
@@ -1219,13 +1285,8 @@ static unsigned int __bcmgenet_tx_reclaim(struct net_device *dev, · 45 unmodified lines
1219	1285	dev->stats.tx_packets += pkts_compl;
1220	1286	dev->stats.tx_bytes += bytes_compl;
1221	1287
1222		− txq = netdev_get_tx_queue(dev, ring->queue);
1223		− netdev_tx_completed_queue(txq, pkts_compl, bytes_compl);
1224		−
1225		− if (ring->free_bds > (MAX_SKB_FRAGS + 1)) {
1226		− if (netif_tx_queue_stopped(txq))
1227		− netif_tx_wake_queue(txq);
1228		− }
	1288	+ netdev_tx_completed_queue(netdev_get_tx_queue(dev, ring->queue),
	1289	+ pkts_compl, bytes_compl);
1229	1290
1230	1291	return pkts_compl;
1231	1292	}
@@ -1248,8 +1309,16 @@ static int bcmgenet_tx_poll(struct napi_struct *napi, int budget) · 16 unmodified lines
1248	1309	struct bcmgenet_tx_ring *ring =
1249	1310	container_of(napi, struct bcmgenet_tx_ring, napi);
1250	1311	unsigned int work_done = 0;
	1312	+ struct netdev_queue *txq;
	1313	+ unsigned long flags;
1251	1314
1252		− work_done = bcmgenet_tx_reclaim(ring->priv->dev, ring);
	1315	+ spin_lock_irqsave(&ring->lock, flags);
	1316	+ work_done = __bcmgenet_tx_reclaim(ring->priv->dev, ring);
	1317	+ if (ring->free_bds > (MAX_SKB_FRAGS + 1)) {
	1318	+ txq = netdev_get_tx_queue(ring->priv->dev, ring->queue);
	1319	+ netif_tx_wake_queue(txq);
	1320	+ }
	1321	+ spin_unlock_irqrestore(&ring->lock, flags);
1253	1322
1254	1323	if (work_done == 0) {
1255	1324	napi_complete(napi);
@@ -2457,24 +2526,28 @@ static int bcmgenet_init_dma(struct bcmgenet_priv *priv) · 1201 unmodified lines
2457	2526	/* Interrupt bottom half */
2458	2527	static void bcmgenet_irq_task(struct work_struct *work)
2459	2528	{
	2529	+ unsigned long flags;
	2530	+ unsigned int status;
2460	2531	struct bcmgenet_priv *priv = container_of(
2461	2532	work, struct bcmgenet_priv, bcmgenet_irq_work);
2462	2533
2463	2534	netif_dbg(priv, intr, priv->dev, "%s\n", __func__);
2464	2535
2465		− if (priv->irq0_stat & UMAC_IRQ_MPD_R) {
2466		− priv->irq0_stat &= ~UMAC_IRQ_MPD_R;
	2536	+ spin_lock_irqsave(&priv->lock, flags);
	2537	+ status = priv->irq0_stat;
	2538	+ priv->irq0_stat = 0;
	2539	+ spin_unlock_irqrestore(&priv->lock, flags);
	2540	+
	2541	+ if (status & UMAC_IRQ_MPD_R) {
2467	2542	netif_dbg(priv, wol, priv->dev,
2468	2543	"magic packet detected, waking up\n");
2469	2544	bcmgenet_power_up(priv, GENET_POWER_WOL_MAGIC);
2470	2545	}
2471	2546
2472	2547	/* Link UP/DOWN event */
2473		− if (priv->irq0_stat & UMAC_IRQ_LINK_EVENT) {
	2548	+ if (status & UMAC_IRQ_LINK_EVENT)
2474	2549	phy_mac_interrupt(priv->phydev,
2475		− !!(priv->irq0_stat & UMAC_IRQ_LINK_UP));
2476		− priv->irq0_stat &= ~UMAC_IRQ_LINK_EVENT;
2477		− }
	2550	+ !!(status & UMAC_IRQ_LINK_UP));
2478	2551	}
2479	2552
2480	2553	/* bcmgenet_isr1: handle Rx and Tx priority queues */
@@ -2483,22 +2556,21 @@ static irqreturn_t bcmgenet_isr1(int irq, void *dev_id) · 2 unmodified lines
2483	2556	struct bcmgenet_priv *priv = dev_id;
2484	2557	struct bcmgenet_rx_ring *rx_ring;
2485	2558	struct bcmgenet_tx_ring *tx_ring;
2486		− unsigned int index;
	2559	+ unsigned int index, status;
2487	2560
2488		− /* Save irq status for bottom-half processing. */
2489		− priv->irq1_stat =
2490		− bcmgenet_intrl2_1_readl(priv, INTRL2_CPU_STAT) &
	2561	+ /* Read irq status */
	2562	+ status = bcmgenet_intrl2_1_readl(priv, INTRL2_CPU_STAT) &
2491	2563	~bcmgenet_intrl2_1_readl(priv, INTRL2_CPU_MASK_STATUS);
2492	2564
2493	2565	/* clear interrupts */
2494		− bcmgenet_intrl2_1_writel(priv, priv->irq1_stat, INTRL2_CPU_CLEAR);
	2566	+ bcmgenet_intrl2_1_writel(priv, status, INTRL2_CPU_CLEAR);
2495	2567
2496	2568	netif_dbg(priv, intr, priv->dev,
2497		− "%s: IRQ=0x%x\n", __func__, priv->irq1_stat);
	2569	+ "%s: IRQ=0x%x\n", __func__, status);
2498	2570
2499	2571	/* Check Rx priority queue interrupts */
2500	2572	for (index = 0; index < priv->hw_params->rx_queues; index++) {
2501		− if (!(priv->irq1_stat & BIT(UMAC_IRQ1_RX_INTR_SHIFT + index)))
	2573	+ if (!(status & BIT(UMAC_IRQ1_RX_INTR_SHIFT + index)))
2502	2574	continue;
2503	2575
2504	2576	rx_ring = &priv->rx_rings[index];
@@ -2511,7 +2583,7 @@ static irqreturn_t bcmgenet_isr1(int irq, void *dev_id) · 6 unmodified lines
2511	2583
2512	2584	/* Check Tx priority queue interrupts */
2513	2585	for (index = 0; index < priv->hw_params->tx_queues; index++) {
2514		− if (!(priv->irq1_stat & BIT(index)))
	2586	+ if (!(status & BIT(index)))
2515	2587	continue;
2516	2588
2517	2589	tx_ring = &priv->tx_rings[index];
@@ -2531,19 +2603,20 @@ static irqreturn_t bcmgenet_isr0(int irq, void *dev_id) · 13 unmodified lines
2531	2603	struct bcmgenet_priv *priv = dev_id;
2532	2604	struct bcmgenet_rx_ring *rx_ring;
2533	2605	struct bcmgenet_tx_ring *tx_ring;
	2606	+ unsigned int status;
	2607	+ unsigned long flags;
2534	2608
2535		− /* Save irq status for bottom-half processing. */
2536		− priv->irq0_stat =
2537		− bcmgenet_intrl2_0_readl(priv, INTRL2_CPU_STAT) &
	2609	+ /* Read irq status */
	2610	+ status = bcmgenet_intrl2_0_readl(priv, INTRL2_CPU_STAT) &
2538	2611	~bcmgenet_intrl2_0_readl(priv, INTRL2_CPU_MASK_STATUS);
2539	2612
2540	2613	/* clear interrupts */
2541		− bcmgenet_intrl2_0_writel(priv, priv->irq0_stat, INTRL2_CPU_CLEAR);
	2614	+ bcmgenet_intrl2_0_writel(priv, status, INTRL2_CPU_CLEAR);
2542	2615
2543	2616	netif_dbg(priv, intr, priv->dev,
2544		− "IRQ=0x%x\n", priv->irq0_stat);
	2617	+ "IRQ=0x%x\n", status);
2545	2618
2546		− if (priv->irq0_stat & UMAC_IRQ_RXDMA_DONE) {
	2619	+ if (status & UMAC_IRQ_RXDMA_DONE) {
2547	2620	rx_ring = &priv->rx_rings[DESC_INDEX];
2548	2621
2549	2622	if (likely(napi_schedule_prep(&rx_ring->napi))) {
@@ -2552,7 +2625,7 @@ static irqreturn_t bcmgenet_isr0(int irq, void *dev_id) · 2 unmodified lines
2552	2625	}
2553	2626	}
2554	2627
2555		− if (priv->irq0_stat & UMAC_IRQ_TXDMA_DONE) {
	2628	+ if (status & UMAC_IRQ_TXDMA_DONE) {
2556	2629	tx_ring = &priv->tx_rings[DESC_INDEX];
2557	2630
2558	2631	if (likely(napi_schedule_prep(&tx_ring->napi))) {
@@ -2561,22 +2634,23 @@ static irqreturn_t bcmgenet_isr0(int irq, void *dev_id) · 2 unmodified lines
2561	2634	}
2562	2635	}
2563	2636
2564		− if (priv->irq0_stat & (UMAC_IRQ_PHY_DET_R \|
2565		− UMAC_IRQ_PHY_DET_F \|
2566		− UMAC_IRQ_LINK_EVENT \|
2567		− UMAC_IRQ_HFB_SM \|
2568		− UMAC_IRQ_HFB_MM \|
2569		− UMAC_IRQ_MPD_R)) {
2570		− /* all other interested interrupts handled in bottom half */
2571		− schedule_work(&priv->bcmgenet_irq_work);
2572		− }
2573		−
2574	2637	if ((priv->hw_params->flags & GENET_HAS_MDIO_INTR) &&
2575		− priv->irq0_stat & (UMAC_IRQ_MDIO_DONE \| UMAC_IRQ_MDIO_ERROR)) {
2576		− priv->irq0_stat &= ~(UMAC_IRQ_MDIO_DONE \| UMAC_IRQ_MDIO_ERROR);
	2638	+ status & (UMAC_IRQ_MDIO_DONE \| UMAC_IRQ_MDIO_ERROR)) {
2577	2639	wake_up(&priv->wq);
2578	2640	}
2579	2641
	2642	+ /* all other interested interrupts handled in bottom half */
	2643	+ status &= (UMAC_IRQ_LINK_EVENT \|
	2644	+ UMAC_IRQ_MPD_R);
	2645	+ if (status) {
	2646	+ /* Save irq status for bottom-half processing. */
	2647	+ spin_lock_irqsave(&priv->lock, flags);
	2648	+ priv->irq0_stat \|= status;
	2649	+ spin_unlock_irqrestore(&priv->lock, flags);
	2650	+
	2651	+ schedule_work(&priv->bcmgenet_irq_work);
	2652	+ }
	2653	+
2580	2654	return IRQ_HANDLED;
2581	2655	}
2582	2656
@@ -2801,6 +2875,8 @@ err_irq0: · 218 unmodified lines
2801	2875	err_fini_dma:
2802	2876	bcmgenet_fini_dma(priv);
2803	2877	err_clk_disable:
	2878	+ if (priv->internal_phy)
	2879	+ bcmgenet_power_down(priv, GENET_POWER_PASSIVE);
2804	2880	clk_disable_unprepare(priv->clk);
2805	2881	return ret;
2806	2882	}
@@ -3177,6 +3253,12 @@ static void bcmgenet_set_hw_params(struct bcmgenet_priv *priv) · 370 unmodified lines
3177	3253	*/
3178	3254	gphy_rev = reg & 0xffff;
3179	3255
	3256	+ /* This is reserved so should require special treatment */
	3257	+ if (gphy_rev == 0 \|\| gphy_rev == 0x01ff) {
	3258	+ pr_warn("Invalid GPHY revision detected: 0x%04x\n", gphy_rev);
	3259	+ return;
	3260	+ }
	3261	+
3180	3262	/* This is the good old scheme, just GPHY major, no minor nor patch */
3181	3263	if ((gphy_rev & 0xf0) != 0)
3182	3264	priv->gphy_rev = gphy_rev << 8;
@@ -3185,12 +3267,6 @@ static void bcmgenet_set_hw_params(struct bcmgenet_priv *priv) · 2 unmodified lines
3185	3267	else if ((gphy_rev & 0xff00) != 0)
3186	3268	priv->gphy_rev = gphy_rev;
3187	3269
3188		− /* This is reserved so should require special treatment */
3189		− else if (gphy_rev == 0 \|\| gphy_rev == 0x01ff) {
3190		− pr_warn("Invalid GPHY revision detected: 0x%04x\n", gphy_rev);
3191		− return;
3192		− }
3193		−
3194	3270	#ifdef CONFIG_PHYS_ADDR_T_64BIT
3195	3271	if (!(params->flags & GENET_HAS_40BITS))
3196	3272	pr_warn("GENET does not support 40-bits PA\n");
@@ -3233,6 +3309,7 @@ static int bcmgenet_probe(struct platform_device *pdev) · 36 unmodified lines
3233	3309	const void *macaddr;
3234	3310	struct resource *r;
3235	3311	int err = -EIO;
	3312	+ const char *phy_mode_str;
3236	3313
3237	3314	/* Up to GENET_MAX_MQ_CNT + 1 TX queues and RX queues */
3238	3315	dev = alloc_etherdev_mqs(sizeof(*priv), GENET_MAX_MQ_CNT + 1,
@@ -3276,6 +3353,8 @@ static int bcmgenet_probe(struct platform_device *pdev) · 37 unmodified lines
3276	3353	goto err;
3277	3354	}
3278	3355
	3356	+ spin_lock_init(&priv->lock);
	3357	+
3279	3358	SET_NETDEV_DEV(dev, &pdev->dev);
3280	3359	dev_set_drvdata(&pdev->dev, dev);
3281	3360	ether_addr_copy(dev->dev_addr, macaddr);
@@ -3338,6 +3417,13 @@ static int bcmgenet_probe(struct platform_device *pdev) · 56 unmodified lines
3338	3417	priv->clk_eee = NULL;
3339	3418	}
3340	3419
	3420	+ /* If this is an internal GPHY, power it on now, before UniMAC is
	3421	+ * brought out of reset as absolutely no UniMAC activity is allowed
	3422	+ */
	3423	+ if (dn && !of_property_read_string(dn, "phy-mode", &phy_mode_str) &&
	3424	+ !strcasecmp(phy_mode_str, "internal"))
	3425	+ bcmgenet_power_up(priv, GENET_POWER_PASSIVE);
	3426	+
3341	3427	err = reset_umac(priv);
3342	3428	if (err)
3343	3429	goto err_clk_disable;
@@ -3502,6 +3588,8 @@ static int bcmgenet_resume(struct device *d) · 158 unmodified lines
3502	3588	return 0;
3503	3589
3504	3590	out_clk_disable:
	3591	+ if (priv->internal_phy)
	3592	+ bcmgenet_power_down(priv, GENET_POWER_PASSIVE);
3505	3593	clk_disable_unprepare(priv->clk);
3506	3594	return ret;
3507	3595	}

drivers/net/ethernet/broadcom/genet/bcmgenet.h+11−5

@@ -1,5 +1,5 @@
1	1	/*
2		− * Copyright (c) 2014 Broadcom Corporation
	2	+ * Copyright (c) 2014-2017 Broadcom
3	3	*
4	4	* This program is free software; you can redistribute it and/or modify
5	5	* it under the terms of the GNU General Public License version 2 as
@@ -214,7 +214,9 @@ struct bcmgenet_mib_counters { · 208 unmodified lines
214	214	#define MDIO_REG_SHIFT 16
215	215	#define MDIO_REG_MASK 0x1F
216	216
217		−#define UMAC_RBUF_OVFL_CNT 0x61C
	217	+#define UMAC_RBUF_OVFL_CNT_V1 0x61C
	218	+#define RBUF_OVFL_CNT_V2 0x80
	219	+#define RBUF_OVFL_CNT_V3PLUS 0x94
218	220
219	221	#define UMAC_MPD_CTRL 0x620
220	222	#define MPD_EN (1 << 0)
@@ -224,7 +226,9 @@ struct bcmgenet_mib_counters { · 3 unmodified lines
224	226
225	227	#define UMAC_MPD_PW_MS 0x624
226	228	#define UMAC_MPD_PW_LS 0x628
227		−#define UMAC_RBUF_ERR_CNT 0x634
	229	+#define UMAC_RBUF_ERR_CNT_V1 0x634
	230	+#define RBUF_ERR_CNT_V2 0x84
	231	+#define RBUF_ERR_CNT_V3PLUS 0x98
228	232	#define UMAC_MDF_ERR_CNT 0x638
229	233	#define UMAC_MDF_CTRL 0x650
230	234	#define UMAC_MDF_ADDR 0x654
@@ -619,11 +623,13 @@ struct bcmgenet_priv { · 388 unmodified lines
619	623	struct work_struct bcmgenet_irq_work;
620	624	int irq0;
621	625	int irq1;
622		− unsigned int irq0_stat;
623		− unsigned int irq1_stat;
624	626	int wol_irq;
625	627	bool wol_irq_disabled;
626	628
	629	+ /* shared status */
	630	+ spinlock_t lock;
	631	+ unsigned int irq0_stat;
	632	+
627	633	/* HW descriptors/checksum variables */
628	634	bool desc_64b_en;
629	635	bool desc_rxchk_en;

drivers/net/ethernet/cavium/liquidio/lio_main.c+55−55

@@ -152,7 +152,7 @@ struct octnic_gather { · 151 unmodified lines
152	152	*/
153	153	struct octeon_sg_entry *sg;
154	154
155		− u64 sg_dma_ptr;
	155	+ dma_addr_t sg_dma_ptr;
156	156	};
157	157
158	158	struct handshake {
@@ -734,6 +734,9 @@ static void delete_glists(struct lio *lio) · 575 unmodified lines
734	734	struct octnic_gather *g;
735	735	int i;
736	736
	737	+ kfree(lio->glist_lock);
	738	+ lio->glist_lock = NULL;
	739	+
737	740	if (!lio->glist)
738	741	return;
739	742
@@ -741,23 +744,26 @@ static void delete_glists(struct lio *lio) · 1 unmodified line
741	744	do {
742	745	g = (struct octnic_gather *)
743	746	list_delete_head(&lio->glist[i]);
744		− if (g) {
745		− if (g->sg) {
746		− dma_unmap_single(&lio->oct_dev->
747		− pci_dev->dev,
748		− g->sg_dma_ptr,
749		− g->sg_size,
750		− DMA_TO_DEVICE);
751		− kfree((void *)((unsigned long)g->sg -
752		− g->adjust));
753		− }
	747	+ if (g)
754	748	kfree(g);
755		− }
756	749	} while (g);
	750	+
	751	+ if (lio->glists_virt_base && lio->glists_virt_base[i]) {
	752	+ lio_dma_free(lio->oct_dev,
	753	+ lio->glist_entry_size * lio->tx_qsize,
	754	+ lio->glists_virt_base[i],
	755	+ lio->glists_dma_base[i]);
	756	+ }
757	757	}
758	758
759		− kfree((void *)lio->glist);
760		− kfree((void *)lio->glist_lock);
	759	+ kfree(lio->glists_virt_base);
	760	+ lio->glists_virt_base = NULL;
	761	+
	762	+ kfree(lio->glists_dma_base);
	763	+ lio->glists_dma_base = NULL;
	764	+
	765	+ kfree(lio->glist);
	766	+ lio->glist = NULL;
761	767	}
762	768
763	769	/**
@@ -772,13 +778,30 @@ static int setup_glists(struct octeon_device oct, struct lio lio, int num_iqs) · 8 unmodified lines
772	778	lio->glist_lock = kcalloc(num_iqs, sizeof(*lio->glist_lock),
773	779	GFP_KERNEL);
774	780	if (!lio->glist_lock)
775		− return 1;
	781	+ return -ENOMEM;
776	782
777	783	lio->glist = kcalloc(num_iqs, sizeof(*lio->glist),
778	784	GFP_KERNEL);
779	785	if (!lio->glist) {
780		− kfree((void *)lio->glist_lock);
781		− return 1;
	786	+ kfree(lio->glist_lock);
	787	+ lio->glist_lock = NULL;
	788	+ return -ENOMEM;
	789	+ }
	790	+
	791	+ lio->glist_entry_size =
	792	+ ROUNDUP8((ROUNDUP4(OCTNIC_MAX_SG) >> 2) * OCT_SG_ENTRY_SIZE);
	793	+
	794	+ /* allocate memory to store virtual and dma base address of
	795	+ * per glist consistent memory
	796	+ */
	797	+ lio->glists_virt_base = kcalloc(num_iqs, sizeof(*lio->glists_virt_base),
	798	+ GFP_KERNEL);
	799	+ lio->glists_dma_base = kcalloc(num_iqs, sizeof(*lio->glists_dma_base),
	800	+ GFP_KERNEL);
	801	+
	802	+ if (!lio->glists_virt_base \|\| !lio->glists_dma_base) {
	803	+ delete_glists(lio);
	804	+ return -ENOMEM;
782	805	}
783	806
784	807	for (i = 0; i < num_iqs; i++) {
@@ -788,6 +811,16 @@ static int setup_glists(struct octeon_device oct, struct lio lio, int num_iqs) · 3 unmodified lines
788	811
789	812	INIT_LIST_HEAD(&lio->glist[i]);
790	813
	814	+ lio->glists_virt_base[i] =
	815	+ lio_dma_alloc(oct,
	816	+ lio->glist_entry_size * lio->tx_qsize,
	817	+ &lio->glists_dma_base[i]);
	818	+
	819	+ if (!lio->glists_virt_base[i]) {
	820	+ delete_glists(lio);
	821	+ return -ENOMEM;
	822	+ }
	823	+
791	824	for (j = 0; j < lio->tx_qsize; j++) {
792	825	g = kzalloc_node(sizeof(*g), GFP_KERNEL,
793	826	numa_node);
@@ -796,43 +829,18 @@ static int setup_glists(struct octeon_device oct, struct lio lio, int num_iqs) · 2 unmodified lines
796	829	if (!g)
797	830	break;
798	831
799		− g->sg_size = ((ROUNDUP4(OCTNIC_MAX_SG) >> 2) *
800		− OCT_SG_ENTRY_SIZE);
	832	+ g->sg = lio->glists_virt_base[i] +
	833	+ (j * lio->glist_entry_size);
801	834
802		− g->sg = kmalloc_node(g->sg_size + 8,
803		− GFP_KERNEL, numa_node);
804		− if (!g->sg)
805		− g->sg = kmalloc(g->sg_size + 8, GFP_KERNEL);
806		− if (!g->sg) {
807		− kfree(g);
808		− break;
809		− }
810		−
811		− /* The gather component should be aligned on 64-bit
812		− * boundary
813		− */
814		− if (((unsigned long)g->sg) & 7) {
815		− g->adjust = 8 - (((unsigned long)g->sg) & 7);
816		− g->sg = (struct octeon_sg_entry *)
817		− ((unsigned long)g->sg + g->adjust);
818		− }
819		− g->sg_dma_ptr = dma_map_single(&oct->pci_dev->dev,
820		− g->sg, g->sg_size,
821		− DMA_TO_DEVICE);
822		− if (dma_mapping_error(&oct->pci_dev->dev,
823		− g->sg_dma_ptr)) {
824		− kfree((void *)((unsigned long)g->sg -
825		− g->adjust));
826		− kfree(g);
827		− break;
828		− }
	835	+ g->sg_dma_ptr = lio->glists_dma_base[i] +
	836	+ (j * lio->glist_entry_size);
829	837
830	838	list_add_tail(&g->list, &lio->glist[i]);
831	839	}
832	840
833	841	if (j != lio->tx_qsize) {
834	842	delete_glists(lio);
835		− return 1;
	843	+ return -ENOMEM;
836	844	}
837	845	}
838	846
@@ -1885,9 +1893,6 @@ static void free_netsgbuf(void *buf) · 1046 unmodified lines
1885	1893	i++;
1886	1894	}
1887	1895
1888		− dma_sync_single_for_cpu(&lio->oct_dev->pci_dev->dev,
1889		− g->sg_dma_ptr, g->sg_size, DMA_TO_DEVICE);
1890		−
1891	1896	iq = skb_iq(lio, skb);
1892	1897	spin_lock(&lio->glist_lock[iq]);
1893	1898	list_add_tail(&g->list, &lio->glist[iq]);
@@ -1933,9 +1938,6 @@ static void free_netsgbuf_with_resp(void *buf) · 39 unmodified lines
1933	1938	i++;
1934	1939	}
1935	1940
1936		− dma_sync_single_for_cpu(&lio->oct_dev->pci_dev->dev,
1937		− g->sg_dma_ptr, g->sg_size, DMA_TO_DEVICE);
1938		−
1939	1941	iq = skb_iq(lio, skb);
1940	1942
1941	1943	spin_lock(&lio->glist_lock[iq]);
@@ -3273,8 +3275,6 @@ static int liquidio_xmit(struct sk_buff skb, struct net_device netdev) · 1331 unmodified lines
3273	3275	i++;
3274	3276	}
3275	3277
3276		− dma_sync_single_for_device(&oct->pci_dev->dev, g->sg_dma_ptr,
3277		− g->sg_size, DMA_TO_DEVICE);
3278	3278	dptr = g->sg_dma_ptr;
3279	3279
3280	3280	if (OCTEON_CN23XX_PF(oct))

drivers/net/ethernet/cavium/liquidio/lio_vf_main.c+55−49

@@ -108,6 +108,8 @@ struct octnic_gather { · 107 unmodified lines
108	108	* received from the IP layer.
109	109	*/
110	110	struct octeon_sg_entry *sg;
	111	+
	112	+ dma_addr_t sg_dma_ptr;
111	113	};
112	114
113	115	struct octeon_device_priv {
@@ -490,6 +492,9 @@ static void delete_glists(struct lio *lio) · 376 unmodified lines
490	492	struct octnic_gather *g;
491	493	int i;
492	494
	495	+ kfree(lio->glist_lock);
	496	+ lio->glist_lock = NULL;
	497	+
493	498	if (!lio->glist)
494	499	return;
495	500
@@ -497,17 +502,26 @@ static void delete_glists(struct lio *lio) · 1 unmodified line
497	502	do {
498	503	g = (struct octnic_gather *)
499	504	list_delete_head(&lio->glist[i]);
500		− if (g) {
501		− if (g->sg)
502		− kfree((void *)((unsigned long)g->sg -
503		− g->adjust));
	505	+ if (g)
504	506	kfree(g);
505		− }
506	507	} while (g);
	508	+
	509	+ if (lio->glists_virt_base && lio->glists_virt_base[i]) {
	510	+ lio_dma_free(lio->oct_dev,
	511	+ lio->glist_entry_size * lio->tx_qsize,
	512	+ lio->glists_virt_base[i],
	513	+ lio->glists_dma_base[i]);
	514	+ }
507	515	}
508	516
	517	+ kfree(lio->glists_virt_base);
	518	+ lio->glists_virt_base = NULL;
	519	+
	520	+ kfree(lio->glists_dma_base);
	521	+ lio->glists_dma_base = NULL;
	522	+
509	523	kfree(lio->glist);
510		− kfree(lio->glist_lock);
	524	+ lio->glist = NULL;
511	525	}
512	526
513	527	/**
@@ -522,13 +536,30 @@ static int setup_glists(struct lio *lio, int num_iqs) · 8 unmodified lines
522	536	lio->glist_lock =
523	537	kzalloc(sizeof(lio->glist_lock) num_iqs, GFP_KERNEL);
524	538	if (!lio->glist_lock)
525		− return 1;
	539	+ return -ENOMEM;
526	540
527	541	lio->glist =
528	542	kzalloc(sizeof(lio->glist) num_iqs, GFP_KERNEL);
529	543	if (!lio->glist) {
530	544	kfree(lio->glist_lock);
531		− return 1;
	545	+ lio->glist_lock = NULL;
	546	+ return -ENOMEM;
	547	+ }
	548	+
	549	+ lio->glist_entry_size =
	550	+ ROUNDUP8((ROUNDUP4(OCTNIC_MAX_SG) >> 2) * OCT_SG_ENTRY_SIZE);
	551	+
	552	+ /* allocate memory to store virtual and dma base address of
	553	+ * per glist consistent memory
	554	+ */
	555	+ lio->glists_virt_base = kcalloc(num_iqs, sizeof(*lio->glists_virt_base),
	556	+ GFP_KERNEL);
	557	+ lio->glists_dma_base = kcalloc(num_iqs, sizeof(*lio->glists_dma_base),
	558	+ GFP_KERNEL);
	559	+
	560	+ if (!lio->glists_virt_base \|\| !lio->glists_dma_base) {
	561	+ delete_glists(lio);
	562	+ return -ENOMEM;
532	563	}
533	564
534	565	for (i = 0; i < num_iqs; i++) {
@@ -536,34 +567,33 @@ static int setup_glists(struct lio *lio, int num_iqs) · 1 unmodified line
536	567
537	568	INIT_LIST_HEAD(&lio->glist[i]);
538	569
	570	+ lio->glists_virt_base[i] =
	571	+ lio_dma_alloc(lio->oct_dev,
	572	+ lio->glist_entry_size * lio->tx_qsize,
	573	+ &lio->glists_dma_base[i]);
	574	+
	575	+ if (!lio->glists_virt_base[i]) {
	576	+ delete_glists(lio);
	577	+ return -ENOMEM;
	578	+ }
	579	+
539	580	for (j = 0; j < lio->tx_qsize; j++) {
540	581	g = kzalloc(sizeof(*g), GFP_KERNEL);
541	582	if (!g)
542	583	break;
543	584
544		− g->sg_size = ((ROUNDUP4(OCTNIC_MAX_SG) >> 2) *
545		− OCT_SG_ENTRY_SIZE);
	585	+ g->sg = lio->glists_virt_base[i] +
	586	+ (j * lio->glist_entry_size);
546	587
547		− g->sg = kmalloc(g->sg_size + 8, GFP_KERNEL);
548		− if (!g->sg) {
549		− kfree(g);
550		− break;
551		− }
	588	+ g->sg_dma_ptr = lio->glists_dma_base[i] +
	589	+ (j * lio->glist_entry_size);
552	590
553		− /* The gather component should be aligned on 64-bit
554		− * boundary
555		− */
556		− if (((unsigned long)g->sg) & 7) {
557		− g->adjust = 8 - (((unsigned long)g->sg) & 7);
558		− g->sg = (struct octeon_sg_entry *)
559		− ((unsigned long)g->sg + g->adjust);
560		− }
561	591	list_add_tail(&g->list, &lio->glist[i]);
562	592	}
563	593
564	594	if (j != lio->tx_qsize) {
565	595	delete_glists(lio);
566		− return 1;
	596	+ return -ENOMEM;
567	597	}
568	598	}
569	599
@@ -1324,10 +1354,6 @@ static void free_netsgbuf(void *buf) · 754 unmodified lines
1324	1354	i++;
1325	1355	}
1326	1356
1327		− dma_unmap_single(&lio->oct_dev->pci_dev->dev,
1328		− finfo->dptr, g->sg_size,
1329		− DMA_TO_DEVICE);
1330		−
1331	1357	iq = skb_iq(lio, skb);
1332	1358
1333	1359	spin_lock(&lio->glist_lock[iq]);
@@ -1374,10 +1400,6 @@ static void free_netsgbuf_with_resp(void *buf) · 40 unmodified lines
1374	1400	i++;
1375	1401	}
1376	1402
1377		− dma_unmap_single(&lio->oct_dev->pci_dev->dev,
1378		− finfo->dptr, g->sg_size,
1379		− DMA_TO_DEVICE);
1380		−
1381	1403	iq = skb_iq(lio, skb);
1382	1404
1383	1405	spin_lock(&lio->glist_lock[iq]);
@@ -2382,23 +2404,7 @@ static int liquidio_xmit(struct sk_buff skb, struct net_device netdev) · 998 unmodified lines
2382	2404	i++;
2383	2405	}
2384	2406
2385		− dptr = dma_map_single(&oct->pci_dev->dev,
2386		− g->sg, g->sg_size,
2387		− DMA_TO_DEVICE);
2388		− if (dma_mapping_error(&oct->pci_dev->dev, dptr)) {
2389		− dev_err(&oct->pci_dev->dev, "%s DMA mapping error 4\n",
2390		− __func__);
2391		− dma_unmap_single(&oct->pci_dev->dev, g->sg[0].ptr[0],
2392		− skb->len - skb->data_len,
2393		− DMA_TO_DEVICE);
2394		− for (j = 1; j <= frags; j++) {
2395		− frag = &skb_shinfo(skb)->frags[j - 1];
2396		− dma_unmap_page(&oct->pci_dev->dev,
2397		− g->sg[j >> 2].ptr[j & 3],
2398		− frag->size, DMA_TO_DEVICE);
2399		− }
2400		− return NETDEV_TX_BUSY;
2401		− }
	2407	+ dptr = g->sg_dma_ptr;
2402	2408
2403	2409	ndata.cmd.cmd3.dptr = dptr;
2404	2410	finfo->dptr = dptr;

drivers/net/ethernet/cavium/liquidio/octeon_config.h+3−3

@@ -71,17 +71,17 @@ · 70 unmodified lines
71	71	#define CN23XX_MAX_RINGS_PER_VF 8
72	72
73	73	#define CN23XX_MAX_INPUT_QUEUES CN23XX_MAX_RINGS_PER_PF
74		−#define CN23XX_MAX_IQ_DESCRIPTORS 2048
	74	+#define CN23XX_MAX_IQ_DESCRIPTORS 512
75	75	#define CN23XX_DB_MIN 1
76	76	#define CN23XX_DB_MAX 8
77	77	#define CN23XX_DB_TIMEOUT 1
78	78
79	79	#define CN23XX_MAX_OUTPUT_QUEUES CN23XX_MAX_RINGS_PER_PF
80		−#define CN23XX_MAX_OQ_DESCRIPTORS 2048
	80	+#define CN23XX_MAX_OQ_DESCRIPTORS 512
81	81	#define CN23XX_OQ_BUF_SIZE 1536
82	82	#define CN23XX_OQ_PKTSPER_INTR 128
83	83	/#define CAVIUM_ONLY_CN23XX_RX_PERF/
84		−#define CN23XX_OQ_REFIL_THRESHOLD 128
	84	+#define CN23XX_OQ_REFIL_THRESHOLD 16
85	85
86	86	#define CN23XX_OQ_INTR_PKT 64
87	87	#define CN23XX_OQ_INTR_TIME 100

drivers/net/ethernet/cavium/liquidio/octeon_droq.c+2−15

@@ -155,11 +155,6 @@ octeon_droq_destroy_ring_buffers(struct octeon_device *oct, · 154 unmodified lines
155	155	recv_buffer_destroy(droq->recv_buf_list[i].buffer,
156	156	pg_info);
157	157
158		− if (droq->desc_ring && droq->desc_ring[i].info_ptr)
159		− lio_unmap_ring_info(oct->pci_dev,
160		− (u64)droq->
161		− desc_ring[i].info_ptr,
162		− OCT_DROQ_INFO_SIZE);
163	158	droq->recv_buf_list[i].buffer = NULL;
164	159	}
165	160
@@ -211,10 +206,7 @@ int octeon_delete_droq(struct octeon_device *oct, u32 q_no) · 45 unmodified lines
211	206	vfree(droq->recv_buf_list);
212	207
213	208	if (droq->info_base_addr)
214		− cnnic_free_aligned_dma(oct->pci_dev, droq->info_list,
215		− droq->info_alloc_size,
216		− droq->info_base_addr,
217		− droq->info_list_dma);
	209	+ lio_free_info_buffer(oct, droq);
218	210
219	211	if (droq->desc_ring)
220	212	lio_dma_free(oct, (droq->max_count * OCT_DROQ_DESC_SIZE),
@@ -294,12 +286,7 @@ int octeon_init_droq(struct octeon_device *oct, · 73 unmodified lines
294	286	dev_dbg(&oct->pci_dev->dev, "droq[%d]: num_desc: %d\n", q_no,
295	287	droq->max_count);
296	288
297		− droq->info_list =
298		− cnnic_numa_alloc_aligned_dma((droq->max_count *
299		− OCT_DROQ_INFO_SIZE),
300		− &droq->info_alloc_size,
301		− &droq->info_base_addr,
302		− numa_node);
	289	+ droq->info_list = lio_alloc_info_buffer(oct, droq);
303	290	if (!droq->info_list) {
304	291	dev_err(&oct->pci_dev->dev, "Cannot allocate memory for info list.\n");
305	292	lio_dma_free(oct, (droq->max_count * OCT_DROQ_DESC_SIZE),

drivers/net/ethernet/cavium/liquidio/octeon_droq.h+2−2

@@ -325,10 +325,10 @@ struct octeon_droq { · 324 unmodified lines
325	325	size_t desc_ring_dma;
326	326
327	327	/** Info ptr list are allocated at this virtual address. */
328		− size_t info_base_addr;
	328	+ void *info_base_addr;
329	329
330	330	/** DMA mapped address of the info list */
331		− size_t info_list_dma;
	331	+ dma_addr_t info_list_dma;
332	332
333	333	/** Allocated size of info list. */
334	334	u32 info_alloc_size;

drivers/net/ethernet/cavium/liquidio/octeon_main.h+0−42

@@ -140,48 +140,6 @@ err_release_region: · 139 unmodified lines
140	140	return 1;
141	141	}
142	142
143		−static inline void *
144		−cnnic_numa_alloc_aligned_dma(u32 size,
145		− u32 *alloc_size,
146		− size_t *orig_ptr,
147		− int numa_node)
148		−{
149		− int retries = 0;
150		− void *ptr = NULL;
151		−
152		−#define OCTEON_MAX_ALLOC_RETRIES 1
153		− do {
154		− struct page *page = NULL;
155		−
156		− page = alloc_pages_node(numa_node,
157		− GFP_KERNEL,
158		− get_order(size));
159		− if (!page)
160		− page = alloc_pages(GFP_KERNEL,
161		− get_order(size));
162		− ptr = (void *)page_address(page);
163		− if ((unsigned long)ptr & 0x07) {
164		− __free_pages(page, get_order(size));
165		− ptr = NULL;
166		− /* Increment the size required if the first
167		− * attempt failed.
168		− */
169		− if (!retries)
170		− size += 7;
171		− }
172		− retries++;
173		− } while ((retries <= OCTEON_MAX_ALLOC_RETRIES) && !ptr);
174		−
175		− *alloc_size = size;
176		− *orig_ptr = (unsigned long)ptr;
177		− if ((unsigned long)ptr & 0x07)
178		− ptr = (void *)(((unsigned long)ptr + 7) & ~(7UL));
179		− return ptr;
180		−}
181		−
182		−#define cnnic_free_aligned_dma(pci_dev, ptr, size, orig_ptr, dma_addr) \
183		− free_pages(orig_ptr, get_order(size))
184		−
185	143	static inline int
186	144	sleep_cond(wait_queue_head_t wait_queue, int condition)
187	145	{

drivers/net/ethernet/cavium/liquidio/octeon_network.h+27−16

@@ -62,6 +62,9 @@ struct lio { · 61 unmodified lines
62	62
63	63	/** Array of gather component linked lists */
64	64	struct list_head *glist;
	65	+ void **glists_virt_base;
	66	+ dma_addr_t *glists_dma_base;
	67	+ u32 glist_entry_size;
65	68
66	69	/** Pointer to the NIC properties for the Octeon device this network
67	70	* interface is associated with.
@@ -344,6 +347,29 @@ static inline void tx_buffer_free(void *buffer) · 276 unmodified lines
344	347	#define lio_dma_free(oct, size, virt_addr, dma_addr) \
345	348	dma_free_coherent(&(oct)->pci_dev->dev, size, virt_addr, dma_addr)
346	349
	350	+static inline void *
	351	+lio_alloc_info_buffer(struct octeon_device *oct,
	352	+ struct octeon_droq *droq)
	353	+{
	354	+ void *virt_ptr;
	355	+
	356	+ virt_ptr = lio_dma_alloc(oct, (droq->max_count * OCT_DROQ_INFO_SIZE),
	357	+ &droq->info_list_dma);
	358	+ if (virt_ptr) {
	359	+ droq->info_alloc_size = droq->max_count * OCT_DROQ_INFO_SIZE;
	360	+ droq->info_base_addr = virt_ptr;
	361	+ }
	362	+
	363	+ return virt_ptr;
	364	+}
	365	+
	366	+static inline void lio_free_info_buffer(struct octeon_device *oct,
	367	+ struct octeon_droq *droq)
	368	+{
	369	+ lio_dma_free(oct, droq->info_alloc_size, droq->info_base_addr,
	370	+ droq->info_list_dma);
	371	+}
	372	+
347	373	static inline
348	374	void get_rbd(struct sk_buff skb)
349	375	{
@@ -359,22 +385,7 @@ void get_rbd(struct sk_buff skb) · 9 unmodified lines
359	385	static inline u64
360	386	lio_map_ring_info(struct octeon_droq *droq, u32 i)
361	387	{
362		− dma_addr_t dma_addr;
363		− struct octeon_device *oct = droq->oct_dev;
364		−
365		− dma_addr = dma_map_single(&oct->pci_dev->dev, &droq->info_list[i],
366		− OCT_DROQ_INFO_SIZE, DMA_FROM_DEVICE);
367		−
368		− WARN_ON(dma_mapping_error(&oct->pci_dev->dev, dma_addr));
369		−
370		− return (u64)dma_addr;
371		−}
372		−
373		−static inline void
374		−lio_unmap_ring_info(struct pci_dev *pci_dev,
375		− u64 info_ptr, u32 size)
376		−{
377		− dma_unmap_single(&pci_dev->dev, info_ptr, size, DMA_FROM_DEVICE);
	388	+ return droq->info_list_dma + (i * sizeof(struct octeon_droq_info));
378	389	}
379	390
380	391	static inline u64

drivers/net/ethernet/cavium/thunder/nic.h+1−0

@@ -269,6 +269,7 @@ struct nicvf { · 268 unmodified lines
269	269	#define MAX_QUEUES_PER_QSET 8
270	270	struct queue_set *qs;
271	271	struct nicvf_cq_poll *napi[8];
	272	+ void *iommu_domain;
272	273	u8 vf_id;
273	274	u8 sqs_id;
274	275	bool sqs_mode;

drivers/net/ethernet/cavium/thunder/nicvf_main.c+11−1

@@ -16,6 +16,7 @@ · 15 unmodified lines
16	16	#include <linux/log2.h>
17	17	#include <linux/prefetch.h>
18	18	#include <linux/irq.h>
	19	+#include <linux/iommu.h>
19	20
20	21	#include "nic_reg.h"
21	22	#include "nic.h"
@@ -525,7 +526,12 @@ static void nicvf_snd_pkt_handler(struct net_device *netdev, · 503 unmodified lines
525	526	/* Get actual TSO descriptors and free them */
526	527	tso_sqe =
527	528	(struct sq_hdr_subdesc *)GET_SQ_DESC(sq, hdr->rsvd2);
	529	+ nicvf_unmap_sndq_buffers(nic, sq, hdr->rsvd2,
	530	+ tso_sqe->subdesc_cnt);
528	531	nicvf_put_sq_desc(sq, tso_sqe->subdesc_cnt + 1);
	532	+ } else {
	533	+ nicvf_unmap_sndq_buffers(nic, sq, cqe_tx->sqe_ptr,
	534	+ hdr->subdesc_cnt);
529	535	}
530	536	nicvf_put_sq_desc(sq, hdr->subdesc_cnt + 1);
531	537	prefetch(skb);
@@ -576,6 +582,7 @@ static void nicvf_rcv_pkt_handler(struct net_device *netdev, · 44 unmodified lines
576	582	{
577	583	struct sk_buff *skb;
578	584	struct nicvf *nic = netdev_priv(netdev);
	585	+ struct nicvf *snic = nic;
579	586	int err = 0;
580	587	int rq_idx;
581	588
@@ -592,7 +599,7 @@ static void nicvf_rcv_pkt_handler(struct net_device *netdev, · 10 unmodified lines
592	599	if (err && !cqe_rx->rb_cnt)
593	600	return;
594	601
595		− skb = nicvf_get_rcv_skb(nic, cqe_rx);
	602	+ skb = nicvf_get_rcv_skb(snic, cqe_rx);
596	603	if (!skb) {
597	604	netdev_dbg(nic->netdev, "Packet not received\n");
598	605	return;
@@ -1643,6 +1650,9 @@ static int nicvf_probe(struct pci_dev pdev, const struct pci_device_id ent) · 1044 unmodified lines
1643	1650	if (!pass1_silicon(nic->pdev))
1644	1651	nic->hw_tso = true;
1645	1652
	1653	+ /* Get iommu domain for iova to physical addr conversion */
	1654	+ nic->iommu_domain = iommu_get_domain_for_dev(dev);
	1655	+
1646	1656	pci_read_config_word(nic->pdev, PCI_SUBSYSTEM_ID, &sdevid);
1647	1657	if (sdevid == 0xA134)
1648	1658	nic->t88 = true;

drivers/net/ethernet/cavium/thunder/nicvf_queues.c+145−39

@@ -10,6 +10,7 @@ · 9 unmodified lines
10	10	#include <linux/netdevice.h>
11	11	#include <linux/ip.h>
12	12	#include <linux/etherdevice.h>
	13	+#include <linux/iommu.h>
13	14	#include <net/ip.h>
14	15	#include <net/tso.h>
15	16
@@ -18,6 +19,16 @@ · 2 unmodified lines
18	19	#include "q_struct.h"
19	20	#include "nicvf_queues.h"
20	21
	22	+#define NICVF_PAGE_ORDER ((PAGE_SIZE <= 4096) ? PAGE_ALLOC_COSTLY_ORDER : 0)
	23	+
	24	+static inline u64 nicvf_iova_to_phys(struct nicvf *nic, dma_addr_t dma_addr)
	25	+{
	26	+ /* Translation is installed only when IOMMU is present */
	27	+ if (nic->iommu_domain)
	28	+ return iommu_iova_to_phys(nic->iommu_domain, dma_addr);
	29	+ return dma_addr;
	30	+}
	31	+
21	32	static void nicvf_get_page(struct nicvf *nic)
22	33	{
23	34	if (!nic->rb_pageref \|\| !nic->rb_page)
@@ -87,7 +98,7 @@ static void nicvf_free_q_desc_mem(struct nicvf nic, struct q_desc_mem dmem) · 63 unmodified lines
87	98	static inline int nicvf_alloc_rcv_buffer(struct nicvf *nic, gfp_t gfp,
88	99	u32 buf_len, u64 **rbuf)
89	100	{
90		− int order = (PAGE_SIZE <= 4096) ? PAGE_ALLOC_COSTLY_ORDER : 0;
	101	+ int order = NICVF_PAGE_ORDER;
91	102
92	103	/* Check if request can be accomodated in previous allocated page */
93	104	if (nic->rb_page &&
@@ -97,22 +108,27 @@ static inline int nicvf_alloc_rcv_buffer(struct nicvf *nic, gfp_t gfp, · 3 unmodified lines
97	108	}
98	109
99	110	nicvf_get_page(nic);
100		− nic->rb_page = NULL;
101	111
102	112	/* Allocate a new page */
	113	+ nic->rb_page = alloc_pages(gfp \| __GFP_COMP \| __GFP_NOWARN,
	114	+ order);
103	115	if (!nic->rb_page) {
104		− nic->rb_page = alloc_pages(gfp \| __GFP_COMP \| __GFP_NOWARN,
105		− order);
106		− if (!nic->rb_page) {
107		− this_cpu_inc(nic->pnicvf->drv_stats->
108		− rcv_buffer_alloc_failures);
109		− return -ENOMEM;
110		− }
111		− nic->rb_page_offset = 0;
	116	+ this_cpu_inc(nic->pnicvf->drv_stats->rcv_buffer_alloc_failures);
	117	+ return -ENOMEM;
112	118	}
113		−
	119	+ nic->rb_page_offset = 0;
114	120	ret:
115		− rbuf = (u64 )((u64)page_address(nic->rb_page) + nic->rb_page_offset);
	121	+ /* HW will ensure data coherency, CPU sync not required */
	122	+ rbuf = (u64 )((u64)dma_map_page_attrs(&nic->pdev->dev, nic->rb_page,
	123	+ nic->rb_page_offset, buf_len,
	124	+ DMA_FROM_DEVICE,
	125	+ DMA_ATTR_SKIP_CPU_SYNC));
	126	+ if (dma_mapping_error(&nic->pdev->dev, (dma_addr_t)*rbuf)) {
	127	+ if (!nic->rb_page_offset)
	128	+ __free_pages(nic->rb_page, order);
	129	+ nic->rb_page = NULL;
	130	+ return -ENOMEM;
	131	+ }
116	132	nic->rb_page_offset += buf_len;
117	133
118	134	return 0;
@@ -158,16 +174,21 @@ static int nicvf_init_rbdr(struct nicvf nic, struct rbdr rbdr, · 39 unmodified lines
158	174	rbdr->dma_size = buf_size;
159	175	rbdr->enable = true;
160	176	rbdr->thresh = RBDR_THRESH;
	177	+ rbdr->head = 0;
	178	+ rbdr->tail = 0;
161	179
162	180	nic->rb_page = NULL;
163	181	for (idx = 0; idx < ring_len; idx++) {
164	182	err = nicvf_alloc_rcv_buffer(nic, GFP_KERNEL, RCV_FRAG_LEN,
165	183	&rbuf);
166		− if (err)
	184	+ if (err) {
	185	+ /* To free already allocated and mapped ones */
	186	+ rbdr->tail = idx - 1;
167	187	return err;
	188	+ }
168	189
169	190	desc = GET_RBDR_DESC(rbdr, idx);
170		− desc->buf_addr = virt_to_phys(rbuf) >> NICVF_RCV_BUF_ALIGN;
	191	+ desc->buf_addr = (u64)rbuf >> NICVF_RCV_BUF_ALIGN;
171	192	}
172	193
173	194	nicvf_get_page(nic);
@@ -179,7 +200,7 @@ static int nicvf_init_rbdr(struct nicvf nic, struct rbdr rbdr, · 5 unmodified lines
179	200	static void nicvf_free_rbdr(struct nicvf nic, struct rbdr rbdr)
180	201	{
181	202	int head, tail;
182		− u64 buf_addr;
	203	+ u64 buf_addr, phys_addr;
183	204	struct rbdr_entry_t *desc;
184	205
185	206	if (!rbdr)
@@ -192,18 +213,26 @@ static void nicvf_free_rbdr(struct nicvf nic, struct rbdr rbdr) · 6 unmodified lines
192	213	head = rbdr->head;
193	214	tail = rbdr->tail;
194	215
195		− /* Free SKBs */
	216	+ /* Release page references */
196	217	while (head != tail) {
197	218	desc = GET_RBDR_DESC(rbdr, head);
198		− buf_addr = desc->buf_addr << NICVF_RCV_BUF_ALIGN;
199		− put_page(virt_to_page(phys_to_virt(buf_addr)));
	219	+ buf_addr = ((u64)desc->buf_addr) << NICVF_RCV_BUF_ALIGN;
	220	+ phys_addr = nicvf_iova_to_phys(nic, buf_addr);
	221	+ dma_unmap_page_attrs(&nic->pdev->dev, buf_addr, RCV_FRAG_LEN,
	222	+ DMA_FROM_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
	223	+ if (phys_addr)
	224	+ put_page(virt_to_page(phys_to_virt(phys_addr)));
200	225	head++;
201	226	head &= (rbdr->dmem.q_len - 1);
202	227	}
203		− /* Free SKB of tail desc */
	228	+ /* Release buffer of tail desc */
204	229	desc = GET_RBDR_DESC(rbdr, tail);
205		− buf_addr = desc->buf_addr << NICVF_RCV_BUF_ALIGN;
206		− put_page(virt_to_page(phys_to_virt(buf_addr)));
	230	+ buf_addr = ((u64)desc->buf_addr) << NICVF_RCV_BUF_ALIGN;
	231	+ phys_addr = nicvf_iova_to_phys(nic, buf_addr);
	232	+ dma_unmap_page_attrs(&nic->pdev->dev, buf_addr, RCV_FRAG_LEN,
	233	+ DMA_FROM_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
	234	+ if (phys_addr)
	235	+ put_page(virt_to_page(phys_to_virt(phys_addr)));
207	236
208	237	/* Free RBDR ring */
209	238	nicvf_free_q_desc_mem(nic, &rbdr->dmem);
@@ -250,7 +279,7 @@ refill: · 40 unmodified lines
250	279	break;
251	280
252	281	desc = GET_RBDR_DESC(rbdr, tail);
253		− desc->buf_addr = virt_to_phys(rbuf) >> NICVF_RCV_BUF_ALIGN;
	282	+ desc->buf_addr = (u64)rbuf >> NICVF_RCV_BUF_ALIGN;
254	283	refill_rb_cnt--;
255	284	new_rb++;
256	285	}
@@ -361,9 +390,29 @@ static int nicvf_init_snd_queue(struct nicvf *nic, · 104 unmodified lines
361	390	return 0;
362	391	}
363	392
	393	+void nicvf_unmap_sndq_buffers(struct nicvf nic, struct snd_queue sq,
	394	+ int hdr_sqe, u8 subdesc_cnt)
	395	+{
	396	+ u8 idx;
	397	+ struct sq_gather_subdesc *gather;
	398	+
	399	+ /* Unmap DMA mapped skb data buffers */
	400	+ for (idx = 0; idx < subdesc_cnt; idx++) {
	401	+ hdr_sqe++;
	402	+ hdr_sqe &= (sq->dmem.q_len - 1);
	403	+ gather = (struct sq_gather_subdesc *)GET_SQ_DESC(sq, hdr_sqe);
	404	+ /* HW will ensure data coherency, CPU sync not required */
	405	+ dma_unmap_page_attrs(&nic->pdev->dev, gather->addr,
	406	+ gather->size, DMA_TO_DEVICE,
	407	+ DMA_ATTR_SKIP_CPU_SYNC);
	408	+ }
	409	+}
	410	+
364	411	static void nicvf_free_snd_queue(struct nicvf nic, struct snd_queue sq)
365	412	{
366	413	struct sk_buff *skb;
	414	+ struct sq_hdr_subdesc *hdr;
	415	+ struct sq_hdr_subdesc *tso_sqe;
367	416
368	417	if (!sq)
369	418	return;
@@ -379,8 +428,22 @@ static void nicvf_free_snd_queue(struct nicvf nic, struct snd_queue sq) · 9 unmodified lines
379	428	smp_rmb();
380	429	while (sq->head != sq->tail) {
381	430	skb = (struct sk_buff *)sq->skbuff[sq->head];
382		− if (skb)
383		− dev_kfree_skb_any(skb);
	431	+ if (!skb)
	432	+ goto next;
	433	+ hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, sq->head);
	434	+ /* Check for dummy descriptor used for HW TSO offload on 88xx */
	435	+ if (hdr->dont_send) {
	436	+ /* Get actual TSO descriptors and unmap them */
	437	+ tso_sqe =
	438	+ (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, hdr->rsvd2);
	439	+ nicvf_unmap_sndq_buffers(nic, sq, hdr->rsvd2,
	440	+ tso_sqe->subdesc_cnt);
	441	+ } else {
	442	+ nicvf_unmap_sndq_buffers(nic, sq, sq->head,
	443	+ hdr->subdesc_cnt);
	444	+ }
	445	+ dev_kfree_skb_any(skb);
	446	+next:
384	447	sq->head++;
385	448	sq->head &= (sq->dmem.q_len - 1);
386	449	}
@@ -559,9 +622,11 @@ static void nicvf_rcv_queue_config(struct nicvf nic, struct queue_set qs, · 172 unmodified lines
559	622	nicvf_send_msg_to_pf(nic, &mbx);
560	623
561	624	if (!nic->sqs_mode && (qidx == 0)) {
562		− /* Enable checking L3/L4 length and TCP/UDP checksums */
	625	+ /* Enable checking L3/L4 length and TCP/UDP checksums
	626	+ * Also allow IPv6 pkts with zero UDP checksum.
	627	+ */
563	628	nicvf_queue_reg_write(nic, NIC_QSET_RQ_GEN_CFG, 0,
564		− (BIT(24) \| BIT(23) \| BIT(21)));
	629	+ (BIT(24) \| BIT(23) \| BIT(21) \| BIT(20)));
565	630	nicvf_config_vlan_stripping(nic, nic->netdev->features);
566	631	}
567	632
@@ -882,6 +947,14 @@ static inline int nicvf_get_sq_desc(struct snd_queue *sq, int desc_cnt) · 314 unmodified lines
882	947	return qentry;
883	948	}
884	949
	950	+/* Rollback to previous tail pointer when descriptors not used */
	951	+static inline void nicvf_rollback_sq_desc(struct snd_queue *sq,
	952	+ int qentry, int desc_cnt)
	953	+{
	954	+ sq->tail = qentry;
	955	+ atomic_add(desc_cnt, &sq->free_cnt);
	956	+}
	957	+
885	958	/* Free descriptor back to SQ for future use */
886	959	void nicvf_put_sq_desc(struct snd_queue *sq, int desc_cnt)
887	960	{
@@ -1207,8 +1280,9 @@ int nicvf_sq_append_skb(struct nicvf nic, struct snd_queue sq, · 319 unmodified lines
1207	1280	struct sk_buff *skb, u8 sq_num)
1208	1281	{
1209	1282	int i, size;
1210		− int subdesc_cnt, tso_sqe = 0;
	1283	+ int subdesc_cnt, hdr_sqe = 0;
1211	1284	int qentry;
	1285	+ u64 dma_addr;
1212	1286
1213	1287	subdesc_cnt = nicvf_sq_subdesc_required(nic, skb);
1214	1288	if (subdesc_cnt > atomic_read(&sq->free_cnt))
@@ -1223,12 +1297,21 @@ int nicvf_sq_append_skb(struct nicvf nic, struct snd_queue sq, · 8 unmodified lines
1223	1297	/* Add SQ header subdesc */
1224	1298	nicvf_sq_add_hdr_subdesc(nic, sq, qentry, subdesc_cnt - 1,
1225	1299	skb, skb->len);
1226		− tso_sqe = qentry;
	1300	+ hdr_sqe = qentry;
1227	1301
1228	1302	/* Add SQ gather subdescs */
1229	1303	qentry = nicvf_get_nxt_sqentry(sq, qentry);
1230	1304	size = skb_is_nonlinear(skb) ? skb_headlen(skb) : skb->len;
1231		− nicvf_sq_add_gather_subdesc(sq, qentry, size, virt_to_phys(skb->data));
	1305	+ /* HW will ensure data coherency, CPU sync not required */
	1306	+ dma_addr = dma_map_page_attrs(&nic->pdev->dev, virt_to_page(skb->data),
	1307	+ offset_in_page(skb->data), size,
	1308	+ DMA_TO_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
	1309	+ if (dma_mapping_error(&nic->pdev->dev, dma_addr)) {
	1310	+ nicvf_rollback_sq_desc(sq, qentry, subdesc_cnt);
	1311	+ return 0;
	1312	+ }
	1313	+
	1314	+ nicvf_sq_add_gather_subdesc(sq, qentry, size, dma_addr);
1232	1315
1233	1316	/* Check for scattered buffer */
1234	1317	if (!skb_is_nonlinear(skb))
@@ -1241,15 +1324,26 @@ int nicvf_sq_append_skb(struct nicvf nic, struct snd_queue sq, · 6 unmodified lines
1241	1324
1242	1325	qentry = nicvf_get_nxt_sqentry(sq, qentry);
1243	1326	size = skb_frag_size(frag);
1244		− nicvf_sq_add_gather_subdesc(sq, qentry, size,
1245		− virt_to_phys(
1246		− skb_frag_address(frag)));
	1327	+ dma_addr = dma_map_page_attrs(&nic->pdev->dev,
	1328	+ skb_frag_page(frag),
	1329	+ frag->page_offset, size,
	1330	+ DMA_TO_DEVICE,
	1331	+ DMA_ATTR_SKIP_CPU_SYNC);
	1332	+ if (dma_mapping_error(&nic->pdev->dev, dma_addr)) {
	1333	+ /* Free entire chain of mapped buffers
	1334	+ * here 'i' = frags mapped + above mapped skb->data
	1335	+ */
	1336	+ nicvf_unmap_sndq_buffers(nic, sq, hdr_sqe, i);
	1337	+ nicvf_rollback_sq_desc(sq, qentry, subdesc_cnt);
	1338	+ return 0;
	1339	+ }
	1340	+ nicvf_sq_add_gather_subdesc(sq, qentry, size, dma_addr);
1247	1341	}
1248	1342
1249	1343	doorbell:
1250	1344	if (nic->t88 && skb_shinfo(skb)->gso_size) {
1251	1345	qentry = nicvf_get_nxt_sqentry(sq, qentry);
1252		− nicvf_sq_add_cqe_subdesc(sq, qentry, tso_sqe, skb);
	1346	+ nicvf_sq_add_cqe_subdesc(sq, qentry, hdr_sqe, skb);
1253	1347	}
1254	1348
1255	1349	nicvf_sq_doorbell(nic, skb, sq_num, subdesc_cnt);
@@ -1282,6 +1376,7 @@ struct sk_buff nicvf_get_rcv_skb(struct nicvf nic, struct cqe_rx_t *cqe_rx) · 26 unmodified lines
1282	1376	int offset;
1283	1377	u16 *rb_lens = NULL;
1284	1378	u64 *rb_ptrs = NULL;
	1379	+ u64 phys_addr;
1285	1380
1286	1381	rb_lens = (void )cqe_rx + (3 sizeof(u64));
1287	1382	/* Except 88xx pass1 on all other chips CQE_RX2_S is added to
@@ -1296,15 +1391,23 @@ struct sk_buff nicvf_get_rcv_skb(struct nicvf nic, struct cqe_rx_t *cqe_rx) · 8 unmodified lines
1296	1391	else
1297	1392	rb_ptrs = (void )cqe_rx + (7 sizeof(u64));
1298	1393
1299		− netdev_dbg(nic->netdev, "%s rb_cnt %d rb0_ptr %llx rb0_sz %d\n",
1300		− __func__, cqe_rx->rb_cnt, cqe_rx->rb0_ptr, cqe_rx->rb0_sz);
1301		−
1302	1394	for (frag = 0; frag < cqe_rx->rb_cnt; frag++) {
1303	1395	payload_len = rb_lens[frag_num(frag)];
	1396	+ phys_addr = nicvf_iova_to_phys(nic, *rb_ptrs);
	1397	+ if (!phys_addr) {
	1398	+ if (skb)
	1399	+ dev_kfree_skb_any(skb);
	1400	+ return NULL;
	1401	+ }
	1402	+
1304	1403	if (!frag) {
1305	1404	/* First fragment */
	1405	+ dma_unmap_page_attrs(&nic->pdev->dev,
	1406	+ *rb_ptrs - cqe_rx->align_pad,
	1407	+ RCV_FRAG_LEN, DMA_FROM_DEVICE,
	1408	+ DMA_ATTR_SKIP_CPU_SYNC);
1306	1409	skb = nicvf_rb_ptr_to_skb(nic,
1307		− *rb_ptrs - cqe_rx->align_pad,
	1410	+ phys_addr - cqe_rx->align_pad,
1308	1411	payload_len);
1309	1412	if (!skb)
1310	1413	return NULL;
@@ -1312,8 +1415,11 @@ struct sk_buff nicvf_get_rcv_skb(struct nicvf nic, struct cqe_rx_t *cqe_rx) · 1 unmodified line
1312	1415	skb_put(skb, payload_len);
1313	1416	} else {
1314	1417	/* Add fragments */
1315		− page = virt_to_page(phys_to_virt(*rb_ptrs));
1316		− offset = phys_to_virt(*rb_ptrs) - page_address(page);
	1418	+ dma_unmap_page_attrs(&nic->pdev->dev, *rb_ptrs,
	1419	+ RCV_FRAG_LEN, DMA_FROM_DEVICE,
	1420	+ DMA_ATTR_SKIP_CPU_SYNC);
	1421	+ page = virt_to_page(phys_to_virt(phys_addr));
	1422	+ offset = phys_to_virt(phys_addr) - page_address(page);
1317	1423	skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, page,
1318	1424	offset, payload_len, RCV_FRAG_LEN);
1319	1425	}

drivers/net/ethernet/cavium/thunder/nicvf_queues.h+3−1

@@ -87,7 +87,7 @@ · 86 unmodified lines
87	87	#define RCV_BUF_COUNT (1ULL << (RBDR_SIZE + 13))
88	88	#define MAX_RCV_BUF_COUNT (1ULL << (RBDR_SIZE6 + 13))
89	89	#define RBDR_THRESH (RCV_BUF_COUNT / 2)
90		−#define DMA_BUFFER_LEN 2048 /* In multiples of 128bytes */
	90	+#define DMA_BUFFER_LEN 1536 /* In multiples of 128bytes */
91	91	#define RCV_FRAG_LEN (SKB_DATA_ALIGN(DMA_BUFFER_LEN + NET_SKB_PAD) + \
92	92	SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
93	93
@@ -301,6 +301,8 @@ struct queue_set { · 207 unmodified lines
301	301
302	302	#define CQ_ERR_MASK (CQ_WR_FULL \| CQ_WR_DISABLE \| CQ_WR_FAULT)
303	303
	304	+void nicvf_unmap_sndq_buffers(struct nicvf nic, struct snd_queue sq,
	305	+ int hdr_sqe, u8 subdesc_cnt);
304	306	void nicvf_config_vlan_stripping(struct nicvf *nic,
305	307	netdev_features_t features);
306	308	int nicvf_set_qset_resources(struct nicvf *nic);

drivers/net/ethernet/cavium/thunder/thunder_bgx.c+45−19

@@ -123,14 +123,44 @@ static int bgx_poll_reg(struct bgx *bgx, u8 lmac, u64 reg, u64 mask, bool zero) · 122 unmodified lines
123	123	return 1;
124	124	}
125	125
	126	+static int max_bgx_per_node;
	127	+static void set_max_bgx_per_node(struct pci_dev *pdev)
	128	+{
	129	+ u16 sdevid;
	130	+
	131	+ if (max_bgx_per_node)
	132	+ return;
	133	+
	134	+ pci_read_config_word(pdev, PCI_SUBSYSTEM_ID, &sdevid);
	135	+ switch (sdevid) {
	136	+ case PCI_SUBSYS_DEVID_81XX_BGX:
	137	+ max_bgx_per_node = MAX_BGX_PER_CN81XX;
	138	+ break;
	139	+ case PCI_SUBSYS_DEVID_83XX_BGX:
	140	+ max_bgx_per_node = MAX_BGX_PER_CN83XX;
	141	+ break;
	142	+ case PCI_SUBSYS_DEVID_88XX_BGX:
	143	+ default:
	144	+ max_bgx_per_node = MAX_BGX_PER_CN88XX;
	145	+ break;
	146	+ }
	147	+}
	148	+
	149	+static struct bgx *get_bgx(int node, int bgx_idx)
	150	+{
	151	+ int idx = (node * max_bgx_per_node) + bgx_idx;
	152	+
	153	+ return bgx_vnic[idx];
	154	+}
	155	+
126	156	/* Return number of BGX present in HW */
127	157	unsigned bgx_get_map(int node)
128	158	{
129	159	int i;
130	160	unsigned map = 0;
131	161
132		− for (i = 0; i < MAX_BGX_PER_NODE; i++) {
133		− if (bgx_vnic[(node * MAX_BGX_PER_NODE) + i])
	162	+ for (i = 0; i < max_bgx_per_node; i++) {
	163	+ if (bgx_vnic[(node * max_bgx_per_node) + i])
134	164	map \|= (1 << i);
135	165	}
136	166
@@ -143,7 +173,7 @@ int bgx_get_lmac_count(int node, int bgx_idx) · 6 unmodified lines
143	173	{
144	174	struct bgx *bgx;
145	175
146		− bgx = bgx_vnic[(node * MAX_BGX_PER_NODE) + bgx_idx];
	176	+ bgx = get_bgx(node, bgx_idx);
147	177	if (bgx)
148	178	return bgx->lmac_count;
149	179
@@ -158,7 +188,7 @@ void bgx_get_lmac_link_state(int node, int bgx_idx, int lmacid, void *status) · 8 unmodified lines
158	188	struct bgx *bgx;
159	189	struct lmac *lmac;
160	190
161		− bgx = bgx_vnic[(node * MAX_BGX_PER_NODE) + bgx_idx];
	191	+ bgx = get_bgx(node, bgx_idx);
162	192	if (!bgx)
163	193	return;
164	194
@@ -172,7 +202,7 @@ EXPORT_SYMBOL(bgx_get_lmac_link_state); · 7 unmodified lines
172	202
173	203	const u8 *bgx_get_lmac_mac(int node, int bgx_idx, int lmacid)
174	204	{
175		− struct bgx bgx = bgx_vnic[(node MAX_BGX_PER_NODE) + bgx_idx];
	205	+ struct bgx *bgx = get_bgx(node, bgx_idx);
176	206
177	207	if (bgx)
178	208	return bgx->lmac[lmacid].mac;
@@ -183,7 +213,7 @@ EXPORT_SYMBOL(bgx_get_lmac_mac); · 4 unmodified lines
183	213
184	214	void bgx_set_lmac_mac(int node, int bgx_idx, int lmacid, const u8 *mac)
185	215	{
186		− struct bgx bgx = bgx_vnic[(node MAX_BGX_PER_NODE) + bgx_idx];
	216	+ struct bgx *bgx = get_bgx(node, bgx_idx);
187	217
188	218	if (!bgx)
189	219	return;
@@ -194,7 +224,7 @@ EXPORT_SYMBOL(bgx_set_lmac_mac); · 4 unmodified lines
194	224
195	225	void bgx_lmac_rx_tx_enable(int node, int bgx_idx, int lmacid, bool enable)
196	226	{
197		− struct bgx bgx = bgx_vnic[(node MAX_BGX_PER_NODE) + bgx_idx];
	227	+ struct bgx *bgx = get_bgx(node, bgx_idx);
198	228	struct lmac *lmac;
199	229	u64 cfg;
200	230
@@ -217,7 +247,7 @@ EXPORT_SYMBOL(bgx_lmac_rx_tx_enable); · 16 unmodified lines
217	247	void bgx_lmac_get_pfc(int node, int bgx_idx, int lmacid, void *pause)
218	248	{
219	249	struct pfc pfc = (struct pfc )pause;
220		− struct bgx bgx = bgx_vnic[(node MAX_BGX_PER_CN88XX) + bgx_idx];
	250	+ struct bgx *bgx = get_bgx(node, bgx_idx);
221	251	struct lmac *lmac;
222	252	u64 cfg;
223	253
@@ -237,7 +267,7 @@ EXPORT_SYMBOL(bgx_lmac_get_pfc); · 13 unmodified lines
237	267	void bgx_lmac_set_pfc(int node, int bgx_idx, int lmacid, void *pause)
238	268	{
239	269	struct pfc pfc = (struct pfc )pause;
240		− struct bgx bgx = bgx_vnic[(node MAX_BGX_PER_CN88XX) + bgx_idx];
	270	+ struct bgx *bgx = get_bgx(node, bgx_idx);
241	271	struct lmac *lmac;
242	272	u64 cfg;
243	273
@@ -369,7 +399,7 @@ u64 bgx_get_rx_stats(int node, int bgx_idx, int lmac, int idx) · 125 unmodified lines
369	399	{
370	400	struct bgx *bgx;
371	401
372		− bgx = bgx_vnic[(node * MAX_BGX_PER_NODE) + bgx_idx];
	402	+ bgx = get_bgx(node, bgx_idx);
373	403	if (!bgx)
374	404	return 0;
375	405
@@ -383,7 +413,7 @@ u64 bgx_get_tx_stats(int node, int bgx_idx, int lmac, int idx) · 7 unmodified lines
383	413	{
384	414	struct bgx *bgx;
385	415
386		− bgx = bgx_vnic[(node * MAX_BGX_PER_NODE) + bgx_idx];
	416	+ bgx = get_bgx(node, bgx_idx);
387	417	if (!bgx)
388	418	return 0;
389	419
@@ -411,7 +441,7 @@ void bgx_lmac_internal_loopback(int node, int bgx_idx, · 21 unmodified lines
411	441	struct lmac *lmac;
412	442	u64 cfg;
413	443
414		− bgx = bgx_vnic[(node * MAX_BGX_PER_NODE) + bgx_idx];
	444	+ bgx = get_bgx(node, bgx_idx);
415	445	if (!bgx)
416	446	return;
417	447
@@ -1011,12 +1041,6 @@ static void bgx_print_qlm_mode(struct bgx *bgx, u8 lmacid) · 593 unmodified lines
1011	1041	dev_info(dev, "%s: 40G_KR4\n", (char *)str);
1012	1042	break;
1013	1043	case BGX_MODE_QSGMII:
1014		− if ((lmacid == 0) &&
1015		− (bgx_get_lane2sds_cfg(bgx, lmac) != lmacid))
1016		− return;
1017		− if ((lmacid == 2) &&
1018		− (bgx_get_lane2sds_cfg(bgx, lmac) == lmacid))
1019		− return;
1020	1044	dev_info(dev, "%s: QSGMII\n", (char *)str);
1021	1045	break;
1022	1046	case BGX_MODE_RGMII:
@@ -1334,11 +1358,13 @@ static int bgx_probe(struct pci_dev pdev, const struct pci_device_id ent) · 311 unmodified lines
1334	1358	goto err_release_regions;
1335	1359	}
1336	1360
	1361	+ set_max_bgx_per_node(pdev);
	1362	+
1337	1363	pci_read_config_word(pdev, PCI_DEVICE_ID, &sdevid);
1338	1364	if (sdevid != PCI_DEVICE_ID_THUNDER_RGX) {
1339	1365	bgx->bgx_id = (pci_resource_start(pdev,
1340	1366	PCI_CFG_REG_BAR_NUM) >> 24) & BGX_ID_MASK;
1341		− bgx->bgx_id += nic_get_node_id(pdev) * MAX_BGX_PER_NODE;
	1367	+ bgx->bgx_id += nic_get_node_id(pdev) * max_bgx_per_node;
1342	1368	bgx->max_lmac = MAX_LMAC_PER_BGX;
1343	1369	bgx_vnic[bgx->bgx_id] = bgx;
1344	1370	} else {

drivers/net/ethernet/cavium/thunder/thunder_bgx.h+0−1

@@ -22,7 +22,6 @@ · 21 unmodified lines
22	22	#define MAX_BGX_PER_CN88XX 2
23	23	#define MAX_BGX_PER_CN81XX 3 /* 2 BGXs + 1 RGX */
24	24	#define MAX_BGX_PER_CN83XX 4
25		−#define MAX_BGX_PER_NODE 4
26	25	#define MAX_LMAC_PER_BGX 4
27	26	#define MAX_BGX_CHANS_PER_LMAC 16
28	27	#define MAX_DMAC_PER_LMAC 8

drivers/net/ethernet/ibm/emac/core.c+17−8

@@ -2589,8 +2589,6 @@ static int emac_dt_mdio_probe(struct emac_instance *dev) · 2588 unmodified lines
2589	2589	static int emac_dt_phy_connect(struct emac_instance *dev,
2590	2590	struct device_node *phy_handle)
2591	2591	{
2592		− int res;
2593		−
2594	2592	dev->phy.def = devm_kzalloc(&dev->ofdev->dev, sizeof(*dev->phy.def),
2595	2593	GFP_KERNEL);
2596	2594	if (!dev->phy.def)
@@ -2617,7 +2615,7 @@ static int emac_dt_phy_probe(struct emac_instance *dev) · 20 unmodified lines
2617	2615	{
2618	2616	struct device_node *np = dev->ofdev->dev.of_node;
2619	2617	struct device_node *phy_handle;
2620		− int res = 0;
	2618	+ int res = 1;
2621	2619
2622	2620	phy_handle = of_parse_phandle(np, "phy-handle", 0);
2623	2621
@@ -2714,13 +2712,24 @@ static int emac_init_phy(struct emac_instance *dev) · 90 unmodified lines
2714	2712	if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII)) {
2715	2713	int res = emac_dt_phy_probe(dev);
2716	2714
2717		− mutex_unlock(&emac_phy_map_lock);
2718		− if (!res)
	2715	+ switch (res) {
	2716	+ case 1:
	2717	+ /* No phy-handle property configured.
	2718	+ * Continue with the existing phy probe
	2719	+ * and setup code.
	2720	+ */
	2721	+ break;
	2722	+
	2723	+ case 0:
	2724	+ mutex_unlock(&emac_phy_map_lock);
2719	2725	goto init_phy;
2720	2726
2721		− dev_err(&dev->ofdev->dev, "failed to attach dt phy (%d).\n",
2722		− res);
2723		− return res;
	2727	+ default:
	2728	+ mutex_unlock(&emac_phy_map_lock);
	2729	+ dev_err(&dev->ofdev->dev, "failed to attach dt phy (%d).\n",
	2730	+ res);
	2731	+ return res;
	2732	+ }
2724	2733	}
2725	2734
2726	2735	if (dev->phy_address != 0xffffffff)

drivers/net/ethernet/ibm/ibmvnic.c+34−9

@@ -404,7 +404,7 @@ static int ibmvnic_open(struct net_device *netdev) · 403 unmodified lines
404	404	send_map_query(adapter);
405	405	for (i = 0; i < rxadd_subcrqs; i++) {
406	406	init_rx_pool(adapter, &adapter->rx_pool[i],
407		− IBMVNIC_BUFFS_PER_POOL, i,
	407	+ adapter->req_rx_add_entries_per_subcrq, i,
408	408	be64_to_cpu(size_array[i]), 1);
409	409	if (alloc_rx_pool(adapter, &adapter->rx_pool[i])) {
410	410	dev_err(dev, "Couldn't alloc rx pool\n");
@@ -419,23 +419,23 @@ static int ibmvnic_open(struct net_device *netdev) · 8 unmodified lines
419	419	for (i = 0; i < tx_subcrqs; i++) {
420	420	tx_pool = &adapter->tx_pool[i];
421	421	tx_pool->tx_buff =
422		− kcalloc(adapter->max_tx_entries_per_subcrq,
	422	+ kcalloc(adapter->req_tx_entries_per_subcrq,
423	423	sizeof(struct ibmvnic_tx_buff), GFP_KERNEL);
424	424	if (!tx_pool->tx_buff)
425	425	goto tx_pool_alloc_failed;
426	426
427	427	if (alloc_long_term_buff(adapter, &tx_pool->long_term_buff,
428		− adapter->max_tx_entries_per_subcrq *
	428	+ adapter->req_tx_entries_per_subcrq *
429	429	adapter->req_mtu))
430	430	goto tx_ltb_alloc_failed;
431	431
432	432	tx_pool->free_map =
433		− kcalloc(adapter->max_tx_entries_per_subcrq,
	433	+ kcalloc(adapter->req_tx_entries_per_subcrq,
434	434	sizeof(int), GFP_KERNEL);
435	435	if (!tx_pool->free_map)
436	436	goto tx_fm_alloc_failed;
437	437
438		− for (j = 0; j < adapter->max_tx_entries_per_subcrq; j++)
	438	+ for (j = 0; j < adapter->req_tx_entries_per_subcrq; j++)
439	439	tx_pool->free_map[j] = j;
440	440
441	441	tx_pool->consumer_index = 0;
@@ -705,6 +705,7 @@ static int ibmvnic_xmit(struct sk_buff skb, struct net_device netdev) · 263 unmodified lines
705	705	u8 hdrs = (u8 )&adapter->tx_rx_desc_req;
706	706	struct device *dev = &adapter->vdev->dev;
707	707	struct ibmvnic_tx_buff *tx_buff = NULL;
	708	+ struct ibmvnic_sub_crq_queue *tx_scrq;
708	709	struct ibmvnic_tx_pool *tx_pool;
709	710	unsigned int tx_send_failed = 0;
710	711	unsigned int tx_map_failed = 0;
@@ -724,6 +725,7 @@ static int ibmvnic_xmit(struct sk_buff skb, struct net_device netdev) · 13 unmodified lines
724	725	int ret = 0;
725	726
726	727	tx_pool = &adapter->tx_pool[queue_num];
	728	+ tx_scrq = adapter->tx_scrq[queue_num];
727	729	txq = netdev_get_tx_queue(netdev, skb_get_queue_mapping(skb));
728	730	handle_array = (u64 )((u8 )(adapter->login_rsp_buf) +
729	731	be32_to_cpu(adapter->login_rsp_buf->
@@ -744,7 +746,7 @@ static int ibmvnic_xmit(struct sk_buff skb, struct net_device netdev) · 14 unmodified lines
744	746
745	747	tx_pool->consumer_index =
746	748	(tx_pool->consumer_index + 1) %
747		− adapter->max_tx_entries_per_subcrq;
	749	+ adapter->req_tx_entries_per_subcrq;
748	750
749	751	tx_buff = &tx_pool->tx_buff[index];
750	752	tx_buff->skb = skb;
@@ -817,7 +819,7 @@ static int ibmvnic_xmit(struct sk_buff skb, struct net_device netdev) · 66 unmodified lines
817	819
818	820	if (tx_pool->consumer_index == 0)
819	821	tx_pool->consumer_index =
820		− adapter->max_tx_entries_per_subcrq - 1;
	822	+ adapter->req_tx_entries_per_subcrq - 1;
821	823	else
822	824	tx_pool->consumer_index--;
823	825
@@ -826,6 +828,14 @@ static int ibmvnic_xmit(struct sk_buff skb, struct net_device netdev) · 2 unmodified lines
826	828	ret = NETDEV_TX_BUSY;
827	829	goto out;
828	830	}
	831	+
	832	+ atomic_inc(&tx_scrq->used);
	833	+
	834	+ if (atomic_read(&tx_scrq->used) >= adapter->req_tx_entries_per_subcrq) {
	835	+ netdev_info(netdev, "Stopping queue %d\n", queue_num);
	836	+ netif_stop_subqueue(netdev, queue_num);
	837	+ }
	838	+
829	839	tx_packets++;
830	840	tx_bytes += skb->len;
831	841	txq->trans_start = jiffies;
@@ -1213,6 +1223,7 @@ static struct ibmvnic_sub_crq_queue *init_sub_crq_queue(struct ibmvnic_adapter · 381 unmodified lines
1213	1223	scrq->adapter = adapter;
1214	1224	scrq->size = 4 * PAGE_SIZE / sizeof(*scrq->msgs);
1215	1225	scrq->cur = 0;
	1226	+ atomic_set(&scrq->used, 0);
1216	1227	scrq->rx_skb_top = NULL;
1217	1228	spin_lock_init(&scrq->lock);
1218	1229
@@ -1355,14 +1366,28 @@ restart_loop: · 136 unmodified lines
1355	1366	DMA_TO_DEVICE);
1356	1367	}
1357	1368
1358		− if (txbuff->last_frag)
	1369	+ if (txbuff->last_frag) {
	1370	+ atomic_dec(&scrq->used);
	1371	+
	1372	+ if (atomic_read(&scrq->used) <=
	1373	+ (adapter->req_tx_entries_per_subcrq / 2) &&
	1374	+ netif_subqueue_stopped(adapter->netdev,
	1375	+ txbuff->skb)) {
	1376	+ netif_wake_subqueue(adapter->netdev,
	1377	+ scrq->pool_index);
	1378	+ netdev_dbg(adapter->netdev,
	1379	+ "Started queue %d\n",
	1380	+ scrq->pool_index);
	1381	+ }
	1382	+
1359	1383	dev_kfree_skb_any(txbuff->skb);
	1384	+ }
1360	1385
1361	1386	adapter->tx_pool[pool].free_map[adapter->tx_pool[pool].
1362	1387	producer_index] = index;
1363	1388	adapter->tx_pool[pool].producer_index =
1364	1389	(adapter->tx_pool[pool].producer_index + 1) %
1365		− adapter->max_tx_entries_per_subcrq;
	1390	+ adapter->req_tx_entries_per_subcrq;
1366	1391	}
1367	1392	/* remove tx_comp scrq*/
1368	1393	next->tx_comp.first = 0;

drivers/net/ethernet/ibm/ibmvnic.h+1−0

@@ -863,6 +863,7 @@ struct ibmvnic_sub_crq_queue { · 862 unmodified lines
863	863	spinlock_t lock;
864	864	struct sk_buff *rx_skb_top;
865	865	struct ibmvnic_adapter *adapter;
	866	+ atomic_t used;
866	867	};
867	868
868	869	struct ibmvnic_long_term_buff {

drivers/net/ethernet/mellanox/mlx5/core/Kconfig+1−0

@@ -14,6 +14,7 @@ config MLX5_CORE · 13 unmodified lines
14	14	config MLX5_CORE_EN
15	15	bool "Mellanox Technologies ConnectX-4 Ethernet support"
16	16	depends on NETDEVICES && ETHERNET && PCI && MLX5_CORE
	17	+ depends on IPV6=y \|\| IPV6=n \|\| MLX5_CORE=m
17	18	imply PTP_1588_CLOCK
18	19	default n
19	20	---help---

drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c+5−5

@@ -302,6 +302,9 @@ static u8 mlx5e_dcbnl_setdcbx(struct net_device *dev, u8 mode) · 301 unmodified lines
302	302	struct mlx5e_priv *priv = netdev_priv(dev);
303	303	struct mlx5e_dcbx *dcbx = &priv->dcbx;
304	304
	305	+ if (mode & DCB_CAP_DCBX_LLD_MANAGED)
	306	+ return 1;
	307	+
305	308	if ((!mode) && MLX5_CAP_GEN(priv->mdev, dcbx)) {
306	309	if (dcbx->mode == MLX5E_DCBX_PARAM_VER_OPER_AUTO)
307	310	return 0;
@@ -315,13 +318,10 @@ static u8 mlx5e_dcbnl_setdcbx(struct net_device *dev, u8 mode) · 7 unmodified lines
315	318	return 1;
316	319	}
317	320
318		− if (mlx5e_dcbnl_switch_to_host_mode(netdev_priv(dev)))
	321	+ if (!(mode & DCB_CAP_DCBX_HOST))
319	322	return 1;
320	323
321		− if ((mode & DCB_CAP_DCBX_LLD_MANAGED) \|\|
322		− !(mode & DCB_CAP_DCBX_VER_CEE) \|\|
323		− !(mode & DCB_CAP_DCBX_VER_IEEE) \|\|
324		− !(mode & DCB_CAP_DCBX_HOST))
	324	+ if (mlx5e_dcbnl_switch_to_host_mode(netdev_priv(dev)))
325	325	return 1;
326	326
327	327	return 0;

drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c+1−4

@@ -204,9 +204,6 @@ mlx5e_test_loopback_validate(struct sk_buff *skb, · 203 unmodified lines
204	204	struct iphdr *iph;
205	205
206	206	/* We are only going to peek, no need to clone the SKB */
207		− if (skb->protocol != htons(ETH_P_IP))
208		− goto out;
209		−
210	207	if (MLX5E_TEST_PKT_SIZE - ETH_HLEN > skb_headlen(skb))
211	208	goto out;
212	209
@@ -249,7 +246,7 @@ static int mlx5e_test_loopback_setup(struct mlx5e_priv *priv, · 36 unmodified lines
249	246	lbtp->loopback_ok = false;
250	247	init_completion(&lbtp->comp);
251	248
252		− lbtp->pt.type = htons(ETH_P_ALL);
	249	+ lbtp->pt.type = htons(ETH_P_IP);
253	250	lbtp->pt.func = mlx5e_test_loopback_validate;
254	251	lbtp->pt.dev = priv->netdev;
255	252	lbtp->pt.af_packet_priv = lbtp;

drivers/net/ethernet/mellanox/mlx5/core/en_tc.c+18−15

@@ -48,9 +48,14 @@ · 47 unmodified lines
48	48	#include "eswitch.h"
49	49	#include "vxlan.h"
50	50
	51	+enum {
	52	+ MLX5E_TC_FLOW_ESWITCH = BIT(0),
	53	+};
	54	+
51	55	struct mlx5e_tc_flow {
52	56	struct rhash_head node;
53	57	u64 cookie;
	58	+ u8 flags;
54	59	struct mlx5_flow_handle *rule;
55	60	struct list_head encap; /* flows sharing the same encap */
56	61	struct mlx5_esw_flow_attr *attr;
@@ -177,7 +182,7 @@ static void mlx5e_tc_del_flow(struct mlx5e_priv *priv, · 120 unmodified lines
177	182	mlx5_fc_destroy(priv->mdev, counter);
178	183	}
179	184
180		− if (esw && esw->mode == SRIOV_OFFLOADS) {
	185	+ if (flow->flags & MLX5E_TC_FLOW_ESWITCH) {
181	186	mlx5_eswitch_del_vlan_action(esw, flow->attr);
182	187	if (flow->attr->action & MLX5_FLOW_CONTEXT_ACTION_ENCAP)
183	188	mlx5e_detach_encap(priv, flow);
@@ -598,6 +603,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv, · 414 unmodified lines
598	603	}
599	604
600	605	static int parse_cls_flower(struct mlx5e_priv *priv,
	606	+ struct mlx5e_tc_flow *flow,
601	607	struct mlx5_flow_spec *spec,
602	608	struct tc_cls_flower_offload *f)
603	609	{
@@ -609,7 +615,7 @@ static int parse_cls_flower(struct mlx5e_priv *priv, · 5 unmodified lines
609	615
610	616	err = __parse_cls_flower(priv, spec, f, &min_inline);
611	617
612		− if (!err && esw->mode == SRIOV_OFFLOADS &&
	618	+ if (!err && (flow->flags & MLX5E_TC_FLOW_ESWITCH) &&
613	619	rep->vport != FDB_UPLINK_VPORT) {
614	620	if (min_inline > esw->offloads.inline_mode) {
615	621	netdev_warn(priv->netdev,
@@ -1132,23 +1138,19 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv, __be16 protocol, · 516 unmodified lines
1132	1138	struct tc_cls_flower_offload *f)
1133	1139	{
1134	1140	struct mlx5e_tc_table *tc = &priv->fs.tc;
1135		− int err = 0;
1136		− bool fdb_flow = false;
	1141	+ int err, attr_size = 0;
1137	1142	u32 flow_tag, action;
1138	1143	struct mlx5e_tc_flow *flow;
1139	1144	struct mlx5_flow_spec *spec;
1140	1145	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
	1146	+ u8 flow_flags = 0;
1141	1147
1142		− if (esw && esw->mode == SRIOV_OFFLOADS)
1143		− fdb_flow = true;
1144		−
1145		− if (fdb_flow)
1146		− flow = kzalloc(sizeof(*flow) +
1147		− sizeof(struct mlx5_esw_flow_attr),
1148		− GFP_KERNEL);
1149		− else
1150		− flow = kzalloc(sizeof(*flow), GFP_KERNEL);
	1148	+ if (esw && esw->mode == SRIOV_OFFLOADS) {
	1149	+ flow_flags = MLX5E_TC_FLOW_ESWITCH;
	1150	+ attr_size = sizeof(struct mlx5_esw_flow_attr);
	1151	+ }
1151	1152
	1153	+ flow = kzalloc(sizeof(*flow) + attr_size, GFP_KERNEL);
1152	1154	spec = mlx5_vzalloc(sizeof(*spec));
1153	1155	if (!spec \|\| !flow) {
1154	1156	err = -ENOMEM;
@@ -1156,12 +1158,13 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv, __be16 protocol, · 1 unmodified line
1156	1158	}
1157	1159
1158	1160	flow->cookie = f->cookie;
	1161	+ flow->flags = flow_flags;
1159	1162
1160		− err = parse_cls_flower(priv, spec, f);
	1163	+ err = parse_cls_flower(priv, flow, spec, f);
1161	1164	if (err < 0)
1162	1165	goto err_free;
1163	1166
1164		− if (fdb_flow) {
	1167	+ if (flow->flags & MLX5E_TC_FLOW_ESWITCH) {
1165	1168	flow->attr = (struct mlx5_esw_flow_attr *)(flow + 1);
1166	1169	err = parse_tc_fdb_actions(priv, f->exts, flow);
1167	1170	if (err < 0)

drivers/net/ethernet/mellanox/mlx5/core/fs_core.c+1−1

@@ -1136,7 +1136,7 @@ static struct mlx5_flow_group create_autogroup(struct mlx5_flow_table ft, · 1135 unmodified lines
1136	1136	u32 *match_criteria)
1137	1137	{
1138	1138	int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
1139		− struct list_head *prev = ft->node.children.prev;
	1139	+ struct list_head *prev = &ft->node.children;
1140	1140	unsigned int candidate_index = 0;
1141	1141	struct mlx5_flow_group *fg;
1142	1142	void *match_criteria_addr;

drivers/net/ethernet/mellanox/mlx5/core/main.c+3−2

@@ -1352,6 +1352,7 @@ static int init_one(struct pci_dev *pdev, · 1351 unmodified lines
1352	1352	if (err)
1353	1353	goto clean_load;
1354	1354
	1355	+ pci_save_state(pdev);
1355	1356	return 0;
1356	1357
1357	1358	clean_load:
@@ -1407,9 +1408,8 @@ static pci_ers_result_t mlx5_pci_err_detected(struct pci_dev *pdev, · 49 unmodified lines
1407	1408
1408	1409	mlx5_enter_error_state(dev);
1409	1410	mlx5_unload_one(dev, priv, false);
1410		− /* In case of kernel call save the pci state and drain the health wq */
	1411	+ /* In case of kernel call drain the health wq */
1411	1412	if (state) {
1412		− pci_save_state(pdev);
1413	1413	mlx5_drain_health_wq(dev);
1414	1414	mlx5_pci_disable_device(dev);
1415	1415	}
@@ -1461,6 +1461,7 @@ static pci_ers_result_t mlx5_pci_slot_reset(struct pci_dev *pdev) · 45 unmodified lines
1461	1461
1462	1462	pci_set_master(pdev);
1463	1463	pci_restore_state(pdev);
	1464	+ pci_save_state(pdev);
1464	1465
1465	1466	if (wait_vital(pdev)) {
1466	1467	dev_err(&pdev->dev, "%s: wait_vital timed out\n", __func__);

drivers/net/ethernet/mellanox/mlxsw/reg.h+2−2

@@ -769,7 +769,7 @@ static inline void mlxsw_reg_spvid_pack(char *payload, u8 local_port, u16 pvid) · 768 unmodified lines
769	769	#define MLXSW_REG_SPVM_ID 0x200F
770	770	#define MLXSW_REG_SPVM_BASE_LEN 0x04 /* base length, without records */
771	771	#define MLXSW_REG_SPVM_REC_LEN 0x04 /* record length */
772		−#define MLXSW_REG_SPVM_REC_MAX_COUNT 256
	772	+#define MLXSW_REG_SPVM_REC_MAX_COUNT 255
773	773	#define MLXSW_REG_SPVM_LEN (MLXSW_REG_SPVM_BASE_LEN + \
774	774	MLXSW_REG_SPVM_REC_LEN * MLXSW_REG_SPVM_REC_MAX_COUNT)
775	775
@@ -1702,7 +1702,7 @@ static inline void mlxsw_reg_sfmr_pack(char *payload, · 926 unmodified lines
1702	1702	#define MLXSW_REG_SPVMLR_ID 0x2020
1703	1703	#define MLXSW_REG_SPVMLR_BASE_LEN 0x04 /* base length, without records */
1704	1704	#define MLXSW_REG_SPVMLR_REC_LEN 0x04 /* record length */
1705		−#define MLXSW_REG_SPVMLR_REC_MAX_COUNT 256
	1705	+#define MLXSW_REG_SPVMLR_REC_MAX_COUNT 255
1706	1706	#define MLXSW_REG_SPVMLR_LEN (MLXSW_REG_SPVMLR_BASE_LEN + \
1707	1707	MLXSW_REG_SPVMLR_REC_LEN * \
1708	1708	MLXSW_REG_SPVMLR_REC_MAX_COUNT)

drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c+2−2

@@ -303,11 +303,11 @@ void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, · 302 unmodified lines
303	303	ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, mlxsw_sp_port->dev,
304	304	ingress,
305	305	MLXSW_SP_ACL_PROFILE_FLOWER);
306		− if (WARN_ON(IS_ERR(ruleset)))
	306	+ if (IS_ERR(ruleset))
307	307	return;
308	308
309	309	rule = mlxsw_sp_acl_rule_lookup(mlxsw_sp, ruleset, f->cookie);
310		− if (!WARN_ON(!rule)) {
	310	+ if (rule) {
311	311	mlxsw_sp_acl_rule_del(mlxsw_sp, rule);
312	312	mlxsw_sp_acl_rule_destroy(mlxsw_sp, rule);
313	313	}

drivers/net/ethernet/qlogic/qed/qed_cxt.c+2−1

@@ -422,8 +422,9 @@ static void qed_cxt_set_proto_cid_count(struct qed_hwfn *p_hwfn, · 421 unmodified lines
422	422	u32 page_sz = p_mgr->clients[ILT_CLI_CDUC].p_size.val;
423	423	u32 cxt_size = CONN_CXT_SIZE(p_hwfn);
424	424	u32 elems_per_page = ILT_PAGE_IN_BYTES(page_sz) / cxt_size;
	425	+ u32 align = elems_per_page * DQ_RANGE_ALIGN;
425	426
426		− p_conn->cid_count = roundup(p_conn->cid_count, elems_per_page);
	427	+ p_conn->cid_count = roundup(p_conn->cid_count, align);
427	428	}
428	429	}
429	430

drivers/net/ethernet/qlogic/qed/qed_dev.c+2−3

@@ -2389,9 +2389,8 @@ qed_chain_alloc_sanity_check(struct qed_dev *cdev, · 2388 unmodified lines
2389	2389	* size/capacity fields are of a u32 type.
2390	2390	*/
2391	2391	if ((cnt_type == QED_CHAIN_CNT_TYPE_U16 &&
2392		− chain_size > 0x10000) \|\|
2393		− (cnt_type == QED_CHAIN_CNT_TYPE_U32 &&
2394		− chain_size > 0x100000000ULL)) {
	2392	+ chain_size > ((u32)U16_MAX + 1)) \|\|
	2393	+ (cnt_type == QED_CHAIN_CNT_TYPE_U32 && chain_size > U32_MAX)) {
2395	2394	DP_NOTICE(cdev,
2396	2395	"The actual chain size (0x%llx) is larger than the maximal possible value\n",
2397	2396	chain_size);

drivers/net/ethernet/qlogic/qed/qed_iscsi.c+31−0

@@ -190,6 +190,9 @@ qed_sp_iscsi_func_start(struct qed_hwfn *p_hwfn, · 189 unmodified lines
190	190	p_init->num_sq_pages_in_ring = p_params->num_sq_pages_in_ring;
191	191	p_init->num_r2tq_pages_in_ring = p_params->num_r2tq_pages_in_ring;
192	192	p_init->num_uhq_pages_in_ring = p_params->num_uhq_pages_in_ring;
	193	+ p_init->ooo_enable = p_params->ooo_enable;
	194	+ p_init->ll2_rx_queue_id = p_hwfn->hw_info.resc_start[QED_LL2_QUEUE] +
	195	+ p_params->ll2_ooo_queue_id;
193	196	p_init->func_params.log_page_size = p_params->log_page_size;
194	197	val = p_params->num_tasks;
195	198	p_init->func_params.num_tasks = cpu_to_le16(val);
@@ -786,6 +789,23 @@ static void qed_iscsi_release_connection(struct qed_hwfn *p_hwfn, · 590 unmodified lines
786	789	spin_unlock_bh(&p_hwfn->p_iscsi_info->lock);
787	790	}
788	791
	792	+void qed_iscsi_free_connection(struct qed_hwfn *p_hwfn,
	793	+ struct qed_iscsi_conn *p_conn)
	794	+{
	795	+ qed_chain_free(p_hwfn->cdev, &p_conn->xhq);
	796	+ qed_chain_free(p_hwfn->cdev, &p_conn->uhq);
	797	+ qed_chain_free(p_hwfn->cdev, &p_conn->r2tq);
	798	+ dma_free_coherent(&p_hwfn->cdev->pdev->dev,
	799	+ sizeof(struct tcp_upload_params),
	800	+ p_conn->tcp_upload_params_virt_addr,
	801	+ p_conn->tcp_upload_params_phys_addr);
	802	+ dma_free_coherent(&p_hwfn->cdev->pdev->dev,
	803	+ sizeof(struct scsi_terminate_extra_params),
	804	+ p_conn->queue_cnts_virt_addr,
	805	+ p_conn->queue_cnts_phys_addr);
	806	+ kfree(p_conn);
	807	+}
	808	+
789	809	struct qed_iscsi_info qed_iscsi_alloc(struct qed_hwfn p_hwfn)
790	810	{
791	811	struct qed_iscsi_info *p_iscsi_info;
@@ -807,6 +827,17 @@ void qed_iscsi_setup(struct qed_hwfn *p_hwfn, · 15 unmodified lines
807	827	void qed_iscsi_free(struct qed_hwfn *p_hwfn,
808	828	struct qed_iscsi_info *p_iscsi_info)
809	829	{
	830	+ struct qed_iscsi_conn *p_conn = NULL;
	831	+
	832	+ while (!list_empty(&p_hwfn->p_iscsi_info->free_list)) {
	833	+ p_conn = list_first_entry(&p_hwfn->p_iscsi_info->free_list,
	834	+ struct qed_iscsi_conn, list_entry);
	835	+ if (p_conn) {
	836	+ list_del(&p_conn->list_entry);
	837	+ qed_iscsi_free_connection(p_hwfn, p_conn);
	838	+ }
	839	+ }
	840	+
810	841	kfree(p_iscsi_info);
811	842	}
812	843

drivers/net/ethernet/qlogic/qed/qed_ll2.c+8−5

@@ -211,6 +211,8 @@ static void qed_ll2b_complete_rx_packet(struct qed_hwfn *p_hwfn, · 210 unmodified lines
211	211	/* If need to reuse or there's no replacement buffer, repost this */
212	212	if (rc)
213	213	goto out_post;
	214	+ dma_unmap_single(&cdev->pdev->dev, buffer->phys_addr,
	215	+ cdev->ll2->rx_size, DMA_FROM_DEVICE);
214	216
215	217	skb = build_skb(buffer->data, 0);
216	218	if (!skb) {
@@ -474,7 +476,7 @@ qed_ll2_rxq_completion_gsi(struct qed_hwfn *p_hwfn, · 257 unmodified lines
474	476	static int qed_ll2_rxq_completion_reg(struct qed_hwfn *p_hwfn,
475	477	struct qed_ll2_info *p_ll2_conn,
476	478	union core_rx_cqe_union *p_cqe,
477		− unsigned long lock_flags,
	479	+ unsigned long *p_lock_flags,
478	480	bool b_last_cqe)
479	481	{
480	482	struct qed_ll2_rx_queue *p_rx = &p_ll2_conn->rx_queue;
@@ -495,10 +497,10 @@ static int qed_ll2_rxq_completion_reg(struct qed_hwfn *p_hwfn, · 14 unmodified lines
495	497	"Mismatch between active_descq and the LL2 Rx chain\n");
496	498	list_add_tail(&p_pkt->list_entry, &p_rx->free_descq);
497	499
498		− spin_unlock_irqrestore(&p_rx->lock, lock_flags);
	500	+ spin_unlock_irqrestore(&p_rx->lock, *p_lock_flags);
499	501	qed_ll2b_complete_rx_packet(p_hwfn, p_ll2_conn->my_id,
500	502	p_pkt, &p_cqe->rx_cqe_fp, b_last_cqe);
501		− spin_lock_irqsave(&p_rx->lock, lock_flags);
	503	+ spin_lock_irqsave(&p_rx->lock, *p_lock_flags);
502	504
503	505	return 0;
504	506	}
@@ -538,7 +540,8 @@ static int qed_ll2_rxq_completion(struct qed_hwfn p_hwfn, void cookie) · 33 unmodified lines
538	540	break;
539	541	case CORE_RX_CQE_TYPE_REGULAR:
540	542	rc = qed_ll2_rxq_completion_reg(p_hwfn, p_ll2_conn,
541		− cqe, flags, b_last_cqe);
	543	+ cqe, &flags,
	544	+ b_last_cqe);
542	545	break;
543	546	default:
544	547	rc = -EIO;
@@ -968,7 +971,7 @@ static int qed_ll2_start_ooo(struct qed_dev *cdev, · 423 unmodified lines
968	971	{
969	972	struct qed_hwfn *hwfn = QED_LEADING_HWFN(cdev);
970	973	u8 *handle = &hwfn->pf_params.iscsi_pf_params.ll2_ooo_queue_id;
971		− struct qed_ll2_conn ll2_info;
	974	+ struct qed_ll2_conn ll2_info = { 0 };
972	975	int rc;
973	976
974	977	ll2_info.conn_type = QED_LL2_TYPE_ISCSI_OOO;

drivers/net/ethernet/qlogic/qed/qed_ooo.c+2−0

@@ -159,6 +159,8 @@ struct qed_ooo_info qed_ooo_alloc(struct qed_hwfn p_hwfn) · 158 unmodified lines
159	159	if (!p_ooo_info->ooo_history.p_cqes)
160	160	goto no_history_mem;
161	161
	162	+ p_ooo_info->ooo_history.num_of_cqes = QED_MAX_NUM_OOO_HISTORY_ENTRIES;
	163	+
162	164	return p_ooo_info;
163	165
164	166	no_history_mem:

drivers/net/ethernet/qlogic/qlge/qlge.h+2−2

@@ -1162,8 +1162,8 @@ struct ob_mac_tso_iocb_rsp { · 1161 unmodified lines
1162	1162	struct ib_mac_iocb_rsp {
1163	1163	u8 opcode; /* 0x20 */
1164	1164	u8 flags1;
1165		−#define IB_MAC_IOCB_RSP_OI 0x01 /* Overide intr delay */
1166		−#define IB_MAC_IOCB_RSP_I 0x02 /* Disble Intr Generation */
	1165	+#define IB_MAC_IOCB_RSP_OI 0x01 /* Override intr delay */
	1166	+#define IB_MAC_IOCB_RSP_I 0x02 /* Disable Intr Generation */
1167	1167	#define IB_MAC_CSUM_ERR_MASK 0x1c /* A mask to use for csum errs */
1168	1168	#define IB_MAC_IOCB_RSP_TE 0x04 /* Checksum error */
1169	1169	#define IB_MAC_IOCB_RSP_NU 0x08 /* No checksum rcvd */

drivers/net/ethernet/smsc/smc91x.c+25−22

@@ -1535,32 +1535,33 @@ static int smc_close(struct net_device *dev) · 1534 unmodified lines
1535	1535	* Ethtool support
1536	1536	*/
1537	1537	static int
1538		−smc_ethtool_getsettings(struct net_device dev, struct ethtool_cmd cmd)
	1538	+smc_ethtool_get_link_ksettings(struct net_device *dev,
	1539	+ struct ethtool_link_ksettings *cmd)
1539	1540	{
1540	1541	struct smc_local *lp = netdev_priv(dev);
1541	1542	int ret;
1542	1543
1543		− cmd->maxtxpkt = 1;
1544		− cmd->maxrxpkt = 1;
1545		−
1546	1544	if (lp->phy_type != 0) {
1547	1545	spin_lock_irq(&lp->lock);
1548		− ret = mii_ethtool_gset(&lp->mii, cmd);
	1546	+ ret = mii_ethtool_get_link_ksettings(&lp->mii, cmd);
1549	1547	spin_unlock_irq(&lp->lock);
1550	1548	} else {
1551		− cmd->supported = SUPPORTED_10baseT_Half \|
	1549	+ u32 supported = SUPPORTED_10baseT_Half \|
1552	1550	SUPPORTED_10baseT_Full \|
1553	1551	SUPPORTED_TP \| SUPPORTED_AUI;
1554	1552
1555	1553	if (lp->ctl_rspeed == 10)
1556		− ethtool_cmd_speed_set(cmd, SPEED_10);
	1554	+ cmd->base.speed = SPEED_10;
1557	1555	else if (lp->ctl_rspeed == 100)
1558		− ethtool_cmd_speed_set(cmd, SPEED_100);
	1556	+ cmd->base.speed = SPEED_100;
	1557	+
	1558	+ cmd->base.autoneg = AUTONEG_DISABLE;
	1559	+ cmd->base.port = 0;
	1560	+ cmd->base.duplex = lp->tcr_cur_mode & TCR_SWFDUP ?
	1561	+ DUPLEX_FULL : DUPLEX_HALF;
1559	1562
1560		− cmd->autoneg = AUTONEG_DISABLE;
1561		− cmd->transceiver = XCVR_INTERNAL;
1562		− cmd->port = 0;
1563		− cmd->duplex = lp->tcr_cur_mode & TCR_SWFDUP ? DUPLEX_FULL : DUPLEX_HALF;
	1563	+ ethtool_convert_legacy_u32_to_link_mode(
	1564	+ cmd->link_modes.supported, supported);
1564	1565
1565	1566	ret = 0;
1566	1567	}
@@ -1569,24 +1570,26 @@ smc_ethtool_getsettings(struct net_device dev, struct ethtool_cmd cmd) · 2 unmodified lines
1569	1570	}
1570	1571
1571	1572	static int
1572		−smc_ethtool_setsettings(struct net_device dev, struct ethtool_cmd cmd)
	1573	+smc_ethtool_set_link_ksettings(struct net_device *dev,
	1574	+ const struct ethtool_link_ksettings *cmd)
1573	1575	{
1574	1576	struct smc_local *lp = netdev_priv(dev);
1575	1577	int ret;
1576	1578
1577	1579	if (lp->phy_type != 0) {
1578	1580	spin_lock_irq(&lp->lock);
1579		− ret = mii_ethtool_sset(&lp->mii, cmd);
	1581	+ ret = mii_ethtool_set_link_ksettings(&lp->mii, cmd);
1580	1582	spin_unlock_irq(&lp->lock);
1581	1583	} else {
1582		− if (cmd->autoneg != AUTONEG_DISABLE \|\|
1583		− cmd->speed != SPEED_10 \|\|
1584		− (cmd->duplex != DUPLEX_HALF && cmd->duplex != DUPLEX_FULL) \|\|
1585		− (cmd->port != PORT_TP && cmd->port != PORT_AUI))
	1584	+ if (cmd->base.autoneg != AUTONEG_DISABLE \|\|
	1585	+ cmd->base.speed != SPEED_10 \|\|
	1586	+ (cmd->base.duplex != DUPLEX_HALF &&
	1587	+ cmd->base.duplex != DUPLEX_FULL) \|\|
	1588	+ (cmd->base.port != PORT_TP && cmd->base.port != PORT_AUI))
1586	1589	return -EINVAL;
1587	1590
1588		−// lp->port = cmd->port;
1589		− lp->ctl_rfduplx = cmd->duplex == DUPLEX_FULL;
	1591	+// lp->port = cmd->base.port;
	1592	+ lp->ctl_rfduplx = cmd->base.duplex == DUPLEX_FULL;
1590	1593
1591	1594	// if (netif_running(dev))
1592	1595	// smc_set_port(dev);
@@ -1744,8 +1747,6 @@ static int smc_ethtool_seteeprom(struct net_device *dev, · 151 unmodified lines
1744	1747
1745	1748
1746	1749	static const struct ethtool_ops smc_ethtool_ops = {
1747		− .get_settings = smc_ethtool_getsettings,
1748		− .set_settings = smc_ethtool_setsettings,
1749	1750	.get_drvinfo = smc_ethtool_getdrvinfo,
1750	1751
1751	1752	.get_msglevel = smc_ethtool_getmsglevel,
@@ -1755,6 +1756,8 @@ static const struct ethtool_ops smc_ethtool_ops = { · 3 unmodified lines
1755	1756	.get_eeprom_len = smc_ethtool_geteeprom_len,
1756	1757	.get_eeprom = smc_ethtool_geteeprom,
1757	1758	.set_eeprom = smc_ethtool_seteeprom,
	1759	+ .get_link_ksettings = smc_ethtool_get_link_ksettings,
	1760	+ .set_link_ksettings = smc_ethtool_set_link_ksettings,
1758	1761	};
1759	1762
1760	1763	static const struct net_device_ops smc_netdev_ops = {

drivers/net/hyperv/hyperv_net.h+2−1

@@ -700,6 +700,8 @@ struct net_device_context { · 699 unmodified lines
700	700
701	701	u32 tx_checksum_mask;
702	702
	703	+ u32 tx_send_table[VRSS_SEND_TAB_SIZE];
	704	+
703	705	/* Ethtool settings */
704	706	u8 duplex;
705	707	u32 speed;
@@ -757,7 +759,6 @@ struct netvsc_device { · 51 unmodified lines
757	759
758	760	struct nvsp_message revoke_packet;
759	761
760		− u32 send_table[VRSS_SEND_TAB_SIZE];
761	762	u32 max_chn;
762	763	u32 num_chn;
763	764	spinlock_t sc_lock; /* Protects num_sc_offered variable */

drivers/net/hyperv/netvsc.c+2−6

@@ -1136,15 +1136,11 @@ static void netvsc_receive(struct net_device *ndev, · 1135 unmodified lines
1136	1136	static void netvsc_send_table(struct hv_device *hdev,
1137	1137	struct nvsp_message *nvmsg)
1138	1138	{
1139		− struct netvsc_device *nvscdev;
1140	1139	struct net_device *ndev = hv_get_drvdata(hdev);
	1140	+ struct net_device_context *net_device_ctx = netdev_priv(ndev);
1141	1141	int i;
1142	1142	u32 count, *tab;
1143	1143
1144		− nvscdev = get_outbound_net_device(hdev);
1145		− if (!nvscdev)
1146		− return;
1147		−
1148	1144	count = nvmsg->msg.v5_msg.send_table.count;
1149	1145	if (count != VRSS_SEND_TAB_SIZE) {
1150	1146	netdev_err(ndev, "Received wrong send-table size:%u\n", count);
@@ -1155,7 +1151,7 @@ static void netvsc_send_table(struct hv_device *hdev, · 4 unmodified lines
1155	1151	nvmsg->msg.v5_msg.send_table.offset);
1156	1152
1157	1153	for (i = 0; i < count; i++)
1158		− nvscdev->send_table[i] = tab[i];
	1154	+ net_device_ctx->tx_send_table[i] = tab[i];
1159	1155	}
1160	1156
1161	1157	static void netvsc_send_vf(struct net_device_context *net_device_ctx,

drivers/net/hyperv/netvsc_drv.c+3−8

@@ -206,17 +206,15 @@ static u16 netvsc_select_queue(struct net_device ndev, struct sk_buff skb, · 205 unmodified lines
206	206	void *accel_priv, select_queue_fallback_t fallback)
207	207	{
208	208	struct net_device_context *net_device_ctx = netdev_priv(ndev);
209		− struct netvsc_device *nvsc_dev = net_device_ctx->nvdev;
	209	+ unsigned int num_tx_queues = ndev->real_num_tx_queues;
210	210	struct sock *sk = skb->sk;
211	211	int q_idx = sk_tx_queue_get(sk);
212	212
213		− if (q_idx < 0 \|\| skb->ooo_okay \|\|
214		− q_idx >= ndev->real_num_tx_queues) {
	213	+ if (q_idx < 0 \|\| skb->ooo_okay \|\| q_idx >= num_tx_queues) {
215	214	u16 hash = __skb_tx_hash(ndev, skb, VRSS_SEND_TAB_SIZE);
216	215	int new_idx;
217	216
218		− new_idx = nvsc_dev->send_table[hash]
219		− % nvsc_dev->num_chn;
	217	+ new_idx = net_device_ctx->tx_send_table[hash] % num_tx_queues;
220	218
221	219	if (q_idx != new_idx && sk &&
222	220	sk_fullsock(sk) && rcu_access_pointer(sk->sk_dst_cache))
@@ -225,9 +223,6 @@ static u16 netvsc_select_queue(struct net_device ndev, struct sk_buff skb, · 2 unmodified lines
225	223	q_idx = new_idx;
226	224	}
227	225
228		− if (unlikely(!nvsc_dev->chan_table[q_idx].channel))
229		− q_idx = 0;
230		−
231	226	return q_idx;
232	227	}
233	228

drivers/net/phy/marvell.c+0−15

@@ -1883,17 +1883,6 @@ static int m88e1510_probe(struct phy_device *phydev) · 1882 unmodified lines
1883	1883	return m88e1510_hwmon_probe(phydev);
1884	1884	}
1885	1885
1886		−static void marvell_remove(struct phy_device *phydev)
1887		−{
1888		−#ifdef CONFIG_HWMON
1889		−
1890		− struct marvell_priv *priv = phydev->priv;
1891		−
1892		− if (priv && priv->hwmon_dev)
1893		− hwmon_device_unregister(priv->hwmon_dev);
1894		−#endif
1895		−}
1896		−
1897	1886	static struct phy_driver marvell_drivers[] = {
1898	1887	{
1899	1888	.phy_id = MARVELL_PHY_ID_88E1101,
@@ -1974,7 +1963,6 @@ static struct phy_driver marvell_drivers[] = { · 74 unmodified lines
1974	1963	.features = PHY_GBIT_FEATURES,
1975	1964	.flags = PHY_HAS_INTERRUPT,
1976	1965	.probe = &m88e1121_probe,
1977		− .remove = &marvell_remove,
1978	1966	.config_init = &m88e1121_config_init,
1979	1967	.config_aneg = &m88e1121_config_aneg,
1980	1968	.read_status = &marvell_read_status,
@@ -2087,7 +2075,6 @@ static struct phy_driver marvell_drivers[] = { · 106 unmodified lines
2087	2075	.features = PHY_GBIT_FEATURES \| SUPPORTED_FIBRE,
2088	2076	.flags = PHY_HAS_INTERRUPT,
2089	2077	.probe = &m88e1510_probe,
2090		− .remove = &marvell_remove,
2091	2078	.config_init = &m88e1510_config_init,
2092	2079	.config_aneg = &m88e1510_config_aneg,
2093	2080	.read_status = &marvell_read_status,
@@ -2109,7 +2096,6 @@ static struct phy_driver marvell_drivers[] = { · 15 unmodified lines
2109	2096	.features = PHY_GBIT_FEATURES,
2110	2097	.flags = PHY_HAS_INTERRUPT,
2111	2098	.probe = m88e1510_probe,
2112		− .remove = &marvell_remove,
2113	2099	.config_init = &marvell_config_init,
2114	2100	.config_aneg = &m88e1510_config_aneg,
2115	2101	.read_status = &marvell_read_status,
@@ -2127,7 +2113,6 @@ static struct phy_driver marvell_drivers[] = { · 11 unmodified lines
2127	2113	.phy_id_mask = MARVELL_PHY_ID_MASK,
2128	2114	.name = "Marvell 88E1545",
2129	2115	.probe = m88e1510_probe,
2130		− .remove = &marvell_remove,
2131	2116	.features = PHY_GBIT_FEATURES,
2132	2117	.flags = PHY_HAS_INTERRUPT,
2133	2118	.config_init = &marvell_config_init,

drivers/net/phy/phy_device.c+1−1

@@ -1864,7 +1864,7 @@ static struct phy_driver genphy_driver[] = { · 1863 unmodified lines
1864	1864	.phy_id = 0xffffffff,
1865	1865	.phy_id_mask = 0xffffffff,
1866	1866	.name = "Generic PHY",
1867		− .soft_reset = genphy_soft_reset,
	1867	+ .soft_reset = genphy_no_soft_reset,
1868	1868	.config_init = genphy_config_init,
1869	1869	.features = PHY_GBIT_FEATURES \| SUPPORTED_MII \|
1870	1870	SUPPORTED_AUI \| SUPPORTED_FIBRE \|

drivers/net/phy/spi_ks8995.c+2−1

@@ -491,13 +491,14 @@ static int ks8995_probe(struct spi_device *spi) · 490 unmodified lines
491	491	if (err)
492	492	return err;
493	493
494		− ks->regs_attr.size = ks->chip->regs_size;
495	494	memcpy(&ks->regs_attr, &ks8995_registers_attr, sizeof(ks->regs_attr));
	495	+ ks->regs_attr.size = ks->chip->regs_size;
496	496
497	497	err = ks8995_reset(ks);
498	498	if (err)
499	499	return err;
500	500
	501	+ sysfs_attr_init(&ks->regs_attr.attr);
501	502	err = sysfs_create_bin_file(&spi->dev.kobj, &ks->regs_attr);
502	503	if (err) {
503	504	dev_err(&spi->dev, "unable to create sysfs file, err=%d\n",

drivers/net/team/team.c+1−0

@@ -2072,6 +2072,7 @@ static int team_dev_type_check_change(struct net_device *dev, · 2071 unmodified lines
2072	2072	static void team_setup(struct net_device *dev)
2073	2073	{
2074	2074	ether_setup(dev);
	2075	+ dev->max_mtu = ETH_MAX_MTU;
2075	2076
2076	2077	dev->netdev_ops = &team_netdev_ops;
2077	2078	dev->ethtool_ops = &team_ethtool_ops;

drivers/net/tun.c+15−4

@@ -822,7 +822,18 @@ static void tun_net_uninit(struct net_device *dev) · 821 unmodified lines
822	822	/* Net device open. */
823	823	static int tun_net_open(struct net_device *dev)
824	824	{
	825	+ struct tun_struct *tun = netdev_priv(dev);
	826	+ int i;
	827	+
825	828	netif_tx_start_all_queues(dev);
	829	+
	830	+ for (i = 0; i < tun->numqueues; i++) {
	831	+ struct tun_file *tfile;
	832	+
	833	+ tfile = rtnl_dereference(tun->tfiles[i]);
	834	+ tfile->socket.sk->sk_write_space(tfile->socket.sk);
	835	+ }
	836	+
826	837	return 0;
827	838	}
828	839
@@ -1103,9 +1114,10 @@ static unsigned int tun_chr_poll(struct file file, poll_table wait) · 274 unmodified lines
1103	1114	if (!skb_array_empty(&tfile->tx_array))
1104	1115	mask \|= POLLIN \| POLLRDNORM;
1105	1116
1106		− if (sock_writeable(sk) \|\|
1107		− (!test_and_set_bit(SOCKWQ_ASYNC_NOSPACE, &sk->sk_socket->flags) &&
1108		− sock_writeable(sk)))
	1117	+ if (tun->dev->flags & IFF_UP &&
	1118	+ (sock_writeable(sk) \|\|
	1119	+ (!test_and_set_bit(SOCKWQ_ASYNC_NOSPACE, &sk->sk_socket->flags) &&
	1120	+ sock_writeable(sk))))
1109	1121	mask \|= POLLOUT \| POLLWRNORM;
1110	1122
1111	1123	if (tun->dev->reg_state != NETREG_REGISTERED)
@@ -2570,7 +2582,6 @@ static int __init tun_init(void) · 1458 unmodified lines
2570	2582	int ret = 0;
2571	2583
2572	2584	pr_info("%s, %s\n", DRV_DESCRIPTION, DRV_VERSION);
2573		− pr_info("%s\n", DRV_COPYRIGHT);
2574	2585
2575	2586	ret = rtnl_link_register(&tun_link_ops);
2576	2587	if (ret) {

drivers/net/vrf.c+2−1

@@ -340,6 +340,7 @@ static netdev_tx_t is_ip_tx_frame(struct sk_buff skb, struct net_device dev) · 339 unmodified lines
340	340
341	341	static netdev_tx_t vrf_xmit(struct sk_buff skb, struct net_device dev)
342	342	{
	343	+ int len = skb->len;
343	344	netdev_tx_t ret = is_ip_tx_frame(skb, dev);
344	345
345	346	if (likely(ret == NET_XMIT_SUCCESS \|\| ret == NET_XMIT_CN)) {
@@ -347,7 +348,7 @@ static netdev_tx_t vrf_xmit(struct sk_buff skb, struct net_device dev) · 1 unmodified line
347	348
348	349	u64_stats_update_begin(&dstats->syncp);
349	350	dstats->tx_pkts++;
350		− dstats->tx_bytes += skb->len;
	351	+ dstats->tx_bytes += len;
351	352	u64_stats_update_end(&dstats->syncp);
352	353	} else {
353	354	this_cpu_inc(dev->dstats->tx_drps);

drivers/net/vxlan.c+40−33

@@ -2976,6 +2976,44 @@ static int vxlan_dev_configure(struct net src_net, struct net_device dev, · 2975 unmodified lines
2976	2976	return 0;
2977	2977	}
2978	2978
	2979	+static int __vxlan_dev_create(struct net net, struct net_device dev,
	2980	+ struct vxlan_config *conf)
	2981	+{
	2982	+ struct vxlan_net *vn = net_generic(net, vxlan_net_id);
	2983	+ struct vxlan_dev *vxlan = netdev_priv(dev);
	2984	+ int err;
	2985	+
	2986	+ err = vxlan_dev_configure(net, dev, conf, false);
	2987	+ if (err)
	2988	+ return err;
	2989	+
	2990	+ dev->ethtool_ops = &vxlan_ethtool_ops;
	2991	+
	2992	+ /* create an fdb entry for a valid default destination */
	2993	+ if (!vxlan_addr_any(&vxlan->default_dst.remote_ip)) {
	2994	+ err = vxlan_fdb_create(vxlan, all_zeros_mac,
	2995	+ &vxlan->default_dst.remote_ip,
	2996	+ NUD_REACHABLE \| NUD_PERMANENT,
	2997	+ NLM_F_EXCL \| NLM_F_CREATE,
	2998	+ vxlan->cfg.dst_port,
	2999	+ vxlan->default_dst.remote_vni,
	3000	+ vxlan->default_dst.remote_vni,
	3001	+ vxlan->default_dst.remote_ifindex,
	3002	+ NTF_SELF);
	3003	+ if (err)
	3004	+ return err;
	3005	+ }
	3006	+
	3007	+ err = register_netdevice(dev);
	3008	+ if (err) {
	3009	+ vxlan_fdb_delete_default(vxlan, vxlan->default_dst.remote_vni);
	3010	+ return err;
	3011	+ }
	3012	+
	3013	+ list_add(&vxlan->next, &vn->vxlan_list);
	3014	+ return 0;
	3015	+}
	3016	+
2979	3017	static int vxlan_nl2conf(struct nlattr tb[], struct nlattr data[],
2980	3018	struct net_device dev, struct vxlan_config conf,
2981	3019	bool changelink)
@@ -3172,8 +3210,6 @@ static int vxlan_nl2conf(struct nlattr tb[], struct nlattr data[], · 190 unmodified lines
3172	3210	static int vxlan_newlink(struct net src_net, struct net_device dev,
3173	3211	struct nlattr tb[], struct nlattr data[])
3174	3212	{
3175		− struct vxlan_net *vn = net_generic(src_net, vxlan_net_id);
3176		− struct vxlan_dev *vxlan = netdev_priv(dev);
3177	3213	struct vxlan_config conf;
3178	3214	int err;
3179	3215
@@ -3181,36 +3217,7 @@ static int vxlan_newlink(struct net src_net, struct net_device dev, · 1 unmodified line
3181	3217	if (err)
3182	3218	return err;
3183	3219
3184		− err = vxlan_dev_configure(src_net, dev, &conf, false);
3185		− if (err)
3186		− return err;
3187		−
3188		− dev->ethtool_ops = &vxlan_ethtool_ops;
3189		−
3190		− /* create an fdb entry for a valid default destination */
3191		− if (!vxlan_addr_any(&vxlan->default_dst.remote_ip)) {
3192		− err = vxlan_fdb_create(vxlan, all_zeros_mac,
3193		− &vxlan->default_dst.remote_ip,
3194		− NUD_REACHABLE \| NUD_PERMANENT,
3195		− NLM_F_EXCL \| NLM_F_CREATE,
3196		− vxlan->cfg.dst_port,
3197		− vxlan->default_dst.remote_vni,
3198		− vxlan->default_dst.remote_vni,
3199		− vxlan->default_dst.remote_ifindex,
3200		− NTF_SELF);
3201		− if (err)
3202		− return err;
3203		− }
3204		−
3205		− err = register_netdevice(dev);
3206		− if (err) {
3207		− vxlan_fdb_delete_default(vxlan, vxlan->default_dst.remote_vni);
3208		− return err;
3209		− }
3210		−
3211		− list_add(&vxlan->next, &vn->vxlan_list);
3212		−
3213		− return 0;
	3220	+ return __vxlan_dev_create(src_net, dev, &conf);
3214	3221	}
3215	3222
3216	3223	static int vxlan_changelink(struct net_device dev, struct nlattr tb[],
@@ -3440,7 +3447,7 @@ struct net_device vxlan_dev_create(struct net net, const char *name, · 223 unmodified lines
3440	3447	if (IS_ERR(dev))
3441	3448	return dev;
3442	3449
3443		− err = vxlan_dev_configure(net, dev, conf, false);
	3450	+ err = __vxlan_dev_create(net, dev, conf);
3444	3451	if (err < 0) {
3445	3452	free_netdev(dev);
3446	3453	return ERR_PTR(err);

drivers/net/wan/fsl_ucc_hdlc.c+2−2

@@ -381,8 +381,8 @@ static netdev_tx_t ucc_hdlc_tx(struct sk_buff skb, struct net_device dev) · 380 unmodified lines
381	381	/* set bd status and length */
382	382	bd_status = (bd_status & T_W_S) \| T_R_S \| T_I_S \| T_L_S \| T_TC_S;
383	383
384		− iowrite16be(bd_status, &bd->status);
385	384	iowrite16be(skb->len, &bd->length);
	385	+ iowrite16be(bd_status, &bd->status);
386	386
387	387	/* Move to next BD in the ring */
388	388	if (!(bd_status & T_W_S))
@@ -457,7 +457,7 @@ static int hdlc_rx_done(struct ucc_hdlc_private *priv, int rx_work_limit) · 68 unmodified lines
457	457	struct sk_buff *skb;
458	458	hdlc_device *hdlc = dev_to_hdlc(dev);
459	459	struct qe_bd *bd;
460		− u32 bd_status;
	460	+ u16 bd_status;
461	461	u16 length, howmany = 0;
462	462	u8 *bdbuffer;
463	463	int i;

drivers/net/wimax/i2400m/usb.c+3−0

@@ -467,6 +467,9 @@ int i2400mu_probe(struct usb_interface *iface, · 466 unmodified lines
467	467	struct i2400mu *i2400mu;
468	468	struct usb_device *usb_dev = interface_to_usbdev(iface);
469	469
	470	+ if (iface->cur_altsetting->desc.bNumEndpoints < 4)
	471	+ return -ENODEV;
	472	+
470	473	if (usb_dev->speed != USB_SPEED_HIGH)
471	474	dev_err(dev, "device not connected as high speed\n");
472	475

drivers/net/xen-netback/interface.c+17−9

@@ -165,13 +165,17 @@ static int xenvif_start_xmit(struct sk_buff skb, struct net_device dev) · 164 unmodified lines
165	165	{
166	166	struct xenvif *vif = netdev_priv(dev);
167	167	struct xenvif_queue *queue = NULL;
168		− unsigned int num_queues = vif->num_queues;
	168	+ unsigned int num_queues;
169	169	u16 index;
170	170	struct xenvif_rx_cb *cb;
171	171
172	172	BUG_ON(skb->dev != dev);
173	173
174		− /* Drop the packet if queues are not set up */
	174	+ /* Drop the packet if queues are not set up.
	175	+ * This handler should be called inside an RCU read section
	176	+ * so we don't need to enter it here explicitly.
	177	+ */
	178	+ num_queues = READ_ONCE(vif->num_queues);
175	179	if (num_queues < 1)
176	180	goto drop;
177	181
@@ -222,18 +226,18 @@ static struct net_device_stats xenvif_get_stats(struct net_device dev) · 44 unmodified lines
222	226	{
223	227	struct xenvif *vif = netdev_priv(dev);
224	228	struct xenvif_queue *queue = NULL;
	229	+ unsigned int num_queues;
225	230	u64 rx_bytes = 0;
226	231	u64 rx_packets = 0;
227	232	u64 tx_bytes = 0;
228	233	u64 tx_packets = 0;
229	234	unsigned int index;
230	235
231		− spin_lock(&vif->lock);
232		− if (vif->queues == NULL)
233		− goto out;
	236	+ rcu_read_lock();
	237	+ num_queues = READ_ONCE(vif->num_queues);
234	238
235	239	/* Aggregate tx and rx stats from each queue */
236		− for (index = 0; index < vif->num_queues; ++index) {
	240	+ for (index = 0; index < num_queues; ++index) {
237	241	queue = &vif->queues[index];
238	242	rx_bytes += queue->stats.rx_bytes;
239	243	rx_packets += queue->stats.rx_packets;
@@ -241,8 +245,7 @@ static struct net_device_stats xenvif_get_stats(struct net_device dev) · 1 unmodified line
241	245	tx_packets += queue->stats.tx_packets;
242	246	}
243	247
244		−out:
245		− spin_unlock(&vif->lock);
	248	+ rcu_read_unlock();
246	249
247	250	vif->dev->stats.rx_bytes = rx_bytes;
248	251	vif->dev->stats.rx_packets = rx_packets;
@@ -378,10 +381,13 @@ static void xenvif_get_ethtool_stats(struct net_device *dev, · 129 unmodified lines
378	381	struct ethtool_stats stats, u64 data)
379	382	{
380	383	struct xenvif *vif = netdev_priv(dev);
381		− unsigned int num_queues = vif->num_queues;
	384	+ unsigned int num_queues;
382	385	int i;
383	386	unsigned int queue_index;
384	387
	388	+ rcu_read_lock();
	389	+ num_queues = READ_ONCE(vif->num_queues);
	390	+
385	391	for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++) {
386	392	unsigned long accum = 0;
387	393	for (queue_index = 0; queue_index < num_queues; ++queue_index) {
@@ -390,6 +396,8 @@ static void xenvif_get_ethtool_stats(struct net_device *dev, · 2 unmodified lines
390	396	}
391	397	data[i] = accum;
392	398	}
	399	+
	400	+ rcu_read_unlock();
393	401	}
394	402
395	403	static void xenvif_get_strings(struct net_device dev, u32 stringset, u8 data)

drivers/net/xen-netback/netback.c+1−1

@@ -214,7 +214,7 @@ static void xenvif_fatal_tx_err(struct xenvif *vif) · 213 unmodified lines
214	214	netdev_err(vif->dev, "fatal error; disabling device\n");
215	215	vif->disabled = true;
216	216	/* Disable the vif from queue 0's kthread */
217		− if (vif->queues)
	217	+ if (vif->num_queues)
218	218	xenvif_kick_thread(&vif->queues[0]);
219	219	}
220	220

drivers/net/xen-netback/xenbus.c+10−10

@@ -495,26 +495,26 @@ static void backend_disconnect(struct backend_info *be) · 494 unmodified lines
495	495	struct xenvif *vif = be->vif;
496	496
497	497	if (vif) {
	498	+ unsigned int num_queues = vif->num_queues;
498	499	unsigned int queue_index;
499		− struct xenvif_queue *queues;
500	500
501	501	xen_unregister_watchers(vif);
502	502	#ifdef CONFIG_DEBUG_FS
503	503	xenvif_debugfs_delif(vif);
504	504	#endif /* CONFIG_DEBUG_FS */
505	505	xenvif_disconnect_data(vif);
506		− for (queue_index = 0;
507		− queue_index < vif->num_queues;
508		− ++queue_index)
509		− xenvif_deinit_queue(&vif->queues[queue_index]);
510	506
511		− spin_lock(&vif->lock);
512		− queues = vif->queues;
	507	+ /* At this point some of the handlers may still be active
	508	+ * so we need to have additional synchronization here.
	509	+ */
513	510	vif->num_queues = 0;
514		− vif->queues = NULL;
515		− spin_unlock(&vif->lock);
	511	+ synchronize_net();
516	512
517		− vfree(queues);
	513	+ for (queue_index = 0; queue_index < num_queues; ++queue_index)
	514	+ xenvif_deinit_queue(&vif->queues[queue_index]);
	515	+
	516	+ vfree(vif->queues);
	517	+ vif->queues = NULL;
518	518
519	519	xenvif_disconnect_ctrl(vif);
520	520	}

drivers/pci/dwc/pci-exynos.c+4−4

@@ -132,10 +132,6 @@ static int exynos5440_pcie_get_mem_resources(struct platform_device *pdev, · 131 unmodified lines
132	132	struct device *dev = pci->dev;
133	133	struct resource *res;
134	134
135		− /* If using the PHY framework, doesn't need to get other resource */
136		− if (ep->using_phy)
137		− return 0;
138		−
139	135	ep->mem_res = devm_kzalloc(dev, sizeof(*ep->mem_res), GFP_KERNEL);
140	136	if (!ep->mem_res)
141	137	return -ENOMEM;
@@ -145,6 +141,10 @@ static int exynos5440_pcie_get_mem_resources(struct platform_device *pdev, · 3 unmodified lines
145	141	if (IS_ERR(ep->mem_res->elbi_base))
146	142	return PTR_ERR(ep->mem_res->elbi_base);
147	143
	144	+ /* If using the PHY framework, doesn't need to get other resource */
	145	+ if (ep->using_phy)
	146	+ return 0;
	147	+
148	148	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
149	149	ep->mem_res->phy_base = devm_ioremap_resource(dev, res);
150	150	if (IS_ERR(ep->mem_res->phy_base))

drivers/pci/pcie/aspm.c+2−3

@@ -478,7 +478,7 @@ static void aspm_calc_l1ss_info(struct pcie_link_state *link, · 477 unmodified lines
478	478
479	479	static void pcie_aspm_cap_init(struct pcie_link_state *link, int blacklist)
480	480	{
481		− struct pci_dev child, parent = link->pdev;
	481	+ struct pci_dev child = link->downstream, parent = link->pdev;
482	482	struct pci_bus *linkbus = parent->subordinate;
483	483	struct aspm_register_info upreg, dwreg;
484	484
@@ -491,9 +491,7 @@ static void pcie_aspm_cap_init(struct pcie_link_state *link, int blacklist) · 6 unmodified lines
491	491
492	492	/* Get upstream/downstream components' register state */
493	493	pcie_get_aspm_reg(parent, &upreg);
494		− child = pci_function_0(linkbus);
495	494	pcie_get_aspm_reg(child, &dwreg);
496		− link->downstream = child;
497	495
498	496	/*
499	497	* If ASPM not supported, don't mess with the clocks and link,
@@ -800,6 +798,7 @@ static struct pcie_link_state alloc_pcie_link_state(struct pci_dev pdev) · 300 unmodified lines
800	798	INIT_LIST_HEAD(&link->children);
801	799	INIT_LIST_HEAD(&link->link);
802	800	link->pdev = pdev;
	801	+ link->downstream = pci_function_0(pdev->subordinate);
803	802
804	803	/*
805	804	* Root Ports and PCI/PCI-X to PCIe Bridges are roots of PCIe

drivers/pci/quirks.c+1−0

@@ -2174,6 +2174,7 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_LSI_LOGIC, 0x005d, quirk_blacklist_vpd); · 2173 unmodified lines
2174	2174	DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_LSI_LOGIC, 0x005f, quirk_blacklist_vpd);
2175	2175	DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATTANSIC, PCI_ANY_ID,
2176	2176	quirk_blacklist_vpd);
	2177	+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_QLOGIC, 0x2261, quirk_blacklist_vpd);
2177	2178
2178	2179	/*
2179	2180	* For Broadcom 5706, 5708, 5709 rev. A nics, any read beyond the

drivers/pinctrl/qcom/pinctrl-msm.c+15−0

@@ -422,6 +422,20 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in · 421 unmodified lines
422	422	return 0;
423	423	}
424	424
	425	+static int msm_gpio_get_direction(struct gpio_chip *chip, unsigned int offset)
	426	+{
	427	+ struct msm_pinctrl *pctrl = gpiochip_get_data(chip);
	428	+ const struct msm_pingroup *g;
	429	+ u32 val;
	430	+
	431	+ g = &pctrl->soc->groups[offset];
	432	+
	433	+ val = readl(pctrl->regs + g->ctl_reg);
	434	+
	435	+ /* 0 = output, 1 = input */
	436	+ return val & BIT(g->oe_bit) ? 0 : 1;
	437	+}
	438	+
425	439	static int msm_gpio_get(struct gpio_chip *chip, unsigned offset)
426	440	{
427	441	const struct msm_pingroup *g;
@@ -510,6 +524,7 @@ static void msm_gpio_dbg_show(struct seq_file s, struct gpio_chip chip) · 82 unmodified lines
510	524	static struct gpio_chip msm_gpio_template = {
511	525	.direction_input = msm_gpio_direction_input,
512	526	.direction_output = msm_gpio_direction_output,
	527	+ .get_direction = msm_gpio_get_direction,
513	528	.get = msm_gpio_get,
514	529	.set = msm_gpio_set,
515	530	.request = gpiochip_generic_request,

drivers/pinctrl/uniphier/pinctrl-uniphier-ld11.c+6−6

@@ -390,22 +390,22 @@ static const struct pinctrl_pin_desc uniphier_ld11_pins[] = { · 389 unmodified lines
390	390	UNIPHIER_PINCTRL_PIN(140, "AO1D0", 140,
391	391	140, UNIPHIER_PIN_DRV_1BIT,
392	392	140, UNIPHIER_PIN_PULL_DOWN),
393		− UNIPHIER_PINCTRL_PIN(141, "TCON0", 141,
	393	+ UNIPHIER_PINCTRL_PIN(141, "AO1D1", 141,
394	394	141, UNIPHIER_PIN_DRV_1BIT,
395	395	141, UNIPHIER_PIN_PULL_DOWN),
396		− UNIPHIER_PINCTRL_PIN(142, "TCON1", 142,
	396	+ UNIPHIER_PINCTRL_PIN(142, "AO1D2", 142,
397	397	142, UNIPHIER_PIN_DRV_1BIT,
398	398	142, UNIPHIER_PIN_PULL_DOWN),
399		− UNIPHIER_PINCTRL_PIN(143, "TCON2", 143,
	399	+ UNIPHIER_PINCTRL_PIN(143, "XIRQ9", 143,
400	400	143, UNIPHIER_PIN_DRV_1BIT,
401	401	143, UNIPHIER_PIN_PULL_DOWN),
402		− UNIPHIER_PINCTRL_PIN(144, "TCON3", 144,
	402	+ UNIPHIER_PINCTRL_PIN(144, "XIRQ10", 144,
403	403	144, UNIPHIER_PIN_DRV_1BIT,
404	404	144, UNIPHIER_PIN_PULL_DOWN),
405		− UNIPHIER_PINCTRL_PIN(145, "TCON4", 145,
	405	+ UNIPHIER_PINCTRL_PIN(145, "XIRQ11", 145,
406	406	145, UNIPHIER_PIN_DRV_1BIT,
407	407	145, UNIPHIER_PIN_PULL_DOWN),
408		− UNIPHIER_PINCTRL_PIN(146, "TCON5", 146,
	408	+ UNIPHIER_PINCTRL_PIN(146, "XIRQ13", 146,
409	409	146, UNIPHIER_PIN_DRV_1BIT,
410	410	146, UNIPHIER_PIN_PULL_DOWN),
411	411	UNIPHIER_PINCTRL_PIN(147, "PWMA", 147,

drivers/platform/x86/asus-nb-wmi.c+2−47

@@ -103,15 +103,6 @@ static struct quirk_entry quirk_asus_x200ca = { · 102 unmodified lines
103	103	.wapf = 2,
104	104	};
105	105
106		−static struct quirk_entry quirk_no_rfkill = {
107		− .no_rfkill = true,
108		−};
109		−
110		−static struct quirk_entry quirk_no_rfkill_wapf4 = {
111		− .wapf = 4,
112		− .no_rfkill = true,
113		−};
114		−
115	106	static struct quirk_entry quirk_asus_ux303ub = {
116	107	.wmi_backlight_native = true,
117	108	};
@@ -194,7 +185,7 @@ static const struct dmi_system_id asus_quirks[] = { · 76 unmodified lines
194	185	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
195	186	DMI_MATCH(DMI_PRODUCT_NAME, "X456UA"),
196	187	},
197		− .driver_data = &quirk_no_rfkill_wapf4,
	188	+ .driver_data = &quirk_asus_wapf4,
198	189	},
199	190	{
200	191	.callback = dmi_matched,
@@ -203,7 +194,7 @@ static const struct dmi_system_id asus_quirks[] = { · 2 unmodified lines
203	194	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
204	195	DMI_MATCH(DMI_PRODUCT_NAME, "X456UF"),
205	196	},
206		− .driver_data = &quirk_no_rfkill_wapf4,
	197	+ .driver_data = &quirk_asus_wapf4,
207	198	},
208	199	{
209	200	.callback = dmi_matched,
@@ -367,42 +358,6 @@ static const struct dmi_system_id asus_quirks[] = { · 157 unmodified lines
367	358	},
368	359	.driver_data = &quirk_asus_x200ca,
369	360	},
370		− {
371		− .callback = dmi_matched,
372		− .ident = "ASUSTeK COMPUTER INC. X555UB",
373		− .matches = {
374		− DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
375		− DMI_MATCH(DMI_PRODUCT_NAME, "X555UB"),
376		− },
377		− .driver_data = &quirk_no_rfkill,
378		− },
379		− {
380		− .callback = dmi_matched,
381		− .ident = "ASUSTeK COMPUTER INC. N552VW",
382		− .matches = {
383		− DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
384		− DMI_MATCH(DMI_PRODUCT_NAME, "N552VW"),
385		− },
386		− .driver_data = &quirk_no_rfkill,
387		− },
388		− {
389		− .callback = dmi_matched,
390		− .ident = "ASUSTeK COMPUTER INC. U303LB",
391		− .matches = {
392		− DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
393		− DMI_MATCH(DMI_PRODUCT_NAME, "U303LB"),
394		− },
395		− .driver_data = &quirk_no_rfkill,
396		− },
397		− {
398		− .callback = dmi_matched,
399		− .ident = "ASUSTeK COMPUTER INC. Z550MA",
400		− .matches = {
401		− DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
402		− DMI_MATCH(DMI_PRODUCT_NAME, "Z550MA"),
403		− },
404		− .driver_data = &quirk_no_rfkill,
405		− },
406	361	{
407	362	.callback = dmi_matched,
408	363	.ident = "ASUSTeK COMPUTER INC. UX303UB",

drivers/platform/x86/asus-wmi.c+17−5

@@ -159,6 +159,8 @@ MODULE_LICENSE("GPL"); · 158 unmodified lines
159	159	#define USB_INTEL_XUSB2PR 0xD0
160	160	#define PCI_DEVICE_ID_INTEL_LYNXPOINT_LP_XHCI 0x9c31
161	161
	162	+static const char * const ashs_ids[] = { "ATK4001", "ATK4002", NULL };
	163	+
162	164	struct bios_args {
163	165	u32 arg0;
164	166	u32 arg1;
@@ -2051,6 +2053,16 @@ static int asus_wmi_fan_init(struct asus_wmi *asus) · 1886 unmodified lines
2051	2053	return 0;
2052	2054	}
2053	2055
	2056	+static bool ashs_present(void)
	2057	+{
	2058	+ int i = 0;
	2059	+ while (ashs_ids[i]) {
	2060	+ if (acpi_dev_found(ashs_ids[i++]))
	2061	+ return true;
	2062	+ }
	2063	+ return false;
	2064	+}
	2065	+
2054	2066	/*
2055	2067	* WMI Driver
2056	2068	*/
@@ -2095,7 +2107,11 @@ static int asus_wmi_add(struct platform_device *pdev) · 38 unmodified lines
2095	2107	if (err)
2096	2108	goto fail_leds;
2097	2109
2098		− if (!asus->driver->quirks->no_rfkill) {
	2110	+ asus_wmi_get_devstate(asus, ASUS_WMI_DEVID_WLAN, &result);
	2111	+ if (result & (ASUS_WMI_DSTS_PRESENCE_BIT \| ASUS_WMI_DSTS_USER_BIT))
	2112	+ asus->driver->wlan_ctrl_by_user = 1;
	2113	+
	2114	+ if (!(asus->driver->wlan_ctrl_by_user && ashs_present())) {
2099	2115	err = asus_wmi_rfkill_init(asus);
2100	2116	if (err)
2101	2117	goto fail_rfkill;
@@ -2134,10 +2150,6 @@ static int asus_wmi_add(struct platform_device *pdev) · 32 unmodified lines
2134	2150	if (err)
2135	2151	goto fail_debugfs;
2136	2152
2137		− asus_wmi_get_devstate(asus, ASUS_WMI_DEVID_WLAN, &result);
2138		− if (result & (ASUS_WMI_DSTS_PRESENCE_BIT \| ASUS_WMI_DSTS_USER_BIT))
2139		− asus->driver->wlan_ctrl_by_user = 1;
2140		−
2141	2153	return 0;
2142	2154
2143	2155	fail_debugfs:

drivers/platform/x86/asus-wmi.h+0−1

@@ -39,7 +39,6 @@ struct key_entry; · 38 unmodified lines
39	39	struct asus_wmi;
40	40
41	41	struct quirk_entry {
42		− bool no_rfkill;
43	42	bool hotplug_wireless;
44	43	bool scalar_panel_brightness;
45	44	bool store_backlight_power;

drivers/platform/x86/fujitsu-laptop.c+222−229

@@ -78,18 +78,18 @@ · 77 unmodified lines
78	78
79	79	#define FUJITSU_LCD_N_LEVELS 8
80	80
81		−#define ACPI_FUJITSU_CLASS "fujitsu"
82		−#define ACPI_FUJITSU_HID "FUJ02B1"
83		−#define ACPI_FUJITSU_DRIVER_NAME "Fujitsu laptop FUJ02B1 ACPI brightness driver"
84		−#define ACPI_FUJITSU_DEVICE_NAME "Fujitsu FUJ02B1"
85		−#define ACPI_FUJITSU_HOTKEY_HID "FUJ02E3"
86		−#define ACPI_FUJITSU_HOTKEY_DRIVER_NAME "Fujitsu laptop FUJ02E3 ACPI hotkeys driver"
87		−#define ACPI_FUJITSU_HOTKEY_DEVICE_NAME "Fujitsu FUJ02E3"
	81	+#define ACPI_FUJITSU_CLASS "fujitsu"
	82	+#define ACPI_FUJITSU_BL_HID "FUJ02B1"
	83	+#define ACPI_FUJITSU_BL_DRIVER_NAME "Fujitsu laptop FUJ02B1 ACPI brightness driver"
	84	+#define ACPI_FUJITSU_BL_DEVICE_NAME "Fujitsu FUJ02B1"
	85	+#define ACPI_FUJITSU_LAPTOP_HID "FUJ02E3"
	86	+#define ACPI_FUJITSU_LAPTOP_DRIVER_NAME "Fujitsu laptop FUJ02E3 ACPI hotkeys driver"
	87	+#define ACPI_FUJITSU_LAPTOP_DEVICE_NAME "Fujitsu FUJ02E3"
88	88
89	89	#define ACPI_FUJITSU_NOTIFY_CODE1 0x80
90	90
91	91	/* FUNC interface - command values */
92		−#define FUNC_RFKILL 0x1000
	92	+#define FUNC_FLAGS 0x1000
93	93	#define FUNC_LEDS 0x1001
94	94	#define FUNC_BUTTONS 0x1002
95	95	#define FUNC_BACKLIGHT 0x1004
@@ -97,6 +97,11 @@ · 1 unmodified line
97	97	/* FUNC interface - responses */
98	98	#define UNSUPPORTED_CMD 0x80000000
99	99
	100	+/* FUNC interface - status flags */
	101	+#define FLAG_RFKILL 0x020
	102	+#define FLAG_LID 0x100
	103	+#define FLAG_DOCK 0x200
	104	+
100	105	#if IS_ENABLED(CONFIG_LEDS_CLASS)
101	106	/* FUNC interface - LED control */
102	107	#define FUNC_LED_OFF 0x1
@@ -136,7 +141,7 @@ · 33 unmodified lines
136	141	#endif
137	142
138	143	/* Device controlling the backlight and associated keys */
139		−struct fujitsu_t {
	144	+struct fujitsu_bl {
140	145	acpi_handle acpi_handle;
141	146	struct acpi_device *dev;
142	147	struct input_dev *input;
@@ -150,12 +155,12 @@ struct fujitsu_t { · 7 unmodified lines
150	155	unsigned int brightness_level;
151	156	};
152	157
153		−static struct fujitsu_t *fujitsu;
	158	+static struct fujitsu_bl *fujitsu_bl;
154	159	static int use_alt_lcd_levels = -1;
155	160	static int disable_brightness_adjust = -1;
156	161
157		−/* Device used to access other hotkeys on the laptop */
158		−struct fujitsu_hotkey_t {
	162	+/* Device used to access hotkeys and other features on the laptop */
	163	+struct fujitsu_laptop {
159	164	acpi_handle acpi_handle;
160	165	struct acpi_device *dev;
161	166	struct input_dev *input;
@@ -163,17 +168,15 @@ struct fujitsu_hotkey_t { · 1 unmodified line
163	168	struct platform_device *pf_device;
164	169	struct kfifo fifo;
165	170	spinlock_t fifo_lock;
166		− int rfkill_supported;
167		− int rfkill_state;
	171	+ int flags_supported;
	172	+ int flags_state;
168	173	int logolamp_registered;
169	174	int kblamps_registered;
170	175	int radio_led_registered;
171	176	int eco_led_registered;
172	177	};
173	178
174		−static struct fujitsu_hotkey_t *fujitsu_hotkey;
175		−
176		−static void acpi_fujitsu_hotkey_notify(struct acpi_device *device, u32 event);
	179	+static struct fujitsu_laptop *fujitsu_laptop;
177	180
178	181	#if IS_ENABLED(CONFIG_LEDS_CLASS)
179	182	static enum led_brightness logolamp_get(struct led_classdev *cdev);
@@ -222,8 +225,6 @@ static struct led_classdev eco_led = { · 42 unmodified lines
222	225	static u32 dbg_level = 0x03;
223	226	#endif
224	227
225		−static void acpi_fujitsu_notify(struct acpi_device *device, u32 event);
226		−
227	228	/* Fujitsu ACPI interface function */
228	229
229	230	static int call_fext_func(int cmd, int arg0, int arg1, int arg2)
@@ -239,7 +240,7 @@ static int call_fext_func(int cmd, int arg0, int arg1, int arg2) · 9 unmodified lines
239	240	unsigned long long value;
240	241	acpi_handle handle = NULL;
241	242
242		− status = acpi_get_handle(fujitsu_hotkey->acpi_handle, "FUNC", &handle);
	243	+ status = acpi_get_handle(fujitsu_laptop->acpi_handle, "FUNC", &handle);
243	244	if (ACPI_FAILURE(status)) {
244	245	vdbg_printk(FUJLAPTOP_DBG_ERROR,
245	246	"FUNC interface is not present\n");
@@ -300,9 +301,9 @@ static int radio_led_set(struct led_classdev *cdev, · 54 unmodified lines
300	301	enum led_brightness brightness)
301	302	{
302	303	if (brightness >= LED_FULL)
303		− return call_fext_func(FUNC_RFKILL, 0x5, RADIO_LED_ON, RADIO_LED_ON);
	304	+ return call_fext_func(FUNC_FLAGS, 0x5, RADIO_LED_ON, RADIO_LED_ON);
304	305	else
305		− return call_fext_func(FUNC_RFKILL, 0x5, RADIO_LED_ON, 0x0);
	306	+ return call_fext_func(FUNC_FLAGS, 0x5, RADIO_LED_ON, 0x0);
306	307	}
307	308
308	309	static int eco_led_set(struct led_classdev *cdev,
@@ -346,7 +347,7 @@ static enum led_brightness radio_led_get(struct led_classdev *cdev) · 37 unmodified lines
346	347	{
347	348	enum led_brightness brightness = LED_OFF;
348	349
349		− if (call_fext_func(FUNC_RFKILL, 0x4, 0x0, 0x0) & RADIO_LED_ON)
	350	+ if (call_fext_func(FUNC_FLAGS, 0x4, 0x0, 0x0) & RADIO_LED_ON)
350	351	brightness = LED_FULL;
351	352
352	353	return brightness;
@@ -373,10 +374,10 @@ static int set_lcd_level(int level) · 20 unmodified lines
373	374	vdbg_printk(FUJLAPTOP_DBG_TRACE, "set lcd level via SBLL [%d]\n",
374	375	level);
375	376
376		− if (level < 0 \|\| level >= fujitsu->max_brightness)
	377	+ if (level < 0 \|\| level >= fujitsu_bl->max_brightness)
377	378	return -EINVAL;
378	379
379		− status = acpi_get_handle(fujitsu->acpi_handle, "SBLL", &handle);
	380	+ status = acpi_get_handle(fujitsu_bl->acpi_handle, "SBLL", &handle);
380	381	if (ACPI_FAILURE(status)) {
381	382	vdbg_printk(FUJLAPTOP_DBG_ERROR, "SBLL not present\n");
382	383	return -ENODEV;
@@ -398,10 +399,10 @@ static int set_lcd_level_alt(int level) · 15 unmodified lines
398	399	vdbg_printk(FUJLAPTOP_DBG_TRACE, "set lcd level via SBL2 [%d]\n",
399	400	level);
400	401
401		− if (level < 0 \|\| level >= fujitsu->max_brightness)
	402	+ if (level < 0 \|\| level >= fujitsu_bl->max_brightness)
402	403	return -EINVAL;
403	404
404		− status = acpi_get_handle(fujitsu->acpi_handle, "SBL2", &handle);
	405	+ status = acpi_get_handle(fujitsu_bl->acpi_handle, "SBL2", &handle);
405	406	if (ACPI_FAILURE(status)) {
406	407	vdbg_printk(FUJLAPTOP_DBG_ERROR, "SBL2 not present\n");
407	408	return -ENODEV;
@@ -421,19 +422,19 @@ static int get_lcd_level(void) · 13 unmodified lines
421	422
422	423	vdbg_printk(FUJLAPTOP_DBG_TRACE, "get lcd level via GBLL\n");
423	424
424		− status =
425		− acpi_evaluate_integer(fujitsu->acpi_handle, "GBLL", NULL, &state);
	425	+ status = acpi_evaluate_integer(fujitsu_bl->acpi_handle, "GBLL", NULL,
	426	+ &state);
426	427	if (ACPI_FAILURE(status))
427	428	return 0;
428	429
429		− fujitsu->brightness_level = state & 0x0fffffff;
	430	+ fujitsu_bl->brightness_level = state & 0x0fffffff;
430	431
431	432	if (state & 0x80000000)
432		− fujitsu->brightness_changed = 1;
	433	+ fujitsu_bl->brightness_changed = 1;
433	434	else
434		− fujitsu->brightness_changed = 0;
	435	+ fujitsu_bl->brightness_changed = 0;
435	436
436		− return fujitsu->brightness_level;
	437	+ return fujitsu_bl->brightness_level;
437	438	}
438	439
439	440	static int get_max_brightness(void)
@@ -443,14 +444,14 @@ static int get_max_brightness(void) · 3 unmodified lines
443	444
444	445	vdbg_printk(FUJLAPTOP_DBG_TRACE, "get max lcd level via RBLL\n");
445	446
446		− status =
447		− acpi_evaluate_integer(fujitsu->acpi_handle, "RBLL", NULL, &state);
	447	+ status = acpi_evaluate_integer(fujitsu_bl->acpi_handle, "RBLL", NULL,
	448	+ &state);
448	449	if (ACPI_FAILURE(status))
449	450	return -1;
450	451
451		− fujitsu->max_brightness = state;
	452	+ fujitsu_bl->max_brightness = state;
452	453
453		− return fujitsu->max_brightness;
	454	+ return fujitsu_bl->max_brightness;
454	455	}
455	456
456	457	/* Backlight device stuff */
@@ -483,7 +484,7 @@ static int bl_update_status(struct backlight_device *b) · 26 unmodified lines
483	484	return ret;
484	485	}
485	486
486		−static const struct backlight_ops fujitsubl_ops = {
	487	+static const struct backlight_ops fujitsu_bl_ops = {
487	488	.get_brightness = bl_get_brightness,
488	489	.update_status = bl_update_status,
489	490	};
@@ -511,7 +512,7 @@ show_brightness_changed(struct device *dev, · 21 unmodified lines
511	512
512	513	int ret;
513	514
514		− ret = fujitsu->brightness_changed;
	515	+ ret = fujitsu_bl->brightness_changed;
515	516	if (ret < 0)
516	517	return ret;
517	518
@@ -539,7 +540,7 @@ static ssize_t store_lcd_level(struct device *dev, · 21 unmodified lines
539	540	int level, ret;
540	541
541	542	if (sscanf(buf, "%i", &level) != 1
542		− \|\| (level < 0 \|\| level >= fujitsu->max_brightness))
	543	+ \|\| (level < 0 \|\| level >= fujitsu_bl->max_brightness))
543	544	return -EINVAL;
544	545
545	546	if (use_alt_lcd_levels)
@@ -567,9 +568,9 @@ static ssize_t · 21 unmodified lines
567	568	show_lid_state(struct device *dev,
568	569	struct device_attribute attr, char buf)
569	570	{
570		− if (!(fujitsu_hotkey->rfkill_supported & 0x100))
	571	+ if (!(fujitsu_laptop->flags_supported & FLAG_LID))
571	572	return sprintf(buf, "unknown\n");
572		− if (fujitsu_hotkey->rfkill_state & 0x100)
	573	+ if (fujitsu_laptop->flags_state & FLAG_LID)
573	574	return sprintf(buf, "open\n");
574	575	else
575	576	return sprintf(buf, "closed\n");
@@ -579,9 +580,9 @@ static ssize_t · 3 unmodified lines
579	580	show_dock_state(struct device *dev,
580	581	struct device_attribute attr, char buf)
581	582	{
582		− if (!(fujitsu_hotkey->rfkill_supported & 0x200))
	583	+ if (!(fujitsu_laptop->flags_supported & FLAG_DOCK))
583	584	return sprintf(buf, "unknown\n");
584		− if (fujitsu_hotkey->rfkill_state & 0x200)
	585	+ if (fujitsu_laptop->flags_state & FLAG_DOCK)
585	586	return sprintf(buf, "docked\n");
586	587	else
587	588	return sprintf(buf, "undocked\n");
@@ -591,9 +592,9 @@ static ssize_t · 3 unmodified lines
591	592	show_radios_state(struct device *dev,
592	593	struct device_attribute attr, char buf)
593	594	{
594		− if (!(fujitsu_hotkey->rfkill_supported & 0x20))
	595	+ if (!(fujitsu_laptop->flags_supported & FLAG_RFKILL))
595	596	return sprintf(buf, "unknown\n");
596		− if (fujitsu_hotkey->rfkill_state & 0x20)
	597	+ if (fujitsu_laptop->flags_state & FLAG_RFKILL)
597	598	return sprintf(buf, "on\n");
598	599	else
599	600	return sprintf(buf, "killed\n");
@@ -607,7 +608,7 @@ static DEVICE_ATTR(lid, 0444, show_lid_state, ignore_store); · 7 unmodified lines
607	608	static DEVICE_ATTR(dock, 0444, show_dock_state, ignore_store);
608	609	static DEVICE_ATTR(radios, 0444, show_radios_state, ignore_store);
609	610
610		−static struct attribute *fujitsupf_attributes[] = {
	611	+static struct attribute *fujitsu_pf_attributes[] = {
611	612	&dev_attr_brightness_changed.attr,
612	613	&dev_attr_max_brightness.attr,
613	614	&dev_attr_lcd_level.attr,
@@ -617,11 +618,11 @@ static struct attribute *fujitsupf_attributes[] = { · 3 unmodified lines
617	618	NULL
618	619	};
619	620
620		−static struct attribute_group fujitsupf_attribute_group = {
621		− .attrs = fujitsupf_attributes
	621	+static struct attribute_group fujitsu_pf_attribute_group = {
	622	+ .attrs = fujitsu_pf_attributes
622	623	};
623	624
624		−static struct platform_driver fujitsupf_driver = {
	625	+static struct platform_driver fujitsu_pf_driver = {
625	626	.driver = {
626	627	.name = "fujitsu-laptop",
627	628	}
@@ -630,39 +631,30 @@ static struct platform_driver fujitsupf_driver = { · 2 unmodified lines
630	631	static void __init dmi_check_cb_common(const struct dmi_system_id *id)
631	632	{
632	633	pr_info("Identified laptop model '%s'\n", id->ident);
633		− if (use_alt_lcd_levels == -1) {
634		− if (acpi_has_method(NULL,
635		− "\\_SB.PCI0.LPCB.FJEX.SBL2"))
636		− use_alt_lcd_levels = 1;
637		− else
638		− use_alt_lcd_levels = 0;
639		− vdbg_printk(FUJLAPTOP_DBG_TRACE, "auto-detected usealt as "
640		− "%i\n", use_alt_lcd_levels);
641		− }
642	634	}
643	635
644	636	static int __init dmi_check_cb_s6410(const struct dmi_system_id *id)
645	637	{
646	638	dmi_check_cb_common(id);
647		− fujitsu->keycode1 = KEY_SCREENLOCK; /* "Lock" */
648		− fujitsu->keycode2 = KEY_HELP; /* "Mobility Center" */
	639	+ fujitsu_bl->keycode1 = KEY_SCREENLOCK; /* "Lock" */
	640	+ fujitsu_bl->keycode2 = KEY_HELP; /* "Mobility Center" */
649	641	return 1;
650	642	}
651	643
652	644	static int __init dmi_check_cb_s6420(const struct dmi_system_id *id)
653	645	{
654	646	dmi_check_cb_common(id);
655		− fujitsu->keycode1 = KEY_SCREENLOCK; /* "Lock" */
656		− fujitsu->keycode2 = KEY_HELP; /* "Mobility Center" */
	647	+ fujitsu_bl->keycode1 = KEY_SCREENLOCK; /* "Lock" */
	648	+ fujitsu_bl->keycode2 = KEY_HELP; /* "Mobility Center" */
657	649	return 1;
658	650	}
659	651
660	652	static int __init dmi_check_cb_p8010(const struct dmi_system_id *id)
661	653	{
662	654	dmi_check_cb_common(id);
663		− fujitsu->keycode1 = KEY_HELP; /* "Support" */
664		− fujitsu->keycode3 = KEY_SWITCHVIDEOMODE; /* "Presentation" */
665		− fujitsu->keycode4 = KEY_WWW; /* "Internet" */
	655	+ fujitsu_bl->keycode1 = KEY_HELP; /* "Support" */
	656	+ fujitsu_bl->keycode3 = KEY_SWITCHVIDEOMODE; /* "Presentation" */
	657	+ fujitsu_bl->keycode4 = KEY_WWW; /* "Internet" */
666	658	return 1;
667	659	}
668	660
@@ -693,7 +685,7 @@ static const struct dmi_system_id fujitsu_dmi_table[] __initconst = { · 24 unmodified lines
693	685
694	686	/* ACPI device for LCD brightness control */
695	687
696		−static int acpi_fujitsu_add(struct acpi_device *device)
	688	+static int acpi_fujitsu_bl_add(struct acpi_device *device)
697	689	{
698	690	int state = 0;
699	691	struct input_dev *input;
@@ -702,22 +694,22 @@ static int acpi_fujitsu_add(struct acpi_device *device) · 2 unmodified lines
702	694	if (!device)
703	695	return -EINVAL;
704	696
705		− fujitsu->acpi_handle = device->handle;
706		− sprintf(acpi_device_name(device), "%s", ACPI_FUJITSU_DEVICE_NAME);
	697	+ fujitsu_bl->acpi_handle = device->handle;
	698	+ sprintf(acpi_device_name(device), "%s", ACPI_FUJITSU_BL_DEVICE_NAME);
707	699	sprintf(acpi_device_class(device), "%s", ACPI_FUJITSU_CLASS);
708		− device->driver_data = fujitsu;
	700	+ device->driver_data = fujitsu_bl;
709	701
710		− fujitsu->input = input = input_allocate_device();
	702	+ fujitsu_bl->input = input = input_allocate_device();
711	703	if (!input) {
712	704	error = -ENOMEM;
713	705	goto err_stop;
714	706	}
715	707
716		− snprintf(fujitsu->phys, sizeof(fujitsu->phys),
	708	+ snprintf(fujitsu_bl->phys, sizeof(fujitsu_bl->phys),
717	709	"%s/video/input0", acpi_device_hid(device));
718	710
719	711	input->name = acpi_device_name(device);
720		− input->phys = fujitsu->phys;
	712	+ input->phys = fujitsu_bl->phys;
721	713	input->id.bustype = BUS_HOST;
722	714	input->id.product = 0x06;
723	715	input->dev.parent = &device->dev;
@@ -730,7 +722,7 @@ static int acpi_fujitsu_add(struct acpi_device *device) · 6 unmodified lines
730	722	if (error)
731	723	goto err_free_input_dev;
732	724
733		− error = acpi_bus_update_power(fujitsu->acpi_handle, &state);
	725	+ error = acpi_bus_update_power(fujitsu_bl->acpi_handle, &state);
734	726	if (error) {
735	727	pr_err("Error reading power state\n");
736	728	goto err_unregister_input_dev;
@@ -740,7 +732,7 @@ static int acpi_fujitsu_add(struct acpi_device *device) · 3 unmodified lines
740	732	acpi_device_name(device), acpi_device_bid(device),
741	733	!device->power.state ? "on" : "off");
742	734
743		− fujitsu->dev = device;
	735	+ fujitsu_bl->dev = device;
744	736
745	737	if (acpi_has_method(device->handle, METHOD_NAME__INI)) {
746	738	vdbg_printk(FUJLAPTOP_DBG_INFO, "Invoking _INI\n");
@@ -750,6 +742,15 @@ static int acpi_fujitsu_add(struct acpi_device *device) · 3 unmodified lines
750	742	pr_err("_INI Method failed\n");
751	743	}
752	744
	745	+ if (use_alt_lcd_levels == -1) {
	746	+ if (acpi_has_method(NULL, "\\_SB.PCI0.LPCB.FJEX.SBL2"))
	747	+ use_alt_lcd_levels = 1;
	748	+ else
	749	+ use_alt_lcd_levels = 0;
	750	+ vdbg_printk(FUJLAPTOP_DBG_TRACE, "auto-detected usealt as %i\n",
	751	+ use_alt_lcd_levels);
	752	+ }
	753	+
753	754	/* do config (detect defaults) */
754	755	use_alt_lcd_levels = use_alt_lcd_levels == 1 ? 1 : 0;
755	756	disable_brightness_adjust = disable_brightness_adjust == 1 ? 1 : 0;
@@ -758,7 +759,7 @@ static int acpi_fujitsu_add(struct acpi_device *device) · 2 unmodified lines
758	759	use_alt_lcd_levels, disable_brightness_adjust);
759	760
760	761	if (get_max_brightness() <= 0)
761		− fujitsu->max_brightness = FUJITSU_LCD_N_LEVELS;
	762	+ fujitsu_bl->max_brightness = FUJITSU_LCD_N_LEVELS;
762	763	get_lcd_level();
763	764
764	765	return 0;
@@ -772,38 +773,38 @@ err_stop: · 7 unmodified lines
772	773	return error;
773	774	}
774	775
775		−static int acpi_fujitsu_remove(struct acpi_device *device)
	776	+static int acpi_fujitsu_bl_remove(struct acpi_device *device)
776	777	{
777		− struct fujitsu_t *fujitsu = acpi_driver_data(device);
778		− struct input_dev *input = fujitsu->input;
	778	+ struct fujitsu_bl *fujitsu_bl = acpi_driver_data(device);
	779	+ struct input_dev *input = fujitsu_bl->input;
779	780
780	781	input_unregister_device(input);
781	782
782		− fujitsu->acpi_handle = NULL;
	783	+ fujitsu_bl->acpi_handle = NULL;
783	784
784	785	return 0;
785	786	}
786	787
787	788	/* Brightness notify */
788	789
789		−static void acpi_fujitsu_notify(struct acpi_device *device, u32 event)
	790	+static void acpi_fujitsu_bl_notify(struct acpi_device *device, u32 event)
790	791	{
791	792	struct input_dev *input;
792	793	int keycode;
793	794	int oldb, newb;
794	795
795		− input = fujitsu->input;
	796	+ input = fujitsu_bl->input;
796	797
797	798	switch (event) {
798	799	case ACPI_FUJITSU_NOTIFY_CODE1:
799	800	keycode = 0;
800		− oldb = fujitsu->brightness_level;
	801	+ oldb = fujitsu_bl->brightness_level;
801	802	get_lcd_level();
802		− newb = fujitsu->brightness_level;
	803	+ newb = fujitsu_bl->brightness_level;
803	804
804	805	vdbg_printk(FUJLAPTOP_DBG_TRACE,
805	806	"brightness button event [%i -> %i (%i)]\n",
806		− oldb, newb, fujitsu->brightness_changed);
	807	+ oldb, newb, fujitsu_bl->brightness_changed);
807	808
808	809	if (oldb < newb) {
809	810	if (disable_brightness_adjust != 1) {
@@ -840,7 +841,7 @@ static void acpi_fujitsu_notify(struct acpi_device *device, u32 event) · 30 unmodified lines
840	841
841	842	/* ACPI device for hotkey handling */
842	843
843		−static int acpi_fujitsu_hotkey_add(struct acpi_device *device)
	844	+static int acpi_fujitsu_laptop_add(struct acpi_device *device)
844	845	{
845	846	int result = 0;
846	847	int state = 0;
@@ -851,42 +852,42 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 4 unmodified lines
851	852	if (!device)
852	853	return -EINVAL;
853	854
854		− fujitsu_hotkey->acpi_handle = device->handle;
	855	+ fujitsu_laptop->acpi_handle = device->handle;
855	856	sprintf(acpi_device_name(device), "%s",
856		− ACPI_FUJITSU_HOTKEY_DEVICE_NAME);
	857	+ ACPI_FUJITSU_LAPTOP_DEVICE_NAME);
857	858	sprintf(acpi_device_class(device), "%s", ACPI_FUJITSU_CLASS);
858		− device->driver_data = fujitsu_hotkey;
	859	+ device->driver_data = fujitsu_laptop;
859	860
860	861	/* kfifo */
861		− spin_lock_init(&fujitsu_hotkey->fifo_lock);
862		− error = kfifo_alloc(&fujitsu_hotkey->fifo, RINGBUFFERSIZE * sizeof(int),
	862	+ spin_lock_init(&fujitsu_laptop->fifo_lock);
	863	+ error = kfifo_alloc(&fujitsu_laptop->fifo, RINGBUFFERSIZE * sizeof(int),
863	864	GFP_KERNEL);
864	865	if (error) {
865	866	pr_err("kfifo_alloc failed\n");
866	867	goto err_stop;
867	868	}
868	869
869		− fujitsu_hotkey->input = input = input_allocate_device();
	870	+ fujitsu_laptop->input = input = input_allocate_device();
870	871	if (!input) {
871	872	error = -ENOMEM;
872	873	goto err_free_fifo;
873	874	}
874	875
875		− snprintf(fujitsu_hotkey->phys, sizeof(fujitsu_hotkey->phys),
	876	+ snprintf(fujitsu_laptop->phys, sizeof(fujitsu_laptop->phys),
876	877	"%s/video/input0", acpi_device_hid(device));
877	878
878	879	input->name = acpi_device_name(device);
879		− input->phys = fujitsu_hotkey->phys;
	880	+ input->phys = fujitsu_laptop->phys;
880	881	input->id.bustype = BUS_HOST;
881	882	input->id.product = 0x06;
882	883	input->dev.parent = &device->dev;
883	884
884	885	set_bit(EV_KEY, input->evbit);
885		− set_bit(fujitsu->keycode1, input->keybit);
886		− set_bit(fujitsu->keycode2, input->keybit);
887		− set_bit(fujitsu->keycode3, input->keybit);
888		− set_bit(fujitsu->keycode4, input->keybit);
889		− set_bit(fujitsu->keycode5, input->keybit);
	886	+ set_bit(fujitsu_bl->keycode1, input->keybit);
	887	+ set_bit(fujitsu_bl->keycode2, input->keybit);
	888	+ set_bit(fujitsu_bl->keycode3, input->keybit);
	889	+ set_bit(fujitsu_bl->keycode4, input->keybit);
	890	+ set_bit(fujitsu_bl->keycode5, input->keybit);
890	891	set_bit(KEY_TOUCHPAD_TOGGLE, input->keybit);
891	892	set_bit(KEY_UNKNOWN, input->keybit);
892	893
@@ -894,7 +895,7 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 1 unmodified line
894	895	if (error)
895	896	goto err_free_input_dev;
896	897
897		− error = acpi_bus_update_power(fujitsu_hotkey->acpi_handle, &state);
	898	+ error = acpi_bus_update_power(fujitsu_laptop->acpi_handle, &state);
898	899	if (error) {
899	900	pr_err("Error reading power state\n");
900	901	goto err_unregister_input_dev;
@@ -904,7 +905,7 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 3 unmodified lines
904	905	acpi_device_name(device), acpi_device_bid(device),
905	906	!device->power.state ? "on" : "off");
906	907
907		− fujitsu_hotkey->dev = device;
	908	+ fujitsu_laptop->dev = device;
908	909
909	910	if (acpi_has_method(device->handle, METHOD_NAME__INI)) {
910	911	vdbg_printk(FUJLAPTOP_DBG_INFO, "Invoking _INI\n");
@@ -920,27 +921,27 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 9 unmodified lines
920	921	; /* No action, result is discarded */
921	922	vdbg_printk(FUJLAPTOP_DBG_INFO, "Discarded %i ringbuffer entries\n", i);
922	923
923		− fujitsu_hotkey->rfkill_supported =
924		− call_fext_func(FUNC_RFKILL, 0x0, 0x0, 0x0);
	924	+ fujitsu_laptop->flags_supported =
	925	+ call_fext_func(FUNC_FLAGS, 0x0, 0x0, 0x0);
925	926
926	927	/* Make sure our bitmask of supported functions is cleared if the
927	928	RFKILL function block is not implemented, like on the S7020. */
928		− if (fujitsu_hotkey->rfkill_supported == UNSUPPORTED_CMD)
929		− fujitsu_hotkey->rfkill_supported = 0;
	929	+ if (fujitsu_laptop->flags_supported == UNSUPPORTED_CMD)
	930	+ fujitsu_laptop->flags_supported = 0;
930	931
931		− if (fujitsu_hotkey->rfkill_supported)
932		− fujitsu_hotkey->rfkill_state =
933		− call_fext_func(FUNC_RFKILL, 0x4, 0x0, 0x0);
	932	+ if (fujitsu_laptop->flags_supported)
	933	+ fujitsu_laptop->flags_state =
	934	+ call_fext_func(FUNC_FLAGS, 0x4, 0x0, 0x0);
934	935
935	936	/* Suspect this is a keymap of the application panel, print it */
936	937	pr_info("BTNI: [0x%x]\n", call_fext_func(FUNC_BUTTONS, 0x0, 0x0, 0x0));
937	938
938	939	#if IS_ENABLED(CONFIG_LEDS_CLASS)
939	940	if (call_fext_func(FUNC_LEDS, 0x0, 0x0, 0x0) & LOGOLAMP_POWERON) {
940		− result = led_classdev_register(&fujitsu->pf_device->dev,
	941	+ result = led_classdev_register(&fujitsu_bl->pf_device->dev,
941	942	&logolamp_led);
942	943	if (result == 0) {
943		− fujitsu_hotkey->logolamp_registered = 1;
	944	+ fujitsu_laptop->logolamp_registered = 1;
944	945	} else {
945	946	pr_err("Could not register LED handler for logo lamp, error %i\n",
946	947	result);
@@ -949,10 +950,10 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 2 unmodified lines
949	950
950	951	if ((call_fext_func(FUNC_LEDS, 0x0, 0x0, 0x0) & KEYBOARD_LAMPS) &&
951	952	(call_fext_func(FUNC_BUTTONS, 0x0, 0x0, 0x0) == 0x0)) {
952		− result = led_classdev_register(&fujitsu->pf_device->dev,
	953	+ result = led_classdev_register(&fujitsu_bl->pf_device->dev,
953	954	&kblamps_led);
954	955	if (result == 0) {
955		− fujitsu_hotkey->kblamps_registered = 1;
	956	+ fujitsu_laptop->kblamps_registered = 1;
956	957	} else {
957	958	pr_err("Could not register LED handler for keyboard lamps, error %i\n",
958	959	result);
@@ -966,10 +967,10 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 7 unmodified lines
966	967	* that an RF LED is present.
967	968	*/
968	969	if (call_fext_func(FUNC_BUTTONS, 0x0, 0x0, 0x0) & BIT(24)) {
969		− result = led_classdev_register(&fujitsu->pf_device->dev,
	970	+ result = led_classdev_register(&fujitsu_bl->pf_device->dev,
970	971	&radio_led);
971	972	if (result == 0) {
972		− fujitsu_hotkey->radio_led_registered = 1;
	973	+ fujitsu_laptop->radio_led_registered = 1;
973	974	} else {
974	975	pr_err("Could not register LED handler for radio LED, error %i\n",
975	976	result);
@@ -983,10 +984,10 @@ static int acpi_fujitsu_hotkey_add(struct acpi_device *device) · 7 unmodified lines
983	984	*/
984	985	if ((call_fext_func(FUNC_LEDS, 0x0, 0x0, 0x0) & BIT(14)) &&
985	986	(call_fext_func(FUNC_LEDS, 0x2, ECO_LED, 0x0) != UNSUPPORTED_CMD)) {
986		− result = led_classdev_register(&fujitsu->pf_device->dev,
	987	+ result = led_classdev_register(&fujitsu_bl->pf_device->dev,
987	988	&eco_led);
988	989	if (result == 0) {
989		− fujitsu_hotkey->eco_led_registered = 1;
	990	+ fujitsu_laptop->eco_led_registered = 1;
990	991	} else {
991	992	pr_err("Could not register LED handler for eco LED, error %i\n",
992	993	result);
@@ -1002,47 +1003,47 @@ err_unregister_input_dev: · 9 unmodified lines
1002	1003	err_free_input_dev:
1003	1004	input_free_device(input);
1004	1005	err_free_fifo:
1005		− kfifo_free(&fujitsu_hotkey->fifo);
	1006	+ kfifo_free(&fujitsu_laptop->fifo);
1006	1007	err_stop:
1007	1008	return error;
1008	1009	}
1009	1010
1010		−static int acpi_fujitsu_hotkey_remove(struct acpi_device *device)
	1011	+static int acpi_fujitsu_laptop_remove(struct acpi_device *device)
1011	1012	{
1012		− struct fujitsu_hotkey_t *fujitsu_hotkey = acpi_driver_data(device);
1013		− struct input_dev *input = fujitsu_hotkey->input;
	1013	+ struct fujitsu_laptop *fujitsu_laptop = acpi_driver_data(device);
	1014	+ struct input_dev *input = fujitsu_laptop->input;
1014	1015
1015	1016	#if IS_ENABLED(CONFIG_LEDS_CLASS)
1016		− if (fujitsu_hotkey->logolamp_registered)
	1017	+ if (fujitsu_laptop->logolamp_registered)
1017	1018	led_classdev_unregister(&logolamp_led);
1018	1019
1019		− if (fujitsu_hotkey->kblamps_registered)
	1020	+ if (fujitsu_laptop->kblamps_registered)
1020	1021	led_classdev_unregister(&kblamps_led);
1021	1022
1022		− if (fujitsu_hotkey->radio_led_registered)
	1023	+ if (fujitsu_laptop->radio_led_registered)
1023	1024	led_classdev_unregister(&radio_led);
1024	1025
1025		− if (fujitsu_hotkey->eco_led_registered)
	1026	+ if (fujitsu_laptop->eco_led_registered)
1026	1027	led_classdev_unregister(&eco_led);
1027	1028	#endif
1028	1029
1029	1030	input_unregister_device(input);
1030	1031
1031		− kfifo_free(&fujitsu_hotkey->fifo);
	1032	+ kfifo_free(&fujitsu_laptop->fifo);
1032	1033
1033		− fujitsu_hotkey->acpi_handle = NULL;
	1034	+ fujitsu_laptop->acpi_handle = NULL;
1034	1035
1035	1036	return 0;
1036	1037	}
1037	1038
1038		−static void acpi_fujitsu_hotkey_press(int keycode)
	1039	+static void acpi_fujitsu_laptop_press(int keycode)
1039	1040	{
1040		− struct input_dev *input = fujitsu_hotkey->input;
	1041	+ struct input_dev *input = fujitsu_laptop->input;
1041	1042	int status;
1042	1043
1043		− status = kfifo_in_locked(&fujitsu_hotkey->fifo,
	1044	+ status = kfifo_in_locked(&fujitsu_laptop->fifo,
1044	1045	(unsigned char *)&keycode, sizeof(keycode),
1045		− &fujitsu_hotkey->fifo_lock);
	1046	+ &fujitsu_laptop->fifo_lock);
1046	1047	if (status != sizeof(keycode)) {
1047	1048	vdbg_printk(FUJLAPTOP_DBG_WARN,
1048	1049	"Could not push keycode [0x%x]\n", keycode);
@@ -1054,16 +1055,16 @@ static void acpi_fujitsu_hotkey_press(int keycode) · 5 unmodified lines
1054	1055	"Push keycode into ringbuffer [%d]\n", keycode);
1055	1056	}
1056	1057
1057		−static void acpi_fujitsu_hotkey_release(void)
	1058	+static void acpi_fujitsu_laptop_release(void)
1058	1059	{
1059		− struct input_dev *input = fujitsu_hotkey->input;
	1060	+ struct input_dev *input = fujitsu_laptop->input;
1060	1061	int keycode, status;
1061	1062
1062	1063	while (true) {
1063		− status = kfifo_out_locked(&fujitsu_hotkey->fifo,
	1064	+ status = kfifo_out_locked(&fujitsu_laptop->fifo,
1064	1065	(unsigned char *)&keycode,
1065	1066	sizeof(keycode),
1066		− &fujitsu_hotkey->fifo_lock);
	1067	+ &fujitsu_laptop->fifo_lock);
1067	1068	if (status != sizeof(keycode))
1068	1069	return;
1069	1070	input_report_key(input, keycode, 0);
@@ -1073,14 +1074,14 @@ static void acpi_fujitsu_hotkey_release(void) · 3 unmodified lines
1073	1074	}
1074	1075	}
1075	1076
1076		−static void acpi_fujitsu_hotkey_notify(struct acpi_device *device, u32 event)
	1077	+static void acpi_fujitsu_laptop_notify(struct acpi_device *device, u32 event)
1077	1078	{
1078	1079	struct input_dev *input;
1079	1080	int keycode;
1080	1081	unsigned int irb = 1;
1081	1082	int i;
1082	1083
1083		− input = fujitsu_hotkey->input;
	1084	+ input = fujitsu_laptop->input;
1084	1085
1085	1086	if (event != ACPI_FUJITSU_NOTIFY_CODE1) {
1086	1087	keycode = KEY_UNKNOWN;
@@ -1093,9 +1094,9 @@ static void acpi_fujitsu_hotkey_notify(struct acpi_device *device, u32 event) · 6 unmodified lines
1093	1094	return;
1094	1095	}
1095	1096
1096		− if (fujitsu_hotkey->rfkill_supported)
1097		− fujitsu_hotkey->rfkill_state =
1098		− call_fext_func(FUNC_RFKILL, 0x4, 0x0, 0x0);
	1097	+ if (fujitsu_laptop->flags_supported)
	1098	+ fujitsu_laptop->flags_state =
	1099	+ call_fext_func(FUNC_FLAGS, 0x4, 0x0, 0x0);
1099	1100
1100	1101	i = 0;
1101	1102	while ((irb =
@@ -1103,19 +1104,19 @@ static void acpi_fujitsu_hotkey_notify(struct acpi_device *device, u32 event) · 1 unmodified line
1103	1104	&& (i++) < MAX_HOTKEY_RINGBUFFER_SIZE) {
1104	1105	switch (irb & 0x4ff) {
1105	1106	case KEY1_CODE:
1106		− keycode = fujitsu->keycode1;
	1107	+ keycode = fujitsu_bl->keycode1;
1107	1108	break;
1108	1109	case KEY2_CODE:
1109		− keycode = fujitsu->keycode2;
	1110	+ keycode = fujitsu_bl->keycode2;
1110	1111	break;
1111	1112	case KEY3_CODE:
1112		− keycode = fujitsu->keycode3;
	1113	+ keycode = fujitsu_bl->keycode3;
1113	1114	break;
1114	1115	case KEY4_CODE:
1115		− keycode = fujitsu->keycode4;
	1116	+ keycode = fujitsu_bl->keycode4;
1116	1117	break;
1117	1118	case KEY5_CODE:
1118		− keycode = fujitsu->keycode5;
	1119	+ keycode = fujitsu_bl->keycode5;
1119	1120	break;
1120	1121	case 0:
1121	1122	keycode = 0;
@@ -1128,17 +1129,17 @@ static void acpi_fujitsu_hotkey_notify(struct acpi_device *device, u32 event) · 6 unmodified lines
1128	1129	}
1129	1130
1130	1131	if (keycode > 0)
1131		− acpi_fujitsu_hotkey_press(keycode);
	1132	+ acpi_fujitsu_laptop_press(keycode);
1132	1133	else if (keycode == 0)
1133		− acpi_fujitsu_hotkey_release();
	1134	+ acpi_fujitsu_laptop_release();
1134	1135	}
1135	1136
1136	1137	/* On some models (first seen on the Skylake-based Lifebook
1137	1138	* E736/E746/E756), the touchpad toggle hotkey (Fn+F4) is
1138		− * handled in software; its state is queried using FUNC_RFKILL
	1139	+ * handled in software; its state is queried using FUNC_FLAGS
1139	1140	*/
1140		− if ((fujitsu_hotkey->rfkill_supported & BIT(26)) &&
1141		− (call_fext_func(FUNC_RFKILL, 0x1, 0x0, 0x0) & BIT(26))) {
	1141	+ if ((fujitsu_laptop->flags_supported & BIT(26)) &&
	1142	+ (call_fext_func(FUNC_FLAGS, 0x1, 0x0, 0x0) & BIT(26))) {
1142	1143	keycode = KEY_TOUCHPAD_TOGGLE;
1143	1144	input_report_key(input, keycode, 1);
1144	1145	input_sync(input);
@@ -1150,83 +1151,81 @@ static void acpi_fujitsu_hotkey_notify(struct acpi_device *device, u32 event) · 5 unmodified lines
1150	1151
1151	1152	/* Initialization */
1152	1153
1153		−static const struct acpi_device_id fujitsu_device_ids[] = {
1154		− {ACPI_FUJITSU_HID, 0},
	1154	+static const struct acpi_device_id fujitsu_bl_device_ids[] = {
	1155	+ {ACPI_FUJITSU_BL_HID, 0},
1155	1156	{"", 0},
1156	1157	};
1157	1158
1158		−static struct acpi_driver acpi_fujitsu_driver = {
1159		− .name = ACPI_FUJITSU_DRIVER_NAME,
	1159	+static struct acpi_driver acpi_fujitsu_bl_driver = {
	1160	+ .name = ACPI_FUJITSU_BL_DRIVER_NAME,
1160	1161	.class = ACPI_FUJITSU_CLASS,
1161		− .ids = fujitsu_device_ids,
	1162	+ .ids = fujitsu_bl_device_ids,
1162	1163	.ops = {
1163		− .add = acpi_fujitsu_add,
1164		− .remove = acpi_fujitsu_remove,
1165		− .notify = acpi_fujitsu_notify,
	1164	+ .add = acpi_fujitsu_bl_add,
	1165	+ .remove = acpi_fujitsu_bl_remove,
	1166	+ .notify = acpi_fujitsu_bl_notify,
1166	1167	},
1167	1168	};
1168	1169
1169		−static const struct acpi_device_id fujitsu_hotkey_device_ids[] = {
1170		− {ACPI_FUJITSU_HOTKEY_HID, 0},
	1170	+static const struct acpi_device_id fujitsu_laptop_device_ids[] = {
	1171	+ {ACPI_FUJITSU_LAPTOP_HID, 0},
1171	1172	{"", 0},
1172	1173	};
1173	1174
1174		−static struct acpi_driver acpi_fujitsu_hotkey_driver = {
1175		− .name = ACPI_FUJITSU_HOTKEY_DRIVER_NAME,
	1175	+static struct acpi_driver acpi_fujitsu_laptop_driver = {
	1176	+ .name = ACPI_FUJITSU_LAPTOP_DRIVER_NAME,
1176	1177	.class = ACPI_FUJITSU_CLASS,
1177		− .ids = fujitsu_hotkey_device_ids,
	1178	+ .ids = fujitsu_laptop_device_ids,
1178	1179	.ops = {
1179		− .add = acpi_fujitsu_hotkey_add,
1180		− .remove = acpi_fujitsu_hotkey_remove,
1181		− .notify = acpi_fujitsu_hotkey_notify,
	1180	+ .add = acpi_fujitsu_laptop_add,
	1181	+ .remove = acpi_fujitsu_laptop_remove,
	1182	+ .notify = acpi_fujitsu_laptop_notify,
1182	1183	},
1183	1184	};
1184	1185
1185	1186	static const struct acpi_device_id fujitsu_ids[] __used = {
1186		− {ACPI_FUJITSU_HID, 0},
1187		− {ACPI_FUJITSU_HOTKEY_HID, 0},
	1187	+ {ACPI_FUJITSU_BL_HID, 0},
	1188	+ {ACPI_FUJITSU_LAPTOP_HID, 0},
1188	1189	{"", 0}
1189	1190	};
1190	1191	MODULE_DEVICE_TABLE(acpi, fujitsu_ids);
1191	1192
1192	1193	static int __init fujitsu_init(void)
1193	1194	{
1194		− int ret, result, max_brightness;
	1195	+ int ret, max_brightness;
1195	1196
1196	1197	if (acpi_disabled)
1197	1198	return -ENODEV;
1198	1199
1199		− fujitsu = kzalloc(sizeof(struct fujitsu_t), GFP_KERNEL);
1200		− if (!fujitsu)
	1200	+ fujitsu_bl = kzalloc(sizeof(struct fujitsu_bl), GFP_KERNEL);
	1201	+ if (!fujitsu_bl)
1201	1202	return -ENOMEM;
1202		− fujitsu->keycode1 = KEY_PROG1;
1203		− fujitsu->keycode2 = KEY_PROG2;
1204		− fujitsu->keycode3 = KEY_PROG3;
1205		− fujitsu->keycode4 = KEY_PROG4;
1206		− fujitsu->keycode5 = KEY_RFKILL;
	1203	+ fujitsu_bl->keycode1 = KEY_PROG1;
	1204	+ fujitsu_bl->keycode2 = KEY_PROG2;
	1205	+ fujitsu_bl->keycode3 = KEY_PROG3;
	1206	+ fujitsu_bl->keycode4 = KEY_PROG4;
	1207	+ fujitsu_bl->keycode5 = KEY_RFKILL;
1207	1208	dmi_check_system(fujitsu_dmi_table);
1208	1209
1209		− result = acpi_bus_register_driver(&acpi_fujitsu_driver);
1210		− if (result < 0) {
1211		− ret = -ENODEV;
	1210	+ ret = acpi_bus_register_driver(&acpi_fujitsu_bl_driver);
	1211	+ if (ret)
1212	1212	goto fail_acpi;
1213		− }
1214	1213
1215	1214	/* Register platform stuff */
1216	1215
1217		− fujitsu->pf_device = platform_device_alloc("fujitsu-laptop", -1);
1218		− if (!fujitsu->pf_device) {
	1216	+ fujitsu_bl->pf_device = platform_device_alloc("fujitsu-laptop", -1);
	1217	+ if (!fujitsu_bl->pf_device) {
1219	1218	ret = -ENOMEM;
1220	1219	goto fail_platform_driver;
1221	1220	}
1222	1221
1223		− ret = platform_device_add(fujitsu->pf_device);
	1222	+ ret = platform_device_add(fujitsu_bl->pf_device);
1224	1223	if (ret)
1225	1224	goto fail_platform_device1;
1226	1225
1227	1226	ret =
1228		− sysfs_create_group(&fujitsu->pf_device->dev.kobj,
1229		− &fujitsupf_attribute_group);
	1227	+ sysfs_create_group(&fujitsu_bl->pf_device->dev.kobj,
	1228	+ &fujitsu_pf_attribute_group);
1230	1229	if (ret)
1231	1230	goto fail_platform_device2;
1232	1231
@@ -1236,90 +1235,88 @@ static int __init fujitsu_init(void) · 3 unmodified lines
1236	1235	struct backlight_properties props;
1237	1236
1238	1237	memset(&props, 0, sizeof(struct backlight_properties));
1239		− max_brightness = fujitsu->max_brightness;
	1238	+ max_brightness = fujitsu_bl->max_brightness;
1240	1239	props.type = BACKLIGHT_PLATFORM;
1241	1240	props.max_brightness = max_brightness - 1;
1242		− fujitsu->bl_device = backlight_device_register("fujitsu-laptop",
1243		− NULL, NULL,
1244		− &fujitsubl_ops,
1245		− &props);
1246		− if (IS_ERR(fujitsu->bl_device)) {
1247		− ret = PTR_ERR(fujitsu->bl_device);
1248		− fujitsu->bl_device = NULL;
	1241	+ fujitsu_bl->bl_device = backlight_device_register("fujitsu-laptop",
	1242	+ NULL, NULL,
	1243	+ &fujitsu_bl_ops,
	1244	+ &props);
	1245	+ if (IS_ERR(fujitsu_bl->bl_device)) {
	1246	+ ret = PTR_ERR(fujitsu_bl->bl_device);
	1247	+ fujitsu_bl->bl_device = NULL;
1249	1248	goto fail_sysfs_group;
1250	1249	}
1251		− fujitsu->bl_device->props.brightness = fujitsu->brightness_level;
	1250	+ fujitsu_bl->bl_device->props.brightness = fujitsu_bl->brightness_level;
1252	1251	}
1253	1252
1254		− ret = platform_driver_register(&fujitsupf_driver);
	1253	+ ret = platform_driver_register(&fujitsu_pf_driver);
1255	1254	if (ret)
1256	1255	goto fail_backlight;
1257	1256
1258		− /* Register hotkey driver */
	1257	+ /* Register laptop driver */
1259	1258
1260		− fujitsu_hotkey = kzalloc(sizeof(struct fujitsu_hotkey_t), GFP_KERNEL);
1261		− if (!fujitsu_hotkey) {
	1259	+ fujitsu_laptop = kzalloc(sizeof(struct fujitsu_laptop), GFP_KERNEL);
	1260	+ if (!fujitsu_laptop) {
1262	1261	ret = -ENOMEM;
1263		− goto fail_hotkey;
	1262	+ goto fail_laptop;
1264	1263	}
1265	1264
1266		− result = acpi_bus_register_driver(&acpi_fujitsu_hotkey_driver);
1267		− if (result < 0) {
1268		− ret = -ENODEV;
1269		− goto fail_hotkey1;
1270		− }
	1265	+ ret = acpi_bus_register_driver(&acpi_fujitsu_laptop_driver);
	1266	+ if (ret)
	1267	+ goto fail_laptop1;
1271	1268
1272	1269	/* Sync backlight power status (needs FUJ02E3 device, hence deferred) */
1273	1270	if (acpi_video_get_backlight_type() == acpi_backlight_vendor) {
1274	1271	if (call_fext_func(FUNC_BACKLIGHT, 0x2, 0x4, 0x0) == 3)
1275		− fujitsu->bl_device->props.power = FB_BLANK_POWERDOWN;
	1272	+ fujitsu_bl->bl_device->props.power = FB_BLANK_POWERDOWN;
1276	1273	else
1277		− fujitsu->bl_device->props.power = FB_BLANK_UNBLANK;
	1274	+ fujitsu_bl->bl_device->props.power = FB_BLANK_UNBLANK;
1278	1275	}
1279	1276
1280	1277	pr_info("driver " FUJITSU_DRIVER_VERSION " successfully loaded\n");
1281	1278
1282	1279	return 0;
1283	1280
1284		−fail_hotkey1:
1285		− kfree(fujitsu_hotkey);
1286		−fail_hotkey:
1287		− platform_driver_unregister(&fujitsupf_driver);
	1281	+fail_laptop1:
	1282	+ kfree(fujitsu_laptop);
	1283	+fail_laptop:
	1284	+ platform_driver_unregister(&fujitsu_pf_driver);
1288	1285	fail_backlight:
1289		− backlight_device_unregister(fujitsu->bl_device);
	1286	+ backlight_device_unregister(fujitsu_bl->bl_device);
1290	1287	fail_sysfs_group:
1291		− sysfs_remove_group(&fujitsu->pf_device->dev.kobj,
1292		− &fujitsupf_attribute_group);
	1288	+ sysfs_remove_group(&fujitsu_bl->pf_device->dev.kobj,
	1289	+ &fujitsu_pf_attribute_group);
1293	1290	fail_platform_device2:
1294		− platform_device_del(fujitsu->pf_device);
	1291	+ platform_device_del(fujitsu_bl->pf_device);
1295	1292	fail_platform_device1:
1296		− platform_device_put(fujitsu->pf_device);
	1293	+ platform_device_put(fujitsu_bl->pf_device);
1297	1294	fail_platform_driver:
1298		− acpi_bus_unregister_driver(&acpi_fujitsu_driver);
	1295	+ acpi_bus_unregister_driver(&acpi_fujitsu_bl_driver);
1299	1296	fail_acpi:
1300		− kfree(fujitsu);
	1297	+ kfree(fujitsu_bl);
1301	1298
1302	1299	return ret;
1303	1300	}
1304	1301
1305	1302	static void __exit fujitsu_cleanup(void)
1306	1303	{
1307		− acpi_bus_unregister_driver(&acpi_fujitsu_hotkey_driver);
	1304	+ acpi_bus_unregister_driver(&acpi_fujitsu_laptop_driver);
1308	1305
1309		− kfree(fujitsu_hotkey);
	1306	+ kfree(fujitsu_laptop);
1310	1307
1311		− platform_driver_unregister(&fujitsupf_driver);
	1308	+ platform_driver_unregister(&fujitsu_pf_driver);
1312	1309
1313		− backlight_device_unregister(fujitsu->bl_device);
	1310	+ backlight_device_unregister(fujitsu_bl->bl_device);
1314	1311
1315		− sysfs_remove_group(&fujitsu->pf_device->dev.kobj,
1316		− &fujitsupf_attribute_group);
	1312	+ sysfs_remove_group(&fujitsu_bl->pf_device->dev.kobj,
	1313	+ &fujitsu_pf_attribute_group);
1317	1314
1318		− platform_device_unregister(fujitsu->pf_device);
	1315	+ platform_device_unregister(fujitsu_bl->pf_device);
1319	1316
1320		− acpi_bus_unregister_driver(&acpi_fujitsu_driver);
	1317	+ acpi_bus_unregister_driver(&acpi_fujitsu_bl_driver);
1321	1318
1322		− kfree(fujitsu);
	1319	+ kfree(fujitsu_bl);
1323	1320
1324	1321	pr_info("driver unloaded\n");
1325	1322	}
@@ -1341,7 +1338,3 @@ MODULE_AUTHOR("Jonathan Woithe, Peter Gruber, Tony Vroon"); · 15 unmodified lines
1341	1338	MODULE_DESCRIPTION("Fujitsu laptop extras support");
1342	1339	MODULE_VERSION(FUJITSU_DRIVER_VERSION);
1343	1340	MODULE_LICENSE("GPL");
1344		−
1345		−MODULE_ALIAS("dmi::svnFUJITSUSIEMENS::pvr:rvnFUJITSU:rnFJNB1D3::cvrS6410:");
1346		−MODULE_ALIAS("dmi::svnFUJITSUSIEMENS::pvr:rvnFUJITSU:rnFJNB1E6::cvrS6420:");
1347		−MODULE_ALIAS("dmi::svnFUJITSU::pvr:rvnFUJITSU:rnFJNB19C::cvrS7020:");

drivers/scsi/Kconfig+16−3

@@ -1241,19 +1241,32 @@ config SCSI_LPFC · 1240 unmodified lines
1241	1241	tristate "Emulex LightPulse Fibre Channel Support"
1242	1242	depends on PCI && SCSI
1243	1243	depends on SCSI_FC_ATTRS
1244		− depends on NVME_FC && NVME_TARGET_FC
1245	1244	select CRC_T10DIF
1246		− help
	1245	+ ---help---
1247	1246	This lpfc driver supports the Emulex LightPulse
1248	1247	Family of Fibre Channel PCI host adapters.
1249	1248
1250	1249	config SCSI_LPFC_DEBUG_FS
1251	1250	bool "Emulex LightPulse Fibre Channel debugfs Support"
1252	1251	depends on SCSI_LPFC && DEBUG_FS
1253		− help
	1252	+ ---help---
1254	1253	This makes debugging information from the lpfc driver
1255	1254	available via the debugfs filesystem.
1256	1255
	1256	+config LPFC_NVME_INITIATOR
	1257	+ bool "Emulex LightPulse Fibre Channel NVME Initiator Support"
	1258	+ depends on SCSI_LPFC && NVME_FC
	1259	+ ---help---
	1260	+ This enables NVME Initiator support in the Emulex lpfc driver.
	1261	+
	1262	+config LPFC_NVME_TARGET
	1263	+ bool "Emulex LightPulse Fibre Channel NVME Initiator Support"
	1264	+ depends on SCSI_LPFC && NVME_TARGET_FC
	1265	+ ---help---
	1266	+ This enables NVME Target support in the Emulex lpfc driver.
	1267	+ Target enablement must still be enabled on a per adapter
	1268	+ basis by module parameters.
	1269	+
1257	1270	config SCSI_SIM710
1258	1271	tristate "Simple 53c710 SCSI support (Compaq, NCR machines)"
1259	1272	depends on (EISA \|\| MCA) && SCSI

drivers/scsi/aacraid/src.c+1−1

@@ -468,7 +468,7 @@ err_out: · 467 unmodified lines
468	468	return -1;
469	469
470	470	err_blink:
471		− return (status > 16) & 0xFF;
	471	+ return (status >> 16) & 0xFF;
472	472	}
473	473
474	474	static inline u32 aac_get_vector(struct aac_dev *dev)

drivers/scsi/aic7xxx/aic79xx_core.c+1−1

@@ -6278,7 +6278,7 @@ ahd_reset(struct ahd_softc *ahd, int reinit) · 6277 unmodified lines
6278	6278	* does not disable its parity logic prior to
6279	6279	* the start of the reset. This may cause a
6280	6280	* parity error to be detected and thus a
6281		− * spurious SERR or PERR assertion. Disble
	6281	+ * spurious SERR or PERR assertion. Disable
6282	6282	* PERR and SERR responses during the CHIPRST.
6283	6283	*/
6284	6284	mod_cmd = cmd & ~(PCIM_CMD_PERRESPEN\|PCIM_CMD_SERRESPEN);

drivers/scsi/libiscsi.c+25−1

@@ -561,8 +561,12 @@ static void iscsi_complete_task(struct iscsi_task *task, int state) · 560 unmodified lines
561	561	WARN_ON_ONCE(task->state == ISCSI_TASK_FREE);
562	562	task->state = state;
563	563
564		− if (!list_empty(&task->running))
	564	+ spin_lock_bh(&conn->taskqueuelock);
	565	+ if (!list_empty(&task->running)) {
	566	+ pr_debug_once("%s while task on list", __func__);
565	567	list_del_init(&task->running);
	568	+ }
	569	+ spin_unlock_bh(&conn->taskqueuelock);
566	570
567	571	if (conn->task == task)
568	572	conn->task = NULL;
@@ -784,7 +788,9 @@ __iscsi_conn_send_pdu(struct iscsi_conn conn, struct iscsi_hdr hdr, · 215 unmodified lines
784	788	if (session->tt->xmit_task(task))
785	789	goto free_task;
786	790	} else {
	791	+ spin_lock_bh(&conn->taskqueuelock);
787	792	list_add_tail(&task->running, &conn->mgmtqueue);
	793	+ spin_unlock_bh(&conn->taskqueuelock);
788	794	iscsi_conn_queue_work(conn);
789	795	}
790	796
@@ -1475,8 +1481,10 @@ void iscsi_requeue_task(struct iscsi_task *task) · 684 unmodified lines
1475	1481	* this may be on the requeue list already if the xmit_task callout
1476	1482	* is handling the r2ts while we are adding new ones
1477	1483	*/
	1484	+ spin_lock_bh(&conn->taskqueuelock);
1478	1485	if (list_empty(&task->running))
1479	1486	list_add_tail(&task->running, &conn->requeue);
	1487	+ spin_unlock_bh(&conn->taskqueuelock);
1480	1488	iscsi_conn_queue_work(conn);
1481	1489	}
1482	1490	EXPORT_SYMBOL_GPL(iscsi_requeue_task);
@@ -1513,22 +1521,26 @@ static int iscsi_data_xmit(struct iscsi_conn *conn) · 30 unmodified lines
1513	1521	* only have one nop-out as a ping from us and targets should not
1514	1522	* overflow us with nop-ins
1515	1523	*/
	1524	+ spin_lock_bh(&conn->taskqueuelock);
1516	1525	check_mgmt:
1517	1526	while (!list_empty(&conn->mgmtqueue)) {
1518	1527	conn->task = list_entry(conn->mgmtqueue.next,
1519	1528	struct iscsi_task, running);
1520	1529	list_del_init(&conn->task->running);
	1530	+ spin_unlock_bh(&conn->taskqueuelock);
1521	1531	if (iscsi_prep_mgmt_task(conn, conn->task)) {
1522	1532	/* regular RX path uses back_lock */
1523	1533	spin_lock_bh(&conn->session->back_lock);
1524	1534	__iscsi_put_task(conn->task);
1525	1535	spin_unlock_bh(&conn->session->back_lock);
1526	1536	conn->task = NULL;
	1537	+ spin_lock_bh(&conn->taskqueuelock);
1527	1538	continue;
1528	1539	}
1529	1540	rc = iscsi_xmit_task(conn);
1530	1541	if (rc)
1531	1542	goto done;
	1543	+ spin_lock_bh(&conn->taskqueuelock);
1532	1544	}
1533	1545
1534	1546	/* process pending command queue */
@@ -1536,19 +1548,24 @@ check_mgmt: · 1 unmodified line
1536	1548	conn->task = list_entry(conn->cmdqueue.next, struct iscsi_task,
1537	1549	running);
1538	1550	list_del_init(&conn->task->running);
	1551	+ spin_unlock_bh(&conn->taskqueuelock);
1539	1552	if (conn->session->state == ISCSI_STATE_LOGGING_OUT) {
1540	1553	fail_scsi_task(conn->task, DID_IMM_RETRY);
	1554	+ spin_lock_bh(&conn->taskqueuelock);
1541	1555	continue;
1542	1556	}
1543	1557	rc = iscsi_prep_scsi_cmd_pdu(conn->task);
1544	1558	if (rc) {
1545	1559	if (rc == -ENOMEM \|\| rc == -EACCES) {
	1560	+ spin_lock_bh(&conn->taskqueuelock);
1546	1561	list_add_tail(&conn->task->running,
1547	1562	&conn->cmdqueue);
1548	1563	conn->task = NULL;
	1564	+ spin_unlock_bh(&conn->taskqueuelock);
1549	1565	goto done;
1550	1566	} else
1551	1567	fail_scsi_task(conn->task, DID_ABORT);
	1568	+ spin_lock_bh(&conn->taskqueuelock);
1552	1569	continue;
1553	1570	}
1554	1571	rc = iscsi_xmit_task(conn);
@@ -1559,6 +1576,7 @@ check_mgmt: · 4 unmodified lines
1559	1576	* we need to check the mgmt queue for nops that need to
1560	1577	* be sent to aviod starvation
1561	1578	*/
	1579	+ spin_lock_bh(&conn->taskqueuelock);
1562	1580	if (!list_empty(&conn->mgmtqueue))
1563	1581	goto check_mgmt;
1564	1582	}
@@ -1578,12 +1596,15 @@ check_mgmt: · 13 unmodified lines
1578	1596	conn->task = task;
1579	1597	list_del_init(&conn->task->running);
1580	1598	conn->task->state = ISCSI_TASK_RUNNING;
	1599	+ spin_unlock_bh(&conn->taskqueuelock);
1581	1600	rc = iscsi_xmit_task(conn);
1582	1601	if (rc)
1583	1602	goto done;
	1603	+ spin_lock_bh(&conn->taskqueuelock);
1584	1604	if (!list_empty(&conn->mgmtqueue))
1585	1605	goto check_mgmt;
1586	1606	}
	1607	+ spin_unlock_bh(&conn->taskqueuelock);
1587	1608	spin_unlock_bh(&conn->session->frwd_lock);
1588	1609	return -ENODATA;
1589	1610
@@ -1739,7 +1760,9 @@ int iscsi_queuecommand(struct Scsi_Host host, struct scsi_cmnd sc) · 149 unmodified lines
1739	1760	goto prepd_reject;
1740	1761	}
1741	1762	} else {
	1763	+ spin_lock_bh(&conn->taskqueuelock);
1742	1764	list_add_tail(&task->running, &conn->cmdqueue);
	1765	+ spin_unlock_bh(&conn->taskqueuelock);
1743	1766	iscsi_conn_queue_work(conn);
1744	1767	}
1745	1768
@@ -2897,6 +2920,7 @@ iscsi_conn_setup(struct iscsi_cls_session *cls_session, int dd_size, · 1151 unmodified lines
2897	2920	INIT_LIST_HEAD(&conn->mgmtqueue);
2898	2921	INIT_LIST_HEAD(&conn->cmdqueue);
2899	2922	INIT_LIST_HEAD(&conn->requeue);
	2923	+ spin_lock_init(&conn->taskqueuelock);
2900	2924	INIT_WORK(&conn->xmitwork, iscsi_xmitworker);
2901	2925
2902	2926	/* allocate login_task used for the login/text sequences */

drivers/scsi/lpfc/lpfc.h+3−1

@@ -99,12 +99,13 @@ struct lpfc_sli2_slim; · 98 unmodified lines
99	99	#define FC_MAX_ADPTMSG 64
100	100
101	101	#define MAX_HBAEVT 32
	102	+#define MAX_HBAS_NO_RESET 16
102	103
103	104	/* Number of MSI-X vectors the driver uses */
104	105	#define LPFC_MSIX_VECTORS 2
105	106
106	107	/* lpfc wait event data ready flag */
107		−#define LPFC_DATA_READY (1<<0)
	108	+#define LPFC_DATA_READY 0 /* bit 0 */
108	109
109	110	/* queue dump line buffer size */
110	111	#define LPFC_LBUF_SZ 128
@@ -692,6 +693,7 @@ struct lpfc_hba { · 581 unmodified lines
692	693	* capability
693	694	*/
694	695	#define HBA_NVME_IOQ_FLUSH 0x80000 /* NVME IO queues flushed. */
	696	+#define NVME_XRI_ABORT_EVENT 0x100000
695	697
696	698	uint32_t fcp_ring_in_use; /* When polling test if intr-hndlr active*/
697	699	struct lpfc_dmabuf slim2p;

drivers/scsi/lpfc/lpfc_attr.c+8−1

@@ -3010,6 +3010,12 @@ MODULE_PARM_DESC(lpfc_poll, "FCP ring polling mode control:" · 3009 unmodified lines
3010	3010	static DEVICE_ATTR(lpfc_poll, S_IRUGO \| S_IWUSR,
3011	3011	lpfc_poll_show, lpfc_poll_store);
3012	3012
	3013	+int lpfc_no_hba_reset_cnt;
	3014	+unsigned long lpfc_no_hba_reset[MAX_HBAS_NO_RESET] = {
	3015	+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
	3016	+module_param_array(lpfc_no_hba_reset, ulong, &lpfc_no_hba_reset_cnt, 0444);
	3017	+MODULE_PARM_DESC(lpfc_no_hba_reset, "WWPN of HBAs that should not be reset");
	3018	+
3013	3019	LPFC_ATTR(sli_mode, 0, 0, 3,
3014	3020	"SLI mode selector:"
3015	3021	" 0 - auto (SLI-3 if supported),"
@@ -4451,7 +4457,8 @@ lpfc_fcp_imax_store(struct device dev, struct device_attribute attr, · 1435 unmodified lines
4451	4457	return -EINVAL;
4452	4458
4453	4459	phba->cfg_fcp_imax = (uint32_t)val;
4454		− for (i = 0; i < phba->io_channel_irqs; i++)
	4460	+
	4461	+ for (i = 0; i < phba->io_channel_irqs; i += LPFC_MAX_EQ_DELAY_EQID_CNT)
4455	4462	lpfc_modify_hba_eq_delay(phba, i);
4456	4463
4457	4464	return strlen(buf);

drivers/scsi/lpfc/lpfc_crtn.h+3−1

@@ -384,7 +384,7 @@ void lpfc_free_sysfs_attr(struct lpfc_vport *); · 383 unmodified lines
384	384	extern struct device_attribute *lpfc_hba_attrs[];
385	385	extern struct device_attribute *lpfc_vport_attrs[];
386	386	extern struct scsi_host_template lpfc_template;
387		−extern struct scsi_host_template lpfc_template_s3;
	387	+extern struct scsi_host_template lpfc_template_no_hr;
388	388	extern struct scsi_host_template lpfc_template_nvme;
389	389	extern struct scsi_host_template lpfc_vport_template;
390	390	extern struct fc_function_template lpfc_transport_functions;
@@ -554,3 +554,5 @@ void lpfc_nvme_abort_fcreq_cmpl(struct lpfc_hba *phba, · 163 unmodified lines
554	554	struct lpfc_wcqe_complete *abts_cmpl);
555	555	extern int lpfc_enable_nvmet_cnt;
556	556	extern unsigned long long lpfc_enable_nvmet[];
	557	+extern int lpfc_no_hba_reset_cnt;
	558	+extern unsigned long lpfc_no_hba_reset[];

drivers/scsi/lpfc/lpfc_ct.c+1−1

@@ -939,8 +939,8 @@ lpfc_cmpl_ct_cmd_gft_id(struct lpfc_hba phba, struct lpfc_iocbq cmdiocb, · 938 unmodified lines
939	939	"FC4 x%08x, Data: x%08x x%08x\n",
940	940	ndlp, did, ndlp->nlp_fc4_type,
941	941	FC_TYPE_FCP, FC_TYPE_NVME);
	942	+ ndlp->nlp_prev_state = NLP_STE_REG_LOGIN_ISSUE;
942	943	}
943		− ndlp->nlp_prev_state = NLP_STE_REG_LOGIN_ISSUE;
944	944	lpfc_nlp_set_state(vport, ndlp, NLP_STE_PRLI_ISSUE);
945	945	lpfc_issue_els_prli(vport, ndlp, 0);
946	946	} else

drivers/scsi/lpfc/lpfc_debugfs.c+0−22

@@ -3653,17 +3653,6 @@ lpfc_idiag_queacc_write(struct file file, const char __user buf, · 3652 unmodified lines
3653	3653	idiag.ptr_private = phba->sli4_hba.nvmels_cq;
3654	3654	goto pass_check;
3655	3655	}
3656		− /* NVME LS complete queue */
3657		− if (phba->sli4_hba.nvmels_cq &&
3658		− phba->sli4_hba.nvmels_cq->queue_id == queid) {
3659		− /* Sanity check */
3660		− rc = lpfc_idiag_que_param_check(
3661		− phba->sli4_hba.nvmels_cq, index, count);
3662		− if (rc)
3663		− goto error_out;
3664		− idiag.ptr_private = phba->sli4_hba.nvmels_cq;
3665		− goto pass_check;
3666		− }
3667	3656	/* FCP complete queue */
3668	3657	if (phba->sli4_hba.fcp_cq) {
3669	3658	for (qidx = 0; qidx < phba->cfg_fcp_io_channel;
@@ -3738,17 +3727,6 @@ lpfc_idiag_queacc_write(struct file file, const char __user buf, · 68 unmodified lines
3738	3727	idiag.ptr_private = phba->sli4_hba.nvmels_wq;
3739	3728	goto pass_check;
3740	3729	}
3741		− /* NVME LS work queue */
3742		− if (phba->sli4_hba.nvmels_wq &&
3743		− phba->sli4_hba.nvmels_wq->queue_id == queid) {
3744		− /* Sanity check */
3745		− rc = lpfc_idiag_que_param_check(
3746		− phba->sli4_hba.nvmels_wq, index, count);
3747		− if (rc)
3748		− goto error_out;
3749		− idiag.ptr_private = phba->sli4_hba.nvmels_wq;
3750		− goto pass_check;
3751		− }
3752	3730	/* FCP work queue */
3753	3731	if (phba->sli4_hba.fcp_wq) {
3754	3732	for (qidx = 0; qidx < phba->cfg_fcp_io_channel;

drivers/scsi/lpfc/lpfc_els.c+14−8

@@ -5177,15 +5177,15 @@ lpfc_rdp_res_speed(struct fc_rdp_port_speed_desc desc, struct lpfc_hba phba) · 5176 unmodified lines
5177	5177
5178	5178	static uint32_t
5179	5179	lpfc_rdp_res_diag_port_names(struct fc_rdp_port_name_desc *desc,
5180		− struct lpfc_hba *phba)
	5180	+ struct lpfc_vport *vport)
5181	5181	{
5182	5182
5183	5183	desc->tag = cpu_to_be32(RDP_PORT_NAMES_DESC_TAG);
5184	5184
5185		− memcpy(desc->port_names.wwnn, phba->wwnn,
	5185	+ memcpy(desc->port_names.wwnn, &vport->fc_nodename,
5186	5186	sizeof(desc->port_names.wwnn));
5187	5187
5188		− memcpy(desc->port_names.wwpn, phba->wwpn,
	5188	+ memcpy(desc->port_names.wwpn, &vport->fc_portname,
5189	5189	sizeof(desc->port_names.wwpn));
5190	5190
5191	5191	desc->length = cpu_to_be32(sizeof(desc->port_names));
@@ -5279,7 +5279,7 @@ lpfc_els_rdp_cmpl(struct lpfc_hba phba, struct lpfc_rdp_context rdp_context, · 87 unmodified lines
5279	5279	len += lpfc_rdp_res_link_error((struct fc_rdp_link_error_status_desc *)
5280	5280	(len + pcmd), &rdp_context->link_stat);
5281	5281	len += lpfc_rdp_res_diag_port_names((struct fc_rdp_port_name_desc *)
5282		− (len + pcmd), phba);
	5282	+ (len + pcmd), vport);
5283	5283	len += lpfc_rdp_res_attach_port_names((struct fc_rdp_port_name_desc *)
5284	5284	(len + pcmd), vport, ndlp);
5285	5285	len += lpfc_rdp_res_fec_desc((struct fc_fec_rdp_desc *)(len + pcmd),
@@ -8371,11 +8371,17 @@ lpfc_cmpl_reg_new_vport(struct lpfc_hba phba, LPFC_MBOXQ_t pmb) · 3085 unmodified lines
8371	8371	spin_lock_irq(shost->host_lock);
8372	8372	vport->fc_flag \|= FC_VPORT_NEEDS_REG_VPI;
8373	8373	spin_unlock_irq(shost->host_lock);
8374		− if (vport->port_type == LPFC_PHYSICAL_PORT
8375		− && !(vport->fc_flag & FC_LOGO_RCVD_DID_CHNG))
8376		− lpfc_issue_init_vfi(vport);
8377		− else
	8374	+ if (mb->mbxStatus == MBX_NOT_FINISHED)
	8375	+ break;
	8376	+ if ((vport->port_type == LPFC_PHYSICAL_PORT) &&
	8377	+ !(vport->fc_flag & FC_LOGO_RCVD_DID_CHNG)) {
	8378	+ if (phba->sli_rev == LPFC_SLI_REV4)
	8379	+ lpfc_issue_init_vfi(vport);
	8380	+ else
	8381	+ lpfc_initial_flogi(vport);
	8382	+ } else {
8378	8383	lpfc_initial_fdisc(vport);
	8384	+ }
8379	8385	break;
8380	8386	}
8381	8387	} else {

drivers/scsi/lpfc/lpfc_hbadisc.c+13−11

@@ -313,8 +313,7 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp) · 312 unmodified lines
313	313	ndlp->nlp_state, ndlp->nlp_rpi);
314	314	}
315	315
316		− if (!(vport->load_flag & FC_UNLOADING) &&
317		− !(ndlp->nlp_flag & NLP_DELAY_TMO) &&
	316	+ if (!(ndlp->nlp_flag & NLP_DELAY_TMO) &&
318	317	!(ndlp->nlp_flag & NLP_NPR_2B_DISC) &&
319	318	(ndlp->nlp_state != NLP_STE_UNMAPPED_NODE) &&
320	319	(ndlp->nlp_state != NLP_STE_REG_LOGIN_ISSUE) &&
@@ -641,6 +640,8 @@ lpfc_work_done(struct lpfc_hba *phba) · 320 unmodified lines
641	640	lpfc_handle_rrq_active(phba);
642	641	if (phba->hba_flag & FCP_XRI_ABORT_EVENT)
643	642	lpfc_sli4_fcp_xri_abort_event_proc(phba);
	643	+ if (phba->hba_flag & NVME_XRI_ABORT_EVENT)
	644	+ lpfc_sli4_nvme_xri_abort_event_proc(phba);
644	645	if (phba->hba_flag & ELS_XRI_ABORT_EVENT)
645	646	lpfc_sli4_els_xri_abort_event_proc(phba);
646	647	if (phba->hba_flag & ASYNC_EVENT)
@@ -2173,7 +2174,7 @@ lpfc_mbx_cmpl_fcf_scan_read_fcf_rec(struct lpfc_hba phba, LPFC_MBOXQ_t mboxq) · 1526 unmodified lines
2173	2174	uint32_t boot_flag, addr_mode;
2174	2175	uint16_t fcf_index, next_fcf_index;
2175	2176	struct lpfc_fcf_rec *fcf_rec = NULL;
2176		− uint16_t vlan_id;
	2177	+ uint16_t vlan_id = LPFC_FCOE_NULL_VID;
2177	2178	bool select_new_fcf;
2178	2179	int rc;
2179	2180
@@ -4020,9 +4021,11 @@ lpfc_register_remote_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp) · 1840 unmodified lines
4020	4021	rdata = rport->dd_data;
4021	4022	/* break the link before dropping the ref */
4022	4023	ndlp->rport = NULL;
4023		− if (rdata && rdata->pnode == ndlp)
4024		− lpfc_nlp_put(ndlp);
4025		− rdata->pnode = NULL;
	4024	+ if (rdata) {
	4025	+ if (rdata->pnode == ndlp)
	4026	+ lpfc_nlp_put(ndlp);
	4027	+ rdata->pnode = NULL;
	4028	+ }
4026	4029	/* drop reference for earlier registeration */
4027	4030	put_device(&rport->dev);
4028	4031	}
@@ -4344,9 +4347,8 @@ lpfc_initialize_node(struct lpfc_vport vport, struct lpfc_nodelist ndlp, · 315 unmodified lines
4344	4347	{
4345	4348	INIT_LIST_HEAD(&ndlp->els_retry_evt.evt_listp);
4346	4349	INIT_LIST_HEAD(&ndlp->dev_loss_evt.evt_listp);
4347		− init_timer(&ndlp->nlp_delayfunc);
4348		− ndlp->nlp_delayfunc.function = lpfc_els_retry_delay;
4349		− ndlp->nlp_delayfunc.data = (unsigned long)ndlp;
	4350	+ setup_timer(&ndlp->nlp_delayfunc, lpfc_els_retry_delay,
	4351	+ (unsigned long)ndlp);
4350	4352	ndlp->nlp_DID = did;
4351	4353	ndlp->vport = vport;
4352	4354	ndlp->phba = vport->phba;
@@ -4606,9 +4608,9 @@ lpfc_sli4_dequeue_nport_iocbs(struct lpfc_hba *phba, · 253 unmodified lines
4606	4608	pring = qp->pring;
4607	4609	if (!pring)
4608	4610	continue;
4609		− spin_lock_irq(&pring->ring_lock);
	4611	+ spin_lock(&pring->ring_lock);
4610	4612	__lpfc_dequeue_nport_iocbs(phba, ndlp, pring, dequeue_list);
4611		− spin_unlock_irq(&pring->ring_lock);
	4613	+ spin_unlock(&pring->ring_lock);
4612	4614	}
4613	4615	spin_unlock_irq(&phba->hbalock);
4614	4616	}

drivers/scsi/lpfc/lpfc_hw4.h+2−2

@@ -1001,7 +1001,7 @@ struct eq_delay_info { · 1000 unmodified lines
1001	1001	uint32_t phase;
1002	1002	uint32_t delay_multi;
1003	1003	};
1004		−#define LPFC_MAX_EQ_DELAY 8
	1004	+#define LPFC_MAX_EQ_DELAY_EQID_CNT 8
1005	1005
1006	1006	struct sgl_page_pairs {
1007	1007	uint32_t sgl_pg0_addr_lo;
@@ -1070,7 +1070,7 @@ struct lpfc_mbx_modify_eq_delay { · 62 unmodified lines
1070	1070	union {
1071	1071	struct {
1072	1072	uint32_t num_eq;
1073		− struct eq_delay_info eq[LPFC_MAX_EQ_DELAY];
	1073	+ struct eq_delay_info eq[LPFC_MAX_EQ_DELAY_EQID_CNT];
1074	1074	} request;
1075	1075	struct {
1076	1076	uint32_t word0;

drivers/scsi/lpfc/lpfc_init.c+85−43

@@ -3555,6 +3555,44 @@ out_free_mem: · 3554 unmodified lines
3555	3555	return rc;
3556	3556	}
3557	3557
	3558	+static uint64_t
	3559	+lpfc_get_wwpn(struct lpfc_hba *phba)
	3560	+{
	3561	+ uint64_t wwn;
	3562	+ int rc;
	3563	+ LPFC_MBOXQ_t *mboxq;
	3564	+ MAILBOX_t *mb;
	3565	+
	3566	+
	3567	+ mboxq = (LPFC_MBOXQ_t *) mempool_alloc(phba->mbox_mem_pool,
	3568	+ GFP_KERNEL);
	3569	+ if (!mboxq)
	3570	+ return (uint64_t)-1;
	3571	+
	3572	+ /* First get WWN of HBA instance */
	3573	+ lpfc_read_nv(phba, mboxq);
	3574	+ rc = lpfc_sli_issue_mbox(phba, mboxq, MBX_POLL);
	3575	+ if (rc != MBX_SUCCESS) {
	3576	+ lpfc_printf_log(phba, KERN_ERR, LOG_SLI,
	3577	+ "6019 Mailbox failed , mbxCmd x%x "
	3578	+ "READ_NV, mbxStatus x%x\n",
	3579	+ bf_get(lpfc_mqe_command, &mboxq->u.mqe),
	3580	+ bf_get(lpfc_mqe_status, &mboxq->u.mqe));
	3581	+ mempool_free(mboxq, phba->mbox_mem_pool);
	3582	+ return (uint64_t) -1;
	3583	+ }
	3584	+ mb = &mboxq->u.mb;
	3585	+ memcpy(&wwn, (char *)mb->un.varRDnvp.portname, sizeof(uint64_t));
	3586	+ /* wwn is WWPN of HBA instance */
	3587	+ mempool_free(mboxq, phba->mbox_mem_pool);
	3588	+ if (phba->sli_rev == LPFC_SLI_REV4)
	3589	+ return be64_to_cpu(wwn);
	3590	+ else
	3591	+ return (((wwn & 0xffffffff00000000) >> 32) \|
	3592	+ ((wwn & 0x00000000ffffffff) << 32));
	3593	+
	3594	+}
	3595	+
3558	3596	/**
3559	3597	* lpfc_sli4_nvme_sgl_update - update xri-sgl sizing and mapping
3560	3598	* @phba: pointer to lpfc hba data structure.
@@ -3676,17 +3714,32 @@ lpfc_create_port(struct lpfc_hba phba, int instance, struct device dev) · 115 unmodified lines
3676	3714	struct lpfc_vport *vport;
3677	3715	struct Scsi_Host *shost = NULL;
3678	3716	int error = 0;
	3717	+ int i;
	3718	+ uint64_t wwn;
	3719	+ bool use_no_reset_hba = false;
	3720	+
	3721	+ wwn = lpfc_get_wwpn(phba);
	3722	+
	3723	+ for (i = 0; i < lpfc_no_hba_reset_cnt; i++) {
	3724	+ if (wwn == lpfc_no_hba_reset[i]) {
	3725	+ lpfc_printf_log(phba, KERN_ERR, LOG_SLI,
	3726	+ "6020 Setting use_no_reset port=%llx\n",
	3727	+ wwn);
	3728	+ use_no_reset_hba = true;
	3729	+ break;
	3730	+ }
	3731	+ }
3679	3732
3680	3733	if (phba->cfg_enable_fc4_type & LPFC_ENABLE_FCP) {
3681	3734	if (dev != &phba->pcidev->dev) {
3682	3735	shost = scsi_host_alloc(&lpfc_vport_template,
3683	3736	sizeof(struct lpfc_vport));
3684	3737	} else {
3685		− if (phba->sli_rev == LPFC_SLI_REV4)
	3738	+ if (!use_no_reset_hba)
3686	3739	shost = scsi_host_alloc(&lpfc_template,
3687	3740	sizeof(struct lpfc_vport));
3688	3741	else
3689		− shost = scsi_host_alloc(&lpfc_template_s3,
	3742	+ shost = scsi_host_alloc(&lpfc_template_no_hr,
3690	3743	sizeof(struct lpfc_vport));
3691	3744	}
3692	3745	} else if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
@@ -3734,17 +3787,14 @@ lpfc_create_port(struct lpfc_hba phba, int instance, struct device dev) · 41 unmodified lines
3734	3787	INIT_LIST_HEAD(&vport->rcv_buffer_list);
3735	3788	spin_lock_init(&vport->work_port_lock);
3736	3789
3737		− init_timer(&vport->fc_disctmo);
3738		− vport->fc_disctmo.function = lpfc_disc_timeout;
3739		− vport->fc_disctmo.data = (unsigned long)vport;
	3790	+ setup_timer(&vport->fc_disctmo, lpfc_disc_timeout,
	3791	+ (unsigned long)vport);
3740	3792
3741		− init_timer(&vport->els_tmofunc);
3742		− vport->els_tmofunc.function = lpfc_els_timeout;
3743		− vport->els_tmofunc.data = (unsigned long)vport;
	3793	+ setup_timer(&vport->els_tmofunc, lpfc_els_timeout,
	3794	+ (unsigned long)vport);
3744	3795
3745		− init_timer(&vport->delayed_disc_tmo);
3746		− vport->delayed_disc_tmo.function = lpfc_delayed_disc_tmo;
3747		− vport->delayed_disc_tmo.data = (unsigned long)vport;
	3796	+ setup_timer(&vport->delayed_disc_tmo, lpfc_delayed_disc_tmo,
	3797	+ (unsigned long)vport);
3748	3798
3749	3799	error = scsi_add_host_with_dma(shost, dev, &phba->pcidev->dev);
3750	3800	if (error)
@@ -5406,21 +5456,15 @@ lpfc_setup_driver_resource_phase1(struct lpfc_hba *phba) · 1655 unmodified lines
5406	5456	INIT_LIST_HEAD(&phba->luns);
5407	5457
5408	5458	/* MBOX heartbeat timer */
5409		− init_timer(&psli->mbox_tmo);
5410		− psli->mbox_tmo.function = lpfc_mbox_timeout;
5411		− psli->mbox_tmo.data = (unsigned long) phba;
	5459	+ setup_timer(&psli->mbox_tmo, lpfc_mbox_timeout, (unsigned long)phba);
5412	5460	/* Fabric block timer */
5413		− init_timer(&phba->fabric_block_timer);
5414		− phba->fabric_block_timer.function = lpfc_fabric_block_timeout;
5415		− phba->fabric_block_timer.data = (unsigned long) phba;
	5461	+ setup_timer(&phba->fabric_block_timer, lpfc_fabric_block_timeout,
	5462	+ (unsigned long)phba);
5416	5463	/* EA polling mode timer */
5417		− init_timer(&phba->eratt_poll);
5418		− phba->eratt_poll.function = lpfc_poll_eratt;
5419		− phba->eratt_poll.data = (unsigned long) phba;
	5464	+ setup_timer(&phba->eratt_poll, lpfc_poll_eratt,
	5465	+ (unsigned long)phba);
5420	5466	/* Heartbeat timer */
5421		− init_timer(&phba->hb_tmofunc);
5422		− phba->hb_tmofunc.function = lpfc_hb_timeout;
5423		− phba->hb_tmofunc.data = (unsigned long)phba;
	5467	+ setup_timer(&phba->hb_tmofunc, lpfc_hb_timeout, (unsigned long)phba);
5424	5468
5425	5469	return 0;
5426	5470	}
@@ -5446,9 +5490,8 @@ lpfc_sli_driver_resource_setup(struct lpfc_hba *phba) · 19 unmodified lines
5446	5490	*/
5447	5491
5448	5492	/* FCP polling mode timer */
5449		− init_timer(&phba->fcp_poll_timer);
5450		− phba->fcp_poll_timer.function = lpfc_poll_timeout;
5451		− phba->fcp_poll_timer.data = (unsigned long) phba;
	5493	+ setup_timer(&phba->fcp_poll_timer, lpfc_poll_timeout,
	5494	+ (unsigned long)phba);
5452	5495
5453	5496	/* Host attention work mask setup */
5454	5497	phba->work_ha_mask = (HA_ERATT \| HA_MBATT \| HA_LATT);
@@ -5482,7 +5525,8 @@ lpfc_sli_driver_resource_setup(struct lpfc_hba *phba) · 27 unmodified lines
5482	5525
5483	5526	/* Initialize the host templates the configured values. */
5484	5527	lpfc_vport_template.sg_tablesize = phba->cfg_sg_seg_cnt;
5485		− lpfc_template_s3.sg_tablesize = phba->cfg_sg_seg_cnt;
	5528	+ lpfc_template_no_hr.sg_tablesize = phba->cfg_sg_seg_cnt;
	5529	+ lpfc_template.sg_tablesize = phba->cfg_sg_seg_cnt;
5486	5530
5487	5531	/* There are going to be 2 reserved BDEs: 1 FCP cmnd + 1 FCP rsp */
5488	5532	if (phba->cfg_enable_bg) {
@@ -5617,14 +5661,11 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba) · 128 unmodified lines
5617	5661	* Initialize timers used by driver
5618	5662	*/
5619	5663
5620		− init_timer(&phba->rrq_tmr);
5621		− phba->rrq_tmr.function = lpfc_rrq_timeout;
5622		− phba->rrq_tmr.data = (unsigned long)phba;
	5664	+ setup_timer(&phba->rrq_tmr, lpfc_rrq_timeout, (unsigned long)phba);
5623	5665
5624	5666	/* FCF rediscover timer */
5625		− init_timer(&phba->fcf.redisc_wait);
5626		− phba->fcf.redisc_wait.function = lpfc_sli4_fcf_redisc_wait_tmo;
5627		− phba->fcf.redisc_wait.data = (unsigned long)phba;
	5667	+ setup_timer(&phba->fcf.redisc_wait, lpfc_sli4_fcf_redisc_wait_tmo,
	5668	+ (unsigned long)phba);
5628	5669
5629	5670	/*
5630	5671	* Control structure for handling external multi-buffer mailbox
@@ -5706,6 +5747,7 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba) · 75 unmodified lines
5706	5747	/* Initialize the host templates with the updated values. */
5707	5748	lpfc_vport_template.sg_tablesize = phba->cfg_sg_seg_cnt;
5708	5749	lpfc_template.sg_tablesize = phba->cfg_sg_seg_cnt;
	5750	+ lpfc_template_no_hr.sg_tablesize = phba->cfg_sg_seg_cnt;
5709	5751
5710	5752	if (phba->cfg_sg_dma_buf_size <= LPFC_MIN_SG_SLI4_BUF_SZ)
5711	5753	phba->cfg_sg_dma_buf_size = LPFC_MIN_SG_SLI4_BUF_SZ;
@@ -5736,6 +5778,8 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba) · 24 unmodified lines
5736	5778	/* Initialize the Abort nvme buffer list used by driver */
5737	5779	spin_lock_init(&phba->sli4_hba.abts_nvme_buf_list_lock);
5738	5780	INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvme_buf_list);
	5781	+ /* Fast-path XRI aborted CQ Event work queue list */
	5782	+ INIT_LIST_HEAD(&phba->sli4_hba.sp_nvme_xri_aborted_work_queue);
5739	5783	}
5740	5784
5741	5785	/* This abort list used by worker thread */
@@ -8712,12 +8756,9 @@ lpfc_sli4_queue_setup(struct lpfc_hba *phba) · 2970 unmodified lines
8712	8756	}
8713	8757	}
8714	8758
8715		− /*
8716		− * Configure EQ delay multipier for interrupt coalescing using
8717		− * MODIFY_EQ_DELAY for all EQs created, LPFC_MAX_EQ_DELAY at a time.
8718		− */
8719		− for (qidx = 0; qidx < io_channel; qidx += LPFC_MAX_EQ_DELAY)
	8759	+ for (qidx = 0; qidx < io_channel; qidx += LPFC_MAX_EQ_DELAY_EQID_CNT)
8720	8760	lpfc_modify_hba_eq_delay(phba, qidx);
	8761	+
8721	8762	return 0;
8722	8763
8723	8764	out_destroy:
@@ -8973,6 +9014,11 @@ lpfc_sli4_cq_event_release_all(struct lpfc_hba *phba) · 249 unmodified lines
8973	9014	/* Pending ELS XRI abort events */
8974	9015	list_splice_init(&phba->sli4_hba.sp_els_xri_aborted_work_queue,
8975	9016	&cqelist);
	9017	+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
	9018	+ /* Pending NVME XRI abort events */
	9019	+ list_splice_init(&phba->sli4_hba.sp_nvme_xri_aborted_work_queue,
	9020	+ &cqelist);
	9021	+ }
8976	9022	/* Pending asynnc events */
8977	9023	list_splice_init(&phba->sli4_hba.sp_asynce_work_queue,
8978	9024	&cqelist);
@@ -10400,12 +10446,7 @@ lpfc_pci_remove_one_s3(struct pci_dev *pdev) · 1421 unmodified lines
10400	10446	fc_remove_host(shost);
10401	10447	scsi_remove_host(shost);
10402	10448
10403		− /* Perform ndlp cleanup on the physical port. The nvme and nvmet
10404		− * localports are destroyed after to cleanup all transport memory.
10405		− */
10406	10449	lpfc_cleanup(vport);
10407		− lpfc_nvmet_destroy_targetport(phba);
10408		− lpfc_nvme_destroy_localport(vport);
10409	10450
10410	10451	/*
10411	10452	* Bring down the SLI Layer. This step disable all interrupts,
@@ -12018,6 +12059,7 @@ static struct pci_driver lpfc_driver = { · 1606 unmodified lines
12018	12059	.id_table = lpfc_id_table,
12019	12060	.probe = lpfc_pci_probe_one,
12020	12061	.remove = lpfc_pci_remove_one,
	12062	+ .shutdown = lpfc_pci_remove_one,
12021	12063	.suspend = lpfc_pci_suspend_one,
12022	12064	.resume = lpfc_pci_resume_one,
12023	12065	.err_handler = &lpfc_err_handler,

drivers/scsi/lpfc/lpfc_mem.c+1−1

@@ -646,7 +646,6 @@ lpfc_sli4_nvmet_alloc(struct lpfc_hba *phba) · 645 unmodified lines
646	646	}
647	647
648	648	dma_buf->iocbq = lpfc_sli_get_iocbq(phba);
649		− dma_buf->iocbq->iocb_flag = LPFC_IO_NVMET;
650	649	if (!dma_buf->iocbq) {
651	650	kfree(dma_buf->context);
652	651	pci_pool_free(phba->lpfc_drb_pool, dma_buf->dbuf.virt,
@@ -658,6 +657,7 @@ lpfc_sli4_nvmet_alloc(struct lpfc_hba *phba) · 5 unmodified lines
658	657	"2621 Ran out of nvmet iocb/WQEs\n");
659	658	return NULL;
660	659	}
	660	+ dma_buf->iocbq->iocb_flag = LPFC_IO_NVMET;
661	661	nvmewqe = dma_buf->iocbq;
662	662	wqe = (union lpfc_wqe128 *)&nvmewqe->wqe;
663	663	/* Initialize WQE */

drivers/scsi/lpfc/lpfc_nvme.c+82−25

@@ -316,7 +316,7 @@ lpfc_nvme_gen_req(struct lpfc_vport vport, struct lpfc_dmabuf bmp, · 315 unmodified lines
316	316	bf_set(wqe_dfctl, &wqe->gen_req.wge_ctl, 0);
317	317	bf_set(wqe_si, &wqe->gen_req.wge_ctl, 1);
318	318	bf_set(wqe_la, &wqe->gen_req.wge_ctl, 1);
319		− bf_set(wqe_rctl, &wqe->gen_req.wge_ctl, FC_RCTL_DD_UNSOL_CTL);
	319	+ bf_set(wqe_rctl, &wqe->gen_req.wge_ctl, FC_RCTL_ELS4_REQ);
320	320	bf_set(wqe_type, &wqe->gen_req.wge_ctl, FC_TYPE_NVME);
321	321
322	322	/* Word 6 */
@@ -620,15 +620,15 @@ lpfc_nvme_adj_fcp_sgls(struct lpfc_vport *vport, · 297 unmodified lines
620	620	* Embed the payload in the last half of the WQE
621	621	* WQE words 16-30 get the NVME CMD IU payload
622	622	*
623		− * WQE Word 16 is already setup with flags
624		− * WQE words 17-19 get payload Words 2-4
	623	+ * WQE words 16-19 get payload Words 1-4
625	624	* WQE words 20-21 get payload Words 6-7
626	625	* WQE words 22-29 get payload Words 16-23
627	626	*/
628		− wptr = &wqe->words[17]; /* WQE ptr */
	627	+ wptr = &wqe->words[16]; /* WQE ptr */
629	628	dptr = (uint32_t )nCmd->cmdaddr; / payload ptr */
630		− dptr += 2; /* Skip Words 0-1 in payload */
	629	+ dptr++; /* Skip Word 0 in payload */
631	630
	631	+ wptr++ = dptr++; /* Word 1 */
632	632	wptr++ = dptr++; /* Word 2 */
633	633	wptr++ = dptr++; /* Word 3 */
634	634	wptr++ = dptr++; /* Word 4 */
@@ -978,9 +978,6 @@ lpfc_nvme_prep_io_cmd(struct lpfc_vport *vport, · 343 unmodified lines
978	978	bf_set(wqe_cmd_type, &wqe->generic.wqe_com,
979	979	NVME_WRITE_CMD);
980	980
981		− /* Word 16 */
982		− wqe->words[16] = LPFC_NVME_EMBED_WRITE;
983		−
984	981	phba->fc4NvmeOutputRequests++;
985	982	} else {
986	983	/* Word 7 */
@@ -1002,9 +999,6 @@ lpfc_nvme_prep_io_cmd(struct lpfc_vport *vport, · 15 unmodified lines
1002	999	bf_set(wqe_cmd_type, &wqe->generic.wqe_com,
1003	1000	NVME_READ_CMD);
1004	1001
1005		− /* Word 16 */
1006		− wqe->words[16] = LPFC_NVME_EMBED_READ;
1007		−
1008	1002	phba->fc4NvmeInputRequests++;
1009	1003	}
1010	1004	} else {
@@ -1026,9 +1020,6 @@ lpfc_nvme_prep_io_cmd(struct lpfc_vport *vport, · 15 unmodified lines
1026	1020	/* Word 11 */
1027	1021	bf_set(wqe_cmd_type, &wqe->generic.wqe_com, NVME_READ_CMD);
1028	1022
1029		− /* Word 16 */
1030		− wqe->words[16] = LPFC_NVME_EMBED_CMD;
1031		−
1032	1023	phba->fc4NvmeControlRequests++;
1033	1024	}
1034	1025	/*
@@ -1286,6 +1277,7 @@ lpfc_nvme_fcp_io_submit(struct nvme_fc_local_port *pnvme_lport, · 251 unmodified lines
1286	1277	pnvme_fcreq->private = (void *)lpfc_ncmd;
1287	1278	lpfc_ncmd->nvmeCmd = pnvme_fcreq;
1288	1279	lpfc_ncmd->nrport = rport;
	1280	+ lpfc_ncmd->ndlp = ndlp;
1289	1281	lpfc_ncmd->start_time = jiffies;
1290	1282
1291	1283	lpfc_nvme_prep_io_cmd(vport, lpfc_ncmd, ndlp);
@@ -1319,7 +1311,7 @@ lpfc_nvme_fcp_io_submit(struct nvme_fc_local_port *pnvme_lport, · 27 unmodified lines
1319	1311	"sid: x%x did: x%x oxid: x%x\n",
1320	1312	ret, vport->fc_myDID, ndlp->nlp_DID,
1321	1313	lpfc_ncmd->cur_iocbq.sli4_xritag);
1322		− ret = -EINVAL;
	1314	+ ret = -EBUSY;
1323	1315	goto out_free_nvme_buf;
1324	1316	}
1325	1317
@@ -1821,10 +1813,10 @@ lpfc_post_nvme_sgl_list(struct lpfc_hba *phba, · 495 unmodified lines
1821	1813	pdma_phys_sgl1, cur_xritag);
1822	1814	if (status) {
1823	1815	/* failure, put on abort nvme list */
1824		− lpfc_ncmd->exch_busy = 1;
	1816	+ lpfc_ncmd->flags \|= LPFC_SBUF_XBUSY;
1825	1817	} else {
1826	1818	/* success, put on NVME buffer list */
1827		− lpfc_ncmd->exch_busy = 0;
	1819	+ lpfc_ncmd->flags &= ~LPFC_SBUF_XBUSY;
1828	1820	lpfc_ncmd->status = IOSTAT_SUCCESS;
1829	1821	num_posted++;
1830	1822	}
@@ -1854,10 +1846,10 @@ lpfc_post_nvme_sgl_list(struct lpfc_hba *phba, · 23 unmodified lines
1854	1846	struct lpfc_nvme_buf, list);
1855	1847	if (status) {
1856	1848	/* failure, put on abort nvme list */
1857		− lpfc_ncmd->exch_busy = 1;
	1849	+ lpfc_ncmd->flags \|= LPFC_SBUF_XBUSY;
1858	1850	} else {
1859	1851	/* success, put on NVME buffer list */
1860		− lpfc_ncmd->exch_busy = 0;
	1852	+ lpfc_ncmd->flags &= ~LPFC_SBUF_XBUSY;
1861	1853	lpfc_ncmd->status = IOSTAT_SUCCESS;
1862	1854	num_posted++;
1863	1855	}
@@ -2099,7 +2091,7 @@ lpfc_release_nvme_buf(struct lpfc_hba phba, struct lpfc_nvme_buf lpfc_ncmd) · 235 unmodified lines
2099	2091	unsigned long iflag = 0;
2100	2092
2101	2093	lpfc_ncmd->nonsg_phys = 0;
2102		− if (lpfc_ncmd->exch_busy) {
	2094	+ if (lpfc_ncmd->flags & LPFC_SBUF_XBUSY) {
2103	2095	spin_lock_irqsave(&phba->sli4_hba.abts_nvme_buf_list_lock,
2104	2096	iflag);
2105	2097	lpfc_ncmd->nvmeCmd = NULL;
@@ -2135,11 +2127,12 @@ lpfc_release_nvme_buf(struct lpfc_hba phba, struct lpfc_nvme_buf lpfc_ncmd) · 29 unmodified lines
2135	2127	int
2136	2128	lpfc_nvme_create_localport(struct lpfc_vport *vport)
2137	2129	{
	2130	+ int ret = 0;
2138	2131	struct lpfc_hba *phba = vport->phba;
2139	2132	struct nvme_fc_port_info nfcp_info;
2140	2133	struct nvme_fc_local_port *localport;
2141	2134	struct lpfc_nvme_lport *lport;
2142		− int len, ret = 0;
	2135	+ int len;
2143	2136
2144	2137	/* Initialize this localport instance. The vport wwn usage ensures
2145	2138	* that NPIV is accounted for.
@@ -2156,8 +2149,12 @@ lpfc_nvme_create_localport(struct lpfc_vport *vport) · 10 unmodified lines
2156	2149	/* localport is allocated from the stack, but the registration
2157	2150	* call allocates heap memory as well as the private area.
2158	2151	*/
	2152	+#ifdef CONFIG_LPFC_NVME_INITIATOR
2159	2153	ret = nvme_fc_register_localport(&nfcp_info, &lpfc_nvme_template,
2160	2154	&vport->phba->pcidev->dev, &localport);
	2155	+#else
	2156	+ ret = -ENOMEM;
	2157	+#endif
2161	2158	if (!ret) {
2162	2159	lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME \| LOG_NVME_DISC,
2163	2160	"6005 Successfully registered local "
@@ -2173,10 +2170,10 @@ lpfc_nvme_create_localport(struct lpfc_vport *vport) · 9 unmodified lines
2173	2170	lport->vport = vport;
2174	2171	INIT_LIST_HEAD(&lport->rport_list);
2175	2172	vport->nvmei_support = 1;
	2173	+ len = lpfc_new_nvme_buf(vport, phba->sli4_hba.nvme_xri_max);
	2174	+ vport->phba->total_nvme_bufs += len;
2176	2175	}
2177	2176
2178		− len = lpfc_new_nvme_buf(vport, phba->sli4_hba.nvme_xri_max);
2179		− vport->phba->total_nvme_bufs += len;
2180	2177	return ret;
2181	2178	}
2182	2179
@@ -2193,6 +2190,7 @@ lpfc_nvme_create_localport(struct lpfc_vport *vport) · 10 unmodified lines
2193	2190	void
2194	2191	lpfc_nvme_destroy_localport(struct lpfc_vport *vport)
2195	2192	{
	2193	+#ifdef CONFIG_LPFC_NVME_INITIATOR
2196	2194	struct nvme_fc_local_port *localport;
2197	2195	struct lpfc_nvme_lport *lport;
2198	2196	struct lpfc_nvme_rport rport = NULL, rport_next = NULL;
@@ -2208,7 +2206,6 @@ lpfc_nvme_destroy_localport(struct lpfc_vport *vport) · 9 unmodified lines
2208	2206	lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME,
2209	2207	"6011 Destroying NVME localport %p\n",
2210	2208	localport);
2211		−
2212	2209	list_for_each_entry_safe(rport, rport_next, &lport->rport_list, list) {
2213	2210	/* The last node ref has to get released now before the rport
2214	2211	* private memory area is released by the transport.
@@ -2222,6 +2219,7 @@ lpfc_nvme_destroy_localport(struct lpfc_vport *vport) · 7 unmodified lines
2222	2219	"6008 rport fail destroy %x\n", ret);
2223	2220	wait_for_completion_timeout(&rport->rport_unreg_done, 5);
2224	2221	}
	2222	+
2225	2223	/* lport's rport list is clear. Unregister
2226	2224	* lport and release resources.
2227	2225	*/
@@ -2245,6 +2243,7 @@ lpfc_nvme_destroy_localport(struct lpfc_vport *vport) · 17 unmodified lines
2245	2243	"Failed, status x%x\n",
2246	2244	ret);
2247	2245	}
	2246	+#endif
2248	2247	}
2249	2248
2250	2249	void
@@ -2275,6 +2274,7 @@ lpfc_nvme_update_localport(struct lpfc_vport *vport) · 24 unmodified lines
2275	2274	int
2276	2275	lpfc_nvme_register_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp)
2277	2276	{
	2277	+#ifdef CONFIG_LPFC_NVME_INITIATOR
2278	2278	int ret = 0;
2279	2279	struct nvme_fc_local_port *localport;
2280	2280	struct lpfc_nvme_lport *lport;
@@ -2348,7 +2348,6 @@ lpfc_nvme_register_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp) · 67 unmodified lines
2348	2348	rpinfo.port_role \|= FC_PORT_ROLE_NVME_INITIATOR;
2349	2349	rpinfo.port_name = wwn_to_u64(ndlp->nlp_portname.u.wwn);
2350	2350	rpinfo.node_name = wwn_to_u64(ndlp->nlp_nodename.u.wwn);
2351		−
2352	2351	ret = nvme_fc_register_remoteport(localport, &rpinfo,
2353	2352	&remote_port);
2354	2353	if (!ret) {
@@ -2384,6 +2383,9 @@ lpfc_nvme_register_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp) · 29 unmodified lines
2384	2383	ndlp->nlp_type, ndlp->nlp_DID, ndlp);
2385	2384	}
2386	2385	return ret;
	2386	+#else
	2387	+ return 0;
	2388	+#endif
2387	2389	}
2388	2390
2389	2391	/* lpfc_nvme_unregister_port - unbind the DID and port_role from this rport.
@@ -2401,6 +2403,7 @@ lpfc_nvme_register_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp) · 11 unmodified lines
2401	2403	void
2402	2404	lpfc_nvme_unregister_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp)
2403	2405	{
	2406	+#ifdef CONFIG_LPFC_NVME_INITIATOR
2404	2407	int ret;
2405	2408	struct nvme_fc_local_port *localport;
2406	2409	struct lpfc_nvme_lport *lport;
@@ -2458,7 +2461,61 @@ lpfc_nvme_unregister_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp) · 51 unmodified lines
2458	2461	return;
2459	2462
2460	2463	input_err:
	2464	+#endif
2461	2465	lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
2462	2466	"6168: State error: lport %p, rport%p FCID x%06x\n",
2463	2467	vport->localport, ndlp->rport, ndlp->nlp_DID);
2464	2468	}
	2469	+
	2470	+/**
	2471	+ * lpfc_sli4_nvme_xri_aborted - Fast-path process of NVME xri abort
	2472	+ * @phba: pointer to lpfc hba data structure.
	2473	+ * @axri: pointer to the fcp xri abort wcqe structure.
	2474	+ *
	2475	+ * This routine is invoked by the worker thread to process a SLI4 fast-path
	2476	+ * FCP aborted xri.
	2477	+ **/
	2478	+void
	2479	+lpfc_sli4_nvme_xri_aborted(struct lpfc_hba *phba,
	2480	+ struct sli4_wcqe_xri_aborted *axri)
	2481	+{
	2482	+ uint16_t xri = bf_get(lpfc_wcqe_xa_xri, axri);
	2483	+ uint16_t rxid = bf_get(lpfc_wcqe_xa_remote_xid, axri);
	2484	+ struct lpfc_nvme_buf lpfc_ncmd, next_lpfc_ncmd;
	2485	+ struct lpfc_nodelist *ndlp;
	2486	+ unsigned long iflag = 0;
	2487	+ int rrq_empty = 0;
	2488	+
	2489	+ if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
	2490	+ return;
	2491	+ spin_lock_irqsave(&phba->hbalock, iflag);
	2492	+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
	2493	+ list_for_each_entry_safe(lpfc_ncmd, next_lpfc_ncmd,
	2494	+ &phba->sli4_hba.lpfc_abts_nvme_buf_list,
	2495	+ list) {
	2496	+ if (lpfc_ncmd->cur_iocbq.sli4_xritag == xri) {
	2497	+ list_del(&lpfc_ncmd->list);
	2498	+ lpfc_ncmd->flags &= ~LPFC_SBUF_XBUSY;
	2499	+ lpfc_ncmd->status = IOSTAT_SUCCESS;
	2500	+ spin_unlock(
	2501	+ &phba->sli4_hba.abts_nvme_buf_list_lock);
	2502	+
	2503	+ rrq_empty = list_empty(&phba->active_rrq_list);
	2504	+ spin_unlock_irqrestore(&phba->hbalock, iflag);
	2505	+ ndlp = lpfc_ncmd->ndlp;
	2506	+ if (ndlp) {
	2507	+ lpfc_set_rrq_active(
	2508	+ phba, ndlp,
	2509	+ lpfc_ncmd->cur_iocbq.sli4_lxritag,
	2510	+ rxid, 1);
	2511	+ lpfc_sli4_abts_err_handler(phba, ndlp, axri);
	2512	+ }
	2513	+ lpfc_release_nvme_buf(phba, lpfc_ncmd);
	2514	+ if (rrq_empty)
	2515	+ lpfc_worker_wake_up(phba);
	2516	+ return;
	2517	+ }
	2518	+ }
	2519	+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
	2520	+ spin_unlock_irqrestore(&phba->hbalock, iflag);
	2521	+}

drivers/scsi/lpfc/lpfc_nvme.h+1−0

@@ -57,6 +57,7 @@ struct lpfc_nvme_buf { · 56 unmodified lines
57	57	struct list_head list;
58	58	struct nvmefc_fcp_req *nvmeCmd;
59	59	struct lpfc_nvme_rport *nrport;
	60	+ struct lpfc_nodelist *ndlp;
60	61
61	62	uint32_t timeout;
62	63

drivers/scsi/lpfc/lpfc_nvmet.c+35−8

@@ -571,6 +571,7 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport, · 570 unmodified lines
571	571	lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
572	572	"6102 Bad state IO x%x aborted\n",
573	573	ctxp->oxid);
	574	+ rc = -ENXIO;
574	575	goto aerr;
575	576	}
576	577
@@ -580,6 +581,7 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport, · 3 unmodified lines
580	581	lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
581	582	"6152 FCP Drop IO x%x: Prep\n",
582	583	ctxp->oxid);
	584	+ rc = -ENXIO;
583	585	goto aerr;
584	586	}
585	587
@@ -618,8 +620,9 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport, · 32 unmodified lines
618	620	ctxp->wqeq->hba_wqidx = 0;
619	621	nvmewqeq->context2 = NULL;
620	622	nvmewqeq->context3 = NULL;
	623	+ rc = -EBUSY;
621	624	aerr:
622		− return -ENXIO;
	625	+ return rc;
623	626	}
624	627
625	628	static void
@@ -668,9 +671,13 @@ lpfc_nvmet_create_targetport(struct lpfc_hba *phba) · 42 unmodified lines
668	671	lpfc_tgttemplate.target_features = NVMET_FCTGTFEAT_READDATA_RSP \|
669	672	NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED;
670	673
	674	+#ifdef CONFIG_LPFC_NVME_TARGET
671	675	error = nvmet_fc_register_targetport(&pinfo, &lpfc_tgttemplate,
672	676	&phba->pcidev->dev,
673	677	&phba->targetport);
	678	+#else
	679	+ error = -ENOMEM;
	680	+#endif
674	681	if (error) {
675	682	lpfc_printf_log(phba, KERN_ERR, LOG_NVME_DISC,
676	683	"6025 Cannot register NVME targetport "
@@ -731,9 +738,25 @@ lpfc_nvmet_update_targetport(struct lpfc_hba *phba) · 54 unmodified lines
731	738	return 0;
732	739	}
733	740
	741	+/**
	742	+ * lpfc_sli4_nvmet_xri_aborted - Fast-path process of nvmet xri abort
	743	+ * @phba: pointer to lpfc hba data structure.
	744	+ * @axri: pointer to the nvmet xri abort wcqe structure.
	745	+ *
	746	+ * This routine is invoked by the worker thread to process a SLI4 fast-path
	747	+ * NVMET aborted xri.
	748	+ **/
	749	+void
	750	+lpfc_sli4_nvmet_xri_aborted(struct lpfc_hba *phba,
	751	+ struct sli4_wcqe_xri_aborted *axri)
	752	+{
	753	+ /* TODO: work in progress */
	754	+}
	755	+
734	756	void
735	757	lpfc_nvmet_destroy_targetport(struct lpfc_hba *phba)
736	758	{
	759	+#ifdef CONFIG_LPFC_NVME_TARGET
737	760	struct lpfc_nvmet_tgtport *tgtp;
738	761
739	762	if (phba->nvmet_support == 0)
@@ -745,6 +768,7 @@ lpfc_nvmet_destroy_targetport(struct lpfc_hba *phba) · 5 unmodified lines
745	768	wait_for_completion_timeout(&tgtp->tport_unreg_done, 5);
746	769	}
747	770	phba->targetport = NULL;
	771	+#endif
748	772	}
749	773
750	774	/**
@@ -764,6 +788,7 @@ static void · 13 unmodified lines
764	788	lpfc_nvmet_unsol_ls_buffer(struct lpfc_hba phba, struct lpfc_sli_ring pring,
765	789	struct hbq_dmabuf *nvmebuf)
766	790	{
	791	+#ifdef CONFIG_LPFC_NVME_TARGET
767	792	struct lpfc_nvmet_tgtport *tgtp;
768	793	struct fc_frame_header *fc_hdr;
769	794	struct lpfc_nvmet_rcv_ctx *ctxp;
@@ -844,6 +869,7 @@ dropit: · 74 unmodified lines
844	869
845	870	atomic_inc(&tgtp->xmt_ls_abort);
846	871	lpfc_nvmet_unsol_ls_issue_abort(phba, ctxp, sid, oxid);
	872	+#endif
847	873	}
848	874
849	875	/**
@@ -865,6 +891,7 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba, · 15 unmodified lines
865	891	struct rqb_dmabuf *nvmebuf,
866	892	uint64_t isr_timestamp)
867	893	{
	894	+#ifdef CONFIG_LPFC_NVME_TARGET
868	895	struct lpfc_nvmet_rcv_ctx *ctxp;
869	896	struct lpfc_nvmet_tgtport *tgtp;
870	897	struct fc_frame_header *fc_hdr;
@@ -955,7 +982,7 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba, · 84 unmodified lines
955	982
956	983	atomic_inc(&tgtp->rcv_fcp_cmd_drop);
957	984	lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
958		− "6159 FCP Drop IO x%x: nvmet_fc_rcv_fcp_req x%x\n",
	985	+ "6159 FCP Drop IO x%x: err x%x\n",
959	986	ctxp->oxid, rc);
960	987	dropit:
961	988	lpfc_nvmeio_data(phba, "NVMET FCP DROP: xri x%x sz %d from %06x\n",
@@ -970,6 +997,7 @@ dropit: · 8 unmodified lines
970	997	/* We assume a rcv'ed cmd ALWAYs fits into 1 buffer */
971	998	lpfc_nvmet_rq_post(phba, NULL, &nvmebuf->hbuf);
972	999	}
	1000	+#endif
973	1001	}
974	1002
975	1003	/**
@@ -1114,7 +1142,7 @@ lpfc_nvmet_prep_ls_wqe(struct lpfc_hba *phba, · 138 unmodified lines
1114	1142	bf_set(wqe_dfctl, &wqe->xmit_sequence.wge_ctl, 0);
1115	1143	bf_set(wqe_ls, &wqe->xmit_sequence.wge_ctl, 1);
1116	1144	bf_set(wqe_la, &wqe->xmit_sequence.wge_ctl, 0);
1117		− bf_set(wqe_rctl, &wqe->xmit_sequence.wge_ctl, FC_RCTL_DD_SOL_CTL);
	1145	+ bf_set(wqe_rctl, &wqe->xmit_sequence.wge_ctl, FC_RCTL_ELS4_REP);
1118	1146	bf_set(wqe_type, &wqe->xmit_sequence.wge_ctl, FC_TYPE_NVME);
1119	1147
1120	1148	/* Word 6 */
@@ -1445,7 +1473,6 @@ lpfc_nvmet_prep_fcp_wqe(struct lpfc_hba *phba, · 324 unmodified lines
1445	1473
1446	1474	case NVMET_FCOP_RSP:
1447	1475	/* Words 0 - 2 */
1448		− sgel = &rsp->sg[0];
1449	1476	physaddr = rsp->rspdma;
1450	1477	wqe->fcp_trsp.bde.tus.f.bdeFlags = BUFF_TYPE_BDE_64;
1451	1478	wqe->fcp_trsp.bde.tus.f.bdeSize = rsp->rsplen;
@@ -1681,8 +1708,8 @@ lpfc_nvmet_unsol_issue_abort(struct lpfc_hba *phba, · 229 unmodified lines
1681	1708	struct lpfc_nodelist *ndlp;
1682	1709
1683	1710	lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
1684		− "6067 %s: Entrypoint: sid %x xri %x\n", __func__,
1685		− sid, xri);
	1711	+ "6067 Abort: sid %x xri x%x/x%x\n",
	1712	+ sid, xri, ctxp->wqeq->sli4_xritag);
1686	1713
1687	1714	tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
1688	1715
@@ -1693,7 +1720,7 @@ lpfc_nvmet_unsol_issue_abort(struct lpfc_hba *phba, · 4 unmodified lines
1693	1720	atomic_inc(&tgtp->xmt_abort_rsp_error);
1694	1721	lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
1695	1722	"6134 Drop ABTS - wrong NDLP state x%x.\n",
1696		− ndlp->nlp_state);
	1723	+ (ndlp) ? ndlp->nlp_state : NLP_STE_MAX_STATE);
1697	1724
1698	1725	/* No failure to an ABTS request. */
1699	1726	return 0;
@@ -1791,7 +1818,7 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba, · 91 unmodified lines
1791	1818	atomic_inc(&tgtp->xmt_abort_rsp_error);
1792	1819	lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
1793	1820	"6160 Drop ABTS - wrong NDLP state x%x.\n",
1794		− ndlp->nlp_state);
	1821	+ (ndlp) ? ndlp->nlp_state : NLP_STE_MAX_STATE);
1795	1822
1796	1823	/* No failure to an ABTS request. */
1797	1824	return 0;

drivers/scsi/lpfc/lpfc_scsi.c+2−2

@@ -5953,12 +5953,13 @@ struct scsi_host_template lpfc_template_nvme = { · 5952 unmodified lines
5953	5953	.track_queue_depth = 0,
5954	5954	};
5955	5955
5956		−struct scsi_host_template lpfc_template_s3 = {
	5956	+struct scsi_host_template lpfc_template_no_hr = {
5957	5957	.module = THIS_MODULE,
5958	5958	.name = LPFC_DRIVER_NAME,
5959	5959	.proc_name = LPFC_DRIVER_NAME,
5960	5960	.info = lpfc_info,
5961	5961	.queuecommand = lpfc_queuecommand,
	5962	+ .eh_timed_out = fc_eh_timed_out,
5962	5963	.eh_abort_handler = lpfc_abort_handler,
5963	5964	.eh_device_reset_handler = lpfc_device_reset_handler,
5964	5965	.eh_target_reset_handler = lpfc_target_reset_handler,
@@ -6015,7 +6016,6 @@ struct scsi_host_template lpfc_vport_template = { · 50 unmodified lines
6015	6016	.eh_abort_handler = lpfc_abort_handler,
6016	6017	.eh_device_reset_handler = lpfc_device_reset_handler,
6017	6018	.eh_target_reset_handler = lpfc_target_reset_handler,
6018		− .eh_bus_reset_handler = lpfc_bus_reset_handler,
6019	6019	.slave_alloc = lpfc_slave_alloc,
6020	6020	.slave_configure = lpfc_slave_configure,
6021	6021	.slave_destroy = lpfc_slave_destroy,

drivers/scsi/lpfc/lpfc_sli.c+59−9

@@ -1,3 +1,4 @@
	1	+
1	2	/*******************************************************************
2	3	* This file is part of the Emulex Linux Device Driver for *
3	4	* Fibre Channel Host Bus Adapters. *
@@ -952,7 +953,7 @@ __lpfc_sli_get_els_sglq(struct lpfc_hba phba, struct lpfc_iocbq piocbq) · 948 unmodified lines
952	953	start_sglq = sglq;
953	954	while (!found) {
954	955	if (!sglq)
955		− return NULL;
	956	+ break;
956	957	if (ndlp && ndlp->active_rrqs_xri_bitmap &&
957	958	test_bit(sglq->sli4_lxritag,
958	959	ndlp->active_rrqs_xri_bitmap)) {
@@ -12212,6 +12213,41 @@ void lpfc_sli4_fcp_xri_abort_event_proc(struct lpfc_hba *phba) · 11253 unmodified lines
12212	12213	}
12213	12214	}
12214	12215
	12216	+/**
	12217	+ * lpfc_sli4_nvme_xri_abort_event_proc - Process nvme xri abort event
	12218	+ * @phba: pointer to lpfc hba data structure.
	12219	+ *
	12220	+ * This routine is invoked by the worker thread to process all the pending
	12221	+ * SLI4 NVME abort XRI events.
	12222	+ **/
	12223	+void lpfc_sli4_nvme_xri_abort_event_proc(struct lpfc_hba *phba)
	12224	+{
	12225	+ struct lpfc_cq_event *cq_event;
	12226	+
	12227	+ /* First, declare the fcp xri abort event has been handled */
	12228	+ spin_lock_irq(&phba->hbalock);
	12229	+ phba->hba_flag &= ~NVME_XRI_ABORT_EVENT;
	12230	+ spin_unlock_irq(&phba->hbalock);
	12231	+ /* Now, handle all the fcp xri abort events */
	12232	+ while (!list_empty(&phba->sli4_hba.sp_nvme_xri_aborted_work_queue)) {
	12233	+ /* Get the first event from the head of the event queue */
	12234	+ spin_lock_irq(&phba->hbalock);
	12235	+ list_remove_head(&phba->sli4_hba.sp_nvme_xri_aborted_work_queue,
	12236	+ cq_event, struct lpfc_cq_event, list);
	12237	+ spin_unlock_irq(&phba->hbalock);
	12238	+ /* Notify aborted XRI for NVME work queue */
	12239	+ if (phba->nvmet_support) {
	12240	+ lpfc_sli4_nvmet_xri_aborted(phba,
	12241	+ &cq_event->cqe.wcqe_axri);
	12242	+ } else {
	12243	+ lpfc_sli4_nvme_xri_aborted(phba,
	12244	+ &cq_event->cqe.wcqe_axri);
	12245	+ }
	12246	+ /* Free the event processed back to the free pool */
	12247	+ lpfc_sli4_cq_event_release(phba, cq_event);
	12248	+ }
	12249	+}
	12250	+
12215	12251	/**
12216	12252	* lpfc_sli4_els_xri_abort_event_proc - Process els xri abort event
12217	12253	* @phba: pointer to lpfc hba data structure.
@@ -12709,10 +12745,22 @@ lpfc_sli4_sp_handle_abort_xri_wcqe(struct lpfc_hba *phba, · 491 unmodified lines
12709	12745	spin_unlock_irqrestore(&phba->hbalock, iflags);
12710	12746	workposted = true;
12711	12747	break;
	12748	+ case LPFC_NVME:
	12749	+ spin_lock_irqsave(&phba->hbalock, iflags);
	12750	+ list_add_tail(&cq_event->list,
	12751	+ &phba->sli4_hba.sp_nvme_xri_aborted_work_queue);
	12752	+ /* Set the nvme xri abort event flag */
	12753	+ phba->hba_flag \|= NVME_XRI_ABORT_EVENT;
	12754	+ spin_unlock_irqrestore(&phba->hbalock, iflags);
	12755	+ workposted = true;
	12756	+ break;
12712	12757	default:
12713	12758	lpfc_printf_log(phba, KERN_ERR, LOG_SLI,
12714		− "0603 Invalid work queue CQE subtype (x%x)\n",
12715		− cq->subtype);
	12759	+ "0603 Invalid CQ subtype %d: "
	12760	+ "%08x %08x %08x %08x\n",
	12761	+ cq->subtype, wcqe->word0, wcqe->parameter,
	12762	+ wcqe->word2, wcqe->word3);
	12763	+ lpfc_sli4_cq_event_release(phba, cq_event);
12716	12764	workposted = false;
12717	12765	break;
12718	12766	}
@@ -13827,6 +13875,8 @@ lpfc_dual_chute_pci_bar_map(struct lpfc_hba *phba, uint16_t pci_barset) · 1108 unmodified lines
13827	13875	* @startq: The starting FCP EQ to modify
13828	13876	*
13829	13877	* This function sends an MODIFY_EQ_DELAY mailbox command to the HBA.
	13878	+ * The command allows up to LPFC_MAX_EQ_DELAY_EQID_CNT EQ ID's to be
	13879	+ * updated in one mailbox command.
13830	13880	*
13831	13881	* The @phba struct is used to send mailbox command to HBA. The @startq
13832	13882	* is used to get the starting FCP EQ to change.
@@ -13879,7 +13929,7 @@ lpfc_modify_hba_eq_delay(struct lpfc_hba *phba, uint32_t startq) · 46 unmodified lines
13879	13929	eq_delay->u.request.eq[cnt].phase = 0;
13880	13930	eq_delay->u.request.eq[cnt].delay_multi = dmult;
13881	13931	cnt++;
13882		− if (cnt >= LPFC_MAX_EQ_DELAY)
	13932	+ if (cnt >= LPFC_MAX_EQ_DELAY_EQID_CNT)
13883	13933	break;
13884	13934	}
13885	13935	eq_delay->u.request.num_eq = cnt;
@@ -15185,17 +15235,17 @@ lpfc_mrq_create(struct lpfc_hba phba, struct lpfc_queue *hrqp, · 1299 unmodified lines
15185	15235	drq = drqp[idx];
15186	15236	cq = cqp[idx];
15187	15237
15188		− if (hrq->entry_count != drq->entry_count) {
15189		− status = -EINVAL;
15190		− goto out;
15191		− }
15192		−
15193	15238	/* sanity check on queue memory */
15194	15239	if (!hrq \|\| !drq \|\| !cq) {
15195	15240	status = -ENODEV;
15196	15241	goto out;
15197	15242	}
15198	15243
	15244	+ if (hrq->entry_count != drq->entry_count) {
	15245	+ status = -EINVAL;
	15246	+ goto out;
	15247	+ }
	15248	+
15199	15249	if (idx == 0) {
15200	15250	bf_set(lpfc_mbx_rq_create_num_pages,
15201	15251	&rq_create->u.request,

drivers/scsi/lpfc/lpfc_sli4.h+6−0

@@ -642,6 +642,7 @@ struct lpfc_sli4_hba { · 641 unmodified lines
642	642	struct list_head sp_asynce_work_queue;
643	643	struct list_head sp_fcp_xri_aborted_work_queue;
644	644	struct list_head sp_els_xri_aborted_work_queue;
	645	+ struct list_head sp_nvme_xri_aborted_work_queue;
645	646	struct list_head sp_unsol_work_queue;
646	647	struct lpfc_sli4_link link_state;
647	648	struct lpfc_sli4_lnk_info lnk_info;
@@ -794,9 +795,14 @@ void lpfc_sli4_fcf_redisc_event_proc(struct lpfc_hba *); · 146 unmodified lines
794	795	int lpfc_sli4_resume_rpi(struct lpfc_nodelist *,
795	796	void ()(struct lpfc_hba , LPFC_MBOXQ_t ), void );
796	797	void lpfc_sli4_fcp_xri_abort_event_proc(struct lpfc_hba *);
	798	+void lpfc_sli4_nvme_xri_abort_event_proc(struct lpfc_hba *phba);
797	799	void lpfc_sli4_els_xri_abort_event_proc(struct lpfc_hba *);
798	800	void lpfc_sli4_fcp_xri_aborted(struct lpfc_hba *,
799	801	struct sli4_wcqe_xri_aborted *);
	802	+void lpfc_sli4_nvme_xri_aborted(struct lpfc_hba *phba,
	803	+ struct sli4_wcqe_xri_aborted *axri);
	804	+void lpfc_sli4_nvmet_xri_aborted(struct lpfc_hba *phba,
	805	+ struct sli4_wcqe_xri_aborted *axri);
800	806	void lpfc_sli4_els_xri_aborted(struct lpfc_hba *,
801	807	struct sli4_wcqe_xri_aborted *);
802	808	void lpfc_sli4_vport_delete_els_xri_aborted(struct lpfc_vport *);

drivers/scsi/lpfc/lpfc_version.h+1−1

@@ -20,7 +20,7 @@ · 19 unmodified lines
20	20	* included with this package. *
21	21	*******************************************************************/
22	22
23		−#define LPFC_DRIVER_VERSION "11.2.0.7"
	23	+#define LPFC_DRIVER_VERSION "11.2.0.10"
24	24	#define LPFC_DRIVER_NAME "lpfc"
25	25
26	26	/* Used for SLI 2/3 */

drivers/scsi/mpt3sas/mpt3sas_base.h+0−3

@@ -1442,9 +1442,6 @@ void mpt3sas_transport_update_links(struct MPT3SAS_ADAPTER *ioc, · 1441 unmodified lines
1442	1442	u64 sas_address, u16 handle, u8 phy_number, u8 link_rate);
1443	1443	extern struct sas_function_template mpt3sas_transport_functions;
1444	1444	extern struct scsi_transport_template *mpt3sas_transport_template;
1445		−extern int scsi_internal_device_block(struct scsi_device *sdev);
1446		−extern int scsi_internal_device_unblock(struct scsi_device *sdev,
1447		− enum scsi_device_state new_state);
1448	1445	/* trigger data externs */
1449	1446	void mpt3sas_send_trigger_data_event(struct MPT3SAS_ADAPTER *ioc,
1450	1447	struct SL_WH_TRIGGERS_EVENT_DATA_T *event_data);

drivers/scsi/mpt3sas/mpt3sas_scsih.c+2−17

@@ -2859,7 +2859,7 @@ _scsih_internal_device_block(struct scsi_device *sdev, · 2858 unmodified lines
2859	2859	sas_device_priv_data->sas_target->handle);
2860	2860	sas_device_priv_data->block = 1;
2861	2861
2862		− r = scsi_internal_device_block(sdev);
	2862	+ r = scsi_internal_device_block(sdev, false);
2863	2863	if (r == -EINVAL)
2864	2864	sdev_printk(KERN_WARNING, sdev,
2865	2865	"device_block failed with return(%d) for handle(0x%04x)\n",
@@ -2895,7 +2895,7 @@ _scsih_internal_device_unblock(struct scsi_device *sdev, · 29 unmodified lines
2895	2895	"performing a block followed by an unblock\n",
2896	2896	r, sas_device_priv_data->sas_target->handle);
2897	2897	sas_device_priv_data->block = 1;
2898		− r = scsi_internal_device_block(sdev);
	2898	+ r = scsi_internal_device_block(sdev, false);
2899	2899	if (r)
2900	2900	sdev_printk(KERN_WARNING, sdev, "retried device_block "
2901	2901	"failed with return(%d) for handle(0x%04x)\n",
@@ -4677,7 +4677,6 @@ _scsih_io_done(struct MPT3SAS_ADAPTER *ioc, u16 smid, u8 msix_index, u32 reply) · 1775 unmodified lines
4677	4677	struct MPT3SAS_DEVICE *sas_device_priv_data;
4678	4678	u32 response_code = 0;
4679	4679	unsigned long flags;
4680		− unsigned int sector_sz;
4681	4680
4682	4681	mpi_reply = mpt3sas_base_get_reply_virt_addr(ioc, reply);
4683	4682
@@ -4742,20 +4741,6 @@ _scsih_io_done(struct MPT3SAS_ADAPTER *ioc, u16 smid, u8 msix_index, u32 reply) · 58 unmodified lines
4742	4741	}
4743	4742
4744	4743	xfer_cnt = le32_to_cpu(mpi_reply->TransferCount);
4745		−
4746		− /* In case of bogus fw or device, we could end up having
4747		− * unaligned partial completion. We can force alignment here,
4748		− * then scsi-ml does not need to handle this misbehavior.
4749		− */
4750		− sector_sz = scmd->device->sector_size;
4751		− if (unlikely(!blk_rq_is_passthrough(scmd->request) && sector_sz &&
4752		− xfer_cnt % sector_sz)) {
4753		− sdev_printk(KERN_INFO, scmd->device,
4754		− "unaligned partial completion avoided (xfer_cnt=%u, sector_sz=%u)\n",
4755		− xfer_cnt, sector_sz);
4756		− xfer_cnt = round_down(xfer_cnt, sector_sz);
4757		− }
4758		−
4759	4744	scsi_set_resid(scmd, scsi_bufflen(scmd) - xfer_cnt);
4760	4745	if (ioc_status & MPI2_IOCSTATUS_FLAG_LOG_INFO_AVAILABLE)
4761	4746	log_info = le32_to_cpu(mpi_reply->IOCLogInfo);

drivers/scsi/qedf/qedf_dbg.h+8−5

@@ -81,14 +81,17 @@ struct qedf_dbg_ctx { · 80 unmodified lines
81	81	#define QEDF_INFO(pdev, level, fmt, ...) \
82	82	qedf_dbg_info(pdev, __func__, __LINE__, level, fmt, \
83	83	## __VA_ARGS__)
84		−
85		−extern void qedf_dbg_err(struct qedf_dbg_ctx qedf, const char func, u32 line,
	84	+__printf(4, 5)
	85	+void qedf_dbg_err(struct qedf_dbg_ctx qedf, const char func, u32 line,
86	86	const char *fmt, ...);
87		−extern void qedf_dbg_warn(struct qedf_dbg_ctx qedf, const char func, u32 line,
	87	+__printf(4, 5)
	88	+void qedf_dbg_warn(struct qedf_dbg_ctx qedf, const char func, u32 line,
88	89	const char *, ...);
89		−extern void qedf_dbg_notice(struct qedf_dbg_ctx qedf, const char func,
	90	+__printf(4, 5)
	91	+void qedf_dbg_notice(struct qedf_dbg_ctx qedf, const char func,
90	92	u32 line, const char *, ...);
91		−extern void qedf_dbg_info(struct qedf_dbg_ctx qedf, const char func, u32 line,
	93	+__printf(5, 6)
	94	+void qedf_dbg_info(struct qedf_dbg_ctx qedf, const char func, u32 line,
92	95	u32 info, const char *fmt, ...);
93	96
94	97	/* GRC Dump related defines */

drivers/scsi/qedf/qedf_fip.c+1−1

@@ -203,7 +203,7 @@ void qedf_fip_recv(struct qedf_ctx qedf, struct sk_buff skb) · 202 unmodified lines
203	203	case FIP_DT_MAC:
204	204	mp = (struct fip_mac_desc *)desc;
205	205	QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
206		− "fd_mac=%pM.\n", __func__, mp->fd_mac);
	206	+ "fd_mac=%pM\n", mp->fd_mac);
207	207	ether_addr_copy(cvl_mac, mp->fd_mac);
208	208	break;
209	209	case FIP_DT_NAME:

drivers/scsi/qedf/qedf_io.c+2−2

@@ -1342,7 +1342,7 @@ void qedf_scsi_completion(struct qedf_ctx qedf, struct fcoe_cqe cqe, · 1341 unmodified lines
1342	1342	} else {
1343	1343	refcount = kref_read(&io_req->refcount);
1344	1344	QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
1345		− "%d:0:%d:%d xid=0x%0x op=0x%02x "
	1345	+ "%d:0:%d:%lld xid=0x%0x op=0x%02x "
1346	1346	"lba=%02x%02x%02x%02x cdb_status=%d "
1347	1347	"fcp_resid=0x%x refcount=%d.\n",
1348	1348	qedf->lport->host->host_no, sc_cmd->device->id,
@@ -1426,7 +1426,7 @@ void qedf_scsi_done(struct qedf_ctx qedf, struct qedf_ioreq io_req, · 77 unmodified lines
1426	1426
1427	1427	sc_cmd->result = result << 16;
1428	1428	refcount = kref_read(&io_req->refcount);
1429		− QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "%d:0:%d:%d: Completing "
	1429	+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "%d:0:%d:%lld: Completing "
1430	1430	"sc_cmd=%p result=0x%08x op=0x%02x lba=0x%02x%02x%02x%02x, "
1431	1431	"allowed=%d retries=%d refcount=%d.\n",
1432	1432	qedf->lport->host->host_no, sc_cmd->device->id,

drivers/scsi/qedf/qedf_main.c+2−2

@@ -2456,8 +2456,8 @@ static int qedf_alloc_bdq(struct qedf_ctx *qedf) · 2455 unmodified lines
2456	2456	}
2457	2457
2458	2458	QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
2459		− "BDQ PBL addr=0x%p dma=0x%llx.\n", qedf->bdq_pbl,
2460		− qedf->bdq_pbl_dma);
	2459	+ "BDQ PBL addr=0x%p dma=%pad\n",
	2460	+ qedf->bdq_pbl, &qedf->bdq_pbl_dma);
2461	2461
2462	2462	/*
2463	2463	* Populate BDQ PBL with physical and virtual address of individual

drivers/scsi/qedi/qedi_debugfs.c+8−8

@@ -14,7 +14,7 @@ · 13 unmodified lines
14	14	#include <linux/debugfs.h>
15	15	#include <linux/module.h>
16	16
17		−int do_not_recover;
	17	+int qedi_do_not_recover;
18	18	static struct dentry *qedi_dbg_root;
19	19
20	20	void
@@ -74,22 +74,22 @@ qedi_dbg_exit(void) · 53 unmodified lines
74	74	static ssize_t
75	75	qedi_dbg_do_not_recover_enable(struct qedi_dbg_ctx *qedi_dbg)
76	76	{
77		− if (!do_not_recover)
78		− do_not_recover = 1;
	77	+ if (!qedi_do_not_recover)
	78	+ qedi_do_not_recover = 1;
79	79
80	80	QEDI_INFO(qedi_dbg, QEDI_LOG_DEBUGFS, "do_not_recover=%d\n",
81		− do_not_recover);
	81	+ qedi_do_not_recover);
82	82	return 0;
83	83	}
84	84
85	85	static ssize_t
86	86	qedi_dbg_do_not_recover_disable(struct qedi_dbg_ctx *qedi_dbg)
87	87	{
88		− if (do_not_recover)
89		− do_not_recover = 0;
	88	+ if (qedi_do_not_recover)
	89	+ qedi_do_not_recover = 0;
90	90
91	91	QEDI_INFO(qedi_dbg, QEDI_LOG_DEBUGFS, "do_not_recover=%d\n",
92		− do_not_recover);
	92	+ qedi_do_not_recover);
93	93	return 0;
94	94	}
95	95
@@ -141,7 +141,7 @@ qedi_dbg_do_not_recover_cmd_read(struct file filp, char __user buffer, · 45 unmodified lines
141	141	if (*ppos)
142	142	return 0;
143	143
144		− cnt = sprintf(buffer, "do_not_recover=%d\n", do_not_recover);
	144	+ cnt = sprintf(buffer, "do_not_recover=%d\n", qedi_do_not_recover);
145	145	cnt = min_t(int, count, cnt - *ppos);
146	146	*ppos += cnt;
147	147	return cnt;

drivers/scsi/qedi/qedi_fw.c+2−2

@@ -1461,9 +1461,9 @@ static void qedi_tmf_work(struct work_struct *work) · 1460 unmodified lines
1461	1461	get_itt(tmf_hdr->rtt), get_itt(ctask->itt), cmd->task_id,
1462	1462	qedi_conn->iscsi_conn_id);
1463	1463
1464		− if (do_not_recover) {
	1464	+ if (qedi_do_not_recover) {
1465	1465	QEDI_ERR(&qedi->dbg_ctx, "DONT SEND CLEANUP/ABORT %d\n",
1466		− do_not_recover);
	1466	+ qedi_do_not_recover);
1467	1467	goto abort_ret;
1468	1468	}
1469	1469

drivers/scsi/qedi/qedi_gbl.h+7−1

@@ -12,8 +12,14 @@ · 11 unmodified lines
12	12
13	13	#include "qedi_iscsi.h"
14	14
	15	+#ifdef CONFIG_DEBUG_FS
	16	+extern int qedi_do_not_recover;
	17	+#else
	18	+#define qedi_do_not_recover (0)
	19	+#endif
	20	+
15	21	extern uint qedi_io_tracing;
16		−extern int do_not_recover;
	22	+
17	23	extern struct scsi_host_template qedi_host_template;
18	24	extern struct iscsi_transport qedi_iscsi_transport;
19	25	extern const struct qed_iscsi_ops *qedi_ops;

drivers/scsi/qedi/qedi_iscsi.c+4−4

@@ -833,7 +833,7 @@ qedi_ep_connect(struct Scsi_Host shost, struct sockaddr dst_addr, · 832 unmodified lines
833	833	return ERR_PTR(ret);
834	834	}
835	835
836		− if (do_not_recover) {
	836	+ if (qedi_do_not_recover) {
837	837	ret = -ENOMEM;
838	838	return ERR_PTR(ret);
839	839	}
@@ -957,7 +957,7 @@ static int qedi_ep_poll(struct iscsi_endpoint *ep, int timeout_ms) · 117 unmodified lines
957	957	struct qedi_endpoint *qedi_ep;
958	958	int ret = 0;
959	959
960		− if (do_not_recover)
	960	+ if (qedi_do_not_recover)
961	961	return 1;
962	962
963	963	qedi_ep = ep->dd_data;
@@ -1025,7 +1025,7 @@ static void qedi_ep_disconnect(struct iscsi_endpoint *ep) · 61 unmodified lines
1025	1025	}
1026	1026
1027	1027	if (test_bit(QEDI_IN_RECOVERY, &qedi->flags)) {
1028		− if (do_not_recover) {
	1028	+ if (qedi_do_not_recover) {
1029	1029	QEDI_INFO(&qedi->dbg_ctx, QEDI_LOG_INFO,
1030	1030	"Do not recover cid=0x%x\n",
1031	1031	qedi_ep->iscsi_cid);
@@ -1039,7 +1039,7 @@ static void qedi_ep_disconnect(struct iscsi_endpoint *ep) · 7 unmodified lines
1039	1039	}
1040	1040	}
1041	1041
1042		− if (do_not_recover)
	1042	+ if (qedi_do_not_recover)
1043	1043	goto ep_exit_recover;
1044	1044
1045	1045	switch (qedi_ep->state) {

drivers/scsi/qedi/qedi_main.c+1−1

@@ -1805,7 +1805,7 @@ static int __qedi_probe(struct pci_dev *pdev, int mode) · 1804 unmodified lines
1805	1805	*/
1806	1806	qedi_ops->common->update_pf_params(qedi->cdev, &qedi->pf_params);
1807	1807
1808		− qedi_setup_int(qedi);
	1808	+ rc = qedi_setup_int(qedi);
1809	1809	if (rc)
1810	1810	goto stop_iscsi_func;
1811	1811

drivers/scsi/qla2xxx/qla_dbg.c+4−8

@@ -2707,13 +2707,9 @@ ql_dump_buffer(uint32_t level, scsi_qla_host_t *vha, int32_t id, · 2706 unmodified lines
2707	2707	"%-+5d 0 1 2 3 4 5 6 7 8 9 A B C D E F\n", size);
2708	2708	ql_dbg(level, vha, id,
2709	2709	"----- -----------------------------------------------\n");
2710		− for (cnt = 0; cnt < size; cnt++, buf++) {
2711		− if (cnt % 16 == 0)
2712		− ql_dbg(level, vha, id, "%04x:", cnt & ~0xFU);
2713		− printk(" %02x", *buf);
2714		− if (cnt % 16 == 15)
2715		− printk("\n");
	2710	+ for (cnt = 0; cnt < size; cnt += 16) {
	2711	+ ql_dbg(level, vha, id, "%04x: ", cnt);
	2712	+ print_hex_dump(KERN_CONT, "", DUMP_PREFIX_NONE, 16, 1,
	2713	+ buf + cnt, min(16U, size - cnt), false);
2716	2714	}
2717		− if (cnt % 16 != 0)
2718		− printk("\n");
2719	2715	}

drivers/scsi/scsi_lib.c+10−4

@@ -2932,6 +2932,8 @@ EXPORT_SYMBOL(scsi_target_resume); · 2931 unmodified lines
2932	2932	/**
2933	2933	* scsi_internal_device_block - internal function to put a device temporarily into the SDEV_BLOCK state
2934	2934	* @sdev: device to block
	2935	+ * @wait: Whether or not to wait until ongoing .queuecommand() /
	2936	+ * .queue_rq() calls have finished.
2935	2937	*
2936	2938	* Block request made by scsi lld's to temporarily stop all
2937	2939	* scsi commands on the specified device. May sleep.
@@ -2949,7 +2951,7 @@ EXPORT_SYMBOL(scsi_target_resume); · 11 unmodified lines
2949	2951	* remove the rport mutex lock and unlock calls from srp_queuecommand().
2950	2952	*/
2951	2953	int
2952		−scsi_internal_device_block(struct scsi_device *sdev)
	2954	+scsi_internal_device_block(struct scsi_device *sdev, bool wait)
2953	2955	{
2954	2956	struct request_queue *q = sdev->request_queue;
2955	2957	unsigned long flags;
@@ -2969,12 +2971,16 @@ scsi_internal_device_block(struct scsi_device *sdev) · 13 unmodified lines
2969	2971	* request queue.
2970	2972	*/
2971	2973	if (q->mq_ops) {
2972		− blk_mq_quiesce_queue(q);
	2974	+ if (wait)
	2975	+ blk_mq_quiesce_queue(q);
	2976	+ else
	2977	+ blk_mq_stop_hw_queues(q);
2973	2978	} else {
2974	2979	spin_lock_irqsave(q->queue_lock, flags);
2975	2980	blk_stop_queue(q);
2976	2981	spin_unlock_irqrestore(q->queue_lock, flags);
2977		− scsi_wait_for_queuecommand(sdev);
	2982	+ if (wait)
	2983	+ scsi_wait_for_queuecommand(sdev);
2978	2984	}
2979	2985
2980	2986	return 0;
@@ -3036,7 +3042,7 @@ EXPORT_SYMBOL_GPL(scsi_internal_device_unblock); · 55 unmodified lines
3036	3042	static void
3037	3043	device_block(struct scsi_device sdev, void data)
3038	3044	{
3039		− scsi_internal_device_block(sdev);
	3045	+ scsi_internal_device_block(sdev, true);
3040	3046	}
3041	3047
3042	3048	static int

drivers/scsi/scsi_priv.h+0−3

@@ -188,8 +188,5 @@ static inline void scsi_dh_remove_device(struct scsi_device *sdev) { } · 187 unmodified lines
188	188	*/
189	189
190	190	#define SCSI_DEVICE_BLOCK_MAX_TIMEOUT 600 /* units in seconds */
191		−extern int scsi_internal_device_block(struct scsi_device *sdev);
192		−extern int scsi_internal_device_unblock(struct scsi_device *sdev,
193		− enum scsi_device_state new_state);
194	191
195	192	#endif /* _SCSI_PRIV_H */

drivers/scsi/sd.c+25−33

@@ -1783,6 +1783,8 @@ static int sd_done(struct scsi_cmnd *SCpnt) · 1782 unmodified lines
1783	1783	{
1784	1784	int result = SCpnt->result;
1785	1785	unsigned int good_bytes = result ? 0 : scsi_bufflen(SCpnt);
	1786	+ unsigned int sector_size = SCpnt->device->sector_size;
	1787	+ unsigned int resid;
1786	1788	struct scsi_sense_hdr sshdr;
1787	1789	struct scsi_disk *sdkp = scsi_disk(SCpnt->request->rq_disk);
1788	1790	struct request *req = SCpnt->request;
@@ -1813,6 +1815,21 @@ static int sd_done(struct scsi_cmnd *SCpnt) · 24 unmodified lines
1813	1815	scsi_set_resid(SCpnt, blk_rq_bytes(req));
1814	1816	}
1815	1817	break;
	1818	+ default:
	1819	+ /*
	1820	+ * In case of bogus fw or device, we could end up having
	1821	+ * an unaligned partial completion. Check this here and force
	1822	+ * alignment.
	1823	+ */
	1824	+ resid = scsi_get_resid(SCpnt);
	1825	+ if (resid & (sector_size - 1)) {
	1826	+ sd_printk(KERN_INFO, sdkp,
	1827	+ "Unaligned partial completion (resid=%u, sector_sz=%u)\n",
	1828	+ resid, sector_size);
	1829	+ resid = min(scsi_bufflen(SCpnt),
	1830	+ round_up(resid, sector_size));
	1831	+ scsi_set_resid(SCpnt, resid);
	1832	+ }
1816	1833	}
1817	1834
1818	1835	if (result) {
@@ -3075,23 +3092,6 @@ static void sd_probe_async(void *data, async_cookie_t cookie) · 1256 unmodified lines
3075	3092	put_device(&sdkp->dev);
3076	3093	}
3077	3094
3078		−struct sd_devt {
3079		− int idx;
3080		− struct disk_devt disk_devt;
3081		−};
3082		−
3083		−static void sd_devt_release(struct disk_devt *disk_devt)
3084		−{
3085		− struct sd_devt *sd_devt = container_of(disk_devt, struct sd_devt,
3086		− disk_devt);
3087		−
3088		− spin_lock(&sd_index_lock);
3089		− ida_remove(&sd_index_ida, sd_devt->idx);
3090		− spin_unlock(&sd_index_lock);
3091		−
3092		− kfree(sd_devt);
3093		−}
3094		−
3095	3095	/**
3096	3096	* sd_probe - called during driver initialization and whenever a
3097	3097	* new scsi device is attached to the system. It is called once
@@ -3113,7 +3113,6 @@ static void sd_devt_release(struct disk_devt *disk_devt) · 15 unmodified lines
3113	3113	static int sd_probe(struct device *dev)
3114	3114	{
3115	3115	struct scsi_device *sdp = to_scsi_device(dev);
3116		− struct sd_devt *sd_devt;
3117	3116	struct scsi_disk *sdkp;
3118	3117	struct gendisk *gd;
3119	3118	int index;
@@ -3139,13 +3138,9 @@ static int sd_probe(struct device *dev) · 19 unmodified lines
3139	3138	if (!sdkp)
3140	3139	goto out;
3141	3140
3142		− sd_devt = kzalloc(sizeof(*sd_devt), GFP_KERNEL);
3143		− if (!sd_devt)
3144		− goto out_free;
3145		−
3146	3141	gd = alloc_disk(SD_MINORS);
3147	3142	if (!gd)
3148		− goto out_free_devt;
	3143	+ goto out_free;
3149	3144
3150	3145	do {
3151	3146	if (!ida_pre_get(&sd_index_ida, GFP_KERNEL))
@@ -3161,11 +3156,6 @@ static int sd_probe(struct device *dev) · 9 unmodified lines
3161	3156	goto out_put;
3162	3157	}
3163	3158
3164		− atomic_set(&sd_devt->disk_devt.count, 1);
3165		− sd_devt->disk_devt.release = sd_devt_release;
3166		− sd_devt->idx = index;
3167		− gd->disk_devt = &sd_devt->disk_devt;
3168		−
3169	3159	error = sd_format_disk_name("sd", index, gd->disk_name, DISK_NAME_LEN);
3170	3160	if (error) {
3171	3161	sdev_printk(KERN_WARNING, sdp, "SCSI disk (sd) name length exceeded.\n");
@@ -3205,12 +3195,11 @@ static int sd_probe(struct device *dev) · 33 unmodified lines
3205	3195	return 0;
3206	3196
3207	3197	out_free_index:
3208		− put_disk_devt(&sd_devt->disk_devt);
3209		− sd_devt = NULL;
	3198	+ spin_lock(&sd_index_lock);
	3199	+ ida_remove(&sd_index_ida, index);
	3200	+ spin_unlock(&sd_index_lock);
3210	3201	out_put:
3211	3202	put_disk(gd);
3212		− out_free_devt:
3213		− kfree(sd_devt);
3214	3203	out_free:
3215	3204	kfree(sdkp);
3216	3205	out:
@@ -3271,7 +3260,10 @@ static void scsi_disk_release(struct device *dev) · 54 unmodified lines
3271	3260	struct scsi_disk *sdkp = to_scsi_disk(dev);
3272	3261	struct gendisk *disk = sdkp->disk;
3273	3262
3274		− put_disk_devt(disk->disk_devt);
	3263	+ spin_lock(&sd_index_lock);
	3264	+ ida_remove(&sd_index_ida, sdkp->index);
	3265	+ spin_unlock(&sd_index_lock);
	3266	+
3275	3267	disk->private_data = NULL;
3276	3268	put_disk(disk);
3277	3269	put_device(&sdkp->device->sdev_gendev);

drivers/scsi/storvsc_drv.c+17−10

@@ -400,8 +400,6 @@ MODULE_PARM_DESC(storvsc_vcpus_per_sub_channel, "Ratio of VCPUs to subchannels") · 399 unmodified lines
400	400	*/
401	401	static int storvsc_timeout = 180;
402	402
403		−static int msft_blist_flags = BLIST_TRY_VPD_PAGES;
404		−
405	403	#if IS_ENABLED(CONFIG_SCSI_FC_ATTRS)
406	404	static struct scsi_transport_template *fc_transport_template;
407	405	#endif
@@ -1383,6 +1381,22 @@ static int storvsc_do_io(struct hv_device *device, · 975 unmodified lines
1383	1381	return ret;
1384	1382	}
1385	1383
	1384	+static int storvsc_device_alloc(struct scsi_device *sdevice)
	1385	+{
	1386	+ /*
	1387	+ * Set blist flag to permit the reading of the VPD pages even when
	1388	+ * the target may claim SPC-2 compliance. MSFT targets currently
	1389	+ * claim SPC-2 compliance while they implement post SPC-2 features.
	1390	+ * With this flag we can correctly handle WRITE_SAME_16 issues.
	1391	+ *
	1392	+ * Hypervisor reports SCSI_UNKNOWN type for DVD ROM device but
	1393	+ * still supports REPORT LUN.
	1394	+ */
	1395	+ sdevice->sdev_bflags = BLIST_REPORTLUN2 \| BLIST_TRY_VPD_PAGES;
	1396	+
	1397	+ return 0;
	1398	+}
	1399	+
1386	1400	static int storvsc_device_configure(struct scsi_device *sdevice)
1387	1401	{
1388	1402
@@ -1395,14 +1409,6 @@ static int storvsc_device_configure(struct scsi_device *sdevice) · 6 unmodified lines
1395	1409
1396	1410	sdevice->no_write_same = 1;
1397	1411
1398		− /*
1399		− * Add blist flags to permit the reading of the VPD pages even when
1400		− * the target may claim SPC-2 compliance. MSFT targets currently
1401		− * claim SPC-2 compliance while they implement post SPC-2 features.
1402		− * With this patch we can correctly handle WRITE_SAME_16 issues.
1403		− */
1404		− sdevice->sdev_bflags \|= msft_blist_flags;
1405		−
1406	1412	/*
1407	1413	* If the host is WIN8 or WIN8 R2, claim conformance to SPC-3
1408	1414	* if the device is a MSFT virtual device. If the host is
@@ -1661,6 +1667,7 @@ static struct scsi_host_template scsi_driver = { · 252 unmodified lines
1661	1667	.eh_host_reset_handler = storvsc_host_reset_handler,
1662	1668	.proc_name = "storvsc_host",
1663	1669	.eh_timed_out = storvsc_eh_timed_out,
	1670	+ .slave_alloc = storvsc_device_alloc,
1664	1671	.slave_configure = storvsc_device_configure,
1665	1672	.cmd_per_lun = 255,
1666	1673	.this_id = -1,

drivers/scsi/ufs/ufs.h+8−14

@@ -146,7 +146,7 @@ enum attr_idn { · 145 unmodified lines
146	146	/* Descriptor idn for Query requests */
147	147	enum desc_idn {
148	148	QUERY_DESC_IDN_DEVICE = 0x0,
149		− QUERY_DESC_IDN_CONFIGURAION = 0x1,
	149	+ QUERY_DESC_IDN_CONFIGURATION = 0x1,
150	150	QUERY_DESC_IDN_UNIT = 0x2,
151	151	QUERY_DESC_IDN_RFU_0 = 0x3,
152	152	QUERY_DESC_IDN_INTERCONNECT = 0x4,
@@ -162,19 +162,13 @@ enum desc_header_offset { · 9 unmodified lines
162	162	QUERY_DESC_DESC_TYPE_OFFSET = 0x01,
163	163	};
164	164
165		−enum ufs_desc_max_size {
166		− QUERY_DESC_DEVICE_MAX_SIZE = 0x40,
167		− QUERY_DESC_CONFIGURAION_MAX_SIZE = 0x90,
168		− QUERY_DESC_UNIT_MAX_SIZE = 0x23,
169		− QUERY_DESC_INTERCONNECT_MAX_SIZE = 0x06,
170		− /*
171		− * Max. 126 UNICODE characters (2 bytes per character) plus 2 bytes
172		− * of descriptor header.
173		− */
174		− QUERY_DESC_STRING_MAX_SIZE = 0xFE,
175		− QUERY_DESC_GEOMETRY_MAX_SIZE = 0x44,
176		− QUERY_DESC_POWER_MAX_SIZE = 0x62,
177		− QUERY_DESC_RFU_MAX_SIZE = 0x00,
	165	+enum ufs_desc_def_size {
	166	+ QUERY_DESC_DEVICE_DEF_SIZE = 0x40,
	167	+ QUERY_DESC_CONFIGURATION_DEF_SIZE = 0x90,
	168	+ QUERY_DESC_UNIT_DEF_SIZE = 0x23,
	169	+ QUERY_DESC_INTERCONNECT_DEF_SIZE = 0x06,
	170	+ QUERY_DESC_GEOMETRY_DEF_SIZE = 0x44,
	171	+ QUERY_DESC_POWER_DEF_SIZE = 0x62,
178	172	};
179	173
180	174	/* Unit descriptor parameters offsets in bytes*/

drivers/scsi/ufs/ufshcd.c+173−58

@@ -100,19 +100,6 @@ · 99 unmodified lines
100	100	#define ufshcd_hex_dump(prefix_str, buf, len) \
101	101	print_hex_dump(KERN_ERR, prefix_str, DUMP_PREFIX_OFFSET, 16, 4, buf, len, false)
102	102
103		−static u32 ufs_query_desc_max_size[] = {
104		− QUERY_DESC_DEVICE_MAX_SIZE,
105		− QUERY_DESC_CONFIGURAION_MAX_SIZE,
106		− QUERY_DESC_UNIT_MAX_SIZE,
107		− QUERY_DESC_RFU_MAX_SIZE,
108		− QUERY_DESC_INTERCONNECT_MAX_SIZE,
109		− QUERY_DESC_STRING_MAX_SIZE,
110		− QUERY_DESC_RFU_MAX_SIZE,
111		− QUERY_DESC_GEOMETRY_MAX_SIZE,
112		− QUERY_DESC_POWER_MAX_SIZE,
113		− QUERY_DESC_RFU_MAX_SIZE,
114		−};
115		−
116	103	enum {
117	104	UFSHCD_MAX_CHANNEL = 0,
118	105	UFSHCD_MAX_ID = 1,
@@ -2857,7 +2844,7 @@ static int __ufshcd_query_descriptor(struct ufs_hba *hba, · 2738 unmodified lines
2857	2844	goto out;
2858	2845	}
2859	2846
2860		− if (buf_len <= QUERY_DESC_MIN_SIZE \|\| buf_len > QUERY_DESC_MAX_SIZE) {
	2847	+ if (buf_len < QUERY_DESC_MIN_SIZE \|\| buf_len > QUERY_DESC_MAX_SIZE) {
2861	2848	dev_err(hba->dev, "%s: descriptor buffer size (%d) is out of range\n",
2862	2849	__func__, *buf_len);
2863	2850	err = -EINVAL;
@@ -2937,6 +2924,92 @@ static int ufshcd_query_descriptor_retry(struct ufs_hba *hba, · 73 unmodified lines
2937	2924	return err;
2938	2925	}
2939	2926
	2927	+/**
	2928	+ * ufshcd_read_desc_length - read the specified descriptor length from header
	2929	+ * @hba: Pointer to adapter instance
	2930	+ * @desc_id: descriptor idn value
	2931	+ * @desc_index: descriptor index
	2932	+ * @desc_length: pointer to variable to read the length of descriptor
	2933	+ *
	2934	+ * Return 0 in case of success, non-zero otherwise
	2935	+ */
	2936	+static int ufshcd_read_desc_length(struct ufs_hba *hba,
	2937	+ enum desc_idn desc_id,
	2938	+ int desc_index,
	2939	+ int *desc_length)
	2940	+{
	2941	+ int ret;
	2942	+ u8 header[QUERY_DESC_HDR_SIZE];
	2943	+ int header_len = QUERY_DESC_HDR_SIZE;
	2944	+
	2945	+ if (desc_id >= QUERY_DESC_IDN_MAX)
	2946	+ return -EINVAL;
	2947	+
	2948	+ ret = ufshcd_query_descriptor_retry(hba, UPIU_QUERY_OPCODE_READ_DESC,
	2949	+ desc_id, desc_index, 0, header,
	2950	+ &header_len);
	2951	+
	2952	+ if (ret) {
	2953	+ dev_err(hba->dev, "%s: Failed to get descriptor header id %d",
	2954	+ __func__, desc_id);
	2955	+ return ret;
	2956	+ } else if (desc_id != header[QUERY_DESC_DESC_TYPE_OFFSET]) {
	2957	+ dev_warn(hba->dev, "%s: descriptor header id %d and desc_id %d mismatch",
	2958	+ __func__, header[QUERY_DESC_DESC_TYPE_OFFSET],
	2959	+ desc_id);
	2960	+ ret = -EINVAL;
	2961	+ }
	2962	+
	2963	+ *desc_length = header[QUERY_DESC_LENGTH_OFFSET];
	2964	+ return ret;
	2965	+
	2966	+}
	2967	+
	2968	+/**
	2969	+ * ufshcd_map_desc_id_to_length - map descriptor IDN to its length
	2970	+ * @hba: Pointer to adapter instance
	2971	+ * @desc_id: descriptor idn value
	2972	+ * @desc_len: mapped desc length (out)
	2973	+ *
	2974	+ * Return 0 in case of success, non-zero otherwise
	2975	+ */
	2976	+int ufshcd_map_desc_id_to_length(struct ufs_hba *hba,
	2977	+ enum desc_idn desc_id, int *desc_len)
	2978	+{
	2979	+ switch (desc_id) {
	2980	+ case QUERY_DESC_IDN_DEVICE:
	2981	+ *desc_len = hba->desc_size.dev_desc;
	2982	+ break;
	2983	+ case QUERY_DESC_IDN_POWER:
	2984	+ *desc_len = hba->desc_size.pwr_desc;
	2985	+ break;
	2986	+ case QUERY_DESC_IDN_GEOMETRY:
	2987	+ *desc_len = hba->desc_size.geom_desc;
	2988	+ break;
	2989	+ case QUERY_DESC_IDN_CONFIGURATION:
	2990	+ *desc_len = hba->desc_size.conf_desc;
	2991	+ break;
	2992	+ case QUERY_DESC_IDN_UNIT:
	2993	+ *desc_len = hba->desc_size.unit_desc;
	2994	+ break;
	2995	+ case QUERY_DESC_IDN_INTERCONNECT:
	2996	+ *desc_len = hba->desc_size.interc_desc;
	2997	+ break;
	2998	+ case QUERY_DESC_IDN_STRING:
	2999	+ *desc_len = QUERY_DESC_MAX_SIZE;
	3000	+ break;
	3001	+ case QUERY_DESC_IDN_RFU_0:
	3002	+ case QUERY_DESC_IDN_RFU_1:
	3003	+ *desc_len = 0;
	3004	+ break;
	3005	+ default:
	3006	+ *desc_len = 0;
	3007	+ return -EINVAL;
	3008	+ }
	3009	+ return 0;
	3010	+}
	3011	+EXPORT_SYMBOL(ufshcd_map_desc_id_to_length);
	3012	+
2940	3013	/**
2941	3014	* ufshcd_read_desc_param - read the specified descriptor parameter
2942	3015	* @hba: Pointer to adapter instance
@@ -2951,42 +3024,49 @@ static int ufshcd_query_descriptor_retry(struct ufs_hba *hba, · 8 unmodified lines
2951	3024	static int ufshcd_read_desc_param(struct ufs_hba *hba,
2952	3025	enum desc_idn desc_id,
2953	3026	int desc_index,
2954		− u32 param_offset,
	3027	+ u8 param_offset,
2955	3028	u8 *param_read_buf,
2956		− u32 param_size)
	3029	+ u8 param_size)
2957	3030	{
2958	3031	int ret;
2959	3032	u8 *desc_buf;
2960		− u32 buff_len;
	3033	+ int buff_len;
2961	3034	bool is_kmalloc = true;
2962	3035
2963		− /* safety checks */
2964		− if (desc_id >= QUERY_DESC_IDN_MAX)
	3036	+ /* Safety check */
	3037	+ if (desc_id >= QUERY_DESC_IDN_MAX \|\| !param_size)
2965	3038	return -EINVAL;
2966	3039
2967		− buff_len = ufs_query_desc_max_size[desc_id];
2968		− if ((param_offset + param_size) > buff_len)
2969		− return -EINVAL;
	3040	+ /* Get the max length of descriptor from structure filled up at probe
	3041	+ * time.
	3042	+ */
	3043	+ ret = ufshcd_map_desc_id_to_length(hba, desc_id, &buff_len);
2970	3044
2971		− if (!param_offset && (param_size == buff_len)) {
2972		− /* memory space already available to hold full descriptor */
2973		− desc_buf = param_read_buf;
2974		− is_kmalloc = false;
2975		− } else {
2976		− /* allocate memory to hold full descriptor */
	3045	+ /* Sanity checks */
	3046	+ if (ret \|\| !buff_len) {
	3047	+ dev_err(hba->dev, "%s: Failed to get full descriptor length",
	3048	+ __func__);
	3049	+ return ret;
	3050	+ }
	3051	+
	3052	+ /* Check whether we need temp memory */
	3053	+ if (param_offset != 0 \|\| param_size < buff_len) {
2977	3054	desc_buf = kmalloc(buff_len, GFP_KERNEL);
2978	3055	if (!desc_buf)
2979	3056	return -ENOMEM;
	3057	+ } else {
	3058	+ desc_buf = param_read_buf;
	3059	+ is_kmalloc = false;
2980	3060	}
2981	3061
	3062	+ /* Request for full descriptor */
2982	3063	ret = ufshcd_query_descriptor_retry(hba, UPIU_QUERY_OPCODE_READ_DESC,
2983		− desc_id, desc_index, 0, desc_buf,
2984		− &buff_len);
	3064	+ desc_id, desc_index, 0,
	3065	+ desc_buf, &buff_len);
2985	3066
2986	3067	if (ret) {
2987	3068	dev_err(hba->dev, "%s: Failed reading descriptor. desc_id %d, desc_index %d, param_offset %d, ret %d",
2988	3069	__func__, desc_id, desc_index, param_offset, ret);
2989		−
2990	3070	goto out;
2991	3071	}
2992	3072
@@ -2998,25 +3078,9 @@ static int ufshcd_read_desc_param(struct ufs_hba *hba, · 5 unmodified lines
2998	3078	goto out;
2999	3079	}
3000	3080
3001		− /*
3002		− * While reading variable size descriptors (like string descriptor),
3003		− * some UFS devices may report the "LENGTH" (field in "Transaction
3004		− * Specific fields" of Query Response UPIU) same as what was requested
3005		− * in Query Request UPIU instead of reporting the actual size of the
3006		− * variable size descriptor.
3007		− * Although it's safe to ignore the "LENGTH" field for variable size
3008		− * descriptors as we can always derive the length of the descriptor from
3009		− * the descriptor header fields. Hence this change impose the length
3010		− * match check only for fixed size descriptors (for which we always
3011		− * request the correct size as part of Query Request UPIU).
3012		− */
3013		− if ((desc_id != QUERY_DESC_IDN_STRING) &&
3014		− (buff_len != desc_buf[QUERY_DESC_LENGTH_OFFSET])) {
3015		− dev_err(hba->dev, "%s: desc_buf length mismatch: buff_len %d, buff_len(desc_header) %d",
3016		− __func__, buff_len, desc_buf[QUERY_DESC_LENGTH_OFFSET]);
3017		− ret = -EINVAL;
3018		− goto out;
3019		− }
	3081	+ /* Check wherher we will not copy more data, than available */
	3082	+ if (is_kmalloc && param_size > buff_len)
	3083	+ param_size = buff_len;
3020	3084
3021	3085	if (is_kmalloc)
3022	3086	memcpy(param_read_buf, &desc_buf[param_offset], param_size);
@@ -5919,8 +5983,8 @@ static int ufshcd_set_icc_levels_attr(struct ufs_hba *hba, u32 icc_level) · 2896 unmodified lines
5919	5983	static void ufshcd_init_icc_levels(struct ufs_hba *hba)
5920	5984	{
5921	5985	int ret;
5922		− int buff_len = QUERY_DESC_POWER_MAX_SIZE;
5923		− u8 desc_buf[QUERY_DESC_POWER_MAX_SIZE];
	5986	+ int buff_len = hba->desc_size.pwr_desc;
	5987	+ u8 desc_buf[hba->desc_size.pwr_desc];
5924	5988
5925	5989	ret = ufshcd_read_power_desc(hba, desc_buf, buff_len);
5926	5990	if (ret) {
@@ -6017,11 +6081,10 @@ static int ufs_get_device_desc(struct ufs_hba *hba, · 90 unmodified lines
6017	6081	{
6018	6082	int err;
6019	6083	u8 model_index;
6020		− u8 str_desc_buf[QUERY_DESC_STRING_MAX_SIZE + 1] = {0};
6021		− u8 desc_buf[QUERY_DESC_DEVICE_MAX_SIZE];
	6084	+ u8 str_desc_buf[QUERY_DESC_MAX_SIZE + 1] = {0};
	6085	+ u8 desc_buf[hba->desc_size.dev_desc];
6022	6086
6023		− err = ufshcd_read_device_desc(hba, desc_buf,
6024		− QUERY_DESC_DEVICE_MAX_SIZE);
	6087	+ err = ufshcd_read_device_desc(hba, desc_buf, hba->desc_size.dev_desc);
6025	6088	if (err) {
6026	6089	dev_err(hba->dev, "%s: Failed reading Device Desc. err = %d\n",
6027	6090	__func__, err);
@@ -6038,14 +6101,14 @@ static int ufs_get_device_desc(struct ufs_hba *hba, · 10 unmodified lines
6038	6101	model_index = desc_buf[DEVICE_DESC_PARAM_PRDCT_NAME];
6039	6102
6040	6103	err = ufshcd_read_string_desc(hba, model_index, str_desc_buf,
6041		− QUERY_DESC_STRING_MAX_SIZE, ASCII_STD);
	6104	+ QUERY_DESC_MAX_SIZE, ASCII_STD);
6042	6105	if (err) {
6043	6106	dev_err(hba->dev, "%s: Failed reading Product Name. err = %d\n",
6044	6107	__func__, err);
6045	6108	goto out;
6046	6109	}
6047	6110
6048		− str_desc_buf[QUERY_DESC_STRING_MAX_SIZE] = '\0';
	6111	+ str_desc_buf[QUERY_DESC_MAX_SIZE] = '\0';
6049	6112	strlcpy(dev_desc->model, (str_desc_buf + QUERY_DESC_HDR_SIZE),
6050	6113	min_t(u8, str_desc_buf[QUERY_DESC_LENGTH_OFFSET],
6051	6114	MAX_MODEL_LEN));
@@ -6251,6 +6314,51 @@ static void ufshcd_clear_dbg_ufs_stats(struct ufs_hba *hba) · 199 unmodified lines
6251	6314	hba->req_abort_count = 0;
6252	6315	}
6253	6316
	6317	+static void ufshcd_init_desc_sizes(struct ufs_hba *hba)
	6318	+{
	6319	+ int err;
	6320	+
	6321	+ err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_DEVICE, 0,
	6322	+ &hba->desc_size.dev_desc);
	6323	+ if (err)
	6324	+ hba->desc_size.dev_desc = QUERY_DESC_DEVICE_DEF_SIZE;
	6325	+
	6326	+ err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_POWER, 0,
	6327	+ &hba->desc_size.pwr_desc);
	6328	+ if (err)
	6329	+ hba->desc_size.pwr_desc = QUERY_DESC_POWER_DEF_SIZE;
	6330	+
	6331	+ err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_INTERCONNECT, 0,
	6332	+ &hba->desc_size.interc_desc);
	6333	+ if (err)
	6334	+ hba->desc_size.interc_desc = QUERY_DESC_INTERCONNECT_DEF_SIZE;
	6335	+
	6336	+ err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_CONFIGURATION, 0,
	6337	+ &hba->desc_size.conf_desc);
	6338	+ if (err)
	6339	+ hba->desc_size.conf_desc = QUERY_DESC_CONFIGURATION_DEF_SIZE;
	6340	+
	6341	+ err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_UNIT, 0,
	6342	+ &hba->desc_size.unit_desc);
	6343	+ if (err)
	6344	+ hba->desc_size.unit_desc = QUERY_DESC_UNIT_DEF_SIZE;
	6345	+
	6346	+ err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_GEOMETRY, 0,
	6347	+ &hba->desc_size.geom_desc);
	6348	+ if (err)
	6349	+ hba->desc_size.geom_desc = QUERY_DESC_GEOMETRY_DEF_SIZE;
	6350	+}
	6351	+
	6352	+static void ufshcd_def_desc_sizes(struct ufs_hba *hba)
	6353	+{
	6354	+ hba->desc_size.dev_desc = QUERY_DESC_DEVICE_DEF_SIZE;
	6355	+ hba->desc_size.pwr_desc = QUERY_DESC_POWER_DEF_SIZE;
	6356	+ hba->desc_size.interc_desc = QUERY_DESC_INTERCONNECT_DEF_SIZE;
	6357	+ hba->desc_size.conf_desc = QUERY_DESC_CONFIGURATION_DEF_SIZE;
	6358	+ hba->desc_size.unit_desc = QUERY_DESC_UNIT_DEF_SIZE;
	6359	+ hba->desc_size.geom_desc = QUERY_DESC_GEOMETRY_DEF_SIZE;
	6360	+}
	6361	+
6254	6362	/**
6255	6363	* ufshcd_probe_hba - probe hba to detect device and initialize
6256	6364	* @hba: per-adapter instance
@@ -6285,6 +6393,9 @@ static int ufshcd_probe_hba(struct ufs_hba *hba) · 28 unmodified lines
6285	6393	if (ret)
6286	6394	goto out;
6287	6395
	6396	+ /* Init check for device descriptor sizes */
	6397	+ ufshcd_init_desc_sizes(hba);
	6398	+
6288	6399	ret = ufs_get_device_desc(hba, &card);
6289	6400	if (ret) {
6290	6401	dev_err(hba->dev, "%s: Failed getting device info. err = %d\n",
@@ -6320,6 +6431,7 @@ static int ufshcd_probe_hba(struct ufs_hba *hba) · 29 unmodified lines
6320	6431
6321	6432	/* set the state as operational after switching to desired gear */
6322	6433	hba->ufshcd_state = UFSHCD_STATE_OPERATIONAL;
	6434	+
6323	6435	/*
6324	6436	* If we are in error handling context or in power management callbacks
6325	6437	* context, no need to scan the host
@@ -7774,6 +7886,9 @@ int ufshcd_init(struct ufs_hba hba, void __iomem mmio_base, unsigned int irq) · 1448 unmodified lines
7774	7886	hba->mmio_base = mmio_base;
7775	7887	hba->irq = irq;
7776	7888
	7889	+ /* Set descriptor lengths to specification defaults */
	7890	+ ufshcd_def_desc_sizes(hba);
	7891	+
7777	7892	err = ufshcd_hba_init(hba);
7778	7893	if (err)
7779	7894	goto out_error;

drivers/scsi/ufs/ufshcd.h+15−0

@@ -220,6 +220,15 @@ struct ufs_dev_cmd { · 219 unmodified lines
220	220	struct ufs_query query;
221	221	};
222	222
	223	+struct ufs_desc_size {
	224	+ int dev_desc;
	225	+ int pwr_desc;
	226	+ int geom_desc;
	227	+ int interc_desc;
	228	+ int unit_desc;
	229	+ int conf_desc;
	230	+};
	231	+
223	232	/**
224	233	* struct ufs_clk_info - UFS clock related info
225	234	* @list: list headed by hba->clk_list_head
@@ -483,6 +492,7 @@ struct ufs_stats { · 257 unmodified lines
483	492	* @clk_list_head: UFS host controller clocks list node head
484	493	* @pwr_info: holds current power mode
485	494	* @max_pwr_info: keeps the device max valid pwm
	495	+ * @desc_size: descriptor sizes reported by device
486	496	* @urgent_bkops_lvl: keeps track of urgent bkops level for device
487	497	* @is_urgent_bkops_lvl_checked: keeps track if the urgent bkops level for
488	498	* device is known or not.
@@ -666,6 +676,7 @@ struct ufs_hba { · 177 unmodified lines
666	676	bool is_urgent_bkops_lvl_checked;
667	677
668	678	struct rw_semaphore clk_scaling_lock;
	679	+ struct ufs_desc_size desc_size;
669	680	};
670	681
671	682	/* Returns true if clocks can be gated. Otherwise false */
@@ -832,6 +843,10 @@ int ufshcd_query_flag(struct ufs_hba *hba, enum query_opcode opcode, · 160 unmodified lines
832	843	enum flag_idn idn, bool *flag_res);
833	844	int ufshcd_hold(struct ufs_hba *hba, bool async);
834	845	void ufshcd_release(struct ufs_hba *hba);
	846	+
	847	+int ufshcd_map_desc_id_to_length(struct ufs_hba *hba, enum desc_idn desc_id,
	848	+ int *desc_length);
	849	+
835	850	u32 ufshcd_get_local_unipro_ver(struct ufs_hba *hba);
836	851
837	852	/* Wrapper functions for safely calling variant operations */

drivers/scsi/vmw_pvscsi.c+1−1

@@ -1487,7 +1487,7 @@ static int pvscsi_probe(struct pci_dev pdev, const struct pci_device_id id) · 1486 unmodified lines
1487	1487	irq_flag &= ~PCI_IRQ_MSI;
1488	1488
1489	1489	error = pci_alloc_irq_vectors(adapter->dev, 1, 1, irq_flag);
1490		− if (error)
	1490	+ if (error < 0)
1491	1491	goto out_reset_adapter;
1492	1492
1493	1493	adapter->use_req_threshold = pvscsi_setup_req_threshold(adapter, true);

drivers/staging/lustre/lnet/lnet/lib-socket.c+2−2

@@ -532,7 +532,7 @@ lnet_sock_accept(struct socket *newsockp, struct socket sock) · 531 unmodified lines
532	532
533	533	newsock->ops = sock->ops;
534	534
535		− rc = sock->ops->accept(sock, newsock, O_NONBLOCK);
	535	+ rc = sock->ops->accept(sock, newsock, O_NONBLOCK, false);
536	536	if (rc == -EAGAIN) {
537	537	/* Nothing ready, so wait for activity */
538	538	init_waitqueue_entry(&wait, current);
@@ -540,7 +540,7 @@ lnet_sock_accept(struct socket *newsockp, struct socket sock) · 1 unmodified line
540	540	set_current_state(TASK_INTERRUPTIBLE);
541	541	schedule();
542	542	remove_wait_queue(sk_sleep(sock->sk), &wait);
543		− rc = sock->ops->accept(sock, newsock, O_NONBLOCK);
	543	+ rc = sock->ops->accept(sock, newsock, O_NONBLOCK, false);
544	544	}
545	545
546	546	if (rc)

drivers/staging/octeon/ethernet-rx.c+0−1

@@ -336,7 +336,6 @@ static int cvm_oct_poll(struct oct_rx_group *rx_group, int budget) · 335 unmodified lines
336	336	if (likely((port < TOTAL_NUMBER_OF_PORTS) &&
337	337	cvm_oct_device[port])) {
338	338	struct net_device *dev = cvm_oct_device[port];
339		− struct octeon_ethernet *priv = netdev_priv(dev);
340	339
341	340	/*
342	341	* Only accept packets for devices that are

drivers/staging/vc04_services/Kconfig+1−0

@@ -1,6 +1,7 @@
1	1	config BCM2835_VCHIQ
2	2	tristate "Videocore VCHIQ"
3	3	depends on HAS_DMA
	4	+ depends on OF
4	5	depends on RASPBERRYPI_FIRMWARE \|\| (COMPILE_TEST && !RASPBERRYPI_FIRMWARE)
5	6	default y
6	7	help

drivers/tty/n_hdlc.c+69−63

@@ -114,7 +114,7 @@ · 113 unmodified lines
114	114	#define DEFAULT_TX_BUF_COUNT 3
115	115
116	116	struct n_hdlc_buf {
117		− struct n_hdlc_buf *link;
	117	+ struct list_head list_item;
118	118	int count;
119	119	char buf[1];
120	120	};
@@ -122,8 +122,7 @@ struct n_hdlc_buf { · 1 unmodified line
122	122	#define N_HDLC_BUF_SIZE (sizeof(struct n_hdlc_buf) + maxframe)
123	123
124	124	struct n_hdlc_buf_list {
125		− struct n_hdlc_buf *head;
126		− struct n_hdlc_buf *tail;
	125	+ struct list_head list;
127	126	int count;
128	127	spinlock_t spinlock;
129	128	};
@@ -136,7 +135,6 @@ struct n_hdlc_buf_list { · 6 unmodified lines
136	135	* @backup_tty - TTY to use if tty gets closed
137	136	* @tbusy - reentrancy flag for tx wakeup code
138	137	* @woke_up - FIXME: describe this field
139		− * @tbuf - currently transmitting tx buffer
140	138	* @tx_buf_list - list of pending transmit frame buffers
141	139	* @rx_buf_list - list of received frame buffers
142	140	* @tx_free_buf_list - list unused transmit frame buffers
@@ -149,7 +147,6 @@ struct n_hdlc { · 6 unmodified lines
149	147	struct tty_struct *backup_tty;
150	148	int tbusy;
151	149	int woke_up;
152		− struct n_hdlc_buf *tbuf;
153	150	struct n_hdlc_buf_list tx_buf_list;
154	151	struct n_hdlc_buf_list rx_buf_list;
155	152	struct n_hdlc_buf_list tx_free_buf_list;
@@ -159,6 +156,8 @@ struct n_hdlc { · 3 unmodified lines
159	156	/*
160	157	* HDLC buffer list manipulation functions
161	158	*/
	159	+static void n_hdlc_buf_return(struct n_hdlc_buf_list *buf_list,
	160	+ struct n_hdlc_buf *buf);
162	161	static void n_hdlc_buf_put(struct n_hdlc_buf_list *list,
163	162	struct n_hdlc_buf *buf);
164	163	static struct n_hdlc_buf n_hdlc_buf_get(struct n_hdlc_buf_list list);
@@ -208,16 +207,9 @@ static void flush_tx_queue(struct tty_struct *tty) · 43 unmodified lines
208	207	{
209	208	struct n_hdlc *n_hdlc = tty2n_hdlc(tty);
210	209	struct n_hdlc_buf *buf;
211		− unsigned long flags;
212	210
213	211	while ((buf = n_hdlc_buf_get(&n_hdlc->tx_buf_list)))
214	212	n_hdlc_buf_put(&n_hdlc->tx_free_buf_list, buf);
215		− spin_lock_irqsave(&n_hdlc->tx_buf_list.spinlock, flags);
216		− if (n_hdlc->tbuf) {
217		− n_hdlc_buf_put(&n_hdlc->tx_free_buf_list, n_hdlc->tbuf);
218		− n_hdlc->tbuf = NULL;
219		− }
220		− spin_unlock_irqrestore(&n_hdlc->tx_buf_list.spinlock, flags);
221	213	}
222	214
223	215	static struct tty_ldisc_ops n_hdlc_ldisc = {
@@ -283,7 +275,6 @@ static void n_hdlc_release(struct n_hdlc *n_hdlc) · 59 unmodified lines
283	275	} else
284	276	break;
285	277	}
286		− kfree(n_hdlc->tbuf);
287	278	kfree(n_hdlc);
288	279
289	280	} /* end of n_hdlc_release() */
@@ -402,13 +393,7 @@ static void n_hdlc_send_frames(struct n_hdlc n_hdlc, struct tty_struct tty) · 112 unmodified lines
402	393	n_hdlc->woke_up = 0;
403	394	spin_unlock_irqrestore(&n_hdlc->tx_buf_list.spinlock, flags);
404	395
405		− /* get current transmit buffer or get new transmit */
406		− /* buffer from list of pending transmit buffers */
407		−
408		− tbuf = n_hdlc->tbuf;
409		− if (!tbuf)
410		− tbuf = n_hdlc_buf_get(&n_hdlc->tx_buf_list);
411		−
	396	+ tbuf = n_hdlc_buf_get(&n_hdlc->tx_buf_list);
412	397	while (tbuf) {
413	398	if (debuglevel >= DEBUG_LEVEL_INFO)
414	399	printk("%s(%d)sending frame %p, count=%d\n",
@@ -420,7 +405,7 @@ static void n_hdlc_send_frames(struct n_hdlc n_hdlc, struct tty_struct tty) · 5 unmodified lines
420	405
421	406	/* rollback was possible and has been done */
422	407	if (actual == -ERESTARTSYS) {
423		− n_hdlc->tbuf = tbuf;
	408	+ n_hdlc_buf_return(&n_hdlc->tx_buf_list, tbuf);
424	409	break;
425	410	}
426	411	/* if transmit error, throw frame away by */
@@ -435,10 +420,7 @@ static void n_hdlc_send_frames(struct n_hdlc n_hdlc, struct tty_struct tty) · 8 unmodified lines
435	420
436	421	/* free current transmit buffer */
437	422	n_hdlc_buf_put(&n_hdlc->tx_free_buf_list, tbuf);
438		−
439		− /* this tx buffer is done */
440		− n_hdlc->tbuf = NULL;
441		−
	423	+
442	424	/* wait up sleeping writers */
443	425	wake_up_interruptible(&tty->write_wait);
444	426
@@ -448,10 +430,12 @@ static void n_hdlc_send_frames(struct n_hdlc n_hdlc, struct tty_struct tty) · 3 unmodified lines
448	430	if (debuglevel >= DEBUG_LEVEL_INFO)
449	431	printk("%s(%d)frame %p pending\n",
450	432	__FILE__,__LINE__,tbuf);
451		−
452		− /* buffer not accepted by driver */
453		− /* set this buffer as pending buffer */
454		− n_hdlc->tbuf = tbuf;
	433	+
	434	+ /*
	435	+ * the buffer was not accepted by driver,
	436	+ * return it back into tx queue
	437	+ */
	438	+ n_hdlc_buf_return(&n_hdlc->tx_buf_list, tbuf);
455	439	break;
456	440	}
457	441	}
@@ -749,7 +733,8 @@ static int n_hdlc_tty_ioctl(struct tty_struct tty, struct file file, · 291 unmodified lines
749	733	int error = 0;
750	734	int count;
751	735	unsigned long flags;
752		−
	736	+ struct n_hdlc_buf *buf = NULL;
	737	+
753	738	if (debuglevel >= DEBUG_LEVEL_INFO)
754	739	printk("%s(%d)n_hdlc_tty_ioctl() called %d\n",
755	740	__FILE__,__LINE__,cmd);
@@ -763,8 +748,10 @@ static int n_hdlc_tty_ioctl(struct tty_struct tty, struct file file, · 7 unmodified lines
763	748	/* report count of read data available */
764	749	/* in next available frame (if any) */
765	750	spin_lock_irqsave(&n_hdlc->rx_buf_list.spinlock,flags);
766		− if (n_hdlc->rx_buf_list.head)
767		− count = n_hdlc->rx_buf_list.head->count;
	751	+ buf = list_first_entry_or_null(&n_hdlc->rx_buf_list.list,
	752	+ struct n_hdlc_buf, list_item);
	753	+ if (buf)
	754	+ count = buf->count;
768	755	else
769	756	count = 0;
770	757	spin_unlock_irqrestore(&n_hdlc->rx_buf_list.spinlock,flags);
@@ -776,8 +763,10 @@ static int n_hdlc_tty_ioctl(struct tty_struct tty, struct file file, · 5 unmodified lines
776	763	count = tty_chars_in_buffer(tty);
777	764	/* add size of next output frame in queue */
778	765	spin_lock_irqsave(&n_hdlc->tx_buf_list.spinlock,flags);
779		− if (n_hdlc->tx_buf_list.head)
780		− count += n_hdlc->tx_buf_list.head->count;
	766	+ buf = list_first_entry_or_null(&n_hdlc->tx_buf_list.list,
	767	+ struct n_hdlc_buf, list_item);
	768	+ if (buf)
	769	+ count += buf->count;
781	770	spin_unlock_irqrestore(&n_hdlc->tx_buf_list.spinlock,flags);
782	771	error = put_user(count, (int __user *)arg);
783	772	break;
@@ -825,14 +814,14 @@ static unsigned int n_hdlc_tty_poll(struct tty_struct tty, struct file filp, · 41 unmodified lines
825	814	poll_wait(filp, &tty->write_wait, wait);
826	815
827	816	/* set bits for operations that won't block */
828		− if (n_hdlc->rx_buf_list.head)
	817	+ if (!list_empty(&n_hdlc->rx_buf_list.list))
829	818	mask \|= POLLIN \| POLLRDNORM; /* readable */
830	819	if (test_bit(TTY_OTHER_CLOSED, &tty->flags))
831	820	mask \|= POLLHUP;
832	821	if (tty_hung_up_p(filp))
833	822	mask \|= POLLHUP;
834	823	if (!tty_is_writelocked(tty) &&
835		− n_hdlc->tx_free_buf_list.head)
	824	+ !list_empty(&n_hdlc->tx_free_buf_list.list))
836	825	mask \|= POLLOUT \| POLLWRNORM; /* writable */
837	826	}
838	827	return mask;
@@ -856,7 +845,12 @@ static struct n_hdlc *n_hdlc_alloc(void) · 17 unmodified lines
856	845	spin_lock_init(&n_hdlc->tx_free_buf_list.spinlock);
857	846	spin_lock_init(&n_hdlc->rx_buf_list.spinlock);
858	847	spin_lock_init(&n_hdlc->tx_buf_list.spinlock);
859		−
	848	+
	849	+ INIT_LIST_HEAD(&n_hdlc->rx_free_buf_list.list);
	850	+ INIT_LIST_HEAD(&n_hdlc->tx_free_buf_list.list);
	851	+ INIT_LIST_HEAD(&n_hdlc->rx_buf_list.list);
	852	+ INIT_LIST_HEAD(&n_hdlc->tx_buf_list.list);
	853	+
860	854	/* allocate free rx buffer list */
861	855	for(i=0;i<DEFAULT_RX_BUF_COUNT;i++) {
862	856	buf = kmalloc(N_HDLC_BUF_SIZE, GFP_KERNEL);
@@ -883,54 +877,66 @@ static struct n_hdlc *n_hdlc_alloc(void) · 20 unmodified lines
883	877
884	878	} /* end of n_hdlc_alloc() */
885	879
	880	+/**
	881	+ * n_hdlc_buf_return - put the HDLC buffer after the head of the specified list
	882	+ * @buf_list - pointer to the buffer list
	883	+ * @buf - pointer to the buffer
	884	+ */
	885	+static void n_hdlc_buf_return(struct n_hdlc_buf_list *buf_list,
	886	+ struct n_hdlc_buf *buf)
	887	+{
	888	+ unsigned long flags;
	889	+
	890	+ spin_lock_irqsave(&buf_list->spinlock, flags);
	891	+
	892	+ list_add(&buf->list_item, &buf_list->list);
	893	+ buf_list->count++;
	894	+
	895	+ spin_unlock_irqrestore(&buf_list->spinlock, flags);
	896	+}
	897	+
886	898	/**
887	899	* n_hdlc_buf_put - add specified HDLC buffer to tail of specified list
888		− * @list - pointer to buffer list
	900	+ * @buf_list - pointer to buffer list
889	901	* @buf - pointer to buffer
890	902	*/
891		−static void n_hdlc_buf_put(struct n_hdlc_buf_list *list,
	903	+static void n_hdlc_buf_put(struct n_hdlc_buf_list *buf_list,
892	904	struct n_hdlc_buf *buf)
893	905	{
894	906	unsigned long flags;
895		− spin_lock_irqsave(&list->spinlock,flags);
896		−
897		− buf->link=NULL;
898		− if (list->tail)
899		− list->tail->link = buf;
900		− else
901		− list->head = buf;
902		− list->tail = buf;
903		− (list->count)++;
904		−
905		− spin_unlock_irqrestore(&list->spinlock,flags);
906		−
	907	+
	908	+ spin_lock_irqsave(&buf_list->spinlock, flags);
	909	+
	910	+ list_add_tail(&buf->list_item, &buf_list->list);
	911	+ buf_list->count++;
	912	+
	913	+ spin_unlock_irqrestore(&buf_list->spinlock, flags);
907	914	} /* end of n_hdlc_buf_put() */
908	915
909	916	/**
910	917	* n_hdlc_buf_get - remove and return an HDLC buffer from list
911		− * @list - pointer to HDLC buffer list
	918	+ * @buf_list - pointer to HDLC buffer list
912	919	*
913	920	* Remove and return an HDLC buffer from the head of the specified HDLC buffer
914	921	* list.
915	922	* Returns a pointer to HDLC buffer if available, otherwise %NULL.
916	923	*/
917		−static struct n_hdlc_buf* n_hdlc_buf_get(struct n_hdlc_buf_list *list)
	924	+static struct n_hdlc_buf n_hdlc_buf_get(struct n_hdlc_buf_list buf_list)
918	925	{
919	926	unsigned long flags;
920	927	struct n_hdlc_buf *buf;
921		− spin_lock_irqsave(&list->spinlock,flags);
922		−
923		− buf = list->head;
	928	+
	929	+ spin_lock_irqsave(&buf_list->spinlock, flags);
	930	+
	931	+ buf = list_first_entry_or_null(&buf_list->list,
	932	+ struct n_hdlc_buf, list_item);
924	933	if (buf) {
925		− list->head = buf->link;
926		− (list->count)--;
	934	+ list_del(&buf->list_item);
	935	+ buf_list->count--;
927	936	}
928		− if (!list->head)
929		− list->tail = NULL;
930		−
931		− spin_unlock_irqrestore(&list->spinlock,flags);
	937	+
	938	+ spin_unlock_irqrestore(&buf_list->spinlock, flags);
932	939	return buf;
933		−
934	940	} /* end of n_hdlc_buf_get() */
935	941
936	942	static char hdlc_banner[] __initdata =

drivers/tty/serial/samsung.c+4−2

@@ -1031,8 +1031,10 @@ static int s3c64xx_serial_startup(struct uart_port *port) · 1030 unmodified lines
1031	1031	if (ourport->dma) {
1032	1032	ret = s3c24xx_serial_request_dma(ourport);
1033	1033	if (ret < 0) {
1034		− dev_warn(port->dev, "DMA request failed\n");
1035		− return ret;
	1034	+ dev_warn(port->dev,
	1035	+ "DMA request failed, DMA will not be used\n");
	1036	+ devm_kfree(port->dev, ourport->dma);
	1037	+ ourport->dma = NULL;
1036	1038	}
1037	1039	}
1038	1040

drivers/usb/dwc3/dwc3-omap.c+2−1

@@ -250,6 +250,7 @@ static void dwc3_omap_set_mailbox(struct dwc3_omap *omap, · 249 unmodified lines
250	250	val = dwc3_omap_read_utmi_ctrl(omap);
251	251	val \|= USBOTGSS_UTMI_OTG_CTRL_IDDIG;
252	252	dwc3_omap_write_utmi_ctrl(omap, val);
	253	+ break;
253	254
254	255	case OMAP_DWC3_VBUS_OFF:
255	256	val = dwc3_omap_read_utmi_ctrl(omap);
@@ -392,7 +393,7 @@ static void dwc3_omap_set_utmi_mode(struct dwc3_omap *omap) · 136 unmodified lines
392	393	{
393	394	u32 reg;
394	395	struct device_node *node = omap->dev->of_node;
395		− int utmi_mode = 0;
	396	+ u32 utmi_mode = 0;
396	397
397	398	reg = dwc3_omap_read_utmi_ctrl(omap);
398	399

drivers/usb/dwc3/gadget.c+67−9

@@ -1342,6 +1342,68 @@ static int dwc3_gadget_ep_dequeue(struct usb_ep *ep, · 1341 unmodified lines
1342	1342	if (r == req) {
1343	1343	/* wait until it is processed */
1344	1344	dwc3_stop_active_transfer(dwc, dep->number, true);
	1345	+
	1346	+ /*
	1347	+ * If request was already started, this means we had to
	1348	+ * stop the transfer. With that we also need to ignore
	1349	+ * all TRBs used by the request, however TRBs can only
	1350	+ * be modified after completion of END_TRANSFER
	1351	+ * command. So what we do here is that we wait for
	1352	+ * END_TRANSFER completion and only after that, we jump
	1353	+ * over TRBs by clearing HWO and incrementing dequeue
	1354	+ * pointer.
	1355	+ *
	1356	+ * Note that we have 2 possible types of transfers here:
	1357	+ *
	1358	+ * i) Linear buffer request
	1359	+ * ii) SG-list based request
	1360	+ *
	1361	+ * SG-list based requests will have r->num_pending_sgs
	1362	+ * set to a valid number (> 0). Linear requests,
	1363	+ * normally use a single TRB.
	1364	+ *
	1365	+ * For each of these two cases, if r->unaligned flag is
	1366	+ * set, one extra TRB has been used to align transfer
	1367	+ * size to wMaxPacketSize.
	1368	+ *
	1369	+ * All of these cases need to be taken into
	1370	+ * consideration so we don't mess up our TRB ring
	1371	+ * pointers.
	1372	+ */
	1373	+ wait_event_lock_irq(dep->wait_end_transfer,
	1374	+ !(dep->flags & DWC3_EP_END_TRANSFER_PENDING),
	1375	+ dwc->lock);
	1376	+
	1377	+ if (!r->trb)
	1378	+ goto out1;
	1379	+
	1380	+ if (r->num_pending_sgs) {
	1381	+ struct dwc3_trb *trb;
	1382	+ int i = 0;
	1383	+
	1384	+ for (i = 0; i < r->num_pending_sgs; i++) {
	1385	+ trb = r->trb + i;
	1386	+ trb->ctrl &= ~DWC3_TRB_CTRL_HWO;
	1387	+ dwc3_ep_inc_deq(dep);
	1388	+ }
	1389	+
	1390	+ if (r->unaligned) {
	1391	+ trb = r->trb + r->num_pending_sgs + 1;
	1392	+ trb->ctrl &= ~DWC3_TRB_CTRL_HWO;
	1393	+ dwc3_ep_inc_deq(dep);
	1394	+ }
	1395	+ } else {
	1396	+ struct dwc3_trb *trb = r->trb;
	1397	+
	1398	+ trb->ctrl &= ~DWC3_TRB_CTRL_HWO;
	1399	+ dwc3_ep_inc_deq(dep);
	1400	+
	1401	+ if (r->unaligned) {
	1402	+ trb = r->trb + 1;
	1403	+ trb->ctrl &= ~DWC3_TRB_CTRL_HWO;
	1404	+ dwc3_ep_inc_deq(dep);
	1405	+ }
	1406	+ }
1345	1407	goto out1;
1346	1408	}
1347	1409	dev_err(dwc->dev, "request %p was not queued to %s\n",
@@ -1352,6 +1414,7 @@ static int dwc3_gadget_ep_dequeue(struct usb_ep *ep, · 4 unmodified lines
1352	1414
1353	1415	out1:
1354	1416	/* giveback the request */
	1417	+ dep->queued_requests--;
1355	1418	dwc3_gadget_giveback(dep, req, -ECONNRESET);
1356	1419
1357	1420	out0:
@@ -2126,12 +2189,12 @@ static int __dwc3_cleanup_done_trbs(struct dwc3 dwc, struct dwc3_ep dep, · 768 unmodified lines
2126	2189	return 1;
2127	2190	}
2128	2191
2129		− if ((trb->ctrl & DWC3_TRB_CTRL_HWO) && status != -ESHUTDOWN)
2130		− return 1;
2131		−
2132	2192	count = trb->size & DWC3_TRB_SIZE_MASK;
2133	2193	req->remaining += count;
2134	2194
	2195	+ if ((trb->ctrl & DWC3_TRB_CTRL_HWO) && status != -ESHUTDOWN)
	2196	+ return 1;
	2197	+
2135	2198	if (dep->direction) {
2136	2199	if (count) {
2137	2200	trb_status = DWC3_TRB_SIZE_TRBSTS(trb->size);
@@ -3228,15 +3291,10 @@ void dwc3_gadget_exit(struct dwc3 *dwc) · 1090 unmodified lines
3228	3291
3229	3292	int dwc3_gadget_suspend(struct dwc3 *dwc)
3230	3293	{
3231		− int ret;
3232		−
3233	3294	if (!dwc->gadget_driver)
3234	3295	return 0;
3235	3296
3236		− ret = dwc3_gadget_run_stop(dwc, false, false);
3237		− if (ret < 0)
3238		− return ret;
3239		−
	3297	+ dwc3_gadget_run_stop(dwc, false, false);
3240	3298	dwc3_disconnect_gadget(dwc);
3241	3299	__dwc3_gadget_stop(dwc);
3242	3300

drivers/usb/dwc3/gadget.h+7−7

@@ -28,23 +28,23 @@ struct dwc3; · 27 unmodified lines
28	28	#define gadget_to_dwc(g) (container_of(g, struct dwc3, gadget))
29	29
30	30	/* DEPCFG parameter 1 */
31		−#define DWC3_DEPCFG_INT_NUM(n) ((n) << 0)
	31	+#define DWC3_DEPCFG_INT_NUM(n) (((n) & 0x1f) << 0)
32	32	#define DWC3_DEPCFG_XFER_COMPLETE_EN (1 << 8)
33	33	#define DWC3_DEPCFG_XFER_IN_PROGRESS_EN (1 << 9)
34	34	#define DWC3_DEPCFG_XFER_NOT_READY_EN (1 << 10)
35	35	#define DWC3_DEPCFG_FIFO_ERROR_EN (1 << 11)
36	36	#define DWC3_DEPCFG_STREAM_EVENT_EN (1 << 13)
37		−#define DWC3_DEPCFG_BINTERVAL_M1(n) ((n) << 16)
	37	+#define DWC3_DEPCFG_BINTERVAL_M1(n) (((n) & 0xff) << 16)
38	38	#define DWC3_DEPCFG_STREAM_CAPABLE (1 << 24)
39		−#define DWC3_DEPCFG_EP_NUMBER(n) ((n) << 25)
	39	+#define DWC3_DEPCFG_EP_NUMBER(n) (((n) & 0x1f) << 25)
40	40	#define DWC3_DEPCFG_BULK_BASED (1 << 30)
41	41	#define DWC3_DEPCFG_FIFO_BASED (1 << 31)
42	42
43	43	/* DEPCFG parameter 0 */
44		−#define DWC3_DEPCFG_EP_TYPE(n) ((n) << 1)
45		−#define DWC3_DEPCFG_MAX_PACKET_SIZE(n) ((n) << 3)
46		−#define DWC3_DEPCFG_FIFO_NUMBER(n) ((n) << 17)
47		−#define DWC3_DEPCFG_BURST_SIZE(n) ((n) << 22)
	44	+#define DWC3_DEPCFG_EP_TYPE(n) (((n) & 0x3) << 1)
	45	+#define DWC3_DEPCFG_MAX_PACKET_SIZE(n) (((n) & 0x7ff) << 3)
	46	+#define DWC3_DEPCFG_FIFO_NUMBER(n) (((n) & 0x1f) << 17)
	47	+#define DWC3_DEPCFG_BURST_SIZE(n) (((n) & 0xf) << 22)
48	48	#define DWC3_DEPCFG_DATA_SEQ_NUM(n) ((n) << 26)
49	49	/* This applies for core versions earlier than 1.94a */
50	50	#define DWC3_DEPCFG_IGN_SEQ_NUM (1 << 31)

drivers/usb/gadget/configfs.c+1−0

@@ -269,6 +269,7 @@ static ssize_t gadget_dev_desc_UDC_store(struct config_item *item, · 268 unmodified lines
269	269	ret = unregister_gadget(gi);
270	270	if (ret)
271	271	goto err;
	272	+ kfree(name);
272	273	} else {
273	274	if (gi->composite.gadget_driver.udc_name) {
274	275	ret = -EBUSY;

drivers/usb/gadget/function/f_fs.c+14−3

@@ -1834,11 +1834,14 @@ static int ffs_func_eps_enable(struct ffs_function *func) · 1833 unmodified lines
1834	1834	spin_lock_irqsave(&func->ffs->eps_lock, flags);
1835	1835	while(count--) {
1836	1836	struct usb_endpoint_descriptor *ds;
	1837	+ struct usb_ss_ep_comp_descriptor *comp_desc = NULL;
	1838	+ int needs_comp_desc = false;
1837	1839	int desc_idx;
1838	1840
1839		− if (ffs->gadget->speed == USB_SPEED_SUPER)
	1841	+ if (ffs->gadget->speed == USB_SPEED_SUPER) {
1840	1842	desc_idx = 2;
1841		− else if (ffs->gadget->speed == USB_SPEED_HIGH)
	1843	+ needs_comp_desc = true;
	1844	+ } else if (ffs->gadget->speed == USB_SPEED_HIGH)
1842	1845	desc_idx = 1;
1843	1846	else
1844	1847	desc_idx = 0;
@@ -1855,6 +1858,14 @@ static int ffs_func_eps_enable(struct ffs_function *func) · 10 unmodified lines
1855	1858
1856	1859	ep->ep->driver_data = ep;
1857	1860	ep->ep->desc = ds;
	1861	+
	1862	+ comp_desc = (struct usb_ss_ep_comp_descriptor *)(ds +
	1863	+ USB_DT_ENDPOINT_SIZE);
	1864	+ ep->ep->maxburst = comp_desc->bMaxBurst + 1;
	1865	+
	1866	+ if (needs_comp_desc)
	1867	+ ep->ep->comp_desc = comp_desc;
	1868	+
1858	1869	ret = usb_ep_enable(ep->ep);
1859	1870	if (likely(!ret)) {
1860	1871	epfile->ep = ep;
@@ -2253,7 +2264,7 @@ static int __ffs_data_do_os_desc(enum ffs_os_desc_type type, · 392 unmodified lines
2253	2264
2254	2265	if (len < sizeof(*d) \|\|
2255	2266	d->bFirstInterfaceNumber >= ffs->interfaces_count \|\|
2256		− d->Reserved1)
	2267	+ !d->Reserved1)
2257	2268	return -EINVAL;
2258	2269	for (i = 0; i < ARRAY_SIZE(d->Reserved2); ++i)
2259	2270	if (d->Reserved2[i])

drivers/usb/gadget/function/f_uvc.c+0−7

@@ -258,13 +258,6 @@ uvc_function_setup(struct usb_function f, const struct usb_ctrlrequest ctrl) · 257 unmodified lines
258	258	memcpy(&uvc_event->req, ctrl, sizeof(uvc_event->req));
259	259	v4l2_event_queue(&uvc->vdev, &v4l2_event);
260	260
261		− /* Pass additional setup data to userspace */
262		− if (uvc->event_setup_out && uvc->event_length) {
263		− uvc->control_req->length = uvc->event_length;
264		− return usb_ep_queue(uvc->func.config->cdev->gadget->ep0,
265		− uvc->control_req, GFP_ATOMIC);
266		− }
267		−
268	261	return 0;
269	262	}
270	263

drivers/usb/gadget/legacy/inode.c+4−3

@@ -84,8 +84,7 @@ static int ep_open(struct inode , struct file ); · 83 unmodified lines
84	84
85	85	/* /dev/gadget/$CHIP represents ep0 and the whole device */
86	86	enum ep0_state {
87		− /* DISBLED is the initial state.
88		− */
	87	+ /* DISABLED is the initial state. */
89	88	STATE_DEV_DISABLED = 0,
90	89
91	90	/* Only one open() of /dev/gadget/$CHIP; only one file tracks
@@ -1782,8 +1781,10 @@ dev_config (struct file fd, const char __user buf, size_t len, loff_t *ptr) · 1690 unmodified lines
1782	1781
1783	1782	spin_lock_irq (&dev->lock);
1784	1783	value = -EINVAL;
1785		− if (dev->buf)
	1784	+ if (dev->buf) {
	1785	+ kfree(kbuf);
1786	1786	goto fail;
	1787	+ }
1787	1788	dev->buf = kbuf;
1788	1789
1789	1790	/* full or low speed config */

drivers/usb/gadget/udc/atmel_usba_udc.c+2−2

@@ -610,7 +610,7 @@ usba_ep_enable(struct usb_ep _ep, const struct usb_endpoint_descriptor desc) · 609 unmodified lines
610	610	{
611	611	struct usba_ep *ep = to_usba_ep(_ep);
612	612	struct usba_udc *udc = ep->udc;
613		− unsigned long flags, ept_cfg, maxpacket;
	613	+ unsigned long flags, maxpacket;
614	614	unsigned int nr_trans;
615	615
616	616	DBG(DBG_GADGET, "%s: ep_enable: desc=%p\n", ep->ep.name, desc);
@@ -630,7 +630,7 @@ usba_ep_enable(struct usb_ep _ep, const struct usb_endpoint_descriptor desc) · 13 unmodified lines
630	630	ep->is_in = 0;
631	631
632	632	DBG(DBG_ERR, "%s: EPT_CFG = 0x%lx (maxpacket = %lu)\n",
633		− ep->ep.name, ept_cfg, maxpacket);
	633	+ ep->ep.name, ep->ept_cfg, maxpacket);
634	634
635	635	if (usb_endpoint_dir_in(desc)) {
636	636	ep->is_in = 1;

drivers/usb/gadget/udc/dummy_hcd.c+2−0

@@ -1031,6 +1031,8 @@ static int dummy_udc_probe(struct platform_device *pdev) · 1030 unmodified lines
1031	1031	int rc;
1032	1032
1033	1033	dum = ((void *)dev_get_platdata(&pdev->dev));
	1034	+ /* Clear usb_gadget region for new registration to udc-core */
	1035	+ memzero_explicit(&dum->gadget, sizeof(struct usb_gadget));
1034	1036	dum->gadget.name = gadget_name;
1035	1037	dum->gadget.ops = &dummy_ops;
1036	1038	dum->gadget.max_speed = USB_SPEED_SUPER;

drivers/usb/gadget/udc/net2280.c+13−12

@@ -1146,15 +1146,15 @@ static int scan_dma_completions(struct net2280_ep *ep) · 1145 unmodified lines
1146	1146	*/
1147	1147	while (!list_empty(&ep->queue)) {
1148	1148	struct net2280_request *req;
1149		− u32 tmp;
	1149	+ u32 req_dma_count;
1150	1150
1151	1151	req = list_entry(ep->queue.next,
1152	1152	struct net2280_request, queue);
1153	1153	if (!req->valid)
1154	1154	break;
1155	1155	rmb();
1156		− tmp = le32_to_cpup(&req->td->dmacount);
1157		− if ((tmp & BIT(VALID_BIT)) != 0)
	1156	+ req_dma_count = le32_to_cpup(&req->td->dmacount);
	1157	+ if ((req_dma_count & BIT(VALID_BIT)) != 0)
1158	1158	break;
1159	1159
1160	1160	/* SHORT_PACKET_TRANSFERRED_INTERRUPT handles "usb-short"
@@ -1163,40 +1163,41 @@ static int scan_dma_completions(struct net2280_ep *ep) · 2 unmodified lines
1163	1163	*/
1164	1164	if (unlikely(req->td->dmadesc == 0)) {
1165	1165	/* paranoia */
1166		− tmp = readl(&ep->dma->dmacount);
1167		− if (tmp & DMA_BYTE_COUNT_MASK)
	1166	+ u32 const ep_dmacount = readl(&ep->dma->dmacount);
	1167	+
	1168	+ if (ep_dmacount & DMA_BYTE_COUNT_MASK)
1168	1169	break;
1169	1170	/* single transfer mode */
1170		− dma_done(ep, req, tmp, 0);
	1171	+ dma_done(ep, req, req_dma_count, 0);
1171	1172	num_completed++;
1172	1173	break;
1173	1174	} else if (!ep->is_in &&
1174	1175	(req->req.length % ep->ep.maxpacket) &&
1175	1176	!(ep->dev->quirks & PLX_PCIE)) {
1176	1177
1177		− tmp = readl(&ep->regs->ep_stat);
	1178	+ u32 const ep_stat = readl(&ep->regs->ep_stat);
1178	1179	/* AVOID TROUBLE HERE by not issuing short reads from
1179	1180	* your gadget driver. That helps avoids errata 0121,
1180	1181	* 0122, and 0124; not all cases trigger the warning.
1181	1182	*/
1182		− if ((tmp & BIT(NAK_OUT_PACKETS)) == 0) {
	1183	+ if ((ep_stat & BIT(NAK_OUT_PACKETS)) == 0) {
1183	1184	ep_warn(ep->dev, "%s lost packet sync!\n",
1184	1185	ep->ep.name);
1185	1186	req->req.status = -EOVERFLOW;
1186	1187	} else {
1187		− tmp = readl(&ep->regs->ep_avail);
1188		− if (tmp) {
	1188	+ u32 const ep_avail = readl(&ep->regs->ep_avail);
	1189	+ if (ep_avail) {
1189	1190	/* fifo gets flushed later */
1190	1191	ep->out_overflow = 1;
1191	1192	ep_dbg(ep->dev,
1192	1193	"%s dma, discard %d len %d\n",
1193		− ep->ep.name, tmp,
	1194	+ ep->ep.name, ep_avail,
1194	1195	req->req.length);
1195	1196	req->req.status = -EOVERFLOW;
1196	1197	}
1197	1198	}
1198	1199	}
1199		− dma_done(ep, req, tmp, 0);
	1200	+ dma_done(ep, req, req_dma_count, 0);
1200	1201	num_completed++;
1201	1202	}
1202	1203

drivers/usb/gadget/udc/pxa27x_udc.c+3−2

@@ -2534,9 +2534,10 @@ static int pxa_udc_remove(struct platform_device *_dev) · 2533 unmodified lines
2534	2534	usb_del_gadget_udc(&udc->gadget);
2535	2535	pxa_cleanup_debugfs(udc);
2536	2536
2537		− if (!IS_ERR_OR_NULL(udc->transceiver))
	2537	+ if (!IS_ERR_OR_NULL(udc->transceiver)) {
2538	2538	usb_unregister_notifier(udc->transceiver, &pxa27x_udc_phy);
2539		− usb_put_phy(udc->transceiver);
	2539	+ usb_put_phy(udc->transceiver);
	2540	+ }
2540	2541
2541	2542	udc->transceiver = NULL;
2542	2543	the_controller = NULL;

drivers/usb/host/ohci-at91.c+2−2

@@ -350,7 +350,7 @@ static int ohci_at91_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, · 349 unmodified lines
350	350
351	351	case USB_PORT_FEAT_SUSPEND:
352	352	dev_dbg(hcd->self.controller, "SetPortFeat: SUSPEND\n");
353		− if (valid_port(wIndex)) {
	353	+ if (valid_port(wIndex) && ohci_at91->sfr_regmap) {
354	354	ohci_at91_port_suspend(ohci_at91->sfr_regmap,
355	355	1);
356	356	return 0;
@@ -393,7 +393,7 @@ static int ohci_at91_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, · 36 unmodified lines
393	393
394	394	case USB_PORT_FEAT_SUSPEND:
395	395	dev_dbg(hcd->self.controller, "ClearPortFeature: SUSPEND\n");
396		− if (valid_port(wIndex)) {
	396	+ if (valid_port(wIndex) && ohci_at91->sfr_regmap) {
397	397	ohci_at91_port_suspend(ohci_at91->sfr_regmap,
398	398	0);
399	399	return 0;

drivers/usb/host/xhci-dbg.c+1−1

@@ -109,7 +109,7 @@ static void xhci_print_cap_regs(struct xhci_hcd *xhci) · 108 unmodified lines
109	109	xhci_dbg(xhci, "RTSOFF 0x%x:\n", temp & RTSOFF_MASK);
110	110
111	111	/* xhci 1.1 controllers have the HCCPARAMS2 register */
112		− if (hci_version > 100) {
	112	+ if (hci_version > 0x100) {
113	113	temp = readl(&xhci->cap_regs->hcc_params2);
114	114	xhci_dbg(xhci, "HCC PARAMS2 0x%x:\n", (unsigned int) temp);
115	115	xhci_dbg(xhci, " HC %s Force save context capability",

drivers/usb/host/xhci-mtk.c+3−4

@@ -382,7 +382,6 @@ static int usb_wakeup_of_property_parse(struct xhci_hcd_mtk *mtk, · 381 unmodified lines
382	382
383	383	static int xhci_mtk_setup(struct usb_hcd *hcd);
384	384	static const struct xhci_driver_overrides xhci_mtk_overrides __initconst = {
385		− .extra_priv_size = sizeof(struct xhci_hcd),
386	385	.reset = xhci_mtk_setup,
387	386	};
388	387
@@ -678,13 +677,13 @@ static int xhci_mtk_probe(struct platform_device *pdev) · 289 unmodified lines
678	677	goto power_off_phys;
679	678	}
680	679
681		− if (HCC_MAX_PSA(xhci->hcc_params) >= 4)
682		− xhci->shared_hcd->can_do_streams = 1;
683		−
684	680	ret = usb_add_hcd(hcd, irq, IRQF_SHARED);
685	681	if (ret)
686	682	goto put_usb3_hcd;
687	683
	684	+ if (HCC_MAX_PSA(xhci->hcc_params) >= 4)
	685	+ xhci->shared_hcd->can_do_streams = 1;
	686	+
688	687	ret = usb_add_hcd(xhci->shared_hcd, irq, IRQF_SHARED);
689	688	if (ret)
690	689	goto dealloc_usb2_hcd;

drivers/usb/host/xhci-plat.c+2−0

@@ -286,6 +286,8 @@ static int xhci_plat_remove(struct platform_device *dev) · 285 unmodified lines
286	286	struct xhci_hcd *xhci = hcd_to_xhci(hcd);
287	287	struct clk *clk = xhci->clk;
288	288
	289	+ xhci->xhc_state \|= XHCI_STATE_REMOVING;
	290	+
289	291	usb_remove_hcd(xhci->shared_hcd);
290	292	usb_phy_shutdown(hcd->usb_phy);
291	293

drivers/usb/host/xhci-tegra.c+0−1

@@ -1308,7 +1308,6 @@ static int tegra_xhci_setup(struct usb_hcd *hcd) · 1307 unmodified lines
1308	1308	}
1309	1309
1310	1310	static const struct xhci_driver_overrides tegra_xhci_overrides __initconst = {
1311		− .extra_priv_size = sizeof(struct xhci_hcd),
1312	1311	.reset = tegra_xhci_setup,
1313	1312	};
1314	1313

drivers/usb/host/xhci.c+2−2

@@ -868,7 +868,7 @@ static void xhci_disable_port_wake_on_bits(struct xhci_hcd *xhci) · 867 unmodified lines
868	868
869	869	spin_lock_irqsave(&xhci->lock, flags);
870	870
871		− /* disble usb3 ports Wake bits*/
	871	+ /* disable usb3 ports Wake bits */
872	872	port_index = xhci->num_usb3_ports;
873	873	port_array = xhci->usb3_ports;
874	874	while (port_index--) {
@@ -879,7 +879,7 @@ static void xhci_disable_port_wake_on_bits(struct xhci_hcd *xhci) · 4 unmodified lines
879	879	writel(t2, port_array[port_index]);
880	880	}
881	881
882		− /* disble usb2 ports Wake bits*/
	882	+ /* disable usb2 ports Wake bits */
883	883	port_index = xhci->num_usb2_ports;
884	884	port_array = xhci->usb2_ports;
885	885	while (port_index--) {

drivers/usb/misc/iowarrior.c+15−6

@@ -781,12 +781,6 @@ static int iowarrior_probe(struct usb_interface *interface, · 780 unmodified lines
781	781	iface_desc = interface->cur_altsetting;
782	782	dev->product_id = le16_to_cpu(udev->descriptor.idProduct);
783	783
784		− if (iface_desc->desc.bNumEndpoints < 1) {
785		− dev_err(&interface->dev, "Invalid number of endpoints\n");
786		− retval = -EINVAL;
787		− goto error;
788		− }
789		−
790	784	/* set up the endpoint information */
791	785	for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) {
792	786	endpoint = &iface_desc->endpoint[i].desc;
@@ -797,6 +791,21 @@ static int iowarrior_probe(struct usb_interface *interface, · 4 unmodified lines
797	791	/* this one will match for the IOWarrior56 only */
798	792	dev->int_out_endpoint = endpoint;
799	793	}
	794	+
	795	+ if (!dev->int_in_endpoint) {
	796	+ dev_err(&interface->dev, "no interrupt-in endpoint found\n");
	797	+ retval = -ENODEV;
	798	+ goto error;
	799	+ }
	800	+
	801	+ if (dev->product_id == USB_DEVICE_ID_CODEMERCS_IOW56) {
	802	+ if (!dev->int_out_endpoint) {
	803	+ dev_err(&interface->dev, "no interrupt-out endpoint found\n");
	804	+ retval = -ENODEV;
	805	+ goto error;
	806	+ }
	807	+ }
	808	+
800	809	/* we have to check the report_size often, so remember it in the endianness suitable for our machine */
801	810	dev->report_size = usb_endpoint_maxp(dev->int_in_endpoint);
802	811	if ((dev->interface->cur_altsetting->desc.bInterfaceNumber == 0) &&

drivers/usb/misc/usb251xb.c+24−35

@@ -375,18 +375,24 @@ static int usb251xb_get_ofdata(struct usb251xb *hub, · 374 unmodified lines
375	375	if (of_get_property(np, "dynamic-power-switching", NULL))
376	376	hub->conf_data2 \|= BIT(7);
377	377
378		− if (of_get_property(np, "oc-delay-100us", NULL)) {
379		− hub->conf_data2 &= ~BIT(5);
380		− hub->conf_data2 &= ~BIT(4);
381		− } else if (of_get_property(np, "oc-delay-4ms", NULL)) {
382		− hub->conf_data2 &= ~BIT(5);
383		− hub->conf_data2 \|= BIT(4);
384		− } else if (of_get_property(np, "oc-delay-8ms", NULL)) {
385		− hub->conf_data2 \|= BIT(5);
386		− hub->conf_data2 &= ~BIT(4);
387		− } else if (of_get_property(np, "oc-delay-16ms", NULL)) {
388		− hub->conf_data2 \|= BIT(5);
389		− hub->conf_data2 \|= BIT(4);
	378	+ if (!of_property_read_u32(np, "oc-delay-us", property_u32)) {
	379	+ if (*property_u32 == 100) {
	380	+ /* 100 us*/
	381	+ hub->conf_data2 &= ~BIT(5);
	382	+ hub->conf_data2 &= ~BIT(4);
	383	+ } else if (*property_u32 == 4000) {
	384	+ /* 4 ms */
	385	+ hub->conf_data2 &= ~BIT(5);
	386	+ hub->conf_data2 \|= BIT(4);
	387	+ } else if (*property_u32 == 16000) {
	388	+ /* 16 ms */
	389	+ hub->conf_data2 \|= BIT(5);
	390	+ hub->conf_data2 \|= BIT(4);
	391	+ } else {
	392	+ /* 8 ms (DEFAULT) */
	393	+ hub->conf_data2 \|= BIT(5);
	394	+ hub->conf_data2 &= ~BIT(4);
	395	+ }
390	396	}
391	397
392	398	if (of_get_property(np, "compound-device", NULL))
@@ -432,30 +438,9 @@ static int usb251xb_get_ofdata(struct usb251xb *hub, · 39 unmodified lines
432	438	}
433	439	}
434	440
435		− hub->max_power_sp = USB251XB_DEF_MAX_POWER_SELF;
436		− if (!of_property_read_u32(np, "max-sp-power", property_u32))
437		− hub->max_power_sp = min_t(u8, be32_to_cpu(*property_u32) / 2,
438		− 250);
439		−
440		− hub->max_power_bp = USB251XB_DEF_MAX_POWER_BUS;
441		− if (!of_property_read_u32(np, "max-bp-power", property_u32))
442		− hub->max_power_bp = min_t(u8, be32_to_cpu(*property_u32) / 2,
443		− 250);
444		−
445		− hub->max_current_sp = USB251XB_DEF_MAX_CURRENT_SELF;
446		− if (!of_property_read_u32(np, "max-sp-current", property_u32))
447		− hub->max_current_sp = min_t(u8, be32_to_cpu(*property_u32) / 2,
448		− 250);
449		−
450		− hub->max_current_bp = USB251XB_DEF_MAX_CURRENT_BUS;
451		− if (!of_property_read_u32(np, "max-bp-current", property_u32))
452		− hub->max_current_bp = min_t(u8, be32_to_cpu(*property_u32) / 2,
453		− 250);
454		−
455	441	hub->power_on_time = USB251XB_DEF_POWER_ON_TIME;
456		− if (!of_property_read_u32(np, "power-on-time", property_u32))
457		− hub->power_on_time = min_t(u8, be32_to_cpu(*property_u32) / 2,
458		− 255);
	442	+ if (!of_property_read_u32(np, "power-on-time-ms", property_u32))
	443	+ hub->power_on_time = min_t(u8, *property_u32 / 2, 255);
459	444
460	445	if (of_property_read_u16_array(np, "language-id", &hub->lang_id, 1))
461	446	hub->lang_id = USB251XB_DEF_LANGUAGE_ID;
@@ -492,6 +477,10 @@ static int usb251xb_get_ofdata(struct usb251xb *hub, · 30 unmodified lines
492	477	/* The following parameters are currently not exposed to devicetree, but
493	478	* may be as soon as needed.
494	479	*/
	480	+ hub->max_power_sp = USB251XB_DEF_MAX_POWER_SELF;
	481	+ hub->max_power_bp = USB251XB_DEF_MAX_POWER_BUS;
	482	+ hub->max_current_sp = USB251XB_DEF_MAX_CURRENT_SELF;
	483	+ hub->max_current_bp = USB251XB_DEF_MAX_CURRENT_BUS;
495	484	hub->bat_charge_en = USB251XB_DEF_BATTERY_CHARGING_ENABLE;
496	485	hub->boost_up = USB251XB_DEF_BOOST_UP;
497	486	hub->boost_x = USB251XB_DEF_BOOST_X;

drivers/usb/phy/phy-isp1301.c+7−0

@@ -33,6 +33,12 @@ static const struct i2c_device_id isp1301_id[] = { · 32 unmodified lines
33	33	};
34	34	MODULE_DEVICE_TABLE(i2c, isp1301_id);
35	35
	36	+static const struct of_device_id isp1301_of_match[] = {
	37	+ {.compatible = "nxp,isp1301" },
	38	+ { },
	39	+};
	40	+MODULE_DEVICE_TABLE(of, isp1301_of_match);
	41	+
36	42	static struct i2c_client *isp1301_i2c_client;
37	43
38	44	static int __isp1301_write(struct isp1301 *isp, u8 reg, u8 value, u8 clear)
@@ -130,6 +136,7 @@ static int isp1301_remove(struct i2c_client *client) · 91 unmodified lines
130	136	static struct i2c_driver isp1301_driver = {
131	137	.driver = {
132	138	.name = DRV_NAME,
	139	+ .of_match_table = of_match_ptr(isp1301_of_match),
133	140	},
134	141	.probe = isp1301_probe,
135	142	.remove = isp1301_remove,

drivers/usb/serial/digi_acceleport.c+1−1

@@ -1500,7 +1500,7 @@ static int digi_read_oob_callback(struct urb *urb) · 1499 unmodified lines
1500	1500	return -1;
1501	1501
1502	1502	/* handle each oob command */
1503		− for (i = 0; i < urb->actual_length - 4; i += 4) {
	1503	+ for (i = 0; i < urb->actual_length - 3; i += 4) {
1504	1504	opcode = buf[i];
1505	1505	line = buf[i + 1];
1506	1506	status = buf[i + 2];

drivers/usb/serial/io_ti.c+7−1

@@ -1674,6 +1674,12 @@ static void edge_interrupt_callback(struct urb *urb) · 1673 unmodified lines
1674	1674	function = TIUMP_GET_FUNC_FROM_CODE(data[0]);
1675	1675	dev_dbg(dev, "%s - port_number %d, function %d, info 0x%x\n", __func__,
1676	1676	port_number, function, data[1]);
	1677	+
	1678	+ if (port_number >= edge_serial->serial->num_ports) {
	1679	+ dev_err(dev, "bad port number %d\n", port_number);
	1680	+ goto exit;
	1681	+ }
	1682	+
1677	1683	port = edge_serial->serial->port[port_number];
1678	1684	edge_port = usb_get_serial_port_data(port);
1679	1685	if (!edge_port) {
@@ -1755,7 +1761,7 @@ static void edge_bulk_in_callback(struct urb *urb) · 75 unmodified lines
1755	1761
1756	1762	port_number = edge_port->port->port_number;
1757	1763
1758		− if (edge_port->lsr_event) {
	1764	+ if (urb->actual_length > 0 && edge_port->lsr_event) {
1759	1765	edge_port->lsr_event = 0;
1760	1766	dev_dbg(dev, "%s ===== Port %u LSR Status = %02x, Data = %02x ======\n",
1761	1767	__func__, port_number, edge_port->lsr_mask, *data);

drivers/usb/serial/omninet.c+0−13

@@ -31,7 +31,6 @@ · 30 unmodified lines
31	31	#define BT_IGNITIONPRO_ID 0x2000
32	32
33	33	/* function prototypes */
34		−static int omninet_open(struct tty_struct tty, struct usb_serial_port port);
35	34	static void omninet_process_read_urb(struct urb *urb);
36	35	static void omninet_write_bulk_callback(struct urb *urb);
37	36	static int omninet_write(struct tty_struct tty, struct usb_serial_port port,
@@ -60,7 +59,6 @@ static struct usb_serial_driver zyxel_omninet_device = { · 22 unmodified lines
60	59	.attach = omninet_attach,
61	60	.port_probe = omninet_port_probe,
62	61	.port_remove = omninet_port_remove,
63		− .open = omninet_open,
64	62	.write = omninet_write,
65	63	.write_room = omninet_write_room,
66	64	.write_bulk_callback = omninet_write_bulk_callback,
@@ -140,17 +138,6 @@ static int omninet_port_remove(struct usb_serial_port *port) · 73 unmodified lines
140	138	return 0;
141	139	}
142	140
143		−static int omninet_open(struct tty_struct tty, struct usb_serial_port port)
144		−{
145		− struct usb_serial *serial = port->serial;
146		− struct usb_serial_port *wport;
147		−
148		− wport = serial->port[1];
149		− tty_port_tty_set(&wport->port, tty);
150		−
151		− return usb_serial_generic_open(tty, port);
152		−}
153		−
154	141	#define OMNINET_HEADERLEN 4
155	142	#define OMNINET_BULKOUTSIZE 64
156	143	#define OMNINET_PAYLOADSIZE (OMNINET_BULKOUTSIZE - OMNINET_HEADERLEN)

drivers/usb/serial/safe_serial.c+5−0

@@ -200,6 +200,11 @@ static void safe_process_read_urb(struct urb *urb) · 199 unmodified lines
200	200	if (!safe)
201	201	goto out;
202	202
	203	+ if (length < 2) {
	204	+ dev_err(&port->dev, "malformed packet\n");
	205	+ return;
	206	+ }
	207	+
203	208	fcs = fcs_compute10(data, length, CRC10_INITFCS);
204	209	if (fcs) {
205	210	dev_err(&port->dev, "%s - bad CRC %x\n", __func__, fcs);

drivers/usb/storage/unusual_devs.h+14−0

@@ -2071,6 +2071,20 @@ UNUSUAL_DEV( 0x1370, 0x6828, 0x0110, 0x0110, · 2070 unmodified lines
2071	2071	USB_SC_DEVICE, USB_PR_DEVICE, NULL,
2072	2072	US_FL_IGNORE_RESIDUE ),
2073	2073
	2074	+/*
	2075	+ * Reported by Tobias Jakobi <tjakobi@math.uni-bielefeld.de>
	2076	+ * The INIC-3619 bridge is used in the StarTech SLSODDU33B
	2077	+ * SATA-USB enclosure for slimline optical drives.
	2078	+ *
	2079	+ * The quirk enables MakeMKV to properly exchange keys with
	2080	+ * an installed BD drive.
	2081	+ */
	2082	+UNUSUAL_DEV( 0x13fd, 0x3609, 0x0209, 0x0209,
	2083	+ "Initio Corporation",
	2084	+ "INIC-3619",
	2085	+ USB_SC_DEVICE, USB_PR_DEVICE, NULL,
	2086	+ US_FL_IGNORE_RESIDUE ),
	2087	+
2074	2088	/* Reported by Qinglin Ye <yestyle@gmail.com> */
2075	2089	UNUSUAL_DEV( 0x13fe, 0x3600, 0x0100, 0x0100,
2076	2090	"Kingston",

drivers/xen/swiotlb-xen.c+47−0

@@ -681,3 +681,50 @@ xen_swiotlb_set_dma_mask(struct device *dev, u64 dma_mask) · 680 unmodified lines
681	681	return 0;
682	682	}
683	683	EXPORT_SYMBOL_GPL(xen_swiotlb_set_dma_mask);
	684	+
	685	+/*
	686	+ * Create userspace mapping for the DMA-coherent memory.
	687	+ * This function should be called with the pages from the current domain only,
	688	+ * passing pages mapped from other domains would lead to memory corruption.
	689	+ */
	690	+int
	691	+xen_swiotlb_dma_mmap(struct device dev, struct vm_area_struct vma,
	692	+ void *cpu_addr, dma_addr_t dma_addr, size_t size,
	693	+ unsigned long attrs)
	694	+{
	695	+#if defined(CONFIG_ARM) \|\| defined(CONFIG_ARM64)
	696	+ if (__generic_dma_ops(dev)->mmap)
	697	+ return __generic_dma_ops(dev)->mmap(dev, vma, cpu_addr,
	698	+ dma_addr, size, attrs);
	699	+#endif
	700	+ return dma_common_mmap(dev, vma, cpu_addr, dma_addr, size);
	701	+}
	702	+EXPORT_SYMBOL_GPL(xen_swiotlb_dma_mmap);
	703	+
	704	+/*
	705	+ * This function should be called with the pages from the current domain only,
	706	+ * passing pages mapped from other domains would lead to memory corruption.
	707	+ */
	708	+int
	709	+xen_swiotlb_get_sgtable(struct device dev, struct sg_table sgt,
	710	+ void *cpu_addr, dma_addr_t handle, size_t size,
	711	+ unsigned long attrs)
	712	+{
	713	+#if defined(CONFIG_ARM) \|\| defined(CONFIG_ARM64)
	714	+ if (__generic_dma_ops(dev)->get_sgtable) {
	715	+#if 0
	716	+ /*
	717	+ * This check verifies that the page belongs to the current domain and
	718	+ * is not one mapped from another domain.
	719	+ * This check is for debug only, and should not go to production build
	720	+ */
	721	+ unsigned long bfn = PHYS_PFN(dma_to_phys(dev, handle));
	722	+ BUG_ON (!page_is_ram(bfn));
	723	+#endif
	724	+ return __generic_dma_ops(dev)->get_sgtable(dev, sgt, cpu_addr,
	725	+ handle, size, attrs);
	726	+ }
	727	+#endif
	728	+ return dma_common_get_sgtable(dev, sgt, cpu_addr, handle, size);
	729	+}
	730	+EXPORT_SYMBOL_GPL(xen_swiotlb_get_sgtable);

drivers/xen/xenbus/xenbus_dev_frontend.c+0−1

@@ -55,7 +55,6 @@ · 54 unmodified lines
55	55	#include <linux/string.h>
56	56	#include <linux/slab.h>
57	57	#include <linux/miscdevice.h>
58		−#include <linux/init.h>
59	58
60	59	#include <xen/xenbus.h>
61	60	#include <xen/xen.h>

fs/dlm/lowcomms.c+1−1

@@ -743,7 +743,7 @@ static int tcp_accept_from_sock(struct connection *con) · 742 unmodified lines
743	743	newsock->type = con->sock->type;
744	744	newsock->ops = con->sock->ops;
745	745
746		− result = con->sock->ops->accept(con->sock, newsock, O_NONBLOCK);
	746	+ result = con->sock->ops->accept(con->sock, newsock, O_NONBLOCK, true);
747	747	if (result < 0)
748	748	goto accept_err;
749	749

fs/fat/inode.c+12−1

@@ -1359,6 +1359,16 @@ out: · 1358 unmodified lines
1359	1359	return 0;
1360	1360	}
1361	1361
	1362	+static void fat_dummy_inode_init(struct inode *inode)
	1363	+{
	1364	+ /* Initialize this dummy inode to work as no-op. */
	1365	+ MSDOS_I(inode)->mmu_private = 0;
	1366	+ MSDOS_I(inode)->i_start = 0;
	1367	+ MSDOS_I(inode)->i_logstart = 0;
	1368	+ MSDOS_I(inode)->i_attrs = 0;
	1369	+ MSDOS_I(inode)->i_pos = 0;
	1370	+}
	1371	+
1362	1372	static int fat_read_root(struct inode *inode)
1363	1373	{
1364	1374	struct msdos_sb_info *sbi = MSDOS_SB(inode->i_sb);
@@ -1803,12 +1813,13 @@ int fat_fill_super(struct super_block sb, void data, int silent, int isvfat, · 438 unmodified lines
1803	1813	fat_inode = new_inode(sb);
1804	1814	if (!fat_inode)
1805	1815	goto out_fail;
1806		− MSDOS_I(fat_inode)->i_pos = 0;
	1816	+ fat_dummy_inode_init(fat_inode);
1807	1817	sbi->fat_inode = fat_inode;
1808	1818
1809	1819	fsinfo_inode = new_inode(sb);
1810	1820	if (!fsinfo_inode)
1811	1821	goto out_fail;
	1822	+ fat_dummy_inode_init(fsinfo_inode);
1812	1823	fsinfo_inode->i_ino = MSDOS_FSINFO_INO;
1813	1824	sbi->fsinfo_inode = fsinfo_inode;
1814	1825	insert_inode_hash(fsinfo_inode);

fs/fs-writeback.c+21−14

@@ -173,19 +173,33 @@ static void wb_wakeup(struct bdi_writeback *wb) · 172 unmodified lines
173	173	spin_unlock_bh(&wb->work_lock);
174	174	}
175	175
	176	+static void finish_writeback_work(struct bdi_writeback *wb,
	177	+ struct wb_writeback_work *work)
	178	+{
	179	+ struct wb_completion *done = work->done;
	180	+
	181	+ if (work->auto_free)
	182	+ kfree(work);
	183	+ if (done && atomic_dec_and_test(&done->cnt))
	184	+ wake_up_all(&wb->bdi->wb_waitq);
	185	+}
	186	+
176	187	static void wb_queue_work(struct bdi_writeback *wb,
177	188	struct wb_writeback_work *work)
178	189	{
179	190	trace_writeback_queue(wb, work);
180	191
181		− spin_lock_bh(&wb->work_lock);
182		− if (!test_bit(WB_registered, &wb->state))
183		− goto out_unlock;
184	192	if (work->done)
185	193	atomic_inc(&work->done->cnt);
186		− list_add_tail(&work->list, &wb->work_list);
187		− mod_delayed_work(bdi_wq, &wb->dwork, 0);
188		−out_unlock:
	194	+
	195	+ spin_lock_bh(&wb->work_lock);
	196	+
	197	+ if (test_bit(WB_registered, &wb->state)) {
	198	+ list_add_tail(&work->list, &wb->work_list);
	199	+ mod_delayed_work(bdi_wq, &wb->dwork, 0);
	200	+ } else
	201	+ finish_writeback_work(wb, work);
	202	+
189	203	spin_unlock_bh(&wb->work_lock);
190	204	}
191	205
@@ -1873,16 +1887,9 @@ static long wb_do_writeback(struct bdi_writeback *wb) · 1681 unmodified lines
1873	1887
1874	1888	set_bit(WB_writeback_running, &wb->state);
1875	1889	while ((work = get_next_work_item(wb)) != NULL) {
1876		− struct wb_completion *done = work->done;
1877		−
1878	1890	trace_writeback_exec(wb, work);
1879		−
1880	1891	wrote += wb_writeback(wb, work);
1881		−
1882		− if (work->auto_free)
1883		− kfree(work);
1884		− if (done && atomic_dec_and_test(&done->cnt))
1885		− wake_up_all(&wb->bdi->wb_waitq);
	1892	+ finish_writeback_work(wb, work);
1886	1893	}
1887	1894
1888	1895	/*

fs/gfs2/incore.h+1−1

@@ -207,7 +207,7 @@ struct lm_lockname { · 206 unmodified lines
207	207	struct gfs2_sbd *ln_sbd;
208	208	u64 ln_number;
209	209	unsigned int ln_type;
210		−};
	210	+} __packed __aligned(sizeof(int));
211	211
212	212	#define lm_name_equal(name1, name2) \
213	213	(((name1)->ln_number == (name2)->ln_number) && \

fs/iomap.c+10−7

@@ -846,7 +846,8 @@ iomap_dio_rw(struct kiocb iocb, struct iov_iter iter, · 845 unmodified lines
846	846	struct address_space *mapping = iocb->ki_filp->f_mapping;
847	847	struct inode *inode = file_inode(iocb->ki_filp);
848	848	size_t count = iov_iter_count(iter);
849		− loff_t pos = iocb->ki_pos, end = iocb->ki_pos + count - 1, ret = 0;
	849	+ loff_t pos = iocb->ki_pos, start = pos;
	850	+ loff_t end = iocb->ki_pos + count - 1, ret = 0;
850	851	unsigned int flags = IOMAP_DIRECT;
851	852	struct blk_plug plug;
852	853	struct iomap_dio *dio;
@@ -887,12 +888,12 @@ iomap_dio_rw(struct kiocb iocb, struct iov_iter iter, · 34 unmodified lines
887	888	}
888	889
889	890	if (mapping->nrpages) {
890		− ret = filemap_write_and_wait_range(mapping, iocb->ki_pos, end);
	891	+ ret = filemap_write_and_wait_range(mapping, start, end);
891	892	if (ret)
892	893	goto out_free_dio;
893	894
894	895	ret = invalidate_inode_pages2_range(mapping,
895		− iocb->ki_pos >> PAGE_SHIFT, end >> PAGE_SHIFT);
	896	+ start >> PAGE_SHIFT, end >> PAGE_SHIFT);
896	897	WARN_ON_ONCE(ret);
897	898	ret = 0;
898	899	}
@@ -941,6 +942,8 @@ iomap_dio_rw(struct kiocb iocb, struct iov_iter iter, · 42 unmodified lines
941	942	__set_current_state(TASK_RUNNING);
942	943	}
943	944
	945	+ ret = iomap_dio_complete(dio);
	946	+
944	947	/*
945	948	* Try again to invalidate clean pages which might have been cached by
946	949	* non-direct readahead, or faulted in by get_user_pages() if the source
@@ -949,12 +952,12 @@ iomap_dio_rw(struct kiocb iocb, struct iov_iter iter, · 2 unmodified lines
949	952	* this invalidation fails, tough, the write still worked...
950	953	*/
951	954	if (iov_iter_rw(iter) == WRITE && mapping->nrpages) {
952		− ret = invalidate_inode_pages2_range(mapping,
953		− iocb->ki_pos >> PAGE_SHIFT, end >> PAGE_SHIFT);
954		− WARN_ON_ONCE(ret);
	955	+ int err = invalidate_inode_pages2_range(mapping,
	956	+ start >> PAGE_SHIFT, end >> PAGE_SHIFT);
	957	+ WARN_ON_ONCE(err);
955	958	}
956	959
957		− return iomap_dio_complete(dio);
	960	+ return ret;
958	961
959	962	out_free_dio:
960	963	kfree(dio);

fs/ocfs2/cluster/tcp.c+1−1

@@ -1863,7 +1863,7 @@ static int o2net_accept_one(struct socket sock, int more) · 1862 unmodified lines
1863	1863
1864	1864	new_sock->type = sock->type;
1865	1865	new_sock->ops = sock->ops;
1866		− ret = sock->ops->accept(sock, new_sock, O_NONBLOCK);
	1866	+ ret = sock->ops->accept(sock, new_sock, O_NONBLOCK, false);
1867	1867	if (ret < 0)
1868	1868	goto out;
1869	1869

fs/overlayfs/util.c+0−1

@@ -12,7 +12,6 @@ · 11 unmodified lines
12	12	#include <linux/slab.h>
13	13	#include <linux/cred.h>
14	14	#include <linux/xattr.h>
15		−#include <linux/sched/signal.h>
16	15	#include "overlayfs.h"
17	16	#include "ovl_entry.h"
18	17

fs/timerfd.c+4−4

@@ -400,9 +400,9 @@ SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags) · 399 unmodified lines
400	400	clockid != CLOCK_BOOTTIME_ALARM))
401	401	return -EINVAL;
402	402
403		− if (!capable(CAP_WAKE_ALARM) &&
404		− (clockid == CLOCK_REALTIME_ALARM \|\|
405		− clockid == CLOCK_BOOTTIME_ALARM))
	403	+ if ((clockid == CLOCK_REALTIME_ALARM \|\|
	404	+ clockid == CLOCK_BOOTTIME_ALARM) &&
	405	+ !capable(CAP_WAKE_ALARM))
406	406	return -EPERM;
407	407
408	408	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
@@ -449,7 +449,7 @@ static int do_timerfd_settime(int ufd, int flags, · 40 unmodified lines
449	449	return ret;
450	450	ctx = f.file->private_data;
451	451
452		− if (!capable(CAP_WAKE_ALARM) && isalarm(ctx)) {
	452	+ if (isalarm(ctx) && !capable(CAP_WAKE_ALARM)) {
453	453	fdput(f);
454	454	return -EPERM;
455	455	}

fs/userfaultfd.c+26−49

@@ -138,8 +138,6 @@ out: · 137 unmodified lines
138	138	* userfaultfd_ctx_get - Acquires a reference to the internal userfaultfd
139	139	* context.
140	140	* @ctx: [in] Pointer to the userfaultfd context.
141		− *
142		− * Returns: In case of success, returns not zero.
143	141	*/
144	142	static void userfaultfd_ctx_get(struct userfaultfd_ctx *ctx)
145	143	{
@@ -267,6 +265,7 @@ static inline bool userfaultfd_must_wait(struct userfaultfd_ctx *ctx, · 121 unmodified lines
267	265	{
268	266	struct mm_struct *mm = ctx->mm;
269	267	pgd_t *pgd;
	268	+ p4d_t *p4d;
270	269	pud_t *pud;
271	270	pmd_t *pmd, _pmd;
272	271	pte_t *pte;
@@ -277,7 +276,10 @@ static inline bool userfaultfd_must_wait(struct userfaultfd_ctx *ctx, · 4 unmodified lines
277	276	pgd = pgd_offset(mm, address);
278	277	if (!pgd_present(*pgd))
279	278	goto out;
280		− pud = pud_offset(pgd, address);
	279	+ p4d = p4d_offset(pgd, address);
	280	+ if (!p4d_present(*p4d))
	281	+ goto out;
	282	+ pud = pud_offset(p4d, address);
281	283	if (!pud_present(*pud))
282	284	goto out;
283	285	pmd = pmd_offset(pud, address);
@@ -490,7 +492,7 @@ int handle_userfault(struct vm_fault *vmf, unsigned long reason) · 206 unmodified lines
490	492	* in such case.
491	493	*/
492	494	down_read(&mm->mmap_sem);
493		− ret = 0;
	495	+ ret = VM_FAULT_NOPAGE;
494	496	}
495	497	}
496	498
@@ -527,10 +529,11 @@ out: · 30 unmodified lines
527	529	return ret;
528	530	}
529	531
530		−static int userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx,
531		− struct userfaultfd_wait_queue *ewq)
	532	+static void userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx,
	533	+ struct userfaultfd_wait_queue *ewq)
532	534	{
533		− int ret = 0;
	535	+ if (WARN_ON_ONCE(current->flags & PF_EXITING))
	536	+ goto out;
534	537
535	538	ewq->ctx = ctx;
536	539	init_waitqueue_entry(&ewq->wq, current);
@@ -547,8 +550,16 @@ static int userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx, · 10 unmodified lines
547	550	break;
548	551	if (ACCESS_ONCE(ctx->released) \|\|
549	552	fatal_signal_pending(current)) {
550		− ret = -1;
551	553	__remove_wait_queue(&ctx->event_wqh, &ewq->wq);
	554	+ if (ewq->msg.event == UFFD_EVENT_FORK) {
	555	+ struct userfaultfd_ctx *new;
	556	+
	557	+ new = (struct userfaultfd_ctx *)
	558	+ (unsigned long)
	559	+ ewq->msg.arg.reserved.reserved1;
	560	+
	561	+ userfaultfd_ctx_put(new);
	562	+ }
552	563	break;
553	564	}
554	565
@@ -566,9 +577,8 @@ static int userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx, · 11 unmodified lines
566	577	* ctx may go away after this if the userfault pseudo fd is
567	578	* already released.
568	579	*/
569		−
	580	+out:
570	581	userfaultfd_ctx_put(ctx);
571		− return ret;
572	582	}
573	583
574	584	static void userfaultfd_event_complete(struct userfaultfd_ctx *ctx,
@@ -626,7 +636,7 @@ int dup_userfaultfd(struct vm_area_struct vma, struct list_head fcs) · 51 unmodified lines
626	636	return 0;
627	637	}
628	638
629		−static int dup_fctx(struct userfaultfd_fork_ctx *fctx)
	639	+static void dup_fctx(struct userfaultfd_fork_ctx *fctx)
630	640	{
631	641	struct userfaultfd_ctx *ctx = fctx->orig;
632	642	struct userfaultfd_wait_queue ewq;
@@ -636,17 +646,15 @@ static int dup_fctx(struct userfaultfd_fork_ctx *fctx) · 3 unmodified lines
636	646	ewq.msg.event = UFFD_EVENT_FORK;
637	647	ewq.msg.arg.reserved.reserved1 = (unsigned long)fctx->new;
638	648
639		− return userfaultfd_event_wait_completion(ctx, &ewq);
	649	+ userfaultfd_event_wait_completion(ctx, &ewq);
640	650	}
641	651
642	652	void dup_userfaultfd_complete(struct list_head *fcs)
643	653	{
644		− int ret = 0;
645	654	struct userfaultfd_fork_ctx fctx, n;
646	655
647	656	list_for_each_entry_safe(fctx, n, fcs, list) {
648		− if (!ret)
649		− ret = dup_fctx(fctx);
	657	+ dup_fctx(fctx);
650	658	list_del(&fctx->list);
651	659	kfree(fctx);
652	660	}
@@ -689,8 +697,7 @@ void mremap_userfaultfd_complete(struct vm_userfaultfd_ctx *vm_ctx, · 36 unmodified lines
689	697	userfaultfd_event_wait_completion(ctx, &ewq);
690	698	}
691	699
692		−void userfaultfd_remove(struct vm_area_struct *vma,
693		− struct vm_area_struct **prev,
	700	+bool userfaultfd_remove(struct vm_area_struct *vma,
694	701	unsigned long start, unsigned long end)
695	702	{
696	703	struct mm_struct *mm = vma->vm_mm;
@@ -699,13 +706,11 @@ void userfaultfd_remove(struct vm_area_struct *vma, · 2 unmodified lines
699	706
700	707	ctx = vma->vm_userfaultfd_ctx.ctx;
701	708	if (!ctx \|\| !(ctx->features & UFFD_FEATURE_EVENT_REMOVE))
702		− return;
	709	+ return true;
703	710
704	711	userfaultfd_ctx_get(ctx);
705	712	up_read(&mm->mmap_sem);
706	713
707		− prev = NULL; / We wait for ACK w/o the mmap semaphore */
708		−
709	714	msg_init(&ewq.msg);
710	715
711	716	ewq.msg.event = UFFD_EVENT_REMOVE;
@@ -714,7 +719,7 @@ void userfaultfd_remove(struct vm_area_struct *vma, · 2 unmodified lines
714	719
715	720	userfaultfd_event_wait_completion(ctx, &ewq);
716	721
717		− down_read(&mm->mmap_sem);
	722	+ return false;
718	723	}
719	724
720	725	static bool has_unmap_ctx(struct userfaultfd_ctx ctx, struct list_head unmaps,
@@ -775,34 +780,6 @@ void userfaultfd_unmap_complete(struct mm_struct mm, struct list_head uf) · 54 unmodified lines
775	780	}
776	781	}
777	782
778		−void userfaultfd_exit(struct mm_struct *mm)
779		−{
780		− struct vm_area_struct *vma = mm->mmap;
781		−
782		− /*
783		− * We can do the vma walk without locking because the caller
784		− * (exit_mm) knows it now has exclusive access
785		− */
786		− while (vma) {
787		− struct userfaultfd_ctx *ctx = vma->vm_userfaultfd_ctx.ctx;
788		−
789		− if (ctx && (ctx->features & UFFD_FEATURE_EVENT_EXIT)) {
790		− struct userfaultfd_wait_queue ewq;
791		−
792		− userfaultfd_ctx_get(ctx);
793		−
794		− msg_init(&ewq.msg);
795		− ewq.msg.event = UFFD_EVENT_EXIT;
796		−
797		− userfaultfd_event_wait_completion(ctx, &ewq);
798		−
799		− ctx->features &= ~UFFD_FEATURE_EVENT_EXIT;
800		− }
801		−
802		− vma = vma->vm_next;
803		− }
804		−}
805		−
806	783	static int userfaultfd_release(struct inode inode, struct file file)
807	784	{
808	785	struct userfaultfd_ctx *ctx = file->private_data;

fs/xfs/kmem.c+0−18

@@ -25,24 +25,6 @@ · 24 unmodified lines
25	25	#include "kmem.h"
26	26	#include "xfs_message.h"
27	27
28		−/*
29		− * Greedy allocation. May fail and may return vmalloced memory.
30		− */
31		−void *
32		−kmem_zalloc_greedy(size_t *size, size_t minsize, size_t maxsize)
33		−{
34		− void *ptr;
35		− size_t kmsize = maxsize;
36		−
37		− while (!(ptr = vzalloc(kmsize))) {
38		− if ((kmsize >>= 1) <= minsize)
39		− kmsize = minsize;
40		− }
41		− if (ptr)
42		− *size = kmsize;
43		− return ptr;
44		−}
45		−
46	28	void *
47	29	kmem_alloc(size_t size, xfs_km_flags_t flags)
48	30	{

fs/xfs/kmem.h+0−2

@@ -69,8 +69,6 @@ static inline void kmem_free(const void *ptr) · 68 unmodified lines
69	69	}
70	70
71	71
72		−extern void kmem_zalloc_greedy(size_t , size_t, size_t);
73		−
74	72	static inline void *
75	73	kmem_zalloc(size_t size, xfs_km_flags_t flags)
76	74	{

fs/xfs/libxfs/xfs_bmap.c+21−13

@@ -763,8 +763,8 @@ xfs_bmap_extents_to_btree( · 762 unmodified lines
763	763	args.type = XFS_ALLOCTYPE_START_BNO;
764	764	args.fsbno = XFS_INO_TO_FSB(mp, ip->i_ino);
765	765	} else if (dfops->dop_low) {
766		−try_another_ag:
767	766	args.type = XFS_ALLOCTYPE_START_BNO;
	767	+try_another_ag:
768	768	args.fsbno = *firstblock;
769	769	} else {
770	770	args.type = XFS_ALLOCTYPE_NEAR_BNO;
@@ -790,13 +790,17 @@ try_another_ag: · 19 unmodified lines
790	790	if (xfs_sb_version_hasreflink(&cur->bc_mp->m_sb) &&
791	791	args.fsbno == NULLFSBLOCK &&
792	792	args.type == XFS_ALLOCTYPE_NEAR_BNO) {
793		− dfops->dop_low = true;
	793	+ args.type = XFS_ALLOCTYPE_FIRST_AG;
794	794	goto try_another_ag;
795	795	}
	796	+ if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
	797	+ xfs_iroot_realloc(ip, -1, whichfork);
	798	+ xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
	799	+ return -ENOSPC;
	800	+ }
796	801	/*
797	802	* Allocation can't fail, the space was reserved.
798	803	*/
799		− ASSERT(args.fsbno != NULLFSBLOCK);
800	804	ASSERT(*firstblock == NULLFSBLOCK \|\|
801	805	args.agno >= XFS_FSB_TO_AGNO(mp, *firstblock));
802	806	*firstblock = cur->bc_private.b.firstblock = args.fsbno;
@@ -4150,6 +4154,19 @@ xfs_bmapi_read( · 3347 unmodified lines
4150	4154	return 0;
4151	4155	}
4152	4156
	4157	+/*
	4158	+ * Add a delayed allocation extent to an inode. Blocks are reserved from the
	4159	+ * global pool and the extent inserted into the inode in-core extent tree.
	4160	+ *
	4161	+ * On entry, got refers to the first extent beyond the offset of the extent to
	4162	+ * allocate or eof is specified if no such extent exists. On return, got refers
	4163	+ * to the extent record that was inserted to the inode fork.
	4164	+ *
	4165	+ * Note that the allocated extent may have been merged with contiguous extents
	4166	+ * during insertion into the inode fork. Thus, got does not reflect the current
	4167	+ * state of the inode fork on return. If necessary, the caller can use lastx to
	4168	+ * look up the updated record in the inode fork.
	4169	+ */
4153	4170	int
4154	4171	xfs_bmapi_reserve_delalloc(
4155	4172	struct xfs_inode *ip,
@@ -4236,13 +4253,8 @@ xfs_bmapi_reserve_delalloc( · 80 unmodified lines
4236	4253	got->br_startblock = nullstartblock(indlen);
4237	4254	got->br_blockcount = alen;
4238	4255	got->br_state = XFS_EXT_NORM;
4239		− xfs_bmap_add_extent_hole_delay(ip, whichfork, lastx, got);
4240	4256
4241		− /*
4242		− * Update our extent pointer, given that xfs_bmap_add_extent_hole_delay
4243		− * might have merged it into one of the neighbouring ones.
4244		− */
4245		− xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *lastx), got);
	4257	+ xfs_bmap_add_extent_hole_delay(ip, whichfork, lastx, got);
4246	4258
4247	4259	/*
4248	4260	* Tag the inode if blocks were preallocated. Note that COW fork
@@ -4254,10 +4266,6 @@ xfs_bmapi_reserve_delalloc( · 5 unmodified lines
4254	4266	if (whichfork == XFS_COW_FORK && (prealloc \|\| aoff < off \|\| alen > len))
4255	4267	xfs_inode_set_cowblocks_tag(ip);
4256	4268
4257		− ASSERT(got->br_startoff <= aoff);
4258		− ASSERT(got->br_startoff + got->br_blockcount >= aoff + alen);
4259		− ASSERT(isnullstartblock(got->br_startblock));
4260		− ASSERT(got->br_state == XFS_EXT_NORM);
4261	4269	return 0;
4262	4270
4263	4271	out_unreserve_blocks:

fs/xfs/libxfs/xfs_bmap_btree.c+3−3

@@ -447,8 +447,8 @@ xfs_bmbt_alloc_block( · 446 unmodified lines
447	447
448	448	if (args.fsbno == NULLFSBLOCK) {
449	449	args.fsbno = be64_to_cpu(start->l);
450		−try_another_ag:
451	450	args.type = XFS_ALLOCTYPE_START_BNO;
	451	+try_another_ag:
452	452	/*
453	453	* Make sure there is sufficient room left in the AG to
454	454	* complete a full tree split for an extent insert. If
@@ -488,8 +488,8 @@ try_another_ag: · 33 unmodified lines
488	488	if (xfs_sb_version_hasreflink(&cur->bc_mp->m_sb) &&
489	489	args.fsbno == NULLFSBLOCK &&
490	490	args.type == XFS_ALLOCTYPE_NEAR_BNO) {
491		− cur->bc_private.b.dfops->dop_low = true;
492	491	args.fsbno = cur->bc_private.b.firstblock;
	492	+ args.type = XFS_ALLOCTYPE_FIRST_AG;
493	493	goto try_another_ag;
494	494	}
495	495
@@ -506,7 +506,7 @@ try_another_ag: · 10 unmodified lines
506	506	goto error0;
507	507	cur->bc_private.b.dfops->dop_low = true;
508	508	}
509		− if (args.fsbno == NULLFSBLOCK) {
	509	+ if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
510	510	XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT);
511	511	*stat = 0;
512	512	return 0;

fs/xfs/xfs_aops.c+27−32

@@ -274,54 +274,49 @@ xfs_end_io( · 273 unmodified lines
274	274	struct xfs_ioend *ioend =
275	275	container_of(work, struct xfs_ioend, io_work);
276	276	struct xfs_inode *ip = XFS_I(ioend->io_inode);
	277	+ xfs_off_t offset = ioend->io_offset;
	278	+ size_t size = ioend->io_size;
277	279	int error = ioend->io_bio->bi_error;
278	280
279	281	/*
280		− * Set an error if the mount has shut down and proceed with end I/O
281		− * processing so it can perform whatever cleanups are necessary.
	282	+ * Just clean up the in-memory strutures if the fs has been shut down.
282	283	*/
283		− if (XFS_FORCED_SHUTDOWN(ip->i_mount))
	284	+ if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
284	285	error = -EIO;
	286	+ goto done;
	287	+ }
285	288
286	289	/*
287		− * For a CoW extent, we need to move the mapping from the CoW fork
288		− * to the data fork. If instead an error happened, just dump the
289		− * new blocks.
	290	+ * Clean up any COW blocks on an I/O error.
290	291	*/
291		− if (ioend->io_type == XFS_IO_COW) {
292		− if (error)
293		− goto done;
294		− if (ioend->io_bio->bi_error) {
295		− error = xfs_reflink_cancel_cow_range(ip,
296		− ioend->io_offset, ioend->io_size);
297		− goto done;
	292	+ if (unlikely(error)) {
	293	+ switch (ioend->io_type) {
	294	+ case XFS_IO_COW:
	295	+ xfs_reflink_cancel_cow_range(ip, offset, size, true);
	296	+ break;
298	297	}
299		− error = xfs_reflink_end_cow(ip, ioend->io_offset,
300		− ioend->io_size);
301		− if (error)
302		− goto done;
	298	+
	299	+ goto done;
303	300	}
304	301
305	302	/*
306		− * For unwritten extents we need to issue transactions to convert a
307		− * range to normal written extens after the data I/O has finished.
308		− * Detecting and handling completion IO errors is done individually
309		− * for each case as different cleanup operations need to be performed
310		− * on error.
	303	+ * Success: commit the COW or unwritten blocks if needed.
311	304	*/
312		− if (ioend->io_type == XFS_IO_UNWRITTEN) {
313		− if (error)
314		− goto done;
315		− error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
316		− ioend->io_size);
317		− } else if (ioend->io_append_trans) {
318		− error = xfs_setfilesize_ioend(ioend, error);
319		− } else {
320		− ASSERT(!xfs_ioend_is_append(ioend) \|\|
321		− ioend->io_type == XFS_IO_COW);
	305	+ switch (ioend->io_type) {
	306	+ case XFS_IO_COW:
	307	+ error = xfs_reflink_end_cow(ip, offset, size);
	308	+ break;
	309	+ case XFS_IO_UNWRITTEN:
	310	+ error = xfs_iomap_write_unwritten(ip, offset, size);
	311	+ break;
	312	+ default:
	313	+ ASSERT(!xfs_ioend_is_append(ioend) \|\| ioend->io_append_trans);
	314	+ break;
322	315	}
323	316
324	317	done:
	318	+ if (ioend->io_append_trans)
	319	+ error = xfs_setfilesize_ioend(ioend, error);
325	320	xfs_destroy_ioend(ioend, error);
326	321	}
327	322

fs/xfs/xfs_icache.c+1−1

@@ -1608,7 +1608,7 @@ xfs_inode_free_cowblocks( · 1607 unmodified lines
1608	1608	xfs_ilock(ip, XFS_IOLOCK_EXCL);
1609	1609	xfs_ilock(ip, XFS_MMAPLOCK_EXCL);
1610	1610
1611		− ret = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF);
	1611	+ ret = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, false);
1612	1612
1613	1613	xfs_iunlock(ip, XFS_MMAPLOCK_EXCL);
1614	1614	xfs_iunlock(ip, XFS_IOLOCK_EXCL);

fs/xfs/xfs_inode.c+1−1

@@ -1615,7 +1615,7 @@ xfs_itruncate_extents( · 1614 unmodified lines
1615	1615
1616	1616	/* Remove all pending CoW reservations. */
1617	1617	error = xfs_reflink_cancel_cow_blocks(ip, &tp, first_unmap_block,
1618		− last_block);
	1618	+ last_block, true);
1619	1619	if (error)
1620	1620	goto out;
1621	1621

fs/xfs/xfs_iomap.c+18−7

@@ -630,6 +630,11 @@ retry: · 629 unmodified lines
630	630	goto out_unlock;
631	631	}
632	632
	633	+ /*
	634	+ * Flag newly allocated delalloc blocks with IOMAP_F_NEW so we punch
	635	+ * them out if the write happens to fail.
	636	+ */
	637	+ iomap->flags = IOMAP_F_NEW;
633	638	trace_xfs_iomap_alloc(ip, offset, count, 0, &got);
634	639	done:
635	640	if (isnullstartblock(got.br_startblock))
@@ -1071,16 +1076,22 @@ xfs_file_iomap_end_delalloc( · 435 unmodified lines
1071	1076	struct xfs_inode *ip,
1072	1077	loff_t offset,
1073	1078	loff_t length,
1074		− ssize_t written)
	1079	+ ssize_t written,
	1080	+ struct iomap *iomap)
1075	1081	{
1076	1082	struct xfs_mount *mp = ip->i_mount;
1077	1083	xfs_fileoff_t start_fsb;
1078	1084	xfs_fileoff_t end_fsb;
1079	1085	int error = 0;
1080	1086
1081		− /* behave as if the write failed if drop writes is enabled */
1082		− if (xfs_mp_drop_writes(mp))
	1087	+ /*
	1088	+ * Behave as if the write failed if drop writes is enabled. Set the NEW
	1089	+ * flag to force delalloc cleanup.
	1090	+ */
	1091	+ if (xfs_mp_drop_writes(mp)) {
	1092	+ iomap->flags \|= IOMAP_F_NEW;
1083	1093	written = 0;
	1094	+ }
1084	1095
1085	1096	/*
1086	1097	* start_fsb refers to the first unused block after a short write. If
@@ -1094,14 +1105,14 @@ xfs_file_iomap_end_delalloc( · 7 unmodified lines
1094	1105	end_fsb = XFS_B_TO_FSB(mp, offset + length);
1095	1106
1096	1107	/*
1097		− * Trim back delalloc blocks if we didn't manage to write the whole
1098		− * range reserved.
	1108	+ * Trim delalloc blocks if they were allocated by this write and we
	1109	+ * didn't manage to write the whole range.
1099	1110	*
1100	1111	* We don't need to care about racing delalloc as we hold i_mutex
1101	1112	* across the reserve/allocate/unreserve calls. If there are delalloc
1102	1113	* blocks in the range, they are ours.
1103	1114	*/
1104		− if (start_fsb < end_fsb) {
	1115	+ if ((iomap->flags & IOMAP_F_NEW) && start_fsb < end_fsb) {
1105	1116	truncate_pagecache_range(VFS_I(ip), XFS_FSB_TO_B(mp, start_fsb),
1106	1117	XFS_FSB_TO_B(mp, end_fsb) - 1);
1107	1118
@@ -1131,7 +1142,7 @@ xfs_file_iomap_end( · 23 unmodified lines
1131	1142	{
1132	1143	if ((flags & IOMAP_WRITE) && iomap->type == IOMAP_DELALLOC)
1133	1144	return xfs_file_iomap_end_delalloc(XFS_I(inode), offset,
1134		− length, written);
	1145	+ length, written, iomap);
1135	1146	return 0;
1136	1147	}
1137	1148

fs/xfs/xfs_itable.c+2−4

@@ -361,7 +361,6 @@ xfs_bulkstat( · 360 unmodified lines
361	361	xfs_agino_t agino; /* inode # in allocation group */
362	362	xfs_agnumber_t agno; /* allocation group number */
363	363	xfs_btree_cur_t cur; / btree cursor for ialloc btree */
364		− size_t irbsize; /* size of irec buffer in bytes */
365	364	xfs_inobt_rec_incore_t irbuf; / start of irec buffer */
366	365	int nirbuf; /* size of irbuf */
367	366	int ubcount; /* size of user's buffer */
@@ -388,11 +387,10 @@ xfs_bulkstat( · 20 unmodified lines
388	387	*ubcountp = 0;
389	388	*done = 0;
390	389
391		− irbuf = kmem_zalloc_greedy(&irbsize, PAGE_SIZE, PAGE_SIZE * 4);
	390	+ irbuf = kmem_zalloc_large(PAGE_SIZE * 4, KM_SLEEP);
392	391	if (!irbuf)
393	392	return -ENOMEM;
394		−
395		− nirbuf = irbsize / sizeof(*irbuf);
	393	+ nirbuf = (PAGE_SIZE * 4) / sizeof(*irbuf);
396	394
397	395	/*
398	396	* Loop over the allocation groups, starting from the last

fs/xfs/xfs_mount.c+1−2

@@ -513,8 +513,7 @@ STATIC void · 512 unmodified lines
513	513	xfs_set_inoalignment(xfs_mount_t *mp)
514	514	{
515	515	if (xfs_sb_version_hasalign(&mp->m_sb) &&
516		− mp->m_sb.sb_inoalignmt >=
517		− XFS_B_TO_FSBT(mp, mp->m_inode_cluster_size))
	516	+ mp->m_sb.sb_inoalignmt >= xfs_icluster_size_fsb(mp))
518	517	mp->m_inoalign_mask = mp->m_sb.sb_inoalignmt - 1;
519	518	else
520	519	mp->m_inoalign_mask = 0;

fs/xfs/xfs_reflink.c+16−7

@@ -548,14 +548,18 @@ xfs_reflink_trim_irec_to_next_cow( · 547 unmodified lines
548	548	}
549	549
550	550	/*
551		− * Cancel all pending CoW reservations for some block range of an inode.
	551	+ * Cancel CoW reservations for some block range of an inode.
	552	+ *
	553	+ * If cancel_real is true this function cancels all COW fork extents for the
	554	+ * inode; if cancel_real is false, real extents are not cleared.
552	555	*/
553	556	int
554	557	xfs_reflink_cancel_cow_blocks(
555	558	struct xfs_inode *ip,
556	559	struct xfs_trans **tpp,
557	560	xfs_fileoff_t offset_fsb,
558		− xfs_fileoff_t end_fsb)
	561	+ xfs_fileoff_t end_fsb,
	562	+ bool cancel_real)
559	563	{
560	564	struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK);
561	565	struct xfs_bmbt_irec got, del;
@@ -579,7 +583,7 @@ xfs_reflink_cancel_cow_blocks( · 17 unmodified lines
579	583	&idx, &got, &del);
580	584	if (error)
581	585	break;
582		− } else {
	586	+ } else if (del.br_state == XFS_EXT_UNWRITTEN \|\| cancel_real) {
583	587	xfs_trans_ijoin(*tpp, ip, 0);
584	588	xfs_defer_init(&dfops, &firstfsb);
585	589
@@ -621,13 +625,17 @@ xfs_reflink_cancel_cow_blocks( · 35 unmodified lines
621	625	}
622	626
623	627	/*
624		− * Cancel all pending CoW reservations for some byte range of an inode.
	628	+ * Cancel CoW reservations for some byte range of an inode.
	629	+ *
	630	+ * If cancel_real is true this function cancels all COW fork extents for the
	631	+ * inode; if cancel_real is false, real extents are not cleared.
625	632	*/
626	633	int
627	634	xfs_reflink_cancel_cow_range(
628	635	struct xfs_inode *ip,
629	636	xfs_off_t offset,
630		− xfs_off_t count)
	637	+ xfs_off_t count,
	638	+ bool cancel_real)
631	639	{
632	640	struct xfs_trans *tp;
633	641	xfs_fileoff_t offset_fsb;
@@ -653,7 +661,8 @@ xfs_reflink_cancel_cow_range( · 19 unmodified lines
653	661	xfs_trans_ijoin(tp, ip, 0);
654	662
655	663	/* Scrape out the old CoW reservations */
656		− error = xfs_reflink_cancel_cow_blocks(ip, &tp, offset_fsb, end_fsb);
	664	+ error = xfs_reflink_cancel_cow_blocks(ip, &tp, offset_fsb, end_fsb,
	665	+ cancel_real);
657	666	if (error)
658	667	goto out_cancel;
659	668
@@ -1450,7 +1459,7 @@ next: · 790 unmodified lines
1450	1459	* We didn't find any shared blocks so turn off the reflink flag.
1451	1460	* First, get rid of any leftover CoW mappings.
1452	1461	*/
1453		− error = xfs_reflink_cancel_cow_blocks(ip, tpp, 0, NULLFILEOFF);
	1462	+ error = xfs_reflink_cancel_cow_blocks(ip, tpp, 0, NULLFILEOFF, true);
1454	1463	if (error)
1455	1464	return error;
1456	1465

fs/xfs/xfs_reflink.h+2−2

@@ -39,9 +39,9 @@ extern void xfs_reflink_trim_irec_to_next_cow(struct xfs_inode *ip, · 38 unmodified lines
39	39
40	40	extern int xfs_reflink_cancel_cow_blocks(struct xfs_inode *ip,
41	41	struct xfs_trans **tpp, xfs_fileoff_t offset_fsb,
42		− xfs_fileoff_t end_fsb);
	42	+ xfs_fileoff_t end_fsb, bool cancel_real);
43	43	extern int xfs_reflink_cancel_cow_range(struct xfs_inode *ip, xfs_off_t offset,
44		− xfs_off_t count);
	44	+ xfs_off_t count, bool cancel_real);
45	45	extern int xfs_reflink_end_cow(struct xfs_inode *ip, xfs_off_t offset,
46	46	xfs_off_t count);
47	47	extern int xfs_reflink_recover_cow(struct xfs_mount *mp);

fs/xfs/xfs_super.c+1−1

@@ -953,7 +953,7 @@ xfs_fs_destroy_inode( · 952 unmodified lines
953	953	XFS_STATS_INC(ip->i_mount, vn_remove);
954	954
955	955	if (xfs_is_reflink_inode(ip)) {
956		− error = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF);
	956	+ error = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, true);
957	957	if (error && !XFS_FORCED_SHUTDOWN(ip->i_mount))
958	958	xfs_warn(ip->i_mount,
959	959	"Error %d while evicting CoW blocks for inode %llu.",

include/asm-generic/4level-fixup.h+2−1

@@ -15,7 +15,6 @@ · 14 unmodified lines
15	15	((unlikely(pgd_none(*(pud))) && __pmd_alloc(mm, pud, address))? \
16	16	NULL: pmd_offset(pud, address))
17	17
18		−#define pud_alloc(mm, pgd, address) (pgd)
19	18	#define pud_offset(pgd, start) (pgd)
20	19	#define pud_none(pud) 0
21	20	#define pud_bad(pud) 0
@@ -35,4 +34,6 @@ · 13 unmodified lines
35	34	#undef pud_addr_end
36	35	#define pud_addr_end(addr, end) (end)
37	36
	37	+#include <asm-generic/5level-fixup.h>
	38	+
38	39	#endif

include/asm-generic/5level-fixup.h+41−0

@@ -0,0 +1,41 @@
	1	+#ifndef _5LEVEL_FIXUP_H
	2	+#define _5LEVEL_FIXUP_H
	3	+
	4	+#define __ARCH_HAS_5LEVEL_HACK
	5	+#define __PAGETABLE_P4D_FOLDED
	6	+
	7	+#define P4D_SHIFT PGDIR_SHIFT
	8	+#define P4D_SIZE PGDIR_SIZE
	9	+#define P4D_MASK PGDIR_MASK
	10	+#define PTRS_PER_P4D 1
	11	+
	12	+#define p4d_t pgd_t
	13	+
	14	+#define pud_alloc(mm, p4d, address) \
	15	+ ((unlikely(pgd_none(*(p4d))) && __pud_alloc(mm, p4d, address)) ? \
	16	+ NULL : pud_offset(p4d, address))
	17	+
	18	+#define p4d_alloc(mm, pgd, address) (pgd)
	19	+#define p4d_offset(pgd, start) (pgd)
	20	+#define p4d_none(p4d) 0
	21	+#define p4d_bad(p4d) 0
	22	+#define p4d_present(p4d) 1
	23	+#define p4d_ERROR(p4d) do { } while (0)
	24	+#define p4d_clear(p4d) pgd_clear(p4d)
	25	+#define p4d_val(p4d) pgd_val(p4d)
	26	+#define p4d_populate(mm, p4d, pud) pgd_populate(mm, p4d, pud)
	27	+#define p4d_page(p4d) pgd_page(p4d)
	28	+#define p4d_page_vaddr(p4d) pgd_page_vaddr(p4d)
	29	+
	30	+#define __p4d(x) __pgd(x)
	31	+#define set_p4d(p4dp, p4d) set_pgd(p4dp, p4d)
	32	+
	33	+#undef p4d_free_tlb
	34	+#define p4d_free_tlb(tlb, x, addr) do { } while (0)
	35	+#define p4d_free(mm, x) do { } while (0)
	36	+#define __p4d_free_tlb(tlb, x, addr) do { } while (0)
	37	+
	38	+#undef p4d_addr_end
	39	+#define p4d_addr_end(addr, end) (end)
	40	+
	41	+#endif

include/asm-generic/pgtable-nop4d-hack.h+62−0

@@ -0,0 +1,62 @@
	1	+#ifndef _PGTABLE_NOP4D_HACK_H
	2	+#define _PGTABLE_NOP4D_HACK_H
	3	+
	4	+#ifndef __ASSEMBLY__
	5	+#include <asm-generic/5level-fixup.h>
	6	+
	7	+#define __PAGETABLE_PUD_FOLDED
	8	+
	9	+/*
	10	+ * Having the pud type consist of a pgd gets the size right, and allows
	11	+ * us to conceptually access the pgd entry that this pud is folded into
	12	+ * without casting.
	13	+ */
	14	+typedef struct { pgd_t pgd; } pud_t;
	15	+
	16	+#define PUD_SHIFT PGDIR_SHIFT
	17	+#define PTRS_PER_PUD 1
	18	+#define PUD_SIZE (1UL << PUD_SHIFT)
	19	+#define PUD_MASK (~(PUD_SIZE-1))
	20	+
	21	+/*
	22	+ * The "pgd_xxx()" functions here are trivial for a folded two-level
	23	+ * setup: the pud is never bad, and a pud always exists (as it's folded
	24	+ * into the pgd entry)
	25	+ */
	26	+static inline int pgd_none(pgd_t pgd) { return 0; }
	27	+static inline int pgd_bad(pgd_t pgd) { return 0; }
	28	+static inline int pgd_present(pgd_t pgd) { return 1; }
	29	+static inline void pgd_clear(pgd_t *pgd) { }
	30	+#define pud_ERROR(pud) (pgd_ERROR((pud).pgd))
	31	+
	32	+#define pgd_populate(mm, pgd, pud) do { } while (0)
	33	+/*
	34	+ * (puds are folded into pgds so this doesn't get actually called,
	35	+ * but the define is needed for a generic inline function.)
	36	+ */
	37	+#define set_pgd(pgdptr, pgdval) set_pud((pud_t *)(pgdptr), (pud_t) { pgdval })
	38	+
	39	+static inline pud_t pud_offset(pgd_t pgd, unsigned long address)
	40	+{
	41	+ return (pud_t *)pgd;
	42	+}
	43	+
	44	+#define pud_val(x) (pgd_val((x).pgd))
	45	+#define __pud(x) ((pud_t) { __pgd(x) })
	46	+
	47	+#define pgd_page(pgd) (pud_page((pud_t){ pgd }))
	48	+#define pgd_page_vaddr(pgd) (pud_page_vaddr((pud_t){ pgd }))
	49	+
	50	+/*
	51	+ * allocating and freeing a pud is trivial: the 1-entry pud is
	52	+ * inside the pgd, so has no extra memory associated with it.
	53	+ */
	54	+#define pud_alloc_one(mm, address) NULL
	55	+#define pud_free(mm, x) do { } while (0)
	56	+#define __pud_free_tlb(tlb, x, a) do { } while (0)
	57	+
	58	+#undef pud_addr_end
	59	+#define pud_addr_end(addr, end) (end)
	60	+
	61	+#endif /* __ASSEMBLY__ */
	62	+#endif /* _PGTABLE_NOP4D_HACK_H */

include/asm-generic/pgtable-nop4d.h+56−0

@@ -0,0 +1,56 @@
	1	+#ifndef _PGTABLE_NOP4D_H
	2	+#define _PGTABLE_NOP4D_H
	3	+
	4	+#ifndef __ASSEMBLY__
	5	+
	6	+#define __PAGETABLE_P4D_FOLDED
	7	+
	8	+typedef struct { pgd_t pgd; } p4d_t;
	9	+
	10	+#define P4D_SHIFT PGDIR_SHIFT
	11	+#define PTRS_PER_P4D 1
	12	+#define P4D_SIZE (1UL << P4D_SHIFT)
	13	+#define P4D_MASK (~(P4D_SIZE-1))
	14	+
	15	+/*
	16	+ * The "pgd_xxx()" functions here are trivial for a folded two-level
	17	+ * setup: the p4d is never bad, and a p4d always exists (as it's folded
	18	+ * into the pgd entry)
	19	+ */
	20	+static inline int pgd_none(pgd_t pgd) { return 0; }
	21	+static inline int pgd_bad(pgd_t pgd) { return 0; }
	22	+static inline int pgd_present(pgd_t pgd) { return 1; }
	23	+static inline void pgd_clear(pgd_t *pgd) { }
	24	+#define p4d_ERROR(p4d) (pgd_ERROR((p4d).pgd))
	25	+
	26	+#define pgd_populate(mm, pgd, p4d) do { } while (0)
	27	+/*
	28	+ * (p4ds are folded into pgds so this doesn't get actually called,
	29	+ * but the define is needed for a generic inline function.)
	30	+ */
	31	+#define set_pgd(pgdptr, pgdval) set_p4d((p4d_t *)(pgdptr), (p4d_t) { pgdval })
	32	+
	33	+static inline p4d_t p4d_offset(pgd_t pgd, unsigned long address)
	34	+{
	35	+ return (p4d_t *)pgd;
	36	+}
	37	+
	38	+#define p4d_val(x) (pgd_val((x).pgd))
	39	+#define __p4d(x) ((p4d_t) { __pgd(x) })
	40	+
	41	+#define pgd_page(pgd) (p4d_page((p4d_t){ pgd }))
	42	+#define pgd_page_vaddr(pgd) (p4d_page_vaddr((p4d_t){ pgd }))
	43	+
	44	+/*
	45	+ * allocating and freeing a p4d is trivial: the 1-entry p4d is
	46	+ * inside the pgd, so has no extra memory associated with it.
	47	+ */
	48	+#define p4d_alloc_one(mm, address) NULL
	49	+#define p4d_free(mm, x) do { } while (0)
	50	+#define __p4d_free_tlb(tlb, x, a) do { } while (0)
	51	+
	52	+#undef p4d_addr_end
	53	+#define p4d_addr_end(addr, end) (end)
	54	+
	55	+#endif /* __ASSEMBLY__ */
	56	+#endif /* _PGTABLE_NOP4D_H */

include/asm-generic/pgtable-nopud.h+27−21

@@ -3,52 +3,57 @@ · 2 unmodified lines
3	3
4	4	#ifndef __ASSEMBLY__
5	5
	6	+#ifdef __ARCH_USE_5LEVEL_HACK
	7	+#include <asm-generic/pgtable-nop4d-hack.h>
	8	+#else
	9	+#include <asm-generic/pgtable-nop4d.h>
	10	+
6	11	#define __PAGETABLE_PUD_FOLDED
7	12
8	13	/*
9		− * Having the pud type consist of a pgd gets the size right, and allows
10		− * us to conceptually access the pgd entry that this pud is folded into
	14	+ * Having the pud type consist of a p4d gets the size right, and allows
	15	+ * us to conceptually access the p4d entry that this pud is folded into
11	16	* without casting.
12	17	*/
13		−typedef struct { pgd_t pgd; } pud_t;
	18	+typedef struct { p4d_t p4d; } pud_t;
14	19
15		−#define PUD_SHIFT PGDIR_SHIFT
	20	+#define PUD_SHIFT P4D_SHIFT
16	21	#define PTRS_PER_PUD 1
17	22	#define PUD_SIZE (1UL << PUD_SHIFT)
18	23	#define PUD_MASK (~(PUD_SIZE-1))
19	24
20	25	/*
21		− * The "pgd_xxx()" functions here are trivial for a folded two-level
	26	+ * The "p4d_xxx()" functions here are trivial for a folded two-level
22	27	* setup: the pud is never bad, and a pud always exists (as it's folded
23		− * into the pgd entry)
	28	+ * into the p4d entry)
24	29	*/
25		−static inline int pgd_none(pgd_t pgd) { return 0; }
26		−static inline int pgd_bad(pgd_t pgd) { return 0; }
27		−static inline int pgd_present(pgd_t pgd) { return 1; }
28		−static inline void pgd_clear(pgd_t *pgd) { }
29		−#define pud_ERROR(pud) (pgd_ERROR((pud).pgd))
	30	+static inline int p4d_none(p4d_t p4d) { return 0; }
	31	+static inline int p4d_bad(p4d_t p4d) { return 0; }
	32	+static inline int p4d_present(p4d_t p4d) { return 1; }
	33	+static inline void p4d_clear(p4d_t *p4d) { }
	34	+#define pud_ERROR(pud) (p4d_ERROR((pud).p4d))
30	35
31		−#define pgd_populate(mm, pgd, pud) do { } while (0)
	36	+#define p4d_populate(mm, p4d, pud) do { } while (0)
32	37	/*
33		− * (puds are folded into pgds so this doesn't get actually called,
	38	+ * (puds are folded into p4ds so this doesn't get actually called,
34	39	* but the define is needed for a generic inline function.)
35	40	*/
36		−#define set_pgd(pgdptr, pgdval) set_pud((pud_t *)(pgdptr), (pud_t) { pgdval })
	41	+#define set_p4d(p4dptr, p4dval) set_pud((pud_t *)(p4dptr), (pud_t) { p4dval })
37	42
38		−static inline pud_t * pud_offset(pgd_t * pgd, unsigned long address)
	43	+static inline pud_t pud_offset(p4d_t p4d, unsigned long address)
39	44	{
40		− return (pud_t *)pgd;
	45	+ return (pud_t *)p4d;
41	46	}
42	47
43		−#define pud_val(x) (pgd_val((x).pgd))
44		−#define __pud(x) ((pud_t) { __pgd(x) } )
	48	+#define pud_val(x) (p4d_val((x).p4d))
	49	+#define __pud(x) ((pud_t) { __p4d(x) })
45	50
46		−#define pgd_page(pgd) (pud_page((pud_t){ pgd }))
47		−#define pgd_page_vaddr(pgd) (pud_page_vaddr((pud_t){ pgd }))
	51	+#define p4d_page(p4d) (pud_page((pud_t){ p4d }))
	52	+#define p4d_page_vaddr(p4d) (pud_page_vaddr((pud_t){ p4d }))
48	53
49	54	/*
50	55	* allocating and freeing a pud is trivial: the 1-entry pud is
51		− * inside the pgd, so has no extra memory associated with it.
	56	+ * inside the p4d, so has no extra memory associated with it.
52	57	*/
53	58	#define pud_alloc_one(mm, address) NULL
54	59	#define pud_free(mm, x) do { } while (0)
@@ -58,4 +63,5 @@ static inline pud_t * pud_offset(pgd_t * pgd, unsigned long address) · 3 unmodified lines
58	63	#define pud_addr_end(addr, end) (end)
59	64
60	65	#endif /* __ASSEMBLY__ */
	66	+#endif /* !__ARCH_USE_5LEVEL_HACK */
61	67	#endif /* _PGTABLE_NOPUD_H */

include/asm-generic/pgtable.h+45−3

@@ -10,9 +10,9 @@ · 9 unmodified lines
10	10	#include <linux/bug.h>
11	11	#include <linux/errno.h>
12	12
13		−#if 4 - defined(__PAGETABLE_PUD_FOLDED) - defined(__PAGETABLE_PMD_FOLDED) != \
14		− CONFIG_PGTABLE_LEVELS
15		−#error CONFIG_PGTABLE_LEVELS is not consistent with __PAGETABLE_{PUD,PMD}_FOLDED
	13	+#if 5 - defined(__PAGETABLE_P4D_FOLDED) - defined(__PAGETABLE_PUD_FOLDED) - \
	14	+ defined(__PAGETABLE_PMD_FOLDED) != CONFIG_PGTABLE_LEVELS
	15	+#error CONFIG_PGTABLE_LEVELS is not consistent with __PAGETABLE_{P4D,PUD,PMD}_FOLDED
16	16	#endif
17	17
18	18	/*
@@ -424,6 +424,13 @@ static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot) · 405 unmodified lines
424	424	(__boundary - 1 < (end) - 1)? __boundary: (end); \
425	425	})
426	426
	427	+#ifndef p4d_addr_end
	428	+#define p4d_addr_end(addr, end) \
	429	+({ unsigned long __boundary = ((addr) + P4D_SIZE) & P4D_MASK; \
	430	+ (__boundary - 1 < (end) - 1)? __boundary: (end); \
	431	+})
	432	+#endif
	433	+
427	434	#ifndef pud_addr_end
428	435	#define pud_addr_end(addr, end) \
429	436	({ unsigned long __boundary = ((addr) + PUD_SIZE) & PUD_MASK; \
@@ -444,6 +451,7 @@ static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot) · 14 unmodified lines
444	451	* Do the tests inline, but report and clear the bad entry in mm/memory.c.
445	452	*/
446	453	void pgd_clear_bad(pgd_t *);
	454	+void p4d_clear_bad(p4d_t *);
447	455	void pud_clear_bad(pud_t *);
448	456	void pmd_clear_bad(pmd_t *);
449	457
@@ -458,6 +466,17 @@ static inline int pgd_none_or_clear_bad(pgd_t *pgd) · 8 unmodified lines
458	466	return 0;
459	467	}
460	468
	469	+static inline int p4d_none_or_clear_bad(p4d_t *p4d)
	470	+{
	471	+ if (p4d_none(*p4d))
	472	+ return 1;
	473	+ if (unlikely(p4d_bad(*p4d))) {
	474	+ p4d_clear_bad(p4d);
	475	+ return 1;
	476	+ }
	477	+ return 0;
	478	+}
	479	+
461	480	static inline int pud_none_or_clear_bad(pud_t *pud)
462	481	{
463	482	if (pud_none(*pud))
@@ -844,11 +863,30 @@ static inline int pmd_protnone(pmd_t pmd) · 380 unmodified lines
844	863	#endif /* CONFIG_MMU */
845	864
846	865	#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
	866	+
	867	+#ifndef __PAGETABLE_P4D_FOLDED
	868	+int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot);
	869	+int p4d_clear_huge(p4d_t *p4d);
	870	+#else
	871	+static inline int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
	872	+{
	873	+ return 0;
	874	+}
	875	+static inline int p4d_clear_huge(p4d_t *p4d)
	876	+{
	877	+ return 0;
	878	+}
	879	+#endif /* !__PAGETABLE_P4D_FOLDED */
	880	+
847	881	int pud_set_huge(pud_t *pud, phys_addr_t addr, pgprot_t prot);
848	882	int pmd_set_huge(pmd_t *pmd, phys_addr_t addr, pgprot_t prot);
849	883	int pud_clear_huge(pud_t *pud);
850	884	int pmd_clear_huge(pmd_t *pmd);
851	885	#else /* !CONFIG_HAVE_ARCH_HUGE_VMAP */
	886	+static inline int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
	887	+{
	888	+ return 0;
	889	+}
852	890	static inline int pud_set_huge(pud_t *pud, phys_addr_t addr, pgprot_t prot)
853	891	{
854	892	return 0;
@@ -857,6 +895,10 @@ static inline int pmd_set_huge(pmd_t *pmd, phys_addr_t addr, pgprot_t prot) · 2 unmodified lines
857	895	{
858	896	return 0;
859	897	}
	898	+static inline int p4d_clear_huge(p4d_t *p4d)
	899	+{
	900	+ return 0;
	901	+}
860	902	static inline int pud_clear_huge(pud_t *pud)
861	903	{
862	904	return 0;

include/asm-generic/tlb.h+11−3

@@ -270,6 +270,12 @@ static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb, · 269 unmodified lines
270	270	__pte_free_tlb(tlb, ptep, address); \
271	271	} while (0)
272	272
	273	+#define pmd_free_tlb(tlb, pmdp, address) \
	274	+ do { \
	275	+ __tlb_adjust_range(tlb, address, PAGE_SIZE); \
	276	+ __pmd_free_tlb(tlb, pmdp, address); \
	277	+ } while (0)
	278	+
273	279	#ifndef __ARCH_HAS_4LEVEL_HACK
274	280	#define pud_free_tlb(tlb, pudp, address) \
275	281	do { \
@@ -278,11 +284,13 @@ static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb, · 2 unmodified lines
278	284	} while (0)
279	285	#endif
280	286
281		−#define pmd_free_tlb(tlb, pmdp, address) \
	287	+#ifndef __ARCH_HAS_5LEVEL_HACK
	288	+#define p4d_free_tlb(tlb, pudp, address) \
282	289	do { \
283		− __tlb_adjust_range(tlb, address, PAGE_SIZE); \
284		− __pmd_free_tlb(tlb, pmdp, address); \
	290	+ __tlb_adjust_range(tlb, address, PAGE_SIZE); \
	291	+ __p4d_free_tlb(tlb, pudp, address); \
285	292	} while (0)
	293	+#endif
286	294
287	295	#define tlb_migrate_finish(mm) do {} while (0)
288	296

include/crypto/if_alg.h+1−1

@@ -73,7 +73,7 @@ int af_alg_unregister_type(const struct af_alg_type *type); · 72 unmodified lines
73	73
74	74	int af_alg_release(struct socket *sock);
75	75	void af_alg_release_parent(struct sock *sk);
76		−int af_alg_accept(struct sock sk, struct socket newsock);
	76	+int af_alg_accept(struct sock sk, struct socket newsock, bool kern);
77	77
78	78	int af_alg_make_sg(struct af_alg_sgl sgl, struct iov_iter iter, int len);
79	79	void af_alg_free_sg(struct af_alg_sgl *sgl);

include/dt-bindings/sound/cs42l42.h+1−1

@@ -20,7 +20,7 @@ · 19 unmodified lines
20	20	#define CS42L42_HPOUT_LOAD_1NF 0
21	21	#define CS42L42_HPOUT_LOAD_10NF 1
22	22
23		−/* HPOUT Clamp to GND Overide */
	23	+/* HPOUT Clamp to GND Override */
24	24	#define CS42L42_HPOUT_CLAMP_EN 0
25	25	#define CS42L42_HPOUT_CLAMP_DIS 1
26	26

include/linux/blkdev.h+0−1

@@ -435,7 +435,6 @@ struct request_queue { · 434 unmodified lines
435	435	struct delayed_work delay_work;
436	436
437	437	struct backing_dev_info *backing_dev_info;
438		− struct disk_devt *disk_devt;
439	438
440	439	/*
441	440	* The queue owner gets to use this for whatever they like.

include/linux/ceph/libceph.h+2−0

@@ -48,6 +48,7 @@ struct ceph_options { · 47 unmodified lines
48	48	unsigned long mount_timeout; /* jiffies */
49	49	unsigned long osd_idle_ttl; /* jiffies */
50	50	unsigned long osd_keepalive_timeout; /* jiffies */
	51	+ unsigned long osd_request_timeout; /* jiffies */
51	52
52	53	/*
53	54	* any type that can't be simply compared or doesn't need need
@@ -68,6 +69,7 @@ struct ceph_options { · 14 unmodified lines
68	69	#define CEPH_MOUNT_TIMEOUT_DEFAULT msecs_to_jiffies(60 * 1000)
69	70	#define CEPH_OSD_KEEPALIVE_DEFAULT msecs_to_jiffies(5 * 1000)
70	71	#define CEPH_OSD_IDLE_TTL_DEFAULT msecs_to_jiffies(60 * 1000)
	72	+#define CEPH_OSD_REQUEST_TIMEOUT_DEFAULT 0 /* no timeout */
71	73
72	74	#define CEPH_MONC_HUNT_INTERVAL msecs_to_jiffies(3 * 1000)
73	75	#define CEPH_MONC_PING_INTERVAL msecs_to_jiffies(10 * 1000)

include/linux/ceph/osd_client.h+1−0

@@ -189,6 +189,7 @@ struct ceph_osd_request { · 188 unmodified lines
189	189
190	190	/* internal */
191	191	unsigned long r_stamp; /* jiffies, send or check time */
	192	+ unsigned long r_start_stamp; /* jiffies */
192	193	int r_attempts;
193	194	struct ceph_eversion r_replay_version; /* aka reassert_version */
194	195	u32 r_last_force_resend;

include/linux/dccp.h+1−0

@@ -163,6 +163,7 @@ struct dccp_request_sock { · 162 unmodified lines
163	163	__u64 dreq_isr;
164	164	__u64 dreq_gsr;
165	165	__be32 dreq_service;
	166	+ spinlock_t dreq_lock;
166	167	struct list_head dreq_featneg;
167	168	__u32 dreq_timestamp_echo;
168	169	__u32 dreq_timestamp_time;

include/linux/filter.h+12−4

@@ -409,6 +409,7 @@ struct bpf_prog { · 408 unmodified lines
409	409	u16 pages; /* Number of allocated pages */
410	410	kmemcheck_bitfield_begin(meta);
411	411	u16 jited:1, /* Is our filter JIT'ed? */
	412	+ locked:1, /* Program image locked? */
412	413	gpl_compatible:1, /* Is filter GPL compatible? */
413	414	cb_access:1, /* Is control block accessed? */
414	415	dst_needed:1, /* Do we need dst entry? */
@@ -554,22 +555,29 @@ static inline bool bpf_prog_was_classic(const struct bpf_prog *prog) · 139 unmodified lines
554	555	#ifdef CONFIG_ARCH_HAS_SET_MEMORY
555	556	static inline void bpf_prog_lock_ro(struct bpf_prog *fp)
556	557	{
557		− set_memory_ro((unsigned long)fp, fp->pages);
	558	+ fp->locked = 1;
	559	+ WARN_ON_ONCE(set_memory_ro((unsigned long)fp, fp->pages));
558	560	}
559	561
560	562	static inline void bpf_prog_unlock_ro(struct bpf_prog *fp)
561	563	{
562		− set_memory_rw((unsigned long)fp, fp->pages);
	564	+ if (fp->locked) {
	565	+ WARN_ON_ONCE(set_memory_rw((unsigned long)fp, fp->pages));
	566	+ /* In case set_memory_rw() fails, we want to be the first
	567	+ * to crash here instead of some random place later on.
	568	+ */
	569	+ fp->locked = 0;
	570	+ }
563	571	}
564	572
565	573	static inline void bpf_jit_binary_lock_ro(struct bpf_binary_header *hdr)
566	574	{
567		− set_memory_ro((unsigned long)hdr, hdr->pages);
	575	+ WARN_ON_ONCE(set_memory_ro((unsigned long)hdr, hdr->pages));
568	576	}
569	577
570	578	static inline void bpf_jit_binary_unlock_ro(struct bpf_binary_header *hdr)
571	579	{
572		− set_memory_rw((unsigned long)hdr, hdr->pages);
	580	+ WARN_ON_ONCE(set_memory_rw((unsigned long)hdr, hdr->pages));
573	581	}
574	582	#else
575	583	static inline void bpf_prog_lock_ro(struct bpf_prog *fp)

include/linux/fs.h+1−1

@@ -2678,7 +2678,7 @@ static const char * const kernel_read_file_str[] = { · 2677 unmodified lines
2678	2678
2679	2679	static inline const char *kernel_read_file_id_str(enum kernel_read_file_id id)
2680	2680	{
2681		− if (id < 0 \|\| id >= READING_MAX_ID)
	2681	+ if ((unsigned)id >= READING_MAX_ID)
2682	2682	return kernel_read_file_str[READING_UNKNOWN];
2683	2683
2684	2684	return kernel_read_file_str[id];

include/linux/genhd.h+0−8

@@ -167,13 +167,6 @@ struct blk_integrity { · 166 unmodified lines
167	167	};
168	168
169	169	#endif /* CONFIG_BLK_DEV_INTEGRITY */
170		−struct disk_devt {
171		− atomic_t count;
172		− void (release)(struct disk_devt disk_devt);
173		−};
174		−
175		−void put_disk_devt(struct disk_devt *disk_devt);
176		−void get_disk_devt(struct disk_devt *disk_devt);
177	170
178	171	struct gendisk {
179	172	/* major, first_minor and minors are input parameters only,
@@ -183,7 +176,6 @@ struct gendisk { · 3 unmodified lines
183	176	int first_minor;
184	177	int minors; /* maximum number of minors, =1 for
185	178	* disks that can't be partitioned. */
186		− struct disk_devt *disk_devt;
187	179
188	180	char disk_name[DISK_NAME_LEN]; /* name of major driver */
189	181	char (devnode)(struct gendisk gd, umode_t mode);

include/linux/hugetlb.h+4−1

@@ -122,7 +122,7 @@ struct page follow_huge_pmd(struct mm_struct mm, unsigned long address, · 121 unmodified lines
122	122	struct page follow_huge_pud(struct mm_struct mm, unsigned long address,
123	123	pud_t *pud, int flags);
124	124	int pmd_huge(pmd_t pmd);
125		−int pud_huge(pud_t pmd);
	125	+int pud_huge(pud_t pud);
126	126	unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
127	127	unsigned long address, unsigned long end, pgprot_t newprot);
128	128
@@ -197,6 +197,9 @@ static inline void __unmap_hugepage_range(struct mmu_gather *tlb, · 68 unmodified lines
197	197	#ifndef pgd_huge
198	198	#define pgd_huge(x) 0
199	199	#endif
	200	+#ifndef p4d_huge
	201	+#define p4d_huge(x) 0
	202	+#endif
200	203
201	204	#ifndef pgd_write
202	205	static inline int pgd_write(pgd_t pgd)

include/linux/irqchip/arm-gic-v3.h+2−0

@@ -373,6 +373,8 @@ · 372 unmodified lines
373	373	#define ICC_IGRPEN0_EL1_MASK (1 << ICC_IGRPEN0_EL1_SHIFT)
374	374	#define ICC_IGRPEN1_EL1_SHIFT 0
375	375	#define ICC_IGRPEN1_EL1_MASK (1 << ICC_IGRPEN1_EL1_SHIFT)
	376	+#define ICC_SRE_EL1_DIB (1U << 2)
	377	+#define ICC_SRE_EL1_DFB (1U << 1)
376	378	#define ICC_SRE_EL1_SRE (1U << 0)
377	379
378	380	/*

include/linux/irqdomain.h+4−0

@@ -524,6 +524,10 @@ static inline struct irq_domain *irq_find_matching_fwnode( · 523 unmodified lines
524	524	{
525	525	return NULL;
526	526	}
	527	+static inline bool irq_domain_check_msi_remap(void)
	528	+{
	529	+ return false;
	530	+}
527	531	#endif /* !CONFIG_IRQ_DOMAIN */
528	532
529	533	#endif /* _LINUX_IRQDOMAIN_H */

include/linux/jump_label.h+9−2

@@ -90,6 +90,13 @@ extern bool static_key_initialized; · 89 unmodified lines
90	90	struct static_key {
91	91	atomic_t enabled;
92	92	/*
	93	+ * Note:
	94	+ * To make anonymous unions work with old compilers, the static
	95	+ * initialization of them requires brackets. This creates a dependency
	96	+ * on the order of the struct with the initializers. If any fields
	97	+ * are added, STATIC_KEY_INIT_TRUE and STATIC_KEY_INIT_FALSE may need
	98	+ * to be modified.
	99	+ *
93	100	* bit 0 => 1 if key is initially true
94	101	* 0 if initially false
95	102	* bit 1 => 1 if points to struct static_key_mod
@@ -166,10 +173,10 @@ extern void static_key_disable(struct static_key *key); · 70 unmodified lines
166	173	*/
167	174	#define STATIC_KEY_INIT_TRUE \
168	175	{ .enabled = { 1 }, \
169		− .entries = (void *)JUMP_TYPE_TRUE }
	176	+ { .entries = (void *)JUMP_TYPE_TRUE } }
170	177	#define STATIC_KEY_INIT_FALSE \
171	178	{ .enabled = { 0 }, \
172		− .entries = (void *)JUMP_TYPE_FALSE }
	179	+ { .entries = (void *)JUMP_TYPE_FALSE } }
173	180
174	181	#else /* !HAVE_JUMP_LABEL */
175	182

include/linux/kasan.h+1−0

@@ -18,6 +18,7 @@ extern unsigned char kasan_zero_page[PAGE_SIZE]; · 17 unmodified lines
18	18	extern pte_t kasan_zero_pte[PTRS_PER_PTE];
19	19	extern pmd_t kasan_zero_pmd[PTRS_PER_PMD];
20	20	extern pud_t kasan_zero_pud[PTRS_PER_PUD];
	21	+extern p4d_t kasan_zero_p4d[PTRS_PER_P4D];
21	22
22	23	void kasan_populate_zero_shadow(const void *shadow_start,
23	24	const void *shadow_end);

include/linux/list_nulls.h+5−0

@@ -29,6 +29,11 @@ struct hlist_nulls_node { · 28 unmodified lines
29	29	((ptr)->first = (struct hlist_nulls_node *) NULLS_MARKER(nulls))
30	30
31	31	#define hlist_nulls_entry(ptr, type, member) container_of(ptr,type,member)
	32	+
	33	+#define hlist_nulls_entry_safe(ptr, type, member) \
	34	+ ({ typeof(ptr) ____ptr = (ptr); \
	35	+ !is_a_nulls(____ptr) ? hlist_nulls_entry(____ptr, type, member) : NULL; \
	36	+ })
32	37	/**
33	38	* ptr_is_a_nulls - Test if a ptr is a nulls
34	39	* @ptr: ptr to be tested

include/linux/mm.h+28−6

@@ -1560,14 +1560,24 @@ static inline pte_t get_locked_pte(struct mm_struct mm, unsigned long addr, · 1559 unmodified lines
1560	1560	return ptep;
1561	1561	}
1562	1562
	1563	+#ifdef __PAGETABLE_P4D_FOLDED
	1564	+static inline int __p4d_alloc(struct mm_struct mm, pgd_t pgd,
	1565	+ unsigned long address)
	1566	+{
	1567	+ return 0;
	1568	+}
	1569	+#else
	1570	+int __p4d_alloc(struct mm_struct mm, pgd_t pgd, unsigned long address);
	1571	+#endif
	1572	+
1563	1573	#ifdef __PAGETABLE_PUD_FOLDED
1564		−static inline int __pud_alloc(struct mm_struct mm, pgd_t pgd,
	1574	+static inline int __pud_alloc(struct mm_struct mm, p4d_t p4d,
1565	1575	unsigned long address)
1566	1576	{
1567	1577	return 0;
1568	1578	}
1569	1579	#else
1570		−int __pud_alloc(struct mm_struct mm, pgd_t pgd, unsigned long address);
	1580	+int __pud_alloc(struct mm_struct mm, p4d_t p4d, unsigned long address);
1571	1581	#endif
1572	1582
1573	1583	#if defined(__PAGETABLE_PMD_FOLDED) \|\| !defined(CONFIG_MMU)
@@ -1619,11 +1629,22 @@ int __pte_alloc_kernel(pmd_t *pmd, unsigned long address); · 45 unmodified lines
1619	1629	* Remove it when 4level-fixup.h has been removed.
1620	1630	*/
1621	1631	#if defined(CONFIG_MMU) && !defined(__ARCH_HAS_4LEVEL_HACK)
1622		−static inline pud_t pud_alloc(struct mm_struct mm, pgd_t *pgd, unsigned long address)
	1632	+
	1633	+#ifndef __ARCH_HAS_5LEVEL_HACK
	1634	+static inline p4d_t p4d_alloc(struct mm_struct mm, pgd_t *pgd,
	1635	+ unsigned long address)
	1636	+{
	1637	+ return (unlikely(pgd_none(*pgd)) && __p4d_alloc(mm, pgd, address)) ?
	1638	+ NULL : p4d_offset(pgd, address);
	1639	+}
	1640	+
	1641	+static inline pud_t pud_alloc(struct mm_struct mm, p4d_t *p4d,
	1642	+ unsigned long address)
1623	1643	{
1624		− return (unlikely(pgd_none(*pgd)) && __pud_alloc(mm, pgd, address))?
1625		− NULL: pud_offset(pgd, address);
	1644	+ return (unlikely(p4d_none(*p4d)) && __pud_alloc(mm, p4d, address)) ?
	1645	+ NULL : pud_offset(p4d, address);
1626	1646	}
	1647	+#endif /* !__ARCH_HAS_5LEVEL_HACK */
1627	1648
1628	1649	static inline pmd_t pmd_alloc(struct mm_struct mm, pud_t *pud, unsigned long address)
1629	1650	{
@@ -2385,7 +2406,8 @@ void sparse_mem_maps_populate_node(struct page **map_map, · 755 unmodified lines
2385	2406
2386	2407	struct page *sparse_mem_map_populate(unsigned long pnum, int nid);
2387	2408	pgd_t *vmemmap_pgd_populate(unsigned long addr, int node);
2388		−pud_t vmemmap_pud_populate(pgd_t pgd, unsigned long addr, int node);
	2409	+p4d_t vmemmap_p4d_populate(pgd_t pgd, unsigned long addr, int node);
	2410	+pud_t vmemmap_pud_populate(p4d_t p4d, unsigned long addr, int node);
2389	2411	pmd_t vmemmap_pmd_populate(pud_t pud, unsigned long addr, int node);
2390	2412	pte_t vmemmap_pte_populate(pmd_t pmd, unsigned long addr, int node);
2391	2413	void *vmemmap_alloc_block(unsigned long size, int node);

include/linux/net.h+1−1

@@ -146,7 +146,7 @@ struct proto_ops { · 145 unmodified lines
146	146	int (socketpair)(struct socket sock1,
147	147	struct socket *sock2);
148	148	int (accept) (struct socket sock,
149		− struct socket *newsock, int flags);
	149	+ struct socket *newsock, int flags, bool kern);
150	150	int (getname) (struct socket sock,
151	151	struct sockaddr *addr,
152	152	int *sockaddr_len, int peer);

include/linux/phy.h+4−0

@@ -837,6 +837,10 @@ int genphy_read_status(struct phy_device *phydev); · 836 unmodified lines
837	837	int genphy_suspend(struct phy_device *phydev);
838	838	int genphy_resume(struct phy_device *phydev);
839	839	int genphy_soft_reset(struct phy_device *phydev);
	840	+static inline int genphy_no_soft_reset(struct phy_device *phydev)
	841	+{
	842	+ return 0;
	843	+}
840	844	void phy_driver_unregister(struct phy_driver *drv);
841	845	void phy_drivers_unregister(struct phy_driver *drv, int n);
842	846	int phy_driver_register(struct phy_driver new_driver, struct module owner);

include/linux/purgatory.h+23−0

@@ -0,0 +1,23 @@
	1	+#ifndef _LINUX_PURGATORY_H
	2	+#define _LINUX_PURGATORY_H
	3	+
	4	+#include <linux/types.h>
	5	+#include <crypto/sha.h>
	6	+#include <uapi/linux/kexec.h>
	7	+
	8	+struct kexec_sha_region {
	9	+ unsigned long start;
	10	+ unsigned long len;
	11	+};
	12	+
	13	+/*
	14	+ * These forward declarations serve two purposes:
	15	+ *
	16	+ * 1) Make sparse happy when checking arch/purgatory
	17	+ * 2) Document that these are required to be global so the symbol
	18	+ * lookup in kexec works
	19	+ */
	20	+extern struct kexec_sha_region purgatory_sha_regions[KEXEC_SEGMENT_MAX];
	21	+extern u8 purgatory_sha256_digest[SHA256_DIGEST_SIZE];
	22	+
	23	+#endif

include/linux/random.h+15−3

@@ -37,14 +37,26 @@ extern void get_random_bytes(void *buf, int nbytes); · 36 unmodified lines
37	37	extern int add_random_ready_callback(struct random_ready_callback *rdy);
38	38	extern void del_random_ready_callback(struct random_ready_callback *rdy);
39	39	extern void get_random_bytes_arch(void *buf, int nbytes);
40		−extern int random_int_secret_init(void);
41	40
42	41	#ifndef MODULE
43	42	extern const struct file_operations random_fops, urandom_fops;
44	43	#endif
45	44
46		−unsigned int get_random_int(void);
47		−unsigned long get_random_long(void);
	45	+u32 get_random_u32(void);
	46	+u64 get_random_u64(void);
	47	+static inline unsigned int get_random_int(void)
	48	+{
	49	+ return get_random_u32();
	50	+}
	51	+static inline unsigned long get_random_long(void)
	52	+{
	53	+#if BITS_PER_LONG == 64
	54	+ return get_random_u64();
	55	+#else
	56	+ return get_random_u32();
	57	+#endif
	58	+}
	59	+
48	60	unsigned long randomize_page(unsigned long start, unsigned long range);
49	61
50	62	u32 prandom_u32(void);

include/linux/rculist_nulls.h+14−0

@@ -156,5 +156,19 @@ static inline void hlist_nulls_add_tail_rcu(struct hlist_nulls_node *n, · 155 unmodified lines
156	156	({ tpos = hlist_nulls_entry(pos, typeof(*tpos), member); 1; }); \
157	157	pos = rcu_dereference_raw(hlist_nulls_next_rcu(pos)))
158	158
	159	+/**
	160	+ * hlist_nulls_for_each_entry_safe -
	161	+ * iterate over list of given type safe against removal of list entry
	162	+ * @tpos: the type * to use as a loop cursor.
	163	+ * @pos: the &struct hlist_nulls_node to use as a loop cursor.
	164	+ * @head: the head for your list.
	165	+ * @member: the name of the hlist_nulls_node within the struct.
	166	+ */
	167	+#define hlist_nulls_for_each_entry_safe(tpos, pos, head, member) \
	168	+ for (({barrier();}), \
	169	+ pos = rcu_dereference_raw(hlist_nulls_first_rcu(head)); \
	170	+ (!is_a_nulls(pos)) && \
	171	+ ({ tpos = hlist_nulls_entry(pos, typeof(*tpos), member); \
	172	+ pos = rcu_dereference_raw(hlist_nulls_next_rcu(pos)); 1; });)
159	173	#endif
160	174	#endif

include/linux/regulator/machine.h+1−1

@@ -65,7 +65,7 @@ struct regulator_state { · 64 unmodified lines
65	65	int uV; /* suspend voltage */
66	66	unsigned int mode; /* suspend regulator operating mode */
67	67	int enabled; /* is regulator enabled in this suspend state */
68		− int disabled; /* is the regulator disbled in this suspend state */
	68	+ int disabled; /* is the regulator disabled in this suspend state */
69	69	};
70	70
71	71	/**

include/linux/user_namespace.h+1−1

@@ -72,7 +72,7 @@ struct ucounts { · 71 unmodified lines
72	72	struct hlist_node node;
73	73	struct user_namespace *ns;
74	74	kuid_t uid;
75		− atomic_t count;
	75	+ int count;
76	76	atomic_t ucount[UCOUNT_COUNTS];
77	77	};
78	78

include/linux/userfaultfd_k.h+3−10

@@ -61,8 +61,7 @@ extern void mremap_userfaultfd_complete(struct vm_userfaultfd_ctx *, · 60 unmodified lines
61	61	unsigned long from, unsigned long to,
62	62	unsigned long len);
63	63
64		−extern void userfaultfd_remove(struct vm_area_struct *vma,
65		− struct vm_area_struct **prev,
	64	+extern bool userfaultfd_remove(struct vm_area_struct *vma,
66	65	unsigned long start,
67	66	unsigned long end);
68	67
@@ -72,8 +71,6 @@ extern int userfaultfd_unmap_prep(struct vm_area_struct *vma, · 3 unmodified lines
72	71	extern void userfaultfd_unmap_complete(struct mm_struct *mm,
73	72	struct list_head *uf);
74	73
75		−extern void userfaultfd_exit(struct mm_struct *mm);
76		−
77	74	#else /* CONFIG_USERFAULTFD */
78	75
79	76	/* mm helpers */
@@ -120,11 +117,11 @@ static inline void mremap_userfaultfd_complete(struct vm_userfaultfd_ctx *ctx, · 40 unmodified lines
120	117	{
121	118	}
122	119
123		−static inline void userfaultfd_remove(struct vm_area_struct *vma,
124		− struct vm_area_struct **prev,
	120	+static inline bool userfaultfd_remove(struct vm_area_struct *vma,
125	121	unsigned long start,
126	122	unsigned long end)
127	123	{
	124	+ return true;
128	125	}
129	126
130	127	static inline int userfaultfd_unmap_prep(struct vm_area_struct *vma,
@@ -139,10 +136,6 @@ static inline void userfaultfd_unmap_complete(struct mm_struct *mm, · 8 unmodified lines
139	136	{
140	137	}
141	138
142		−static inline void userfaultfd_exit(struct mm_struct *mm)
143		−{
144		−}
145		−
146	139	#endif /* CONFIG_USERFAULTFD */
147	140
148	141	#endif /* _LINUX_USERFAULTFD_K_H */

include/linux/vm_event_item.h+3−0

@@ -79,6 +79,9 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT, · 78 unmodified lines
79	79	THP_SPLIT_PAGE_FAILED,
80	80	THP_DEFERRED_SPLIT_PAGE,
81	81	THP_SPLIT_PMD,
	82	+#ifdef CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
	83	+ THP_SPLIT_PUD,
	84	+#endif
82	85	THP_ZERO_PAGE_ALLOC,
83	86	THP_ZERO_PAGE_ALLOC_FAILED,
84	87	#endif

include/linux/wait.h+10−21

@@ -620,30 +620,19 @@ do { \ · 619 unmodified lines
620	620	__ret; \
621	621	})
622	622
	623	+extern int do_wait_intr(wait_queue_head_t , wait_queue_t );
	624	+extern int do_wait_intr_irq(wait_queue_head_t , wait_queue_t );
623	625
624		−#define __wait_event_interruptible_locked(wq, condition, exclusive, irq) \
	626	+#define __wait_event_interruptible_locked(wq, condition, exclusive, fn) \
625	627	({ \
626		− int __ret = 0; \
	628	+ int __ret; \
627	629	DEFINE_WAIT(__wait); \
628	630	if (exclusive) \
629	631	__wait.flags \|= WQ_FLAG_EXCLUSIVE; \
630	632	do { \
631		− if (likely(list_empty(&__wait.task_list))) \
632		− __add_wait_queue_tail(&(wq), &__wait); \
633		− set_current_state(TASK_INTERRUPTIBLE); \
634		− if (signal_pending(current)) { \
635		− __ret = -ERESTARTSYS; \
	633	+ __ret = fn(&(wq), &__wait); \
	634	+ if (__ret) \
636	635	break; \
637		− } \
638		− if (irq) \
639		− spin_unlock_irq(&(wq).lock); \
640		− else \
641		− spin_unlock(&(wq).lock); \
642		− schedule(); \
643		− if (irq) \
644		− spin_lock_irq(&(wq).lock); \
645		− else \
646		− spin_lock(&(wq).lock); \
647	636	} while (!(condition)); \
648	637	__remove_wait_queue(&(wq), &__wait); \
649	638	__set_current_state(TASK_RUNNING); \
@@ -676,7 +665,7 @@ do { \ · 26 unmodified lines
676	665	*/
677	666	#define wait_event_interruptible_locked(wq, condition) \
678	667	((condition) \
679		− ? 0 : __wait_event_interruptible_locked(wq, condition, 0, 0))
	668	+ ? 0 : __wait_event_interruptible_locked(wq, condition, 0, do_wait_intr))
680	669
681	670	/**
682	671	* wait_event_interruptible_locked_irq - sleep until a condition gets true
@@ -703,7 +692,7 @@ do { \ · 20 unmodified lines
703	692	*/
704	693	#define wait_event_interruptible_locked_irq(wq, condition) \
705	694	((condition) \
706		− ? 0 : __wait_event_interruptible_locked(wq, condition, 0, 1))
	695	+ ? 0 : __wait_event_interruptible_locked(wq, condition, 0, do_wait_intr_irq))
707	696
708	697	/**
709	698	* wait_event_interruptible_exclusive_locked - sleep exclusively until a condition gets true
@@ -734,7 +723,7 @@ do { \ · 24 unmodified lines
734	723	*/
735	724	#define wait_event_interruptible_exclusive_locked(wq, condition) \
736	725	((condition) \
737		− ? 0 : __wait_event_interruptible_locked(wq, condition, 1, 0))
	726	+ ? 0 : __wait_event_interruptible_locked(wq, condition, 1, do_wait_intr))
738	727
739	728	/**
740	729	* wait_event_interruptible_exclusive_locked_irq - sleep until a condition gets true
@@ -765,7 +754,7 @@ do { \ · 24 unmodified lines
765	754	*/
766	755	#define wait_event_interruptible_exclusive_locked_irq(wq, condition) \
767	756	((condition) \
768		− ? 0 : __wait_event_interruptible_locked(wq, condition, 1, 1))
	757	+ ? 0 : __wait_event_interruptible_locked(wq, condition, 1, do_wait_intr_irq))
769	758
770	759
771	760	#define __wait_event_killable(wq, condition) \

include/media/vsp1.h+11−2

@@ -20,8 +20,17 @@ struct device; · 19 unmodified lines
20	20
21	21	int vsp1_du_init(struct device *dev);
22	22
23		−int vsp1_du_setup_lif(struct device *dev, unsigned int width,
24		− unsigned int height);
	23	+/**
	24	+ * struct vsp1_du_lif_config - VSP LIF configuration
	25	+ * @width: output frame width
	26	+ * @height: output frame height
	27	+ */
	28	+struct vsp1_du_lif_config {
	29	+ unsigned int width;
	30	+ unsigned int height;
	31	+};
	32	+
	33	+int vsp1_du_setup_lif(struct device dev, const struct vsp1_du_lif_config cfg);
25	34
26	35	struct vsp1_du_atomic_config {
27	36	u32 pixelformat;

include/net/inet_common.h+2−1

@@ -20,7 +20,8 @@ int __inet_stream_connect(struct socket sock, struct sockaddr uaddr, · 19 unmodified lines
20	20	int addr_len, int flags, int is_sendmsg);
21	21	int inet_dgram_connect(struct socket sock, struct sockaddr uaddr,
22	22	int addr_len, int flags);
23		−int inet_accept(struct socket sock, struct socket newsock, int flags);
	23	+int inet_accept(struct socket sock, struct socket newsock, int flags,
	24	+ bool kern);
24	25	int inet_sendmsg(struct socket sock, struct msghdr msg, size_t size);
25	26	ssize_t inet_sendpage(struct socket sock, struct page page, int offset,
26	27	size_t size, int flags);

include/net/inet_connection_sock.h+1−1

@@ -258,7 +258,7 @@ inet_csk_rto_backoff(const struct inet_connection_sock *icsk, · 257 unmodified lines
258	258	return (unsigned long)min_t(u64, when, max_when);
259	259	}
260	260
261		−struct sock inet_csk_accept(struct sock sk, int flags, int *err);
	261	+struct sock inet_csk_accept(struct sock sk, int flags, int *err, bool kern);
262	262
263	263	int inet_csk_get_port(struct sock *sk, unsigned short snum);
264	264

include/net/irda/timer.h+1−1

@@ -59,7 +59,7 @@ struct lap_cb; · 58 unmodified lines
59	59	* Slot timer must never exceed 85 ms, and must always be at least 25 ms,
60	60	* suggested to 75-85 msec by IrDA lite. This doesn't work with a lot of
61	61	* devices, and other stackes uses a lot more, so it's best we do it as well
62		− * (Note : this is the default value and sysctl overides it - Jean II)
	62	+ * (Note : this is the default value and sysctl overrides it - Jean II)
63	63	*/
64	64	#define SLOT_TIMEOUT (90*HZ/1000)
65	65

include/net/sctp/structs.h+2−1

@@ -476,7 +476,8 @@ struct sctp_pf { · 475 unmodified lines
476	476	int (send_verify) (struct sctp_sock , union sctp_addr *);
477	477	int (supported_addrs)(const struct sctp_sock , __be16 *);
478	478	struct sock (create_accept_sk) (struct sock *sk,
479		− struct sctp_association *asoc);
	479	+ struct sctp_association *asoc,
	480	+ bool kern);
480	481	int (addr_to_user)(struct sctp_sock sk, union sctp_addr *addr);
481	482	void (to_sk_saddr)(union sctp_addr , struct sock *sk);
482	483	void (to_sk_daddr)(union sctp_addr , struct sock *sk);

include/net/sock.h+6−3

@@ -236,6 +236,7 @@ struct sock_common { · 235 unmodified lines
236	236	* @sk_shutdown: mask of %SEND_SHUTDOWN and/or %RCV_SHUTDOWN
237	237	* @sk_userlocks: %SO_SNDBUF and %SO_RCVBUF settings
238	238	* @sk_lock: synchronizer
	239	+ * @sk_kern_sock: True if sock is using kernel lock classes
239	240	* @sk_rcvbuf: size of receive buffer in bytes
240	241	* @sk_wq: sock wait queue and async head
241	242	* @sk_rx_dst: receive input route used by early demux
@@ -430,7 +431,8 @@ struct sock { · 188 unmodified lines
430	431	#endif
431	432
432	433	kmemcheck_bitfield_begin(flags);
433		− unsigned int sk_padding : 2,
	434	+ unsigned int sk_padding : 1,
	435	+ sk_kern_sock : 1,
434	436	sk_no_check_tx : 1,
435	437	sk_no_check_rx : 1,
436	438	sk_userlocks : 4,
@@ -1015,7 +1017,8 @@ struct proto { · 578 unmodified lines
1015	1017	int addr_len);
1016	1018	int (disconnect)(struct sock sk, int flags);
1017	1019
1018		− struct sock * (accept)(struct sock sk, int flags, int *err);
	1020	+ struct sock * (accept)(struct sock sk, int flags, int *err,
	1021	+ bool kern);
1019	1022
1020	1023	int (ioctl)(struct sock sk, int cmd,
1021	1024	unsigned long arg);
@@ -1573,7 +1576,7 @@ int sock_cmsg_send(struct sock sk, struct msghdr msg, · 551 unmodified lines
1573	1576	int sock_no_bind(struct socket , struct sockaddr , int);
1574	1577	int sock_no_connect(struct socket , struct sockaddr , int, int);
1575	1578	int sock_no_socketpair(struct socket , struct socket );
1576		−int sock_no_accept(struct socket , struct socket , int);
	1579	+int sock_no_accept(struct socket , struct socket , int, bool);
1577	1580	int sock_no_getname(struct socket , struct sockaddr , int *, int);
1578	1581	unsigned int sock_no_poll(struct file , struct socket ,
1579	1582	struct poll_table_struct *);

include/scsi/libiscsi.h+1−0

@@ -196,6 +196,7 @@ struct iscsi_conn { · 195 unmodified lines
196	196	struct iscsi_task task; / xmit task in progress */
197	197
198	198	/* xmit */
	199	+ spinlock_t taskqueuelock; /* protects the next three lists */
199	200	struct list_head mgmtqueue; /* mgmt (control) xmit queue */
200	201	struct list_head cmdqueue; /* data-path cmd queue */
201	202	struct list_head requeue; /* tasks needing another run */

include/scsi/scsi_device.h+4−0

@@ -472,6 +472,10 @@ static inline int scsi_device_created(struct scsi_device *sdev) · 471 unmodified lines
472	472	sdev->sdev_state == SDEV_CREATED_BLOCK;
473	473	}
474	474
	475	+int scsi_internal_device_block(struct scsi_device *sdev, bool wait);
	476	+int scsi_internal_device_unblock(struct scsi_device *sdev,
	477	+ enum scsi_device_state new_state);
	478	+
475	479	/* accessor functions for the SCSI parameters */
476	480	static inline int scsi_device_sync(struct scsi_device *sdev)
477	481	{

include/trace/events/syscalls.h+1−0

@@ -1,5 +1,6 @@
1	1	#undef TRACE_SYSTEM
2	2	#define TRACE_SYSTEM raw_syscalls
	3	+#undef TRACE_INCLUDE_FILE
3	4	#define TRACE_INCLUDE_FILE syscalls
4	5
5	6	#if !defined(_TRACE_EVENTS_SYSCALLS_H) \|\| defined(TRACE_HEADER_MULTI_READ)

include/uapi/linux/packet_diag.h+1−1

@@ -64,7 +64,7 @@ struct packet_diag_mclist { · 63 unmodified lines
64	64	__u32 pdmc_count;
65	65	__u16 pdmc_type;
66	66	__u16 pdmc_alen;
67		− __u8 pdmc_addr[MAX_ADDR_LEN];
	67	+ __u8 pdmc_addr[32]; /* MAX_ADDR_LEN */
68	68	};
69	69
70	70	struct packet_diag_ring {

include/uapi/linux/userfaultfd.h+1−4

@@ -18,8 +18,7 @@ · 17 unmodified lines
18	18	* means the userland is reading).
19	19	*/
20	20	#define UFFD_API ((__u64)0xAA)
21		−#define UFFD_API_FEATURES (UFFD_FEATURE_EVENT_EXIT \| \
22		− UFFD_FEATURE_EVENT_FORK \| \
	21	+#define UFFD_API_FEATURES (UFFD_FEATURE_EVENT_FORK \| \
23	22	UFFD_FEATURE_EVENT_REMAP \| \
24	23	UFFD_FEATURE_EVENT_REMOVE \| \
25	24	UFFD_FEATURE_EVENT_UNMAP \| \
@@ -113,7 +112,6 @@ struct uffd_msg { · 87 unmodified lines
113	112	#define UFFD_EVENT_REMAP 0x14
114	113	#define UFFD_EVENT_REMOVE 0x15
115	114	#define UFFD_EVENT_UNMAP 0x16
116		−#define UFFD_EVENT_EXIT 0x17
117	115
118	116	/* flags for UFFD_EVENT_PAGEFAULT */
119	117	#define UFFD_PAGEFAULT_FLAG_WRITE (1<<0) /* If this was a write fault */
@@ -163,7 +161,6 @@ struct uffdio_api { · 43 unmodified lines
163	161	#define UFFD_FEATURE_MISSING_HUGETLBFS (1<<4)
164	162	#define UFFD_FEATURE_MISSING_SHMEM (1<<5)
165	163	#define UFFD_FEATURE_EVENT_UNMAP (1<<6)
166		−#define UFFD_FEATURE_EVENT_EXIT (1<<7)
167	164	__u64 features;
168	165
169	166	__u64 ioctls;

include/xen/swiotlb-xen.h+11−0

@@ -2,6 +2,7 @@ · 1 unmodified line
2	2	#define __LINUX_SWIOTLB_XEN_H
3	3
4	4	#include <linux/dma-direction.h>
	5	+#include <linux/scatterlist.h>
5	6	#include <linux/swiotlb.h>
6	7
7	8	extern int xen_swiotlb_init(int verbose, bool early);
@@ -55,4 +56,14 @@ xen_swiotlb_dma_supported(struct device *hwdev, u64 mask); · 47 unmodified lines
55	56
56	57	extern int
57	58	xen_swiotlb_set_dma_mask(struct device *dev, u64 dma_mask);
	59	+
	60	+extern int
	61	+xen_swiotlb_dma_mmap(struct device dev, struct vm_area_struct vma,
	62	+ void *cpu_addr, dma_addr_t dma_addr, size_t size,
	63	+ unsigned long attrs);
	64	+
	65	+extern int
	66	+xen_swiotlb_get_sgtable(struct device dev, struct sg_table sgt,
	67	+ void *cpu_addr, dma_addr_t handle, size_t size,
	68	+ unsigned long attrs);
58	69	#endif /* __LINUX_SWIOTLB_XEN_H */

init/main.c+0−1

@@ -882,7 +882,6 @@ static void __init do_basic_setup(void) · 881 unmodified lines
882	882	do_ctors();
883	883	usermodehelper_enable();
884	884	do_initcalls();
885		− random_int_secret_init();
886	885	}
887	886
888	887	static void __init do_pre_smp_initcalls(void)

kernel/bpf/hashtab.c+80−39

@@ -13,11 +13,12 @@ · 12 unmodified lines
13	13	#include <linux/bpf.h>
14	14	#include <linux/jhash.h>
15	15	#include <linux/filter.h>
	16	+#include <linux/rculist_nulls.h>
16	17	#include "percpu_freelist.h"
17	18	#include "bpf_lru_list.h"
18	19
19	20	struct bucket {
20		− struct hlist_head head;
	21	+ struct hlist_nulls_head head;
21	22	raw_spinlock_t lock;
22	23	};
23	24
@@ -44,9 +45,14 @@ enum extra_elem_state { · 20 unmodified lines
44	45	/* each htab element is struct htab_elem + key + value */
45	46	struct htab_elem {
46	47	union {
47		− struct hlist_node hash_node;
48		− struct bpf_htab *htab;
49		− struct pcpu_freelist_node fnode;
	48	+ struct hlist_nulls_node hash_node;
	49	+ struct {
	50	+ void *padding;
	51	+ union {
	52	+ struct bpf_htab *htab;
	53	+ struct pcpu_freelist_node fnode;
	54	+ };
	55	+ };
50	56	};
51	57	union {
52	58	struct rcu_head rcu;
@@ -162,7 +168,8 @@ skip_percpu_elems: · 109 unmodified lines
162	168	offsetof(struct htab_elem, lru_node),
163	169	htab->elem_size, htab->map.max_entries);
164	170	else
165		− pcpu_freelist_populate(&htab->freelist, htab->elems,
	171	+ pcpu_freelist_populate(&htab->freelist,
	172	+ htab->elems + offsetof(struct htab_elem, fnode),
166	173	htab->elem_size, htab->map.max_entries);
167	174
168	175	return 0;
@@ -217,6 +224,11 @@ static struct bpf_map htab_map_alloc(union bpf_attr attr) · 48 unmodified lines
217	224	int err, i;
218	225	u64 cost;
219	226
	227	+ BUILD_BUG_ON(offsetof(struct htab_elem, htab) !=
	228	+ offsetof(struct htab_elem, hash_node.pprev));
	229	+ BUILD_BUG_ON(offsetof(struct htab_elem, fnode.next) !=
	230	+ offsetof(struct htab_elem, hash_node.pprev));
	231	+
220	232	if (lru && !capable(CAP_SYS_ADMIN))
221	233	/* LRU implementation is much complicated than other
222	234	* maps. Hence, limit to CAP_SYS_ADMIN for now.
@@ -326,7 +338,7 @@ static struct bpf_map htab_map_alloc(union bpf_attr attr) · 103 unmodified lines
326	338	goto free_htab;
327	339
328	340	for (i = 0; i < htab->n_buckets; i++) {
329		− INIT_HLIST_HEAD(&htab->buckets[i].head);
	341	+ INIT_HLIST_NULLS_HEAD(&htab->buckets[i].head, i);
330	342	raw_spin_lock_init(&htab->buckets[i].lock);
331	343	}
332	344
@@ -366,20 +378,44 @@ static inline struct bucket __select_bucket(struct bpf_htab htab, u32 hash) · 33 unmodified lines
366	378	return &htab->buckets[hash & (htab->n_buckets - 1)];
367	379	}
368	380
369		−static inline struct hlist_head select_bucket(struct bpf_htab htab, u32 hash)
	381	+static inline struct hlist_nulls_head select_bucket(struct bpf_htab htab, u32 hash)
370	382	{
371	383	return &__select_bucket(htab, hash)->head;
372	384	}
373	385
374		−static struct htab_elem lookup_elem_raw(struct hlist_head head, u32 hash,
	386	+/* this lookup function can only be called with bucket lock taken */
	387	+static struct htab_elem lookup_elem_raw(struct hlist_nulls_head head, u32 hash,
375	388	void *key, u32 key_size)
376	389	{
	390	+ struct hlist_nulls_node *n;
	391	+ struct htab_elem *l;
	392	+
	393	+ hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
	394	+ if (l->hash == hash && !memcmp(&l->key, key, key_size))
	395	+ return l;
	396	+
	397	+ return NULL;
	398	+}
	399	+
	400	+/* can be called without bucket lock. it will repeat the loop in
	401	+ * the unlikely event when elements moved from one bucket into another
	402	+ * while link list is being walked
	403	+ */
	404	+static struct htab_elem lookup_nulls_elem_raw(struct hlist_nulls_head head,
	405	+ u32 hash, void *key,
	406	+ u32 key_size, u32 n_buckets)
	407	+{
	408	+ struct hlist_nulls_node *n;
377	409	struct htab_elem *l;
378	410
379		− hlist_for_each_entry_rcu(l, head, hash_node)
	411	+again:
	412	+ hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
380	413	if (l->hash == hash && !memcmp(&l->key, key, key_size))
381	414	return l;
382	415
	416	+ if (unlikely(get_nulls_value(n) != (hash & (n_buckets - 1))))
	417	+ goto again;
	418	+
383	419	return NULL;
384	420	}
385	421
@@ -387,7 +423,7 @@ static struct htab_elem lookup_elem_raw(struct hlist_head head, u32 hash, · 1 unmodified line
387	423	static void __htab_map_lookup_elem(struct bpf_map map, void *key)
388	424	{
389	425	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
390		− struct hlist_head *head;
	426	+ struct hlist_nulls_head *head;
391	427	struct htab_elem *l;
392	428	u32 hash, key_size;
393	429
@@ -400,7 +436,7 @@ static void __htab_map_lookup_elem(struct bpf_map map, void *key) · 6 unmodified lines
400	436
401	437	head = select_bucket(htab, hash);
402	438
403		− l = lookup_elem_raw(head, hash, key, key_size);
	439	+ l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
404	440
405	441	return l;
406	442	}
@@ -433,8 +469,9 @@ static void htab_lru_map_lookup_elem(struct bpf_map map, void *key) · 26 unmodified lines
433	469	static bool htab_lru_map_delete_node(void arg, struct bpf_lru_node node)
434	470	{
435	471	struct bpf_htab htab = (struct bpf_htab )arg;
436		− struct htab_elem l, tgt_l;
437		− struct hlist_head *head;
	472	+ struct htab_elem l = NULL, tgt_l;
	473	+ struct hlist_nulls_head *head;
	474	+ struct hlist_nulls_node *n;
438	475	unsigned long flags;
439	476	struct bucket *b;
440	477
@@ -444,9 +481,9 @@ static bool htab_lru_map_delete_node(void arg, struct bpf_lru_node node) · 3 unmodified lines
444	481
445	482	raw_spin_lock_irqsave(&b->lock, flags);
446	483
447		− hlist_for_each_entry_rcu(l, head, hash_node)
	484	+ hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
448	485	if (l == tgt_l) {
449		− hlist_del_rcu(&l->hash_node);
	486	+ hlist_nulls_del_rcu(&l->hash_node);
450	487	break;
451	488	}
452	489
@@ -459,7 +496,7 @@ static bool htab_lru_map_delete_node(void arg, struct bpf_lru_node node) · 6 unmodified lines
459	496	static int htab_map_get_next_key(struct bpf_map map, void key, void *next_key)
460	497	{
461	498	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
462		− struct hlist_head *head;
	499	+ struct hlist_nulls_head *head;
463	500	struct htab_elem l, next_l;
464	501	u32 hash, key_size;
465	502	int i;
@@ -473,7 +510,7 @@ static int htab_map_get_next_key(struct bpf_map map, void key, void *next_key) · 7 unmodified lines
473	510	head = select_bucket(htab, hash);
474	511
475	512	/* lookup the key */
476		− l = lookup_elem_raw(head, hash, key, key_size);
	513	+ l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
477	514
478	515	if (!l) {
479	516	i = 0;
@@ -481,7 +518,7 @@ static int htab_map_get_next_key(struct bpf_map map, void key, void *next_key) · 1 unmodified line
481	518	}
482	519
483	520	/* key was found, get next key in the same bucket */
484		− next_l = hlist_entry_safe(rcu_dereference_raw(hlist_next_rcu(&l->hash_node)),
	521	+ next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_next_rcu(&l->hash_node)),
485	522	struct htab_elem, hash_node);
486	523
487	524	if (next_l) {
@@ -500,7 +537,7 @@ find_first_elem: · 12 unmodified lines
500	537	head = select_bucket(htab, i);
501	538
502	539	/* pick first element in the bucket */
503		− next_l = hlist_entry_safe(rcu_dereference_raw(hlist_first_rcu(head)),
	540	+ next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_first_rcu(head)),
504	541	struct htab_elem, hash_node);
505	542	if (next_l) {
506	543	/* if it's not empty, just return it */
@@ -582,9 +619,13 @@ static struct htab_elem alloc_htab_elem(struct bpf_htab htab, void *key, · 75 unmodified lines
582	619	int err = 0;
583	620
584	621	if (prealloc) {
585		− l_new = (struct htab_elem *)pcpu_freelist_pop(&htab->freelist);
586		− if (!l_new)
	622	+ struct pcpu_freelist_node *l;
	623	+
	624	+ l = pcpu_freelist_pop(&htab->freelist);
	625	+ if (!l)
587	626	err = -E2BIG;
	627	+ else
	628	+ l_new = container_of(l, struct htab_elem, fnode);
588	629	} else {
589	630	if (atomic_inc_return(&htab->count) > htab->map.max_entries) {
590	631	atomic_dec(&htab->count);
@@ -661,7 +702,7 @@ static int htab_map_update_elem(struct bpf_map map, void key, void *value, · 70 unmodified lines
661	702	{
662	703	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
663	704	struct htab_elem l_new = NULL, l_old;
664		− struct hlist_head *head;
	705	+ struct hlist_nulls_head *head;
665	706	unsigned long flags;
666	707	struct bucket *b;
667	708	u32 key_size, hash;
@@ -700,9 +741,9 @@ static int htab_map_update_elem(struct bpf_map map, void key, void *value, · 32 unmodified lines
700	741	/* add new element to the head of the list, so that
701	742	* concurrent search will find it before old elem
702	743	*/
703		− hlist_add_head_rcu(&l_new->hash_node, head);
	744	+ hlist_nulls_add_head_rcu(&l_new->hash_node, head);
704	745	if (l_old) {
705		− hlist_del_rcu(&l_old->hash_node);
	746	+ hlist_nulls_del_rcu(&l_old->hash_node);
706	747	free_htab_elem(htab, l_old);
707	748	}
708	749	ret = 0;
@@ -716,7 +757,7 @@ static int htab_lru_map_update_elem(struct bpf_map map, void key, void *value, · 7 unmodified lines
716	757	{
717	758	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
718	759	struct htab_elem l_new, l_old = NULL;
719		− struct hlist_head *head;
	760	+ struct hlist_nulls_head *head;
720	761	unsigned long flags;
721	762	struct bucket *b;
722	763	u32 key_size, hash;
@@ -757,10 +798,10 @@ static int htab_lru_map_update_elem(struct bpf_map map, void key, void *value, · 34 unmodified lines
757	798	/* add new element to the head of the list, so that
758	799	* concurrent search will find it before old elem
759	800	*/
760		− hlist_add_head_rcu(&l_new->hash_node, head);
	801	+ hlist_nulls_add_head_rcu(&l_new->hash_node, head);
761	802	if (l_old) {
762	803	bpf_lru_node_set_ref(&l_new->lru_node);
763		− hlist_del_rcu(&l_old->hash_node);
	804	+ hlist_nulls_del_rcu(&l_old->hash_node);
764	805	}
765	806	ret = 0;
766	807
@@ -781,7 +822,7 @@ static int __htab_percpu_map_update_elem(struct bpf_map map, void key, · 14 unmodified lines
781	822	{
782	823	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
783	824	struct htab_elem l_new = NULL, l_old;
784		− struct hlist_head *head;
	825	+ struct hlist_nulls_head *head;
785	826	unsigned long flags;
786	827	struct bucket *b;
787	828	u32 key_size, hash;
@@ -820,7 +861,7 @@ static int __htab_percpu_map_update_elem(struct bpf_map map, void key, · 32 unmodified lines
820	861	ret = PTR_ERR(l_new);
821	862	goto err;
822	863	}
823		− hlist_add_head_rcu(&l_new->hash_node, head);
	864	+ hlist_nulls_add_head_rcu(&l_new->hash_node, head);
824	865	}
825	866	ret = 0;
826	867	err:
@@ -834,7 +875,7 @@ static int __htab_lru_percpu_map_update_elem(struct bpf_map map, void key, · 7 unmodified lines
834	875	{
835	876	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
836	877	struct htab_elem l_new = NULL, l_old;
837		− struct hlist_head *head;
	878	+ struct hlist_nulls_head *head;
838	879	unsigned long flags;
839	880	struct bucket *b;
840	881	u32 key_size, hash;
@@ -882,7 +923,7 @@ static int __htab_lru_percpu_map_update_elem(struct bpf_map map, void key, · 41 unmodified lines
882	923	} else {
883	924	pcpu_copy_value(htab, htab_elem_get_ptr(l_new, key_size),
884	925	value, onallcpus);
885		− hlist_add_head_rcu(&l_new->hash_node, head);
	926	+ hlist_nulls_add_head_rcu(&l_new->hash_node, head);
886	927	l_new = NULL;
887	928	}
888	929	ret = 0;
@@ -910,7 +951,7 @@ static int htab_lru_percpu_map_update_elem(struct bpf_map map, void key, · 21 unmodified lines
910	951	static int htab_map_delete_elem(struct bpf_map map, void key)
911	952	{
912	953	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
913		− struct hlist_head *head;
	954	+ struct hlist_nulls_head *head;
914	955	struct bucket *b;
915	956	struct htab_elem *l;
916	957	unsigned long flags;
@@ -930,7 +971,7 @@ static int htab_map_delete_elem(struct bpf_map map, void key) · 13 unmodified lines
930	971	l = lookup_elem_raw(head, hash, key, key_size);
931	972
932	973	if (l) {
933		− hlist_del_rcu(&l->hash_node);
	974	+ hlist_nulls_del_rcu(&l->hash_node);
934	975	free_htab_elem(htab, l);
935	976	ret = 0;
936	977	}
@@ -942,7 +983,7 @@ static int htab_map_delete_elem(struct bpf_map map, void key) · 5 unmodified lines
942	983	static int htab_lru_map_delete_elem(struct bpf_map map, void key)
943	984	{
944	985	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
945		− struct hlist_head *head;
	986	+ struct hlist_nulls_head *head;
946	987	struct bucket *b;
947	988	struct htab_elem *l;
948	989	unsigned long flags;
@@ -962,7 +1003,7 @@ static int htab_lru_map_delete_elem(struct bpf_map map, void key) · 13 unmodified lines
962	1003	l = lookup_elem_raw(head, hash, key, key_size);
963	1004
964	1005	if (l) {
965		− hlist_del_rcu(&l->hash_node);
	1006	+ hlist_nulls_del_rcu(&l->hash_node);
966	1007	ret = 0;
967	1008	}
968	1009
@@ -977,12 +1018,12 @@ static void delete_all_elements(struct bpf_htab *htab) · 8 unmodified lines
977	1018	int i;
978	1019
979	1020	for (i = 0; i < htab->n_buckets; i++) {
980		− struct hlist_head *head = select_bucket(htab, i);
981		− struct hlist_node *n;
	1021	+ struct hlist_nulls_head *head = select_bucket(htab, i);
	1022	+ struct hlist_nulls_node *n;
982	1023	struct htab_elem *l;
983	1024
984		− hlist_for_each_entry_safe(l, n, head, hash_node) {
985		− hlist_del_rcu(&l->hash_node);
	1025	+ hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
	1026	+ hlist_nulls_del_rcu(&l->hash_node);
986	1027	if (l->state != HTAB_EXTRA_ELEM_USED)
987	1028	htab_elem_free(htab, l);
988	1029	}

kernel/bpf/lpm_trie.c+6−0

@@ -500,9 +500,15 @@ unlock: · 499 unmodified lines
500	500	raw_spin_unlock(&trie->lock);
501	501	}
502	502
	503	+static int trie_get_next_key(struct bpf_map map, void key, void *next_key)
	504	+{
	505	+ return -ENOTSUPP;
	506	+}
	507	+
503	508	static const struct bpf_map_ops trie_ops = {
504	509	.map_alloc = trie_alloc,
505	510	.map_free = trie_free,
	511	+ .map_get_next_key = trie_get_next_key,
506	512	.map_lookup_elem = trie_lookup_elem,
507	513	.map_update_elem = trie_update_elem,
508	514	.map_delete_elem = trie_delete_elem,

kernel/cgroup/cgroup-v1.c+1−1

@@ -1329,7 +1329,7 @@ static int cgroup_css_links_read(struct seq_file seq, void v) · 1328 unmodified lines
1329	1329	struct task_struct *task;
1330	1330	int count = 0;
1331	1331
1332		− seq_printf(seq, "css_set %p\n", cset);
	1332	+ seq_printf(seq, "css_set %pK\n", cset);
1333	1333
1334	1334	list_for_each_entry(task, &cset->tasks, cg_list) {
1335	1335	if (count++ > MAX_TASKS_SHOWN_PER_CSS)

kernel/cgroup/cgroup.c+1−1

@@ -2669,7 +2669,7 @@ static bool css_visible(struct cgroup_subsys_state *css) · 2668 unmodified lines
2669	2669	*
2670	2670	* Returns 0 on success, -errno on failure. On failure, csses which have
2671	2671	* been processed already aren't cleaned up. The caller is responsible for
2672		− * cleaning up with cgroup_apply_control_disble().
	2672	+ * cleaning up with cgroup_apply_control_disable().
2673	2673	*/
2674	2674	static int cgroup_apply_control_enable(struct cgroup *cgrp)
2675	2675	{

kernel/cgroup/pids.c+1−1

@@ -229,7 +229,7 @@ static int pids_can_fork(struct task_struct *task) · 228 unmodified lines
229	229	/* Only log the first time events_limit is incremented. */
230	230	if (atomic64_inc_return(&pids->events_limit) == 1) {
231	231	pr_info("cgroup: fork rejected by pids controller in ");
232		− pr_cont_cgroup_path(task_cgroup(current, pids_cgrp_id));
	232	+ pr_cont_cgroup_path(css->cgroup);
233	233	pr_cont("\n");
234	234	}
235	235	cgroup_file_notify(&pids->events_file);

kernel/events/core.c+1−1

@@ -1001,7 +1001,7 @@ list_update_cgroup_event(struct perf_event *event, · 1000 unmodified lines
1001	1001	*/
1002	1002	#define PERF_CPU_HRTIMER (1000 / HZ)
1003	1003	/*
1004		− * function must be called with interrupts disbled
	1004	+ * function must be called with interrupts disabled
1005	1005	*/
1006	1006	static enum hrtimer_restart perf_mux_hrtimer_handler(struct hrtimer *hr)
1007	1007	{

kernel/exit.c+0−1

@@ -554,7 +554,6 @@ static void exit_mm(void) · 553 unmodified lines
554	554	enter_lazy_tlb(mm, current);
555	555	task_unlock(current);
556	556	mm_update_next_owner(mm);
557		− userfaultfd_exit(mm);
558	557	mmput(mm);
559	558	if (test_thread_flag(TIF_MEMDIE))
560	559	exit_oom_victim();

kernel/kexec_file.c+4−4

@@ -614,13 +614,13 @@ static int kexec_calculate_store_digests(struct kimage *image) · 613 unmodified lines
614	614	ret = crypto_shash_final(desc, digest);
615	615	if (ret)
616	616	goto out_free_digest;
617		− ret = kexec_purgatory_get_set_symbol(image, "sha_regions",
618		− sha_regions, sha_region_sz, 0);
	617	+ ret = kexec_purgatory_get_set_symbol(image, "purgatory_sha_regions",
	618	+ sha_regions, sha_region_sz, 0);
619	619	if (ret)
620	620	goto out_free_digest;
621	621
622		− ret = kexec_purgatory_get_set_symbol(image, "sha256_digest",
623		− digest, SHA256_DIGEST_SIZE, 0);
	622	+ ret = kexec_purgatory_get_set_symbol(image, "purgatory_sha256_digest",
	623	+ digest, SHA256_DIGEST_SIZE, 0);
624	624	if (ret)
625	625	goto out_free_digest;
626	626	}

kernel/kexec_internal.h+1−5

@@ -15,11 +15,7 @@ int kimage_is_destination_range(struct kimage *image, · 14 unmodified lines
15	15	extern struct mutex kexec_mutex;
16	16
17	17	#ifdef CONFIG_KEXEC_FILE
18		−struct kexec_sha_region {
19		− unsigned long start;
20		− unsigned long len;
21		−};
22		−
	18	+#include <linux/purgatory.h>
23	19	void kimage_file_post_load_cleanup(struct kimage *image);
24	20	#else /* CONFIG_KEXEC_FILE */
25	21	static inline void kimage_file_post_load_cleanup(struct kimage *image) { }

kernel/locking/lockdep.c+9−2

@@ -3262,10 +3262,17 @@ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, · 3261 unmodified lines
3262	3262	if (depth) {
3263	3263	hlock = curr->held_locks + depth - 1;
3264	3264	if (hlock->class_idx == class_idx && nest_lock) {
3265		− if (hlock->references)
	3265	+ if (hlock->references) {
	3266	+ /*
	3267	+ * Check: unsigned int references:12, overflow.
	3268	+ */
	3269	+ if (DEBUG_LOCKS_WARN_ON(hlock->references == (1 << 12)-1))
	3270	+ return 0;
	3271	+
3266	3272	hlock->references++;
3267		− else
	3273	+ } else {
3268	3274	hlock->references = 2;
	3275	+ }
3269	3276
3270	3277	return 1;
3271	3278	}

kernel/locking/test-ww_mutex.c+3−3

@@ -50,7 +50,7 @@ static void test_mutex_work(struct work_struct *work) · 49 unmodified lines
50	50
51	51	if (mtx->flags & TEST_MTX_TRY) {
52	52	while (!ww_mutex_trylock(&mtx->mutex))
53		− cpu_relax();
	53	+ cond_resched();
54	54	} else {
55	55	ww_mutex_lock(&mtx->mutex, NULL);
56	56	}
@@ -88,7 +88,7 @@ static int __test_mutex(unsigned int flags) · 31 unmodified lines
88	88	ret = -EINVAL;
89	89	break;
90	90	}
91		− cpu_relax();
	91	+ cond_resched();
92	92	} while (time_before(jiffies, timeout));
93	93	} else {
94	94	ret = wait_for_completion_timeout(&mtx.done, TIMEOUT);
@@ -627,7 +627,7 @@ static int __init test_ww_mutex_init(void) · 532 unmodified lines
627	627	if (ret)
628	628	return ret;
629	629
630		− ret = stress(4096, hweight32(STRESS_ALL)*ncpus, 1<<12, STRESS_ALL);
	630	+ ret = stress(4095, hweight32(STRESS_ALL)*ncpus, 1<<12, STRESS_ALL);
631	631	if (ret)
632	632	return ret;
633	633

kernel/sched/core.c+8−3

@@ -3287,10 +3287,15 @@ pick_next_task(struct rq rq, struct task_struct prev, struct rq_flags *rf) · 3286 unmodified lines
3287	3287	struct task_struct *p;
3288	3288
3289	3289	/*
3290		− * Optimization: we know that if all tasks are in
3291		− * the fair class we can call that function directly:
	3290	+ * Optimization: we know that if all tasks are in the fair class we can
	3291	+ * call that function directly, but only if the @prev task wasn't of a
	3292	+ * higher scheduling class, because otherwise those loose the
	3293	+ * opportunity to pull in more work from other CPUs.
3292	3294	*/
3293		− if (likely(rq->nr_running == rq->cfs.h_nr_running)) {
	3295	+ if (likely((prev->sched_class == &idle_sched_class \|\|
	3296	+ prev->sched_class == &fair_sched_class) &&
	3297	+ rq->nr_running == rq->cfs.h_nr_running)) {
	3298	+
3294	3299	p = fair_sched_class.pick_next_task(rq, prev, rf);
3295	3300	if (unlikely(p == RETRY_TASK))
3296	3301	goto again;

kernel/sched/cpufreq_schedutil.c+9−10

@@ -36,6 +36,7 @@ struct sugov_policy { · 35 unmodified lines
36	36	u64 last_freq_update_time;
37	37	s64 freq_update_delay_ns;
38	38	unsigned int next_freq;
	39	+ unsigned int cached_raw_freq;
39	40
40	41	/* The next fields are only needed if fast switch cannot be used. */
41	42	struct irq_work irq_work;
@@ -52,7 +53,6 @@ struct sugov_cpu { · 10 unmodified lines
52	53	struct update_util_data update_util;
53	54	struct sugov_policy *sg_policy;
54	55
55		− unsigned int cached_raw_freq;
56	56	unsigned long iowait_boost;
57	57	unsigned long iowait_boost_max;
58	58	u64 last_update;
@@ -116,7 +116,7 @@ static void sugov_update_commit(struct sugov_policy *sg_policy, u64 time, · 57 unmodified lines
116	116
117	117	/**
118	118	* get_next_freq - Compute a new frequency for a given cpufreq policy.
119		− * @sg_cpu: schedutil cpu object to compute the new frequency for.
	119	+ * @sg_policy: schedutil policy object to compute the new frequency for.
120	120	* @util: Current CPU utilization.
121	121	* @max: CPU capacity.
122	122	*
@@ -136,19 +136,18 @@ static void sugov_update_commit(struct sugov_policy *sg_policy, u64 time, · 13 unmodified lines
136	136	* next_freq (as calculated above) is returned, subject to policy min/max and
137	137	* cpufreq driver limitations.
138	138	*/
139		−static unsigned int get_next_freq(struct sugov_cpu *sg_cpu, unsigned long util,
140		− unsigned long max)
	139	+static unsigned int get_next_freq(struct sugov_policy *sg_policy,
	140	+ unsigned long util, unsigned long max)
141	141	{
142		− struct sugov_policy *sg_policy = sg_cpu->sg_policy;
143	142	struct cpufreq_policy *policy = sg_policy->policy;
144	143	unsigned int freq = arch_scale_freq_invariant() ?
145	144	policy->cpuinfo.max_freq : policy->cur;
146	145
147	146	freq = (freq + (freq >> 2)) * util / max;
148	147
149		− if (freq == sg_cpu->cached_raw_freq && sg_policy->next_freq != UINT_MAX)
	148	+ if (freq == sg_policy->cached_raw_freq && sg_policy->next_freq != UINT_MAX)
150	149	return sg_policy->next_freq;
151		− sg_cpu->cached_raw_freq = freq;
	150	+ sg_policy->cached_raw_freq = freq;
152	151	return cpufreq_driver_resolve_freq(policy, freq);
153	152	}
154	153
@@ -213,7 +212,7 @@ static void sugov_update_single(struct update_util_data *hook, u64 time, · 58 unmodified lines
213	212	} else {
214	213	sugov_get_util(&util, &max);
215	214	sugov_iowait_boost(sg_cpu, &util, &max);
216		− next_f = get_next_freq(sg_cpu, util, max);
	215	+ next_f = get_next_freq(sg_policy, util, max);
217	216	}
218	217	sugov_update_commit(sg_policy, time, next_f);
219	218	}
@@ -267,7 +266,7 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, · 47 unmodified lines
267	266	sugov_iowait_boost(j_sg_cpu, &util, &max);
268	267	}
269	268
270		− return get_next_freq(sg_cpu, util, max);
	269	+ return get_next_freq(sg_policy, util, max);
271	270	}
272	271
273	272	static void sugov_update_shared(struct update_util_data *hook, u64 time,
@@ -580,6 +579,7 @@ static int sugov_start(struct cpufreq_policy *policy) · 306 unmodified lines
580	579	sg_policy->next_freq = UINT_MAX;
581	580	sg_policy->work_in_progress = false;
582	581	sg_policy->need_freq_update = false;
	582	+ sg_policy->cached_raw_freq = 0;
583	583
584	584	for_each_cpu(cpu, policy->cpus) {
585	585	struct sugov_cpu *sg_cpu = &per_cpu(sugov_cpu, cpu);
@@ -590,7 +590,6 @@ static int sugov_start(struct cpufreq_policy *policy) · 4 unmodified lines
590	590	sg_cpu->max = 0;
591	591	sg_cpu->flags = SCHED_CPUFREQ_RT;
592	592	sg_cpu->last_update = 0;
593		− sg_cpu->cached_raw_freq = 0;
594	593	sg_cpu->iowait_boost = 0;
595	594	sg_cpu->iowait_boost_max = policy->cpuinfo.max_freq;
596	595	cpufreq_add_update_util_hook(cpu, &sg_cpu->update_util,

kernel/sched/fair.c+1−1

@@ -5799,7 +5799,7 @@ static int select_idle_cpu(struct task_struct p, struct sched_domain sd, int t · 5798 unmodified lines
5799	5799	* Due to large variance we need a large fuzz factor; hackbench in
5800	5800	* particularly is sensitive here.
5801	5801	*/
5802		− if ((avg_idle / 512) < avg_cost)
	5802	+ if (sched_feat(SIS_AVG_CPU) && (avg_idle / 512) < avg_cost)
5803	5803	return -1;
5804	5804
5805	5805	time = local_clock();

kernel/sched/features.h+5−0

@@ -51,6 +51,11 @@ SCHED_FEAT(NONTASK_CAPACITY, true) · 50 unmodified lines
51	51	*/
52	52	SCHED_FEAT(TTWU_QUEUE, true)
53	53
	54	+/*
	55	+ * When doing wakeups, attempt to limit superfluous scans of the LLC domain.
	56	+ */
	57	+SCHED_FEAT(SIS_AVG_CPU, false)
	58	+
54	59	#ifdef HAVE_RT_PUSH_IPI
55	60	/*
56	61	* In order to avoid a thundering herd attack of CPUs that are

kernel/sched/wait.c+39−0

@@ -242,6 +242,45 @@ long prepare_to_wait_event(wait_queue_head_t q, wait_queue_t wait, int state) · 241 unmodified lines
242	242	}
243	243	EXPORT_SYMBOL(prepare_to_wait_event);
244	244
	245	+/*
	246	+ * Note! These two wait functions are entered with the
	247	+ * wait-queue lock held (and interrupts off in the _irq
	248	+ * case), so there is no race with testing the wakeup
	249	+ * condition in the caller before they add the wait
	250	+ * entry to the wake queue.
	251	+ */
	252	+int do_wait_intr(wait_queue_head_t wq, wait_queue_t wait)
	253	+{
	254	+ if (likely(list_empty(&wait->task_list)))
	255	+ __add_wait_queue_tail(wq, wait);
	256	+
	257	+ set_current_state(TASK_INTERRUPTIBLE);
	258	+ if (signal_pending(current))
	259	+ return -ERESTARTSYS;
	260	+
	261	+ spin_unlock(&wq->lock);
	262	+ schedule();
	263	+ spin_lock(&wq->lock);
	264	+ return 0;
	265	+}
	266	+EXPORT_SYMBOL(do_wait_intr);
	267	+
	268	+int do_wait_intr_irq(wait_queue_head_t wq, wait_queue_t wait)
	269	+{
	270	+ if (likely(list_empty(&wait->task_list)))
	271	+ __add_wait_queue_tail(wq, wait);
	272	+
	273	+ set_current_state(TASK_INTERRUPTIBLE);
	274	+ if (signal_pending(current))
	275	+ return -ERESTARTSYS;
	276	+
	277	+ spin_unlock_irq(&wq->lock);
	278	+ schedule();
	279	+ spin_lock_irq(&wq->lock);
	280	+ return 0;
	281	+}
	282	+EXPORT_SYMBOL(do_wait_intr_irq);
	283	+
245	284	/**
246	285	* finish_wait - clean up after waiting in a queue
247	286	* @q: waitqueue waited on

kernel/time/jiffies.c+1−1

@@ -125,7 +125,7 @@ int register_refined_jiffies(long cycles_per_second) · 124 unmodified lines
125	125	shift_hz += cycles_per_tick/2;
126	126	do_div(shift_hz, cycles_per_tick);
127	127	/* Calculate nsec_per_tick using shift_hz */
128		− nsec_per_tick = (u64)TICK_NSEC << 8;
	128	+ nsec_per_tick = (u64)NSEC_PER_SEC << 8;
129	129	nsec_per_tick += (u32)shift_hz/2;
130	130	do_div(nsec_per_tick, (u32)shift_hz);
131	131

kernel/trace/ftrace.c+18−5

@@ -4416,16 +4416,24 @@ static int __init set_graph_notrace_function(char *str) · 4415 unmodified lines
4416	4416	}
4417	4417	__setup("ftrace_graph_notrace=", set_graph_notrace_function);
4418	4418
	4419	+static int __init set_graph_max_depth_function(char *str)
	4420	+{
	4421	+ if (!str)
	4422	+ return 0;
	4423	+ fgraph_max_depth = simple_strtoul(str, NULL, 0);
	4424	+ return 1;
	4425	+}
	4426	+__setup("ftrace_graph_max_depth=", set_graph_max_depth_function);
	4427	+
4419	4428	static void __init set_ftrace_early_graph(char *buf, int enable)
4420	4429	{
4421	4430	int ret;
4422	4431	char *func;
4423	4432	struct ftrace_hash *hash;
4424	4433
4425		− if (enable)
4426		− hash = ftrace_graph_hash;
4427		− else
4428		− hash = ftrace_graph_notrace_hash;
	4434	+ hash = alloc_ftrace_hash(FTRACE_HASH_DEFAULT_BITS);
	4435	+ if (WARN_ON(!hash))
	4436	+ return;
4429	4437
4430	4438	while (buf) {
4431	4439	func = strsep(&buf, ",");
@@ -4435,6 +4443,11 @@ static void __init set_ftrace_early_graph(char *buf, int enable) · 3 unmodified lines
4435	4443	printk(KERN_DEBUG "ftrace: function %s not "
4436	4444	"traceable\n", func);
4437	4445	}
	4446	+
	4447	+ if (enable)
	4448	+ ftrace_graph_hash = hash;
	4449	+ else
	4450	+ ftrace_graph_notrace_hash = hash;
4438	4451	}
4439	4452	#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
4440	4453
@@ -5488,7 +5501,7 @@ static void ftrace_ops_assist_func(unsigned long ip, unsigned long parent_ip, · 1047 unmodified lines
5488	5501	* Normally the mcount trampoline will call the ops->func, but there
5489	5502	* are times that it should not. For example, if the ops does not
5490	5503	* have its own recursion protection, then it should call the
5491		− * ftrace_ops_recurs_func() instead.
	5504	+ * ftrace_ops_assist_func() instead.
5492	5505	*
5493	5506	* Returns the function that the trampoline should call for @ops.
5494	5507	*/

kernel/trace/trace_stack.c+1−1

@@ -65,7 +65,7 @@ void stack_trace_print(void) · 64 unmodified lines
65	65	}
66	66
67	67	/*
68		− * When arch-specific code overides this function, the following
	68	+ * When arch-specific code overrides this function, the following
69	69	* data should be filled up, assuming stack_trace_max_lock is held to
70	70	* prevent concurrent updates.
71	71	* stack_trace_index[]

kernel/ucount.c+11−7

@@ -144,7 +144,7 @@ static struct ucounts get_ucounts(struct user_namespace ns, kuid_t uid) · 143 unmodified lines
144	144
145	145	new->ns = ns;
146	146	new->uid = uid;
147		− atomic_set(&new->count, 0);
	147	+ new->count = 0;
148	148
149	149	spin_lock_irq(&ucounts_lock);
150	150	ucounts = find_ucounts(ns, uid, hashent);
@@ -155,8 +155,10 @@ static struct ucounts get_ucounts(struct user_namespace ns, kuid_t uid) · 4 unmodified lines
155	155	ucounts = new;
156	156	}
157	157	}
158		− if (!atomic_add_unless(&ucounts->count, 1, INT_MAX))
	158	+ if (ucounts->count == INT_MAX)
159	159	ucounts = NULL;
	160	+ else
	161	+ ucounts->count += 1;
160	162	spin_unlock_irq(&ucounts_lock);
161	163	return ucounts;
162	164	}
@@ -165,13 +167,15 @@ static void put_ucounts(struct ucounts *ucounts) · 2 unmodified lines
165	167	{
166	168	unsigned long flags;
167	169
168		− if (atomic_dec_and_test(&ucounts->count)) {
169		− spin_lock_irqsave(&ucounts_lock, flags);
	170	+ spin_lock_irqsave(&ucounts_lock, flags);
	171	+ ucounts->count -= 1;
	172	+ if (!ucounts->count)
170	173	hlist_del_init(&ucounts->node);
171		− spin_unlock_irqrestore(&ucounts_lock, flags);
	174	+ else
	175	+ ucounts = NULL;
	176	+ spin_unlock_irqrestore(&ucounts_lock, flags);
172	177
173		− kfree(ucounts);
174		− }
	178	+ kfree(ucounts);
175	179	}
176	180
177	181	static inline bool atomic_inc_below(atomic_t *v, int u)

kernel/workqueue.c+1−0

@@ -1507,6 +1507,7 @@ static void __queue_delayed_work(int cpu, struct workqueue_struct *wq, · 1506 unmodified lines
1507	1507	struct timer_list *timer = &dwork->timer;
1508	1508	struct work_struct *work = &dwork->work;
1509	1509
	1510	+ WARN_ON_ONCE(!wq);
1510	1511	WARN_ON_ONCE(timer->function != delayed_work_timer_fn \|\|
1511	1512	timer->data != (unsigned long)dwork);
1512	1513	WARN_ON_ONCE(timer_pending(timer));

lib/ioremap.c+36−3

@@ -14,6 +14,7 @@ · 13 unmodified lines
14	14	#include <asm/pgtable.h>
15	15
16	16	#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
	17	+static int __read_mostly ioremap_p4d_capable;
17	18	static int __read_mostly ioremap_pud_capable;
18	19	static int __read_mostly ioremap_pmd_capable;
19	20	static int __read_mostly ioremap_huge_disabled;
@@ -35,6 +36,11 @@ void __init ioremap_huge_init(void) · 15 unmodified lines
35	36	}
36	37	}
37	38
	39	+static inline int ioremap_p4d_enabled(void)
	40	+{
	41	+ return ioremap_p4d_capable;
	42	+}
	43	+
38	44	static inline int ioremap_pud_enabled(void)
39	45	{
40	46	return ioremap_pud_capable;
@@ -46,6 +52,7 @@ static inline int ioremap_pmd_enabled(void) · 5 unmodified lines
46	52	}
47	53
48	54	#else /* !CONFIG_HAVE_ARCH_HUGE_VMAP */
	55	+static inline int ioremap_p4d_enabled(void) { return 0; }
49	56	static inline int ioremap_pud_enabled(void) { return 0; }
50	57	static inline int ioremap_pmd_enabled(void) { return 0; }
51	58	#endif /* CONFIG_HAVE_ARCH_HUGE_VMAP */
@@ -94,14 +101,14 @@ static inline int ioremap_pmd_range(pud_t *pud, unsigned long addr, · 42 unmodified lines
94	101	return 0;
95	102	}
96	103
97		−static inline int ioremap_pud_range(pgd_t *pgd, unsigned long addr,
	104	+static inline int ioremap_pud_range(p4d_t *p4d, unsigned long addr,
98	105	unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
99	106	{
100	107	pud_t *pud;
101	108	unsigned long next;
102	109
103	110	phys_addr -= addr;
104		− pud = pud_alloc(&init_mm, pgd, addr);
	111	+ pud = pud_alloc(&init_mm, p4d, addr);
105	112	if (!pud)
106	113	return -ENOMEM;
107	114	do {
@@ -120,6 +127,32 @@ static inline int ioremap_pud_range(pgd_t *pgd, unsigned long addr, · 12 unmodified lines
120	127	return 0;
121	128	}
122	129
	130	+static inline int ioremap_p4d_range(pgd_t *pgd, unsigned long addr,
	131	+ unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
	132	+{
	133	+ p4d_t *p4d;
	134	+ unsigned long next;
	135	+
	136	+ phys_addr -= addr;
	137	+ p4d = p4d_alloc(&init_mm, pgd, addr);
	138	+ if (!p4d)
	139	+ return -ENOMEM;
	140	+ do {
	141	+ next = p4d_addr_end(addr, end);
	142	+
	143	+ if (ioremap_p4d_enabled() &&
	144	+ ((next - addr) == P4D_SIZE) &&
	145	+ IS_ALIGNED(phys_addr + addr, P4D_SIZE)) {
	146	+ if (p4d_set_huge(p4d, phys_addr + addr, prot))
	147	+ continue;
	148	+ }
	149	+
	150	+ if (ioremap_pud_range(p4d, addr, next, phys_addr + addr, prot))
	151	+ return -ENOMEM;
	152	+ } while (p4d++, addr = next, addr != end);
	153	+ return 0;
	154	+}
	155	+
123	156	int ioremap_page_range(unsigned long addr,
124	157	unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
125	158	{
@@ -135,7 +168,7 @@ int ioremap_page_range(unsigned long addr, · 9 unmodified lines
135	168	pgd = pgd_offset_k(addr);
136	169	do {
137	170	next = pgd_addr_end(addr, end);
138		− err = ioremap_pud_range(pgd, addr, next, phys_addr+addr, prot);
	171	+ err = ioremap_p4d_range(pgd, addr, next, phys_addr+addr, prot);
139	172	if (err)
140	173	break;
141	174	} while (pgd++, addr = next, addr != end);

lib/radix-tree.c+2−2

@@ -2129,8 +2129,8 @@ int ida_pre_get(struct ida *ida, gfp_t gfp) · 2128 unmodified lines
2129	2129	struct ida_bitmap bitmap = kmalloc(sizeof(bitmap), gfp);
2130	2130	if (!bitmap)
2131	2131	return 0;
2132		− bitmap = this_cpu_cmpxchg(ida_bitmap, NULL, bitmap);
2133		− kfree(bitmap);
	2132	+ if (this_cpu_cmpxchg(ida_bitmap, NULL, bitmap))
	2133	+ kfree(bitmap);
2134	2134	}
2135	2135
2136	2136	return 1;

lib/refcount.c+7−7

@@ -58,7 +58,7 @@ bool refcount_add_not_zero(unsigned int i, refcount_t *r) · 57 unmodified lines
58	58	val = old;
59	59	}
60	60
61		− WARN(new == UINT_MAX, "refcount_t: saturated; leaking memory.\n");
	61	+ WARN_ONCE(new == UINT_MAX, "refcount_t: saturated; leaking memory.\n");
62	62
63	63	return true;
64	64	}
@@ -66,7 +66,7 @@ EXPORT_SYMBOL_GPL(refcount_add_not_zero); · 1 unmodified line
66	66
67	67	void refcount_add(unsigned int i, refcount_t *r)
68	68	{
69		− WARN(!refcount_add_not_zero(i, r), "refcount_t: addition on 0; use-after-free.\n");
	69	+ WARN_ONCE(!refcount_add_not_zero(i, r), "refcount_t: addition on 0; use-after-free.\n");
70	70	}
71	71	EXPORT_SYMBOL_GPL(refcount_add);
72	72
@@ -97,7 +97,7 @@ bool refcount_inc_not_zero(refcount_t *r) · 24 unmodified lines
97	97	val = old;
98	98	}
99	99
100		− WARN(new == UINT_MAX, "refcount_t: saturated; leaking memory.\n");
	100	+ WARN_ONCE(new == UINT_MAX, "refcount_t: saturated; leaking memory.\n");
101	101
102	102	return true;
103	103	}
@@ -111,7 +111,7 @@ EXPORT_SYMBOL_GPL(refcount_inc_not_zero); · 7 unmodified lines
111	111	*/
112	112	void refcount_inc(refcount_t *r)
113	113	{
114		− WARN(!refcount_inc_not_zero(r), "refcount_t: increment on 0; use-after-free.\n");
	114	+ WARN_ONCE(!refcount_inc_not_zero(r), "refcount_t: increment on 0; use-after-free.\n");
115	115	}
116	116	EXPORT_SYMBOL_GPL(refcount_inc);
117	117
@@ -125,7 +125,7 @@ bool refcount_sub_and_test(unsigned int i, refcount_t *r) · 7 unmodified lines
125	125
126	126	new = val - i;
127	127	if (new > val) {
128		− WARN(new > val, "refcount_t: underflow; use-after-free.\n");
	128	+ WARN_ONCE(new > val, "refcount_t: underflow; use-after-free.\n");
129	129	return false;
130	130	}
131	131
@@ -164,7 +164,7 @@ EXPORT_SYMBOL_GPL(refcount_dec_and_test); · 32 unmodified lines
164	164
165	165	void refcount_dec(refcount_t *r)
166	166	{
167		− WARN(refcount_dec_and_test(r), "refcount_t: decrement hit 0; leaking memory.\n");
	167	+ WARN_ONCE(refcount_dec_and_test(r), "refcount_t: decrement hit 0; leaking memory.\n");
168	168	}
169	169	EXPORT_SYMBOL_GPL(refcount_dec);
170	170
@@ -204,7 +204,7 @@ bool refcount_dec_not_one(refcount_t *r) · 33 unmodified lines
204	204
205	205	new = val - 1;
206	206	if (new > val) {
207		− WARN(new > val, "refcount_t: underflow; use-after-free.\n");
	207	+ WARN_ONCE(new > val, "refcount_t: underflow; use-after-free.\n");
208	208	return true;
209	209	}
210	210

mm/backing-dev.c+28−15

@@ -683,33 +683,26 @@ static int cgwb_bdi_init(struct backing_dev_info *bdi) · 682 unmodified lines
683	683	static void cgwb_bdi_destroy(struct backing_dev_info *bdi)
684	684	{
685	685	struct radix_tree_iter iter;
686		− struct rb_node *rbn;
687	686	void **slot;
688	687
689	688	WARN_ON(test_bit(WB_registered, &bdi->wb.state));
690	689
691	690	spin_lock_irq(&cgwb_lock);
692		−
693	691	radix_tree_for_each_slot(slot, &bdi->cgwb_tree, &iter, 0)
694	692	cgwb_kill(*slot);
695		−
696		− while ((rbn = rb_first(&bdi->cgwb_congested_tree))) {
697		− struct bdi_writeback_congested *congested =
698		− rb_entry(rbn, struct bdi_writeback_congested, rb_node);
699		−
700		− rb_erase(rbn, &bdi->cgwb_congested_tree);
701		− congested->bdi = NULL; /* mark @congested unlinked */
702		− }
703		−
704	693	spin_unlock_irq(&cgwb_lock);
705	694
706	695	/*
707		− * All cgwb's and their congested states must be shutdown and
708		− * released before returning. Drain the usage counter to wait for
709		− * all cgwb's and cgwb_congested's ever created on @bdi.
	696	+ * All cgwb's must be shutdown and released before returning. Drain
	697	+ * the usage counter to wait for all cgwb's ever created on @bdi.
710	698	*/
711	699	atomic_dec(&bdi->usage_cnt);
712	700	wait_event(cgwb_release_wait, !atomic_read(&bdi->usage_cnt));
	701	+ /*
	702	+ * Grab back our reference so that we hold it when @bdi gets
	703	+ * re-registered.
	704	+ */
	705	+ atomic_inc(&bdi->usage_cnt);
713	706	}
714	707
715	708	/**
@@ -749,6 +742,21 @@ void wb_blkcg_offline(struct blkcg *blkcg) · 33 unmodified lines
749	742	spin_unlock_irq(&cgwb_lock);
750	743	}
751	744
	745	+static void cgwb_bdi_exit(struct backing_dev_info *bdi)
	746	+{
	747	+ struct rb_node *rbn;
	748	+
	749	+ spin_lock_irq(&cgwb_lock);
	750	+ while ((rbn = rb_first(&bdi->cgwb_congested_tree))) {
	751	+ struct bdi_writeback_congested *congested =
	752	+ rb_entry(rbn, struct bdi_writeback_congested, rb_node);
	753	+
	754	+ rb_erase(rbn, &bdi->cgwb_congested_tree);
	755	+ congested->bdi = NULL; /* mark @congested unlinked */
	756	+ }
	757	+ spin_unlock_irq(&cgwb_lock);
	758	+}
	759	+
752	760	#else /* CONFIG_CGROUP_WRITEBACK */
753	761
754	762	static int cgwb_bdi_init(struct backing_dev_info *bdi)
@@ -769,7 +777,9 @@ static int cgwb_bdi_init(struct backing_dev_info *bdi) · 14 unmodified lines
769	777	return 0;
770	778	}
771	779
772		−static void cgwb_bdi_destroy(struct backing_dev_info *bdi)
	780	+static void cgwb_bdi_destroy(struct backing_dev_info *bdi) { }
	781	+
	782	+static void cgwb_bdi_exit(struct backing_dev_info *bdi)
773	783	{
774	784	wb_congested_put(bdi->wb_congested);
775	785	}
@@ -857,6 +867,8 @@ int bdi_register_owner(struct backing_dev_info bdi, struct device owner) · 81 unmodified lines
857	867	MINOR(owner->devt));
858	868	if (rc)
859	869	return rc;
	870	+ /* Leaking owner reference... */
	871	+ WARN_ON(bdi->owner);
860	872	bdi->owner = owner;
861	873	get_device(owner);
862	874	return 0;
@@ -898,6 +910,7 @@ static void bdi_exit(struct backing_dev_info *bdi) · 35 unmodified lines
898	910	{
899	911	WARN_ON_ONCE(bdi->dev);
900	912	wb_exit(&bdi->wb);
	913	+ cgwb_bdi_exit(bdi);
901	914	}
902	915
903	916	static void release_bdi(struct kref *ref)

mm/gup.c+40−6

@@ -226,6 +226,7 @@ struct page follow_page_mask(struct vm_area_struct vma, · 225 unmodified lines
226	226	unsigned int *page_mask)
227	227	{
228	228	pgd_t *pgd;
	229	+ p4d_t *p4d;
229	230	pud_t *pud;
230	231	pmd_t *pmd;
231	232	spinlock_t *ptl;
@@ -243,8 +244,13 @@ struct page follow_page_mask(struct vm_area_struct vma, · 11 unmodified lines
243	244	pgd = pgd_offset(mm, address);
244	245	if (pgd_none(pgd) \|\| unlikely(pgd_bad(pgd)))
245	246	return no_page_table(vma, flags);
246		−
247		− pud = pud_offset(pgd, address);
	247	+ p4d = p4d_offset(pgd, address);
	248	+ if (p4d_none(*p4d))
	249	+ return no_page_table(vma, flags);
	250	+ BUILD_BUG_ON(p4d_huge(*p4d));
	251	+ if (unlikely(p4d_bad(*p4d)))
	252	+ return no_page_table(vma, flags);
	253	+ pud = pud_offset(p4d, address);
248	254	if (pud_none(*pud))
249	255	return no_page_table(vma, flags);
250	256	if (pud_huge(*pud) && vma->vm_flags & VM_HUGETLB) {
@@ -325,6 +331,7 @@ static int get_gate_page(struct mm_struct *mm, unsigned long address, · 74 unmodified lines
325	331	struct page **page)
326	332	{
327	333	pgd_t *pgd;
	334	+ p4d_t *p4d;
328	335	pud_t *pud;
329	336	pmd_t *pmd;
330	337	pte_t *pte;
@@ -338,7 +345,9 @@ static int get_gate_page(struct mm_struct *mm, unsigned long address, · 7 unmodified lines
338	345	else
339	346	pgd = pgd_offset_gate(mm, address);
340	347	BUG_ON(pgd_none(*pgd));
341		− pud = pud_offset(pgd, address);
	348	+ p4d = p4d_offset(pgd, address);
	349	+ BUG_ON(p4d_none(*p4d));
	350	+ pud = pud_offset(p4d, address);
342	351	BUG_ON(pud_none(*pud));
343	352	pmd = pmd_offset(pud, address);
344	353	if (pmd_none(*pmd))
@@ -1400,13 +1409,13 @@ static int gup_pmd_range(pud_t pud, unsigned long addr, unsigned long end, · 1055 unmodified lines
1400	1409	return 1;
1401	1410	}
1402	1411
1403		−static int gup_pud_range(pgd_t pgd, unsigned long addr, unsigned long end,
	1412	+static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
1404	1413	int write, struct page *pages, int nr)
1405	1414	{
1406	1415	unsigned long next;
1407	1416	pud_t *pudp;
1408	1417
1409		− pudp = pud_offset(&pgd, addr);
	1418	+ pudp = pud_offset(&p4d, addr);
1410	1419	do {
1411	1420	pud_t pud = READ_ONCE(*pudp);
1412	1421
@@ -1428,6 +1437,31 @@ static int gup_pud_range(pgd_t pgd, unsigned long addr, unsigned long end, · 15 unmodified lines
1428	1437	return 1;
1429	1438	}
1430	1439
	1440	+static int gup_p4d_range(pgd_t pgd, unsigned long addr, unsigned long end,
	1441	+ int write, struct page *pages, int nr)
	1442	+{
	1443	+ unsigned long next;
	1444	+ p4d_t *p4dp;
	1445	+
	1446	+ p4dp = p4d_offset(&pgd, addr);
	1447	+ do {
	1448	+ p4d_t p4d = READ_ONCE(*p4dp);
	1449	+
	1450	+ next = p4d_addr_end(addr, end);
	1451	+ if (p4d_none(p4d))
	1452	+ return 0;
	1453	+ BUILD_BUG_ON(p4d_huge(p4d));
	1454	+ if (unlikely(is_hugepd(__hugepd(p4d_val(p4d))))) {
	1455	+ if (!gup_huge_pd(__hugepd(p4d_val(p4d)), addr,
	1456	+ P4D_SHIFT, next, write, pages, nr))
	1457	+ return 0;
	1458	+ } else if (!gup_pud_range(p4d, addr, next, write, pages, nr))
	1459	+ return 0;
	1460	+ } while (p4dp++, addr = next, addr != end);
	1461	+
	1462	+ return 1;
	1463	+}
	1464	+
1431	1465	/*
1432	1466	* Like get_user_pages_fast() except it's IRQ-safe in that it won't fall back to
1433	1467	* the regular GUP. It will only return non-negative values.
@@ -1478,7 +1512,7 @@ int __get_user_pages_fast(unsigned long start, int nr_pages, int write, · 44 unmodified lines
1478	1512	if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
1479	1513	PGDIR_SHIFT, next, write, pages, &nr))
1480	1514	break;
1481		− } else if (!gup_pud_range(pgd, addr, next, write, pages, &nr))
	1515	+ } else if (!gup_p4d_range(pgd, addr, next, write, pages, &nr))
1482	1516	break;
1483	1517	} while (pgdp++, addr = next, addr != end);
1484	1518	local_irq_restore(flags);

mm/huge_memory.c+7−2

@@ -1828,7 +1828,7 @@ static void __split_huge_pud_locked(struct vm_area_struct vma, pud_t pud, · 1827 unmodified lines
1828	1828	VM_BUG_ON_VMA(vma->vm_end < haddr + HPAGE_PUD_SIZE, vma);
1829	1829	VM_BUG_ON(!pud_trans_huge(pud) && !pud_devmap(pud));
1830	1830
1831		− count_vm_event(THP_SPLIT_PMD);
	1831	+ count_vm_event(THP_SPLIT_PUD);
1832	1832
1833	1833	pudp_huge_clear_flush_notify(vma, haddr, pud);
1834	1834	}
@@ -2048,6 +2048,7 @@ void split_huge_pmd_address(struct vm_area_struct *vma, unsigned long address, · 213 unmodified lines
2048	2048	bool freeze, struct page *page)
2049	2049	{
2050	2050	pgd_t *pgd;
	2051	+ p4d_t *p4d;
2051	2052	pud_t *pud;
2052	2053	pmd_t *pmd;
2053	2054
@@ -2055,7 +2056,11 @@ void split_huge_pmd_address(struct vm_area_struct *vma, unsigned long address, · 1 unmodified line
2055	2056	if (!pgd_present(*pgd))
2056	2057	return;
2057	2058
2058		− pud = pud_offset(pgd, address);
	2059	+ p4d = p4d_offset(pgd, address);
	2060	+ if (!p4d_present(*p4d))
	2061	+ return;
	2062	+
	2063	+ pud = pud_offset(p4d, address);
2059	2064	if (!pud_present(*pud))
2060	2065	return;
2061	2066

mm/hugetlb.c+18−11

@@ -4555,7 +4555,8 @@ out: · 4554 unmodified lines
4555	4555	int huge_pmd_unshare(struct mm_struct mm, unsigned long addr, pte_t *ptep)
4556	4556	{
4557	4557	pgd_t pgd = pgd_offset(mm, addr);
4558		− pud_t pud = pud_offset(pgd, addr);
	4558	+ p4d_t p4d = p4d_offset(pgd, addr);
	4559	+ pud_t pud = pud_offset(p4d, addr);
4559	4560
4560	4561	BUG_ON(page_count(virt_to_page(ptep)) == 0);
4561	4562	if (page_count(virt_to_page(ptep)) == 1)
@@ -4586,11 +4587,13 @@ pte_t huge_pte_alloc(struct mm_struct mm, · 24 unmodified lines
4586	4587	unsigned long addr, unsigned long sz)
4587	4588	{
4588	4589	pgd_t *pgd;
	4590	+ p4d_t *p4d;
4589	4591	pud_t *pud;
4590	4592	pte_t *pte = NULL;
4591	4593
4592	4594	pgd = pgd_offset(mm, addr);
4593		− pud = pud_alloc(mm, pgd, addr);
	4595	+ p4d = p4d_offset(pgd, addr);
	4596	+ pud = pud_alloc(mm, p4d, addr);
4594	4597	if (pud) {
4595	4598	if (sz == PUD_SIZE) {
4596	4599	pte = (pte_t *)pud;
@@ -4610,18 +4613,22 @@ pte_t huge_pte_alloc(struct mm_struct mm, · 13 unmodified lines
4610	4613	pte_t huge_pte_offset(struct mm_struct mm, unsigned long addr)
4611	4614	{
4612	4615	pgd_t *pgd;
	4616	+ p4d_t *p4d;
4613	4617	pud_t *pud;
4614		− pmd_t *pmd = NULL;
	4618	+ pmd_t *pmd;
4615	4619
4616	4620	pgd = pgd_offset(mm, addr);
4617		− if (pgd_present(*pgd)) {
4618		− pud = pud_offset(pgd, addr);
4619		− if (pud_present(*pud)) {
4620		− if (pud_huge(*pud))
4621		− return (pte_t *)pud;
4622		− pmd = pmd_offset(pud, addr);
4623		− }
4624		− }
	4621	+ if (!pgd_present(*pgd))
	4622	+ return NULL;
	4623	+ p4d = p4d_offset(pgd, addr);
	4624	+ if (!p4d_present(*p4d))
	4625	+ return NULL;
	4626	+ pud = pud_offset(p4d, addr);
	4627	+ if (!pud_present(*pud))
	4628	+ return NULL;
	4629	+ if (pud_huge(*pud))
	4630	+ return (pte_t *)pud;
	4631	+ pmd = pmd_offset(pud, addr);
4625	4632	return (pte_t *) pmd;
4626	4633	}
4627	4634

mm/kasan/kasan_init.c+39−5

@@ -30,6 +30,9 @@ · 29 unmodified lines
30	30	*/
31	31	unsigned char kasan_zero_page[PAGE_SIZE] __page_aligned_bss;
32	32
	33	+#if CONFIG_PGTABLE_LEVELS > 4
	34	+p4d_t kasan_zero_p4d[PTRS_PER_P4D] __page_aligned_bss;
	35	+#endif
33	36	#if CONFIG_PGTABLE_LEVELS > 3
34	37	pud_t kasan_zero_pud[PTRS_PER_PUD] __page_aligned_bss;
35	38	#endif
@@ -82,10 +85,10 @@ static void __init zero_pmd_populate(pud_t *pud, unsigned long addr, · 46 unmodified lines
82	85	} while (pmd++, addr = next, addr != end);
83	86	}
84	87
85		−static void __init zero_pud_populate(pgd_t *pgd, unsigned long addr,
	88	+static void __init zero_pud_populate(p4d_t *p4d, unsigned long addr,
86	89	unsigned long end)
87	90	{
88		− pud_t *pud = pud_offset(pgd, addr);
	91	+ pud_t *pud = pud_offset(p4d, addr);
89	92	unsigned long next;
90	93
91	94	do {
@@ -107,6 +110,23 @@ static void __init zero_pud_populate(pgd_t *pgd, unsigned long addr, · 15 unmodified lines
107	110	} while (pud++, addr = next, addr != end);
108	111	}
109	112
	113	+static void __init zero_p4d_populate(pgd_t *pgd, unsigned long addr,
	114	+ unsigned long end)
	115	+{
	116	+ p4d_t *p4d = p4d_offset(pgd, addr);
	117	+ unsigned long next;
	118	+
	119	+ do {
	120	+ next = p4d_addr_end(addr, end);
	121	+
	122	+ if (p4d_none(*p4d)) {
	123	+ p4d_populate(&init_mm, p4d,
	124	+ early_alloc(PAGE_SIZE, NUMA_NO_NODE));
	125	+ }
	126	+ zero_pud_populate(p4d, addr, next);
	127	+ } while (p4d++, addr = next, addr != end);
	128	+}
	129	+
110	130	/**
111	131	* kasan_populate_zero_shadow - populate shadow memory region with
112	132	* kasan_zero_page
@@ -125,6 +145,7 @@ void __init kasan_populate_zero_shadow(const void *shadow_start, · 12 unmodified lines
125	145	next = pgd_addr_end(addr, end);
126	146
127	147	if (IS_ALIGNED(addr, PGDIR_SIZE) && end - addr >= PGDIR_SIZE) {
	148	+ p4d_t *p4d;
128	149	pud_t *pud;
129	150	pmd_t *pmd;
130	151
@@ -135,9 +156,22 @@ void __init kasan_populate_zero_shadow(const void *shadow_start, · 4 unmodified lines
135	156	* 3,2 - level page tables where we don't have
136	157	* puds,pmds, so pgd_populate(), pud_populate()
137	158	* is noops.
	159	+ *
	160	+ * The ifndef is required to avoid build breakage.
	161	+ *
	162	+ * With 5level-fixup.h, pgd_populate() is not nop and
	163	+ * we reference kasan_zero_p4d. It's not defined
	164	+ * unless 5-level paging enabled.
	165	+ *
	166	+ * The ifndef can be dropped once all KASAN-enabled
	167	+ * architectures will switch to pgtable-nop4d.h.
138	168	*/
139		− pgd_populate(&init_mm, pgd, lm_alias(kasan_zero_pud));
140		− pud = pud_offset(pgd, addr);
	169	+#ifndef __ARCH_HAS_5LEVEL_HACK
	170	+ pgd_populate(&init_mm, pgd, lm_alias(kasan_zero_p4d));
	171	+#endif
	172	+ p4d = p4d_offset(pgd, addr);
	173	+ p4d_populate(&init_mm, p4d, lm_alias(kasan_zero_pud));
	174	+ pud = pud_offset(p4d, addr);
141	175	pud_populate(&init_mm, pud, lm_alias(kasan_zero_pmd));
142	176	pmd = pmd_offset(pud, addr);
143	177	pmd_populate_kernel(&init_mm, pmd, lm_alias(kasan_zero_pte));
@@ -148,6 +182,6 @@ void __init kasan_populate_zero_shadow(const void *shadow_start, · 4 unmodified lines
148	182	pgd_populate(&init_mm, pgd,
149	183	early_alloc(PAGE_SIZE, NUMA_NO_NODE));
150	184	}
151		− zero_pud_populate(pgd, addr, next);
	185	+ zero_p4d_populate(pgd, addr, next);
152	186	} while (pgd++, addr = next, addr != end);
153	187	}

mm/kasan/quarantine.c+44−7

@@ -25,6 +25,7 @@ · 24 unmodified lines
25	25	#include <linux/printk.h>
26	26	#include <linux/shrinker.h>
27	27	#include <linux/slab.h>
	28	+#include <linux/srcu.h>
28	29	#include <linux/string.h>
29	30	#include <linux/types.h>
30	31
@@ -103,6 +104,7 @@ static int quarantine_tail; · 72 unmodified lines
103	104	/* Total size of all objects in global_quarantine across all batches. */
104	105	static unsigned long quarantine_size;
105	106	static DEFINE_SPINLOCK(quarantine_lock);
	107	+DEFINE_STATIC_SRCU(remove_cache_srcu);
106	108
107	109	/* Maximum size of the global queue. */
108	110	static unsigned long quarantine_max_size;
@@ -173,17 +175,22 @@ void quarantine_put(struct kasan_free_meta info, struct kmem_cache cache) · 64 unmodified lines
173	175	struct qlist_head *q;
174	176	struct qlist_head temp = QLIST_INIT;
175	177
	178	+ /*
	179	+ * Note: irq must be disabled until after we move the batch to the
	180	+ * global quarantine. Otherwise quarantine_remove_cache() can miss
	181	+ * some objects belonging to the cache if they are in our local temp
	182	+ * list. quarantine_remove_cache() executes on_each_cpu() at the
	183	+ * beginning which ensures that it either sees the objects in per-cpu
	184	+ * lists or in the global quarantine.
	185	+ */
176	186	local_irq_save(flags);
177	187
178	188	q = this_cpu_ptr(&cpu_quarantine);
179	189	qlist_put(q, &info->quarantine_link, cache->size);
180		− if (unlikely(q->bytes > QUARANTINE_PERCPU_SIZE))
	190	+ if (unlikely(q->bytes > QUARANTINE_PERCPU_SIZE)) {
181	191	qlist_move_all(q, &temp);
182	192
183		− local_irq_restore(flags);
184		−
185		− if (unlikely(!qlist_empty(&temp))) {
186		− spin_lock_irqsave(&quarantine_lock, flags);
	193	+ spin_lock(&quarantine_lock);
187	194	WRITE_ONCE(quarantine_size, quarantine_size + temp.bytes);
188	195	qlist_move_all(&temp, &global_quarantine[quarantine_tail]);
189	196	if (global_quarantine[quarantine_tail].bytes >=
@@ -196,20 +203,33 @@ void quarantine_put(struct kasan_free_meta info, struct kmem_cache cache) · 6 unmodified lines
196	203	if (new_tail != quarantine_head)
197	204	quarantine_tail = new_tail;
198	205	}
199		− spin_unlock_irqrestore(&quarantine_lock, flags);
	206	+ spin_unlock(&quarantine_lock);
200	207	}
	208	+
	209	+ local_irq_restore(flags);
201	210	}
202	211
203	212	void quarantine_reduce(void)
204	213	{
205	214	size_t total_size, new_quarantine_size, percpu_quarantines;
206	215	unsigned long flags;
	216	+ int srcu_idx;
207	217	struct qlist_head to_free = QLIST_INIT;
208	218
209	219	if (likely(READ_ONCE(quarantine_size) <=
210	220	READ_ONCE(quarantine_max_size)))
211	221	return;
212	222
	223	+ /*
	224	+ * srcu critical section ensures that quarantine_remove_cache()
	225	+ * will not miss objects belonging to the cache while they are in our
	226	+ * local to_free list. srcu is chosen because (1) it gives us private
	227	+ * grace period domain that does not interfere with anything else,
	228	+ * and (2) it allows synchronize_srcu() to return without waiting
	229	+ * if there are no pending read critical sections (which is the
	230	+ * expected case).
	231	+ */
	232	+ srcu_idx = srcu_read_lock(&remove_cache_srcu);
213	233	spin_lock_irqsave(&quarantine_lock, flags);
214	234
215	235	/*
@@ -237,6 +257,7 @@ void quarantine_reduce(void) · 21 unmodified lines
237	257	spin_unlock_irqrestore(&quarantine_lock, flags);
238	258
239	259	qlist_free_all(&to_free, NULL);
	260	+ srcu_read_unlock(&remove_cache_srcu, srcu_idx);
240	261	}
241	262
242	263	static void qlist_move_cache(struct qlist_head *from,
@@ -280,12 +301,28 @@ void quarantine_remove_cache(struct kmem_cache *cache) · 37 unmodified lines
280	301	unsigned long flags, i;
281	302	struct qlist_head to_free = QLIST_INIT;
282	303
	304	+ /*
	305	+ * Must be careful to not miss any objects that are being moved from
	306	+ * per-cpu list to the global quarantine in quarantine_put(),
	307	+ * nor objects being freed in quarantine_reduce(). on_each_cpu()
	308	+ * achieves the first goal, while synchronize_srcu() achieves the
	309	+ * second.
	310	+ */
283	311	on_each_cpu(per_cpu_remove_cache, cache, 1);
284	312
285	313	spin_lock_irqsave(&quarantine_lock, flags);
286		− for (i = 0; i < QUARANTINE_BATCHES; i++)
	314	+ for (i = 0; i < QUARANTINE_BATCHES; i++) {
	315	+ if (qlist_empty(&global_quarantine[i]))
	316	+ continue;
287	317	qlist_move_cache(&global_quarantine[i], &to_free, cache);
	318	+ /* Scanning whole quarantine can take a while. */
	319	+ spin_unlock_irqrestore(&quarantine_lock, flags);
	320	+ cond_resched();
	321	+ spin_lock_irqsave(&quarantine_lock, flags);
	322	+ }
288	323	spin_unlock_irqrestore(&quarantine_lock, flags);
289	324
290	325	qlist_free_all(&to_free, cache);
	326	+
	327	+ synchronize_srcu(&remove_cache_srcu);
291	328	}

mm/madvise.c+41−3

@@ -513,7 +513,43 @@ static long madvise_dontneed(struct vm_area_struct *vma, · 512 unmodified lines
513	513	if (!can_madv_dontneed_vma(vma))
514	514	return -EINVAL;
515	515
516		− userfaultfd_remove(vma, prev, start, end);
	516	+ if (!userfaultfd_remove(vma, start, end)) {
	517	+ prev = NULL; / mmap_sem has been dropped, prev is stale */
	518	+
	519	+ down_read(&current->mm->mmap_sem);
	520	+ vma = find_vma(current->mm, start);
	521	+ if (!vma)
	522	+ return -ENOMEM;
	523	+ if (start < vma->vm_start) {
	524	+ /*
	525	+ * This "vma" under revalidation is the one
	526	+ * with the lowest vma->vm_start where start
	527	+ * is also < vma->vm_end. If start <
	528	+ * vma->vm_start it means an hole materialized
	529	+ * in the user address space within the
	530	+ * virtual range passed to MADV_DONTNEED.
	531	+ */
	532	+ return -ENOMEM;
	533	+ }
	534	+ if (!can_madv_dontneed_vma(vma))
	535	+ return -EINVAL;
	536	+ if (end > vma->vm_end) {
	537	+ /*
	538	+ * Don't fail if end > vma->vm_end. If the old
	539	+ * vma was splitted while the mmap_sem was
	540	+ * released the effect of the concurrent
	541	+ * operation may not cause MADV_DONTNEED to
	542	+ * have an undefined result. There may be an
	543	+ * adjacent next vma that we'll walk
	544	+ * next. userfaultfd_remove() will generate an
	545	+ * UFFD_EVENT_REMOVE repetition on the
	546	+ * end-vma->vm_end range, but the manager can
	547	+ * handle a repetition fine.
	548	+ */
	549	+ end = vma->vm_end;
	550	+ }
	551	+ VM_WARN_ON(start >= end);
	552	+ }
517	553	zap_page_range(vma, start, end - start);
518	554	return 0;
519	555	}
@@ -554,8 +590,10 @@ static long madvise_remove(struct vm_area_struct *vma, · 34 unmodified lines
554	590	* mmap_sem.
555	591	*/
556	592	get_file(f);
557		− userfaultfd_remove(vma, prev, start, end);
558		− up_read(&current->mm->mmap_sem);
	593	+ if (userfaultfd_remove(vma, start, end)) {
	594	+ /* mmap_sem was not released by userfaultfd_remove() */
	595	+ up_read(&current->mm->mmap_sem);
	596	+ }
559	597	error = vfs_fallocate(f,
560	598	FALLOC_FL_PUNCH_HOLE \| FALLOC_FL_KEEP_SIZE,
561	599	offset, end - start);

mm/memblock.c+4−1

@@ -1118,7 +1118,10 @@ unsigned long __init_memblock memblock_next_valid_pfn(unsigned long pfn, · 1117 unmodified lines
1118	1118	}
1119	1119	} while (left < right);
1120	1120
1121		− return min(PHYS_PFN(type->regions[right].base), max_pfn);
	1121	+ if (right == type->cnt)
	1122	+ return max_pfn;
	1123	+ else
	1124	+ return min(PHYS_PFN(type->regions[right].base), max_pfn);
1122	1125	}
1123	1126
1124	1127	/**

mm/memcontrol.c+13−5

@@ -466,6 +466,8 @@ static void mem_cgroup_update_tree(struct mem_cgroup memcg, struct page page) · 465 unmodified lines
466	466	struct mem_cgroup_tree_per_node *mctz;
467	467
468	468	mctz = soft_limit_tree_from_page(page);
	469	+ if (!mctz)
	470	+ return;
469	471	/*
470	472	* Necessary to update all ancestors when hierarchy is used.
471	473	* because their event counter is not touched.
@@ -503,7 +505,8 @@ static void mem_cgroup_remove_from_trees(struct mem_cgroup *memcg) · 31 unmodified lines
503	505	for_each_node(nid) {
504	506	mz = mem_cgroup_nodeinfo(memcg, nid);
505	507	mctz = soft_limit_tree_node(nid);
506		− mem_cgroup_remove_exceeded(mz, mctz);
	508	+ if (mctz)
	509	+ mem_cgroup_remove_exceeded(mz, mctz);
507	510	}
508	511	}
509	512
@@ -2558,7 +2561,7 @@ unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order, · 2048 unmodified lines
2558	2561	* is empty. Do it lockless to prevent lock bouncing. Races
2559	2562	* are acceptable as soft limit is best effort anyway.
2560	2563	*/
2561		− if (RB_EMPTY_ROOT(&mctz->rb_root))
	2564	+ if (!mctz \|\| RB_EMPTY_ROOT(&mctz->rb_root))
2562	2565	return 0;
2563	2566
2564	2567	/*
@@ -4135,17 +4138,22 @@ static void free_mem_cgroup_per_node_info(struct mem_cgroup *memcg, int node) · 1570 unmodified lines
4135	4138	kfree(memcg->nodeinfo[node]);
4136	4139	}
4137	4140
4138		−static void mem_cgroup_free(struct mem_cgroup *memcg)
	4141	+static void __mem_cgroup_free(struct mem_cgroup *memcg)
4139	4142	{
4140	4143	int node;
4141	4144
4142		− memcg_wb_domain_exit(memcg);
4143	4145	for_each_node(node)
4144	4146	free_mem_cgroup_per_node_info(memcg, node);
4145	4147	free_percpu(memcg->stat);
4146	4148	kfree(memcg);
4147	4149	}
4148	4150
	4151	+static void mem_cgroup_free(struct mem_cgroup *memcg)
	4152	+{
	4153	+ memcg_wb_domain_exit(memcg);
	4154	+ __mem_cgroup_free(memcg);
	4155	+}
	4156	+
4149	4157	static struct mem_cgroup *mem_cgroup_alloc(void)
4150	4158	{
4151	4159	struct mem_cgroup *memcg;
@@ -4196,7 +4204,7 @@ static struct mem_cgroup *mem_cgroup_alloc(void) · 44 unmodified lines
4196	4204	fail:
4197	4205	if (memcg->id.id > 0)
4198	4206	idr_remove(&mem_cgroup_idr, memcg->id.id);
4199		− mem_cgroup_free(memcg);
	4207	+ __mem_cgroup_free(memcg);
4200	4208	return NULL;
4201	4209	}
4202	4210

mm/memory.c+196−34

@@ -445,7 +445,7 @@ static inline void free_pmd_range(struct mmu_gather tlb, pud_t pud, · 444 unmodified lines
445	445	mm_dec_nr_pmds(tlb->mm);
446	446	}
447	447
448		−static inline void free_pud_range(struct mmu_gather tlb, pgd_t pgd,
	448	+static inline void free_pud_range(struct mmu_gather tlb, p4d_t p4d,
449	449	unsigned long addr, unsigned long end,
450	450	unsigned long floor, unsigned long ceiling)
451	451	{
@@ -454,7 +454,7 @@ static inline void free_pud_range(struct mmu_gather tlb, pgd_t pgd, · 2 unmodified lines
454	454	unsigned long start;
455	455
456	456	start = addr;
457		− pud = pud_offset(pgd, addr);
	457	+ pud = pud_offset(p4d, addr);
458	458	do {
459	459	next = pud_addr_end(addr, end);
460	460	if (pud_none_or_clear_bad(pud))
@@ -462,6 +462,39 @@ static inline void free_pud_range(struct mmu_gather tlb, pgd_t pgd, · 1 unmodified line
462	462	free_pmd_range(tlb, pud, addr, next, floor, ceiling);
463	463	} while (pud++, addr = next, addr != end);
464	464
	465	+ start &= P4D_MASK;
	466	+ if (start < floor)
	467	+ return;
	468	+ if (ceiling) {
	469	+ ceiling &= P4D_MASK;
	470	+ if (!ceiling)
	471	+ return;
	472	+ }
	473	+ if (end - 1 > ceiling - 1)
	474	+ return;
	475	+
	476	+ pud = pud_offset(p4d, start);
	477	+ p4d_clear(p4d);
	478	+ pud_free_tlb(tlb, pud, start);
	479	+}
	480	+
	481	+static inline void free_p4d_range(struct mmu_gather tlb, pgd_t pgd,
	482	+ unsigned long addr, unsigned long end,
	483	+ unsigned long floor, unsigned long ceiling)
	484	+{
	485	+ p4d_t *p4d;
	486	+ unsigned long next;
	487	+ unsigned long start;
	488	+
	489	+ start = addr;
	490	+ p4d = p4d_offset(pgd, addr);
	491	+ do {
	492	+ next = p4d_addr_end(addr, end);
	493	+ if (p4d_none_or_clear_bad(p4d))
	494	+ continue;
	495	+ free_pud_range(tlb, p4d, addr, next, floor, ceiling);
	496	+ } while (p4d++, addr = next, addr != end);
	497	+
465	498	start &= PGDIR_MASK;
466	499	if (start < floor)
467	500	return;
@@ -473,9 +506,9 @@ static inline void free_pud_range(struct mmu_gather tlb, pgd_t pgd, · 5 unmodified lines
473	506	if (end - 1 > ceiling - 1)
474	507	return;
475	508
476		− pud = pud_offset(pgd, start);
	509	+ p4d = p4d_offset(pgd, start);
477	510	pgd_clear(pgd);
478		− pud_free_tlb(tlb, pud, start);
	511	+ p4d_free_tlb(tlb, p4d, start);
479	512	}
480	513
481	514	/*
@@ -539,7 +572,7 @@ void free_pgd_range(struct mmu_gather *tlb, · 57 unmodified lines
539	572	next = pgd_addr_end(addr, end);
540	573	if (pgd_none_or_clear_bad(pgd))
541	574	continue;
542		− free_pud_range(tlb, pgd, addr, next, floor, ceiling);
	575	+ free_p4d_range(tlb, pgd, addr, next, floor, ceiling);
543	576	} while (pgd++, addr = next, addr != end);
544	577	}
545	578
@@ -658,7 +691,8 @@ static void print_bad_pte(struct vm_area_struct *vma, unsigned long addr, · 112 unmodified lines
658	691	pte_t pte, struct page *page)
659	692	{
660	693	pgd_t *pgd = pgd_offset(vma->vm_mm, addr);
661		− pud_t *pud = pud_offset(pgd, addr);
	694	+ p4d_t *p4d = p4d_offset(pgd, addr);
	695	+ pud_t *pud = pud_offset(p4d, addr);
662	696	pmd_t *pmd = pmd_offset(pud, addr);
663	697	struct address_space *mapping;
664	698	pgoff_t index;
@@ -1023,16 +1057,16 @@ static inline int copy_pmd_range(struct mm_struct dst_mm, struct mm_struct src · 358 unmodified lines
1023	1057	}
1024	1058
1025	1059	static inline int copy_pud_range(struct mm_struct dst_mm, struct mm_struct src_mm,
1026		− pgd_t dst_pgd, pgd_t src_pgd, struct vm_area_struct *vma,
	1060	+ p4d_t dst_p4d, p4d_t src_p4d, struct vm_area_struct *vma,
1027	1061	unsigned long addr, unsigned long end)
1028	1062	{
1029	1063	pud_t src_pud, dst_pud;
1030	1064	unsigned long next;
1031	1065
1032		− dst_pud = pud_alloc(dst_mm, dst_pgd, addr);
	1066	+ dst_pud = pud_alloc(dst_mm, dst_p4d, addr);
1033	1067	if (!dst_pud)
1034	1068	return -ENOMEM;
1035		− src_pud = pud_offset(src_pgd, addr);
	1069	+ src_pud = pud_offset(src_p4d, addr);
1036	1070	do {
1037	1071	next = pud_addr_end(addr, end);
1038	1072	if (pud_trans_huge(src_pud) \|\| pud_devmap(src_pud)) {
@@ -1056,6 +1090,28 @@ static inline int copy_pud_range(struct mm_struct dst_mm, struct mm_struct src · 17 unmodified lines
1056	1090	return 0;
1057	1091	}
1058	1092
	1093	+static inline int copy_p4d_range(struct mm_struct dst_mm, struct mm_struct src_mm,
	1094	+ pgd_t dst_pgd, pgd_t src_pgd, struct vm_area_struct *vma,
	1095	+ unsigned long addr, unsigned long end)
	1096	+{
	1097	+ p4d_t src_p4d, dst_p4d;
	1098	+ unsigned long next;
	1099	+
	1100	+ dst_p4d = p4d_alloc(dst_mm, dst_pgd, addr);
	1101	+ if (!dst_p4d)
	1102	+ return -ENOMEM;
	1103	+ src_p4d = p4d_offset(src_pgd, addr);
	1104	+ do {
	1105	+ next = p4d_addr_end(addr, end);
	1106	+ if (p4d_none_or_clear_bad(src_p4d))
	1107	+ continue;
	1108	+ if (copy_pud_range(dst_mm, src_mm, dst_p4d, src_p4d,
	1109	+ vma, addr, next))
	1110	+ return -ENOMEM;
	1111	+ } while (dst_p4d++, src_p4d++, addr = next, addr != end);
	1112	+ return 0;
	1113	+}
	1114	+
1059	1115	int copy_page_range(struct mm_struct dst_mm, struct mm_struct src_mm,
1060	1116	struct vm_area_struct *vma)
1061	1117	{
@@ -1111,7 +1167,7 @@ int copy_page_range(struct mm_struct dst_mm, struct mm_struct src_mm, · 49 unmodified lines
1111	1167	next = pgd_addr_end(addr, end);
1112	1168	if (pgd_none_or_clear_bad(src_pgd))
1113	1169	continue;
1114		− if (unlikely(copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd,
	1170	+ if (unlikely(copy_p4d_range(dst_mm, src_mm, dst_pgd, src_pgd,
1115	1171	vma, addr, next))) {
1116	1172	ret = -ENOMEM;
1117	1173	break;
@@ -1267,14 +1323,14 @@ next: · 149 unmodified lines
1267	1323	}
1268	1324
1269	1325	static inline unsigned long zap_pud_range(struct mmu_gather *tlb,
1270		− struct vm_area_struct vma, pgd_t pgd,
	1326	+ struct vm_area_struct vma, p4d_t p4d,
1271	1327	unsigned long addr, unsigned long end,
1272	1328	struct zap_details *details)
1273	1329	{
1274	1330	pud_t *pud;
1275	1331	unsigned long next;
1276	1332
1277		− pud = pud_offset(pgd, addr);
	1333	+ pud = pud_offset(p4d, addr);
1278	1334	do {
1279	1335	next = pud_addr_end(addr, end);
1280	1336	if (pud_trans_huge(pud) \|\| pud_devmap(pud)) {
@@ -1295,6 +1351,25 @@ next: · 14 unmodified lines
1295	1351	return addr;
1296	1352	}
1297	1353
	1354	+static inline unsigned long zap_p4d_range(struct mmu_gather *tlb,
	1355	+ struct vm_area_struct vma, pgd_t pgd,
	1356	+ unsigned long addr, unsigned long end,
	1357	+ struct zap_details *details)
	1358	+{
	1359	+ p4d_t *p4d;
	1360	+ unsigned long next;
	1361	+
	1362	+ p4d = p4d_offset(pgd, addr);
	1363	+ do {
	1364	+ next = p4d_addr_end(addr, end);
	1365	+ if (p4d_none_or_clear_bad(p4d))
	1366	+ continue;
	1367	+ next = zap_pud_range(tlb, vma, p4d, addr, next, details);
	1368	+ } while (p4d++, addr = next, addr != end);
	1369	+
	1370	+ return addr;
	1371	+}
	1372	+
1298	1373	void unmap_page_range(struct mmu_gather *tlb,
1299	1374	struct vm_area_struct *vma,
1300	1375	unsigned long addr, unsigned long end,
@@ -1310,7 +1385,7 @@ void unmap_page_range(struct mmu_gather *tlb, · 9 unmodified lines
1310	1385	next = pgd_addr_end(addr, end);
1311	1386	if (pgd_none_or_clear_bad(pgd))
1312	1387	continue;
1313		− next = zap_pud_range(tlb, vma, pgd, addr, next, details);
	1388	+ next = zap_p4d_range(tlb, vma, pgd, addr, next, details);
1314	1389	} while (pgd++, addr = next, addr != end);
1315	1390	tlb_end_vma(tlb, vma);
1316	1391	}
@@ -1465,16 +1540,24 @@ EXPORT_SYMBOL_GPL(zap_vma_ptes); · 148 unmodified lines
1465	1540	pte_t __get_locked_pte(struct mm_struct mm, unsigned long addr,
1466	1541	spinlock_t **ptl)
1467	1542	{
1468		− pgd_t *pgd = pgd_offset(mm, addr);
1469		− pud_t *pud = pud_alloc(mm, pgd, addr);
1470		− if (pud) {
1471		− pmd_t *pmd = pmd_alloc(mm, pud, addr);
1472		− if (pmd) {
1473		− VM_BUG_ON(pmd_trans_huge(*pmd));
1474		− return pte_alloc_map_lock(mm, pmd, addr, ptl);
1475		− }
1476		− }
1477		− return NULL;
	1543	+ pgd_t *pgd;
	1544	+ p4d_t *p4d;
	1545	+ pud_t *pud;
	1546	+ pmd_t *pmd;
	1547	+
	1548	+ pgd = pgd_offset(mm, addr);
	1549	+ p4d = p4d_alloc(mm, pgd, addr);
	1550	+ if (!p4d)
	1551	+ return NULL;
	1552	+ pud = pud_alloc(mm, p4d, addr);
	1553	+ if (!pud)
	1554	+ return NULL;
	1555	+ pmd = pmd_alloc(mm, pud, addr);
	1556	+ if (!pmd)
	1557	+ return NULL;
	1558	+
	1559	+ VM_BUG_ON(pmd_trans_huge(*pmd));
	1560	+ return pte_alloc_map_lock(mm, pmd, addr, ptl);
1478	1561	}
1479	1562
1480	1563	/*
@@ -1740,7 +1823,7 @@ static inline int remap_pmd_range(struct mm_struct mm, pud_t pud, · 259 unmodified lines
1740	1823	return 0;
1741	1824	}
1742	1825
1743		−static inline int remap_pud_range(struct mm_struct mm, pgd_t pgd,
	1826	+static inline int remap_pud_range(struct mm_struct mm, p4d_t p4d,
1744	1827	unsigned long addr, unsigned long end,
1745	1828	unsigned long pfn, pgprot_t prot)
1746	1829	{
@@ -1748,7 +1831,7 @@ static inline int remap_pud_range(struct mm_struct mm, pgd_t pgd, · 1 unmodified line
1748	1831	unsigned long next;
1749	1832
1750	1833	pfn -= addr >> PAGE_SHIFT;
1751		− pud = pud_alloc(mm, pgd, addr);
	1834	+ pud = pud_alloc(mm, p4d, addr);
1752	1835	if (!pud)
1753	1836	return -ENOMEM;
1754	1837	do {
@@ -1760,6 +1843,26 @@ static inline int remap_pud_range(struct mm_struct mm, pgd_t pgd, · 5 unmodified lines
1760	1843	return 0;
1761	1844	}
1762	1845
	1846	+static inline int remap_p4d_range(struct mm_struct mm, pgd_t pgd,
	1847	+ unsigned long addr, unsigned long end,
	1848	+ unsigned long pfn, pgprot_t prot)
	1849	+{
	1850	+ p4d_t *p4d;
	1851	+ unsigned long next;
	1852	+
	1853	+ pfn -= addr >> PAGE_SHIFT;
	1854	+ p4d = p4d_alloc(mm, pgd, addr);
	1855	+ if (!p4d)
	1856	+ return -ENOMEM;
	1857	+ do {
	1858	+ next = p4d_addr_end(addr, end);
	1859	+ if (remap_pud_range(mm, p4d, addr, next,
	1860	+ pfn + (addr >> PAGE_SHIFT), prot))
	1861	+ return -ENOMEM;
	1862	+ } while (p4d++, addr = next, addr != end);
	1863	+ return 0;
	1864	+}
	1865	+
1763	1866	/**
1764	1867	* remap_pfn_range - remap kernel memory to userspace
1765	1868	* @vma: user vma to map to
@@ -1816,7 +1919,7 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr, · 50 unmodified lines
1816	1919	flush_cache_range(vma, addr, end);
1817	1920	do {
1818	1921	next = pgd_addr_end(addr, end);
1819		− err = remap_pud_range(mm, pgd, addr, next,
	1922	+ err = remap_p4d_range(mm, pgd, addr, next,
1820	1923	pfn + (addr >> PAGE_SHIFT), prot);
1821	1924	if (err)
1822	1925	break;
@@ -1932,7 +2035,7 @@ static int apply_to_pmd_range(struct mm_struct mm, pud_t pud, · 109 unmodified lines
1932	2035	return err;
1933	2036	}
1934	2037
1935		−static int apply_to_pud_range(struct mm_struct mm, pgd_t pgd,
	2038	+static int apply_to_pud_range(struct mm_struct mm, p4d_t p4d,
1936	2039	unsigned long addr, unsigned long end,
1937	2040	pte_fn_t fn, void *data)
1938	2041	{
@@ -1940,7 +2043,7 @@ static int apply_to_pud_range(struct mm_struct mm, pgd_t pgd, · 1 unmodified line
1940	2043	unsigned long next;
1941	2044	int err;
1942	2045
1943		− pud = pud_alloc(mm, pgd, addr);
	2046	+ pud = pud_alloc(mm, p4d, addr);
1944	2047	if (!pud)
1945	2048	return -ENOMEM;
1946	2049	do {
@@ -1952,6 +2055,26 @@ static int apply_to_pud_range(struct mm_struct mm, pgd_t pgd, · 5 unmodified lines
1952	2055	return err;
1953	2056	}
1954	2057
	2058	+static int apply_to_p4d_range(struct mm_struct mm, pgd_t pgd,
	2059	+ unsigned long addr, unsigned long end,
	2060	+ pte_fn_t fn, void *data)
	2061	+{
	2062	+ p4d_t *p4d;
	2063	+ unsigned long next;
	2064	+ int err;
	2065	+
	2066	+ p4d = p4d_alloc(mm, pgd, addr);
	2067	+ if (!p4d)
	2068	+ return -ENOMEM;
	2069	+ do {
	2070	+ next = p4d_addr_end(addr, end);
	2071	+ err = apply_to_pud_range(mm, p4d, addr, next, fn, data);
	2072	+ if (err)
	2073	+ break;
	2074	+ } while (p4d++, addr = next, addr != end);
	2075	+ return err;
	2076	+}
	2077	+
1955	2078	/*
1956	2079	* Scan a region of virtual memory, filling in page tables as necessary
1957	2080	* and calling a provided function on each leaf page table.
@@ -1970,7 +2093,7 @@ int apply_to_page_range(struct mm_struct *mm, unsigned long addr, · 12 unmodified lines
1970	2093	pgd = pgd_offset(mm, addr);
1971	2094	do {
1972	2095	next = pgd_addr_end(addr, end);
1973		− err = apply_to_pud_range(mm, pgd, addr, next, fn, data);
	2096	+ err = apply_to_p4d_range(mm, pgd, addr, next, fn, data);
1974	2097	if (err)
1975	2098	break;
1976	2099	} while (pgd++, addr = next, addr != end);
@@ -3653,11 +3776,15 @@ static int __handle_mm_fault(struct vm_area_struct *vma, unsigned long address, · 1676 unmodified lines
3653	3776	};
3654	3777	struct mm_struct *mm = vma->vm_mm;
3655	3778	pgd_t *pgd;
	3779	+ p4d_t *p4d;
3656	3780	int ret;
3657	3781
3658	3782	pgd = pgd_offset(mm, address);
	3783	+ p4d = p4d_alloc(mm, pgd, address);
	3784	+ if (!p4d)
	3785	+ return VM_FAULT_OOM;
3659	3786
3660		− vmf.pud = pud_alloc(mm, pgd, address);
	3787	+ vmf.pud = pud_alloc(mm, p4d, address);
3661	3788	if (!vmf.pud)
3662	3789	return VM_FAULT_OOM;
3663	3790	if (pud_none(*vmf.pud) && transparent_hugepage_enabled(vma)) {
@@ -3779,12 +3906,35 @@ int handle_mm_fault(struct vm_area_struct *vma, unsigned long address, · 115 unmodified lines
3779	3906	}
3780	3907	EXPORT_SYMBOL_GPL(handle_mm_fault);
3781	3908
	3909	+#ifndef __PAGETABLE_P4D_FOLDED
	3910	+/*
	3911	+ * Allocate p4d page table.
	3912	+ * We've already handled the fast-path in-line.
	3913	+ */
	3914	+int __p4d_alloc(struct mm_struct mm, pgd_t pgd, unsigned long address)
	3915	+{
	3916	+ p4d_t *new = p4d_alloc_one(mm, address);
	3917	+ if (!new)
	3918	+ return -ENOMEM;
	3919	+
	3920	+ smp_wmb(); /* See comment in __pte_alloc */
	3921	+
	3922	+ spin_lock(&mm->page_table_lock);
	3923	+ if (pgd_present(pgd)) / Another has populated it */
	3924	+ p4d_free(mm, new);
	3925	+ else
	3926	+ pgd_populate(mm, pgd, new);
	3927	+ spin_unlock(&mm->page_table_lock);
	3928	+ return 0;
	3929	+}
	3930	+#endif /* __PAGETABLE_P4D_FOLDED */
	3931	+
3782	3932	#ifndef __PAGETABLE_PUD_FOLDED
3783	3933	/*
3784	3934	* Allocate page upper directory.
3785	3935	* We've already handled the fast-path in-line.
3786	3936	*/
3787		−int __pud_alloc(struct mm_struct mm, pgd_t pgd, unsigned long address)
	3937	+int __pud_alloc(struct mm_struct mm, p4d_t p4d, unsigned long address)
3788	3938	{
3789	3939	pud_t *new = pud_alloc_one(mm, address);
3790	3940	if (!new)
@@ -3793,10 +3943,17 @@ int __pud_alloc(struct mm_struct mm, pgd_t pgd, unsigned long address) · 2 unmodified lines
3793	3943	smp_wmb(); /* See comment in __pte_alloc */
3794	3944
3795	3945	spin_lock(&mm->page_table_lock);
3796		− if (pgd_present(pgd)) / Another has populated it */
	3946	+#ifndef __ARCH_HAS_5LEVEL_HACK
	3947	+ if (p4d_present(p4d)) / Another has populated it */
3797	3948	pud_free(mm, new);
3798	3949	else
3799		− pgd_populate(mm, pgd, new);
	3950	+ p4d_populate(mm, p4d, new);
	3951	+#else
	3952	+ if (pgd_present(p4d)) / Another has populated it */
	3953	+ pud_free(mm, new);
	3954	+ else
	3955	+ pgd_populate(mm, p4d, new);
	3956	+#endif /* __ARCH_HAS_5LEVEL_HACK */
3800	3957	spin_unlock(&mm->page_table_lock);
3801	3958	return 0;
3802	3959	}
@@ -3839,6 +3996,7 @@ static int __follow_pte_pmd(struct mm_struct *mm, unsigned long address, · 36 unmodified lines
3839	3996	pte_t ptepp, pmd_t pmdpp, spinlock_t **ptlp)
3840	3997	{
3841	3998	pgd_t *pgd;
	3999	+ p4d_t *p4d;
3842	4000	pud_t *pud;
3843	4001	pmd_t *pmd;
3844	4002	pte_t *ptep;
@@ -3847,7 +4005,11 @@ static int __follow_pte_pmd(struct mm_struct *mm, unsigned long address, · 2 unmodified lines
3847	4005	if (pgd_none(pgd) \|\| unlikely(pgd_bad(pgd)))
3848	4006	goto out;
3849	4007
3850		− pud = pud_offset(pgd, address);
	4008	+ p4d = p4d_offset(pgd, address);
	4009	+ if (p4d_none(p4d) \|\| unlikely(p4d_bad(p4d)))
	4010	+ goto out;
	4011	+

4 more lines aren’t shown here.

mm/mlock.c+5−5

This diff is large: its 38 lines aren’t shown here.

mm/mprotect.c+23−3

This diff is large: its 53 lines aren’t shown here.

mm/mremap.c+11−2

This diff is large: its 38 lines aren’t shown here.

mm/page_alloc.c+2−1

This diff is large: its 10 lines aren’t shown here.

mm/page_vma_mapped.c+5−1

This diff is large: its 20 lines aren’t shown here.

mm/pagewalk.c+29−3

This diff is large: its 59 lines aren’t shown here.

mm/percpu-vm.c+3−4

This diff is large: its 35 lines aren’t shown here.

mm/percpu.c+4−1

This diff is large: its 13 lines aren’t shown here.

mm/pgtable-generic.c+6−0

This diff is large: its 13 lines aren’t shown here.

mm/rmap.c+13−7

This diff is large: its 48 lines aren’t shown here.

mm/sparse-vmemmap.c+19−3

This diff is large: its 51 lines aren’t shown here.

mm/swapfile.c+23−3

This diff is large: its 54 lines aren’t shown here.

mm/userfaultfd.c+13−10

This diff is large: its 33 lines aren’t shown here.

mm/vmalloc.c+60−21

This diff is large: its 146 lines aren’t shown here.

mm/vmstat.c+3−0

This diff is large: its 10 lines aren’t shown here.

net/atm/svc.c+3−2

This diff is large: its 19 lines aren’t shown here.

net/ax25/af_ax25.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/bluetooth/l2cap_sock.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/bluetooth/rfcomm/sock.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/bluetooth/sco.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/bridge/br_input.c+1−0

This diff is large: its 8 lines aren’t shown here.

net/bridge/br_netfilter_hooks.c+0−21

This diff is large: its 35 lines aren’t shown here.

net/ceph/ceph_common.c+15−0

This diff is large: its 50 lines aren’t shown here.

net/ceph/osd_client.c+35−1

This diff is large: its 79 lines aren’t shown here.

net/ceph/osdmap.c+1−3

This diff is large: its 19 lines aren’t shown here.

net/core/dev.c+1−0

This diff is large: its 8 lines aren’t shown here.

net/core/net-sysfs.c+3−3

This diff is large: its 27 lines aren’t shown here.

net/core/skbuff.c+16−14

This diff is large: its 53 lines aren’t shown here.

net/core/sock.c+57−49

This diff is large: its 162 lines aren’t shown here.

net/dccp/ccids/ccid2.c+1−0

This diff is large: its 8 lines aren’t shown here.

net/dccp/ipv4.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/dccp/ipv6.c+5−3

This diff is large: its 16 lines aren’t shown here.

net/dccp/minisocks.c+16−8

This diff is large: its 59 lines aren’t shown here.

net/decnet/af_decnet.c+3−2

This diff is large: its 19 lines aren’t shown here.

net/ipv4/af_inet.c+6−3

This diff is large: its 27 lines aren’t shown here.

net/ipv4/inet_connection_sock.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/ipv4/ip_output.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/ipv4/tcp_ipv4.c+7−3

This diff is large: its 26 lines aren’t shown here.

net/ipv4/tcp_timer.c+4−2

This diff is large: its 20 lines aren’t shown here.

net/ipv6/af_inet6.c+5−5

This diff is large: its 29 lines aren’t shown here.

net/ipv6/ip6_fib.c+2−0

This diff is large: its 9 lines aren’t shown here.

net/ipv6/ip6_offload.c+3−1

This diff is large: its 12 lines aren’t shown here.

net/ipv6/ip6_output.c+7−2

This diff is large: its 36 lines aren’t shown here.

net/ipv6/ip6_vti.c+6−2

This diff is large: its 18 lines aren’t shown here.

net/ipv6/route.c+6−5

This diff is large: its 53 lines aren’t shown here.

net/ipv6/tcp_ipv6.c+5−3

This diff is large: its 16 lines aren’t shown here.

net/irda/af_irda.c+3−2

This diff is large: its 19 lines aren’t shown here.

net/iucv/af_iucv.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/llc/af_llc.c+3−1

This diff is large: its 15 lines aren’t shown here.

net/mpls/af_mpls.c+3−1

This diff is large: its 18 lines aren’t shown here.

net/netrom/af_netrom.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/nfc/llcp_sock.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/phonet/pep.c+4−2

This diff is large: its 20 lines aren’t shown here.

net/phonet/socket.c+2−2

This diff is large: its 18 lines aren’t shown here.

net/rds/connection.c+1−0

This diff is large: its 8 lines aren’t shown here.

net/rds/ib_cm.c+36−11

This diff is large: its 121 lines aren’t shown here.

net/rds/rds.h+3−3

This diff is large: its 25 lines aren’t shown here.

net/rds/tcp.c+21−17

This diff is large: its 92 lines aren’t shown here.

net/rds/tcp.h+1−1

This diff is large: its 9 lines aren’t shown here.

net/rds/tcp_listen.c+8−3

This diff is large: its 45 lines aren’t shown here.

net/rose/af_rose.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/rxrpc/input.c+19−8

This diff is large: its 81 lines aren’t shown here.

net/rxrpc/recvmsg.c+2−2

This diff is large: its 18 lines aren’t shown here.

net/rxrpc/sendmsg.c+31−18

This diff is large: its 91 lines aren’t shown here.

net/sched/act_connmark.c+3−0

This diff is large: its 10 lines aren’t shown here.

net/sched/act_skbmod.c+0−1

This diff is large: its 8 lines aren’t shown here.

net/sctp/ipv6.c+3−2

This diff is large: its 18 lines aren’t shown here.

net/sctp/protocol.c+3−2

This diff is large: its 14 lines aren’t shown here.

net/sctp/socket.c+2−2

This diff is large: its 18 lines aren’t shown here.

net/smc/af_smc.c+1−1

This diff is large: its 9 lines aren’t shown here.

net/socket.c+3−2

This diff is large: its 26 lines aren’t shown here.

net/tipc/socket.c+5−3

This diff is large: its 29 lines aren’t shown here.

net/unix/af_unix.c+3−2

This diff is large: its 19 lines aren’t shown here.

net/vmw_vsock/af_vsock.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/x25/af_x25.c+2−1

This diff is large: its 10 lines aren’t shown here.

net/xfrm/xfrm_policy.c+9−10

This diff is large: its 61 lines aren’t shown here.

scripts/gcc-plugins/sancov_plugin.c+1−1

This diff is large: its 9 lines aren’t shown here.

scripts/module-common.lds+2−0

This diff is large: its 7 lines aren’t shown here.

scripts/spelling.txt+3−0

This diff is large: its 17 lines aren’t shown here.

sound/soc/amd/acp-pcm-dma.c+1−1

This diff is large: its 9 lines aren’t shown here.

tools/include/uapi/linux/bpf_perf_event.h+18−0

This diff is large: its 19 lines aren’t shown here.

tools/lguest/lguest.c+1−1

This diff is large: its 9 lines aren’t shown here.

tools/lib/bpf/Makefile+1−1

This diff is large: its 9 lines aren’t shown here.

tools/lib/traceevent/Makefile+1−1

This diff is large: its 9 lines aren’t shown here.

tools/lib/traceevent/event-parse.h+1−1

This diff is large: its 9 lines aren’t shown here.

tools/objtool/builtin-check.c+12−3

This diff is large: its 24 lines aren’t shown here.

tools/objtool/elf.c+12−0

This diff is large: its 19 lines aren’t shown here.

tools/objtool/elf.h+1−0

This diff is large: its 8 lines aren’t shown here.

tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c+1−1

This diff is large: its 9 lines aren’t shown here.

tools/testing/ktest/ktest.pl+13−8

This diff is large: its 51 lines aren’t shown here.

tools/testing/radix-tree/Makefile+9−6

This diff is large: its 48 lines aren’t shown here.

tools/testing/radix-tree/benchmark.c+166−7

This diff is large: its 206 lines aren’t shown here.

tools/testing/radix-tree/idr-test.c+75−3

This diff is large: its 129 lines aren’t shown here.

tools/testing/radix-tree/main.c+1−0

This diff is large: its 8 lines aren’t shown here.

tools/testing/radix-tree/tag_check.c+29−0

This diff is large: its 41 lines aren’t shown here.

tools/testing/radix-tree/test.h+1−0

This diff is large: its 8 lines aren’t shown here.

tools/testing/selftests/bpf/Makefile+3−1

This diff is large: its 16 lines aren’t shown here.

tools/testing/selftests/bpf/test_verifier.c+4−0

This diff is large: its 22 lines aren’t shown here.

tools/testing/selftests/powerpc/harness.c+4−2

This diff is large: its 14 lines aren’t shown here.

tools/testing/selftests/powerpc/include/vsx_asm.h+24−24

This diff is large: its 81 lines aren’t shown here.

tools/testing/selftests/vm/Makefile+4−0

This diff is large: its 10 lines aren’t shown here.

tools/testing/selftests/x86/fsgsbase.c+1−1

This diff is large: its 9 lines aren’t shown here.

tools/testing/selftests/x86/ldt_gdt.c+11−5

This diff is large: its 58 lines aren’t shown here.

tools/testing/selftests/x86/ptrace_syscall.c+2−1

This diff is large: its 10 lines aren’t shown here.

tools/testing/selftests/x86/single_step_syscall.c+4−1

This diff is large: its 22 lines aren’t shown here.

virt/kvm/arm/vgic/vgic-its.c+65−44

This diff is large: its 143 lines aren’t shown here.

virt/kvm/arm/vgic/vgic-mmio.c+24−8

This diff is large: its 46 lines aren’t shown here.

virt/kvm/arm/vgic/vgic-v3.c+4−1

This diff is large: its 15 lines aren’t shown here.