Pull perf updates from Thomas Gleixner: "Core: - Better handling of page table leaves on archictectures which have architectures have non-pagetable aligned huge/large pages. For such architectures a leaf can actually be part of a larger entry. - Prevent a deadlock vs exec_update_mutex Architectures: - The related updates for page size calculation of leaf entries - The usual churn to support new CPUs - Small fixes and improvements all over the place" * tag 'perf-core-2020-12-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (24 commits) perf/x86/intel: Add Tremont Topdown support uprobes/x86: Fix fall-through warnings for Clang perf/x86: Fix fall-through warnings for Clang kprobes/x86: Fix fall-through warnings for Clang perf/x86/intel/lbr: Fix the return type of get_lbr_cycles() perf/x86/intel: Fix rtm_abort_event encoding on Ice Lake x86/kprobes: Restore BTF if the single-stepping is cancelled perf: Break deadlock involving exec_update_mutex sparc64/mm: Implement pXX_leaf_size() support powerpc/8xx: Implement pXX_leaf_size() support arm64/mm: Implement pXX_leaf_size() support perf/core: Fix arch_perf_get_page_size() mm: Introduce pXX_leaf_size() mm/gup: Provide gup_get_pte() more generic perf/x86/intel: Add event constraint for CYCLE_ACTIVITY.STALLS_MEM_ANY perf/x86/intel/uncore: Add Rocket Lake support perf/x86/msr: Add Rocket Lake CPU support perf/x86/cstate: Add Rocket Lake CPU support perf/x86/intel: Add Rocket Lake CPU support perf,mm: Handle non-page-table-aligned hugetlbfs ...
Showing the changes to include/uapi/linux (1 of 21 files). Show all
Unified · Split
| @@ -143,8 +143,10 @@ enum perf_event_sample_format { · 142 unmodified lines | |||
| 143 | PERF_SAMPLE_PHYS_ADDR = 1U << 19, | 143 | PERF_SAMPLE_PHYS_ADDR = 1U << 19, |
| 144 | PERF_SAMPLE_AUX = 1U << 20, | 144 | PERF_SAMPLE_AUX = 1U << 20, |
| 145 | PERF_SAMPLE_CGROUP = 1U << 21, | 145 | PERF_SAMPLE_CGROUP = 1U << 21, |
| 146 | PERF_SAMPLE_DATA_PAGE_SIZE = 1U << 22, | ||
| 147 | PERF_SAMPLE_CODE_PAGE_SIZE = 1U << 23, | ||
| 146 | 148 | ||
| 147 | PERF_SAMPLE_MAX = 1U << 22, /* non-ABI */ | 149 | PERF_SAMPLE_MAX = 1U << 24, /* non-ABI */ |
| 148 | 150 | ||
| 149 | __PERF_SAMPLE_CALLCHAIN_EARLY = 1ULL << 63, /* non-ABI; internal use */ | 151 | __PERF_SAMPLE_CALLCHAIN_EARLY = 1ULL << 63, /* non-ABI; internal use */ |
| 150 | }; | 152 | }; |
| @@ -896,6 +898,8 @@ enum perf_event_type { · 745 unmodified lines | |||
| 896 | * { u64 phys_addr;} && PERF_SAMPLE_PHYS_ADDR | 898 | * { u64 phys_addr;} && PERF_SAMPLE_PHYS_ADDR |
| 897 | * { u64 size; | 899 | * { u64 size; |
| 898 | * char data[size]; } && PERF_SAMPLE_AUX | 900 | * char data[size]; } && PERF_SAMPLE_AUX |
| 901 | * { u64 data_page_size;} && PERF_SAMPLE_DATA_PAGE_SIZE | ||
| 902 | * { u64 code_page_size;} && PERF_SAMPLE_CODE_PAGE_SIZE | ||
| 899 | * }; | 903 | * }; |
| 900 | */ | 904 | */ |
| 901 | PERF_RECORD_SAMPLE = 9, | 905 | PERF_RECORD_SAMPLE = 9, |