Pull nfsd updates from Bruce Fields: "Highlights: - Add a new knfsd file cache, so that we don't have to open and close on each (NFSv2/v3) READ or WRITE. This can speed up read and write in some cases. It also replaces our readahead cache. - Prevent silent data loss on write errors, by treating write errors like server reboots for the purposes of write caching, thus forcing clients to resend their writes. - Tweak the code that allocates sessions to be more forgiving, so that NFSv4.1 mounts are less likely to hang when a server already has a lot of clients. - Eliminate an arbitrary limit on NFSv4 ACL sizes; they should now be limited only by the backend filesystem and the maximum RPC size. - Allow the server to enforce use of the correct kerberos credentials when a client reclaims state after a reboot. And some miscellaneous smaller bugfixes and cleanup" * tag 'nfsd-5.4' of git://linux-nfs.org/~bfields/linux: (34 commits) sunrpc: clean up indentation issue nfsd: fix nfs read eof detection nfsd: Make nfsd_reset_boot_verifier_locked static nfsd: degraded slot-count more gracefully as allocation nears exhaustion. nfsd: handle drc over-allocation gracefully. nfsd: add support for upcall version 2 nfsd: add a "GetVersion" upcall for nfsdcld nfsd: Reset the boot verifier on all write I/O errors nfsd: Don't garbage collect files that might contain write errors nfsd: Support the server resetting the boot verifier nfsd: nfsd_file cache entries should be per net namespace nfsd: eliminate an unnecessary acl size limit Deprecate nfsd fault injection nfsd: remove duplicated include from filecache.c nfsd: Fix the documentation for svcxdr_tmpalloc() nfsd: Fix up some unused variable warnings nfsd: close cached files prior to a REMOVE or RENAME that would replace target nfsd: rip out the raparms cache nfsd: have nfsd_test_lock use the nfsd_file cache nfsd: hook up nfs4_preprocess_stateid_op to the nfsd_file cache ...
Showing the changes to include (5 of 40 files). Show all
Unified · Split
| @@ -1168,6 +1168,11 @@ extern void lease_get_mtime(struct inode *, struct timespec64 *time); · 1167 unmodified lines | ||
| 1168 | 1168 | extern int generic_setlease(struct file *, long, struct file_lock **, void **priv); |
| 1169 | 1169 | extern int vfs_setlease(struct file *, long, struct file_lock **, void **); |
| 1170 | 1170 | extern int lease_modify(struct file_lock *, int, struct list_head *); |
| 1171 | + | |
| 1172 | +struct notifier_block; | |
| 1173 | +extern int lease_register_notifier(struct notifier_block *); | |
| 1174 | +extern void lease_unregister_notifier(struct notifier_block *); | |
| 1175 | + | |
| 1171 | 1176 | struct files_struct; |
| 1172 | 1177 | extern void show_fd_locks(struct seq_file *f, |
| 1173 | 1178 | struct file *filp, struct files_struct *files); |
| @@ -475,6 +475,8 @@ extern void fsnotify_destroy_mark(struct fsnotify_mark *mark, · 474 unmodified lines | ||
| 475 | 475 | extern void fsnotify_detach_mark(struct fsnotify_mark *mark); |
| 476 | 476 | /* free mark */ |
| 477 | 477 | extern void fsnotify_free_mark(struct fsnotify_mark *mark); |
| 478 | +/* Wait until all marks queued for destruction are destroyed */ | |
| 479 | +extern void fsnotify_wait_marks_destroyed(void); | |
| 478 | 480 | /* run all the marks in a group, and clear all of the marks attached to given object type */ |
| 479 | 481 | extern void fsnotify_clear_marks_by_group(struct fsnotify_group *group, unsigned int type); |
| 480 | 482 | /* run all the marks in a group, and clear all of the vfsmount marks */ |
| @@ -87,6 +87,7 @@ struct cache_detail { · 86 unmodified lines | ||
| 87 | 87 | int has_died); |
| 88 | 88 | |
| 89 | 89 | struct cache_head * (*alloc)(void); |
| 90 | + void (*flush)(void); | |
| 90 | 91 | int (*match)(struct cache_head *orig, struct cache_head *new); |
| 91 | 92 | void (*init)(struct cache_head *orig, struct cache_head *new); |
| 92 | 93 | void (*update)(struct cache_head *orig, struct cache_head *new); |
| @@ -107,9 +108,9 @@ struct cache_detail { · 14 unmodified lines | ||
| 107 | 108 | /* fields for communication over channel */ |
| 108 | 109 | struct list_head queue; |
| 109 | 110 | |
| 110 | − atomic_t readers; /* how many time is /chennel open */ | |
| 111 | − time_t last_close; /* if no readers, when did last close */ | |
| 112 | − time_t last_warn; /* when we last warned about no readers */ | |
| 111 | + atomic_t writers; /* how many time is /channel open */ | |
| 112 | + time_t last_close; /* if no writers, when did last close */ | |
| 113 | + time_t last_warn; /* when we last warned about no writers */ | |
| 113 | 114 | |
| 114 | 115 | union { |
| 115 | 116 | struct proc_dir_entry *procfs; |
| @@ -42,6 +42,7 @@ · 41 unmodified lines | ||
| 42 | 42 | |
| 43 | 43 | #ifndef SVC_RDMA_H |
| 44 | 44 | #define SVC_RDMA_H |
| 45 | +#include <linux/llist.h> | |
| 45 | 46 | #include <linux/sunrpc/xdr.h> |
| 46 | 47 | #include <linux/sunrpc/svcsock.h> |
| 47 | 48 | #include <linux/sunrpc/rpc_rdma.h> |
| @@ -107,8 +108,7 @@ struct svcxprt_rdma { · 59 unmodified lines | ||
| 107 | 108 | struct list_head sc_read_complete_q; |
| 108 | 109 | struct work_struct sc_work; |
| 109 | 110 | |
| 110 | − spinlock_t sc_recv_lock; | |
| 111 | − struct list_head sc_recv_ctxts; | |
| 111 | + struct llist_head sc_recv_ctxts; | |
| 112 | 112 | }; |
| 113 | 113 | /* sc_flags */ |
| 114 | 114 | #define RDMAXPRT_CONN_PENDING 3 |
| @@ -125,6 +125,7 @@ enum { · 10 unmodified lines | ||
| 125 | 125 | #define RPCSVC_MAXPAYLOAD_RDMA RPCSVC_MAXPAYLOAD |
| 126 | 126 | |
| 127 | 127 | struct svc_rdma_recv_ctxt { |
| 128 | + struct llist_node rc_node; | |
| 128 | 129 | struct list_head rc_list; |
| 129 | 130 | struct ib_recv_wr rc_recv_wr; |
| 130 | 131 | struct ib_cqe rc_cqe; |
| @@ -200,7 +201,6 @@ extern struct svc_xprt_class svc_rdma_bc_class; · 69 unmodified lines | ||
| 200 | 201 | #endif |
| 201 | 202 | |
| 202 | 203 | /* svc_rdma.c */ |
| 203 | −extern struct workqueue_struct *svc_rdma_wq; | |
| 204 | 204 | extern int svc_rdma_init(void); |
| 205 | 205 | extern void svc_rdma_cleanup(void); |
| 206 | 206 | |
| @@ -26,17 +26,22 @@ · 25 unmodified lines | ||
| 26 | 26 | #include <linux/types.h> |
| 27 | 27 | |
| 28 | 28 | /* latest upcall version available */ |
| 29 | −#define CLD_UPCALL_VERSION 1 | |
| 29 | +#define CLD_UPCALL_VERSION 2 | |
| 30 | 30 | |
| 31 | 31 | /* defined by RFC3530 */ |
| 32 | 32 | #define NFS4_OPAQUE_LIMIT 1024 |
| 33 | 33 | |
| 34 | +#ifndef SHA256_DIGEST_SIZE | |
| 35 | +#define SHA256_DIGEST_SIZE 32 | |
| 36 | +#endif | |
| 37 | + | |
| 34 | 38 | enum cld_command { |
| 35 | 39 | Cld_Create, /* create a record for this cm_id */ |
| 36 | 40 | Cld_Remove, /* remove record of this cm_id */ |
| 37 | 41 | Cld_Check, /* is this cm_id allowed? */ |
| 38 | 42 | Cld_GraceDone, /* grace period is complete */ |
| 39 | − Cld_GraceStart, | |
| 43 | + Cld_GraceStart, /* grace start (upload client records) */ | |
| 44 | + Cld_GetVersion, /* query max supported upcall version */ | |
| 40 | 45 | }; |
| 41 | 46 | |
| 42 | 47 | /* representation of long-form NFSv4 client ID */ |
| @@ -45,6 +50,17 @@ struct cld_name { · 2 unmodified lines | ||
| 45 | 50 | unsigned char cn_id[NFS4_OPAQUE_LIMIT]; /* client-provided */ |
| 46 | 51 | } __attribute__((packed)); |
| 47 | 52 | |
| 53 | +/* sha256 hash of the kerberos principal */ | |
| 54 | +struct cld_princhash { | |
| 55 | + __u8 cp_len; /* length of cp_data */ | |
| 56 | + unsigned char cp_data[SHA256_DIGEST_SIZE]; /* hash of principal */ | |
| 57 | +} __attribute__((packed)); | |
| 58 | + | |
| 59 | +struct cld_clntinfo { | |
| 60 | + struct cld_name cc_name; | |
| 61 | + struct cld_princhash cc_princhash; | |
| 62 | +} __attribute__((packed)); | |
| 63 | + | |
| 48 | 64 | /* message struct for communication with userspace */ |
| 49 | 65 | struct cld_msg { |
| 50 | 66 | __u8 cm_vers; /* upcall version */ |
| @@ -54,7 +70,28 @@ struct cld_msg { · 3 unmodified lines | ||
| 54 | 70 | union { |
| 55 | 71 | __s64 cm_gracetime; /* grace period start time */ |
| 56 | 72 | struct cld_name cm_name; |
| 73 | + __u8 cm_version; /* for getting max version */ | |
| 74 | + } __attribute__((packed)) cm_u; | |
| 75 | +} __attribute__((packed)); | |
| 76 | + | |
| 77 | +/* version 2 message can include hash of kerberos principal */ | |
| 78 | +struct cld_msg_v2 { | |
| 79 | + __u8 cm_vers; /* upcall version */ | |
| 80 | + __u8 cm_cmd; /* upcall command */ | |
| 81 | + __s16 cm_status; /* return code */ | |
| 82 | + __u32 cm_xid; /* transaction id */ | |
| 83 | + union { | |
| 84 | + struct cld_name cm_name; | |
| 85 | + __u8 cm_version; /* for getting max version */ | |
| 86 | + struct cld_clntinfo cm_clntinfo; /* name & princ hash */ | |
| 57 | 87 | } __attribute__((packed)) cm_u; |
| 58 | 88 | } __attribute__((packed)); |
| 59 | 89 | |
| 90 | +struct cld_msg_hdr { | |
| 91 | + __u8 cm_vers; /* upcall version */ | |
| 92 | + __u8 cm_cmd; /* upcall command */ | |
| 93 | + __s16 cm_status; /* return code */ | |
| 94 | + __u32 cm_xid; /* transaction id */ | |
| 95 | +} __attribute__((packed)); | |
| 96 | + | |
| 60 | 97 | #endif /* !_NFSD_CLD_H */ |