linux/fs/nfs/fscache.h

245 lines
6.9 KiB
C
Raw Normal View History

/* NFS filesystem cache interface definitions
*
* Copyright (C) 2008 Red Hat, Inc. All Rights Reserved.
* Written by David Howells (dhowells@redhat.com)
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public Licence
* as published by the Free Software Foundation; either version
* 2 of the Licence, or (at your option) any later version.
*/
#ifndef _NFS_FSCACHE_H
#define _NFS_FSCACHE_H
#include <linux/nfs_fs.h>
#include <linux/nfs_mount.h>
#include <linux/nfs4_mount.h>
#include <linux/fscache.h>
#ifdef CONFIG_NFS_FSCACHE
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 23:42:42 +08:00
/*
* set of NFS FS-Cache objects that form a superblock key
*/
struct nfs_fscache_key {
struct rb_node node;
struct nfs_client *nfs_client; /* the server */
/* the elements of the unique key - as used by nfs_compare_super() and
* nfs_compare_mount_options() to distinguish superblocks */
struct {
struct {
unsigned long s_flags; /* various flags
* (& NFS_MS_MASK) */
} super;
struct {
struct nfs_fsid fsid;
int flags;
unsigned int rsize; /* read size */
unsigned int wsize; /* write size */
unsigned int acregmin; /* attr cache timeouts */
unsigned int acregmax;
unsigned int acdirmin;
unsigned int acdirmax;
} nfs_server;
struct {
rpc_authflavor_t au_flavor;
} rpc_auth;
/* uniquifier - can be used if nfs_server.flags includes
* NFS_MOUNT_UNSHARED */
u8 uniq_len;
char uniquifier[0];
} key;
};
/*
* Definition of the auxiliary data attached to NFS inode storage objects
* within the cache.
*
* The contents of this struct are recorded in the on-disk local cache in the
* auxiliary data attached to the data storage object backing an inode. This
* permits coherency to be managed when a new inode binds to an already extant
* cache object.
*/
struct nfs_fscache_inode_auxdata {
struct timespec mtime;
struct timespec ctime;
u64 change_attr;
};
/*
* fscache-index.c
*/
extern struct fscache_netfs nfs_fscache_netfs;
extern const struct fscache_cookie_def nfs_fscache_server_index_def;
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 23:42:42 +08:00
extern const struct fscache_cookie_def nfs_fscache_super_index_def;
extern const struct fscache_cookie_def nfs_fscache_inode_object_def;
extern int nfs_fscache_register(void);
extern void nfs_fscache_unregister(void);
/*
* fscache.c
*/
extern void nfs_fscache_get_client_cookie(struct nfs_client *);
extern void nfs_fscache_release_client_cookie(struct nfs_client *);
extern void nfs_fscache_get_super_cookie(struct super_block *, const char *, int);
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 23:42:42 +08:00
extern void nfs_fscache_release_super_cookie(struct super_block *);
NFS: Use i_writecount to control whether to get an fscache cookie in nfs_open() Use i_writecount to control whether to get an fscache cookie in nfs_open() as NFS does not do write caching yet. I *think* this is the cause of a problem encountered by Mark Moseley whereby __fscache_uncache_page() gets a NULL pointer dereference because cookie->def is NULL: BUG: unable to handle kernel NULL pointer dereference at 0000000000000010 IP: [<ffffffff812a1903>] __fscache_uncache_page+0x23/0x160 PGD 0 Thread overran stack, or stack corrupted Oops: 0000 [#1] SMP Modules linked in: ... CPU: 7 PID: 18993 Comm: php Not tainted 3.11.1 #1 Hardware name: Dell Inc. PowerEdge R420/072XWF, BIOS 1.3.5 08/21/2012 task: ffff8804203460c0 ti: ffff880420346640 RIP: 0010:[<ffffffff812a1903>] __fscache_uncache_page+0x23/0x160 RSP: 0018:ffff8801053af878 EFLAGS: 00210286 RAX: 0000000000000000 RBX: ffff8800be2f8780 RCX: ffff88022ffae5e8 RDX: 0000000000004c66 RSI: ffffea00055ff440 RDI: ffff8800be2f8780 RBP: ffff8801053af898 R08: 0000000000000001 R09: 0000000000000003 R10: 0000000000000000 R11: 0000000000000000 R12: ffffea00055ff440 R13: 0000000000001000 R14: ffff8800c50be538 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffff88042fc60000(0063) knlGS:00000000e439c700 CS: 0010 DS: 002b ES: 002b CR0: 0000000080050033 CR2: 0000000000000010 CR3: 0000000001d8f000 CR4: 00000000000607f0 Stack: ... Call Trace: [<ffffffff81365a72>] __nfs_fscache_invalidate_page+0x42/0x70 [<ffffffff813553d5>] nfs_invalidate_page+0x75/0x90 [<ffffffff811b8f5e>] truncate_inode_page+0x8e/0x90 [<ffffffff811b90ad>] truncate_inode_pages_range.part.12+0x14d/0x620 [<ffffffff81d6387d>] ? __mutex_lock_slowpath+0x1fd/0x2e0 [<ffffffff811b95d3>] truncate_inode_pages_range+0x53/0x70 [<ffffffff811b969d>] truncate_inode_pages+0x2d/0x40 [<ffffffff811b96ff>] truncate_pagecache+0x4f/0x70 [<ffffffff81356840>] nfs_setattr_update_inode+0xa0/0x120 [<ffffffff81368de4>] nfs3_proc_setattr+0xc4/0xe0 [<ffffffff81357f78>] nfs_setattr+0xc8/0x150 [<ffffffff8122d95b>] notify_change+0x1cb/0x390 [<ffffffff8120a55b>] do_truncate+0x7b/0xc0 [<ffffffff8121f96c>] do_last+0xa4c/0xfd0 [<ffffffff8121ffbc>] path_openat+0xcc/0x670 [<ffffffff81220a0e>] do_filp_open+0x4e/0xb0 [<ffffffff8120ba1f>] do_sys_open+0x13f/0x2b0 [<ffffffff8126aaf6>] compat_SyS_open+0x36/0x50 [<ffffffff81d7204c>] sysenter_dispatch+0x7/0x24 The code at the instruction pointer was disassembled: > (gdb) disas __fscache_uncache_page > Dump of assembler code for function __fscache_uncache_page: > ... > 0xffffffff812a18ff <+31>: mov 0x48(%rbx),%rax > 0xffffffff812a1903 <+35>: cmpb $0x0,0x10(%rax) > 0xffffffff812a1907 <+39>: je 0xffffffff812a19cd <__fscache_uncache_page+237> These instructions make up: ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX); That cmpb is the faulting instruction (%rax is 0). So cookie->def is NULL - which presumably means that the cookie has already been at least partway through __fscache_relinquish_cookie(). What I think may be happening is something like a three-way race on the same file: PROCESS 1 PROCESS 2 PROCESS 3 =============== =============== =============== open(O_TRUNC|O_WRONLY) open(O_RDONLY) open(O_WRONLY) -->nfs_open() -->nfs_fscache_set_inode_cookie() nfs_fscache_inode_lock() nfs_fscache_disable_inode_cookie() __fscache_relinquish_cookie() nfs_inode->fscache = NULL <--nfs_fscache_set_inode_cookie() -->nfs_open() -->nfs_fscache_set_inode_cookie() nfs_fscache_inode_lock() nfs_fscache_enable_inode_cookie() __fscache_acquire_cookie() nfs_inode->fscache = cookie <--nfs_fscache_set_inode_cookie() <--nfs_open() -->nfs_setattr() ... ... -->nfs_invalidate_page() -->__nfs_fscache_invalidate_page() cookie = nfsi->fscache -->nfs_open() -->nfs_fscache_set_inode_cookie() nfs_fscache_inode_lock() nfs_fscache_disable_inode_cookie() -->__fscache_relinquish_cookie() -->__fscache_uncache_page(cookie) <crash> <--__fscache_relinquish_cookie() nfs_inode->fscache = NULL <--nfs_fscache_set_inode_cookie() What is needed is something to prevent process #2 from reacquiring the cookie - and I think checking i_writecount should do the trick. It's also possible to have a two-way race on this if the file is opened O_TRUNC|O_RDONLY instead. Reported-by: Mark Moseley <moseleymark@gmail.com> Signed-off-by: David Howells <dhowells@redhat.com>
2013-09-27 18:20:03 +08:00
extern void nfs_fscache_init_inode(struct inode *);
extern void nfs_fscache_clear_inode(struct inode *);
extern void nfs_fscache_open_file(struct inode *, struct file *);
extern void __nfs_fscache_invalidate_page(struct page *, struct inode *);
extern int nfs_fscache_release_page(struct page *, gfp_t);
extern int __nfs_readpage_from_fscache(struct nfs_open_context *,
struct inode *, struct page *);
extern int __nfs_readpages_from_fscache(struct nfs_open_context *,
struct inode *, struct address_space *,
struct list_head *, unsigned *);
extern void __nfs_readpage_to_fscache(struct inode *, struct page *, int);
/*
* wait for a page to complete writing to the cache
*/
static inline void nfs_fscache_wait_on_page_write(struct nfs_inode *nfsi,
struct page *page)
{
if (PageFsCache(page))
fscache_wait_on_page_write(nfsi->fscache, page);
}
/*
* release the caching state associated with a page if undergoing complete page
* invalidation
*/
static inline void nfs_fscache_invalidate_page(struct page *page,
struct inode *inode)
{
if (PageFsCache(page))
__nfs_fscache_invalidate_page(page, inode);
}
/*
* Retrieve a page from an inode data storage object.
*/
static inline int nfs_readpage_from_fscache(struct nfs_open_context *ctx,
struct inode *inode,
struct page *page)
{
if (NFS_I(inode)->fscache)
return __nfs_readpage_from_fscache(ctx, inode, page);
return -ENOBUFS;
}
/*
* Retrieve a set of pages from an inode data storage object.
*/
static inline int nfs_readpages_from_fscache(struct nfs_open_context *ctx,
struct inode *inode,
struct address_space *mapping,
struct list_head *pages,
unsigned *nr_pages)
{
if (NFS_I(inode)->fscache)
return __nfs_readpages_from_fscache(ctx, inode, mapping, pages,
nr_pages);
return -ENOBUFS;
}
/*
* Store a page newly fetched from the server in an inode data storage object
* in the cache.
*/
static inline void nfs_readpage_to_fscache(struct inode *inode,
struct page *page,
int sync)
{
if (PageFsCache(page))
__nfs_readpage_to_fscache(inode, page, sync);
}
NFS: Use FS-Cache invalidation Use the new FS-Cache invalidation facility from NFS to deal with foreign changes being detected on the server rather than attempting to retire the old cookie and get a new one. The problem with the old method was that NFS did not wait for all outstanding storage and retrieval ops on the cache to complete. There was no automatic wait between the calls to ->readpages() and calls to invalidate_inode_pages2() as the latter can only wait on locked pages that have been added to the pagecache (which they haven't yet on entry to ->readpages()). This was leading to oopses like the one below when an outstanding read got cut off from its cookie by a premature release. BUG: unable to handle kernel NULL pointer dereference at 00000000000000a8 IP: [<ffffffffa0075118>] __fscache_read_or_alloc_pages+0x1dd/0x315 [fscache] PGD 15889067 PUD 15890067 PMD 0 Oops: 0000 [#1] SMP CPU 0 Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc Pid: 4544, comm: tar Not tainted 3.1.0-rc4-fsdevel+ #1064 /DG965RY RIP: 0010:[<ffffffffa0075118>] [<ffffffffa0075118>] __fscache_read_or_alloc_pages+0x1dd/0x315 [fscache] RSP: 0018:ffff8800158799e8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8800070d41e0 RCX: ffff8800083dc1b0 RDX: 0000000000000000 RSI: ffff880015879960 RDI: ffff88003e627b90 RBP: ffff880015879a28 R08: 0000000000000002 R09: 0000000000000002 R10: 0000000000000001 R11: ffff880015879950 R12: ffff880015879aa4 R13: 0000000000000000 R14: ffff8800083dc158 R15: ffff880015879be8 FS: 00007f671e9d87c0(0000) GS:ffff88003bc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00000000000000a8 CR3: 000000001587f000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process tar (pid: 4544, threadinfo ffff880015878000, task ffff880015875040) Stack: ffffffffa00b1759 ffff8800070dc158 ffff8800000213da ffff88002a286508 ffff880015879aa4 ffff880015879be8 0000000000000001 ffff88002a2866e8 ffff880015879a88 ffffffffa00b20be 00000000000200da ffff880015875040 Call Trace: [<ffffffffa00b1759>] ? nfs_fscache_wait_bit+0xd/0xd [nfs] [<ffffffffa00b20be>] __nfs_readpages_from_fscache+0x7e/0x13f [nfs] [<ffffffff81095fe7>] ? __alloc_pages_nodemask+0x156/0x662 [<ffffffffa0098763>] nfs_readpages+0xee/0x187 [nfs] [<ffffffff81098a5e>] __do_page_cache_readahead+0x1be/0x267 [<ffffffff81098942>] ? __do_page_cache_readahead+0xa2/0x267 [<ffffffff81098d7b>] ra_submit+0x1c/0x20 [<ffffffff8109900a>] ondemand_readahead+0x28b/0x29a [<ffffffff810990ce>] page_cache_sync_readahead+0x38/0x3a [<ffffffff81091d8a>] generic_file_aio_read+0x2ab/0x67e [<ffffffffa008cfbe>] nfs_file_read+0xa4/0xc9 [nfs] [<ffffffff810c22c4>] do_sync_read+0xba/0xfa [<ffffffff810a62c9>] ? might_fault+0x4e/0x9e [<ffffffff81177a47>] ? security_file_permission+0x7b/0x84 [<ffffffff810c25dd>] ? rw_verify_area+0xab/0xc8 [<ffffffff810c29a4>] vfs_read+0xaa/0x13a [<ffffffff810c2a79>] sys_read+0x45/0x6c [<ffffffff813ac37b>] system_call_fastpath+0x16/0x1b Reported-by: Mark Moseley <moseleymark@gmail.com> Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 05:52:38 +08:00
/*
* Invalidate the contents of fscache for this inode. This will not sleep.
*/
static inline void nfs_fscache_invalidate(struct inode *inode)
{
fscache_invalidate(NFS_I(inode)->fscache);
}
/*
* Wait for an object to finish being invalidated.
*/
static inline void nfs_fscache_wait_on_invalidate(struct inode *inode)
{
fscache_wait_on_invalidate(NFS_I(inode)->fscache);
}
/*
* indicate the client caching state as readable text
*/
static inline const char *nfs_server_fscache_state(struct nfs_server *server)
{
if (server->fscache && (server->options & NFS_OPTION_FSCACHE))
return "yes";
return "no ";
}
#else /* CONFIG_NFS_FSCACHE */
static inline int nfs_fscache_register(void) { return 0; }
static inline void nfs_fscache_unregister(void) {}
static inline void nfs_fscache_get_client_cookie(struct nfs_client *clp) {}
static inline void nfs_fscache_release_client_cookie(struct nfs_client *clp) {}
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 23:42:42 +08:00
static inline void nfs_fscache_release_super_cookie(struct super_block *sb) {}
NFS: Use i_writecount to control whether to get an fscache cookie in nfs_open() Use i_writecount to control whether to get an fscache cookie in nfs_open() as NFS does not do write caching yet. I *think* this is the cause of a problem encountered by Mark Moseley whereby __fscache_uncache_page() gets a NULL pointer dereference because cookie->def is NULL: BUG: unable to handle kernel NULL pointer dereference at 0000000000000010 IP: [<ffffffff812a1903>] __fscache_uncache_page+0x23/0x160 PGD 0 Thread overran stack, or stack corrupted Oops: 0000 [#1] SMP Modules linked in: ... CPU: 7 PID: 18993 Comm: php Not tainted 3.11.1 #1 Hardware name: Dell Inc. PowerEdge R420/072XWF, BIOS 1.3.5 08/21/2012 task: ffff8804203460c0 ti: ffff880420346640 RIP: 0010:[<ffffffff812a1903>] __fscache_uncache_page+0x23/0x160 RSP: 0018:ffff8801053af878 EFLAGS: 00210286 RAX: 0000000000000000 RBX: ffff8800be2f8780 RCX: ffff88022ffae5e8 RDX: 0000000000004c66 RSI: ffffea00055ff440 RDI: ffff8800be2f8780 RBP: ffff8801053af898 R08: 0000000000000001 R09: 0000000000000003 R10: 0000000000000000 R11: 0000000000000000 R12: ffffea00055ff440 R13: 0000000000001000 R14: ffff8800c50be538 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffff88042fc60000(0063) knlGS:00000000e439c700 CS: 0010 DS: 002b ES: 002b CR0: 0000000080050033 CR2: 0000000000000010 CR3: 0000000001d8f000 CR4: 00000000000607f0 Stack: ... Call Trace: [<ffffffff81365a72>] __nfs_fscache_invalidate_page+0x42/0x70 [<ffffffff813553d5>] nfs_invalidate_page+0x75/0x90 [<ffffffff811b8f5e>] truncate_inode_page+0x8e/0x90 [<ffffffff811b90ad>] truncate_inode_pages_range.part.12+0x14d/0x620 [<ffffffff81d6387d>] ? __mutex_lock_slowpath+0x1fd/0x2e0 [<ffffffff811b95d3>] truncate_inode_pages_range+0x53/0x70 [<ffffffff811b969d>] truncate_inode_pages+0x2d/0x40 [<ffffffff811b96ff>] truncate_pagecache+0x4f/0x70 [<ffffffff81356840>] nfs_setattr_update_inode+0xa0/0x120 [<ffffffff81368de4>] nfs3_proc_setattr+0xc4/0xe0 [<ffffffff81357f78>] nfs_setattr+0xc8/0x150 [<ffffffff8122d95b>] notify_change+0x1cb/0x390 [<ffffffff8120a55b>] do_truncate+0x7b/0xc0 [<ffffffff8121f96c>] do_last+0xa4c/0xfd0 [<ffffffff8121ffbc>] path_openat+0xcc/0x670 [<ffffffff81220a0e>] do_filp_open+0x4e/0xb0 [<ffffffff8120ba1f>] do_sys_open+0x13f/0x2b0 [<ffffffff8126aaf6>] compat_SyS_open+0x36/0x50 [<ffffffff81d7204c>] sysenter_dispatch+0x7/0x24 The code at the instruction pointer was disassembled: > (gdb) disas __fscache_uncache_page > Dump of assembler code for function __fscache_uncache_page: > ... > 0xffffffff812a18ff <+31>: mov 0x48(%rbx),%rax > 0xffffffff812a1903 <+35>: cmpb $0x0,0x10(%rax) > 0xffffffff812a1907 <+39>: je 0xffffffff812a19cd <__fscache_uncache_page+237> These instructions make up: ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX); That cmpb is the faulting instruction (%rax is 0). So cookie->def is NULL - which presumably means that the cookie has already been at least partway through __fscache_relinquish_cookie(). What I think may be happening is something like a three-way race on the same file: PROCESS 1 PROCESS 2 PROCESS 3 =============== =============== =============== open(O_TRUNC|O_WRONLY) open(O_RDONLY) open(O_WRONLY) -->nfs_open() -->nfs_fscache_set_inode_cookie() nfs_fscache_inode_lock() nfs_fscache_disable_inode_cookie() __fscache_relinquish_cookie() nfs_inode->fscache = NULL <--nfs_fscache_set_inode_cookie() -->nfs_open() -->nfs_fscache_set_inode_cookie() nfs_fscache_inode_lock() nfs_fscache_enable_inode_cookie() __fscache_acquire_cookie() nfs_inode->fscache = cookie <--nfs_fscache_set_inode_cookie() <--nfs_open() -->nfs_setattr() ... ... -->nfs_invalidate_page() -->__nfs_fscache_invalidate_page() cookie = nfsi->fscache -->nfs_open() -->nfs_fscache_set_inode_cookie() nfs_fscache_inode_lock() nfs_fscache_disable_inode_cookie() -->__fscache_relinquish_cookie() -->__fscache_uncache_page(cookie) <crash> <--__fscache_relinquish_cookie() nfs_inode->fscache = NULL <--nfs_fscache_set_inode_cookie() What is needed is something to prevent process #2 from reacquiring the cookie - and I think checking i_writecount should do the trick. It's also possible to have a two-way race on this if the file is opened O_TRUNC|O_RDONLY instead. Reported-by: Mark Moseley <moseleymark@gmail.com> Signed-off-by: David Howells <dhowells@redhat.com>
2013-09-27 18:20:03 +08:00
static inline void nfs_fscache_init_inode(struct inode *inode) {}
static inline void nfs_fscache_clear_inode(struct inode *inode) {}
static inline void nfs_fscache_open_file(struct inode *inode,
struct file *filp) {}
static inline int nfs_fscache_release_page(struct page *page, gfp_t gfp)
{
return 1; /* True: may release page */
}
static inline void nfs_fscache_invalidate_page(struct page *page,
struct inode *inode) {}
static inline void nfs_fscache_wait_on_page_write(struct nfs_inode *nfsi,
struct page *page) {}
static inline int nfs_readpage_from_fscache(struct nfs_open_context *ctx,
struct inode *inode,
struct page *page)
{
return -ENOBUFS;
}
static inline int nfs_readpages_from_fscache(struct nfs_open_context *ctx,
struct inode *inode,
struct address_space *mapping,
struct list_head *pages,
unsigned *nr_pages)
{
return -ENOBUFS;
}
static inline void nfs_readpage_to_fscache(struct inode *inode,
struct page *page, int sync) {}
NFS: Use FS-Cache invalidation Use the new FS-Cache invalidation facility from NFS to deal with foreign changes being detected on the server rather than attempting to retire the old cookie and get a new one. The problem with the old method was that NFS did not wait for all outstanding storage and retrieval ops on the cache to complete. There was no automatic wait between the calls to ->readpages() and calls to invalidate_inode_pages2() as the latter can only wait on locked pages that have been added to the pagecache (which they haven't yet on entry to ->readpages()). This was leading to oopses like the one below when an outstanding read got cut off from its cookie by a premature release. BUG: unable to handle kernel NULL pointer dereference at 00000000000000a8 IP: [<ffffffffa0075118>] __fscache_read_or_alloc_pages+0x1dd/0x315 [fscache] PGD 15889067 PUD 15890067 PMD 0 Oops: 0000 [#1] SMP CPU 0 Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc Pid: 4544, comm: tar Not tainted 3.1.0-rc4-fsdevel+ #1064 /DG965RY RIP: 0010:[<ffffffffa0075118>] [<ffffffffa0075118>] __fscache_read_or_alloc_pages+0x1dd/0x315 [fscache] RSP: 0018:ffff8800158799e8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8800070d41e0 RCX: ffff8800083dc1b0 RDX: 0000000000000000 RSI: ffff880015879960 RDI: ffff88003e627b90 RBP: ffff880015879a28 R08: 0000000000000002 R09: 0000000000000002 R10: 0000000000000001 R11: ffff880015879950 R12: ffff880015879aa4 R13: 0000000000000000 R14: ffff8800083dc158 R15: ffff880015879be8 FS: 00007f671e9d87c0(0000) GS:ffff88003bc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00000000000000a8 CR3: 000000001587f000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process tar (pid: 4544, threadinfo ffff880015878000, task ffff880015875040) Stack: ffffffffa00b1759 ffff8800070dc158 ffff8800000213da ffff88002a286508 ffff880015879aa4 ffff880015879be8 0000000000000001 ffff88002a2866e8 ffff880015879a88 ffffffffa00b20be 00000000000200da ffff880015875040 Call Trace: [<ffffffffa00b1759>] ? nfs_fscache_wait_bit+0xd/0xd [nfs] [<ffffffffa00b20be>] __nfs_readpages_from_fscache+0x7e/0x13f [nfs] [<ffffffff81095fe7>] ? __alloc_pages_nodemask+0x156/0x662 [<ffffffffa0098763>] nfs_readpages+0xee/0x187 [nfs] [<ffffffff81098a5e>] __do_page_cache_readahead+0x1be/0x267 [<ffffffff81098942>] ? __do_page_cache_readahead+0xa2/0x267 [<ffffffff81098d7b>] ra_submit+0x1c/0x20 [<ffffffff8109900a>] ondemand_readahead+0x28b/0x29a [<ffffffff810990ce>] page_cache_sync_readahead+0x38/0x3a [<ffffffff81091d8a>] generic_file_aio_read+0x2ab/0x67e [<ffffffffa008cfbe>] nfs_file_read+0xa4/0xc9 [nfs] [<ffffffff810c22c4>] do_sync_read+0xba/0xfa [<ffffffff810a62c9>] ? might_fault+0x4e/0x9e [<ffffffff81177a47>] ? security_file_permission+0x7b/0x84 [<ffffffff810c25dd>] ? rw_verify_area+0xab/0xc8 [<ffffffff810c29a4>] vfs_read+0xaa/0x13a [<ffffffff810c2a79>] sys_read+0x45/0x6c [<ffffffff813ac37b>] system_call_fastpath+0x16/0x1b Reported-by: Mark Moseley <moseleymark@gmail.com> Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 05:52:38 +08:00
static inline void nfs_fscache_invalidate(struct inode *inode) {}
static inline void nfs_fscache_wait_on_invalidate(struct inode *inode) {}
NFS: Use FS-Cache invalidation Use the new FS-Cache invalidation facility from NFS to deal with foreign changes being detected on the server rather than attempting to retire the old cookie and get a new one. The problem with the old method was that NFS did not wait for all outstanding storage and retrieval ops on the cache to complete. There was no automatic wait between the calls to ->readpages() and calls to invalidate_inode_pages2() as the latter can only wait on locked pages that have been added to the pagecache (which they haven't yet on entry to ->readpages()). This was leading to oopses like the one below when an outstanding read got cut off from its cookie by a premature release. BUG: unable to handle kernel NULL pointer dereference at 00000000000000a8 IP: [<ffffffffa0075118>] __fscache_read_or_alloc_pages+0x1dd/0x315 [fscache] PGD 15889067 PUD 15890067 PMD 0 Oops: 0000 [#1] SMP CPU 0 Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc Pid: 4544, comm: tar Not tainted 3.1.0-rc4-fsdevel+ #1064 /DG965RY RIP: 0010:[<ffffffffa0075118>] [<ffffffffa0075118>] __fscache_read_or_alloc_pages+0x1dd/0x315 [fscache] RSP: 0018:ffff8800158799e8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff8800070d41e0 RCX: ffff8800083dc1b0 RDX: 0000000000000000 RSI: ffff880015879960 RDI: ffff88003e627b90 RBP: ffff880015879a28 R08: 0000000000000002 R09: 0000000000000002 R10: 0000000000000001 R11: ffff880015879950 R12: ffff880015879aa4 R13: 0000000000000000 R14: ffff8800083dc158 R15: ffff880015879be8 FS: 00007f671e9d87c0(0000) GS:ffff88003bc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00000000000000a8 CR3: 000000001587f000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process tar (pid: 4544, threadinfo ffff880015878000, task ffff880015875040) Stack: ffffffffa00b1759 ffff8800070dc158 ffff8800000213da ffff88002a286508 ffff880015879aa4 ffff880015879be8 0000000000000001 ffff88002a2866e8 ffff880015879a88 ffffffffa00b20be 00000000000200da ffff880015875040 Call Trace: [<ffffffffa00b1759>] ? nfs_fscache_wait_bit+0xd/0xd [nfs] [<ffffffffa00b20be>] __nfs_readpages_from_fscache+0x7e/0x13f [nfs] [<ffffffff81095fe7>] ? __alloc_pages_nodemask+0x156/0x662 [<ffffffffa0098763>] nfs_readpages+0xee/0x187 [nfs] [<ffffffff81098a5e>] __do_page_cache_readahead+0x1be/0x267 [<ffffffff81098942>] ? __do_page_cache_readahead+0xa2/0x267 [<ffffffff81098d7b>] ra_submit+0x1c/0x20 [<ffffffff8109900a>] ondemand_readahead+0x28b/0x29a [<ffffffff810990ce>] page_cache_sync_readahead+0x38/0x3a [<ffffffff81091d8a>] generic_file_aio_read+0x2ab/0x67e [<ffffffffa008cfbe>] nfs_file_read+0xa4/0xc9 [nfs] [<ffffffff810c22c4>] do_sync_read+0xba/0xfa [<ffffffff810a62c9>] ? might_fault+0x4e/0x9e [<ffffffff81177a47>] ? security_file_permission+0x7b/0x84 [<ffffffff810c25dd>] ? rw_verify_area+0xab/0xc8 [<ffffffff810c29a4>] vfs_read+0xaa/0x13a [<ffffffff810c2a79>] sys_read+0x45/0x6c [<ffffffff813ac37b>] system_call_fastpath+0x16/0x1b Reported-by: Mark Moseley <moseleymark@gmail.com> Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 05:52:38 +08:00
static inline const char *nfs_server_fscache_state(struct nfs_server *server)
{
return "no ";
}
#endif /* CONFIG_NFS_FSCACHE */
#endif /* _NFS_FSCACHE_H */