Revert bd_mount_mutex back to a semaphore so that xfs_freeze -f /mnt/newtest; xfs_freeze -u /mnt/newtest works safely and doesn't produce lockdep warnings. (XFS unlocks the semaphore from a different task, by design. The mutex code warns about this) Signed-off-by: Dave Chinner <dgc@sgi.com> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
NFS: Fix race in nfs_release_page()
invalidate_inode_pages2() may find the dirty bit has been set on a page
owing to the fact that the page may still be mapped after it was locked.
Only after the call to unmap_mapping_range() are we sure that the page
can no longer be dirtied.
In order to fix this, NFS has hooked the releasepage() method and tries
to write the page out between the call to unmap_mapping_range() and the
call to remove_mapping(). This, however leads to deadlocks in the page
reclaim code, where the page may be locked without holding a reference
to the inode or dentry.
Fix is to add a new address_space_operation, launder_page(), which will
attempt to write out a dirty page without releasing the page lock.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Also, the bare SetPageDirty() can skew all sort of accounting leading to
other nasties.
[akpm@osdl.org: cleanup]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add "relatime" (relative atime) support. Relative atime only updates the atime if the previous atime is older than the mtime or ctime. Like noatime, but useful for applications like mutt that need to know when a file has been read since it was last modified. A corresponding patch against mount(8) is available at http://userweb.kernel.org/~akpm/mount-relative-atime.txt Signed-off-by: Valerie Henson <val_henson@linux.intel.com> Cc: Mark Fasheh <mark.fasheh@oracle.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Christoph Hellwig <hch@lst.de> Cc: Karel Zak <kzak@redhat.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch changes struct file to use struct path instead of having
independent pointers to struct dentry and struct vfsmount, and converts all
users of f_{dentry,vfsmnt} in fs/ to use f_path.{dentry,mnt}.
Additionally, it adds two #define's to make the transition easier for users of
the f_dentry and f_vfsmnt.
Signed-off-by: Josef "Jeff" Sipek <jsipek@cs.sunysb.edu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Remove the old complex and crufty bd_mutex annotation. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Neil Brown <neilb@cse.unsw.edu.au> Cc: Ingo Molnar <mingo@elte.hu> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: Jason Baron <jbaron@redhat.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
[acme@newtoy net-2.6.20]$ pahole --cacheline 64 fs/inode.o inode
/* /pub/scm/linux/kernel/git/acme/net-2.6.20/include/linux/dcache.h:86 */
struct inode {
struct hlist_node i_hash; /* 0 8 */
struct list_head i_list; /* 8 8 */
struct list_head i_sb_list; /* 16 8 */
struct list_head i_dentry; /* 24 8 */
long unsigned int i_ino; /* 32 4 */
atomic_t i_count; /* 36 4 */
umode_t i_mode; /* 40 2 */
/* XXX 2 bytes hole, try to pack */
unsigned int i_nlink; /* 44 4 */
uid_t i_uid; /* 48 4 */
gid_t i_gid; /* 52 4 */
dev_t i_rdev; /* 56 4 */
loff_t i_size; /* 60 8 */
struct timespec i_atime; /* 68 8 */
struct timespec i_mtime; /* 76 8 */
struct timespec i_ctime; /* 84 8 */
unsigned int i_blkbits; /* 92 4 */
long unsigned int i_version; /* 96 4 */
blkcnt_t i_blocks; /* 100 4 */
short unsigned int i_bytes; /* 104 2 */
/* XXX 2 bytes hole, try to pack */
spinlock_t i_lock; /* 108 40 */
struct mutex i_mutex; /* 148 76 */
struct rw_semaphore i_alloc_sem; /* 224 64 */
struct inode_operations * i_op; /* 288 4 */
const struct file_operations * i_fop; /* 292 4 */
struct super_block * i_sb; /* 296 4 */
struct file_lock * i_flock; /* 300 4 */
struct address_space * i_mapping; /* 304 4 */
struct address_space i_data; /* 308 188 */
struct list_head i_devices; /* 496 8 */
union ; /* 504 4 */
int i_cindex; /* 508 4 */
__u32 i_generation; /* 512 4 */
/* ---------- cacheline 8 boundary ---------- */
long unsigned int i_dnotify_mask; /* 516 4 */
struct dnotify_struct * i_dnotify; /* 520 4 */
struct list_head inotify_watches; /* 524 8 */
struct mutex inotify_mutex; /* 532 76 */
long unsigned int i_state; /* 608 4 */
long unsigned int dirtied_when; /* 612 4 */
unsigned int i_flags; /* 616 4 */
atomic_t i_writecount; /* 620 4 */
void * i_security; /* 624 4 */
void * i_private; /* 628 4 */
}; /* size: 632, sum members: 628, holes: 2, sum holes: 4 */
[acme@newtoy net-2.6.20]$
So just moving i_mode to after i_bytes we save 4 bytes by nuking both holes:
[acme@newtoy net-2.6.20]$ codiff -V /tmp/inode.o.before fs/inode.o
/pub/scm/linux/kernel/git/acme/net-2.6.20/fs/inode.c:
struct inode | -4
i_mode;
from: umode_t /* 40(0) 2(0) */
to: umode_t /* 102(0) 2(0) */
1 struct changed
[acme@newtoy net-2.6.20]$
I've prunned all the other offset changes, only this one is of interest here.
So now we have:
[acme@newtoy net-2.6.20]$ pahole --cacheline 64 ../OUTPUT/qemu/net-2.6.20/fs/inode.o inode
/* /pub/scm/linux/kernel/git/acme/net-2.6.20/include/linux/dcache.h:86 */
struct inode {
struct hlist_node i_hash; /* 0 8 */
struct list_head i_list; /* 8 8 */
struct list_head i_sb_list; /* 16 8 */
struct list_head i_dentry; /* 24 8 */
long unsigned int i_ino; /* 32 4 */
atomic_t i_count; /* 36 4 */
unsigned int i_nlink; /* 40 4 */
uid_t i_uid; /* 44 4 */
gid_t i_gid; /* 48 4 */
dev_t i_rdev; /* 52 4 */
loff_t i_size; /* 56 8 */
/* ---------- cacheline 1 boundary ---------- */
struct timespec i_atime; /* 64 8 */
struct timespec i_mtime; /* 72 8 */
struct timespec i_ctime; /* 80 8 */
unsigned int i_blkbits; /* 88 4 */
long unsigned int i_version; /* 92 4 */
blkcnt_t i_blocks; /* 96 4 */
short unsigned int i_bytes; /* 100 2 */
umode_t i_mode; /* 102 2 */
spinlock_t i_lock; /* 104 40 */
struct mutex i_mutex; /* 144 76 */
struct rw_semaphore i_alloc_sem; /* 220 64 */
struct inode_operations * i_op; /* 284 4 */
const struct file_operations * i_fop; /* 288 4 */
struct super_block * i_sb; /* 292 4 */
struct file_lock * i_flock; /* 296 4 */
struct address_space * i_mapping; /* 300 4 */
struct address_space i_data; /* 304 188 */
struct list_head i_devices; /* 492 8 */
union ; /* 500 4 */
int i_cindex; /* 504 4 */
__u32 i_generation; /* 508 4 */
/* ---------- cacheline 8 boundary ---------- */
long unsigned int i_dnotify_mask; /* 512 4 */
struct dnotify_struct * i_dnotify; /* 516 4 */
struct list_head inotify_watches; /* 520 8 */
struct mutex inotify_mutex; /* 528 76 */
long unsigned int i_state; /* 604 4 */
long unsigned int dirtied_when; /* 608 4 */
unsigned int i_flags; /* 612 4 */
atomic_t i_writecount; /* 616 4 */
void * i_security; /* 620 4 */
void * i_private; /* 624 4 */
}; /* size: 628 */
[acme@newtoy net-2.6.20]$
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
On 32bits SMP platforms, 64bits i_size is protected by a seqcount (i_size_seqcount). When i_size is read or written, i_size_seqcount is read/written as well, so it make sense to group these two fields together in the same cache line. This patch moves i_size_seqcount next to i_size, and also moves i_version to let offsetof(struct inode, i_size) being 0x40 instead of 0x3c (for 32bits platforms). For 64 bits platforms, i_size_seqcount doesnt exist, and the move of a 'long i_version' should not introduce a new hole because of padding. Signed-off-by: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Change the signature of i_size_read(), IMINOR() and IMAJOR() because they, or the functions they call, will never modify the argument. Signed-off-by: Jan Engelhardt <jengelh@gmx.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
SLAB_KERNEL is an alias of GFP_KERNEL. Signed-off-by: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The names_cachep is used for getname() and putname(). So lets put it into fs.h near those two definitions. Signed-off-by: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>