diff options
author | Eric Dumazet <eric.dumazet@gmail.com> | 2010-10-23 19:18:01 +0400 |
---|---|---|
committer | Al Viro <viro@zeniv.linux.org.uk> | 2010-10-26 05:26:11 +0400 |
commit | f991bd2e14210fb93d722cb23e54991de20e8a3d (patch) | |
tree | 60643fb9d828c388d963ceb26504dce9890eabbb /fs/inode.c | |
parent | 7de9c6ee3ecffd99e1628e81a5ea5468f7581a1f (diff) | |
download | linux-f991bd2e14210fb93d722cb23e54991de20e8a3d.tar.xz |
fs: introduce a per-cpu last_ino allocator
new_inode() dirties a contended cache line to get increasing
inode numbers. This limits performance on workloads that cause
significant parallel inode allocation.
Solve this problem by using a per_cpu variable fed by the shared
last_ino in batches of 1024 allocations. This reduces contention on
the shared last_ino, and give same spreading ino numbers than before
(i.e. same wraparound after 2^32 allocations).
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Diffstat (limited to 'fs/inode.c')
-rw-r--r-- | fs/inode.c | 45 |
1 files changed, 38 insertions, 7 deletions
diff --git a/fs/inode.c b/fs/inode.c index 05ea293d5f32..46a3e120b196 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -717,6 +717,43 @@ repeat: return NULL; } +/* + * Each cpu owns a range of LAST_INO_BATCH numbers. + * 'shared_last_ino' is dirtied only once out of LAST_INO_BATCH allocations, + * to renew the exhausted range. + * + * This does not significantly increase overflow rate because every CPU can + * consume at most LAST_INO_BATCH-1 unused inode numbers. So there is + * NR_CPUS*(LAST_INO_BATCH-1) wastage. At 4096 and 1024, this is ~0.1% of the + * 2^32 range, and is a worst-case. Even a 50% wastage would only increase + * overflow rate by 2x, which does not seem too significant. + * + * On a 32bit, non LFS stat() call, glibc will generate an EOVERFLOW + * error if st_ino won't fit in target struct field. Use 32bit counter + * here to attempt to avoid that. + */ +#define LAST_INO_BATCH 1024 +static DEFINE_PER_CPU(unsigned int, last_ino); + +static unsigned int get_next_ino(void) +{ + unsigned int *p = &get_cpu_var(last_ino); + unsigned int res = *p; + +#ifdef CONFIG_SMP + if (unlikely((res & (LAST_INO_BATCH-1)) == 0)) { + static atomic_t shared_last_ino; + int next = atomic_add_return(LAST_INO_BATCH, &shared_last_ino); + + res = next - LAST_INO_BATCH; + } +#endif + + *p = ++res; + put_cpu_var(last_ino); + return res; +} + /** * new_inode - obtain an inode * @sb: superblock @@ -731,12 +768,6 @@ repeat: */ struct inode *new_inode(struct super_block *sb) { - /* - * On a 32bit, non LFS stat() call, glibc will generate an EOVERFLOW - * error if st_ino won't fit in target struct field. Use 32bit counter - * here to attempt to avoid that. - */ - static unsigned int last_ino; struct inode *inode; spin_lock_prefetch(&inode_lock); @@ -745,7 +776,7 @@ struct inode *new_inode(struct super_block *sb) if (inode) { spin_lock(&inode_lock); __inode_sb_list_add(inode); - inode->i_ino = ++last_ino; + inode->i_ino = get_next_ino(); inode->i_state = 0; spin_unlock(&inode_lock); } |