ext4: allocate ->s_blockgroup_lock separately
authorPekka Enberg <penberg@cs.helsinki.fi>
Sun, 15 Feb 2009 23:07:52 +0000 (18:07 -0500)
committerTheodore Ts'o <tytso@mit.edu>
Sun, 15 Feb 2009 23:07:52 +0000 (18:07 -0500)
commit705895b61133ef43d106fe6a6bbdb2eec923867e
treeea05232098c0425fda052240a8bb0e6218bb5e82
parent3d0518f4758eca4339e75e5b9dbb7e06a5ce08b4
ext4: allocate ->s_blockgroup_lock separately

As spotted by kmemtrace, struct ext4_sb_info is 17664 bytes on 64-bit
which makes it a very bad fit for SLAB allocators.  The culprit of the
wasted memory is ->s_blockgroup_lock which can be as big as 16 KB when
NR_CPUS >= 32.

To fix that, allocate ->s_blockgroup_lock, which fits nicely in a order 2
page in the worst case, separately.  This shinks down struct ext4_sb_info
enough to fit a 2 KB slab cache so now we allocate 16 KB + 2 KB instead of
32 KB saving 14 KB of memory.

Acked-by: Andreas Dilger <adilger@sun.com>
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
fs/ext4/ext4_sb.h
fs/ext4/super.c