[lkml]   [2011]   [Apr]   [19]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Patch in this message
Subject[43/70] vfs: Fix absolute RCU path walk failures due to uninitialized seq number
2.6.38-stable review patch.  If anyone has any objections, please let us know.


From: Tim Chen <>

commit c1530019e311c91d14b24d8e74d233152d806e45 upstream.

During RCU walk in path_lookupat and path_openat, the rcu lookup
frequently failed if looking up an absolute path, because when root
directory was looked up, seq number was not properly set in nameidata.

We dropped out of RCU walk in nameidata_drop_rcu due to mismatch in
directory entry's seq number. We reverted to slow path walk that need
to take references.

With the following patch, I saw a 50% increase in an exim mail server
benchmark throughput on a 4-socket Nehalem-EX system.

Signed-off-by: Tim Chen <>
Reviewed-by: Andi Kleen <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>

fs/namei.c | 1 +
1 file changed, 1 insertion(+)

--- a/fs/namei.c
+++ b/fs/namei.c
@@ -712,6 +712,7 @@ static __always_inline void set_root_rcu
do {
seq = read_seqcount_begin(&fs->seq);
nd->root = fs->root;
+ nd->seq = __read_seqcount_begin(&nd->root.dentry->d_seq);
} while (read_seqcount_retry(&fs->seq, seq));

 \ /
  Last update: 2011-04-19 22:27    [W:0.182 / U:0.364 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site