NFSD: memory corruption due to writing beyond the stat array
authorKonstantin Khorenko <khorenko@parallels.com>
Tue, 1 Feb 2011 14:16:29 +0000 (17:16 +0300)
committerJ. Bruce Fields <bfields@redhat.com>
Mon, 14 Feb 2011 15:35:18 +0000 (10:35 -0500)
If nfsd fails to find an exported via NFS file in the readahead cache, it
should increment corresponding nfsdstats counter (ra_depth[10]), but due to a
bug it may instead write to ra_depth[11], corrupting the following field.

In a kernel with NFSDv4 compiled in the corruption takes the form of an
increment of a counter of the number of NFSv4 operation 0's received; since
there is no operation 0, this is harmless.

In a kernel with NFSDv4 disabled it corrupts whatever happens to be in the
memory beyond nfsdstats.

Signed-off-by: Konstantin Khorenko <khorenko@openvz.org>
Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
fs/nfsd/vfs.c

index fda3be237773a83076a443a97de5d582c90e0d27..30c73f8a57913dd6540fb9d7690410c8c01c9eb8 100644 (file)
@@ -808,7 +808,7 @@ nfsd_get_raparms(dev_t dev, ino_t ino)
                if (ra->p_count == 0)
                        frap = rap;
        }
-       depth = nfsdstats.ra_size*11/10;
+       depth = nfsdstats.ra_size;
        if (!frap) {    
                spin_unlock(&rab->pb_lock);
                return NULL;