Files
ubports_kernel_google_msm/include/linux
Trond Myklebust c63c7b0513 NFS: Fix a race when doing NFS write coalescing
Currently we do write coalescing in a very inefficient manner: one pass in
generic_writepages() in order to lock the pages for writing, then one pass
in nfs_flush_mapping() and/or nfs_sync_mapping_wait() in order to gather
the locked pages for coalescing into RPC requests of size "wsize".

In fact, it turns out there is actually a deadlock possible here since we
only start I/O on the second pass. If the user signals the process while
we're in nfs_sync_mapping_wait(), for instance, then we may exit before
starting I/O on all the requests that have been queued up.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2007-04-30 22:17:06 -07:00
..
2007-02-20 17:10:15 -08:00
2007-02-20 17:10:14 -08:00
2007-04-27 10:57:31 -07:00
2007-04-27 10:57:34 -07:00
2007-03-27 09:05:15 -07:00
2007-04-28 11:01:07 -04:00
2007-04-19 14:56:12 +02:00
2007-04-07 10:03:43 -07:00
2007-04-25 22:30:01 -07:00
2007-04-25 22:29:10 -07:00
2007-04-25 22:24:41 -07:00
2007-04-25 22:29:10 -07:00
2007-03-01 14:53:38 -08:00
2007-04-25 22:25:52 -07:00
2007-04-30 16:40:39 -07:00
2007-04-30 16:40:40 -07:00
2007-04-25 22:29:49 -07:00
2007-03-04 11:12:42 +02:00
2007-02-17 19:17:37 +01:00
2007-03-01 14:53:35 -08:00
2007-03-12 16:31:50 -07:00
2007-03-16 19:25:05 -07:00
2007-03-12 16:31:50 -07:00
2007-04-28 14:15:58 -04:00
2007-03-27 08:55:47 +02:00
2007-04-17 16:36:26 -07:00
2007-04-30 16:40:41 -07:00
2007-04-27 10:46:51 -07:00
2007-02-20 17:10:14 -08:00
2007-02-20 17:10:14 -08:00
2007-02-20 17:10:13 -08:00
2007-02-16 08:13:57 -08:00
2007-02-16 08:13:56 -08:00
2007-03-05 07:57:51 -08:00
2007-04-28 21:20:32 -07:00