Files
mirror_zfs/module/zfs
Brian Behlendorf b0319c1faa OpenZFS 7143 - dbuf_read() creates unnecessary zio_root() for bonus buf
dbuf_read() creates a zio_root() to track and wait for all the
zio's that may happen as part of this call. However, if the blkptr_t
for this buffer is NULL or a hole, we will not create any more zio's,
so this zio_root() is unnecessary. This is always the case when calling
dbuf_read() on a bonus buffer, because it has no blkptr (it's part of
the containing dnode). For workloads that read a lot of bonus buffers
(e.g. file creation and removal), creating and destroying these
unnecessary zio's can decrease performance by around 3%.

The fix is to only create/destroy the zio_root() in dbuf_read() if
the blkptr is not NULL and not a hole.

Changes sponsored by Intel Corp.

Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by:  Alex Zhuravlev <alexey.zhuravlev@intel.com>
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue openzfs/openzfs#137
Closes #4803 
Closes #5382
2016-12-01 16:50:11 -07:00
..
2015-01-16 14:41:26 -08:00
2016-01-08 15:08:19 -08:00
2015-01-16 14:41:26 -08:00
2016-11-02 12:14:45 -07:00
2016-06-28 13:47:02 -07:00
2016-10-08 14:57:56 -07:00
2013-12-18 16:46:35 -08:00
2015-01-16 14:41:26 -08:00
2015-01-16 14:41:26 -08:00
2015-12-22 10:21:33 -08:00
2016-04-21 09:49:25 -07:00
2016-06-21 09:58:37 -07:00
2016-12-01 16:39:06 -07:00
2010-08-31 13:41:58 -07:00
2016-08-11 11:23:49 -07:00
2016-11-30 16:18:20 -07:00
2016-11-29 14:34:32 -08:00
2016-11-29 14:34:33 -08:00
2016-11-29 14:34:33 -08:00
2016-11-29 14:34:33 -08:00
2013-10-31 14:58:04 -07:00
2016-11-30 16:18:20 -07:00
2010-08-31 13:41:58 -07:00
2015-09-04 16:08:14 -07:00
2016-09-21 18:09:00 -07:00
2016-09-21 18:09:00 -07:00
2016-11-30 16:18:20 -07:00
2016-11-07 11:04:44 -08:00
2016-11-10 08:50:32 -08:00
2016-11-04 10:46:40 -07:00
2016-05-31 16:04:26 -07:00
2016-11-02 12:14:45 -07:00