cfq: merge cooperating cfq_queues
commitd34507d090ff70f0663659e70b5980248a5d6774
authorJeff Moyer <jmoyer@redhat.com>
Thu, 29 Sep 2011 09:57:11 +0000 (29 15:27 +0530)
committerGreg Kroah-Hartman <gregkh@suse.de>
Mon, 7 Nov 2011 20:32:39 +0000 (7 12:32 -0800)
tree717db9f0e082f8d2bd3fa50fe79b4f39e03cfee4
parent4a78a06ce2ecea224c55eb2e171ea020110ecaa9
cfq: merge cooperating cfq_queues

commit df5fe3e8e13883f58dc97489076bbcc150789a21 upstream.

When cooperating cfq_queues are detected currently, they are allowed to
skip ahead in the scheduling order.  It is much more efficient to
automatically share the cfq_queue data structure between cooperating processes.
Performance of the read-test2 benchmark (which is written to emulate the
dump(8) utility) went from 12MB/s to 90MB/s on my SATA disk.  NFS servers
with multiple nfsd threads also saw performance increases.

Signed-off-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Suresh Jayaraman <sjayaraman@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
block/cfq-iosched.c