From 080d676de095a14ecba14c0b9a91acb5bbb634df Mon Sep 17 00:00:00 2001 From: Jeff Moyer Date: Wed, 2 Nov 2011 13:40:10 -0700 Subject: aio: allocate kiocbs in batches In testing aio on a fast storage device, I found that the context lock takes up a fair amount of cpu time in the I/O submission path. The reason is that we take it for every I/O submitted (see __aio_get_req). Since we know how many I/Os are passed to io_submit, we can preallocate the kiocbs in batches, reducing the number of times we take and release the lock. In my testing, I was able to reduce the amount of time spent in _raw_spin_lock_irq by .56% (average of 3 runs). The command I used to test this was: aio-stress -O -o 2 -o 3 -r 8 -d 128 -b 32 -i 32 -s 16384 I also tested the patch with various numbers of events passed to io_submit, and I ran the xfstests aio group of tests to ensure I didn't break anything. Signed-off-by: Jeff Moyer Cc: Daniel Ehrenberg Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/aio.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/aio.h b/include/linux/aio.h index 2dcb72bff4b6..2314ad8b3c9c 100644 --- a/include/linux/aio.h +++ b/include/linux/aio.h @@ -117,6 +117,7 @@ struct kiocb { struct list_head ki_list; /* the aio core uses this * for cancellation */ + struct list_head ki_batch; /* batch allocation */ /* * If the aio_resfd field of the userspace iocb is not zero, -- cgit v1.2.3-71-gd317