linux-kernel-test/block
Jens Axboe b31dc66a54 [PATCH] Kill PF_SYNCWRITE flag
A process flag to indicate whether we are doing sync io is incredibly
ugly. It also causes performance problems when one does a lot of async
io and then proceeds to sync it. Part of the io will go out as async,
and the other part as sync. This causes a disconnect between the
previously submitted io and the synced io. For io schedulers such as CFQ,
this will cause us lost merges and suboptimal behaviour in scheduling.

Remove PF_SYNCWRITE completely from the fsync/msync paths, and let
the O_DIRECT path just directly indicate that the writes are sync
by using WRITE_SYNC instead.

Signed-off-by: Jens Axboe <axboe@suse.de>
2006-06-23 17:10:39 +02:00
..
as-iosched.c [PATCH] Kill PF_SYNCWRITE flag 2006-06-23 17:10:39 +02:00
blktrace.c [PATCH] Block queue IO tracing support (blktrace) as of 2006-03-23 2006-03-23 20:00:26 +01:00
cfq-iosched.c [PATCH] Kill PF_SYNCWRITE flag 2006-06-23 17:10:39 +02:00
deadline-iosched.c [PATCH] iosched: use hlist for request hashtable 2006-06-23 17:10:38 +02:00
elevator.c [PATCH] remove dead code from elevator switching 2006-06-23 17:10:38 +02:00
genhd.c [PATCH] Driver core: add generic "subsystem" link to all devices 2006-06-21 12:40:49 -07:00
ioctl.c [PATCH] Block queue IO tracing support (blktrace) as of 2006-03-23 2006-03-23 20:00:26 +01:00
Kconfig [PATCH] config: fix CONFIG_LFS option 2006-03-31 12:18:55 -08:00
Kconfig.iosched [BLOCK] iosched: fix setting of default io scheduler 2005-11-04 08:44:58 +01:00
ll_rw_blk.c [PATCH] Kill PF_SYNCWRITE flag 2006-06-23 17:10:39 +02:00
Makefile [PATCH] Block queue IO tracing support (blktrace) as of 2006-03-23 2006-03-23 20:00:26 +01:00
noop-iosched.c [PATCH] elevator switching race 2006-06-08 15:14:23 -07:00
scsi_ioctl.c [SCSI] unify SCSI_IOCTL_SEND_COMMAND implementations 2006-04-13 10:13:15 -05:00