[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <cover.1430414609.git.shli@fb.com>
Date: Thu, 30 Apr 2015 10:45:13 -0700
From: Shaohua Li <shli@...com>
To: <linux-kernel@...r.kernel.org>
CC: <axboe@...com>, <jmoyer@...hat.com>, <hch@....de>, <neilb@...e.de>
Subject: [PATCH 0/5] blk plug fixes
These are some block plug related patches. The first 3 are independent and can
apply separately.
For the multiple queue case, I only handled !BLK_MQ_F_DEFER_ISSUE case. The
DEFER_ISSUE is really confusing, I'm not sure what's the correct behavior.
BTW, calling blk_mq_merge_queue_io is confusing too to me. The request will be
dispatched immediately or offload to a workqueue to dispatch. In either case,
the request will be dispatched soon. There is a tiny window there are requests
in the queue and requests can be merged. I thought we can delete the code with
plug merge. And if out of order request merge is important, we really should
add a ELEVATOR_INSERT_SORT_MERGE like mechanism. This makes me more confusing
about BLK_MQ_F_DEFER_ISSUE too.
Thanks,
Shaohua
Jeff Moyer (1):
blk-mq: fix plugging in blk_sq_make_request
Shaohua Li (4):
blk: clean up plug
sched: always use blk_schedule_flush_plug in io_schedule_out
blk-mq: do limited block plug for multiple queue case
blk-mq: make plug work for mutiple disks and queues
block/blk-core.c | 34 ++++++++-------
block/blk-mq.c | 120 ++++++++++++++++++++++++++++++++--------------------
block/blk.h | 3 +-
kernel/sched/core.c | 8 +---
4 files changed, 98 insertions(+), 67 deletions(-)
--
1.8.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists