[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <6e73aaab-fd2e-f9c4-0826-16643717694e@kernel.dk>
Date: Thu, 13 Jul 2023 12:13:58 -0600
From: Jens Axboe <axboe@...nel.dk>
To: Tejun Heo <tj@...nel.org>,
Chengming Zhou <chengming.zhou@...ux.dev>
Cc: hch@....de, linux-block@...r.kernel.org,
linux-kernel@...r.kernel.org, ming.lei@...hat.com,
zhouchengming@...edance.com
Subject: Re: [PATCH v5] blk-mq: fix start_time_ns and alloc_time_ns for
pre-allocated rq
On 7/13/23 11:58?AM, Tejun Heo wrote:
> It's a bit out of scope for this patchset but I think it might make
> sense to build a timestamp caching infrastructure. The cached
> timestamp can be invalidated on context switches (block layer already
> hooks into them) and issue and other path boundaries (e.g. at the end
> of plug flush).
This is a great idea! Have the plug init start with the timestamp
invalid, and use some blk_get_time() helpers that return the time for no
plug, and set it in the plug if not set. Flushing the plug would mark it
invalid again.
This obviously won't help the no plugging cases, but those are not that
interesting in comparison.
--
Jens Axboe
Powered by blists - more mailing lists