Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1443905

Re: [PATCH 5/5] f2fs: use blk_plug in all the possible paths

From Jaegeuk Kim <jaegeuk@kernel.org>
Newsgroups linux.kernel
Subject Re: [PATCH 5/5] f2fs: use blk_plug in all the possible paths
Date 2016-07-15 05:10 +0200
Message-ID <rV1FL-1xr-3@gated-at.bofh.it> (permalink)
References <rUZuh-8n3-3@gated-at.bofh.it> <rUZuh-8n3-1@gated-at.bofh.it> <rV133-14j-9@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


On Thu, Jul 14, 2016 at 07:21:32PM -0700, Christoph Hellwig wrote:
> On Thu, Jul 14, 2016 at 05:46:29PM -0700, Jaegeuk Kim wrote:
> > The main reason is that blk_start_plug can be used to wake up from low-power
> > mode before submitting further bios.
> 
> Where does that myth come from?

From kernel guys working on android.
And, I'm in also doubt they may have done something regarding to power mode
change, by the fact that I've seen annoying performance jitters comparing to
ext4's ones all the time in some android products. 

Frankly speaking, I cannot find a concrete reason to remove plugging entirely,
since I can't see big performance differences in all my test results in severs
as well.

Thanks,

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread


Thread

[PATCH 5/5] f2fs: use blk_plug in all the possible paths Jaegeuk Kim <jaegeuk@kernel.org> - 2016-07-15 02:50 +0200
  Re: [PATCH 5/5] f2fs: use blk_plug in all the possible paths Christoph Hellwig <hch@infradead.org> - 2016-07-15 04:30 +0200
    Re: [PATCH 5/5] f2fs: use blk_plug in all the possible paths Jaegeuk Kim <jaegeuk@kernel.org> - 2016-07-15 05:10 +0200
      Re: [PATCH 5/5] f2fs: use blk_plug in all the possible paths Christoph Hellwig <hch@infradead.org> - 2016-07-19 06:10 +0200
        Re: [PATCH 5/5] f2fs: use blk_plug in all the possible paths Jaegeuk Kim <jaegeuk@kernel.org> - 2016-07-19 07:50 +0200

csiph-web