linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Jeffrey Vanhoof <jdv1029@gmail.com>
To: linux-mm@kvack.org
Subject: issue with direct reclaims and kswapd reclaims on 2.6.35.7
Date: Mon, 15 Aug 2011 00:08:23 -0500	[thread overview]
Message-ID: <CAML7nqd9_F4L0M7ynLFz4HKET94n2mwsk42Z7g2EjAfYnD-JgQ@mail.gmail.com> (raw)

On a 2.6.35.7 based kernel, on a portable device with 512MB RAM, I am
seeing the following issues while consuming 20Mbps video content:
1) direct reclaims occurring quite frequently, resulting in delayed
file read requests
2) direct reclaims falling into congestion_wait() even though no
congestion at the time, this results in video jitter.
3) kswapd not reclaiming pages quickly enough due to falling into
congestion_wait() very often.
4) power consumption is degraded as a result of time being spent in
io_schedule_timeout() called within congestion_wait(). (the power
c-state will stay in C0)

Are there specific patches which can be easily back-ported to K35
which may address most of these issues?

For file read performance, I believe it is better for kswapd to
reclaim memory instead of hitting a direct reclaim, and for power it
would be best that while reclaiming memory in kswapd that
io_schedule()/io_schedule_timeout() is never called unless absolutely
required.

Are any of the workarounds listed below appropriate to use?
1) change the congestion_wait() timeout value in balance_pgdat() from
HZ/10 to HZ/50. This allows for faster reclaims in kswapd and limits
the time spend in congestion_wait().
2) change SWAP_CLUSTER_MAX from 32 to 128 or higher (swap is enabled,
but there is no swap).
3) change DEF_PRIORITY from 12 to 9. This results in a larger scan and
pages are reclaimed quicker. Also, this causes congestion_wait() to be
called less frequently due to the likelyhood of pages being found with
increase priority.

Any help would be appreciated.

Thanks,
Jeff V.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

             reply	other threads:[~2011-08-15  5:08 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-08-15  5:08 Jeffrey Vanhoof [this message]
2011-08-18  6:47 ` Jeffrey Vanhoof

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAML7nqd9_F4L0M7ynLFz4HKET94n2mwsk42Z7g2EjAfYnD-JgQ@mail.gmail.com \
    --to=jdv1029@gmail.com \
    --cc=linux-mm@kvack.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox