linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pavel Machek <pavel@ucw.cz>
To: Giuseppe CAVALLARO <peppe.cavallaro@st.com>
Cc: alexandre.torgue@st.com, David Miller <davem@davemloft.net>,
	netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: stmmac ethernet in kernel 4.9-rc6: coalescing related pauses.
Date: Mon, 5 Dec 2016 13:01:36 +0100	[thread overview]
Message-ID: <20161205120136.GC16126@amd> (raw)
In-Reply-To: <c4748c8f-ae4b-9bbc-7d20-9c7b7ce433c8@st.com>

[-- Attachment #1: Type: text/plain, Size: 1870 bytes --]

Hi!

> >>We had experimented this tuning on STB IP where just datagrams
> >>had to send externally. To be honest, although we had seen
> >>better results w/o any timer, we kept this approach enabled
> >>because the timer was fast enough to cover our tests on SH4 boxes.
> >
> >Please reply to David, and explain how it is supposed to
> >work... because right now it does not. 40 msec delays are not
> >acceptable in default configuration.
> 
> I mean, that on UP and SMP system this schema helped
> to improve the performance saving CPU on my side and this has been
> tested since a long time (~4 years).
> I tested something similar to yours where unidirectional traffic
> with limited throughput was needed and I can confirm you that
> tuning/removing coalesce parameters this helped. The tuning I decided
> to keep in the driver was suitable in several user cases and if now
> you have problems or you want to review it I can just confirm that
> there are no problems on my side. If you want to simply the logic
> around the tx process and remove timer on official driver I can accept
> that. I will just ask you uto double check if the throughput and
> CPU usage when request max throughput (better if on GiGa setup) has
> no regressions.
> Otherwise we could start thinking about adaptive schema if feasible.

Ok, so you see the issue. Good.

See the other email to description how it could be fixed... the logic
is broken. How it was not discovered for 4 years is mystery to me.
 
> >Agreed that no irqlock protection is needed if we rely on napi and timers.
> 
> ok

Actually I was wrong there. Another reason to disable tx coalescing
until it is fixed.

Best regards,
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

  parent reply	other threads:[~2016-12-05 12:02 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-23 10:51 stmmac ethernet in kernel 4.4: coalescing related pauses? Pavel Machek
2016-11-24  8:55 ` stmmac ethernet in kernel 4.9-rc6: coalescing related pauses Pavel Machek
2016-11-24 10:29   ` Pavel Machek
2016-11-24 10:36     ` Pavel Machek
2016-11-24 10:46       ` [PATCH] stmmac ethernet: unify locking Pavel Machek
2016-11-24 11:05         ` [PATCH] stmmac ethernet: remove cut & paste code Pavel Machek
2016-11-24 20:05           ` Joe Perches
2016-11-24 21:44             ` Pavel Machek
2016-11-24 22:27               ` Joe Perches
2016-11-28 11:50                 ` Pavel Machek
2016-11-28 14:24                   ` Joe Perches
2016-11-28 14:35                     ` Pavel Machek
2016-11-28 16:03                       ` Joe Perches
2016-11-24 16:04   ` stmmac ethernet in kernel 4.9-rc6: coalescing related pauses David Miller
2016-11-24 21:25     ` Pavel Machek
2016-12-02  8:24       ` Giuseppe CAVALLARO
2016-12-02  8:41         ` Giuseppe CAVALLARO
2016-12-02  8:45         ` Pavel Machek
2016-12-02  9:43           ` Giuseppe CAVALLARO
2016-12-02 12:32             ` Pavel Machek
2016-12-02 13:51               ` Giuseppe CAVALLARO
2016-12-02 14:26                 ` Alexandre Torgue
2016-12-02 15:19                   ` Giuseppe CAVALLARO
2016-12-05 12:14                     ` Pavel Machek
2016-12-05 12:01                 ` Pavel Machek [this message]
2016-12-05 10:15             ` Pavel Machek
2016-12-05 11:40               ` Lino Sanfilippo
2016-12-05 22:02                 ` Pavel Machek
2016-12-05 22:37                   ` Lino Sanfilippo
2016-12-05 22:40                     ` Pavel Machek
2016-12-05 22:54                       ` Lino Sanfilippo
2016-12-05 23:11                         ` Lino Sanfilippo
2016-12-02 14:05           ` Aw: " Lino Sanfilippo
2016-12-07 12:31             ` [RFC] " Pavel Machek
2016-12-07 13:18               ` Lino Sanfilippo
2016-12-05 11:56         ` Pavel Machek
2016-11-28 11:55     ` [PATCH] stmmac: fix comments, make debug output consistent Pavel Machek
2016-11-30  0:53       ` David Miller
2016-11-28 12:13     ` stmmac ethernet in kernel 4.9-rc6: coalescing related pauses Pavel Machek
2016-11-28 12:17     ` [PATCH] stmmac: reduce code duplication getting basic descriptors Pavel Machek
2016-11-28 15:25       ` kbuild test robot
2016-12-02 14:09       ` Alexandre Torgue
2016-11-30 11:44     ` [PATCH] stmmac: simplify flag assignment Pavel Machek
2016-12-01 20:23       ` David Miller
2016-12-01 22:48         ` stmmac: turn coalescing / NAPI off in stmmac Pavel Machek
2016-12-02  8:39           ` Giuseppe CAVALLARO
2016-12-02 10:42             ` Pavel Machek
2016-12-02 15:31               ` Giuseppe CAVALLARO
2016-12-05 11:45                 ` Pavel Machek
2016-12-02  8:27       ` [PATCH] stmmac: simplify flag assignment Giuseppe CAVALLARO
2016-12-01 10:32     ` [PATCH] stmmac: cleanup documenation, make it match reality Pavel Machek
2016-12-03 20:07       ` David Miller
2016-12-05 12:27     ` [PATCH] stmmac: disable tx coalescing Pavel Machek
2016-12-11 19:07       ` Pavel Machek
2016-12-11 19:31         ` David Miller
2016-12-11 19:57           ` Pavel Machek
2016-11-28 13:07 ` stmmac ethernet in kernel 4.4: coalescing related pauses? Lino Sanfilippo
2016-11-28 14:54   ` David Miller
2016-11-28 15:31     ` Eric Dumazet
2016-11-28 15:57       ` Lino Sanfilippo
2016-11-28 16:30         ` David Miller
2016-11-28 17:01           ` Lino Sanfilippo
2016-11-30 10:28       ` Pavel Machek
2016-11-28 15:33     ` Lino Sanfilippo

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=20161205120136.GC16126@amd \
    --to=pavel@ucw.cz \
    --cc=alexandre.torgue@st.com \
    --cc=davem@davemloft.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=peppe.cavallaro@st.com \
    /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;
as well as URLs for NNTP newsgroup(s).