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


Groups > linux.debian.kernel > #85692

Bug#972709: Wishlist/RFC: Change to CONFIG_PREEMPT_NONE in linux-image-cloud-*

Path csiph.com!fu-berlin.de!bofh.it!news.nic.it!robomod
From Salvatore Bonaccorso <carnil@debian.org>
Newsgroups linux.debian.bugs.dist, linux.debian.kernel
Subject Bug#972709: Wishlist/RFC: Change to CONFIG_PREEMPT_NONE in linux-image-cloud-*
Date Thu, 20 Feb 2025 10:40:01 +0100
Message-ID <Kibcd-zW2-7@gated-at.bofh.it> (permalink)
References <B2QjL-7oT-1@gated-at.bofh.it> <B3ged-64a-5@gated-at.bofh.it> <Be61P-6mT-5@gated-at.bofh.it> <Be6bv-6q4-1@gated-at.bofh.it> <B2QjL-7oT-1@gated-at.bofh.it> <Be7qV-75I-3@gated-at.bofh.it> <B2QjL-7oT-1@gated-at.bofh.it> <Be9j3-8cF-1@gated-at.bofh.it> <B2QjL-7oT-1@gated-at.bofh.it> <Be9j3-8cF-1@gated-at.bofh.it>
X-Original-To Noah Meyerhans <noahm@debian.org>, 972709@bugs.debian.org
X-Mailbox-Line From debian-bugs-dist-request@lists.debian.org Thu Feb 20 09:33:10 2025
Old-Return-Path <debbugs@buxtehude.debian.org>
X-Spam-Flag NO
X-Spam-Score -3.899
Reply-To Salvatore Bonaccorso <carnil@debian.org>, 972709@bugs.debian.org
Resent-To debian-bugs-dist@lists.debian.org
Resent-Cc Debian Kernel Team <debian-kernel@lists.debian.org>
X-Debian-Pr-Message followup 972709
X-Debian-Pr-Package src:linux
X-Debian-Pr-Source linux
MIME-Version 1.0
Content-Type text/plain; charset=us-ascii
Content-Disposition inline
X-Debian-User carnil
X-Debian-Message from BTS
X-Mailing-List <debian-bugs-dist@lists.debian.org> archive/latest/1888244
List-ID <debian-bugs-dist.lists.debian.org>
List-URL <https://lists.debian.org/debian-bugs-dist/>
Approved robomod@news.nic.it
Lines 36
Organization linux.* mail to news gateway
Sender robomod@news.nic.it
X-Original-Cc Flavio Veloso Soares <flaviovs@magnux.com>, Ben Hutchings <ben@decadent.org.uk>
X-Original-Date Thu, 20 Feb 2025 10:31:43 +0100
X-Original-Message-ID <Z7b2fxgdARSYD99Y@eldamar.lan>
X-Original-References <b7aad91f-31c1-89cf-075a-56402f4b81ce@magnux.com> <303ebe0e4f9a10001c183f6eb6c789aa930c82f7.camel@decadent.org.uk> <e95d04dd-8513-cf52-c1d6-b608d053b3ac@magnux.com> <52725d08fdaaffb773dc4955a4efc8f8ac22a40c.camel@decadent.org.uk> <b7aad91f-31c1-89cf-075a-56402f4b81ce@magnux.com> <0fa1b744-11a4-d459-06e1-2d49e3276ec8@magnux.com> <b7aad91f-31c1-89cf-075a-56402f4b81ce@magnux.com> <20201123014922.GA26417@doom.morgul.net> <b7aad91f-31c1-89cf-075a-56402f4b81ce@magnux.com> <20201123014922.GA26417@doom.morgul.net>
Xref csiph.com linux.debian.bugs.dist:1234173 linux.debian.kernel:85692

Cross-posted to 2 groups.

Show key headers only | View raw


Hi,

On Sun, Nov 22, 2020 at 05:49:22PM -0800, Noah Meyerhans wrote:
> On Sun, Nov 22, 2020 at 03:53:32PM -0800, Flavio Veloso Soares wrote:
> >  Unfortunately, I couldn't find many comprehensive benchmarks of kernel
> >  CONFIG_PREEMPT* options. The one at
> >  [1]https://www.codeblueprint.co.uk/2019/12/23/linux-preemption-latency-throughput.html
> >  seems to be very thorough,
> > 
> >  [...]
> > 
> >  Not particularly.  I'm used to latency benchmarks showing e.g. average,
> >  90th percentile, 99th percentile, as well as worst.
> 
> I don't think Ben was talking about specific benchmarks.  The web page
> you cites lacks basic measurements one would expect to see from *any*
> meaningful performance benchmark.  Comparing maximum latency is fine,
> but it's not really relevant by itself.  If a configuration change
> improves the worst case (100th percentile) but negatively impacts the
> 50th percentile, is that a change worth making?  Maybe.  But without
> having that data at all, the benchmark really isn't worth much at all.
> 
> It's totally reasonable for us to consider making this change, but we
> should have comprehensive data about the impact of doing so.  What
> impact does the change have on different classes of workloads?  e.g.
> high tps, CPU-bound, IO-bound, etc.  It's entirely possible that the
> proposed change improves performance under certain workloads, but
> negatively impacts others.  Without knowing the impact in more in more
> detail, which would allow us to evaluate the tradeoffs, I don't think
> there's a compelling reason to make a change.

I think with setting CONFIG_PREEMPT_LAZY since 6.13.2-1~exp1 we can
close this wishlist bug with that version?

Regards,
Salvatore

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


Thread

Bug#972709: Wishlist/RFC: Change to CONFIG_PREEMPT_NONE in linux-image-cloud-* Salvatore Bonaccorso <carnil@debian.org> - 2025-02-20 10:40 +0100

csiph-web