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


Groups > linux.debian.bugs.rc > #345288

Bug#1041855: linux-image-6.4.0-1-amd64: IO wait regression

From Ben Hutchings <ben@decadent.org.uk>
Newsgroups linux.debian.bugs.rc
Subject Bug#1041855: linux-image-6.4.0-1-amd64: IO wait regression
Date 2023-07-25 01:00 +0200
Message-ID <GVcGZ-2wX0-1@gated-at.bofh.it> (permalink)
References <GV46J-2ro1-9@gated-at.bofh.it> <GV46J-2ro1-9@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


[Multipart message — attachments visible in raw view] - view raw

Control: severity -1 important

On Mon, 2023-07-24 at 15:40 +0200, Christian Göttsche wrote:
> Package: src:linux
> Version: 6.4.4-1
> Severity: serious
> 
> Dear Maintainer,
> 
> Kernel 6.4.4 is affected by a regression causing one core be report
> high IO wait utilization.
> 
> See https://lore.kernel.org/lkml/12251678.O9o76ZdvQC@natalenko.name/

There is no such thing as a CPU having "high IO wait utilization". 
When tasks are in I/O wait state, they are not using a CPU.

Unfortunately, for historical reasons Linux includes such tasks in the
load average and reports a CPU as being in I/O wait state when it's
idle and the last task running on it went into I/O wait state (at least
I think that's what the rule is).

So this bug is about an unwanted change in task/CPU state reporting,
not tasks suddenly using much more CPU time.  For that reason, I don't
think it should "serious", i.e. a blocker for testing propagation.

Ben.

-- 
Ben Hutchings
Never attribute to conspiracy what can adequately be explained
by stupidity.

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


Thread

Bug#1041855: linux-image-6.4.0-1-amd64: IO wait regression Ben Hutchings <ben@decadent.org.uk> - 2023-07-25 01:00 +0200

csiph-web