netdev - Re: Consider switching to WQ_UNBOUND messages (was: Re: [PATCH v2 6/7] workqueue: Report work funcs that trigger automatic CPU

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Tue, 25 Jul 2023 16:46:38 +0200
From: Geert Uytterhoeven <geert@...ux-m68k.org>
To: Tejun Heo <tj@...nel.org>
Cc: Lai Jiangshan <jiangshanlai@...il.com>, 
	"torvalds@...ux-foundation.org" <torvalds@...ux-foundation.org>, Peter Zijlstra <peterz@...radead.org>, 
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>, kernel-team@...a.com, 
	Linux PM list <linux-pm@...r.kernel.org>, 
	DRI Development <dri-devel@...ts.freedesktop.org>, linux-rtc@...r.kernel.org, 
	linux-riscv <linux-riscv@...ts.infradead.org>, netdev <netdev@...r.kernel.org>, 
	Linux Fbdev development list <linux-fbdev@...r.kernel.org>, Linux MMC List <linux-mmc@...r.kernel.org>, 
	"open list:LIBATA SUBSYSTEM (Serial and Parallel ATA drivers)" <linux-ide@...r.kernel.org>, 
	Linux-Renesas <linux-renesas-soc@...r.kernel.org>
Subject: Re: Consider switching to WQ_UNBOUND messages (was: Re: [PATCH v2
 6/7] workqueue: Report work funcs that trigger automatic CPU_INTENSIVE mechanism)

Hi Tejun,

On Wed, Jul 19, 2023 at 12:01 AM Tejun Heo <tj@...nel.org> wrote:
> On Tue, Jul 18, 2023 at 11:54:58AM +0200, Geert Uytterhoeven wrote:
> > I gave it a try on a system with an 800 MHz Cortex A9, only to discover
> > it makes no difference, as that machine has 1600 BogoMIPS:
>
> Oops.
>
> > workqueue: blk_mq_run_work_fn hogged CPU for >10000us 4 times,
> > consider switching to WQ_UNBOUND
>
> It could be that we actually want to switch to UNBOUND for some reports but
> the above triggering most likely indicates that the threshold is too
> aggressive.
>
> > Artificially low BogoMIPS numbers only happen on systems that have
> > the related timers (Cortex A7/A15 and later, Cortex A9 MPCore,
> > and arm64).
>
> Ah, I see. Thanks for the explanation.
>
> > I will test on more systems, but that will probably not happen until
> > next week...
>
> Thanks, really appreciate it. Can you try the following instead when you
> have time? I just pushed up the lower boundary to 4000 MIPS. The scaling is
> still capped at 1s.

Thanks, with the below, I see no more WQ_UNBOUND messages.

> From 8555cbd4b22e5f85eb2bdcb84fd1d1f519a0a0d3 Mon Sep 17 00:00:00 2001
> From: Tejun Heo <tj@...nel.org>
> Date: Mon, 17 Jul 2023 12:50:02 -1000
> Subject: [PATCH] workqueue: Scale up wq_cpu_intensive_thresh_us if BogoMIPS is
>  below 4000

> --- a/kernel/workqueue.c
> +++ b/kernel/workqueue.c

> @@ -6513,6 +6516,42 @@ void __init workqueue_init_early(void)
>                !system_freezable_power_efficient_wq);
>  }
>
> +static void __init wq_cpu_intensive_thresh_init(void)
> +{
> +       unsigned long thresh;
> +       unsigned long mips;

This fails to build on mips.
Apparently mips is a predefined preprocessor macro:

$ echo | mipsel-linux-gnu-gcc -dM -E - | grep -w mips
#define mips 1

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...ux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds