All of lore.kernel.org
 help / color / mirror / Atom feed
From: Philippe Gerum <rpm@xenomai.org>
To: Gabriel Durante <gabriel@neuraldsp.com>
Cc: xenomai@lists.linux.dev
Subject: Re: Xenomai 4 IMX8MM
Date: Thu, 04 Aug 2022 16:39:44 +0200	[thread overview]
Message-ID: <877d3oys2r.fsf@xenomai.org> (raw)
In-Reply-To: <CAANaHT9E=73w7_sOugpw7JBToTzfRtDy+aTeW5HgPKZjB_rKQQ@mail.gmail.com>


Gabriel Durante <gabriel@neuraldsp.com> writes:

> Thanks Philippe,
>
> IRQCHIP_PIPELINE_SAFE flag on gpio-pca953x.c did the trick, no issues on dmesg.
>
> Now only the issue (kernel hangs) with edt_ft5x06 loaded while running
> sched-quota-accuracy.
>
> [ 2341.500412] edt_ft5x06 2-0038: Unable to fetch data, error: -110
>

I suspect this may be the outcome of the test overconsuming the CPU
bandwidth on the same processor which runs the (threaded?) IRQ handler
serving the touchpad events. IOW, the IRQ handler cannot run quickly
enough for the touchpad hw and/or driver to be happy, because
sched-quota-accuracy is keeping the CPU (too) busy.

Three options to check this:

- move the test away from CPU1, which might be shared with the IRQ
  handler, using the -c <cpu> option (you can run sched-quota-accuracy
  manually, outside of the 'evl test' scripting).

- change the CPU affinity of the IRQ for the touchpad device, away from
  CPU1. Fixing up /proc/irq/<irqnum>/smp_affinity accordingly may help,
  with <irqnum> matching the proper device in /proc/interrupts.

- try reducing the load imposed on the system by the test, e.g. by
  lowering the number of concurrent threads using the -n <nr_threads>
  option. Defaults to 3, maybe 1 would allow the test to pass in that
  configuration.

You also may want to make sure to switch off
CONFIG_PROVE_LOCKING->CONFIG_DEBUG_LOCK_ALLOC, CONFIG_DEBUG_LIST and
friends which slow down the (regular) kernel code a lot.

-- 
Philippe.

  reply	other threads:[~2022-08-04 14:57 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-08-03 21:13 Xenomai 4 IMX8MM Gabriel Durante
2022-08-04 12:44 ` Philippe Gerum
2022-08-04 13:40   ` Gabriel Durante
2022-08-04 14:39     ` Philippe Gerum [this message]
2022-08-04 17:03       ` Gabriel Durante

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=877d3oys2r.fsf@xenomai.org \
    --to=rpm@xenomai.org \
    --cc=gabriel@neuraldsp.com \
    --cc=xenomai@lists.linux.dev \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.