All of lore.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Adrián Larumbe" <adrian.larumbe@collabora.com>
Cc: dri-devel@lists.freedesktop.org
Subject: Re: [PATCH v15 13/15] drm/panfrost: Avoid cache flush after perfcnt sample in fully coherent systems
Date: Mon, 12 Oct 2026 15:05:44 +0000	[thread overview]
Message-ID: <sashiko-outbox-168312@kernel.org> (raw)
In-Reply-To: <20261012-claude-fixes-v15-13-bc0a6c498553@collabora.com>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] drm/panfrost: UAF of perfcnt during device unbind
--

commit 1164f77fb530d75cfd2e6bee453335ff4ea53de6
Author: Adrián Larumbe <adrian.larumbe@collabora.com>

drm/panfrost: Avoid cache flush after perfcnt sample in fully coherent systems
    
If the GPU and CPU are fully coherent, then there's no need to flush the
GPU caches after a perfcnt sample is made available.

> diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> index ffc77121070e2..b3f71d7fd82a9 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> @@ -52,7 +52,10 @@ void panfrost_perfcnt_clean_cache_done(struct panfrost_device *pfdev)
>  
>  void panfrost_perfcnt_sample_done(struct panfrost_device *pfdev)
>  {
> -	gpu_write(pfdev, GPU_CMD, GPU_CMD_CLEAN_CACHES);
> +	if (pfdev->features.selected_coherency != COHERENCY_ACE)
> +		gpu_write(pfdev, GPU_CMD, GPU_CMD_CLEAN_CACHES);
> +	else
> +		complete(&pfdev->perfcnt->dump_comp);
                ^^^^^^^^

[Severity: High]
Does this introduce a use-after-free regression during device unbind?

If a user thread calls the perfcnt dump ioctl and blocks in
wait_for_completion_interruptible_timeout() inside
panfrost_perfcnt_dump_locked(), and the device is concurrently unbound:

panfrost_remove() will return without cancelling the wait, and devres cleans
up allocations.

Because devm_kzalloc() for perfcnt in panfrost_perfcnt_init() occurs after
devm_request_irq(), devres frees pfdev->perfcnt while the thread is still
sleeping and the IRQ is still registered.

When the thread wakes up, it will access the freed completion structure.

Additionally, if a shared interrupt fires in this window and GPU_INT_STAT
returns 0xffffffff (because clocks are off), panfrost_perfcnt_sample_done()
will dereference the freed pfdev->perfcnt pointer here.

>  }
>  
>  static int panfrost_perfcnt_dump_locked(struct panfrost_device *pfdev)

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261012-claude-fixes-v15-0-bc0a6c498553@collabora.com?part=13

  reply	other threads:[~2026-10-12 15:05 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-12 14:47 [PATCH v15 00/15] Collection of fixes for Panfrost: Perfcnt, RPM, refactorings Adrián Larumbe
2026-10-12 14:47 ` [PATCH v15 01/15] drm/panfrost: Move shrinker initialization and unplug one level down Adrián Larumbe
2026-10-12 14:47 ` [PATCH v15 02/15] drm/panfrost: Move lock and modparam initialisations into their subsystems Adrián Larumbe
2026-10-12 14:47 ` [PATCH v15 03/15] drm/panfrost: Move debugfs initialisation to relevant subsystems Adrián Larumbe
2026-10-12 14:47 ` [PATCH v15 04/15] drm/panfrost: Skip NULL checks for clock enable/disabling Adrián Larumbe
2026-10-12 14:47 ` [PATCH v15 05/15] drm/panfrost: Consolidate device clock management and reset Adrián Larumbe
2026-10-12 14:47 ` [PATCH v15 06/15] drm/panfrost: Fix PM refcnt and autosuspend issues at device probe/remove Adrián Larumbe
2026-10-12 15:01   ` sashiko-bot
2026-10-12 14:47 ` [PATCH v15 07/15] drm/panfrost: Explicitly enable MMU interrupts at device init Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 08/15] drm/panfrost: Move all DRM device initialisation into device_init() Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 09/15] drm/panfrost: Warn of RPM get failure at device remove time Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 10/15] drm/panfrost: Add debugfs knob for manually triggering a GPU reset Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 11/15] drm/panfrost: Move perfcnt GPU disable sequence into a helper Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 12/15] drm/panfrost: Skip cache flush/invalidate when enabling perfcnt Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 13/15] drm/panfrost: Avoid cache flush after perfcnt sample in fully coherent systems Adrián Larumbe
2026-10-12 15:05   ` sashiko-bot [this message]
2026-10-12 14:48 ` [PATCH v15 14/15] drm/panfrost: Introduce a reset lock Adrián Larumbe
2026-10-12 14:48 ` [PATCH v15 15/15] drm/panfrost: Fix races between perfcnt and reset sequence Adrián Larumbe
2026-10-12 15:03   ` sashiko-bot

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=sashiko-outbox-168312@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=adrian.larumbe@collabora.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=sashiko-reviews@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.