From: Philipp Stanner <phasta@mailbox.org>
To: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>,
Danilo Krummrich <dakr@kernel.org>
Cc: phasta@kernel.org, "Lyude Paul" <lyude@redhat.com>,
"David Airlie" <airlied@gmail.com>,
"Simona Vetter" <simona@ffwll.ch>,
"Matthew Brost" <matthew.brost@intel.com>,
"Christian König" <ckoenig.leichtzumerken@gmail.com>,
"Maarten Lankhorst" <maarten.lankhorst@linux.intel.com>,
"Maxime Ripard" <mripard@kernel.org>,
"Thomas Zimmermann" <tzimmermann@suse.de>,
dri-devel@lists.freedesktop.org, nouveau@lists.freedesktop.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/5] drm/sched: Warn if pending list is not empty
Date: Tue, 22 Apr 2025 14:21:07 +0200 [thread overview]
Message-ID: <9566a923b606e16f7725c79bd8e0a2c7c80daf9e.camel@mailbox.org> (raw)
In-Reply-To: <d3c0f721-2d19-4a1c-a086-33e8d6bd7be6@igalia.com>
On Tue, 2025-04-22 at 13:07 +0100, Tvrtko Ursulin wrote:
>
> On 22/04/2025 12:13, Danilo Krummrich wrote:
> > On Tue, Apr 22, 2025 at 11:39:11AM +0100, Tvrtko Ursulin wrote:
> > > Question I raised is if there are other drivers which manage to
> > > clean up
> > > everything correctly (like the mock scheduler does), but trigger
> > > that
> > > warning. Maybe there are not and maybe mock scheduler is the only
> > > false
> > > positive.
> >
> > So far the scheduler simply does not give any guideline on how to
> > address the
> > problem, hence every driver simply does something (or nothing,
> > effectively
> > ignoring the problem). This is what we want to fix.
> >
> > The mock scheduler keeps it's own list of pending jobs and on tear
> > down stops
> > the scheduler's workqueues, traverses it's own list and eventually
> > frees the
> > pending jobs without updating the scheduler's internal pending
> > list.
> >
> > So yes, it does avoid memory leaks, but it also leaves the
> > schedulers internal
> > structures with an invalid state, i.e. the pending list of the
> > scheduler has
> > pointers to already freed memory.
> >
> > What if the drm_sched_fini() starts touching the pending list? Then
> > you'd end up
> > with UAF bugs with this implementation. We cannot invalidate the
> > schedulers
> > internal structures and yet call scheduler functions - e.g.
> > drm_sched_fini() -
> > subsequently.
> >
> > Hence, the current implementation of the mock scheduler is
> > fundamentally flawed.
> > And so would be *every* driver that still has entries within the
> > scheduler's
> > pending list.
> >
> > This is not a false positive, it already caught a real bug -- in
> > the mock
> > scheduler.
>
> To avoid furher splitting hairs on whether real bugs need to be able
> to
> manifest or not, lets move past this with a conclusion that there are
> two potential things to do here:
>
> First one is to either send separately or include in this series
> something like:
>
> diff --git a/drivers/gpu/drm/scheduler/tests/mock_scheduler.c
> b/drivers/gpu/drm/scheduler/tests/mock_scheduler.c
> index f999c8859cf7..7c4df0e890ac 100644
> --- a/drivers/gpu/drm/scheduler/tests/mock_scheduler.c
> +++ b/drivers/gpu/drm/scheduler/tests/mock_scheduler.c
> @@ -300,6 +300,8 @@ void drm_mock_sched_fini(struct
> drm_mock_scheduler
> *sched)
> drm_mock_sched_job_complete(job);
> spin_unlock_irqrestore(&sched->lock, flags);
>
> + drm_sched_fini(&sched->base);
> +
> /*
> * Free completed jobs and jobs not yet processed by the DRM
> scheduler
> * free worker.
> @@ -311,8 +313,6 @@ void drm_mock_sched_fini(struct
> drm_mock_scheduler
> *sched)
>
> list_for_each_entry_safe(job, next, &list, link)
> mock_sched_free_job(&job->base);
> -
> - drm_sched_fini(&sched->base);
> }
>
> /**
>
> That should satisfy the requirement to "clear" memory about to be
> freed
> and be 100% compliant with drm_sched_fini() kerneldoc (guideline b).
>
> But the new warning from 3/5 here will still be there AFAICT and
> would
> you then agree it is a false positive?
>
> Secondly, the series should modify all drivers (including the unit
> tests) which are known to trigger this false positive.
There is no false positive. I just stated that in my other answer.
That said, I agree that the unit tests should be a second user of this
code in addition to Nouveau. They then should also provide the
fence_context_kill() callback, though.
And it seems we also agree that the memory leaks must be solved
centrally in drm_sched_fini().
P.
>
> Regards,
>
> Tvrtko
>
next prev parent reply other threads:[~2025-04-22 12:21 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-07 15:22 [PATCH 0/5] drm/sched: Fix memory leaks in drm_sched_fini() Philipp Stanner
2025-04-07 15:22 ` [PATCH 1/5] drm/sched: Fix teardown leaks with waitqueue Philipp Stanner
2025-04-17 7:49 ` Philipp Stanner
2025-04-07 15:22 ` [PATCH 2/5] drm/sched: Prevent teardown waitque from blocking too long Philipp Stanner
2025-04-07 15:22 ` [PATCH 3/5] drm/sched: Warn if pending list is not empty Philipp Stanner
2025-04-17 7:45 ` Philipp Stanner
2025-04-17 11:27 ` Tvrtko Ursulin
2025-04-17 12:11 ` Danilo Krummrich
2025-04-17 14:20 ` Tvrtko Ursulin
2025-04-17 14:48 ` Danilo Krummrich
2025-04-17 16:08 ` Tvrtko Ursulin
2025-04-17 17:07 ` Danilo Krummrich
2025-04-22 6:06 ` Philipp Stanner
2025-04-22 10:39 ` Tvrtko Ursulin
2025-04-22 11:13 ` Danilo Krummrich
2025-04-22 12:00 ` Philipp Stanner
2025-04-22 13:25 ` Tvrtko Ursulin
2025-04-22 12:07 ` Tvrtko Ursulin
2025-04-22 12:21 ` Philipp Stanner [this message]
2025-04-22 12:32 ` Danilo Krummrich
2025-04-22 13:39 ` Tvrtko Ursulin
2025-04-22 13:46 ` Philipp Stanner
2025-04-22 14:08 ` Danilo Krummrich
2025-04-22 14:16 ` Philipp Stanner
2025-04-22 14:52 ` Danilo Krummrich
2025-04-23 7:34 ` Tvrtko Ursulin
2025-04-23 8:48 ` Danilo Krummrich
2025-04-23 10:10 ` Tvrtko Ursulin
2025-04-23 10:26 ` Danilo Krummrich
2025-04-07 15:22 ` [PATCH 4/5] drm/nouveau: Add new callback for scheduler teardown Philipp Stanner
2025-04-07 15:22 ` [PATCH 5/5] drm/nouveau: Remove waitque for sched teardown Philipp Stanner
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=9566a923b606e16f7725c79bd8e0a2c7c80daf9e.camel@mailbox.org \
--to=phasta@mailbox.org \
--cc=airlied@gmail.com \
--cc=ckoenig.leichtzumerken@gmail.com \
--cc=dakr@kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lyude@redhat.com \
--cc=maarten.lankhorst@linux.intel.com \
--cc=matthew.brost@intel.com \
--cc=mripard@kernel.org \
--cc=nouveau@lists.freedesktop.org \
--cc=phasta@kernel.org \
--cc=simona@ffwll.ch \
--cc=tvrtko.ursulin@igalia.com \
--cc=tzimmermann@suse.de \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®