All of lore.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Maxime Ripard" <mripard@kernel.org>
Cc: dri-devel@lists.freedesktop.org
Subject: Re: [PATCH v12 3/9] drm/atomic: Allow filling a commit with pristine object states
Date: Fri, 20 Nov 2026 13:38:17 +0000	[thread overview]
Message-ID: <sashiko-outbox-225256@kernel.org> (raw)
In-Reply-To: <20261120-drm-reset-state-flag-v12-3-42d63d7417fc@kernel.org>

Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [High] drm/atomic: lost driver state in
  drm_atomic_commit_fill_with_defaults()
--

commit 5edb49cfe4c7062b503d63439cf904e87048fe2c
Author: Maxime Ripard <mripard@kernel.org>

drm/atomic: Allow filling a commit with pristine object states

This commit adds drm_atomic_commit_fill_with_defaults() to populate an
atomic commit with freshly created default object states via the
atomic_create_state() hooks while carrying over adjusted_mode and
best_encoder.

> diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
> index 7a77051096c91..43df13aaef1eb 100644
> --- a/drivers/gpu/drm/drm_atomic.c
> +++ b/drivers/gpu/drm/drm_atomic.c
> @@ -1649,6 +1649,327 @@ bool drm_atomic_implements_create_state(struct drm_device *dev)
>  }
>  EXPORT_SYMBOL(drm_atomic_implements_create_state);
>  
[ ... ]
> +int drm_atomic_commit_fill_with_defaults(struct drm_file *file_priv,
> +					 struct drm_atomic_commit *commit)
> +{
[ ... ]
> +	drm_for_each_crtc(crtc, dev) {
> +		struct drm_crtc_state *crtc_state;
[ ... ]
> +		crtc_state = crtc->funcs->atomic_create_state(crtc);
> +		if (IS_ERR(crtc_state))
> +			return PTR_ERR(crtc_state);
[ ... ]
> +		drm_mode_copy(&crtc_state->adjusted_mode,
> +			      &crtc->state->adjusted_mode);
> +
> +		/*
> +		 * The planes come back with the default zpos, which the
> +		 * atomic helpers only normalize when the planes bound to a
> +		 * CRTC or one of their zpos changed. Re-applying the same
> +		 * configuration on top of the reset wouldn't trigger
> +		 * either, so ask for a normalization explicitly.
> +		 */
> +		crtc_state->zpos_changed = true;
> +
> +		drm_atomic_commit_set_crtc_state(commit, crtc, crtc_state);
> +	}
> +
> +	drm_for_each_plane(plane, dev) {
> +		struct drm_plane_state *plane_state;
[ ... ]
> +		plane_state = plane->funcs->atomic_create_state(plane);
> +		if (IS_ERR(plane_state))
> +			return PTR_ERR(plane_state);
[ ... ]
> +		drm_atomic_commit_set_plane_state(commit, plane, plane_state);
> +	}
> +
> +	ret = drm_modeset_lock(&config->connection_mutex, commit->acquire_ctx);
> +	if (ret)
> +		return ret;
> +
> +	drm_connector_list_iter_begin(dev, &conn_iter);
> +	drm_for_each_connector_iter(connector, &conn_iter) {
> +		struct drm_connector_state *connector_state;
[ ... ]
> +		connector_state = connector->funcs->atomic_create_state(connector);
> +		if (IS_ERR(connector_state)) {
> +			drm_connector_list_iter_end(&conn_iter);
> +			ret = PTR_ERR(connector_state);
> +			return ret;
> +		}
> +
> +		/*
> +		 * best_encoder is a routing information maintained by the
> +		 * atomic helpers rather than a userspace property, and it
> +		 * is only carried over to a new state when that state is a
> +		 * duplicate of the current one. Since
> +		 * update_connector_routing() uses it to detect the commits
> +		 * that keep the connector routing unchanged, carry the
> +		 * encoder currently in use over so that a reset on which
> +		 * userspace re-applies the exact same configuration doesn't
> +		 * require a modeset.
> +		 */
> +		connector_state->best_encoder = connector->state->best_encoder;

[Severity: High]
Does building the commit from atomic_create_state() while carrying over
adjusted_mode and best_encoder lose driver-private state in drivers where
drm_atomic_implements_create_state() is true?

When userspace re-applies the current mode and connector routing on top of
a reset, carrying over adjusted_mode and best_encoder keeps mode_changed
and connectors_changed false, while the new CRTC, plane, and connector
states have their driver-private fields zero-initialized instead of copied
from the current state.

In msm/mdp5 (which implements mdp5_crtc_create_state(),
mdp5_plane_create_state(), and drm_bridge_connector_create_state()),
mdp5_plane_atomic_check_with_state() releases the old hardware pipes from
mdp5_state->hwpipe and mdp5_state->r_hwpipe on the new plane state.
Because mdp5_plane_create_state() leaves those pointers NULL, resetting an
active plane passes NULL to mdp5_pipe_release() and permanently leaks the
assigned pipes in mdp5_global_state->hwpipe.hwpipe_to_plane[].  If a reset
keeps a CRTC active with no visible planes, mdp5_crtc_atomic_check() also
returns before mdp5_crtc_setup_pipeline(), leaving
mdp5_cstate->pipeline.mixer NULL and causing a NULL pointer dereference in
blend_setup() during mdp5_crtc_atomic_flush().

In nouveau dispnv50 (nv50_head_create_state(), nv50_wndw_create_state(),
and nouveau_conn_atomic_create_state()), skipping a modeset leaves
asyh->mode, asyh->view, asyh->or, and asyh->wndw.owned zeroed in the CRTC
state because nv50_outp_atomic_check() and nv50_head_atomic_check() only
recompute them when mode_changed or connectors_changed is set.  On planes
kept enabled with the same framebuffer, nv50_wndw_atomic_check_acquire()
skips populating asyw->image when modeset is false, causing
curs507a_acquire() to reject cursor commits with -EINVAL (asyw->image.w ==
0) and zeroing asyw->ntfy.handle, asyw->sema.handle, and asyw->xlut.handle
so they are not cleaned up on later disables.  Resetting a modified
connector property back to its default without a modeset also leaves
asyc->set.mask at 0 because nouveau_conn_atomic_set_property() compares
against the newly created default state, skipping hardware reprogramming.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261120-drm-reset-state-flag-v12-0-42d63d7417fc@kernel.org?part=3

  reply	other threads:[~2026-11-20 13:38 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-20 13:10 [PATCH v12 0/9] drm: Add DRM_MODE_ATOMIC_RESET flag Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 1/9] drm/lease: Add a helper to tell whether a lease was revoked Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 2/9] drm/atomic: Add drm_atomic_implements_create_state() helper Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 3/9] drm/atomic: Allow filling a commit with pristine object states Maxime Ripard
2026-11-20 13:38   ` sashiko-bot [this message]
2026-11-20 13:10 ` [PATCH v12 4/9] drm/tests: atomic: Add tests for drm_atomic_commit_fill_with_defaults() Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 5/9] drm/amd/display: Take over the current stream and dc_state Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 6/9] drm/vc4: Carry the HVS channel over to created CRTC states Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 7/9] drm/msm/dpu1: Carry the mixers " Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 8/9] drm/atomic-uapi: Add DRM_MODE_ATOMIC_RESET flag Maxime Ripard
2026-11-20 13:10 ` [PATCH v12 9/9] drm/vkms: Add driver-specific plane property for testing Maxime Ripard

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-225256@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=mripard@kernel.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.