Skip to content

drm: backport drm_bridge changes#765

Open
quic-vishsain wants to merge 2 commits into
qualcomm-linux:qcom-6.18.yfrom
quic-vishsain:next_bridge_backport
Open

drm: backport drm_bridge changes#765
quic-vishsain wants to merge 2 commits into
qualcomm-linux:qcom-6.18.yfrom
quic-vishsain:next_bridge_backport

Conversation

@quic-vishsain

@quic-vishsain quic-vishsain commented Jun 23, 2026

Copy link
Copy Markdown

lucaceresoli and others added 2 commits June 24, 2026 01:52
Many bridge drivers store a next_bridge pointer in their private data and
use it for attach and sometimes other purposes. This is going to be risky
when bridge hot-unplug is used.

Considering this example scenario:

  1. pipeline: encoder --> bridge A --> bridge B --> bridge C
  2. encoder takes a reference to bridge B
  3. bridge B takes a next_bridge reference to bridge C
  4. encoder calls (bridge B)->b_foo(), which in turns references
     next_bridge, e.g.:

       b_foo() {
           bar(b->next_bridge);
       }

If bridges B and C are removed, bridge C can be freed but B is still
allocated because the encoder holds a reference to B. So when step 4
happens, 'b->next-bridge' would be a use-after-free.

Calling drm_bridge_put() in the B bridge .remove function does not solve
the problem as it leaves a (potentially long) risk window between B removal
and the final deallocation of B. A safe moment to put the B reference is in
__drm_bridge_free(), when the last reference has been put. This can be done
by drivers in the .destroy func. However to avoid the need for so many
drivers to implement a .destroy func, just offer a next_bridge pointer to
all bridges that is automatically put it in __drm_bridge_free(), exactly
when the .destroy func is called.

Suggested-by: Maxime Ripard <mripard@kernel.org>
Link: https://lore.kernel.org/all/20251201-thick-jasmine-oarfish-1eceb0@houat/
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Link: https://patch.msgid.link/20251216-drm-bridge-alloc-getput-drm_of_find_bridge-v3-6-b5165fab8058@bootlin.com
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Vishnu Saini <vishnu.saini@oss.qualcomm.com>
Having only a single set of callbacks, hdmi_clear_infoframe and
hdmi_write_infoframe, bridge drivers don't have an easy way to signal to
the DRM framework, which InfoFrames are actually supported by the
hardware and by the driver and which are not. Also, it makes it
extremely easy for HDMI bridge drivers to skip implementing the
seemingly required InfoFrames (e.g. HDMI VSI). Last, but not least,
those callbacks take a single 'type' parameter, which makes it
impossible to implement support for multiple VSIs (which will be
required once we start working on HDMI Forum VSI).

Split the callbacks into a per-InfoFrame-kind pairs, letting the bridge
drivers actually signal supported features. The implementation follows
the overall drm_bridge design, where the bridge has a single
drm_bridge_funcs implementation and signals, which functions are to be
called using the drm_bridge->ops flags.

The AVI and HDMI VSI are assumed to be required for a normal HDMI
operation (with the drivers getting a drm_warn_once() stub
implementation if one is missing). The Audio InfoFrame is handled by the
existing DRM_BRIDGE_OP_HDMI_AUDIO, while the SPD and HDR DRM InfoFrames
got new drm_bridge_ops values.

Acked-by: Maxime Ripard <mripard@kernel.org>
Link: https://patch.msgid.link/20260107-limit-infoframes-2-v4-5-213d0d3bd490@oss.qualcomm.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Vishnu Saini <vishnu.saini@oss.qualcomm.com>
@quic-vishsain quic-vishsain requested review from a team, Komal-Bajaj, aiquny and shashim-quic June 23, 2026 20:36
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants