[SRU][N][PATCH 1/1] drm/amd/display: Fix out-of-bound accesses

Cengiz Can cengiz.can at canonical.com
Wed Sep 9 16:48:20 UTC 2026


From: Alex Hung <alex.hung at amd.com>

[WHAT & HOW]
hpo_stream_to_link_encoder_mapping has size MAX_HPO_DP2_ENCODERS(=4),
but location can have size up to 6. As a result, it is necessary to
check location against MAX_HPO_DP2_ENCODERS.

Similiarly, disp_cfg_stream_location can be used as an array index which
should be 0..5, so the ASSERT's conditions should be less without equal.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3904
Reviewed-by: Austin Zheng <Austin.Zheng at amd.com>
Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira at amd.com>
Signed-off-by: Alex Hung <alex.hung at amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
(backported from commit 8adbb2a98b00926315fd513b5fe2596b5716b82d)
[bot_kybele: Dropped the dml21_translation_helper.c hunk (that subsystem/file is
 absent from this tree) and kept HEAD's is_dp2p0_output_encoder(pipe) logic for
 the output-encoder case since this tree lacks
 hpo_stream_to_link_encoder_mapping and has no out-of-bound access there; the
 two ASSERT <= to < fixes in dml2_translation_helper.c applied cleanly.]
CVE-2025-21985
Assisted-by: kybele:claude-opus-4.8
Signed-off-by: Cengiz Can <cengiz.can at canonical.com>
---
 drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c
index 33a628bc29ba..330ac7097f48 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c
@@ -1105,7 +1105,7 @@ void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_stat
 		if (disp_cfg_stream_location < 0)
 			disp_cfg_stream_location = dml_dispcfg->num_timings++;
 
-		ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
+		ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location < __DML2_WRAPPER_MAX_STREAMS_PLANES__);
 
 		populate_dml_timing_cfg_from_stream_state(&dml_dispcfg->timing, disp_cfg_stream_location, context->streams[i]);
 		populate_dml_output_cfg_from_stream_state(&dml_dispcfg->output, disp_cfg_stream_location, context->streams[i], current_pipe_context);
@@ -1141,7 +1141,7 @@ void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_stat
 				if (disp_cfg_plane_location < 0)
 					disp_cfg_plane_location = dml_dispcfg->num_surfaces++;
 
-				ASSERT(disp_cfg_plane_location >= 0 && disp_cfg_plane_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
+				ASSERT(disp_cfg_plane_location >= 0 && disp_cfg_plane_location < __DML2_WRAPPER_MAX_STREAMS_PLANES__);
 
 				populate_dml_surface_cfg_from_plane_state(dml2->v20.dml_core_ctx.project, &dml_dispcfg->surface, disp_cfg_plane_location, context->stream_status[i].plane_states[j]);
 				populate_dml_plane_cfg_from_plane_state(
-- 
2.53.0




More information about the kernel-team mailing list