]> git.ipfire.org Git - thirdparty/kernel/linux.git/commitdiff
drm/amd/display: Fix 8K Mode Not Parsed by EDID
authorFangzhi Zuo <Jerry.Zuo@amd.com>
Wed, 24 Jun 2026 20:54:30 +0000 (16:54 -0400)
committerAlex Deucher <alexander.deucher@amd.com>
Fri, 17 Jul 2026 21:37:33 +0000 (17:37 -0400)
[why]
The 8K120/8K240 timings live in DisplayID extension blocks 2 and 3
of this EDID. The EDID is a 4-block (512-byte) HDMI 2.1 EDID
that uses HF-EEODB.
drm core reads and parses this correctly, but amdgpu rebuilds its own copy.
Only 2 of 4 blocks were copied into sink->dc_edid, that leads to
drm_edid_connector_add_modes() never sees blocks 2 and 3.

[how]
Directly populate edid_blob_ptr with a blob whose length is the full,
and HF-EEODB-aware size.

Reviewed-by: Sun peng (Leo) Li <sunpeng.li@amd.com>
Signed-off-by: Fangzhi Zuo <Jerry.Zuo@amd.com>
Signed-off-by: George Zhang <george.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 11a90eaf5c808ba800249dda0d481c35d0888589)

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c

index c6f94eb71ffadcef2e8b8e5d65468b78c42b6fed..6be7f6edd0b29366b3ce906e525f058a68db8303 100644 (file)
@@ -1201,11 +1201,25 @@ enum dc_edid_status dm_helpers_read_local_edid(
                        continue;
 
                edid = drm_edid_raw(drm_edid); // FIXME: Get rid of drm_edid_raw()
-               if (!edid ||
-                   edid->extensions >= sizeof(sink->dc_edid.raw_edid) / EDID_LENGTH)
+               /*
+                * Use the length of the EDID property blob populated by
+                * drm_edid_connector_update() above. It reflects the true number
+                * of EDID blocks, including any HDMI Forum EDID Extension Override
+                * Data Block (HF-EEODB) count, which the raw byte 0x7e extension
+                * count can hide (e.g. HDMI 8K sinks).
+                */
+               if (!edid || !connector->edid_blob_ptr ||
+                   connector->edid_blob_ptr->length > sizeof(sink->dc_edid.raw_edid))
                        return EDID_BAD_INPUT;
 
-               sink->dc_edid.length = EDID_LENGTH * (edid->extensions + 1);
+               /*
+                * FIXME: amdgpu_dm today does not consider the HF-EEODB, which
+                * may contain additional mode info for sinks. This is a
+                * workaround until dc_edid is refactored out from DC into
+                * amdgpu_dm's ownership, allowing amdgpu_dm to use drm_edid
+                * directly
+                */
+               sink->dc_edid.length = connector->edid_blob_ptr->length;
                memmove(sink->dc_edid.raw_edid, (uint8_t *)edid, sink->dc_edid.length);
 
                /* We don't need the original edid anymore */