[Bug 216774] New: On amdgpu the drm subsystem lists additional modes beyond what EDID has available

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=216774

            Bug ID: 216774
           Summary: On amdgpu the drm subsystem lists additional modes
                    beyond what EDID has available
           Product: Drivers
           Version: 2.5
    Kernel Version: 5.15.81
          Hardware: All
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: low
          Priority: P1
         Component: Video(DRI - non Intel)
          Assignee: drivers_video-dri@kernel-bugs.osdl.org
          Reporter: wolfwings@gmail.com
        Regression: No

Created attachment 303357
  --> https://bugzilla.kernel.org/attachment.cgi?id=303357&action=edit
Single-resolution EDID (1080p120) that demonstrates the problem

The attached EDID is one I've used across multiple operating systems, however
on Linux it creates the following additional modes at the DRM level visible
with drm_info outside of X/Wayland from the CLI:

1680x1050@120.00
1280x1024@120.00
1440x900@120.00
1280x800@120.00
1280x720@120.00
1024x768@120.00
800x600@120.00
640x480@120.00

This EDID is being loaded via "drm.edid_firmware=eDP:edid.bin" and is present
in the initramfs, and this happens both with the stock Ubuntu kernel and when I
compiled the latest 5.15.81 kernel from source, so I'm reporting it here.

This might be a misconfiguration or some additional option I need to pass to
the drm subsystem to disable some built-in mode-list it's adding by default, or
even fixed in later kernel versions and I was simply unaware, if so I apologize
for taking up your time!

This EDID is meant to only have a single mode available, 1080p120. Changing the
EDID to use 1080p60 (only involves changing the pixel clock value for the
single mode and EDID checksum) still has the same problem, and looking at the
modes it appears it's just replacing the visible-pixel values with the smaller
values and retaining all the other timing details so these seem to be
auto-generated somewhere I can't find.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.