[Bug 101166] New: option to disable deep color (or set HDMI bandwidth limit)

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Wed May 24 13:09:45 UTC 2017


https://bugs.freedesktop.org/show_bug.cgi?id=101166

            Bug ID: 101166
           Summary: option to disable deep color (or set HDMI bandwidth
                    limit)
           Product: DRI
           Version: unspecified
          Hardware: All
                OS: Linux (All)
            Status: NEW
          Severity: minor
          Priority: medium
         Component: DRM/Intel
          Assignee: intel-gfx-bugs at lists.freedesktop.org
          Reporter: axkibe at gmail.com
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
                CC: intel-gfx-bugs at lists.freedesktop.org

Steps to reproduce the problem:

* have a kernel (~> 4.6 ) where drm is able to do deep color
* have an chip that uses i915 
* have a deep color able TV screen (1080p)
* connect them via a CAT5-bridge (as in our meeting rooms be done) that can
handle 1060p only on 8bit.

Happens all the time.

This setup breaks with and only with Linux laptops with recent kernels, due to
them turning on deep color which seemingly overloads the CAT5 bridge and thus
the screen starts flickering (turning to black every 10-20 seconds for half a
second).

Following "patch" solves the issue for me, since it forcefully disables all
deep color ability for this notebook:

----
linux-4.11.2$ diff -u drivers/gpu/drm/drm_edid.c.org drivers/gpu/drm/drm_edid.c
--- drivers/gpu/drm/drm_edid.c.org    2017-05-24 14:08:19.162452768 +0200
+++ drivers/gpu/drm/drm_edid.c    2017-05-24 14:10:33.149417356 +0200
@@ -3803,6 +3803,7 @@

     if (cea_db_payload_len(hdmi) < 6)
         return;
+    return;

     if (hdmi[6] & DRM_EDID_HDMI_DC_30) {
         dc_bpc = 10;
------

I'd wish there was a module parameter to force this off, other than hacking the
source. (Or I kept overlooking an existing one)

PS: As far I realize, unless this is a more sophisticated EDID bug, at fault
here is that stupid CAT5-bridge that is supposed to modify the EDID to
something it can handle, which it doesn't.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20170524/1e6bf948/attachment.html>


More information about the intel-gfx-bugs mailing list