Bug 199085 - Multiple [drm] Got external EDID base block and 0 extensions messages
Summary: Multiple [drm] Got external EDID base block and 0 extensions messages
Status: RESOLVED WILL_FIX_LATER
Alias: None
Product: Drivers
Classification: Unclassified
Component: Video(DRI - non Intel) (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: drivers_video-dri
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-03-12 09:57 UTC by Stuart Foster
Modified: 2018-04-12 08:40 UTC (History)
0 users

See Also:
Kernel Version: 4.14.25
Subsystem:
Regression: No
Bisected commit-id:


Attachments
dmesg log (224.62 KB, text/plain)
2018-03-12 10:00 UTC, Stuart Foster
Details

Description Stuart Foster 2018-03-12 09:57:56 UTC

    
Comment 1 Stuart Foster 2018-03-12 10:00:56 UTC
Created attachment 274677 [details]
dmesg log

Started getting these "[drm] Got external EDID base block and 0 extensions..." messages with recent 4.14 kernels. Is this a bug or a configuration issue on my part ?
Comment 2 Stuart Foster 2018-03-12 13:45:14 UTC
Also present with recent 4.15 kernels.
Comment 3 Jani Nikula 2018-03-19 08:31:58 UTC
That's an info level message from your use of the edid_firmware option. Been there since 2012.
Comment 4 Stuart Foster 2018-03-19 08:55:38 UTC
Yes thanks, the message is not new to me but the repeated loading of the firmware edid blob does appear new to me. Why the repeated loading of the firmware blob once should be sufficient ??
Comment 5 Stuart Foster 2018-04-01 10:42:02 UTC
If I remove drm_kms_helper.edid_firmware=edid/hp_lp2065.bin from the kernel boot line the messages go away and as far as I can seen no edid data is loaded by the kernel is this the correct behaviour ?

thanks
Comment 6 Jani Nikula 2018-04-11 15:01:09 UTC
The firmware EDID handling was moved at a lower level in the stack to be a more transparent replacement for the display EDID. Without this, many features were erroneously parsed from the display EDID, which was in conflict with the firmware EDID.

This means that any time the driver would read the EDID, it gets loaded as the firmware EDID.

It's arguably something that should be improved in the driver to not keep reading the EDID so many times over, whether it's display or firmware EDID.
Comment 7 Stuart Foster 2018-04-12 08:40:09 UTC
Thanks for the explanation, hopefully there will be an improvement in the future.

Note You need to log in before you can comment on or make changes to this bug.