Bug 103080

Summary: DC driver from drm-next-4.15-dc branch (upstream PR) not working
Product: DRI Reporter: Carlo Caione <carlo>
Component: DRM/AMDgpuAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact:
Severity: normal    
Priority: medium CC: alexdeucher, fdsfgs, harry.wentland
Version: unspecified   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
drm-next-4.15-dc log
none
CONFIG_DRM_AMD_DC_PRE_VEGA kernel Oops none

Description Carlo Caione 2017-10-03 16:53:04 UTC
Created attachment 134642 [details]
drm-next-4.15-dc log

Hardware is a new laptops that has the AMD Bristol Ridge chipset with this SoC:

AMD A10-9620P RADEON R5, 10 COMPUTE CORES 4C+6G

00:01.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 9874 (rev ca)
03:00.0 Display controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 699f (rev ff)

I tried the DC support using the https://cgit.freedesktop.org/~agd5f/linux/log/?h=drm-next-4.15-dc branch and the firmware from https://people.freedesktop.org/~agd5f/radeon_ucode/polaris/ but I keep having no audio at all for HDMI.

In attachment the log.

The only way to have the DC and the HDMI audio output working is by using the branch https://cgit.freedesktop.org/~agd5f/linux/log/?h=amd-staging-4.13 and reverting both https://lists.freedesktop.org/archives/amd-gfx/2017-May/009720.html and https://cgit.freedesktop.org/~agd5f/linux/commit/?id=8f5ed2ed31944eea5c2746ec66417afac043eabd (the first one to have the HDMI output working at all, the second one to have the audio working fine).

Any idea?
Comment 1 Harry Wentland 2017-10-03 17:28:30 UTC
Hi Carlo,

please pass amdgpu.dc=1 module parameter or set CONFIG_DRM_AMD_DC_PRE_VEGA flag at build.

With this branch we decided to only enable it by default for Vega ASICs. With anything older than Vega you'll have to force enable it.

Harry
Comment 2 Carlo Caione 2017-10-03 17:59:54 UTC
Created attachment 134643 [details]
CONFIG_DRM_AMD_DC_PRE_VEGA kernel Oops

(In reply to Harry Wentland from comment #1)
> Hi Carlo,
> 
> please pass amdgpu.dc=1 module parameter or set CONFIG_DRM_AMD_DC_PRE_VEGA
> flag at build.

This is interesting. When compiling with CONFIG_DRM_AMD_DC_PRE_VEGA I get a kernel Oops when connecting the HDMI cable, see the attached log.

But (good news) if I revert https://cgit.freedesktop.org/~agd5f/linux/commit/?id=8f5ed2ed31944eea5c2746ec66417afac043eabd I have no kernel Oops and HDMI works fine. So something is definitely wrong with that commit.
Comment 3 Harry Wentland 2017-10-04 21:05:41 UTC
We've got a fix incoming in a day or two for that kernel oops with the next set of DC patches.
Comment 4 Carlo Caione 2017-10-05 07:40:27 UTC
Great, thank you.

Any special reason why you decided to disable the DC support for pre-VEGA ASICs?
Comment 5 Harry Wentland 2017-10-05 14:48:26 UTC
We want to avoid any regression since pre-Vega is already supported by non-DC. Generally I have a lot more faith in DC but there could always be corner cases. Vega is a new ASIC so by definition there won't be regressions (unless we mess with non-display stuff).
Comment 6 Daniel Drake 2017-10-06 02:20:48 UTC
Harry, HDMI audio is not working for us on the non-DC driver on this hardware (AMD A10-9620P RADEON R5, 10 COMPUTE CORES 4C+6G). Are you saying non-DC should be working fine there?
Comment 7 Dave Airlie 2017-10-06 02:25:28 UTC
Daniel,

not working HDMI audio is a missing feature, not a regression. Not enabling dc by default on older GPUs is to avoid regressions in working features. After some testing it will get enabled.
Comment 8 Daniel Drake 2017-10-06 02:49:23 UTC
Got it. We have a number of affected systems here, so we will help with testing efforts, reporting bugs and successes.
Comment 9 Martin Peres 2019-11-19 08:24:22 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/242.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.