amdgpu hangs on boot or shutdown on AMD Raven Ridge CPU (Engineer Sample)

[Chris Chiu]

On Thu, Feb 1, 2018 at 9:13 PM, Chris Chiu <chiu at endlessm.com> wrote:
> On Thu, Feb 1, 2018 at 12:08 AM, Harry Wentland <harry.wentland at amd.com> wrote:
>> On 2018-01-31 09:31 AM, Chris Chiu wrote:
>>> Hi,
>>>     We are working with new laptops that have the AMD Ravenl Ridge
>>> chipset with this `/proc/cpuinfo`
>>> https://gist.github.com/mschiu77/b06dba574e89b9a30cf4c450eaec49bc
>>>
>>>     With the latest kernel 4.15, there're lots of different
>>> panics/oops during boot so no chance to get into X. It also happens
>>> during shutdown. Then I tried to build kernel from
>>> git://people.freedesktop.org/~agd5f/linux on branch
>>> amd-staging-drm-next with head on commit "drm: Fix trailing semicolon"
>>> and update the linux-firmware. Things seem to get better, only 1 oops
>>> observed. Here's the oops
>>> https://gist.github.com/mschiu77/1a68f27272b24775b2040acdb474cdd3.
>>
>> Hi Chris,
>>
>> what are the steps to reproduce this oops?
>>
>> Does it reproduce all the time or is it intermittent?
>>
>> Can you send a dmesg with amdgpu.dc_log=1, in addition to drm.debug=0xe?
>>
>> Thanks,
>> Harry
>>
>
> I did nothing special to reproduce the oops. Boot and sometimes it
> just shows blank
> screen but still responds to magic sysrq. So I reboot and take the journal log.
>
> It's intermittent, I ran into it 2 times during 13 reboots.
> The logs are listed as follows
> https://gist.github.com/mschiu77/9307d1ca0acd046cc6817f8cad63d79c
> https://gist.github.com/mschiu77/fa81110f93428721f017cb9fbfd06fbe
>
> One more log here. It enters X OK but after few minutes the display
> went black and
> only a mouse cursor left. But the mouse cursor can't even move. So I do a sysrq
> reboot again.
> The last error is
> ""
> [  636.312759] endless kernel:
> [drm:drm_atomic_helper_wait_for_flip_done [drm_kms_helper]] *ERROR*
> [CRTC:41:crtc-0] flip_done timed out
> [  646.552344] endless kernel:
> [drm:drm_atomic_helper_wait_for_dependencies [drm_kms_helper]] *ERROR*
> [CRTC:41:crtc-0] flip_done timed out
> ""
> full log here https://gist.github.com/mschiu77/c8696e5fefb17bb1c53598214fb4e382
>
> Only 4 times I can login X, blank screen or hangs w/o responding to
> magic sysrq for
> the rest. I took a picture of the only panic although I think it's not
> about amdgpu.
> It's here.
> https://pasteboard.co/H5CUvxk.jpg
>
> Hope they can be helpful.
>
> Chris
>
>>> However, I still get stuck on the following messages during boot very
>>> often
>>> ""
>>> [    4.998241] endless kernel: [drm] amdgpu kernel modesetting enabled.
>>> [    4.998288] endless kernel: checking generic (e0000000 7f0000) vs
>>> hw (e0000000 10000000)
>>> [    4.998289] endless kernel: fb: switching to amdgpudrmfb from EFI VGA
>>> ""
>>> I turned on drm.debug=0xe while booting, but no more information at this point.
>>> Anything I can do at this point?
>>>
>>>     And there's 1 more information may be helpful. Sometimes the
>>> system boots OK with the blank screen, I can't switch to virtual
>>> console, but it did respond to the magic sys-rq key. The dmesg with
>>> drm.debug=0xe is here
>>> https://gist.github.com/mschiu77/291e47b1f07dc52be9461c55c820464c.
>>>
>>>     I'm pretty sure it's due to the amdgpu driver. Because when I boot
>>> with my own kernel which disables the amdgpu driver, all these
>>> symptoms went away. Please suggest anything I can do for this. Thanks
>>>
>>> Chris
>>> _______________________________________________
>>> amd-gfx mailing list
>>> amd-gfx at lists.freedesktop.org
>>> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
>>>

Gentle ping, cheers.

Chris