Nouveau - Nov 2024 - [REGRESSION] GM20B pmu timeout

[CCing Danilo, who committed the culprit]

On 04.11.24 13:11, Diogo Ivo wrote:
> On Tue Oct 15, 2024 at 7:13 PM WEST, Linux regression tracking (Thorsten
Leemhuis) wrote:
>> On 10.10.24 15:32, Diogo Ivo wrote:
>>>
>>> Somewhere between 6.11-rc4 and 6.11-rc5 the following error message
is displayed
>>> when trying to initialize a nvc0_screen on the Tegra X1's
GM20B:
>>>
>>> [ 34.431210] nouveau 57000000.gpu: pmu:hpq: timeout waiting for
queue ready
>>> [ 34.438145] nouveau 57000000.gpu: gr: init failed, -110
>>> nvc0_screen_create:1075 - Error allocating PGRAPH context for M2MF:
-110
>>> failed to create GPU screen
>>
>> Thx for the report. Hmmm. No reply so far. :-/
>>
>> Diogo, maybe report this here as well:
>> https://gitlab.freedesktop.org/drm/nouveau/-/issues/
>>
>> Afterwards drop a link to the ticket here. Reporting nouveau issues via
>> email should work, but maybe you have more luck there.
> 
> Gentle ping on this topic.
@airlied: I missed this earlier (and I assume you did, too), but seems
this regression is caused by the following change of yours:
9b340aeb26d50e ("nouveau/firmware: use dma non-coherent allocator")

For the full report see the start of the thread here:
https://lore.kernel.org/all/20241010133253.30311-1-diogo.ivo at
tecnico.ulisboa.pt/

Diogo submitted an issue to the nouveau tracker as well:
https://gitlab.freedesktop.org/drm/nouveau/-/issues/391

Another user just joined and reported seeing similar problems.

And I just noticed there is another regression report bisected to the
same change:
https://gitlab.freedesktop.org/drm/nouveau/-/issues/388

Two users there are affected by the change; one of them in the 6.6.y
series and a revert there was able to fix this.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker'
hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

#regzbot poke

On Tue, 12 Nov 2024 at 22:34, Linux regression tracking (Thorsten
Leemhuis) <regressions at leemhuis.info> wrote:
>
> [CCing Danilo, who committed the culprit]
>
> On 04.11.24 13:11, Diogo Ivo wrote:
> > On Tue Oct 15, 2024 at 7:13 PM WEST, Linux regression tracking
(Thorsten Leemhuis) wrote:
> >> On 10.10.24 15:32, Diogo Ivo wrote:
> >>>
> >>> Somewhere between 6.11-rc4 and 6.11-rc5 the following error
message is displayed
> >>> when trying to initialize a nvc0_screen on the Tegra X1's
GM20B:
> >>>
> >>> [ 34.431210] nouveau 57000000.gpu: pmu:hpq: timeout waiting
for queue ready
> >>> [ 34.438145] nouveau 57000000.gpu: gr: init failed, -110
> >>> nvc0_screen_create:1075 - Error allocating PGRAPH context for
M2MF: -110
> >>> failed to create GPU screen
> >>
> >> Thx for the report. Hmmm. No reply so far. :-/
Apologies for the delay and thanks to Thorsten for bringing it up again,

Does the attached patch fix it?

if you say it does I'll send it out properly.

Dave.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-nouveau-sync-dma-after-setup-is-called.patch
Type: text/x-patch
Size: 1406 bytes
Desc: not available
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20241113/c1f7da36/attachment.bin>