Hi Dmitry,

As I stated below, what I observed with the two patches were these kernel messages, which did not seem like it was setting a single GPU frequency (133MHz, 315MHz and 624MHz):
> I see some messages print:
> [   34.745966] adreno b00000.gpu: GPU setting to 133000000
> [   34.754545] msm_hdmi_hdcp_read_validate_aksv: AKSV QFPROM doesn't have 20 1's, 20 0's
> [   34.754630] msm_hdmi_hdcp_read_validate_aksv: QFPROM AKSV chk failed (AKSV=6800000000)
> [   34.761428] [drm:mdp5_irq_error_handler] *ERROR* errors: 40000000
> [   34.769257] msm_hdmi_hdcp_auth_prepare: ASKV validation failed
> [   34.775762] msm_hdmi_hdcp_auth_work: auth prepare failed -524
> [   34.781566] msm_hdmi_hdcp_auth_work: hdcp is not supported
> [   34.998064] adreno b00000.gpu: GPU setting to 315000000
> [   35.262050] adreno b00000.gpu: GPU setting to 624000000

Are you saying these frequencies are not getting set?
Maybe the patches got mixed up? Can you create a patch from our repo since you have access?  Then I can try that patch?

Thanks,
Kim



-----Original Message-----
From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Sent: Wednesday, April 5, 2023 5:07 AM
To: Kim Steiner <kim.steiner@sightlineapplications.com>; Jordan Holt <jordan.holt@sightlineapplications.com>
Cc: Paul Neuhardt <paul.neuhardt@linaro.org>; sightlineapplications@lists.linaro.org
Subject: Re: Low temperature GPU hang testing

Hi Kim

On 30/03/2023 19:40, Kim Steiner wrote:
> Hi Dmitry,


Please excuse me for not responding earlier. Somehow this email got into a different folder and thus I missed it. Please accept my humble apologies.

> Anything new for us to test?  Are you still working on a single GPU frequency patch for us?

Maybe there was some kind of misunderstanding. Does this last patch I have sent you fix the issue on two boards that were resetting with the previous single-frequency patch? Which speed bin do they report (look for the "Detected speed bin" kernel log entry.

 From the software perspective point of view the freq.patch I have send you in the last email should be as good as a the single-freq: it powers on the GPU and then switches to the highest possible frequency, taking into account GPU speed classification, possible overheating, etc. It user the performance governor, so it will not lower the frequency unless required by the CPU/GPU being close to the melting point.

>
> Thanks,
> Kim
>
> -----Original Message-----
> From: Kim Steiner
> Sent: Thursday, March 23, 2023 12:42 PM
> To: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>; Jordan Holt
> <jordan.holt@sightlineapplications.com>
> Cc: Paul Neuhardt <paul.neuhardt@linaro.org>;
> sightlineapplications@lists.linaro.org
> Subject: RE: Low temperature GPU hang testing
>
> Hi Dmitry,
>
> I tested the combined patch (I had to change the dev_info statement because I got a build error) and the board appears to boot ok.
>
> I see some messages print:
> [   34.745966] adreno b00000.gpu: GPU setting to 133000000
> [   34.754545] msm_hdmi_hdcp_read_validate_aksv: AKSV QFPROM doesn't have 20 1's, 20 0's
> [   34.754630] msm_hdmi_hdcp_read_validate_aksv: QFPROM AKSV chk failed (AKSV=6800000000)
> [   34.761428] [drm:mdp5_irq_error_handler] *ERROR* errors: 40000000
> [   34.769257] msm_hdmi_hdcp_auth_prepare: ASKV validation failed
> [   34.775762] msm_hdmi_hdcp_auth_work: auth prepare failed -524
> [   34.781566] msm_hdmi_hdcp_auth_work: hdcp is not supported
> [   34.998064] adreno b00000.gpu: GPU setting to 315000000
> [   35.262050] adreno b00000.gpu: GPU setting to 624000000
>
> Thanks,
> Kim
>
> -----Original Message-----
> From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
> Sent: Tuesday, March 21, 2023 9:38 AM
> To: Kim Steiner <kim.steiner@sightlineapplications.com>; Jordan Holt
> <jordan.holt@sightlineapplications.com>
> Cc: Paul Neuhardt <paul.neuhardt@linaro.org>;
> sightlineapplications@lists.linaro.org
> Subject: Re: Low temperature GPU hang testing
>
> Hi Kim
>
> On 15/03/2023 00:13, Kim Steiner wrote:
>> Hi Dmitry,
>>
>> Anything new to report on your end?  Anything for us to test?  Did you try building the kernel from our repo?
>
> After spending some time testing it locally, I still was not able to reproduce the latest issue you have reported (some of the boards rebooting with the single-frequency patch). Could you please test the attached patch on top of the single-freq patch? I would be very interested in the dmesg output both if the attached patch fixes the issue on your side and if it does not.
>
>>
>> Thanks,
>> Kim
>>
>> -----Original Message-----
>> From: Kim Steiner
>> Sent: Thursday, March 9, 2023 7:57 AM
>> To: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>; Jordan Holt
>> <jordan.holt@sightlineapplications.com>
>> Cc: Paul Neuhardt <paul.neuhardt@linaro.org>;
>> sightlineapplications@lists.linaro.org
>> Subject: RE: Low temperature GPU hang testing
>>
>> Hi Dmitry,
>>
>> I added you to the 5.15Linarokrn repository.
>>
>> git clone
>> https://<username>%40sightlineapplications.com@git7.projectlocker.com
>> / SightLineApplications/5.15Linarokrn.git
>>
>> The username should be the same as for the 4.14 kernel login.
>>
>> Thanks,
>> Kim
>
> --
> With best wishes
> Dmitry

--
With best wishes
Dmitry