Hi Kim,
On 08/02/2023 22:42, Kim Steiner wrote:
Hi Dmitry,
I tried a full recovery to 4.14 on the SLA-4000. Then flashed the latest 5.15 kernel and modules to the board. The board boots just fine. Something must have gotten changed since earlier testing of 5.15 😊
Thanks for the confirmation!
We had a debugging session with Tamotsu-san. He assisted me with the license file and checking of the boards, but we still were not able to enable the HD-SDI camera (switching to the replaced HD-SDI board and camera). I have provided him with the logs of the Sightline software service. Hopefully we can solve the HD-SDI camera issue at some point.
Regarding the load testing.
Could you please compare the following items between 4.14 and 5.15:
- cat /proc/loadavg (this is the load average of the CPUs)
- sudo cat /sys/kernel/debug/dri/0/perf (this outpus current load of the GPU)
- measure the actual frequencies of the CPU clusters:
git clone https://github.com/andersson/debugcc.git cd debugcc make sudo ./msm8996-debugcc -p msm8996 -a -b cpu
Thanks, Kim
-----Original Message----- From: Kim Steiner Sent: Wednesday, January 25, 2023 9:48 AM To: Dmitry Baryshkov dmitry.baryshkov@linaro.org Cc: Jordan Holt jordan.holt@sightlineapplications.com; Paul Neuhardt paul.neuhardt@linaro.org; sightlineapplications@lists.linaro.org Subject: RE: [Sightlineapplications] Re: Project Status and Burn Report
Hi Dmitry,
One other thought in terms of getting the hd-sdi camera to work is to try the analog out connector. It is the yellow connector on the hd-sdi camera cabling. Would you have a way to display this?
Thanks, Kim
-----Original Message----- From: Kim Steiner Sent: Tuesday, January 24, 2023 3:31 PM To: Dmitry Baryshkov dmitry.baryshkov@linaro.org Cc: Jordan Holt jordan.holt@sightlineapplications.com; Paul Neuhardt paul.neuhardt@linaro.org; sightlineapplications@lists.linaro.org Subject: RE: [Sightlineapplications] Re: Project Status and Burn Report
Resending with basically no message then this one:
Hi Dmitry,
In regards to our testing, we have seen a GPU fault crash with our custom board (with 5.15 kernel) during temp testing. On one unit, we have seen the GPU fault crash at 118F (47C) while our custom board is sitting on a desk. The output of the GPU fault crash occurs once and our application continues to run. What do you think of this output?:
[ 156.660355] msm 900000.mdss: CP | opcode error | possible opcode=0x00000000 [ 156.662139] msm 900000.mdss: [drm:a5xx_irq] *ERROR* gpu fault ring 0 fence 5e47 status 800001C1 rb 1820/17e6 ib1 0000000003C1B000/0000 ib2 0000000001483000/0000 [ 156.666297] msm 900000.mdss: [drm:recover_worker] *ERROR* A530: hangcheck recover! [ 156.680832] msm 900000.mdss: [drm:recover_worker] *ERROR* A530: offending task: VTNext_ARM:sq0 (./VTNext_ARM64_Release.out -FULL) [ 156.699196] revision: 530 (5.3.0.2) [ 156.699341] msm 900000.mdss: [drm:a5xx_irq] *ERROR* gpu fault ring 0 fence 5e47 status 800001C1 rb 1820/17e6 ib1 0000000003C1B000/0000 ib2 0000000001483000/0000 [ 156.699833] rb 0: fence: 24134/24135 [ 156.717750] rptr: 6081 [ 156.721300] rb wptr: 6118 [ 156.724047] rb 1: fence: 0/0 [ 156.726786] rptr: 0 [ 156.729795] rb wptr: 0 [ 156.732207] rb 2: fence: 0/0 [ 156.734670] rptr: 0 [ 156.737779] rb wptr: 0 [ 156.740191] rb 3: fence: 0/0 [ 156.742649] rptr: 0 [ 156.745768] rb wptr: 0 [ 156.748186] CP_SCRATCH_REG0: 0 [ 156.750641] CP_SCRATCH_REG1: 0 [ 156.753762] CP_SCRATCH_REG2: 24134 [ 156.756812] CP_SCRATCH_REG3: 0 [ 156.760169] CP_SCRATCH_REG4: 0 [ 156.763222] CP_SCRATCH_REG5: 0 [ 156.766262] CP_SCRATCH_REG6: 0 [ 156.769300] CP_SCRATCH_REG7: 0 [ 156.773675] msm 900000.mdss: [drm:recover_worker] *ERROR* A530: hangcheck recover!
Thanks, Kim