You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[BUG][MTL-004][Chrome]IPC timeout when set pipeline state to running for DMIC pipeline #7228
changed the title [-][BUG][MTL-Chromebook][MTL-004] IPC timeout when set pipeline state to running[/-][+][BUG][MTL-Chromebook][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/+]on Mar 8, 2023
changed the title [-][BUG][MTL-Chromebook][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/-][+][BUG][Chrome][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/+]on Mar 8, 2023
changed the title [-][BUG][Chrome][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/-][+][BUG][MTL-004][Chrome]IPC timeout when set pipeline state to running for DMIC pipeline[/+]on Mar 8, 2023
Analyzed today the problem with -d option, we are crashing because of wrong clock selected for this load and this results in timer overruns. Without -d option DSP can still handle it and crashing only with asserts enabled.
However, the clock selection mechanism and manifest data are correct. The true root cause is incorrect data in IPC for gain module. For gain FW receives 0 CPC instead of 10183 from manifest and this results in wrong calculations for clock selection.
All other modules receives normal CPC.
Gain receives correct CPC in python tests also.
Activity
[-][BUG][MTL-Chromebook][MTL-004] IPC timeout when set pipeline state to running[/-][+][BUG][MTL-Chromebook][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/+][-][BUG][MTL-Chromebook][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/-][+][BUG][Chrome][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/+]keqiaozhang commentedon Mar 8, 2023
This issue also can be reproduced on MTL-RVP. I filed a similar issue #7231
[-][BUG][Chrome][MTL-004] IPC timeout when set pipeline state to running for DMIC pipeline[/-][+][BUG][MTL-004][Chrome]IPC timeout when set pipeline state to running for DMIC pipeline[/+]jairaj-arava commentedon Mar 8, 2023
@aiChaoSONG I see the same issue on the device at my end with MTL-004 release with both DMIC and Headset MIC.
DMIC:
[ 187.862623] sof-audio-pci-intel-mtl 0000:00:1f.3: ipc timed out for 0x13020003|0x0
[ 187.871121] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ IPC dump start ]------------
[ 187.880602] sof-audio-pci-intel-mtl 0000:00:1f.3: Host IPC initiator: 0x13020003|0x0|0x0, target: 0x33000000|0x0|0x0, ctl: 0x3
[ 187.893337] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ IPC dump end ]------------
[ 187.902574] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ DSP dump start ]------------
[ 187.912003] sof-audio-pci-intel-mtl 0000:00:1f.3: IPC timeout
[ 187.918424] sof-audio-pci-intel-mtl 0000:00:1f.3: fw_state: SOF_FW_BOOT_COMPLETE (6)
[ 187.927272] sof-audio-pci-intel-mtl 0000:00:1f.3: ROM status: 0x5, ROM error: 0x0
[ 187.935665] sof-audio-pci-intel-mtl 0000:00:1f.3: ROM debug status: 0x50000005, ROM debug error: 0x0
[ 187.946010] sof-audio-pci-intel-mtl 0000:00:1f.3: ROM feature bit not enabled
[ 187.954014] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ DSP dump end ]------------
[ 187.963253] sof-audio-pci-intel-mtl 0000:00:1f.3: failed to pause pipeline 14
[ 187.971227] sof-audio-pci-intel-mtl 0000:00:1f.3: ASoC: error at soc_component_trigger on 0000:00:1f.3: -110
[ 187.982218] DMIC: ASoC: trigger FE cmd: 1 failed: -110
Headset:
[ 74.643167] init: ml-service (mojo_service) post-start process (5057) terminated with status 2
[ 77.403116] sof-audio-pci-intel-mtl 0000:00:1f.3: ipc timed out for 0x13020003|0x0
[ 77.411612] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ IPC dump start ]------------
[ 77.421065] sof-audio-pci-intel-mtl 0000:00:1f.3: Host IPC initiator: 0x13020003|0x0|0x0, target: 0x33000000|0x0|0x0, ctl: 0x3
[ 77.433823] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ IPC dump end ]------------
[ 77.443060] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ DSP dump start ]------------
[ 77.452491] sof-audio-pci-intel-mtl 0000:00:1f.3: IPC timeout
[ 77.458911] sof-audio-pci-intel-mtl 0000:00:1f.3: fw_state: SOF_FW_BOOT_COMPLETE (6)
[ 77.467569] sof-audio-pci-intel-mtl 0000:00:1f.3: ROM status: 0x5, ROM error: 0x0
[ 77.475933] sof-audio-pci-intel-mtl 0000:00:1f.3: ROM debug status: 0x50000005, ROM debug error: 0x0
[ 77.486217] sof-audio-pci-intel-mtl 0000:00:1f.3: ROM feature bit not enabled
[ 77.494192] sof-audio-pci-intel-mtl 0000:00:1f.3: ------------[ DSP dump end ]------------
[ 77.503429] sof-audio-pci-intel-mtl 0000:00:1f.3: failed to pause pipeline 6
[ 77.511318] sof-audio-pci-intel-mtl 0000:00:1f.3: ASoC: error at soc_component_trigger on 0000:00:1f.3: -110
[ 77.522375] Headset: ASoC: trigger FE cmd: 1 failed: -110
abonislawski commentedon Mar 9, 2023
Analyzed today the problem with -d option, we are crashing because of wrong clock selected for this load and this results in timer overruns. Without -d option DSP can still handle it and crashing only with asserts enabled.
However, the clock selection mechanism and manifest data are correct.
The true root cause is incorrect data in IPC for gain module. For gain FW receives 0 CPC instead of 10183 from manifest and this results in wrong calculations for clock selection.
All other modules receives normal CPC.
Gain receives correct CPC in python tests also.
topology2: add cpc value for gain
topology2: add cpc value for gain
topology2: add cpc value for gain
topology2: add cpc value for gain
mengdonglin commentedon Mar 22, 2023
Closing as not reproduced on MTL Chrome book.
aiChaoSONG commentedon Mar 22, 2023
can not be reproduced on Intel RVP