Repository navigation
[BUG] LNL has FW reported error: 7 with multiple pipeline playback #8275
Description
Describe the bug
First LNL issue.
For LNLM_RVP_NOCODEC, multiple pipeline playback fails with FW reported error: 7 - Unsupported operation requested.
[ 1328.146968] kernel: snd_sof:sof_ipc4_set_pipeline_state: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc4 set pipeline instance 6 state 3
[ 1328.146983] kernel: snd_sof:sof_ipc4_log_header: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc tx : 0x13060003|0x0: GLB_SET_PIPELINE_STATE
[ 1328.147386] kernel: snd_sof:sof_ipc4_log_header: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc tx reply: 0x33000000|0x0: GLB_SET_PIPELINE_STATE
[ 1328.147493] kernel: snd_sof:sof_ipc4_log_header: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc tx done : 0x13060003|0x0: GLB_SET_PIPELINE_STATE
[ 1328.147521] kernel: snd_sof:sof_ipc4_set_pipeline_state: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc4 set pipeline instance 6 state 4
[ 1328.147532] kernel: snd_sof:sof_ipc4_log_header: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc tx : 0x13060004|0x0: GLB_SET_PIPELINE_STATE
[ 1328.149150] kernel: snd_sof:sof_ipc4_log_header: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc tx reply: 0x33000007|0x0: GLB_SET_PIPELINE_STATE
[ 1328.149162] kernel: sof-audio-pci-intel-lnl 0000:00:1f.3: FW reported error: 7 - Unsupported operation requested
[ 1328.149388] kernel: sof-audio-pci-intel-lnl 0000:00:1f.3: ipc error for msg 0x13060004|0x0
[ 1328.149420] kernel: sof-audio-pci-intel-lnl 0000:00:1f.3: ASoC: error at soc_dai_trigger on SSP1 Pin: -22
[ 1328.149439] kernel: Port1: ASoC: trigger FE cmd: 1 failed: -22
To Reproduce
TPLG=/lib/firmware/intel/sof-ace-tplg/sof-lnl-nocodec.tplg ~/sof-test/test-case/multiple-pipeline-playback.sh -l 10
Reproduction Rate
100%
Environment
- kernelBranch : topic/sof-dev
- kernelCommit : b9c99093f5ed-3
- sofBranch : main
- sofCommit : deb7a09
- TPLG: sof-mtl-nocodec.tplg, symbolic link to /lib/firmware/sof-ace-tplg/sof-lnl-nocodec.tplg
NOTE: There is no LNL topology, MTL nocodec tplg is re-used.
Screenshots or console output
dmesg.txt
mtrace.txt
Activity
keqiaozhang commentedon Oct 31, 2023
This issue not only happens on LNL-NOCODEC platform, but also happens on LNL-SDW platform when testing multiple-pipeline or multiple-pause-resume.
abonislawski commentedon Oct 31, 2023
@jxstelter please take a look if this is any known issue and we will assign someone if required
[-][BUG] LNL Nocodec has FW reported error: 7 with multiple pipeline playback[/-][+][BUG] LNL has FW reported error: 7 with multiple pipeline playback[/+]softwarecki commentedon Nov 15, 2023
@fredoh9 Can you provide logs with ipc payloads?
fredoh9 commentedon Nov 15, 2023
let me find how to capture logs with ipc payloads
fredoh9 commentedon Nov 15, 2023
Here you are, this has IPC payloads
dmesg-mtl-aioc.txt
For your information and my memory, this is with LNLM_SDW_AIOC,
cmd: TPLG=/lib/firmware/intel/sof-ace-tplg/sof-lnl-rt711-l0-rt1316-l23-rt714-l1.tplg MODEL=LNLM_SDW_AIOC SOF_TEST_INTERVAL=5 ~/sof-test/test-case/multiple-pipeline.sh -f p -l 50
Intel internal test link:
planresultdetail/34687?model=LNLM_SDW_AIOC&testcase=multiple-pipeline-playback-50
13 remaining items
keqiaozhang commentedon Dec 18, 2023
This issue can still be reproduced with multiple-pipeline testing. Inner test ID:36046.
wszypelt commentedon Dec 22, 2023
@keqiaozhang does this still occur on main or is it different branch?
keqiaozhang commentedon Dec 25, 2023
This issue still exists on main branch, the reproduction rate is very high on LNL platform with external SDW codecs.
WW52.2 https://sof-ci.ostc.intel.com/#/result/planresultdetail/36239
WW52.4 https://sof-ci.ostc.intel.com/#/result/planresultdetail/36335
keqiaozhang commentedon Jan 3, 2024
Now this issue only happens on LNL-SDW platform now, No reproductions on LNL-NOCODEC for weeks.
plbossart commentedon Jan 4, 2024
the logs for the SoundWire case looks very suspicious with a buffer id of either 0 or 65536?
https://sof-ci.ostc.intel.com/#/result/planresultdetail/36335?model=LNLM_SDW_AIOC&testcase=multiple-pipeline-playback-50
plbossart commentedon Jan 4, 2024
@keqiaozhang I wouldn't be surprised if this is an issue with the mixer.
Can we try to put the "Deepbuffer Jack Out" in the deny list to see if independent pipelines work fine?
should do the trick
keqiaozhang commentedon Jan 5, 2024
@plbossart , I tried to skip
Deepbufferpipeline, but it still failed:This issue can be easily triggered by:
plbossart commentedon Jan 5, 2024
Thanks @keqiaozhang that's very useful, One last suggestion would be to try with a passthrough topology with simplified pipelines to see if this is the DSP IOs getting corrupted somehow, or really the DSP graph introducing issues.
bardliao commentedon Jan 25, 2024
@plbossart @keqiaozhang I tested with #8361 and thesofproject/linux#4790 on LNL SDW. And the issue is not reproducible by
aplay -D hw:0,2 -c 2 -r 48000 -f S16_LE /dev/zero -q -d 10 & aplay -D hw:0,0 -c 2 -r 48000 -f S16_LE /dev/zero -q -d 10orTPLG=/lib/firmware/intel/sof-ipc4-tplg/sof-lnl-rt711-l0-rt1316-l23-rt714-l1.tplg MODEL=LNLM_SDW_AIOC SOF_TEST_INTERVAL=5 ~/sof-test/test-case/multiple-pipeline.sh -f p -l 50wszypelt commentedon Jan 26, 2024
root cause: #8361
waiting for merge
wszypelt commentedon Jan 30, 2024
Based on the report from Keqiao
the issue does not reproduce