Repository navigation
[BUG][CML] ipc time out for 0x60050000 size 12 cause HDMI no sound when plug-in and out HDMI cable via MST hub #2055
Description
Describe the bug
the hdmi no sound when playing video file by plug-in and out HDMI cable via MST hub(on board) output. The failure rate is 2/40, sometimes even 1/10.
Summarize the current identification for addressing the issue FYI.
Also capture the intel_audio_dump when issue happened. FYI
Experiment:
PSR1 / PSR2 Panel - Fail
disable PSR (by i915_enable_psr=0) - Pass.
Non PSR Panel - Pass
Disable MST- Pass
[ 1285.792191] sof-audio-pci 0000:00:1f.3: error: ipc timed out for 0x60050000 size 12
[ 1285.792215] sof-audio-pci 0000:00:1f.3: status: fw entered - code 00000005
[ 1285.792568] sof-audio-pci 0000:00:1f.3: error: unexpected fault 0x00000000 trace 0x00004000
[ 1285.792589] sof-audio-pci 0000:00:1f.3: error: hda irq intsts 0x00000000 intlctl 0xc0000080 rirb 00
[ 1285.792595] sof-audio-pci 0000:00:1f.3: error: dsp irq ppsts 0x00000000 adspis 0x00000000
[ 1285.792613] sof-audio-pci 0000:00:1f.3: error: host status 0x00000000 dsp status 0x00000000 mask 0x00000003
[ 1285.792628] HDMI1: ASoC: trigger FE cmd: 0 failed: -110
[ 1285.792215] sof-audio-pci 0000:00:1f.3: status: fw entered - code 00000005
[ 1285.792568] sof-audio-pci 0000:00:1f.3: error: unexpected fault 0x00000000 trace 0x00004000
To Reproduce
- Play a local video.
- Plug in an HDMI / Type-C monitor on daughter board.
- Find external monitor has no sound.
Reproduction Rate
2/40 , 1/40
Expected behavior
HDMI playback works normally.
Impact
HDMI playback is unable to recover when issue happened. Only one case, pause the video playback pause/stop, waiting 25s~ until dsp poweroff. Then resume the playback is normal.
Environment
Firmware info: version 1:1:0-04852
Firmware: ABI 3:9:0 Kernel ABI 3:8:0
Activity
macchian commentedon Apr 27, 2020
hdmi_no_sound[intel_audio_dump].txt
Uploading hdmi_sound_recovery_after_30s[intel_audio_dump].txt…
ipc_timeout_when_hdmi_no_sound[kernel].txt
macchian commentedon Apr 27, 2020
@tlauda gentle pin after searching for HDMI ipc timed out comments. Or invite someone can help. Thanks.
macchian commentedon May 6, 2020
update new logs. When issue happens we saw the logs below, is that a underrun case ? @mengdonglin @tlauda
pcm_avail returned frames larger than buf_size: sof-sof_rt5682: :0,2: 22148 > 16368
2020-04-30T11:17:37.303424+00:00 ERR cras_server[1961]: pcm_avail returned frames larger than buf_size: sof-sof_rt5682: :0,2: 22148 > 16368
2020-04-30T11:18:08.340780+00:00 ERR cras_server[1961]: pcm_avail returned frames larger than buf_size: sof-sof_rt5682: :0,2: 22147 > 16368
2020-04-30T11:18:23.653193+00:00 NOTICE temp_logger[5588]: 00:41C 01:20C 02:41C 03:37C 04:34C 05:41C 06:37C PL1:15.000W
2020-04-30T11:18:39.378051+00:00 ERR cras_server[1961]: pcm_avail returned frames larger than buf_size: sof-sof_rt5682: :0,2: 22148 > 16368
2020-04-30T11:19:02.186113+00:00 ERR kernel: [ 710.091567] sof-audio-pci 0000:00:1f.3: error: ipc error for 0x60050000 size 12
2020-04-30T11:19:02.186143+00:00 ERR kernel: [ 710.092181] HDMI1: ASoC: trigger FE cmd: 0 failed: -22
2020-04-30T11:19:23.719089+00:00 NOTICE temp_logger[5768]: 00:40C 01:20C 02:41C 03:37C 04:34C 05:40C 06:36C PL1:15.000W
2020-04-30T11:20:23.786523+00:00 NOTICE temp_logger[5946]: 00:40C 01:20C 02:40C 03:36C 04:33C 05:40C 06:36C PL1:15.000W
2020-04-30T11:21:23.851321+00:00 NOTICE temp_logger[6123]: 00:39C 01:20C 02:40C 03:36C 04:33C 05:39C 06:36C PL1:15.000W
2020-04-30T11:22:00.605434+00:00 ERR kernel: [ 888.421498] HDMI HDA Codec ehdaudio0D2: hdac_hdmi_present_sense: disconnect for pin:port 6:1
2020-04-30T11:22:08.749083+00:00 INFO kernel: [ 896.561827] sof-audio-pci 0000:00:1f.3: firmware boot complete
2020-04-30T11:22:08.801752+00:00 ERR cras_server[1961]: Unable to find the best channel map
kv2019i commentedon May 13, 2020
@macchian I think the problem could be two-fold. As PSR-panel/MST impacts the occurence rate, I think the rootcause is a bug on i915 driver side that causes streaming to fail.
Second, once streaming fails on the HDA iDisp link, I think we have an additional problem in SOF FW to handle the XRUN. I note the FW used is somewhat old. Would it be possible to retry the test with a newer firmware version? @tlauda Other suggestions?
macchian commentedon May 14, 2020
@keyonjie , thanks for recommendation. I agree it's XRUN issue but no clue so far... I will try the following fw branch and see if any improvement. Any further debug suggestion/log is welcome. I can help address fw and catch sof-logger.
remotes/origin/cml-009-drop-stable
remotes/origin/cml-010-drop-stable
remotes/origin/cml-011-drop-stable
plbossart commentedon Aug 12, 2020
@macchian @kv2019i no update since May 13, is this a real issue?
macchian commentedon Aug 18, 2020
As per customer's report, they can not reproduce on specific OS image above and the bisect result indicate the regression issue on Chromium changes not kernel changes. We can close it until any clue pointing to kernel side by reopen.
staging: vme_user: Fix checkpatch warnings in vme_tsi148.c