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][v1.5.1]shows IPC error and sof error "ERROR eq_iir_init_coef(), ch 0 is set to bypass" shows during the check-audio-equalizer test #3035
Describe the bug
"ERROR eq_iir_init_coef(), ch 0 is set to bypass" shows during the check-audio-equalizer test. Issue caused by commit: 88707ca based on the bisect result. Could you please have a look @abonislawski
All pipline can work after issue occurred. Seems has no functional side effect.
To Reproduce
Boot up system with topology has a EQ component in it
cd sof-test and run " export TPLG=/lib/firmware/intel/sof-tplg/sof-cml-rt1011-rt5682.tplg"
changed the title [-][BUG][v1.5.1]"ERROR eq_iir_init_coef(), ch 0 is set to bypass" shows during the check-audio-equalizer test[/-][+][BUG][v1.5.1]shows IPC error and sof error "ERROR eq_iir_init_coef(), ch 0 is set to bypass" shows during the check-audio-equalizer test[/+]on Jun 9, 2020
The code should work as gcc build too, despite lower optimization level, I'll still see what happens. Even if the process would require to close this I'll still keep this as action for myself to study what happens.
Also re-confirm that with xt-xcc build the above check for filter length (to prevent larger fail of SOF) does not trigger and the test case passes, no errors in trace. OK to close in my opinion too.
Actually there's a minor bug in above "Filter length 0", the printed value (not set yet) is not the same as tested value. I'll make a small PR to fix the error message.
(Thanks for well prepared error report @YvonneYang2 )
Activity
[-][BUG][v1.5.1]"ERROR eq_iir_init_coef(), ch 0 is set to bypass" shows during the check-audio-equalizer test[/-][+][BUG][v1.5.1]shows IPC error and sof error "ERROR eq_iir_init_coef(), ch 0 is set to bypass" shows during the check-audio-equalizer test[/+]abonislawski commentedon Jun 9, 2020
Are you sure about 88707ca?
There is no trace from this commit
trace_ipc_error("ipc: invalid size 0x%x", hdr->size);
YvonneYang2 commentedon Jun 9, 2020
Tested it with fw compiled via xcc, issue cannot be reproduced.
Close it for it's gcc specific issue.
singalsu commentedon Jun 9, 2020
The code should work as gcc build too, despite lower optimization level, I'll still see what happens. Even if the process would require to close this I'll still keep this as action for myself to study what happens.
singalsu commentedon Jun 9, 2020
I'm not able to repeat this same error on a CML device with gcc build. The fail with too long FIR filter is normal and it's known.
Also re-confirm that with xt-xcc build the above check for filter length (to prevent larger fail of SOF) does not trigger and the test case passes, no errors in trace. OK to close in my opinion too.
singalsu commentedon Jun 9, 2020
Actually there's a minor bug in above "Filter length 0", the printed value (not set yet) is not the same as tested value. I'll make a small PR to fix the error message.
(Thanks for well prepared error report @YvonneYang2 )