[BUG] Zephyr panic excause 63 in nocodec capture test check-capture-50rounds

[kv2019i]

Describe the bug

DSP panic seen in daily tests on Intel PTL platform running check-capture-50rounds. First seen on daily test of 21st Feb with following dump in FW trace:

[00:01:42.280,021]  os: print_fatal_exception:  ** FATAL EXCEPTION
[00:01:42.280,028]  os: print_fatal_exception:  ** CPU 2 EXCCAUSE 63 (zephyr exception)
[00:01:42.280,031]  os: print_fatal_exception:  **  PC 0xa0030577 VADDR 0xa005e9a0
[00:01:42.280,035]  os: print_fatal_exception:  **  PS 0x60023
[00:01:42.280,036]  os: print_fatal_exception:  **    (INTLEVEL:3 EXCM: 0 UM:1 RING:0 WOE:1 OWB:0 CALLINC:2)
[00:01:42.280,040]  os: xtensa_dump_stack:  **  A0 0xa005cbc0  SP 0xa00b01d0  A2 0x4  A3 0xa00b01e0
[00:01:42.280,043]  os: xtensa_dump_stack:  **  A4 0xa00b01c0  A5 0x4  A6 0x47f  A7 0x4
[00:01:42.280,046]  os: xtensa_dump_stack:  **  A8 0xa005cb0a  A9 0xa00b0160 A10 0xa0083d85 A11 0xa00b01e0
[00:01:42.280,050]  os: xtensa_dump_stack:  ** A12 0xa00b01c0 A13 0x4 A14 0x1 A15 0x480
[00:01:42.280,051]  os: xtensa_dump_stack:  ** LBEG 0xa00388a5 LEND 0xa00388b4 LCOUNT 0xa005d510
[00:01:42.280,055]  os: xtensa_dump_stack:  ** SAR 0x1
[00:01:42.280,058]  os: xtensa_dump_stack:  **  THREADPTR 0xfffffff4

--
Backtrace:0xa0030574:0xa00b01d0  |<-CORRUPTED

To Reproduce
sof-test check-capture-50rounds

Reproduction Rate
open -- in multiple daily test plans since 21st Feb

Expected behavior
No DSP panics

Impact
Loss of audio functionality

Environment

1. SOF main , latest sof-dev kernel
2. Name of the topology file
   • Topology: sof-ptl-nocodec.tplg
3. Name of the platform(s) on which the bug is observed.
   • Platform: ptl

[kv2019i]

PRs merged before 21st: #9850 (21st), #9906 (20th), #9848 (20th), #9900 (21st)

[kv2019i]

Not seen in pull-requests tests as the panic usually triggers after at least 20 iterations of the check-capture test (i.e. PR CI test is too short to capture this fail).

[lyakh]

#9946 should fix this, merged, next daily test will tell

[lyakh]

Haven't been able to find any more of these failures since #9946 was merged on Apr. 2nd. Propose to close

[wszypelt]

@kv2019i request for verification and close

[kv2019i]

Aa, sorry, I tried to close this already before, but hit a github glitch. So based on PR/daily test results, this issue is fixed and can be closed.