[BUG] multiple-pipeline-all.sh sof-test failing on ptl-nocodec

[kv2019i]

Describe the bug Regression with https://github.com/thesofproject/sof/pull/10408 on the multiple-pipeline-all.sh test.

This went unnoticed as we had https://github.com/thesofproject/sof/issues/10414 open in main (now fix merged).

To Reproduce run the affected test

Reproduction Rate 100% (seen on PR and daily test)

Expected behavior no error

Impact

Environment see https://sof-ci.01.org/sofpr/PR10408/build17839/devicetest/index.html

Screenshots or console output

2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: Port0 [hw:0,0]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,0 -c 2 -r 48000 -f S16_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: DMIC Raw [hw:0,10]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,10 -c 4 -r 48000 -f S32_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: Port0 2nd Capture [hw:0,12]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,12 -c 2 -r 48000 -f S16_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: DMIC SFX1 [hw:0,27]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,27 -c 4 -r 48000 -f S32_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: DMIC SFX2 [hw:0,28]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,28 -c 4 -r 48000 -f S32_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: Port2 [hw:0,2]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,2 -c 2 -r 48000 -f S16_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] Testing: Port1 [hw:0,1]
2025-12-03 19:17:51 UTC [REMOTE_COMMAND] arecord   -D hw:0,1 -c 2 -r 48000 -f S16_LE /dev/null -q
2025-12-03 19:17:51 UTC [REMOTE_INFO] wait 0.5s for aplay_opts()
arecord: set_params:1435: Unable to install hw params:

[ 2822.466195] <inf> ipc: ipc_cmd: rx	: 0x40100003|0x10d0074
[ 2822.466523] <err> module_adapter: module_adapter_mem_alloc: failed to allocate memory for module
[ 2822.466726] <err> ipc: ipc4_init_module_instance: error: failed to init module 3 : 10
[ 2822.466785] <err> ipc: ipc_cmd: ipc4: MODULE_MSG failed with err 104
[ 2822.467111] <inf> ipc: ipc_cmd: rx	: 0x120d0000|0x0
[ 2822.581020] <inf> ipc: ipc_cmd: rx	: 0x13060003|0x0
[ 2822.581133] <inf> pipe: pipeline_trigger: pipe:6 0x0 pipe trigger cmd 2

[kv2019i]

@softwarecki can you take a look?

[kv2019i]

Tests continue successfully after the fail (and smaller multi-pipeline tests for playback and capture work), so could this be just system running out of memory with user-space enabled for the build?

[lgirdwood]

Tests continue successfully after the fail (and smaller multi-pipeline tests for playback and capture work), so could this be just system running out of memory with user-space enabled for the build?

Perhaps heap fragmentation leading to later larger allocation failing if so, vregion will prevent this.

[kv2019i]

@softwarecki @abonislawski @lgirdwood @lyakh still triggering in each PR

[lyakh]

@softwarecki @abonislawski @lgirdwood @lyakh still triggering in each PR

@kv2019i #10447

[lgirdwood]

@softwarecki @abonislawski @lgirdwood @lyakh still triggering in each PR

@kv2019i #10447

Lets monitor for a few days and close.