Re-introducing state change in dai_comp_trigger_internal() breaks some tests from Intel CI

[LaurentiuM1234]

Attempting to bring back the state transition logic in dai-zephyr's dai_comp_trigger_internal() that was initially removed by 6ffb052 seems to break some tests from the Intel CI. The logic is needed for NXP platforms and is re-introduced by #8859 via 59d05a3.

As a workaround, the state transition logic has been made IPC3-specific but it might be worth looking into this and trying to make the code IPC version-agnostic if possible.

Links to the aforementioned failing CI builds:

• https://sof-ci.01.org/sofpr/PR8859/build3069/devicetest/index.html
• https://sof-ci.01.org/sofpr/PR8859/build3070/devicetest/index.html
• https://sof-ci.01.org/sof-pr-viewer/#/build/PR8859/build13666634 (internal build)

[lgirdwood]

@LaurentiuM1234 probably the best thing todo is to break down the changes into small bi-sectable patches and upstream them incrementally. This would allow more specific targeting when the issue is introduced.

[lgirdwood]

@LaurentiuM1234 are we good to close now that #8859 is merged or more still todo ?

[LaurentiuM1234]

@LaurentiuM1234 are we good to close now that #8859 is merged or more still todo ?

This issue was actually created by the introduction of #8859 (1bf6b8c). I think the question still stands: why does that code bit break Intel test(s) if made IPC version-agnostic?

[kv2019i]

Stable branched for 2.10, nobody assigned to work on this, so moving this to 2.11.

[kv2019i]

No owner -> v2.12