alsa_settings: add alsa settings for IPC4 platforms #992
Conversation
keqiaozhang
requested review from
aiChaoSONG,
fredoh9,
greg-intel,
marc-hb and
miRoox
miRoox
left a comment
miRoox
left a comment
There was a problem hiding this comment.
Why don't we ignore the difference between IPC3/IPC4, like trimming the model name to exclude the 4th part?
Because the tplgs are different, so the volume control names are different between IPC3 and IPC4. |
miRoox
dismissed
their stale review
I don't like this solution, but it's acceptable for me.
|
@ranj063 I think most of the settings will have to be changed with your topology cleanups, no? |
|
@ranj063 , please help to confirm if we need to change the amixer setting after your topology cleanups? |
@keqiaozhang @plbossart , Im going to punt this to @jsarha. He's working on cleaning up the mixer names. EDIT: |
keqiaozhang
force-pushed
the
alsasetting_ipc4
branch
from
e0d9e9b to
0e2290c
Compare
Signed-off-by: Keqiao Zhang <keqiao.zhang@intel.com>
|
I have verified this patch with PR991, the results are as expected:https://sof-ci.sh.intel.com/#/result/planresultdetail/19443
|
|
@aiChaoSONG please help to review this PR, since we have enabled alsabat test on all our IPC4 platforms, so we need this PR to set the amixer settings before the test. This PR is well tested. |
| amixer -c sofnocodec cset name='gain.3.1 Playback Volume 3' 45 | ||
| amixer -c sofnocodec cset name='gain.4.1 Main Playback Volume 4' 45 | ||
| amixer -c sofnocodec cset name='gain.5.1 Playback Volume 5' 45 | ||
| amixer -c sofnocodec cset name='gain.6.1 Main Playback Volume 6' 45 |
There was a problem hiding this comment.
Can you remind me why we should care about the volume settings in the firmware? This should be 0dB by default, always. the way this is implemented will break for every topology change or rename of controls.
There was a problem hiding this comment.
Yes, the default volume should be 0dB, but we cannot ensure that these volumes will not be modified before testing alsabat.
There was a problem hiding this comment.
@keqiaozhang you mean because other tests may change them? Why?
I've seen a large number of discussions about ALSA settings and UCM for a few years now and I still don't know how the "ideal state" should look like. Exceptions and quirks are sometimes necessary but I still can't tell what is a quirk versus what is a Good Thing. Is this already documented somewhere or would a new https://github.com/thesofproject/sof-test/discussions help?
There was a problem hiding this comment.
Yes, the default volume should be 0dB, but we cannot ensure that these volumes will not be modified before testing alsabat.
then use a script to find all gain/PGA for the firmware using regexps and reset the value separately from codec stuff.
There was a problem hiding this comment.
then use a script to find all gain/PGA for the firmware using regexps and reset the value separately from codec stuff
Can we merge this PR first? I will try your suggestion next. It's not only related to volume, but also related to switch.
There was a problem hiding this comment.
then use a script to find all gain/PGA for the firmware using regexps and reset the value separately from codec stuff.
@keqiaozhang is this what you're trying to do in #1010?
There was a problem hiding this comment.
@keqiaozhang is this what you're trying to do in #1010?
Yes, I'm working on it.
|
Closing. New PR:#1012 |
6 participants
Signed-off-by: Keqiao Zhang keqiao.zhang@intel.com