On 29/05/2024 15:17, Mark Brown wrote:
> On Wed, May 29, 2024 at 02:57:02PM +0300, Péter Ujfalusi wrote:
>> On 29/05/2024 14:48, Mark Brown wrote:
>
>>> Sorry, it's already been applied and published with more stuff in CI now
>>> based on it.
>
>> Right, I don't see it in for-next, for-6.10 or for-linus. This patch
>> should be sent for 6.10 and I can prepare and send the rest (4 more
>> patches) to close the remaining holes regarding to NHLT blob selection
>> logic.
>
> If something is a bug fix it shouldn't have a subject saying "Add
> support", that is very obvioulsy a new feature. Have these systems ever
> worked?
Well, it is adding support for something which has never been consider
as a valid configuration (16bit only DMIC blob in NHLT table of a laptop
[1]).
I have added to fixes tag to carry this patch along with the patch that
implemented support for preferring 32bit configuration in firmware (the
[1] would never worked, so that is unrelated).
The mentioned additional patches would future proof this whole DMIC blob
lookup by extending it beyond the bytes per sample.
These are future proofing and a would be nice fixes for the -rc, but if
not than 6.11 is OK.
In short, yes it worked well but we were presented with unimaginable
vendor creativity.
[1] https://github.com/thesofproject/linux/issues/4973
--
Péter
[Index of Archives]
[Pulseaudio]
[Linux Audio Users]
[ALSA Devel]
[Fedora Desktop]
[Fedora SELinux]
[Big List of Linux Books]
[Yosemite News]
[KDE Users]