You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[FEATURE] Add support for Samus (2015 Pixelbook) #2157
Is your feature request related to a problem? Please describe.
A common machine for Google to test on is the pixelbooks, it would be excellent to add samus to the mix
Additional context
Uses the bdw-rt5677 machine driver
@plbossart do you think the SSP L/R bit is inverted here ? or incorrectly documented ? It seems like the L/R swap fix is probably a SSP bit change (unless you have already tried this) ?
@plbossart do you think the SSP L/R bit is inverted here ? or incorrectly documented ? It seems like the L/R swap fix is probably a SSP bit change (unless you have already tried this) ?
Yeah, I tried and it's not a polarity issue (happens also with DSP_A/B). I have pictures somewhere.
@plbossart thanks got you. Fwiw, I wonder (since swapping is consistent) if we were to write a single 32 bit word to the SSP FIFO as part of trigger for > 16 bit data output to re-order the FIFO output on the link. Alternatively we could fill the FIFO with 16 * 0s as part of trigger start meaning the FIFO is always in a consistent state at start has send data without underruning (which may mean we drop a sample and invert L/R). Just guessing....
Closing, Samus is well supported at the firmware level. The driver is still missing suspend/resume support but that's a different request that will probably never be implemented.
Activity
lgirdwood commentedon Dec 16, 2019
@wenqingfu @mengdonglin do we have any Samus HW ?
cujomalainey commentedon Dec 16, 2019
I think we found the issue in the DAI links in another bug, we are patching both samus and buddy to support SOF
cujomalainey commentedon Dec 16, 2019
thesofproject/linux#1484
[-][FEATURE] Add support for Samus (2015 Pixelbook[/-][+][FEATURE] Add support for Samus (2015 Pixelbook)[/+]xiulipan commentedon Dec 18, 2019
@cujomalainey I have test on one device with #2188 and #2095
Will try to redo the test after rebase and update thesofproject/linux#1484
plbossart commentedon Jul 29, 2020
Samus is well supported now, the only bug is #2459 at the firmware level, and we need to sort out suspend-resume + improve the position reports without IPC:
https://github.com/thesofproject/linux/issues?q=is%3Aissue+is%3Aopen+label%3ABDW
lgirdwood commentedon Aug 4, 2020
@plbossart do you think the SSP L/R bit is inverted here ? or incorrectly documented ? It seems like the L/R swap fix is probably a SSP bit change (unless you have already tried this) ?
plbossart commentedon Aug 5, 2020
Yeah, I tried and it's not a polarity issue (happens also with DSP_A/B). I have pictures somewhere.
plbossart commentedon Aug 5, 2020
@lgirdwood see screen grabs in #2459
lgirdwood commentedon Aug 6, 2020
@plbossart thanks got you. Fwiw, I wonder (since swapping is consistent) if we were to write a single 32 bit word to the SSP FIFO as part of trigger for > 16 bit data output to re-order the FIFO output on the link. Alternatively we could fill the FIFO with 16 * 0s as part of trigger start meaning the FIFO is always in a consistent state at start has send data without underruning (which may mean we drop a sample and invert L/R). Just guessing....
plbossart commentedon Jun 10, 2021
Closing, Samus is well supported at the firmware level. The driver is still missing suspend/resume support but that's a different request that will probably never be implemented.