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
[BUG] glitch in test 02_00_TestPtpSimpleLoopback48000Hz24b32b8ch #4076
@wszypelt@zrombel I didn't know we supported the same buffer for host DMA and DAI DMA ? I thought the simplest would be the passthrough volume that places a volume in the middle to make sure we dont use the same buffer for host and DAI DMA ?
This image can be a little bit confusing as "Host plb" and "Host cap" component are a part of pipelines (test documentation needs a fix). Regarding buffers: buffers shown on topology are buffers that are explicitly created by test logic by IPC "TplgBuffNew" but as far as I know one another buffer is being created implicitly by FW for data from host DMA - those buffers are not included in tests topology drawings. Please correct me if I'm wrong.
There is a lot of very similar loopback test failures using different audio pipelines with different processing options now marked with "duplicate" label.
All these tests have loopback in common so this simple example should be solved first as it's test case architecture is shared with the other more complex examples. Once this is solved , then the others can be retested and closed/updated as necessary.
Activity
lgirdwood commentedon Apr 27, 2021
@wszypelt @zrombel I didn't know we supported the same buffer for host DMA and DAI DMA ? I thought the simplest would be the passthrough volume that places a volume in the middle to make sure we dont use the same buffer for host and DAI DMA ?
zrombel commentedon Apr 27, 2021
This image can be a little bit confusing as "Host plb" and "Host cap" component are a part of pipelines (test documentation needs a fix). Regarding buffers: buffers shown on topology are buffers that are explicitly created by test logic by IPC "TplgBuffNew" but as far as I know one another buffer is being created implicitly by FW for data from host DMA - those buffers are not included in tests topology drawings. Please correct me if I'm wrong.
lgirdwood commentedon May 13, 2021
There is a lot of very similar loopback test failures using different audio pipelines with different processing options now marked with "duplicate" label.
All these tests have loopback in common so this simple example should be solved first as it's test case architecture is shared with the other more complex examples. Once this is solved , then the others can be retested and closed/updated as necessary.
wszypelt commentedon Oct 13, 2022
No longer reproduced