topology2: add nocodec support for real case validation #8748
Conversation
RanderWang
requested review from
keqiaozhang,
kv2019i,
lgirdwood,
mengdonglin and
plbossart
|
|
|
Thank you @RanderWang ! This help us simulate Chrome use scenario in our testing. Maybe we can name it "sof-mtl-mock-chrome-nocodec"? But I feel current name is also good. |
|
Not following @RanderWang, why not use a 'nocodec' topology we already have? Why do we need multiple flavors of nocodec? Also what does this bring in terms of validation? You haven't quite described the problem statement and alternatives.... |
|
@plbossart The background is that we want to evaluate the impact of multicore timeout issue #8642 for Chrome by running a similar topology as Chrome on MTL RVP. That issue is reproduced on MTL RVP with nocodec topology (3 cores, dummy smart amp, 11 concurrent streams). This PR made a topology for 2 cores, no smart amp, AEC on core 2, and less streams. |
Ack - this will differentiate and allow stress and developer debug/test. |
Sof-mtl-max98357a-rt5682 is used by chrome project. This patch uses nocodec backend for validation so that our CI test can check real production case. Currently we use cavs-nocodec.tplg for validation, but it is far from real usage and makes no sense. Signed-off-by: Rander Wang <rander.wang@intel.com>
RanderWang
force-pushed
the
chrome-nocodec
branch
from
c4647ea to
f293847
Compare
4 participants
Sof-mtl-max98357a-rt5682 is used by chrome project. This patch uses nocodec backend for validation so that our CI test can check real production case.