[DNM] Zephyr smp rework #8764
[DNM] Zephyr smp rework #8764
Conversation
RanderWang
requested review from
tmleman,
lgirdwood,
plbossart,
mmaka1,
lbetlej,
dbaluta and
kv2019i
as code owners
RanderWang
force-pushed
the
zephyr-smp-rework
branch
from
00231ae to
f55f280
Compare
There was a problem hiding this comment.
@RanderWang can you check why TGL builds are failing? https://sof-ci.01.org/sofpr/PR8764/build2107/build/index.html
This shouldn't happen and now we don't get any test results for this version. If this would pass the CI, we could just merge this and drop the older Zephyr baseline update PR.
RanderWang
force-pushed
the
zephyr-smp-rework
branch
from
f55f280 to
fd8bcbb
Compare
This changes the secondary core power up routine to use the newly
introduced k_smp_cpu_start() and k_smp_cpu_resume(). This removes
the need to mirror part of the SMP start up code from Zephyr, and
no longer need to call into Zephyr private kernel code.
West update includes :
eefaeee061c8 kernel: smp: introduce k_smp_cpu_resume
042cb6ac4e00 soc: intel_adsp: enable DfTTS-based time stamping
on ACE platforms
6a0b1da158a4 soc: intel_adsp: call framework callback function for restore
e7217925c93e ace: use a 'switch' statement in pm_state_set()
c99a604bbf2c ace: remove superfluous variable initialisation
a0ac2faf9bde intel_adsp: ace: enable power domain
4204ca9bcb3f ace: fix DSP panic during startup (fixes c3a6274bf5e4)
d4b0273ab0c4 cmake: sparse.template: add COMMAND_ERROR_IS_FATAL
ca12fd13c6d3 xtensa: intel_adsp: fix a cache handling error
0ee1e28a2f5f xtensa: polish doxygen and add to missing doc
035c8d8ceb4b xtensa: remove sys_define_gpr_with_alias()
a64eec6aaeec xtensa: remove XTENSA_ERR_NORET
Signed-off-by: Daniel Leung <[email protected]>
Signed-off-by: Rander Wang <[email protected]>
RanderWang
force-pushed
the
zephyr-smp-rework
branch
from
fd8bcbb to
8084711
Compare
|
The failed cases in three types QuickBuild test case pass, but got failed result, why ? |
@RanderWang I've reviewed the last four builds for this pull request on QB. In the older ones, TGL doesn't build, but in the latest one, this issue is no longer present. However, there is a failure in the gain_quality test (also TGL). The log indicates that there was an issue in the final step during the audio quality verification. @wszypelt , could you check if this is a problem with the test? PS. I've scheduled a rebuild and rerun of tests. |
There was a problem hiding this comment.
Mandatory tests are now passing. The MTL error is known, so only open now is the cavs2.5 fails and whether there is more of them with this update than we have in baseline (we do have fails with pause-resume in mainline as well).
UPDATE: e.g. in this recent PR test, all these cavs2.5 tests pass:
https://sof-ci.01.org/sofpr/PR8756/build1998/devicetest/index.html
Compared to this PR:
https://sof-ci.01.org/sofpr/PR8764/build2195/devicetest/index.html
| @@ -43,7 +43,7 @@ manifest: | |||
|
|
|||
| - name: zephyr | |||
| repo-path: zephyr | |||
| revision: d7af6f371034f31b9440b27c694b0be3c87491d3 | |||
| revision: 6a0b1da158a4f8bc5f2ca5058637dce26a38660e | |||
There was a problem hiding this comment.
can you make this a separate patch for bisect-ability.
There was a problem hiding this comment.
I think that if we separate the west update into a separate commit, then we'll lose the ability to bisect because these changes won't compile without updating Zephyr.
@RanderWang correct me if I'm wrong.
There was a problem hiding this comment.
@lgirdwood west update needs to be in same commit, otherwise bisect is broken.
|
Marking as DNM so this doesn't get accidentally merged. We need to understand first why we have so many fails in https://sof-ci.01.org/sofpr/PR8764/build2195/devicetest/index.html |
|
#8792 filed for the chain-dma failures seen. I'll submit a fix soon to Zephyr. |
|
#8790 submitted to fix the double-free. |
Update secondary core setting based on zephyr smp changes zephyrproject-rtos/zephyr#64755.
Waiting for : #8747