debug: debug_stream: Access debug slot directly using cavstool.py #9498
Conversation
jsarha
requested review from
lgirdwood,
plbossart,
mmaka1,
lbetlej,
dbaluta and
kv2019i
as code owners
jsarha
force-pushed
the
debug_stream_thread_info_cavstool_access
branch
from
e3ebc10
to
cafdef9
Compare
|
My latest cleanups were still un committed, thus the last update. |
jsarha
force-pushed
the
debug_stream_thread_info_cavstool_access
branch
from
cafdef9
to
5c1957e
Compare
|
Latest push added some robustness improvements to the python script and config overlay for building debug_stream_thread_info. |
jsarha
force-pushed
the
debug_stream_thread_info_cavstool_access
branch
from
5c1957e
to
4bfe83a
Compare
|
The important part of this PR - the debug_stream.py direct access - is now split in two:
Then there is the debug_stream_overlay.conf added that is identical to the previous version and on simple typo fix commit. |
|
@jsarha does this PR depend on the Zephyr PRs ? if so , lets mark as DNM until Zephyr parts merged. |
Access debug slot directly using cavstool.py, by specifying the debug slot number where the debug_stream data is transferred. Adds one command line parameter for selecting the debug slot directly and adds an alternative mainloop for polling the slot through cavstool.py direct access. The commit also adds quite a few data consistency checks and error handling improvements as with the direct memory access its much more likely to get inconsistent data when the DSP is booting up, and code needs to be robust enough not to crash in such a situation. Also the logging messages about those checks failing has been lowered so that they are not too noisy. Signed-off-by: Jyri Sarha <[email protected]>
jsarha
force-pushed
the
debug_stream_thread_info_cavstool_access
branch
from
4bfe83a
to
e040178
Compare
|
Fixed couple of simple codestyle issues found by checkpatch. |
The script actually does not stop from working even if the cavstool.py changes are not available, that is, if the debugfs file is there. But maybe the error messages if trying to use -c flag are too confusing. I'll add the tag. |
jsarha
changed the title
jsarha
changed the title
@jsarha Both Zephyr PRs look very similar - best to close one PR if its been deprecated by a new PR. |
@lgirdwood , the first PR [1] is the base of the second PR [2]. Actually the second [2] has only one commit on top of [1]. I was expecting the first PR to get merged soon as it already had two approvals (and I did not want it to get back to square one). Once the first is merged, I'll rebase the second, and it becomes only one commit PR. [1] zephyrproject-rtos/zephyr#78753 |
Add overlay to build debug_stream protocol over a debug window slot, make room for the slot, and send thread info data through it. Signed-off-by: Jyri Sarha <[email protected]>
A simple typo fix to a logging message. Signed-off-by: Jyri Sarha <[email protected]>
jsarha
force-pushed
the
debug_stream_thread_info_cavstool_access
branch
from
e040178
to
49e74b9
Compare
jsarha
changed the title
|
I dropped the commit that depends on zephyrproject-rtos/zephyr#79061 so this should be good for merging now. I'll make another PR that depends on the above PR. The new PR will be rebased on top of this, so it will contain the same commit as this. |
Access debug slot directly using cavstool.py, by specifying the debug slot number where the debug_stream data is transferred.
Adds one command line parameter for selecting the debug slot directly, adds an alternative mainloop for polling the slot through cavstool.py direct access, and relaxes some often occurring warning print. The warning prints happen when the FW boots and the cavstool code wakes up before the init code has initialized the slot.
NOTE:
This code depends from this PR: zephyrproject-rtos/zephyr#78753 . However, since the import line is in the function that is only called if the --direct-access-slot flag is present, the old code will still work without cavstool.py, but that would in turn need this PR: thesofproject/linux#5154