instruments/trace_cmd: Add tracing mode support to TraceCmdInstrument() #1276
Conversation
|
LGTM however it needs to wait until this devlib PR is merged: EDIT: strikethrough |
There was a problem hiding this comment.
Also looks good to me.
And thanks @douglas-raillard-arm for the poke on the corresponding devlib PR. I think we can avoid having to wait for a PyPi release but you're correct we'll need at least a dev bump to align these PR's when merging.
Just came back to remove that part from my comment. As long as WA's PyPI release is not pushed before devlib's one it's probably ok to only support devlib's master in WA's master branch |
| @@ -162,6 +162,12 @@ class TraceCmdInstrument(Instrument): | |||
| installed on the host (the one in your | |||
| distribution's repos may be too old). | |||
| """), | |||
| Parameter('mode', kind=str, default='write-to-memory', | |||
There was a problem hiding this comment.
should specify allowed_values=['write-to-disk', 'write-to-memory']
|
This PR supersedes #1232 |
|
Added the branch to our vendored version for dogfooding |
metin-arm
force-pushed
the
tracing-mode
branch
from
7952451 to
86aabab
Compare
Implement tracing mode support (mainly for write-to-disk mode) in TraceCmdInstrument, enabling efficient collection of large trace datasets without encountering memory limitations. This feature is particularly useful for scenarios requiring extensive trace data. Additional changes: - Replace hardcoded strings with corresponding string literals for improved maintainability. Signed-off-by: Metin Kaya <[email protected]>
metin-arm
force-pushed
the
tracing-mode
branch
from
86aabab to
cdf5576
Compare
Implement tracing mode support (mainly for write-to-disk mode) in TraceCmdInstrument, enabling efficient collection of large trace datasets without encountering memory limitations.
This feature is particularly useful for scenarios requiring extensive trace data.
Additional changes: