cmake: targets: Allow adding additional arguments to puncover - #115155
cmake: targets: Allow adding additional arguments to puncover#115155paulwuertz wants to merge 1 commit into
Conversation
|
You have been identified as a likely reviewer for the code this pull request changes, but could not be added to its review request automatically. Please review it if you are able to. |
|
In the documentation I was thinking if to add, that stack usage in puncover is limited to ARM architecture and to some extend to RISC-V (I find it not working for some applications HBehrens/puncover#151). I opened an proposal to consume the .ci files generated by There is also an open PR with the interesting option to add a file with a list of dynamic / indirect calls, so the stack use estimation views more of the real call tree and yields better results, this would be another option to provide potentially soon :) |
paulwuertz
changed the title
d024f2d to
c22f7ea
Compare
paulwuertz
changed the title
d355e12 to
e713dc2
Compare
This allows i.e. non-interactive use and generating JSON reports of given list of functions with an associated max stack sizes and flexibility to add any other options. Relates to zephyrproject-rtos#112163 Signed-off-by: Paul Würtz <paulwuertz@posteo.de>
|
Sleeping a night over it I am undecided, weather adding to the tools docs example of a config file, or if this should better go as a ref to the puncover repo and there should be some more documentation :? - added it here for feedback :) Metioned also in the RFC - here is a puncover config with this MR aiming to print some feedback during/after building about stack usage: elf_file: ~/zephyrproject/zephyr/samples/net/mqtt_publisher/build/zephyr/zephyr.elf
gcc-tools-base: ~/zephyr-sdk-1.0.1/gnu/arm-zephyr-eabi/bin/arm-zephyr-eabi-
src_root: ~/zephyrproject/zephyr
build_dir: ~/zephyrproject/zephyr/samples/net/mqtt_publisher/build
generate-report: true
non-interactive: true
report-type: json
print-stack-size-report: true
report-max-static-stack-usage:
log_process_thread_func:::832
- shell_thread:::4160
- bg_thread_main:::2112
- work_queue_main:::1088
- mgmt_event_work_handler:::896then exporting a detailed report as a JSON for all symbols and printing an overview of stack usages of given threads: ┌─────────────────────────────────────────────────────────────────────────┐
│ log_process_thread_func - 42.3% - 352/ 832b - ████████------------│
│ shell_thread - 7.3% - 304/ 4160b - █-------------------│
│ mgmt_event_work_handler - 31.2% - 280/ 896b - ██████--------------│
│ bg_thread_main - 58.3% - 1232/ 2112b - ███████████---------│
│ work_queue_main - 25.0% - 272/ 1088b - █████---------------│
└─────────────────────────────────────────────────────────────────────────┘So enabling some use in CIs for continous feedback and help hint memory changes in reviews :) |
|
3 participants
This allows to use puncover i.e. non-interactively generating a JSON reports of given list of functions with an associated max stack sizes and adds flexibility to add any other options.
Relates to #112163
Signed-off-by: Paul Würtz paulwuertz@posteo.de