v01d opened a new pull request #2335:
URL: https://github.com/apache/incubator-nuttx/pull/2335
## Summary
This is a WIP to paralellize generation of Make.dep files. It does so by
generating temporary per-file dep file as a separate make target (which allows
parallelization) which are collected into the final Make.dep.
This is a first naive attempt for testing (and only applied to sim arch
Makefile, every other arch would need the same treatment). I would like to know
about anything that may not work in other platforms (I test on Linux) and a
better way to handle all this repetition. Ideally, most of these rules could be
centralized somewhere but there are many slight variations of the dependency
call which I'm not sure what would be the best way to handle compactly.
## Impact
Dependency generation speed improvment.
* Original version:
* Single job:
make depend 16.47s user 8.86s system 102% cpu 24.790 total
* Multiple jobs (will only paralellize across directories):
make depend -j8 16.11s user 9.02s system 118% cpu 21.163 total
* Modified version:
* Multiple jobs (I have four actual cores, eight via hyperthreading):
make depend -j8 19.35s user 7.85s system 323% cpu 8.399 total
## Testing
sim:nsh
Ideally testing should verify we get the same Make.dep files.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
