Closed jayzhuang closed 3 years ago
Note for the PATH hack, I was only able to test it in Linux, since I don't have access to a Windows environment.
Also on a side note, I had some trouble aligning things while changing some of the files, if you don't mind I'm happy to clang-format everything and set up a make format
target to automate that in the future.
Nice.
Don't worry about format. I use clang-format
myself everywhere nowadays, I'll bring my setup at some point.
Built on Windows and seems to work.
Thanks for testing on Windows! I squashed everything into one commit, PTAL.
Sorry for the delay.
Can't remember the exact details, but yes, the README is wrong now. I don't have a use for the dynamic size myself, all the commands I trace fit in 16MB. If you need more just change the LOGSZ value in fsatrace.h.
I think Google has a fork of this tool with more features, maybe you can try it. Maybe @jayzhuang has more details about that one?
This enables users to optionally specify a larger buffer size in case size of trace output exceeds default. Also enables a use case, although not likely, where users want to limit memory footprint.
Also addressed compiler warning, updated
make clean
to catch everything, and bumped optimizations from O2 to O3.Test: make test