-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Refactor update script #1596
Refactor update script #1596
Conversation
So regarding how to integrate this, we have two options. 1) create support for users just requesting compile time in which case the workflow would look like:
Or would could just add those two steps to the default case. Note the:
Will only need to run-once per update of llvm-project. |
NB: Didn't test V2. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please make sure that the llvm submodule version is the same with the llvm version in the base branch. You can run git submodule update
after git pull
:)
This primary purpose is to support machine specific compile times and move more of the codes to a common python script. Tested the functionality (without pushing) locally. This does not come with the ci scripts: New functionalities: Update baseline compile time: ``` python3 scripts/update.py --baseline --bench bench --comptime --no-diff ``` Get compile time for a patch + create diffs: ``` python3 scripts/update.py --bench bench --comptime ``` Update baseline stats + diffs: ``` python3 scripts/update.py --baseline --bench bench --stats ``` Update stats for a patch + create diffs: ``` python3 scripts/update.py --bench bench --stats ```
f8accf2
to
6d04fa3
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LG. I will land this patch at my convenience.
"-vectorize-loops=false", "-vectorize-slp=false", input_file, "-S" | ||
] | ||
tmp_output = output_file + ".bench_tmp.ll" | ||
cmd += ["-o", tmp_output] |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
cmd += ["-o", tmp_output] | |
cmd += ["-o", tmp_output] if COMPTIME_OUT is None else ['--disable-output'] |
It would make the instruction count more stable.
I will land this patch on Friday. |
Okay, any plans for the benchmarks support scripts? |
Do you mean the scripts for automatic version bumping of benchmarks? |
Err, did you see my above comment, but I mean integration into the ci scripts |
Oh, I got it. I will implement the first option after this patch is verified on the upstream. |
This primary purpose is to support machine specific compile times and move more of the codes to a common python script.
Tested the functionality (without pushing) locally.
This does not come with the ci scripts:
New functionalities:
Update baseline compile time:
Get compile time for a patch + create diffs:
Update baseline stats + diffs:
Update stats for a patch + create diffs: