Open inglor opened 6 months ago
Thanks for opening your first issue at git-cliff! Be sure to follow the issue template! ⛰️
I don't have much experience with MUSL but I ran a profiler for comparison:
There isn't a big difference here. It might be a good idea to investigate the calls though.
Can you share the repo you are running git-cliff
with?
Unfortunately no since it's internal at work. But can run profiler myself potentially.
Here are some instructions:
git clone https://github.com/orhun/git-cliff && cd git-cliff/
cargo build
samply record target/debug/git-cliff --no-exec -w "$workdir"
rustup target add x86_64-unknown-linux-musl
cargo build --target x86_64-unknown-linux-musl
samply record target/x86_64-unknown-linux-musl/debug/git-cliff --no-exec -w "$workdir"
Then on the browser you can upload profile.json
at top right.
Had any chance to try out the profiler?
I did but showed nothing obvious. If the flamegraph seems the same I'm thinking it has to do with the loading of shared libs before the execution of the program itself. The environment it's running on is in AWS Linux 2 (with an old glibc version that's why I'm using musl) and has some 'enterprise' daemons running so I'm assuming it's scanning everything before it tries to execute them? Speculation from my side.
Yeah, that might be the case. Can you share some execution times? I'm just curious about how slow it is.
Also there is this blog post from 2020 FWIW: https://andygrove.io/2020/05/why-musl-extremely-slow/
Is there an existing issue for this?
Description of the bug
The musl verison of the application is slow compared to the one using glibc one.
Steps To Reproduce
Execute git-cliff musl and observe being slow
Expected behavior
Same performance as glibc
Screenshots / Logs
No response
Software information
Additional context
No response