How to reproduce the benchmark result? #742
-
|
I'd like to try if I can do some optimizations or not. Is there any instruction on how to conduct a benchmark as you show in the README? |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments 2 replies
-
|
Just build programs with a verbose option (which depends on build system) to see what is actual linker command line arguments. Then invoke the linker with the same arguments with |
Beta Was this translation helpful? Give feedback.
-
|
Any data regarding memory usage? vs lld or gold? I tried 1.6.0 on our project (private; distcc + ccache; mold -run make -j72) and near the end of build process it failed “mmap: failed”. |
Beta Was this translation helpful? Give feedback.
-
|
I have made a bunch of comparative measurements, including memory use, on a mid-sized developer laptop. You can find the results here: https://github.com/sicherha/linker-benchmarks/ |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, so slightly worse than lld. |
Beta Was this translation helpful? Give feedback.
-
|
Can you copy-n-paste the error message? |
Beta Was this translation helpful? Give feedback.
Just build programs with a verbose option (which depends on build system) to see what is actual linker command line arguments. Then invoke the linker with the same arguments with
time.