LLVM Clang 15 Delivers Some Small x86_64 Performance Improvements But Mostly Flat

[phoronix]

Phoronix: LLVM Clang 15 Delivers Some Small x86_64 Performance Improvements But Mostly Flat

Released last month was LLVM/Clang 15 and since then a number of Phoronix readers have been inquiring about Clang 15 compiler benchmarks or there the lack of on Phoronix. I've been testing various Intel and AMD systems with Clang 15, but it hasn't been particularly exciting -- but that's not a bad thing, just that it doesn't make for particularly compelling benchmark articles...

https://www.phoronix.com/news/LLVM-Clang-15-Benchmarks

[scottishduck]

I’m mostly interested in seeing a feature complete release of llvm-libc. The release of clang forced GCC to make improvements the project otherwise probably wouldn’t have made. Hopefully some competition in the fully featured libc space also creates a competitive mutual improvement with glibc.

[FireBurn]

I’m mostly interested in seeing a feature complete release of llvm-libc. The release of clang forced GCC to make improvements the project otherwise probably wouldn’t have made. Hopefully some competition in the fully featured libc space also creates a competitive mutual improvement with glibc.

I keep looking at github to watch it's progress, but it's difficult to see how far along they are to a fully working libc

[ms178]

What do you think comes first: A fully working libc or a Clang-built glibc?

[brad0]

What do you think comes first: A fully working libc or a Clang-built glibc?

The scope of "fully working" libc is a pretty broad scope and depends where and what it's being used for. Still quite a lot of work to do. The situation with glibc is a typical GNU shitshow.