Chris Elrod

Results 837 comments of Chris Elrod

`@fastmath` let's LLVM use fma instructions, when the host architecture has them, and gives it a little more license to reorder floating point computations. For my own libraries, I have...

I look forward to trying them. So far, I haven't been able to successfully build Julia with LLVM 10.

This should be fixed on Julia master: https://github.com/JuliaLang/julia/pull/34473 Julia and LLVM didn't always agree on alignment of structs containing tuples of `VecElement`s, which would often cause crashes (due to an...

> Is this issue still active? The current benchmark shows: Yes. ```julia julia> @benchmark ($(Ref(A3))[]*$(Ref(A3))[]) BenchmarkTools.Trial: 10000 samples with 1000 evaluations. Range (min … max): 4.645 ns … 16.075 ns...

This comment is sort of a meandering mess, so I at least labeled sections. ## -fveclib= and Numba Clang/LLVM has the optional flag [-fveclib=](https://clang.llvm.org/docs/ClangCommandLineReference.html#cmdoption-clang-fveclib). This is probably what Numba is...

BTW, [SLEEFPirates now](https://github.com/chriselrod/SLEEFPirates.jl/blob/master/deps/build.jl) tries to find `libmvec.so.1` on Linus systems. This is the glibc shared library that contains some SVML functions. If it can find it, it'll use it for...

> That is nice! Does Windows have that too? Not that I know of. Macs have "AppleAccelerate" which may have some functions. I'd accept PRs wrapping it (as well as...

The major improvement these provide is that they're vectorized. If `x` is a scalar, then there isn't much benefit, if there is any at all. Version of LoopVectorization provided an...

@RoyiAvital Technically the code is in [glibc](https://github.com/bminor/glibc/tree/5cb226d7e4e710939cff7288bf9970cb52ec0dfa/sysdeps/x86_64/fpu/multiarch). You could also look at the code for the specific functions you do want to compile. Ubuntu WSL on my Windows laptop has...

The license is just the GPL. > The GNU C Library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public...