julia: add cpu_target variant to let user customise JULIA_CPU_TARGET

[giordano]

This is useful when compiling julia on heterogeneous systems where we want to provide a single sysimage usable on different nodes.

Note that this currently doesn't work particularly well because values of variants don't seem to allow commas, and JULIA_CPU_TARGET needs commas to separate target features.

[giordano]

Note that this currently doesn't work particularly well because values of variants don't seem to allow commas, and JULIA_CPU_TARGET needs commas to separate target features.

The variant now takes pipes, and they're internally replaced to commas. I verified I could successfully and correctly build the spec

julia cpu_target='generic;sandybridge|-xsaveopt|clone_all;haswell|-rdrnd|base(1);znver4|-rdrnd|base(1)' ^llvm~clang~gold~polly ^openblas threads=openmp

and then

julia> Base.parse_image_targets(Base.parse_cache_header(joinpath(dirname(Sys.BINDIR), "share", "julia", "compiled", "v1.11", "Pkg", "tUTdb_liHqa.ji"))[7])
4-element Vector{Base.ImageTarget}:
 generic; flags=0; features_en=(cx16)
 sandybridge; flags=0; features_en=(sse3, pclmul, ssse3, cx16, sse4.1, sse4.2, popcnt, xsave, avx, sahf)
 haswell; flags=0; features_en=(sse3, pclmul, ssse3, fma, cx16, sse4.1, sse4.2, movbe, popcnt, xsave, avx, f16c, fsgsbase, bmi, avx2, bmi2, sahf, lzcnt)
 znver4; flags=0; features_en=(sse3, pclmul, ssse3, fma, cx16, sse4.1, sse4.2, movbe, popcnt, aes, xsave, avx, f16c, fsgsbase, bmi, avx2, bmi2, avx512f, avx512dq, adx, avx512ifma, clflushopt, clwb, avx512cd, sha, avx512bw, avx512vl, avx512vbmi, pku, avx512vbmi2, shstk, gfni, vaes, vpclmulqdq, avx512vnni, avx512bitalg, avx512vpopcntdq, rdpid, sahf, lzcnt, sse4a, prfchw, mwaitx, xsavec, xsaves, clzero, wbnoinvd, avx512bf16)

which confirms the pkgimage of the the Pkg.jl standard library was indeed compiled for the requested targets.

[haampie]

Hi @giordano, this package-related pull request can be migrated from spack/spack to spack/spack-packages using the migrate-pkg-prs utility.

We encourage authors to run the migration script themselves to preserve author attribution.

You can migrate all your open pull requests at once, following the steps in the documentation.

---

This comment was posted automatically.