Backport: add --trace-compile-timing arg to add compile timing comments (#54662)

[kpamnany]

PR Description

% ./julia --trace-compile=stderr --trace-compile-timing --start=no -e "using ProfileView"
#=   0.0 ms =# precompile(Tuple{typeof(Requires.__init__)})
#=  19.4 ms =# precompile(Tuple{typeof(Requires.isprecompiling)})
#=   0.0 ms =# precompile(Tuple{typeof(Requires.listenpkg), Any, Base.PkgId})
#=   0.0 ms =# precompile(Tuple{typeof(Requires.loaded), Base.PkgId})
#=  13.4 ms =# precompile(Tuple{typeof(Base.haskey), Base.Dict{Base.PkgId, Module}, Base.PkgId})
#=   0.0 ms =# precompile(Tuple{typeof(Requires.callbacks), Base.PkgId})
#=   0.0 ms =# precompile(Tuple{typeof(Libiconv_jll.__init__)})
#=   0.0 ms =# precompile(Tuple{typeof(Libiconv_jll.find_artifact_dir)})
#=  32.0 ms =# precompile(Tuple{typeof(Artifacts._artifact_str), Module, String, Base.SubString{String}, String, Base.Dict{String, Any}, Base.SHA1, Base.BinaryPlatforms.Platform, Any})
#=   1.5 ms =# precompile(Tuple{typeof(Base.invokelatest), Any})

Checklist

Requirements for merging:

• I have opened an issue or PR upstream on JuliaLang/julia: add --trace-compile-timing arg to add compile timing comments JuliaLang/julia#54662
• I have removed the port-to-* labels that don't apply.
• I have opened a PR on raicode to test these changes: https://github.com/RelationalAI/raicode/pull/21153

[NHDaly]

I'm surprised the original PR had to add a measurement here... We already have a metric for compilation time -- why isn't this incremented there?

[NHDaly]

For example, here: https://github.com/JuliaLang/julia/blob/fe864374d50b121aeb3ce5473c206338b21df5e5/src/gf.c#L4390-L4396

[NHDaly]

Asked here: JuliaLang#54662 (review)

[NHDaly]

Anyway LGTM for now! 👍

I'll raise this question upstream