Re: [PATCH] compiler-gcc: get back Clang build

[Nick Desaulniers]

On Tue, Aug 21, 2018 at 9:45 AM Joe Perches <joe@xxxxxxxxxxx> wrote:
>
> On Tue, 2018-08-21 at 09:32 -0700, Nick Desaulniers wrote:
> > On Tue, Aug 21, 2018 at 5:38 AM Dominique Martinet
> > <asmadeus@xxxxxxxxxxxxx> wrote:
> > >
> > > Nick Desaulniers Aug. 21, 2018, 8:09 a.m. UTC:
> > > > Thanks for noticing, and sending this patch.  I'm happy to see others
> > > > testing with Clang.  I noticed this too near the end of the day
> > > > https://github.com/ClangBuiltLinux/linux/issues/27.
> > >
> > > FWIW libbcc so many BPF users also use clang, so this has more impact
> > > than just testing to build linux with clang (not that this would be any
> > > reason to delay fixing either way)
> > >
> > > I would tend to agree havin a compiler-common + make clang/intel not
> > > include compiler-gcc would probably be best in the long run but we might
> > > want a quick fix for 4.19 meanwhile..
> >
> > That's fair. SOP here is quick (full) revert, then come up with a
> > better fix.  And I do prefer Masahiro's partial revert to a full
> > revert of Joe's patch.  That will give us more time to develop the
> > proper fix rather than rush.  I'll try to see how we can more properly
> > split the compiler specific headers.
> >
> > Tested with gcc-7 and clang-8.
>
> clang-8? Isn't the latest officlal clang 6.0.1 ?

Yes, but I have a local llvm tree that I work out of, that's in my
$PATH, so my version of clang is never too far behind Top of Tree.
For android, we're using clang-5, but currently staging an upgrade to
clang 6.0.1.

> So if something other than 6.0.x is required,
> then some additional check should probably be
> added to compiler-clang.h as well.
>

Sure, but that doesn't need to go in Mashiro's patch today.  That can
wait for a proper separation between compiler headers where we can
then implement improved version checks.

-- 
Thanks,
~Nick Desaulniers