[PATCH] D28791: [compiler-rt][crt] Simple crtbegin and crtend implementation

[Joerg Sonnenberger via llvm-commits]

On Wed, Dec 19, 2018 at 12:57:15PM -0800, Petr Hosek wrote:
> On Fri, Dec 14, 2018 at 10:47 AM Joerg Sonnenberger via llvm-commits <
> llvm-commits at lists.llvm.org> wrote:
> 
> > On Fri, Dec 14, 2018 at 01:27:14AM +0000, Eric Christopher via Phabricator
> > via llvm-commits wrote:
> > > Joerg: We asked some very specific questions to which there don't
> > > appear to be answers coming from you and this patch doesn't affect any
> > > other target negatively here. Unless you come up with a compelling
> > > argument not to commit this other than "some platforms have support
> > > for this in other places" I think we need to commit this.
> >
> > I have no idea what else I can say other than "This is not a generic
> > implementation. Please do not pretend it to be one".
> 
> 
> Can you provide more details? Which part of the implementation is not
> generic? I've deliberately omitted support for GCJ since Clang doesn't
> support it, but aside from that it should be.

Read the history of the review? I gave the list. I don't really care
about the Java support, that's mostly a historical outlier. The rest is
quite different.

> > Even more so that I
> > strongly believe that especially GNU should gets its act together and
> > ship a working copy with glibc.
> >
> 
> None of the popular Linux libc implementations (I looked at glibc, musl,
> newlib, uclibc and dietlibc) provide crtbegin/crtend aside from Bionic.
> This is how crtbegin/crtend is handled on Linux. You may disagree with it,
> but I don't think this is the place to argue about that design. My goal is
> to ensure that Clang supports this platform without relying on other
> compilers as it does today.

"All the world is GCC" has been the default assumption for ages. Short
of complaining nothing is going to change...

Joerg