[Bug target/78336] powerpc-darwin bootstrap broken (probably by 241930)

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78336

--- Comment #2 from Iain Sandoe <iains at gcc dot gnu.org> ---
(In reply to David Edelsohn from comment #1)
> I thought that Darwin specifically picked up the definition from
> config/darwin.h.
> 
> I don't understand the failure because the rs6000.h logic did not change and
> rs6000_asm_weaken_decl is declared in rs6000-protos.h.
> 
> This bug report does not report the actual failure mode, it only provides a
> patch, so there is no way to review the patch.  I only can guess at the
> problem.
> 
> Am I to infer that rs6000_asm_weaken_decl function doesn't compile on Darwin
> because SET_ASM_OP is not defined?  Again, Darwin will not actually use the
> function because the logic was not changed.

It doesn't compile because ASM_OUTPUT_DEF is not defined for darwin.

My first inclination was define ASM_OUTPUT_DEF (to say gcc_unreachable())
however, then the compilation breaks elsewhere (since varasm.c has different
code paths enabled on the availability of ASM_OUTPUT_DEF).