On Thu, Jul 11, 2019 at 06:09:10PM +0100, Gaius Mulley wrote: > Segher Boessenkool <seg...@kernel.crashing.org> writes: > > > On Thu, Jul 11, 2019 at 12:49:44PM +0100, Gaius Mulley wrote: > >> Matthias Klose <d...@ubuntu.com> writes: > >> > powerpc64le-linux-gnu fails to build (search for "unfinished" in the > >> > build log) > >> > > >> > during RTL pass: final > >> > ../../src/gcc/gm2/gm2-libs-coroutines/SYSTEM.def: In function > >> > '_M2_SYSTEM_init': > >> > ../../src/gcc/gm2/gm2-libs-coroutines/SYSTEM.def:20: internal compiler > >> > error: in > >> > rs6000_output_function_epilogue, at conf > >> > ig/rs6000/rs6000.c:29169 > > > > We don't yet support Modula2 in the traceback tables. I'll add it. > > What is the exact spelling in lang_hooks.name? "GNU Modula 2"? Or > > "GNU Modula2"? Or what else :-) > > Hi Segher, > > very close ! > > #define LANG_HOOKS_NAME "GNU Modula-2" > > >> Hi Matthias, > >> > >> many thanks for the build results. There is a newer version of > >> gm2/Make-lang.in on the 9.1.0 branch which (quietens the output for mc) > >> and fixes the gm2l execvp (on other platforms). > >> > >> The ICE is very interesting! > > > > Should I add this to the GCC 9 branch as well? Does that help testing, > > I mean? > > yes please - all that helps is good,
I committed support for this to both trunk and 9 yesterday, fwiw. Segher