Re: [Gm2] Build Problems on Mac

[Gaius Mulley]

Michael Lambert <address@hidden> writes:

> Build environment:  MacOS 10.4.8 using the latest Apple compiler tools
> (Xcode 2.4.1).  CVS sources as of about 0000 GMT today.  gcc-4.1.1.
> 
> PowerPC G5:
> 
> ./xgm2 -Wnil -Wcase -Wbounds -Wreturn -g  -funbounded-by-reference
> -Wverbose-unbounded -Wpim -Wpedantic-param-names -Wextended-opaque
> -Wpedantic-cast -c -B./stage1/gm2
> -I../../gcc-4.1.1/gcc/gm2/ulm-lib-gm2/std:../../gcc-4.1.1/gcc/gm2/ulm-lib-gm2/sys:../../gcc-4.1.1/gcc/gm2/gm2-libs
>  ../../gcc-4.1.1/gcc/gm2/ulm-lib-gm2/std/TimeIO.mod
> -o gm2/ulm-lib-gm2/std/TimeIO.o 
> Internal error: in make_decl_rtl, at varasm.c:890
> Please report this crash to the GNU Modula-2 mailing list
> <address@hidden>
> make[2]: *** [gm2/ulm-lib-gm2/std/TimeIO.o] Error 1
> make[1]: *** [all-gcc] Error 2 
> make: *** [all] Error 2
> 
> Intel Core 2 Duo:
> 
> ./xgm2 -Wnil -Wcase -Wbounds -Wreturn -g  -funbounded-by-reference
> -Wverbose-unbounded -Wpim -Wpedantic-param-names -Wextended-opaque
> -Wpedantic-cast -c -B./stage1/gm2
> -I../../gcc-4.1.1/gcc/gm2/ulm-lib-gm2/std:../../gcc-4.1.1/gcc/gm2/ulm-lib-gm2/sys:../../gcc-4.1.1/gcc/gm2/gm2-libs
>  ../../gcc-4.1.1/gcc/gm2/ulm-lib-gm2/std/TimeIO.mod
> -o gm2/ulm-lib-gm2/std/TimeIO.o 
> Internal error: in make_decl_rtl, at varasm.c:890 
> Please report this crash to the GNU Modula-2 mailing list
> <address@hidden> 
> make[2]: *** [gm2/ulm-lib-gm2/std/TimeIO.o] Error 1
> make[1]: *** [all-gcc] Error 2 
> make: *** [all] Error 2
> 
> I get exactly the same error (on the Intel at least) when building
> installing the FSF gcc-4.1.1 C compiler first and using it for the gm2
> build.
> 
> Michael

Hi Michael,

thanks for the report, I've checked in some changes which appear to
have broken the above.  The changes fix a few optimisation test cases
but they have broken the building of the ulm libraries (which use
multilevel exports among other features).  The good news is that
unit_at_a_time optimisation is working for simple modules.
I also think I've uncovered some of the linking bugs people were
reporting recently - which manifested themselves on non GNU systems.

I'll try and fix the above promptly, tomorrow (for the non optimised
case).

regards,