Re: gnat-gps_5.0-13_m68k FTBFS

2012-12-08 Thread Thorsten Glaser
Ludovic Brenta dixit: > and monitor memory usage of "gnat1" under top. No luck there, it came not even close to swapping, it only went up to some 150M or so. Apparently, there are thus still bugs in GCC (who would have thought ;-). How to debug this on the GCC/GNAT side? Can I force it to write

Re: gnat-gps_5.0-13_m68k FTBFS

2012-12-06 Thread Thorsten Glaser
Ludovic Brenta dixit: > Maybe the compiler is really getting out of memory? How much RAM and > swap does this machine have? You need several hundred megabytes. Ouch, okay. It’s got 768 MiB RAM and a bit of swap, I’ll add more. >> Just running the gcc-4.6 line throws an error much earlier, namel

Re: gnat-gps_5.0-13_m68k FTBFS

2012-12-06 Thread Ludovic Brenta
Thorsten Glaser wrote: gcc-4.6 -c -O0 -gnatafo -gnatVa -gnatn -I- -gnatA /tmp/buildd/gnat-gps-5.0/src_editor/src/src_editor_buffer-line_information.adb gcc-4.6: internal compiler error: Segmentation fault (program gnat1) Maybe the compiler is really getting out of memory? How much RAM and swap

gnat-gps_5.0-13_m68k FTBFS

2012-12-06 Thread Thorsten Glaser
Hi, gnat-gps doesn’t like to be built, it ICEs the compiler; full build log attached. Even if I force it to build with -O1 or -O0, e.g. via (pbuild27944)2|root@aranym:~/gnat-gps-5.0 # debian/rules build-arch CFLAGS=-O0 in the chroot, I get messages like this: […] "/tmp/buildd/gnat-gps-5.0/obj