Re: Possible problems again with alpha and the fsirl patch (fwd)

["Christopher C. Chimelis" <chris@debian.org>]

One more follow-up...with the fsirl patch, compilation of gcc still fails 
(see below).   I *might* be able to correct this *if* it's meant to
produce proper DWARF debugging output.  Otherwise, I'm faced with a
terrible dilemma: leave 2.11.90.0.7-2 in woody for the freeze or leave a
compiler in woody for the freeze that will break subsequent binutils.

C

Here's the failure:


for name in _eh; \
do \
  echo ${name}; \
  ./xgcc -B/usr/alpha-linux/bin/ -B./ -I/usr/alpha-linux/include -O2
-DIN_GCC -DHAIFA    -g -I./include  -fPIC -g1 -DHAVE_GTHR_DEFAULT
-DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED    -fexceptions
-I. -I/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc
-I/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc/config
-I/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc/../include
-c \
      -DL${name}
/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc/libgcc2.c
-o ${name}.o; \
  if [ $? -eq 0 ] ; then true; else exit 1; fi; \
  ar rc tmplibgcc2.a ${name}.o; \
  rm -f ${name}.o; \
done
_eh
/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc/libgcc2.c: In
function `__throw':
/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc/libgcc2.c:3935: Internal
compiler error:
/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/src-native/gcc/libgcc2.c:3935: internal
error--insn does not satisfy its constraints:
(insn 204 176 178 (set (reg:DI 23 $23)
        (plus:DI (reg:DI 30 $30)
            (reg:DI 42 $f10))) 9 {adddi3} (insn_list 77 (insn_list 70
(nil)))
    (nil))
make[6]: *** [libgcc2.a] Error 1
make[6]: Leaving directory
`/src/gcc/2.95.4/2.95.4.ds3-0.010629/gcc-2.95-2.95.4.ds3/build-native/gcc'
make[5]: *** [stmp-multilib-sub] Error 2



On Tue, 17 Jul 2001, Christopher C. Chimelis wrote:

> Ack!  Didn't realise that we still weren't including that patch....
> 
> Well, I'm trying to compile WITH the patch again :-P  Either way, I'll
> figure this out...
> 
> C
> 
> On Tue, 17 Jul 2001, Christopher C. Chimelis wrote:
> 
> > 
> > Oops...forgot to cc the list :-P
> > 
> > 
> > ---------- Forwarded message ----------
> > Date: Tue, 17 Jul 2001 01:10:39 -0400 (EDT)
> > From: Christopher C. Chimelis <chris@debian.org>
> > To: Matthias Klose <doko@cs.tu-berlin.de>
> > Subject: Possible problems again with alpha and the fsirl patch
> > 
> > 
> > I just spent at least two hours trying to debug a segfault with a binutils
> > version that I'm packaging.  About 10 mins ago, I tried compiling with
> > gcc-3.0 and the segfault went away.  From what I gather, it's the DWARF
> > output causing problems, which is only modified (for non-arch-specific
> > files) by the fsirl patch.  I can't tell for sure (yet) if that's the
> > problem since I'm still recompiling gcc without the patch to find out, but
> > it's almost definite that the dwarf output is what's causing the problem.
> > 
> > I'll keep you posted...just wanted to let you know that I'm investigating
> > a problem and that it is probably gcc-related (2.95.4).
> > 
> > C
> > 
> > 
> > 
> 
>