Failed build on mips-sgi-irix5.3 for 1998-06-28 snapshot

Failed build on mips-sgi-irix5.3 for 1998-06-28 snapshot

[Daniel Karipides]

Here's my latest attempt on trying to build a snapshot on irix5.3.
I've noticed that the regular build reports have stopped including
SGI targets.  Fingers crossed that this gets addressed before
egcs1.1. :)

The details:

config.guess: mips-sgi-irix5.3
configure options: --with-gnu-as
binutils version: 2.9.1
make: gnumake 3.76

build directory completely separate from tarball directory.

Tail of the buildlog shown below.  I may try to revert to
binutils 2.8.1 (which some on the list says will help), but
I kind of wanted to try some exception handling.  I will
also try the 1.0.3 release.

As usual, hope this info is helpful.

-Dan
-----
karipid@mae.cornell.edu

-----------------------------------------------------------
case "compare" in compare | compare-lean ) stage=2 ;; * ) stage=`echo compare |
sed -e 's,^compare\([0-9][0-9]*\).*,\1,'` ;; esac; \
for dir in tmp-foo cp f objc; do \
  if [ "`echo $dir/*.o`" != "$dir/*.o" ] ; then \
    for file in $dir/*.o; do \
      tail +16c ./$file > tmp-foo1; \
      tail +16c stage$stage/$file > tmp-foo2 \
        && (cmp tmp-foo1 tmp-foo2 > /dev/null 2>&1 || echo $file differs >> .bad
_compare) || true; \
    done; \
  fi; \
done
rm -f tmp-foo*
case "compare" in compare | compare-lean ) stage=2 ;; * ) stage=`echo compare |
sed -e 's,^compare\([0-9][0-9]*\).*,\1,'` ;; esac; \
if [ -f .bad_compare ]; then \
  echo "Bootstrap comparison failure!"; \
  cat .bad_compare; \
  exit 1; \
else \
  case "compare" in \
    *-lean ) rm -rf stage$stage ;; \
    *) ;; \
  esac; true; \
fi
Bootstrap comparison failure!
xcoffout.o differs
f/storag.o differs
make[1]: *** [compare] Error 1
make[1]: Leaving directory `/scratch/egcsobjdir/gcc'
make: *** [bootstrap] Error 2

Re: Failed build on mips-sgi-irix5.3 for 1998-06-28 snapshot

[Carlo Wood]

| Bootstrap comparison failure!
| xcoffout.o differs
| f/storag.o differs

I had a comparison failure once and it was
caused by a different version string in the
object code;
A 'make clean' should fix troubles like this
of course - you might want to take this
possibility into account when looking for
what is wrong.

-- 
 Carlo Wood  <carlo@runaway.xs4all.nl>

PS The version string I am refering to is
   also visible in a *.s file:

.ident  "GCC: (GNU) egcs-2.91.42 19980626 (gcc2 ss-980502 experimental)"

   For some reason my `ident' doesn't pick these up - but with
   some hacking...

>hexdump -e '"%_p"' gcc/stage1/xcoffout.o | grep -B4 experimental
GC*
: (GNU) egcs-2.91.43 19*
80628 (gc*
2 s*
-980502 experimental).*

>hexdump -e '"%_p"' gcc/stage2/xcoffout.o | grep -B4 experimental
GC*
: (GNU) egcs-2.91.45 19*
80702 (gc*
2 s*
-980609 experimental).*

Can you say "different" ?  Hmm.

Re: Failed build on mips-sgi-irix5.3 for 1998-06-28 snapshot

[Jeffrey A Law]

  In message < 199807011341.JAA17053@scotch.mae.cornell.edu >you write:
  > 
  > Here's my latest attempt on trying to build a snapshot on irix5.3.
  > I've noticed that the regular build reports have stopped including
  > SGI targets.  Fingers crossed that this gets addressed before
  > egcs1.1. :)
It is possible that the recent aliasing changes are causing this failure;
I saw bootstrap failures on the PA recently due to problems in that code.

We've fixed those problems, so hopefully you'll be able to bootstrap the
next snapshot :-)
jeff