Why does sparc build for you? or, why are make check results reported when it hasn't bootstrapped since May 5?

[Brad Lucier <lucier at math dot purdue dot edu>]

With the following configure and build script:

#!/bin/tcsh
/bin/rm -rf *; ../configure ; make -j 8 bootstrap >& build.log && (make -j 12 -k check >& check.log ; make mail-report-with-warnings.log; ./mail-report-with-warnings.log)

today's gcc-3.4 fails bootstrap on sparc-sun-solaris2.8 with

/export/mdd0/users/lucier/programs/gcc/gcc-3.4/objdir-sparc/gcc/xgcc -shared-libgcc -B/export/mdd0/users/lucier/programs/gcc/gcc-3.4/objdir-sparc/gcc/ -nostdinc++ -L/export/mdd0/users/lucier/programs/gcc/gcc-3.4/objdir-sparc/sparc-sun-solaris2.8/sparcv9/libstdc++-v3/src -L/export/mdd0/users/lucier/programs/gcc/gcc-3.4/objdir-sparc/sparc-sun-solaris2.8/sparcv9/libstdc++-v3/src/.libs -B/usr/local/sparc-sun-solaris2.8/bin/ -B/usr/local/sparc-sun-solaris2.8/lib/ -isystem /usr/local/sparc-sun-solaris2.8/include -isystem /usr/local/sparc-sun-solaris2.8/sys-include -m64 -DHAVE_CONFIG_H -I. -I../../../../libjava -I./include -I./gcj -I../../../../libjava -Iinclude -I../../../../libjava/include -I../../../../libjava/../boehm-gc/include -DGC_SOLARIS_THREADS=1 -DGC_SOLARIS_PTHREADS=1 -DSILENT=1 -DNO_SIGNALS=1 -DNO_EXECUTE_PERMISSION=1 -DALL_INTERIOR_POINTERS=1 -DJAVA_FINALIZATION=1 -DGC_GCJ_SUPPORT=1 -DATOMIC_UNCOLLECTABLE=1 -I../../../../libjava/libltdl -I../../../../libjava/libltdl -I!
../../../../libjava/.././libjava/../gcc -I../../../../libjava/../zlib -I../../../../libjava/../libffi/include -I../libffi/include -O2 -g -O2 -fno-rtti -fnon-call-exceptions -fdollars-in-identifiers -Wswitch-enum -I/pkgs/XFree86-3.3.3.1/include -W -Wall -D_GNU_SOURCE -DPREFIX=\"/usr/local\" -DLIBDIR=\"/usr/local/lib\" -g -O2 -m64 -MD -MT java/lang/natFloat.lo -MF java/lang/natFloat.pp -c ../../../../libjava/java/lang/natFloat.cc  -fPIC -DPIC -o java/lang/.libs/natFloat.o
../../../../libjava/java/lang/natFloat.cc: In static member function `static 
   jint java::lang::Float::floatToRawIntBits(__java_float)':
../../../../libjava/java/lang/natFloat.cc:42: internal compiler error: in 
   verify_wide_reg, at flow.c:552
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
make[5]: *** [java/lang/natFloat.lo] Error 1

It's stopped at this failure for about a month now.  (I first reported it on
May 5.)  Since that time Mike Stump has asked this list about it, and other
people have claimed that sparc doesn't get enough testing.

At the same time, make check reports have been pouring in for various sparc
architectures.

So, what is going on?  Does sparc work?  I'm using gcc-3.3 as my bootstrap
compiler, and solaris ld/as.  I've tried at least 15 times since May 5
to check out, build, and test the compiler on solaris, and it hasn't
bootstrapped in that period.

Solaris builds have failed for long periods in the past, too, and seem to
get fixed just before a release.  Solaris would get more testing if the
compiler would bootstrap on the architecture.

Brad