This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

gcc-2.95.3, but not gcc-2.95.2, requires gas on hppa2.0-hp-hpux10.20


>>>>> "Dave" == John David Anglin <dave@hiauly1.hia.nrc.ca> writes:

>>>>> "Bernd" == Bernd Schmidt <bernds@redhat.com> writes:
Bernd> On Mon, 22 Jan 2001, Martin Buchholz wrote:
>>> Executive summary: gcc-2.95.3.test2 has a C++ regression on hppa2.0-hp-hpux10.20

>>> I tried to bootstrap gcc-2.95.2 on hppa2.0-hp-hpux10.20 in the same
>>> way, with objdir != srcdir, and srcdir pristine.
>>> 
>>> It fails in compare:
>>> 
>>> Bootstrap comparison failure!
>>> alias.o differs
>>> alloca.o differs
>>> bitmap.o differs
>>> c-aux-info.o differs

Dave> My first guess would be a mismatch in bison version.  

It is a gcc release engineering bug if bison (or yacc) is ever invoked
when building the release tarball (as opposed to CVS extracts.)  The
architecture-independent bison output should be pre-built.

Dave> Anther known issue is some versions of HP's sed.  

It is a gcc bug if gcc's configure does not detect broken seds and
either work around them or abort the build.

Dave> I recommend the latest binutils for the assembler (configure
Dave> with --with-gnu-as).

I reran the build.  Here is what I get this time:

make[3]: Entering directory `/usr/local/xemacs/gcc-2.95.3/gcc'
./xgcc -B/usr/local/hppa2.0-hp-hpux10.20/bin/ -B./ -I/usr/local/hppa2.0-hp-hpux10.20/include -O2   -DIN_GCC -DHAIFA    -O2 -O2 -I./include  -fPIC -g1  -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED   -O2 -I. -I/project/xemacs/import/gcc-2.95.3.test2/gcc -I/project/xemacs/import/gcc-2.95.3.test2/gcc/config -I/project/xemacs/import/gcc-2.95.3.test2/gcc/../include \
  -c /project/xemacs/import/gcc-2.95.3.test2/gcc/cp/tinfo.cc
cc1plus: warning: -g is only supported when using GAS on this processor,
cc1plus: warning: -g option disabled.
as: error 7403: undefined label - L$0175 (7403)

It does seem like a gasism has infected the generated assembler code.

To test this hypothesis, I built a binutils-2.10 for my
hppa2.0-hp-hpux10.20 system, and then retried bootstrapping
gcc-2.95.3.test2 again, adding --with-as=/path/to/as to the
configure command.

This time, tinfo.cc built just fine.

$cdi/gcc-2.95.3.test2/configure --prefix=/usr/local --with-as=$cdxw/binutils/bin/as --enable-languages='c,c++' --disable-nls  && make bootstrap

make[3]: Entering directory `/usr/local/xemacs/gcc-2.95.3/gcc'
./xgcc -B/usr/local/hppa2.0-hp-hpux10.20/bin/ -B./ -I/usr/local/hppa2.0-hp-hpux10.20/include -O2   -DIN_GCC -DHAIFA    -O2 -O2 -I./include  -fPIC -g1  -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED   -O2 -I. -I/project/xemacs/import/gcc-2.95.3.test2/gcc -I/project/xemacs/import/gcc-2.95.3.test2/gcc/config -I/project/xemacs/import/gcc-2.95.3.test2/gcc/../include \
  -c /project/xemacs/import/gcc-2.95.3.test2/gcc/cp/tinfo.cc
make[3]: Leaving directory `/usr/local/xemacs/gcc-2.95.3/gcc'

And `make bootstrap' succeeded completely, including the compare step.

I know the Red Hat folks tend to build gcc as part of a complete tool
chain including GNU as, but I recommend that allowing builds using
just HP-UX as continue to be supported.  Else, one might have to
bootstrap gcc on HP-UX using:

build gcc using --enable-languages=c and install
build binutils using the installed gcc C compiler
build gcc again using --enable-languages='c,c++' -with-as=/path/to/gnu-as

In any case, if HP-UX as is not supported, configure should let the
user know.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]