https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68115
John David Anglin <danglin at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|WAITING |NEW
--- Comment #3 from John David Anglin <danglin at gcc dot gnu.org> ---
The problem comes from the configure command used to build libbacktrace:
$ /test/gnu/gcc/gcc/libbacktrace/configure --srcdir=../../../gcc/libbacktrace
--cache-file=./config.cache --enable-multilib --with-gnu-as
--with-as=/opt/gnu64/bin/as --with-ld=/usr/ccs/bin/ld --enable-shared
--with-local-prefix=/opt/gnu64 --prefix=/opt/gnu64/gcc/gcc-6
--enable-threads=posix --with-gmp=/opt/gnu64/gcc/gmp --enable-checking=release
--enable-languages=c,c++,fortran,lto,objc,obj-c++
--program-transform-name=s,y,y, --disable-option-checking
--with-target-subdir=hppa64-hp-hpux11.11 --build=hppa64-hp-hpux11.11
--host=hppa64-hp-hpux11.11 --target=hppa64-hp-hpux11.11
The --with-target-subdir option overrides the configure test for __sync
support:
# Test for __sync support.
AC_CACHE_CHECK([__sync extensions],
[libbacktrace_cv_sys_sync],
[if test -n "${with_target_subdir}"; then
libbacktrace_cv_sys_sync=yes
else
AC_LINK_IFELSE(
[AC_LANG_PROGRAM([int i;],
[__sync_bool_compare_and_swap (&i, i, i);
__sync_lock_test_and_set (&i, 1);
__sync_lock_release (&i);])],
[libbacktrace_cv_sys_sync=yes],
[libbacktrace_cv_sys_sync=no])
fi])
BACKTRACE_SUPPORTS_THREADS=0
if test "$libbacktrace_cv_sys_sync" = "yes"; then
BACKTRACE_SUPPORTS_THREADS=1
AC_DEFINE([HAVE_SYNC_FUNCTIONS], 1,
[Define to 1 if you have the __sync functions])
fi
AC_SUBST(BACKTRACE_SUPPORTS_THREADS)
Probably introduced by:
2015-08-23 Francois-Xavier Coudert <[email protected]>
PR libfortran/54572
* Makefile.def: Make libgfortran depend on libbacktrace.
* Makefile.in: Regenerate.
Guess following would fix:
case "${host}" in
hppa*-*-hpux*) libbacktrace_cv_sys_sync=no ;;
*) libbacktrace_cv_sys_sync=yes ;;
esac
Still it would seem test could be done on a native build.
