[Bug target/78056] [7 Regression] build failure on Power7
amodra at gmail dot com
gcc-bugzilla@gcc.gnu.org
Sun Oct 23 02:27:00 GMT 2016
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78056
--- Comment #13 from Alan Modra <amodra at gmail dot com> ---
Thanks Martin, so it is HAVE_AS_POWER9 being undefined, which then results in
TARGET_P9_VECTOR and others being defined to zero. Kelvin's fix of not
defining the builtins when no power9 support is available sounds right to me
(but best to use derived TARGET_P9_VECTOR etc. rather than HAS_AS_POWER9).
I also think that invoking gcc with -mcpu=power9 ought to complain if
HAVE_AS_POWER9 is not defined, ie. tests like
if (TARGET_P9_VECTOR != ((rs6000_isa_flags & OPTION_MASK_P9_VECTOR) != 0))
error ("-mpower9-vector requires assembler support for power9");
be added to rs6000_override_options.
More information about the Gcc-bugs
mailing list