Hi Kewen, On Wed, Jan 16, 2019 at 10:08:46PM +0800, Kewen.Lin wrote: > On target rs6000, for all types, use of vec_ld/vec_st historically permits > the dereferenced pointer to be of a scalar type or the corresponding vector > type. But for vector unsigned/signed long long and double, we have an > omission in our table for vec_ld/vec_st. > > Ok for trunk, and eventual backport to GCC 7 and 8 aftersome burn-in time?
This looks fine, thanks! Okay for trunk. Also okay for 8 and 7 (do the backports in that order please) after a week or so. Segher > 2019-01-16 Kewen Lin <li...@gcc.gnu.org> > > * doc/extend.texi: Add four new prototypes for vec_ld and seven new > prototypes for vec_st. > * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add entries > for scalar address type variants of altivec_vec_ld/altivec_vec_st, > mainly on signed/unsigned long long and double. > > gcc/testsuite/ChangeLog > > 2019-01-16 Kewen Lin <li...@gcc.gnu.org> > > * gcc.target/powerpc/altivec_vld_vst_addr.c: New test.
