On Thu, Mar 28, 2019 at 09:15:54PM +1030, Alan Modra wrote: > On Thu, Mar 28, 2019 at 12:10:06AM +1030, Alan Modra wrote: > > Also, using a register move cost of 2 for for power9 direct moves > > gives these fails, even with the .md file tweaks: > > +FAIL: gcc.target/powerpc/vec-set-char.c scan-assembler-not mtvsrwz > > +FAIL: gcc.target/powerpc/vec-set-char.c scan-assembler vspltisb > > +FAIL: gcc.target/powerpc/vec-set-char.c scan-assembler xxspltib > > +FAIL: gcc.target/powerpc/vec-set-int.c scan-assembler-not mtvsrwz > > +FAIL: gcc.target/powerpc/vec-set-int.c scan-assembler vspltisw > > +FAIL: gcc.target/powerpc/vec-set-short.c scan-assembler-not mtvsrwz > > +FAIL: gcc.target/powerpc/vec-set-short.c scan-assembler vspltish > > +FAIL: gcc.target/powerpc/vec-set-short.c scan-assembler xxspltib > > +FAIL: gcc.target/powerpc/vsx-himode3.c scan-assembler lxsihzx > > +FAIL: gcc.target/powerpc/vsx-qimode3.c scan-assembler lxsibzx > > These can be all be fixed by removing "?"s disparaging vector > > alternatives in movsi_internal1 and mov<mode>_internal. > > Like this. Bootstrapped and regression tested powerpc64le-linux. > OK for stage1?
Yup, together with the previous patch. Thanks, Segher
