Lines Matching refs:ACC3
107 my $ACC3="%ymm3";
204 vmovdqu 32*3-128($np), $ACC3
214 vmovdqu $ACC3, 32*3-128($np)
227 vmovdqu 32*3-128($ap), $ACC3
250 vpaddq $ACC3, $ACC3, $ACC3
253 vmovdqa $ACC3, 32*2-128($aap)
272 vpmuludq $B1, $ACC3, $ACC3
302 vpmuludq 32*2-128($aap), $B1, $ACC3
303 vpaddq 32*3-192($tp0), $ACC3, $ACC3
322 vpaddq $TEMP1, $ACC3, $ACC3
338 vmovdqu $ACC3, 32*3-192($tp0)
384 vpmuludq 32*7-128($aap), $B1, $ACC3
386 vpaddq 32*12-448($tp1), $ACC3, $ACC3
397 vpaddq $TEMP1, $ACC3, $ACC3
406 vpaddq $TEMP0, $ACC3, $ACC3
414 vmovdqu $ACC3, 32*12-448($tp1)
469 vmovdqu 32*3-192($tp0), $ACC3
523 vpaddq $TEMP2, $ACC3, $ACC3
560 vpaddq $TEMP1, $ACC3, $ACC3
604 vpaddq $TEMP0, $ACC3, $ACC3
646 vpaddq $TEMP2, $ACC3, $ACC2
652 vpaddq $TEMP0, $ACC4, $ACC3
688 vpaddq 32*12-448($tp1), $ACC3, $ACC3
702 vpsrlq \$29, $ACC3, $TEMP4
704 vpand $AND_MASK, $ACC3, $ACC3
716 vpaddq $TEMP3, $ACC3, $ACC3
726 vpsrlq \$29, $ACC3, $TEMP4
728 vpand $AND_MASK, $ACC3, $ACC3
742 vpaddq $TEMP3, $ACC3, $ACC3
745 vmovdqu $ACC3, 32*3-128($rp)
864 my $ACC3="%ymm3";
971 vmovdqu 32*3-128($np), $ACC3
984 vmovdqu $ACC3, 32*3-128($np)
985 vpxor $ACC3, $ACC3, $ACC3
1016 vpsrlq \$29, $ACC3, $ACC9 # correct $ACC3(*)
1042 vpand $AND_MASK, $ACC3, $ACC3 # correct $ACC3
1043 vpaddq $TEMP2,$ACC3,$ACC3
1051 vpermq \$0x93, $ACC9, $ACC9 # correct $ACC3
1077 vpaddq $TEMP1,$ACC3,$ACC3
1085 vpblendd \$3, $ZERO, $ACC9, $ACC9 # correct $ACC3
1088 vpaddq $ACC9, $ACC3, $ACC3 # correct $ACC3
1116 vpaddq $TEMP0,$ACC3,$ACC3
1156 vpaddq $TEMP2,$ACC3,$ACC3
1196 vpaddq $TEMP2,$ACC3,$ACC3
1233 vpaddq $TEMP2,$ACC3,$ACC3
1271 vpaddq $TEMP2,$ACC3,$ACC3
1307 vpaddq $TEMP2,$ACC3,$ACC2
1310 vpaddq $TEMP0,$ACC4,$ACC3
1332 # (*) Original implementation was correcting ACC1-ACC3 for overflow
1355 vpsrlq \$29, $ACC3, $TEMP4
1357 vpand $AND_MASK, $ACC3, $ACC3
1369 vpaddq $TEMP3, $ACC3, $ACC3
1379 vpsrlq \$29, $ACC3, $TEMP4
1381 vpand $AND_MASK, $ACC3, $ACC3
1393 vpaddq $TEMP3, $ACC3, $ACC3
1399 vmovdqu $ACC3, 96-128($rp)
