@@ -1804,7 +1804,7 @@ vunpcklps (%rax){1to16}, %zmm17, %zmm19 {z}{k1}
1804
1804
# CHECK-NEXT: 2 10 1.00 * vpermq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z}
1805
1805
# CHECK-NEXT: 36 8 8.00 * vpscatterdd %zmm1, (%rdx,%zmm0,4) {%k1}
1806
1806
# CHECK-NEXT: 19 7 4.00 * vpscatterdq %zmm1, (%rdx,%ymm0,4) {%k1}
1807
- # CHECK-NEXT: 19 7 4.00 * vpscatterqd %ymm1, (%rdx,%zmm0,4) {%k1}
1807
+ # CHECK-NEXT: 1 1 0.50 * vpscatterqd %ymm1, (%rdx,%zmm0,4) {%k1}
1808
1808
# CHECK-NEXT: 19 7 4.00 * vpscatterqq %zmm1, (%rdx,%zmm0,4) {%k1}
1809
1809
# CHECK-NEXT: 1 1 1.00 vpshufd $0, %zmm16, %zmm19
1810
1810
# CHECK-NEXT: 2 8 1.00 * vpshufd $0, (%rax), %zmm19
@@ -1871,7 +1871,7 @@ vunpcklps (%rax){1to16}, %zmm17, %zmm19 {z}{k1}
1871
1871
# CHECK-NEXT: 2 8 1.00 * vpunpcklqdq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z}
1872
1872
# CHECK-NEXT: 36 7 8.00 * vscatterdps %zmm1, (%rdx,%zmm0,4) {%k1}
1873
1873
# CHECK-NEXT: 19 7 4.00 * vscatterdpd %zmm1, (%rdx,%ymm0,4) {%k1}
1874
- # CHECK-NEXT: 19 7 4.00 * vscatterqps %ymm1, (%rdx,%zmm0,4) {%k1}
1874
+ # CHECK-NEXT: 1 1 0.50 * vscatterqps %ymm1, (%rdx,%zmm0,4) {%k1}
1875
1875
# CHECK-NEXT: 19 7 4.00 * vscatterqpd %zmm1, (%rdx,%zmm0,4) {%k1}
1876
1876
# CHECK-NEXT: 1 3 1.00 vshuff32x4 $0, %zmm16, %zmm17, %zmm19
1877
1877
# CHECK-NEXT: 2 10 1.00 * vshuff32x4 $0, (%rax), %zmm17, %zmm19
@@ -2054,7 +2054,7 @@ vunpcklps (%rax){1to16}, %zmm17, %zmm19 {z}{k1}
2054
2054
2055
2055
# CHECK: Resource pressure per iteration:
2056
2056
# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]
2057
- # CHECK-NEXT: - 612.00 411 .17 103 .67 327.50 327.50 48 .50 593 .17 6 .00 48 .50 48 .50 48 .50
2057
+ # CHECK-NEXT: - 612.00 408 .17 102 .67 327.50 327.50 41 .50 592 .17 5 .00 41 .50 41 .50 41 .50
2058
2058
2059
2059
# CHECK: Resource pressure by instruction:
2060
2060
# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] Instructions:
@@ -2774,7 +2774,7 @@ vunpcklps (%rax){1to16}, %zmm17, %zmm19 {z}{k1}
2774
2774
# CHECK-NEXT: - - - - 0.50 0.50 - 1.00 - - - - vpermq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z}
2775
2775
# CHECK-NEXT: - - 1.50 0.50 - - 8.00 1.50 0.50 8.00 8.00 8.00 vpscatterdd %zmm1, (%rdx,%zmm0,4) {%k1}
2776
2776
# CHECK-NEXT: - - 1.50 0.50 - - 4.00 0.50 0.50 4.00 4.00 4.00 vpscatterdq %zmm1, (%rdx,%ymm0,4) {%k1}
2777
- # CHECK-NEXT: - - 1.50 0.50 - - 4.00 0.50 0.50 4.00 4.00 4.00 vpscatterqd %ymm1, (%rdx,%zmm0,4) {%k1}
2777
+ # CHECK-NEXT: - - - - - - 0.50 - - 0.50 0.50 0.50 vpscatterqd %ymm1, (%rdx,%zmm0,4) {%k1}
2778
2778
# CHECK-NEXT: - - 1.50 0.50 - - 4.00 0.50 0.50 4.00 4.00 4.00 vpscatterqq %zmm1, (%rdx,%zmm0,4) {%k1}
2779
2779
# CHECK-NEXT: - - - - - - - 1.00 - - - - vpshufd $0, %zmm16, %zmm19
2780
2780
# CHECK-NEXT: - - - - 0.50 0.50 - 1.00 - - - - vpshufd $0, (%rax), %zmm19
@@ -2841,7 +2841,7 @@ vunpcklps (%rax){1to16}, %zmm17, %zmm19 {z}{k1}
2841
2841
# CHECK-NEXT: - - - - 0.50 0.50 - 1.00 - - - - vpunpcklqdq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z}
2842
2842
# CHECK-NEXT: - - 1.50 0.50 - - 8.00 1.50 0.50 8.00 8.00 8.00 vscatterdps %zmm1, (%rdx,%zmm0,4) {%k1}
2843
2843
# CHECK-NEXT: - - 1.50 0.50 - - 4.00 0.50 0.50 4.00 4.00 4.00 vscatterdpd %zmm1, (%rdx,%ymm0,4) {%k1}
2844
- # CHECK-NEXT: - - 1.50 0.50 - - 4.00 0.50 0.50 4.00 4.00 4.00 vscatterqps %ymm1, (%rdx,%zmm0,4) {%k1}
2844
+ # CHECK-NEXT: - - - - - - 0.50 - - 0.50 0.50 0.50 vscatterqps %ymm1, (%rdx,%zmm0,4) {%k1}
2845
2845
# CHECK-NEXT: - - 1.50 0.50 - - 4.00 0.50 0.50 4.00 4.00 4.00 vscatterqpd %zmm1, (%rdx,%zmm0,4) {%k1}
2846
2846
# CHECK-NEXT: - - - - - - - 1.00 - - - - vshuff32x4 $0, %zmm16, %zmm17, %zmm19
2847
2847
# CHECK-NEXT: - - - - 0.50 0.50 - 1.00 - - - - vshuff32x4 $0, (%rax), %zmm17, %zmm19
0 commit comments