@@ -1344,6 +1344,16 @@ vpmulld %ymm16, %ymm17, %ymm19 {z}{k1}
13441344vpmulld (%rax ), %ymm17, %ymm19 {z}{k1}
13451345vpmulld (%rax ){1to8}, %ymm17, %ymm19 {z}{k1}
13461346
1347+ vpscatterdd %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1348+ vpscatterdq %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1349+ vpscatterqd %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1350+ vpscatterqq %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1351+
1352+ vpscatterdd %ymm1 , (%rdx ,%ymm0 ,4 ) {%k1 }
1353+ vpscatterdq %ymm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1354+ vpscatterqd %xmm1 , (%rdx ,%ymm0 ,4 ) {%k1 }
1355+ vpscatterqq %ymm1 , (%rdx ,%ymm0 ,4 ) {%k1 }
1356+
13471357vpshufd $0 , %xmm16, %xmm19
13481358vpshufd $0 , (%rax ), %xmm19
13491359vpshufd $0 , (%rax ){1to4}, %xmm19
@@ -1500,6 +1510,16 @@ vpunpckldq %ymm16, %ymm17, %ymm19 {z}{k1}
15001510vpunpckldq (%rax ), %ymm17, %ymm19 {z}{k1}
15011511vpunpckldq (%rax ){1to8}, %ymm17, %ymm19 {z}{k1}
15021512
1513+ vscatterdps %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1514+ vscatterdpd %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1515+ vscatterqps %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1516+ vscatterqpd %xmm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1517+
1518+ vscatterdps %ymm1 , (%rdx ,%ymm0 ,4 ) {%k1 }
1519+ vscatterdpd %ymm1 , (%rdx ,%xmm0 ,4 ) {%k1 }
1520+ vscatterqps %xmm1 , (%rdx ,%ymm0 ,4 ) {%k1 }
1521+ vscatterqpd %ymm1 , (%rdx ,%ymm0 ,4 ) {%k1 }
1522+
15031523vshuff32x4 $0 , %ymm16, %ymm17, %ymm19
15041524vshuff32x4 $0 , (%rax ), %ymm17, %ymm19
15051525vshuff32x4 $0 , (%rax ){1to8}, %ymm17, %ymm19
@@ -2897,6 +2917,14 @@ vunpcklps (%rax){1to8}, %ymm17, %ymm19 {z}{k1}
28972917# CHECK-NEXT: 1 5 1.00 vpmulld %ymm16, %ymm17, %ymm19 {%k1} {z}
28982918# CHECK-NEXT: 2 12 1.00 * vpmulld (%rax), %ymm17, %ymm19 {%k1} {z}
28992919# CHECK-NEXT: 2 12 1.00 * vpmulld (%rax){1to8}, %ymm17, %ymm19 {%k1} {z}
2920+ # CHECK-NEXT: 1 1 1.00 * vpscatterdd %xmm1, (%rdx,%xmm0,4) {%k1}
2921+ # CHECK-NEXT: 1 1 1.00 * vpscatterdq %xmm1, (%rdx,%xmm0,4) {%k1}
2922+ # CHECK-NEXT: 1 1 1.00 * vpscatterqd %xmm1, (%rdx,%xmm0,4) {%k1}
2923+ # CHECK-NEXT: 1 1 1.00 * vpscatterqq %xmm1, (%rdx,%xmm0,4) {%k1}
2924+ # CHECK-NEXT: 1 1 1.00 * vpscatterdd %ymm1, (%rdx,%ymm0,4) {%k1}
2925+ # CHECK-NEXT: 1 1 1.00 * vpscatterdq %ymm1, (%rdx,%xmm0,4) {%k1}
2926+ # CHECK-NEXT: 1 1 1.00 * vpscatterqd %xmm1, (%rdx,%ymm0,4) {%k1}
2927+ # CHECK-NEXT: 1 1 1.00 * vpscatterqq %ymm1, (%rdx,%ymm0,4) {%k1}
29002928# CHECK-NEXT: 1 1 0.50 vpshufd $0, %xmm16, %xmm19
29012929# CHECK-NEXT: 2 7 0.50 * vpshufd $0, (%rax), %xmm19
29022930# CHECK-NEXT: 2 7 0.50 * vpshufd $0, (%rax){1to4}, %xmm19
@@ -3035,6 +3063,14 @@ vunpcklps (%rax){1to8}, %ymm17, %ymm19 {z}{k1}
30353063# CHECK-NEXT: 1 1 1.00 vpunpckldq %ymm16, %ymm17, %ymm19 {%k1} {z}
30363064# CHECK-NEXT: 2 8 1.00 * vpunpckldq (%rax), %ymm17, %ymm19 {%k1} {z}
30373065# CHECK-NEXT: 2 8 1.00 * vpunpckldq (%rax){1to8}, %ymm17, %ymm19 {%k1} {z}
3066+ # CHECK-NEXT: 1 1 1.00 * vscatterdps %xmm1, (%rdx,%xmm0,4) {%k1}
3067+ # CHECK-NEXT: 1 1 1.00 * vscatterdpd %xmm1, (%rdx,%xmm0,4) {%k1}
3068+ # CHECK-NEXT: 1 1 1.00 * vscatterqps %xmm1, (%rdx,%xmm0,4) {%k1}
3069+ # CHECK-NEXT: 1 1 1.00 * vscatterqpd %xmm1, (%rdx,%xmm0,4) {%k1}
3070+ # CHECK-NEXT: 1 1 1.00 * vscatterdps %ymm1, (%rdx,%ymm0,4) {%k1}
3071+ # CHECK-NEXT: 1 1 1.00 * vscatterdpd %ymm1, (%rdx,%xmm0,4) {%k1}
3072+ # CHECK-NEXT: 1 1 1.00 * vscatterqps %xmm1, (%rdx,%ymm0,4) {%k1}
3073+ # CHECK-NEXT: 1 1 1.00 * vscatterqpd %ymm1, (%rdx,%ymm0,4) {%k1}
30383074# CHECK-NEXT: 1 1 1.00 vshuff32x4 $0, %ymm16, %ymm17, %ymm19
30393075# CHECK-NEXT: 2 8 1.00 * vshuff32x4 $0, (%rax), %ymm17, %ymm19
30403076# CHECK-NEXT: 2 8 1.00 * vshuff32x4 $0, (%rax){1to8}, %ymm17, %ymm19
@@ -3228,7 +3264,7 @@ vunpcklps (%rax){1to8}, %ymm17, %ymm19 {z}{k1}
32283264
32293265# CHECK: Resource pressure per iteration:
32303266# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1]
3231- # CHECK-NEXT: - 1935.00 278.00 579.50 32 .00 738.50 486 .50 486 .50
3267+ # CHECK-NEXT: - 1935.00 278.00 579.50 48 .00 738.50 494 .50 494 .50
32323268
32333269# CHECK: Resource pressure by instruction:
32343270# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions:
@@ -4420,6 +4456,14 @@ vunpcklps (%rax){1to8}, %ymm17, %ymm19 {z}{k1}
44204456# CHECK-NEXT: - - 1.00 - - - - - vpmulld %ymm16, %ymm17, %ymm19 {%k1} {z}
44214457# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vpmulld (%rax), %ymm17, %ymm19 {%k1} {z}
44224458# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vpmulld (%rax){1to8}, %ymm17, %ymm19 {%k1} {z}
4459+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterdd %xmm1, (%rdx,%xmm0,4) {%k1}
4460+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterdq %xmm1, (%rdx,%xmm0,4) {%k1}
4461+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterqd %xmm1, (%rdx,%xmm0,4) {%k1}
4462+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterqq %xmm1, (%rdx,%xmm0,4) {%k1}
4463+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterdd %ymm1, (%rdx,%ymm0,4) {%k1}
4464+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterdq %ymm1, (%rdx,%xmm0,4) {%k1}
4465+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterqd %xmm1, (%rdx,%ymm0,4) {%k1}
4466+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vpscatterqq %ymm1, (%rdx,%ymm0,4) {%k1}
44234467# CHECK-NEXT: - - - 0.50 - 0.50 - - vpshufd $0, %xmm16, %xmm19
44244468# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpshufd $0, (%rax), %xmm19
44254469# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpshufd $0, (%rax){1to4}, %xmm19
@@ -4558,6 +4602,14 @@ vunpcklps (%rax){1to8}, %ymm17, %ymm19 {z}{k1}
45584602# CHECK-NEXT: - - - - - 1.00 - - vpunpckldq %ymm16, %ymm17, %ymm19 {%k1} {z}
45594603# CHECK-NEXT: - - - - - 1.00 0.50 0.50 vpunpckldq (%rax), %ymm17, %ymm19 {%k1} {z}
45604604# CHECK-NEXT: - - - - - 1.00 0.50 0.50 vpunpckldq (%rax){1to8}, %ymm17, %ymm19 {%k1} {z}
4605+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterdps %xmm1, (%rdx,%xmm0,4) {%k1}
4606+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterdpd %xmm1, (%rdx,%xmm0,4) {%k1}
4607+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterqps %xmm1, (%rdx,%xmm0,4) {%k1}
4608+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterqpd %xmm1, (%rdx,%xmm0,4) {%k1}
4609+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterdps %ymm1, (%rdx,%ymm0,4) {%k1}
4610+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterdpd %ymm1, (%rdx,%xmm0,4) {%k1}
4611+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterqps %xmm1, (%rdx,%ymm0,4) {%k1}
4612+ # CHECK-NEXT: - - - - 1.00 - 0.50 0.50 vscatterqpd %ymm1, (%rdx,%ymm0,4) {%k1}
45614613# CHECK-NEXT: - - - - - 1.00 - - vshuff32x4 $0, %ymm16, %ymm17, %ymm19
45624614# CHECK-NEXT: - - - - - 1.00 0.50 0.50 vshuff32x4 $0, (%rax), %ymm17, %ymm19
45634615# CHECK-NEXT: - - - - - 1.00 0.50 0.50 vshuff32x4 $0, (%rax){1to8}, %ymm17, %ymm19
0 commit comments