[llvm] f7b2d4a - [AArch64] Add patterns to fold zext(cmpeq(x, splat(0)))

[Cullen Rhodes via llvm-commits]

Author: Cullen Rhodes
Date: 2022-07-19T08:14:38Z
New Revision: f7b2d4aac66576ce9d091b748141f5cfe5100f91

URL: https://github.com/llvm/llvm-project/commit/f7b2d4aac66576ce9d091b748141f5cfe5100f91
DIFF: https://github.com/llvm/llvm-project/commit/f7b2d4aac66576ce9d091b748141f5cfe5100f91.diff

LOG: [AArch64] Add patterns to fold zext(cmpeq(x, splat(0)))

Reviewed By: paulwalker-arm

Differential Revision: https://reviews.llvm.org/D129626

Added: 
    

Modified: 
    llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
    llvm/test/CodeGen/AArch64/sve-cmp-folds.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
index e49b5b8901d7..424b24a7a30b 100644
--- a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
+++ b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
@@ -455,6 +455,16 @@ let Predicates = [HasSVEorSME] in {
   defm FABS_ZPmZ : sve_int_un_pred_arit_1_fp<0b100, "fabs", AArch64fabs_mt>;
   defm FNEG_ZPmZ : sve_int_un_pred_arit_1_fp<0b101, "fneg", AArch64fneg_mt>;
 
+  // zext(cmpeq(x, splat(0))) -> cnot(x)
+  def : Pat<(nxv16i8 (zext (nxv16i1 (AArch64setcc_z (nxv16i1 (SVEAllActive):$Pg), nxv16i8:$Op2, (SVEDup0), SETEQ)))),
+            (CNOT_ZPmZ_B $Op2, $Pg, $Op2)>;
+  def : Pat<(nxv8i16 (zext (nxv8i1 (AArch64setcc_z (nxv8i1 (SVEAllActive):$Pg), nxv8i16:$Op2, (SVEDup0), SETEQ)))),
+            (CNOT_ZPmZ_H $Op2, $Pg, $Op2)>;
+  def : Pat<(nxv4i32 (zext (nxv4i1 (AArch64setcc_z (nxv4i1 (SVEAllActive):$Pg), nxv4i32:$Op2, (SVEDup0), SETEQ)))),
+            (CNOT_ZPmZ_S $Op2, $Pg, $Op2)>;
+  def : Pat<(nxv2i64 (zext (nxv2i1 (AArch64setcc_z (nxv2i1 (SVEAllActive):$Pg), nxv2i64:$Op2, (SVEDup0), SETEQ)))),
+            (CNOT_ZPmZ_D $Op2, $Pg, $Op2)>;
+
   defm SMAX_ZPmZ : sve_int_bin_pred_arit_1<0b000, "smax", "SMAX_ZPZZ", int_aarch64_sve_smax, DestructiveBinaryComm>;
   defm UMAX_ZPmZ : sve_int_bin_pred_arit_1<0b001, "umax", "UMAX_ZPZZ", int_aarch64_sve_umax, DestructiveBinaryComm>;
   defm SMIN_ZPmZ : sve_int_bin_pred_arit_1<0b010, "smin", "SMIN_ZPZZ", int_aarch64_sve_smin, DestructiveBinaryComm>;

diff  --git a/llvm/test/CodeGen/AArch64/sve-cmp-folds.ll b/llvm/test/CodeGen/AArch64/sve-cmp-folds.ll
index 81039307fd0d..beded623272c 100644
--- a/llvm/test/CodeGen/AArch64/sve-cmp-folds.ll
+++ b/llvm/test/CodeGen/AArch64/sve-cmp-folds.ll
@@ -57,8 +57,7 @@ define <vscale x 16 x i8> @icmp_cnot_nxv16i8(<vscale x 16 x i8> %a) {
 ; CHECK-LABEL: icmp_cnot_nxv16i8:
 ; CHECK:       // %bb.0:
 ; CHECK-NEXT:    ptrue p0.b
-; CHECK-NEXT:    cmpeq p0.b, p0/z, z0.b, #0
-; CHECK-NEXT:    mov z0.b, p0/z, #1 // =0x1
+; CHECK-NEXT:    cnot z0.b, p0/m, z0.b
 ; CHECK-NEXT:    ret
   %mask = icmp eq <vscale x 16 x i8> %a, zeroinitializer
   %zext = zext <vscale x 16 x i1> %mask to <vscale x 16 x i8>
@@ -69,8 +68,7 @@ define <vscale x 8 x i16> @icmp_cnot_nxv8i16(<vscale x 8 x i16> %a) {
 ; CHECK-LABEL: icmp_cnot_nxv8i16:
 ; CHECK:       // %bb.0:
 ; CHECK-NEXT:    ptrue p0.h
-; CHECK-NEXT:    cmpeq p0.h, p0/z, z0.h, #0
-; CHECK-NEXT:    mov z0.h, p0/z, #1 // =0x1
+; CHECK-NEXT:    cnot z0.h, p0/m, z0.h
 ; CHECK-NEXT:    ret
   %mask = icmp eq <vscale x 8 x i16> %a, zeroinitializer
   %zext = zext <vscale x 8 x i1> %mask to <vscale x 8 x i16>
@@ -81,8 +79,7 @@ define <vscale x 4 x i32> @icmp_cnot_nxv4i32(<vscale x 4 x i32> %a) {
 ; CHECK-LABEL: icmp_cnot_nxv4i32:
 ; CHECK:       // %bb.0:
 ; CHECK-NEXT:    ptrue p0.s
-; CHECK-NEXT:    cmpeq p0.s, p0/z, z0.s, #0
-; CHECK-NEXT:    mov z0.s, p0/z, #1 // =0x1
+; CHECK-NEXT:    cnot z0.s, p0/m, z0.s
 ; CHECK-NEXT:    ret
   %mask = icmp eq <vscale x 4 x i32> %a, zeroinitializer
   %zext = zext <vscale x 4 x i1> %mask to <vscale x 4 x i32>
@@ -93,8 +90,7 @@ define <vscale x 2 x i64> @icmp_cnot_nxv2i64(<vscale x 2 x i64> %a) {
 ; CHECK-LABEL: icmp_cnot_nxv2i64:
 ; CHECK:       // %bb.0:
 ; CHECK-NEXT:    ptrue p0.d
-; CHECK-NEXT:    cmpeq p0.d, p0/z, z0.d, #0
-; CHECK-NEXT:    mov z0.d, p0/z, #1 // =0x1
+; CHECK-NEXT:    cnot z0.d, p0/m, z0.d
 ; CHECK-NEXT:    ret
   %mask = icmp eq <vscale x 2 x i64> %a, zeroinitializer
   %zext = zext <vscale x 2 x i1> %mask to <vscale x 2 x i64>