}
}
+ // If the target supports masking y in (shl, y),
+ // fold (shl x, (and y, ((1 << numbits(x)) - 1))) -> (shl x, y)
+ if (TLI.isOperationLegal(ISD::SHL, VT) &&
+ TLI.supportsModuloShift(ISD::SHL, VT) && N1->getOpcode() == ISD::AND) {
+ if (ConstantSDNode *Mask = isConstOrConstSplat(N1->getOperand(1))) {
+ if (Mask->getZExtValue() == OpSizeInBits - 1) {
+ return DAG.getNode(ISD::SHL, SDLoc(N), VT, N0, N1->getOperand(0));
+ }
+ }
+ }
+
ConstantSDNode *N1C = isConstOrConstSplat(N1);
// fold (shl c1, c2) -> c1<<c2
EVT VT = N0.getValueType();
unsigned OpSizeInBits = VT.getScalarSizeInBits();
+ // If the target supports masking y in (sra, y),
+ // fold (sra x, (and y, ((1 << numbits(x)) - 1))) -> (sra x, y)
+ if (TLI.isOperationLegal(ISD::SRA, VT) &&
+ TLI.supportsModuloShift(ISD::SRA, VT) && N1->getOpcode() == ISD::AND) {
+ if (ConstantSDNode *Mask = isConstOrConstSplat(N1->getOperand(1))) {
+ if (Mask->getZExtValue() == OpSizeInBits - 1) {
+ return DAG.getNode(ISD::SRA, SDLoc(N), VT, N0, N1->getOperand(0));
+ }
+ }
+ }
+
// Arithmetic shifting an all-sign-bit value is a no-op.
if (DAG.ComputeNumSignBits(N0) == OpSizeInBits)
return N0;
EVT VT = N0.getValueType();
unsigned OpSizeInBits = VT.getScalarSizeInBits();
+ // If the target supports masking y in (srl, y),
+ // fold (srl x, (and y, ((1 << numbits(x)) - 1))) -> (srl x, y)
+ if (TLI.isOperationLegal(ISD::SRL, VT) &&
+ TLI.supportsModuloShift(ISD::SRL, VT) && N1->getOpcode() == ISD::AND) {
+ if (ConstantSDNode *Mask = isConstOrConstSplat(N1->getOperand(1))) {
+ if (Mask->getZExtValue() == OpSizeInBits - 1) {
+ return DAG.getNode(ISD::SRL, SDLoc(N), VT, N0, N1->getOperand(0));
+ }
+ }
+ }
+
// fold vector ops
if (VT.isVector())
if (SDValue FoldedVOp = SimplifyVBinOp(N))
define <16 x i8> @test010(<16 x i8> %a, <16 x i8> %b) {
; CHECK-LABEL: test010:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltisb 4, 7
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vslb 2, 2, 3
; CHECK-NEXT: blr
%rem = and <16 x i8> %b, <i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7>
define <8 x i16> @test011(<8 x i16> %a, <8 x i16> %b) {
; CHECK-LABEL: test011:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltish 4, 15
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vslh 2, 2, 3
; CHECK-NEXT: blr
%rem = and <8 x i16> %b, <i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15>
define <4 x i32> @test012(<4 x i32> %a, <4 x i32> %b) {
; CHECK-LABEL: test012:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltisw 4, -16
-; CHECK-NEXT: vspltisw 5, 15
-; CHECK-NEXT: vsubuwm 4, 5, 4
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vslw 2, 2, 3
; CHECK-NEXT: blr
%rem = and <4 x i32> %b, <i32 31, i32 31, i32 31, i32 31>
define <2 x i64> @test013(<2 x i64> %a, <2 x i64> %b) {
; CHECK-LABEL: test013:
; CHECK: # BB#0:
-; CHECK-NEXT: addis 3, 2, .LCPI7_0@toc@ha
-; CHECK-NEXT: addi 3, 3, .LCPI7_0@toc@l
-; CHECK-NEXT: lxvd2x 0, 0, 3
-; CHECK-NEXT: xxswapd 36, 0
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsld 2, 2, 3
; CHECK-NEXT: blr
%rem = and <2 x i64> %b, <i64 63, i64 63>
define <16 x i8> @test110(<16 x i8> %a, <16 x i8> %b) {
; CHECK-LABEL: test110:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltisb 4, 7
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrb 2, 2, 3
; CHECK-NEXT: blr
%rem = and <16 x i8> %b, <i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7>
define <8 x i16> @test111(<8 x i16> %a, <8 x i16> %b) {
; CHECK-LABEL: test111:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltish 4, 15
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrh 2, 2, 3
; CHECK-NEXT: blr
%rem = and <8 x i16> %b, <i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15>
define <4 x i32> @test112(<4 x i32> %a, <4 x i32> %b) {
; CHECK-LABEL: test112:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltisw 4, -16
-; CHECK-NEXT: vspltisw 5, 15
-; CHECK-NEXT: vsubuwm 4, 5, 4
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrw 2, 2, 3
; CHECK-NEXT: blr
%rem = and <4 x i32> %b, <i32 31, i32 31, i32 31, i32 31>
define <2 x i64> @test113(<2 x i64> %a, <2 x i64> %b) {
; CHECK-LABEL: test113:
; CHECK: # BB#0:
-; CHECK-NEXT: addis 3, 2, .LCPI15_0@toc@ha
-; CHECK-NEXT: addi 3, 3, .LCPI15_0@toc@l
-; CHECK-NEXT: lxvd2x 0, 0, 3
-; CHECK-NEXT: xxswapd 36, 0
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrd 2, 2, 3
; CHECK-NEXT: blr
%rem = and <2 x i64> %b, <i64 63, i64 63>
define <16 x i8> @test210(<16 x i8> %a, <16 x i8> %b) {
; CHECK-LABEL: test210:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltisb 4, 7
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrab 2, 2, 3
; CHECK-NEXT: blr
%rem = and <16 x i8> %b, <i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7>
define <8 x i16> @test211(<8 x i16> %a, <8 x i16> %b) {
; CHECK-LABEL: test211:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltish 4, 15
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrah 2, 2, 3
; CHECK-NEXT: blr
%rem = and <8 x i16> %b, <i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15, i16 15>
define <4 x i32> @test212(<4 x i32> %a, <4 x i32> %b) {
; CHECK-LABEL: test212:
; CHECK: # BB#0:
-; CHECK-NEXT: vspltisw 4, -16
-; CHECK-NEXT: vspltisw 5, 15
-; CHECK-NEXT: vsubuwm 4, 5, 4
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsraw 2, 2, 3
; CHECK-NEXT: blr
%rem = and <4 x i32> %b, <i32 31, i32 31, i32 31, i32 31>
define <2 x i64> @test213(<2 x i64> %a, <2 x i64> %b) {
; CHECK-LABEL: test213:
; CHECK: # BB#0:
-; CHECK-NEXT: addis 3, 2, .LCPI23_0@toc@ha
-; CHECK-NEXT: addi 3, 3, .LCPI23_0@toc@l
-; CHECK-NEXT: lxvd2x 0, 0, 3
-; CHECK-NEXT: xxswapd 36, 0
-; CHECK-NEXT: xxland 35, 35, 36
; CHECK-NEXT: vsrad 2, 2, 3
; CHECK-NEXT: blr
%rem = and <2 x i64> %b, <i64 63, i64 63>