[dev.simd] cmd/compile: make compares between NaNs all false.

This CL updates the predicate immediate value of Equal, GreaterEqual,
Greater.

This CL is generated by Cl 686215.

Change-Id: I77fc411f40f5c790a1be7f3d5ffd11f12df50ec7
Reviewed-on: https://go-review.googlesource.com/c/go/+/686235
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>

diff --git a/src/cmd/compile/internal/ssa/_gen/simdAMD64.rules b/src/cmd/compile/internal/ssa/_gen/simdAMD64.rules
index 6ba52a9e9c9f93100c630b3d058be162962d69d3..757020b6c966b1ee21adb563cf5573401a026ac8 100644 (file)
--- a/src/cmd/compile/internal/ssa/_gen/simdAMD64.rules
+++ b/src/cmd/compile/internal/ssa/_gen/simdAMD64.rules
@@ -242,66 +242,66 @@
 (GetElemUint16x8 [a] x) => (VPEXTRW128 [a] x)
 (GetElemUint32x4 [a] x) => (VPEXTRD128 [a] x)
 (GetElemUint64x2 [a] x) => (VPEXTRQ128 [a] x)
-(GreaterFloat32x4 x y) => (VCMPPS128 [6] x y)
-(GreaterFloat32x8 x y) => (VCMPPS256 [6] x y)
-(GreaterFloat32x16 x y) => (VPMOVMToVec32x16 (VCMPPS512 [6] x y))
-(GreaterFloat64x2 x y) => (VCMPPD128 [6] x y)
-(GreaterFloat64x4 x y) => (VCMPPD256 [6] x y)
-(GreaterFloat64x8 x y) => (VPMOVMToVec64x8 (VCMPPD512 [6] x y))
+(GreaterFloat32x4 x y) => (VCMPPS128 [14] x y)
+(GreaterFloat32x8 x y) => (VCMPPS256 [14] x y)
+(GreaterFloat32x16 x y) => (VPMOVMToVec32x16 (VCMPPS512 [14] x y))
+(GreaterFloat64x2 x y) => (VCMPPD128 [14] x y)
+(GreaterFloat64x4 x y) => (VCMPPD256 [14] x y)
+(GreaterFloat64x8 x y) => (VPMOVMToVec64x8 (VCMPPD512 [14] x y))
 (GreaterInt8x16 ...) => (VPCMPGTB128 ...)
 (GreaterInt8x32 ...) => (VPCMPGTB256 ...)
-(GreaterInt8x64 x y) => (VPMOVMToVec8x64 (VPCMPB512 [6] x y))
+(GreaterInt8x64 x y) => (VPMOVMToVec8x64 (VPCMPB512 [14] x y))
 (GreaterInt16x8 ...) => (VPCMPGTW128 ...)
 (GreaterInt16x16 ...) => (VPCMPGTW256 ...)
-(GreaterInt16x32 x y) => (VPMOVMToVec16x32 (VPCMPW512 [6] x y))
+(GreaterInt16x32 x y) => (VPMOVMToVec16x32 (VPCMPW512 [14] x y))
 (GreaterInt32x4 ...) => (VPCMPGTD128 ...)
 (GreaterInt32x8 ...) => (VPCMPGTD256 ...)
-(GreaterInt32x16 x y) => (VPMOVMToVec32x16 (VPCMPD512 [6] x y))
-(GreaterInt64x2 x y) => (VPMOVMToVec64x2 (VPCMPQ128 [6] x y))
+(GreaterInt32x16 x y) => (VPMOVMToVec32x16 (VPCMPD512 [14] x y))
+(GreaterInt64x2 x y) => (VPMOVMToVec64x2 (VPCMPQ128 [14] x y))
 (GreaterInt64x4 ...) => (VPCMPGTQ256 ...)
-(GreaterInt64x8 x y) => (VPMOVMToVec64x8 (VPCMPQ512 [6] x y))
-(GreaterUint8x16 x y) => (VPMOVMToVec8x16 (VPCMPUB128 [6] x y))
-(GreaterUint8x32 x y) => (VPMOVMToVec8x32 (VPCMPUB256 [6] x y))
-(GreaterUint8x64 x y) => (VPMOVMToVec8x64 (VPCMPUB512 [6] x y))
-(GreaterUint16x8 x y) => (VPMOVMToVec16x8 (VPCMPUW128 [6] x y))
-(GreaterUint16x16 x y) => (VPMOVMToVec16x16 (VPCMPUW256 [6] x y))
-(GreaterUint16x32 x y) => (VPMOVMToVec16x32 (VPCMPUW512 [6] x y))
-(GreaterUint32x4 x y) => (VPMOVMToVec32x4 (VPCMPUD128 [6] x y))
-(GreaterUint32x8 x y) => (VPMOVMToVec32x8 (VPCMPUD256 [6] x y))
-(GreaterUint32x16 x y) => (VPMOVMToVec32x16 (VPCMPUD512 [6] x y))
-(GreaterUint64x2 x y) => (VPMOVMToVec64x2 (VPCMPUQ128 [6] x y))
-(GreaterUint64x4 x y) => (VPMOVMToVec64x4 (VPCMPUQ256 [6] x y))
-(GreaterUint64x8 x y) => (VPMOVMToVec64x8 (VPCMPUQ512 [6] x y))
-(GreaterEqualFloat32x4 x y) => (VCMPPS128 [5] x y)
-(GreaterEqualFloat32x8 x y) => (VCMPPS256 [5] x y)
-(GreaterEqualFloat32x16 x y) => (VPMOVMToVec32x16 (VCMPPS512 [5] x y))
-(GreaterEqualFloat64x2 x y) => (VCMPPD128 [5] x y)
-(GreaterEqualFloat64x4 x y) => (VCMPPD256 [5] x y)
-(GreaterEqualFloat64x8 x y) => (VPMOVMToVec64x8 (VCMPPD512 [5] x y))
-(GreaterEqualInt8x16 x y) => (VPMOVMToVec8x16 (VPCMPB128 [5] x y))
-(GreaterEqualInt8x32 x y) => (VPMOVMToVec8x32 (VPCMPB256 [5] x y))
-(GreaterEqualInt8x64 x y) => (VPMOVMToVec8x64 (VPCMPB512 [5] x y))
-(GreaterEqualInt16x8 x y) => (VPMOVMToVec16x8 (VPCMPW128 [5] x y))
-(GreaterEqualInt16x16 x y) => (VPMOVMToVec16x16 (VPCMPW256 [5] x y))
-(GreaterEqualInt16x32 x y) => (VPMOVMToVec16x32 (VPCMPW512 [5] x y))
-(GreaterEqualInt32x4 x y) => (VPMOVMToVec32x4 (VPCMPD128 [5] x y))
-(GreaterEqualInt32x8 x y) => (VPMOVMToVec32x8 (VPCMPD256 [5] x y))
-(GreaterEqualInt32x16 x y) => (VPMOVMToVec32x16 (VPCMPD512 [5] x y))
-(GreaterEqualInt64x2 x y) => (VPMOVMToVec64x2 (VPCMPQ128 [5] x y))
-(GreaterEqualInt64x4 x y) => (VPMOVMToVec64x4 (VPCMPQ256 [5] x y))
-(GreaterEqualInt64x8 x y) => (VPMOVMToVec64x8 (VPCMPQ512 [5] x y))
-(GreaterEqualUint8x16 x y) => (VPMOVMToVec8x16 (VPCMPUB128 [5] x y))
-(GreaterEqualUint8x32 x y) => (VPMOVMToVec8x32 (VPCMPUB256 [5] x y))
-(GreaterEqualUint8x64 x y) => (VPMOVMToVec8x64 (VPCMPUB512 [5] x y))
-(GreaterEqualUint16x8 x y) => (VPMOVMToVec16x8 (VPCMPUW128 [5] x y))
-(GreaterEqualUint16x16 x y) => (VPMOVMToVec16x16 (VPCMPUW256 [5] x y))
-(GreaterEqualUint16x32 x y) => (VPMOVMToVec16x32 (VPCMPUW512 [5] x y))
-(GreaterEqualUint32x4 x y) => (VPMOVMToVec32x4 (VPCMPUD128 [5] x y))
-(GreaterEqualUint32x8 x y) => (VPMOVMToVec32x8 (VPCMPUD256 [5] x y))
-(GreaterEqualUint32x16 x y) => (VPMOVMToVec32x16 (VPCMPUD512 [5] x y))
-(GreaterEqualUint64x2 x y) => (VPMOVMToVec64x2 (VPCMPUQ128 [5] x y))
-(GreaterEqualUint64x4 x y) => (VPMOVMToVec64x4 (VPCMPUQ256 [5] x y))
-(GreaterEqualUint64x8 x y) => (VPMOVMToVec64x8 (VPCMPUQ512 [5] x y))
+(GreaterInt64x8 x y) => (VPMOVMToVec64x8 (VPCMPQ512 [14] x y))
+(GreaterUint8x16 x y) => (VPMOVMToVec8x16 (VPCMPUB128 [14] x y))
+(GreaterUint8x32 x y) => (VPMOVMToVec8x32 (VPCMPUB256 [14] x y))
+(GreaterUint8x64 x y) => (VPMOVMToVec8x64 (VPCMPUB512 [14] x y))
+(GreaterUint16x8 x y) => (VPMOVMToVec16x8 (VPCMPUW128 [14] x y))
+(GreaterUint16x16 x y) => (VPMOVMToVec16x16 (VPCMPUW256 [14] x y))
+(GreaterUint16x32 x y) => (VPMOVMToVec16x32 (VPCMPUW512 [14] x y))
+(GreaterUint32x4 x y) => (VPMOVMToVec32x4 (VPCMPUD128 [14] x y))
+(GreaterUint32x8 x y) => (VPMOVMToVec32x8 (VPCMPUD256 [14] x y))
+(GreaterUint32x16 x y) => (VPMOVMToVec32x16 (VPCMPUD512 [14] x y))
+(GreaterUint64x2 x y) => (VPMOVMToVec64x2 (VPCMPUQ128 [14] x y))
+(GreaterUint64x4 x y) => (VPMOVMToVec64x4 (VPCMPUQ256 [14] x y))
+(GreaterUint64x8 x y) => (VPMOVMToVec64x8 (VPCMPUQ512 [14] x y))
+(GreaterEqualFloat32x4 x y) => (VCMPPS128 [13] x y)
+(GreaterEqualFloat32x8 x y) => (VCMPPS256 [13] x y)
+(GreaterEqualFloat32x16 x y) => (VPMOVMToVec32x16 (VCMPPS512 [13] x y))
+(GreaterEqualFloat64x2 x y) => (VCMPPD128 [13] x y)
+(GreaterEqualFloat64x4 x y) => (VCMPPD256 [13] x y)
+(GreaterEqualFloat64x8 x y) => (VPMOVMToVec64x8 (VCMPPD512 [13] x y))
+(GreaterEqualInt8x16 x y) => (VPMOVMToVec8x16 (VPCMPB128 [13] x y))
+(GreaterEqualInt8x32 x y) => (VPMOVMToVec8x32 (VPCMPB256 [13] x y))
+(GreaterEqualInt8x64 x y) => (VPMOVMToVec8x64 (VPCMPB512 [13] x y))
+(GreaterEqualInt16x8 x y) => (VPMOVMToVec16x8 (VPCMPW128 [13] x y))
+(GreaterEqualInt16x16 x y) => (VPMOVMToVec16x16 (VPCMPW256 [13] x y))
+(GreaterEqualInt16x32 x y) => (VPMOVMToVec16x32 (VPCMPW512 [13] x y))
+(GreaterEqualInt32x4 x y) => (VPMOVMToVec32x4 (VPCMPD128 [13] x y))
+(GreaterEqualInt32x8 x y) => (VPMOVMToVec32x8 (VPCMPD256 [13] x y))
+(GreaterEqualInt32x16 x y) => (VPMOVMToVec32x16 (VPCMPD512 [13] x y))
+(GreaterEqualInt64x2 x y) => (VPMOVMToVec64x2 (VPCMPQ128 [13] x y))
+(GreaterEqualInt64x4 x y) => (VPMOVMToVec64x4 (VPCMPQ256 [13] x y))
+(GreaterEqualInt64x8 x y) => (VPMOVMToVec64x8 (VPCMPQ512 [13] x y))
+(GreaterEqualUint8x16 x y) => (VPMOVMToVec8x16 (VPCMPUB128 [13] x y))
+(GreaterEqualUint8x32 x y) => (VPMOVMToVec8x32 (VPCMPUB256 [13] x y))
+(GreaterEqualUint8x64 x y) => (VPMOVMToVec8x64 (VPCMPUB512 [13] x y))
+(GreaterEqualUint16x8 x y) => (VPMOVMToVec16x8 (VPCMPUW128 [13] x y))
+(GreaterEqualUint16x16 x y) => (VPMOVMToVec16x16 (VPCMPUW256 [13] x y))
+(GreaterEqualUint16x32 x y) => (VPMOVMToVec16x32 (VPCMPUW512 [13] x y))
+(GreaterEqualUint32x4 x y) => (VPMOVMToVec32x4 (VPCMPUD128 [13] x y))
+(GreaterEqualUint32x8 x y) => (VPMOVMToVec32x8 (VPCMPUD256 [13] x y))
+(GreaterEqualUint32x16 x y) => (VPMOVMToVec32x16 (VPCMPUD512 [13] x y))
+(GreaterEqualUint64x2 x y) => (VPMOVMToVec64x2 (VPCMPUQ128 [13] x y))
+(GreaterEqualUint64x4 x y) => (VPMOVMToVec64x4 (VPCMPUQ256 [13] x y))
+(GreaterEqualUint64x8 x y) => (VPMOVMToVec64x8 (VPCMPUQ512 [13] x y))
 (IsNanFloat32x4 x y) => (VCMPPS128 [3] x y)
 (IsNanFloat32x8 x y) => (VCMPPS256 [3] x y)
 (IsNanFloat32x16 x y) => (VPMOVMToVec32x16 (VCMPPS512 [3] x y))
@@ -563,66 +563,66 @@
 (MaskedGaloisFieldMulUint8x16 x y mask) => (VGF2P8MULBMasked128 x y (VPMOVVec8x16ToM <types.TypeMask> mask))
 (MaskedGaloisFieldMulUint8x32 x y mask) => (VGF2P8MULBMasked256 x y (VPMOVVec8x32ToM <types.TypeMask> mask))
 (MaskedGaloisFieldMulUint8x64 x y mask) => (VGF2P8MULBMasked512 x y (VPMOVVec8x64ToM <types.TypeMask> mask))
-(MaskedGreaterFloat32x4 x y mask) => (VPMOVMToVec32x4 (VCMPPSMasked128 [6] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
-(MaskedGreaterFloat32x8 x y mask) => (VPMOVMToVec32x8 (VCMPPSMasked256 [6] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
-(MaskedGreaterFloat32x16 x y mask) => (VPMOVMToVec32x16 (VCMPPSMasked512 [6] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
-(MaskedGreaterFloat64x2 x y mask) => (VPMOVMToVec64x2 (VCMPPDMasked128 [6] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
-(MaskedGreaterFloat64x4 x y mask) => (VPMOVMToVec64x4 (VCMPPDMasked256 [6] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
-(MaskedGreaterFloat64x8 x y mask) => (VPMOVMToVec64x8 (VCMPPDMasked512 [6] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
-(MaskedGreaterInt8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPBMasked128 [6] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
-(MaskedGreaterInt8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPBMasked256 [6] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
-(MaskedGreaterInt8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPBMasked512 [6] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
-(MaskedGreaterInt16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPWMasked128 [6] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
-(MaskedGreaterInt16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPWMasked256 [6] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
-(MaskedGreaterInt16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPWMasked512 [6] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
-(MaskedGreaterInt32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPDMasked128 [6] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
-(MaskedGreaterInt32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPDMasked256 [6] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
-(MaskedGreaterInt32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPDMasked512 [6] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
-(MaskedGreaterInt64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPQMasked128 [6] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
-(MaskedGreaterInt64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPQMasked256 [6] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
-(MaskedGreaterInt64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPQMasked512 [6] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
-(MaskedGreaterUint8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPUBMasked128 [6] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
-(MaskedGreaterUint8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPUBMasked256 [6] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
-(MaskedGreaterUint8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPUBMasked512 [6] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
-(MaskedGreaterUint16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPUWMasked128 [6] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
-(MaskedGreaterUint16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPUWMasked256 [6] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
-(MaskedGreaterUint16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPUWMasked512 [6] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
-(MaskedGreaterUint32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPUDMasked128 [6] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
-(MaskedGreaterUint32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPUDMasked256 [6] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
-(MaskedGreaterUint32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPUDMasked512 [6] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
-(MaskedGreaterUint64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPUQMasked128 [6] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
-(MaskedGreaterUint64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPUQMasked256 [6] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
-(MaskedGreaterUint64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPUQMasked512 [6] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualFloat32x4 x y mask) => (VPMOVMToVec32x4 (VCMPPSMasked128 [5] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualFloat32x8 x y mask) => (VPMOVMToVec32x8 (VCMPPSMasked256 [5] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualFloat32x16 x y mask) => (VPMOVMToVec32x16 (VCMPPSMasked512 [5] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualFloat64x2 x y mask) => (VPMOVMToVec64x2 (VCMPPDMasked128 [5] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualFloat64x4 x y mask) => (VPMOVMToVec64x4 (VCMPPDMasked256 [5] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualFloat64x8 x y mask) => (VPMOVMToVec64x8 (VCMPPDMasked512 [5] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPBMasked128 [5] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPBMasked256 [5] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPBMasked512 [5] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPWMasked128 [5] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPWMasked256 [5] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPWMasked512 [5] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPDMasked128 [5] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPDMasked256 [5] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPDMasked512 [5] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPQMasked128 [5] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPQMasked256 [5] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualInt64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPQMasked512 [5] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPUBMasked128 [5] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPUBMasked256 [5] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPUBMasked512 [5] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPUWMasked128 [5] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPUWMasked256 [5] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPUWMasked512 [5] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPUDMasked128 [5] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPUDMasked256 [5] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPUDMasked512 [5] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPUQMasked128 [5] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPUQMasked256 [5] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
-(MaskedGreaterEqualUint64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPUQMasked512 [5] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+(MaskedGreaterFloat32x4 x y mask) => (VPMOVMToVec32x4 (VCMPPSMasked128 [14] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+(MaskedGreaterFloat32x8 x y mask) => (VPMOVMToVec32x8 (VCMPPSMasked256 [14] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+(MaskedGreaterFloat32x16 x y mask) => (VPMOVMToVec32x16 (VCMPPSMasked512 [14] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+(MaskedGreaterFloat64x2 x y mask) => (VPMOVMToVec64x2 (VCMPPDMasked128 [14] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+(MaskedGreaterFloat64x4 x y mask) => (VPMOVMToVec64x4 (VCMPPDMasked256 [14] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+(MaskedGreaterFloat64x8 x y mask) => (VPMOVMToVec64x8 (VCMPPDMasked512 [14] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+(MaskedGreaterInt8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPBMasked128 [14] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+(MaskedGreaterInt8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPBMasked256 [14] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+(MaskedGreaterInt8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPBMasked512 [14] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+(MaskedGreaterInt16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPWMasked128 [14] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+(MaskedGreaterInt16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPWMasked256 [14] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+(MaskedGreaterInt16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPWMasked512 [14] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+(MaskedGreaterInt32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPDMasked128 [14] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+(MaskedGreaterInt32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPDMasked256 [14] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+(MaskedGreaterInt32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPDMasked512 [14] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+(MaskedGreaterInt64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPQMasked128 [14] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+(MaskedGreaterInt64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPQMasked256 [14] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+(MaskedGreaterInt64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPQMasked512 [14] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+(MaskedGreaterUint8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPUBMasked128 [14] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+(MaskedGreaterUint8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPUBMasked256 [14] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+(MaskedGreaterUint8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPUBMasked512 [14] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+(MaskedGreaterUint16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPUWMasked128 [14] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+(MaskedGreaterUint16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPUWMasked256 [14] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+(MaskedGreaterUint16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPUWMasked512 [14] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+(MaskedGreaterUint32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPUDMasked128 [14] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+(MaskedGreaterUint32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPUDMasked256 [14] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+(MaskedGreaterUint32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPUDMasked512 [14] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+(MaskedGreaterUint64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPUQMasked128 [14] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+(MaskedGreaterUint64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPUQMasked256 [14] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+(MaskedGreaterUint64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPUQMasked512 [14] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualFloat32x4 x y mask) => (VPMOVMToVec32x4 (VCMPPSMasked128 [13] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualFloat32x8 x y mask) => (VPMOVMToVec32x8 (VCMPPSMasked256 [13] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualFloat32x16 x y mask) => (VPMOVMToVec32x16 (VCMPPSMasked512 [13] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualFloat64x2 x y mask) => (VPMOVMToVec64x2 (VCMPPDMasked128 [13] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualFloat64x4 x y mask) => (VPMOVMToVec64x4 (VCMPPDMasked256 [13] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualFloat64x8 x y mask) => (VPMOVMToVec64x8 (VCMPPDMasked512 [13] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPBMasked128 [13] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPBMasked256 [13] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPBMasked512 [13] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPWMasked128 [13] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPWMasked256 [13] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPWMasked512 [13] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPDMasked128 [13] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPDMasked256 [13] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPDMasked512 [13] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPQMasked128 [13] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPQMasked256 [13] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualInt64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPQMasked512 [13] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint8x16 x y mask) => (VPMOVMToVec8x16 (VPCMPUBMasked128 [13] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint8x32 x y mask) => (VPMOVMToVec8x32 (VPCMPUBMasked256 [13] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint8x64 x y mask) => (VPMOVMToVec8x64 (VPCMPUBMasked512 [13] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint16x8 x y mask) => (VPMOVMToVec16x8 (VPCMPUWMasked128 [13] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint16x16 x y mask) => (VPMOVMToVec16x16 (VPCMPUWMasked256 [13] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint16x32 x y mask) => (VPMOVMToVec16x32 (VPCMPUWMasked512 [13] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint32x4 x y mask) => (VPMOVMToVec32x4 (VPCMPUDMasked128 [13] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint32x8 x y mask) => (VPMOVMToVec32x8 (VPCMPUDMasked256 [13] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint32x16 x y mask) => (VPMOVMToVec32x16 (VPCMPUDMasked512 [13] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint64x2 x y mask) => (VPMOVMToVec64x2 (VPCMPUQMasked128 [13] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint64x4 x y mask) => (VPMOVMToVec64x4 (VPCMPUQMasked256 [13] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+(MaskedGreaterEqualUint64x8 x y mask) => (VPMOVMToVec64x8 (VPCMPUQMasked512 [13] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
 (MaskedIsNanFloat32x4 x y mask) => (VPMOVMToVec32x4 (VCMPPSMasked128 [3] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
 (MaskedIsNanFloat32x8 x y mask) => (VPMOVMToVec32x8 (VCMPPSMasked256 [3] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
 (MaskedIsNanFloat32x16 x y mask) => (VPMOVMToVec32x16 (VCMPPSMasked512 [3] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))

diff --git a/src/cmd/compile/internal/ssa/rewriteAMD64.go b/src/cmd/compile/internal/ssa/rewriteAMD64.go
index 1cf23c4ec5b0e49f3bb43dca1b3e3648ba78802f..6e0726de9b30d227d6df3caf47126b2d8adc2f1b 100644 (file)
--- a/src/cmd/compile/internal/ssa/rewriteAMD64.go
+++ b/src/cmd/compile/internal/ssa/rewriteAMD64.go
@@ -31275,13 +31275,13 @@ func rewriteValueAMD64_OpGreaterEqualFloat32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualFloat32x16 x y)
-       // result: (VPMOVMToVec32x16 (VCMPPS512 [5] x y))
+       // result: (VPMOVMToVec32x16 (VCMPPS512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPS512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31291,12 +31291,12 @@ func rewriteValueAMD64_OpGreaterEqualFloat32x4(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterEqualFloat32x4 x y)
-       // result: (VCMPPS128 [5] x y)
+       // result: (VCMPPS128 [13] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPS128)
-               v.AuxInt = int8ToAuxInt(5)
+               v.AuxInt = int8ToAuxInt(13)
                v.AddArg2(x, y)
                return true
        }
@@ -31305,12 +31305,12 @@ func rewriteValueAMD64_OpGreaterEqualFloat32x8(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterEqualFloat32x8 x y)
-       // result: (VCMPPS256 [5] x y)
+       // result: (VCMPPS256 [13] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPS256)
-               v.AuxInt = int8ToAuxInt(5)
+               v.AuxInt = int8ToAuxInt(13)
                v.AddArg2(x, y)
                return true
        }
@@ -31319,12 +31319,12 @@ func rewriteValueAMD64_OpGreaterEqualFloat64x2(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterEqualFloat64x2 x y)
-       // result: (VCMPPD128 [5] x y)
+       // result: (VCMPPD128 [13] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPD128)
-               v.AuxInt = int8ToAuxInt(5)
+               v.AuxInt = int8ToAuxInt(13)
                v.AddArg2(x, y)
                return true
        }
@@ -31333,12 +31333,12 @@ func rewriteValueAMD64_OpGreaterEqualFloat64x4(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterEqualFloat64x4 x y)
-       // result: (VCMPPD256 [5] x y)
+       // result: (VCMPPD256 [13] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPD256)
-               v.AuxInt = int8ToAuxInt(5)
+               v.AuxInt = int8ToAuxInt(13)
                v.AddArg2(x, y)
                return true
        }
@@ -31349,13 +31349,13 @@ func rewriteValueAMD64_OpGreaterEqualFloat64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualFloat64x8 x y)
-       // result: (VPMOVMToVec64x8 (VCMPPD512 [5] x y))
+       // result: (VPMOVMToVec64x8 (VCMPPD512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPD512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31367,13 +31367,13 @@ func rewriteValueAMD64_OpGreaterEqualInt16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt16x16 x y)
-       // result: (VPMOVMToVec16x16 (VPCMPW256 [5] x y))
+       // result: (VPMOVMToVec16x16 (VPCMPW256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPW256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31385,13 +31385,13 @@ func rewriteValueAMD64_OpGreaterEqualInt16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt16x32 x y)
-       // result: (VPMOVMToVec16x32 (VPCMPW512 [5] x y))
+       // result: (VPMOVMToVec16x32 (VPCMPW512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPW512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31403,13 +31403,13 @@ func rewriteValueAMD64_OpGreaterEqualInt16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt16x8 x y)
-       // result: (VPMOVMToVec16x8 (VPCMPW128 [5] x y))
+       // result: (VPMOVMToVec16x8 (VPCMPW128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPW128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31421,13 +31421,13 @@ func rewriteValueAMD64_OpGreaterEqualInt32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt32x16 x y)
-       // result: (VPMOVMToVec32x16 (VPCMPD512 [5] x y))
+       // result: (VPMOVMToVec32x16 (VPCMPD512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPD512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31439,13 +31439,13 @@ func rewriteValueAMD64_OpGreaterEqualInt32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt32x4 x y)
-       // result: (VPMOVMToVec32x4 (VPCMPD128 [5] x y))
+       // result: (VPMOVMToVec32x4 (VPCMPD128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPD128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31457,13 +31457,13 @@ func rewriteValueAMD64_OpGreaterEqualInt32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt32x8 x y)
-       // result: (VPMOVMToVec32x8 (VPCMPD256 [5] x y))
+       // result: (VPMOVMToVec32x8 (VPCMPD256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPD256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31475,13 +31475,13 @@ func rewriteValueAMD64_OpGreaterEqualInt64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt64x2 x y)
-       // result: (VPMOVMToVec64x2 (VPCMPQ128 [5] x y))
+       // result: (VPMOVMToVec64x2 (VPCMPQ128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQ128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31493,13 +31493,13 @@ func rewriteValueAMD64_OpGreaterEqualInt64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt64x4 x y)
-       // result: (VPMOVMToVec64x4 (VPCMPQ256 [5] x y))
+       // result: (VPMOVMToVec64x4 (VPCMPQ256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQ256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31511,13 +31511,13 @@ func rewriteValueAMD64_OpGreaterEqualInt64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt64x8 x y)
-       // result: (VPMOVMToVec64x8 (VPCMPQ512 [5] x y))
+       // result: (VPMOVMToVec64x8 (VPCMPQ512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQ512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31529,13 +31529,13 @@ func rewriteValueAMD64_OpGreaterEqualInt8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt8x16 x y)
-       // result: (VPMOVMToVec8x16 (VPCMPB128 [5] x y))
+       // result: (VPMOVMToVec8x16 (VPCMPB128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPB128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31547,13 +31547,13 @@ func rewriteValueAMD64_OpGreaterEqualInt8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt8x32 x y)
-       // result: (VPMOVMToVec8x32 (VPCMPB256 [5] x y))
+       // result: (VPMOVMToVec8x32 (VPCMPB256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPB256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31565,13 +31565,13 @@ func rewriteValueAMD64_OpGreaterEqualInt8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualInt8x64 x y)
-       // result: (VPMOVMToVec8x64 (VPCMPB512 [5] x y))
+       // result: (VPMOVMToVec8x64 (VPCMPB512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPB512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31583,13 +31583,13 @@ func rewriteValueAMD64_OpGreaterEqualUint16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint16x16 x y)
-       // result: (VPMOVMToVec16x16 (VPCMPUW256 [5] x y))
+       // result: (VPMOVMToVec16x16 (VPCMPUW256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUW256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31601,13 +31601,13 @@ func rewriteValueAMD64_OpGreaterEqualUint16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint16x32 x y)
-       // result: (VPMOVMToVec16x32 (VPCMPUW512 [5] x y))
+       // result: (VPMOVMToVec16x32 (VPCMPUW512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUW512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31619,13 +31619,13 @@ func rewriteValueAMD64_OpGreaterEqualUint16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint16x8 x y)
-       // result: (VPMOVMToVec16x8 (VPCMPUW128 [5] x y))
+       // result: (VPMOVMToVec16x8 (VPCMPUW128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUW128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31637,13 +31637,13 @@ func rewriteValueAMD64_OpGreaterEqualUint32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint32x16 x y)
-       // result: (VPMOVMToVec32x16 (VPCMPUD512 [5] x y))
+       // result: (VPMOVMToVec32x16 (VPCMPUD512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUD512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31655,13 +31655,13 @@ func rewriteValueAMD64_OpGreaterEqualUint32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint32x4 x y)
-       // result: (VPMOVMToVec32x4 (VPCMPUD128 [5] x y))
+       // result: (VPMOVMToVec32x4 (VPCMPUD128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUD128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31673,13 +31673,13 @@ func rewriteValueAMD64_OpGreaterEqualUint32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint32x8 x y)
-       // result: (VPMOVMToVec32x8 (VPCMPUD256 [5] x y))
+       // result: (VPMOVMToVec32x8 (VPCMPUD256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUD256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31691,13 +31691,13 @@ func rewriteValueAMD64_OpGreaterEqualUint64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint64x2 x y)
-       // result: (VPMOVMToVec64x2 (VPCMPUQ128 [5] x y))
+       // result: (VPMOVMToVec64x2 (VPCMPUQ128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQ128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31709,13 +31709,13 @@ func rewriteValueAMD64_OpGreaterEqualUint64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint64x4 x y)
-       // result: (VPMOVMToVec64x4 (VPCMPUQ256 [5] x y))
+       // result: (VPMOVMToVec64x4 (VPCMPUQ256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQ256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31727,13 +31727,13 @@ func rewriteValueAMD64_OpGreaterEqualUint64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint64x8 x y)
-       // result: (VPMOVMToVec64x8 (VPCMPUQ512 [5] x y))
+       // result: (VPMOVMToVec64x8 (VPCMPUQ512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQ512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31745,13 +31745,13 @@ func rewriteValueAMD64_OpGreaterEqualUint8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint8x16 x y)
-       // result: (VPMOVMToVec8x16 (VPCMPUB128 [5] x y))
+       // result: (VPMOVMToVec8x16 (VPCMPUB128 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUB128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31763,13 +31763,13 @@ func rewriteValueAMD64_OpGreaterEqualUint8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint8x32 x y)
-       // result: (VPMOVMToVec8x32 (VPCMPUB256 [5] x y))
+       // result: (VPMOVMToVec8x32 (VPCMPUB256 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUB256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31781,13 +31781,13 @@ func rewriteValueAMD64_OpGreaterEqualUint8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterEqualUint8x64 x y)
-       // result: (VPMOVMToVec8x64 (VPCMPUB512 [5] x y))
+       // result: (VPMOVMToVec8x64 (VPCMPUB512 [13] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUB512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31799,13 +31799,13 @@ func rewriteValueAMD64_OpGreaterFloat32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterFloat32x16 x y)
-       // result: (VPMOVMToVec32x16 (VCMPPS512 [6] x y))
+       // result: (VPMOVMToVec32x16 (VCMPPS512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPS512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31815,12 +31815,12 @@ func rewriteValueAMD64_OpGreaterFloat32x4(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterFloat32x4 x y)
-       // result: (VCMPPS128 [6] x y)
+       // result: (VCMPPS128 [14] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPS128)
-               v.AuxInt = int8ToAuxInt(6)
+               v.AuxInt = int8ToAuxInt(14)
                v.AddArg2(x, y)
                return true
        }
@@ -31829,12 +31829,12 @@ func rewriteValueAMD64_OpGreaterFloat32x8(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterFloat32x8 x y)
-       // result: (VCMPPS256 [6] x y)
+       // result: (VCMPPS256 [14] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPS256)
-               v.AuxInt = int8ToAuxInt(6)
+               v.AuxInt = int8ToAuxInt(14)
                v.AddArg2(x, y)
                return true
        }
@@ -31843,12 +31843,12 @@ func rewriteValueAMD64_OpGreaterFloat64x2(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterFloat64x2 x y)
-       // result: (VCMPPD128 [6] x y)
+       // result: (VCMPPD128 [14] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPD128)
-               v.AuxInt = int8ToAuxInt(6)
+               v.AuxInt = int8ToAuxInt(14)
                v.AddArg2(x, y)
                return true
        }
@@ -31857,12 +31857,12 @@ func rewriteValueAMD64_OpGreaterFloat64x4(v *Value) bool {
        v_1 := v.Args[1]
        v_0 := v.Args[0]
        // match: (GreaterFloat64x4 x y)
-       // result: (VCMPPD256 [6] x y)
+       // result: (VCMPPD256 [14] x y)
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VCMPPD256)
-               v.AuxInt = int8ToAuxInt(6)
+               v.AuxInt = int8ToAuxInt(14)
                v.AddArg2(x, y)
                return true
        }
@@ -31873,13 +31873,13 @@ func rewriteValueAMD64_OpGreaterFloat64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterFloat64x8 x y)
-       // result: (VPMOVMToVec64x8 (VCMPPD512 [6] x y))
+       // result: (VPMOVMToVec64x8 (VCMPPD512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPD512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31891,13 +31891,13 @@ func rewriteValueAMD64_OpGreaterInt16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterInt16x32 x y)
-       // result: (VPMOVMToVec16x32 (VPCMPW512 [6] x y))
+       // result: (VPMOVMToVec16x32 (VPCMPW512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPW512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31909,13 +31909,13 @@ func rewriteValueAMD64_OpGreaterInt32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterInt32x16 x y)
-       // result: (VPMOVMToVec32x16 (VPCMPD512 [6] x y))
+       // result: (VPMOVMToVec32x16 (VPCMPD512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPD512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31927,13 +31927,13 @@ func rewriteValueAMD64_OpGreaterInt64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterInt64x2 x y)
-       // result: (VPMOVMToVec64x2 (VPCMPQ128 [6] x y))
+       // result: (VPMOVMToVec64x2 (VPCMPQ128 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQ128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31945,13 +31945,13 @@ func rewriteValueAMD64_OpGreaterInt64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterInt64x8 x y)
-       // result: (VPMOVMToVec64x8 (VPCMPQ512 [6] x y))
+       // result: (VPMOVMToVec64x8 (VPCMPQ512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQ512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31963,13 +31963,13 @@ func rewriteValueAMD64_OpGreaterInt8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterInt8x64 x y)
-       // result: (VPMOVMToVec8x64 (VPCMPB512 [6] x y))
+       // result: (VPMOVMToVec8x64 (VPCMPB512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPB512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31981,13 +31981,13 @@ func rewriteValueAMD64_OpGreaterUint16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint16x16 x y)
-       // result: (VPMOVMToVec16x16 (VPCMPUW256 [6] x y))
+       // result: (VPMOVMToVec16x16 (VPCMPUW256 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUW256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -31999,13 +31999,13 @@ func rewriteValueAMD64_OpGreaterUint16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint16x32 x y)
-       // result: (VPMOVMToVec16x32 (VPCMPUW512 [6] x y))
+       // result: (VPMOVMToVec16x32 (VPCMPUW512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUW512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32017,13 +32017,13 @@ func rewriteValueAMD64_OpGreaterUint16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint16x8 x y)
-       // result: (VPMOVMToVec16x8 (VPCMPUW128 [6] x y))
+       // result: (VPMOVMToVec16x8 (VPCMPUW128 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUW128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32035,13 +32035,13 @@ func rewriteValueAMD64_OpGreaterUint32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint32x16 x y)
-       // result: (VPMOVMToVec32x16 (VPCMPUD512 [6] x y))
+       // result: (VPMOVMToVec32x16 (VPCMPUD512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUD512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32053,13 +32053,13 @@ func rewriteValueAMD64_OpGreaterUint32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint32x4 x y)
-       // result: (VPMOVMToVec32x4 (VPCMPUD128 [6] x y))
+       // result: (VPMOVMToVec32x4 (VPCMPUD128 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUD128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32071,13 +32071,13 @@ func rewriteValueAMD64_OpGreaterUint32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint32x8 x y)
-       // result: (VPMOVMToVec32x8 (VPCMPUD256 [6] x y))
+       // result: (VPMOVMToVec32x8 (VPCMPUD256 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUD256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32089,13 +32089,13 @@ func rewriteValueAMD64_OpGreaterUint64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint64x2 x y)
-       // result: (VPMOVMToVec64x2 (VPCMPUQ128 [6] x y))
+       // result: (VPMOVMToVec64x2 (VPCMPUQ128 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQ128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32107,13 +32107,13 @@ func rewriteValueAMD64_OpGreaterUint64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint64x4 x y)
-       // result: (VPMOVMToVec64x4 (VPCMPUQ256 [6] x y))
+       // result: (VPMOVMToVec64x4 (VPCMPUQ256 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQ256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32125,13 +32125,13 @@ func rewriteValueAMD64_OpGreaterUint64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint64x8 x y)
-       // result: (VPMOVMToVec64x8 (VPCMPUQ512 [6] x y))
+       // result: (VPMOVMToVec64x8 (VPCMPUQ512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQ512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32143,13 +32143,13 @@ func rewriteValueAMD64_OpGreaterUint8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint8x16 x y)
-       // result: (VPMOVMToVec8x16 (VPCMPUB128 [6] x y))
+       // result: (VPMOVMToVec8x16 (VPCMPUB128 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUB128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32161,13 +32161,13 @@ func rewriteValueAMD64_OpGreaterUint8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint8x32 x y)
-       // result: (VPMOVMToVec8x32 (VPCMPUB256 [6] x y))
+       // result: (VPMOVMToVec8x32 (VPCMPUB256 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUB256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -32179,13 +32179,13 @@ func rewriteValueAMD64_OpGreaterUint8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (GreaterUint8x64 x y)
-       // result: (VPMOVMToVec8x64 (VPCMPUB512 [6] x y))
+       // result: (VPMOVMToVec8x64 (VPCMPUB512 [14] x y))
        for {
                x := v_0
                y := v_1
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUB512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v0.AddArg2(x, y)
                v.AddArg(v0)
                return true
@@ -38162,14 +38162,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualFloat32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualFloat32x16 x y mask)
-       // result: (VPMOVMToVec32x16 (VCMPPSMasked512 [5] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x16 (VCMPPSMasked512 [13] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPSMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38184,14 +38184,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualFloat32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualFloat32x4 x y mask)
-       // result: (VPMOVMToVec32x4 (VCMPPSMasked128 [5] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x4 (VCMPPSMasked128 [13] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPSMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38206,14 +38206,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualFloat32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualFloat32x8 x y mask)
-       // result: (VPMOVMToVec32x8 (VCMPPSMasked256 [5] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x8 (VCMPPSMasked256 [13] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPSMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38228,14 +38228,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualFloat64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualFloat64x2 x y mask)
-       // result: (VPMOVMToVec64x2 (VCMPPDMasked128 [5] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x2 (VCMPPDMasked128 [13] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPDMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x2ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38250,14 +38250,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualFloat64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualFloat64x4 x y mask)
-       // result: (VPMOVMToVec64x4 (VCMPPDMasked256 [5] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x4 (VCMPPDMasked256 [13] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPDMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38272,14 +38272,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualFloat64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualFloat64x8 x y mask)
-       // result: (VPMOVMToVec64x8 (VCMPPDMasked512 [5] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x8 (VCMPPDMasked512 [13] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPDMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38294,14 +38294,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt16x16 x y mask)
-       // result: (VPMOVMToVec16x16 (VPCMPWMasked256 [5] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x16 (VPCMPWMasked256 [13] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPWMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38316,14 +38316,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt16x32 x y mask)
-       // result: (VPMOVMToVec16x32 (VPCMPWMasked512 [5] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x32 (VPCMPWMasked512 [13] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPWMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38338,14 +38338,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt16x8 x y mask)
-       // result: (VPMOVMToVec16x8 (VPCMPWMasked128 [5] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x8 (VPCMPWMasked128 [13] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPWMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38360,14 +38360,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt32x16 x y mask)
-       // result: (VPMOVMToVec32x16 (VPCMPDMasked512 [5] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x16 (VPCMPDMasked512 [13] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPDMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38382,14 +38382,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt32x4 x y mask)
-       // result: (VPMOVMToVec32x4 (VPCMPDMasked128 [5] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x4 (VPCMPDMasked128 [13] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPDMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38404,14 +38404,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt32x8 x y mask)
-       // result: (VPMOVMToVec32x8 (VPCMPDMasked256 [5] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x8 (VPCMPDMasked256 [13] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPDMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38426,14 +38426,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt64x2 x y mask)
-       // result: (VPMOVMToVec64x2 (VPCMPQMasked128 [5] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x2 (VPCMPQMasked128 [13] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x2ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38448,14 +38448,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt64x4 x y mask)
-       // result: (VPMOVMToVec64x4 (VPCMPQMasked256 [5] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x4 (VPCMPQMasked256 [13] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38470,14 +38470,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt64x8 x y mask)
-       // result: (VPMOVMToVec64x8 (VPCMPQMasked512 [5] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x8 (VPCMPQMasked512 [13] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38492,14 +38492,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt8x16 x y mask)
-       // result: (VPMOVMToVec8x16 (VPCMPBMasked128 [5] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x16 (VPCMPBMasked128 [13] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPBMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38514,14 +38514,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt8x32 x y mask)
-       // result: (VPMOVMToVec8x32 (VPCMPBMasked256 [5] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x32 (VPCMPBMasked256 [13] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPBMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38536,14 +38536,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualInt8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualInt8x64 x y mask)
-       // result: (VPMOVMToVec8x64 (VPCMPBMasked512 [5] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x64 (VPCMPBMasked512 [13] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPBMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x64ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38558,14 +38558,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint16x16 x y mask)
-       // result: (VPMOVMToVec16x16 (VPCMPUWMasked256 [5] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x16 (VPCMPUWMasked256 [13] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUWMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38580,14 +38580,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint16x32 x y mask)
-       // result: (VPMOVMToVec16x32 (VPCMPUWMasked512 [5] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x32 (VPCMPUWMasked512 [13] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUWMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38602,14 +38602,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint16x8 x y mask)
-       // result: (VPMOVMToVec16x8 (VPCMPUWMasked128 [5] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x8 (VPCMPUWMasked128 [13] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUWMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38624,14 +38624,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint32x16 x y mask)
-       // result: (VPMOVMToVec32x16 (VPCMPUDMasked512 [5] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x16 (VPCMPUDMasked512 [13] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUDMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38646,14 +38646,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint32x4 x y mask)
-       // result: (VPMOVMToVec32x4 (VPCMPUDMasked128 [5] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x4 (VPCMPUDMasked128 [13] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUDMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38668,14 +38668,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint32x8 x y mask)
-       // result: (VPMOVMToVec32x8 (VPCMPUDMasked256 [5] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x8 (VPCMPUDMasked256 [13] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUDMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38690,14 +38690,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint64x2 x y mask)
-       // result: (VPMOVMToVec64x2 (VPCMPUQMasked128 [5] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x2 (VPCMPUQMasked128 [13] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x2ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38712,14 +38712,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint64x4 x y mask)
-       // result: (VPMOVMToVec64x4 (VPCMPUQMasked256 [5] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x4 (VPCMPUQMasked256 [13] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38734,14 +38734,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint64x8 x y mask)
-       // result: (VPMOVMToVec64x8 (VPCMPUQMasked512 [5] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x8 (VPCMPUQMasked512 [13] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38756,14 +38756,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint8x16 x y mask)
-       // result: (VPMOVMToVec8x16 (VPCMPUBMasked128 [5] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x16 (VPCMPUBMasked128 [13] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUBMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38778,14 +38778,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint8x32 x y mask)
-       // result: (VPMOVMToVec8x32 (VPCMPUBMasked256 [5] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x32 (VPCMPUBMasked256 [13] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUBMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38800,14 +38800,14 @@ func rewriteValueAMD64_OpMaskedGreaterEqualUint8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterEqualUint8x64 x y mask)
-       // result: (VPMOVMToVec8x64 (VPCMPUBMasked512 [5] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x64 (VPCMPUBMasked512 [13] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUBMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(5)
+               v0.AuxInt = int8ToAuxInt(13)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x64ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38822,14 +38822,14 @@ func rewriteValueAMD64_OpMaskedGreaterFloat32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterFloat32x16 x y mask)
-       // result: (VPMOVMToVec32x16 (VCMPPSMasked512 [6] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x16 (VCMPPSMasked512 [14] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPSMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38844,14 +38844,14 @@ func rewriteValueAMD64_OpMaskedGreaterFloat32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterFloat32x4 x y mask)
-       // result: (VPMOVMToVec32x4 (VCMPPSMasked128 [6] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x4 (VCMPPSMasked128 [14] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPSMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38866,14 +38866,14 @@ func rewriteValueAMD64_OpMaskedGreaterFloat32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterFloat32x8 x y mask)
-       // result: (VPMOVMToVec32x8 (VCMPPSMasked256 [6] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x8 (VCMPPSMasked256 [14] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPSMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38888,14 +38888,14 @@ func rewriteValueAMD64_OpMaskedGreaterFloat64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterFloat64x2 x y mask)
-       // result: (VPMOVMToVec64x2 (VCMPPDMasked128 [6] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x2 (VCMPPDMasked128 [14] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPDMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x2ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38910,14 +38910,14 @@ func rewriteValueAMD64_OpMaskedGreaterFloat64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterFloat64x4 x y mask)
-       // result: (VPMOVMToVec64x4 (VCMPPDMasked256 [6] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x4 (VCMPPDMasked256 [14] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPDMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38932,14 +38932,14 @@ func rewriteValueAMD64_OpMaskedGreaterFloat64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterFloat64x8 x y mask)
-       // result: (VPMOVMToVec64x8 (VCMPPDMasked512 [6] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x8 (VCMPPDMasked512 [14] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VCMPPDMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38954,14 +38954,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt16x16 x y mask)
-       // result: (VPMOVMToVec16x16 (VPCMPWMasked256 [6] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x16 (VPCMPWMasked256 [14] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPWMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38976,14 +38976,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt16x32 x y mask)
-       // result: (VPMOVMToVec16x32 (VPCMPWMasked512 [6] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x32 (VPCMPWMasked512 [14] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPWMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -38998,14 +38998,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt16x8 x y mask)
-       // result: (VPMOVMToVec16x8 (VPCMPWMasked128 [6] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x8 (VPCMPWMasked128 [14] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPWMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39020,14 +39020,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt32x16 x y mask)
-       // result: (VPMOVMToVec32x16 (VPCMPDMasked512 [6] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x16 (VPCMPDMasked512 [14] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPDMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39042,14 +39042,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt32x4 x y mask)
-       // result: (VPMOVMToVec32x4 (VPCMPDMasked128 [6] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x4 (VPCMPDMasked128 [14] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPDMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39064,14 +39064,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt32x8 x y mask)
-       // result: (VPMOVMToVec32x8 (VPCMPDMasked256 [6] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x8 (VPCMPDMasked256 [14] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPDMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39086,14 +39086,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt64x2 x y mask)
-       // result: (VPMOVMToVec64x2 (VPCMPQMasked128 [6] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x2 (VPCMPQMasked128 [14] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x2ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39108,14 +39108,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt64x4 x y mask)
-       // result: (VPMOVMToVec64x4 (VPCMPQMasked256 [6] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x4 (VPCMPQMasked256 [14] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39130,14 +39130,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt64x8 x y mask)
-       // result: (VPMOVMToVec64x8 (VPCMPQMasked512 [6] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x8 (VPCMPQMasked512 [14] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPQMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39152,14 +39152,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt8x16 x y mask)
-       // result: (VPMOVMToVec8x16 (VPCMPBMasked128 [6] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x16 (VPCMPBMasked128 [14] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPBMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39174,14 +39174,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt8x32 x y mask)
-       // result: (VPMOVMToVec8x32 (VPCMPBMasked256 [6] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x32 (VPCMPBMasked256 [14] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPBMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39196,14 +39196,14 @@ func rewriteValueAMD64_OpMaskedGreaterInt8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterInt8x64 x y mask)
-       // result: (VPMOVMToVec8x64 (VPCMPBMasked512 [6] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x64 (VPCMPBMasked512 [14] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPBMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x64ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39218,14 +39218,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint16x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint16x16 x y mask)
-       // result: (VPMOVMToVec16x16 (VPCMPUWMasked256 [6] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x16 (VPCMPUWMasked256 [14] x y (VPMOVVec16x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUWMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39240,14 +39240,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint16x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint16x32 x y mask)
-       // result: (VPMOVMToVec16x32 (VPCMPUWMasked512 [6] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x32 (VPCMPUWMasked512 [14] x y (VPMOVVec16x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUWMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39262,14 +39262,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint16x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint16x8 x y mask)
-       // result: (VPMOVMToVec16x8 (VPCMPUWMasked128 [6] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec16x8 (VPCMPUWMasked128 [14] x y (VPMOVVec16x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec16x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUWMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec16x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39284,14 +39284,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint32x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint32x16 x y mask)
-       // result: (VPMOVMToVec32x16 (VPCMPUDMasked512 [6] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x16 (VPCMPUDMasked512 [14] x y (VPMOVVec32x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUDMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39306,14 +39306,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint32x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint32x4 x y mask)
-       // result: (VPMOVMToVec32x4 (VPCMPUDMasked128 [6] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x4 (VPCMPUDMasked128 [14] x y (VPMOVVec32x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUDMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39328,14 +39328,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint32x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint32x8 x y mask)
-       // result: (VPMOVMToVec32x8 (VPCMPUDMasked256 [6] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec32x8 (VPCMPUDMasked256 [14] x y (VPMOVVec32x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec32x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUDMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec32x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39350,14 +39350,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint64x2(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint64x2 x y mask)
-       // result: (VPMOVMToVec64x2 (VPCMPUQMasked128 [6] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x2 (VPCMPUQMasked128 [14] x y (VPMOVVec64x2ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x2)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x2ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39372,14 +39372,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint64x4(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint64x4 x y mask)
-       // result: (VPMOVMToVec64x4 (VPCMPUQMasked256 [6] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x4 (VPCMPUQMasked256 [14] x y (VPMOVVec64x4ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x4)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x4ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39394,14 +39394,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint64x8(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint64x8 x y mask)
-       // result: (VPMOVMToVec64x8 (VPCMPUQMasked512 [6] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec64x8 (VPCMPUQMasked512 [14] x y (VPMOVVec64x8ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec64x8)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUQMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec64x8ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39416,14 +39416,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint8x16(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint8x16 x y mask)
-       // result: (VPMOVMToVec8x16 (VPCMPUBMasked128 [6] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x16 (VPCMPUBMasked128 [14] x y (VPMOVVec8x16ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x16)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUBMasked128, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x16ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39438,14 +39438,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint8x32(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint8x32 x y mask)
-       // result: (VPMOVMToVec8x32 (VPCMPUBMasked256 [6] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x32 (VPCMPUBMasked256 [14] x y (VPMOVVec8x32ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x32)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUBMasked256, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x32ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)
@@ -39460,14 +39460,14 @@ func rewriteValueAMD64_OpMaskedGreaterUint8x64(v *Value) bool {
        b := v.Block
        typ := &b.Func.Config.Types
        // match: (MaskedGreaterUint8x64 x y mask)
-       // result: (VPMOVMToVec8x64 (VPCMPUBMasked512 [6] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
+       // result: (VPMOVMToVec8x64 (VPCMPUBMasked512 [14] x y (VPMOVVec8x64ToM <types.TypeMask> mask)))
        for {
                x := v_0
                y := v_1
                mask := v_2
                v.reset(OpAMD64VPMOVMToVec8x64)
                v0 := b.NewValue0(v.Pos, OpAMD64VPCMPUBMasked512, typ.Mask)
-               v0.AuxInt = int8ToAuxInt(6)
+               v0.AuxInt = int8ToAuxInt(14)
                v1 := b.NewValue0(v.Pos, OpAMD64VPMOVVec8x64ToM, types.TypeMask)
                v1.AddArg(mask)
                v0.AddArg3(x, y, v1)