[SVE] Lower fixed length vector integer subtract operations.

[efriedma]

llvm/lib/Target/AArch64/AArch64ISelLowering.cpp

3633

Is the change to ISD::ADD here related somehow?

[paulwalker-arm]

llvm/lib/Target/AArch64/AArch64ISelLowering.cpp

3633

Only in that ADD and SUB should look pretty much identical and I didn't want to copy the "useSVE..." test, which I probably shouldn't have added in the first place for ADD.

Should I move this change into a separate patch?

[efriedma]

It's fine. LGTM

llvm/lib/Target/AArch64/AArch64ISelLowering.h

enum NodeType : unsigned {
   FSUB_PRED,
   MUL_PRED,
   SDIV_PRED,
   SHL_PRED,
   SMAX_PRED,
   SMIN_PRED,
   SRA_PRED,
   SRL_PRED,
+  SUB_PRED,
   UDIV_PRED,
   UMAX_PRED,
   UMIN_PRED,
 
   SETCC_MERGE_ZERO,
 
   // Arithmetic instructions which write flags.
   ADDS,

llvm/lib/Target/AArch64/AArch64ISelLowering.cpp

void AArch64TargetLowering::addTypeForFixedLengthSVE(MVT VT) {
   setOperationAction(ISD::FDIV, VT, Custom);
   setOperationAction(ISD::FMA, VT, Custom);
   setOperationAction(ISD::FMUL, VT, Custom);
   setOperationAction(ISD::FSUB, VT, Custom);
   setOperationAction(ISD::LOAD, VT, Custom);
   setOperationAction(ISD::MUL, VT, Custom);
   setOperationAction(ISD::OR, VT, Custom);
   setOperationAction(ISD::STORE, VT, Custom);
+  setOperationAction(ISD::SUB, VT, Custom);
   setOperationAction(ISD::TRUNCATE, VT, Custom);
   setOperationAction(ISD::XOR, VT, Custom);
 }
 
 void AArch64TargetLowering::addDRTypeForNEON(MVT VT) {
   addRegisterClass(VT, &AArch64::FPR64RegClass);
   addTypeForNEON(VT, MVT::v2i32);
 }
case AArch64ISD::FIRST_NUMBER:
     MAKE_CASE(AArch64ISD::ADD_PRED)
     MAKE_CASE(AArch64ISD::MUL_PRED)
     MAKE_CASE(AArch64ISD::SDIV_PRED)
     MAKE_CASE(AArch64ISD::SHL_PRED)
     MAKE_CASE(AArch64ISD::SMAX_PRED)
     MAKE_CASE(AArch64ISD::SMIN_PRED)
     MAKE_CASE(AArch64ISD::SRA_PRED)
     MAKE_CASE(AArch64ISD::SRL_PRED)
+    MAKE_CASE(AArch64ISD::SUB_PRED)
     MAKE_CASE(AArch64ISD::UDIV_PRED)
     MAKE_CASE(AArch64ISD::UMAX_PRED)
     MAKE_CASE(AArch64ISD::UMIN_PRED)
     MAKE_CASE(AArch64ISD::SETCC_MERGE_ZERO)
     MAKE_CASE(AArch64ISD::ADC)
     MAKE_CASE(AArch64ISD::SBC)
     MAKE_CASE(AArch64ISD::ADDS)
     MAKE_CASE(AArch64ISD::SUBS)
case ISD::VSCALE:
     return LowerVSCALE(Op, DAG);
   case ISD::TRUNCATE:
     return LowerTRUNCATE(Op, DAG);
   case ISD::LOAD:
     if (useSVEForFixedLengthVectorVT(Op.getValueType()))
       return LowerFixedLengthVectorLoadToSVE(Op, DAG);
     llvm_unreachable("Unexpected request to lower ISD::LOAD");
   case ISD::ADD:
-    if (useSVEForFixedLengthVectorVT(Op.getValueType()))
-      return LowerToPredicatedOp(Op, DAG, AArch64ISD::ADD_PRED);
-    llvm_unreachable("Unexpected request to lower ISD::ADD");
   case ISD::AND:
     return LowerToScalableOp(Op, DAG);
+  case ISD::SUB:
+    return LowerToPredicatedOp(Op, DAG, AArch64ISD::SUB_PRED);
   }
 }
 
 bool AArch64TargetLowering::useSVEForFixedLengthVectors() const {
   // Prefer NEON unless larger SVE registers are available.
   return Subtarget->hasSVE() && Subtarget->getMinSVEVectorSizeInBits() >= 256;
 }
 

llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td

 def AArch64fmul_p : SDNode<"AArch64ISD::FMUL_PRED", SDT_AArch64Arith>;
 def AArch64fsub_p : SDNode<"AArch64ISD::FSUB_PRED", SDT_AArch64Arith>;
 def AArch64lsl_p  : SDNode<"AArch64ISD::SHL_PRED",  SDT_AArch64Arith>;
 def AArch64lsr_p  : SDNode<"AArch64ISD::SRL_PRED",  SDT_AArch64Arith>;
 def AArch64mul_p  : SDNode<"AArch64ISD::MUL_PRED",  SDT_AArch64Arith>;
 def AArch64sdiv_p : SDNode<"AArch64ISD::SDIV_PRED", SDT_AArch64Arith>;
 def AArch64smax_p : SDNode<"AArch64ISD::SMAX_PRED", SDT_AArch64Arith>;
 def AArch64smin_p : SDNode<"AArch64ISD::SMIN_PRED", SDT_AArch64Arith>;
+def AArch64sub_p  : SDNode<"AArch64ISD::SUB_PRED",  SDT_AArch64Arith>;
 def AArch64udiv_p : SDNode<"AArch64ISD::UDIV_PRED", SDT_AArch64Arith>;
 def AArch64umax_p : SDNode<"AArch64ISD::UMAX_PRED", SDT_AArch64Arith>;
 def AArch64umin_p : SDNode<"AArch64ISD::UMIN_PRED", SDT_AArch64Arith>;
 
 def SDT_AArch64ReduceWithInit : SDTypeProfile<1, 3, [SDTCisVec<1>, SDTCisVec<3>]>;
 def AArch64clasta_n   : SDNode<"AArch64ISD::CLASTA_N",   SDT_AArch64ReduceWithInit>;
 def AArch64clastb_n   : SDNode<"AArch64ISD::CLASTB_N",   SDT_AArch64ReduceWithInit>;
 def AArch64fadda_p    : SDNode<"AArch64ISD::FADDA_PRED", SDT_AArch64ReduceWithInit>;
let Predicates = [HasSVE] in {
   defm EOR_ZZZ : sve_int_bin_cons_log<0b10, "eor", xor>;
   defm BIC_ZZZ : sve_int_bin_cons_log<0b11, "bic", null_frag>;
 
   defm ADD_ZPmZ  : sve_int_bin_pred_arit_0<0b000, "add",  "ADD_ZPZZ", int_aarch64_sve_add, DestructiveBinaryComm>;
   defm SUB_ZPmZ  : sve_int_bin_pred_arit_0<0b001, "sub",  "SUB_ZPZZ", int_aarch64_sve_sub, DestructiveBinaryCommWithRev, "SUBR_ZPmZ">;
   defm SUBR_ZPmZ : sve_int_bin_pred_arit_0<0b011, "subr", "SUBR_ZPZZ", int_aarch64_sve_subr, DestructiveBinaryCommWithRev, "SUB_ZPmZ", /*isReverseInstr*/ 1>;
 
   defm ADD_ZPZZ  : sve_int_bin_pred_bhsd<AArch64add_p>;
+  defm SUB_ZPZZ  : sve_int_bin_pred_bhsd<AArch64sub_p>;
 
   let Predicates = [HasSVE, UseExperimentalZeroingPseudos] in {
     defm ADD_ZPZZ  : sve_int_bin_pred_zeroing_bhsd<int_aarch64_sve_add>;
     defm SUB_ZPZZ  : sve_int_bin_pred_zeroing_bhsd<int_aarch64_sve_sub>;
     defm SUBR_ZPZZ : sve_int_bin_pred_zeroing_bhsd<int_aarch64_sve_subr>;
   }
 
   defm ORR_ZPmZ : sve_int_bin_pred_log<0b000, "orr", int_aarch64_sve_orr>;

llvm/test/CodeGen/AArch64/sve-fixed-length-int-arith.ll

 ; CHECK: ret
   %op1 = load <32 x i64>, <32 x i64>* %a
   %op2 = load <32 x i64>, <32 x i64>* %b
   %res = mul <32 x i64> %op1, %op2
   store <32 x i64> %res, <32 x i64>* %a
   ret void
 }
 
+;
+; SUB
+;
+
+; Don't use SVE for 64-bit vectors.
+define <8 x i8> @sub_v8i8(<8 x i8> %op1, <8 x i8> %op2) #0 {
+; CHECK-LABEL: sub_v8i8:
+; CHECK: sub v0.8b, v0.8b, v1.8b
+; CHECK: ret
+  %res = sub <8 x i8> %op1, %op2
+  ret <8 x i8> %res
+}
+
+; Don't use SVE for 128-bit vectors.
+define <16 x i8> @sub_v16i8(<16 x i8> %op1, <16 x i8> %op2) #0 {
+; CHECK-LABEL: sub_v16i8:
+; CHECK: sub v0.16b, v0.16b, v1.16b
+; CHECK: ret
+  %res = sub <16 x i8> %op1, %op2
+  ret <16 x i8> %res
+}
+
+define void @sub_v32i8(<32 x i8>* %a, <32 x i8>* %b) #0 {
+; CHECK-LABEL: sub_v32i8:
+; CHECK: ptrue [[PG:p[0-9]+]].b, vl[[#min(VBYTES,32)]]
+; CHECK-DAG: ld1b { [[OP1:z[0-9]+]].b }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1b { [[OP2:z[0-9]+]].b }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].b, [[PG]]/m, [[OP1]].b, [[OP2]].b
+; CHECK: st1b { [[RES]].b }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <32 x i8>, <32 x i8>* %a
+  %op2 = load <32 x i8>, <32 x i8>* %b
+  %res = sub <32 x i8> %op1, %op2
+  store <32 x i8> %res, <32 x i8>* %a
+  ret void
+}
+
+define void @sub_v64i8(<64 x i8>* %a, <64 x i8>* %b) #0 {
+; CHECK-LABEL: sub_v64i8:
+; CHECK: ptrue [[PG:p[0-9]+]].b, vl[[#min(VBYTES,64)]]
+; CHECK-DAG: ld1b { [[OP1:z[0-9]+]].b }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1b { [[OP2:z[0-9]+]].b }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].b, [[PG]]/m, [[OP1]].b, [[OP2]].b
+; CHECK: st1b { [[RES]].b }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <64 x i8>, <64 x i8>* %a
+  %op2 = load <64 x i8>, <64 x i8>* %b
+  %res = sub <64 x i8> %op1, %op2
+  store <64 x i8> %res, <64 x i8>* %a
+  ret void
+}
+
+define void @sub_v128i8(<128 x i8>* %a, <128 x i8>* %b) #0 {
+; CHECK-LABEL: sub_v128i8:
+; CHECK: ptrue [[PG:p[0-9]+]].b, vl[[#min(VBYTES,128)]]
+; CHECK-DAG: ld1b { [[OP1:z[0-9]+]].b }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1b { [[OP2:z[0-9]+]].b }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].b, [[PG]]/m, [[OP1]].b, [[OP2]].b
+; CHECK: st1b { [[RES]].b }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <128 x i8>, <128 x i8>* %a
+  %op2 = load <128 x i8>, <128 x i8>* %b
+  %res = sub <128 x i8> %op1, %op2
+  store <128 x i8> %res, <128 x i8>* %a
+  ret void
+}
+
+define void @sub_v256i8(<256 x i8>* %a, <256 x i8>* %b) #0 {
+; CHECK-LABEL: sub_v256i8:
+; CHECK: ptrue [[PG:p[0-9]+]].b, vl[[#min(VBYTES,256)]]
+; CHECK-DAG: ld1b { [[OP1:z[0-9]+]].b }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1b { [[OP2:z[0-9]+]].b }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].b, [[PG]]/m, [[OP1]].b, [[OP2]].b
+; CHECK: st1b { [[RES]].b }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <256 x i8>, <256 x i8>* %a
+  %op2 = load <256 x i8>, <256 x i8>* %b
+  %res = sub <256 x i8> %op1, %op2
+  store <256 x i8> %res, <256 x i8>* %a
+  ret void
+}
+
+; Don't use SVE for 64-bit vectors.
+define <4 x i16> @sub_v4i16(<4 x i16> %op1, <4 x i16> %op2) #0 {
+; CHECK-LABEL: sub_v4i16:
+; CHECK: sub v0.4h, v0.4h, v1.4h
+; CHECK: ret
+  %res = sub <4 x i16> %op1, %op2
+  ret <4 x i16> %res
+}
+
+; Don't use SVE for 128-bit vectors.
+define <8 x i16> @sub_v8i16(<8 x i16> %op1, <8 x i16> %op2) #0 {
+; CHECK-LABEL: sub_v8i16:
+; CHECK: sub v0.8h, v0.8h, v1.8h
+; CHECK: ret
+  %res = sub <8 x i16> %op1, %op2
+  ret <8 x i16> %res
+}
+
+define void @sub_v16i16(<16 x i16>* %a, <16 x i16>* %b) #0 {
+; CHECK-LABEL: sub_v16i16:
+; CHECK: ptrue [[PG:p[0-9]+]].h, vl[[#min(div(VBYTES,2),16)]]
+; CHECK-DAG: ld1h { [[OP1:z[0-9]+]].h }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1h { [[OP2:z[0-9]+]].h }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].h, [[PG]]/m, [[OP1]].h, [[OP2]].h
+; CHECK: st1h { [[RES]].h }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <16 x i16>, <16 x i16>* %a
+  %op2 = load <16 x i16>, <16 x i16>* %b
+  %res = sub <16 x i16> %op1, %op2
+  store <16 x i16> %res, <16 x i16>* %a
+  ret void
+}
+
+define void @sub_v32i16(<32 x i16>* %a, <32 x i16>* %b) #0 {
+; CHECK-LABEL: sub_v32i16:
+; CHECK: ptrue [[PG:p[0-9]+]].h, vl[[#min(div(VBYTES,2),32)]]
+; CHECK-DAG: ld1h { [[OP1:z[0-9]+]].h }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1h { [[OP2:z[0-9]+]].h }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].h, [[PG]]/m, [[OP1]].h, [[OP2]].h
+; CHECK: st1h { [[RES]].h }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <32 x i16>, <32 x i16>* %a
+  %op2 = load <32 x i16>, <32 x i16>* %b
+  %res = sub <32 x i16> %op1, %op2
+  store <32 x i16> %res, <32 x i16>* %a
+  ret void
+}
+
+define void @sub_v64i16(<64 x i16>* %a, <64 x i16>* %b) #0 {
+; CHECK-LABEL: sub_v64i16:
+; CHECK: ptrue [[PG:p[0-9]+]].h, vl[[#min(div(VBYTES,2),64)]]
+; CHECK-DAG: ld1h { [[OP1:z[0-9]+]].h }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1h { [[OP2:z[0-9]+]].h }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].h, [[PG]]/m, [[OP1]].h, [[OP2]].h
+; CHECK: st1h { [[RES]].h }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <64 x i16>, <64 x i16>* %a
+  %op2 = load <64 x i16>, <64 x i16>* %b
+  %res = sub <64 x i16> %op1, %op2
+  store <64 x i16> %res, <64 x i16>* %a
+  ret void
+}
+
+define void @sub_v128i16(<128 x i16>* %a, <128 x i16>* %b) #0 {
+; CHECK-LABEL: sub_v128i16:
+; CHECK: ptrue [[PG:p[0-9]+]].h, vl[[#min(div(VBYTES,2),128)]]
+; CHECK-DAG: ld1h { [[OP1:z[0-9]+]].h }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1h { [[OP2:z[0-9]+]].h }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].h, [[PG]]/m, [[OP1]].h, [[OP2]].h
+; CHECK: st1h { [[RES]].h }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <128 x i16>, <128 x i16>* %a
+  %op2 = load <128 x i16>, <128 x i16>* %b
+  %res = sub <128 x i16> %op1, %op2
+  store <128 x i16> %res, <128 x i16>* %a
+  ret void
+}
+
+; Don't use SVE for 64-bit vectors.
+define <2 x i32> @sub_v2i32(<2 x i32> %op1, <2 x i32> %op2) #0 {
+; CHECK-LABEL: sub_v2i32:
+; CHECK: sub v0.2s, v0.2s, v1.2s
+; CHECK: ret
+  %res = sub <2 x i32> %op1, %op2
+  ret <2 x i32> %res
+}
+
+; Don't use SVE for 128-bit vectors.
+define <4 x i32> @sub_v4i32(<4 x i32> %op1, <4 x i32> %op2) #0 {
+; CHECK-LABEL: sub_v4i32:
+; CHECK: sub v0.4s, v0.4s, v1.4s
+; CHECK: ret
+  %res = sub <4 x i32> %op1, %op2
+  ret <4 x i32> %res
+}
+
+define void @sub_v8i32(<8 x i32>* %a, <8 x i32>* %b) #0 {
+; CHECK-LABEL: sub_v8i32:
+; CHECK: ptrue [[PG:p[0-9]+]].s, vl[[#min(div(VBYTES,4),8)]]
+; CHECK-DAG: ld1w { [[OP1:z[0-9]+]].s }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1w { [[OP2:z[0-9]+]].s }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].s, [[PG]]/m, [[OP1]].s, [[OP2]].s
+; CHECK: st1w { [[RES]].s }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <8 x i32>, <8 x i32>* %a
+  %op2 = load <8 x i32>, <8 x i32>* %b
+  %res = sub <8 x i32> %op1, %op2
+  store <8 x i32> %res, <8 x i32>* %a
+  ret void
+}
+
+define void @sub_v16i32(<16 x i32>* %a, <16 x i32>* %b) #0 {
+; CHECK-LABEL: sub_v16i32:
+; CHECK: ptrue [[PG:p[0-9]+]].s, vl[[#min(div(VBYTES,4),16)]]
+; CHECK-DAG: ld1w { [[OP1:z[0-9]+]].s }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1w { [[OP2:z[0-9]+]].s }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].s, [[PG]]/m, [[OP1]].s, [[OP2]].s
+; CHECK: st1w { [[RES]].s }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <16 x i32>, <16 x i32>* %a
+  %op2 = load <16 x i32>, <16 x i32>* %b
+  %res = sub <16 x i32> %op1, %op2
+  store <16 x i32> %res, <16 x i32>* %a
+  ret void
+}
+
+define void @sub_v32i32(<32 x i32>* %a, <32 x i32>* %b) #0 {
+; CHECK-LABEL: sub_v32i32:
+; CHECK: ptrue [[PG:p[0-9]+]].s, vl[[#min(div(VBYTES,4),32)]]
+; CHECK-DAG: ld1w { [[OP1:z[0-9]+]].s }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1w { [[OP2:z[0-9]+]].s }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].s, [[PG]]/m, [[OP1]].s, [[OP2]].s
+; CHECK: st1w { [[RES]].s }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <32 x i32>, <32 x i32>* %a
+  %op2 = load <32 x i32>, <32 x i32>* %b
+  %res = sub <32 x i32> %op1, %op2
+  store <32 x i32> %res, <32 x i32>* %a
+  ret void
+}
+
+define void @sub_v64i32(<64 x i32>* %a, <64 x i32>* %b) #0 {
+; CHECK-LABEL: sub_v64i32:
+; CHECK: ptrue [[PG:p[0-9]+]].s, vl[[#min(div(VBYTES,4),64)]]
+; CHECK-DAG: ld1w { [[OP1:z[0-9]+]].s }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1w { [[OP2:z[0-9]+]].s }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].s, [[PG]]/m, [[OP1]].s, [[OP2]].s
+; CHECK: st1w { [[RES]].s }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <64 x i32>, <64 x i32>* %a
+  %op2 = load <64 x i32>, <64 x i32>* %b
+  %res = sub <64 x i32> %op1, %op2
+  store <64 x i32> %res, <64 x i32>* %a
+  ret void
+}
+
+; Don't use SVE for 64-bit vectors.
+define <1 x i64> @sub_v1i64(<1 x i64> %op1, <1 x i64> %op2) #0 {
+; CHECK-LABEL: sub_v1i64:
+; CHECK: sub d0, d0, d1
+; CHECK: ret
+  %res = sub <1 x i64> %op1, %op2
+  ret <1 x i64> %res
+}
+
+; Don't use SVE for 128-bit vectors.
+define <2 x i64> @sub_v2i64(<2 x i64> %op1, <2 x i64> %op2) #0 {
+; CHECK-LABEL: sub_v2i64:
+; CHECK: sub v0.2d, v0.2d, v1.2d
+; CHECK: ret
+  %res = sub <2 x i64> %op1, %op2
+  ret <2 x i64> %res
+}
+
+define void @sub_v4i64(<4 x i64>* %a, <4 x i64>* %b) #0 {
+; CHECK-LABEL: sub_v4i64:
+; CHECK: ptrue [[PG:p[0-9]+]].d, vl[[#min(div(VBYTES,8),4)]]
+; CHECK-DAG: ld1d { [[OP1:z[0-9]+]].d }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1d { [[OP2:z[0-9]+]].d }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].d, [[PG]]/m, [[OP1]].d, [[OP2]].d
+; CHECK: st1d { [[RES]].d }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <4 x i64>, <4 x i64>* %a
+  %op2 = load <4 x i64>, <4 x i64>* %b
+  %res = sub <4 x i64> %op1, %op2
+  store <4 x i64> %res, <4 x i64>* %a
+  ret void
+}
+
+define void @sub_v8i64(<8 x i64>* %a, <8 x i64>* %b) #0 {
+; CHECK-LABEL: sub_v8i64:
+; CHECK: ptrue [[PG:p[0-9]+]].d, vl[[#min(div(VBYTES,8),8)]]
+; CHECK-DAG: ld1d { [[OP1:z[0-9]+]].d }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1d { [[OP2:z[0-9]+]].d }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].d, [[PG]]/m, [[OP1]].d, [[OP2]].d
+; CHECK: st1d { [[RES]].d }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <8 x i64>, <8 x i64>* %a
+  %op2 = load <8 x i64>, <8 x i64>* %b
+  %res = sub <8 x i64> %op1, %op2
+  store <8 x i64> %res, <8 x i64>* %a
+  ret void
+}
+
+define void @sub_v16i64(<16 x i64>* %a, <16 x i64>* %b) #0 {
+; CHECK-LABEL: sub_v16i64:
+; CHECK: ptrue [[PG:p[0-9]+]].d, vl[[#min(div(VBYTES,8),16)]]
+; CHECK-DAG: ld1d { [[OP1:z[0-9]+]].d }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1d { [[OP2:z[0-9]+]].d }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].d, [[PG]]/m, [[OP1]].d, [[OP2]].d
+; CHECK: st1d { [[RES]].d }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <16 x i64>, <16 x i64>* %a
+  %op2 = load <16 x i64>, <16 x i64>* %b
+  %res = sub <16 x i64> %op1, %op2
+  store <16 x i64> %res, <16 x i64>* %a
+  ret void
+}
+
+define void @sub_v32i64(<32 x i64>* %a, <32 x i64>* %b) #0 {
+; CHECK-LABEL: sub_v32i64:
+; CHECK: ptrue [[PG:p[0-9]+]].d, vl[[#min(div(VBYTES,8),32)]]
+; CHECK-DAG: ld1d { [[OP1:z[0-9]+]].d }, [[PG]]/z, [x0]
+; CHECK-DAG: ld1d { [[OP2:z[0-9]+]].d }, [[PG]]/z, [x1]
+; CHECK: sub [[RES:z[0-9]+]].d, [[PG]]/m, [[OP1]].d, [[OP2]].d
+; CHECK: st1d { [[RES]].d }, [[PG]], [x0]
+; CHECK: ret
+  %op1 = load <32 x i64>, <32 x i64>* %a
+  %op2 = load <32 x i64>, <32 x i64>* %b
+  %res = sub <32 x i64> %op1, %op2
+  store <32 x i64> %res, <32 x i64>* %a
+  ret void
+}
+
 attributes #0 = { "target-features"="+sve" }