// instructions for splatting repeating bit patterns across the immediate.
def logical_imm32_XFORM : SDNodeXForm<imm, [{
uint64_t enc = AArch64_AM::encodeLogicalImmediate(N->getZExtValue(), 32);
- return CurDAG->getTargetConstant(enc, MVT::i32);
+ return CurDAG->getTargetConstant(enc, SDLoc(N), MVT::i32);
}]>;
def logical_imm64_XFORM : SDNodeXForm<imm, [{
uint64_t enc = AArch64_AM::encodeLogicalImmediate(N->getZExtValue(), 64);
- return CurDAG->getTargetConstant(enc, MVT::i32);
+ return CurDAG->getTargetConstant(enc, SDLoc(N), MVT::i32);
}]>;
let DiagnosticType = "LogicalSecondSource" in {
let ParserMatchClass = Imm0_31Operand;
}
+// True if the 32-bit immediate is in the range [0,31]
+def imm32_0_31 : Operand<i32>, ImmLeaf<i32, [{
+ return ((uint64_t)Imm) < 32;
+}]> {
+ let ParserMatchClass = Imm0_31Operand;
+}
+
// imm0_15 predicate - True if the immediate is in the range [0,15]
def imm0_15 : Operand<i64>, ImmLeaf<i64, [{
return ((uint64_t)Imm) < 16;
// imm32_0_15 predicate - True if the 32-bit immediate is in the range [0,15]
def imm32_0_15 : Operand<i32>, ImmLeaf<i32, [{
return ((uint32_t)Imm) < 16;
-}]>;
+}]> {
+ let ParserMatchClass = Imm0_15Operand;
+}
// An arithmetic shifter operand:
// {7-6} - shift type: 00 = lsl, 01 = lsr, 10 = asr
let ParserMatchClass = MoveVecShifterOperand;
}
-def AddSubImmOperand : AsmOperandClass {
- let Name = "AddSubImm";
- let ParserMethod = "tryParseAddSubImm";
- let DiagnosticType = "AddSubSecondSource";
+let DiagnosticType = "AddSubSecondSource" in {
+ def AddSubImmOperand : AsmOperandClass {
+ let Name = "AddSubImm";
+ let ParserMethod = "tryParseAddSubImm";
+ }
+ def AddSubImmNegOperand : AsmOperandClass {
+ let Name = "AddSubImmNeg";
+ let ParserMethod = "tryParseAddSubImm";
+ }
}
// An ADD/SUB immediate shifter operand:
// second operand:
let MIOperandInfo = (ops i32imm, i32imm);
}
+class addsub_shifted_imm_neg<ValueType Ty>
+ : Operand<Ty> {
+ let EncoderMethod = "getAddSubImmOpValue";
+ let ParserMatchClass = AddSubImmNegOperand;
+ let MIOperandInfo = (ops i32imm, i32imm);
+}
+
def addsub_shifted_imm32 : addsub_shifted_imm<i32>;
def addsub_shifted_imm64 : addsub_shifted_imm<i64>;
+def addsub_shifted_imm32_neg : addsub_shifted_imm_neg<i32>;
+def addsub_shifted_imm64_neg : addsub_shifted_imm_neg<i64>;
class neg_addsub_shifted_imm<ValueType Ty>
: Operand<Ty>, ComplexPattern<Ty, 2, "SelectNegArithImmed", [imm]> {
}], SDNodeXForm<fpimm, [{
APFloat InVal = N->getValueAPF();
uint32_t enc = AArch64_AM::getFP32Imm(InVal);
- return CurDAG->getTargetConstant(enc, MVT::i32);
+ return CurDAG->getTargetConstant(enc, SDLoc(N), MVT::i32);
}]>> {
let ParserMatchClass = FPImmOperand;
let PrintMethod = "printFPImmOperand";
}], SDNodeXForm<fpimm, [{
APFloat InVal = N->getValueAPF();
uint32_t enc = AArch64_AM::getFP64Imm(InVal);
- return CurDAG->getTargetConstant(enc, MVT::i32);
+ return CurDAG->getTargetConstant(enc, SDLoc(N), MVT::i32);
}]>> {
let ParserMatchClass = FPImmOperand;
let PrintMethod = "printFPImmOperand";
uint32_t enc = AArch64_AM::encodeAdvSIMDModImmType10(N->getValueAPF()
.bitcastToAPInt()
.getZExtValue());
- return CurDAG->getTargetConstant(enc, MVT::i32);
+ return CurDAG->getTargetConstant(enc, SDLoc(N), MVT::i32);
}]>> {
let ParserMatchClass = SIMDImmType10Operand;
let PrintMethod = "printSIMDType10Operand";
let Inst{7-5} = pstatefield{2-0};
let DecoderMethod = "DecodeSystemPStateInstruction";
+ // MSRpstateI aliases with MSRI. When the MSRpstateI decoder method returns
+ // Fail the decoder should attempt to decode the instruction as MSRI.
+ let hasCompleteDecoder = 0;
}
// SYS and SYSL generic system instructions.
(inst dstRegtype:$dst, src1Regtype:$src1, src2Regtype:$src2,
shiftExt)>;
-multiclass AddSub<bit isSub, string mnemonic,
+multiclass AddSub<bit isSub, string mnemonic, string alias,
SDPatternOperator OpNode = null_frag> {
let hasSideEffects = 0, isReMaterializable = 1, isAsCheapAsAMove = 1 in {
// Add/Subtract immediate
+ // Increase the weight of the immediate variant to try to match it before
+ // the extended register variant.
+ // We used to match the register variant before the immediate when the
+ // register argument could be implicitly zero-extended.
let AddedComplexity = 6 in
def Wri : BaseAddSubImm<isSub, 0, GPR32sp, GPR32sp, addsub_shifted_imm32,
mnemonic, OpNode> {
let Inst{31} = 1;
}
+ // add Rd, Rb, -imm -> sub Rd, Rn, imm
+ def : InstAlias<alias#" $Rd, $Rn, $imm",
+ (!cast<Instruction>(NAME # "Wri") GPR32sp:$Rd, GPR32sp:$Rn,
+ addsub_shifted_imm32_neg:$imm), 0>;
+ def : InstAlias<alias#" $Rd, $Rn, $imm",
+ (!cast<Instruction>(NAME # "Xri") GPR64sp:$Rd, GPR64sp:$Rn,
+ addsub_shifted_imm64_neg:$imm), 0>;
+
// Register/register aliases with no shift when SP is not used.
def : AddSubRegAlias<mnemonic, !cast<Instruction>(NAME#"Wrs"),
GPR32, GPR32, GPR32, 0>;
GPR64sp, GPR64sponly, GPR64, 24>; // UXTX #0
}
-multiclass AddSubS<bit isSub, string mnemonic, SDNode OpNode, string cmp> {
+multiclass AddSubS<bit isSub, string mnemonic, SDNode OpNode, string cmp,
+ string alias, string cmpAlias> {
let isCompare = 1, Defs = [NZCV] in {
// Add/Subtract immediate
def Wri : BaseAddSubImm<isSub, 1, GPR32, GPR32sp, addsub_shifted_imm32,
}
} // Defs = [NZCV]
+ // Support negative immediates, e.g. adds Rd, Rn, -imm -> subs Rd, Rn, imm
+ def : InstAlias<alias#" $Rd, $Rn, $imm",
+ (!cast<Instruction>(NAME # "Wri") GPR32:$Rd, GPR32sp:$Rn,
+ addsub_shifted_imm32_neg:$imm), 0>;
+ def : InstAlias<alias#" $Rd, $Rn, $imm",
+ (!cast<Instruction>(NAME # "Xri") GPR64:$Rd, GPR64sp:$Rn,
+ addsub_shifted_imm64_neg:$imm), 0>;
+
// Compare aliases
def : InstAlias<cmp#" $src, $imm", (!cast<Instruction>(NAME#"Wri")
WZR, GPR32sp:$src, addsub_shifted_imm32:$imm), 5>;
def : InstAlias<cmp#" $src1, $src2$sh", (!cast<Instruction>(NAME#"Xrs")
XZR, GPR64:$src1, GPR64:$src2, arith_shift64:$sh), 4>;
+ // Support negative immediates, e.g. cmp Rn, -imm -> cmn Rn, imm
+ def : InstAlias<cmpAlias#" $src, $imm", (!cast<Instruction>(NAME#"Wri")
+ WZR, GPR32sp:$src, addsub_shifted_imm32_neg:$imm), 0>;
+ def : InstAlias<cmpAlias#" $src, $imm", (!cast<Instruction>(NAME#"Xri")
+ XZR, GPR64sp:$src, addsub_shifted_imm64_neg:$imm), 0>;
+
// Compare shorthands
def : InstAlias<cmp#" $src1, $src2", (!cast<Instruction>(NAME#"Wrs")
WZR, GPR32:$src1, GPR32:$src2, 0), 5>;
//---
let mayLoad = 0, mayStore = 0, hasSideEffects = 0 in
-class BaseCondSetFlagsImm<bit op, RegisterClass regtype, string asm>
- : I<(outs), (ins regtype:$Rn, imm0_31:$imm, imm0_15:$nzcv, ccode:$cond),
- asm, "\t$Rn, $imm, $nzcv, $cond", "", []>,
+class BaseCondComparisonImm<bit op, RegisterClass regtype, ImmLeaf immtype,
+ string mnemonic, SDNode OpNode>
+ : I<(outs), (ins regtype:$Rn, immtype:$imm, imm32_0_15:$nzcv, ccode:$cond),
+ mnemonic, "\t$Rn, $imm, $nzcv, $cond", "",
+ [(set NZCV, (OpNode regtype:$Rn, immtype:$imm, (i32 imm:$nzcv),
+ (i32 imm:$cond), NZCV))]>,
Sched<[WriteI, ReadI]> {
let Uses = [NZCV];
let Defs = [NZCV];
let Inst{3-0} = nzcv;
}
-multiclass CondSetFlagsImm<bit op, string asm> {
- def Wi : BaseCondSetFlagsImm<op, GPR32, asm> {
- let Inst{31} = 0;
- }
- def Xi : BaseCondSetFlagsImm<op, GPR64, asm> {
- let Inst{31} = 1;
- }
-}
-
let mayLoad = 0, mayStore = 0, hasSideEffects = 0 in
-class BaseCondSetFlagsReg<bit op, RegisterClass regtype, string asm>
- : I<(outs), (ins regtype:$Rn, regtype:$Rm, imm0_15:$nzcv, ccode:$cond),
- asm, "\t$Rn, $Rm, $nzcv, $cond", "", []>,
+class BaseCondComparisonReg<bit op, RegisterClass regtype, string mnemonic,
+ SDNode OpNode>
+ : I<(outs), (ins regtype:$Rn, regtype:$Rm, imm32_0_15:$nzcv, ccode:$cond),
+ mnemonic, "\t$Rn, $Rm, $nzcv, $cond", "",
+ [(set NZCV, (OpNode regtype:$Rn, regtype:$Rm, (i32 imm:$nzcv),
+ (i32 imm:$cond), NZCV))]>,
Sched<[WriteI, ReadI, ReadI]> {
let Uses = [NZCV];
let Defs = [NZCV];
let Inst{3-0} = nzcv;
}
-multiclass CondSetFlagsReg<bit op, string asm> {
- def Wr : BaseCondSetFlagsReg<op, GPR32, asm> {
+multiclass CondComparison<bit op, string mnemonic, SDNode OpNode> {
+ // immediate operand variants
+ def Wi : BaseCondComparisonImm<op, GPR32, imm32_0_31, mnemonic, OpNode> {
+ let Inst{31} = 0;
+ }
+ def Xi : BaseCondComparisonImm<op, GPR64, imm0_31, mnemonic, OpNode> {
+ let Inst{31} = 1;
+ }
+ // register operand variants
+ def Wr : BaseCondComparisonReg<op, GPR32, mnemonic, OpNode> {
let Inst{31} = 0;
}
- def Xr : BaseCondSetFlagsReg<op, GPR64, asm> {
+ def Xr : BaseCondComparisonReg<op, GPR64, mnemonic, OpNode> {
let Inst{31} = 1;
}
}
def inv_cond_XFORM : SDNodeXForm<imm, [{
AArch64CC::CondCode CC = static_cast<AArch64CC::CondCode>(N->getZExtValue());
- return CurDAG->getTargetConstant(AArch64CC::getInvertedCondCode(CC), MVT::i32);
+ return CurDAG->getTargetConstant(AArch64CC::getInvertedCondCode(CC), SDLoc(N),
+ MVT::i32);
}]>;
multiclass CondSelectOp<bit op, bits<2> op2, string asm, PatFrag frag> {
: BaseLoadStoreExclusive<sz, o2, L, o1, o0, oops, iops, asm, operands> {
bits<5> Rt;
bits<5> Rn;
+ let Inst{20-16} = 0b11111;
+ let Unpredictable{20-16} = 0b11111;
+ let Inst{14-10} = 0b11111;
+ let Unpredictable{14-10} = 0b11111;
let Inst{9-5} = Rn;
let Inst{4-0} = Rt;
//---
let mayLoad = 0, mayStore = 0, hasSideEffects = 0 in
-class BaseFPCondComparison<bit signalAllNans,
- RegisterClass regtype, string asm>
- : I<(outs), (ins regtype:$Rn, regtype:$Rm, imm0_15:$nzcv, ccode:$cond),
- asm, "\t$Rn, $Rm, $nzcv, $cond", "", []>,
+class BaseFPCondComparison<bit signalAllNans, RegisterClass regtype,
+ string mnemonic, list<dag> pat>
+ : I<(outs), (ins regtype:$Rn, regtype:$Rm, imm32_0_15:$nzcv, ccode:$cond),
+ mnemonic, "\t$Rn, $Rm, $nzcv, $cond", "", pat>,
Sched<[WriteFCmp]> {
+ let Uses = [NZCV];
+ let Defs = [NZCV];
+
bits<5> Rn;
bits<5> Rm;
bits<4> nzcv;
let Inst{3-0} = nzcv;
}
-multiclass FPCondComparison<bit signalAllNans, string asm> {
- let Defs = [NZCV], Uses = [NZCV] in {
- def Srr : BaseFPCondComparison<signalAllNans, FPR32, asm> {
+multiclass FPCondComparison<bit signalAllNans, string mnemonic,
+ SDPatternOperator OpNode = null_frag> {
+ def Srr : BaseFPCondComparison<signalAllNans, FPR32, mnemonic,
+ [(set NZCV, (OpNode (f32 FPR32:$Rn), (f32 FPR32:$Rm), (i32 imm:$nzcv),
+ (i32 imm:$cond), NZCV))]> {
let Inst{22} = 0;
}
-
- def Drr : BaseFPCondComparison<signalAllNans, FPR64, asm> {
+ def Drr : BaseFPCondComparison<signalAllNans, FPR64, mnemonic,
+ [(set NZCV, (OpNode (f64 FPR64:$Rn), (f64 FPR64:$Rm), (i32 imm:$nzcv),
+ (i32 imm:$cond), NZCV))]> {
let Inst{22} = 1;
}
- } // Defs = [NZCV], Uses = [NZCV]
}
//---
let Inst{4-0} = Rd;
}
+let mayStore = 0, mayLoad = 0, hasSideEffects = 0 in
+class BaseSIMDThreeScalarTied<bit U, bits<2> size, bit R, bits<5> opcode,
+ dag oops, dag iops, string asm,
+ list<dag> pattern>
+ : I<oops, iops, asm, "\t$Rd, $Rn, $Rm", "$Rd = $dst", pattern>,
+ Sched<[WriteV]> {
+ bits<5> Rd;
+ bits<5> Rn;
+ bits<5> Rm;
+ let Inst{31-30} = 0b01;
+ let Inst{29} = U;
+ let Inst{28-24} = 0b11110;
+ let Inst{23-22} = size;
+ let Inst{21} = R;
+ let Inst{20-16} = Rm;
+ let Inst{15-11} = opcode;
+ let Inst{10} = 1;
+ let Inst{9-5} = Rn;
+ let Inst{4-0} = Rd;
+}
+
multiclass SIMDThreeScalarD<bit U, bits<5> opc, string asm,
SDPatternOperator OpNode> {
def v1i64 : BaseSIMDThreeScalar<U, 0b11, opc, FPR64, asm,
def v1i16 : BaseSIMDThreeScalar<U, 0b01, opc, FPR16, asm, []>;
}
+multiclass SIMDThreeScalarHSTied<bit U, bit R, bits<5> opc, string asm,
+ SDPatternOperator OpNode = null_frag> {
+ def v1i32: BaseSIMDThreeScalarTied<U, 0b10, R, opc, (outs FPR32:$dst),
+ (ins FPR32:$Rd, FPR32:$Rn, FPR32:$Rm),
+ asm, []>;
+ def v1i16: BaseSIMDThreeScalarTied<U, 0b01, R, opc, (outs FPR16:$dst),
+ (ins FPR16:$Rd, FPR16:$Rn, FPR16:$Rm),
+ asm, []>;
+}
+
multiclass SIMDThreeScalarSD<bit U, bit S, bits<5> opc, string asm,
SDPatternOperator OpNode = null_frag> {
let mayLoad = 0, mayStore = 0, hasSideEffects = 0 in {
(!cast<Instruction>(NAME # v1i64rz) FPR64:$Rn)>;
}
-multiclass SIMDCmpTwoScalarSD<bit U, bit S, bits<5> opc, string asm,
+multiclass SIMDFPCmpTwoScalar<bit U, bit S, bits<5> opc, string asm,
SDPatternOperator OpNode> {
def v1i64rz : BaseSIMDCmpTwoScalar<U, {S,1}, opc, FPR64, asm, "0.0">;
def v1i32rz : BaseSIMDCmpTwoScalar<U, {S,0}, opc, FPR32, asm, "0.0">;
(!cast<Instruction>(NAME # "v1i64") FPR64:$Rn)>;
}
-multiclass SIMDTwoScalarSD<bit U, bit S, bits<5> opc, string asm> {
+multiclass SIMDFPTwoScalar<bit U, bit S, bits<5> opc, string asm> {
def v1i64 : BaseSIMDTwoScalar<U, {S,1}, opc, FPR64, FPR64, asm,[]>;
def v1i32 : BaseSIMDTwoScalar<U, {S,0}, opc, FPR32, FPR32, asm,[]>;
}
asm, ".2d">;
}
-multiclass SIMDPairwiseScalarSD<bit U, bit S, bits<5> opc, string asm> {
+multiclass SIMDFPPairwiseScalar<bit U, bit S, bits<5> opc, string asm> {
def v2i32p : BaseSIMDPairwiseScalar<U, {S,0}, opc, FPR32Op, V64,
asm, ".2s">;
def v2i64p : BaseSIMDPairwiseScalar<U, {S,1}, opc, FPR64Op, V128,
let Inst{20-18} = idx;
let Inst{17-16} = 0b10;
let Inst{14-12} = idx2;
- let Inst{11} = 0;
+ let Inst{11} = {?};
}
def vi32lane : SIMDInsFromElement<".s", v4i32, i32, VectorIndexS> {
bits<2> idx;
let Inst{20-19} = idx;
let Inst{18-16} = 0b100;
let Inst{14-13} = idx2;
- let Inst{12-11} = 0;
+ let Inst{12-11} = {?,?};
}
def vi64lane : SIMDInsFromElement<".d", v2i64, i64, VectorIndexD> {
bits<1> idx;
let Inst{20} = idx;
let Inst{19-16} = 0b1000;
let Inst{14} = idx2;
- let Inst{13-11} = 0;
+ let Inst{13-11} = {?,?,?};
}
// For all forms of the INS instruction, the "mov" mnemonic is the
let Inst{4-0} = Rd;
}
-multiclass SIMDFPIndexedSD<bit U, bits<4> opc, string asm,
- SDPatternOperator OpNode> {
+multiclass SIMDFPIndexed<bit U, bits<4> opc, string asm,
+ SDPatternOperator OpNode> {
def v2i32_indexed : BaseSIMDIndexed<0, U, 0, 0b10, opc,
V64, V64,
V128, VectorIndexS,
}
}
-multiclass SIMDFPIndexedSDTiedPatterns<string INST, SDPatternOperator OpNode> {
+multiclass SIMDFPIndexedTiedPatterns<string INST, SDPatternOperator OpNode> {
// 2 variants for the .2s version: DUPLANE from 128-bit and DUP scalar.
def : Pat<(v2f32 (OpNode (v2f32 V64:$Rd), (v2f32 V64:$Rn),
(AArch64duplane32 (v4f32 V128:$Rm),
V128:$Rm, VectorIndexD:$idx)>;
}
-multiclass SIMDFPIndexedSDTied<bit U, bits<4> opc, string asm> {
+multiclass SIMDFPIndexedTied<bit U, bits<4> opc, string asm> {
def v2i32_indexed : BaseSIMDIndexedTied<0, U, 0, 0b10, opc, V64, V64,
V128, VectorIndexS,
asm, ".2s", ".2s", ".2s", ".s", []> {
}
} // end of 'let Predicates = [HasNEON]'
+//----------------------------------------------------------------------------
+// AdvSIMD v8.1 Rounding Double Multiply Add/Subtract
+//----------------------------------------------------------------------------
+
+let Predicates = [HasNEON, HasV8_1a] in {
+
+class BaseSIMDThreeSameVectorTiedR0<bit Q, bit U, bits<2> size, bits<5> opcode,
+ RegisterOperand regtype, string asm,
+ string kind, list<dag> pattern>
+ : BaseSIMDThreeSameVectorTied<Q, U, size, opcode, regtype, asm, kind,
+ pattern> {
+ let Inst{21}=0;
+}
+multiclass SIMDThreeSameVectorSQRDMLxHTiedHS<bit U, bits<5> opc, string asm,
+ SDPatternOperator Accum> {
+ def v4i16 : BaseSIMDThreeSameVectorTiedR0<0, U, 0b01, opc, V64, asm, ".4h",
+ [(set (v4i16 V64:$dst),
+ (Accum (v4i16 V64:$Rd),
+ (v4i16 (int_aarch64_neon_sqrdmulh (v4i16 V64:$Rn),
+ (v4i16 V64:$Rm)))))]>;
+ def v8i16 : BaseSIMDThreeSameVectorTiedR0<1, U, 0b01, opc, V128, asm, ".8h",
+ [(set (v8i16 V128:$dst),
+ (Accum (v8i16 V128:$Rd),
+ (v8i16 (int_aarch64_neon_sqrdmulh (v8i16 V128:$Rn),
+ (v8i16 V128:$Rm)))))]>;
+ def v2i32 : BaseSIMDThreeSameVectorTiedR0<0, U, 0b10, opc, V64, asm, ".2s",
+ [(set (v2i32 V64:$dst),
+ (Accum (v2i32 V64:$Rd),
+ (v2i32 (int_aarch64_neon_sqrdmulh (v2i32 V64:$Rn),
+ (v2i32 V64:$Rm)))))]>;
+ def v4i32 : BaseSIMDThreeSameVectorTiedR0<1, U, 0b10, opc, V128, asm, ".4s",
+ [(set (v4i32 V128:$dst),
+ (Accum (v4i32 V128:$Rd),
+ (v4i32 (int_aarch64_neon_sqrdmulh (v4i32 V128:$Rn),
+ (v4i32 V128:$Rm)))))]>;
+}
+
+multiclass SIMDIndexedSQRDMLxHSDTied<bit U, bits<4> opc, string asm,
+ SDPatternOperator Accum> {
+ def v4i16_indexed : BaseSIMDIndexedTied<0, U, 0, 0b01, opc,
+ V64, V64, V128_lo, VectorIndexH,
+ asm, ".4h", ".4h", ".4h", ".h",
+ [(set (v4i16 V64:$dst),
+ (Accum (v4i16 V64:$Rd),
+ (v4i16 (int_aarch64_neon_sqrdmulh
+ (v4i16 V64:$Rn),
+ (v4i16 (AArch64duplane16 (v8i16 V128_lo:$Rm),
+ VectorIndexH:$idx))))))]> {
+ bits<3> idx;
+ let Inst{11} = idx{2};
+ let Inst{21} = idx{1};
+ let Inst{20} = idx{0};
+ }
+
+ def v8i16_indexed : BaseSIMDIndexedTied<1, U, 0, 0b01, opc,
+ V128, V128, V128_lo, VectorIndexH,
+ asm, ".8h", ".8h", ".8h", ".h",
+ [(set (v8i16 V128:$dst),
+ (Accum (v8i16 V128:$Rd),
+ (v8i16 (int_aarch64_neon_sqrdmulh
+ (v8i16 V128:$Rn),
+ (v8i16 (AArch64duplane16 (v8i16 V128_lo:$Rm),
+ VectorIndexH:$idx))))))]> {
+ bits<3> idx;
+ let Inst{11} = idx{2};
+ let Inst{21} = idx{1};
+ let Inst{20} = idx{0};
+ }
+
+ def v2i32_indexed : BaseSIMDIndexedTied<0, U, 0, 0b10, opc,
+ V64, V64, V128, VectorIndexS,
+ asm, ".2s", ".2s", ".2s", ".s",
+ [(set (v2i32 V64:$dst),
+ (Accum (v2i32 V64:$Rd),
+ (v2i32 (int_aarch64_neon_sqrdmulh
+ (v2i32 V64:$Rn),
+ (v2i32 (AArch64duplane32 (v4i32 V128:$Rm),
+ VectorIndexS:$idx))))))]> {
+ bits<2> idx;
+ let Inst{11} = idx{1};
+ let Inst{21} = idx{0};
+ }
+
+ // FIXME: it would be nice to use the scalar (v1i32) instruction here, but
+ // an intermediate EXTRACT_SUBREG would be untyped.
+ // FIXME: direct EXTRACT_SUBREG from v2i32 to i32 is illegal, that's why we
+ // got it lowered here as (i32 vector_extract (v4i32 insert_subvector(..)))
+ def : Pat<(i32 (Accum (i32 FPR32Op:$Rd),
+ (i32 (vector_extract
+ (v4i32 (insert_subvector
+ (undef),
+ (v2i32 (int_aarch64_neon_sqrdmulh
+ (v2i32 V64:$Rn),
+ (v2i32 (AArch64duplane32
+ (v4i32 V128:$Rm),
+ VectorIndexS:$idx)))),
+ (i32 0))),
+ (i64 0))))),
+ (EXTRACT_SUBREG
+ (v2i32 (!cast<Instruction>(NAME # v2i32_indexed)
+ (v2i32 (INSERT_SUBREG (v2i32 (IMPLICIT_DEF)),
+ FPR32Op:$Rd,
+ ssub)),
+ V64:$Rn,
+ V128:$Rm,
+ VectorIndexS:$idx)),
+ ssub)>;
+
+ def v4i32_indexed : BaseSIMDIndexedTied<1, U, 0, 0b10, opc,
+ V128, V128, V128, VectorIndexS,
+ asm, ".4s", ".4s", ".4s", ".s",
+ [(set (v4i32 V128:$dst),
+ (Accum (v4i32 V128:$Rd),
+ (v4i32 (int_aarch64_neon_sqrdmulh
+ (v4i32 V128:$Rn),
+ (v4i32 (AArch64duplane32 (v4i32 V128:$Rm),
+ VectorIndexS:$idx))))))]> {
+ bits<2> idx;
+ let Inst{11} = idx{1};
+ let Inst{21} = idx{0};
+ }
+
+ // FIXME: it would be nice to use the scalar (v1i32) instruction here, but
+ // an intermediate EXTRACT_SUBREG would be untyped.
+ def : Pat<(i32 (Accum (i32 FPR32Op:$Rd),
+ (i32 (vector_extract
+ (v4i32 (int_aarch64_neon_sqrdmulh
+ (v4i32 V128:$Rn),
+ (v4i32 (AArch64duplane32
+ (v4i32 V128:$Rm),
+ VectorIndexS:$idx)))),
+ (i64 0))))),
+ (EXTRACT_SUBREG
+ (v4i32 (!cast<Instruction>(NAME # v4i32_indexed)
+ (v4i32 (INSERT_SUBREG (v4i32 (IMPLICIT_DEF)),
+ FPR32Op:$Rd,
+ ssub)),
+ V128:$Rn,
+ V128:$Rm,
+ VectorIndexS:$idx)),
+ ssub)>;
+
+ def i16_indexed : BaseSIMDIndexedTied<1, U, 1, 0b01, opc,
+ FPR16Op, FPR16Op, V128_lo,
+ VectorIndexH, asm, ".h", "", "", ".h",
+ []> {
+ bits<3> idx;
+ let Inst{11} = idx{2};
+ let Inst{21} = idx{1};
+ let Inst{20} = idx{0};
+ }
+
+ def i32_indexed : BaseSIMDIndexedTied<1, U, 1, 0b10, opc,
+ FPR32Op, FPR32Op, V128, VectorIndexS,
+ asm, ".s", "", "", ".s",
+ [(set (i32 FPR32Op:$dst),
+ (Accum (i32 FPR32Op:$Rd),
+ (i32 (int_aarch64_neon_sqrdmulh
+ (i32 FPR32Op:$Rn),
+ (i32 (vector_extract (v4i32 V128:$Rm),
+ VectorIndexS:$idx))))))]> {
+ bits<2> idx;
+ let Inst{11} = idx{1};
+ let Inst{21} = idx{0};
+ }
+}
+} // let Predicates = [HasNeon, HasV8_1a]
+
//----------------------------------------------------------------------------
// Crypto extensions
//----------------------------------------------------------------------------
[(set (i32 FPR32:$Rd), (OpNode (i32 FPR32:$Rn)))]>;
} // end of 'let Predicates = [HasCrypto]'
+//----------------------------------------------------------------------------
+// v8.1 atomic instructions extension:
+// * CAS
+// * CASP
+// * SWP
+// * LDOPregister<OP>, and aliases STOPregister<OP>
+
+// Instruction encodings:
+//
+// 31 30|29 24|23|22|21|20 16|15|14 10|9 5|4 0
+// CAS SZ |001000|1 |A |1 |Rs |R |11111 |Rn |Rt
+// CASP 0|SZ|001000|0 |A |1 |Rs |R |11111 |Rn |Rt
+// SWP SZ |111000|A |R |1 |Rs |1 |OPC|00|Rn |Rt
+// LD SZ |111000|A |R |1 |Rs |0 |OPC|00|Rn |Rt
+// ST SZ |111000|A |R |1 |Rs |0 |OPC|00|Rn |11111
+
+// Instruction syntax:
+//
+// CAS{<order>}[<size>] <Ws>, <Wt>, [<Xn|SP>]
+// CAS{<order>} <Xs>, <Xt>, [<Xn|SP>]
+// CASP{<order>} <Ws>, <W(s+1)>, <Wt>, <W(t+1)>, [<Xn|SP>]
+// CASP{<order>} <Xs>, <X(s+1)>, <Xt>, <X(t+1)>, [<Xn|SP>]
+// SWP{<order>}[<size>] <Ws>, <Wt>, [<Xn|SP>]
+// SWP{<order>} <Xs>, <Xt>, [<Xn|SP>]
+// LD<OP>{<order>}[<size>] <Ws>, <Wt>, [<Xn|SP>]
+// LD<OP>{<order>} <Xs>, <Xt>, [<Xn|SP>]
+// ST<OP>{<order>}[<size>] <Ws>, [<Xn|SP>]
+// ST<OP>{<order>} <Xs>, [<Xn|SP>]
+
+let Predicates = [HasV8_1a], mayLoad = 1, mayStore = 1, hasSideEffects = 1 in
+class BaseCASEncoding<dag oops, dag iops, string asm, string operands,
+ string cstr, list<dag> pattern>
+ : I<oops, iops, asm, operands, cstr, pattern> {
+ bits<2> Sz;
+ bit NP;
+ bit Acq;
+ bit Rel;
+ bits<5> Rs;
+ bits<5> Rn;
+ bits<5> Rt;
+ let Inst{31-30} = Sz;
+ let Inst{29-24} = 0b001000;
+ let Inst{23} = NP;
+ let Inst{22} = Acq;
+ let Inst{21} = 0b1;
+ let Inst{20-16} = Rs;
+ let Inst{15} = Rel;
+ let Inst{14-10} = 0b11111;
+ let Inst{9-5} = Rn;
+ let Inst{4-0} = Rt;
+}
+
+class BaseCAS<string order, string size, RegisterClass RC>
+ : BaseCASEncoding<(outs RC:$out),(ins RC:$Rs, RC:$Rt, GPR64sp:$Rn),
+ "cas" # order # size, "\t$Rs, $Rt, [$Rn]",
+ "$out = $Rs",[]> {
+ let NP = 1;
+}
+
+multiclass CompareAndSwap<bits<1> Acq, bits<1> Rel, string order> {
+ let Sz = 0b00, Acq = Acq, Rel = Rel in def b : BaseCAS<order, "b", GPR32>;
+ let Sz = 0b01, Acq = Acq, Rel = Rel in def h : BaseCAS<order, "h", GPR32>;
+ let Sz = 0b10, Acq = Acq, Rel = Rel in def s : BaseCAS<order, "", GPR32>;
+ let Sz = 0b11, Acq = Acq, Rel = Rel in def d : BaseCAS<order, "", GPR64>;
+}
+
+class BaseCASP<string order, string size, RegisterOperand RC>
+ : BaseCASEncoding<(outs RC:$out),(ins RC:$Rs, RC:$Rt, GPR64sp:$Rn),
+ "casp" # order # size, "\t$Rs, $Rt, [$Rn]",
+ "$out = $Rs",[]> {
+ let NP = 0;
+}
+
+multiclass CompareAndSwapPair<bits<1> Acq, bits<1> Rel, string order> {
+ let Sz = 0b00, Acq = Acq, Rel = Rel in
+ def s : BaseCASP<order, "", WSeqPairClassOperand>;
+ let Sz = 0b01, Acq = Acq, Rel = Rel in
+ def d : BaseCASP<order, "", XSeqPairClassOperand>;
+}
+
+let Predicates = [HasV8_1a] in
+class BaseSWP<string order, string size, RegisterClass RC>
+ : I<(outs RC:$Rt),(ins RC:$Rs, GPR64sp:$Rn), "swp" # order # size,
+ "\t$Rs, $Rt, [$Rn]","",[]> {
+ bits<2> Sz;
+ bit Acq;
+ bit Rel;
+ bits<5> Rs;
+ bits<3> opc = 0b000;
+ bits<5> Rn;
+ bits<5> Rt;
+ let Inst{31-30} = Sz;
+ let Inst{29-24} = 0b111000;
+ let Inst{23} = Acq;
+ let Inst{22} = Rel;
+ let Inst{21} = 0b1;
+ let Inst{20-16} = Rs;
+ let Inst{15} = 0b1;
+ let Inst{14-12} = opc;
+ let Inst{11-10} = 0b00;
+ let Inst{9-5} = Rn;
+ let Inst{4-0} = Rt;
+}
+
+multiclass Swap<bits<1> Acq, bits<1> Rel, string order> {
+ let Sz = 0b00, Acq = Acq, Rel = Rel in def b : BaseSWP<order, "b", GPR32>;
+ let Sz = 0b01, Acq = Acq, Rel = Rel in def h : BaseSWP<order, "h", GPR32>;
+ let Sz = 0b10, Acq = Acq, Rel = Rel in def s : BaseSWP<order, "", GPR32>;
+ let Sz = 0b11, Acq = Acq, Rel = Rel in def d : BaseSWP<order, "", GPR64>;
+}
+
+let Predicates = [HasV8_1a], mayLoad = 1, mayStore = 1, hasSideEffects = 1 in
+class BaseLDOPregister<string op, string order, string size, RegisterClass RC>
+ : I<(outs RC:$Rt),(ins RC:$Rs, GPR64sp:$Rn), "ld" # op # order # size,
+ "\t$Rs, $Rt, [$Rn]","",[]> {
+ bits<2> Sz;
+ bit Acq;
+ bit Rel;
+ bits<5> Rs;
+ bits<3> opc;
+ bits<5> Rn;
+ bits<5> Rt;
+ let Inst{31-30} = Sz;
+ let Inst{29-24} = 0b111000;
+ let Inst{23} = Acq;
+ let Inst{22} = Rel;
+ let Inst{21} = 0b1;
+ let Inst{20-16} = Rs;
+ let Inst{15} = 0b0;
+ let Inst{14-12} = opc;
+ let Inst{11-10} = 0b00;
+ let Inst{9-5} = Rn;
+ let Inst{4-0} = Rt;
+}
+
+multiclass LDOPregister<bits<3> opc, string op, bits<1> Acq, bits<1> Rel,
+ string order> {
+ let Sz = 0b00, Acq = Acq, Rel = Rel, opc = opc in
+ def b : BaseLDOPregister<op, order, "b", GPR32>;
+ let Sz = 0b01, Acq = Acq, Rel = Rel, opc = opc in
+ def h : BaseLDOPregister<op, order, "h", GPR32>;
+ let Sz = 0b10, Acq = Acq, Rel = Rel, opc = opc in
+ def s : BaseLDOPregister<op, order, "", GPR32>;
+ let Sz = 0b11, Acq = Acq, Rel = Rel, opc = opc in
+ def d : BaseLDOPregister<op, order, "", GPR64>;
+}
+
+let Predicates = [HasV8_1a] in
+class BaseSTOPregister<string asm, RegisterClass OP, Register Reg,
+ Instruction inst> :
+ InstAlias<asm # "\t$Rs, [$Rn]", (inst Reg, OP:$Rs, GPR64sp:$Rn)>;
+
+multiclass STOPregister<string asm, string instr> {
+ def : BaseSTOPregister<asm # "lb", GPR32, WZR,
+ !cast<Instruction>(instr # "Lb")>;
+ def : BaseSTOPregister<asm # "lh", GPR32, WZR,
+ !cast<Instruction>(instr # "Lh")>;
+ def : BaseSTOPregister<asm # "l", GPR32, WZR,
+ !cast<Instruction>(instr # "Ls")>;
+ def : BaseSTOPregister<asm # "l", GPR64, XZR,
+ !cast<Instruction>(instr # "Ld")>;
+ def : BaseSTOPregister<asm # "b", GPR32, WZR,
+ !cast<Instruction>(instr # "b")>;
+ def : BaseSTOPregister<asm # "h", GPR32, WZR,
+ !cast<Instruction>(instr # "h")>;
+ def : BaseSTOPregister<asm, GPR32, WZR,
+ !cast<Instruction>(instr # "s")>;
+ def : BaseSTOPregister<asm, GPR64, XZR,
+ !cast<Instruction>(instr # "d")>;
+}
+
+//----------------------------------------------------------------------------
// Allow the size specifier tokens to be upper case, not just lower.
def : TokenAlias<".8B", ".8b">;
def : TokenAlias<".4H", ".4h">;
projects / oota-llvm.git / blobdiff