// NEON load / store instructions
//===----------------------------------------------------------------------===//
-let mayLoad = 1, neverHasSideEffects = 1 in {
// Use vldmia to load a Q register as a D register pair.
// This is equivalent to VLDMD except that it has a Q register operand
// instead of a pair of D registers.
def VLDMQ
- : AXDI5<(outs QPR:$dst), (ins addrmode5:$addr, pred:$p),
+ : AXDI5<(outs QPR:$dst), (ins addrmode4:$addr, pred:$p),
IndexModeNone, IIC_fpLoadm,
- "vldm${addr:submode}${p}\t${addr:base}, ${dst:dregpair}", "", []>;
+ "vldm${addr:submode}${p}\t$addr, ${dst:dregpair}", "",
+ [(set QPR:$dst, (v2f64 (load addrmode4:$addr)))]>;
+let mayLoad = 1, neverHasSideEffects = 1 in {
// Use vld1 to load a Q register as a D register pair.
// This alternative to VLDMQ allows an alignment to be specified.
// This is equivalent to VLD1q64 except that it has a Q register operand.
IIC_VLD1, "vld1", "64", "${dst:dregpair}, $addr", "", []>;
} // mayLoad = 1, neverHasSideEffects = 1
-let mayStore = 1, neverHasSideEffects = 1 in {
// Use vstmia to store a Q register as a D register pair.
// This is equivalent to VSTMD except that it has a Q register operand
// instead of a pair of D registers.
def VSTMQ
- : AXDI5<(outs), (ins QPR:$src, addrmode5:$addr, pred:$p),
+ : AXDI5<(outs), (ins QPR:$src, addrmode4:$addr, pred:$p),
IndexModeNone, IIC_fpStorem,
- "vstm${addr:submode}${p}\t${addr:base}, ${src:dregpair}", "", []>;
+ "vstm${addr:submode}${p}\t$addr, ${src:dregpair}", "",
+ [(store (v2f64 QPR:$src), addrmode4:$addr)]>;
+let mayStore = 1, neverHasSideEffects = 1 in {
// Use vst1 to store a Q register as a D register pair.
// This alternative to VSTMQ allows an alignment to be specified.
// This is equivalent to VST1q64 except that it has a Q register operand.
// Classes for VST* pseudo-instructions with multi-register operands.
// These are expanded to real instructions after register allocation.
+class VSTQPseudo
+ : PseudoNLdSt<(outs), (ins addrmode6:$addr, QPR:$src), IIC_VST, "">;
+class VSTQWBPseudo
+ : PseudoNLdSt<(outs GPR:$wb),
+ (ins addrmode6:$addr, am6offset:$offset, QPR:$src), IIC_VST,
+ "$addr.addr = $wb">;
class VSTQQPseudo
: PseudoNLdSt<(outs), (ins addrmode6:$addr, QQPR:$src), IIC_VST, "">;
class VSTQQWBPseudo
def VST1q32 : VST1Q<0b1000, "32">;
def VST1q64 : VST1Q<0b1100, "64">;
+def VST1q8Pseudo : VSTQPseudo;
+def VST1q16Pseudo : VSTQPseudo;
+def VST1q32Pseudo : VSTQPseudo;
+def VST1q64Pseudo : VSTQPseudo;
+
// ...with address register writeback:
class VST1DWB<bits<4> op7_4, string Dt>
: NLdSt<0, 0b00, 0b0111, op7_4, (outs GPR:$wb),
def VST1q32_UPD : VST1QWB<0b1000, "32">;
def VST1q64_UPD : VST1QWB<0b1100, "64">;
+def VST1q8Pseudo_UPD : VSTQWBPseudo;
+def VST1q16Pseudo_UPD : VSTQWBPseudo;
+def VST1q32Pseudo_UPD : VSTQWBPseudo;
+def VST1q64Pseudo_UPD : VSTQWBPseudo;
+
// ...with 3 registers (some of these are only for the disassembler):
class VST1D3<bits<4> op7_4, string Dt>
: NLdSt<0, 0b00, 0b0110, op7_4, (outs),
def VST2q16 : VST2Q<0b0100, "16">;
def VST2q32 : VST2Q<0b1000, "32">;
+def VST2d8Pseudo : VSTQPseudo;
+def VST2d16Pseudo : VSTQPseudo;
+def VST2d32Pseudo : VSTQPseudo;
+
+def VST2q8Pseudo : VSTQQPseudo;
+def VST2q16Pseudo : VSTQQPseudo;
+def VST2q32Pseudo : VSTQQPseudo;
+
// ...with address register writeback:
class VST2DWB<bits<4> op11_8, bits<4> op7_4, string Dt>
: NLdSt<0, 0b00, op11_8, op7_4, (outs GPR:$wb),
def VST2q16_UPD : VST2QWB<0b0100, "16">;
def VST2q32_UPD : VST2QWB<0b1000, "32">;
+def VST2d8Pseudo_UPD : VSTQWBPseudo;
+def VST2d16Pseudo_UPD : VSTQWBPseudo;
+def VST2d32Pseudo_UPD : VSTQWBPseudo;
+
+def VST2q8Pseudo_UPD : VSTQQWBPseudo;
+def VST2q16Pseudo_UPD : VSTQQWBPseudo;
+def VST2q32Pseudo_UPD : VSTQQWBPseudo;
+
// ...with double-spaced registers (for disassembly only):
def VST2b8 : VST2D<0b1001, 0b0000, "8">;
def VST2b16 : VST2D<0b1001, 0b0100, "16">;
(ins QPR:$src), itin, OpcodeStr, Dt, "$dst, $src", "",
[(set QPR:$dst, (ResTy (IntOp (OpTy QPR:$src))))]>;
+// Narrow 2-register operations.
+class N2VN<bits<2> op24_23, bits<2> op21_20, bits<2> op19_18,
+ bits<2> op17_16, bits<5> op11_7, bit op6, bit op4,
+ InstrItinClass itin, string OpcodeStr, string Dt,
+ ValueType TyD, ValueType TyQ, SDNode OpNode>
+ : N2V<op24_23, op21_20, op19_18, op17_16, op11_7, op6, op4, (outs DPR:$dst),
+ (ins QPR:$src), itin, OpcodeStr, Dt, "$dst, $src", "",
+ [(set DPR:$dst, (TyD (OpNode (TyQ QPR:$src))))]>;
+
// Narrow 2-register intrinsics.
class N2VNInt<bits<2> op24_23, bits<2> op21_20, bits<2> op19_18,
bits<2> op17_16, bits<5> op11_7, bit op6, bit op4,
let isCommutable = Commutable;
}
+// Long 3-register operations.
+class N3VL<bit op24, bit op23, bits<2> op21_20, bits<4> op11_8, bit op4,
+ InstrItinClass itin, string OpcodeStr, string Dt,
+ ValueType TyQ, ValueType TyD, SDNode OpNode, SDNode ExtOp,
+ bit Commutable>
+ : N3V<op24, op23, op21_20, op11_8, 0, op4,
+ (outs QPR:$dst), (ins DPR:$src1, DPR:$src2), N3RegFrm, itin,
+ OpcodeStr, Dt, "$dst, $src1, $src2", "",
+ [(set QPR:$dst, (OpNode (TyQ (ExtOp (TyD DPR:$src1))),
+ (TyQ (ExtOp (TyD DPR:$src2)))))]> {
+ let isCommutable = Commutable;
+}
+
// Long 3-register intrinsics.
class N3VLInt<bit op24, bit op23, bits<2> op21_20, bits<4> op11_8, bit op4,
InstrItinClass itin, string OpcodeStr, string Dt,
(OpTy (NEONvduplane (OpTy DPR_8:$src2),
imm:$lane)))))]>;
-// Wide 3-register intrinsics.
-class N3VWInt<bit op24, bit op23, bits<2> op21_20, bits<4> op11_8, bit op4,
- string OpcodeStr, string Dt, ValueType TyQ, ValueType TyD,
- Intrinsic IntOp, bit Commutable>
+// Wide 3-register operations.
+class N3VW<bit op24, bit op23, bits<2> op21_20, bits<4> op11_8, bit op4,
+ string OpcodeStr, string Dt, ValueType TyQ, ValueType TyD,
+ SDNode OpNode, SDNode ExtOp, bit Commutable>
: N3V<op24, op23, op21_20, op11_8, 0, op4,
(outs QPR:$dst), (ins QPR:$src1, DPR:$src2), N3RegFrm, IIC_VSUBiD,
OpcodeStr, Dt, "$dst, $src1, $src2", "",
- [(set QPR:$dst, (TyQ (IntOp (TyQ QPR:$src1), (TyD DPR:$src2))))]> {
+ [(set QPR:$dst, (OpNode (TyQ QPR:$src1),
+ (TyQ (ExtOp (TyD DPR:$src2)))))]> {
let isCommutable = Commutable;
}
}
+// Neon Narrowing 2-register vector operations,
+// source operand element sizes of 16, 32 and 64 bits:
+multiclass N2VN_HSD<bits<2> op24_23, bits<2> op21_20, bits<2> op17_16,
+ bits<5> op11_7, bit op6, bit op4,
+ InstrItinClass itin, string OpcodeStr, string Dt,
+ SDNode OpNode> {
+ def v8i8 : N2VN<op24_23, op21_20, 0b00, op17_16, op11_7, op6, op4,
+ itin, OpcodeStr, !strconcat(Dt, "16"),
+ v8i8, v8i16, OpNode>;
+ def v4i16 : N2VN<op24_23, op21_20, 0b01, op17_16, op11_7, op6, op4,
+ itin, OpcodeStr, !strconcat(Dt, "32"),
+ v4i16, v4i32, OpNode>;
+ def v2i32 : N2VN<op24_23, op21_20, 0b10, op17_16, op11_7, op6, op4,
+ itin, OpcodeStr, !strconcat(Dt, "64"),
+ v2i32, v2i64, OpNode>;
+}
+
// Neon Narrowing 2-register vector intrinsics,
// source operand element sizes of 16, 32 and 64 bits:
multiclass N2VNInt_HSD<bits<2> op24_23, bits<2> op21_20, bits<2> op17_16,
}
+// Neon Long 3-register vector operations.
+
+multiclass N3VL_QHS<bit op24, bit op23, bits<4> op11_8, bit op4,
+ InstrItinClass itin16, InstrItinClass itin32,
+ string OpcodeStr, string Dt,
+ SDNode OpNode, SDNode ExtOp, bit Commutable = 0> {
+ def v4i32 : N3VL<op24, op23, 0b01, op11_8, op4, itin16,
+ OpcodeStr, !strconcat(Dt, "16"),
+ v4i32, v4i16, OpNode, ExtOp, Commutable>;
+ def v2i64 : N3VL<op24, op23, 0b10, op11_8, op4, itin32,
+ OpcodeStr, !strconcat(Dt, "32"),
+ v2i64, v2i32, OpNode, ExtOp, Commutable>;
+ def v8i16 : N3VL<op24, op23, 0b00, op11_8, op4, itin16,
+ OpcodeStr, !strconcat(Dt, "8"),
+ v8i16, v8i8, OpNode, ExtOp, Commutable>;
+}
+
// Neon Long 3-register vector intrinsics.
// First with only element sizes of 16 and 32 bits:
// Neon Wide 3-register vector intrinsics,
// source operand element sizes of 8, 16 and 32 bits:
-multiclass N3VWInt_QHS<bit op24, bit op23, bits<4> op11_8, bit op4,
- string OpcodeStr, string Dt,
- Intrinsic IntOp, bit Commutable = 0> {
- def v8i16 : N3VWInt<op24, op23, 0b00, op11_8, op4,
- OpcodeStr, !strconcat(Dt, "8"),
- v8i16, v8i8, IntOp, Commutable>;
- def v4i32 : N3VWInt<op24, op23, 0b01, op11_8, op4,
- OpcodeStr, !strconcat(Dt, "16"),
- v4i32, v4i16, IntOp, Commutable>;
- def v2i64 : N3VWInt<op24, op23, 0b10, op11_8, op4,
- OpcodeStr, !strconcat(Dt, "32"),
- v2i64, v2i32, IntOp, Commutable>;
+multiclass N3VW_QHS<bit op24, bit op23, bits<4> op11_8, bit op4,
+ string OpcodeStr, string Dt,
+ SDNode OpNode, SDNode ExtOp, bit Commutable = 0> {
+ def v8i16 : N3VW<op24, op23, 0b00, op11_8, op4,
+ OpcodeStr, !strconcat(Dt, "8"),
+ v8i16, v8i8, OpNode, ExtOp, Commutable>;
+ def v4i32 : N3VW<op24, op23, 0b01, op11_8, op4,
+ OpcodeStr, !strconcat(Dt, "16"),
+ v4i32, v4i16, OpNode, ExtOp, Commutable>;
+ def v2i64 : N3VW<op24, op23, 0b10, op11_8, op4,
+ OpcodeStr, !strconcat(Dt, "32"),
+ v2i64, v2i32, OpNode, ExtOp, Commutable>;
}
def VADDfq : N3VQ<0, 0, 0b00, 0b1101, 0, IIC_VBINQ, "vadd", "f32",
v4f32, v4f32, fadd, 1>;
// VADDL : Vector Add Long (Q = D + D)
-defm VADDLs : N3VLInt_QHS<0,1,0b0000,0, IIC_VSHLiD, IIC_VSHLiD,
- "vaddl", "s", int_arm_neon_vaddls, 1>;
-defm VADDLu : N3VLInt_QHS<1,1,0b0000,0, IIC_VSHLiD, IIC_VSHLiD,
- "vaddl", "u", int_arm_neon_vaddlu, 1>;
+defm VADDLs : N3VL_QHS<0,1,0b0000,0, IIC_VSHLiD, IIC_VSHLiD,
+ "vaddl", "s", add, sext, 1>;
+defm VADDLu : N3VL_QHS<1,1,0b0000,0, IIC_VSHLiD, IIC_VSHLiD,
+ "vaddl", "u", add, zext, 1>;
// VADDW : Vector Add Wide (Q = Q + D)
-defm VADDWs : N3VWInt_QHS<0,1,0b0001,0, "vaddw", "s", int_arm_neon_vaddws, 0>;
-defm VADDWu : N3VWInt_QHS<1,1,0b0001,0, "vaddw", "u", int_arm_neon_vaddwu, 0>;
+defm VADDWs : N3VW_QHS<0,1,0b0001,0, "vaddw", "s", add, sext, 0>;
+defm VADDWu : N3VW_QHS<1,1,0b0001,0, "vaddw", "u", add, zext, 0>;
// VHADD : Vector Halving Add
defm VHADDs : N3VInt_QHS<0, 0, 0b0000, 0, N3RegFrm,
IIC_VBINi4D, IIC_VBINi4D, IIC_VBINi4Q, IIC_VBINi4Q,
def VSUBfq : N3VQ<0, 0, 0b10, 0b1101, 0, IIC_VBINQ, "vsub", "f32",
v4f32, v4f32, fsub, 0>;
// VSUBL : Vector Subtract Long (Q = D - D)
-defm VSUBLs : N3VLInt_QHS<0,1,0b0010,0, IIC_VSHLiD, IIC_VSHLiD,
- "vsubl", "s", int_arm_neon_vsubls, 1>;
-defm VSUBLu : N3VLInt_QHS<1,1,0b0010,0, IIC_VSHLiD, IIC_VSHLiD,
- "vsubl", "u", int_arm_neon_vsublu, 1>;
+defm VSUBLs : N3VL_QHS<0,1,0b0010,0, IIC_VSHLiD, IIC_VSHLiD,
+ "vsubl", "s", sub, sext, 0>;
+defm VSUBLu : N3VL_QHS<1,1,0b0010,0, IIC_VSHLiD, IIC_VSHLiD,
+ "vsubl", "u", sub, zext, 0>;
// VSUBW : Vector Subtract Wide (Q = Q - D)
-defm VSUBWs : N3VWInt_QHS<0,1,0b0011,0, "vsubw", "s", int_arm_neon_vsubws, 0>;
-defm VSUBWu : N3VWInt_QHS<1,1,0b0011,0, "vsubw", "u", int_arm_neon_vsubwu, 0>;
+defm VSUBWs : N3VW_QHS<0,1,0b0011,0, "vsubw", "s", sub, sext, 0>;
+defm VSUBWu : N3VW_QHS<1,1,0b0011,0, "vsubw", "u", sub, zext, 0>;
// VHSUB : Vector Halving Subtract
defm VHSUBs : N3VInt_QHS<0, 0, 0b0010, 0, N3RegFrm,
IIC_VSUBi4D, IIC_VSUBi4D, IIC_VSUBi4Q, IIC_VSUBi4Q,
defm VABDLs : N3VLInt_QHS<0,1,0b0111,0, IIC_VSUBi4Q, IIC_VSUBi4Q,
"vabdl", "s", int_arm_neon_vabdls, 0>;
defm VABDLu : N3VLInt_QHS<1,1,0b0111,0, IIC_VSUBi4Q, IIC_VSUBi4Q,
- "vabdl", "u", int_arm_neon_vabdlu, 0>;
+ "vabdl", "u", int_arm_neon_vabdlu, 0>;
// VABA : Vector Absolute Difference and Accumulate
defm VABAs : N3VInt3_QHS<0,0,0b0111,1, IIC_VABAD, IIC_VABAQ,
[(set QPR:$dst, (v4f32 (NEONvdup (f32 SPR:$src))))]>;
// VMOVN : Vector Narrowing Move
-defm VMOVN : N2VNInt_HSD<0b11,0b11,0b10,0b00100,0,0, IIC_VMOVD,
- "vmovn", "i", int_arm_neon_vmovn>;
+defm VMOVN : N2VN_HSD<0b11,0b11,0b10,0b00100,0,0, IIC_VMOVD,
+ "vmovn", "i", trunc>;
// VQMOVN : Vector Saturating Narrowing Move
defm VQMOVNs : N2VNInt_HSD<0b11,0b11,0b10,0b00101,0,0, IIC_VQUNAiD,
"vqmovn", "s", int_arm_neon_vqmovns>;
projects / oota-llvm.git / blobdiff