X-Git-Url: http://demsky.eecs.uci.edu/git/?a=blobdiff_plain;f=lib%2FTarget%2FX86%2FX86InstrFPStack.td;h=2224a08d59f46b0261e8cf6718b4295f6a2e17da;hb=e0364b64d12330f6f8c47ef98fc658468e2b72e4;hp=57634d3c0fb1694fda011b5f7777d0575ca79ae3;hpb=0b9325c97d031ab0e9a240d69a2be11ec1559e37;p=oota-llvm.git diff --git a/lib/Target/X86/X86InstrFPStack.td b/lib/Target/X86/X86InstrFPStack.td index 57634d3c0fb..2224a08d59f 100644 --- a/lib/Target/X86/X86InstrFPStack.td +++ b/lib/Target/X86/X86InstrFPStack.td @@ -1,10 +1,10 @@ -//==- X86InstrFPStack.td - Describe the X86 Instruction Set --*- tablegen -*-=// -// +//===- X86InstrFPStack.td - FPU Instruction Set ------------*- tablegen -*-===// +// // The LLVM Compiler Infrastructure // // This file is distributed under the University of Illinois Open Source // License. See LICENSE.TXT for details. -// +// //===----------------------------------------------------------------------===// // // This file describes the X86 x87 FPU instruction set, defining the @@ -27,6 +27,7 @@ def SDTX86Fst : SDTypeProfile<0, 3, [SDTCisFP<0>, SDTCisVT<2, OtherVT>]>; def SDTX86Fild : SDTypeProfile<1, 2, [SDTCisFP<0>, SDTCisPtrTy<1>, SDTCisVT<2, OtherVT>]>; +def SDTX86Fnstsw : SDTypeProfile<1, 1, [SDTCisVT<0, i16>, SDTCisVT<1, i16>]>; def SDTX86FpToIMem : SDTypeProfile<0, 2, [SDTCisFP<0>, SDTCisPtrTy<1>]>; def SDTX86CwdStore : SDTypeProfile<0, 1, [SDTCisPtrTy<0>]>; @@ -34,13 +35,14 @@ def SDTX86CwdStore : SDTypeProfile<0, 1, [SDTCisPtrTy<0>]>; def X86fld : SDNode<"X86ISD::FLD", SDTX86Fld, [SDNPHasChain, SDNPMayLoad, SDNPMemOperand]>; def X86fst : SDNode<"X86ISD::FST", SDTX86Fst, - [SDNPHasChain, SDNPInFlag, SDNPMayStore, + [SDNPHasChain, SDNPInGlue, SDNPMayStore, SDNPMemOperand]>; def X86fild : SDNode<"X86ISD::FILD", SDTX86Fild, [SDNPHasChain, SDNPMayLoad, SDNPMemOperand]>; def X86fildflag : SDNode<"X86ISD::FILD_FLAG", SDTX86Fild, - [SDNPHasChain, SDNPOutFlag, SDNPMayLoad, + [SDNPHasChain, SDNPOutGlue, SDNPMayLoad, SDNPMemOperand]>; +def X86fp_stsw : SDNode<"X86ISD::FNSTSW16r", SDTX86Fnstsw>; def X86fp_to_i16mem : SDNode<"X86ISD::FP_TO_INT16_IN_MEM", SDTX86FpToIMem, [SDNPHasChain, SDNPMayStore, SDNPMemOperand]>; def X86fp_to_i32mem : SDNode<"X86ISD::FP_TO_INT32_IN_MEM", SDTX86FpToIMem, @@ -73,41 +75,23 @@ def fpimmneg1 : PatLeaf<(fpimm), [{ // Some 'special' instructions let usesCustomInserter = 1 in { // Expanded after instruction selection. - def FP32_TO_INT16_IN_MEM : I<0, Pseudo, - (outs), (ins i16mem:$dst, RFP32:$src), - "##FP32_TO_INT16_IN_MEM PSEUDO!", + def FP32_TO_INT16_IN_MEM : PseudoI<(outs), (ins i16mem:$dst, RFP32:$src), [(X86fp_to_i16mem RFP32:$src, addr:$dst)]>; - def FP32_TO_INT32_IN_MEM : I<0, Pseudo, - (outs), (ins i32mem:$dst, RFP32:$src), - "##FP32_TO_INT32_IN_MEM PSEUDO!", + def FP32_TO_INT32_IN_MEM : PseudoI<(outs), (ins i32mem:$dst, RFP32:$src), [(X86fp_to_i32mem RFP32:$src, addr:$dst)]>; - def FP32_TO_INT64_IN_MEM : I<0, Pseudo, - (outs), (ins i64mem:$dst, RFP32:$src), - "##FP32_TO_INT64_IN_MEM PSEUDO!", + def FP32_TO_INT64_IN_MEM : PseudoI<(outs), (ins i64mem:$dst, RFP32:$src), [(X86fp_to_i64mem RFP32:$src, addr:$dst)]>; - def FP64_TO_INT16_IN_MEM : I<0, Pseudo, - (outs), (ins i16mem:$dst, RFP64:$src), - "##FP64_TO_INT16_IN_MEM PSEUDO!", + def FP64_TO_INT16_IN_MEM : PseudoI<(outs), (ins i16mem:$dst, RFP64:$src), [(X86fp_to_i16mem RFP64:$src, addr:$dst)]>; - def FP64_TO_INT32_IN_MEM : I<0, Pseudo, - (outs), (ins i32mem:$dst, RFP64:$src), - "##FP64_TO_INT32_IN_MEM PSEUDO!", + def FP64_TO_INT32_IN_MEM : PseudoI<(outs), (ins i32mem:$dst, RFP64:$src), [(X86fp_to_i32mem RFP64:$src, addr:$dst)]>; - def FP64_TO_INT64_IN_MEM : I<0, Pseudo, - (outs), (ins i64mem:$dst, RFP64:$src), - "##FP64_TO_INT64_IN_MEM PSEUDO!", + def FP64_TO_INT64_IN_MEM : PseudoI<(outs), (ins i64mem:$dst, RFP64:$src), [(X86fp_to_i64mem RFP64:$src, addr:$dst)]>; - def FP80_TO_INT16_IN_MEM : I<0, Pseudo, - (outs), (ins i16mem:$dst, RFP80:$src), - "##FP80_TO_INT16_IN_MEM PSEUDO!", + def FP80_TO_INT16_IN_MEM : PseudoI<(outs), (ins i16mem:$dst, RFP80:$src), [(X86fp_to_i16mem RFP80:$src, addr:$dst)]>; - def FP80_TO_INT32_IN_MEM : I<0, Pseudo, - (outs), (ins i32mem:$dst, RFP80:$src), - "##FP80_TO_INT32_IN_MEM PSEUDO!", + def FP80_TO_INT32_IN_MEM : PseudoI<(outs), (ins i32mem:$dst, RFP80:$src), [(X86fp_to_i32mem RFP80:$src, addr:$dst)]>; - def FP80_TO_INT64_IN_MEM : I<0, Pseudo, - (outs), (ins i64mem:$dst, RFP80:$src), - "##FP80_TO_INT64_IN_MEM PSEUDO!", + def FP80_TO_INT64_IN_MEM : PseudoI<(outs), (ins i64mem:$dst, RFP80:$src), [(X86fp_to_i64mem RFP80:$src, addr:$dst)]>; } @@ -130,31 +114,8 @@ let usesCustomInserter = 1 in { // Expanded after instruction selection. // a pattern) and the FPI instruction should have emission info (e.g. opcode // encoding and asm printing info). -// Pseudo Instructions for FP stack return values. -def FpGET_ST0_32 : FpI_<(outs RFP32:$dst), (ins), SpecialFP, []>; // FPR = ST(0) -def FpGET_ST0_64 : FpI_<(outs RFP64:$dst), (ins), SpecialFP, []>; // FPR = ST(0) -def FpGET_ST0_80 : FpI_<(outs RFP80:$dst), (ins), SpecialFP, []>; // FPR = ST(0) - -// FpGET_ST1* should only be issued *after* an FpGET_ST0* has been issued when -// there are two values live out on the stack from a call or inlineasm. This -// magic is handled by the stackifier. It is not valid to emit FpGET_ST1* and -// then FpGET_ST0*. In addition, it is invalid for any FP-using operations to -// occur between them. -def FpGET_ST1_32 : FpI_<(outs RFP32:$dst), (ins), SpecialFP, []>; // FPR = ST(1) -def FpGET_ST1_64 : FpI_<(outs RFP64:$dst), (ins), SpecialFP, []>; // FPR = ST(1) -def FpGET_ST1_80 : FpI_<(outs RFP80:$dst), (ins), SpecialFP, []>; // FPR = ST(1) - -let Defs = [ST0] in { -def FpSET_ST0_32 : FpI_<(outs), (ins RFP32:$src), SpecialFP, []>; // ST(0) = FPR -def FpSET_ST0_64 : FpI_<(outs), (ins RFP64:$src), SpecialFP, []>; // ST(0) = FPR -def FpSET_ST0_80 : FpI_<(outs), (ins RFP80:$src), SpecialFP, []>; // ST(0) = FPR -} - -let Defs = [ST1] in { -def FpSET_ST1_32 : FpI_<(outs), (ins RFP32:$src), SpecialFP, []>; // ST(1) = FPR -def FpSET_ST1_64 : FpI_<(outs), (ins RFP64:$src), SpecialFP, []>; // ST(1) = FPR -def FpSET_ST1_80 : FpI_<(outs), (ins RFP80:$src), SpecialFP, []>; // ST(1) = FPR -} +// Pseudo Instruction for FP stack return values. +def FpPOP_RETVAL : FpI_<(outs RFP80:$dst), (ins), SpecialFP, []>; // FpIf32, FpIf64 - Floating Point Pseudo Instruction template. // f32 instructions can use SSE1 and are predicated on FPStackf32 == !SSE1. @@ -165,19 +126,6 @@ class FpIf32 pattern> : class FpIf64 pattern> : FpI_, Requires<[FPStackf64]>; -// Register copies. Just copies, the shortening ones do not truncate. -let neverHasSideEffects = 1 in { - def MOV_Fp3232 : FpIf32<(outs RFP32:$dst), (ins RFP32:$src), SpecialFP, []>; - def MOV_Fp3264 : FpIf32<(outs RFP64:$dst), (ins RFP32:$src), SpecialFP, []>; - def MOV_Fp6432 : FpIf32<(outs RFP32:$dst), (ins RFP64:$src), SpecialFP, []>; - def MOV_Fp6464 : FpIf64<(outs RFP64:$dst), (ins RFP64:$src), SpecialFP, []>; - def MOV_Fp8032 : FpIf32<(outs RFP32:$dst), (ins RFP80:$src), SpecialFP, []>; - def MOV_Fp3280 : FpIf32<(outs RFP80:$dst), (ins RFP32:$src), SpecialFP, []>; - def MOV_Fp8064 : FpIf64<(outs RFP64:$dst), (ins RFP80:$src), SpecialFP, []>; - def MOV_Fp6480 : FpIf64<(outs RFP80:$dst), (ins RFP64:$src), SpecialFP, []>; - def MOV_Fp8080 : FpI_ <(outs RFP80:$dst), (ins RFP80:$src), SpecialFP, []>; -} - // Factoring for arithmetic. multiclass FPBinary_rr { // Register op register -> register @@ -257,6 +205,7 @@ def _FI32m : FPI<0xDA, fp, (outs), (ins i32mem:$src), } } +let Defs = [FPSW] in { defm ADD : FPBinary_rr; defm SUB : FPBinary_rr; defm MUL : FPBinary_rr; @@ -267,6 +216,7 @@ defm SUBR: FPBinary; defm MUL : FPBinary; defm DIV : FPBinary; defm DIVR: FPBinary; +} class FPST0rInst o, string asm> : FPI, D8; @@ -279,22 +229,22 @@ class FPrST0PInst o, string asm> // of some of the 'reverse' forms of the fsub and fdiv instructions. As such, // we have to put some 'r's in and take them out of weird places. def ADD_FST0r : FPST0rInst <0xC0, "fadd\t$op">; -def ADD_FrST0 : FPrST0Inst <0xC0, "fadd\t{%st(0), $op|$op, %ST(0)}">; +def ADD_FrST0 : FPrST0Inst <0xC0, "fadd\t{%st(0), $op|$op, ST(0)}">; def ADD_FPrST0 : FPrST0PInst<0xC0, "faddp\t$op">; def SUBR_FST0r : FPST0rInst <0xE8, "fsubr\t$op">; -def SUB_FrST0 : FPrST0Inst <0xE8, "fsub{r}\t{%st(0), $op|$op, %ST(0)}">; +def SUB_FrST0 : FPrST0Inst <0xE8, "fsub{r}\t{%st(0), $op|$op, ST(0)}">; def SUB_FPrST0 : FPrST0PInst<0xE8, "fsub{r}p\t$op">; def SUB_FST0r : FPST0rInst <0xE0, "fsub\t$op">; -def SUBR_FrST0 : FPrST0Inst <0xE0, "fsub{|r}\t{%st(0), $op|$op, %ST(0)}">; +def SUBR_FrST0 : FPrST0Inst <0xE0, "fsub{|r}\t{%st(0), $op|$op, ST(0)}">; def SUBR_FPrST0 : FPrST0PInst<0xE0, "fsub{|r}p\t$op">; def MUL_FST0r : FPST0rInst <0xC8, "fmul\t$op">; -def MUL_FrST0 : FPrST0Inst <0xC8, "fmul\t{%st(0), $op|$op, %ST(0)}">; +def MUL_FrST0 : FPrST0Inst <0xC8, "fmul\t{%st(0), $op|$op, ST(0)}">; def MUL_FPrST0 : FPrST0PInst<0xC8, "fmulp\t$op">; def DIVR_FST0r : FPST0rInst <0xF8, "fdivr\t$op">; -def DIV_FrST0 : FPrST0Inst <0xF8, "fdiv{r}\t{%st(0), $op|$op, %ST(0)}">; +def DIV_FrST0 : FPrST0Inst <0xF8, "fdiv{r}\t{%st(0), $op|$op, ST(0)}">; def DIV_FPrST0 : FPrST0PInst<0xF8, "fdiv{r}p\t$op">; def DIV_FST0r : FPST0rInst <0xF0, "fdiv\t$op">; -def DIVR_FrST0 : FPrST0Inst <0xF0, "fdiv{|r}\t{%st(0), $op|$op, %ST(0)}">; +def DIVR_FrST0 : FPrST0Inst <0xF0, "fdiv{|r}\t{%st(0), $op|$op, ST(0)}">; def DIVR_FPrST0 : FPrST0PInst<0xF0, "fdiv{|r}p\t$op">; def COM_FST0r : FPST0rInst <0xD0, "fcom\t$op">; @@ -311,6 +261,7 @@ def _Fp80 : FpI_<(outs RFP80:$dst), (ins RFP80:$src), OneArgFPRW, def _F : FPI, D9; } +let Defs = [FPSW] in { defm CHS : FPUnary; defm ABS : FPUnary; defm SQRT: FPUnary; @@ -323,6 +274,7 @@ def TST_Fp64 : FpIf64<(outs), (ins RFP64:$src), OneArgFP, []>; def TST_Fp80 : FpI_<(outs), (ins RFP80:$src), OneArgFP, []>; } def TST_F : FPI<0xE4, RawFrm, (outs), (ins), "ftst">, D9; +} // Defs = [FPSW] // Versions of FP instructions that take a single memory operand. Added for the // disassembler; remove as they are included with patterns elsewhere. @@ -340,7 +292,7 @@ def FCOMP64m : FPI<0xDC, MRM3m, (outs), (ins f64mem:$src), "fcomp{l}\t$src">; def FRSTORm : FPI<0xDD, MRM4m, (outs f32mem:$dst), (ins), "frstor\t$dst">; def FSAVEm : FPI<0xDD, MRM6m, (outs f32mem:$dst), (ins), "fnsave\t$dst">; -def FNSTSWm : FPI<0xDD, MRM7m, (outs f32mem:$dst), (ins), "fnstsw{w}\t$dst">; +def FNSTSWm : FPI<0xDD, MRM7m, (outs f32mem:$dst), (ins), "fnstsw\t$dst">; def FICOM16m : FPI<0xDE, MRM2m, (outs), (ins i16mem:$src), "ficom{s}\t$src">; def FICOMP16m: FPI<0xDE, MRM3m, (outs), (ins i16mem:$src), "ficomp{s}\t$src">; @@ -370,6 +322,7 @@ multiclass FPCMov { Requires<[HasCMov]>; } +let Defs = [FPSW] in { let Uses = [EFLAGS], Constraints = "$src1 = $dst" in { defm CMOVB : FPCMov; defm CMOVBE : FPCMov; @@ -384,21 +337,21 @@ defm CMOVNP : FPCMov; let Predicates = [HasCMov] in { // These are not factored because there's no clean way to pass DA/DB. def CMOVB_F : FPI<0xC0, AddRegFrm, (outs RST:$op), (ins), - "fcmovb\t{$op, %st(0)|%ST(0), $op}">, DA; + "fcmovb\t{$op, %st(0)|ST(0), $op}">, DA; def CMOVBE_F : FPI<0xD0, AddRegFrm, (outs RST:$op), (ins), - "fcmovbe\t{$op, %st(0)|%ST(0), $op}">, DA; + "fcmovbe\t{$op, %st(0)|ST(0), $op}">, DA; def CMOVE_F : FPI<0xC8, AddRegFrm, (outs RST:$op), (ins), - "fcmove\t{$op, %st(0)|%ST(0), $op}">, DA; + "fcmove\t{$op, %st(0)|ST(0), $op}">, DA; def CMOVP_F : FPI<0xD8, AddRegFrm, (outs RST:$op), (ins), - "fcmovu\t {$op, %st(0)|%ST(0), $op}">, DA; + "fcmovu\t {$op, %st(0)|ST(0), $op}">, DA; def CMOVNB_F : FPI<0xC0, AddRegFrm, (outs RST:$op), (ins), - "fcmovnb\t{$op, %st(0)|%ST(0), $op}">, DB; + "fcmovnb\t{$op, %st(0)|ST(0), $op}">, DB; def CMOVNBE_F: FPI<0xD0, AddRegFrm, (outs RST:$op), (ins), - "fcmovnbe\t{$op, %st(0)|%ST(0), $op}">, DB; + "fcmovnbe\t{$op, %st(0)|ST(0), $op}">, DB; def CMOVNE_F : FPI<0xC8, AddRegFrm, (outs RST:$op), (ins), - "fcmovne\t{$op, %st(0)|%ST(0), $op}">, DB; + "fcmovne\t{$op, %st(0)|ST(0), $op}">, DB; def CMOVNP_F : FPI<0xD8, AddRegFrm, (outs RST:$op), (ins), - "fcmovnu\t{$op, %st(0)|%ST(0), $op}">, DB; + "fcmovnu\t{$op, %st(0)|ST(0), $op}">, DB; } // Predicates = [HasCMov] // Floating point loads & stores. @@ -469,68 +422,85 @@ def IST_Fp32m80 : FpI_<(outs), (ins i32mem:$op, RFP80:$src), OneArgFP, []>; def IST_Fp64m80 : FpI_<(outs), (ins i64mem:$op, RFP80:$src), OneArgFP, []>; } -let mayLoad = 1 in { -def LD_F32m : FPI<0xD9, MRM0m, (outs), (ins f32mem:$src), "fld{s}\t$src">; -def LD_F64m : FPI<0xDD, MRM0m, (outs), (ins f64mem:$src), "fld{l}\t$src">; -def LD_F80m : FPI<0xDB, MRM5m, (outs), (ins f80mem:$src), "fld{t}\t$src">; -def ILD_F16m : FPI<0xDF, MRM0m, (outs), (ins i16mem:$src), "fild{s}\t$src">; -def ILD_F32m : FPI<0xDB, MRM0m, (outs), (ins i32mem:$src), "fild{l}\t$src">; -def ILD_F64m : FPI<0xDF, MRM5m, (outs), (ins i64mem:$src), "fild{ll}\t$src">; +let mayLoad = 1, SchedRW = [WriteLoad] in { +def LD_F32m : FPI<0xD9, MRM0m, (outs), (ins f32mem:$src), "fld{s}\t$src", + IIC_FLD>; +def LD_F64m : FPI<0xDD, MRM0m, (outs), (ins f64mem:$src), "fld{l}\t$src", + IIC_FLD>; +def LD_F80m : FPI<0xDB, MRM5m, (outs), (ins f80mem:$src), "fld{t}\t$src", + IIC_FLD80>; +def ILD_F16m : FPI<0xDF, MRM0m, (outs), (ins i16mem:$src), "fild{s}\t$src", + IIC_FILD>; +def ILD_F32m : FPI<0xDB, MRM0m, (outs), (ins i32mem:$src), "fild{l}\t$src", + IIC_FILD>; +def ILD_F64m : FPI<0xDF, MRM5m, (outs), (ins i64mem:$src), "fild{ll}\t$src", + IIC_FILD>; } -let mayStore = 1 in { -def ST_F32m : FPI<0xD9, MRM2m, (outs), (ins f32mem:$dst), "fst{s}\t$dst">; -def ST_F64m : FPI<0xDD, MRM2m, (outs), (ins f64mem:$dst), "fst{l}\t$dst">; -def ST_FP32m : FPI<0xD9, MRM3m, (outs), (ins f32mem:$dst), "fstp{s}\t$dst">; -def ST_FP64m : FPI<0xDD, MRM3m, (outs), (ins f64mem:$dst), "fstp{l}\t$dst">; -def ST_FP80m : FPI<0xDB, MRM7m, (outs), (ins f80mem:$dst), "fstp{t}\t$dst">; -def IST_F16m : FPI<0xDF, MRM2m, (outs), (ins i16mem:$dst), "fist{s}\t$dst">; -def IST_F32m : FPI<0xDB, MRM2m, (outs), (ins i32mem:$dst), "fist{l}\t$dst">; -def IST_FP16m : FPI<0xDF, MRM3m, (outs), (ins i16mem:$dst), "fistp{s}\t$dst">; -def IST_FP32m : FPI<0xDB, MRM3m, (outs), (ins i32mem:$dst), "fistp{l}\t$dst">; -def IST_FP64m : FPI<0xDF, MRM7m, (outs), (ins i64mem:$dst), "fistp{ll}\t$dst">; +let mayStore = 1, SchedRW = [WriteStore] in { +def ST_F32m : FPI<0xD9, MRM2m, (outs), (ins f32mem:$dst), "fst{s}\t$dst", + IIC_FST>; +def ST_F64m : FPI<0xDD, MRM2m, (outs), (ins f64mem:$dst), "fst{l}\t$dst", + IIC_FST>; +def ST_FP32m : FPI<0xD9, MRM3m, (outs), (ins f32mem:$dst), "fstp{s}\t$dst", + IIC_FST>; +def ST_FP64m : FPI<0xDD, MRM3m, (outs), (ins f64mem:$dst), "fstp{l}\t$dst", + IIC_FST>; +def ST_FP80m : FPI<0xDB, MRM7m, (outs), (ins f80mem:$dst), "fstp{t}\t$dst", + IIC_FST80>; +def IST_F16m : FPI<0xDF, MRM2m, (outs), (ins i16mem:$dst), "fist{s}\t$dst", + IIC_FIST>; +def IST_F32m : FPI<0xDB, MRM2m, (outs), (ins i32mem:$dst), "fist{l}\t$dst", + IIC_FIST>; +def IST_FP16m : FPI<0xDF, MRM3m, (outs), (ins i16mem:$dst), "fistp{s}\t$dst", + IIC_FIST>; +def IST_FP32m : FPI<0xDB, MRM3m, (outs), (ins i32mem:$dst), "fistp{l}\t$dst", + IIC_FIST>; +def IST_FP64m : FPI<0xDF, MRM7m, (outs), (ins i64mem:$dst), "fistp{ll}\t$dst", + IIC_FIST>; } // FISTTP requires SSE3 even though it's a FPStack op. +let Predicates = [HasSSE3] in { def ISTT_Fp16m32 : FpI_<(outs), (ins i16mem:$op, RFP32:$src), OneArgFP, - [(X86fp_to_i16mem RFP32:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i16mem RFP32:$src, addr:$op)]>; def ISTT_Fp32m32 : FpI_<(outs), (ins i32mem:$op, RFP32:$src), OneArgFP, - [(X86fp_to_i32mem RFP32:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i32mem RFP32:$src, addr:$op)]>; def ISTT_Fp64m32 : FpI_<(outs), (ins i64mem:$op, RFP32:$src), OneArgFP, - [(X86fp_to_i64mem RFP32:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i64mem RFP32:$src, addr:$op)]>; def ISTT_Fp16m64 : FpI_<(outs), (ins i16mem:$op, RFP64:$src), OneArgFP, - [(X86fp_to_i16mem RFP64:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i16mem RFP64:$src, addr:$op)]>; def ISTT_Fp32m64 : FpI_<(outs), (ins i32mem:$op, RFP64:$src), OneArgFP, - [(X86fp_to_i32mem RFP64:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i32mem RFP64:$src, addr:$op)]>; def ISTT_Fp64m64 : FpI_<(outs), (ins i64mem:$op, RFP64:$src), OneArgFP, - [(X86fp_to_i64mem RFP64:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i64mem RFP64:$src, addr:$op)]>; def ISTT_Fp16m80 : FpI_<(outs), (ins i16mem:$op, RFP80:$src), OneArgFP, - [(X86fp_to_i16mem RFP80:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i16mem RFP80:$src, addr:$op)]>; def ISTT_Fp32m80 : FpI_<(outs), (ins i32mem:$op, RFP80:$src), OneArgFP, - [(X86fp_to_i32mem RFP80:$src, addr:$op)]>, - Requires<[HasSSE3]>; + [(X86fp_to_i32mem RFP80:$src, addr:$op)]>; def ISTT_Fp64m80 : FpI_<(outs), (ins i64mem:$op, RFP80:$src), OneArgFP, - [(X86fp_to_i64mem RFP80:$src, addr:$op)]>, - Requires<[HasSSE3]>; - -let mayStore = 1 in { -def ISTT_FP16m : FPI<0xDF, MRM1m, (outs), (ins i16mem:$dst), "fisttp{s}\t$dst">; -def ISTT_FP32m : FPI<0xDB, MRM1m, (outs), (ins i32mem:$dst), "fisttp{l}\t$dst">; + [(X86fp_to_i64mem RFP80:$src, addr:$op)]>; +} // Predicates = [HasSSE3] + +let mayStore = 1, SchedRW = [WriteStore] in { +def ISTT_FP16m : FPI<0xDF, MRM1m, (outs), (ins i16mem:$dst), "fisttp{s}\t$dst", + IIC_FST>; +def ISTT_FP32m : FPI<0xDB, MRM1m, (outs), (ins i32mem:$dst), "fisttp{l}\t$dst", + IIC_FST>; def ISTT_FP64m : FPI<0xDD, MRM1m, (outs), (ins i64mem:$dst), - "fisttp{ll}\t$dst">; + "fisttp{ll}\t$dst", IIC_FST>; } // FP Stack manipulation instructions. -def LD_Frr : FPI<0xC0, AddRegFrm, (outs), (ins RST:$op), "fld\t$op">, D9; -def ST_Frr : FPI<0xD0, AddRegFrm, (outs), (ins RST:$op), "fst\t$op">, DD; -def ST_FPrr : FPI<0xD8, AddRegFrm, (outs), (ins RST:$op), "fstp\t$op">, DD; -def XCH_F : FPI<0xC8, AddRegFrm, (outs), (ins RST:$op), "fxch\t$op">, D9; +let SchedRW = [WriteMove] in { +def LD_Frr : FPI<0xC0, AddRegFrm, (outs), (ins RST:$op), "fld\t$op", + IIC_FLD>, D9; +def ST_Frr : FPI<0xD0, AddRegFrm, (outs), (ins RST:$op), "fst\t$op", + IIC_FST>, DD; +def ST_FPrr : FPI<0xD8, AddRegFrm, (outs), (ins RST:$op), "fstp\t$op", + IIC_FST>, DD; +def XCH_F : FPI<0xC8, AddRegFrm, (outs), (ins RST:$op), "fxch\t$op", + IIC_FXCH>, D9; +} // Floating point constant loads. let isReMaterializable = 1 in { @@ -548,20 +518,25 @@ def LD_Fp180 : FpI_<(outs RFP80:$dst), (ins), ZeroArgFP, [(set RFP80:$dst, fpimm1)]>; } -def LD_F0 : FPI<0xEE, RawFrm, (outs), (ins), "fldz">, D9; -def LD_F1 : FPI<0xE8, RawFrm, (outs), (ins), "fld1">, D9; - +let SchedRW = [WriteZero] in { +def LD_F0 : FPI<0xEE, RawFrm, (outs), (ins), "fldz", IIC_FLDZ>, D9; +def LD_F1 : FPI<0xE8, RawFrm, (outs), (ins), "fld1", IIC_FIST>, D9; +} // Floating point compares. -let Defs = [EFLAGS] in { +let SchedRW = [WriteFAdd] in { def UCOM_Fpr32 : FpIf32<(outs), (ins RFP32:$lhs, RFP32:$rhs), CompareFP, - []>; // FPSW = cmp ST(0) with ST(i) + [(set FPSW, (trunc (X86cmp RFP32:$lhs, RFP32:$rhs)))]>; def UCOM_Fpr64 : FpIf64<(outs), (ins RFP64:$lhs, RFP64:$rhs), CompareFP, - []>; // FPSW = cmp ST(0) with ST(i) + [(set FPSW, (trunc (X86cmp RFP64:$lhs, RFP64:$rhs)))]>; def UCOM_Fpr80 : FpI_ <(outs), (ins RFP80:$lhs, RFP80:$rhs), CompareFP, - []>; // FPSW = cmp ST(0) with ST(i) - + [(set FPSW, (trunc (X86cmp RFP80:$lhs, RFP80:$rhs)))]>; +} // SchedRW +} // Defs = [FPSW] + +let SchedRW = [WriteFAdd] in { // CC = ST(0) cmp ST(i) +let Defs = [EFLAGS, FPSW] in { def UCOM_FpIr32: FpIf32<(outs), (ins RFP32:$lhs, RFP32:$rhs), CompareFP, [(set EFLAGS, (X86cmp RFP32:$lhs, RFP32:$rhs))]>; def UCOM_FpIr64: FpIf64<(outs), (ins RFP64:$lhs, RFP64:$rhs), CompareFP, @@ -570,81 +545,100 @@ def UCOM_FpIr80: FpI_<(outs), (ins RFP80:$lhs, RFP80:$rhs), CompareFP, [(set EFLAGS, (X86cmp RFP80:$lhs, RFP80:$rhs))]>; } -let Defs = [EFLAGS], Uses = [ST0] in { +let Defs = [FPSW], Uses = [ST0] in { def UCOM_Fr : FPI<0xE0, AddRegFrm, // FPSW = cmp ST(0) with ST(i) (outs), (ins RST:$reg), - "fucom\t$reg">, DD; + "fucom\t$reg", IIC_FUCOM>, DD; def UCOM_FPr : FPI<0xE8, AddRegFrm, // FPSW = cmp ST(0) with ST(i), pop (outs), (ins RST:$reg), - "fucomp\t$reg">, DD; + "fucomp\t$reg", IIC_FUCOM>, DD; def UCOM_FPPr : FPI<0xE9, RawFrm, // cmp ST(0) with ST(1), pop, pop (outs), (ins), - "fucompp">, DA; + "fucompp", IIC_FUCOM>, DA; +} +let Defs = [EFLAGS, FPSW], Uses = [ST0] in { def UCOM_FIr : FPI<0xE8, AddRegFrm, // CC = cmp ST(0) with ST(i) (outs), (ins RST:$reg), - "fucomi\t{$reg, %st(0)|%ST(0), $reg}">, DB; + "fucomi\t$reg", IIC_FUCOMI>, DB; def UCOM_FIPr : FPI<0xE8, AddRegFrm, // CC = cmp ST(0) with ST(i), pop (outs), (ins RST:$reg), - "fucomip\t{$reg, %st(0)|%ST(0), $reg}">, DF; + "fucompi\t$reg", IIC_FUCOMI>, DF; } +let Defs = [EFLAGS, FPSW] in { def COM_FIr : FPI<0xF0, AddRegFrm, (outs), (ins RST:$reg), - "fcomi\t{$reg, %st(0)|%ST(0), $reg}">, DB; + "fcomi\t$reg", IIC_FCOMI>, DB; def COM_FIPr : FPI<0xF0, AddRegFrm, (outs), (ins RST:$reg), - "fcomip\t{$reg, %st(0)|%ST(0), $reg}">, DF; + "fcompi\t$reg", IIC_FCOMI>, DF; +} +} // SchedRW // Floating point flag ops. -let Defs = [AX] in -def FNSTSW8r : I<0xE0, RawFrm, // AX = fp flags - (outs), (ins), "fnstsw %ax", []>, DF; +let SchedRW = [WriteALU] in { +let Defs = [AX], Uses = [FPSW] in +def FNSTSW16r : I<0xE0, RawFrm, // AX = fp flags + (outs), (ins), "fnstsw %ax", + [(set AX, (X86fp_stsw FPSW))], IIC_FNSTSW>, DF; def FNSTCW16m : I<0xD9, MRM7m, // [mem16] = X87 control world - (outs), (ins i16mem:$dst), "fnstcw{w}\t$dst", - [(X86fp_cwd_get16 addr:$dst)]>; - + (outs), (ins i16mem:$dst), "fnstcw\t$dst", + [(X86fp_cwd_get16 addr:$dst)], IIC_FNSTCW>; +} // SchedRW let mayLoad = 1 in def FLDCW16m : I<0xD9, MRM5m, // X87 control world = [mem16] - (outs), (ins i16mem:$dst), "fldcw{w}\t$dst", []>; + (outs), (ins i16mem:$dst), "fldcw\t$dst", [], IIC_FLDCW>, + Sched<[WriteLoad]>; // FPU control instructions -def FNINIT : I<0xE3, RawFrm, (outs), (ins), "fninit", []>, DB; +let SchedRW = [WriteMicrocoded] in { +let Defs = [FPSW] in +def FNINIT : I<0xE3, RawFrm, (outs), (ins), "fninit", [], IIC_FNINIT>, DB; def FFREE : FPI<0xC0, AddRegFrm, (outs), (ins RST:$reg), - "ffree\t$reg">, DD; - + "ffree\t$reg", IIC_FFREE>, DD; // Clear exceptions -def FNCLEX : I<0xE2, RawFrm, (outs), (ins), "fnclex", []>, DB; +let Defs = [FPSW] in +def FNCLEX : I<0xE2, RawFrm, (outs), (ins), "fnclex", [], IIC_FNCLEX>, DB; +} // SchedRW // Operandless floating-point instructions for the disassembler. -def WAIT : I<0x9B, RawFrm, (outs), (ins), "wait", []>; - -def FNOP : I<0xD0, RawFrm, (outs), (ins), "fnop", []>, D9; -def FXAM : I<0xE5, RawFrm, (outs), (ins), "fxam", []>, D9; -def FLDL2T : I<0xE9, RawFrm, (outs), (ins), "fldl2t", []>, D9; -def FLDL2E : I<0xEA, RawFrm, (outs), (ins), "fldl2e", []>, D9; -def FLDPI : I<0xEB, RawFrm, (outs), (ins), "fldpi", []>, D9; -def FLDLG2 : I<0xEC, RawFrm, (outs), (ins), "fldlg2", []>, D9; -def FLDLN2 : I<0xED, RawFrm, (outs), (ins), "fldln2", []>, D9; -def F2XM1 : I<0xF0, RawFrm, (outs), (ins), "f2xm1", []>, D9; -def FYL2X : I<0xF1, RawFrm, (outs), (ins), "fyl2x", []>, D9; -def FPTAN : I<0xF2, RawFrm, (outs), (ins), "fptan", []>, D9; -def FPATAN : I<0xF3, RawFrm, (outs), (ins), "fpatan", []>, D9; -def FXTRACT : I<0xF4, RawFrm, (outs), (ins), "fxtract", []>, D9; -def FPREM1 : I<0xF5, RawFrm, (outs), (ins), "fprem1", []>, D9; -def FDECSTP : I<0xF6, RawFrm, (outs), (ins), "fdecstp", []>, D9; -def FINCSTP : I<0xF7, RawFrm, (outs), (ins), "fincstp", []>, D9; -def FPREM : I<0xF8, RawFrm, (outs), (ins), "fprem", []>, D9; -def FYL2XP1 : I<0xF9, RawFrm, (outs), (ins), "fyl2xp1", []>, D9; -def FSINCOS : I<0xFB, RawFrm, (outs), (ins), "fsincos", []>, D9; -def FRNDINT : I<0xFC, RawFrm, (outs), (ins), "frndint", []>, D9; -def FSCALE : I<0xFD, RawFrm, (outs), (ins), "fscale", []>, D9; -def FCOMPP : I<0xD9, RawFrm, (outs), (ins), "fcompp", []>, DE; +let SchedRW = [WriteMicrocoded] in { +def WAIT : I<0x9B, RawFrm, (outs), (ins), "wait", [], IIC_WAIT>; + +def FNOP : I<0xD0, RawFrm, (outs), (ins), "fnop", [], IIC_FNOP>, D9; +def FXAM : I<0xE5, RawFrm, (outs), (ins), "fxam", [], IIC_FXAM>, D9; +def FLDL2T : I<0xE9, RawFrm, (outs), (ins), "fldl2t", [], IIC_FLDL>, D9; +def FLDL2E : I<0xEA, RawFrm, (outs), (ins), "fldl2e", [], IIC_FLDL>, D9; +def FLDPI : I<0xEB, RawFrm, (outs), (ins), "fldpi", [], IIC_FLDL>, D9; +def FLDLG2 : I<0xEC, RawFrm, (outs), (ins), "fldlg2", [], IIC_FLDL>, D9; +def FLDLN2 : I<0xED, RawFrm, (outs), (ins), "fldln2", [], IIC_FLDL>, D9; +def F2XM1 : I<0xF0, RawFrm, (outs), (ins), "f2xm1", [], IIC_F2XM1>, D9; +def FYL2X : I<0xF1, RawFrm, (outs), (ins), "fyl2x", [], IIC_FYL2X>, D9; +def FPTAN : I<0xF2, RawFrm, (outs), (ins), "fptan", [], IIC_FPTAN>, D9; +def FPATAN : I<0xF3, RawFrm, (outs), (ins), "fpatan", [], IIC_FPATAN>, D9; +def FXTRACT : I<0xF4, RawFrm, (outs), (ins), "fxtract", [], IIC_FXTRACT>, D9; +def FPREM1 : I<0xF5, RawFrm, (outs), (ins), "fprem1", [], IIC_FPREM1>, D9; +def FDECSTP : I<0xF6, RawFrm, (outs), (ins), "fdecstp", [], IIC_FPSTP>, D9; +def FINCSTP : I<0xF7, RawFrm, (outs), (ins), "fincstp", [], IIC_FPSTP>, D9; +def FPREM : I<0xF8, RawFrm, (outs), (ins), "fprem", [], IIC_FPREM>, D9; +def FYL2XP1 : I<0xF9, RawFrm, (outs), (ins), "fyl2xp1", [], IIC_FYL2XP1>, D9; +def FSINCOS : I<0xFB, RawFrm, (outs), (ins), "fsincos", [], IIC_FSINCOS>, D9; +def FRNDINT : I<0xFC, RawFrm, (outs), (ins), "frndint", [], IIC_FRNDINT>, D9; +def FSCALE : I<0xFD, RawFrm, (outs), (ins), "fscale", [], IIC_FSCALE>, D9; +def FCOMPP : I<0xD9, RawFrm, (outs), (ins), "fcompp", [], IIC_FCOMPP>, DE; def FXSAVE : I<0xAE, MRM0m, (outs opaque512mem:$dst), (ins), - "fxsave\t$dst", []>, TB; + "fxsave\t$dst", [], IIC_FXSAVE>, TB; +def FXSAVE64 : I<0xAE, MRM0m, (outs opaque512mem:$dst), (ins), + "fxsaveq\t$dst", [], IIC_FXSAVE>, TB, REX_W, + Requires<[In64BitMode]>; def FXRSTOR : I<0xAE, MRM1m, (outs), (ins opaque512mem:$src), - "fxrstor\t$src", []>, TB; + "fxrstor\t$src", [], IIC_FXRSTOR>, TB; +def FXRSTOR64 : I<0xAE, MRM1m, (outs), (ins opaque512mem:$src), + "fxrstorq\t$src", [], IIC_FXRSTOR>, TB, REX_W, + Requires<[In64BitMode]>; +} // SchedRW //===----------------------------------------------------------------------===// // Non-Instruction Patterns