RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-31 20:51:52 +01:00
Code Issues Projects Releases Wiki Activity

Change the x86 assembly output to use tab characters to separate the

Browse Source 
mnemonics from their operands instead of single spaces. This makes the
assembly output a little more consistent with various other compilers
(f.e. GCC), and slightly easier to read. Also, update the regression
tests accordingly.

llvm-svn: 40648
This commit is contained in:
Dan Gohman 2007-07-31 20:11:57 +00:00
parent 4251f42689
commit e3464e6bec
35 changed files with 1163 additions and 1163 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

114
lib/Target/X86/X86InstrFPStack.td
View File

@ -165,9 +165,9 @@ def _Fp64m32: FpI<(outs RFP64:$dst), (ins RFP64:$src1, f32mem:$src2), OneArgFPRW
[(set RFP64:$dst,
(OpNode RFP64:$src1, (extloadf32 addr:$src2)))]>;
def _F32m : FPI<0xD8, fp, (outs), (ins f32mem:$src),
!strconcat("f", !strconcat(asmstring, "{s} $src"))>;
!strconcat("f", !strconcat(asmstring, "{s}\t$src"))>;
def _F64m : FPI<0xDC, fp, (outs), (ins f64mem:$src),
!strconcat("f", !strconcat(asmstring, "{l} $src"))>;
!strconcat("f", !strconcat(asmstring, "{l}\t$src"))>;
// ST(0) = ST(0) + [memint]
def _FpI16m32 : FpI<(outs RFP32:$dst), (ins RFP32:$src1, i16mem:$src2), OneArgFPRW,
[(set RFP32:$dst, (OpNode RFP32:$src1,
@ -182,9 +182,9 @@ def _FpI32m64 : FpI<(outs RFP64:$dst), (ins RFP64:$src1, i32mem:$src2), OneArgFP
[(set RFP64:$dst, (OpNode RFP64:$src1,
(X86fild addr:$src2, i32)))]>;
def _FI16m : FPI<0xDE, fp, (outs), (ins i16mem:$src),
!strconcat("fi", !strconcat(asmstring, "{s} $src"))>;
!strconcat("fi", !strconcat(asmstring, "{s}\t$src"))>;
def _FI32m : FPI<0xDA, fp, (outs), (ins i32mem:$src),
!strconcat("fi", !strconcat(asmstring, "{l} $src"))>;
!strconcat("fi", !strconcat(asmstring, "{l}\t$src"))>;
}
defm ADD : FPBinary_rr<fadd>;
@ -208,24 +208,24 @@ class FPrST0PInst<bits<8> o, string asm>
// NOTE: GAS and apparently all other AT&T style assemblers have a broken notion
// 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 $op">;
def ADD_FrST0 : FPrST0Inst <0xC0, "fadd {%st(0), $op|$op, %ST(0)}">;
def ADD_FPrST0 : FPrST0PInst<0xC0, "faddp $op">;
def SUBR_FST0r : FPST0rInst <0xE8, "fsubr $op">;
def SUB_FrST0 : FPrST0Inst <0xE8, "fsub{r} {%st(0), $op|$op, %ST(0)}">;
def SUB_FPrST0 : FPrST0PInst<0xE8, "fsub{r}p $op">;
def SUB_FST0r : FPST0rInst <0xE0, "fsub $op">;
def SUBR_FrST0 : FPrST0Inst <0xE0, "fsub{|r} {%st(0), $op|$op, %ST(0)}">;
def SUBR_FPrST0 : FPrST0PInst<0xE0, "fsub{|r}p $op">;
def MUL_FST0r : FPST0rInst <0xC8, "fmul $op">;
def MUL_FrST0 : FPrST0Inst <0xC8, "fmul {%st(0), $op|$op, %ST(0)}">;
def MUL_FPrST0 : FPrST0PInst<0xC8, "fmulp $op">;
def DIVR_FST0r : FPST0rInst <0xF8, "fdivr $op">;
def DIV_FrST0 : FPrST0Inst <0xF8, "fdiv{r} {%st(0), $op|$op, %ST(0)}">;
def DIV_FPrST0 : FPrST0PInst<0xF8, "fdiv{r}p $op">;
def DIV_FST0r : FPST0rInst <0xF0, "fdiv $op">;
def DIVR_FrST0 : FPrST0Inst <0xF0, "fdiv{|r} {%st(0), $op|$op, %ST(0)}">;
def DIVR_FPrST0 : FPrST0PInst<0xF0, "fdiv{|r}p $op">;
def ADD_FST0r : FPST0rInst <0xC0, "fadd\t$op">;
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_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_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_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_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_FPrST0 : FPrST0PInst<0xF0, "fdiv{|r}p\t$op">;
// Unary operations.
multiclass FPUnary<SDNode OpNode, bits<8> opcode, string asmstring> {
@ -270,21 +270,21 @@ defm CMOVNP : FPCMov<X86_COND_NP>;
// These are not factored because there's no clean way to pass DA/DB.
def CMOVB_F : FPI<0xC0, AddRegFrm, (outs RST:$op), (ins),
"fcmovb {$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 {$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 {$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 {$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 {$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 {$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 {$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 {$op, %st(0)|%ST(0), $op}">, DB;
"fcmovnu\t{$op, %st(0)|%ST(0), $op}">, DB;
// Floating point loads & stores.
def LD_Fp32m : FpI<(outs RFP32:$dst), (ins f32mem:$src), ZeroArgFP,
@ -321,20 +321,20 @@ def IST_Fp16m64 : FpI<(outs), (ins i16mem:$op, RFP64:$src), OneArgFP, []>;
def IST_Fp32m64 : FpI<(outs), (ins i32mem:$op, RFP64:$src), OneArgFP, []>;
def IST_Fp64m64 : FpI<(outs), (ins i64mem:$op, RFP64:$src), OneArgFP, []>;
def LD_F32m : FPI<0xD9, MRM0m, (outs), (ins f32mem:$src), "fld{s} $src">;
def LD_F64m : FPI<0xDD, MRM0m, (outs), (ins f64mem:$src), "fld{l} $src">;
def ILD_F16m : FPI<0xDF, MRM0m, (outs), (ins i16mem:$src), "fild{s} $src">;
def ILD_F32m : FPI<0xDB, MRM0m, (outs), (ins i32mem:$src), "fild{l} $src">;
def ILD_F64m : FPI<0xDF, MRM5m, (outs), (ins i64mem:$src), "fild{ll} $src">;
def ST_F32m : FPI<0xD9, MRM2m, (outs), (ins f32mem:$dst), "fst{s} $dst">;
def ST_F64m : FPI<0xDD, MRM2m, (outs), (ins f64mem:$dst), "fst{l} $dst">;
def ST_FP32m : FPI<0xD9, MRM3m, (outs), (ins f32mem:$dst), "fstp{s} $dst">;
def ST_FP64m : FPI<0xDD, MRM3m, (outs), (ins f64mem:$dst), "fstp{l} $dst">;
def IST_F16m : FPI<0xDF, MRM2m, (outs), (ins i16mem:$dst), "fist{s} $dst">;
def IST_F32m : FPI<0xDB, MRM2m, (outs), (ins i32mem:$dst), "fist{l} $dst">;
def IST_FP16m : FPI<0xDF, MRM3m, (outs), (ins i16mem:$dst), "fistp{s} $dst">;
def IST_FP32m : FPI<0xDB, MRM3m, (outs), (ins i32mem:$dst), "fistp{l} $dst">;
def IST_FP64m : FPI<0xDF, MRM7m, (outs), (ins i64mem:$dst), "fistp{ll} $dst">;
def LD_F32m : FPI<0xD9, MRM0m, (outs), (ins f32mem:$src), "fld{s}\t$src">;
def LD_F64m : FPI<0xDD, MRM0m, (outs), (ins f64mem:$src), "fld{l}\t$src">;
def ILD_F16m : FPI<0xDF, MRM0m, (outs), (ins i16mem:$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$src">;
def ST_F32m : FPI<0xD9, MRM2m, (outs), (ins f32mem:$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$dst">;
def IST_F16m : FPI<0xDF, MRM2m, (outs), (ins i16mem:$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$dst">;
// FISTTP requires SSE3 even though it's a FPStack op.
def ISTT_Fp16m32 : FpI_<(outs), (ins i16mem:$op, RFP32:$src), OneArgFP,
@ -356,15 +356,15 @@ def ISTT_Fp64m64 : FpI_<(outs), (ins i64mem:$op, RFP64:$src), OneArgFP,
[(X86fp_to_i64mem RFP64:$src, addr:$op)]>,
Requires<[HasSSE3]>;
def ISTT_FP16m : FPI<0xDF, MRM1m, (outs), (ins i16mem:$dst), "fisttp{s} $dst">;
def ISTT_FP32m : FPI<0xDB, MRM1m, (outs), (ins i32mem:$dst), "fisttp{l} $dst">;
def ISTT_FP64m : FPI<0xDD, MRM1m, (outs), (ins i64mem:$dst), "fisttp{ll} $dst">;
def ISTT_FP16m : FPI<0xDF, MRM1m, (outs), (ins i16mem:$dst), "fisttp{s}\t$dst">;
def ISTT_FP32m : FPI<0xDB, MRM1m, (outs), (ins i32mem:$dst), "fisttp{l}\t$dst">;
def ISTT_FP64m : FPI<0xDD, MRM1m, (outs), (ins i64mem:$dst), "fisttp{ll}\t$dst">;
// FP Stack manipulation instructions.
def LD_Frr : FPI<0xC0, AddRegFrm, (outs), (ins RST:$op), "fld $op">, D9;
def ST_Frr : FPI<0xD0, AddRegFrm, (outs), (ins RST:$op), "fst $op">, DD;
def ST_FPrr : FPI<0xD8, AddRegFrm, (outs), (ins RST:$op), "fstp $op">, DD;
def XCH_F : FPI<0xC8, AddRegFrm, (outs), (ins RST:$op), "fxch $op">, D9;
def LD_Frr : FPI<0xC0, AddRegFrm, (outs), (ins RST:$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$op">, D9;
// Floating point constant loads.
let isReMaterializable = 1 in {
@ -394,29 +394,29 @@ def UCOM_FpIr64: FpI<(outs), (ins RFP64:$lhs, RFP64:$rhs), CompareFP,
def UCOM_Fr : FPI<0xE0, AddRegFrm, // FPSW = cmp ST(0) with ST(i)
(outs), (ins RST:$reg),
"fucom $reg">, DD, Imp<[ST0],[]>;
"fucom\t$reg">, DD, Imp<[ST0],[]>;
def UCOM_FPr : FPI<0xE8, AddRegFrm, // FPSW = cmp ST(0) with ST(i), pop
(outs), (ins RST:$reg),
"fucomp $reg">, DD, Imp<[ST0],[]>;
"fucomp\t$reg">, DD, Imp<[ST0],[]>;
def UCOM_FPPr : FPI<0xE9, RawFrm, // cmp ST(0) with ST(1), pop, pop
(outs), (ins),
"fucompp">, DA, Imp<[ST0],[]>;
def UCOM_FIr : FPI<0xE8, AddRegFrm, // CC = cmp ST(0) with ST(i)
(outs), (ins RST:$reg),
"fucomi {$reg, %st(0)|%ST(0), $reg}">, DB, Imp<[ST0],[]>;
"fucomi\t{$reg, %st(0)|%ST(0), $reg}">, DB, Imp<[ST0],[]>;
def UCOM_FIPr : FPI<0xE8, AddRegFrm, // CC = cmp ST(0) with ST(i), pop
(outs), (ins RST:$reg),
"fucomip {$reg, %st(0)|%ST(0), $reg}">, DF, Imp<[ST0],[]>;
"fucomip\t{$reg, %st(0)|%ST(0), $reg}">, DF, Imp<[ST0],[]>;
// Floating point flag ops.
def FNSTSW8r : I<0xE0, RawFrm, // AX = fp flags
(outs), (ins), "fnstsw", []>, DF, Imp<[],[AX]>;
def FNSTCW16m : I<0xD9, MRM7m, // [mem16] = X87 control world
(outs), (ins i16mem:$dst), "fnstcw $dst", []>;
(outs), (ins i16mem:$dst), "fnstcw\t$dst", []>;
def FLDCW16m : I<0xD9, MRM5m, // X87 control world = [mem16]
(outs), (ins i16mem:$dst), "fldcw $dst", []>;
(outs), (ins i16mem:$dst), "fldcw\t$dst", []>;
//===----------------------------------------------------------------------===//
// Non-Instruction Patterns

974
lib/Target/X86/X86InstrInfo.td
View File

File diff suppressed because it is too large Load Diff

112
lib/Target/X86/X86InstrMMX.td
View File

@ -85,12 +85,12 @@ let isTwoAddress = 1 in {
multiclass MMXI_binop_rm<bits<8> opc, string OpcodeStr, SDNode OpNode,
ValueType OpVT, bit Commutable = 0> {
def rr : MMXI<opc, MRMSrcReg, (outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (OpVT (OpNode VR64:$src1, VR64:$src2)))]> {
let isCommutable = Commutable;
}
def rm : MMXI<opc, MRMSrcMem, (outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (OpVT (OpNode VR64:$src1,
(bitconvert
(load_mmx addr:$src2)))))]>;
@ -99,12 +99,12 @@ let isTwoAddress = 1 in {
multiclass MMXI_binop_rm_int<bits<8> opc, string OpcodeStr, Intrinsic IntId,
bit Commutable = 0> {
def rr : MMXI<opc, MRMSrcReg, (outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (IntId VR64:$src1, VR64:$src2))]> {
let isCommutable = Commutable;
}
def rm : MMXI<opc, MRMSrcMem, (outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (IntId VR64:$src1,
(bitconvert (load_mmx addr:$src2))))]>;
}
@ -117,12 +117,12 @@ let isTwoAddress = 1 in {
multiclass MMXI_binop_rm_v1i64<bits<8> opc, string OpcodeStr, SDNode OpNode,
bit Commutable = 0> {
def rr : MMXI<opc, MRMSrcReg, (outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (v1i64 (OpNode VR64:$src1, VR64:$src2)))]> {
let isCommutable = Commutable;
}
def rm : MMXI<opc, MRMSrcMem, (outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst,
(OpNode VR64:$src1,(load_mmx addr:$src2)))]>;
}
@ -130,14 +130,14 @@ let isTwoAddress = 1 in {
multiclass MMXI_binop_rmi_int<bits<8> opc, bits<8> opc2, Format ImmForm,
string OpcodeStr, Intrinsic IntId> {
def rr : MMXI<opc, MRMSrcReg, (outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (IntId VR64:$src1, VR64:$src2))]>;
def rm : MMXI<opc, MRMSrcMem, (outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (IntId VR64:$src1,
(bitconvert (load_mmx addr:$src2))))]>;
def ri : MMXIi8<opc2, ImmForm, (outs VR64:$dst), (ins VR64:$src1, i32i8imm:$src2),
!strconcat(OpcodeStr, " {$src2, $dst|$dst, $src2}"),
!strconcat(OpcodeStr, "\t{$src2, $dst|$dst, $src2}"),
[(set VR64:$dst, (IntId VR64:$src1,
(scalar_to_vector (i32 imm:$src2))))]>;
}
@ -156,50 +156,50 @@ def MMX_FEMMS : MMXI<0x0E, RawFrm, (outs), (ins), "femms", [(int_x86_mmx_femms)]
// Data Transfer Instructions
def MMX_MOVD64rr : MMXI<0x6E, MRMSrcReg, (outs VR64:$dst), (ins GR32:$src),
"movd {$src, $dst|$dst, $src}", []>;
"movd\t{$src, $dst|$dst, $src}", []>;
def MMX_MOVD64rm : MMXI<0x6E, MRMSrcMem, (outs VR64:$dst), (ins i32mem:$src),
"movd {$src, $dst|$dst, $src}", []>;
"movd\t{$src, $dst|$dst, $src}", []>;
def MMX_MOVD64mr : MMXI<0x7E, MRMDestMem, (outs), (ins i32mem:$dst, VR64:$src),
"movd {$src, $dst|$dst, $src}", []>;
"movd\t{$src, $dst|$dst, $src}", []>;
def MMX_MOVD64to64rr : MMXRI<0x6E, MRMSrcReg, (outs VR64:$dst), (ins GR64:$src),
"movd {$src, $dst|$dst, $src}", []>;
"movd\t{$src, $dst|$dst, $src}", []>;
def MMX_MOVQ64rr : MMXI<0x6F, MRMSrcReg, (outs VR64:$dst), (ins VR64:$src),
"movq {$src, $dst|$dst, $src}", []>;
"movq\t{$src, $dst|$dst, $src}", []>;
def MMX_MOVQ64rm : MMXI<0x6F, MRMSrcMem, (outs VR64:$dst), (ins i64mem:$src),
"movq {$src, $dst|$dst, $src}",
"movq\t{$src, $dst|$dst, $src}",
[(set VR64:$dst, (load_mmx addr:$src))]>;
def MMX_MOVQ64mr : MMXI<0x7F, MRMDestMem, (outs), (ins i64mem:$dst, VR64:$src),
"movq {$src, $dst|$dst, $src}",
"movq\t{$src, $dst|$dst, $src}",
[(store (v1i64 VR64:$src), addr:$dst)]>;
def MMX_MOVDQ2Qrr : MMXID<0xD6, MRMDestMem, (outs VR64:$dst), (ins VR128:$src),
"movdq2q {$src, $dst|$dst, $src}",
"movdq2q\t{$src, $dst|$dst, $src}",
[(set VR64:$dst,
(v1i64 (vector_extract (v2i64 VR128:$src),
(iPTR 0))))]>;
def MMX_MOVQ2DQrr : MMXIS<0xD6, MRMDestMem, (outs VR128:$dst), (ins VR64:$src),
"movq2dq {$src, $dst|$dst, $src}",
"movq2dq\t{$src, $dst|$dst, $src}",
[(set VR128:$dst,
(bitconvert (v1i64 VR64:$src)))]>;
def MMX_MOVNTQmr : MMXI<0xE7, MRMDestMem, (outs), (ins i64mem:$dst, VR64:$src),
"movntq {$src, $dst|$dst, $src}",
"movntq\t{$src, $dst|$dst, $src}",
[(int_x86_mmx_movnt_dq addr:$dst, VR64:$src)]>;
let AddedComplexity = 15 in
// movd to MMX register zero-extends
def MMX_MOVZDI2PDIrr : MMX2I<0x6E, MRMSrcReg, (outs VR64:$dst), (ins GR32:$src),
"movd {$src, $dst|$dst, $src}",
"movd\t{$src, $dst|$dst, $src}",
[(set VR64:$dst,
(v2i32 (vector_shuffle immAllZerosV,
(v2i32 (scalar_to_vector GR32:$src)),
MMX_MOVL_shuffle_mask)))]>;
let AddedComplexity = 20 in
def MMX_MOVZDI2PDIrm : MMX2I<0x6E, MRMSrcMem, (outs VR64:$dst), (ins i32mem:$src),
"movd {$src, $dst|$dst, $src}",
"movd\t{$src, $dst|$dst, $src}",
[(set VR64:$dst,
(v2i32 (vector_shuffle immAllZerosV,
(v2i32 (scalar_to_vector
@ -261,12 +261,12 @@ defm MMX_PXOR : MMXI_binop_rm_v1i64<0xEF, "pxor", xor, 1>;
let isTwoAddress = 1 in {
def MMX_PANDNrr : MMXI<0xDF, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"pandn {$src2, $dst|$dst, $src2}",
"pandn\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst, (v1i64 (and (vnot VR64:$src1),
VR64:$src2)))]>;
def MMX_PANDNrm : MMXI<0xDF, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"pandn {$src2, $dst|$dst, $src2}",
"pandn\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst, (v1i64 (and (vnot VR64:$src1),
(load addr:$src2))))]>;
}
@ -307,13 +307,13 @@ let isTwoAddress = 1 in {
// Unpack High Packed Data Instructions
def MMX_PUNPCKHBWrr : MMXI<0x68, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"punpckhbw {$src2, $dst|$dst, $src2}",
"punpckhbw\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v8i8 (vector_shuffle VR64:$src1, VR64:$src2,
MMX_UNPCKH_shuffle_mask)))]>;
def MMX_PUNPCKHBWrm : MMXI<0x68, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"punpckhbw {$src2, $dst|$dst, $src2}",
"punpckhbw\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v8i8 (vector_shuffle VR64:$src1,
(bc_v8i8 (load_mmx addr:$src2)),
@ -321,13 +321,13 @@ let isTwoAddress = 1 in {
def MMX_PUNPCKHWDrr : MMXI<0x69, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"punpckhwd {$src2, $dst|$dst, $src2}",
"punpckhwd\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v4i16 (vector_shuffle VR64:$src1, VR64:$src2,
MMX_UNPCKH_shuffle_mask)))]>;
def MMX_PUNPCKHWDrm : MMXI<0x69, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"punpckhwd {$src2, $dst|$dst, $src2}",
"punpckhwd\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v4i16 (vector_shuffle VR64:$src1,
(bc_v4i16 (load_mmx addr:$src2)),
@ -335,13 +335,13 @@ let isTwoAddress = 1 in {
def MMX_PUNPCKHDQrr : MMXI<0x6A, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"punpckhdq {$src2, $dst|$dst, $src2}",
"punpckhdq\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v2i32 (vector_shuffle VR64:$src1, VR64:$src2,
MMX_UNPCKH_shuffle_mask)))]>;
def MMX_PUNPCKHDQrm : MMXI<0x6A, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"punpckhdq {$src2, $dst|$dst, $src2}",
"punpckhdq\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v2i32 (vector_shuffle VR64:$src1,
(bc_v2i32 (load_mmx addr:$src2)),
@ -350,13 +350,13 @@ let isTwoAddress = 1 in {
// Unpack Low Packed Data Instructions
def MMX_PUNPCKLBWrr : MMXI<0x60, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"punpcklbw {$src2, $dst|$dst, $src2}",
"punpcklbw\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v8i8 (vector_shuffle VR64:$src1, VR64:$src2,
MMX_UNPCKL_shuffle_mask)))]>;
def MMX_PUNPCKLBWrm : MMXI<0x60, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"punpcklbw {$src2, $dst|$dst, $src2}",
"punpcklbw\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v8i8 (vector_shuffle VR64:$src1,
(bc_v8i8 (load_mmx addr:$src2)),
@ -364,13 +364,13 @@ let isTwoAddress = 1 in {
def MMX_PUNPCKLWDrr : MMXI<0x61, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"punpcklwd {$src2, $dst|$dst, $src2}",
"punpcklwd\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v4i16 (vector_shuffle VR64:$src1, VR64:$src2,
MMX_UNPCKL_shuffle_mask)))]>;
def MMX_PUNPCKLWDrm : MMXI<0x61, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"punpcklwd {$src2, $dst|$dst, $src2}",
"punpcklwd\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v4i16 (vector_shuffle VR64:$src1,
(bc_v4i16 (load_mmx addr:$src2)),
@ -378,13 +378,13 @@ let isTwoAddress = 1 in {
def MMX_PUNPCKLDQrr : MMXI<0x62, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, VR64:$src2),
"punpckldq {$src2, $dst|$dst, $src2}",
"punpckldq\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v2i32 (vector_shuffle VR64:$src1, VR64:$src2,
MMX_UNPCKL_shuffle_mask)))]>;
def MMX_PUNPCKLDQrm : MMXI<0x62, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i64mem:$src2),
"punpckldq {$src2, $dst|$dst, $src2}",
"punpckldq\t{$src2, $dst|$dst, $src2}",
[(set VR64:$dst,
(v2i32 (vector_shuffle VR64:$src1,
(bc_v2i32 (load_mmx addr:$src2)),
@ -399,14 +399,14 @@ defm MMX_PACKUSWB : MMXI_binop_rm_int<0x67, "packuswb", int_x86_mmx_packuswb>;
// -- Shuffle Instructions
def MMX_PSHUFWri : MMXIi8<0x70, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, i8imm:$src2),
"pshufw {$src2, $src1, $dst|$dst, $src1, $src2}",
"pshufw\t{$src2, $src1, $dst|$dst, $src1, $src2}",
[(set VR64:$dst,
(v4i16 (vector_shuffle
VR64:$src1, (undef),
MMX_PSHUFW_shuffle_mask:$src2)))]>;
def MMX_PSHUFWmi : MMXIi8<0x70, MRMSrcMem,
(outs VR64:$dst), (ins i64mem:$src1, i8imm:$src2),
"pshufw {$src2, $src1, $dst|$dst, $src1, $src2}",
"pshufw\t{$src2, $src1, $dst|$dst, $src1, $src2}",
[(set VR64:$dst,
(v4i16 (vector_shuffle
(bc_v4i16 (load_mmx addr:$src1)),
@ -415,34 +415,34 @@ def MMX_PSHUFWmi : MMXIi8<0x70, MRMSrcMem,
// -- Conversion Instructions
def MMX_CVTPD2PIrr : MMX2I<0x2D, MRMSrcReg, (outs VR64:$dst), (ins VR128:$src),
"cvtpd2pi {$src, $dst|$dst, $src}", []>;
"cvtpd2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPD2PIrm : MMX2I<0x2D, MRMSrcMem, (outs VR64:$dst), (ins f128mem:$src),
"cvtpd2pi {$src, $dst|$dst, $src}", []>;
"cvtpd2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPI2PDrr : MMX2I<0x2A, MRMSrcReg, (outs VR128:$dst), (ins VR64:$src),
"cvtpi2pd {$src, $dst|$dst, $src}", []>;
"cvtpi2pd\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPI2PDrm : MMX2I<0x2A, MRMSrcMem, (outs VR128:$dst), (ins i64mem:$src),
"cvtpi2pd {$src, $dst|$dst, $src}", []>;
"cvtpi2pd\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPI2PSrr : MMXI<0x2A, MRMSrcReg, (outs VR128:$dst), (ins VR64:$src),
"cvtpi2ps {$src, $dst|$dst, $src}", []>;
"cvtpi2ps\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPI2PSrm : MMXI<0x2A, MRMSrcMem, (outs VR128:$dst), (ins i64mem:$src),
"cvtpi2ps {$src, $dst|$dst, $src}", []>;
"cvtpi2ps\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPS2PIrr : MMXI<0x2D, MRMSrcReg, (outs VR64:$dst), (ins VR128:$src),
"cvtps2pi {$src, $dst|$dst, $src}", []>;
"cvtps2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTPS2PIrm : MMXI<0x2D, MRMSrcMem, (outs VR64:$dst), (ins f64mem:$src),
"cvtps2pi {$src, $dst|$dst, $src}", []>;
"cvtps2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTTPD2PIrr : MMX2I<0x2C, MRMSrcReg, (outs VR64:$dst), (ins VR128:$src),
"cvttpd2pi {$src, $dst|$dst, $src}", []>;
"cvttpd2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTTPD2PIrm : MMX2I<0x2C, MRMSrcMem, (outs VR64:$dst), (ins f128mem:$src),
"cvttpd2pi {$src, $dst|$dst, $src}", []>;
"cvttpd2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTTPS2PIrr : MMXI<0x2C, MRMSrcReg, (outs VR64:$dst), (ins VR128:$src),
"cvttps2pi {$src, $dst|$dst, $src}", []>;
"cvttps2pi\t{$src, $dst|$dst, $src}", []>;
def MMX_CVTTPS2PIrm : MMXI<0x2C, MRMSrcMem, (outs VR64:$dst), (ins f64mem:$src),
"cvttps2pi {$src, $dst|$dst, $src}", []>;
"cvttps2pi\t{$src, $dst|$dst, $src}", []>;
// Extract / Insert
def MMX_X86pextrw : SDNode<"X86ISD::PEXTRW", SDTypeProfile<1, 2, []>, []>;
@ -450,18 +450,18 @@ def MMX_X86pinsrw : SDNode<"X86ISD::PINSRW", SDTypeProfile<1, 3, []>, []>;
def MMX_PEXTRWri : MMXIi8<0xC5, MRMSrcReg,
(outs GR32:$dst), (ins VR64:$src1, i16i8imm:$src2),
"pextrw {$src2, $src1, $dst|$dst, $src1, $src2}",
"pextrw\t{$src2, $src1, $dst|$dst, $src1, $src2}",
[(set GR32:$dst, (MMX_X86pextrw (v4i16 VR64:$src1),
(iPTR imm:$src2)))]>;
let isTwoAddress = 1 in {
def MMX_PINSRWrri : MMXIi8<0xC4, MRMSrcReg,
(outs VR64:$dst), (ins VR64:$src1, GR32:$src2, i16i8imm:$src3),
"pinsrw {$src3, $src2, $dst|$dst, $src2, $src3}",
"pinsrw\t{$src3, $src2, $dst|$dst, $src2, $src3}",
[(set VR64:$dst, (v4i16 (MMX_X86pinsrw (v4i16 VR64:$src1),
GR32:$src2, (iPTR imm:$src3))))]>;
def MMX_PINSRWrmi : MMXIi8<0xC4, MRMSrcMem,
(outs VR64:$dst), (ins VR64:$src1, i16mem:$src2, i16i8imm:$src3),
"pinsrw {$src3, $src2, $dst|$dst, $src2, $src3}",
"pinsrw\t{$src3, $src2, $dst|$dst, $src2, $src3}",
[(set VR64:$dst,
(v4i16 (MMX_X86pinsrw (v4i16 VR64:$src1),
(i32 (anyext (loadi16 addr:$src2))),
@ -470,12 +470,12 @@ let isTwoAddress = 1 in {
// Mask creation
def MMX_PMOVMSKBrr : MMXI<0xD7, MRMSrcReg, (outs GR32:$dst), (ins VR64:$src),
"pmovmskb {$src, $dst|$dst, $src}",
"pmovmskb\t{$src, $dst|$dst, $src}",
[(set GR32:$dst, (int_x86_mmx_pmovmskb VR64:$src))]>;
// Misc.
def MMX_MASKMOVQ : MMXI<0xF7, MRMDestMem, (outs), (ins VR64:$src, VR64:$mask),
"maskmovq {$mask, $src|$src, $mask}",
"maskmovq\t{$mask, $src|$src, $mask}",
[(int_x86_mmx_maskmovq VR64:$src, VR64:$mask, EDI)]>,
Imp<[EDI],[]>;
@ -487,10 +487,10 @@ def MMX_MASKMOVQ : MMXI<0xF7, MRMDestMem, (outs), (ins VR64:$src, VR64:$mask),
// FIXME: remove when we can teach regalloc that xor reg, reg is ok.
let isReMaterializable = 1 in {
def MMX_V_SET0 : MMXI<0xEF, MRMInitReg, (outs VR64:$dst), (ins),
"pxor $dst, $dst",
"pxor\t$dst, $dst",
[(set VR64:$dst, (v1i64 immAllZerosV))]>;
def MMX_V_SETALLONES : MMXI<0x76, MRMInitReg, (outs VR64:$dst), (ins),
"pcmpeqd $dst, $dst",
"pcmpeqd\t$dst, $dst",
[(set VR64:$dst, (v1i64 immAllOnesV))]>;
}

604
lib/Target/X86/X86InstrSSE.td
View File

File diff suppressed because it is too large Load Diff

430
lib/Target/X86/X86InstrX86-64.td
View File

File diff suppressed because it is too large Load Diff

2
test/CodeGen/X86/commute-two-addr.ll
View File

@ -3,7 +3,7 @@
; Make sure there are only 3 mov's for each testcase
; RUN: llvm-upgrade < %s | llvm-as | llc -march=x86 -x86-asm-syntax=intel | \
; RUN: grep {mov } | wc -l | grep 6
; RUN: grep {\\\<mov\\\>} | wc -l | grep 6
target triple = "i686-pc-linux-gnu"

2
test/Transforms/TailDup/if-tail-dup.ll
View File

@ -1,6 +1,6 @@
; RUN: llvm-upgrade < %s | llvm-as | opt -tailduplicate | \
; RUN: llc -march=x86 -o %t -f
; RUN: grep {je } %t
; RUN: grep {\\\<je\\\>} %t
; RUN: not grep jmp %t
; END.
; This should have no unconditional jumps in it. The C source is: