maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

Add correct encodings for NEON vabal. 

llvm-svn: 117315
This commit is contained in:
Owen Anderson 2010-10-25 21:29:04 +00:00
parent 82b2e0da9c
commit 1f6aad053d
2 changed files with 77 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
llvm/lib/Target/ARM/ARMInstrNEON.td
View File

@ -1483,11 +1483,11 @@ class N3VLIntExtOp<bit op24, bit op23, bits<2> op21_20, bits<4> op11_8, bit op4,
ValueType TyQ, ValueType TyD, Intrinsic IntOp, SDNode ExtOp,
SDNode OpNode>
: N3V<op24, op23, op21_20, op11_8, 0, op4,
(outs QPR:$dst), (ins QPR:$src1, DPR:$src2, DPR:$src3), N3RegFrm, itin,
OpcodeStr, Dt, "$dst, $src2, $src3", "$src1 = $dst",
[(set QPR:$dst, (OpNode (TyQ QPR:$src1),
(TyQ (ExtOp (TyD (IntOp (TyD DPR:$src2),
(TyD DPR:$src3)))))))]>;
(outs QPR:$Vd), (ins QPR:$src1, DPR:$Vn, DPR:$Vm), N3RegFrm, itin,
OpcodeStr, Dt, "$Vd, $Vn, $Vm", "$src1 = $Vd",
[(set QPR:$Vd, (OpNode (TyQ QPR:$src1),
(TyQ (ExtOp (TyD (IntOp (TyD DPR:$Vn),
(TyD DPR:$Vm)))))))]>;
// Neon Long 3-argument intrinsic. The destination register is
// a quad-register and is also used as the first source operand register.

72
llvm/test/MC/ARM/neon-absdiff-encoding.ll
View File

@ -336,3 +336,75 @@ define <4 x i32> @vabau_4xi32(<4 x i32>* %A, <4 x i32>* %B, <4 x i32>* %C) nounw
%tmp5 = add <4 x i32> %tmp1, %tmp4
ret <4 x i32> %tmp5
}
; CHECK: vabals_8xi8
define <8 x i16> @vabals_8xi8(<8 x i16>* %A, <8 x i8>* %B, <8 x i8>* %C) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = load <8 x i8>* %B
%tmp3 = load <8 x i8>* %C
; CHECK: vabal.s8 q8, d19, d18 @ encoding: [0xa2,0x05,0xc3,0xf2]
%tmp4 = call <8 x i8> @llvm.arm.neon.vabds.v8i8(<8 x i8> %tmp2, <8 x i8> %tmp3)
%tmp5 = zext <8 x i8> %tmp4 to <8 x i16>
%tmp6 = add <8 x i16> %tmp1, %tmp5
ret <8 x i16> %tmp6
}
; CHECK: vabals_4xi16
define <4 x i32> @vabals_4xi16(<4 x i32>* %A, <4 x i16>* %B, <4 x i16>* %C) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = load <4 x i16>* %B
%tmp3 = load <4 x i16>* %C
; CHECK: vabal.s16 q8, d19, d18 @ encoding: [0xa2,0x05,0xd3,0xf2]
%tmp4 = call <4 x i16> @llvm.arm.neon.vabds.v4i16(<4 x i16> %tmp2, <4 x i16> %tmp3)
%tmp5 = zext <4 x i16> %tmp4 to <4 x i32>
%tmp6 = add <4 x i32> %tmp1, %tmp5
ret <4 x i32> %tmp6
}
; CHECK: vabals_2xi32
define <2 x i64> @vabals_2xi32(<2 x i64>* %A, <2 x i32>* %B, <2 x i32>* %C) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = load <2 x i32>* %B
%tmp3 = load <2 x i32>* %C
; CHECK: vabal.s32 q8, d19, d18 @ encoding: [0xa2,0x05,0xe3,0xf2]
%tmp4 = call <2 x i32> @llvm.arm.neon.vabds.v2i32(<2 x i32> %tmp2, <2 x i32> %tmp3)
%tmp5 = zext <2 x i32> %tmp4 to <2 x i64>
%tmp6 = add <2 x i64> %tmp1, %tmp5
ret <2 x i64> %tmp6
}
; CHECK: vabalu_8xi8
define <8 x i16> @vabalu_8xi8(<8 x i16>* %A, <8 x i8>* %B, <8 x i8>* %C) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = load <8 x i8>* %B
%tmp3 = load <8 x i8>* %C
; CHECK: vabal.u8 q8, d19, d18 @ encoding: [0xa2,0x05,0xc3,0xf3]
%tmp4 = call <8 x i8> @llvm.arm.neon.vabdu.v8i8(<8 x i8> %tmp2, <8 x i8> %tmp3)
%tmp5 = zext <8 x i8> %tmp4 to <8 x i16>
%tmp6 = add <8 x i16> %tmp1, %tmp5
ret <8 x i16> %tmp6
}
; CHECK: vabalu_4xi16
define <4 x i32> @vabalu_4xi16(<4 x i32>* %A, <4 x i16>* %B, <4 x i16>* %C) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = load <4 x i16>* %B
%tmp3 = load <4 x i16>* %C
; CHECK: vabal.u16 q8, d19, d18 @ encoding: [0xa2,0x05,0xd3,0xf3]
%tmp4 = call <4 x i16> @llvm.arm.neon.vabdu.v4i16(<4 x i16> %tmp2, <4 x i16> %tmp3)
%tmp5 = zext <4 x i16> %tmp4 to <4 x i32>
%tmp6 = add <4 x i32> %tmp1, %tmp5
ret <4 x i32> %tmp6
}
; CHECK: vabalu_2xi32
define <2 x i64> @vabalu_2xi32(<2 x i64>* %A, <2 x i32>* %B, <2 x i32>* %C) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = load <2 x i32>* %B
%tmp3 = load <2 x i32>* %C
; CHECK: vabal.u32 q8, d19, d18 @ encoding: [0xa2,0x05,0xe3,0xf3]
%tmp4 = call <2 x i32> @llvm.arm.neon.vabdu.v2i32(<2 x i32> %tmp2, <2 x i32> %tmp3)
%tmp5 = zext <2 x i32> %tmp4 to <2 x i64>
%tmp6 = add <2 x i64> %tmp1, %tmp5
ret <2 x i64> %tmp6
}