From 5258b619667c54d3f07c12031fa0d75595a25527 Mon Sep 17 00:00:00 2001
From: Owen Anderson <resistor@mac.com>
Date: Mon, 25 Oct 2010 21:29:04 +0000
Subject: [PATCH] Add correct encodings for NEON vabal.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@117315 91177308-0d34-0410-b5e6-96231b3b80d8
---
 lib/Target/ARM/ARMInstrNEON.td       | 10 ++--
 test/MC/ARM/neon-absdiff-encoding.ll | 72 ++++++++++++++++++++++++++++
 2 files changed, 77 insertions(+), 5 deletions(-)

diff --git a/lib/Target/ARM/ARMInstrNEON.td b/lib/Target/ARM/ARMInstrNEON.td
index 11296bc9f9e..a38b482fcc0 100644
--- a/lib/Target/ARM/ARMInstrNEON.td
+++ b/lib/Target/ARM/ARMInstrNEON.td
@@ -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.
diff --git a/test/MC/ARM/neon-absdiff-encoding.ll b/test/MC/ARM/neon-absdiff-encoding.ll
index 4f0e64a819b..f0adf39467e 100644
--- a/test/MC/ARM/neon-absdiff-encoding.ll
+++ b/test/MC/ARM/neon-absdiff-encoding.ll
@@ -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
+}