[mips][msa] Added support for matching mod_[us] from normal IR (i.e. not intrinsics)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191737 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/Mips/MipsMSAInstrInfo.td b/lib/Target/Mips/MipsMSAInstrInfo.td
index 88687443bf1d9b6b2a421a5048ca6137c69106de..48c204af92e3cd7b22ce4419a6b0aa575bc8b1c7 100644 (file)
--- a/lib/Target/Mips/MipsMSAInstrInfo.td
+++ b/lib/Target/Mips/MipsMSAInstrInfo.td
@@ -2102,15 +2102,15 @@ class MINI_U_W_DESC : MSA_I5_DESC_BASE<"mini_u.w", MipsVUMin, vsplati32_uimm5,
 class MINI_U_D_DESC : MSA_I5_DESC_BASE<"mini_u.d", MipsVUMin, vsplati64_uimm5,
                                        MSA128DOpnd>;
 
-class MOD_S_B_DESC : MSA_3R_DESC_BASE<"mod_s.b", int_mips_mod_s_b, MSA128BOpnd>;
-class MOD_S_H_DESC : MSA_3R_DESC_BASE<"mod_s.h", int_mips_mod_s_h, MSA128HOpnd>;
-class MOD_S_W_DESC : MSA_3R_DESC_BASE<"mod_s.w", int_mips_mod_s_w, MSA128WOpnd>;
-class MOD_S_D_DESC : MSA_3R_DESC_BASE<"mod_s.d", int_mips_mod_s_d, MSA128DOpnd>;
-
-class MOD_U_B_DESC : MSA_3R_DESC_BASE<"mod_u.b", int_mips_mod_u_b, MSA128BOpnd>;
-class MOD_U_H_DESC : MSA_3R_DESC_BASE<"mod_u.h", int_mips_mod_u_h, MSA128HOpnd>;
-class MOD_U_W_DESC : MSA_3R_DESC_BASE<"mod_u.w", int_mips_mod_u_w, MSA128WOpnd>;
-class MOD_U_D_DESC : MSA_3R_DESC_BASE<"mod_u.d", int_mips_mod_u_d, MSA128DOpnd>;
+class MOD_S_B_DESC : MSA_3R_DESC_BASE<"mod_s.b", srem, MSA128BOpnd>;
+class MOD_S_H_DESC : MSA_3R_DESC_BASE<"mod_s.h", srem, MSA128HOpnd>;
+class MOD_S_W_DESC : MSA_3R_DESC_BASE<"mod_s.w", srem, MSA128WOpnd>;
+class MOD_S_D_DESC : MSA_3R_DESC_BASE<"mod_s.d", srem, MSA128DOpnd>;
+
+class MOD_U_B_DESC : MSA_3R_DESC_BASE<"mod_u.b", urem, MSA128BOpnd>;
+class MOD_U_H_DESC : MSA_3R_DESC_BASE<"mod_u.h", urem, MSA128HOpnd>;
+class MOD_U_W_DESC : MSA_3R_DESC_BASE<"mod_u.w", urem, MSA128WOpnd>;
+class MOD_U_D_DESC : MSA_3R_DESC_BASE<"mod_u.d", urem, MSA128DOpnd>;
 
 class MOVE_V_DESC {
   dag OutOperandList = (outs MSA128B:$wd);

diff --git a/lib/Target/Mips/MipsSEISelLowering.cpp b/lib/Target/Mips/MipsSEISelLowering.cpp
index 52509b72d1b3ad86f447751c8de9a9eaef7bc77a..82cb7a84935202f9e53dfbefaff59407258fb328 100644 (file)
--- a/lib/Target/Mips/MipsSEISelLowering.cpp
+++ b/lib/Target/Mips/MipsSEISelLowering.cpp
@@ -177,11 +177,13 @@ addMSAIntType(MVT::SimpleValueType Ty, const TargetRegisterClass *RC) {
   setOperationAction(ISD::MUL, Ty, Legal);
   setOperationAction(ISD::OR, Ty, Legal);
   setOperationAction(ISD::SDIV, Ty, Legal);
+  setOperationAction(ISD::SREM, Ty, Legal);
   setOperationAction(ISD::SHL, Ty, Legal);
   setOperationAction(ISD::SRA, Ty, Legal);
   setOperationAction(ISD::SRL, Ty, Legal);
   setOperationAction(ISD::SUB, Ty, Legal);
   setOperationAction(ISD::UDIV, Ty, Legal);
+  setOperationAction(ISD::UREM, Ty, Legal);
   setOperationAction(ISD::VECTOR_SHUFFLE, Ty, Custom);
   setOperationAction(ISD::VSELECT, Ty, Legal);
   setOperationAction(ISD::XOR, Ty, Legal);
@@ -1424,6 +1426,18 @@ SDValue MipsSETargetLowering::lowerINTRINSIC_WO_CHAIN(SDValue Op,
   case Intrinsic::mips_mini_u_d:
     return DAG.getNode(MipsISD::VUMIN, DL, Op->getValueType(0),
                        Op->getOperand(1), lowerMSASplatImm(Op, 2, DAG));
+  case Intrinsic::mips_mod_s_b:
+  case Intrinsic::mips_mod_s_h:
+  case Intrinsic::mips_mod_s_w:
+  case Intrinsic::mips_mod_s_d:
+    return DAG.getNode(ISD::SREM, DL, Op->getValueType(0), Op->getOperand(1),
+                       Op->getOperand(2));
+  case Intrinsic::mips_mod_u_b:
+  case Intrinsic::mips_mod_u_h:
+  case Intrinsic::mips_mod_u_w:
+  case Intrinsic::mips_mod_u_d:
+    return DAG.getNode(ISD::UREM, DL, Op->getValueType(0), Op->getOperand(1),
+                       Op->getOperand(2));
   case Intrinsic::mips_mulv_b:
   case Intrinsic::mips_mulv_h:
   case Intrinsic::mips_mulv_w:

diff --git a/test/CodeGen/Mips/msa/arithmetic.ll b/test/CodeGen/Mips/msa/arithmetic.ll
index dc655e184abf6584eede41aa2fc88a0c0cf5fe58..9f290e29a5526965e184443ac046fb58a50a4f9d 100644 (file)
--- a/test/CodeGen/Mips/msa/arithmetic.ll
+++ b/test/CodeGen/Mips/msa/arithmetic.ll
@@ -435,3 +435,131 @@ define void @div_u_v2i64(<2 x i64>* %c, <2 x i64>* %a, <2 x i64>* %b) nounwind {
   ret void
   ; CHECK: .size div_u_v2i64
 }
+
+define void @mod_s_v16i8(<16 x i8>* %c, <16 x i8>* %a, <16 x i8>* %b) nounwind {
+  ; CHECK: mod_s_v16i8:
+
+  %1 = load <16 x i8>* %a
+  ; CHECK-DAG: ld.b [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <16 x i8>* %b
+  ; CHECK-DAG: ld.b [[R2:\$w[0-9]+]], 0($6)
+  %3 = srem <16 x i8> %1, %2
+  ; CHECK-DAG: mod_s.b [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <16 x i8> %3, <16 x i8>* %c
+  ; CHECK-DAG: st.b [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_s_v16i8
+}
+
+define void @mod_s_v8i16(<8 x i16>* %c, <8 x i16>* %a, <8 x i16>* %b) nounwind {
+  ; CHECK: mod_s_v8i16:
+
+  %1 = load <8 x i16>* %a
+  ; CHECK-DAG: ld.h [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <8 x i16>* %b
+  ; CHECK-DAG: ld.h [[R2:\$w[0-9]+]], 0($6)
+  %3 = srem <8 x i16> %1, %2
+  ; CHECK-DAG: mod_s.h [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <8 x i16> %3, <8 x i16>* %c
+  ; CHECK-DAG: st.h [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_s_v8i16
+}
+
+define void @mod_s_v4i32(<4 x i32>* %c, <4 x i32>* %a, <4 x i32>* %b) nounwind {
+  ; CHECK: mod_s_v4i32:
+
+  %1 = load <4 x i32>* %a
+  ; CHECK-DAG: ld.w [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <4 x i32>* %b
+  ; CHECK-DAG: ld.w [[R2:\$w[0-9]+]], 0($6)
+  %3 = srem <4 x i32> %1, %2
+  ; CHECK-DAG: mod_s.w [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <4 x i32> %3, <4 x i32>* %c
+  ; CHECK-DAG: st.w [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_s_v4i32
+}
+
+define void @mod_s_v2i64(<2 x i64>* %c, <2 x i64>* %a, <2 x i64>* %b) nounwind {
+  ; CHECK: mod_s_v2i64:
+
+  %1 = load <2 x i64>* %a
+  ; CHECK-DAG: ld.d [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <2 x i64>* %b
+  ; CHECK-DAG: ld.d [[R2:\$w[0-9]+]], 0($6)
+  %3 = srem <2 x i64> %1, %2
+  ; CHECK-DAG: mod_s.d [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <2 x i64> %3, <2 x i64>* %c
+  ; CHECK-DAG: st.d [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_s_v2i64
+}
+
+define void @mod_u_v16i8(<16 x i8>* %c, <16 x i8>* %a, <16 x i8>* %b) nounwind {
+  ; CHECK: mod_u_v16i8:
+
+  %1 = load <16 x i8>* %a
+  ; CHECK-DAG: ld.b [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <16 x i8>* %b
+  ; CHECK-DAG: ld.b [[R2:\$w[0-9]+]], 0($6)
+  %3 = urem <16 x i8> %1, %2
+  ; CHECK-DAG: mod_u.b [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <16 x i8> %3, <16 x i8>* %c
+  ; CHECK-DAG: st.b [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_u_v16i8
+}
+
+define void @mod_u_v8i16(<8 x i16>* %c, <8 x i16>* %a, <8 x i16>* %b) nounwind {
+  ; CHECK: mod_u_v8i16:
+
+  %1 = load <8 x i16>* %a
+  ; CHECK-DAG: ld.h [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <8 x i16>* %b
+  ; CHECK-DAG: ld.h [[R2:\$w[0-9]+]], 0($6)
+  %3 = urem <8 x i16> %1, %2
+  ; CHECK-DAG: mod_u.h [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <8 x i16> %3, <8 x i16>* %c
+  ; CHECK-DAG: st.h [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_u_v8i16
+}
+
+define void @mod_u_v4i32(<4 x i32>* %c, <4 x i32>* %a, <4 x i32>* %b) nounwind {
+  ; CHECK: mod_u_v4i32:
+
+  %1 = load <4 x i32>* %a
+  ; CHECK-DAG: ld.w [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <4 x i32>* %b
+  ; CHECK-DAG: ld.w [[R2:\$w[0-9]+]], 0($6)
+  %3 = urem <4 x i32> %1, %2
+  ; CHECK-DAG: mod_u.w [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <4 x i32> %3, <4 x i32>* %c
+  ; CHECK-DAG: st.w [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_u_v4i32
+}
+
+define void @mod_u_v2i64(<2 x i64>* %c, <2 x i64>* %a, <2 x i64>* %b) nounwind {
+  ; CHECK: mod_u_v2i64:
+
+  %1 = load <2 x i64>* %a
+  ; CHECK-DAG: ld.d [[R1:\$w[0-9]+]], 0($5)
+  %2 = load <2 x i64>* %b
+  ; CHECK-DAG: ld.d [[R2:\$w[0-9]+]], 0($6)
+  %3 = urem <2 x i64> %1, %2
+  ; CHECK-DAG: mod_u.d [[R3:\$w[0-9]+]], [[R1]], [[R2]]
+  store <2 x i64> %3, <2 x i64>* %c
+  ; CHECK-DAG: st.d [[R3]], 0($4)
+
+  ret void
+  ; CHECK: .size mod_u_v2i64
+}