[8/11][POC][Clang][RISCV] Define tuple type variant of vloxseg2ei32 v…

…luxseg2ei32

For the cover letter of this patch-set, please checkout D146872.

Depends on D147913.

This is the 8th patch of the patch-set.

This patch is a proof-of-concept and will be extended to full coverage
in the future. Currently, the old non-tuple indexed segment load is
not removed, and only signed integer indexed segment load of NF=2,
EEW=32 is defined here.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D147914

clang/include/clang/Basic/riscv_vector.td

@@ -1744,13 +1744,66 @@ multiclass RVVStridedSegStoreTuple<string op> {
   }
 }
 
+multiclass RVVIndexedSegLoadTuple<string op> {
+  foreach type = ["i"] in {
+    foreach eew_info = [["32", "(Log2EEW:5)"]] in {
+      defvar eew = eew_info[0];
+      defvar eew_type = eew_info[1];
+      foreach nf = [2] in {
+        let Name = op # nf # "ei" # eew # "_v_tuple",
+            OverloadedName = op # nf # "ei" # eew # "_tuple",
+            IRName = op # nf,
+            MaskedIRName = op # nf # "_mask",
+            NF = nf,
+            ManualCodegen = [{
+    {
+      assert(((IsMasked && (PolicyAttrs & RVV_VTA) && (PolicyAttrs & RVV_VMA)) ||
+             (!IsMasked && (PolicyAttrs & RVV_VTA))) &&
+             "FIXME: Only handling default policy (TAMA) for now");
+
+      llvm::Type *ElementVectorType = cast<StructType>(ResultType)->elements()[0];
+
+      SmallVector<llvm::Value*, 12> Operands;
+
+      Operands.append(NF, llvm::PoisonValue::get(ElementVectorType));
+
+      unsigned Offset = IsMasked ? 1 : 0;
+      Operands.push_back(Ops[Offset]); // Ptr
+      Operands.push_back(Ops[Offset + 1]); // Idx
+      if (IsMasked)
+        Operands.push_back(Ops[0]);
+      Operands.push_back(Ops[Offset + 2]); // VL
+      if (IsMasked)
+        Operands.push_back(ConstantInt::get(Ops.back()->getType(), PolicyAttrs));
+
+      IntrinsicTypes = {ElementVectorType, Ops[Offset + 1]->getType(),
+                        Ops.back()->getType()};
+      llvm::Function *F = CGM.getIntrinsic(ID, IntrinsicTypes);
+      llvm::Value *LoadValue = Builder.CreateCall(F, Operands, "");
+
+      if (ReturnValue.isNull())
+        return LoadValue;
+      else
+        return Builder.CreateStore(LoadValue, ReturnValue.getValue());
+    }
+    }] in {
+          defvar T = "(Tuple:" # nf # ")";
+          def : RVVBuiltin<"v", T # "vPCe" # eew_type # "Uv", type>;
+        }
+      }
+    }
+  }
+}
+
 // TODO: Extend for policy
 let UnMaskedPolicyScheme = NonePolicy,
     MaskedPolicyScheme = NonePolicy,
     IsTuple = true in {
 defm : RVVUnitStridedSegLoadTuple<"vlseg">;
 defm : RVVUnitStridedSegLoadFFTuple<"vlseg">;
 defm : RVVStridedSegLoadTuple<"vlsseg">;
+defm : RVVIndexedSegLoadTuple<"vluxseg">;
+defm : RVVIndexedSegLoadTuple<"vloxseg">;
 }
 
 let UnMaskedPolicyScheme = NonePolicy,

clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/vloxseg2ei32_tuple.c

@@ -0,0 +1,28 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 2
+// REQUIRES: riscv-registered-target
+// RUN: %clang_cc1 -triple riscv64 -target-feature +v -target-feature +zfh \
+// RUN:   -target-feature +experimental-zvfh -disable-O0-optnone  \
+// RUN:   -emit-llvm %s -o - | opt -S -passes=mem2reg | \
+// RUN:   FileCheck --check-prefix=CHECK-RV64 %s
+
+#include <riscv_vector.h>
+
+// CHECK-RV64-LABEL: define dso_local { <vscale x 2 x i32>, <vscale x 2 x i32> } @test_vloxseg2ei32_v_tuple_i32m1
+// CHECK-RV64-SAME: (ptr noundef [[BASE:%.*]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call { <vscale x 2 x i32>, <vscale x 2 x i32> } @llvm.riscv.vloxseg2.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> poison, <vscale x 2 x i32> poison, ptr [[BASE]], <vscale x 2 x i32> [[BINDEX]], i64 [[VL]])
+// CHECK-RV64-NEXT:    ret { <vscale x 2 x i32>, <vscale x 2 x i32> } [[TMP0]]
+//
+vint32m1x2_t test_vloxseg2ei32_v_tuple_i32m1(const int32_t *base, vuint32m1_t bindex, size_t vl) {
+  return __riscv_vloxseg2ei32_v_tuple_i32m1(base, bindex, vl);
+}
+
+// CHECK-RV64-LABEL: define dso_local { <vscale x 2 x i32>, <vscale x 2 x i32> } @test_vloxseg2ei32_v_tuple_i32m1_m
+// CHECK-RV64-SAME: (<vscale x 2 x i1> [[MASK:%.*]], ptr noundef [[BASE:%.*]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call { <vscale x 2 x i32>, <vscale x 2 x i32> } @llvm.riscv.vloxseg2.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> poison, <vscale x 2 x i32> poison, ptr [[BASE]], <vscale x 2 x i32> [[BINDEX]], <vscale x 2 x i1> [[MASK]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    ret { <vscale x 2 x i32>, <vscale x 2 x i32> } [[TMP0]]
+//
+vint32m1x2_t test_vloxseg2ei32_v_tuple_i32m1_m(vbool32_t mask, const int32_t *base, vuint32m1_t bindex, size_t vl) {
+  return __riscv_vloxseg2ei32_v_tuple_i32m1_m(mask, base, bindex, vl);
+}

clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/vluxseg2ei32_tuple.c

@@ -0,0 +1,28 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 2
+// REQUIRES: riscv-registered-target
+// RUN: %clang_cc1 -triple riscv64 -target-feature +v -target-feature +zfh \
+// RUN:   -target-feature +experimental-zvfh -disable-O0-optnone  \
+// RUN:   -emit-llvm %s -o - | opt -S -passes=mem2reg | \
+// RUN:   FileCheck --check-prefix=CHECK-RV64 %s
+
+#include <riscv_vector.h>
+
+// CHECK-RV64-LABEL: define dso_local { <vscale x 2 x i32>, <vscale x 2 x i32> } @test_vluxseg2ei32_v_tuple_i32m1
+// CHECK-RV64-SAME: (ptr noundef [[BASE:%.*]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call { <vscale x 2 x i32>, <vscale x 2 x i32> } @llvm.riscv.vluxseg2.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> poison, <vscale x 2 x i32> poison, ptr [[BASE]], <vscale x 2 x i32> [[BINDEX]], i64 [[VL]])
+// CHECK-RV64-NEXT:    ret { <vscale x 2 x i32>, <vscale x 2 x i32> } [[TMP0]]
+//
+vint32m1x2_t test_vluxseg2ei32_v_tuple_i32m1(const int32_t *base, vuint32m1_t bindex, size_t vl) {
+  return __riscv_vluxseg2ei32_v_tuple_i32m1(base, bindex, vl);
+}
+
+// CHECK-RV64-LABEL: define dso_local { <vscale x 2 x i32>, <vscale x 2 x i32> } @test_vluxseg2ei32_v_tuple_i32m1_m
+// CHECK-RV64-SAME: (<vscale x 2 x i1> [[MASK:%.*]], ptr noundef [[BASE:%.*]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call { <vscale x 2 x i32>, <vscale x 2 x i32> } @llvm.riscv.vluxseg2.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> poison, <vscale x 2 x i32> poison, ptr [[BASE]], <vscale x 2 x i32> [[BINDEX]], <vscale x 2 x i1> [[MASK]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    ret { <vscale x 2 x i32>, <vscale x 2 x i32> } [[TMP0]]
+//
+vint32m1x2_t test_vluxseg2ei32_v_tuple_i32m1_m(vbool32_t mask, const int32_t *base, vuint32m1_t bindex, size_t vl) {
+  return __riscv_vluxseg2ei32_v_tuple_i32m1_m(mask, base, bindex, vl);
+}