[SVE] Add folds for truncation of vscale

Reviewed By: david-arm

Differential Revision: https://reviews.llvm.org/D107453

Author: DylanFleming-arm

llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp

@@ -961,6 +961,20 @@ Instruction *InstCombinerImpl::visitTrunc(TruncInst &Trunc) {
       return BinaryOperator::CreateAdd(NarrowCtlz, WidthDiff);
     }
   }
+
+  if (match(Src, m_VScale(DL))) {
+    if (Trunc.getFunction()->hasFnAttribute(Attribute::VScaleRange)) {
+      unsigned MaxVScale = Trunc.getFunction()
+                               ->getFnAttribute(Attribute::VScaleRange)
+                               .getVScaleRangeArgs()
+                               .second;
+      if (MaxVScale > 0 && Log2_32(MaxVScale) < DestWidth) {
+        Value *VScale = Builder.CreateVScale(ConstantInt::get(DestTy, 1));
+        return replaceInstUsesWith(Trunc, VScale);
+      }
+    }
+  }
+
   return nullptr;
 }
 

llvm/test/Transforms/InstCombine/vscale_trunc.ll

@@ -0,0 +1,40 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -instcombine -S | FileCheck %s
+
+define i8 @vscale_trunc_i32toi8() vscale_range(0, 255) {
+; CHECK-LABEL: @vscale_trunc_i32toi8(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[TMP0:%.*]] = call i8 @llvm.vscale.i8()
+; CHECK-NEXT:    ret i8 [[TMP0]]
+entry:
+  %0 = call i32 @llvm.vscale.i32()
+  %1 = trunc i32 %0 to i8
+  ret i8 %1
+}
+
+
+define i8 @vscale_trunc_i32toi8_poison() vscale_range(0, 256) {
+; CHECK-LABEL: @vscale_trunc_i32toi8_poison(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.vscale.i32()
+; CHECK-NEXT:    [[TMP1:%.*]] = trunc i32 [[TMP0]] to i8
+; CHECK-NEXT:    ret i8 [[TMP1]]
+  entry:
+  %0 = call i32 @llvm.vscale.i32()
+  %1 = trunc i32 %0 to i8
+  ret i8 %1
+}
+
+define i8 @vscale_trunc_i32toi8_noAttr() {
+; CHECK-LABEL: @vscale_trunc_i32toi8_noAttr(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.vscale.i32()
+; CHECK-NEXT:    [[TMP1:%.*]] = trunc i32 [[TMP0]] to i8
+; CHECK-NEXT:    ret i8 [[TMP1]]
+  entry:
+  %0 = call i32 @llvm.vscale.i32()
+  %1 = trunc i32 %0 to i8
+  ret i8 %1
+}
+
+declare i32 @llvm.vscale.i32()