}
case Instruction::Load: {
// Check if the loads are consecutive or of we need to swizzle them.
- for (unsigned i = 0, e = VL.size() - 1; i < e; ++i)
- if (!isConsecutiveAccess(VL[i], VL[i + 1])) {
+ for (unsigned i = 0, e = VL.size() - 1; i < e; ++i) {
+ LoadInst *L = cast<LoadInst>(VL[i]);
+ if (!L->isSimple() || !isConsecutiveAccess(VL[i], VL[i + 1])) {
newTreeEntry(VL, false);
DEBUG(dbgs() << "SLP: Need to swizzle loads.\n");
return;
}
-
+ }
newTreeEntry(VL, true);
DEBUG(dbgs() << "SLP: added a vector of loads.\n");
return;
if (!SI)
continue;
+ // Don't touch volatile stores.
+ if (!SI->isSimple())
+ continue;
+
// Check that the pointer points to scalars.
Type *Ty = SI->getValueOperand()->getType();
if (Ty->isAggregateType() || Ty->isVectorTy())
store double %mul5, double* %arrayidx5, align 8
ret void
}
+
+; Don't vectorize volatile loads.
+; CHECK: test_volatile_load
+; CHECK-NOT: load <2 x double>
+; CHECK: store <2 x double>
+; CHECK: ret
+define void @test_volatile_load(double* %a, double* %b, double* %c) {
+entry:
+ %i0 = load volatile double* %a, align 8
+ %i1 = load volatile double* %b, align 8
+ %mul = fmul double %i0, %i1
+ %arrayidx3 = getelementptr inbounds double* %a, i64 1
+ %i3 = load double* %arrayidx3, align 8
+ %arrayidx4 = getelementptr inbounds double* %b, i64 1
+ %i4 = load double* %arrayidx4, align 8
+ %mul5 = fmul double %i3, %i4
+ store double %mul, double* %c, align 8
+ %arrayidx5 = getelementptr inbounds double* %c, i64 1
+ store double %mul5, double* %arrayidx5, align 8
+ ret void
+}
+
+; Don't vectorize volatile stores.
+; CHECK: test_volatile_store
+; CHECK-NOT: store <2 x double>
+; CHECK: ret
+define void @test_volatile_store(double* %a, double* %b, double* %c) {
+entry:
+ %i0 = load double* %a, align 8
+ %i1 = load double* %b, align 8
+ %mul = fmul double %i0, %i1
+ %arrayidx3 = getelementptr inbounds double* %a, i64 1
+ %i3 = load double* %arrayidx3, align 8
+ %arrayidx4 = getelementptr inbounds double* %b, i64 1
+ %i4 = load double* %arrayidx4, align 8
+ %mul5 = fmul double %i3, %i4
+ store volatile double %mul, double* %c, align 8
+ %arrayidx5 = getelementptr inbounds double* %c, i64 1
+ store volatile double %mul5, double* %arrayidx5, align 8
+ ret void
+}
+
+