================
@@ -4,9 +4,8 @@
void test() {
// CHECK: VarDecl {{.*}} used f3 'vector<float, 3>':'float
__attribute__((ext_vector_type(3)))' cinit
- // CHECK-NEXt: ImplicitCastExpr {{.*}} 'vector<float, 3>':'float
__attribute__((ext_vector_type(3)))' <VectorSplat>
- // CHECK-NEXt: ImplicitCastExpr {{.*}} 'float' <FloatingCast>
- // CHECK-NEXt: FloatingLiteral {{.*}} 'double' 1.000000e+00
+ // CHECK-NEXT: ImplicitCastExpr {{.*}} 'vector<float, 3>':'float
__attribute__((ext_vector_type(3)))' <VectorSplat>
----------------
bogner wrote:
Yes this is intentional, as the test doesn't pass without the change. Since
#91015 the literal is a `float`, not a `double`, so the implicit cast to float
doesn't happen.
https://github.com/llvm/llvm-project/pull/95155
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
