On 6/10/20 5:11 PM, Marek Polacek wrote:
Since r10-7096 convert_like, when called in a template, creates an
IMPLICIT_CONV_EXPR when we're converting to/from array type.
In this test, we have e[f], and we're converting f (of type class A) to
int, so convert_like in build_new_op_1 created the IMPLICIT_CONV_EXPR
that got into cp_build_array_ref which calls maybe_constant_value. My
patch above failed to adjust this spot to call fold_non_dependent_expr
instead, which can handle codes like I_C_E in a template.
Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk/10?
gcc/cp/ChangeLog:
PR c++/95508
* typeck.c (cp_build_array_ref): Call fold_non_dependent_expr instead
of maybe_constant_value.
gcc/testsuite/ChangeLog:
PR c++/95508
* g++.dg/template/conv16.C: New test.
---
gcc/cp/typeck.c | 2 +-
gcc/testsuite/g++.dg/template/conv16.C | 17 +++++++++++++++++
2 files changed, 18 insertions(+), 1 deletion(-)
create mode 100644 gcc/testsuite/g++.dg/template/conv16.C
diff --git a/gcc/cp/typeck.c b/gcc/cp/typeck.c
index f01ae656254..07144d4c1fc 100644
--- a/gcc/cp/typeck.c
+++ b/gcc/cp/typeck.c
@@ -3565,7 +3565,7 @@ cp_build_array_ref (location_t loc, tree array, tree idx,
pointer arithmetic.) */
idx = cp_perform_integral_promotions (idx, complain);
- idx = maybe_constant_value (idx);
+ idx = fold_non_dependent_expr (idx, complain);
Hmm, if idx isn't constant this will leave us with non-template trees in
the ARRAY_REF. I think what we want is a function that will fold a
non-dependent expr to a constant or return the argument.
maybe_fold_non_dependent_expr?
Jason
