Hi!
On this invalid testcase we ICE starting with PR47956 changes, because
a static data member is VAR_HAD_UNKNOWN_BOUND, but
maybe_deduce_size_from_array_init changes its type from incomplete
array to error_mark_node. tsubst_decl assumes that VAR_HAD_UNKNOWN_BOUND
has ARRAY_TYPE type. Here is one possible fix, another possible fix
would be to add if (type == error_mark_node) return type;
to the beginning of strip_array_domain, another option perhaps adjust
the SET_VAR_HAD_UNKNOWN_BOUND macro to take another argument and
be able to clear that flag and clear it in
maybe_deduce_size_from_array_init.
This variant has been bootstrapped/regtested on x86_64-linux and i686-linux.
2011-12-06 Jakub Jelinek <ja...@redhat.com>
PR c++/51430
* pt.c (tsubst_decl): Don't call strip_array_domain on
error_mark_node.
--- gcc/cp/pt.c.jj 2011-12-01 11:45:04.000000000 +0100
+++ gcc/cp/pt.c 2011-12-06 09:11:29.715888859 +0100
@@ -10622,7 +10622,9 @@ tsubst_decl (tree t, tree args, tsubst_f
type = DECL_ORIGINAL_TYPE (t);
else
type = TREE_TYPE (t);
- if (TREE_CODE (t) == VAR_DECL && VAR_HAD_UNKNOWN_BOUND (t))
+ if (TREE_CODE (t) == VAR_DECL
+ && VAR_HAD_UNKNOWN_BOUND (t)
+ && type != error_mark_node)
type = strip_array_domain (type);
type = tsubst (type, args, complain, in_decl);
}
--- gcc/testsuite/g++.dg/template/static32.C.jj 2011-12-06 09:16:02.606219171
+0100
+++ gcc/testsuite/g++.dg/template/static32.C 2011-12-06 09:16:28.001068818
+0100
@@ -0,0 +1,9 @@
+// PR c++/51430
+// { dg-do compile }
+
+template<int> struct A
+{
+ static const int x[] = 0; // { dg-error "in-class
initialization|initializer fails" }
+};
+
+A<0> a;
Jakub
