[Bug c++/69023] bitset whose name is used in constant-expression rejected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69023 --- Comment #5 from Martin Sebor --- Author: msebor Date: Wed Dec 23 21:52:50 2015 New Revision: 231938 URL: https://gcc.gnu.org/viewcvs?rev=231938&root=gcc&view=rev Log: PR c++/69023 - bitset whose name is used in constant-expression rejected * g++.dg/lookup/name-clash11.C: New test. Added: trunk/gcc/testsuite/g++.dg/lookup/name-clash11.C Modified: trunk/gcc/testsuite/ChangeLog
[Bug c++/69023] bitset whose name is used in constant-expression rejected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69023 Martin Sebor changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |INVALID --- Comment #4 from Martin Sebor --- Thanks for the reference! I now see it quoted in the code as well. (I should have looked it up before opening the bug.) I'll close it as invalid, though perhaps adding a test case might be useful since I don't see this case covered in the test suite. Let me submit one.
[Bug c++/69023] bitset whose name is used in constant-expression rejected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69023 TC changed: What|Removed |Added CC||rs2740 at gmail dot com --- Comment #3 from TC --- [basic.scope.class]/1, item 2: A name N used in a class S shall refer to the same declaration in its context and when re-evaluated in the completed scope of S. No diagnostic is required for a violation of this rule.
[Bug c++/69023] bitset whose name is used in constant-expression rejected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69023
--- Comment #2 from Martin Sebor ---
I'm not sure. I did read the relevant text in the C++ and C standards and I
can't say I see a clear answer. But then it rarely seems to be in these corner
cases.
GCC (in C mode) accepts the code. All other C and C++ compilers I have access
to (about half a dozen) do as well. The one interesting thing I noticed that
the EDG front end issues the following warnings in C89 and C++ modes but not in
C99 and later.
"b.c", line 4: warning: reference to local variable of enclosing function is
not allowed
struct { int : sizeof x; } b0; (void)&b0;
^
"b.c", line 5: warning: reference to local variable of enclosing function is
not allowed
struct { int x: sizeof x; } b1; (void)&b1;
^
[Bug c++/69023] bitset whose name is used in constant-expression rejected
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69023 Andrew Pinski changed: What|Removed |Added Keywords||diagnostic --- Comment #1 from Andrew Pinski --- I think GCC is correct with its error message for b1 even if it does not do it also for b2. The meaning of x does change after the struct is finalized. Note this kind of diagnostic is not required to happen. That is the standard says the code is invalid but no diagnostic is required. This is one case where the two compilers can disagree but both be correct even though the one of them rejects the code.
