Bug 33378

Summary:	[4.3 regression] Recent gcc 4.3 snapshots think comparison operators are ambiguous
Product:	gcc	Reporter:	bero
Component:	c++	Assignee:	Not yet assigned to anyone <unassigned>
Status:	RESOLVED WORKSFORME
Severity:	normal	CC:	fang, gcc-bugs, tbm
Priority:	P3
Version:	4.3.0
Target Milestone:	---
Host:	i586-pc-linux-gnu	Target:	i586-pc-linux-gnu
Build:	i586-pc-linux-gnu	Known to work:	3.4.0 4.0.0 4.2.0 4.2.1
Known to fail:	4.3.0	Last reconfirmed:
Attachments:	test case

Description bero 2007-09-10 14:11:09 UTC

The attached test case has 2 code branches (#ifdef BREAKME and #ifndef BREAKME) that are treated as identical in all prior gcc versions.

4.3 snapshots barf with:

gcc-4.3-testcase.cpp: In function 'int main()':
gcc-4.3-testcase.cpp:28: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
gcc-4.3-testcase.cpp:14: note: candidate 1: bool operator==(const junk&, const char*)
gcc-4.3-testcase.cpp:7: note: candidate 2: bool junk::operator==(const foo&) const


since even in the BREAKME version, no implicit casting is involved, I tend to think older gcc versions are right.

Comment 1 bero 2007-09-10 14:11:45 UTC

Created attachment 14184 [details]
test case

Add test case

Comment 2 Paolo Carlini 2007-09-17 17:22:20 UTC

I can't reproduce the problem with current (128551) mainline. Likely a transient issue, otherwise, please reopen.