--- Additional Comments From bangerth at dealii dot org 2004-12-17 16:34
---
Nathan, if this isn't a regression but a patch has been applied to the
3.4 branch, then you should also apply it to mainline. Otherwise you have
just created a regression (3.4.4 will work as expected, but 4.0.
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2004-12-17
16:19 ---
Subject: Bug 18975
CVSROOT:/cvs/gcc
Module name:gcc
Branch: gcc-3_4-branch
Changes by: [EMAIL PROTECTED] 2004-12-17 16:19:24
Modified files:
gcc/cp : Change
--- Additional Comments From nathan at gcc dot gnu dot org 2004-12-17
16:10 ---
Fix for 3.4 branch
2004-12-17 Nathan Sidwell <[EMAIL PROTECTED]>
PR c++/18975
* method.c (do_build_copy_constructor): Refactor. Don't const
qualify a mutable field.
(do_buil
--- Additional Comments From nathan at gcc dot gnu dot org 2004-12-17
13:44 ---
I now think the code is ok, and both compilers are incorrect to reject any of
the versions.
The versions that do not contain user defined
XYZ (const XYZ& src)
XYZ& operator= (const XYZ& src)
functions a
--- Additional Comments From giovannibajo at libero dot it 2004-12-17
13:19 ---
Wolfgang, I suggest you to bring this up in comp.std.c++, to get some official
answer about this.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18975
--- Additional Comments From nathan at gcc dot gnu dot org 2004-12-15
14:44 ---
I am dubious about my own analysis of this. Wolfgang sent me more detailed
comparison between g++ and comeau
Hi Nathan,
I try to explain what I meant with my last comment. I have three variations
of my code
--- Additional Comments From nathan at gcc dot gnu dot org 2004-12-14
16:55 ---
(In reply to comment #4)
I can't tell what's being said here. IIUC, you're saying that
Comeau (which uses the edg frontend) also rejects the code.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18975
--- Additional Comments From wolfgang dot roehrl at de dot gi-de dot com
2004-12-14 14:19 ---
Subject: Antwort: Copying objects with mutable non-static data
members
Hi all,
I'm responding to Comments From nathan at gcc dot gnu dot org 2004-12-14
13:57 (bug 18975):
I just tried
--- Additional Comments From pinskia at gcc dot gnu dot org 2004-12-14
13:25 ---
XYZ (PTR& p) : ptr(p) {}
That will always call the constructor for ptr so this is invalid.
--
What|Removed |Added
---
--- Additional Comments From nathan at gcc dot gnu dot org 2004-12-14
13:57 ---
The code is invalid and the compiler correct.
The following line of reasoning is incorrect
- The implicitly defined copy constructor should use the copy constructor of
PTR (12.8/8). Since the member ptr of
--- Additional Comments From giovannibajo at libero dot it 2004-12-14
13:41 ---
No, this is a subtler issue. I need to think about it. Nathan may want to
express an opinion on this issue too.
--
What|Removed |Added
---
11 matches
Mail list logo
