public inbox for gcc-prs@sourceware.org
help / color / mirror / Atom feed
* Re: c++/10591: [pch] Members of anonymous namespaces get name mangled with header name
@ 2003-05-01 23:38 bangerth
0 siblings, 0 replies; 2+ messages in thread
From: bangerth @ 2003-05-01 23:38 UTC (permalink / raw)
To: bangerth, gcc-bugs, gcc-prs, geoffk, nobody
Synopsis: [pch] Members of anonymous namespaces get name mangled with header name
Responsible-Changed-From-To: unassigned->geoffk
Responsible-Changed-By: bangerth
Responsible-Changed-When: Thu May 1 23:38:51 2003
Responsible-Changed-Why:
Master of pch and author of a very much related patch for
non-pch quite recently -- what a lucky coincidence :-)
http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=10591
^ permalink raw reply [flat|nested] 2+ messages in thread
* c++/10591: [pch] Members of anonymous namespaces get name mangled with header name
@ 2003-05-01 23:26 bangerth
0 siblings, 0 replies; 2+ messages in thread
From: bangerth @ 2003-05-01 23:26 UTC (permalink / raw)
To: gcc-gnats
>Number: 10591
>Category: c++
>Synopsis: [pch] Members of anonymous namespaces get name mangled with header name
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: unassigned
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Thu May 01 23:26:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator: Wolfgang Bangerth
>Release: unknown-1.0
>Organization:
>Environment:
present mainline (3.4)
>Description:
Members of anonymous namespaces are very much like static
variables/functions, except that they are given external
linkage but a name that is mangled in a way so as to
make them unique. This is not the case if it happens in
a header file that is then precompiled, subsequently
leading to linker errors due to duplicate symbols.
Example:
echo 'namespace { int i; }' > x.h
echo '#include "x.h"' > a.cc
echo '#include "x.h"' > b.cc
echo 'int main () {}' >> b.cc
/home/bangerth/bin/gcc-3.4-pre/bin/c++ -c x.h
/home/bangerth/bin/gcc-3.4-pre/bin/c++ -c a.cc
/home/bangerth/bin/gcc-3.4-pre/bin/c++ -c b.cc
/home/bangerth/bin/gcc-3.4-pre/bin/c++ a.o b.o
b.o(.bss+0x0): multiple definition of `(anonymous namespace)::i'
a.o(.bss+0x0): first defined here
collect2: ld returned 1 exit status
The reason is the mangled name this variable is given:
g/x> nm a.o
00000000 B _ZN20_GLOBAL__N_x.huNZudc1iE
g/x> nm b.o
00000000 B _ZN20_GLOBAL__N_x.huNZudc1iE
00000000 T main
The standard says that they must be unique every time,
but obviously they are not.
This situation is a little unfortunate, since for example
boost uses such variables in anonymous namespaces in some
headers, meaning that one cannot use precompiled headers
in conjunction with boost.
W.
>How-To-Repeat:
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted:
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2003-05-01 23:38 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-05-01 23:38 c++/10591: [pch] Members of anonymous namespaces get name mangled with header name bangerth
-- strict thread matches above, loose matches on Subject: below --
2003-05-01 23:26 bangerth
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).