Bug#516727: gcc segfaults on simple wrong c++ code
Package: g++ (Debian 4.3.3-4) 4.3.3
The current c++ compiler segfaults on the following simple code mistake, where
a destructor doesn't have the same id as that the class.
template< typename T >
class some_class {
some_class() {}
// segfault
~some() {}
// no segfault
// ~some_class() {}
};
Kind regards,
Rutger ter Borg
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Processed: Re: Bug#516727: gcc segfaults on simple wrong c++ code
Processing commands for cont...@bugs.debian.org: > reassign 516727 g++-4.3 Bug#516727: gcc segfaults on simple wrong c++ code Warning: Unknown package 'debian' Warning: Unknown package '4.3.3-4' Warning: Unknown package '4.3.3' Bug reassigned from package `g++ (debian 4.3.3-4) 4.3.3' to `g++-4.3'. > -- Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#516727: gcc segfaults on simple wrong c++ code
* Rutger ter Borg [2009-02-23 12:14]: > Package: g++ (Debian 4.3.3-4) 4.3.3 > > The current c++ compiler segfaults on the following simple code mistake, > where > a destructor doesn't have the same id as that the class. Confirmed. Note that this didn't happen with 4.3.3-1 -- Martin Michlmayr http://www.cyrius.com/ -- To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#516727: gcc segfaults on simple wrong c++ code
* Rutger ter Borg [2009-02-23 12:14]: > The current c++ compiler segfaults on the following simple code mistake, > where > a destructor doesn't have the same id as that the class. I don't get this bug with the current 4.3 from SVN, so this has either been fixed or has been introduced by a Debian-specific patch. I'll leave it to doko. (sid)3863:t...@em64t: ..4.3-2009-02-23-r144380/gcc] ./g++ -B. ~/src/t.cc /home/tbm/src/t.cc:7: error: declaration of '~some' as member of 'some_class' zsh: exit 1 ./g++ -B. ~/src/t.cc (sid)3864:t...@em64t: ..4.3-2009-02-23-r144380/gcc] ./g++ -v Using built-in specs. Target: x86_64-unknown-linux-gnu Configured with: /home/tbm/scratch/gcc-4.3/configure --disable-bootstrap --enable-languages=c,c++ Thread model: posix gcc version 4.3.4 20090223 (prerelease) (GCC) -- Martin Michlmayr http://www.cyrius.com/ -- To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#516419: marked as done (ppl_0.10-3(mips/unstable):)
Your message dated Mon, 23 Feb 2009 14:40:45 +0100 with message-id <20090223134045.ga...@l04.local> and subject line Re: Bug#516419: ppl_0.10-3(mips/unstable): has caused the Debian Bug report #516419, regarding ppl_0.10-3(mips/unstable): to be marked as done. This means that you claim that the problem has been dealt with. If this is not the case it is now your responsibility to reopen the Bug report if necessary, and/or fix the problem forthwith. (NB: If you are a system administrator and have no idea what this message is talking about, this may indicate a serious mail system misconfiguration somewhere. Please contact ow...@bugs.debian.org immediately.) -- 516419: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=516419 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems --- Begin Message --- Package: ppl Version: 0.10-3 Severity: serious There was an error while trying to autobuild your package: > Automatic build of ppl_0.10-3 on mayr by sbuild/mips 99.999 > Build started at 20090219-0952 [...] > ** Using build dependencies supplied by package: > Build-Depends: debhelper (>= 6.0.7~), quilt, libgmp3-dev (>= 4.1.3), > autoconf, automake, libtool, autotools-dev [...] > make[3]: Entering directory `/build/buildd/ppl-0.10' > make[3]: Nothing to be done for `all-am'. > make[3]: Leaving directory `/build/buildd/ppl-0.10' > make[2]: Leaving directory `/build/buildd/ppl-0.10' > make[1]: Leaving directory `/build/buildd/ppl-0.10' > touch build-arch-stamp > export save_size=1 ; /usr/bin/make -C doc user-configured > make[1]: Entering directory `/build/buildd/ppl-0.10/doc' > rm -rf ppl-user-0.10-html > TEXINPUTS=/build/buildd/ppl-0.10/doc: doxygen user.doxyconf-html > /bin/sh: doxygen: command not found > make[1]: *** [ppl-user-0.10-html] Error 127 > make[1]: Leaving directory `/build/buildd/ppl-0.10/doc' > make: *** [build-indep-stamp] Error 2 > dpkg-buildpackage: failure: debian/rules build gave error exit status 2 A full build log can be found at: http://buildd.debian.org/build.php?arch=mips&pkg=ppl&ver=0.10-3 --- End Message --- --- Begin Message --- Version: 0.10-4 [...] > > Yes, thanks, already noticed; in fact, I uploaded a version supposed to fix > this > an hour ago. I'll close this bug tomorrow, once autobuilding has confirmed > that > my fix is appropriate. > Ok, that issue (which affected all the platforms) is solved, but there are some more failures which seem to be platform/buildd specifc, I'll investigate later... Best, Michael pgp4HA7SvpqfM.pgp Description: PGP signature --- End Message ---
Processed: [bts-link] source package gcc-defaults
Processing commands for cont...@bugs.debian.org: > # > # bts-link upstream status pull for source package gcc-defaults > # see http://lists.debian.org/debian-devel-announce/2006/05/msg1.html > # > user bts-link-upstr...@lists.alioth.debian.org Setting user to bts-link-upstr...@lists.alioth.debian.org (was bts-link-de...@lists.alioth.debian.org). > # remote status report for #231748 > # * http://gcc.gnu.org/PR11931 > # * remote status changed: NEW -> RESOLVED > # * remote resolution changed: (?) -> FIXED > # * closed upstream > tags 231748 + fixed-upstream Bug#231748: libc6-dev: programs using HUGE_VAL will issue a warning when compiled with -pedantic on the g++-3.3 Tags were: upstream Tags added: fixed-upstream > usertags 231748 - status-NEW Bug#231748: libc6-dev: programs using HUGE_VAL will issue a warning when compiled with -pedantic on the g++-3.3 Usertags were: status-NEW. Usertags are now: . > usertags 231748 + status-RESOLVED resolution-FIXED Bug#231748: libc6-dev: programs using HUGE_VAL will issue a warning when compiled with -pedantic on the g++-3.3 There were no usertags set. Usertags are now: status-RESOLVED resolution-FIXED. > # remote status report for #231748 > # * http://gcc.gnu.org/PR11931 > # * remote status changed: NEW -> RESOLVED > # * remote resolution changed: (?) -> FIXED > # * closed upstream > tags 231748 + fixed-upstream Bug#231748: libc6-dev: programs using HUGE_VAL will issue a warning when compiled with -pedantic on the g++-3.3 Tags were: fixed-upstream upstream Tags added: fixed-upstream > usertags 231748 - status-NEW Bug#231748: libc6-dev: programs using HUGE_VAL will issue a warning when compiled with -pedantic on the g++-3.3 Usertags were: status-RESOLVED resolution-FIXED. Usertags are now: status-RESOLVED resolution-FIXED. > usertags 231748 + status-RESOLVED resolution-FIXED Bug#231748: libc6-dev: programs using HUGE_VAL will issue a warning when compiled with -pedantic on the g++-3.3 Usertags were: status-RESOLVED resolution-FIXED. Usertags are now: status-RESOLVED resolution-FIXED. > thanks Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
[bts-link] source package gcc-defaults
# # bts-link upstream status pull for source package gcc-defaults # see http://lists.debian.org/debian-devel-announce/2006/05/msg1.html # user bts-link-upstr...@lists.alioth.debian.org # remote status report for #231748 # * http://gcc.gnu.org/PR11931 # * remote status changed: NEW -> RESOLVED # * remote resolution changed: (?) -> FIXED # * closed upstream tags 231748 + fixed-upstream usertags 231748 - status-NEW usertags 231748 + status-RESOLVED resolution-FIXED # remote status report for #231748 # * http://gcc.gnu.org/PR11931 # * remote status changed: NEW -> RESOLVED # * remote resolution changed: (?) -> FIXED # * closed upstream tags 231748 + fixed-upstream usertags 231748 - status-NEW usertags 231748 + status-RESOLVED resolution-FIXED thanks -- To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#516862: optimization can prevent terminate() from being called on throw from destructor while unwinding
Package: g++-4.3
Version: 4.3.2-1.1 , 4.3.3-4
This is also in at least g++-4.1 (4.1.2-25 and 4.1.1-21) and g++-4.2
(4.2.4-6). I don't have a g++-3.x installed to test with.
This is on i386.
The program at the end of this report should always print
"my_terminate()" and exit. If the "WTF dont_call_terminate;" line is
commented, that's what it does. If -O0 is given, that's what it does.
If that line is not commented and -O2, -O1, or -Os is given, it breaks
differently depending on whether the "inline" is commented.
With the "inline" commented, both throw_twice_no_terminate() and
throw_twice() result in "caught a 'std::exception'". With the inline
uncommented, throw_twice_no_terminate() results in "inner handler",
while throw_twice() results in "my_terminate()".
#include
#include
#include
void my_terminate() {
std::cerr<<"my_terminate()\n";
exit(1);
}
class WTF {
public:
~WTF(){}
};
class throwing_dtor
{
WTF dont_call_terminate;
public:
throwing_dtor() {}
~throwing_dtor()
{
throw std::exception();
}
};
inline // it breaks differently if this is commented
void do_throw_twice() {
throwing_dtor td;
throw std::exception();
}
void throw_twice() {
try {
do_throw_twice();
} catch (std::exception & e) {
std::cerr<<"caught a '"<
