I made the const_string destructor virtual and here is what happened (gdb
message):
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1283466336 (LWP 7230)]
0x004ccb29 in __gnu_cxx::__exchange_and_add () from /usr/lib/libstdc++.so.6
(gdb) where
#0 0x004ccb29 in __gnu_cxx::__exchange_and_add () from
/usr/lib/libstdc++.so.6
#1 0x004b0f24 in std::basic_string<char, std::char_traits<char>,
std::allocator<char> >::~basic_string () from /usr/lib/libstdc++.so.6
#2 0x08068564 in ~ColData_Tmpl (this=0xb37fbff8)
at /usr/include/mysql++/coldata.h:87
#3 0x08062258 in btau::DBSession::retrieveLastLinkingitem (this=0x87f8738,
linkingItemList=@0xb37fc18c) at ../DatabaseModule/DBSession.cpp:607
#4 0x080764a2 in
btau::ApplicationDataLogicManager::saveTimestampAndLinkingItem
(this=0xb37fd2c4, tst=@0xb37fd2d8)
at ../ApplicationLogicModule/ApplicationDataLogicManager.cpp:388
#5 0x08089543 in
btau::TSUMessageProcessingThread::manageTimeStampsaveRequest
(this=0x87f69e8, tsuReq=0x87f6d58)
at ../ApplicationLogicModule/TSUMessageProcessingThread.cpp:526
#6 0x08089e7a in btau::TSUMessageProcessingThread::Run (this=0x87f69e8)
at ../ApplicationLogicModule/TSUMessageProcessingThread.cpp:402
#7 0x0809f513 in Thread::StartThread (zz=0x87f69e8) at Thread.cpp:106
#8 0x00369433 in start_thread () from /lib/libpthread.so.0
#9 0x001c9a1e in clone () from /lib/libc.so.6
I do not know if that can be interesting to you, I reported it just in case.
I will restore the standard version. I'll have some tries with a simpler
code: I will "valgrind" a small main and see what happens.
I will try and find some workaround, or the new version which you mentioned.
Thanks.
--------- Original Message --------
Da: plusplus@stripped
To: MySQL++ Mailing List <plusplus@stripped>
Oggetto: Re: Is method "const ColData Row::at(size_type i) const" leaking
memory?
Data: 26/11/07 21:58
>
>
>
> Thunder0 wrote:
> >
> > (const_string.h:110)==10964==
>
> Try making ~const_string() virtual and rebuilding.
>
> I don't really expect that to fix it, because if it does, it means
> someone somewhere is holding references to ColData<const_string> objects
> by pointers to the base class and deleting them through that pointer.
> That shouldn't be happening, because as far as I know, everyone keeps
> ColData objects as ColData. (This lack of use of the is-a property of
> inheritance is one reason why ColData no longer uses inheritance in v3.)
>
> If, as seems more likely, making the base class dtor virtual doesn't
> help, I'm at a total loss. You can clearly see in const_string.h that
> memory allocated in the ctors gets deallocated in the dtor. When a
> ColData_Tmpl<const_string> object goes away, ~const_string() should be
> called, freeing up the memory.
>
> --
> MySQL++ Mailing List
> For list archives: http://lists.mysql.com/plusplus
> To unsubscribe: http://lists.mysql.com/plusplus?unsub=1
>
>
>
>
--
Email.it, the professional e-mail, gratis per te: http://www.email.it/f
Sponsor:
Stai cercando l'email marketing di successo ed economica? Prova con
Email.it le tue campagne di Email Marketing a partire da 250 Euro!
Clicca qui: http://adv.email.it/cgi-bin/foclick.cgi?mid=7149&d=20071127
