Re: Patch: meta_data.c

it doesn't hurt whether gwlist_destroy(keys, NULL) or gwlist_destroy(keys, octstr_get_cstr)

Am 03.12.2009 um 14:16 schrieb Nikos Balkanas:

> Yeap. That's why valgrind continued to complain... In fact my line:
>  
> gwlist_destroy(keys, octstr_destroy_item)
>  
> is meaningless since by that time the list is empty. It doesn't hurt, but would you be kind enough to revert it to the original one?
>  
> gwlist_destroy(keys, NULL)
>  
> BR,
> Nikos
> ----- Original Message -----
> From: Alexander Malysh
> To: Nikos Balkanas
> Cc: [email protected]
> Sent: Thursday, December 03, 2009 11:45 AM
> Subject: Re: Patch: meta_data.c
> 
> Hi Nikos,
> 
> thanks for patch but your patch is not enough.
> Full fix commited to cvs and attached.
> 
> Thanks,
> Alexander Malysh
> 
> 
> 
> Am 03.12.2009 um 08:46 schrieb Nikos Balkanas:
> 
>> Hi,
>>  
>> This patch fixes a memory leak each time an smpp_pdu is processed. Reported first by valgrind, source code 1 liner, very obvious. However, valgrind still reports it as leak after patch.
>>  
>> BR,
>> Nikos
>> <meta_data.diff>
> 
> 
> 
> Hi Nikos,
> 
> thanks for patch but your patch is not enough.
> Full fix commited to cvs and attached.
> 
> Thanks,
> Alexander Malysh
> 
> Am 03.12.2009 um 08:46 schrieb Nikos Balkanas:
> 
> > Hi,
> >  
> > This patch fixes a memory leak each time an smpp_pdu is processed. Reported first by valgrind, source code 1 liner, very obvious. However, valgrind still reports it as leak after patch.
> >  
> > BR,
> > Nikos
> > <meta_data.diff>
>