memory leak in res_nsend()

[Antony Dovgal]

On 24.10.2005 21:57, Daniel Jacobowitz wrote:
> On Mon, Oct 24, 2005 at 09:45:42PM +0400, Antony Dovgal wrote:
>> On 24.10.2005 21:14, Daniel Jacobowitz wrote:
>> >On Mon, Oct 24, 2005 at 02:28:03PM +0400, Antony Dovgal wrote:
>> >>But res_nclose() doesn't free those segments.
>> >>That's what I've found in res_thread_freeres () function 
>> >>(resolv/res_init.c):
>> >
>> >Well, when you're running under valgrind, the freeres functions are
>> >supposed to be called - for exactly this reason. 
>> 
>> Could you please elaborate: why should valgrind call something?
>> *I'm* calling res_nsend() and I suppose valgrind shouldn't call anything to 
>> free resources allocated by the user function calls.
>> Am I wrong?
> 
> Partly.  valgrind frees things that would ordinarily be freed by
> exit(), so that it can detect leaks better.
> 
> If it's doing that and still reporting a leak, then perhaps it should
> be calling __libc_thread_freeres in still-alive threads at exit.

__libc_thread_freeres(void) frees global _res, while I need to free resources associated with my local __res_state struct.
I honestly do not understand how can valgrind help me with that calling __libc_thread_freeres().

-- 
Wbr, 
Antony Dovgal