Re: getaddrinfo_a memory leaks

[Martin Sustrik <sustrik at 250bpm dot com>]

Hi OndÅej,

Thanks for prompt answer.

I've forgot to attach my libc version:

GNU C Library (Ubuntu EGLIBC 2.13-0ubuntu13) stable release version 
2.13, by Roland McGrath et al.
Copyright (C) 2011 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.
Compiled by GNU CC version 4.5.2.
Compiled on a Linux 2.6.38 system on 2011-04-11.
Available extensions:
	crypt add-on version 2.1 by Michael Glad and others
	GNU Libidn by Simon Josefsson
	Native POSIX Threads Library by Ulrich Drepper et al
	BIND-8.2.3-T5B
libc ABIs: UNIQUE IFUNC

Anyway, I am seeing several additional problems with getaddrinfo_a in 
this version:

1. Documentation says: "The gai_cancel() function cancels the request 
req. If the request has been canceled successfully, the error status of 
the request will be set to EAI_CANCELLED and normal asynchronous 
notification will be performed." What I see in the reality is that no 
async notification is performed in this case.
2. There are memory leaks, as mentioned above.
3. Documentation mentioned EAI_CANCELLED and EAI_NOTCANCELLED whereas 
header files define EAI_CANCELED and EAI_NOTCANCELED.

So, my overall question is: Given the above, is getaddrinfo_a 
considered an unstable/experimental functionality that I should not even 
try to use? Or is it that I am using an old version of the library and 
getaddrinfo_a in new versions should actually work as expected?

Martin


On 2013-05-31 10:49, OndÅej BÃlka wrote:
> On Fri, May 31, 2013 at 10:27:51AM +0200, Martin Sustrik wrote:
> > Hi all,
> >
> > Just a quick question: Is getaddrinfo_a supposed to leak memory?
> It should not.
> > When run with valgrind following leaks are reported:
> >
> > ==1959== 3,200 bytes in 50 blocks are definitely lost in loss record
> > 3 of 4
> > ==1959==    at 0x4C28FAC: malloc (vg_replace_malloc.c:236)
> > ==1959==    by 0x573D00E: gaih_inet (getaddrinfo.c:1001)
> Code at line 1001 is following which is out of sync with info above.
>
>           malloc_canonbuf = true;
>             }
>           char *s;
>
>           if (DL_CALL_FCT (cfct, (at->name ?: name,
>                 canonbuf,
>                 max_fqdn_len,
>                 &s, &rc, &herrno))
>               == NSS_STATUS_SUCCESS)
>
>
> > ==1959==    by 0x57406B1: getaddrinfo (getaddrinfo.c:2161)
> > ==1959==    by 0x5262ED8: handle_requests (gai_misc.c:321)
> > ==1959==    by 0x546CD8B: start_thread (pthread_create.c:304)
> > ==1959==    by 0x576A04C: clone (clone.S:112)
> > ==1959==
> > ==1959== 5,200 bytes in 50 blocks are possibly lost in loss record 4
> > of 4
> > ==1959==    at 0x4C28FAC: malloc (vg_replace_malloc.c:236)
> > ==1959==    by 0x5263C08: getaddrinfo_a (getaddrinfo_a.c:150)
> > ==1959==    my application stack follows here
> >
> > Thanks!
> > Martin