RE: [Valgrind-users] How do I suppress this?

[Burton M. S. I. <BSt...@ac...>]

No redirection.  I took that out, thinking it was the problem, although the
tee I was using had worked for every other case.  This one just prints the
line and keeps going....

As for Addr2 vs. Value2, I missed that in the docs.  Thanks... My problem
was that I didn't understand the mapping between the error message from
Valgrind, e.g. "Use of uninitialised value of size 4" and the skin:code,
e.g. Memcheck:Value4.

I wonder if it would be possible to have a more detailed examples section.
A page or two of here's the Valgrind output and here's how to suppress it
would help.  Here's a contribution to start you off.

Valgrind suppression examples...

==31251== Thread 4:
==31251== Use of uninitialised value of size 4
==31251==    at 0x4155398F: _IO_vfprintf_internal (in /lib/libc-2.3.2.so)
==31251==    by 0x415E94F1: __GI_vsyslog (in /lib/libc-2.3.2.so)
==31251==    by 0x415E935E: __GI_syslog (in /lib/libc-2.3.2.so)
==31251==    by 0x402EE312: ntop_gdbm_nextkey (util.c:4008)

{
   example1
   Memcheck:Value4
   fun:_IO_vfprintf_internal
   fun:__GI_vsyslog
   fun:__GI_syslog
   fun:ntop_gdbm_nextkey
}

==31251== Thread 4:
==31251== Conditional jump or move depends on uninitialised value(s)
==31251==    at 0x41553997: _IO_vfprintf_internal (in /lib/libc-2.3.2.so)
==31251==    by 0x415E94F1: __GI_vsyslog (in /lib/libc-2.3.2.so)
==31251==    by 0x415E935E: __GI_syslog (in /lib/libc-2.3.2.so)
==31251==    by 0x402EE312: ntop_gdbm_nextkey (util.c:4008)

{
   example2
   Memcheck:Cond
   fun:_IO_vfprintf_internal
   fun:__GI_vsyslog
   fun:__GI_syslog
   fun:ntop_gdbm_nextkey
}


==31251== 400 bytes in 2 blocks are definitely lost in loss record 17 of 25
==31251==    at 0x4148FD6C: my_malloc (vg_libpthread.c:263)
==31251==    by 0x414916FF: get_or_allocate_specifics_ptr
(vg_libpthread.c:1405)
==31251==    by 0x414919DE: __pthread_getspecific (vg_libpthread.c:298)
==31251==    by 0x418783CB: _dlerror_run (in /lib/libdl-2.3.2.so)

{
   example3
   Memcheck:Leak
   fun:my_malloc
   fun:get_or_allocate_specifics_ptr
   fun:__pthread_getspecific
   fun:_dlerror_run
}


Thanks again!

-----Burton

PS: you can add ntop (http://www.ntop.org) to the list of projects using
Valgrind.  I've actually got Valgrind integrated into my startup script on
the development box and kick it off every couple of weeks to run a few hours
or days.


> -----Original Message-----
> From: Nicholas Nethercote [mailto:nj...@he...]On Behalf Of
> Nicholas Nethercote
> Sent: Sunday, February 01, 2004 7:39 AM
> To: Burton M. Strauss III
> Cc: Valgrind Users
> Subject: Re: [Valgrind-users] How do I suppress this?
>
>
> On Sat, 31 Jan 2004, Burton M. Strauss III wrote:
>
> > ==11919== Thread 7:
> > ==11919== Invalid read of size 4
> > ==11919==    at 0x4152E322: __GI_getenv (in /lib/i686/libc-2.3.2.so)
> > ==11919==    by 0x4159B7EE: tzset_internal (in /lib/i686/libc-2.3.2.so)
> > ==11919==    by 0x4159C3F7: __tzset (in /lib/i686/libc-2.3.2.so)
> > ==11919==    by 0x415A33CA: __GI_strftime (in /lib/i686/libc-2.3.2.so)
> > ==11919==  Address 0xBFFFEFF8 is not stack'd, malloc'd or free'd
> > ==11919==
> > ==11919== ---- Print suppression ? --- [Return/N/n/Y/y/C/c]
> ---- ==11919==
> > ==11919== Thread 7:
> > ==11919== Invalid read of size 2
> > ==11919==    at 0x4152E330: __GI_getenv (in /lib/i686/libc-2.3.2.so)
> > ==11919==    by 0x4159B7EE: tzset_internal (in /lib/i686/libc-2.3.2.so)
> > ==11919==    by 0x4159C3F7: __tzset (in /lib/i686/libc-2.3.2.so)
> > ==11919==    by 0x415A33CA: __GI_strftime (in /lib/i686/libc-2.3.2.so)
> > ==11919==  Address 0xBFFFFA32 is not stack'd, malloc'd or free'd
> >
> > You'll note that it never stops after the 'Print suppression' line, so I
> > can't automatically generate this.
>
> That's odd... are you doing something unusual with stdin, like redirecting
> it?
>
> > I've tried:
> >
> > {
> >    strftime_tz_getenv2
> >    Memcheck:Value2
> >    fun:__GI_getenv
> >    fun:tzset_internal
> >    fun:__tzset
> >    fun:__GI_strftime
> > }
> > {
> >    strftime_tz_getenv4
> >    Memcheck:Value4
> >    fun:__GI_getenv
> >    fun:tzset_internal
> >    fun:__tzset
> >    fun:__GI_strftime
> > }
> >
> > To no effect.
>
> I think you want "Memcheck:Addr2" and "Memcheck:Addr4" for the suppression
> kinds.
>
> N