Re: [Valgrind-users] dlopen in GLIBC 2.1

[Jarrod K G. <gr...@cs...>]

Interesting. The dlopen() errors that I get are of the following sort:

==6004== 16 bytes in 1 blocks are still reachable in loss record 1 of 3
==6004==    at 0x4015E780: calloc (vg_clientfuncs.c:245)
==6004==    by 0x4022A308: _dlerror_run (in /lib/libdl-2.2.93.so)
==6004==    by 0x40229EC3: dlopen@@GLIBC_2.1 (in /lib/libdl-2.2.93.so)
==6004==    by 0x804BD38: LoadLibraries (Decode.c:43)
==6004==    by 0x8049C2A: main (main.c:102)
==6004==    by 0x420158D3: __libc_start_main (in /lib/i686/libc-2.2.93.so)
==6004==    by 0x8049978: (within /home/bob/test)
==6004==
==6004== 176 bytes in 12 blocks are still reachable in loss record 2 of 3
==6004==    at 0x4015E310: malloc (vg_clientfuncs.c:103)
==6004==    by 0x4000948B: _dl_map_object_deps_internal (in
/lib/ld-2.2.93.so)
==6004==    by 0x4210951A: dl_open_worker (in /lib/i686/libc-2.2.93.so)
==6004==    by 0x4000A425: _dl_catch_error_internal (in /lib/ld-2.2.93.so)
==6004==    by 0x421092BE: __GI__dl_open (in /lib/i686/libc-2.2.93.so)
==6004==    by 0x40229F18: dlopen_doit (in /lib/libdl-2.2.93.so)
==6004==    by 0x4000A425: _dl_catch_error_internal (in /lib/ld-2.2.93.so)
==6004==    by 0x4022A2C3: _dlerror_run (in /lib/libdl-2.2.93.so)
==6004==    by 0x40229EC3: dlopen@@GLIBC_2.1 (in /lib/libdl-2.2.93.so)
==6004==    by 0x804BD38: LoadLibraries (Decode.c:43)
==6004==    by 0x8049C2A: main (main.c:102)
==6004==    by 0x420158D3: __libc_start_main (in /lib/i686/libc-2.2.93.so)
==6004==    by 0x8049978: (within /home/bob/test)

Jarrod

On Wed, 11 Jun 2003, Leonard mckinley wrote:

> Anyone seen this with dlopen() in GLIBC?  Don't think this is my
> app's fault, but would like to be sure.
>
> The more general question would be: is there are has anyone thought of
> having a database (web directory) of valgrind suppression rules for
> more commonly used libraries on Linux?