From: Alexander P. <gl...@go...> - 2010-04-20 11:07:07
|
This should be tuned in gcc, not Valgrind. Valgrind just prints the debug info generated by gcc. For example: $ cat leak.cc int foobar() { int *foo = new int; delete[] foo; return 0; } int main() { foobar(); return 0; } $ g++ ../leak/leak.cc -o leak -g $ valgrind ./leak ... ==11525== Mismatched free() / delete / delete [] ==11525== at 0x4C23530: operator delete[](void*) /tmp/vg/coregrind/m_replacemalloc/vg_replace_malloc.c:446 ==11525== by 0x40060D: foobar() /home/glider/src/leak/../leak/leak.cc:3 ==11525== by 0x40061E: main /home/glider/src/leak/../leak/leak.cc:8 ==11525== Address 0x5926040 is 0 bytes inside a block of size 4 alloc'd ==11525== at 0x4C255C7: operator new(unsigned long) /tmp/vg/coregrind/m_replacemalloc/vg_replace_malloc.c:281 ==11525== by 0x4005F9: foobar() /home/glider/src/leak/../leak/leak.cc:2 ==11525== by 0x40061E: main /home/glider/src/leak/../leak/leak.cc:8 ==11525== On Tue, Apr 20, 2010 at 2:18 PM, Stefano Sabatini <ste...@gm...> wrote: > Hi all, > > I'm currently using valgrind-3.5.0-Debian. > > I see in the commentary: > ... > ==7197== by 0x80688A3: XYZ::Foo::Bar() (FooBar.cxx:361) > ==7197== by 0x4326435: PThread::PX_ThreadStart(void*) (tlibthrd.cxx:1341) > ==7197== by 0x4042584: start_thread (pthread_create.c:300) > ==7197== by 0x47CA29D: clone (clone.S:130) > > Since I'm using emacs, which has support for the > path/to/file:line_num syntax, I'd like valgrind to print the complete > pathname and allow me to jump to the source with just one click. > > So I wonder if it would be possible for example for valgrind to print: > ==7197== by 0x80688A3: XYZ::Foo::Bar() (/home/stefano/src/foobar-proj/libfoobar/FooBar.cxx:361) > > Is it currently already implemented? (from the reading of the docs it seems no) > > Should I file a feat request on bugzilla? > > Regards. > > ------------------------------------------------------------------------------ > Download Intel® Parallel Studio Eval > Try the new software tools for yourself. Speed compiling, find bugs > proactively, and fine-tune applications for parallel performance. > See why Intel Parallel Studio got high marks during beta. > http://p.sf.net/sfu/intel-sw-dev > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users > -- Alexander Potapenko Software Engineer Google Moscow |