On Sat, Mar 15, 2014 at 11:36 AM, Yuri Gribov <[email protected]> wrote:

> > Does
> > anyone have an idea why PC is 0x00000000?
>
> I think that Asan initialization code (__asan_init) didn't run so
> glibc interceptors were not set up.
>
> > The only un-ordinary
> > part was I had to add some of the sanitizer libraries to LIBS due to
> > undefined symbols like __asan_report8 when building the Python archive
> > libpython3.4m.a:
>
> I don't think this is the right way to go (actually, I believe this is
> the cause of error). You need to make sure that python executable is
> linked with -fsanitize-address - this flag will pass all the necessary
> libs and flags to linker (it's non-trivial!).
>

Yep.


>
> @Kostya: can we add a check in SEGV handler and detect the case when
> pc == NULL? We could report something like "Looks like Asan runtime
> wasn't linked properly, please check your link flags"?
>

We see the stack trace from asan's SEGV handler, so __asan_init was already
called.
PC could be 0 in different cases (just some user function pointer is NULL)

--kcc


>
> -Y
>
> --
> You received this message because you are subscribed to the Google Groups
> "address-sanitizer" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"address-sanitizer" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to