https://bugs.kde.org/show_bug.cgi?id=456171
--- Comment #7 from obi...@gmail.com <obi...@gmail.com> ---
> So presume the _elf_aux_info call will fail
Yeah it does, that's the raison d'être of this patch.
> My question is where does that memory come from?
It should be copied over on exec ('sys/kern/kern_exec.c', that's also where the
'kern.ps_strings' sysctl is implemented) to userspace at
'PROC_PS_STRINGS(proc)' (which resides somewhere at the beginning of that
process' userstack, cf 'sys/sys/exec.h').
> And I presume (again) that it causes problems because it came from somewhere
> and memcheck didn't see it being allocated.
I'm not entirely sure why Valgrind doesn't track this memory, but in any case
there should be a guarantee that it's okay to access this memory, which is why
I attempted to do what I attempted to do with this patch - ignore accesses to
that memory.
Again though, there may be (and actually probably is) a much more elegant
solution to this which I'm not seeing right now!
> Also this code does look a bit leaky
It indeed shouldn't be much of an issue because 'buf' & 'obuf' are both static
and checked to be NULL before being allocating. I can't test with
'--show-reachable=yes' until Monday, but I also expect it to make memcheck
complain.
Have a nice weekend!
--
You are receiving this mail because:
You are watching all bug changes.
