This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: libsanitizer merge from upstream r191666


On Wed, Nov 6, 2013 at 3:53 AM, Jakub Jelinek <jakub@redhat.com> wrote:
> On Tue, Nov 05, 2013 at 01:49:43PM -0800, Evgeniy Stepanov wrote:
>> This way we can't test kernel interfaces that are not used in glibc,
>> like linux aio.
>
> So you just test what you can test.  Why do you need to intercept kernel
> aio when hardly anything uses it?  Also, what is the reason why say
> stat* interceptors couldn't just use the glibc stat* (using dlsym
> obtained pointers) and just do some work before and/or after)?

stat* can, should, and do use glibc stat*. But we also have stat*
system call handlers. Those are not wired anywhere and provided as
annotations for any system calls used directly from the user code.

We need aio handlers for MSan, where not having them results in false
positives. ASan gets them for free.

>
>> Most of our team is travelling, and we won't be able to submit a
>> proper fix until next week. Is this blocking anyone? Are you OK with
>> disabling broken parts with #ifdefs for now? Most of this is used in
>> the interceptors, and they can be disabled on an individual basis.
>
> Bootstrap failures are certainly blocking lots of people, so if you could
> disable known problematic parts with #ifdefs for now and resolve later, it
> would be great.
>
>         Jakub


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]