Hi Simon,

My problem looks similar to what he is facing. He has generalized the issue.
In our implementations it looks like tcpip_thread is accessing the stack of
application thread.

I tried printing the dataptr value before every tcp_write call in
do_writemore function. It is always the same till the crash. I don't think
it is an issue of corrupted memory/pointers.

Thanks,
Anil


On Tue, Sep 11, 2012 at 2:13 PM, Simon Goldschmidt <[email protected]> wrote:

> Anil kumar wrote:
> > Yes, i am using lwip with LWIP_TCPIP_CORE_LOCKING=1.
> >
> > How to use the default one, is it the same code base with different
> > options
> > or is it completely a different code base?
>
> It's a compile-time option only.
>
> > Regarding the crash, i am getting kernel oops with the message "unable to
> > handle kernel paging request at virtual address xxx".  xxx is the address
> > of the data (dataptr).
>
> Could it be that you have stumbled accross the same problem as Artem
> Pisarenko two weeks ago (see this thread:
> http://lists.nongnu.org/archive/html/lwip-users/2012-08/msg00143.html)?
>
> His problem was that his implementation of lwIP threads did not allow
> different threads to access other thread's memory, which is normally not
> the case for threads (e.g. for pthreads), only for processes and lwIP is
> not designed for this.
>
> It could, however, also be that the pointer just got invalid by corrupted
> memory?
>
>
> Simon
>
> _______________________________________________
> lwip-users mailing list
> [email protected]
> https://lists.nongnu.org/mailman/listinfo/lwip-users
>
_______________________________________________
lwip-users mailing list
[email protected]
https://lists.nongnu.org/mailman/listinfo/lwip-users

Reply via email to