> > If the changes don't make it into the stack, I think the minimum 
> > change for that should be to check alignment requirements are 
> > satisfied before casting e.g. p->payload to struct ip_hdr.
> 
> That seems like an excellent debug change.  Could you submit 
> a patch for it.  I promise to give the patch list some 
> attention in the coming days/weeks!

I just noticed my compiler (nios2-elf-gcc v3.4.1) automatically loads
the non-aligned-values (IP address, TCP seqno) automatically using
byte-access if they are located in a packed struct (which explains why
this was working for me when I tested it :-)
So, if adding the check mentioned above, this would break compatibility
again (at least on gcc). Although I don't know if what gcc does here is
standard behaviour...

Knowing this, I don't know if submitting a patch for this is a good
idea.

Oh, and I'm looking forward to the patch attention! *g*

Simon.


_______________________________________________
lwip-users mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/lwip-users

Reply via email to