> The SOCKHASH_LOCK() addition is not necessary, in fact the correct
> change is to remove them from the ipv4 UDP version. Just like in
> udp_v4_lookup/udp_v6_lookup we know that the code is always called
> from within a BH handler, so the bh atomicity would be spurious.
OK, thanks for the correction. I'll remove the locks and add a comment to make
it clear that the context is always a BH handler.
BTW, it would be very helpful from the standpoint of finding races if the
networking code would include comments indicating the calling context. This is
especially true if the code can be called from multiple contexts.
Regards,
Bill
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu