Herbert Xu wrote:
In fact this is the reason why this whole problem is non-existant.
Unless there is a kernel user that sends netlink messages with a timeout
value other than 0, this patch does not resolve any dead-locks at all
since there is none to start with.
yes, you are totally right, but as I told you before, I'm not focusing
my efforts in improving netlink sockets for tools like iproute, for
those MSG_DONTWAIT is fairly ok.
So it looks to me as if this entire patch is simply papering over bugs
in the user-space application where it either isn't processing the
messages fast enough or it needs to enlarge its queue size.
I don't agree, enlarging the queue size for applications that send a lot
of information in a short period of time is just a workaround.
Here is a tip, use separate sockets for unicast and multicast messages.
That way your unicast socket is very unlikely to overrun.
So I'd like to see the whole thing reverted.
Herbert, I don't agree, my patch isn't related to stuff that you've
mentioned. It's fairly true that most çof application use socket timeout
0, but my patch is not for those! :-) well, still disagree?