re: On softints, softnet_lock and sleeping (aka ipv6 vs USB network interfaces)

[matthew green <mrg%eterna.com.au@localhost>]

Manuel Bouyer writes:
> On Sun, Dec 06, 2015 at 04:01:22PM +0000, Taylor R Campbell wrote:
> > [...]
> >    2) CPU 0 takes clock interrupt and nd6_timer expires.  nd6_timer starts and
> >    tries to take softnet lock and blocks
> 
> a callout shout never sleep. When it does, this delays all other callouts.
> I've already fixed deadlocks like this (the last one being related to
> I2C).

yes - we know this.  but how do you best avoid it in this particular
case is the question here.  Nick and i discussed this elsewhere, but
we don't know the networking layer well enough to answer that..


.mrg.