Hello David,
> if people are ok with this, i'll go through and do the same for the link
> state and address change hooks.
I like your approach, and I think same change should be done linkstate
and addrchange hooks as a follow up commit.
I'm OK with your diff.
thanks and
regards
sashan
hook_establish can fail, but drivers are inconsistent about checking for
that. apparently there's also a requirement that detach hooks are
run in opposite order to the one they were established in, but that
is also applied inconsistently by drivers.
this replaces if_detachhooks with a task_list, a
