On Tue, May 16, 2017 at 10:53 AM, Cong Wang <xiyou.wangc...@gmail.com> wrote: > How about adding a check for ->fib_dead in these work's? > If the last treeref is gone, probably it is no longer needed > to continue to do the offloading work.
Hmm, this doesn't look correct either, we may miss an offloading work if fib_release_info() comes first.
