On 09/03/2016 09:13, Fam Zheng wrote: >> > @@ -352,7 +352,9 @@ static void qed_start_need_check_timer(BDRVQEDState *s) >> > static void qed_cancel_need_check_timer(BDRVQEDState *s) >> > { >> > trace_qed_cancel_need_check_timer(s); >> > - timer_del(s->need_check_timer); >> > + if (s->need_check_timer) { >> > + timer_del(s->need_check_timer); >> > + } >> > } > Not clear why this change is needed in this patch, but it is obviously not > wrong. If this is to mask a bug, it at least deserves a comment. > > The other parts of the patch looks good to me.
I need to check. :) Paolo