Re: [PATCH 2/2] task_work: check callback if it's NULL

[li guang]

å 2013-03-15äç 09:43 +0800ïLi Zefanåéï
> On 2013/3/15 9:26, li guang wrote:
> > å 2013-03-15äç 09:01 +0800ïLi Zefanåéï
> >> On 2013/3/15 8:20, li guang wrote:
> >>> å 2013-03-14åç 15:43 +0100ïOleg Nesterovåéï
> >>>> On 03/14, liguang wrote:
> >>>>>
> >>>>> Signed-off-by: liguang <lig.fnst@xxxxxxxxxxxxxx>
> >>>>> ---
> >>>>>  kernel/task_work.c |    3 ++-
> >>>>>  1 files changed, 2 insertions(+), 1 deletions(-)
> >>>>>
> >>>>> diff --git a/kernel/task_work.c b/kernel/task_work.c
> >>>>> index 0bf4258..f458b08 100644
> >>>>> --- a/kernel/task_work.c
> >>>>> +++ b/kernel/task_work.c
> >>>>> @@ -75,7 +75,8 @@ void task_work_run(void)
> >>>>>
> >>>>>  		do {
> >>>>>  			next = work->next;
> >>>>> -			work->func(work);
> >>>>> +			if (unlikely(work->func))
> >>>>> +				work->func(work);
> >>>>
> >>>> Why?
> >>>>
> >>>> Oleg.
> >>>>
> >>>
> >>> can we believe a callback always be call-able?
> >>> can it happened to be 0? e.g. wrong initialized.
> >>> of course, we can complain the caller, be why don't
> >>> we easily make it more safer?
> >>>
> >>
> >> Because you're not making things safer, but your're trying
> >> to cover up bugs...
> >>
> > 
> > Oh, that's a little harsh to a normal programmer like me :-)
> > for it seems you are asking me to be coding without any bug.
> > are you? or it is the theory of kernel coding?
> > 
> 
> And you make a bug, and you want the kernel to cover up the bug
> instead of crash on a null pointer deref so you'll know you've
> made a bug?
> 
> Why we check if a callback is NULL before calling it? Because
> it's allowed to be. Why we don't check if a callback is NULL?
> Because it's not supposed to be.
> 

OK, Thanks for your reminder.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/