Re: [PATCH 04/16] ptrace: kill tracehook_notify_jctl()

[Tejun Heo <tj@kernel.org>]

On Mon, Dec 20, 2010 at 03:59:56PM +0100, Oleg Nesterov wrote:
> On 12/06, Tejun Heo wrote:
> >
> > tracehook_notify_jctl() aids in determining whether and what to report
> > to the parent when a task is stopped or continued.  The function also
> > adds an extra requirement that siglock may be released across it,
> > which is currently unused and quite difficult to satisfy in
> > well-defined manner.
> 
> OK. I agree, tracehook_notify_jctl() looks very unobvious, especially
> because it is not really used currently.
> 
> The patch looks correct, except
> 
> > @@ -1853,21 +1850,19 @@ relock:
> >  	if (unlikely(signal->flags & SIGNAL_CLD_MASK)) {
> >  		int why;
> >
> > -		if (signal->flags & SIGNAL_CLD_CONTINUED)
> > +		if (task_ptrace(current) ||
> > +		    (signal->flags & SIGNAL_CLD_CONTINUED))
> >  			why = CLD_CONTINUED;
> >  		else
> >  			why = CLD_STOPPED;
> 
> Hmm, I can't understand this.
> 
> task_ptrace() should not turn CLD_STOPPED in CLD_CONTINUED?
> 
> Looking ahead, it _seems_ that the next patches keep this logic,
> could you explain?

That's the logic from tracehook_notify_jctl() or I think it is
incorrectly.  Yes, the latter.  I got confused the two parameters.  I
thought tracehook_notify_jctl() always returned CLD_CONTINUED when
traced.  The @why is @notified and CLD_CONTINUED is @why.  :-)

I'll drop the above chunk.  Thanks.

-- 
tejun