Re: [PATCH v4 3.0-rc2-tip 4/22] 4: Uprobes: register/unregister probes.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



* Peter Zijlstra <peterz@xxxxxxxxxxxxx> [2011-06-16 11:46:22]:

> On Thu, 2011-06-16 at 09:41 +0530, Srikar Dronamraju wrote:
> > * Peter Zijlstra <peterz@xxxxxxxxxxxxx> [2011-06-15 19:41:59]:
> > 
> > > On Tue, 2011-06-07 at 18:29 +0530, Srikar Dronamraju wrote:
> > > > 1. Use mm->owner and walk thro the thread_group of mm->owner, siblings
> > > > of mm->owner, siblings of parent of mm->owner.  This should be
> > > > good list to traverse. Not sure if this is an exhaustive
> > > > enough list that all tasks that have a mm set to this mm_struct are
> > > > walked through. 
> > > 
> > > As per copy_process():
> > > 
> > > 	/*
> > > 	 * Thread groups must share signals as well, and detached threads
> > > 	 * can only be started up within the thread group.
> > > 	 */
> > > 	if ((clone_flags & CLONE_THREAD) && !(clone_flags & CLONE_SIGHAND))
> > > 		return ERR_PTR(-EINVAL);
> > > 
> > > 	/*
> > > 	 * Shared signal handlers imply shared VM. By way of the above,
> > > 	 * thread groups also imply shared VM. Blocking this case allows
> > > 	 * for various simplifications in other code.
> > > 	 */
> > > 	if ((clone_flags & CLONE_SIGHAND) && !(clone_flags & CLONE_VM))
> > > 		return ERR_PTR(-EINVAL);
> > > 
> > > CLONE_THREAD implies CLONE_VM, but not the other way around, we
> > > therefore would be able to CLONE_VM and not be part of the primary
> > > owner's thread group.
> > > 
> > > This is of course all terribly sad..
> > 
> > Agree, 
> > 
> > If clone(CLONE_VM) were to be done by a thread_group leader, we can walk
> > thro the siblings of parent of mm->owner.
> > 
> > However if clone(CLONE_VM) were to be done by non thread_group_leader
> > thread, then we dont even seem to add it to the init_task. i.e I dont
> > think we can refer to such a thread even when we walk thro
> > do_each_thread(g,t) { .. } while_each_thread(g,t);
> > 
> > right?
> 
> No, we initialize p->group_leader = p; and only change that for
> CLONE_THREAD, so a clone without CLONE_THREAD always results in a new
> thread group leader, which are always added to the init_task list.
> 

Ahh .. I missed the p->group_leader = p thing.

In which case, shouldnt traversing all the tasks of all siblings of 
parent of mm->owner should provide us all the the tasks that have linked
to mm. Right?

Agree that we can bother about this a little later.

-- 
Thanks and Regards
Srikar

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>


[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]