Re: [PATCH]: exec: avoid propagating PF_NO_SETAFFINITY into userspace child

[Oleg Nesterov <oleg@redhat.com>]

On 11/28, Peter Zijlstra wrote:
>
> On Wed, Nov 27, 2013 at 01:31:17PM -0500, Tejun Heo wrote:
> > On Wed, Nov 27, 2013 at 10:07:03AM +0800, zhang.yi20@zte.com.cn wrote:
> > > Userspace process doesn't want the PF_NO_SETAFFINITY, but its parent may be
> > > a kernel worker thread which has PF_NO_SETAFFINITY set, and this worker thread
> > > can do kernel_thread() to create the child.
> > > Clearing this flag in usersapce child to enable its migrating capability.
> > >
> > >
> > > Signed-off-by: Zhang Yi <zhang.yi20@zte.com.cn>
> >
> > Acked-by: Tejun Heo <tj@kernel.org>
> >
> > cc'ing Ingo and Peter.  Ingo, I think this one doesn't really suit the
> > workqueue tree.  Can you please pick this one up w/ Oleg's ack added
> > and stable cc'd?  The original patch is
> >
> >   http://article.gmane.org/gmane.linux.kernel/1602429/raw
>
> So I don't get the problem; aren't all usermode helper thingies spawned
> by the khelper task, which doesn't have PG_NO_SETAFFINITY set?

It has? khelper is a workqueue thread, this flag is set by create_worker().

And it does kernel_thread() (not kthread_create()) so the child gets this
flag too.

> The Changelog is not explaining anything much -- so no I will not take
> this patch.

Well, perhaps the changelog can be more verbose and clear...

Zhang, I'm afraid you have to send v3 ;)

Oleg.