Re: [PATCH RFC] use of activate_mm in fs/aio.c:use_mm()?

Re: [PATCH RFC] use of activate_mm in fs/aio.c:use_mm()?

[Andrew Morton]

On Wed, 06 Dec 2006 17:36:02 -0800
Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Jeremy Fitzhardinge wrote:
> > I'm wondering if activate_mm() is the right thing to be using in
> > use_mm(); shouldn't this be switch_mm()?
> >
> > On normal x86, they're synonymous, but for the Xen patches I'm adding a
> > hook which assumes that activate_mm is only used the first time a new mm
> > is used after creation (I have another hook for dealing with dup_mm).  I
> > think this use of activate_mm() is the only place where it could be used
> > a second time on an mm.
> >
> > From a quick look at the other architectures I think this is OK (most
> > simply implement one in terms of the other), but some are doing some
> > subtly different stuff between the two.
> >
> > Thanks,
> >     J
> >
> >
> >   
> Er, lets try that again:
> 
> diff -r 455b71ed4525 fs/aio.c
> --- a/fs/aio.c	Wed Dec 06 13:16:42 2006 -0800
> +++ b/fs/aio.c	Wed Dec 06 17:17:43 2006 -0800
> @@ -588,7 +588,7 @@ static void use_mm(struct mm_struct *mm)
>  	 * Note that on UML this *requires* PF_BORROWED_MM to be set, otherwise
>  	 * it won't work. Update it accordingly if you change it here
>  	 */
> -	activate_mm(active_mm, mm);
> +	switch_mm(active_mm, mm, tsk);
>  	task_unlock(tsk);
>  
>  	mmdrop(active_mm);

That to me sounds like a reasonable description of the difference between
activate_mm() and switch_mm().  And the change appears reasonable as well.

But it is a change which the architecture maintainers would need to have a
think about, please.

Re: [PATCH RFC] use of activate_mm in fs/aio.c:use_mm()?

[David Miller]

From: Andrew Morton <akpm@osdl.org>
Date: Fri, 8 Dec 2006 15:45:22 -0800

> On Wed, 06 Dec 2006 17:36:02 -0800
> Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> 
> > Jeremy Fitzhardinge wrote:
> > > I'm wondering if activate_mm() is the right thing to be using in
> > > use_mm(); shouldn't this be switch_mm()?
> > >
> > > On normal x86, they're synonymous, but for the Xen patches I'm adding a
> > > hook which assumes that activate_mm is only used the first time a new mm
> > > is used after creation (I have another hook for dealing with dup_mm).  I
> > > think this use of activate_mm() is the only place where it could be used
> > > a second time on an mm.
> > >
> > > From a quick look at the other architectures I think this is OK (most
> > > simply implement one in terms of the other), but some are doing some
> > > subtly different stuff between the two.
> > >
> > > Thanks,
> > >     J
> > >
> > >
> > >   
> > Er, lets try that again:
> > 
> > diff -r 455b71ed4525 fs/aio.c
> > --- a/fs/aio.c	Wed Dec 06 13:16:42 2006 -0800
> > +++ b/fs/aio.c	Wed Dec 06 17:17:43 2006 -0800
> > @@ -588,7 +588,7 @@ static void use_mm(struct mm_struct *mm)
> >  	 * Note that on UML this *requires* PF_BORROWED_MM to be set, otherwise
> >  	 * it won't work. Update it accordingly if you change it here
> >  	 */
> > -	activate_mm(active_mm, mm);
> > +	switch_mm(active_mm, mm, tsk);
> >  	task_unlock(tsk);
> >  
> >  	mmdrop(active_mm);
> 
> That to me sounds like a reasonable description of the difference between
> activate_mm() and switch_mm().  And the change appears reasonable as well.
> 
> But it is a change which the architecture maintainers would need to have a
> think about, please.

This looks absolutely correct to me.