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

Re: [Xen-devel] [PATCH RFC 01/55] x86/mm: defer clearing page in virt_to_xen_lXe



On Fri, Mar 15, 2019 at 08:38:38AM -0600, Jan Beulich wrote:
> >>> On 07.02.19 at 17:44, <wei.liu2@xxxxxxxxxx> wrote:
> > --- a/xen/arch/x86/mm.c
> > +++ b/xen/arch/x86/mm.c
> > @@ -4752,13 +4752,13 @@ static l3_pgentry_t *virt_to_xen_l3e(unsigned long 
> > v)
> >  
> >          if ( !pl3e )
> >              return NULL;
> > -        clear_page(pl3e);
> >          if ( locking )
> >              spin_lock(&map_pgdir_lock);
> >          if ( !(l4e_get_flags(*pl4e) & _PAGE_PRESENT) )
> >          {
> >              l4_pgentry_t l4e = l4e_from_paddr(__pa(pl3e), 
> > __PAGE_HYPERVISOR);
> >  
> > +            clear_page(pl3e);
> 
> Is this really an optimization? You treat avoiding to clear the page
> in a hopefully infrequent case of a race for holding the spin lock
> for quite a bit longer.

I was actually two minded while writing this patch. It is difficult to
prove one way or the other. To avoid distracting from the main issue at
hand, dropping this patch is the best way forward.

Wei.

> 
> Jan
> 
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.