Re: [RFC PATCH 12/15] kmap: Add stray write protection for device pages

[Ira Weiny]

On Tue, Jul 14, 2020 at 10:44:51AM +0200, Peter Zijlstra wrote:
> On Tue, Jul 14, 2020 at 12:02:17AM -0700, ira.weiny@xxxxxxxxx wrote:
> > From: Ira Weiny <ira.weiny@xxxxxxxxx>
> > 
> > Device managed pages may have additional protections.  These protections
> > need to be removed prior to valid use by kernel users.
> > 
> > Check for special treatment of device managed pages in kmap and take
> > action if needed.  We use kmap as an interface for generic kernel code
> > because under normal circumstances it would be a bug for general kernel
> > code to not use kmap prior to accessing kernel memory.  Therefore, this
> > should allow any valid kernel users to seamlessly use these pages
> > without issues.
> > 
> > Signed-off-by: Ira Weiny <ira.weiny@xxxxxxxxx>
> > ---
> >  include/linux/highmem.h | 32 +++++++++++++++++++++++++++++++-
> >  1 file changed, 31 insertions(+), 1 deletion(-)
> > 
> > diff --git a/include/linux/highmem.h b/include/linux/highmem.h
> > index d6e82e3de027..7f809d8d5a94 100644
> > --- a/include/linux/highmem.h
> > +++ b/include/linux/highmem.h
> > @@ -8,6 +8,7 @@
> >  #include <linux/mm.h>
> >  #include <linux/uaccess.h>
> >  #include <linux/hardirq.h>
> > +#include <linux/memremap.h>
> >  
> >  #include <asm/cacheflush.h>
> >  
> > @@ -31,6 +32,20 @@ static inline void invalidate_kernel_vmap_range(void *vaddr, int size)
> >  
> >  #include <asm/kmap_types.h>
> >  
> > +static inline void enable_access(struct page *page)
> > +{
> > +	if (!page_is_access_protected(page))
> > +		return;
> > +	dev_access_enable();
> > +}
> > +
> > +static inline void disable_access(struct page *page)
> > +{
> > +	if (!page_is_access_protected(page))
> > +		return;
> > +	dev_access_disable();
> > +}
> 
> So, if I followed along correctly, you're proposing to do a WRMSR per
> k{,un}map{_atomic}(), sounds like excellent performance all-round :-(

Only to pages which have this additional protection, ie not DRAM.

User mappings of this memory is not affected (would be covered by User PKeys if
desired).  User mappings to persistent memory are the primary use case and the
performant path.

Ira