Re: [RFC v5 02/11] mm: Prepare for FAULT_FLAG_SPECULATIVE

[Peter Zijlstra <peterz@infradead.org>]

On Tue, Aug 08, 2017 at 03:54:01PM +0530, Anshuman Khandual wrote:
> On 06/16/2017 11:22 PM, Laurent Dufour wrote:
> > From: Peter Zijlstra <peterz@infradead.org>
> > 
> > When speculating faults (without holding mmap_sem) we need to validate
> > that the vma against which we loaded pages is still valid when we're
> > ready to install the new PTE.
> > 
> > Therefore, replace the pte_offset_map_lock() calls that (re)take the
> > PTL with pte_map_lock() which can fail in case we find the VMA changed
> > since we started the fault.
> 
> Where we are checking if VMA has changed or not since the fault ?

Not there yet, this is what you call a preparatory patch. They help
review in that you can consider smaller steps.

> > diff --git a/mm/memory.c b/mm/memory.c
> > index fd952f05e016..40834444ea0d 100644
> > --- a/mm/memory.c
> > +++ b/mm/memory.c
> > @@ -2240,6 +2240,12 @@ static inline void wp_page_reuse(struct vm_fault *vmf)
> >  	pte_unmap_unlock(vmf->pte, vmf->ptl);
> >  }
> >  
> > +static bool pte_map_lock(struct vm_fault *vmf)
> > +{
> > +	vmf->pte = pte_offset_map_lock(vmf->vma->vm_mm, vmf->pmd, vmf->address, &vmf->ptl);
> > +	return true;
> > +}
> 
> This is always true ? Then we should not have all these if (!pte_map_lock(vmf))
> check blocks down below.

Later patches will make it possible to return false. This patch is about
the placing this call. Having this in a separate patch makes it easier
to review all those new error conditions.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>