Re: [RFC PATCH 5/8] mm: Pass 'address' to map to do_set_pte() and drop FAULT_FLAG_PREFAULT

[Will Deacon]

On Thu, Jan 14, 2021 at 10:17:22AM -0800, Linus Torvalds wrote:
> On Thu, Jan 14, 2021 at 10:01 AM Will Deacon <will@xxxxxxxxxx> wrote:
> >
> > Rather than modifying the 'address' field of the 'struct vm_fault_info'
> > passed to do_set_pte(), leave that to identify the real faulting address
> > and pass in the virtual address to be mapped by the new pte as a
> > separate argument.
> 
> Ack.
> 
> In fact - apart from the question I had about the 'info' sub-structure
> - ack on the whole series. But this one struck me particularly as
> "that's simpler and clearer" even if that finish_fault() case is now
> not as pretty (but with an unnamed structure it would be slightly
> simpler, at least).

Thanks. I'll post another spin next week to avoid the named structure, but
if you (or anybody else) get a chance to glance over some of the other
things I mentioned in the cover letter beforehand that would be really
great.

Cheers,

Will