[Intel-gfx] [FIXES 2/3] drm/i915/userptr: Handle unlocked gup retries

[Chris Wilson]

Quoting Chris Wilson (2019-11-11 14:27:16)
> Quoting Tvrtko Ursulin (2019-11-11 14:19:31)
> > 
> > On 11/11/2019 13:32, Chris Wilson wrote:
> > > Enable gup to retry and fault the pages outside of the mmap_sem lock in
> > > our worker. As we are inside our worker, outside of any critical path,
> > > we can allow the mmap_sem lock to be dropped in order to service a page
> > > fault; this in turn allows the mm to populate the page using a slow
> > > fault handler.
> > > 
> > > Testcase: igt/gem_userptr/userfault
> > 
> > There are no references or explanation on what does this fix?
> 
> gup simply fails if it is not allowed to drop the lock for some faults,
> 
> __get_user_pages_locked:
>                 ret = __get_user_pages(tsk, mm, start, nr_pages, flags, pages,
>                                        vmas, locked);
>                 if (!locked)
>                         /* VM_FAULT_RETRY couldn't trigger, bypass */
>                         return ret;
> 
> userfault being the first time I discovered this even existed. Since we
> are only holding the mmap_sem for the gup (and not protecting anything
> else) we can simply allow gup to drop the lock if it needs to.

Fixes: 5b56d49fc31d ("mm: add locked parameter to get_user_pages_remote()")
-Chris