Re: [PATCH v3] userfaultfd: opportunistic TLB-flush batching for present pages in MOVE

[Lokesh Gidra]

On Thu, Aug 7, 2025 at 3:54 PM Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> On Thu, 7 Aug 2025 15:16:57 -0400 Peter Xu <peterx@xxxxxxxxxx> wrote:
>
> > Hi, Lokesh,
> >
> > On Thu, Aug 07, 2025 at 03:39:02AM -0700, Lokesh Gidra wrote:
> > > MOVE ioctl's runtime is dominated by TLB-flush cost, which is required
> > > for moving present pages. Mitigate this cost by opportunistically
> > > batching present contiguous pages for TLB flushing.
> > >
> > > Without batching, in our testing on an arm64 Android device with UFFD GC,
> > > which uses MOVE ioctl for compaction, we observed that out of the total
> > > time spent in move_pages_pte(), over 40% is in ptep_clear_flush(), and
> > > ~20% in vm_normal_folio().
> > >
> > > With batching, the proportion of vm_normal_folio() increases to over
> > > 70% of move_pages_pte() without any changes to vm_normal_folio().
> >
> > Do you know why vm_normal_folio() could be expensive? I still see quite
> > some other things this path needs to do.
>
> Maybe as explained here?
> https://lkml.kernel.org/r/20250807185819.199865-1-lorenzo.stoakes@xxxxxxxxxx
>
Thanks for sharing this, Andrew. IMHO, this seems like the most likely
reason to me. There is nothing there other than a cold access to the
page struct.

>