Re: [PATCH v6 0/5] mm: batch TLB flushing for dirty folios in vmscan
From: Zhang Peng
Date: Mon Nov 16 2026 - 07:47:46 EST
On Mon, Nov 16, 2026 at 01:10:22PM +0100, David Hildenbrand (Arm) wrote:
> On 10/31/26 11:35, Zhang Peng wrote:
> > This series introduces batch TLB flushing optimization for dirty folios
> > during memory reclaim, aiming to reduce IPI overhead on multi-core systems.
> >
> > Background
> > ----------
> > Reclaim calls try_to_unmap_flush_dirty() before writing out each dirty
> > folio. When writable TLB entries are pending, this triggers a flush.
> > On systems that use IPIs for TLB shootdowns, frequent flushes can add
> > significant overhead.
> >
>
> I will send out a series soon that fixes various issues with current delayed TLB
> flushing.
>
> It's all such a mess.
>
> There will be one issue that is harder to resolve, which will interact with your
> changes.
>
> I think something along the lines of your approach might come in handy to
> resolve it. I'll CC you once I send it out and we can discuss there.
>
> --
> Cheers,
>
> David
Thanks for the heads-up. I'll take a look at your series once it's posted,
and I'd be happy to discuss how this approach might help.
Thanks,
Zhang Peng