Re: [RFC PATCH v3 12/25] block, mm: Support PISA for disk I/O and scrub migration

From: Christoph Hellwig

Date: Fri Nov 20 2026 - 11:17:32 EST


On Fri, Nov 20, 2026 at 01:06:42AM +0000, Luigi Rizzo wrote:
> Extend PISA user/page-cache allocators to support O_DIRECT and buffered
> disk I/O, gated per block/controller device through pisa_rx_bufs (read)
> and pisa_tx_bufs (write):
>
> - Split per-node user_pa into DMA_BIDIRECTIONAL, DMA_TO_DEVICE, and
> DMA_FROM_DEVICE allocators, promoting pages to DMA_BIDIRECTIONAL if
> later accessed in the opposite direction.
> - Upgrade pinned O_DIRECT user pages in bio_iov_iter_get_pages(), and
> allocate/promote order-0 page-cache folios for buffered I/O when PISA
> is enabled on the target block device.
> - Add pisa.scrub_migrate to compact or demote lingering folios off old,
> sparse user_pa 2MB pages.

No, doing MM migrations in an (any) I/O path makes no sense at all.

Can we all please go back to actually thinking about problems instead
of generating giants amounts of shit code without the slightest
reflection.

Than you very much for your attention.