[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CACGkMEuEOOkv7pkSo21mdqPpRZJm59SQX_ZGG1Br_v_iN-vWiA@mail.gmail.com>
Date: Mon, 22 Sep 2025 09:06:02 +0800
From: Jason Wang <jasowang@...hat.com>
To: "Michael S. Tsirkin" <mst@...hat.com>
Cc: xuanzhuo@...ux.alibaba.com, eperezma@...hat.com,
virtualization@...ts.linux.dev, linux-kernel@...r.kernel.org,
hch@...radead.org
Subject: Re: [PATCH V6 0/9] Refine virtio mapping API
On Mon, Sep 22, 2025 at 6:15 AM Michael S. Tsirkin <mst@...hat.com> wrote:
>
> On Thu, Aug 21, 2025 at 02:46:32PM +0800, Jason Wang wrote:
> > Hi all:
> >
> > Virtio used to be coupled with DMA API. This works fine for the device
> > that do real DMA but not the others. For example, VDUSE nees to craft
> > with DMA API in order to let the virtio-vdpa driver to work.
> >
> > This series tries to solve this issue by introducing the mapping API
> > in the virtio core. So transport like vDPA can implement their own
> > mapping logic without the need to hack with DMA API. The mapping API
> > are abstracted with a new map operations in order to be re-used by
> > transprot or device. So device like VDUSE can implement its own
> > mapping loigc.
> >
> > For device that uses DMA (for example PCI device), the virtio core
> > will still call DMA API directly without the need of implementing map
> > ops per device/transport.
> >
> > Please review.
>
> I sent some typos i found.
> Pls send fixups on top to fix them so testing in linux-next is not invalidated.
> Thanks!
>
Will do, I see the HEAD in linux-next is
vdpa: support virtio_map
Is this intended? Would you want me to send fixes on top?
Thanks
>
>
> > Changes since V5:
> >
> > - Rename mapping_token to virtio_map
> > - Do not use opaque void * pointer, just use a forward decalration of
> > vduse_iova_domain
> > - Remove unused variable and typo fixes
> >
> > Changes since V4:
> >
> > - Rename map_token to mapping_token
> > - Introduce a union container for opaque token as well as the DMA
> > device so we won't lose the type safety
> > - Do not try to set DMA mask for VDUSE device
> > - Introduce a new mapper_error op for API completeness
> >
> > Changes since V3:
> >
> > - Fix build error of PDS vDPA driver
> >
> > Changes since V2:
> >
> > - Drop VDUSE dependenct for HAS_DMA and ARCH_HAS_DMA_OPS
> >
> > Changes since V1:
> >
> > - Fix build error of mlx5_vdpa driver
> >
> >
> > Jason Wang (9):
> > virtio_ring: constify virtqueue pointer for DMA helpers
> > virtio_ring: switch to use dma_{map|unmap}_page()
> > virtio: rename dma helpers
> > virtio: introduce virtio_map container union
> > virtio_ring: rename dma_handle to map_handle
> > virtio: introduce map ops in virtio core
> > vdpa: support virtio_map
> > vdpa: introduce map ops
> > vduse: switch to use virtio map API instead of DMA API
> >
> > drivers/net/virtio_net.c | 28 +-
> > drivers/vdpa/Kconfig | 8 +-
> > drivers/vdpa/alibaba/eni_vdpa.c | 5 +-
> > drivers/vdpa/ifcvf/ifcvf_main.c | 5 +-
> > drivers/vdpa/mlx5/core/mr.c | 4 +-
> > drivers/vdpa/mlx5/net/mlx5_vnet.c | 15 +-
> > drivers/vdpa/octeon_ep/octep_vdpa_main.c | 6 +-
> > drivers/vdpa/pds/vdpa_dev.c | 5 +-
> > drivers/vdpa/solidrun/snet_main.c | 8 +-
> > drivers/vdpa/vdpa.c | 5 +-
> > drivers/vdpa/vdpa_sim/vdpa_sim.c | 4 +-
> > drivers/vdpa/vdpa_user/iova_domain.c | 2 +-
> > drivers/vdpa/vdpa_user/iova_domain.h | 2 +-
> > drivers/vdpa/vdpa_user/vduse_dev.c | 79 ++--
> > drivers/vdpa/virtio_pci/vp_vdpa.c | 5 +-
> > drivers/vhost/vdpa.c | 6 +-
> > drivers/virtio/virtio_ring.c | 459 ++++++++++++++---------
> > drivers/virtio/virtio_vdpa.c | 20 +-
> > include/linux/vdpa.h | 25 +-
> > include/linux/virtio.h | 46 ++-
> > include/linux/virtio_config.h | 72 ++++
> > include/linux/virtio_ring.h | 7 +-
> > 22 files changed, 531 insertions(+), 285 deletions(-)
> >
> > --
> > 2.31.1
>
Powered by blists - more mailing lists