[Intel-gfx] [PATCH v3 0/7] Make the rest of the VFIO driver interface use vfio_device

Jason Gunthorpe jgg at nvidia.com
Thu May 5 19:44:57 UTC 2022


On Thu, May 05, 2022 at 12:56:14PM -0600, Alex Williamson wrote:

> I wasn't cc'd on that version.  It looks to me like we'd still prefer
> to see acks from GVT-g maintainers, Zhenyu and Zhi.

Somehow the entire To line got wiped out, still unclear why. Maybe
get_maintainers had a bad time. I'm watching for this now

> Also, I was thinking of posting the below cleanup patch unless you'd
> prefer to roll it in.

If it is the style, I will roll it in

> Regarding your other outstanding patches, I think all of these depend
> on the IOMMU changes, please correct if there are any that can be
> queued with only the GVT-g topic branch dependency:
> 
> Subject: [PATCH v3 0/8] Remove vfio_group from the struct file facing VFIO API
> Date: Wed,  4 May 2022 16:14:38 -0300
> https://lore.kernel.org/all/0-v3-f7729924a7ea+25e33-vfio_kvm_no_group_jgg@nvidia.com/

This one applies cleanly without the iommu series, I just confirmed it.
 
> Subject: [PATCH] vfio: Delete container_q
> Date: Fri, 29 Apr 2022 15:46:17 -0300
> https://lore.kernel.org/all/0-v1-a1e8791d795b+6b-vfio_container_q_jgg@nvidia.com/

This one needs the iommu series

> And I'm waiting for a respin based on comments for:
> 
> Subject: [PATCH v3 0/2] Remove vfio_device_get_from_dev()
> Date: Wed,  4 May 2022 16:01:46 -0300
> https://lore.kernel.org/all/0-v3-4adf6c1b8e7c+170-vfio_get_from_dev_jgg@nvidia.com/

I will do this in a few hours hopefully

> If there are others I should be tracking, please let me know.  Thanks,

Nothing more for your tree, but FYI:

The enforced coherent series is going through Joerg's tree
https://lore.kernel.org/linux-iommu/0-v3-2cf356649677+a32-intel_no_snoop_jgg@nvidia.com/

The bug fix for the iommu series is pending Joerg:
https://lore.kernel.org/linux-iommu/0-v2-f62259511ac0+6-iommu_dma_block_jgg@nvidia.com/

I've got one more series to post which just needs a last pass over
now, maybe today, but that is looking grim again.

It is alot of series, it is hard to keep all this organized, thanks.

Jason


More information about the Intel-gfx mailing list