From: Alex Williamson <alex.williamson@redhat.com>
To: Yi Liu <yi.l.liu@intel.com>
Cc: linux-s390@vger.kernel.org, yi.y.sun@linux.intel.com,
mjrosato@linux.ibm.com, kvm@vger.kernel.org,
intel-gvt-dev@lists.freedesktop.org, joro@8bytes.org,
cohuck@redhat.com, peterx@redhat.com,
suravee.suthikulpanit@amd.com, eric.auger@redhat.com,
nicolinc@nvidia.com, shameerali.kolothum.thodi@huawei.com,
jgg@nvidia.com, intel-gfx@lists.freedesktop.org,
chao.p.peng@linux.intel.com, lulu@redhat.com,
robin.murphy@arm.com, jasowang@redhat.com
Subject: Re: [Intel-gfx] [PATCH v3 00/15] Add vfio_device cdev for iommufd support
Date: Mon, 13 Feb 2023 12:47:19 -0700 [thread overview]
Message-ID: <20230213124719.126eb828.alex.williamson@redhat.com> (raw)
In-Reply-To: <20230213151348.56451-1-yi.l.liu@intel.com>
On Mon, 13 Feb 2023 07:13:33 -0800
Yi Liu <yi.l.liu@intel.com> wrote:
> Existing VFIO provides group-centric user APIs for userspace. Userspace
> opens the /dev/vfio/$group_id first before getting device fd and hence
> getting access to device. This is not the desired model for iommufd. Per
> the conclusion of community discussion[1], iommufd provides device-centric
> kAPIs and requires its consumer (like VFIO) to be device-centric user
> APIs. Such user APIs are used to associate device with iommufd and also
> the I/O address spaces managed by the iommufd.
>
> This series first introduces a per device file structure to be prepared
> for further enhancement and refactors the kvm-vfio code to be prepared
> for accepting device file from userspace. Then refactors the vfio to be
> able to handle iommufd binding. This refactor includes the mechanism of
> blocking device access before iommufd bind, making vfio_device_open() be
> exclusive between the group path and the cdev path. Eventually, adds the
> cdev support for vfio device, and makes group infrastructure optional as
> it is not needed when vfio device cdev is compiled.
>
> This is also a prerequisite for iommu nesting for vfio device[2].
>
> The complete code can be found in below branch, simple test done with the
> legacy group path and the cdev path. Draft QEMU branch can be found at[3]
>
> https://github.com/yiliu1765/iommufd/tree/vfio_device_cdev_v3
> (config CONFIG_IOMMUFD=y CONFIG_VFIO_DEVICE_CDEV=y)
Even using your branch[1], it seems like this has not been tested
except with cdev support enabled:
/home/alwillia/Work/linux.git/drivers/vfio/vfio_main.c: In function ‘vfio_device_add’:
/home/alwillia/Work/linux.git/drivers/vfio/vfio_main.c:253:48: error: ‘struct vfio_device’ has no member named ‘cdev’; did you mean ‘dev’?
253 | ret = cdev_device_add(&device->cdev, &device->device);
| ^~~~
| dev
/home/alwillia/Work/linux.git/drivers/vfio/vfio_main.c: In function ‘vfio_device_del’:
/home/alwillia/Work/linux.git/drivers/vfio/vfio_main.c:262:42: error: ‘struct vfio_device’ has no member named ‘cdev’; did you mean ‘dev’?
262 | cdev_device_del(&device->cdev, &device->device);
| ^~~~
| dev
Additionally the VFIO_ENABLE_GROUP Kconfig option doesn't make much
sense to me, it seems entirely redundant to VFIO_GROUP.
I think it's too late for v6.3 already, but given this needs at least
one more spin, let's set expectations of this being v6.4 material. Thanks,
Alex
[1] 98491da60ae1 cover-letter: Add vfio_device cdev for iommufd support
next prev parent reply other threads:[~2023-02-13 19:47 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-13 15:13 [Intel-gfx] [PATCH v3 00/15] Add vfio_device cdev for iommufd support Yi Liu
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 01/15] vfio: Allocate per device file structure Yi Liu
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 02/15] vfio: Refine vfio file kAPIs Yi Liu
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 03/15] vfio: Accept vfio device file in the driver facing kAPI Yi Liu
2023-02-13 23:21 ` Alex Williamson
2023-02-14 2:19 ` Liu, Yi L
2023-02-13 23:43 ` Jason Gunthorpe
2023-02-14 2:02 ` Liu, Yi L
2023-02-14 7:19 ` Liu, Yi L
2023-02-17 10:55 ` Liu, Yi L
2023-02-17 15:59 ` Jason Gunthorpe
2023-02-18 2:54 ` Liu, Yi L
2023-02-15 12:38 ` Jason Gunthorpe
2023-02-15 14:43 ` Liu, Yi L
2023-02-15 14:46 ` Jason Gunthorpe
2023-02-15 15:32 ` Alex Williamson
2023-02-15 17:04 ` Jason Gunthorpe
2023-02-15 17:19 ` Alex Williamson
2023-02-15 17:33 ` Jason Gunthorpe
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 04/15] kvm/vfio: Rename kvm_vfio_group to prepare for accepting vfio device fd Yi Liu
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 05/15] kvm/vfio: Accept vfio device file from userspace Yi Liu
2023-02-14 22:26 ` Alex Williamson
2023-02-14 23:25 ` Jason Gunthorpe
2023-02-14 23:42 ` Alex Williamson
2023-02-15 0:17 ` Jason Gunthorpe
2023-02-15 0:27 ` Timothy Pearson
2023-02-17 5:34 ` Liu, Yi L
2023-02-17 5:48 ` Liu, Yi L
2023-02-17 16:00 ` Jason Gunthorpe
2023-02-15 7:37 ` Liu, Yi L
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 06/15] vfio: Pass struct vfio_device_file * to vfio_device_open/close() Yi Liu
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 07/15] vfio: Block device access via device fd until device is opened Yi Liu
2023-02-14 22:46 ` Alex Williamson
2023-02-15 6:12 ` Liu, Yi L
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 08/15] vfio: Add infrastructure for bind_iommufd from userspace Yi Liu
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 09/15] vfio-iommufd: Add detach_ioas support for physical VFIO devices Yi Liu
2023-02-14 8:05 ` Tian, Kevin
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 10/15] vfio-iommufd: Add detach_ioas for emulated " Yi Liu
2023-02-14 8:06 ` Tian, Kevin
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 11/15] vfio: Add cdev_device_open_cnt to vfio_group Yi Liu
2023-02-14 8:18 ` Tian, Kevin
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 12/15] vfio: Make vfio_device_open() single open for device cdev path Yi Liu
2023-02-14 8:25 ` Tian, Kevin
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 13/15] vfio: Add cdev for vfio_device Yi Liu
2023-02-14 8:32 ` Tian, Kevin
2023-02-14 8:35 ` Liu, Yi L
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 14/15] vfio: Add ioctls for device cdev using iommufd Yi Liu
2023-02-14 8:53 ` Tian, Kevin
2023-02-14 23:39 ` Yan Zhao
2023-02-15 2:04 ` Tian, Kevin
2023-02-15 7:37 ` Liu, Yi L
2023-02-16 8:24 ` Yan Zhao
2023-02-16 9:10 ` Liu, Yi L
2023-02-16 9:23 ` Yan Zhao
2023-02-16 10:28 ` Liu, Yi L
2023-02-16 14:24 ` Jason Gunthorpe
2023-02-13 15:13 ` [Intel-gfx] [PATCH v3 15/15] vfio: Compile group optionally Yi Liu
2023-02-13 15:30 ` [Intel-gfx] ✗ Fi.CI.BUILD: failure for Add vfio_device cdev for iommufd support (rev2) Patchwork
2023-02-13 19:47 ` Alex Williamson [this message]
2023-02-13 23:21 ` [Intel-gfx] [PATCH v3 00/15] Add vfio_device cdev for iommufd support Jason Gunthorpe
2023-02-14 15:15 ` Liu, Yi L
2023-02-14 15:54 ` Alex Williamson
2023-02-14 16:48 ` Jason Gunthorpe
2023-02-14 1:55 ` Liu, Yi L
2023-02-14 15:47 ` Alex Williamson
2023-02-15 7:54 ` Liu, Yi L
2023-02-15 20:09 ` Alex Williamson
2023-02-16 2:53 ` Liu, Yi L
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20230213124719.126eb828.alex.williamson@redhat.com \
--to=alex.williamson@redhat.com \
--cc=chao.p.peng@linux.intel.com \
--cc=cohuck@redhat.com \
--cc=eric.auger@redhat.com \
--cc=intel-gfx@lists.freedesktop.org \
--cc=intel-gvt-dev@lists.freedesktop.org \
--cc=jasowang@redhat.com \
--cc=jgg@nvidia.com \
--cc=joro@8bytes.org \
--cc=kvm@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=lulu@redhat.com \
--cc=mjrosato@linux.ibm.com \
--cc=nicolinc@nvidia.com \
--cc=peterx@redhat.com \
--cc=robin.murphy@arm.com \
--cc=shameerali.kolothum.thodi@huawei.com \
--cc=suravee.suthikulpanit@amd.com \
--cc=yi.l.liu@intel.com \
--cc=yi.y.sun@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox