Commit Graph

20 Commits

Author SHA1 Message Date
Xie Yongji
c8a6153b6c vduse: Introduce VDUSE - vDPA Device in Userspace
This VDUSE driver enables implementing software-emulated vDPA
devices in userspace. The vDPA device is created by
ioctl(VDUSE_CREATE_DEV) on /dev/vduse/control. Then a char device
interface (/dev/vduse/$NAME) is exported to userspace for device
emulation.

In order to make the device emulation more secure, the device's
control path is handled in kernel. A message mechnism is introduced
to forward some dataplane related control messages to userspace.

And in the data path, the DMA buffer will be mapped into userspace
address space through different ways depending on the vDPA bus to
which the vDPA device is attached. In virtio-vdpa case, the MMU-based
software IOTLB is used to achieve that. And in vhost-vdpa case, the
DMA buffer is reside in a userspace memory region which can be shared
to the VDUSE userspace processs via transferring the shmfd.

For more details on VDUSE design and usage, please see the follow-on
Documentation commit.

NB(mst): when merging this with
b542e383d8 ("eventfd: Make signal recursion protection a task bit")
replace eventfd_signal_count with eventfd_signal_allowed,
and drop the previous
("eventfd: Export eventfd_wake_count to modules").

Signed-off-by: Xie Yongji <xieyongji@bytedance.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Link: https://lore.kernel.org/r/20210831103634.33-13-xieyongji@bytedance.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-09-06 07:20:58 -04:00
Eli Cohen
5262912ef3 vdpa/mlx5: Add support for control VQ and MAC setting
Add support to handle control virtqueue configurations per virtio
specification. The control virtqueue is implemented in software and no
hardware offloading is involved.

Control VQ configuration need task context, therefore all configurations
are handled in a workqueue created for the purpose.

Modifications are made to the memory registration code to allow for
saving a copy of itolb to be used by the control VQ to access the vring.

The max number of data virtqueus supported by the driver has been
updated to 2 since multiqueue is not supported at this stage and we need
to ensure consistency of VQ indices mapping to either data or control
VQ.

Signed-off-by: Eli Cohen <elic@nvidia.com>
Link: https://lore.kernel.org/r/20210823052123.14909-6-elic@nvidia.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-09-05 16:23:08 -04:00
Max Gurtovoy
0c853c2c29 vdpa: add vdpa simulator for block device
This will allow running vDPA for virtio block protocol.

It's a preliminary implementation with a simple request handling:
for each request, only the status (last byte) is set.
It's always set to VIRTIO_BLK_S_OK.

Also input validation is missing and will be added in the next commits.

Signed-off-by: Max Gurtovoy <mgurtovoy@nvidia.com>
[sgarzare: various cleanups/fixes]
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Link: https://lore.kernel.org/r/20210315163450.254396-12-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-05-03 04:55:53 -04:00
Stefano Garzarella
4080fc1067 vdpa_sim: use iova module to allocate IOVA addresses
The identical mapping used until now created issues when mapping
different virtual pages with the same physical address.
To solve this issue, we can use the iova module, to handle the IOVA
allocation.
For simplicity we use an IOVA allocator with byte granularity.

We add two new functions, vdpasim_map_range() and vdpasim_unmap_range(),
to handle the IOVA allocation and the registration into the IOMMU/IOTLB.
These functions are used by dma_map_ops callbacks.

Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Link: https://lore.kernel.org/r/20210315163450.254396-2-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-05-03 04:55:52 -04:00
Jason Wang
64b9f64f80 vdpa: introduce virtio pci driver
This patch introduce a vDPA driver for virtio-pci device. It bridges
the virtio-pci control command to the vDPA bus. This will be used for
features prototyping and testing.

Note that get/restore virtqueue state is not supported which needs
extension on the virtio specification.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Link: https://lore.kernel.org/r/20210223061905.422659-4-jasowang@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-05-03 04:55:51 -04:00
Parav Pandit
33b347503f vdpa: Define vdpa mgmt device, ops and a netlink interface
To add one or more VDPA devices, define a management device which
allows adding or removing vdpa device. A management device defines
set of callbacks to manage vdpa devices.

To begin with, it defines add and remove callbacks through which a user
defined vdpa device can be added or removed.

A unique management device is identified by its unique handle identified
by management device name and optionally the bus name.

Hence, introduce routine through which driver can register a
management device and its callback operations for adding and remove
a vdpa device.

Introduce vdpa netlink socket family so that user can query management
device and its attributes.

Example of show vdpa management device which allows creating vdpa device of
networking class (device id = 0x1) of virtio specification 1.1
section 5.1.1.

$ vdpa mgmtdev show
vdpasim_net:
  supported_classes:
    net

Example of showing vdpa management device in JSON format.

$ vdpa mgmtdev show -jp
{
    "show": {
        "vdpasim_net": {
            "supported_classes": [ "net" ]
        }
    }
}

Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Eli Cohen <elic@nvidia.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
Link: https://lore.kernel.org/r/20210105103203.82508-4-parav@nvidia.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

Including a bugfix:

vpda: correctly size vdpa_nl_policy

We need to ensure last entry of vdpa_nl_policy[]
is zero, otherwise out-of-bounds access is hurting us.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Parav Pandit <parav@nvidia.com>
Cc: Eli Cohen <elic@nvidia.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Link: https://lore.kernel.org/r/20210210134911.4119555-1-eric.dumazet@gmail.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-02-23 07:52:56 -05:00
Max Gurtovoy
db1e8bb6c6 vdpa: split vdpasim to core and net modules
Introduce new vdpa_sim_net and vdpa_sim (core) drivers. This is a
preparation for adding a vdpa simulator module for block devices.

Signed-off-by: Max Gurtovoy <mgurtovoy@nvidia.com>
[sgarzare: various cleanups/fixes]
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Link: https://lore.kernel.org/r/20201215144256.155342-19-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-12-18 16:14:30 -05:00
Stefano Garzarella
29b90f92ee vdpa: remove unnecessary 'default n' in Kconfig entries
'default n' is not necessary since it is already the default when
nothing is specified.

Suggested-by: Jason Wang <jasowang@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Link: https://lore.kernel.org/r/20201215144256.155342-2-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-12-18 16:14:28 -05:00
Randy Dunlap
98701a2a86 vdpa: mlx5: fix vdpa/vhost dependencies
drivers/vdpa/mlx5/ uses vhost_iotlb*() interfaces, so select
VHOST_IOTLB to make them be built.

However, if VHOST_IOTLB is the only VHOST symbol that is
set/enabled, the object file still won't be built because
drivers/Makefile won't descend into drivers/vhost/ to build it,
so make drivers/Makefile build the needed binary whenever
VHOST_IOTLB is set, like it does for VHOST_RING.

Fixes these build errors:
ERROR: modpost: "vhost_iotlb_itree_next" [drivers/vdpa/mlx5/mlx5_vdpa.ko] undefined!
ERROR: modpost: "vhost_iotlb_itree_first" [drivers/vdpa/mlx5/mlx5_vdpa.ko] undefined!

Fixes: 29064bfdab ("vdpa/mlx5: Add support library for mlx5 VDPA implementation")
Fixes: aff90770e5 ("vdpa/mlx5: Fix dependency on MLX5_CORE")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Eli Cohen <eli@mellanox.com>
Cc: Parav Pandit <parav@mellanox.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: virtualization@lists.linux-foundation.org
Cc: Saeed Mahameed <saeedm@nvidia.com>
Cc: Leon Romanovsky <leonro@nvidia.com>
Cc: netdev@vger.kernel.org
Link: https://lore.kernel.org/r/20201128213905.27409-1-rdunlap@infradead.org
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
2020-12-02 04:09:56 -05:00
Laurent Vivier
a312db697c vdpasim: fix "mac_pton" undefined error
ERROR: modpost: "mac_pton" [drivers/vdpa/vdpa_sim/vdpa_sim.ko] undefined!

mac_pton() is defined in lib/net_utils.c and is not built if NET is not set.

Select GENERIC_NET_UTILS as vdpasim doesn't depend on NET.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Link: https://lore.kernel.org/r/20201113155706.599434-1-lvivier@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
2020-11-15 17:30:54 -05:00
Eli Cohen
aff90770e5 vdpa/mlx5: Fix dependency on MLX5_CORE
Remove propmt for selecting MLX5_VDPA by the user and modify
MLX5_VDPA_NET to select MLX5_VDPA. Also modify MLX5_VDPA_NET to depend
on mlx5_core.

This fixes an issue where configuration sets 'y' for MLX5_VDPA_NET while
MLX5_CORE is compiled as a module causing link errors.

Reported-by: kernel test robot <lkp@intel.com>
Fixes: 1a86b377aa ("vdpa/mlx5: Add VDPA driver for supported mlx5 device")s
Signed-off-by: Eli Cohen <elic@nvidia.com>
Link: https://lore.kernel.org/r/20201007064011.GA50074@mtl-vdi-166.wap.labs.mlnx
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-10-08 16:02:00 -04:00
Linus Torvalds
57b0779392 virtio: fixes, features
IRQ bypass support for vdpa and IFC
 MLX5 vdpa driver
 Endian-ness fixes for virtio drivers
 Misc other fixes
 
 Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
 -----BEGIN PGP SIGNATURE-----
 
 iQFDBAABCAAtFiEEXQn9CHHI+FuUyooNKB8NuNKNVGkFAl8yVEwPHG1zdEByZWRo
 YXQuY29tAAoJECgfDbjSjVRpNPEH/0Dtq1s1V4r/kxtLUoMophv9wuORpWCr98BQ
 2aOveTmwTOVdZVOiw2tzTgO9nbWx+cL2HvkU7Aajfpz5hh93Z2VOo2n4a7hBC79f
 rlc3GXiG+pMk5RfmqGofIHTU+D6ony4D5SXlUDurLdtEwunyuqZwABiWkZjdclZJ
 bv90IL8Upzbz0rxYr7k3z8UepdOCt7r4QS/o7STHZBjJRyylxmO/R2yTnh6PtpRK
 Q/z35wJBJ3SKc8X3Fi0VOOSeGNZOiypkkl9ZnLVY5lExNAU1+2MMn2UK119SlCDV
 MSxb7quYFF4cksXH1g77GMBNi1uADRh1dtFMZdkKhZGljGxKLxo=
 =6VTZ
 -----END PGP SIGNATURE-----

Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost

Pull virtio updates from Michael Tsirkin:

 - IRQ bypass support for vdpa and IFC

 - MLX5 vdpa driver

 - Endianness fixes for virtio drivers

 - Misc other fixes

* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost: (71 commits)
  vdpa/mlx5: fix up endian-ness for mtu
  vdpa: Fix pointer math bug in vdpasim_get_config()
  vdpa/mlx5: Fix pointer math in mlx5_vdpa_get_config()
  vdpa/mlx5: fix memory allocation failure checks
  vdpa/mlx5: Fix uninitialised variable in core/mr.c
  vdpa_sim: init iommu lock
  virtio_config: fix up warnings on parisc
  vdpa/mlx5: Add VDPA driver for supported mlx5 devices
  vdpa/mlx5: Add shared memory registration code
  vdpa/mlx5: Add support library for mlx5 VDPA implementation
  vdpa/mlx5: Add hardware descriptive header file
  vdpa: Modify get_vq_state() to return error code
  net/vdpa: Use struct for set/get vq state
  vdpa: remove hard coded virtq num
  vdpasim: support batch updating
  vhost-vdpa: support IOTLB batching hints
  vhost-vdpa: support get/set backend features
  vhost: generialize backend features setting/getting
  vhost-vdpa: refine ioctl pre-processing
  vDPA: dont change vq irq after DRIVER_OK
  ...
2020-08-11 14:34:17 -07:00
Eli Cohen
1a86b377aa vdpa/mlx5: Add VDPA driver for supported mlx5 devices
Add a front end VDPA driver that registers in the VDPA bus and provides
networking to a guest. The VDPA driver creates the necessary resources
on the VF it is driving such that data path will be offloaded.

Notifications are being communicated through the driver.

Currently, only VFs are supported. In subsequent patches we will have
devlink support to control which VF is used for VDPA and which function
is used for regular networking.

Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Eli Cohen <eli@mellanox.com>
Link: https://lore.kernel.org/r/20200804162048.22587-13-eli@mellanox.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
2020-08-05 19:00:24 -04:00
Eli Cohen
29064bfdab vdpa/mlx5: Add support library for mlx5 VDPA implementation
Following patches introduce VDPA network driver for Mellanox Connectx6
devices. This patch provides functionality that will be used by those
patches.

Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Eli Cohen <eli@mellanox.com>
Link: https://lore.kernel.org/r/20200804162048.22587-11-eli@mellanox.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-08-05 19:00:23 -04:00
Christoph Hellwig
2f9237d4f6 dma-mapping: make support for dma ops optional
Avoid the overhead of the dma ops support for tiny builds that only
use the direct mapping.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
2020-07-19 09:29:23 +02:00
Michael S. Tsirkin
213e772138 vhost: revert "vhost: disable for OABI"
This reverts commit d085eb8ce7 ("vhost: disable for OABI")
With commit "virtio: force spec specified alignment on types"
in place, we force proper alignment for all structures,
so there's no longer a reason to blacklist OABI.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-06-02 02:45:13 -04:00
Michael S. Tsirkin
d085eb8ce7 vhost: disable for OABI
vhost is currently broken on the some ARM configs.

The reason is that the ring element addresses are passed between
components with different alignments assumptions. Thus, if
guest selects a pointer and host then gets and dereferences
it, then alignment assumed by the host's compiler might be
greater than the actual alignment of the pointer.
compiler on the host from assuming pointer is aligned.

This actually triggers on ARM with -mabi=apcs-gnu - which is a
deprecated configuration. With this OABI, compiler assumes that
all structures are 4 byte aligned - which is stronger than
virtio guarantees for available and used rings, which are
merely 2 bytes. Thus a guest without -mabi=apcs-gnu running
on top of host with -mabi=apcs-gnu will be broken.

The correct fix is to force alignment of structures - however
that is an intrusive fix that's best deferred until the next release.

We didn't previously support such ancient systems at all - this surfaced
after vdpa support prompted removing dependency of vhost on
VIRTULIZATION. So for now, let's just add something along the lines of

	depends on !ARM || AEABI

to the virtio Kconfig declaration, and add a comment that it has to do
with struct member alignment.

Note: we can't make VHOST and VHOST_RING themselves have
a dependency since these are selected. Add a new symbol for that.

We should be able to drop this dependency down the road.

Fixes: 20c384f1ea ("vhost: refine vhost and vringh kconfig")
Suggested-by: Ard Biesheuvel <ardb@kernel.org>
Suggested-by: Richard Earnshaw <Richard.Earnshaw@arm.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-04-20 10:19:22 -04:00
Michael S. Tsirkin
58ad13729a vdpa: make vhost, virtio depend on menu
If user did not configure any vdpa drivers, neither vhost
nor virtio vdpa are going to be useful. So there's no point
in prompting for these and selecting vdpa core automatically.
Simplify configuration by making virtio and vhost vdpa
drivers depend on vdpa menu entry. Once done, we no longer
need a separate menu entry, so also get rid of this.
While at it, fix up the IFC entry: VDPA->vDPA for consistency
with other places.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
2020-04-17 06:05:30 -04:00
Michael S. Tsirkin
355b36e6bf vdpa-sim: depend on HAS_DMA
set_dma_ops isn't available on all architectures:

        make ARCH=um
...

   drivers/vdpa/vdpa_sim/vdpa_sim.c: In function 'vdpasim_create':
>> drivers/vdpa/vdpa_sim/vdpa_sim.c:324:2: error: implicit declaration of function 'set_dma_ops'; did you mean 'set_groups'?
+[-Werror=implicit-function-declaration]
     set_dma_ops(dev, &vdpasim_dma_ops);
     ^~~~~~~~~~~
     set_groups

Disable vdpa-sim on architectures where it isn't.

Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-04-16 18:31:01 -04:00
Michael S. Tsirkin
c9b9f5f8c0 vdpa: move to drivers/vdpa
We have both vhost and virtio drivers that depend on vdpa.
It's easier to locate it at a top level directory otherwise
we run into issues e.g. if vhost is built-in but virtio
is modular.  Let's just move it up a level.

Reported-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-04-02 10:41:40 -04:00