linux.git/drivers/vhost/vsock.c, branch v5.17

vsock: each transport cycles only on its own sockets

2022-03-12T07:14:19+00:00

When iterating over sockets using vsock_for_each_connected_socket, make
sure that a transport filters out sockets that don't belong to the
transport.

There actually was an issue caused by this; in a nested VM
configuration, destroying the nested VM (which often involves the
closing of /dev/vhost-vsock if there was h2g connections to the nested
VM) kills not only the h2g connections, but also all existing g2h
connections to the (outmost) host which are totally unrelated.

Tested: Executed the following steps on Cuttlefish (Android running on a
VM) [1]: (1) Enter into an `adb shell` session - to have a g2h
connection inside the VM, (2) open and then close /dev/vhost-vsock by
`exec 3< /dev/vhost-vsock && exec 3<&-`, (3) observe that the adb
session is not reset.

[1] https://android.googlesource.com/device/google/cuttlefish/

Fixes: c0cfa2d8a788 ("vsock: add multi-transports support")
Reviewed-by: Stefano Garzarella 
Acked-by: Michael S. Tsirkin 
Signed-off-by: Jiyong Park 
Link: https://lore.kernel.org/r/20220311020017.1509316-1-jiyong@google.com
Signed-off-by: Jakub Kicinski

vhost/vsock: don't check owner in vhost_vsock_stop() while releasing

2022-02-23T12:32:33+00:00

vhost_vsock_stop() calls vhost_dev_check_owner() to check the device
ownership. It expects current->mm to be valid.

vhost_vsock_stop() is also called by vhost_vsock_dev_release() when
the user has not done close(), so when we are in do_exit(). In this
case current->mm is invalid and we're releasing the device, so we
should clean it anyway.

Let's check the owner only when vhost_vsock_stop() is called
by an ioctl.

When invoked from release we can not fail so we don't check return
code of vhost_vsock_stop(). We need to stop vsock even if it's not
the owner.

Fixes: 433fc58e6bf2 ("VSOCK: Introduce vhost_vsock.ko")
Cc: stable@vger.kernel.org
Reported-by: syzbot+1e3ea63db39f2b4440e0@syzkaller.appspotmail.com
Reported-and-tested-by: syzbot+3140b17cb44a7b174008@syzkaller.appspotmail.com
Signed-off-by: Stefano Garzarella 
Acked-by: Jason Wang 
Signed-off-by: David S. Miller

vhost/vsock: cleanup removing `len` variable

2021-11-25T00:00:28+00:00

We can increment `total_len` directly and remove `len` since it
is no longer used for vhost_add_used().

Signed-off-by: Stefano Garzarella 
Link: https://lore.kernel.org/r/20211122163525.294024-3-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin 
Reviewed-by: Stefan Hajnoczi

vhost/vsock: fix incorrect used length reported to the guest

2021-11-25T00:00:28+00:00

The "used length" reported by calling vhost_add_used() must be the
number of bytes written by the device (using "in" buffers).

In vhost_vsock_handle_tx_kick() the device only reads the guest
buffers (they are all "out" buffers), without writing anything,
so we must pass 0 as "used length" to comply virtio spec.

Fixes: 433fc58e6bf2 ("VSOCK: Introduce vhost_vsock.ko")
Cc: stable@vger.kernel.org
Reported-by: Halil Pasic 
Suggested-by: Jason Wang 
Signed-off-by: Stefano Garzarella 
Link: https://lore.kernel.org/r/20211122163525.294024-2-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin 
Reviewed-by: Stefan Hajnoczi 
Reviewed-by: Halil Pasic

vhost/vsock: support MSG_EOR bit processing

2021-09-05T20:23:09+00:00

'MSG_EOR' handling has similar logic as 'MSG_EOM' - if bit present
in packet's header, reset it to 0. Then restore it back if packet
processing wasn't completed. Instead of bool variable for each
flag, bit mask variable was added: it has logical OR of 'MSG_EOR'
and 'MSG_EOM' if needed, to restore flags, this variable is ORed
with flags field of packet.

Signed-off-by: Arseny Krasnov 
Link: https://lore.kernel.org/r/20210903123238.3273526-1-arseny.krasnov@kaspersky.com
Signed-off-by: Michael S. Tsirkin 
Reviewed-by: Stefano Garzarella

virtio/vsock: rename 'EOR' to 'EOM' bit.

2021-09-05T20:23:08+00:00

This current implemented bit is used to mark end of messages
('EOM' - end of message), not records('EOR' - end of record).
Also rename 'record' to 'message' in implementation as it is
different things.

Signed-off-by: Arseny Krasnov 
Reviewed-by: Stefano Garzarella 
Link: https://lore.kernel.org/r/20210903123109.3273053-1-arseny.krasnov@kaspersky.com
Signed-off-by: Michael S. Tsirkin

vhost: remove work arg from vhost_work_flush

2021-07-03T08:50:54+00:00

vhost_work_flush doesn't do anything with the work arg. This patch drops
it and then renames vhost_work_flush to vhost_work_dev_flush to reflect
that the function flushes all the works in the dev and not just a
specific queue or work item.

Signed-off-by: Mike Christie 
Acked-by: Jason Wang 
Reviewed-by: Chaitanya Kulkarni 
Link: https://lore.kernel.org/r/20210525174733.6212-2-michael.christie@oracle.com
Reviewed-by: Stefano Garzarella 
Signed-off-by: Michael S. Tsirkin

net: sock: introduce sk_error_report

2021-06-29T18:28:21+00:00

This patch introduces a function wrapper to call the sk_error_report
callback. That will prepare to add additional handling whenever
sk_error_report is called, for example to trace socket errors.

Signed-off-by: Alexander Aring 
Signed-off-by: David S. Miller

vhost/vsock: support SEQPACKET for transport

2021-06-11T20:32:47+00:00

When received packet is copied to guests's rx queue, data buffers
of rx queue could be smaller that data buffer of input packet, so
data of input packet is copied to each rx buffer, thus each rx
buffer will be a packet with dynamically created header. Fields
of such header are initialized from header of input packet(except
length field which value is depends on number of bytes copied to
rx buffer). But in SEQPACKET case, we also need to take care of
record delimeter bit: if input packet has this bit set, we don't
copy it to header of packet in rx buffer, except case when such
rx buffer is last part of input packet. Otherwise, we will get
sequence of packets with delimeter bit set, thus braking record
bounds.
Also remove ignore of non-stream type of packets, handle SEQPACKET
feature bit.

Signed-off-by: Arseny Krasnov 
Signed-off-by: David S. Miller

vhost/vsock: add IOTLB API support

2020-12-27T10:49:01+00:00

This patch enables the IOTLB API support for vhost-vsock devices,
allowing the userspace to emulate an IOMMU for the guest.

These changes were made following vhost-net, in details this patch:
- exposes VIRTIO_F_ACCESS_PLATFORM feature and inits the iotlb
  device if the feature is acked
- implements VHOST_GET_BACKEND_FEATURES and
  VHOST_SET_BACKEND_FEATURES ioctls
- calls vq_meta_prefetch() before vq processing to prefetch vq
  metadata address in IOTLB
- provides .read_iter, .write_iter, and .poll callbacks for the
  chardev; they are used by the userspace to exchange IOTLB messages

This patch was tested specifying "intel_iommu=strict" in the guest
kernel command line. I used QEMU with a patch applied [1] to fix a
simple issue (that patch was merged in QEMU v5.2.0):
    $ qemu -M q35,accel=kvm,kernel-irqchip=split \
           -drive file=fedora.qcow2,format=qcow2,if=virtio \
           -device intel-iommu,intremap=on,device-iotlb=on \
           -device vhost-vsock-pci,guest-cid=3,iommu_platform=on,ats=on

[1] https://lists.gnu.org/archive/html/qemu-devel/2020-10/msg09077.html

Reviewed-by: Stefan Hajnoczi 
Signed-off-by: Stefano Garzarella 
Link: https://lore.kernel.org/r/20201223143638.123417-1-sgarzare@redhat.com
Signed-off-by: Michael S. Tsirkin 
Acked-by: Jason Wang