mbox series

[RFC,0/3] vsock: add support for sockmap

Message ID 20230118-support-vsock-sockmap-connectible-v1-0-d47e6294827b@bytedance.com (mailing list archive)
Headers show
Series vsock: add support for sockmap | expand

Message

Bobby Eshleman Jan. 18, 2023, 8:27 p.m. UTC 
Add support for sockmap to vsock.

We're testing usage of vsock as a way to redirect guest-local UDS requests to
the host and this patch series greatly improves the performance of such a
setup.

Compared to copying packets via userspace, this improves throughput by 221% in
basic testing.

Tested as follows.

Setup: guest unix dgram sender -> guest vsock redirector -> host vsock server
Threads: 1
Payload: 64k
No sockmap:
- 76.3 MB/s
- The guest vsock redirector was
  "socat VSOCK-CONNECT:2:1234 UNIX-RECV:/path/to/sock"
Using sockmap (this patch):
- 168.8 MB/s (+221%)
- The guest redirector was a simple sockmap echo server,
  redirecting unix ingress to vsock 2:1234 egress.
- Same sender and server programs

Only the virtio transport has been tested. The loopback transport was used in
writing bpf/selftests, but not thoroughly tested otherwise.

This series requires the skb patch.

To: Stefan Hajnoczi <stefanha@redhat.com>
To: Stefano Garzarella <sgarzare@redhat.com>
To: "Michael S. Tsirkin" <mst@redhat.com>
To: Jason Wang <jasowang@redhat.com>
To: "David S. Miller" <davem@davemloft.net>
To: Eric Dumazet <edumazet@google.com>
To: Jakub Kicinski <kuba@kernel.org>
To: Paolo Abeni <pabeni@redhat.com>
To: Andrii Nakryiko <andrii@kernel.org>
To: Mykola Lysenko <mykolal@fb.com>
To: Alexei Starovoitov <ast@kernel.org>
To: Daniel Borkmann <daniel@iogearbox.net>
To: Martin KaFai Lau <martin.lau@linux.dev>
To: Song Liu <song@kernel.org>
To: Yonghong Song <yhs@fb.com>
To: John Fastabend <john.fastabend@gmail.com>
To: KP Singh <kpsingh@kernel.org>
To: Stanislav Fomichev <sdf@google.com>
To: Hao Luo <haoluo@google.com>
To: Jiri Olsa <jolsa@kernel.org>
To: Shuah Khan <shuah@kernel.org>
Cc: linux-kernel@vger.kernel.org
Cc: kvm@vger.kernel.org
Cc: virtualization@lists.linux-foundation.org
Cc: netdev@vger.kernel.org
Cc: bpf@vger.kernel.org
Cc: linux-kselftest@vger.kernel.org
Signed-off-by: Bobby Eshleman <bobby.eshleman@bytedance.com>

---
Bobby Eshleman (3):
      vsock: support sockmap
      selftests/bpf: add vsock to vmtest.sh
      selftests/bpf: Add a test case for vsock sockmap

 drivers/vhost/vsock.c                              |   1 +
 include/linux/virtio_vsock.h                       |   1 +
 include/net/af_vsock.h                             |  17 ++
 net/vmw_vsock/Makefile                             |   1 +
 net/vmw_vsock/af_vsock.c                           |  59 ++++++-
 net/vmw_vsock/virtio_transport.c                   |   2 +
 net/vmw_vsock/virtio_transport_common.c            |  22 +++
 net/vmw_vsock/vsock_bpf.c                          | 180 +++++++++++++++++++++
 net/vmw_vsock/vsock_loopback.c                     |   2 +
 tools/testing/selftests/bpf/config.x86_64          |   4 +
 .../selftests/bpf/prog_tests/sockmap_listen.c      | 163 +++++++++++++++++++
 tools/testing/selftests/bpf/vmtest.sh              |   1 +
 12 files changed, 447 insertions(+), 6 deletions(-)
---
base-commit: f12f4326c6a75a74e908714be6d2f0e2f0fd0d76
change-id: 20230118-support-vsock-sockmap-connectible-2e1297d2111a

Best regards,

Comments

Bobby Eshleman Jan. 18, 2023, 3:10 p.m. UTC | #1 
On Thu, Jan 19, 2023 at 11:49:02AM +0100, Stefano Garzarella wrote:
> Hi Bobby,
> 
> On Wed, Jan 18, 2023 at 12:27:39PM -0800, Bobby Eshleman wrote:
> > Add support for sockmap to vsock.
> > 
> > We're testing usage of vsock as a way to redirect guest-local UDS requests to
> > the host and this patch series greatly improves the performance of such a
> > setup.
> > 
> > Compared to copying packets via userspace, this improves throughput by 221% in
> > basic testing.
> 
> Cool, nice series!
> 
> > 
> > Tested as follows.
> > 
> > Setup: guest unix dgram sender -> guest vsock redirector -> host vsock server
> > Threads: 1
> > Payload: 64k
> > No sockmap:
> > - 76.3 MB/s
> > - The guest vsock redirector was
> >  "socat VSOCK-CONNECT:2:1234 UNIX-RECV:/path/to/sock"
> > Using sockmap (this patch):
> > - 168.8 MB/s (+221%)
> 
> Assuming the absolute value is correct, there is a typo here, it would be
> +121% right?
> 

Lol, yes.

> > - The guest redirector was a simple sockmap echo server,
> >  redirecting unix ingress to vsock 2:1234 egress.
> > - Same sender and server programs
> > 
> > Only the virtio transport has been tested.
> 
> I think is fine for now.
> 
> > The loopback transport was used in
> > writing bpf/selftests, but not thoroughly tested otherwise.
> 
> I did a quick review mainly for vsock stuff.
> Hoping others can take a better look at net/vmw_vsock/vsock_bpf.c, since I'm
> not very familiar with that subsystem.
> 
> FYI I will be off the next two weeks (till Feb 7) with limited internet
> access.
> 

Roger that.

Thanks,
Bobby

> Thanks,
> Stefano
> 
> > 
> > This series requires the skb patch.
> > 
> > To: Stefan Hajnoczi <stefanha@redhat.com>
> > To: Stefano Garzarella <sgarzare@redhat.com>
> > To: "Michael S. Tsirkin" <mst@redhat.com>
> > To: Jason Wang <jasowang@redhat.com>
> > To: "David S. Miller" <davem@davemloft.net>
> > To: Eric Dumazet <edumazet@google.com>
> > To: Jakub Kicinski <kuba@kernel.org>
> > To: Paolo Abeni <pabeni@redhat.com>
> > To: Andrii Nakryiko <andrii@kernel.org>
> > To: Mykola Lysenko <mykolal@fb.com>
> > To: Alexei Starovoitov <ast@kernel.org>
> > To: Daniel Borkmann <daniel@iogearbox.net>
> > To: Martin KaFai Lau <martin.lau@linux.dev>
> > To: Song Liu <song@kernel.org>
> > To: Yonghong Song <yhs@fb.com>
> > To: John Fastabend <john.fastabend@gmail.com>
> > To: KP Singh <kpsingh@kernel.org>
> > To: Stanislav Fomichev <sdf@google.com>
> > To: Hao Luo <haoluo@google.com>
> > To: Jiri Olsa <jolsa@kernel.org>
> > To: Shuah Khan <shuah@kernel.org>
> > Cc: linux-kernel@vger.kernel.org
> > Cc: kvm@vger.kernel.org
> > Cc: virtualization@lists.linux-foundation.org
> > Cc: netdev@vger.kernel.org
> > Cc: bpf@vger.kernel.org
> > Cc: linux-kselftest@vger.kernel.org
> > Signed-off-by: Bobby Eshleman <bobby.eshleman@bytedance.com>
> > 
> > ---
> > Bobby Eshleman (3):
> >      vsock: support sockmap
> >      selftests/bpf: add vsock to vmtest.sh
> >      selftests/bpf: Add a test case for vsock sockmap
> > 
> > drivers/vhost/vsock.c                              |   1 +
> > include/linux/virtio_vsock.h                       |   1 +
> > include/net/af_vsock.h                             |  17 ++
> > net/vmw_vsock/Makefile                             |   1 +
> > net/vmw_vsock/af_vsock.c                           |  59 ++++++-
> > net/vmw_vsock/virtio_transport.c                   |   2 +
> > net/vmw_vsock/virtio_transport_common.c            |  22 +++
> > net/vmw_vsock/vsock_bpf.c                          | 180 +++++++++++++++++++++
> > net/vmw_vsock/vsock_loopback.c                     |   2 +
> > tools/testing/selftests/bpf/config.x86_64          |   4 +
> > .../selftests/bpf/prog_tests/sockmap_listen.c      | 163 +++++++++++++++++++
> > tools/testing/selftests/bpf/vmtest.sh              |   1 +
> > 12 files changed, 447 insertions(+), 6 deletions(-)
> > ---
> > base-commit: f12f4326c6a75a74e908714be6d2f0e2f0fd0d76
> > change-id: 20230118-support-vsock-sockmap-connectible-2e1297d2111a
> > 
> > Best regards,
> > -- 
> > Bobby Eshleman <bobby.eshleman@bytedance.com>
> > 
> 
> _______________________________________________
> Virtualization mailing list
> Virtualization@lists.linux-foundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/virtualization
Stefano Garzarella Jan. 19, 2023, 10:49 a.m. UTC | #2 
Hi Bobby,

On Wed, Jan 18, 2023 at 12:27:39PM -0800, Bobby Eshleman wrote:
>Add support for sockmap to vsock.
>
>We're testing usage of vsock as a way to redirect guest-local UDS requests to
>the host and this patch series greatly improves the performance of such a
>setup.
>
>Compared to copying packets via userspace, this improves throughput by 221% in
>basic testing.

Cool, nice series!

>
>Tested as follows.
>
>Setup: guest unix dgram sender -> guest vsock redirector -> host vsock server
>Threads: 1
>Payload: 64k
>No sockmap:
>- 76.3 MB/s
>- The guest vsock redirector was
>  "socat VSOCK-CONNECT:2:1234 UNIX-RECV:/path/to/sock"
>Using sockmap (this patch):
>- 168.8 MB/s (+221%)

Assuming the absolute value is correct, there is a typo here, it would 
be +121% right?

>- The guest redirector was a simple sockmap echo server,
>  redirecting unix ingress to vsock 2:1234 egress.
>- Same sender and server programs
>
>Only the virtio transport has been tested.

I think is fine for now.

>The loopback transport was used in
>writing bpf/selftests, but not thoroughly tested otherwise.

I did a quick review mainly for vsock stuff.
Hoping others can take a better look at net/vmw_vsock/vsock_bpf.c, since 
I'm not very familiar with that subsystem.

FYI I will be off the next two weeks (till Feb 7) with limited internet 
access.

Thanks,
Stefano

>
>This series requires the skb patch.
>
>To: Stefan Hajnoczi <stefanha@redhat.com>
>To: Stefano Garzarella <sgarzare@redhat.com>
>To: "Michael S. Tsirkin" <mst@redhat.com>
>To: Jason Wang <jasowang@redhat.com>
>To: "David S. Miller" <davem@davemloft.net>
>To: Eric Dumazet <edumazet@google.com>
>To: Jakub Kicinski <kuba@kernel.org>
>To: Paolo Abeni <pabeni@redhat.com>
>To: Andrii Nakryiko <andrii@kernel.org>
>To: Mykola Lysenko <mykolal@fb.com>
>To: Alexei Starovoitov <ast@kernel.org>
>To: Daniel Borkmann <daniel@iogearbox.net>
>To: Martin KaFai Lau <martin.lau@linux.dev>
>To: Song Liu <song@kernel.org>
>To: Yonghong Song <yhs@fb.com>
>To: John Fastabend <john.fastabend@gmail.com>
>To: KP Singh <kpsingh@kernel.org>
>To: Stanislav Fomichev <sdf@google.com>
>To: Hao Luo <haoluo@google.com>
>To: Jiri Olsa <jolsa@kernel.org>
>To: Shuah Khan <shuah@kernel.org>
>Cc: linux-kernel@vger.kernel.org
>Cc: kvm@vger.kernel.org
>Cc: virtualization@lists.linux-foundation.org
>Cc: netdev@vger.kernel.org
>Cc: bpf@vger.kernel.org
>Cc: linux-kselftest@vger.kernel.org
>Signed-off-by: Bobby Eshleman <bobby.eshleman@bytedance.com>
>
>---
>Bobby Eshleman (3):
>      vsock: support sockmap
>      selftests/bpf: add vsock to vmtest.sh
>      selftests/bpf: Add a test case for vsock sockmap
>
> drivers/vhost/vsock.c                              |   1 +
> include/linux/virtio_vsock.h                       |   1 +
> include/net/af_vsock.h                             |  17 ++
> net/vmw_vsock/Makefile                             |   1 +
> net/vmw_vsock/af_vsock.c                           |  59 ++++++-
> net/vmw_vsock/virtio_transport.c                   |   2 +
> net/vmw_vsock/virtio_transport_common.c            |  22 +++
> net/vmw_vsock/vsock_bpf.c                          | 180 +++++++++++++++++++++
> net/vmw_vsock/vsock_loopback.c                     |   2 +
> tools/testing/selftests/bpf/config.x86_64          |   4 +
> .../selftests/bpf/prog_tests/sockmap_listen.c      | 163 +++++++++++++++++++
> tools/testing/selftests/bpf/vmtest.sh              |   1 +
> 12 files changed, 447 insertions(+), 6 deletions(-)
>---
>base-commit: f12f4326c6a75a74e908714be6d2f0e2f0fd0d76
>change-id: 20230118-support-vsock-sockmap-connectible-2e1297d2111a
>
>Best regards,
>-- 
>Bobby Eshleman <bobby.eshleman@bytedance.com>
>