search for: sockmap

On Tue, Feb 28, 2023 at 07:04:33PM +0000, 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 > 121% in basic test...

...h a setup. > > Compared to copying packets via userspace, this improves throughput by > 121% 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 (+121%) > - The guest redirector was a simple sockmap echo server, > redirecting unix ingress to vsock 2:1234 egre...

...a comment. I agree, since I had the same doubt in previous versions, I think it's worth putting a comment in the code to explain why. Since there may be a v4, I'll leave some small comments in a separate email. Thanks, Stefano > >The real reason is due to this piece of logic in sockmap: >https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/net/core/sock_map.c?h=v6.2#n531 > >And because of it, you see the same thing in (for example) >unix_dgram_connect(): >https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/net/unix/af_unix.c?h...