On Tue, Feb 18, 2020 at 09:53:19AM +0000, Stefan Hajnoczi wrote: > On Mon, Feb 17, 2020 at 06:31:21PM +0100, Stefano Garzarella wrote: > > Linux 5.6 added the new well-known VMADDR_CID_LOCAL for > > local communication. > > > > This patch explains how to use it and remove the legacy > > VMADDR_CID_RESERVED no longer available. > > > > Reviewed-by: Jorgen Hansen <jhansen@xxxxxxxxxx> > > Signed-off-by: Stefano Garzarella <sgarzare@xxxxxxxxxx> > > --- > > v3: > > * rephrased "Previous versions" part [Jorgen] > > v2: > > * rephrased "Local communication" description [Stefan] > > * added a mention of previous versions that supported > > loopback only in the guest [Stefan] > > --- > > man7/vsock.7 | 19 +++++++++++++++++-- > > 1 file changed, 17 insertions(+), 2 deletions(-) > > > > diff --git a/man7/vsock.7 b/man7/vsock.7 > > index c5ffcf07d..219e3505f 100644 > > --- a/man7/vsock.7 > > +++ b/man7/vsock.7 > > @@ -127,8 +127,8 @@ There are several special addresses: > > means any address for binding; > > .B VMADDR_CID_HYPERVISOR > > (0) is reserved for services built into the hypervisor; > > -.B VMADDR_CID_RESERVED > > -(1) must not be used; > > +.B VMADDR_CID_LOCAL > > +(1) is the well-known address for local communication (loopback); > > .B VMADDR_CID_HOST > > (2) > > is the well-known address of the host. > > @@ -164,6 +164,16 @@ Consider using > > .B VMADDR_CID_ANY > > when binding instead of getting the local CID with > > .BR IOCTL_VM_SOCKETS_GET_LOCAL_CID . > > +.SS Local communication > > +The > > +.B VMADDR_CID_LOCAL > > +(1) directs packets to the same host that generated them. This is useful > > Please see my comment on v2. "The VMADDR_CID_LOCAL (1) directs packets > ..." sounds unnatural. Please drop "The". I received it just now :O but you sent it 4 days ago! Sorry for that! I'll send v4 dropping "The". Thanks, Stefano
