vmm: detect dead connections during live migration

During a live migration, connections may get interrupted silently,
without being reset or closed. Interrupted idle connections may stay
alive indefinitely, for example if a connection dies during prefaulting
on the receiver side.

To detect dead idle connections, we enable
`SO_KEEPALIVE`. With `SO_KEEPALIVE`, the kernel will send keepalive
probes if a connection is idle and close the connection if the probes
remain unacknowledged.

To detect dead connections when actively sending data in a timely
manner, we enable `TCP_USER_TIMEOUT`, to reduce the timeout for closing
a connection where the peer doesn't acknowledge sent data.

On-behalf-of: SAP julian.schindel@sap.com
Signed-off-by: Julian Schindel <julian.schindel@cyberus-technology.de>

Cargo.lock

@@ -2310,6 +2310,16 @@ version = "1.15.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8ed6a63f02c8539c91a8685a86f4099661ba3da017932f6ebbea6de3f0fa7c90"
 
+[[package]]
+name = "socket2"
+version = "0.6.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c3d1e2c7f27f8d4cb10542a02c49005dbd6e93095799d6f3be745fae9f8fedd4"
+dependencies = [
+ "libc",
+ "windows-sys 0.61.2",
+]
+
 [[package]]
 name = "spin"
 version = "0.9.9"
@@ -2852,6 +2862,7 @@ dependencies = [
  "serial_buffer",
  "sha2",
  "signal-hook",
+ "socket2",
  "tempfile",
  "thiserror",
  "tracer",