Hold a window of each output's sends in its transport, not the outbox

A link's ring and a wire's queue took every saved send of their output
and kept it until delivered: with the receiver down they held the
outbox a second time, 178 B and 313 B more a send (104). Each outlet now
holds a window, --window bytes (4 MiB) and 65,536 sends past what its
receiver has, at least one; the outbox keeps the rest.

The owner remembers each outlet's place in the outbox and looks on from
there after a save and after its transports deliver. Places move back by
the sends that left the outbox since (PureApp.sends_gone), so a trim
before a place skips nothing, and sends queued before are not looked at
again however many wait. pure_app_sent reads a send by its place (the
newest-first visit it replaces is gone); pure_link_held and
pure_wire_held say what a transport holds.

With the receiver down a node now grows by the outbox alone once the
window is full: 219 and 218 B a send over the link and HTTP (399 and
534 B before). check.py server_windows keeps the receivers of a sender's
two outputs, one linked and one over HTTP, down through 300 buys with a
2,000-byte window, then starts them: each applies all 300 once. Without
the refill after deliveries the link stops at 117.

1 files changed, 2 insertions, 0 deletions

diff --git a/link/link.h b/link/link.h
index 3390a82..64846ce 100644
--- a/link/link.h
+++ b/link/link.h
@@ -22,6 +22,8 @@ bool pure_link_queue(PureLink *, uint64_t seq, const unsigned char *bytes, size_
 uint64_t pure_link_queued(PureLink *);
 /* The last number the receiver saved; any thread. */
 uint64_t pure_link_delivered(PureLink *);
+/* The bytes of the sends queued and not yet saved; any thread. */
+size_t pure_link_held(PureLink *);
 /* Packets go in bursts with UDP GSO where the system has it (at first);
  * off, each in a call of its own: for a path whose queues take a burst as
  * one packet (netem does). Before pure_link_start. */