Sixteen 9P connections, and the next one is told there is no room instead of being reset

Four slots ran out under scripts plus a mount, and the fifth client was closed without a word, which through a mount looks exactly like a dead session. pardes now serves sixteen, and builds against cloud9 09b77cf, whose runner answers the Tversion of a client it cannot seat with an Rerror "too many connections" and tells pardes, which logs `err - 9p: too many connections`. QUIC keeps its own sixteen and still just closes.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

1 files changed, 16 insertions, 0 deletions

diff --git a/test/fs.py b/test/fs.py
index 72c52b7e..62e78655 100644
--- a/test/fs.py
+++ b/test/fs.py
@@ -371,6 +371,22 @@ def discovery(binary, embedded=False):
             assert f'err {unsaved} ctl: Save /nonexistent-pardes-dir/x.txt: '.encode() in client.read('/log'), client.read('/log')
             client.remove(f'/pane/{unsaved}')
 
+            # Sixteen connections fill the slots; the next is told why.
+            extra = [Client(address) for _ in range(15)]
+            try:
+                try:
+                    Client(address).__exit__()
+                    raise AssertionError('a seventeenth connection was served')
+                except OSError as refused:
+                    assert 'too many connections' in str(refused), refused
+            finally:
+                for other in extra:
+                    other.__exit__()
+            deadline = time.monotonic() + 5
+            while b'err - 9p: too many connections' not in client.read('/log'):
+                assert time.monotonic() < deadline, client.read('/log')
+                time.sleep(.01)
+
     print('9P discovery: listing/stat/find are inert; new, remove, look, exec, name, sel, log, ctl lock, focus, the ctl split and commands behave')