From 36926af2665bf9e7cf2b97313e1765a0182ce3eb Mon Sep 17 00:00:00 2001
From: teknium1 <127238744+teknium1@users.noreply.github.com>
Date: Sat, 25 Jul 2026 22:43:53 -0700
Subject: [PATCH] test(sessions): prove the connector reached the lock before
 asserting blocked

Closes the last false-pass window @helix4u flagged on #71779. He explicitly
said not to hold the PR for it; it is two lines, so worth doing rather than
leaving a known-soft assertion in a concurrency test.

connection_opened.wait(1.0) proved the connection had not opened, but not
that the connector thread had actually reached connect_tracked() -- an
unscheduled thread produces the same observation. The connector now sets
connect_attempted immediately before the blocking call, and the test waits
on that first, so "still blocked" means blocked at the lock rather than
not yet started.

15/15 stable at ~1.1s. Removed-lock sabotage still fails.
---
 tests/hermes_cli/test_session_recovery.py | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/tests/hermes_cli/test_session_recovery.py b/tests/hermes_cli/test_session_recovery.py
index cb6f9c05ef31..612f13abf16f 100644
--- a/tests/hermes_cli/test_session_recovery.py
+++ b/tests/hermes_cli/test_session_recovery.py
@@ -315,6 +315,7 @@ def test_snapshot_blocks_connections_opened_during_the_copy(
 
     inside_copy = threading.Event()
     release_copy = threading.Event()
+    connect_attempted = threading.Event()
     connection_opened = threading.Event()
     errors: list[str] = []
     real_copy2 = recovery_module.shutil.copy2
@@ -333,6 +334,10 @@ def test_snapshot_blocks_connections_opened_during_the_copy(
             errors.append(f"copy failed: {exc}")
 
     def do_connect():
+        # Signal immediately before the blocking call so a timed "still
+        # blocked" assertion cannot pass merely because this thread had not
+        # been scheduled yet.
+        connect_attempted.set()
         try:
             conn = connect_tracked(source, isolation_level=None, timeout=30.0)
             connection_opened.set()
@@ -348,7 +353,9 @@ def test_snapshot_blocks_connections_opened_during_the_copy(
         assert inside_copy.wait(30), "copy never reached the patched operation"
 
         connector.start()
-        # While the copy holds the lifecycle lock the connection must not open.
+        assert connect_attempted.wait(30), "connector thread never started"
+        # The connector is at the lock. While the copy holds it, the
+        # connection must not open.
         assert not connection_opened.wait(1.0), (
             "connect_tracked() completed while raw copy descriptors were open "
             "— the guard is not holding the lifecycle lock across the copy"