[Bug 53464] New: conhost.exe:tty sometimes times out on Windows 10

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=53464

            Bug ID: 53464
           Summary: conhost.exe:tty sometimes times out on Windows 10
           Product: Wine
           Version: unspecified
          Hardware: x86-64
                OS: Windows
            Status: NEW
          Severity: normal
          Priority: P2
         Component: programs
          Assignee: wine-bugs at winehq.org
          Reporter: fgouget at codeweavers.com

conhost.exe:tty sometimes times out on Windows 10:

conhost.exe:tty start programs/conhost/tests/tty.c
conhost.exe:tty:1734 done (258) in 120s

https://test.winehq.org/data/patterns.html#conhost.exe:tty

The timeouts are very frequent on Windows 10 1909 as per bug 51182.

However they also happen on other Windows 10 versions like 2004 and 21H1 where
they cannot be blamed on conhost.exe being a buggy early version.

Adding more traces shows that the test gets stuck in
test_read_console_control(). Depending on the run this happens in one of two
locations:

1350: skip_sequence("\x1b[25l");  /* broken hide cursor */
tty.c:1370:4.047 Test succeeded
tty.c:1373:4.047 Test succeeded
tty.c:226:4.047 Test succeeded
tty.c:1705:3.984 Test succeeded
tty.c:1350:4.110 Test succeeded
tty.c:603:4.tty.c:1602:4.047 Test 110 Test succeeded
succeeded
ttytty.c:60.c:6:4.110 T16est succeeded
04:4tty.c:.043477:4.110  Test succeeded
Test succeeded
tty.c:1245:4.110 Test succeeded
conhost.exe:tty:1dbc done (258) in 120s

1406: skip_sequence("\x1b[?25l"); /* hide cursor */
tty.c:1647:6.578 Test succeeded
tty.c:1650:tty.c6.578:140 Test5:6.6 succeeded87 Tes
t succeeded
tty.c:1405:6.687 Test succeeded
tty.c:1405:6.687 Test succeeded
tty.c:1405:6.687 Test succeeded
tty.c:1406:6.687 Test succeeded
tty.c:437:6.750 Test succeeded
tty.c:1245:6.750 Test succeeded
conhost.exe:tty:13d4 done (258) in 120s

The traces are mangled because the main test process and the child process
write to stdout at the same time.

In all cases the last trace is from write_console_pipe():
1245: ok(res, "WriteFile failed: %lu\n", GetLastError());

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.