After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 747676 - gio/tests/socket fails: test_fd_roundtrip
gio/tests/socket fails: test_fd_roundtrip
Status: RESOLVED FIXED
Product: glib
Classification: Platform
Component: gio
2.44.x
Other Mac OS
: Normal normal
: ---
Assigned To: gtkdev
gtkdev
Depends on:
Blocks:
 
 
Reported: 2015-04-11 08:15 UTC by Daniel Macks
Modified: 2015-06-30 01:44 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Daniel Macks 2015-04-11 08:15:13 UTC
On OS X 10.8, glib-2.44.0...

PASS: socket 1 /socket/ipv4_sync
PASS: socket 2 /socket/ipv4_async
PASS: socket 3 /socket/ipv6_sync
PASS: socket 4 /socket/ipv6_async
PASS: socket 5 /socket/ipv6_v4mapped
PASS: socket 6 /socket/close_graceful
PASS: socket 7 /socket/timed_wait
ERROR: socket - missing test plan
ERROR: socket - exited with status 134 (terminated by signal 6?)

gio/tests/socket.log reports:

GLib-GIO:ERROR:socket.c:1007:test_fd_roundtrip: assertion failed (error == NULL): Unable to shutdown socket: Socket is not connected (g-io-error-quark, 45)

Looks like a new failure vs 2.42.2
Comment 1 Dan Winship 2015-06-21 16:10:01 UTC
test was added in bug 741707

The test does seem slightly dubious; we can dup the socket fd, but there's still only one actual TCP connection, so it's plausible that calling shutdown() on either fd would also effectively shutdown the other.

Does the test pass if you remove the failing assertion? (And pass NULL instead of &error in the g_socket_shutdown(client2) call)
Comment 2 Daniel Macks 2015-06-29 17:34:19 UTC
Yup...

/socket/fd_roundtrip: OK