On 1/17/23 12:32, Philippe Mathieu-Daudé wrote:
On 17/1/23 12:03, Laurent Vivier wrote:
On 1/17/23 11:00, Thomas Huth wrote:
On 16/01/2023 09.40, Thomas Huth wrote:
On 16/01/2023 09.29, Laurent Vivier wrote:
ping

On 1/5/23 10:37, Laurent Vivier wrote:
Signed-off-by: Laurent Vivier <lviv...@redhat.com>
Acked-by: Michael S. Tsirkin <m...@redhat.com>

Acked-by: Thomas Huth <th...@redhat.com>

I'll queue it for my next pull request (unless someone else wants to take this 
first)

Sorry, but I have to unqueue it again. I'm still seeing failures
in the Windows Cirrus-CI:

  https://cirrus-ci.com/task/5867407370092544

For example:

218/556 qemu:qtest+qtest-aarch64 / qtest-aarch64/netdev-socket ERROR           0.02s exit status 3
------------------------------------- 8< -------------------------------------
stderr:
socket_check_protocol_support() failed

(C:/Users/ContainerAdministrator/AppData/Local/Temp/cirrus-ci-build/build/tests/qtest/netdev-socket.exe:3300):
 GLib-CRITICAL **: 09:08:00.984: g_utf8_to_utf16: assertion 'str != NULL' failed

(test program exited with status code 3)

No clue where this comes from, though, I don't see a call
to g_utf8_to_utf16() in your code?

OK, there is an error in the log:

socket_check_protocol_support() failed

So tmpdir is NULL and we try an g_rmdir(tmpdir)

"In case of errors, NULL is returned and error will be set."

I don't see the ""Can't create temporary directory" string in
Thomas' failing job.

In such case, are the g_strconcat(tmpdir) calls returning "\0"?



I'm going to use g_error() on socket_check_protocol_support() failure to exit. Without the problem with g_utf8_to_utf16() I would not know the socket_init() function was needed.

Thanks,
Laurent


Reply via email to