| Subject: | Killing undef in t/client.t Kills the Parent |
Dear Chris Dent,
Hi. This is imacat from Taiwan. I found that the t/client.t in
your Purple-0.9 may kill the parent if test fails. I tried to
investigate further. It seems that, in t/client.t, if
$server->background() fails, "kill 15, $pid" will become "kill 15,
undef" and kill the parent process accidently, at least on Linux. This
parent process may be the test suite that I'm running, or the make or
even the shell.
I made a simple patch against Purple-0.9, in the hope that it helps.
Please tell me if there is any problem, or if I could be of any help.
Thank you.
| Subject: | Purple-0.9-killundef.diff |
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
diff -u -r Purple-0.9.orig/t/client.t Purple-0.9/t/client.t
- --- Purple-0.9.orig/t/client.t 2006-02-19 11:16:09.000000000 +0800
+++ Purple-0.9/t/client.t 2008-04-14 10:00:33.000000000 +0800
@@ -65,5 +65,5 @@
}
END {
- - kill 15, $pid;
+ kill 15, $pid if defined $pid;
}
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
iEYEARECAAYFAkgCutkACgkQi9gubzC5S1xaywCePnikx3w29EXpDcHMqKMvATlO
QksAoKrZW4yIVOHf/LgITdIG4VA7Al3s
=D8qy
-----END PGP SIGNATURE-----