]> granicus.if.org Git - python/commitdiff
XXX about extreme expense of test_no_leaking() on Windows. I'm not sure
authorTim Peters <tim.peters@gmail.com>
Wed, 13 Oct 2004 03:14:40 +0000 (03:14 +0000)
committerTim Peters <tim.peters@gmail.com>
Wed, 13 Oct 2004 03:14:40 +0000 (03:14 +0000)
what this is trying to do.  If it's necessary for it to create > 1000
processes, it should be controlled by a new resource and not run by
default on Windows.

Lib/test/test_subprocess.py

index ba8ffa21992991a2aff7c6d6661f4db53ada8cbc..3312f7cdbdc1855c4f8e232bbd3a1c2cd8f6f49b 100644 (file)
@@ -317,6 +317,7 @@ class ProcessTestCase(unittest.TestCase):
             # Interpreter without universal newline support
             self.assertEqual(stdout, "line1\nline2\rline3\r\nline4\r\nline5\nline6")
 
+    # XXX test_no_leaking takes > a minute to run on a high-end WinXP Pro box
     def test_no_leaking(self):
         # Make sure we leak no resources
         for i in range(1026):