Added another crasher, which hit me today (I was not intentionally

writing such code, of course, but it took some gdb time to figure out
what my bug was).

diff --git a/Lib/test/crashers/recursion_limit_too_high.py b/Lib/test/crashers/recursion_limit_too_high.py
new file mode 100644 (file)
index 0000000..1fa4d32
--- /dev/null
+++ b/Lib/test/crashers/recursion_limit_too_high.py
@@ -0,0 +1,16 @@
+# The following example may crash or not depending on the platform.
+# E.g. on 32-bit Intel Linux in a "standard" configuration it seems to
+# crash on Python 2.5 (but not 2.4 nor 2.3).  On Windows the import
+# eventually fails to find the module, possibly because we run out of
+# file handles.
+
+# The point of this example is to show that sys.setrecursionlimit() is a
+# hack, and not a robust solution.  This example simply exercices a path
+# where it takes many C-level recursions, consuming a lot of stack
+# space, for each Python-level recursion.  So 1000 times this amount of
+# stack space may be too much for standard platforms already.
+
+import sys
+if 'recursion_limit_too_high' in sys.modules:
+    del sys.modules['recursion_limit_too_high']
+import recursion_limit_too_high