+2007-05-21 Jan Schmidt <thaytan@mad.scientist.com>
+
+ * tests/check/gst/gstbuffer.c: (GST_START_TEST):
+
+ Disable part of the gst_buffer_try_new_and_alloc test, because
+ it can happily succeed on 64-bit systems where there's more address
+ space available.
+
2007-05-21 Sebastian Dröge <slomo@circular-chaos.org>
* tests/check/gst/gstpad.c: (GST_START_TEST), (gst_pad_suite):
GST_BUFFER_DATA (buf)[640 * 479 * 4 + 479] = 0xff;
gst_buffer_unref (buf);
+#if 0
+ /* Disabled this part of the test, because it happily succeeds on 64-bit
+ * machines that have enough memory+swap, because the address space is large
+ * enough. There's not really any way to test the failure case except by
+ * allocating chunks of memory until it fails, which would suck. */
+
/* now this better fail (don't run in valgrind, it will abort
* or warn when passing silly arguments to malloc) */
if (!RUNNING_ON_VALGRIND) {
buf = gst_buffer_try_new_and_alloc ((guint) - 1);
fail_unless (buf == NULL);
}
+#endif
}
GST_END_TEST;