425068 - test_bug392567.html (and other jar: tests) make the mochitest server stop responding

Status: RESOLVED DUPLICATE of bug 418009

(Testing :: Mochitest, defect)

Description

[Jesse Ruderman]

Steps to reproduce:
    python runtests.py --test-path=content/html/content/test/ --autorun
or
    python runtests.py --test-path=content/html/content/test/test_bug392567.html

Result:
* The mochitest server on port 8888 stops responding (even to Safari).  This causes it and all subsequent mochitests to time out.
* The assertions in bug 406905 fire.

Is this only happening on my machine?  (MBP running Tiger)

I don't see any debugging output from the server, but that might be a result of bug 418009.  I tried using runtests.pl (instead of runtests.py) to work around that bug, but then I hit bug 420150!  (It's frustrating to run into mazes of bugs where every potential workaround reveals another bug.)

docshell/test/test_bug369814.html also triggers the bug.  Both tests involve the jar: protocol.

Comment 1

[Jeff Walden [:Waldo]]

Attachment: runtests.pl output

Doing a rebuild with |make -f client.mk build| got rid of the CC asserts for whatever reason; when I did so, I get the output in the attachment.  Not sure what went wrong there, but I'm pretty sure it's not a server problem (or at least it's not caused by a bug in the server).  Assuming the same thing happens with the Python one, the server hang is all but certainly just the stdout/stderr buffer Python's keeping for the server filling up and not being emptied, the bug mentioned in comment 0.

Comment 2

[Jeff Walden [:Waldo]]

Everything works fine with the patch in bug 418009, and we were pretty sure the problem was the same anyway (and comment 1 seems to confirm), so marking as a dup...

Comment 3

[Jeff Walden [:Waldo]]

Moving a bunch of Core :: Testing bugs to Testing :: Mochitest to clear out the former, which is obsolete now that we have more specialized categories for such bugs; filter on the string "MochitestMmMm" to delete all these notifications.