862051 - Intermittent B2G emulator test_descriptor_storage.html | application timed out after 330 seconds with no output | Test timed out. | [SimpleTest.finish()] this test already called finish! | called finish() multiple times

Status: RESOLVED WORKSFORME

(Core :: CSS Parsing and Computation, defect)

Description

[Ryan VanderMeulen [:RyanVM]]

https://tbpl.mozilla.org/php/getParsedLog.php?id=21821414&tree=Mozilla-Inbound

b2g_ics_armv7a_gecko_emulator_vm mozilla-inbound opt test mochitest-9 on 2013-04-15 10:29:20 PDT for push c626d0a4b3ac
slave: tst-linux64-ec2-158

11:22:42  WARNING -  TEST-UNEXPECTED-FAIL | /tests/layout/style/test/test_descriptor_storage.html | application timed out after 330 seconds with no output
11:22:42     INFO -  INFO | automation.py | Application ran for: 0:47:11.138276
11:22:42     INFO -  INFO | zombiecheck | Reading PID log: /tmp/tmpxZwQG8pidlog
11:23:46     INFO -  checking for crashes in '/data/local/tests/profile/minidumps'
11:23:46     INFO -  WARNING | leakcheck | refcount logging is off, so leaks can't be detected!
11:23:46     INFO -  INFO | runtests.py | Running tests: end.
11:24:01    ERROR - Return code: 1

Comment 1

[Daniel Holbert [:dholbert]]

More context from log:
{
11:17:11     INFO -  9292 INFO TEST-PASS | /tests/layout/style/test/test_descriptor_storage.html | setting 'seriff' on 'font-family'
11:22:42     INFO -  9293 INFO TEST-PASS | /tests/layout/style/test/test_descriptor_storage.html | serialization should match descriptor value
[...]
11:22:42     INFO -  9318 INFO TEST-PASS | /tests/layout/style/test/test_descriptor_storage.html | parse+serialize should be idempotent for 'font-stretch: semi-condensed'
11:22:42     INFO -  9319 INFO TEST-PASS | /tests/layout/style/test/test_descriptor_storage.html | setting 'semi-expanded' on 'font-stretch'
11:22:42  WARNING -  TEST-UNEXPECTED-FAIL | /tests/layout/style/test/test_descriptor_storage.html | application timed out after 330 seconds with no output
}

So there's a 5 minute 31 second gap between
> setting 'seriff' on 'font-family'
...and the next line of output...
> serialization should match descriptor value

From that point on, it *looks* like everything proceeds correctly, except that the timeout-detector kicks in and declares the test as having timed out, even though it's actually running.  (Note that the "INFO TEST-PASS" line right before the timeout has the same timestamp as the timeout message.)

So the question is, why do we have that huge delay after "setting 'seriff'", I wonder.

Comment 2

[Daniel Holbert [:dholbert]]

CC'ing dbaron -- it looks like he wrote the test, so he's more likely than I am to know what might be going on here.

Comment 3

[Daniel Holbert [:dholbert]]

(The "called finish multiple times" stuff looks like a harness bug, rather than a test bug -- it's because the harness declared the test as having timed out, but it apparently allowed the test to keep running, and the test called finish() after it had been declared as timed-out.)

Comment 4

[Ed Morley [:emorley]]

(OrangeWFM for bugs not modified in > 2 months)

Comment 5

[Carsten Book [:Tomcat]]

reopen, happend again

Comment 6

[Carsten Book [:Tomcat]]

err different issue , sorry