819138 - Intermittent test_aboutmemory.xul | application timed out after 330 seconds with no output

Status: RESOLVED WORKSFORME

(Toolkit :: about:memory, defect)

Description

[Ed Morley [:emorley]]

Rev4 MacOSX Snow Leopard 10.6 mozilla-inbound debug test mochitest-other on 2012-12-03 11:29:57 PST for push d8c788ff2545

slave: talos-r4-snow-045

https://tbpl.mozilla.org/php/getParsedLog.php?id=17563255&tree=Mozilla-Inbound

Just the highlights, since quite a lot of output...

{
###!!! ASSERTION: Invalid value for explicit/js/compartment(http://too-big.com/)/stuff: 'false', file aboutMemory.js, line 0
...
###!!! ASSERTION: Invalid value for explicit/(2 tiny): 'false', file aboutMemory.js, line 0
...
###!!! ASSERTION: Invalid value for explicit/(2 tiny)/neg1: 'false', file aboutMemory.js, line 0
...
###!!! ASSERTION: Invalid value for explicit/(2 tiny)/heap-unclassified: 'false', file aboutMemory.js, line 0
...
###!!! ASSERTION: Invalid value for other1: 'false', file aboutMemory.js, line 0
...
###!!! ASSERTION: Invalid value for other2: 'false', file aboutMemory.js, line 0
...
###!!! ASSERTION: Invalid value for other3: 'false', file aboutMemory.js, line 0
...
<and so on>
...
TEST-UNEXPECTED-FAIL | chrome://mochitests/content/chrome/toolkit/components/aboutmemory/tests/test_aboutmemory.xul | application timed out after 330 seconds with no output
}

Comment 1

[Nicholas Nethercote [inactive]]

These assertion failures are expected, they arise from deliberately bogus values in the test.  They're triggered like this:

  let debug = Cc["@mozilla.org/xpcom/debug;1"].getService(Ci.nsIDebug2);
  if (debug.isDebugBuild) {
    debug.assertion(aMsg, "false", "aboutMemory.js", 0);
  }

So perhaps the assertions aren't the problem, but they look suspicious so you included them?  The real issue is the hang...

Comment 2

[Ed Morley [:emorley]]

(In reply to Nicholas Nethercote [:njn] from comment #1)
> So perhaps the assertions aren't the problem, but they look suspicious so
> you included them?  The real issue is the hang...

Correct :-)

Comment 3

[Nicholas Nethercote [inactive]]

The hang happens (IIUC) when the actual output doesn't match the expected output.  I've seen this occasionally happen when a new memory reporter gets registered mid-test, which screws things up.  It's possible that patch 2 in bug 848560 will fix this, by preventing new reporters from being reported mid-test.

Comment 4

[Ed Morley [:emorley]]

(OrangeWFM bugs not modified in > 2 months)