We've got this orange on m-i TBPL right now: https://tbpl.mozilla.org/php/getParsedLog.php?id=15798566&tree=Mozilla-Inbound Android no-ionmonkey Tegra 250 mozilla-inbound opt test mochitest-5 on 2012-10-03 21:56:27 PDT for push 1cceedd227a7 slave: tegra-269 builder: mozilla-inbound_tegra_android-noion_test-mochitest-5 slave: tegra-269 starttime: 1349326587.59 results: warnings (1) buildid: 20121003211916 builduid: 65a92f95b95146608340b72d17b38dc9 revision: 1cceedd227a7 ...and the log gives no clue as to what went wrong. All of the "exit code" lines are "exit code 0" (success), and it in fact looks essentially identical to this successful mochitest-5 log: https://tbpl.mozilla.org/php/getParsedLog.php?id=15798117&tree=Mozilla-Inbound (that one is successful, just for comparison ^)

(gah, accidentally hit "enter" while the wrong field was focused & filed bug before I was done) So I was saying... it looks like the logs are basically the same, except the failing log has this... > DeviceManager: pulling file '/mnt/sdcard/tests/logs/mochitest.log' unsuccessful: No such file or directory ...in place of all the actual mochitest output. ...and it looks like philor already has a bug on this, cool.

this looks like Fennec crashed at startup: First we see: W/GeckoApp( 1652): unexpected exception, passing url directly to Gecko but we should explicitly catch this W/GeckoApp( 1652): java.net.UnknownHostException: Host is unresolved: mochi.test:8888 W/GeckoApp( 1652): at java.net.Socket.connect(Socket.java:1038) then a bit later I/DEBUG ( 937): Build fingerprint: 'nvidia/harmony/harmony/harmony:2.2/FRF91/20110202.102810:eng/test-keys' I/DEBUG ( 937): pid: 1652, tid: 1652 >>> org.mozilla.fennec <<< I/DEBUG ( 937): signal 11 (SIGSEGV), fault addr 0028a006 Which then dumps out a whole (on android logcat) stack. I'm not sure if this is a case of the devices NIC dying or something else.