Closed Bug 917508 Opened 6 years ago Closed 4 years ago

green up jsreftest-5 - test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js fails on Android x86 emulator

Categories

(Core :: JavaScript Engine, defect)

defect
Not set

Tracking

()

RESOLVED WONTFIX

People

(Reporter: gbrown, Unassigned)

References

Details

(Whiteboard: [leave open] [test disabled])

Attachments

(1 file)

+++ This bug was initially created as a clone of Bug #854320 +++

We are trying to get our test suites running in an Android x86 environment. All but one of the js-reftests succeed; the single, consistent failure is seen here:

https://tbpl.mozilla.org/php/getParsedLog.php?id=27944067&tree=Cedar&full=1#error0

18:45:08     INFO -  REFTEST TEST-START | http://10.0.2.2:8858/jsreftest/tests/jsreftest.html?test=test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js
18:45:08     INFO -  REFTEST TEST-LOAD | http://10.0.2.2:8858/jsreftest/tests/jsreftest.html?test=test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js | 879 / 1006 (87%)
18:45:08     INFO -  REFTEST TEST-UNEXPECTED-FAIL | http://10.0.2.2:8858/jsreftest/tests/jsreftest.html?test=test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js | Unknown http://10.0.2.2:8858/jsreftest/tests/test262/shell.js:921: Error: Test262 error: #2.2: -1 % -1 === - 0. Actual: +0 item 1
18:45:08     INFO -  REFTEST INFO | Saved log: START http://10.0.2.2:8858/jsreftest/tests/jsreftest.html?test=test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] OnDocumentLoad triggering AfterOnLoadScripts
18:45:08     INFO -  REFTEST INFO | Saved log: Initializing canvas snapshot
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] AfterOnLoadScripts belatedly entering WaitForTestEnd
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] WaitForTestEnd: Adding listeners
18:45:08     INFO -  REFTEST INFO | Saved log: Initializing canvas snapshot
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] MakeProgress: STATE_WAITING_TO_FIRE_INVALIDATE_EVENT
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] MakeProgress: dispatching MozReftestInvalidate
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] MakeProgress: STATE_WAITING_FOR_REFTEST_WAIT_REMOVAL
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] MakeProgress: STATE_WAITING_FOR_SPELL_CHECKS
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] MakeProgress: STATE_WAITING_TO_FINISH
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] MakeProgress: Completed
18:45:08     INFO -  REFTEST INFO | Saved log: [CONTENT] RecordResult fired
18:45:08     INFO -  REFTEST INFO | Saved log: RecordResult fired
18:45:08     INFO -  REFTEST INFO | Loading a blank page
18:45:08     INFO -  REFTEST TEST-END | http://10.0.2.2:8858/jsreftest/tests/jsreftest.html?test=test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js
Marty suggests that Android x86's libc might be at fault for this.  I'm too busy getting Intl stuff shipping on desktop to take this on now/immediately, but I might be able to look soonish.

The expedient short-term fix -- although I suspect there are more failure modes than just this one, if libc is indeed at fault -- is to add the appropriate skip-if to js/src/tests/jstests.list.  Feel free to poke me for a review on that if you want to go that route for now.
Here is that short-term fix. I am disabling the test for AndroidVersion>=17 -- that avoids affecting the existing tegra and panda jobs. The x86 emulator is currently running android 17 with plans to upgrade to android 18 in the near future.

Try run confirming that this does not affect the test on existing test platforms: https://tbpl.mozilla.org/?tree=Try&rev=b5edec60f6d4
Attachment #806751 - Flags: review?(jwalden+bmo)
Review ping? (Simple patch -- just trying to get this test disabled so that we can move forward with Android x86 unit tests).
Attachment #806751 - Flags: review?(jwalden+bmo) → review+
No longer blocks: 892688
Blocks: 891959
Summary: test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js fails on Android x86 emulator → green up jsreftest-5 - test262/ch11/11.5/11.5.3/S11.5.3_A4_T2.js fails on Android x86 emulator
This bug is actually a follow up from the work that has happened on bug 891959.
No longer blocks: 891959
Depends on: 891959
Assignee: general → nobody
We're not running this suite on Android x86 now.  Test manifest updated in bug 1257948.
Status: NEW → RESOLVED
Closed: 4 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.