From what I can tell, this is caused by images decoding asynchronously. Usually it comes back soon enough, but sometimes it doesn't and we end up trying to scroll the full frame, rather than the underlying image. We can fix this easily if bug 1416328 was implemented. In turn, I have an incentive to see this fixed so I can land bug 1436247 which was backed out due to slightly changing the timing for this bug and permafailing it.
Assignee: nobody → aosmond
Test is failing frequently on Linux. (~127 failures in the last 7days) Here is a recent log: https://treeherder.mozilla.org/logviewer.html#?repo=mozilla-inbound&job_id=164083783&lineNumber=2684 [task 2018-02-24T02:07:42.371Z] 02:07:42 INFO - TEST-START | dom/events/test/test_DataTransferItemList.html [task 2018-02-24T02:07:43.409Z] 02:07:43 INFO - TEST-INFO | started process screentopng [task 2018-02-24T02:07:43.842Z] 02:07:43 INFO - TEST-INFO | screentopng: exit 0 [task 2018-02-24T02:07:43.842Z] 02:07:43 INFO - Buffered messages logged at 02:07:43 [task 2018-02-24T02:07:43.842Z] 02:07:43 INFO - SpawnTask.js | Entering test [task 2018-02-24T02:07:43.842Z] 02:07:43 INFO - Buffered messages finished [task 2018-02-24T02:07:43.844Z] 02:07:43 INFO - TEST-UNEXPECTED-FAIL | dom/events/test/test_DataTransferItemList.html | TypeError: _EU_maybeWrap(...) is null - Should not throw any errors [task 2018-02-24T02:07:43.845Z] 02:07:43 INFO - createDragEventObject@chrome://mochikit/content/tests/SimpleTest/EventUtils.js:2428:38 [task 2018-02-24T02:07:43.846Z] 02:07:43 INFO - synthesizeDragOver@chrome://mochikit/content/tests/SimpleTest/EventUtils.js:2517:15 [task 2018-02-24T02:07:43.847Z] 02:07:43 INFO - synthesizeDrop@chrome://mochikit/content/tests/SimpleTest/EventUtils.js:2596:34 [task 2018-02-24T02:07:43.848Z] 02:07:43 INFO - @chrome://mochitests/content/chrome/dom/events/test/test_DataTransferItemList.html:204:5 [task 2018-02-24T02:07:43.849Z] 02:07:43 INFO - async*add_task/</<@chrome://mochikit/content/tests/SimpleTest/SpawnTask.js:298:21 [task 2018-02-24T02:07:43.850Z] 02:07:43 INFO - onFulfilled@chrome://mochikit/content/tests/SimpleTest/SpawnTask.js:69:15 [task 2018-02-24T02:07:43.851Z] 02:07:43 INFO - co/<@chrome://mochikit/content/tests/SimpleTest/SpawnTask.js:58:5 [task 2018-02-24T02:07:43.852Z] 02:07:43 INFO - co@chrome://mochikit/content/tests/SimpleTest/SpawnTask.js:54:10 [task 2018-02-24T02:07:43.853Z] 02:07:43 INFO - setTimeout handler*SimpleTest_setTimeoutShim@chrome://mochikit/content/tests/SimpleTest/SimpleTest.js:676:12 [task 2018-02-24T02:07:43.854Z] 02:07:43 INFO - add_task@chrome://mochikit/content/tests/SimpleTest/SpawnTask.js:271:7 [task 2018-02-24T02:07:43.854Z] 02:07:43 INFO - @chrome://mochitests/content/chrome/dom/events/test/test_DataTransferItemList.html:27:3 [task 2018-02-24T02:07:43.855Z] 02:07:43 INFO - GECKO(1981) | MEMORY STAT vsizeMaxContiguous not supported in this build configuration. [task 2018-02-24T02:07:43.857Z] 02:07:43 INFO - GECKO(1981) | MEMORY STAT | vsize 825MB | residentFast 257MB | heapAllocated 102MB [task 2018-02-24T02:07:43.858Z] 02:07:43 INFO - TEST-OK | dom/events/test/test_DataTransferItemList.html | took 1054ms :overholt, can you please take a look?
Hello! Joel, can you please review this patch? I didn't disabled it on debug because the failures I saw in Orange factor were only on try and the one on inbound was missclassified. There was also a file named chrome.ini where there was this test name in it. Should I have edited that ini file or it's ok if I created one with this test name? And also what happens with the new file we created? It gets added to the remote repo when the patch is landed? Thanks.
Attachment #8953915 - Flags: review?(jmaher)
Comment on attachment 8953915 [details] [diff] [review] Disable on Linux !debug Review of attachment 8953915 [details] [diff] [review]: ----------------------------------------------------------------- sorry for the confusion here- this would work great if this were a web-platform-test (testing/web-platform/tests/...), but in this case it is a mochitest chrome (hence the chrome.ini), so you can modify the existing chrome.ini file. In general when you see mochitest.ini, chrome.ini, browser.ini, or xpcshell.ini- you modify in the existing file (make note that we have a lot of other variations like browser2.ini, or mochitest-oop.ini, etc.)
Attachment #8953915 - Flags: review?(jmaher) → review-
Oh, initially that was what I thought but eventually did it another .ini. No worries, I've now modified chrome.ini with those changes. Thanks.
Comment on attachment 8953976 [details] [diff] [review] Disable on Linux !debug Review of attachment 8953976 [details] [diff] [review]: ----------------------------------------------------------------- looks great
Attachment #8953976 - Flags: review?(jmaher) → review+
Pushed by email@example.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/5684c810fb4b Disable dom/events/test/test_DataTransferItemList.html for frequent failures. r=jmaher
Pushed by firstname.lastname@example.org: https://hg.mozilla.org/integration/mozilla-inbound/rev/723b390519de Fix the syntax for disabling dom/events/test/test_DataTransferItemList.html on a CLOSED TREE
I think Andrew's working on this given comment 7.
Flags: needinfo?(overholt) → needinfo?(aosmond)
Whiteboard: [stockwell disable-recommended] → [stockwell disabled]
You need to log in before you can comment on or make changes to this bug.