1055483 - Intermittent font-inflation/height-constraint-percent-7.html | image comparison (==), max difference: 255, number of differing pixels: 3

Status: RESOLVED WORKSFORME

(Testing :: Reftest, defect)

Description

[Ed Morley [:emorley]]

Attachment: reftest log

Ubuntu VM 12.04 mozilla-inbound debug test reftest on 2014-08-18 23:57:37 PDT for push 1db9d4b9d740

slave: tst-linux32-spot-699

https://tbpl.mozilla.org/php/getParsedLog.php?id=46243130&tree=Mozilla-Inbound

{
00:51:21     INFO -  REFTEST TEST-START | file:///builds/slave/test/build/tests/reftest/tests/layout/reftests/font-inflation/height-constraint-percent-7.html
00:51:21     INFO -  SET PREFERENCE pref(font.size.inflation.emPerLine,15)
00:51:21     INFO -  SET PREFERENCE pref(font.size.inflation.forceEnabled,true)
00:51:21     INFO -  SET PREFERENCE pref(font.size.inflation.lineThreshold,0)
00:51:21     INFO -  REFTEST TEST-LOAD | file:///builds/slave/test/build/tests/reftest/tests/layout/reftests/font-inflation/height-constraint-percent-7.html | 5623 / 11220 (50%)
00:51:21     INFO -  ++DOMWINDOW == 47 (0xa158a850) [pid = 1695] [serial = 15435] [outer = 0xa5a33760]
00:51:22     INFO -  RESTORE PREFERENCE pref(font.size.inflation.lineThreshold,400)
00:51:22     INFO -  RESTORE PREFERENCE pref(font.size.inflation.forceEnabled,false)
00:51:22     INFO -  RESTORE PREFERENCE pref(font.size.inflation.emPerLine,0)
00:51:22     INFO -  REFTEST TEST-LOAD | file:///builds/slave/test/build/tests/reftest/tests/layout/reftests/font-inflation/height-constraint-percent-7-ref.html | 5623 / 11220 (50%)
00:51:22     INFO -  ++DOMWINDOW == 48 (0xa158aa40) [pid = 1695] [serial = 15436] [outer = 0xa5a33760]
00:51:22     INFO -  REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/test/build/tests/reftest/tests/layout/reftests/font-inflation/height-constraint-percent-7.html | image comparison (==), max difference: 255, number of differing pixels: 3
}

Comment 1

[Jonathan Kew [:jfkthame]]

There's something fishy about this: the data URLs for the "test" and "reference" images in that log are identical, AFAICT. Which suggests some kind of flakiness in the test framework or the logging, I guess.

Comment 2

[Jonathan Kew [:jfkthame]]

Moving this to Testing:Reftest, as it's apparently reporting a difference that doesn't exist. Or else it's not logging the right images. Either way, it seems to be broken.

Comment 3

[Ed Morley [:emorley]]

Mass-closing intermittent-failure bugs filed by me, that have not occurred recently and do not have the leave-open keyword set.