1413567 - (android-rdl) Enable Retained Display Lists for Android

Status: RESOLVED FIXED

(Core :: Web Painting, enhancement, P2)

Description

[Miko Mynttinen]

At the moment, the content process check in http://searchfox.org/mozilla-central/rev/aa1343961fca52e8c7dab16788530da31aab7c8d/layout/generic/nsFrame.cpp#974 disables retained display lists on Android.

This bug tracks enabling retained display lists on Android.

Comment 1

[Botond Ballo [:botond]]

I understand this to be blocked on bug 1459312, which I hope to tackle in the near future.

Comment 2

[Chris Peterson [:cpeterson]]

Kats, will Retained Display List still be useful on Android after we have WebRender?

Comment 3

[Kartikaya Gupta (email:[e-mail])]

I believe so, yes. They affect different parts of the rendering pipeline, so we can have both.

Comment 4

[Chris Peterson [:cpeterson]]

Jessie says that the gfx team plans to work on Android RDL in Q2.

Comment 5

[Timothy Nikkel (:tnikkel)]

Attachment: Bug 1413567. Adjust test expectations for RDL on android. r=mattwoodrow

All three failures are small antialiasing differences in the text in the test. The text in the tests aren't even testing anything, they just describe what to expect from the test.

Until we get fuzzy matching in web platform tests (bug 1478472) we just have to mark them as fails, but we can be very specific about the platform, so we're probably still passing these tests on other android jobs if they are run there (not sure).

Comment 6

[Timothy Nikkel (:tnikkel)]

Attachment: Bug 1413567. Enable retained display lists for android. r=mattwoodrow

Even though we don't look at the chrome value on android in nsLayoutUtils::AreRetainedDisplayListsEnabled(), I flipped it just in case someone looks at that pref somewhere else.

Depends on D25280

Comment 7

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/a18b96a8205d
Adjust test expectations for RDL on android. r=mattwoodrow
https://hg.mozilla.org/integration/autoland/rev/ac5ec27136d7
Enable retained display lists for android. r=mattwoodrow

Comment 8

[Raul Gurzau (:RaulG)]

Backed out 2 changesets (bug 1413567) for xpcshell failures on Android platform on a CLOSED TREE.

Backout link: https://hg.mozilla.org/integration/autoland/rev/01c0722546ab4497e48d9404f2281176fb9796bc

Push with failures: https://treeherder.mozilla.org/#/jobs?repo=autoland&resultStatus=testfailed%2Cbusted%2Cexception&selectedJob=236753454&revision=ac5ec27136d703282444075027fd58de6be378a5

Log link: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=236756984&repo=autoland&lineNumber=1390

Log snippet:

[task 2019-03-28T22:04:44.853Z] 22:04:44 INFO - TEST-START | security/manager/ssl/tests/unit/test_ocsp_stapling.js
[task 2019-03-28T22:04:46.363Z] 22:04:46 WARNING - TEST-UNEXPECTED-FAIL | security/manager/ssl/tests/unit/test_ocsp_stapling.js | xpcshell return code: 139
[task 2019-03-28T22:04:46.364Z] 22:04:46 INFO - TEST-INFO took 1511ms
[task 2019-03-28T22:04:46.365Z] 22:04:46 INFO - >>>>>>>
[task 2019-03-28T22:04:46.365Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | xpcw: cd /sdcard/tests/xpc/security/manager/ssl/tests/unit
[task 2019-03-28T22:04:46.365Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | xpcw: xpcshell -r /sdcard/tests/xpc/c/httpd.manifest --greomni /data/local/xpcb/target.apk -m -s -e const _HEAD_JS_PATH = "/sdcard/tests/xpc/head.js"; -e const _MOZINFO_JS_PATH = "/sdcard/tests/xpc/p/mozinfo.json"; -e const _PREFS_FILE = "/sdcard/tests/xpc/user.js"; -e const _TESTING_MODULES_DIR = "/sdcard/tests/xpc/m"; -f /sdcard/tests/xpc/head.js -e const _HEAD_FILES = ["/sdcard/tests/xpc/security/manager/ssl/tests/unit/head_psm.js"]; -e const _JSDEBUGGER_PORT = 0; -e const _TEST_FILE = ["test_ocsp_stapling.js"]; -e const _TEST_NAME = "security/manager/ssl/tests/unit/test_ocsp_stapling.js"; -e _execute_test(); quit(0);
[task 2019-03-28T22:04:46.365Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | [829, Unnamed thread 47a0e080] WARNING: XPCOM objects created/destroyed from static ctor/dtor: file /builds/worker/workspace/build/src/xpcom/base/nsTraceRefcnt.cpp, line 194
[task 2019-03-28T22:04:46.366Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | [829, Unnamed thread 47a0e080] WARNING: XPCOM objects created/destroyed from static ctor/dtor: file /builds/worker/workspace/build/src/xpcom/base/nsTraceRefcnt.cpp, line 194
[task 2019-03-28T22:04:46.366Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | [829, Unnamed thread 47a0e080] WARNING: XPCOM objects created/destroyed from static ctor/dtor: file /builds/worker/workspace/build/src/xpcom/base/nsTraceRefcnt.cpp, line 194
[task 2019-03-28T22:04:46.366Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | [829, Unnamed thread 47a0e080] WARNING: XPCOM objects created/destroyed from static ctor/dtor: file /builds/worker/workspace/build/src/xpcom/base/nsTraceRefcnt.cpp, line 194
[task 2019-03-28T22:04:46.366Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | [829, Main Thread] WARNING: No Android crash handler set: file /builds/worker/workspace/build/src/toolkit/crashreporter/nsExceptionHandler.cpp, line 1463
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | \x07[829, Main Thread] ###!!! ASSERTION: greprefs.js:3181: prefs parse error: expected ',' or ')' after pref value: 'Error', file /builds/worker/workspace/build/src/modules/libpref/Preferences.cpp, line 1740
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #01: ???[libxul.so +0x3a57122]
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #02: ???[libxul.so +0x3a56efa]
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #03: ???[libxul.so +0x6b4abe]
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #04: ???[libxul.so +0x6ad9ea]
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #05: ???[libxul.so +0x69fc2e]
[task 2019-03-28T22:04:46.368Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #06: ???[libxul.so +0x69f86a]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #07: ???[libxul.so +0x65d602]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #08: ???[libxul.so +0x667720]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #09: ???[libxul.so +0x665472]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #10: ???[libxul.so +0x669434]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #11: ???[libxul.so +0x5dfc86]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #12: ???[libxul.so +0x6912d4]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #13: ???[libxul.so +0x6a5778]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #14: ???[libxul.so +0x6aedea]
[task 2019-03-28T22:04:46.369Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #15: ???[libxul.so +0x6aedbe]
[task 2019-03-28T22:04:46.370Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #16: ???[libxul.so +0xcdbcfe]
[task 2019-03-28T22:04:46.370Z] 22:04:46 INFO - security/manager/ssl/tests/unit/test_ocsp_stapling.js | #17: ???[libxul.so +0xcdb812]

Comment 9

[Timothy Nikkel (:tnikkel)]

Argh, forgot a closing parenthesis on the pref.

Comment 10

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/7050f5bf674a
Adjust test expectations for RDL on android. r=mattwoodrow
https://hg.mozilla.org/integration/autoland/rev/47dd44501599
Enable retained display lists for android. r=mattwoodrow

Comment 11

[Cristina Coroiu [:ccoroiu]]

https://hg.mozilla.org/mozilla-central/rev/7050f5bf674a
https://hg.mozilla.org/mozilla-central/rev/47dd44501599

Comment 12

[Andreea Pavel [:apavel]]

Backed out for causing bug 1523833 to permafail

Push with failures: https://treeherder.mozilla.org/#/jobs?repo=autoland&searchStr=r42&tochange=bb3a9a19e1083bf1418b72bc9617fce9de83178e&fromchange=d33c1566c2d9a0958fcbd48f5a275241519b10e4&selectedJob=236788842

Failure log (tier1): https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=236788842&repo=autoland&lineNumber=2331
tier2: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=236797833&repo=autoland&lineNumber=3589

Backout: https://hg.mozilla.org/mozilla-central/rev/579cc975268f03bcc335f099a25215b87836a4ef

Comment 13

[Timothy Nikkel (:tnikkel)]

Weird, I've got dozens of green r42 debug android 4.3 on my try runs without a single failure. Wondering if something else landed that caused an bad interaction.

Comment 14

[Andreea Pavel [:apavel]]

Timothy, this is the range between the green job and the perma-failure: https://treeherder.mozilla.org/#/jobs?repo=autoland&searchStr=r42&tochange=7cfcdf742d19ef55f45ef9150a15908247112f03&fromchange=01c0722546ab4497e48d9404f2281176fb9796bc

Comment 15

[Timothy Nikkel (:tnikkel)]

Oh sorry, I meant if something landed between my last try push and pushing this to autoland.

Comment 16

[Timothy Nikkel (:tnikkel)]

Bisecting on try points to

https://hg.mozilla.org/mozilla-central/rev/685208691386
bug 1535507

Comment 17

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/53f0254dfc7e
Adjust test expectations for RDL on android. r=mattwoodrow
https://hg.mozilla.org/integration/mozilla-inbound/rev/08f895aaa82a
Enable retained display lists for android. r=mattwoodrow

Comment 18

[Arthur Iakab [arthur_iakab]]

https://hg.mozilla.org/mozilla-central/rev/53f0254dfc7e
https://hg.mozilla.org/mozilla-central/rev/08f895aaa82a

Comment 19

[Pascal Chevrel:pascalc]

Should it be mentioned in our Android 68 Beta release notes as we did for our Desktop beta release notes in version 61?

Comment 20

[Timothy Nikkel (:tnikkel)]

(In reply to Pascal Chevrel:pascalc from comment #19)

Should it be mentioned in our Android 68 Beta release notes as we did for
our Desktop beta release notes in version 61?

Sounds like it should!

Comment 21

[Julien Cristau [:jcristau]]

This is in the 68.0beta android release notes as "Improved web page painting performance by avoiding redundant calculations during paint".