695772d145a062aea963c2aaa35f50c802cdc98b: Bug 1107299: Only block shared surface usage when that is broken. r=jrmuizel
Bas Schouten <bschouten@mozilla.com> - Sat, 06 Dec 2014 00:02:18 +0000 - rev 218610
Bug 1107299: Only block shared surface usage when that is broken. r=jrmuizel Currently when the usage of surface sharing is broken on certain drivers, be block D3D11 entirely and fall back to Basic composition. There's no real benefit in this case from disabling D3D11 entirely, and we might as well just use it with software composition like we do when we don't have Direct2D, and in addition to that disable texture sharing for webGL, falling back to slow WebGL which again is no worse than the situation when using Basic or D3D9 composition.
823227372483eb3b066c36df5f93036918560880: Bug 1107297: Only recomposite the damaged rect with D3D11. r=jrmuizel
Bas Schouten <bschouten@mozilla.com> - Sat, 06 Dec 2014 00:02:18 +0000 - rev 218609
Bug 1107297: Only recomposite the damaged rect with D3D11. r=jrmuizel This patch makes us behave similarly to when we're using the BasicCompositor, essentially we will clip all drawing to the final display to the area that was labeled as invalid. When DXGI 1.2 is available we will then also report the damaged area to the Present call so that can then be used to minimize the amount of bits that actually need to be blitted to the screen. Since we're no longer recompositing the whole screen this means we should also only clear the damaged area of the window.
2a61df4eaa2d95c62aaa6ec6d411afd0c9c897f1: Bug 1089417 patch 8 - Only drop MediumFeaturesChanged on the floor if we've never computed style before, rather than never computed style using this rule processor. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:01 -0800 - rev 218608
Bug 1089417 patch 8 - Only drop MediumFeaturesChanged on the floor if we've never computed style before, rather than never computed style using this rule processor. r=heycam This depends on patches 4 and 7.
38feece7ff7373043c150bb2f4441cc56c282b13: Bug 1089417 patch 7 - Save the previous media query cache key on the rule processor. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218607
Bug 1089417 patch 7 - Save the previous media query cache key on the rule processor. r=heycam This depends on patches 5 and 6, and is needed for patch 8.
8347130b79e809f0fd1b085a824a10407da9600a: Bug 1089417 patch 6 - Pass the previous CSS rule processor to the constructor of the new one (when we replace one with another). r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218606
Bug 1089417 patch 6 - Pass the previous CSS rule processor to the constructor of the new one (when we replace one with another). r=heycam This is needed for patch 7.
7abee9c623d974bb1c1302551f2754de4153b224: Bug 1089417 patch 5 - Add method to save the current media query result cache key from a rule processor. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218605
Bug 1089417 patch 5 - Add method to save the current media query result cache key from a rule processor. r=heycam This is needed for patch 7.
f4cea909daca0fd8bed90f2d5d7eaeb3b894fa1f: Bug 1089417 patch 4 - Define equality operators on nsMediaQueryResultCacheKey and the types it contains. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218604
Bug 1089417 patch 4 - Define equality operators on nsMediaQueryResultCacheKey and the types it contains. r=heycam This is needed for the equality comparison in nsCSSRuleProcessor::MediumFeaturesChanged in patch 8.
cfe005e3ad7675b7ef43d272b654e869bfcb33c3: Bug 1089417 patch 3 - Add mochitest. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218603
Bug 1089417 patch 3 - Add mochitest. r=heycam Note that if I make the style element currently /* empty */ have a rule in it, the test passes. Patch 8 also makes the test pass.
e1e2b51647f2ef1804ab433d357467d4c48cede2: Bug 1089417 patch 2 - Add assertion that RecordStyleSheetChange is called inside an update. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218602
Bug 1089417 patch 2 - Add assertion that RecordStyleSheetChange is called inside an update. r=heycam This was just something that seemed worth asserting in the process of debugging, since I wanted to see if it was the problem.
7043b617f43549d04c58faf6209b8151f6d70231: Bug 1089417 patch 1 - Add assertions that rule processors are valid before we access them. r=heycam
L. David Baron <dbaron@dbaron.org> - Fri, 05 Dec 2014 15:50:00 -0800 - rev 218601
Bug 1089417 patch 1 - Add assertions that rule processors are valid before we access them. r=heycam This was just something that seemed worth asserting in the process of debugging, since I wanted to see if it was the problem.
0ea4d63621093a65aa950030fbaf51337bc53f8c: Bug 1102048 (Part 18) - Make image/src files comply with the Mozilla Coding Style Guide. r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 04 Dec 2014 11:43:00 -0800 - rev 218600
Bug 1102048 (Part 18) - Make image/src files comply with the Mozilla Coding Style Guide. r=seth
cbe0505a8c9d038aeefaf4f3a008966f7f5089fd: Bug 1102048 (Part 16) - Make image/src files comply with the Mozilla Coding Style Guide. r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Thu, 04 Dec 2014 11:23:00 -0800 - rev 218599
Bug 1102048 (Part 16) - Make image/src files comply with the Mozilla Coding Style Guide. r=seth
dde067dbfe2739942b5fa202e04dd8b055b83600: Bug 1102048 (Part 22) - Make image/src files comply with the Mozilla Coding Style Guide. r=seth
Glenn Randers-Pehrson <glennrp+bmo@gmail.com> - Fri, 05 Dec 2014 15:12:08 -0800 - rev 218598
Bug 1102048 (Part 22) - Make image/src files comply with the Mozilla Coding Style Guide. r=seth
8878e674973db4a6b8ae046dda700b7a520675cf: Bug 1108059 - Enable mp4, not webm, for MediaSource on MacOS X. r=ajones
Ralph Giles <giles@mozilla.com> - Fri, 05 Dec 2014 14:33:00 -0800 - rev 218597
Bug 1108059 - Enable mp4, not webm, for MediaSource on MacOS X. r=ajones This makes the setting the same as on Windows. This will reduce the amount of testing we get for WebM, since only Linux will be left with that enabled, but reduces the exposed code surface while we're trying to stabilize things enough to ship support.
4d850020e23a51c41f116c204fe4a1ee370512e8: Bug 1108023 - update idb path in mozrunner, r=ahal
Malini Das <mdas@mozilla.com> - Fri, 05 Dec 2014 14:32:38 -0800 - rev 218596
Bug 1108023 - update idb path in mozrunner, r=ahal
7cf7a6ea28c251d2fafb4e71c158a0a7d1fc4fc8: Bug 1098126 - Disable MoofParser for EME; r=rillian
Anthony Jones <ajones@mozilla.com> - Fri, 05 Dec 2014 14:17:23 -0800 - rev 218595
Bug 1098126 - Disable MoofParser for EME; r=rillian
21aab02fb41c511bb7ef8ed1cbd76cc9f6649111: Bug 1098126 - Add CENC support to MoofParser; r=edwin
Anthony Jones <ajones@mozilla.com> - Fri, 05 Dec 2014 14:17:23 -0800 - rev 218594
Bug 1098126 - Add CENC support to MoofParser; r=edwin
55dba2167b365f6b98af513ff57ed79740271170: Bug 1098126 - MoofParser forced moof read; r=mattwoodrow
Anthony Jones <ajones@mozilla.com> - Fri, 05 Dec 2014 14:17:23 -0800 - rev 218593
Bug 1098126 - MoofParser forced moof read; r=mattwoodrow
73c2707c4f6028dc5cf63f8862ad5f4329d31f35: Bug 1098126 - Use MoofParser to read fragmented MP4 data; r=mattwoodrow
Anthony Jones <ajones@mozilla.com> - Fri, 05 Dec 2014 14:17:23 -0800 - rev 218592
Bug 1098126 - Use MoofParser to read fragmented MP4 data; r=mattwoodrow
4d6a52c5fb75f84a8f52245cb53e2d17e9a0a7cb: Bug 1098126 - Remove duplication in MP4 demuxer seek; r=mattwoodrow
Anthony Jones <ajones@mozilla.com> - Fri, 05 Dec 2014 14:17:22 -0800 - rev 218591
Bug 1098126 - Remove duplication in MP4 demuxer seek; r=mattwoodrow
(0) -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 tip