Commit Graph

1752 Commits

Author SHA1 Message Date
Jeff Gilbert
891a5c7f86 Bug 1144906 - Add accel E10S backend for WebGL compositing. - r=jrmuizel,mattwoodrow,nical,sotaro 2015-06-04 14:21:32 -07:00
David Anderson
b2eb29a9c5 Use widgets for APZ checks, rather than gfxPrefs. (bug 1162064, r=kats,mstange) 2015-06-04 16:51:10 -04:00
Amanda Sambath
376d6d03c3 Bug 1158120 - Replace nsIntSize by mozilla::gfx::IntSize in gfx/layers. r=nical 2015-06-03 13:45:24 +02:00
Sotaro Ikeda
ab0826ea13 Bug 1169200 - Send OpRemoveTextureAsync before actor destroyed r=nical 2015-06-02 07:31:05 -07:00
Amanda Sambath
e5ab8210f6 Bug 1158120 - Edit include and comments that contained gfxIntSize and nsIntSize. r=nical 2015-05-29 17:01:46 +02:00
Wes Kocher
5ee1a8327e Backed out changeset 29bd04fc57f0 (bug 1144906) for b2g m12 orange
--HG--
extra : amend_source : 79faafb4230715cf6af79b452d121d1e7b58e665
2015-05-28 17:08:53 -07:00
Jeff Gilbert
ec631286de Bug 1144906 - Fast WebGL compositing on E10S. - r=jrmuizel,mattwoodrow,nical 2015-05-28 14:51:39 -07:00
Nicolas Silva
82d4c31667 Bug 1150549 - Simplify TiledContentHost. r=jrmuizel 2015-05-22 13:38:13 +02:00
Jeremy Chen
dedec71f0f Bug 1155495 - Part 2: Remove unnecessary class inheritance and casting. r=nical, r=sotaro 2015-05-20 01:43:00 -04:00
Ryan VanderMeulen
6b55a23433 Backed out changesets 4b6aa5c0a1bf and fdf38a41d92b (bug 1150549) for Mulet crashes.
CLOSED TREE
2015-05-25 14:05:05 -04:00
Nicolas Silva
17c3851bc7 Bug 1150549 - Simplify TiledContentHost. r=jrmuizel 2015-05-22 13:38:13 +02:00
Timothy Nikkel
f3507bdc3b Bug 1160642. Add reftest-async-zoom to apply an async zoom before taking snapshot. r=dbaron 2015-05-24 16:48:26 -04:00
Bas Schouten
09e0edbcc8 Bug 1158575: Support using GDI rendering for opaque surfaces when using cross-process layers. r=jrmuizel 2015-05-22 17:54:24 +02:00
Ryan VanderMeulen
4ba8667b9c Backed out 2 changesets (bug 1150549) for OSX asserts.
Backed out changeset ef2faa9a0845 (bug 1150549)
Backed out changeset b35c88257a9c (bug 1150549)

CLOSED TREE
2015-05-22 10:28:52 -04:00
Kartikaya Gupta
b54dc47f24 Back out 2 csets for bug 1163832 for emulator crashtest failures. r=me
CLOSED TREE
2015-05-22 07:51:55 -04:00
Nicolas Silva
f15ae7d768 Bug 1150549 - Simplify TiledContentHost. r=jrmuizel 2015-05-22 13:38:13 +02:00
Kartikaya Gupta
86ed888daa Bug 1163832 - Add an API to flush pending APZ repaint requests and dispatch a notification upon completion. r=botond 2015-05-22 06:42:08 -04:00
Nathan Froyd
db188ea282 Bug 1160485 - remove implicit conversion from RefPtr<T> to TemporaryRef<T>; r=ehsan
Having this implicit conversion means that we can silently do extra
refcounting when it's completely unnecessary.  It's also an obstacle to
making RefPtr more nsRefPtr-like, so let's get rid of it.
2015-05-01 09:14:16 -04:00
Nathan Froyd
c24f5f4e7c Bug 1116905 - part 3 - remove dependence on implicit conversion from T* to TemporaryRef<T>, gfx changes; r=jrmuizel 2015-04-30 15:20:30 -04:00
Kan-Ru Chen
3e9e269e92 Bug 1154231 - Part 1. Use LRU to manage remote layers. r=mattwoodrow 2015-05-22 07:08:20 +08:00
Ethan Lin
aac29696f5 Bug 1155498 - Part 2 - Use new FenceHandle to handle the fence in TextureHost. r=sotaro
--HG--
extra : rebase_source : 5e619c27d9157dcc379a7a3f721aa44aa7b42370
2015-05-13 00:36:00 +02:00
Ethan Lin
0e78b95a1c Bug 1155498 - Use FdObj to replace FenceHandle's android fence. r=sotaro
--HG--
extra : rebase_source : df49b5e482064e30338a19748371bb70550cdd13
2015-05-17 07:21:00 +02:00
Phil Ringnalda
f7950798e0 Back out 530555a2d6d4 (bug 1162064) for debug b2g emulator assertions and failure to thrive
CLOSED TREE
2015-05-17 22:42:05 -07:00
David Anderson
4b20fa6b94 Use widgets for APZ checks, rather than gfxPrefs. (bug 1162064, r=kats,mstange) 2015-05-17 23:11:44 -04:00
David Anderson
cadd78802e Backout changeset e5d045d49b40 for bustage on an a=CLOSED TREE 2015-05-17 21:46:30 -04:00
David Anderson
502eb5d3ca Use widgets for APZ checks, rather than gfxPrefs. (bug 1162064, r=kats,mstange) 2015-05-17 20:44:43 -04:00
Phil Ringnalda
a5df436c2e Backed out 2 changesets (bug 1154231) for M-oth crashes
CLOSED TREE

Backed out changeset 5e8145eff457 (bug 1154231)
Backed out changeset 951fb8ed6f25 (bug 1154231)
2015-05-15 20:29:15 -07:00
Kan-Ru Chen
10fec84bca Bug 1154231 - Part 2. Use LRU to manage remote layers. r=mattwoodrow 2015-05-16 08:43:39 +08:00
Sotaro Ikeda
59b114892e Bug 1164513 - Add nullptr check r=milan 2015-05-14 12:09:18 -07:00
David Anderson
8a9daf3775 Enable APZ for all Desktop windows if E10s is enabled. (bug 1163115, r=kats) 2015-05-12 01:26:01 -04:00
Paul Rouget
4ca5b03bc0 Bug 1161407 - Only check for IsMultiProcessWindow on non-B2G & non-Android platforms. r=kats 2015-05-11 14:59:07 -04:00
Eric Rahm
67b91ff53a Bug 1162748 - Remove instances of #ifdef PR_LOGGING in graphics. r=froydnj
PR_LOGGING is now always defined, we can remove #ifdefs checking for it.
2015-05-08 14:37:01 -07:00
Sotaro Ikeda
e2d6f1ea99 Bug 1091037 - Raise ImageBridge thread priority r=gsvelto,nical 2015-05-07 06:53:47 -07:00
Thibaud Backenstrass
f8d4cbad76 Bug 1158122 - Remove the remaining occurences of nsIntRect in gfx/layers. r=nical 2015-05-07 11:07:53 +02:00
Phil Ringnalda
24bb7f3b9f Back out dc2fed1aa0af (bug 1144906) for ASan, Android debug, and OS X 10.10 mochitest-gl bustage
CLOSED TREE
2015-05-06 19:55:40 -07:00
Jeff Gilbert
b0077b976e Bug 1144906 - Fast WebGL on E10S. - r=mattwoodrow,nical 2015-05-06 16:35:29 -07:00
Kartikaya Gupta
85f09c0ea6 Bug 1162179 - Don't create APZCTreeManagers for windows which aren't multiprocess. r=dvander 2015-05-06 19:15:10 -04:00
Wes Kocher
b6f7abf05b Backed out changeset 93af6e2a390d (bug 1144906) for build bustage 2015-05-06 16:01:47 -07:00
Jeff Gilbert
587246e6fd Bug 1144906 - Support fast WebGL on E10S. - r=mattwoodrow,nical 2015-05-06 13:48:28 -07:00
Daniel Holbert
3899eae2cc Bug 1161731: Remove newline characters from the ends of NS_WARNING messages in /gfx and /layout. r=mstange 2015-05-05 13:57:47 -07:00
Nicolas Silva
66177fae1c Bug 1155649 - XFlush at the end of frames when OMTC is enabled on Linux. r=jrmuizel 2015-05-04 12:35:18 +02:00
Sotaro Ikeda
fb70b24e9f Bug 1156981 - Split CompositorParent's scheduling of composition to CompositorScheduler r=mchang 2015-04-30 11:35:13 -07:00
Timothy Nikkel
499081808b Bug 1157066. When reading back the OMTA transfrom of a layer don't include transforms from async panning or zooming. r=kats,birtles
Containerless scrolling means that the pan zoom controller applies it's transforms (to compensate for differences between the state of layout the last time we painted and the current state as composited to the screen) to the layers that are scrolled instead of the container layer that contains the layers that scroll.

When running test_animations_omta.html there is a zoom of 1.306122 applied, and the page is scrolled down to 67 screen pixels (before the test starts, not sure why exactly). Gecko scrolls as close to 67 screen pixels as it can: 67/1.306122 = 51.29689 css pixels, which is 3077.813 appunits. Gecko scrolls to 3078 app units. When AsyncCompositionManager::TransformScrollableLayer runs we calculate the scroll position of gecko and the current scroll position that the pan zoom controller is using. Since there are no async pan or zoom operations taking place these should match. However when the gecko scroll position is calculated we get 3078/60*1.306122 = 67.0040586. So it applies a transform of 0.0040586 to the container layer for the transform that test_animations_omta.html is animating off main thread. When test_animations_omta.html reads the transform of this layer it fails because it's expecting 0 and 0.0040586 is outside of it's epsilon for considering it to be close enough.
2015-04-24 19:05:52 -05:00
Wes Kocher
936a2fdf3f Backed out changeset 517b9b831685 (bug 1157066) for build bustage 2015-04-24 14:56:55 -07:00
Timothy Nikkel
e488294393 Bug 1157066. When reading back the OMTA transfrom of a layer don't include transforms from async panning or zooming. r=kats,birtles
Containerless scrolling means that the pan zoom controller applies it's transforms (to compensate for differences between the state of layout the last time we painted and the current state as composited to the screen) to the layers that are scrolled instead of the container layer that contains the layers that scroll.

When running test_animations_omta.html there is a zoom of 1.306122 applied, and the page is scrolled down to 67 screen pixels (before the test starts, not sure why exactly). Gecko scrolls as close to 67 screen pixels as it can: 67/1.306122 = 51.29689 css pixels, which is 3077.813 appunits. Gecko scrolls to 3078 app units. When AsyncCompositionManager::TransformScrollableLayer runs we calculate the scroll position of gecko and the current scroll position that the pan zoom controller is using. Since there are no async pan or zoom operations taking place these should match. However when the gecko scroll position is calculated we get 3078/60*1.306122 = 67.0040586. So it applies a transform of 0.0040586 to the container layer for the transform that test_animations_omta.html is animating off main thread. When test_animations_omta.html reads the transform of this layer it fails because it's expecting 0 and 0.0040586 is outside of it's epsilon for considering it to be close enough.
2015-04-24 16:41:56 -05:00
L. David Baron
20dd52e0ad Bug 1157455 patch 2 - Send angles (in CSS transform functions) to the compositor thread with their units rather than sending all such angles in radians. r=birtles
This avoids accumulating floating point error from conversion, so that
when we switch to doubles at the start of
nsCSSValue::GetAngleValueInRadians we're using the original unit, rather
than a different one that will round differently.
2015-04-23 09:18:12 -07:00
L. David Baron
cc896f3f84 Bug 1156456 patch 3 - Send transform origin and perspective origin to layer in device pixels rather than CSS pixels. r=birtles 2015-04-22 19:37:04 -07:00
Andrea Marchesini
ade2831035 Bug 1156632 - Remove unused forward class declarations - patch 6 - the rest of the tree, r=ehsan 2015-04-22 08:29:24 +02:00
Nicolas Silva
cfff5e52c5 Bug 1155621 - Make nsIntRect and nsIntPoint typedefs of mozilla::gfx::IntRect and mozilla::gfx::IntPoint. r=Bas 2015-04-21 17:04:57 +02:00
Kartikaya Gupta
96eba4649d Back out cset 1a68731ea592 (bug 1145295) because the crash it was attempting to shed more light on seems to have gone away. r=me 2015-04-21 10:19:16 -04:00