summaryrefslogtreecommitdiffstats
path: root/Source/WebCore/platform/graphics
Commit message (Collapse)AuthorAgeFilesLines
...
| * Make the Gpu Upload path fasterTeng-Hui Zhu2011-09-291-8/+18
| | | | | | | | | | | | | | | | | | Basically, move unnecessary function calls out from the loop. And move the clear related code into DEBUG only. This is saving us some CPU overhead. bug:5347539 Change-Id: I237162fd6a502e93ea32c9856f907c5a9ddb2e78
* | Merge "Use state machine to track tile state"Chris Craik2011-09-297-66/+187
|\ \
| * | Use state machine to track tile stateChris Craik2011-09-297-66/+187
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5369978 Manage invalidations, painting, transferring, and swapping with a state machine. notes: * readyFor shouldn't be needed if all of the events that would cause a readyfor fail mark the tile dirty (such as stealing, scale change) * changing the scale of a page should discard all textures * m_dirty should be more gracefully worked into the state machine * a tile may drop dirtiness notifications if it's already painting, since upon completion m_dirty will be cleared - (issues for layers only) Change-Id: I5909fb5d208da2fb276e223c56bf143741a9a24c
* | Add debug informations to keep better track of layers allocationNicolas Roard2011-09-295-7/+61
| | | | | | | | | | | | | | | | | | Fix coherency issue in ClassTracker (it's accessed from two different threads and we were not locking the datastructure) bug:5366807 Change-Id: I7dbd4417a49f4b3af3e09d614cec6b4568b09f7a
* | Fix a memory leak with LayerAndroidNicolas Roard2011-09-297-24/+78
|/ | | | | | | (we had a cycle between LayerAndroid and PaintedSurface) bug:5366807 Change-Id: I1b7341984b3cc6fd998b9ab54205bb963eb9223d
* Dynamically set the upper limit of layer tiles' texture numberTeng-Hui Zhu2011-09-272-21/+9
| | | | | | | | | | Instead of using a static number, now the upper limit of the layer tiles is also dependent on the viewport size. At the same time, get rid of the old code only useful without tiled layer. bug:5347539 Change-Id: Ie720449ea15043d42a56f4c8af57820fa7f6e93c
* Increase the texture number for layersTeng-Hui Zhu2011-09-271-1/+1
| | | | | | | | This will break the Gpu upload code path on phones which is turned off by default, but it will work on the tablet. bug:5347539 Change-Id: Idc9889a8681035addd85843191322039a69110e5
* Merge "Egl sync code path should be only activated when we are in GpuUpload ↵Teng-Hui Zhu2011-09-261-1/+2
|\ | | | | | | mode"
| * Egl sync code path should be only activated when we are in GpuUpload modeTeng-Hui Zhu2011-09-261-1/+2
| | | | | | | | | | bug:5347539 Change-Id: I12b8b45792d5aa684295b897be308181a7d44631
* | Merge "Add shared images for layers"Nicolas Roard2011-09-267-43/+322
|\ \ | |/ |/|
| * Add shared images for layersNicolas Roard2011-09-267-43/+322
| | | | | | | | | | | | bug:5242595 bug:5218173 Change-Id: I37d395e85441671312aac3e236cc8276019aa990
* | am fab9df83: Rename Lohit_Hindi to Lohit-DevanagariRussell Brenner2011-09-231-1/+1
|\ \ | | | | | | | | | | | | * commit 'fab9df838471ec0516f4ce0360c6bc2f7248f53f': Rename Lohit_Hindi to Lohit-Devanagari
| * | Rename Lohit_Hindi to Lohit-DevanagariRussell Brenner2011-09-231-1/+1
| | | | | | | | | | | | | | | Bug: 5364436 Change-Id: I0ce921e330ab892d8334b3b8bf4539e6a175a78c
| * | Merge "DO NOT MERGE:Share the display from the UI thread" into ics-factoryromTeng-Hui Zhu2011-09-202-13/+27
| |\ \
| | * | DO NOT MERGE:Share the display from the UI threadTeng-Hui Zhu2011-09-192-13/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently disable the wait b/c b/5332112. For now it should be the same functionality as the original workaround by using glDraw. cherry pick from master c/135811/ bug:5270526 Change-Id: I1973188d8f86ea28e684d14bca794b2a61dac34c
| * | | Cherry-pick "Fix missing tiles" from master - do not mergeChris Craik2011-09-192-16/+28
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5288530 Fix issue where stealing a back texture from a tile (while the tile was ready to swap) wouldn't mark tile as dirty. Now, mark tile dirty whenever texture is stolen from it, and avoid trying to steal busy textures. Cherry-pick of CL: 6aa089bded6a1abbf8d61264588e1e3a32be31b3 Change-Id: Iacbda17ba4b4c6382ceb1c84a6e5f02e609a76a5
* | | Merge "Crash if stack is corrupted"Chris Craik2011-09-231-1/+3
|\ \ \
| * | | Crash if stack is corruptedChris Craik2011-09-231-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5362098 For tracking down very rare, very wrong scale values Change-Id: I00a5dd8d75c47c4c13c903e587b3983ea23496d7
* | | | Merge "Enable the CPU upload path for Browser."Teng-Hui Zhu2011-09-234-106/+154
|\ \ \ \ | |/ / / |/| | |
| * | | Enable the CPU upload path for Browser.Teng-Hui Zhu2011-09-234-106/+154
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The default for WebView should be still be GPU upload. That means Gmail will be default to using GPU upload. bug:5347539 The browser change is in https://android-git.corp.google.com/g/#/c/137606/ Change-Id: Ic7d42331511b24cf8a58f2f5fb64aaffc18cae5f
* | | | Merge "Avoid infinite prepare loop if bad scale provided"Chris Craik2011-09-232-1/+14
|\ \ \ \
| * | | | Avoid infinite prepare loop if bad scale providedChris Craik2011-09-232-1/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5362098 Note: we shouldn't be getting bad scales, now that the scale corruption issue has been reverted. Added logging for these to wrap transfer queue as well to detect fp corruption regression. Change-Id: I5e6d2afc1d483452140fab5390395c9581db86ca
* | | | | Merge "If we have a bad framework inval rect, request entire screen redraw"Chris Craik2011-09-231-6/+13
|\ \ \ \ \
| * | | | | If we have a bad framework inval rect, request entire screen redrawChris Craik2011-09-231-6/+13
| |/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5365139 This problem is occurring because we convert webkit invals to framework ones, but webkit invals don't account for stolen/discarded tiles/textures. This is just a temporary fix, see also bug:5321078 which describes the root of the problem. Change-Id: I6d823315dba8d669171c9c7c9ed3ca20fbed07ae
* | | | | Merge "Prioritize ugly tiles only when scrolling"Chris Craik2011-09-231-6/+13
|\ \ \ \ \
| * | | | | Prioritize ugly tiles only when scrollingChris Craik2011-09-231-6/+13
| |/ / / / | | | | | | | | | | | | | | | | | | | | bug:5365139 Change-Id: Iad181cd1a6f1c9b30f3ac64527c84036bd57ac2c
* | | | | Merge "Discard BaseTileTextures entirely when they delete their GL textures"Chris Craik2011-09-221-7/+1
|\ \ \ \ \
| * | | | | Discard BaseTileTextures entirely when they delete their GL texturesChris Craik2011-09-221-7/+1
| |/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5361337 Fixes issue with garbage being drawn in unpainted tiles, and also allocates tiles very lazily. Change-Id: I9799b63670a80748a4809b61506da96a739d3a65
* | | | | Merge "Fix find-in-page to scroll scrollable layers"Steve Block2011-09-225-7/+81
|\ \ \ \ \
| * | | | | Fix find-in-page to scroll scrollable layersSteve Block2011-09-225-7/+81
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This requires the addition of the following methods ... - Layer::contentIsScrollable() - Layer::localToParent() - ScrollableLayerAndroid::scrollRectIntoView() Bug: 5262656 Change-Id: I2f1cf3342f73890f98a172f1b4e3f440c02dd9f4
* | | | | | Merge "Get rid of unnecessary glFinish"Teng-Hui Zhu2011-09-221-2/+0
|\ \ \ \ \ \
| * | | | | | Get rid of unnecessary glFinishTeng-Hui Zhu2011-09-211-2/+0
| | |_|/ / / | |/| | | | | | | | | | | | | | | | | | | | | | bug:5347539 Change-Id: Iac3888afe1130043e7800e623dd6e061ecf6f40d
* | | | | | Port Webkit r74089 to GraphicsContextAndroid.Ben Murdoch2011-09-221-2/+7
| |_|/ / / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Port the change made to PlatformContextSkia in http://trac.webkit.org/changeset/74089 to GraphicsContext Android. This fixes the rendering of inset box shadows. Bug: 5351383 Change-Id: I7b68c965aec87e610d853a8220468fd8113351e6
* | | | | Merge "Clean up style and add some comments in Layer"Steve Block2011-09-223-59/+36
|\ \ \ \ \ | | |/ / / | |/| | |
| * | | | Clean up style and add some comments in LayerSteve Block2011-09-213-59/+36
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is preparation for https://android-git.corp.google.com/g/#/c/134488/4 Refactoring only, no functional change. Bug: 5262656 Change-Id: I44e362cf35fc5080f7d9fba34183188d3a2a6331
* | | | Merge "Prefetch browser content with tiled page"Chris Craik2011-09-215-0/+74
|\ \ \ \
| * | | | Prefetch browser content with tiled pageChris Craik2011-09-215-0/+74
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5262519 Use the tiled page not used by content rendering to render much fewer tiles for the same content, at an inflated scale. These prefetched tiles are prioritized for painting above all others, so that content is (almost) always visible on the base layer. Change-Id: I598b7925cb68beef632f828df3ae522a0b21e2b4
* | | | | Merge "Prevent rings from being clipped"John Reck2011-09-212-3/+10
|\ \ \ \ \ | |_|/ / / |/| | | |
| * | | | Prevent rings from being clippedJohn Reck2011-09-202-3/+10
| |/ / / | | | | | | | | | | | | | | | | | | | | Bug: 5348386 Change-Id: I6fd9c1d856be257a1a757649aa85e55d4a14a8d1
* | | | Merge "Actually update the texture filter."Teng-Hui Zhu2011-09-201-2/+2
|\ \ \ \ | |_|_|/ |/| | |
| * | | Actually update the texture filter.Teng-Hui Zhu2011-09-201-2/+2
| | | | | | | | | | | | | | | | | | | | bug:5347539 Change-Id: Ic70f9a3d3d780ddf8c60bf51ee94e0b36f56c22e
* | | | Merge "Support drawing the quad in GL_NEAREST mode."Teng-Hui Zhu2011-09-203-7/+10
|\ \ \ \ | |/ / /
| * | | Support drawing the quad in GL_NEAREST mode.Teng-Hui Zhu2011-09-203-7/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This can save the draw call perf up to about 20-40% for each quad. We should make more draw calls using NEAREST mode when it is 1 to 1 mapping. bug:5347539 Change-Id: I1ae206716f2b1352775e0079e25e54f46bc18578
* | | | Merge "A better CPU upload path"Teng-Hui Zhu2011-09-201-7/+13
|\ \ \ \ | |/ / / | | / / | |/ / |/| |
| * | A better CPU upload pathTeng-Hui Zhu2011-09-201-7/+13
| | | | | | | | | | | | | | | | | | | | | | | | Clean up the unnecessary Surface Texture operation in CPU upload code path. bug:5347539 Change-Id: Id8cf1d3e472dce2470bc62a844b95e8d5d2a6371
* | | Merge "Fix missing tiles"Chris Craik2011-09-192-16/+28
|\ \ \
| * | | Fix missing tilesChris Craik2011-09-192-16/+28
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | bug:5335634 Fix issue where stealing a back texture from a tile (while the tile was ready to swap) wouldn't mark tile as dirty. Now, mark tile dirty whenever texture is stolen from it, and avoid trying to steal busy textures. Change-Id: Iacbda17ba4b4c6382ceb1c84a6e5f02e609a76a5
* | | Merge changes Ifb6cc64e,I4d819cc5Steve Block2011-09-195-23/+37
|\ \ \ | | | | | | | | | | | | | | | | | | | | * changes: Simplify logic in ScrollableLayerAndroid::scrollTo() Fixes style and adds some comments to layers code
| * | | Simplify logic in ScrollableLayerAndroid::scrollTo()Steve Block2011-09-191-4/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | No functional change Bug: 5262656 Change-Id: Ifb6cc64e7578625acc1d5e91b8296b9e4cf1d449
| * | | Fixes style and adds some comments to layers codeSteve Block2011-09-194-19/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Also removes FrameCachePermission::AllowNewest. This should have been removed as part of https://android-git.corp.google.com/g/#/c/2712. Bug: 5262656 Change-Id: I4d819cc5087653e4a1180d04ca0b5b20db7e2305
generated by cgit v1.1 at 2024-09-13 18:59:02 +0000