summaryrefslogtreecommitdiffstats
path: root/cc
Commit message (Collapse)AuthorAgeFilesLines
* cc: Recycle tilings to make SetLiveTilesRect cheaperenne@chromium.org2013-07-0111-31/+429
| | | | | | | | | | | | | | | | | | | | | | | | | | | Rather than recreating new tilings every frame, use the tilings and old tiles from the recycle tree in order to prevent excessive amounts of tile allocations (1 per tile) during the first update. In order to reuse a tiling from two frames ago, the tilings must be resized, invalidated, have their pile updated, and also any missing tiles that can now be rastered need to be filled in. This brings back a bunch of functions that were removed in https://codereview.chromium.org/14600003/ to enable tilings to be resized and updated at times other than SetLiveTilesRect. This patch makes SetLiveTilesRect on the pending tree go from 2.2ms to 0.16ms on average on an N4 for the heavy page linked in the bug with ~700 tiles. On a page with about 300 tiles and fewer commits, this patch is a slight win but is mostly in the noise. This is an updated reland of https://codereview.chromium.org/16888016, which got rolled out previously. R=danakj@chromium.org BUG=248954 Review URL: https://chromiumcodereview.appspot.com/18326013 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209525 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Replace LayerTreeHostCommon::CallFunctionForSubtree() with Callback versionegraether@chromium.org2013-07-014-80/+29
| | | | | | | | | | | This change switches all uses of LayerTreeHostCommon::CallFunctionForSubtree() to the base::Callback version and removes the functor implementation. BUG= Review URL: https://chromiumcodereview.appspot.com/18089014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209478 0039d316-1c4b-4281-b951-d872f2087c98
* Use fuzzy pixel comparer on Windows for GLRendererPixelTest.AntiAliasing.skaslev@chromium.org2013-07-011-10/+2
| | | | | | | | R=enne@chromium.org Review URL: https://codereview.chromium.org/18328012 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209467 0039d316-1c4b-4281-b951-d872f2087c98
* Fix clang build + marked test revisions.skaslev@chromium.org2013-07-011-6/+12
| | | | | | | | TBR=jamesr,zmo Review URL: https://codereview.chromium.org/18083020 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209456 0039d316-1c4b-4281-b951-d872f2087c98
* Optimized antialiasing fragment shaders by computing quad edge distance per ↵skaslev@chromium.org2013-07-0110-213/+433
| | | | | | | | | | | | | | | | | | | | | vertex instead of per pixel. In particular for tile AA fragment shaders: before: vs: 26 cycles fs: 12 cycles after: vs: 47 cycles fs: 6 cycles which is a good performance win given there are only 4 vertices but ~64k pixels per tile. BUG=233101 R=epenner@chromium.org, reveman@chromium.org Review URL: https://codereview.chromium.org/17770004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209445 0039d316-1c4b-4281-b951-d872f2087c98
* cc: request redraw when debug state is changedcaseq@chromium.org2013-07-011-1/+1
| | | | | | | | | | | | We exclude HUD from normal damage tracking logic, so request full redraw when HUD settings change. BUG=252664 R=enne@chromium.org Review URL: https://codereview.chromium.org/17321004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209417 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Use base::Callback approach instead of using WebKit interface.tfarina@chromium.org2013-07-014-26/+19
| | | | | | | | | | | | | | Instead of having cc/ talk directly to the WebKit::WebLayerScrollClient, we define a callback in cc/ that it can talk to. In the compositor_bindings, we pass this callback down to cc, so it can call us back later. BUG=None TEST=cc_unittests, webkit_compositor_bindings_unittests R=jamesr@chromium.org Review URL: https://chromiumcodereview.appspot.com/18139004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209386 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Stop assuming LayerImpl::WillDraw returns trueboliu@chromium.org2013-06-292-4/+6
| | | | | | | | BUG= Review URL: https://chromiumcodereview.appspot.com/18176010 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209341 0039d316-1c4b-4281-b951-d872f2087c98
* CC: Force uploads even if animations are present.epenner@chromium.org2013-06-296-24/+1
| | | | | | | | | | | | This is the nasty 'middle ground' case. If we don't force uploads we will get smoother animations, but it will be at the expense of very bad main-thread performance. BUG=251990 Review URL: https://chromiumcodereview.appspot.com/17157014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209296 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Pipe visibility and memory allocation to DelegatingRendererpiman@chromium.org2013-06-2927-209/+343
| | | | | | | | | | | | | | | | | | The GPU process makes memory allocation decisions based on the visibility of surface. In the DelegatingRenderer case we don't use the surface directly, but we still have it - just like composite-to-mailbox. So we can use setVisibilityCHROMIUM to signal the GPU process, and listen to the memory allocation callback. This CL refactors a bunch of the logic that was hardcoded in GLRenderer into OutputSurface which is shared. That also opens the door to implementing a memory manager for the software renderer. BUG=123444 Review URL: https://chromiumcodereview.appspot.com/18121005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209287 0039d316-1c4b-4281-b951-d872f2087c98
* gpu: Add more context labelling.epenner@chromium.org2013-06-293-2/+34
| | | | | | | | | | | | | | Make sure names are unique in different renderers, and add one for offscreen renderer context (used for filters). The offscreen context is created lazily, so the label is also applied lazily. BUG=242999,233101 Review URL: https://chromiumcodereview.appspot.com/17350003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209284 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Add debug rects for touch and wheel event handler regions and slow ↵egraether@chromium.org2013-06-298-1/+228
| | | | | | | | | | | | | | | scrolling areas This change adds new debug rect types to show regions with touch or wheel event handlers as well as areas that scroll slowly. Each rect type also displays a label in the top left corner to make them easier to identify. BUG=253552 Review URL: https://chromiumcodereview.appspot.com/17599002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209252 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Remove RenderingStats passed to Layer::Update()egraether@chromium.org2013-06-2934-159/+99
| | | | | | | | | | | depends on: https://codereview.chromium.org/13245007/ and: https://codereview.chromium.org/13132005/ BUG=181319 Review URL: https://chromiumcodereview.appspot.com/13258003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209251 0039d316-1c4b-4281-b951-d872f2087c98
* Premultiply alpha in shader for non-premultiplied textures.alokp@chromium.org2013-06-296-45/+65
| | | | | | | | | | | | | | | | | | Instead of using different blending equations we can perform the pre-multiplication in the shader itself. I am working on improving the performance of compositing canvas elements with opaque background. This requires blending the canvas texture with the background color in the shader. The implementation of this blending equation becomes much easier if we can assume that textures are always premultiplied. BUG=236982 TEST=compositing/webgl/webgl-nonpremultiplied-blend.html Review URL: https://chromiumcodereview.appspot.com/16943021 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209246 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Fix incorrect LTHI shutdown logicenne@chromium.org2013-06-294-31/+4
| | | | | | | | | | | | | | | | | | | | | | | | | It's possible that if there has never been an activation (implying no root layer on the active tree), then the trees will not be cleared during LayerTreeHostImpl's destructor. Given the comment in the destructor about animation, this seems to be problematic enough. Issue 251722 is about a case where the tile manager tries to activate the tree during its destruction but because the trees haven't been cleared yet, it thinks this is a supported operation. This is fixed by always destroying the trees in the destructor. Related cleanup is to remove the CleanupRenderSurfaces() call which isn't needed there. When you destroy a tree, you destroy its render surfaces. And, it looked like none of the other CleanupRenderSurfaces() calls were really needed either except for losing the output surface, so one call got moved there and all the rest of the functions got removed. R=danakj@chromium.org BUG=251722 Review URL: https://chromiumcodereview.appspot.com/16831016 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209242 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Pre multiplying alpha value in YUVA Fragment Shadervigneshv@chromium.org2013-06-282-5/+33
| | | | | | | | | | | | | | The alpha value from the video's alpha plane should be pre-multiplied in the YUVA Fragment Shader. This fixes videos not rendering properly when the alpha plane is completely transparent (i.e.) alpha value is zero. Also, adding a pixel test to test the correct behavior. BUG=147355 Review URL: https://chromiumcodereview.appspot.com/18122007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209234 0039d316-1c4b-4281-b951-d872f2087c98
* cc: PostTask the deleteTexture call to the impl thread.danakj@chromium.org2013-06-283-8/+34
| | | | | | | | | | | | | | | | | | After an async texture copy is returned to the embedder on the main thread, it will call the TextureMailbox's ReleaseCallback from the main thread, which was bound to a function that will delete the texture id since it is no longer needed. This function needs to run on the impl thread though, since it makes use of the compositor context. So, add a PostTask over to the impl thread in the call stack to get the deleteTexture on the right thread. R=piman@chromium.org BUG=254451 Review URL: https://codereview.chromium.org/17918003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209218 0039d316-1c4b-4281-b951-d872f2087c98
* Call EnsureBackbuffer before Reshape.aelias@chromium.org2013-06-284-6/+7
| | | | | | | | | | | | | | r204650 switched EnsureBackbuffer to occur after Reshape as an unintended side effect of other refactoring. Arguably this isn't correct and has been causing checkfailures on the server side, so reorder them back. NOTRY=true BUG=254687 Review URL: https://chromiumcodereview.appspot.com/17984002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209202 0039d316-1c4b-4281-b951-d872f2087c98
* Delete VertexShaderPosTexTransformFlip.alokp@chromium.org2013-06-285-73/+19
| | | | | | | | | | The flip can be easily done using the texture transform in VertexShaderPosTexTransform. There is no need for a different shader just to do the flip. Review URL: https://chromiumcodereview.appspot.com/18084004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209066 0039d316-1c4b-4281-b951-d872f2087c98
* Remove execute bit on image file.eroman@chromium.org2013-06-281-0/+0
| | | | | | | | BUG=147355 Review URL: https://codereview.chromium.org/18166004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209053 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Replace FIXMEs with TODO(danakj)enne@chromium.org2013-06-2819-69/+68
| | | | | | | | | R=danakj@chromium.org BUG=none Review URL: https://chromiumcodereview.appspot.com/18050012 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209050 0039d316-1c4b-4281-b951-d872f2087c98
* Add test file for YUVA fragment shader.fgalligan@chromium.org2013-06-271-0/+0
| | | | | | | | | BUG=147355 TBR=vignesh,enne Review URL: https://codereview.chromium.org/18160003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209028 0039d316-1c4b-4281-b951-d872f2087c98
* Use a direct include of time headers in cc/.avi@chromium.org2013-06-2733-33/+33
| | | | | | | | | | BUG=254986 TEST=none TBR=ben@chromium.org Review URL: https://chromiumcodereview.appspot.com/18120002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@209027 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Remove RenderingStats passed to ContentLayerUpdater during Layer::Update()egraether@chromium.org2013-06-2718-60/+33
| | | | | | | | | | depends on: https://codereview.chromium.org/13265003/ BUG=181319 Review URL: https://chromiumcodereview.appspot.com/13245007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208980 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Switch Picture to use RenderingStatsInstrumentationegraether@chromium.org2013-06-2711-48/+63
| | | | | | | | | | depends on: https://codereview.chromium.org/13266002/ BUG=181319 Review URL: https://chromiumcodereview.appspot.com/13132005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208976 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Don't forbid TODO (no parens) in presubmitenne@chromium.org2013-06-271-2/+2
| | | | | | | | | | | | This makes it impossible to refer to a TODO in a comment. It seems like cpplint.py properly warns about this, so this check is redundant anyway. R=danakj@chromium.org BUG=none Review URL: https://chromiumcodereview.appspot.com/17968002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208908 0039d316-1c4b-4281-b951-d872f2087c98
* Update Linux to use scoped_refptr<T>::get() rather than implicit "operator T*"rsleevi@chromium.org2013-06-278-18/+16
| | | | | | | | | | | | Fixes a number of cases either not caught by the automated tool or that have recently regressed. BUG=206189 TBR=darin Review URL: https://chromiumcodereview.appspot.com/17948002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208882 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Reset tile priority before deleting tiling.reveman@chromium.org2013-06-275-0/+14
| | | | | | | | BUG=251938 Review URL: https://chromiumcodereview.appspot.com/17891002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208880 0039d316-1c4b-4281-b951-d872f2087c98
* re-land cc: Switched ContentLayerUpdaters to use RenderingStatsInstrumentationernstm@chromium.org2013-06-277-28/+48
| | | | | | | | | | | - fixed root cause of failing interactive_ui_test (dangling pointer in ContentLayerUpdater) BUG=181319, 223693 Review URL: https://chromiumcodereview.appspot.com/17762002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208828 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Create a AnimationDelegate interface and an adapter class in ↵tfarina@chromium.org2013-06-2610-36/+53
| | | | | | | | | | | | | | | | | compositor_bindings. Instead of having cc/ talk directly to the WebKit::WebAnimationDelegate, we create an abstract interface in cc/ that it can talk to. In the compositor_bindings, we implement this class, and forward the function calls made into it over to the WebKit implementation. BUG=None TEST=cc_unittests, webkit_compositor_bindings_unittests R=jamesr@chromium.org Review URL: https://codereview.chromium.org/17755003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208800 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Add FIXME presubmit warningenne@chromium.org2013-06-261-0/+17
| | | | | | | | | R=danakj@chromium.org BUG=none Review URL: https://chromiumcodereview.appspot.com/17904003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208772 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Use multiple resources to switch LCD statusvmpstr@chromium.org2013-06-2622-366/+290
| | | | | | | | | | | | | Currently we destroy and recreate tiles with text if the LCD status changes. With tile manager now supporting multiple versions of tiles, we can use this to update a tile version with new LCD status in place. BUG=249048 Review URL: https://chromiumcodereview.appspot.com/17167003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208769 0039d316-1c4b-4281-b951-d872f2087c98
* Optimize shorthand CSS filter application. When the filter chain can be ↵senorblanco@chromium.org2013-06-262-7/+17
| | | | | | | | | | | | expressed as a single color matrix, apply it directly in the compositor and skip the intermediate FBO. This is similar to an optimization that has already been applied for the SkImageFilter DAG path. R=epenner,jamesr BUG=233101 Review URL: https://chromiumcodereview.appspot.com/17693002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208751 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Fix infinite BeginFrame that Scheduler causes.dongseong.hwang@intel.com2013-06-265-7/+13
| | | | | | | | | | | | | | | | | | | REGRESSION(r206955): Remove the chance to disable SetNeedsBeginFrame. r206955 made SetNeedsBeginFrame be called proactively, but the proactive condition mostly was true. This issue makes proactive_begin_frame_wanted condition more restrict. There are two changes. 1. Do not be proactive when invisible. 2. Do not be proactive when throttling frame production. This patch is mostly based on what Brian Anderson consults. BUG=253543 Review URL: https://chromiumcodereview.appspot.com/17587014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208662 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Use a vector for completed tasksvmpstr@chromium.org2013-06-2610-45/+61
| | | | | | | | | | | Instead of using a deque, this patch uses a vector for completed tasks. BUG=253971 Review URL: https://chromiumcodereview.appspot.com/17587019 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208624 0039d316-1c4b-4281-b951-d872f2087c98
* Remove GpuMemoryAllocationForRenderer::enforce_but_do_not_keep_as_policypiman@chromium.org2013-06-2610-26/+13
| | | | | | | | | | It's always false, and we're not planning on using the feature at this point. BUG=None Review URL: https://chromiumcodereview.appspot.com/16972012 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208614 0039d316-1c4b-4281-b951-d872f2087c98
* Send hardware video frames with mailboxes.danakj@chromium.org2013-06-255-107/+75
| | | | | | | | | | | | | | | | | | | | Change media::VideoFrame from holding a raw texture id to instead hold a gpu::Mailbox. The hardware video decoders produce a mailbox on their context from the texture id before putting it in the VideoFrame, and consume the mailbox again when they receive the texture id back. In the compositor, we hold onto the VideoFrame::MailboxHolder as long as the mailboxed texture is in use in the compositor (or its parent). This keeps the video decoders from writing to the texture id, as they register a callback in the VideoFrame to be called when it is destroyed. BUG=179729 Review URL: https://chromiumcodereview.appspot.com/14199002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208568 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Use priority_queue for worker_pool's ready to run tasksvmpstr@chromium.org2013-06-251-15/+21
| | | | | | | | | | This changes std::map to be a priority queue with an underlying vector. BUG=253968 Review URL: https://chromiumcodereview.appspot.com/17605004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208549 0039d316-1c4b-4281-b951-d872f2087c98
* Move implementation of WebFilterOperations into ccajuma@chromium.org2013-06-2537-282/+829
| | | | | | | | | | | | | | | | This moves the implementation of WebFilterOperations into cc, and (behind an #ifdef) defines a WebFilterOperationsImpl class that implements the WebFilterOperations interface by wrapping a cc::FilterOperations. With this change, cc and ui/compositor no longer need to include WebFilterOperations.h. BUG=181613 Review URL: https://chromiumcodereview.appspot.com/16968002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208531 0039d316-1c4b-4281-b951-d872f2087c98
* CC: Force uploads in reverse.epenner@chromium.org2013-06-251-3/+11
| | | | | | | | | For performance downstream, we would rather wait once than multiple times. Review URL: https://chromiumcodereview.appspot.com/17639002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208420 0039d316-1c4b-4281-b951-d872f2087c98
* Add signalQuery, to allow for a callback when the result of a query is ↵hubbe@chromium.org2013-06-252-2/+22
| | | | | | | | | | available. BUG=249925 Review URL: https://chromiumcodereview.appspot.com/16983007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208376 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Fix incorrect LTHI shutdown logicboliu@chromium.org2013-06-241-9/+7
| | | | | | | | | | | | | | | | | | | | | Landing this fix for enne@. Original description by enne@: It's possible that if there has never been an activation (implying no root layer on the active tree), then the trees will not be cleared during LayerTreeHostImpl's destructor. Given the comment in the destructor about animation, this seems to be problematic enough. Issue 251722 is about a case where the tile manager tries to activate the tree during its destruction but because the trees haven't been cleared yet, it thinks this is a supported operation. This is fixed by always destroying the trees in the destructor. BUG=251722 NOTRY=true Review URL: https://chromiumcodereview.appspot.com/16983013 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208251 0039d316-1c4b-4281-b951-d872f2087c98
* Add a hide_layer_and_subtree() flag to cc::Layerdanakj@chromium.org2013-06-249-3/+378
| | | | | | | | | | | | | | | | | | | | | | | | This flag does what it says. We can use this flag to simplify ui::Layer as it was recursively calling SetIsDrawable() instead, and now can just SetHideLayerAndSubtree() on a single layer. This flag is needed to allow readbacks of layers that are have hidden visibility (possibly due to an ancestor). Tests: LayerTest.CheckPropertyChangeCausesCorrectBehavior LayerImplTest.VerifyLayerChangesAreTrackedProperly LayerTreeHostCommonTest.SubtreeHidden_SingleLayer LayerTreeHostCommonTest.SubtreeHidden_SingleLayerImpl LayerTreeHostCommonTest.SubtreeHidden_TwoLayers LayerTreeHostCommonTest.SubtreeHidden_TwoLayersImpl R=enne, piman BUG=242572 Review URL: https://chromiumcodereview.appspot.com/16896017 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208241 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Make tiles that are not required by trees NEVER_BINvmpstr@chromium.org2013-06-222-7/+45
| | | | | | | | | | | | | Currently if we have allow anything mode, we never have a tile in the NEVER_BIN, even if both active and pending trees don't own the tile anymore (ie it's being kept alive by the raster callback). This cleans it up. BUG=251938 Review URL: https://chromiumcodereview.appspot.com/16959022 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@208047 0039d316-1c4b-4281-b951-d872f2087c98
* Add base namespace to more values in sync and elsewhere.brettw@chromium.org2013-06-213-3/+3
| | | | | | | | | | This makes sync and net compile with no "using *Value". BUG= Review URL: https://codereview.chromium.org/17034006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@207907 0039d316-1c4b-4281-b951-d872f2087c98
* Re-revert "Re-land r207660 - "Upgrading Mesa to 9.0.3."""dpranke@chromium.org2013-06-214-0/+0
| | | | | | | | | | | This reverts r207892 - build failure on Linux x64. TBR=kbr@chromium.org BUG=238755 Review URL: https://codereview.chromium.org/16915005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@207896 0039d316-1c4b-4281-b951-d872f2087c98
* Re-land r207660 - "Upgrading Mesa to 9.0.3.""dpranke@chromium.org2013-06-214-0/+0
| | | | | | | | | | | Try again w/ gyp fix. TBR=kbr@chromium.org, mymax@amazon.com BUG=238755 Review URL: https://codereview.chromium.org/17115014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@207892 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Move task graph construction to RasterWorkerPool classes.reveman@chromium.org2013-06-2114-383/+550
| | | | | | | | | | | | | | | | This removes the need to construct a temporary graph by RasterWorkerPool classes. It reduces the number of tasks created and eliminates the need for a dependencies vector as part of the internal::WorkerPoolTask class. This is also a prerequisite to implementing a "ready to activate" signal as that requires as task graph structure that is not a tree. BUG=247677 Review URL: https://chromiumcodereview.appspot.com/17244003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@207864 0039d316-1c4b-4281-b951-d872f2087c98
* Change WGC3DInProcessCBImpl factories to return a base class pointer.sievers@chromium.org2013-06-212-6/+4
| | | | | | | | TBR=sky@chromium.org for ui/aura/bench/bench_main.cc Review URL: https://chromiumcodereview.appspot.com/17447007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@207782 0039d316-1c4b-4281-b951-d872f2087c98
* cc: Produce a single mailbox per video texture in software decode.danakj@chromium.org2013-06-212-49/+36
| | | | | | | | | | | | | | | | | | | | | In the software video decode path, instead of generating a new mailbox for every frame, include the mailbox in the recycled data and reuse it along with the resource/texture. Also, we were reusing an old sync point which was problematic. Instead, don't create a sync point at all, as the ResourceProvider will do so if and when it needs to, to send it to the parent. This will resolve the bug 252053. And, while we're here, drop the wasted extra consume when a resource is recycled back to the VideoResourceUpdater. R=piman BUG=252053 Review URL: https://chromiumcodereview.appspot.com/17094014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@207692 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-02-07 15:39:53 +0000