Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-11-08 12:37:37 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
373,935 Commits 615 Branches 98 Tags 5.9 GiB
3c155637a3
Commit Graph

234 Commits

Author SHA1 Message Date
Oleg Romashin
977fc29345 Bug 612335 - Render remote canvas directly into shared surface r=jones.chris.g 2011-06-09 21:17:02 -04:00
Matt Woodrow
875f244cda Bug 649417 - Part 2 - Add a YUV option to SharedImage and use it to share YUV data across processes. r=cjones,joe 2011-04-21 16:38:39 +12:00
Oleg Romashin
d2f1f3108d Bug 610155 - Use SurfaceDescriptor for ImageLayer and CanvasLayer r=jones.chris.g 2011-04-20 14:45:57 -07:00
Daniel Holbert
c98ac8a3c6 Back out 0736b014d4a5 (Bug 610155) for breaking canvas reftests on android 2011-04-21 14:52:22 -07:00
Oleg Romashin
ddd27efe3d Bug 610155 - Use SurfaceDescriptor for ImageLayer and CanvasLayer r=jones.chris.g 2011-04-20 14:45:57 -07:00
Oleg Romashin
863e12aca2 Bug 607417 - Part1: position:fixed frames mark layers r=roc 
--HG--
extra : rebase_source : dbbff8810bafd165a48c34924d98f5c87702ef3f
 2011-04-05 23:00:25 -06:00
Robert O'Callahan
6e81f6e5b2 Bug 622072. Part 1: Remove rect parameter from Updated() and change implementations to defer updates to render time. r=bas 
The rectangle parameter is currently not used --- all callers always pass the full canvas bounds. In the long term,
we probably won't want this parameter since all implementations should be doing accelerated drawing direct to buffers
with no intermediate copies, hence there will be no need to optimize the size of those copies. Plus, performance-sensitive
testcases tend to paint most or all of the canvas on every frame anyway.
 2011-03-28 12:59:46 +13:00
Ehsan Akhgari
acac18d785 Back out bug 622072 because it caused bug 645987 2011-03-29 14:31:53 -04:00
Robert O'Callahan
173828d394 Bug 622072. Part 1: Remove rect parameter from Updated() and change implementations to defer updates to render time. r=bas 
The rectangle parameter is currently not used --- all callers always pass the full canvas bounds. In the long term,
we probably won't want this parameter since all implementations should be doing accelerated drawing direct to buffers
with no intermediate copies, hence there will be no need to optimize the size of those copies. Plus, performance-sensitive
testcases tend to paint most or all of the canvas on every frame anyway.
 2011-03-28 12:59:46 +13:00
Chris Jones
ee2d946b9f Bug 627273, part 4: Basic impl of ImageLayer tiling for basic layers. r=roc,vlad 2011-01-26 00:26:37 -06:00
Chris Jones
ac5d644f3f Bug 618265: Fix leaking gfxSharedImageSurfaces. r=joe sr=vlad a=b 2011-01-04 10:40:54 -06:00
Oleg Romashin
88c2013756 Bug 619490 - Don't use double-buffered shadowed layers with OGL compositing r=cjones a=b-f 2010-12-31 09:40:31 +02:00
Oleg Romashin
08463ea8b9 Bug 619488 - Inform child processes of the compositor's layer-manager type. r=jones.chris.g a=b-f 2010-12-31 09:40:19 +02:00
Chris Jones
339106b788 Bug 562424: Switch IPDL interfaces to infallible TArray. r=bent a=blocking 2010-11-08 20:49:00 -06:00
Chris Jones
c493eae976 Bug 603885, part 4: Use unsafe-shmem-backed gfxSharedImageSurface for IPC shadow layers. r=joe 2010-11-05 02:17:07 -05:00
Chris Jones
374601de26 Bug 599359, part 2: When we allocate a new front/back buffer pair, pre-fill the new front buffer with what we painted into the back buffer to stabilize the first Swap(). r=vlad a=blocking-fennec 2010-09-28 17:05:30 -05:00
Chris Jones
59ae652264 Bug 599562: Allocate cross-process layers surfaces from SyS V shm instead of /dev/shm to avoid being arbitrarily capped on maemo5/N900. r=stuart a=blocking-fennec 2010-09-27 19:59:43 -05:00
Chris Jones
d4a32df6d8 Bug 591555: Shadow layer resources are scoped to their PLayerParent, of which there may be many for one layer manager, so move resource management down into Shadow*Layer. r=joe sr=roc 2010-09-24 17:41:31 -05:00
Chris Jones
09cd99ac19 Bug 596240: Don't front() on an empty vector. r=joe a=blocking-2.0 2010-09-20 14:40:30 -05:00
Doug Turner
d03c15e824 Bug 593175, part 2: Use 5-6-5 opaque gfxSharedImageSurfaces on mobile platforms. r=karl a=blocking-fennec 2010-09-16 16:24:20 -05:00
Chris Jones
d5d1794564 Bug 593175, part 1: Request backing buffers with gfxContentType rather than gfxImageFormat. r=karl a=blocking-fennec 2010-09-16 16:24:17 -05:00
Chris Jones
b10bbc722f Bug 593175, part 0: Allow forcing use of Shmem-backed cross-process layers in DEBUG builds. r=karl a=blocking-fennec 2010-09-14 00:51:49 -05:00
Chris Jones
166112f466 Bug 596542: Repainting a thebes layer with a non-1.0 resolution shouldn't (always) force buffer reallocs. r=roc 2010-09-15 16:03:13 -05:00
Chris Jones
7c1181632c Bug 570625, part 9: When updating thebes layers, swap out back/front buffers and invalidate the newly-painted region on the old front buffer. r=roc sr=shaver 2010-09-14 00:23:08 -05:00
Chris Jones
0cb7bd0628 Bug 570625, part 5: Mostly mechanically switch BasicShadowable/ShadowThebes over to the new SurfaceDescriptor API. r=roc 2010-09-14 00:23:08 -05:00
Chris Jones
433a837370 Bug 570625, part 4: Put support in place for allocating platform-specific backing buffers for IPC layers, and synchronizing properly during transactions (i.e. XSync() on X11). The platform-specific interfaces just default to Shmem in this patch. r=karl sr=shaver 2010-09-14 00:23:08 -05:00
Chris Jones
7e002c08ab Bug 570625, part 1: Only use back/front buffers for BasicThebesLayer and update back->front in the compositor process. r=roc sr=shaver 2010-09-14 00:23:08 -05:00
Chris Jones
1ef9f416ce Merge m-c --> cedar 
--HG--
rename : layout/style/test/hover_helper.html => layout/style/test/chrome/hover_helper.html
rename : layout/style/test/test_hover.html => layout/style/test/chrome/test_hover.html
 2010-09-03 14:07:55 -05:00
Chris Jones
81e7d1c97a Bug 593248: Release old shmem surfaces on realloc. r=joe 2010-09-02 22:05:01 -05:00
Robert O'Callahan
be2d05f9a8 Bug 579276. Part 3: Change Set/IsOpaqueContent API to a more generic Get/SetContentFlags API. r=cjones,sr=vlad 2010-09-02 21:18:40 +12:00
Chris Jones
23560b1841 Followup to bug 570294: Avoid vector::data(), it's not C++98. 2010-07-21 18:13:24 -05:00
Chris Jones
f652f038cf Bug 570294, part h: Publish BasicCanvasLayer pixels to remote processes (slowly). sr=vlad 2010-07-21 16:17:33 -05:00
Chris Jones
cd6faf2b77 Bug 570294, part f: Implement the "manager" side of IPC layers. r=vlad 2010-07-21 16:17:33 -05:00
Chris Jones
2f28607585 Bug 570294, part e: Implement the "forwarder" side of IPC layers. r=jrmuizel 2010-07-21 16:17:33 -05:00