Hiroyuki Ikezoe
4d526d411d
Bug 1167627 - Part 9: Use mozinfo in services/. r=rnewman
2015-10-11 21:50:00 +02:00
Hiroyuki Ikezoe
dd1c422ef6
Bug 1167627 - Part 8: Use mozinfo in image/.
2015-08-25 15:01:00 +02:00
Hiroyuki Ikezoe
c20868b358
Bug 1167627 - Part 7: Use mozinfo in netwerk/.
2015-08-25 14:58:00 +02:00
Hiroyuki Ikezoe
745b49d160
Bug 1167627 - Part 6: Use mozinfo in security/. r=dkeeler
2015-10-11 21:49:00 +02:00
Hiroyuki Ikezoe
f3b2d1b890
Bug 1167627 - Part 5: Use mozinfo in widget/.
2015-08-25 14:54:00 +02:00
Hiroyuki Ikezoe
0f107a35f9
Bug 1167627 - Part 4: Use mozinfo in uriloader/.
2015-08-25 14:51:00 +02:00
Hiroyuki Ikezoe
14e3e2945e
Bug 1167627 - Part 3: Use mozinfo in xpcom/. r=bsmedberg
2015-10-11 21:48:00 +02:00
Hiroyuki Ikezoe
279cfef515
Bug 1167627 - Part 2: Use mozinfo in modules/libmar/. r=rstrong
2015-10-11 21:48:00 +02:00
Hiroyuki Ikezoe
ab09d590f3
Bug 1167627 - Part 1: Use mozinfo in dom/. r=mrbkap
2015-10-11 21:47:00 +02:00
JW Wang
b114aa9192
Bug 1212701. Part 2 - remove MediaDecoderStateMachine::OnDecodeTaskQueue() which is unused. r=jya.
2015-10-12 12:05:52 +08:00
JW Wang
b66b801b82
Bug 1212701. Part 1 - remove AbstractMediaDecoder::OnDecodeTaskQueue(). r=jya.
2015-10-12 12:05:49 +08:00
Nigel Babu
28a2d5b11d
Backed out changeset ae5f03f140a8 (bug 1209801) on suspicion of Android reftest bustage ON A CLOSED TREE
2015-10-12 09:43:19 +05:30
Nigel Babu
6af640af93
Backed out changeset 9e69375c47ed (bug 1209801) for suspicion of Android reftest bustage ON A CLOSED TREE
2015-10-12 09:43:06 +05:30
Morris Tseng
3a1685318e
Bug 709490 - Part 11: Diabled test_offscreencanvas_many.html on gonk, android, windows and linux. r=jgilbert
2015-10-12 11:21:04 +08:00
Morris Tseng
62ed7ded96
Bug 709490 - Part 10: Using mechanism in RuntimeService to get pref in worker thread instead of gfxPref. r=baku
2015-10-12 11:21:04 +08:00
Morris Tseng
91aa1baea6
Bug 709490 - Part 9: Readback without blocking main thread. r=jgilbert
2015-10-12 11:21:03 +08:00
Morris Tseng
5d86d6193b
Bug 709490 - Part 8: Copy to a temp texture when readback from IOSurface. r=jgilbert
2015-10-12 11:21:03 +08:00
Morris Tseng
9baf236c2f
Bug 709490 - Part 7: If layer is not available, fallback to BasicCanvasLayer. r=roc
2015-10-12 11:21:03 +08:00
Morris Tseng
e0d17ce5bd
Bug 709490 - Part 6: Add frame ID to CanvasClient so compositor could update frame correctly. r=roc
2015-10-12 11:21:03 +08:00
Morris Tseng
3ea0cda3c7
Bug 709490 - Part 5: Add interfaces test. r=ehsan
2015-10-12 11:21:03 +08:00
Morris Tseng
6cd89d555d
Bug 709490 - Part 4: Mochitests for offscreencanvas. r=baku, r=jgilbert
...
Thanks Jon Morton [:jmorton] (jonanin@gmail.com ) for polishing patches.
2015-10-12 11:21:03 +08:00
Morris Tseng
225da9066a
Bug 709490 - Part 3: Transfer OffscreenCanvas from mainthread to workers. r=baku, r=sfink
...
Thanks Jon Morton [:jmorton] (jonanin@gmail.com ) for polishing patches.
2015-10-12 11:21:03 +08:00
Morris Tseng
b1e15627bf
Bug 709490 - Part 2: Introduce OffscreenCanvas and let WebGL context work on workers. r=nical, r=jgilbert, r=jrmuizel, sr=ehsan
...
Thanks Jon Morton [:jmorton] (jonanin@gmail.com ) for polishing patches.
2015-10-12 11:21:03 +08:00
Morris Tseng
76ca7da268
Bug 709490 - Part 1: Let ImageBridge transfer CanvasClient async. r=nical
...
Thanks Jon Morton [:jmorton] (jonanin@gmail.com ) for polishing patches.
2015-10-12 11:21:02 +08:00
Morris Tseng
8ab4182294
Bug 1212724 - Fix compile error in non-unified mode. r=nical
2015-10-12 11:21:02 +08:00
Jean-Yves Avenard
c6a346d745
Bug 1212795: P3. Make AppleDecoderModule detects if the required modules are loaded. r=jwwang
2015-10-12 14:47:50 +11:00
Jean-Yves Avenard
c9902f8f1a
Bug 1212795: P2. Ensure all frameworks required for video decoding on mac are loaded. r=jwwang
...
We used to rely on the compositor to have been started, which is a dangerous assumption..
2015-10-12 14:47:49 +11:00
Jean-Yves Avenard
6c0274e691
Bug 1212795: P1. Ensure the check if HW acceleration is allowed is performed on the main thread. r=jwwang
...
The downside is that a restart is now required if the user change the preference to force HW decoding ; however this is the same behaviour as on Windows.
2015-10-12 14:47:49 +11:00
JW Wang
f597d69750
Bug 1212220 - cache pref values so they are safe to access off the main thread. r=roc.
2015-10-11 20:11:04 +08:00
Heiher
d2173e9c61
Bug 1213532 - IonMonkey: MIPS32: Rename BaseFloatRegisters to FloatRegistersMIPSShared. r=arai
...
---
js/src/jit/mips32/Architecture-mips32.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
2015-10-12 17:14:54 +08:00
Heiher
5754f3aef5
Bug 1213532 - IonMonkey: MIPS32: Fix call FloatRegistersMIPSShared::GetName recursive. r=arai
...
---
js/src/jit/mips32/Architecture-mips32.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
2015-10-12 17:05:02 +08:00
Xidorn Quan
9b48e59080
Bug 1212299 part 3 - Add test for requesting fullscreen from doc inside frame/object. rs=smaug
2015-10-12 11:24:23 +11:00
Xidorn Quan
24b1e70c47
Bug 1212299 part 2 - Rewrite fullscreen-denied test to have a clearer structure. r=smaug
2015-10-12 11:24:23 +11:00
Xidorn Quan
2d2beb4278
Bug 1212299 part 1 - Forbid documents inside elements other than iframe from requesting fullscreen. r=smaug
2015-10-12 11:24:23 +11:00
Jean-Yves Avenard
d41a442db7
Bug 1206845: Prevent overflows in MediaCache. r=roc
2015-10-12 10:01:26 +11:00
Jean-Yves Avenard
d8a8bf0881
Bug 1213173: Always use FFmpeg regardless of version. r=kentuckyfriedtakahe
...
We don't need to prevent using FFmpegDecoderModule on older version anymore
2015-10-12 10:00:04 +11:00
Karl Tomlinson
6c3bd3d921
bug 1210267 remove custom AudioParam::Release and disconnect in destructor r=baku
...
AudioParam has no derived type (and DisconnectFromGraphAndDestroyStream calls
no virtual functions) so no need for special release.
2015-10-02 14:24:51 +13:00
Karl Tomlinson
953b4a9d59
bug 1210267 use DOMEventTargetHelper::LastRelease instead of custom Release r=baku
...
AudioNode already has NS_IMPL_CYCLE_COLLECTING_RELEASE_WITH_LAST_RELEASE,
which performs mRefCnt.incr/decr for LastRelease(), so the kungFuDeathGrip
virtual AddRef/Release in DisconnectFromGraph() is additional noise when
debugging/tracing ownership.
Unlink() already assumes that the caller holds a reference (and it does).
2015-10-01 00:14:26 +13:00
Steve Singer
8aa08a28b7
Bug 1213136 - Declare BrotliPullByte before it is used. r=mcmanus
2015-10-11 18:13:10 +02:00
Tom Klein
9524cbb750
Bug 1212954 - Make BaseSize::IsEmpty return true for negative sizes to avoid NS_ERROR in VectorImage::GetFrameAtSize when root svg doesn't have intrinsic width or height. r=roc
2015-10-11 18:13:10 +02:00
Sebastian Hengst
64ef92adbe
Bug 1212288 - Make ImageContainer::AllocateProducerID callable on all threads; r=roc
2015-10-11 18:13:10 +02:00
Dragana Damjanovic
636a498393
Bug 1211694 - dataLength has been added twice. r=nwgh
2015-10-11 18:13:10 +02:00
Dragana Damjanovic
43c2862c4a
Bug 1211690 - Remove not needed nsILoadInfo get. r=ckerschb
2015-10-11 18:13:10 +02:00
Jamie Nicol
22bfcfd371
Bug 1209801 - Part 2: Do not allow big image textures to be used with TiledLayerBuffers. r=mattwoodrow
...
Big-image textures (where multiple textures are tiled together to act as
a single larger texture, e.g. TiledTextureImage) were in some cases
being used to back the TextureSources associated with each Tile in a
TiledLayerBuffer. CompositorOGL was unaware of this, so when compositing
would only render the first tile in each big image, stretching it to the
size of the entire big image.
It doesn't make much sense to allow tiling-within-tiling, so set the
DISALLOW_BIGIMAGE flag for textures created for use with TiledLayerBuffers.
---
gfx/layers/client/SingleTiledContentClient.cpp | 2 +-
gfx/layers/client/TiledContentClient.cpp | 2 +-
gfx/layers/composite/TextureHost.cpp | 2 +-
3 files changed, 3 insertions(+), 3 deletions(-)
2015-10-11 18:13:10 +02:00
Jamie Nicol
c13ab94c8e
Bug 1209801 - Part 1: Add TextureFlags parameter to TextureClientPool. r=mattwoodrow
...
---
gfx/layers/client/ClientLayerManager.cpp | 19 ++++++++++++-------
gfx/layers/client/ClientLayerManager.h | 2 +-
gfx/layers/client/TextureClientPool.cpp | 6 ++++--
gfx/layers/client/TextureClientPool.h | 8 +++++++-
gfx/layers/client/TiledContentClient.cpp | 3 ++-
5 files changed, 26 insertions(+), 12 deletions(-)
2015-10-11 18:13:10 +02:00
Jorg K
efcc4abcf2
Bug 1209414
- New test for manual dictionary change. r=ehsan
2015-10-11 18:13:10 +02:00
Jakob Olesen
5352e0e73f
Bug 1208674 - Fix ToggleCall to handle constant pools. r=sstangl
...
Handle constant pools inserted anywhere, even at the initial pointer
location. Don't attempt handling constant pools with natural guard
branches. They are not relevant here (and actually never generated
currently).
Stop flushing constant pools in toggledCall(). Since ToggleCall() does
handle constant pools at any point, there is no need to prevent their
insertion.
Replace the NextInstruction() functions with Instruction::skipPool()
which mimics the ARM implementation.
Add an Instruction::IsStackPtrSync() which recognizes the instructions
inserted by syncStackPtr().
2015-10-11 18:13:09 +02:00
Jakob Olesen
8352084037
Bug 1208674 - part 1: Add ToggleCall assertions. r=sstangl
...
Assert that the ToggleCall() function does not overwrite any unexpected
instructions.
2015-10-11 18:13:09 +02:00
sajitk
b4267f6ab9
Bug 1201740 - Prevent interception if the request is a System XHR. r=bkelly
2015-10-11 18:13:09 +02:00
Dragana Damjanovic dd.mozilla@gmail.com
f603d214a0
Bug 1197679 - If nsUnknownDecoder is involved in e10s DivertToParent can break. r=jduell
2015-10-11 18:13:09 +02:00