438c49f67c639f96b4120500d21f3ef278406b02: Merge m-c to graphics
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:52:43 -0400 - rev 407511
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Merge m-c to graphics MozReview-Commit-ID: DVUE3Wys3HV
ee8156686174a5943a63d825e797be091df9f550: Bug 1361357 - Try fixing clip on WRDILayer. r=ethlin
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:51:23 -0400 - rev 407510
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1361357 - Try fixing clip on WRDILayer. r=ethlin This should fix some wrong-looking code that computes the clip for WebRenderDisplayItemLayer instances. As this code is not exercised very much it's hard to know if this change is actually correct. MozReview-Commit-ID: BSsBpBmzWgd
7cb3c4963a3fda3a9ed8fc4ef3e08296c693a53d: Bug 1360246 - Remove the ParentBounds and TransformedVisibleBoundsRelativeToParent functions. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:08 -0400 - rev 407509
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Remove the ParentBounds and TransformedVisibleBoundsRelativeToParent functions. r=nical These functions also have built-in assumptions about the stacking context structure mapping to the layer tree structure which are not necessarily valid. So we remove these functions and update code that uses them. MozReview-Commit-ID: UkXXAF59qv
0f0a5f78c75c6f3be940f20c710b65103401bf20: Bug 1360246 - Remove the RelativeToParent functions by propagating StackingContextHelper chains. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:08 -0400 - rev 407508
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Remove the RelativeToParent functions by propagating StackingContextHelper chains. r=nical The only remaining callers of RelativeToParent() are in StackingContextHelper itself, which we can remove now by having the SCH take a parent SCH and use it instead of RelativeToParent(). This patch implements this change. This makes a failing test pass, because of how preserve-3d container layers work. Specifically, preserve-3d container layers render their descendants in z-order, not in tree order. If those children were assuming that their parent had already pushed a stacking context, that assumption may have been false because the parent might have not yet been rendered because of z-ordering. By using the StackingContextHelper chain instead of the layer tree ancestry, we fix the stacking-context-relative coordinates being used in the descendant subtree of preserve-3d container layers. MozReview-Commit-ID: HzZvBuAlMdB
2e29ab7dc180e0e30423ced43632a9d0a80a3208: Bug 1360246 - Update code to use StackingContextHelper::ToRelativeWr* instead of RelativeToParent. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:08 -0400 - rev 407507
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Update code to use StackingContextHelper::ToRelativeWr* instead of RelativeToParent. r=nical This updates more code that was using RelativeToParent() to use the stacking context helper's ToRelativeWr* functions instead. This get us closer to breaking the assumption that the WR stacking context order maps 1:1 to the layer tree structure. MozReview-Commit-ID: HQrbvCgPOW4
a4b49a2aa1bd45778b594f027fec3e4ba1c7c996: Bug 1360246 - Stop calling RelativeToParent on rects for which we only use the size, since that doesn't change with RelativeToParent. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:08 -0400 - rev 407506
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Stop calling RelativeToParent on rects for which we only use the size, since that doesn't change with RelativeToParent. r=nical Also clean up another untyped Size calculation to use LayoutDeviceSize::FromAppUnits. MozReview-Commit-ID: DvEIhZpBTEi
e291b3dc7825707aea1eb24934c09d089213fc09: Bug 1360246 - Propagate the StackingContextHelper to the rest of the displaylist-building code. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:08 -0400 - rev 407505
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Propagate the StackingContextHelper to the rest of the displaylist-building code. r=nical This propagates the StackingContextHelper to the rest of the code that builds WR display items, because we will need it in future patches to stop using RelativeToParent. MozReview-Commit-ID: 3PlQrJRhH36
4776be9cd57ac5796548d124b46e7ad1a5a42f99: Bug 1360246 - Clean up WebRenderRefLayer::RenderLayer. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:07 -0400 - rev 407504
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Clean up WebRenderRefLayer::RenderLayer. r=nical This cleans up WebRenderRefLayer::RenderLayer to use typed coordinate systems and the StackingContextHelper. Note that this patch contains a functional change, because the clip rect pushed no longer includes the transform on the ref layer itself. It's not clear to me why we were doing that, and didn't seem correct. MozReview-Commit-ID: K7FoeLnXc56
ca74bdc958e1e09b3a83a6eee300adb717c7817c: Bug 1360246 - Clean up WebRenderTextLayer::RenderLayer. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:07 -0400 - rev 407503
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Clean up WebRenderTextLayer::RenderLayer. r=nical This uses the StackingContextHelper and typed helper functions in WebRenderLayer to simplify WebRenderTextLayer::RenderLayer. It also removes the implicit assumption in WebRenderTextLayer that the parent layer pushed a stacking context, which is an assumption we will probably break in the future. MozReview-Commit-ID: CARoGVQd56i
3b25a123e53f9817050d2030014e2291ba6e8d06: Bug 1360246 - Propagate the StackingContextHelper through CreateWebRenderCommands. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:07 -0400 - rev 407502
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Propagate the StackingContextHelper through CreateWebRenderCommands. r=nical This is needed for the next patch, because the bullet frame WR rendering code will need a StackingContextHelper to pass to the PushGlyphs call. It also helps with some of the later patches in this series, when we get rid of the RelativeToParent calls. MozReview-Commit-ID: Is0aciHhy2N
8e529a1e760efca512a14abd01f80849dc44a6f4: Bug 1360246 - Propagate a StackingContextHelper all the way through the RenderLayer traversal. r=nical
Kartikaya Gupta <kgupta@mozilla.com> - Wed, 03 May 2017 08:48:06 -0400 - rev 407501
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360246 - Propagate a StackingContextHelper all the way through the RenderLayer traversal. r=nical This is needed in part 3 to update WebRenderTextLayer::RenderLayer, so that it no longer assumes the parent container layer has pushed a stacking context, and instead explicitly uses the StackingContextHelper. MozReview-Commit-ID: 9twUmDgUipX
1ba304cfe3908fcf53109d2e8435532448887c78: Bug 1360127 - If the background image doesn't have a image container, we shouldn't create a layer for it. r=mchang
Ethan Lin <ethlin@mozilla.com> - Wed, 03 May 2017 18:27:10 +0800 - rev 407500
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1360127 - If the background image doesn't have a image container, we shouldn't create a layer for it. r=mchang
7e836ca5016ae4b4120a376c3cc427d91c88e7c1: Merge m-c to graphics
Kartikaya Gupta <kgupta@mozilla.com> - Tue, 02 May 2017 09:45:11 -0400 - rev 407499
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Merge m-c to graphics MozReview-Commit-ID: 9XwZUWIhGir
ae9f92d425425e98d585b4c905a351da04ed2543: Bug 1358437 - Correct the way to use override pref for WebRender OMTA, r=me
peter chang <pchang@mozilla.com> - Tue, 02 May 2017 11:14:02 +0800 - rev 407498
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1358437 - Correct the way to use override pref for WebRender OMTA, r=me MozReview-Commit-ID: JGrt3GVU0yo
22ed1758f789bc9d1247068f32a52b50d2e06b46: Bug 1358437 - fix compiler error, r=me
peter chang <pchang@mozilla.com> - Tue, 02 May 2017 10:56:52 +0800 - rev 407497
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1358437 - fix compiler error, r=me MozReview-Commit-ID: H2VAWFiONAf
77880d70c8750a3f8f47b0a4bffe987b40e9ae4b: Bug 1358437 - enable WebRender OMTA by default, r?kats
peter chang <pchang@mozilla.com> - Wed, 19 Apr 2017 21:54:11 +0800 - rev 407496
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1358437 - enable WebRender OMTA by default, r?kats MozReview-Commit-ID: DG2eK8lMTBi
6ba01d83eb61a11d5c58deb5f50b2bd1675a2600: Bug 1358437 - pass layer's transform/opacity to compositor, r?kats
peter chang <pchang@mozilla.com> - Tue, 25 Apr 2017 11:52:44 +0800 - rev 407495
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1358437 - pass layer's transform/opacity to compositor, r?kats Need to pass the default transform/opacity to compositor if animations exist because it is possible that gecko fails to get animated value after animation sampling, like an animation with delay. MozReview-Commit-ID: IK06hWvaSPf
53f5de187374b12528a57e5619f14fb6a9d54aa5: Bug 1358437 - Setup correct opacity/transform in stacking context when there exists opacity/transform aniamtions. r?kats
peter chang <pchang@mozilla.com> - Fri, 28 Apr 2017 16:01:00 +0800 - rev 407494
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1358437 - Setup correct opacity/transform in stacking context when there exists opacity/transform aniamtions. r?kats Pass empty opacity(transform) in stacking context when there exists opacity(transform) animations and the final opacity(transform) value will be resolved on the compositor after animation sampling MozReview-Commit-ID: 6pF9Oe8Ks2I
f877064cf25c2eafef32df87deb02f16107ce705: Bug 1358437 - pass layer's transform attributes for transform animation, r?kats
peter chang <pchang@mozilla.com> - Mon, 17 Apr 2017 11:35:42 +0800 - rev 407493
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1358437 - pass layer's transform attributes for transform animation, r?kats MozReview-Commit-ID: J7JHuwvWuet
588a50945a7f67fd7932aa3bbb7f86b64802bbaf: Bug 1357320 - Dispatches the discarded compositor animations id list in one async call, r?kats
peter chang <pchang@mozilla.com> - Wed, 26 Apr 2017 10:58:50 +0800 - rev 407492
Push 1490 by mtabara@mozilla.com at Mon, 31 Jul 2017 14:08:16 +0000
Bug 1357320 - Dispatches the discarded compositor animations id list in one async call, r?kats MozReview-Commit-ID: 8cCNCzxMKpY
(0) -300000 -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 tip