searching for reviewer(gw)
f0c87ec1d694: Bug 1544374 - Make FilterOp not copy. r=gw
Nicolas Silva <nsilva@mozilla.com> - Tue, 16 Apr 2019 07:43:04 +0000 - rev 469658
Push 112806 by apavel@mozilla.com at Tue, 16 Apr 2019 15:49:42 +0000
Bug 1544374 - Make FilterOp not copy. r=gw Differential Revision: https://phabricator.services.mozilla.com/D27480
9feaaddc3761: Bug 1444434 - Bind pixel pack buffer before trying to map it for readback r=gw
Barret Rennie <barret@brennie.ca> - Mon, 15 Apr 2019 23:39:10 +0000 - rev 469601
Push 112803 by ccoroiu@mozilla.com at Tue, 16 Apr 2019 04:10:43 +0000
Bug 1444434 - Bind pixel pack buffer before trying to map it for readback r=gw Differential Revision: https://phabricator.services.mozilla.com/D26946
4febe6f32b23: Bug 1525642 - ensure that w~=1 when assuming a transform is just a translation. r=gw
Alexis Beingessner <a.beingessner@gmail.com> - Thu, 11 Apr 2019 20:52:21 +0000 - rev 469116
Push 112769 by csabou@mozilla.com at Fri, 12 Apr 2019 03:51:12 +0000
Bug 1525642 - ensure that w~=1 when assuming a transform is just a translation. r=gw Differential Revision: https://phabricator.services.mozilla.com/D27163
c1073f5dd499: Bug 1538711 - Fix tile cache clip chain in WR r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Wed, 03 Apr 2019 16:28:19 +0000 - rev 467847
Push 112658 by aciure@mozilla.com at Thu, 04 Apr 2019 04:41:45 +0000
Bug 1538711 - Fix tile cache clip chain in WR r=gw Differential Revision: https://phabricator.services.mozilla.com/D25667
03ef11b535d4: Bug 1528674 - WR improve picture texture cache allocaton r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Thu, 28 Mar 2019 03:02:45 +0000 - rev 466536
Push 112585 by opoprus@mozilla.com at Thu, 28 Mar 2019 10:26:17 +0000
Bug 1528674 - WR improve picture texture cache allocaton r=gw The change contains a number of incremental improvements with the main goal of: - allocating exactly as many tile as required by the app - respecting the picture caching option Differential Revision: https://phabricator.services.mozilla.com/D24740
bbb6a31ebf07: Bug 1528674 - WR improve picture texture cache allocaton r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Tue, 26 Mar 2019 21:18:09 +0000 - rev 466251
Push 112571 by csabou@mozilla.com at Wed, 27 Mar 2019 05:03:16 +0000
Bug 1528674 - WR improve picture texture cache allocaton r=gw The change contains a number of incremental improvements with the main goal of: - allocating exactly as many tile as required by the app - respecting the picture caching option Differential Revision: https://phabricator.services.mozilla.com/D24740
1d405e872db4: Bug 1441308 - Map document layers to z ranges r=gw
Doug Thayer <dothayer@mozilla.com> - Fri, 22 Mar 2019 18:28:51 +0000 - rev 465771
Push 112523 by shindli@mozilla.com at Sat, 23 Mar 2019 09:54:29 +0000
Bug 1441308 - Map document layers to z ranges r=gw As discussed in IRC. AFAICT the ORTHO_NEAR|FAR_PLANE should match up with the ranges of valid ZBufferIds, but please double-check me. Differential Revision: https://phabricator.services.mozilla.com/D23599
31eb0acf3a1d: Bug 1441308 - Correct Renderer issues with multiple documents r=gw
Doug Thayer <dothayer@mozilla.com> - Fri, 22 Mar 2019 18:28:46 +0000 - rev 465770
Push 112523 by shindli@mozilla.com at Sat, 23 Mar 2019 09:54:29 +0000
Bug 1441308 - Correct Renderer issues with multiple documents r=gw This corrects A) An issue encountered with our strategy for skipping the end_pass call for all but an offscreen render target. See the comment above the end_pass call for details, and B) An issue with depth clearing where we do not clear the whole rect if there are multiple non-intersecting documents. Differential Revision: https://phabricator.services.mozilla.com/D23056
d8b4d6ec9b40: Bug 1441308 - Map document layers to z ranges r=gw
Doug Thayer <dothayer@mozilla.com> - Thu, 21 Mar 2019 23:15:23 +0000 - rev 465594
Push 112518 by apavel@mozilla.com at Fri, 22 Mar 2019 16:53:03 +0000
Bug 1441308 - Map document layers to z ranges r=gw As discussed in IRC. AFAICT the ORTHO_NEAR|FAR_PLANE should match up with the ranges of valid ZBufferIds, but please double-check me. Differential Revision: https://phabricator.services.mozilla.com/D23599
e2f83e4816dd: Bug 1441308 - Correct Renderer issues with multiple documents r=gw
Doug Thayer <dothayer@mozilla.com> - Fri, 22 Mar 2019 03:15:18 +0000 - rev 465593
Push 112518 by apavel@mozilla.com at Fri, 22 Mar 2019 16:53:03 +0000
Bug 1441308 - Correct Renderer issues with multiple documents r=gw This corrects A) An issue encountered with our strategy for skipping the end_pass call for all but an offscreen render target. See the comment above the end_pass call for details, and B) An issue with depth clearing where we do not clear the whole rect if there are multiple non-intersecting documents. Differential Revision: https://phabricator.services.mozilla.com/D23056
93c9a0e2da6e: Bug 1527751 - properly apply transformed clips in brush shader. r=gw
Alexis Beingessner <a.beingessner@gmail.com> - Mon, 18 Mar 2019 21:05:56 +0000 - rev 464920
Push 112485 by dvarga@mozilla.com at Tue, 19 Mar 2019 09:58:30 +0000
Bug 1527751 - properly apply transformed clips in brush shader. r=gw Differential Revision: https://phabricator.services.mozilla.com/D23905
495ee696b110: Bug 1511433 - Ensure address variables in fragment shaders are highp precision. r=gw
Jamie Nicol <jnicol@mozilla.com> - Sun, 17 Mar 2019 20:54:29 +0000 - rev 464763
Push 112481 by rgurzau@mozilla.com at Mon, 18 Mar 2019 21:51:19 +0000
Bug 1511433 - Ensure address variables in fragment shaders are highp precision. r=gw In GLES 3 GLSL, the default precision for ints is highp (32 bit) in vertex shaders, but only mediump (16 bit) in fragment shaders. To render linear and radial gradients the fragment shader must fetch the gradient stops from the gpu cache, using an address variable. This variable is a 16 bit int, so if the stops data is located at too high an address (row 32 or greater) then this value will have overflown and we fetch from the wrong location. This was resulting in garbage being drawn instead of the correct gradients. To fix this, any address used in a fragment shader must be marked as highp. This includes the varying input which supplies the address, and the arguments to any functions used for the fetch. Differential Revision: https://phabricator.services.mozilla.com/D23669
64f683055202: Bug 1527325 - Don't re-export contents of WR API units r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Thu, 14 Mar 2019 01:44:05 +0000 - rev 463941
Push 112423 by dvarga@mozilla.com at Thu, 14 Mar 2019 10:05:46 +0000
Bug 1527325 - Don't re-export contents of WR API units r=gw The cleans up our WR use statements further, easying the merge conflicts. Note: this PR is subject to instant rot, it is preferred to land quickly. Differential Revision: https://phabricator.services.mozilla.com/D23373
950174a8c708: Bug 1527325 - Don't re-export contents of WR API units r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Wed, 13 Mar 2019 19:30:56 +0000 - rev 463883
Push 112422 by ncsoregi@mozilla.com at Thu, 14 Mar 2019 05:16:19 +0000
Bug 1527325 - Don't re-export contents of WR API units r=gw The cleans up our WR use statements further, easying the merge conflicts. Note: this PR is subject to instant rot, it is preferred to land quickly. Differential Revision: https://phabricator.services.mozilla.com/D23373
2c83d53042c6: Bug 1532174 - Roll back env_logger to 0.5 r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Tue, 12 Mar 2019 02:54:05 +0000 - rev 463588
Push 112400 by btara@mozilla.com at Tue, 12 Mar 2019 10:04:06 +0000
Bug 1532174 - Roll back env_logger to 0.5 r=gw needed to avoid duplicate dependencies Differential Revision: https://phabricator.services.mozilla.com/D23072
c154ee9861ab: Bug 1532174 - Refactor get_relative_transform to not return an option r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Tue, 12 Mar 2019 02:27:29 +0000 - rev 463586
Push 112400 by btara@mozilla.com at Tue, 12 Mar 2019 10:04:06 +0000
Bug 1532174 - Refactor get_relative_transform to not return an option r=gw Just a refactor without any semantical changes. Makes it for cleaner client code. Also prepares for caching of the world transforms. Differential Revision: https://phabricator.services.mozilla.com/D23015
46ad1441e7f2: Bug 1532174 - Strongly typed vertex data stores in WR r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Tue, 12 Mar 2019 02:26:51 +0000 - rev 463585
Push 112400 by btara@mozilla.com at Tue, 12 Mar 2019 10:04:06 +0000
Bug 1532174 - Strongly typed vertex data stores in WR r=gw Establishes a clear link of TransformData with the shader. Associates a vertex data type permanently, which makes the code cleaner and allow finding the relevant bits quickly with search. Differential Revision: https://phabricator.services.mozilla.com/D23006
662d2392cb01: Bug 1532174 - Enable env_logger by default in Wrench r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Tue, 12 Mar 2019 02:26:34 +0000 - rev 463584
Push 112400 by btara@mozilla.com at Tue, 12 Mar 2019 10:04:06 +0000
Bug 1532174 - Enable env_logger by default in Wrench r=gw Env logger is useful for: 1. seeing error! generated by the code (by default). Sometimes it can give us hints on what is wrong with the capture, or a test right away. 2. seeing processed shader compile/link errors, which is very useful when changing them Given that wrench is the main debugging tool at the moment, enabling the logging output by default should increase our productivity, at the cost of slightly longer build times. Differential Revision: https://phabricator.services.mozilla.com/D23000
bb76bbd404c0: Bug 1532245. Factor out component transfer code to a function in brush_blend shader for webreder. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 11 Mar 2019 21:34:30 -0500 - rev 463482
Push 112399 by tnikkel@gmail.com at Tue, 12 Mar 2019 02:34:56 +0000
Bug 1532245. Factor out component transfer code to a function in brush_blend shader for webreder. r=gw Some phones can't compile the shader as is: "ERROR: 0:1448: '' : No default label can be nested inside other control flow nested within their corresponding switch" https://github.com/servo/webrender/wiki/Driver-issues#bug-1532245---switch-statement-inside-control-flow-inside-switch-statement-fails-to-compile-on-some-android-phones Differential Revision: https://phabricator.services.mozilla.com/D22853
c32dcf4e6f92: Bug 1532245. Factor out component transfer code to a function in brush_blend shader for webreder. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 11 Mar 2019 17:01:40 -0500 - rev 463480
Push 112397 by tnikkel@gmail.com at Mon, 11 Mar 2019 22:01:55 +0000
Bug 1532245. Factor out component transfer code to a function in brush_blend shader for webreder. r=gw Some phones can't compile the shader as is: "ERROR: 0:1448: '' : No default label can be nested inside other control flow nested within their corresponding switch" https://github.com/servo/webrender/wiki/Driver-issues#bug-1532245---switch-statement-inside-control-flow-inside-switch-statement-fails-to-compile-on-some-android-phones Differential Revision: https://phabricator.services.mozilla.com/D22853
b958c752bbe6: Bug 1532284 - Build using android-28. r=gw
Kartikaya Gupta <kgupta@mozilla.com> - Thu, 07 Mar 2019 17:34:50 +0000 - rev 463237
Push 112369 by btara@mozilla.com at Fri, 08 Mar 2019 21:57:38 +0000
Bug 1532284 - Build using android-28. r=gw This is the SDK version we use to build other Android stuff in mozilla-central. It's simpler if we build Wrench using the same target since we can reuse the toolchains. Depends on D22376 Differential Revision: https://phabricator.services.mozilla.com/D22377
3dc8bca5dd90: Bug 1532284 - Fix rustc compiler warnings when building wrench for Android. r=gw
Kartikaya Gupta <kgupta@mozilla.com> - Thu, 07 Mar 2019 17:34:50 +0000 - rev 463236
Push 112369 by btara@mozilla.com at Fri, 08 Mar 2019 21:57:38 +0000
Bug 1532284 - Fix rustc compiler warnings when building wrench for Android. r=gw Various variables are unused when building for Android. Differential Revision: https://phabricator.services.mozilla.com/D22376
1972b8836948: Bug 1532810 - Report long-running GL calls to the profiler. r=gw
Bobby Holley <bobbyholley@gmail.com> - Tue, 05 Mar 2019 11:31:49 -0800 - rev 463222
Push 112368 by bholley@mozilla.com at Fri, 08 Mar 2019 21:35:07 +0000
Bug 1532810 - Report long-running GL calls to the profiler. r=gw Differential Revision: https://phabricator.services.mozilla.com/D22202
e8a962d607cd: Bug 1522787 - downscale bitmap glyphs to avoid filtering artifacts r=gw
Lee Salzman <lsalzman@mozilla.com> - Fri, 08 Mar 2019 08:48:38 -0500 - rev 463218
Push 112366 by lsalzman@mozilla.com at Fri, 08 Mar 2019 20:53:22 +0000
Bug 1522787 - downscale bitmap glyphs to avoid filtering artifacts r=gw Differential Revision: https://phabricator.services.mozilla.com/D22704
1cb006402562: Bug 1532810 - Report long-running GL calls to the profiler. r=gw
Bobby Holley <bobbyholley@gmail.com> - Tue, 05 Mar 2019 11:31:49 -0800 - rev 463213
Push 112361 by bholley@mozilla.com at Fri, 08 Mar 2019 17:18:54 +0000
Bug 1532810 - Report long-running GL calls to the profiler. r=gw Differential Revision: https://phabricator.services.mozilla.com/D22202
04e9867c3f73: Bug 1533139 - pass in raster scale to text run shader without inverting r=gw
Lee Salzman <lsalzman@mozilla.com> - Wed, 06 Mar 2019 21:39:04 +0000 - rev 462745
Push 112326 by ccoroiu@mozilla.com at Thu, 07 Mar 2019 04:41:19 +0000
Bug 1533139 - pass in raster scale to text run shader without inverting r=gw Differential Revision: https://phabricator.services.mozilla.com/D22379
690713aece13: Bug 1531196 - Add a "small-screen" debug pref to shrink the overlay a bit. r=gw
Bobby Holley <bobbyholley@gmail.com> - Wed, 06 Mar 2019 20:32:32 +0000 - rev 462724
Push 112326 by ccoroiu@mozilla.com at Thu, 07 Mar 2019 04:41:19 +0000
Bug 1531196 - Add a "small-screen" debug pref to shrink the overlay a bit. r=gw Adaptive sizing would be better, but this is the simplest path to victory. Differential Revision: https://phabricator.services.mozilla.com/D21757
e54a27080037: Bug 1531930 - Interneration refactoring r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Mon, 04 Mar 2019 16:48:40 +0000 - rev 462244
Push 112282 by aciure@mozilla.com at Mon, 04 Mar 2019 21:54:47 +0000
Bug 1531930 - Interneration refactoring r=gw Remove the intern_types module in favor of the associated Internable types that we already have. The only bit of magic I had to do is around serialization bounds, and it's nicely isolated. Differential Revision: https://phabricator.services.mozilla.com/D21797
458edf3460d4: Bug 1531930 - Interneration refactoring r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Sun, 03 Mar 2019 21:14:49 +0000 - rev 462193
Push 112269 by dvarga@mozilla.com at Mon, 04 Mar 2019 09:43:27 +0000
Bug 1531930 - Interneration refactoring r=gw Remove the intern_types module in favor of the associated Internable types that we already have. The only bit of magic I had to do is around serialization bounds, and it's nicely isolated. Differential Revision: https://phabricator.services.mozilla.com/D21797
ee88f4e35d4f: Bug 1531217 - Document origin rewrite and framebuffer coordinates r=gw,nical
Dzmitry Malyshau <dmalyshau@mozilla.com> - Fri, 01 Mar 2019 17:16:59 +0000 - rev 462014
Push 112259 by rmaries@mozilla.com at Sat, 02 Mar 2019 10:10:00 +0000
Bug 1531217 - Document origin rewrite and framebuffer coordinates r=gw,nical The goal of this change was to simplify the semantics of our document placement and split the logical elements inside (display list) from the actual screen rectangle occupied by a document. To achieve that, we introduce the framebuffer space for things Y-flipped on screen. We fix the frame outputs, so that they get produced on the first frame without loopback from the frame building to scene building. Differential Revision: https://phabricator.services.mozilla.com/D21641
93f7dc3084a1: Bug 1531217 - Document origin rewrite and framebuffer coordinates r=gw,nical
Dzmitry Malyshau <dmalyshau@mozilla.com> - Fri, 01 Mar 2019 14:25:36 +0000 - rev 462001
Push 112259 by rmaries@mozilla.com at Sat, 02 Mar 2019 10:10:00 +0000
Bug 1531217 - Document origin rewrite and framebuffer coordinates r=gw,nical The goal of this change was to simplify the semantics of our document placement and split the logical elements inside (display list) from the actual screen rectangle occupied by a document. To achieve that, we introduce the framebuffer space for things Y-flipped on screen. We fix the frame outputs, so that they get produced on the first frame without loopback from the frame building to scene building. Differential Revision: https://phabricator.services.mozilla.com/D21641
ec7068a12a7b: Bug 1530978 - Show profiler stats averaged over half a second instead of 600 frames. r=gw
Nicolas Silva <nsilva@mozilla.com> - Wed, 27 Feb 2019 20:00:56 +0000 - rev 461986
Push 112259 by rmaries@mozilla.com at Sat, 02 Mar 2019 10:10:00 +0000
Bug 1530978 - Show profiler stats averaged over half a second instead of 600 frames. r=gw Differential Revision: https://phabricator.services.mozilla.com/D21392
48bf7d6e2ea7: Bug 1441308 - Fix picture-caching interaction with doc spitting r=gw
Doug Thayer <dothayer@mozilla.com> - Thu, 28 Feb 2019 19:15:15 +0000 - rev 461869
Push 112229 by rgurzau@mozilla.com at Fri, 01 Mar 2019 07:31:26 +0000
Bug 1441308 - Fix picture-caching interaction with doc spitting r=gw This is a stab at what the correct approach to this should be. It seems that we should be using the window size here and not the screen_rect, as the screen_rect is not used to offset what we normally draw, but instead generally for scissoring(?). The end result is if we use an offset screen_rect, we end up applying the offset of the chrome area twice, once because the document's screen rect is offset, and once because of the tile.world_rect offset. Depends on D20696 Differential Revision: https://phabricator.services.mozilla.com/D20698
deb682690025: Bug 1441308 - Ensure lazy init of debug renderer in draw_frame_debug_items r=gw
Doug Thayer <dothayer@mozilla.com> - Tue, 26 Feb 2019 15:29:15 +0000 - rev 461270
Push 112169 by ncsoregi@mozilla.com at Tue, 26 Feb 2019 21:53:47 +0000
Bug 1441308 - Ensure lazy init of debug renderer in draw_frame_debug_items r=gw Document splitting is crashing with early initialization of the debug renderer. Not sure why, and this is just a temporary workaround, but one that I think we want anyway, as we don't want to be unnecessarily lazy-initting the debug renderer. Depends on D20698 Differential Revision: https://phabricator.services.mozilla.com/D20700
8a0710c29415: Bug 1530576 - WR track vao bindings better r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Tue, 26 Feb 2019 06:30:44 +0000 - rev 461249
Push 112169 by ncsoregi@mozilla.com at Tue, 26 Feb 2019 21:53:47 +0000
Bug 1530576 - WR track vao bindings better r=gw this change makes sure that bindings are always synced. Differential Revision: https://phabricator.services.mozilla.com/D21159
55ba29bf61f0: Bug 1505871. Work around a suspected shader miscompilation on Windows. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Feb 2019 00:16:37 -0600 - rev 461076
Push 112149 by tnikkel@gmail.com at Tue, 26 Feb 2019 06:17:00 +0000
Bug 1505871. Work around a suspected shader miscompilation on Windows. r=gw On Windows the vFuncs array is always 0 in the fragment shader. If we move the computation of the vFuncs array outside of the switch (so that it is computed for every type of shader, even when it is not needed) then it works.
1562b43107f2: Bug 1505871. Implementation of shaders. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Feb 2019 00:16:37 -0600 - rev 461075
Push 112149 by tnikkel@gmail.com at Tue, 26 Feb 2019 06:17:00 +0000
Bug 1505871. Implementation of shaders. r=gw
457a27c7b181: Bug 1505871. Implement putting the required data in the gpu cache for component transfer. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Feb 2019 00:16:37 -0600 - rev 461074
Push 112149 by tnikkel@gmail.com at Tue, 26 Feb 2019 06:17:00 +0000
Bug 1505871. Implement putting the required data in the gpu cache for component transfer. r=gw For table/discrete we just create a lookup table for all 256 possible input values. We should probably switch to just computing the value in the shader, unless the list of value is really long.
461c14f8a14c: Bug 1505871. Pass DataStores to prepare_for_render. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Feb 2019 00:16:37 -0600 - rev 461073
Push 112149 by tnikkel@gmail.com at Tue, 26 Feb 2019 06:17:00 +0000
Bug 1505871. Pass DataStores to prepare_for_render. r=gw We need this to get get filter data from the intern table.
e1f278d901a7: Bug 1505871. Don't borrow frame_state.surface for the duration of prepare_for_render. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Feb 2019 00:16:36 -0600 - rev 461072
Push 112149 by tnikkel@gmail.com at Tue, 26 Feb 2019 06:17:00 +0000
Bug 1505871. Don't borrow frame_state.surface for the duration of prepare_for_render. r=gw We need to call update on the filter data handle which needs to use frame_state (in a later patch), so we can't borrow across that call.
30cb5423d03f: Bug 1505871. Implement the necessary things for interning of filter data. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Tue, 26 Feb 2019 00:16:36 -0600 - rev 461071
Push 112149 by tnikkel@gmail.com at Tue, 26 Feb 2019 06:17:00 +0000
Bug 1505871. Implement the necessary things for interning of filter data. r=gw
1206e1a32e5f: Bug 1505871. Don't borrow frame_state.surface for the duration of prepare_for_render. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 22:45:16 -0600 - rev 461062
Push 112147 by tnikkel@gmail.com at Tue, 26 Feb 2019 04:45:34 +0000
Bug 1505871. Don't borrow frame_state.surface for the duration of prepare_for_render. r=gw We need to call update on the filter data handle which needs to use frame_state (in a later patch), so we can't borrow across that call.
11c9d5a64a30: Bug 1505871. Implement the necessary things for interning of filter data. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 22:45:16 -0600 - rev 461061
Push 112147 by tnikkel@gmail.com at Tue, 26 Feb 2019 04:45:34 +0000
Bug 1505871. Implement the necessary things for interning of filter data. r=gw
6486435a048d: Bug 1505871. Work around a suspected shader miscompilation on Windows. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 19:20:28 -0600 - rev 460925
Push 112142 by tnikkel@gmail.com at Tue, 26 Feb 2019 01:20:51 +0000
Bug 1505871. Work around a suspected shader miscompilation on Windows. r=gw On Windows the vFuncs array is always 0 in the fragment shader. If we move the computation of the vFuncs array outside of the switch (so that it is computed for every type of shader, even when it is not needed) then it works.
9be871042749: Bug 1505871. Implementation of shaders. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 19:20:28 -0600 - rev 460924
Push 112142 by tnikkel@gmail.com at Tue, 26 Feb 2019 01:20:51 +0000
Bug 1505871. Implementation of shaders. r=gw
0007feaf988d: Bug 1505871. Implement putting the required data in the gpu cache for component transfer. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 19:20:28 -0600 - rev 460923
Push 112142 by tnikkel@gmail.com at Tue, 26 Feb 2019 01:20:51 +0000
Bug 1505871. Implement putting the required data in the gpu cache for component transfer. r=gw For table/discrete we just create a lookup table for all 256 possible input values. We should probably switch to just computing the value in the shader, unless the list of value is really long.
3cb8fb01e77e: Bug 1505871. Pass DataStores to prepare_for_render. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 19:20:28 -0600 - rev 460922
Push 112142 by tnikkel@gmail.com at Tue, 26 Feb 2019 01:20:51 +0000
Bug 1505871. Pass DataStores to prepare_for_render. r=gw We need this to get get filter data from the intern table.
2fff213d97e3: Bug 1505871. Don't borrow frame_state.surface for the duration of prepare_for_render. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 19:20:27 -0600 - rev 460921
Push 112142 by tnikkel@gmail.com at Tue, 26 Feb 2019 01:20:51 +0000
Bug 1505871. Don't borrow frame_state.surface for the duration of prepare_for_render. r=gw We need to call update on the filter data handle which needs to use frame_state (in a later patch), so we can't borrow across that call.
1ad20d485eca: Bug 1505871. Implement the necessary things for interning of filter data. r=gw
Timothy Nikkel <tnikkel@gmail.com> - Mon, 25 Feb 2019 19:20:27 -0600 - rev 460920
Push 112142 by tnikkel@gmail.com at Tue, 26 Feb 2019 01:20:51 +0000
Bug 1505871. Implement the necessary things for interning of filter data. r=gw
9bae35dc1471: Bug 1519718 - WR mix-blend rewrite r=gw
Dzmitry Malyshau <dmalyshau@mozilla.com> - Mon, 25 Feb 2019 00:17:22 +0000 - rev 460844
Push 112120 by opoprus@mozilla.com at Mon, 25 Feb 2019 10:33:46 +0000
Bug 1519718 - WR mix-blend rewrite r=gw This is a new implementation of mix-blend compositing that is meant to be more idiomatic to WR and efficient. Previously, mix-blend mode was composed in the following way: 1. parent stacking context was forced to isolate 2. source picture is also isolated 3. when rendering the isolated context, the framebuffer is read upon reaching the source. Both the readback and the source are placed in the RT cache. 4. a mix-blend draw call is issued to read from those cache segments and blend on top of the backdrop The new implementation works by using the picture cutting (intruduced for preserve-3D contexts earlier) and some bits of magic: 1. backdrop stacking context is isolated with a special composition mode that prevents it from actually rendeing unless the suorce stacking context is invisible. 2. source stacking context is isolated with mix-blend composition mode that has a pointer to the backdrop picture 3. the instance of the backdrop picture is placed as a peer of the source picture (not a child) 4. if the backdrop is invisible, the source is drawn as a simple blit 5. otherwise, it's a draw call that reads from the isolated backdrop and source textures Note the differences: - parent stacking context is not isolated, but backdrop is - no framebuffer readback is involved - the source and backdrop pictures are rendered in parallel in a pass, improving the batching - we don't blend onto the backdrop while reading from the backdrop copy at the same time - the depth of the render pass tree is reduced: previously the parent and the source were isolated, now the source and the backdrop, which are siblings Differential Revision: https://phabricator.services.mozilla.com/D20608