Commit Graph

567 Commits

Author SHA1 Message Date
Herb Derby
a08bde606a GrTextContext::Options -> GrSDFTOptions
Rename GrTextContext::Options to GrSDFTOptions. Remove GrTextContext.

TBR=reed@google.com

Change-Id: Ic89966cfe630cc6ee757977a6d63ae1da9e39ef0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/296176
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
2020-06-15 17:14:04 +00:00
Greg Daniel
01f278c834 Better handle backendFormat to CompressionType.
We also fix getReadSwizzle with this change.

Change-Id: I1989d8347dc97d7a4c75aa9094a0146419c6d8fa
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/295819
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
2020-06-12 21:47:14 +00:00
Adlai Holler
1bdbe55135 Cache last-accessed entry in lastRenderTask table.
This table is really hot during drawing and we saw a regression when
it was introduced. This change provides wins locally but I don't have
the regressed hardware handy.

Additionally, we change from SkGoodHash to the MixCheap function used by
some other small tables in the library.

Bug: skia:10372
Change-Id: Ic2dfbf41649515245e82cbcc9765cb29270eb5cf
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/295878
Commit-Queue: Adlai Holler <adlai@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
2020-06-12 16:11:19 +00:00
Chris Dalton
f26e596822 Duplicate ccpr/GrCCStrokeGeometry to tessellate/GrStrokeGeometry
CCPR stroking is not enabled by default and will eventually be
removed. This CL is a step toward migrating it to an MSAA version.

Change-Id: I565bea1c91c51bcd1ea3766aee6e4702adf3b09e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/295608
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-06-11 21:59:23 +00:00
Jim Van Verth
f43da14878 Add D3D descriptor tables and bind to command list.
This CL adds a new descriptor heap manager that handles creation of
descriptor tables, or ranges of shader-visible descriptors. These
are used to set the descriptors that match the setup of the root
signature. Data from non-shader-visible descriptors are copied into
these tables, and then they are bound to the command list.

Change-Id: Ia9fc24cae602244525a196025c044df3c2435d9f
Bug: skia:9935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/294300
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-06-10 01:05:03 +00:00
John Stiles
fe64410954 Remove PremulInput and GrPremulInputFragmentProcessor.fp from Skia.
This FP was unused and not referenced anywhere in the codebase.

Change-Id: I27f963d7c26382101b585f356a75d84cef5ab237
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/295080
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
2020-06-08 21:04:46 +00:00
Michael Ludwig
9cced93b7a Handle color and stencil clears in GrClearOp
With prior changes, it became "clear" that GrClearOp and
GrClearStencilClipOp behaved very similarly, except that the stencil
op did not have any onCombine logic.

This just combines them in to a single clear op that will call the
render pass's clear and stencil clear functions as needed. I also
implemented combine logic to apply color and stencil clears in a single
op if the scissor state was compatible (although there's no render pass
API to combine the two clears into a single GPU function).

Change-Id: I8aa749fe64cc487d187854fd0acf6b03b86f1356
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/290822
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
2020-06-08 14:05:16 +00:00
Herb Derby
64aa5138c7 Remove SkAtlasTextTarget
I can find no references to SkAtlasTextTarget in AOSP or in Chromium.
With google3 CL/314226466 there are no more uses in Google3.

Change-Id: I60b5f06fc17c0e4f8d008886c96645475e3d48e5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/293839
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Herb Derby <herb@google.com>
2020-06-04 03:52:35 +00:00
Jim Van Verth
3eadce263c Create ring buffer for managing D3D uniforms.
* Adds a base class for the ring buffer (to be used by Metal as well),
  which tracks the current available space. APIs will need to
  implement creation of the buffer in the subclass.
* The API implementation will need to store SubmitData on command buffer
  submit, and then pass it to finishSubmit when the command buffer
  finishes.

Change-Id: I4cc5e4a72d259ee9d15dac0e964819d4562da3d7
Bug: skia:9935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291936
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-06-01 20:38:40 +00:00
Brian Osman
273944ac7c Convert GrMatrixEffect to pure CPP/H
Prelude to adding situational optimizations

Change-Id: Ie7994a87380ba6135398c9a1d4b26cea6bf32fa8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/292972
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
2020-06-01 14:33:57 +00:00
Greg Daniel
6c6caf420a Add GrBackendMutableState object to handle shared texture state.
This is will be the main struct used to synchronize changes of certain
texture/image between clients and Skia. With this change we
implement support for the Vulkan shared state as POC.

Bug: skia:10254
Change-Id: I10543357635c347838b193874e4da4496a0dcf06
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/292311
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
2020-05-29 16:53:08 +00:00
Jim Van Verth
afd4113bc3 Change D3DAttachmentViewManager to handle all CPU descriptors
Change-Id: I6af3673a9dedf0a5acfbd588bfbbb447b5c68013
Bug: skia:9935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/292576
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-05-28 15:34:59 +00:00
Chris Dalton
b5391d98b8 Reland "Reland "Add a tessellation mode that uses indirect draws""
This is a reland of 477512c0b3

Original change's description:
> Reland "Add a tessellation mode that uses indirect draws"
>
> This is a reland of 02d7cf79be
>
> Original change's description:
> > Add a tessellation mode that uses indirect draws
> >
> > This mode is oftentimes faster than tessellation, and other times it
> > serves as a polyfill when tessellation just isn't supported.
> >
> > Change-Id: I7b3d57fd0194c6869bfe28ee53ff0ff2e43df479
> > Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291036
> > Commit-Queue: Chris Dalton <csmartdalton@google.com>
> > Reviewed-by: Brian Osman <brianosman@google.com>
>
> TBR=brianosman@google.com
>
> Change-Id: Ia039d7897499a7dad55aff1072f9aa0d01001f9c
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291693
> Reviewed-by: Chris Dalton <csmartdalton@google.com>
> Commit-Queue: Chris Dalton <csmartdalton@google.com>

TBR=brianosman@google.com

Change-Id: I55704b5adf5a7677a5382d07d39cadee9b252731
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291738
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-05-25 02:26:58 +00:00
Chris Dalton
0e6d4f1bcf Revert "Reland "Add a tessellation mode that uses indirect draws""
This reverts commit 477512c0b3.

Reason for revert: Undefined behavior in benchmark

Original change's description:
> Reland "Add a tessellation mode that uses indirect draws"
> 
> This is a reland of 02d7cf79be
> 
> Original change's description:
> > Add a tessellation mode that uses indirect draws
> >
> > This mode is oftentimes faster than tessellation, and other times it
> > serves as a polyfill when tessellation just isn't supported.
> >
> > Change-Id: I7b3d57fd0194c6869bfe28ee53ff0ff2e43df479
> > Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291036
> > Commit-Queue: Chris Dalton <csmartdalton@google.com>
> > Reviewed-by: Brian Osman <brianosman@google.com>
> 
> TBR=brianosman@google.com
> 
> Change-Id: Ia039d7897499a7dad55aff1072f9aa0d01001f9c
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291693
> Reviewed-by: Chris Dalton <csmartdalton@google.com>
> Commit-Queue: Chris Dalton <csmartdalton@google.com>

TBR=egdaniel@google.com,brianosman@google.com,csmartdalton@google.com

Change-Id: I48e8cbff90f5533ba3e2166c819fb8e34bea34ca
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291737
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-05-24 20:49:21 +00:00
Chris Dalton
477512c0b3 Reland "Add a tessellation mode that uses indirect draws"
This is a reland of 02d7cf79be

Original change's description:
> Add a tessellation mode that uses indirect draws
>
> This mode is oftentimes faster than tessellation, and other times it
> serves as a polyfill when tessellation just isn't supported.
>
> Change-Id: I7b3d57fd0194c6869bfe28ee53ff0ff2e43df479
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291036
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Brian Osman <brianosman@google.com>

TBR=brianosman@google.com

Change-Id: Ia039d7897499a7dad55aff1072f9aa0d01001f9c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291693
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-05-24 20:40:19 +00:00
Mike Klein
ec31488ace Revert "Add a tessellation mode that uses indirect draws"
This reverts commit 02d7cf79be.

Reason for revert:

  Perf-Debian10-Clang-GCE-CPU-AVX2-x86_64-Debug-All-ASAN
  Perf-Win2019-Clang-GCE-CPU-AVX2-x86_64-Debug-All-ASAN

Running tessellate_prepareTessellatedCubicWedges	nonrendering
../../../../../../skia/bench/TessellatePathBench.cpp:79: fatal error: "unimplemented."

Signal 5 [Trace/breakpoint trap]:
    /mnt/pd0/s/w/ir/build/nanobench(backtrace+0x3d) [0x34851fd]
    /mnt/pd0/s/w/ir/build/nanobench() [0x3db3d49]
    /lib/x86_64-linux-gnu/libc.so.6(+0x37840) [0x7fdb4722d840]
    sk_abort_no_print()
    BenchmarkTarget::caps() const
    GrTessellatePathOp::prepareTessellatedCubicWedges(GrMeshDrawOp::Target*)
    GrTessellatePathOp::TestingOnly_Benchmark::prepareTessellatedCubicWedges::runBench(GrMeshDrawOp::Target*, GrTessellatePathOp*)
    GrTessellatePathOp::TestingOnly_Benchmark::onDraw(int, SkCanvas*)
    Benchmark::draw(int, SkCanvas*)
    /mnt/pd0/s/w/ir/build/nanobench() [0x3543b81]
    /mnt/pd0/s/w/ir/build/nanobench(main+0x1347) [0x3540a17]
    /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xeb) [0x7fdb4721a09b]
    /mnt/pd0/s/w/ir/build/nanobench(_start+0x2a) [0x346429a]
Command exited with code 5


Original change's description:
> Add a tessellation mode that uses indirect draws
> 
> This mode is oftentimes faster than tessellation, and other times it
> serves as a polyfill when tessellation just isn't supported.
> 
> Change-Id: I7b3d57fd0194c6869bfe28ee53ff0ff2e43df479
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291036
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Brian Osman <brianosman@google.com>

TBR=egdaniel@google.com,brianosman@google.com,csmartdalton@google.com

Change-Id: I4257d5ae5c0aafeebd61d8f1b78eca755ad977ff
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291642
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
2020-05-22 13:46:16 +00:00
Chris Dalton
02d7cf79be Add a tessellation mode that uses indirect draws
This mode is oftentimes faster than tessellation, and other times it
serves as a polyfill when tessellation just isn't supported.

Change-Id: I7b3d57fd0194c6869bfe28ee53ff0ff2e43df479
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/291036
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2020-05-21 22:32:45 +00:00
Chris Dalton
e5a141d0f6 Reland GrRectanizerPow2
This brings back the pow2 rectanizer for use with various ongoing
atlas experiments. If we can further optimize the skyline rectanizer,
then pow2 will be good to have around as a baseline comparison. And
if skyline gets fast enough, then we can delete pow2 again.

Change-Id: I79088c53fba7ba0d120534af99bee7840c135e42
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/290810
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-05-19 18:35:13 +00:00
Jim Van Verth
be259dcce0 Add GrD3DPipelineStateDataManager.
* Creates a common class between GrVkPipelineStateDataManager and
GrD3DPipelineStateDataManager so they can share code (will do
for Metal as well in separate CL)
* Adds means for tracking and setting uniform data.

Change-Id: Ie0dc3a3d26f533201e316d255965a646bcecb842
Bug: skia:9935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/290636
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-05-19 17:33:32 +00:00
Chris Dalton
f6bf516926 Reland "Add an implementation and log2 variants for Wang's formula"
This is a reland of e278e1c1c7

Original change's description:
> Add an implementation and log2 variants for Wang's formula
>
> Wang's formulas for cubics and quadratics (1985) tell us how many line
> segments a curve must be chopped into when tessellating. This CL adds
> an implementation along with optimized log2 variants, as well as tests
> and a benchmark.
>
> Change-Id: I3f777b8d0312c57c3a1cc24307de5945c70be287
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/288321
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Chris Dalton <csmartdalton@google.com>

TBR=bsalomon@google.com

Change-Id: Ie3822c62439fc579a59ea8adb49583224de41aa5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/289680
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-05-14 02:57:59 +00:00
Mike Klein
6c3db04c8b Revert "Add an implementation and log2 variants for Wang's formula"
This reverts commit e278e1c1c7.

Reason for revert: i think we need to do that add with an unsigned, or test instead of always += (1<<23)-1.

Original change's description:
> Add an implementation and log2 variants for Wang's formula
> 
> Wang's formulas for cubics and quadratics (1985) tell us how many line
> segments a curve must be chopped into when tessellating. This CL adds
> an implementation along with optimized log2 variants, as well as tests
> and a benchmark.
> 
> Change-Id: I3f777b8d0312c57c3a1cc24307de5945c70be287
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/288321
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Chris Dalton <csmartdalton@google.com>

TBR=bsalomon@google.com,brianosman@google.com,csmartdalton@google.com

Change-Id: I24dfd8549054b632f38f7b05b4d857b640cf5cd1
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/289658
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
2020-05-13 22:10:47 +00:00
Chris Dalton
e278e1c1c7 Add an implementation and log2 variants for Wang's formula
Wang's formulas for cubics and quadratics (1985) tell us how many line
segments a curve must be chopped into when tessellating. This CL adds
an implementation along with optimized log2 variants, as well as tests
and a benchmark.

Change-Id: I3f777b8d0312c57c3a1cc24307de5945c70be287
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/288321
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-05-13 21:39:24 +00:00
Michael Ludwig
828d341199 Reland "Refactor stencil clip mask generation into helper"
This reverts commit de228e53fe.

Reason for revert: GrReducedClip now assumes context isn't abandoned,
windowrectangles GM abuses GrReducedClip and has to be abandon-aware.

Original change's description:
> Revert "Refactor stencil clip mask generation into helper"
>
> This reverts commit 8b3a8a5238.
>
> Reason for revert: GM assert failure
>
> Original change's description:
> > Refactor stencil clip mask generation into helper
> >
> > Change-Id: If3dc80efde3b44e87ba8e7af3a258896ec5e78e6
> > Reviewed-on: https://skia-review.googlesource.com/c/skia/+/288977
> > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> > Reviewed-by: Chris Dalton <csmartdalton@google.com>
> > Reviewed-by: Brian Salomon <bsalomon@google.com>
>
> TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com
>
> Change-Id: I16559f791601145f57d147cdae345c200af313f1
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/289237
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>

# Not skipping CQ checks because this is a reland.

Change-Id: I6a9372edecd0bdc1a38464ab85f7b7f3ca85e5ed
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/289239
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-05-12 17:49:47 +00:00
Michael Ludwig
de228e53fe Revert "Refactor stencil clip mask generation into helper"
This reverts commit 8b3a8a5238.

Reason for revert: GM assert failure

Original change's description:
> Refactor stencil clip mask generation into helper
> 
> Change-Id: If3dc80efde3b44e87ba8e7af3a258896ec5e78e6
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/288977
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> Reviewed-by: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Brian Salomon <bsalomon@google.com>

TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com

Change-Id: I16559f791601145f57d147cdae345c200af313f1
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/289237
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-05-12 16:17:35 +00:00
Michael Ludwig
8b3a8a5238 Refactor stencil clip mask generation into helper
Change-Id: If3dc80efde3b44e87ba8e7af3a258896ec5e78e6
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/288977
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
2020-05-12 15:27:18 +00:00
Jim Van Verth
7359165e66 Automatically grow D3DRenderTargetView heap as needed.
Change-Id: I55cf2daa48ec694fc9e1939e270f55dd7a3162a7
Bug: skia:9935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/287619
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-05-08 17:49:50 +00:00
Jim Van Verth
2ba8ba26f9 Add GrD3DDescriptorHeap.
Adds a wrapper around ID3D12DescriptorHeap, which manages allocations
of descriptors from the heap.

Change-Id: Idc3bdb43640639114de5d0520c339f9e0173e26f
Bug: skia:9935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/286338
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-04-30 21:34:56 +00:00
Michael Ludwig
f38b711222 Reland "Refactor geometry union capabilities out of GrStyledShape"
This reverts commit af312c9d40.

Reason for revert: improved performance, updated empty point cap behavior
to make chrome happy.

Because of the performance regression in the original CL, this is a bit
more to it than just updating cap behavior. Summary of changes for perf:
1. In asPath(), only call reset() if the type isn't a path or arc.
   Otherwise it was just a wasted realloc of an empty path ref.
2. Rewrote the GrShape::simplify() to not progress through every shape
   type in order, it just jumps to the appropriate type.
3. Have simplify() return whether or not the shape started out closed,
   so we don't have to call GrShape::closed(), which is costly when the
   shape is a path.
4. Expose the GrShape's type enum so GrStyledShape's key writing can use
   switches instead of a giant block of ifs (where path happened to be
   last)

The regressions showed up most heavily on desk_mapsvg and desk_chalkboard
SKPs on the Android skpbench marks. On my system, I was able to
reproduce a similar %-regression from ToT and the original CL on the
chalkboard (but not mapsvg).

Master ranged between 5.1 and 5.3ms, original CL ranged from 5.6-5.8
and after the changes listed above, I got it down to 5.3-5.5. It's not
ideal but I haven't been able to figure out anything more substantial
that it could be. At this point it may just be code layout and/or the
fact that it's now split into two types.


Original change's description:
> Revert "Refactor geometry union capabilities out of GrStyledShape"
>
> This reverts commit 2becdde074.
>
> Reason for revert: likely breaking cc unit test due to empty shape cap change.
>
> Original change's description:
> > Refactor geometry union capabilities out of GrStyledShape
> >
> > The geometry union part of GrStyledShape is now held in GrShape. For the
> > most part, GrShape is entirely style agnostic and focuses on storing
> > the various types of geometry, and destructing them gracefully. It also
> > provides a public API that unifies functionality across all shape types,
> > such as contains() and bounds().
> >
> > GrStyledShape now just owns a GrShape and a GrStyle, and handles the
> > additional simplification logic that relies on knowing the effects of
> > the style on the draw. This is where GrShape makes some allowances for
> > style. Its simplify() function accepts flags that enable/disable various
> > simplification optimizations. Currently these are designed around
> > what is needed to respect path effects and stroking behaviors in
> > GrStyledShape. The main other user of GrShape (the new clip stack) will
> > always provide all flags since it treats every shape as if it were
> > simply filled.
> >
> > Several other related refactorings were taken at the same time:
> > 1. The implementations for asNestedRects, asRRect, etc. were moved out
> >    of the header and into the cpp file for GrStyledShape.
> > 2. GrRenderTargetContext relies on GrStyledShape for its stroke rect
> >    fallbacks.
> > 3. GrShape can hold points, lines, and rects explicitly. This let me
> >    simplify the stroke reasoning.
> >
> > Change-Id: I9fe75613fee51c30b4049b2b5a422daf80a1a86e
> > Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284803
> > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> > Reviewed-by: Brian Salomon <bsalomon@google.com>
> > Reviewed-by: Chris Dalton <csmartdalton@google.com>
>
> TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com
>
> Change-Id: I2af5782e072e0ccb4a87f903bb88cbe335b9613f
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/286039
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>

Change-Id: I8c614573582084f2e9ee0d73f93812e0a7c13983
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/286396
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-04-30 19:36:43 +00:00
Michael Ludwig
af312c9d40 Revert "Refactor geometry union capabilities out of GrStyledShape"
This reverts commit 2becdde074.

Reason for revert: likely breaking cc unit test due to empty shape cap change.

Original change's description:
> Refactor geometry union capabilities out of GrStyledShape
> 
> The geometry union part of GrStyledShape is now held in GrShape. For the
> most part, GrShape is entirely style agnostic and focuses on storing
> the various types of geometry, and destructing them gracefully. It also
> provides a public API that unifies functionality across all shape types,
> such as contains() and bounds().
> 
> GrStyledShape now just owns a GrShape and a GrStyle, and handles the
> additional simplification logic that relies on knowing the effects of
> the style on the draw. This is where GrShape makes some allowances for
> style. Its simplify() function accepts flags that enable/disable various
> simplification optimizations. Currently these are designed around
> what is needed to respect path effects and stroking behaviors in
> GrStyledShape. The main other user of GrShape (the new clip stack) will
> always provide all flags since it treats every shape as if it were
> simply filled.
> 
> Several other related refactorings were taken at the same time:
> 1. The implementations for asNestedRects, asRRect, etc. were moved out
>    of the header and into the cpp file for GrStyledShape.
> 2. GrRenderTargetContext relies on GrStyledShape for its stroke rect
>    fallbacks.
> 3. GrShape can hold points, lines, and rects explicitly. This let me
>    simplify the stroke reasoning.
> 
> Change-Id: I9fe75613fee51c30b4049b2b5a422daf80a1a86e
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284803
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Reviewed-by: Chris Dalton <csmartdalton@google.com>

TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com

Change-Id: I2af5782e072e0ccb4a87f903bb88cbe335b9613f
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/286039
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-04-29 13:32:34 +00:00
Michael Ludwig
2becdde074 Refactor geometry union capabilities out of GrStyledShape
The geometry union part of GrStyledShape is now held in GrShape. For the
most part, GrShape is entirely style agnostic and focuses on storing
the various types of geometry, and destructing them gracefully. It also
provides a public API that unifies functionality across all shape types,
such as contains() and bounds().

GrStyledShape now just owns a GrShape and a GrStyle, and handles the
additional simplification logic that relies on knowing the effects of
the style on the draw. This is where GrShape makes some allowances for
style. Its simplify() function accepts flags that enable/disable various
simplification optimizations. Currently these are designed around
what is needed to respect path effects and stroking behaviors in
GrStyledShape. The main other user of GrShape (the new clip stack) will
always provide all flags since it treats every shape as if it were
simply filled.

Several other related refactorings were taken at the same time:
1. The implementations for asNestedRects, asRRect, etc. were moved out
   of the header and into the cpp file for GrStyledShape.
2. GrRenderTargetContext relies on GrStyledShape for its stroke rect
   fallbacks.
3. GrShape can hold points, lines, and rects explicitly. This let me
   simplify the stroke reasoning.

Change-Id: I9fe75613fee51c30b4049b2b5a422daf80a1a86e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284803
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
2020-04-28 19:17:08 +00:00
Chris Dalton
f5132a05c8 Reland "Optimize GrTessellatePathOp's code to emit inner triangles"
This is a reland of 0f9ebd100e

Original change's description:
> Optimize GrTessellatePathOp's code to emit inner triangles
>
> Previously we used a naive algorithm to generate "middle-out" topologies
> for inner polygons, including copying all endpoints to a new array.
>
> This CL adds a "GrMiddleOutPolygonTriangulator" class that
> accomplishes the same thing in 1/5th the time using a small O(log N)
> stack.
>
> Change-Id: I3a7059e5d133a730b7084a17d8fbaaa3aaa81336
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/285531
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>

TBR=ethannicholas@google.com

Change-Id: I8402e5455e14aef9c5e92ee21100fcd7ace222ff
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/285740
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
2020-04-28 06:09:59 +00:00
Ethan Nicholas
5f56cb1d3b Revert "Optimize GrTessellatePathOp's code to emit inner triangles"
This reverts commit 0f9ebd100e.

Reason for revert: breaking a couple of bots (e.g. https://chromium-swarm.appspot.com/task?id=4bd6ccc785fa3110)

Original change's description:
> Optimize GrTessellatePathOp's code to emit inner triangles
> 
> Previously we used a naive algorithm to generate "middle-out" topologies
> for inner polygons, including copying all endpoints to a new array.
> 
> This CL adds a "GrMiddleOutPolygonTriangulator" class that
> accomplishes the same thing in 1/5th the time using a small O(log N)
> stack.
> 
> Change-Id: I3a7059e5d133a730b7084a17d8fbaaa3aaa81336
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/285531
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>

TBR=csmartdalton@google.com,michaelludwig@google.com

Change-Id: I8ebca7078f2b9c12246447759efa9ce0cbb7e46b
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/285719
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
2020-04-27 23:38:10 +00:00
Chris Dalton
0f9ebd100e Optimize GrTessellatePathOp's code to emit inner triangles
Previously we used a naive algorithm to generate "middle-out" topologies
for inner polygons, including copying all endpoints to a new array.

This CL adds a "GrMiddleOutPolygonTriangulator" class that
accomplishes the same thing in 1/5th the time using a small O(log N)
stack.

Change-Id: I3a7059e5d133a730b7084a17d8fbaaa3aaa81336
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/285531
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
2020-04-27 20:33:09 +00:00
Greg Daniel
d59a91df4d Rename GrSpirv* files to GrSPIRV.
Change-Id: I4a4fc5dbdf6bba78e6ba5d60478b3e109b685a67
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/285098
Reviewed-by: Stephen White <senorblanco@chromium.org>
Commit-Queue: Greg Daniel <egdaniel@google.com>
2020-04-24 19:24:10 +00:00
Chris Dalton
42915c2cee Remove tessellate/GrPathParser
Moves this logic to happen directly ion GrTessellatePathOp instead.

Change-Id: Ia972b8c686bdf8efb8e414fbddc856401fe44920
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284999
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
2020-04-23 16:05:29 +00:00
Greg Daniel
c31edc0b38 Add GrD3DRootSignature class and support.
This is the last thing needed for creating the pipeline state so this
change also starts creating the pipeline state and fixes errors involving
it.

Change-Id: Ifd1b63016d692d98cfa2a931957e857a5c53bec5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284527
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
2020-04-23 15:34:58 +00:00
Greg Daniel
5fc5c8128c Add GrD3DPipelineStateBuilder class.
This allows us to create the binary shaders for d3d. We generate spirv
from sksl, then use spirv-cross to turn it into hlsl. Then that gets
compiled into binary to be used in the pipeline.

Adds hooks GrD3DOpsRenderPass to start creating the pipeline.

Change-Id: Ie731dd945cdd9a00cebd78c1371a3d9784e4e1a3
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284526
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
2020-04-23 15:01:58 +00:00
Brian Salomon
2eb7b57d14 Remove unused GrTextureDomain
Bug: skia:10139

Change-Id: I20cd95fcf5f11832366c32e48ed4d442c82b0719
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284082
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
2020-04-21 18:51:24 +00:00
Greg Daniel
690a450cdc Create a shared spirv uniform and varying handling class.
This moves the dawn spirv implementation of these classes into a shared
class that will be used by d3d backend. We can look into further extending
the class to see if it can be shared with vulkan as well.

Change-Id: I138d403dd55053f534d0c97a55c0fa5d2c5171f0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284525
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
2020-04-21 15:25:14 +00:00
Michael Ludwig
2686d69bf0 Reland "Rename GrShape to GrStyledShape"
This reverts commit 73b86c1ade.

Reason for revert: wasn't the problem, relanding

Original change's description:
> Revert "Rename GrShape to GrStyledShape"
> 
> This reverts commit f3f08af010.
> 
> Reason for revert: maybe this is breaking the ios perf bot; it is the first CL that caused the bot to fail, but I can't really say why this would break them.
> 
> Original change's description:
> > Rename GrShape to GrStyledShape
> > 
> > Change-Id: Ic457e634b4b95356f5615cff3fce1ca7d7677c26
> > Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284036
> > Reviewed-by: Robert Phillips <robertphillips@google.com>
> > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> 
> TBR=bsalomon@google.com,robertphillips@google.com,michaelludwig@google.com
> 
> # Not skipping CQ checks because original CL landed > 1 day ago.
> 
> Change-Id: I2e5adbfc820811fbbde9cb57af28f86a7ba40bd9
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284231
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>

TBR=bsalomon@google.com,robertphillips@google.com,michaelludwig@google.com

Change-Id: Ifdd52f8bacb5d66a7bf58efd328675c4c443ac8a
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284376
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-04-17 20:24:53 +00:00
Michael Ludwig
73b86c1ade Revert "Rename GrShape to GrStyledShape"
This reverts commit f3f08af010.

Reason for revert: maybe this is breaking the ios perf bot; it is the first CL that caused the bot to fail, but I can't really say why this would break them.

Original change's description:
> Rename GrShape to GrStyledShape
> 
> Change-Id: Ic457e634b4b95356f5615cff3fce1ca7d7677c26
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284036
> Reviewed-by: Robert Phillips <robertphillips@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>

TBR=bsalomon@google.com,robertphillips@google.com,michaelludwig@google.com

# Not skipping CQ checks because original CL landed > 1 day ago.

Change-Id: I2e5adbfc820811fbbde9cb57af28f86a7ba40bd9
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284231
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-04-17 19:38:34 +00:00
Stephen White
b353c9bdd3 Reland "Refactor GL and Metal FinishCallback into a helper object."
This is a reland of b641ddbba0

Original change's description:
> Refactor GL and Metal FinishCallback into a helper object.
> 
> Change-Id: I52ede0363016459976e453ef71cafbcc6775f22e
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/283496
> Commit-Queue: Stephen White <senorblanco@chromium.org>
> Reviewed-by: Greg Daniel <egdaniel@google.com>

Change-Id: Ia77c05c0358b6ccab1e8c0f0c06212786f415cb1
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284042
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
2020-04-16 19:20:06 +00:00
Michael Ludwig
f3f08af010 Rename GrShape to GrStyledShape
Change-Id: Ic457e634b4b95356f5615cff3fce1ca7d7677c26
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/284036
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-04-16 18:05:51 +00:00
Michael Ludwig
b15cc56858 Revert "Refactor GL and Metal FinishCallback into a helper object."
This reverts commit b641ddbba0.

Reason for revert: likely breaking command buffer bots

Original change's description:
> Refactor GL and Metal FinishCallback into a helper object.
> 
> Change-Id: I52ede0363016459976e453ef71cafbcc6775f22e
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/283496
> Commit-Queue: Stephen White <senorblanco@chromium.org>
> Reviewed-by: Greg Daniel <egdaniel@google.com>

TBR=egdaniel@google.com,senorblanco@chromium.org

Change-Id: Ifc766e065b6a19cb263c4a3d94f4bbe88c4bce16
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/283782
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
2020-04-15 21:54:58 +00:00
Stephen White
b641ddbba0 Refactor GL and Metal FinishCallback into a helper object.
Change-Id: I52ede0363016459976e453ef71cafbcc6775f22e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/283496
Commit-Queue: Stephen White <senorblanco@chromium.org>
Reviewed-by: Greg Daniel <egdaniel@google.com>
2020-04-15 19:59:48 +00:00
Ethan Nicholas
5843012837 Add sample(child, matrix) to SkSL.
This allows fragment processors to sample their children with their
local coordinate system transformed by a matrix.

Change-Id: Ifa848bbd85b939bbc5751fec5cf8f89ee904bf39
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/282590
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
2020-04-14 16:07:34 +00:00
Jim Van Verth
4f51f47d14 Add GrD3DStencilAttachment
Bug: skia:9935
Change-Id: I249b819d3cc1387aecde200907657ffcd67d1833
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/282591
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
2020-04-13 17:20:48 +00:00
Stephen White
f3d5d44ec9 Dawn: staging manager rewrite.
Using many small staging buffers is quite costly for performance, for
example when there are many small texture uploads (such as MotionMark's
"canvas bouncing clipped rects" test).

Instead, the new manager uses buffers of 32K minimum, and returns
suballocations of those.

For now, allocation is simply iterates through the list of staging
buffers and returns the first one with enough remaining space.
This is O(N) in the number of staging buffers, which should be
ok since there are fewer of them. (If it becomes a hot path, we
can optimize it later.)

This also subsumes the UBO-specific staging buffer, since that one
did much the same thing but did not reuse staging buffers.

Change-Id: I4fddee8bb0fc602c49fe552acc327a640bf6917b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/279921
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
2020-04-08 16:11:48 +00:00
Jim Van Verth
d6ad48086d Add GrD3DBuffer.
Supports buffer creation and destruction.

Bug: skia:9935
Change-Id: I6d8c6d67209a8857989a9384818272e7108177ad
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/281577
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
2020-04-03 19:47:56 +00:00
Chris Dalton
10ee0b2401 Merge GrVkVertexBuffer and GrVkIndexBuffer into GrVkMeshBuffer
Change-Id: Icce3f3dce8581ba1ac3817d62c18e93cee616d3c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/281379
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
2020-04-03 15:54:58 +00:00