Commit Graph

610 Commits

Author SHA1 Message Date
Brian Osman
3e4ef49890 Remove SkBlendMode from SkSL
We've lived without this for a long time. It bloats sksl_gpu (inhibits
some inlining!), and if things go according to plan, we'll be removing
enum support from SkSL entirely soon.

Change-Id: If844bbe5fdae41df7930d5c8ea9b832f9dd1b922
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/419099
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
2021-06-16 21:01:30 +00:00
John Stiles
9c19b9f1ff Fix for fuzzer-discovered out-of-bounds array access.
Change-Id: I4280b5710dd8749ba766ba74d7a8886bc4e024bb
Bug: oss-fuzz:35124
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/417200
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
2021-06-10 14:32:10 +00:00
John Stiles
50d0d097c3 Add dest-color support to the Pipeline stage.
Previously, when converting a program, the pipeline stage could assume
that an input to main() of type half4/float4 was always the input color.
This was a good assumption since the only possible inputs were
coordinates, or the input color.

This CL now recognizes that when a second float4 is passed to the main()
function, it should be a SK_DEST_COLOR_BUILTIN. This will let blend
functions pass in two colors.

ProgramToSkVM now takes a dest-color argument as well, but existing call
sites won't reference it (since they aren't for blend functions). I've
just passed the input-color a second time, since the value will never
actually be accessed.

Change-Id: I4214586bda605c6d287aa25b1b099e6ef5ba15a4
Bug: skia:12080
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/417261
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-06-09 22:09:21 +00:00
John Stiles
f7f36ae077 Add ProgramKind::kRuntimeBlend for Runtime Blend functions.
Runtime Blend effects always take two input colors--source and
destination--instead of one. This CL adds a new ProgramKind for blend
effects, a new program module (empty for now), and adds a test to
confirm that the signature for blend functions is checked. Currently
these are only accessible via skslc; there's no Runtime Effect API to
create one and the dest color isn't hooked up to anything.

Change-Id: I5272a811d2d76b878cfdf3429efa78c9c8b3fd97
Bug: skia:12080
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/416798
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-06-08 19:34:39 +00:00
Chris Dalton
155c33c859 Add a workaround for "ldexp(..., -x)" on Mac/Radeon/GLSL
Bug: skia:12076
Change-Id: Ia7aabe2e048a05475d25811af657228ea39d75ba
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/416561
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
2021-06-08 01:39:38 +00:00
Brian Salomon
3e2fe2b007 One function for computing rt-adjust uniform once flip is determined.
Add some comments explaining why different APIs flip the way they do.

Remove flip from Metal code generator and reverse the way flipping
occurs in Metal backend.

Change-Id: Ibec71e8d96c66d4b34b5ce6417056674090f979c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/413236
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
2021-06-02 19:02:06 +00:00
John Stiles
d774558eb1 Implement compile-time optimization for mix().
mix() has many overloads:
$genType mix($genType x, $genType y, $genType a);
$genType mix($genType x, $genType y, float a);
$genHType mix($genHType x, $genHType y, $genHType a);
$genHType mix($genHType x, $genHType y, half a);
$genType mix($genType x, $genType y, $genBType a);
$genHType mix($genHType x, $genHType y, $genBType a);
$genIType mix($genIType x, $genIType y, $genBType a);
$genBType mix($genBType x, $genBType y, $genBType a);

The top half were simple to implement via `evaluate_3_way_intrinsic`.

The bottom half--`x, y, $genBType`--required adding basic support into
`evaluate_n_way_intrinsic_of_type` for mixed argument types, since `x`
and `y` could be of any numeric type, but `a` is always boolean.
Fortunately, this didn't require major changes.

Change-Id: I015471f053c90d5a5c3ac67cc230d0f90950ff60
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/414443
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-06-01 19:36:05 +00:00
John Stiles
598c5e5b07 Implement compile-time optimization for matrixCompMult().
$mat matrixCompMult($mat x, $mat y);
$hmat matrixCompMult($hmat x, $hmat y);

This required some minor changes to `evaluate_n_way_intrinsic_of_type`
to allow the inputs to be matrices instead of vectors. Fortunately, most
of the moving parts were already generic/flexible enough that this just
worked, but some explicit checks for `x.isVector` needed to become
`!x.isScalar()`, and `columns` needed to become `slotCount`.

Change-Id: I1e22ecad37a7e187a7171e1f590a720f07cf9832
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/414436
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
2021-06-01 19:35:23 +00:00
John Stiles
9df48c6ba6 Add test for literals that overflow int64.
This was handled properly, but lacked a test.

Change-Id: I84adc7cb3d37ab85eef945c1e38fc43c6cd8aa01
Bug: skia:10932
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/414437
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-06-01 17:34:23 +00:00
John Stiles
b0d93ccc90 Detect out-of-range integer assignment.
Now that we know the minimum and maximum values of a given integer Type,
we can check for assigment statements or variable initial-values that
exceed those bounds and report it as an error. This check should work on
anything that can be optimized or folded down to an IntLiteral, but
isn't meant to be 100% exhaustive.

Change-Id: I4473b5b003e1b8e3385943ce60e303e95664e8ba
Bug: skia:10932
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/413437
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-06-01 15:53:43 +00:00
John Stiles
c43dc5e038 Implement compile-time optimization for cross().
We need to be careful to distinguish cross(vec3, vec3)--a real built-in
intrinsic--from cross(vec2, vec2)--an inline function in sksl_gpu, but
not actually a legitimate GLSL intrinsic.

Change-Id: I7e78c99dadfcbb637ae55a2503acfb7e591c932e
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/413440
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
2021-06-01 15:51:53 +00:00
John Stiles
b069582d16 Add regression test for oss-fuzz:32156.
The fuzzer found that it could overflow an int via a properly-crafted
constant-fold expression, leading to a UBSAN error. Constant-fold
expressions now guard against overflow (http://review.skia.org/413138)
and UBSAN is no longer triggered.

Change-Id: I07dba41e87bb9ceed37b84ec6b8922defbdc6550
Bug: oss-fuzz:32156, skia:12050
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/413836
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-06-01 15:19:13 +00:00
John Stiles
e4ac484ac0 Implement compile-time optimization for atan().
This is one of the few compile-time constant functions with a variable
number of arguments, but it's otherwise pretty normal as intrinsics go.

$genType  atan($genType  y, $genType  x);
$genHType atan($genHType y, $genHType x);
$genType  atan($genType  y_over_x);
$genHType atan($genHType y_over_x);

Change-Id: Ie852e10f37d73d53f69e806550872bc015f802d6
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/413439
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
2021-06-01 15:16:43 +00:00
John Stiles
35981296a8 Prevent overflow of integral types during constant-folding.
Expressions which would overflow/wrap the result type are now left
as-is.

Change-Id: I6a942f337e6e5761823f5c9dcd214fa58227a626
Bug: skia:10932, skia:12050
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/413138
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-27 20:12:57 +00:00
John Stiles
b01c18625e Implement compile-time optimization for inverse().
$squareMat inverse($squareMat m);
$squareHMat inverse($squareHMat m);

Change-Id: I1a2b067dd276bb999107712c38d0124811b95e39
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412937
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-27 13:57:00 +00:00
John Stiles
afaddc959c Prevent overflow to infinity in constant-folding.
Expressions which would optimize to a non-finite value are now left
as-is.

Integer math still has problems with overflow, but that can be addressed
separately.

Change-Id: I363a2c42684989062f606186f48246b7ac5b585c
Bug: skia:12050
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412956
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-27 13:23:40 +00:00
John Stiles
b35fbdbb06 Add test demonstrating overflow error in SkSL.
Change-Id: Ic6a3ce722c607d4d3c47d37e5749a01ae5fb193f
Bug: skia:12050
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412668
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 21:52:31 +00:00
John Stiles
f47cfa8be5 Implement compile-time optimization for refract().
$genType refract($genType I, $genType N, float eta);
$genHType refract($genHType I, $genHType N, half eta);

The half form of refract was originally taking a `float eta` in our
headers, which seems wrong (and causes the DSL to break unless you add
casts). I've corrected the headers to use `half eta`.

Change-Id: I74b9ac330e0f7e99622d19cf7365aaa4cc910e57
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412664
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 19:55:39 +00:00
John Stiles
c3ff098020 Implement compile-time optimization for pow().
Change-Id: I1630cf6ecab6a1a18a7318c247f8263d87e2cda9
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412662
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
2021-05-26 19:54:09 +00:00
John Stiles
4ceccaf053 Implement compile-time optimization for reflect().
$genType  reflect($genType  I, $genType  N);
$genHType reflect($genHType I, $genHType N);

Change-Id: I59889ad767829bf7e33838737d7b7f6d1cbc3ece
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412777
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 19:49:29 +00:00
John Stiles
7cfa1ce9db Implement compile-time optimization for faceforward().
This implementation leans heavily into DSL. I also reworked the
`normalize` intrinsic to use more DSL to shrink the implementation.

$genType faceforward($genType N, $genType I, $genType Nref);
$genHType faceforward($genHType N, $genHType I, $genHType Nref);

Change-Id: I73ab11d3fe449d2f2c0ae0d745fc39824fc64771
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412637
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 16:28:48 +00:00
John Stiles
3ef2c61c50 Implement compile-time optimization for normalize().
This one is structured a bit differently; it gets to length() value,
then divides the input by its length using a bit of DSL. Since all the
inputs are constant, the constant-folder will do the right thing.

$genType normalize($genType x);
$genHType normalize($genHType x);

Change-Id: I51e5c65fa9e33738cbe253fcc97ee2160c48cfdd
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412340
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 15:06:08 +00:00
John Stiles
419d2ced2d Implement compile-time optimization for dot().
float dot($genType x, $genType y);
half dot($genHType x, $genHType y);

Change-Id: Ia8b893eaddd8b4325c34143c80fb0ae05b111666
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412339
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 15:05:08 +00:00
John Stiles
e1c2beb3be Implement compile-time optimization for distance().
float distance($genType p0, $genType p1);
half distance($genHType p0, $genHType p1);

Change-Id: I492a90c0fe5c950c878752d45c3b2dc6f2cd866e
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412379
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
2021-05-26 14:11:58 +00:00
John Stiles
9b2baac1d6 Implement compile-time optimization for length().
float length($genType x);
half length($genHType x);

Change-Id: I65b64fdba5f7bd53afba1d6f930217e3f1bd6f6e
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412377
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-26 00:58:53 +00:00
John Stiles
e5240a2498 Implement compile-time optimization for mod().
$genType mod($genType x, float y);
$genType mod($genType x, $genType y);
$genHType mod($genHType x, half y);
$genHType mod($genHType x, $genHType y);

Change-Id: I406eafbe824d7aa55012196521e6f10cac50ed19
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412376
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-25 22:07:53 +00:00
John Stiles
4987c4af49 Implement compile-time optimization for smoothstep().
$genType smoothstep($genType edge0, $genType edge1, $genType x);
$genType smoothstep(float edge0, float edge1, $genType x);
$genHType smoothstep($genHType edge0, $genHType edge1, $genHType x);
$genHType smoothstep(half edge0, half edge1, $genHType x);

Change-Id: Idf6a39fab8d5b81d4a3233444a8b381460cbde71
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412356
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-25 20:32:46 +00:00
John Stiles
017ffd7e05 Implement compile-time optimization for step().
$genType step($genType edge, $genType x);
$genType step(float edge, $genType x);
$genHType step($genHType edge, $genHType x);
$genHType step(half edge, $genHType x);

Change-Id: If52634d7a247772e922cb3fd2e3e8a7c36c275f6
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411842
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-05-25 16:59:22 +00:00
John Stiles
0727a80885 Implement compile-time optimization for clamp().
$genType clamp($genType x, $genType minVal, $genType maxVal);
$genType clamp($genType x, float minVal, float maxVal);
$genHType clamp($genHType x, $genHType minVal, $genHType maxVal);
$genHType clamp($genHType x, half minVal, half maxVal);
$genIType clamp($genIType x, $genIType minVal, $genIType maxVal);
$genIType clamp($genIType x, int minVal, int maxVal);

Change-Id: I726270e84fca781825e59abed0c8c7d099ca7444
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411839
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-25 16:25:42 +00:00
John Stiles
7f17d36c2c Implement compile-time optimization for min() and max().
`evaluate_intrinsic_1_of_type` was rewritten and now supports up to
three arguments. The following APIs are now optimizable:

	$genType min($genType x, $genType y);
	$genType min($genType x, float y);
	$genHType min($genHType x, $genHType y);
	$genHType min($genHType x, half y);
	$genIType min($genIType x, $genIType y);
	$genIType min($genIType x, int y);
	$genType max($genType x, $genType y);
	$genType max($genType x, float y);
	$genHType max($genHType x, $genHType y);
	$genHType max($genHType x, half y);
	$genIType max($genIType x, $genIType y);
	$genIType max($genIType x, int y);

Change-Id: I0a6467fb60b008b61e8b6a7affaebfcb15f5f7a9
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/412057
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-25 14:52:11 +00:00
John Stiles
443fa19832 Implement compile-time optimization for sqrt(constant).
Change-Id: I3427fbaf57787c3051db95ec5882c9292d7985cf
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411312
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-25 04:43:10 +00:00
John Stiles
639e812d9f Remove usage of sqrt() as an optimization barrier in tests.
In the majority of cases, a uniform is an equally good substitute, and
replacing `sqrt(N)` with `unknownInput` actually makes the test clearer.

Change-Id: I7bcb477571972d7aa2ce8c49b3674471f7310748
Bug: skia:12034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411306
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-05-24 17:18:54 +00:00
John Stiles
f061c59792 Rewrite matrix comparisons to work around Adreno bug.
Change-Id: I2f245892e12416decd4f0aee491f16fb4040ffca
Bug: skia:11308
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411036
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
2021-05-24 13:42:19 +00:00
John Stiles
5ee369f623 Rename caps bit for matrix comparison rewrites.
This was originally designated for 2x2 matrices only, but this was not
right--all matrix comparisons actually need to be rewritten to fully
work around the bug.

Change-Id: I743d16a65bc55e93361a3dd8753653384583f063
Bug: skia:11308
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411416
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-21 22:31:53 +00:00
John Stiles
147cfda6d8 Allow function declarations to take (void).
This is allowed by GLSL, so we allow it too.

GLSL ES 1.0, Section 6.1: "The idiom “(void)” as a parameter list is
provided for convenience."

Change-Id: I551c505d3de518a75acd5e306f09f0f0767e43f2
Bug: skia:12025
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/411300
Commit-Queue: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-21 19:21:23 +00:00
John Stiles
36c5796f0b Add caps bit for 2x2 matrix comparison rewrites.
This will allow us to rewrite `mat == mat` on Adreno 5xx/6xx GPUs when
running in GLSL.

Change-Id: I621e918a545a49b7ecb9c944ae59b1e7a7594bae
Bug: skia:11308
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410996
Reviewed-by: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
2021-05-20 23:19:05 +00:00
John Stiles
25be58e24b Reland "Eliminate unreachable code during optimization."
This is a reland of c6260f9742

Problematic DeadReturn.sksl test cases have been moved to DeadReturnES3.

Original change's description:
> Eliminate unreachable code during optimization.
>
> The Adreno 5xx and 6xx previously failed the SkSLStaticSwitchInline
> test; the driver struggled to interpret code with multiple return
> statements in a row. We now detect unreachable statements and eliminate
> them.
>
> Change-Id: I344d632f2488ca65b0635b37bebffe6e4fb607c5
> Bug: skia:12012
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410256
> Auto-Submit: John Stiles <johnstiles@google.com>
> Commit-Queue: Brian Osman <brianosman@google.com>
> Reviewed-by: Brian Osman <brianosman@google.com>

Bug: skia:12012
Change-Id: I748e8761cbc71c811b5ad8fe49186f980261d8b9
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410793
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-20 20:02:15 +00:00
John Stiles
6011bbcf48 Fix comma operator support with matrices in Metal.
MetalCodeGen would incorrectly identify `(someMatrix, someScalar)` as a
math operation between `someMatrix` and `someScalar` and attempt to
convert `someScalar` to a matrix. If `someScalar` was a Boolean type,
this would lead to an assertion.

The binary expression is now checked more thoroughly before converting
the scalar into a matrix.

Change-Id: Id7e104d5533d8c43375927d4815b83e1a3c36be1
Bug: skia:11125
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410682
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-20 18:08:35 +00:00
Brian Salomon
e72fd09ce1 Revert "Eliminate unreachable code during optimization."
This reverts commit c6260f9742.

Reason for revert: Nexus 7 is grumpy, fails SkSLDeadReturn_GPU
https://ci.chromium.org/raw/build/logs.chromium.org/skia/539fff50ff093c11/+/annotations


Original change's description:
> Eliminate unreachable code during optimization.
>
> The Adreno 5xx and 6xx previously failed the SkSLStaticSwitchInline
> test; the driver struggled to interpret code with multiple return
> statements in a row. We now detect unreachable statements and eliminate
> them.
>
> Change-Id: I344d632f2488ca65b0635b37bebffe6e4fb607c5
> Bug: skia:12012
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410256
> Auto-Submit: John Stiles <johnstiles@google.com>
> Commit-Queue: Brian Osman <brianosman@google.com>
> Reviewed-by: Brian Osman <brianosman@google.com>

TBR=brianosman@google.com,ethannicholas@google.com,johnstiles@google.com

Change-Id: Ia3db1f1b28417e479e2d71a4a6ed94a007e47cf9
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:12012
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410780
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
2021-05-20 13:00:36 +00:00
John Stiles
c6260f9742 Eliminate unreachable code during optimization.
The Adreno 5xx and 6xx previously failed the SkSLStaticSwitchInline
test; the driver struggled to interpret code with multiple return
statements in a row. We now detect unreachable statements and eliminate
them.

Change-Id: I344d632f2488ca65b0635b37bebffe6e4fb607c5
Bug: skia:12012
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410256
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-19 23:03:23 +00:00
John Stiles
3c74a5357a Add GLSL workaround for 4x2 diagonal matrix bug.
GLSL now emits 4x2 diagonal matrices as:
	`(mat4x2(1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0) * n)`
instead of
	`mat4x2(n)`.

This works around a long-standing GLSL bug in both Mesa and glslang that
affects several drivers:
https://github.com/KhronosGroup/glslang/issues/2645

Change-Id: If529d5cd150ce720f436cb3634a2fd3423919278
Bug: skia:12003
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410137
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-18 22:50:19 +00:00
Brian Osman
fa6a127310 Large simplification of the CType mapper logic in SkSL
- Arrays were unused
- Each ctype was used with exactly one mapper - remove the SkSL type
  list
- After that, the mapper was just a key + value, so make it just a value
  and use a map (rather than vector)

Change-Id: Ibc150ecc9ac9724ecd1e7056614ca5836ce99d5b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/409998
Commit-Queue: John Stiles <johnstiles@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
2021-05-18 21:40:14 +00:00
Brian Osman
c23386ca39 Remove layout(override_coverage)
This was only used in the context of sk_SampleMask, which was removed
recently.

Change-Id: Id70d7af8b3a100ff157c2984bad4131f1b92f317
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/410056
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
2021-05-18 19:39:54 +00:00
Brian Osman
3adc091af8 Remove layout(tracked) from SkSL
Change-Id: I6019418526def09c6c9f4b22567a2c76542d043c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/409876
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
2021-05-18 17:44:54 +00:00
John Stiles
325c808ebb Enable additional SkSL tests which require non-ES2 features.
Although these tests are simple, it ended up uncovering a legitimate
SPIR-V bug (skia:12009).

Change-Id: Ie89235157256b97626aa6ada4d9d6ba62abc57fa
Bug: skia:12009
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/409299
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-17 20:24:50 +00:00
John Stiles
f2b08cc9c8 Fix SPIR-V code generation of dead return statements.
If we encounter code like `return 1; return 2;` we need to synthesize a
label, even though the second return statement isn't actually reachable.
This is harmless and satisfies the SPIR-V validator.

Ideally we'd eliminate the dead code entirely, but this case is rare and
isn't likely to cause any problems as-is.

Change-Id: I2d6219dff6868011353e19a662301bec44a015d6
Bug: skia:12009
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/409402
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
2021-05-17 20:23:40 +00:00
John Stiles
191b4e2a00 Enable proper testing of matrix-scalar ops.
Now that the various Metal and SPIR-V bugs have been shaken out, we can
enable these tests. Knock on wood.

Change-Id: If4b4e302cfdd91464aaf00bc9639989de5e49aac
Bug: skia:11985
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/408640
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-05-17 17:36:22 +00:00
John Stiles
bdf3bb7b5f Add support for componentwise matrix divide in SPIR-V.
All the pieces of the puzzle were already here to support componentwise
addition and subtraction of matrices. Division was just a forgotten gap
in the implementation and is now patched up to match + and -.

NOTE: if you read the SPIR-V output very closely, you may be surprised
that there are fewer FDiv operations than you'd expect from reading the
input SkSL. As it turns out, a preexisting optimization is rewriting
`mat / 4` into `mat * 0.25` (see line 2689), and this rewritten form can
use the dedicated MatrixTimesScalar op. So we only get componentwise
FDivs for the `4 / mat` lines in the source.

Change-Id: I011c859f5b3a031fbb95a2956f1194a5f3b3794b
Bug: skia:11985
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/408639
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-05-17 16:01:10 +00:00
John Stiles
a91bf055ea Add support for matrix-scalar operations in SPIR-V.
Multiplication is handled just as before. Ops other than multiplication
are handled by splatting the scalar into a matrix, then performing the
op as a componentwise matrix-op-matrix binary expression.

Change-Id: I654715c45bf5c91b8e9660fdf1c1c6d6818b621a
Bug: skia:11985
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/408637
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-05-17 14:19:10 +00:00
John Stiles
ab3e6394a0 Re-enable the nonsquare matrix test.
The mat4x2 portion of the test no longer checks its results, as there
are bugs in the Intel and Radeon GPU drivers that prevent mat4x2s
from being constructed properly.

The SkSL optimizer ends up eliminating the 4x2 matrix entirely because
it is unused by the rest of the code, as well at the 4x4 matrix which is
calculated from the 4x2. At this point, I'm OK with this.

Change-Id: If1464f9e4938b0a37b2ec180c686972389d94e83
Bug: skia:12003
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/408900
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
2021-05-17 13:27:48 +00:00