Commit Graph

315 Commits

Author SHA1 Message Date
Hans-Kristian Arntzen
9b7140e2ba Implement OpAtomicLoad/OpAtomicStore.
Need some emulation on GLSL/HLSL, fix bug with atomic store on MSL.
2020-04-27 12:11:46 +02:00
Hans-Kristian Arntzen
5e5d1c27ce GLSL: Support f16x2 <-> f32 bitcast.
There is no native formulation, so introduce a concept of a "complex"
bitcast to handle odd-ball cases which have no native unary operation.
2020-04-21 23:27:33 +02:00
Hans-Kristian Arntzen
bdb343ea06 Be a bit more careful what nonuniform state is propagated. 2020-04-21 14:48:44 +02:00
Hans-Kristian Arntzen
7b9cba7424 HLSL: Add parens in unpackUint2x32 for clarity. 2020-04-21 11:49:26 +02:00
Hans-Kristian Arntzen
e4e4791c4e HLSL: Only allow 64-bit integers in SM 6.0. 2020-04-21 11:49:26 +02:00
Asuka
55dfbead2f GLSL/HLSL: Support packUint2x32 and unpackUint2x32 2020-04-21 11:34:12 +02:00
Hans-Kristian Arntzen
28bf9057df HLSL: Add support for treating NonWritable UAV texture as SRV instead. 2020-04-03 11:50:50 +02:00
Hans-Kristian Arntzen
b8905bbd95 Add support for forcefully zero-initialized variables.
Useful to better support certain platforms which require all variables
to be initialized to something.
2020-03-26 13:38:27 +01:00
Hans-Kristian Arntzen
04e877df12 GLSL: Implement GL_EXT_shader_framebuffer_fetch. 2020-03-19 14:53:39 +01:00
Hans-Kristian Arntzen
05004a57ea GLSL/HLSL: Fix nonuniform qualifier for SSBO atomics. 2020-03-19 12:05:27 +01:00
Hans-Kristian Arntzen
a3d3c80dd7 GLSL/HLSL: Implement nonuniform qualifier for image atomics. 2020-03-19 11:35:29 +01:00
Hans-Kristian Arntzen
185551bfaf HLSL: Do not emit globallycoherent for SRV ByteAddressBuffer. 2020-03-05 10:37:36 +01:00
Hans-Kristian Arntzen
c27e1efbf1 HLSL: Add option to always treat SSBO as UAV, even with readonly.
This can make codegen more predictable since ByteAddressBuffer is SRV
and not UAV.
2020-03-04 16:42:31 +01:00
Hans-Kristian Arntzen
e81c1b1d98 HLSL: Declare undef variables as static.
Undef variables would somehow become cbuffer variables without any
warning ...
2020-02-08 13:39:50 +01:00
Hans-Kristian Arntzen
f9818f0804 Update license headers to 2020. 2020-01-16 15:24:37 +01:00
Hans-Kristian Arntzen
f79c1e2fed Deal with illegal names in types as well.
- Fixes issue with clip_distance flattening in MSL where member to
  flatten from would come from to_member_name, where it should have used
  the builtin name directly. This member name was modified by this patch
  and broke clip distance test shaders.

- Some cleanups with ir.meta, use ir.find_meta instead to not create
  unnecessary hashmap nodes.
2020-01-16 10:34:49 +01:00
Hans-Kristian Arntzen
172e39f039
Merge pull request #1257 from KhronosGroup/fix-1236
Deal with bitcasting for subgroup Min/Max operations
2020-01-09 15:35:43 +01:00
Hans-Kristian Arntzen
cc153f8d7f HLSL: Add a resource remapping API similar to MSL.
Allows more flexibility of how resources are assigned without having to
remap decorations.
2020-01-09 12:41:06 +01:00
Hans-Kristian Arntzen
88ddeec49a HLSL: Deal with casting for WaveActiveMin/Max. 2020-01-09 12:35:18 +01:00
Hans-Kristian Arntzen
c256525c7b Run format_all.sh. 2020-01-08 14:27:34 +01:00
Hans-Kristian Arntzen
1cbd71b354 HLSL: Fix bug when reading and writing structs from SSBO. 2020-01-08 14:27:02 +01:00
Hans-Kristian Arntzen
151ff1e870 HLSL: Implement stores for complex composites in ByteAddressBuffers. 2020-01-08 14:17:28 +01:00
Hans-Kristian Arntzen
ca9398c122 HLSL: Support loading complex composites from ByteAddressBuffer. 2020-01-08 13:05:56 +01:00
Hans-Kristian Arntzen
b9e5fe01b0 HLSL: Add support to remove register() bindings.
Sometimes it's useful to get automatic binding assignment from the D3D
compiler instead.
2019-11-11 11:23:21 +01:00
Hans-Kristian Arntzen
0b417b586a HLSL: Report more explicitly which member failed validation.
This will be awkward to report in GLSL where we check multiple packing
standards, but for HLSL it should be easy since there's only CBuffer
packing standard to worry about.
2019-11-06 11:21:39 +01:00
Hans-Kristian Arntzen
e73d9bee38 HLSL: Report which cbuffer failed validation. 2019-11-06 11:05:31 +01:00
Hans-Kristian Arntzen
6edbf0c9e9 MSL: Minor cleanups for texture atomic emulation.
Storing pointers to internal objects is generally not done, IDs are
preferred.
2019-10-24 11:30:20 +02:00
Hans-Kristian Arntzen
a9be92569f HLSL: Fix unrolled S/G LE/LT/GE/GT opcodes.
Need to bitcast the unrolled expressions as well.
2019-10-14 16:08:39 +02:00
Hans-Kristian Arntzen
b960ae3b70 HLSL: Partially implement Unordered compare.
We cannot correctly implement unordered equal/ordered not equal without
a lot of extra instructions which slows normal code down.
2019-10-14 15:15:03 +02:00
Hans-Kristian Arntzen
333980ae91 Refactor into stronger types in public API.
Some fallout where internal functions are using stronger types.
Overkill to move everything over to strong types right now, but perhaps
move over to it slowly over time.
2019-09-06 12:29:47 +02:00
Hans-Kristian Arntzen
261b46982a Deal with complex interlock cases in GLSL. 2019-09-04 12:18:04 +02:00
Chip Davis
2eff420d9a Support the SPV_EXT_fragment_shader_interlock extension.
This was straightforward to implement in GLSL. The
`ShadingRateInterlockOrderedEXT` and `ShadingRateInterlockUnorderedEXT`
modes aren't implemented yet, because we don't support
`SPV_NV_shading_rate` or `SPV_EXT_fragment_invocation_density` yet.

HLSL and MSL were more interesting. They don't support this directly,
but they do support marking resources as "rasterizer ordered," which
does roughly the same thing. So this implementation scans all accesses
inside the critical section and marks all storage resources found
therein as rasterizer ordered. They also don't support the fine-grained
controls on pixel- vs. sample-level interlock and disabling ordering
guarantees that GLSL and SPIR-V do, but that's OK. "Unordered" here
merely means the order is undefined; that it just so happens to be the
same as rasterizer order is immaterial. As for pixel- vs. sample-level
interlock, Vulkan explicitly states:

> With sample shading enabled, [the `PixelInterlockOrderedEXT` and
> `PixelInterlockUnorderedEXT`] execution modes are treated like
> `SampleInterlockOrderedEXT` or `SampleInterlockUnorderedEXT`
> respectively.

and:

> If [the `SampleInterlockOrderedEXT` or `SampleInterlockUnorderedEXT`]
> execution modes are used in single-sample mode they are treated like
> `PixelInterlockOrderedEXT` or `PixelInterlockUnorderedEXT`
> respectively.

So this will DTRT for MoltenVK and gfx-rs, at least.

MSL additionally supports multiple raster order groups; resources that
are not accessed together can be placed in different ROGs to allow them
to be synchronized separately. A more sophisticated analysis might be
able to place resources optimally, but that's outside the scope of this
change. For now, we assign all resources to group 0, which should do for
our purposes.

`glslang` doesn't support the `RasterizerOrdered` UAVs this
implementation produces for HLSL, so the test case needs `fxc.exe`.

It also insists on GLSL 4.50 for `GL_ARB_fragment_shader_interlock`,
even though the spec says it needs either 4.20 or
`GL_ARB_shader_image_load_store`; and it doesn't support the
`GL_NV_fragment_shader_interlock` extension at all. So I haven't been
able to test those code paths.

Fixes #1002.
2019-09-02 12:31:10 -05:00
Chip Davis
39dce88d3b MSL: Add support for sampler Y'CbCr conversion.
This change introduces functions and in one case, a class, to support
the `VK_KHR_sampler_ycbcr_conversion` extension. Except in the case of
GBGR8 and BGRG8 formats, for which Metal natively supports implicit
chroma reconstruction, we're on our own here. We have to do everything
ourselves. Much of the complexity comes from the need to support
multiple planes, which must now be passed to functions that use the
corresponding combined image-samplers. The rest is from the actual
Y'CbCr conversion itself, which requires additional post-processing of
the sample retrieved from the image.

Passing sampled images to a function was a particular problem. To
support this, I've added a new class which is emitted to MSL shaders
that pass sampled images with Y'CbCr conversions attached around. It
can handle sampled images with or without Y'CbCr conversion. This is an
awful abomination that should not exist, but I'm worried that there's
some shader out there which does this. This support requires Metal 2.0
to work properly, because it uses default-constructed texture objects,
which were only added in MSL 2. I'm not even going to get into arrays of
combined image-samplers--that's a whole other can of worms.  They are
deliberately unsupported in this change.

I've taken the liberty of refactoring the support for texture swizzling
while I'm at it. It's now treated as a post-processing step similar to
Y'CbCr conversion. I'd like to think this is cleaner than having
everything in `to_function_name()`/`to_function_args()`. It still looks
really hairy, though. I did, however, get rid of the explicit type
arguments to `spvGatherSwizzle()`/`spvGatherCompareSwizzle()`.

Update the C API. In addition to supporting this new functionality, add
some compiler options that I added in previous changes, but for which I
neglected to update the C API.
2019-09-01 18:35:53 -05:00
Hans-Kristian Arntzen
b3305799a8 Deal correctly with sign on bitfield operations.
Need a lot of special purpose implementation functions for these.
2019-08-26 11:36:36 +02:00
Hans-Kristian Arntzen
4bc8729c0e HLSL query lod cleanups. 2019-07-24 11:34:28 +02:00
Hans-Kristian Arntzen
47a18b9f1b Simplify row-major matrix/vector multiplies. 2019-07-23 10:56:57 +02:00
Hans-Kristian Arntzen
dd7ebaf9f7 Start considering how to emit physical type ID. 2019-07-19 10:06:19 +02:00
Hans-Kristian Arntzen
a86308bce1 MSL: Begin rewrite of buffer packing logic. 2019-07-19 10:06:19 +02:00
Chip Davis
50dce10c5d Support the SPV_EXT_demote_to_helper_invocation extension.
This extension provides a new operation which causes a fragment to be
discarded without terminating the fragment shader invocation. The
invocation for the discarded fragment becomes a helper invocation, so
that derivatives will remain defined. The old `HelperInvocation` builtin
becomes undefined when this occurs, so a second new instruction queries
the current helper invocation status.

This is only fully supported for GLSL. HLSL doesn't support the
`IsHelperInvocation` operation and MSL doesn't support the
`DemoteToHelperInvocation` op.

Fixes #1052.
2019-07-17 09:12:22 -05:00
Hans-Kristian Arntzen
c7eda1bce9 Test glsl.std450 more exhaustively.
Make sure to test everything with scalar as well to catch any weird edge
cases.

Not all opcodes are covered here, just the arithmetic ones. FP64 packing
is also ignored.
2019-07-17 11:53:05 +02:00
Hans-Kristian Arntzen
932ee0e328 Deal correctly with return sign of bitscan operations. 2019-07-12 10:57:56 +02:00
Chip Davis
6628ea6e48 MSL: Use the select() function for OpSelect.
This significantly improves codegen for vector `OpSelect` in MSL.
2019-07-11 10:30:37 -05:00
Hans-Kristian Arntzen
d12b54bbb4 Propagate NonUniformEXT to dependent expressions.
This decoration might only be present for the very last ID which is
consumed by a sampling or Load/Store instruction. To make sure our
access chains are emitted correctly, we have to back-propagate this
decoration.
2019-07-08 11:19:38 +02:00
Hans-Kristian Arntzen
4056d0b74e Don't use scalar dot(). 2019-07-03 14:32:06 +02:00
Hans-Kristian Arntzen
041f103d44 MSL/HLSL: Support scalar reflect and refract. 2019-07-03 12:31:52 +02:00
Hans-Kristian Arntzen
f8b084de61 MSL/HLSL: Support OpOuterProduct. 2019-07-01 10:57:27 +02:00
Hans-Kristian Arntzen
ff87419607 Deal with scalar input values for distance/length/normalize.
HLSL and MSL don't support it, so fall back to simpler intrinsics.
2019-06-28 11:20:14 +02:00
Hans-Kristian Arntzen
581ed0fd59 HLSL: Does not support case-fallthrough.
Disable any fallthrough on HLSL. Risky business if fallthrough blocks
had a barrier(), but can't do anything about that ...
2019-06-27 15:10:17 +02:00
Hans-Kristian Arntzen
f171d82590 MSL: Support MinLod operand. 2019-06-19 09:43:03 +02:00
Hans-Kristian Arntzen
96492648d4 MSL: Fix struct declaration order with complex type aliases.
MSL generally emits the aliases, which means we cannot always place the
master type first, unlike GLSL and HLSL. The logic fix is just to
reorder after we have tagged types with packing information, rather than
doing it in the parser fixup.
2019-05-23 14:54:04 +02:00