Commit Graph

4874 Commits

Author SHA1 Message Date
Nikolay Sivov
cf27065b7d vkd3d-shader/hlsl: Add parser support for the RWByteAddressBuffer type.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-28 17:44:28 +01:00
Nikolay Sivov
4195a2f18a vkd3d-shader/hlsl: Use a more compact way to store object method configurations.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-28 17:41:46 +01:00
Elizabeth Figura
390a330b8e vkd3d-shader/hlsl: Avoid leaking the block in the subscript rule. 2024-10-28 15:40:41 +01:00
Elizabeth Figura
bb74180659 vkd3d-shader/hlsl: Return an error expression when an invalid subscript is used. 2024-10-28 15:40:41 +01:00
Elizabeth Figura
7905c47765 vkd3d-shader/hlsl: Handle error expressions in array sizes. 2024-10-28 15:40:41 +01:00
Elizabeth Figura
8b98e483ac vkd3d-shader/hlsl: Handle error expressions in return statements. 2024-10-28 15:40:41 +01:00
Elizabeth Figura
94cf432035 vkd3d-shader/hlsl: Handle error expressions in conditions. 2024-10-28 15:40:41 +01:00
Nikolay Sivov
5eff8bf918 vkd3d-shader/hlsl: Handle SV_GroupIndex.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-24 20:54:04 +02:00
Henri Verbeet
5e9edf25d0 vkd3d-shader/glsl: Implement VKD3DSIH_GATHER4. 2024-10-24 20:52:37 +02:00
Henri Verbeet
d477a00734 vkd3d-shader/glsl: Implement VKD3DSIH_SAMPLE_GRAD. 2024-10-24 20:52:37 +02:00
Henri Verbeet
d58282c6d1 vkd3d-shader/glsl: Implement VKD3DSIH_SAMPLE_LOD. 2024-10-24 20:52:37 +02:00
Henri Verbeet
85697d76be vkd3d-shader/glsl: Implement VKD3DSIH_UMAX and VKD3DSIH_UMIN. 2024-10-24 20:52:37 +02:00
Henri Verbeet
6a119b6946 vkd3d-shader/glsl: Implement VKD3DSGF_FORCE_EARLY_DEPTH_STENCIL. 2024-10-24 20:52:37 +02:00
Francisco Casas
9d82915629 vkd3d-shader/hlsl: Make allocation functions static again. 2024-10-24 20:50:59 +02:00
Francisco Casas
23be6ed0dd vkd3d-shader/hlsl: Store temp declarations in the vsir program.
Move the temp allocation back to hlsl_codegen.c.

Note that the DCL_TEMPS instructions wouldn't be necessary if we had the
capacity to store the temp_count for both the main program and the patch
constant program (or more generally speaking, a temp_count for all
phases).

The plan is to eventually also move the HS_CONTROL_POINT and
HS_FORK_PHASE markers to the vsir_program, making it able to contain
both functions.
2024-10-24 20:50:59 +02:00
Francisco Casas
158bf794e6 vkd3d-shader/hlsl: Introduce hlsl_ir_vsir_instruction_ref, again.
This node type will be deleted (again) once the hlsl->vsir->tpf
translation is complete. It serves the purpose of allowing to keep
both real hlsl_ir_nodes and vsir_instructions in the hlsl_block,
until all the former can be translated into the latter.
2024-10-24 20:50:59 +02:00
Francisco Casas
cf7fade580 vkd3d-shader/hlsl: Store the thread group size in the vsir program. 2024-10-24 20:48:20 +02:00
Giovanni Mascellani
91701f8303 vkd3d-shader: Warn instead of erroring out when failing to dump a shader.
It's a debug operation and shouldn't be considered fatal, even
for development workflows.
2024-10-23 16:14:33 +02:00
Giovanni Mascellani
6a03369f84 vkd3d-shader: Do not dump the target shader if compilation failed.
Otherwise we'd use unitialized data.
2024-10-23 16:14:33 +02:00
Giovanni Mascellani
a1f712d100 vkd3d: Incorporate mutable descriptors in the push descriptor disabling logic.
The computation is also written in the code now instead of in
the comments.
2024-10-23 16:12:50 +02:00
Giovanni Mascellani
bc2b137df9 vkd3d: Only put the mutable descriptor set once in the pipeline layout.
Currently the mutable descriptor set is repeated many times in the
pipeline layout in order to cover the indices for all the
descriptor types that would be present if mutable descriptors were
not used. This is useless and wasteful, but was necessary before
the descriptor sets backing the SRV-UAV-CBV heap were moved at the
end of the allocation table because descriptor set indices are
currently a compile-time constant in many places.

Now this is not needed any more and we can just avoid putting
many copies of the mutable descriptor set in the pipeline layout,
making it easier to meet Vulkan implementation limits.
2024-10-23 16:12:06 +02:00
Giovanni Mascellani
96b324c156 vkd3d: Move descriptor sets backing the SRV-UAV-CBV heap at the end.
So that when mutable descriptors are in use we can avoid putting
the other descriptor sets backing the SRV-UAV-CBV descriptor heap
in the pipeline layout altogether.
2024-10-23 16:07:54 +02:00
Giovanni Mascellani
8d030718ae vkd3d: Introduce an enumerant for the mutable descriptor set.
So we avoid hardcoding that it is number zero. There are two
goals here: first making the code easier to understand and
second allow reshuffling the descriptor set allocation in a
later commit.
2024-10-23 16:07:36 +02:00
Elizabeth Figura
32e6a1bb4d vkd3d-shader: Validate the parsed shader in vsir_parse(). 2024-10-23 16:02:09 +02:00
Elizabeth Figura
3a3f95e1e7 vkd3d-shader: Factor out a vsir_parse() helper. 2024-10-23 16:02:09 +02:00
Conor McCarthy
4889c71857 vkd3d: Zero the pipeline state UAV counter view array when the state is invalidated. (Valgrind)
Otherwise a comparison with uninitialised data occurs in
d3d12_command_list_update_descriptor_table() if virtual heaps are used.
2024-10-22 20:42:06 +02:00
Henri Verbeet
1113d24a70 vkd3d-shader/ir: Store the global flags in struct vsir_program. 2024-10-22 20:40:31 +02:00
Henri Verbeet
21e9029177 vkd3d-shader/glsl: Implement VKD3DSIH_SAMPLE_C_LZ. 2024-10-22 20:39:31 +02:00
Henri Verbeet
5958e527e5 vkd3d-shader/glsl: Implement VKD3DSIH_SAMPLE_C. 2024-10-22 20:39:31 +02:00
Henri Verbeet
19c493722e vkd3d-shader/glsl: Implement VKD3DSIH_SAMPLE_B. 2024-10-22 20:39:31 +02:00
Henri Verbeet
e408f1d984 vkd3d-shader/glsl: Implement support for the VKD3DSPR_THREADID register. 2024-10-22 20:39:31 +02:00
Henri Verbeet
f76dd80cba vkd3d-shader/glsl: Implement VKD3DSIH_LD_UAV_TYPED. 2024-10-22 20:39:31 +02:00
Nikolay Sivov
cf3e251a9f vkd3d-shader/hlsl: Implement the f32tof16() intrinsic.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-22 20:28:33 +02:00
Shaun Ren
f54797bae5 vkd3d-shader/hlsl: Support discard for SM1. 2024-10-22 20:27:34 +02:00
Francisco Casas
2a8c1b2823 vkd3d-shader/hlsl: Sort signature elements by register id. 2024-10-22 20:25:40 +02:00
Francisco Casas
51d05c1844 vkd3d-shader/hlsl: Optimize interstage signatures. 2024-10-22 20:15:15 +02:00
Francisco Casas
88dd082160 vkd3d-shader/hlsl: Allow prioritizing smaller writemasks when allocating signature elements.
For now this doesn't have effect yet, until semantic allocation is
optimized.
2024-10-22 20:06:15 +02:00
Francisco Casas
d562b03c43 vkd3d-shader/hlsl: Allow to force alignment on some semantic vars.
This is required to properly optimize signatures, because these
semantics must be alligned while being packed:
- Array elements.
- The first fields of structs.
- Major vectors of a matrix.

For now this has no effect since semantics are allocated with reg_size
4, but will have effect when optimizing interstage signatures.
2024-10-22 20:04:41 +02:00
Francisco Casas
7fd6c29ee8 vkd3d-shader/hlsl: Allow accounting for interpolation mode when allocating semantics.
For now this has no effect since semantics are allocated with reg_size
4, but will have effect when optimizing interstage signatures.
2024-10-22 20:03:39 +02:00
Francisco Casas
b65f5fb069 vkd3d-shader/hlsl: Also pass field storage modifiers to output signature elements.
The generated pixel shader input signature must be consistent with the
generated vertex shader output signature for the same data type.

Since the interpolation mode affects allocation order, the allocator
needs to know the modifiers for both input and output signature elements.
2024-10-22 19:29:11 +02:00
Francisco Casas
7180a96741 vkd3d-shader/hlsl: Use a register_allocator to allocate semantic registers. 2024-10-22 19:23:33 +02:00
Francisco Casas
9ce7b7806f vkd3d-shader/hlsl: Add mode field to register_allocator allocations. 2024-10-22 19:23:22 +02:00
Nikolay Sivov
fb2b974466 vkd3d-shader/hlsl: Handle snorm/unorm types as resource formats.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-22 19:14:51 +02:00
Elizabeth Figura
a492d64fef vkd3d-shader/ir: Allow controlling point sprite through a parameter. 2024-10-22 19:14:02 +02:00
Nikolay Sivov
8a3fe9cd1e vkd3d-shader/fx: Implement parsing groups and techniques.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-21 18:23:33 +02:00
Giovanni Mascellani
ae772d40cf vkd3d: Disable push descriptors when that's helpful to stay within 8 descriptor sets. 2024-10-21 18:22:15 +02:00
Giovanni Mascellani
3bf114f3c3 vkd3d/device: Trace which descriptor heap implementation is being used.
The descriptor heap implementation is a rather central behavior element
in vkd3d, so it's useful to have all the relevant information logged
in a single place.
2024-10-21 18:22:15 +02:00
Giovanni Mascellani
ed4d45355a vkd3d-shader/ir: Represent the normalisation level with an enumeration. 2024-10-21 18:21:39 +02:00
Shaun Ren
abed6d9cb7 vkd3d-shader/tpf: Implement semantics for domain shaders. 2024-10-21 18:18:25 +02:00
Shaun Ren
0273ff007b vkd3d-shader/tpf: Write domain shader declarations. 2024-10-21 18:18:00 +02:00
Shaun Ren
26efba30b4 vkd3d-shader/tpf: Use vpc input registers for domain shaders. 2024-10-21 18:17:06 +02:00
Shaun Ren
2b897296a1 vkd3d-shader/tpf: Write the input signature of domain shaders as PCSG. 2024-10-21 18:16:50 +02:00
Shaun Ren
768b19d410 vkd3d-shader/ir: Validate control point counts correctly.
Allow zero input control points for domain shaders, and zero output
control points for hull shaders.
2024-10-21 18:10:13 +02:00
Shaun Ren
976fe27836 vkd3d-utils: Return correct use masks during reflection. 2024-10-21 18:09:39 +02:00
Shaun Ren
7408ab145a vkd3d-utils: Implement version reflection. 2024-10-21 18:09:06 +02:00
Giovanni Mascellani
a3b8aaf679 vkd3d: Allow overriding the device capabilities. 2024-10-21 18:06:06 +02:00
Giovanni Mascellani
03ad04c890 vkd3d-shader/ir: Validate OUTCONTROLPOINT registers. 2024-10-17 17:34:55 +02:00
Giovanni Mascellani
1e86f00247 vkd3d-shader/ir: Validate INCONTROLPOINT registers. 2024-10-17 17:34:55 +02:00
Giovanni Mascellani
a4ab53ab7b vkd3d-shader/ir: Validate PATCHCONST registers. 2024-10-17 17:34:55 +02:00
Giovanni Mascellani
62d5bc91f5 vkd3d-shader/ir: Validate OUTPUT registers. 2024-10-17 17:34:55 +02:00
Giovanni Mascellani
349fdd0843 vkd3d-shader/ir: Validate INPUT registers. 2024-10-17 17:34:55 +02:00
Giovanni Mascellani
d9508558a9 vkd3d-shader: Keep track of whether programs have normalised hull shader control point I/O. 2024-10-17 17:34:55 +02:00
Giovanni Mascellani
eba8fd4720 vkd3d-shader: Keep track of whether programs have normalised I/O. 2024-10-17 17:34:55 +02:00
Nikolay Sivov
071130b944 vkd3d-shader/fx: Add support for tracing annotations.
Signed-off-by: Nikolay Sivov <nsivov@codeweavers.com>
2024-10-17 17:32:54 +02:00
Elizabeth Figura
eeb7bbe3c5 vkd3d-shader/hlsl: Handle error expressions in ternary expressions. 2024-10-17 17:28:52 +02:00
Elizabeth Figura
d95962068a vkd3d-shader/hlsl: Handle error expressions in initializers. 2024-10-17 17:28:52 +02:00
Elizabeth Figura
dfe79cb6a9 vkd3d-shader/hlsl: Factor the component count check into initialize_var(). 2024-10-17 17:28:52 +02:00
Elizabeth Figura
d8892a4ca2 vkd3d-shader/hlsl: Store the initializer location in struct parse_initializer. 2024-10-17 17:28:52 +02:00
Elizabeth Figura
44ea7b5e12 vkd3d-shader/hlsl: Factor out an initialize_var() helper. 2024-10-17 17:28:52 +02:00
Giovanni Mascellani
0e43dea42d vkd3d-shader/ir: Check that signature masks are contiguous. 2024-10-17 17:28:52 +02:00
Giovanni Mascellani
083b87c712 vkd3d-shader/d3dbc: Make signature masks contiguous.
The goal is to make a requirement for VSIR that signature element
masks are always contiguous. The SPIR-V backend already implicitly
makes that assumption, since it just consider the LSB and popcount
of the mask.

For example, consider this HLSL pixel shader:

    float4 main(float4 color : COLOR) : SV_Target
    {
        return float4(color.x, 10.0f, 11.0f, color.w);
    }

Currently the parser describes the input signature element
corresponding to semantic COLOR as having mask .xw, which is
sensible. However, the SPIR-V parser will interpret that as
a mask starting at x and with popcount 2, and assuming it is
contiguous it will implicitly act as if it were .xy. This is
not correct, because the wrong component will be loaded from
the vertex stage.
2024-10-17 17:28:52 +02:00
Henri Verbeet
8943999bd2 vkd3d: Slightly simplify the SRV/UAV logic in vk_write_descriptor_set_from_d3d12_desc(). 2024-10-17 17:28:52 +02:00
Conor McCarthy
bfaab9700d vkd3d: Lay out virtual descriptor heap buffer and image bindings consecutively instead of interleaving them.
Slightly simplifies descriptor write addressing, and makes layouts
essentially the same as array layouts, differing only in the binding
details, and therefore easier to understand. This also simplifies the
addition of storage buffer bindings, which can all be added onto the end.
2024-10-17 17:28:46 +02:00
Giovanni Mascellani
0b8a53d75d vkd3d: Create Vulkan bindings in d3d12_root_signature_append_vk_binding(). 2024-10-17 17:23:04 +02:00
Conor McCarthy
74da9ed6fe vkd3d: Create a separate Vulkan descriptor binding array for each descriptor set.
Allows descriptor set layouts to be created after all bindings are
mapped. This is less complex and fragile than the current scheme, and in
a future patch it will support separating descriptor types into different
sets. Descriptors on virtual heaps are currently allocated from pools
which contain an equal number of each descriptor type used by vkd3d, and
this can waste a significant amount of device memory.
2024-10-17 17:23:04 +02:00
Giovanni Mascellani
b5c56a50b5 vkd3d: Use vk_binding_array_add_binding() in d3d12_root_signature_init_root_descriptors(). 2024-10-17 17:23:04 +02:00
Giovanni Mascellani
3aec841a9f vkd3d: Use vk_binding_array_add_binding() in d3d12_root_signature_init_static_samplers(). 2024-10-17 17:23:04 +02:00
Giovanni Mascellani
416ce575f6 vkd3d: Introduce a helper function to add bindings to struct vk_binding_array. 2024-10-17 17:23:04 +02:00
Conor McCarthy
eca344ff56 vkd3d: Introduce a separate structure for storing Vulkan descriptor binding arrays. 2024-10-17 17:23:04 +02:00
Conor McCarthy
3f98ca53ca vkd3d: Access the current range via the declared pointer in d3d12_root_signature_init_root_descriptor_tables(). 2024-10-17 17:23:04 +02:00
Giovanni Mascellani
2b45878ff8 vkd3d-shader/ir: Validate the line density SV_TessFactor signature element for isolines domains. 2024-10-16 21:40:58 +02:00
Giovanni Mascellani
f470eb2a1a vkd3d-shader/ir: Validate the line detail SV_TessFactor signature element for isolines domains. 2024-10-16 21:40:58 +02:00
Giovanni Mascellani
c5c3151414 vkd3d-shader/ir: Validate SV_InsideTessFactor signature elements for tri domains. 2024-10-16 21:40:58 +02:00
Giovanni Mascellani
4e8b34c075 vkd3d-shader/ir: Validate SV_TessFactor signature elements for tri domains. 2024-10-16 21:40:58 +02:00
Giovanni Mascellani
b2460a6794 vkd3d-shader/ir: Validate SV_InsideTessFactor signature elements for quad domains. 2024-10-16 21:40:58 +02:00
Giovanni Mascellani
a8b9a92c10 vkd3d-shader/ir: Validate SV_TessFactor signature elements for quad domains. 2024-10-16 21:40:58 +02:00
Shaun Ren
1ed5f1a4d0 vkd3d-shader/tpf: Write the patch constant function in hull shaders.
We use a single hs_fork_phase here for the patch constant function.
2024-10-16 21:40:40 +02:00
Shaun Ren
ae2d86b99c vkd3d-shader/tpf: Determine SIV from SV and index in write_sm4_dcl_semantic(). 2024-10-16 21:12:29 +02:00
Shaun Ren
df3aaac7cf vkd3d-shader/tpf: Introduce tpf_write_shader_function(). 2024-10-16 21:12:13 +02:00
Shaun Ren
147f8898ca vkd3d-shader/hlsl: Process the patch constant function in hlsl_emit_bytecode().
Also, support patch constant function semantics in sm4_sysval_semantic_from_semantic_name().
2024-10-16 21:10:24 +02:00
Shaun Ren
cbed70c60f vkd3d-shader/tpf: Implement semantics for hull shaders. 2024-10-16 21:07:53 +02:00
Shaun Ren
beb342ed5f vkd3d-shader/tpf: Write hull shader declarations. 2024-10-16 21:07:53 +02:00
Henri Verbeet
47fcbc335b vkd3d-shader/d3dbc: Avoid D3DXSHADER_CONSTANTTABLE. 2024-10-16 21:07:53 +02:00
Shaun Ren
069b8aac64 vkd3d-shader/hlsl: Implement struct single inheritance.
Here, we implement single inheritance by inserting a field at the
beginning of the derived struct with name "$super".

For the following struct declarations

  struct a
  {
      float4 aa;
      float4 bb;
  };

  struct b : a
  {
      float4 cc;
  };

  struct c : b
  {
      float4 bb;
  };

this commit generates the following:

  struct a
  {
      float4 aa;
      float4 bb;
  };

  struct b
  {
      struct a $super;
      float4 cc;
  };

  struct c
  {
      struct b $super;
      float4 bb;
  };
2024-10-16 21:07:53 +02:00
Francisco Casas
7eee877dd4 vkd3d-shader/tpf: Make hlsl_sm4_register_from_semantic() independent of HLSL IR. 2024-10-16 21:07:51 +02:00
Francisco Casas
a243862b8c vkd3d-shader/tpf: Make sysval_semantic_from_hlsl() independent of HLSL IR. 2024-10-16 21:05:10 +02:00
Francisco Casas
7cc8151b81 vkd3d-shader/tpf: Replace uses of ctx->profile with tpf->program->shader_version. 2024-10-16 18:50:15 +02:00
Francisco Casas
c75fbaf94e vkd3d-shader/tpf: Use the I/O signatures from the vsir program in tpf_write_signature(). 2024-10-16 17:40:26 +02:00
Francisco Casas
10442369d8 vkd3d-shader/tpf: Pass a tpf_compiler structure to tpf_compile() callees. 2024-10-16 17:26:47 +02:00
Francisco Casas
16d9a10dd5 vkd3d-shader/tpf: Rename "tpf_writer" to "tpf_compiler".
Analog to d3dbc_compiler.
2024-10-16 17:09:15 +02:00