The hull shader barrier used for this was broken by I/O normalization, since
vocp is no longer exposed to the spirv backend.
Restore this barrier by checking for vocp during normalization instead.
Besides simply avoiding carrying around some uninitialised data, we
check "symbol->descriptor_array" in spirv_compiler_prepare_image(), both
for separate resources and for combined resource/sampler symbols.
The declaration instruction register contains id, range first, and range
last. The backend includes range first in the variable name. After commit
e8b3561252 it was always zero, and since commit 67f0196c33 it is
UINT_MAX, so constant buffers are named, e.g., "%cb0_4294967295".
Element masks can have a left shift in DXIL, but these must start at
bit 0 in the register info. The SPIR-V declaration will either be a
builtin or have SpvDecorationComponent.
We want to be able to remap input signatures based on the signature index, but
signature normalization both reorders the signature, and requires the old
register index, so add a new field for this.
Store it in the shader_desc, and declare temps from that when compiling SPIR-V,
instead of parsing dcl_instructions.
As part of this change, we declare a single, global temps array (with Private
scope instead of Function) which is as large as the maximum of all dcl_temps
instructions. It is not clear to me whether this will improve, hurt, or have no
significant effect on the lower-level compiler. An alternative is to still
redeclare a new temps array every time (although still with a smaller size).
This only affects clip and cull distances. The HLSL compiler emits these using
dcl_input, but the previous shader (vertex or TES) will write them as a SPIRV
builtin, and hence we want to read this as a SPIRV builtin as well.
This fixes validation errors in Wine's test_clip_distance().
In Shader Model 6 each signature element can span a range of register
indices, or 'rows', and system values do not share a register index with
non-system values. Inputs and outputs are referenced by element index
instead of register index. This patch merges multiple signature elements
into a single element under the following conditions:
- The register index in a load or store is specified dynamically by
including a relative address parameter with a base register index. The
dcl_index_range instruction is used to identify these.
- A register declaration is split across multiple elements which declare
different components of the register.
- A patch constant function writes tessellation factors. These are an
array in SPIR-V, but in SM 5.x each factor is declared as a separate
register, and these are dynamically indexed by the fork/join instance
id. Elimination of multiple fork/join phases converts the indices to
constants, but merging the signature elements into a single arrayed
element matches the SPIR-V output.
All references to input/output register indices are converted to element
indices. If a relative address is present, the element index is moved up
a slot so it cannot be confused with a constant offset. Existing code
only handles register index relative addressing for tessellation factors.
This patch adds generic support for it.
It is illegal to match a SPIR-V multisampled resource to a Vulkan resource which
is not multisampled. Vulkan considers a resource to be multisampled if its
sample count is greater than 1 (and SPIR-V does not care about the sample count).
This fixes validation errors in the case where the sample count does actually
match the resource. In order to provide correct behaviour when there is a
mismatch, or when the sample count is missing, we will need yet another
additional interface. In the absence of that it seems best to provide a best
guess.
This fixes a validation error with the not-yet-committed merge request 135, when
the d3d11 runner is run through Wine with the Vulkan backend.
The SPIR-V backend will emit a default control point phase. Inserting
inputs into the IR allows handling of declarations via the usual path
instead of an ad hoc implementation which may not match later changes
to input handling.
In SPIR-V the address must include the invocation id, but in TPF it
is implicit. Move the register index up one slot and insert an
OUTPOINTID relative address.
Normalise the incoming vkd3d_shader_instruction IR to the shader model 6
pattern. This allows generation of a single patch constant function in
SPIR-V.
Atomic ops on images with Unknown type will cause SPIR-V validation failure,
and assertion failure in Mesa debug builds. D3D12 allows atomics on typed
buffers, and this requires a distinction to be made between UAV reads and
atomic ops.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=53874
We would like to generate SPIR-V for input formats other than DXBC.
The "vkd3d_" prefix is dropped, partly to make names shorter, and partly to help
clarify what is an internal function.
I prefer avoiding the vkd3d_* prefix on all internal functions, for these
reasons. However, I'm open to restoring it.
Don't assume that enums and uint32_t parameters are identical. Clang
16 changes the diagonstic for incompatible function pointer types
from a warning into an error by default.
This fixes the following error, when built (for aarch64, but probably
also for other architectures) in MSVC mode:
../src/libs/vkd3d/libs/vkd3d-shader/spirv.c:1083:13: error: incompatible function pointer types passing 'uint32_t (struct vkd3d_spirv_builder *, uint32_t, SpvDim, uint32_t, uint32_t, uint32_t, uint32_t, SpvImageFormat)' (aka 'unsigned int (struct vkd3d_spirv_builder *, unsigned int, enum SpvDim_, unsigned int, unsigned int, unsigned int, unsigned int, enum SpvImageFormat_)') to parameter of type 'vkd3d_spirv_build7_pfn' (aka 'unsigned int (*)(struct vkd3d_spirv_builder *, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int)') [-Wincompatible-function-pointer-types]
vkd3d_spirv_build_op_type_image);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../src/libs/vkd3d/libs/vkd3d-shader/spirv.c:612:68: note: passing argument to parameter 'build_pfn' here
SpvOp op, const uint32_t *operands, vkd3d_spirv_build7_pfn build_pfn)
^
Signed-off-by: Martin Storsjö <martin@martin.st>
Specifying R32 for UAVs created with a vector format, e.g. R32G32B32A32_FLOAT,
results in only the red being loaded/stored, potentially causing images to
contain only the red component.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
The current workaround is broken for texture cube arrays, which already have 4
components.
Sampling with the components not packed together apparently succeeds with newer
NVidia drivers, so just remove the workaround. Tested with 470.103.01 on a GTX
1060.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52886
Signed-off-by: Zebediah Figura <zfigura@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Instead of making the caller convert them.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
SpvCapabilityRuntimeDescriptorArrayEXT requires descriptor indexing and
a validation error will result if it is not enabled.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Runtime descriptor arrays should be used for descriptor bindings with
a variable count.
Support for SpvCapabilityRuntimeDescriptorArrayEXT is required by the
Vulkan spec as part of descriptor indexing support.
The current implementation depends on a separate array declaration for
each range because binding_base_idx is stored in the array data. This
occurs in practice because binding.count is always different. Use of
runtime arrays eliminates the count from the key, so binding_base_idx
must be separated from the array data.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Otherwise if the input is located above position 0 in the private array
it will be swizzled incorrectly, e.g. a.yz instead of a.xy in
test_domain_shader_inputs().
Based on a vkd3d-proton patch by Hans-Kristian Arntzen.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
For isoline tessellation, "density" is specified by OL0, and "detail" by OL1.
Based on a vkd3d-proton patch by Hans-Kristian Arntzen.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Fixes invalid SPIR-V being generated in Monster Hunter: World.
Based on vkd3d-proton patches by Hans-Kristian Arntzen and Philip Rebohle.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Prevents a crash in radv when such inputs are present.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Based on a vkd3d-proton patch by Philip Rebohle.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This is undefined behaviour in SPIR-V, but well-defined in Direct3D, so we
should explicitly 'and' the shift amount with 31.
Based on a vkd3d-proton patch by Philip Rebohle.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Based in part on vkd3d-proton patches by Philip Rebohle and Hans-Kristian Arntzen.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Based in part on a vkd3d-proton patch by Joshua Ashton.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Based on a vkd3d-proton patch by Joshua Ashton.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Fixes NVIDIA HairWorks GPU crash on Metro Exodus, and test failures
in test_line_tessellation().
Based on a vkd3d-proton patch by Hans-Kristian Arntzen.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Source modifiers and immediate constants are not supported.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Based on a vkd3d-proton patch by Joshua Ashton.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Based on a vkd3d-proton patch by Joshua Ashton.
Signed-off-by: Conor McCarthy <cmccarthy@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
We've already set these up in vkd3d_dxbc_compiler_emit_hull_shader_inputs().
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
As opposed to at the end of each fork/join phase. This eliminates a
bunch of redundant SPIR-V, since it's not uncommon for fork/join phases
to only write to individual output components.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Like outputs, when inputs are coalesced, the write mask is modified by
needs_private_io_variable().
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
When outputs are coalesced, the write mask is modified by
needs_private_io_variable(). We want the "dcl_mask" field to reflect
that as well.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Multiple patch constant or output signature elements may be coalesced
into a single output register by needs_private_io_variable() in
vkd3d_dxbc_compiler_emit_output(). When this happens, only the original
element will have a corresponding variable ID assigned in output_info[].
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Much like we do for input components, for much of the same reasons.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
In particular, a shader may declare e.g. v6.x to refer to an output from the
previous stage, and v6.y to refer to a builtin, or the other way around. At
the same time, different hull shader phases may redeclare the same register.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
We need these to be per-vertex. This fixes the Wine d3d11
test_clip_distance() test.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
We will need this for shader model 5.1 resource arrays. However, for the
time being any count other than '1' is unsupported.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
In order to allow it to handle different source types.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
To vkd3d_shader_tessellator_partitioning.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
To vkd3d_shader_tessellator_output_primitive, consistent with vkd3d-shader
naming conventions.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
To vkd3d_shader_spirv_domain_shader_target_info.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
We still use the first index to uniquely identify the register in vkd3d_symbol,
only changing which field is used to declare it.
This patch does not handle descriptor arrays.
Signed-off-by: Zebediah Figura <zfigura@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This would cause CoreValidation-Shader-InterfaceTypeMismatch validation
errors from Wine's test_shader_interstage_interface() d3d11 test. This
reverts parts of commits 1eb7eca411 and
04ec461fb4.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This case needs special care since both VKD3DSPR_INPUT in the
control point phase and VKD3DSPR_INCONTROLPOINT in fork/join
phases refer to the same set of input variables, and we should
not declare input variables with the same location twice.
Encountered in Shadow of the Tomb Raider.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Uses the private patch constant array for tessellation factor built-ins.
Fixes two separate issues encountered in Shadow of the Tomb Raider:
- The output registers that have one component mapped to any of
the TESS_FACTOR sysvals can have their other components mapped
to a regular patch constant output, in which case we need to
use a private io variable.
- The tessellation factor outputs are not necessarily dynamically
indexed within shader code. Previously, this did not work correctly
and lead to invalid store operations in the generated SPIR-V.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Line tessellation factors use two different DXBC semantics that
both map to the same SPIR-V built-in. In this case, we cannot
rely on the semantic index.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Private variables are always vec4, so using a sparse write mask here
will lead to invalid code being generated when accessing the variable.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Fork and join phases in hull shaders allow dynamic indexing for
all output registers, not just the tessellation factor built-ins.
Moreover, the patch constant output register space is shared with
join phases, which can read back the outputs computed in the fork
phases, also allowing dynamic indexing.
In order to support this in a not overly complex way, use a private
array representing the entire patch constant space, and use epilogue
functions to assign them to the actual output variables.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Needed to support dynamically indexed output arrays.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Otherwise, if a private variable is used for the given output,
vkd3d_dxbc_compiler_emit_store_shader_output will write to the
private variable again instead of the actual output, and some
outputs may never be emitted. This is common in hull shaders.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Fixes an assertion when compiling shaders with more than four
clip or cull distances. Output arrays are arrays of scalars,
so shifting the write mask is not very meaningful.
Signed-off-by: Philip Rebohle <philip.rebohle@tu-dortmund.de>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Even if the shader doesn't explicitly declare it.
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This fixes a regression introduced by commit
91820630cb.
We need to take into account the combined write mask returned from
needs_private_io_variable().
Signed-off-by: Józef Kucia <jkucia@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
