This patch fixes a regression caused by 1b6b8bc9 when the implicit buffers
were moved to top of Metal buffer index range.
Fix vertex buffer binding counts when binding indexes are not consecutive,
or when additional synthetic buffer bindings are required to accommodate
vertex attributes that are outside the vertex buffer stride values.
Take into consideration that the app may bind more vertex attribute buffers
than a pipeline may consume, and don't allow these to overwrite any implicit
buffers sent to the shader.
MVKResourcesCommandEncoderState::encodeBinding() clear binding dirty flag
before calling lambda function to allow function operation to possibly
override and leave binding marked dirty.
Set tessellation vertex buffer indexes based on platform maximum vertex stage
buffer count, instead of hardcoding them, to preemptively avoid conflict with
implicit buffers should the platform counts ever change.
renderpass was writing timestamp before renderpass activity was complete.
MVKCommandBuffer tracks whether it contains a stage-based timestamp command,
and MVKCommandEncoder updates the timestamp command fence when ending any
Metal command encoder on such a MVKCommandBuffer.
MVKCommandEncoder reorder member variables to avoid layout gaps (unrelated).
MVKCommandBuffer update _commandCount even for single-use immediate command
encoding (unrelated).
Add VK_KHR_separate_depth_stencil_layouts and enable its features.
Metal generally ignores image layouts, so nothing further needed.
Update Whats_New.md and MoltenVK_Runtime_UserGuide.md documents
with recently-added extensions.
Passes almost all supported separate_layouts CTS tests, with the
few remaining tests failing due to issues with unrelated capabilities.
Unrelated cleanup:
- MVKDevice enable extensions before features.
- MVKDevice reorder ivar declarations to reduce
memory layout gaps, and define default values.
- Rename VkSemaphoreStyle to MVKSemaphoreStyle to
remove potential conflicts with Vulkan Vk name space.
- MVKDevice track enabling VK_KHR_dynamic_rendering extension features.
- Add MVKCmdBeginRendering and MVKCmdEndRendering command objects
and associated command pools.
- Add MVKCommandEncoder::beginRendering() which dynamically instantiates
temporary MVKRenderpass and (imageless) MVKFramebuffer objects to handle
equivalent renderpass operations.
- MVKCommandEncoder retain and release subpass and framebuffer objects
to allow transient dynamic instances to be created and destroyed.
- Add support functions to create MVKRenderpass and (imageless)
MVKFramebuffer objects from VkRenderingInfo.
- MVKRenderpass track VkRenderingFlags.
- MVKRenderSubpass create and track an internal VkPipelineRenderingCreateInfo.
- MVKRenderPassAttachment support LoadOp and StoreOp behavior for
dynamic rendering suspend and resume.
- Refactor viewMask processing to permit operations on viewMask outside of
MVKGraphicsPipeline without requiring the presence of an MVKSubpass.
- MVKGraphicsPipeline extract VkPipelineRenderingCreateInfo from MVKRenderSubpass
if available, or VkGraphicsPipelineCreateInfo::pNext if not.
- MVKCommandBuffer track VkCommandBufferInheritanceRenderingInfo (currently unused).
- (unrelated) vulkan.mm move location of VK_EXT_sample_locations functions
for layout readability consistency.
buffer binding used by a subsequent pipeline that does not use push constants.
This error was previously introduced in 2a17f75, where a push constants
binding could override the Metal buffer binding 0 of a subsequent pipeline
that does not use push constants.
When pipeline binding is encoded, track which stages use push constants
and only encode push constants if the pipeline and stage uses them.
(unrelated) Make use of MVKResourcesCommandEncoderState::getPipeline() consistent.
Update README.md and MoltenVK_Runtime_UserGuide.md to document the impact
of the VK_KHR_portability_enumeration extension during runtime loading on macOS.
Update MoltenVK version to 1.1.10.
Supports only setting custom sample locations in subpasses via
vkBeginRenderpass. Does not support setting custom sample locations via
vkCmdBindPipeline or vkCmdSetSampleLocationsEXT, although collects that
info for possible future enhancements.
- MVKPhysicalDevice track platform support and respond to property queries.
- MVKCmdBeginRenderPassBase collect subpass custom sample locations.
- MVKPipeline support dynamic state values beyond 31.
- MVKPipeline collect custom sample locations.
- Add MVKCmdSetSampleLocations to support vkCmdSetSampleLocations
to collect dynamic custom sample locations.
- MVKCommandEncoder support collecting custom sample positions from subpass
and dynamic, and set into MTLRenderPassDescriptor for each Metal render pass.
- MVKArrayRef add assignment operator.
- Add MVKPhysicalDeviceMetalFeatures::programmableSamplePositions.
- Update VK_MVK_MOLTENVK_SPEC_VERSION to version 34.
- MVKCommandBuffer.h remove obsolete comment documentation.
- Update Whats_New.md.
For pipeline layout compatibility, consume the Metal resource indexes in this order:
- Consume a fixed number of Metal buffer indexes for Metal argument buffers,
but only consume them if Metal argument buffers are not being used.
- Consume push constants Metal buffer index before descriptor set resources,
but only consume it if the stage uses push constants.
- Consume descriptor set bindings.
In MVKPipelineLayout, separate tracking resource counts from push constants
indexes, and move push constant indexes ahead of descriptor bindings.
In MVKPipeline, track which stages use push constants.
Remove unused and obsolete function declaration in MVKDescriptorSet.h.
For a resource object that can be retained by descriptors beyond its lifetime,
release memory resources when the object is destroyed by the app. This includes
objects of type MVKBuffer, MVKBufferView, MVKImageView, and MVKSampler.
When the app destroys an MVKBuffer, also detach from the MVKDeviceMemory,
to fix a potential race condition when the app updates the descriptor on
one thread while also freeing the MVKDeviceMemory on another thread.
MVKImageView guard against detached planes while in descriptor.
Add comment to clarify how destroy() is called from release().
For GPUs that round float clear colors down, a half-ULP adjustment is performed
on normalized formats. But this adjustment should not be performed on SRGB formats,
which Vulkan requires to be treated as linear, with the value managed by the app.
Ensure non-Apple GPU's enable memory barriers.
A previous commit inadvertently disabled GPU memory barriers.
Change tests for memory barriers to runtime test for Apple GPU, instead of
build-time test for Apple Silicon, to accommodate running on Rosetta2, and
refactor tests for Apple Silicon and OS version on some macOS GPU feature settings.
For a combined depth-stencil format in a MVKImageView attachment with
VK_IMAGE_ASPECT_STENCIL_BIT, the attachment format may have been swizzled
to a stencil-only format. In this case, we want to guard against an attempt
to store the non-existent depth component.
Pass MVKImageView attachment to MVKRenderPassAttachment::encodeStoreAction()
and MVKRenderPassAttachment::populateMTLRenderPassAttachmentDescriptor() to
check attachment depth format component.
Consolidate calls to MVKImageView::populateMTLRenderPassAttachmentDescriptor() by calling
it from within MVKRenderPassAttachment::populateMTLRenderPassAttachmentDescriptor().
When flattening shader inputs for stage_in, which are to be read from a buffer
that was populated as nested structs during an earlier stage, the structs will
be aligned according to C++ rules, which can affect the alignment of the first
member of the flattened input struct.
Add SPIRVShaderOutput::firstStructMemberAlignment to track the alignment
requirements of the first member of a nested structure, and recursively
determine the alignment of the first member of each nested output structure.
Move sizeOfOutput() from MVKPipeline.mm to SPIRVReflection.h,
rename to getShaderOutputSize(), and add getShaderOutputAlignment()
to extract member alignment.
If using GPU counters, on all Apple GPUs lock timestampPeriod to 1.0,
since Apple GPUs use nanoseconds, and on non-Apple GPUs, dynamically
adapt value of timestampPeriod by correlating GPU ticks with GPU ticks.
If using CPU sync, set timestampPeriod to OS CPU timestamp tick period.
Previously, on Apple Silicon (iOS, tvOS & macOS M1), we tried to guess
deviceID from GPU parameters, but this is becoming harder as the types
of Apple Silicon is growing, and the actual device SoC itself is less
relevant that the GPU capabilities. So we now set deviceID from the
combination of OS version and GPU type.
Rename MVKDevice::getHighestMTLFeatureSet() to getHighestGPUCapability().
Older Metal does not support rendering without subpass attachments.
In this case, a dummy attachment with a dummy MTLTexture is created
whenever the subpass begins, but was not being correctly used.
Move the creation, retaining, and releasing of the dummy MTLTexture to
MVKFramebuffer, where the extent and layer count is known and can be reused.
Pass framebuffer to MVKCommandEncoder::beginRenderpass() and remember
current framebuffer in MVKCommandEncoder.
Add getFramebufferExtent() and getFramebufferLayerCount() to MVKCommandEncoder.
Pass framebuffer to MVKRenderSubpass:populateMTLRenderPassDescriptor() and
retrieve dummy MTLTexture from framebuffer.
Do not use MTLEvent for VkSemaphore under Rosetta2.
Remove compile test for MVK_MACOS_APPLE_SILICON and MVK_APPLE_SILICON when testing
for Apple GPU families, to allow x86 builds to test for Apple GPU under Rosetta2.
Simplify identifying M1 GPU. All M1 SoCs currently support the A14 (Apple7) GPU.
Support compiling MSL 2.4 in runtime pipelines and MoltenVKShaderConverterTool.
Fix issue where MSL 2.3 only available on Apple Silicon, even on macOS.
Update to latest SPIRV-Cross (unrelated to Rosetta2).
Add build environment variable MVK_HIDE_VULKAN_SYMBOLS. to allow MoltenVK
to be built with static Vulkan API symbols hidden, to avoid library linking
conflicts when bound to a Vulkan Loader that also exports identical symbols.
The default value of MVK_HIDE_VULKAN_SYMBOLS is 0,
meaning Vulkan static symbols are exposed by default.
Add MVK_PUBLIC_VULKAN_SYMBOL directive to mark each Vulkan call symbols
for exporting or hiding.
Update the MoltenVK Xcode project to add the MVK_HIDE_VULKAN_SYMBOLS build
setting, and set the ENABLE_TESTABILITY build setting to NO, because it
conflicts with stripping symbols.
Update MoltenVK version to 1.1.7.
Metal does not provide direct support for BC1_RGB formats
(VK_FORMAT_BC1_RGB_UNORM_BLOCK & VK_FORMAT_BC1_RGB_SRGB_BLOCK).
We have been faking it by mapping these Vulkan formats to Metal
formats containing alpha (MTLPixelFormatBC7_RGBAUnorm &
MTLPixelFormatBC7_RGBAUnorm_sRGB, respectively),
and advertising support for BC1_RGB formats.
However, this triggers CTS failures, because the BC1_RGBA formats can return
an alpha value of 0.0 when constructed that way, whereas the BC1_RGB formats
always expect 1.0 (opaque) to be returned.
This change moves to indirect support for BC1_RGB formats. They will still
be covered by MTLPixelFormatBC7_RGBAUnorm & MTLPixelFormatBC7_RGBAUnorm_sRGB,
and will effectively work (except transparency), but are no longer advertised
through physical device format and image format queries.
Dynamic pipeline state set before the pipeline is set was reading
dynamic flags from previous pipeline. This is fixed here by accepting
the dynamic state, but deferring the decision to use either dynamic
or static state until the pipeline is encoded.
An MVKImageView that renders to only one layer of a multilayer MVKImage
is not performing multilayer-rendering. Only validate multilayer-rendering
when it is definitely requested in an MVKImageView, instead of presuming
when a multilayer MVKImage is marked for rendering.
Check that appropriate GPU counter set is supported on
the device before creating GPU counter sample buffer.
Don't attempt to timestamp using GPU counters unless
a GPU sample counter buffer has been created.
Support maximum point primitive size of 511.
Update to latest SPIRV-Cross version to add support
for OpSpecConstantOp ops OpQuantizeToF16 and OpSRem.
Update MoltenVK version to 1.1.6.
MSL: Support row-major transpose when storing matrix from constant RHS matrix.
MSL: Fix casting in constant expressions with different sizes.
MSL: Fix duplicate gl_Position outputs when gl_Position defined but unused.
Apple GPUs do not support rendering/writing to an attachment and then reading
from that attachment within a single Metal renderpass. On Apple Silicon, restart
the Metal renderpass if an input attachment is cleared inside renderpass.
Don't clear render area when restarting Metal renderpass, as it should not occur,
and itself causes a recursive loop restarting the renderpass as a result.
Add MVKCommandUse::kMVKCommandUseRestartSubpass.
MVKCommandEncoder::beginMetalRenderPass() pass MVKCommandUse to help determine
Metal renderpass attachment load and clearing options.
Rename MVKRenderPassAttachment::shouldUseClearAttachment() to
shouldClearAttachment(), pass whether testing for stencil clearing,
and check stencil clearing distinct from depth clearing.
Refactor MVKRenderSubpass::populateClearAttachments() to work with this.
Remove MVKRenderPassAttachment::getAttachmentStencilLoadOp() as obsolete.
Disable MVK_ALLOW_METAL_FENCES by default.
Disable use of MTLEvent on NVIDIA.
By default, use MTLEvent for VkSemaphore everywhere except NVIDIA.
By default, use CPU synchronization on NVIDIA.
These changes fix a large number of CTS synchronization test failures.
