From 09854a4596019ddb3bb315b8836b5c88e718cdc7 Mon Sep 17 00:00:00 2001
From: Ellie Hermaszewska <ellieh@nvidia.com>
Date: Tue, 12 Sep 2023 11:13:11 +0800
Subject: Add Mesh and Task shader support to GFX (#3190)

* Bump vulkan headers

Also just use vulkan-headers as a submodule

* Add drawMeshTasks to gfx graphics pipelines

* Add DispatchMesh overload with no payload, with GLSL intrinsic

* Require spirv 1.4 for mesh shaders

* Add vulkan mesh shader feature discovery

* Add mesh shader stage bits to vk-util

* Add mesh and task shader support to render-test

* Add mesh and task tests

* Preserve "payload" specifier in task shaders

* Add mesh shader pipeline support to gfx

* Add TODO

* Add numThreads attribute for amplification stage

* Add payload to task shader test

* Drop dependency on d3dx12

* Allow passing payloads from task to mesh shaders

* regenerate vs projects

* check DispatchMesh name correctly

* Add mesh shader tests to failing tests

* Detect wave-ops feature on vulkan

* Add fuse-product to expected failures

This fails because the global varaible `count` is not initialized

* Add required extension to WaveMaskMatch SPIR-V impl

* Remove meshShader member from pipeline desc

* Identify mesh shader support on d3d12
---
 tools/gfx/vulkan/vk-device.cpp | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

(limited to 'tools/gfx/vulkan/vk-device.cpp')

diff --git a/tools/gfx/vulkan/vk-device.cpp b/tools/gfx/vulkan/vk-device.cpp
index 015be5f15..e71337ef6 100644
--- a/tools/gfx/vulkan/vk-device.cpp
+++ b/tools/gfx/vulkan/vk-device.cpp
@@ -430,6 +430,10 @@ Result DeviceImpl::initVulkanInstanceAndDevice(
         extendedFeatures.atomicFloatFeatures.pNext = deviceFeatures2.pNext;
         deviceFeatures2.pNext = &extendedFeatures.atomicFloatFeatures;
 
+        // mesh shader features
+        extendedFeatures.meshShaderFeatures.pNext = deviceFeatures2.pNext;
+        deviceFeatures2.pNext = &extendedFeatures.meshShaderFeatures;
+
         if (VK_MAKE_VERSION(majorVersion, minorVersion, 0) >= VK_API_VERSION_1_2)
         {
             extendedFeatures.vulkan12Features.pNext = deviceFeatures2.pNext;
@@ -566,6 +570,16 @@ Result DeviceImpl::initVulkanInstanceAndDevice(
             m_features.add("realtime-clock");
         }
 
+        if (extendedFeatures.meshShaderFeatures.meshShader)
+        {
+            deviceExtensions.add(VK_EXT_MESH_SHADER_EXTENSION_NAME);
+
+            extendedFeatures.meshShaderFeatures.pNext = (void*)deviceCreateInfo.pNext;
+            deviceCreateInfo.pNext = &extendedFeatures.meshShaderFeatures;
+
+            m_features.add("mesh-shader");
+        }
+
         if (_hasAnySetBits(
                 extendedFeatures.vulkan12Features,
                 offsetof(VkPhysicalDeviceVulkan12Features, pNext) + sizeof(void*)))
@@ -578,10 +592,30 @@ Result DeviceImpl::initVulkanInstanceAndDevice(
             VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PROPERTIES_2 };
         VkPhysicalDeviceRayTracingPipelinePropertiesKHR rtProps = {
             VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_RAY_TRACING_PIPELINE_PROPERTIES_KHR };
+        VkPhysicalDeviceSubgroupProperties subgroupProps = {
+            VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SUBGROUP_PROPERTIES };
+        rtProps.pNext = extendedProps.pNext;
         extendedProps.pNext = &rtProps;
+        subgroupProps.pNext = extendedProps.pNext;
+        extendedProps.pNext = &subgroupProps;
         m_api.vkGetPhysicalDeviceProperties2(m_api.m_physicalDevice, &extendedProps);
         m_api.m_rtProperties = rtProps;
 
+        // Approximate DX12's WaveOps boolean
+        if(subgroupProps.supportedOperations &
+           ( VK_SUBGROUP_FEATURE_BASIC_BIT
+           | VK_SUBGROUP_FEATURE_VOTE_BIT
+           | VK_SUBGROUP_FEATURE_ARITHMETIC_BIT
+           | VK_SUBGROUP_FEATURE_BALLOT_BIT
+           | VK_SUBGROUP_FEATURE_SHUFFLE_BIT
+           | VK_SUBGROUP_FEATURE_SHUFFLE_RELATIVE_BIT
+           | VK_SUBGROUP_FEATURE_CLUSTERED_BIT
+           | VK_SUBGROUP_FEATURE_QUAD_BIT
+           | VK_SUBGROUP_FEATURE_PARTITIONED_BIT_NV))
+        {
+            m_features.add("wave-ops");
+        }
+
         uint32_t extensionCount = 0;
         m_api.vkEnumerateDeviceExtensionProperties(
             m_api.m_physicalDevice, NULL, &extensionCount, NULL);
-- 
cgit v1.2.3