Go to file
2016-03-08 10:02:07 +01:00
tinyxml2@c8dad95d44 Fix issue #7, add #pragma once to generated header 2016-02-17 16:29:56 +01:00
vulkan Changed preliminary vk::Exception to std::system_error. 2016-03-08 10:02:07 +01:00
.gitmodules switch to TinyXML2 which is being fetches as submodule 2016-02-17 15:35:00 +01:00
CMakeLists.txt switch to TinyXML2 which is being fetches as submodule 2016-02-17 15:35:00 +01:00
LICENSE.txt Initial revision for Vulkan 1.0 2016-02-16 15:07:55 +01:00
README.md Added missing </code> in README.md 2016-03-07 09:12:58 +01:00
VkCppGenerator.cpp Changed preliminary vk::Exception to std::system_error. 2016-03-08 10:02:07 +01:00

Open-Source Vulkan C++ API

Vulkan is a C API and as such inherits all common pitfalls of using a general C programming library. The motivation of a low-level Vulkan C++ API is to avoid these common pitfalls by applying commonly known C++ features while keeping the overall structure of a Vulkan program and preserving the full freedom it provides as low-level graphics API. An additional guideline we followed was not to introduce additional runtime overhead by providing a header-only library with inline functions.

Have a look at the following piece of code which creates a VkImage:


  VkImageCreateInfo ci;
  ci.sType = VK_STRUCTURE_TYPE_IMAGE_CREATE_INFO;
  ci.pNext = nullptr;
  ci.flags = ...some flags...;
  ci.imageType = VK_IMAGE_TYPE_2D;
  ci.format = VK_FORMAT_R8G8B8A8_UNORM;
  ci.extent = VkExtent3D { width, height, 1 };
  ci.mipLevels = 1;
  ci.arrayLayers = 1;
  ci.samples = VK_SAMPLE_COUNT_1_BIT;
  ci.tiling = VK_IMAGE_TILING_OPTIMAL;
  ci.usage = VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT;
  ci.sharingMode = VK_SHARING_MODE_EXCLUSIVE;
  ci.queueFamilyIndexCount = 0;
  ci.pQueueFamilyIndices = 0;
  ci.initialLayout = VK_IMAGE_LAYOUT_UNDEFINED;
  vkCreateImage(device, &ci, allocator, &image));

There may be some issues that can happen when filling the structure which cannot be caught at compile time:

  • initialization of ci.sType using wrong enum values
  • uninitialized data fields (e.g. missing initialization of ci.mipLevels)
  • use of invalid bits for ci.flags (no type-safety for bits)
  • use of incorrect enums for fields (no type-safety for enums)

These initializations will most likely show up as random runtime errors, which usually are nasty and time-consuming to debug. Our auto-generated, C++ 11-conform layer uses commonly known C++ features like implicit initialization through constructors to avoid incorrect or missing initializations and introduces type-safety with scoped enums to turn explicit initialization errors into compile errors. Following is a list of features and conventions introduced by our Vulkan C++ layer:

  • works along the official C version of the API
  • defines all symbols within the 'vk' namespace and to avoid redundancy the vk/Vk/VK_ prefixes have been removed from all symbols, i.e. vk::ImageCreateInfo for VkImageCreateInfo.
  • camel case syntax with an 'e' prefix has been introduced for all enums, i.e. vk::ImageType::e2D (the prefix was a compromise, more about that later) removes the 'BIT' suffix from all flag related enums, i.e. vk::ImageUsage::eColorAttachment.
  • introduces constructors for all structs, which by default set the appropriate sType and all other values to zero.
  • encapsulates member variables of the structs with getter and setter functions, i.e. ci.imageType() to get a value and ci.imageType(vk::ImageType::e2D) to set a value.
  • introduces wrapper classes around the vulkan handles, i.e. vk::CommandBuffer for VkCommandBuffer
  • introduces member functions of those wrapper classes, that map to vulkan functions getting the corresponding vulkan handle as its first argument. The type of that handle is stripped from the function name, i.e. vk::Device::getProcAddr for vkGetDeviceProcAddr. Note the special handling for the class CommandBuffer, where most of the vulkan functions would just include "Cmd", instead of "CommandBuffer", i.e. vk::CommandBuffer::bindPipeline for vkCmdBindPipeline.

With those changes applied, the updated code snippet looks like this:


vk::ImageCreateInfo ci;
ci.flags(...some flags...);
ci.imageType(vk::ImageType::e2D);
ci.format(vk::Format::eR8G8B8A8Unorm);
ci.extent(vk::Extent3D { width, height, 1 });
ci.mipLevels(1);
ci.arrayLayers(1);
ci.samples(1);
ci.tiling(vk::ImageTiling::eOptimal);
ci.usage(vk::ImageUsage::eColorAttachment);
ci.sharingMode(vk::SharingMode::eExclusive);
  // ci.queueFamilyIndexCount(0)	// no need to set, already initialized
  // ci.pQueueFamilyIndices(0)	// no need to set, already initialized
ci.initialLayout(vk::ImageLayout::eUndefined);
device.createImage(&ci, allocator, &image);

Which is a total of 13 lines of code, versus 17 lines for the C version. In addition, this code is more robust as described above.

Type-safe Enums

Splitting up the C enums into a namespace and scoped enums resulted in two compilation issues. First some enums started with a digit like vk::ImageType::1D which resulted in a compilation error. Second, there's the risk that upper symbols like vk::CompositeAlphaFlagBitsKHR::OPAQUE do clash with preprocessor defines. In the given example OPAQUE has been defined in win32gdi.h resulting a compilation error.

To overcome those two issues the symbols have been converted to camel case and the prefix 'e' has been added so that each enum starts with a letter.

Improvements to Bit Flags

After those changes the code might look more familiar to C++ developers, but there is still no gain with regards to safety. With C++ features available we replaced all Vulkan enums with scoped enums to achieve type safety which already uncovered a few small issues in our code. The good thing with scoped enums is that there is no implicit casts to integer types anymore. The downside is that OR'ing the bits for the flags does not work anymore without an explicit cast. As a solution to this problem we have introduced a new vk::Flags<T> template which is used for all flags. This class supports the standard operations one usually needs on bitmasks like &=, |=, & and |. Except for the initialization with 0, which is being replaced by the default constructor, the vk::Flags<T> class works exactly like a normal bitmask with the improvement that it is impossible to set bits not specified by the corresponding enum. To generate a bit mask with two bits set write:


ci.usage = vk::ImageUsage::eColorAttachment | vk::ImageUsage::eStorage;

By adding the scoped enums and vk::Flags<T> the C++ API provides type safety for all enums and flags which is a big improvement. This leaves the remaining issue that the compiler might not detect uninitialized fields in structs. As a solution we have added constructors to all structs which accept all values defined by the corresponding struct.


vk::ImageCreateInfo ci(
  ...some flags..., vk::ImageType::e2D, vk::Format::eR8G8B8A8Unorm,
  vk::Extent3D { width, height, 1 }, 1, 1,
  vk::SampleCount::e1, vk::ImageTiling::eOptimal,
  vk::ImageUsage::eColorAttachment, vk::SharingMode::eExclusive,
  0, 0, vk::ImageLayout::eUndefined);

String conversions

At development time it can be quite handy to have a utility function that can convert an enum or flags to a string for debugging purposes. To achieve this, we have implemented getString(type) functions for all enums and flags. Calling getString(vk::SharingMode::eExclusive) will return 'Exclusive' and calling getString(vk::QueueFlagBits::eGraphics | vk::QueueFlagBits::eCompute) will return the concatenated string 'Graphics | Compute'.

Alternative Initialization of Structs

Another nice feature of those constructors is that sType is being initialized internally and thus is always correct.

Finally, we have added a default constructor to each struct which initializes all values to 0 to allow setting the values with the named parameter idiom which is similar to the designated initializer list of C99.


vk::ImageCreateInfo ci = vk::ImageCreateInfo()
  .flags(...some flags...)
  .imageType(vk::ImageType::e2D)
  .format(vk::Format::eR8G8B8A8Unorm)
  .extent(vk::Extent3D { width, height, 1 })
  .mipLevels(1)
  .arrayLayers(1)
  .samples(1)
  .tiling(vk::ImageTiling::eOptimal)
  .usage(vk::ImageUsage::eColorAttachment)
  .sharingMode(vk::SharingMode::eExclusive)
  // .queueFamilyIndexCount(0)	// no need to set, already initialized
  // .pQueueFamilyIndices(0)	// no need to set, already initialized
  .initialLayout(vk::ImageLayout::eUndefined);
device.createImage(&ci, allocator, &image);

Enhancements beyond the API

To provide a more object oriented feeling we're providing classes for each handle which include all Vulkan functions where the first parameter matches the handle. In addition to this we made a few changes to the signatures of the member functions

  • (count, T*) has been replaced by std::vector<T>
  • const char * has been replaced by std::string
  • vk::Result return values have been replaced by exceptions.
  • Functions with a single output value do return this value instead

Here are a few code examples:


  try {
    VkInstance nativeInstance = nullptr; // Fetch the instance from a favorite toolkit

    // create a vkcpp handle from a native handle
    vk::Instance i(nativeInstance);

    // operator=(VkInstance const &) is also supported
    i = nativeInstance;

    // Get VkInstance from vk::Instance 
    nativeInstance = i;

    // Get a std::vector as result of an enumeration call.
    std::vector physicalDevices = i.enumeratePhysicalDevices();
    vk::FormatProperties formatProperties = physicalDevices[0].getFormatProperties(vk::Format::eR8G8B8A8Unorm);

    vk::CommandBuffer commandBuffer = ...;
    vk::Buffer buffer = ...;

    // Accept std::vector as source for updateBuffer
    commandBuffer.updateBuffer(buffer, 0, {some values}); // update buffer with std::vector

    // Sometimes it's necessary to pass a nullptr to a struct. For this case we've added T& null() to all structs T as replacement for nullptr.
    device.allocateMemory(allocateInfo, vk::AllocationCallbacks::null());

  }
  catch (vk::Exception e)
  {
    std::cerr << "Vulkan failure: " << e.what() << std::endl;
  }

Usage

To start with the C++ version of the Vulkan API download header from GIT, put it in a vulkan subdirectory and add #include <vulkan/vk_cpp.h> to your source code.

To build the header for a given vk.xml specification continue with the following steps:

  • Build VkCppGenerator
  • Grab your favourite version vk.xml from Khronos
  • Up to Version 1.0.3 of the API there is a tiny bug in vk.xml. The section of the VK_KHR_display extension is missing one symbol which can easily be fixed by adding the following line
                <type name="VkDisplayPlaneAlphaFlagsKHR"/>

before this line:

                <type name="VkDisplayPlaneAlphaFlagBitsKHR"/>
  • Excute VkCppGenerator <vk.xml> to generate vk_cpp.h in the current working directory.

Build instructions for VkCppGenerator

  • Clone the repository: git clone https://github.com/nvpro-pipeline/vkcpp.git
  • Update submodules: git submodule update --init --recursive
  • Use CMake to generate a solution or makefile for your favourite build environment
  • Launch the build

Providing Pull Requests

NVIDIA is happy to review and consider pull requests for merging into the main tree of vkcpp for bug fixes and features. Before providing a pull request to NVIDIA, please note the following:

  • A pull request provided to this repo by a developer constitutes permission from the developer for NVIDIA to merge the provided changes or any NVIDIA modified version of these changes to the repo. NVIDIA may remove or change the code at any time and in any way deemed appropriate. Due to the required paperwork please refrain from providing pull requests for simple changes and file an issue describing a bug or the desired change instead.
  • Not all pull requests can be or will be accepted. NVIDIA will close pull requests that it does not intend to merge.
  • The modified files and any new files must include the unmodified NVIDIA copyright header seen at the top of all shipping files.