Handle device orientation with Vulkan pre-rotation

This article describes how to efficiently handle device rotation in your Vulkan application by implementing pre-rotation.

With Vulkan, you can specify much more information about rendering state than you can with OpenGL. With Vulkan, you must explicitly implement things that are handled by the driver in OpenGL, such as device orientation and its relationship to render surface orientation. There are three ways that Android can handle reconciling the render surface of the device with the device orientation:

  1. The Android OS can use the device's Display Processing Unit (DPU), which can efficiently handle surface rotation in hardware. Available on supported devices only.
  2. The Android OS can handle surface rotation by adding a compositor pass. This will have a performance cost depending on how the compositor has to deal with rotating the output image.
  3. The application itself can handle the surface rotation by rendering a rotated image onto a render surface that matches the current orientation of the display.

Which of these methods should you use?

Currently, there's no way for an application to know whether surface rotation handled outside of the application will be free. Even if there is a DPU to take care of this for you, there will still likely be a measurable performance penalty to pay. If your application is CPU-bound, this becomes a power issue due to the increased GPU usage by the Android Compositor, which is usually running at a boosted frequency. If your application is GPU bound, then the Android Compositor can also preempt your application's GPU work, causing additional performance loss.

When running shipping titles on the Pixel 4XL, we have seen that SurfaceFlinger (the higher-priority task that drives the Android Compositor):

  • Regularly preempts the application’s work, causing 1-3ms hits to frametimes, and

  • Puts increased pressure on the GPU’s vertex/texture memory, because the Compositor has to read the entire framebuffer to do its composition work.

Handling orientation properly stops GPU preemption by SurfaceFlinger almost entirely, while the GPU frequency drops 40% as the boosted frequency used by the Android Compositor is no longer needed.

To ensure surface rotations are handled properly with as little overhead as possible, as seen in the preceding case, you should implement method 3. This is known as pre-rotation. This tells the Android OS that your app handles the surface rotation. You can do so by passing surface transform flags that specify the orientation during swapchain creation. This stops the Android Compositor from doing the rotation itself.

Knowing how to set the surface transform flag is important for every Vulkan application. Applications tend to either support multiple orientations or support a single orientation where the render surface is in a different orientation to what the device considers its identity orientation. For example, a landscape-only application on a portrait-identity phone, or a portrait-only application on a landscape-identity tablet.

Modify AndroidManifest.xml

To handle device rotation in your app, begin by changing the application’s AndroidManifest.xml file to tell Android that your app will handle orientation and screen size changes. This prevents Android from destroying and recreating the Android Activity and calling the onDestroy() function on the existing window surface when an orientation change occurs. This is done by adding the orientation (to support API level <13) and screenSize attributes to the activity’s configChanges section:

<activity android:name="android.app.NativeActivity"
          android:configChanges="orientation|screenSize">

If your application fixes its screen orientation using the screenOrientation attribute, you don't need to do this. Also, if your application uses a fixed orientation then it will only need to set up the swapchain once on application startup/resume.

Get the Identity Screen Resolution and Camera Parameters

Next, detect the device’s screen resolution associated with the VK_SURFACE_TRANSFORM_IDENTITY_BIT_KHR value. This resolution is associated with the identity orientation of the device, and is therefore the one that the swapchain will always need to be set to. The most reliable way to get this is to make a call to vkGetPhysicalDeviceSurfaceCapabilitiesKHR() at application startup, and store the returned extent. Swap the width and height based on the currentTransform that's also returned in order to ensure that you are storing the identity screen resolution:

VkSurfaceCapabilitiesKHR capabilities;
vkGetPhysicalDeviceSurfaceCapabilitiesKHR(physDevice, surface, &capabilities);

uint32_t width = capabilities.currentExtent.width;
uint32_t height = capabilities.currentExtent.height;
if (capabilities.currentTransform & VK_SURFACE_TRANSFORM_ROTATE_90_BIT_KHR ||
    capabilities.currentTransform & VK_SURFACE_TRANSFORM_ROTATE_270_BIT_KHR) {
  // Swap to get identity width and height
  capabilities.currentExtent.height = width;
  capabilities.currentExtent.width = height;
}

displaySizeIdentity = capabilities.currentExtent;

displaySizeIdentity is a VkExtent2D structure that we use to store said identity resolution of the app's window surface in the display’s natural orientation.

Detect Device Orientation Changes (Android 10+)

The most reliable way to detect an orientation change in your application is to verify whether the vkQueuePresentKHR() function returns VK_SUBOPTIMAL_KHR. For example:

auto res = vkQueuePresentKHR(queue_, &present_info);
if (res == VK_SUBOPTIMAL_KHR){
  orientationChanged = true;
}

Note: This solution only works on devices running Android 10 and later. These versions of Android return VK_SUBOPTIMAL_KHR from vkQueuePresentKHR(). We store the result of this check in orientationChanged, a booleanthat's accessible from the applications' main rendering loop.

Detect Device Orientation Changes (Pre-Android 10)

For devices running Android 10 or older, a different implementation is needed, because VK_SUBOPTIMAL_KHR is not supported.

Using Polling

On pre-Android 10 devices you can poll the current device transform every pollingInterval frames, where pollingInterval is a granularity decided on by the programmer. The way you do this is by calling vkGetPhysicalDeviceSurfaceCapabilitiesKHR() and then comparing the returned currentTransform field with that of the currently stored surface transformation (in this code example stored in pretransformFlag).

currFrameCount++;
if (currFrameCount >= pollInterval){
  VkSurfaceCapabilitiesKHR capabilities;
  vkGetPhysicalDeviceSurfaceCapabilitiesKHR(physDevice, surface, &capabilities);

  if (pretransformFlag != capabilities.currentTransform) {
    window_resized = true;
  }
  currFrameCount = 0;
}

On a Pixel 4 running Android 10, polling vkGetPhysicalDeviceSurfaceCapabilitiesKHR() took between .120-.250ms and on a Pixel 1XL running Android 8, polling took .110-.350ms.

Using Callbacks

A second option for devices running below Android 10 is to register an onNativeWindowResized() callback to call a function that sets the orientationChanged flag, signaling to the application an orientation change has occurred:

void android_main(struct android_app *app) {
  ...
  app->activity->callbacks->onNativeWindowResized = ResizeCallback;
}

Where ResizeCallback is defined as:

void ResizeCallback(ANativeActivity *activity, ANativeWindow *window){
  orientationChanged = true;
}

The problem with this solution is that onNativeWindowResized() only gets called for 90-degree orientation changes, such as going from landscape to portrait or vice versa. Other orientation changes will not trigger the swapchain recreation. For example, a change from landscape to reverse-landscape will not trigger it, requiring the Android compositor to do the flip for your application.

Handling the Orientation Change

To handle the orientation change, call the orientation change routine at the top of the main rendering loop when the orientationChanged variable is set to true. For example:

bool VulkanDrawFrame() {
 if (orientationChanged) {
   OnOrientationChange();
}

You do all the work necessary to recreate the swapchain within the OnOrientationChange() function. This means that you:

  1. Destroy any existing instances of Framebuffer and ImageView,

  2. Recreate the swapchain while destroying the old swapchain (which will be discussed next), and

  3. Recreate the Framebuffers with the new swapchain’s DisplayImages. Note: Attachment images (depth/stencil images, for example) usually don't need to be recreated as they are based on the identity resolution of the pre-rotated swapchain images.

void OnOrientationChange() {
 vkDeviceWaitIdle(getDevice());

 for (int i = 0; i < getSwapchainLength(); ++i) {
   vkDestroyImageView(getDevice(), displayViews_[i], nullptr);
   vkDestroyFramebuffer(getDevice(), framebuffers_[i], nullptr);
 }

 createSwapChain(getSwapchain());
 createFrameBuffers(render_pass, depthBuffer.image_view);
 orientationChanged = false;
}

And at the end of the function you reset the orientationChanged flag to false to show that you have handled the orientation change.

Swapchain Recreation

In the previous section we mention having to recreate the swapchain. The first steps to doing so involves getting the new characteristics of the rendering surface:

void createSwapChain(VkSwapchainKHR oldSwapchain) {
   VkSurfaceCapabilitiesKHR capabilities;
   vkGetPhysicalDeviceSurfaceCapabilitiesKHR(physDevice, surface, &capabilities);
   pretransformFlag = capabilities.currentTransform;

With the VkSurfaceCapabilities struct populated with the new information, you can now check to see whether an orientation change has occurred by checking the currentTransform field. You'll store it for later in the pretransformFlag field as you will be needing it for later when you make adjustments to the MVP matrix.

To do so, specify the following attributes in the VkSwapchainCreateInfo struct:

VkSwapchainCreateInfoKHR swapchainCreateInfo{
  ...
  .sType = VK_STRUCTURE_TYPE_SWAPCHAIN_CREATE_INFO_KHR,
  .imageExtent = displaySizeIdentity,
  .preTransform = pretransformFlag,
  .oldSwapchain = oldSwapchain,
};

vkCreateSwapchainKHR(device_, &swapchainCreateInfo, nullptr, &swapchain_));

if (oldSwapchain != VK_NULL_HANDLE) {
  vkDestroySwapchainKHR(device_, oldSwapchain, nullptr);
}

The imageExtent field will be populated with the displaySizeIdentity extent that you stored at application startup. The preTransform field will be populated with the pretransformFlag variable (which is set to the currentTransform field of the surfaceCapabilities). You also set the oldSwapchain field to the swapchain that will be destroyed.

MVP Matrix Adjustment

The last thing you must do is to apply the pre-transformation by applying a rotation matrix to your MVP matrix. What this essentially does is apply the rotation in clip space so that the resulting image is rotated to the current device orientation. You can then simply pass this updated MVP matrix into your vertex shader and use it as normal without the need to modify your shaders.

glm::mat4 pre_rotate_mat = glm::mat4(1.0f);
glm::vec3 rotation_axis = glm::vec3(0.0f, 0.0f, 1.0f);

if (pretransformFlag & VK_SURFACE_TRANSFORM_ROTATE_90_BIT_KHR) {
  pre_rotate_mat = glm::rotate(pre_rotate_mat, glm::radians(90.0f), rotation_axis);
}

else if (pretransformFlag & VK_SURFACE_TRANSFORM_ROTATE_270_BIT_KHR) {
  pre_rotate_mat = glm::rotate(pre_rotate_mat, glm::radians(270.0f), rotation_axis);
}

else if (pretransformFlag & VK_SURFACE_TRANSFORM_ROTATE_180_BIT_KHR) {
  pre_rotate_mat = glm::rotate(pre_rotate_mat, glm::radians(180.0f), rotation_axis);
}

MVP = pre_rotate_mat * MVP;

Consideration - Non-Full Screen Viewport and Scissor

If your application is using a non-full screen viewport/scissor region, they will need to be updated according to the orientation of the device. This requires that you enable the dynamic Viewport and Scissor options during Vulkan’s pipeline creation:

VkDynamicState dynamicStates[2] = {
  VK_DYNAMIC_STATE_VIEWPORT,
  VK_DYNAMIC_STATE_SCISSOR,
};

VkPipelineDynamicStateCreateInfo dynamicInfo = {
  .sType = VK_STRUCTURE_TYPE_PIPELINE_DYNAMIC_STATE_CREATE_INFO,
  .pNext = nullptr,
  .flags = 0,
  .dynamicStateCount = 2,
  .pDynamicStates = dynamicStates,
};

VkGraphicsPipelineCreateInfo pipelineCreateInfo = {
  .sType = VK_STRUCTURE_TYPE_GRAPHICS_PIPELINE_CREATE_INFO,
  ...
  .pDynamicState = &dynamicInfo,
  ...
};

VkCreateGraphicsPipelines(device, VK_NULL_HANDLE, 1, &pipelineCreateInfo, nullptr, &mPipeline);

The actual computation of the viewport extent during command buffer recording looks like this:

int x = 0, y = 0, w = 500, h = 400;

glm::vec4 viewportData;

switch (device->GetPretransformFlag()) {
  case VK_SURFACE_TRANSFORM_ROTATE_90_BIT_KHR:
    viewportData = {bufferWidth - h - y, x, h, w};
    break;
  case VK_SURFACE_TRANSFORM_ROTATE_180_BIT_KHR:
    viewportData = {bufferWidth - w - x, bufferHeight - h - y, w, h};
    break;
  case VK_SURFACE_TRANSFORM_ROTATE_270_BIT_KHR:
    viewportData = {y, bufferHeight - w - x, h, w};
    break;
  default:
    viewportData = {x, y, w, h};
    break;
}

const VkViewport viewport = {
    .x = viewportData.x,
    .y = viewportData.y,
    .width = viewportData.z,
    .height = viewportData.w,
    .minDepth = 0.0F,
    .maxDepth = 1.0F,
};

vkCmdSetViewport(renderer->GetCurrentCommandBuffer(), 0, 1, &viewport);

The x and y variables define the coordinates of the top left corner of the viewport, while w and h define the width and height of the viewport respectively. The same computation can also be used to set the scissor test, and is included here for completeness:

int x = 0, y = 0, w = 500, h = 400;
glm::vec4 scissorData;

switch (device->GetPretransformFlag()) {
  case VK_SURFACE_TRANSFORM_ROTATE_90_BIT_KHR:
    scissorData = {bufferWidth - h - y, x, h, w};
    break;
  case VK_SURFACE_TRANSFORM_ROTATE_180_BIT_KHR:
    scissorData = {bufferWidth - w - x, bufferHeight - h - y, w, h};
    break;
  case VK_SURFACE_TRANSFORM_ROTATE_270_BIT_KHR:
    scissorData = {y, bufferHeight - w - x, h, w};
    break;
  default:
    scissorData = {x, y, w, h};
    break;
}

const VkRect2D scissor = {
    .offset =
        {
            .x = (int32_t)viewportData.x,
            .y = (int32_t)viewportData.y,
        },
    .extent =
        {
            .width = (uint32_t)viewportData.z,
            .height = (uint32_t)viewportData.w,
        },
};

vkCmdSetScissor(renderer->GetCurrentCommandBuffer(), 0, 1, &scissor);

Consideration - Fragment Shader Derivatives

If your application is using derivative computations such as dFdx and dFdy, additional transformations may be needed to account for the rotated coordinate system as these computations are executed in pixel space. This requires the app to pass some indication of the preTransform into the fragment shader (such as an integer representing the current device orientation) and use that to map the derivative computations properly:

  • For a 90 degree pre-rotated frame
    • dFdx must be mapped to dFdy
    • dFdy must be mapped to -dFdx
  • For a 270 degree pre-rotated frame
    • dFdx must be mapped to -dFdy
    • dFdy must be mapped to dFdx
  • For a 180 degree pre-rotated frame,
    • dFdx must be mapped to -dFdx
    • dFdy must be mapped to -dFdy

Conclusion

In order for your application to get the most out of Vulkan on Android, implementing pre-rotation is a must. The most important takeaways from this article are:

  • Ensure that during swapchain creation or recreation, the pretransform flag is set to match the flag returned by the Android operating system. This will avoid the compositor overhead.
  • Keep the swapchain size fixed to the identity resolution of the app's window surface in the display’s natural orientation.
  • Rotate the MVP matrix in clip space to account for the devices orientation, because the swapchain resolution/extent no longer updates with the orientation of the display.
  • Update viewport and scissor rectangles as needed by your application.

Sample App: Minimal Android pre-rotation