amber
amber copied to clipboard
Amber is a multi-API shader test framework
Currently Vulkan Device::Initialize queries for the physical memory properties. This should be moved to be provided through the engine configuration and the call to `vkGetPhysicalDeviceMemoryProperties` removed.
Currently Vulkan Device::Initialize will query for the physical device properties. This should be changed to accept the properties through the engine config and the call to `vkGetPhysicalDeviceProperties` removed.
We should convert the clang bots to run with -fsanitize=address in all cases. This would catch any memory issues added in unittests.
If you see [tests/cases/probe_without_clear_command.expect_fail.amber](tests/cases/probe_without_clear_command.expect_fail.amber), we expect it fails with "FrameBuffer::ChangeFrameImageLayout new layout cannot be kProbe from kInit" message. We should check the error messages for tests that we expect failures.
Spec says ``` maxUniformBufferRange is the maximum value that can be specified in the range member of any VkDescriptorBufferInfo structures passed to a call to vkUpdateDescriptorSets for descriptors of type...
With #328 we generate either the function pointers or lambdas at build time for the vulkan calls. Since we want to add a run-time flag anyway to enable/disable the logging...
Currently we put all tests into `tests/cases`. We cannot directly use them on Android because we will not put glslang and spirv-as to Android. @dj2 @dneto0 please doublecheck this. Write...
Following options are newly added by #144 : * AMBER_SKIP_TESTS * AMBER_SKIP_SPIRV_TOOLS * AMBER_SKIP_SHADERC @dj2 do you think we should add kokoro tests for them?
When comparing pixel values the absolute errors you care about is strongly determined by the image format. E.g. with 32-bit floats per colour channel, you're in the range of 2**(-23)...
If some libraries in `third_party/` are already installed in the system, current Amber build adopts the system library instead of the one in `third_party/` and it can lead to build...