Overzealous warning on vkGetPhysicalDeviceSurfacePresentModesKHR

[karl-lunarg]

Issue by haasn (MIGRATED)
Sunday Sep 17, 2017 at 05:21 GMT
Originally opened as KhronosGroup/Vulkan-LoaderAndValidationLayers#2071

---

In this code:

VkPresentModeKHR supported_modes[VK_PRESENT_MODE_RANGE_SIZE_KHR];
int num_modes = ARRAY_SIZE(supported_modes);
vkGetPhysicalDeviceSurfacePresentModesKHR(physd, surf, &num_modes, supported_modes);

I get an error as follows:

vkGetPhysicalDeviceSurfacePresentModesKHR() called with non-NULL pPresentModeCount; but no prior positive value has been seen for pPresentModeCount

But I believe common sense dictates that the number of supported swapchain modes can never exceed the number of swapchain modes. I think this warning is a bit too aggressive in cases like these.

[karl-lunarg]

Comment by haasn (MIGRATED)
Sunday Sep 17, 2017 at 05:22 GMT

---

To clarify, the expected behavior would be for this layer check to look at the contents of *pPresentModeCount, and only warn if it's lower than VK_PRESENT_MODE_RANGE_SIZE_KHR.

[karl-lunarg]

Comment by krOoze (MIGRATED)
Tuesday Sep 19, 2017 at 13:16 GMT

---

Problem with common sense is that it is wrong more often than not.
Firstly you should not be using VK_PRESENT_MODE_RANGE_SIZE_KHR. TMK it is deprecated. And always was non-standard anyway.
Secondly it is defined as VK_PRESENT_MODE_RANGE_SIZE_KHR = (VK_PRESENT_MODE_FIFO_RELAXED_KHR - VK_PRESENT_MODE_IMMEDIATE_KHR + 1) which results in 4, but if you look closely, there are 6 present modes including the extended ones.
Thirdly as you can see, the size is not constexpr by nature. New extensions may add to this list at any point, even if your older vulkan.h does not know the enumerants and does not account them into _RANGE_SIZE_ yet.

Such warning simply says you are not using the API as intended. Though as I modified similar commands I used to think it is bit overzealous too (there's also VK_INCOMPLETE). Then again, people seem to keep trying to cheat around the need to query the array size first.

[jzulauf-lunarg]

This warning will be reinstated in #24