Change ifdef guard for VMA_USE_STL_SHARED_MUTEX to use __cplusplus version? about vulkanmemoryallocator

[adam-sawicki-a]

Hi,

First of all, thank you for using and appreciation of VMA. I'm glad to see that despite it is being developed mostly on PC Windows, it is useful also in Vulkan programming for other kinds of GPUs.

This problem shouldn't stop you from upgrading the library. Please note that every such configuration macro is handled like this: #ifndef MACRO #if CONDITION #define MACRO #endif #endif, so that you can override the logic inside even without modifying the source of VMA, by just defining this macro before include:

#define VMA_USE_STL_SHARED_MUTEX 0
#include "vk_mem_alloc.h"

But I agree with you - usage of shared_mutex should be enabled for any C++17 conforming compiler. There is one caveat though: Visual Studio doesn't define __cplusplus macro properly by default, so it has to be handled in other way.

I've submitted my proposed solution to "master" branch. Can you please check if it solves the problem for you?

[adam-sawicki-a]

I assume the issue is fixed now and I close the ticket. Please let me know if you find any problems with it.

[rumblehhh]

Yeah we definitely get good use out of VMA so thanks for the great library! Aha yeah of course we'll do as you suggest and disable STL_SHARED_MUTEX for now. Thanks for the fix in master. It looks fine to me.