[llvm] [Offload] Allow allocations to be freed without knowing their type (PR #140889)

[Joseph Huber via llvm-commits]

================
@@ -639,6 +636,44 @@ struct CUDADeviceTy : public GenericDeviceTy {
     return OFFLOAD_SUCCESS;
   }
 
+  int free_non_blocking(void *TgtPtr) override {
----------------
jhuber6 wrote:

Another option is to only use the NVIDIA virtual memory API and then run a custom allocator on top of it for smaller allocations.

https://github.com/llvm/llvm-project/pull/140889