Legacy resources

Legacy memory resources provide synchronous allocation interfaces backed by the CUDA Runtime’s legacy allocation APIs. They are primarily intended for compatibility with older toolkits or platforms that do not support the newer memory pool-based resources. Prefer the modern memory resources where available.

For the full memory resource model and property system, see Memory Resources (Extended API).

cuda::mr::legacy_pinned_memory_resource

Provides pinned (page-locked) host allocations using cudaMallocHost and cudaFreeHost. This resource is synchronous-only and is intended as a compatibility fallback. For CUDA 12.6 and later, prefer cuda::pinned_memory_resource.

#include <cuda/memory_resource>

void use_legacy_pinned() {
  cuda::mr::legacy_pinned_memory_resource resource{};
  void* ptr = resource.allocate_sync(1024, 64);
  // Use memory...
  resource.deallocate_sync(ptr, 1024, 64);
}

cuda::mr::legacy_managed_memory_resource

Provides managed (unified) allocations using cudaMallocManaged and cudaFree. This resource is synchronous-only and accepts the CUDA attachment flags (cudaMemAttachGlobal / cudaMemAttachHost). Prefer cuda::managed_memory_resource when available.

#include <cuda/memory_resource>

void use_legacy_managed() {
  cuda::mr::legacy_managed_memory_resource resource{cudaMemAttachGlobal};
  void* ptr = resource.allocate_sync(1024, 64);
  // Use memory...
  resource.deallocate_sync(ptr, 1024, 64);
}