API Coverage Report#
These pages provide a diff between the SCALE Runtime API's headers and the NVIDIA documentation.
This resource makes it easy to determine which CUDA APIs are supported by SCALE when cross-compiling to AMD targets.
When compiling for NVIDIA GPUs, NVIDIA's own libraries are used. In that case, you'll have access to the same set of APIs as are provided by that version of the NVIDIA CUDA Toolkit.
The lists are based on the official Nvidia documentation and use the same layout.
Presentation#
The lists are presented using diff syntax. Missing APIs are prefixed with a -,
causing them to highlight red in the listing.
This makes it easy to see which entries are available and which may be missing.
For example:
const char * cudaGetErrorName(cudaError_t);
__device__ const char * cudaGetErrorName(cudaError_t);
const char * cudaGetErrorString(cudaError_t);
__device__ const char * cudaGetErrorString(cudaError_t);
cudaError_t cudaGetLastError();
-__device__ cudaError_t cudaGetLastError();
cudaError_t cudaPeekAtLastError();
-__device__ cudaError_t cudaPeekAtLastError();
In this example, functions cudaGetErrorName and cudaGetErrorString are available on both host and device.
Functions cudaGetLastError and cudaPeekAtLastError are available on host, and are not available on device.
Correctness#
The lists may say that something is unavailable when it's not the case. This may happen for a few reasons.
NVIDIA documentation may differ from what CUDA provides in reality.
An example of that is differences in const-ness of some function arguments.
In such cases SCALE may be forced to maintain "bug compatibility" and the
functions stop matching what NVIDIA documentation promises.
SCALE retains support for some old APIs NVIDIA have since deleted.
Found a mistake?