Function numa_gpu::runtime::cuda_wrapper::memset_async

source · [−]
pub fn memset_async<T: DeviceCopy>(
    mem: LaunchableMutSlice<'_, T>, 
    value: i32, 
    stream: &Stream
) -> Result<()>
Expand description

Fill a launchable slice using the CUDA memset_async function

Limitations

  • The size of T must be an even multiple of a 32-bit integer.
  • Only fill values of type i32 are supported. Fill a T with an i32 value might result in an invalid initialization.