Trait numa_gpu::runtime::memory::PageLock

source · [−]
pub trait PageLock {
    fn page_lock(&mut self) -> Result<()>;

    fn page_unlock(&mut self) -> Result<()>;
}
Expand description

A trait for memory that can be page-locked by CUDA.

Page-locked memory can be copied to the GPU more efficiently.

Invariant

A page-locked memory range must be page-unlocked before it is freed. Thus, the drop() method of the implementation target must call page_unlock().

Required methods

Implementors