Prefetch.h
#ifndef HALIDE_PREFETCH_H
#define HALIDE_PREFETCH_H
/** \file
* Defines the lowering pass that injects prefetch calls when prefetching
* appears in the schedule.
*/
#include <map>
#include <string>
#include <vector>
namespace Halide {
struct Target;
namespace Internal {
class Function;
struct PrefetchDirective;
struct Stmt;
/** Inject placeholder prefetches to 's'. This placholder prefetch
* does not have explicit region to be prefetched yet. It will be computed
* during call to \ref inject_prefetch. */
Stmt inject_placeholder_prefetch(const Stmt &s, const std::map<std::string, Function> &env,
const std::string &prefix,
const std::vector<PrefetchDirective> &prefetches);
/** Compute the actual region to be prefetched and place it to the
* placholder prefetch. Wrap the prefetch call with condition when
* applicable. */
Stmt inject_prefetch(const Stmt &s, const std::map<std::string, Function> &env);
/** Reduce a multi-dimensional prefetch into a prefetch of lower dimension
* (max dimension of the prefetch is specified by target architecture).
* This keeps the 'max_dim' innermost dimensions and adds loops for the rest
* of the dimensions. If maximum prefetched-byte-size is specified (depending
* on the architecture), this also adds an outer loops that tile the prefetches. */
Stmt reduce_prefetch_dimension(Stmt stmt, const Target &t);
/** Hoist all the prefetches in a Block to the beginning of the Block.
* This generally only happens when a loop with prefetches is unrolled;
* in some cases, LLVM's code generation can be suboptimal (unnecessary register spills)
* when prefetches are scattered through the loop. Hoisting to the top of the
* loop is a good way to mitigate this, at the cost of the prefetch calls possibly
* being less useful due to distance from use point. (This is a bit experimental
* and may need revisiting.) See also https://bugs.llvm.org/show_bug.cgi?id=51172 */
Stmt hoist_prefetches(const Stmt &s);
} // namespace Internal
} // namespace Halide
#endif