https://github.com/halide/Halide
Raw File
Tip revision: bf3b66f9d5ce84bd85f0390c9e198b2e8f7bd554 authored by Z Stern on 24 September 2020, 18:00:10 UTC
Add atomic update support to thread_pool_common parallel for implementations.
Tip revision: bf3b66f
LoopCarry.h
#ifndef HALIDE_LOOP_CARRY_H
#define HALIDE_LOOP_CARRY_H

#include "Expr.h"

namespace Halide {
namespace Internal {

/** Reuse loads done on previous loop iterations by stashing them in
 * induction variables instead of redoing the load. If the loads are
 * predicated, the predicates need to match. Can be an optimization or
 * pessimization depending on how good the L1 cache is on the architecture
 * and how many memory issue slots there are. Currently only intended
 * for Hexagon. */
Stmt loop_carry(Stmt, int max_carried_values = 8);

}  // namespace Internal
}  // namespace Halide

#endif
back to top