Disclosed is a system and method for a profiling system wherein the collection of profile data may be enabled and disabled during the execution of an instrumented computer program. The invention utilizes a dedicated bit in a condition register which is checked each time a block of instrumentation code is to be executed. If the bit is enabled, the instrumentation code is executed and profile data is collected. If the bit is not enabled, then the instrumentation block is skipped and not executed and no profile information is collected. The enabling and disabling of the bit in the condition register may be implemented in hardware, in software, or in a combination of the two.

Software profiler with runtime control to enable and disable instrumented executable

The growing gap between the speed of memory access and cache access has made cache misses an influential factor in program efficiency. Much effort has been spent recently on reducing the number of cache misses during program run. This effort includes wise rearranging of program code, cache-conscious data placement, and algorithmic modifications that improve the program cache behavior. In this work we investigate the complexity of finding the optimal placement of objects (or code) in the memory, in the sense that this placement reduces the cache misses to the minimum. We show that this problem is one of the toughest amongst the interesting algorithmic problems in computer science. In particular, suppose one is given a sequence of memory accesses and one has to place the data in the memory so as to minimize the number of cache misses for this sequence. We show that if P ≠ NP, then one cannot efficiently approximate the optimal solution even up to a very liberal approximation ratio. Thus, this problem joins the small family of extremely inapproximable optimization problems. The other two famous members in this family are minimum coloring and maximum clique.

The hardness of cache conscious data placement

A method, apparatus, and computer instructions in a data processing system for processing instructions are provided. Instructions are received at a processor in the data processing system. If a selected indicator is associated with the instruction, counting of each event associated with the execution of the instruction is enabled. In some embodiments, a compiler may obtain performance profile data, annotated by output obtained from the use of the performance indicators and counters, along with the instructions/data of the computer program and use this information to optimize the manner by which the computer program is executed, instructions/data are stored, and the like. The optimizations may be to optimize the instruction paths, optimize the time spent in initial application load, the manner by which the cache and memory is utilized, and the like.

Method and apparatus for optimizing code execution using annotated trace information having performance indicator and counter information

A method and apparatus in a data processing system for measuring events associated with the execution of instructions are provided. Instructions are received at a processor in the data processing system. If a selected indicator is associated with the instruction, counting of each event associated with the execution of the instruction is enabled. In some embodiments, the performance indicators, counters, thresholds, and other performance monitoring structures may be stored in a page table that is used to translate virtual addresses into physical storage addresses. A standard page table is augmented with additional fields for storing the performance monitoring structures. These structures may be set by the performance monitoring application and may be queried and modified as events occur that require access to physical storage.

Method and apparatus for maintaining performance monitoring structures in a page table for use in monitoring performance of a computer program

We investigate the complexity of finding an optimal placement of objects (or code) in the memory, in the sense that this placement reduces the number of cache misses during program execution to the minimum. We show that this problem is one of the toughest amongst the interesting NP optimization problems in computer science, in particular, suppose one is given a sequence of memory accesses and one has to place the data in the memory so as to minimize the number of cache misses for this sequence. We show that if P ≠ NP, then one cannot efficiently approximate the optimal solution even up to a very liberal approximation ratio. Thus, this problem joins the family of extremely inapproximable optimization problems. Two famous members in this family are minimum graph coloring and maximum clique. In light of this harsh lower bound only mild approximation ratios can be obtained. We provide an algorithm that can map arbitrary access sequences within such a mild ratio.Next, we study the information loss when compressing the access sequence, keeping only pairwise relations. We show that the reduced information hides the optimal solution and highlights solutions that are far from optimal. Furthermore, we show that even if one restricts his attention to pairwise information, finding a good placement is computationally difficult.

In this paper we describe how a profiling system can be successfully used to restructure the components of an operating system for improved overall performance. We discuss our choice of a profiling system and how it was applied to the AS/400® (Application System/400®) operating system for the purpose of reordering code. Previous work in the industry has been mainly useful only for application programs. Our work demonstrates how such techniques can be applied to operating system code, while preserving maintainability of the operating system in the customer's environment.

W. J. Schmidt

Papers

Profile-directed restructuring of operating system code