(*58()9$"2#$:0/,;58(03<10/2,>=?33@">"29 #A:0*/,B58(*C2"258/052,D3*>#$,,6-*0'/ 58@F,058*,+HG?!"*0"I"252J58K0/ ,6-*0'/ 030"6N*IO40"58DP)"58QF,058SRUT6252,D<0!2T6252,V52!8("9 "W5X3,06*9E,'Y58(*03C:0'/ X3,06*9E,'Y58(*03C 1622 *'\,20/2XD3Q#$,U-0/269EU,/52,X"58QF,0'58,+ I,2/2-K58X^528-3L2T6252,_0/252/,58('4-*0'2,Y 0C#$,058Z#>58,0@=`58a02T/2*(*C/,':b(/,058c+ \",25C0d@"3,152058[#;58!*03e0/252,/58( 5805f8(""52<00"58>b(3589$3,3*"*58QF058C-02,;"(3T Y2520'58258/,03@20'Q"3+ ] D,Q"...

Object-oriented programming languages confer many benefits, including abstraction, which lets the programmer hide the details of an object’s implementation from the object’s clients. Unfortunately, crossing abstraction boundaries often incurs a substantial run-time overhead in the form of frequent procedure calls. Thus, pervasive use of abstraction, while desirable from a design standpoint, may be impractical when it leads to inefficient programs. Aggressive compiler optimizations can reduce the overhead of abstraction. However, the long compilation times introduced by optimizing compilers delay the programming environment‘s responses to changes in the program. Furthermore, optimization also conflicts with source-level debugging. Thus, programmers are caught on the horns of two dilemmas: they have to choose between abstraction and efficiency, and between responsive programming environments and efficiency. This dissertation shows how to reconcile these seemingly contradictory goals by performing optimizations lazily. Four new techniques work together to achieve high performance and high responsiveness: • Type feedback achieves high performance by allowing the compiler to inline message sends based on information extracted from the runtime system. On average, programs run 1.5 times faster than the previous SELF system; compared to a commercial Smalltalk implementation, two medium-sized benchmarks run about three times faster. This level of performance is obtained with a compiler that is both simpler and faster than previous SELF compilers. • Adaptive optimization achieves high responsiveness without sacrificing performance by using a fast nonoptimizing compiler to generate initial code while automatically recompiling heavily used parts of the program with an optimizing compiler. On a previous-generation workstation like the SPARCstation-2, fewer than 200 pauses exceeded 200 ms during a 50-minute interaction, and 21 pauses exceeded one second. On a currentgeneration workstation, only 13 pauses exceed 400 ms. • Dynamic deoptimization shields the programmer from the complexity of debugging optimized code by transparently recreating non-optimized code as needed. No matter whether a program is optimized or not, it can always be stopped, inspected, and single-stepped. Compared to previous approaches, deoptimization allows more debugging while placing fewer restrictions on the optimizations that can be performed. • Polymorphic inline caching generates type-case sequences on-the-fly to speed up messages sent from the same call site to several different types of object. More significantly, they collect concrete type information for the optimizing compiler. With better performance yet good interactive behavior, these techniques make exploratory programming possible both for pure object-oriented languages and for application domains requiring higher ultimate performance, reconciling exploratory programming, ubiquitous abstraction, and high performance.

Dynamically-dispatched calls often limit the performance of object-oriented programs since object-oriented programming encourages factoring code into small, reusable units, thereby increasing the frequency of these expensive operations. Frequent calls not only slow down execution with the dispatch overhead per se, but more importantly they hinder optimization by limiting the range and effectiveness of standard global optimizations. In particular, dynamicallydispatched calls prevent standard interprocedural optimizations that depend on the availability of a static call graph. The SELF implementation described here offers two novel approaches to optimization. Type feedback speculatively inlines dynamically-dispatched calls based on profile information that predicts likely receiver classes. Adaptive optimization reconciles optimizing compilation with interactive performance by incrementally optimizing only the frequently-executed parts of a program. When combined, these two techniques result in a system that can execute programs significantly faster than previous systems while retaining much of the interactiveness of an interpreted system.

We have designed and implemented an optimizing source-to-source C++ compiler that reduces the frequency of virtual function calls. This technical report describes our preliminary experience with this system. The prototype implementation demonstrates the value of OO-specific optimization of C++. Despite some limitations of our system, and despite the low frequency of virtual function calls in some of the programs, optimization improves the performance of a suite of two small and six large C++ applications totalling over 90,000 lines of code by a median of 20% over the original programs and reduces the number of virtual function calls by a median factor of 5. For more call-intensive versions of the same programs, performance improved by a median of 40 % and the number of virtual calls dropped by a factor of 21. Our measurements indicate that inlining does not necessarily lead to large increases in code size, and that for most programs, the instruction cache miss ratio does not increase significantly.

Run-time code generation is an alternative and complement to compile-time program analysis and optimization. Static analyses are inherently imprecise because most interesting aspects of run-time behavior are uncomputable. By deferring aspects of compilation to run time, more precise information about program behavior can be exploited, leading to greater opportunities for code improvement. The cost of performing optimization at run time is of paramount importance, since it must be repaid by improved performance in order to obtain an overall speedup. This paper describes a lightweight approach to run-time code generation, called deferred compilation, in which compile-time specialization is employed to reduce the cost of optimizing and generating code at run time. Implementation strategies developed for a prototype compiler are discussed, and the results of preliminary experiments demonstrating significant overall speedup are presented. 1 Introduction Many compiler optimizations depend ...

Cecil is a purely object-oriented language intended to support rapid construction of high-quality, extensible software. Cecil combines multi-methods with a simple classless object model, a kind of dynamic inheritance, modules, and optional static type checking. Instance variables in Cecil are accessed solely through messages, allowing instance variables to be replaced or overridden by methods and vice versa. Cecil’s predicate objects mechanism allows an object to be classified automatically based on its run-time (mutable) state. Cecil’s static type system distinguishes between subtyping and code inheritance, but Cecil enables these two graphs to be described with a single set of declarations, streamlining the common case where the two graphs are parallel. Cecil includes a fairly flexible form of parameterization, including explicitly parameterized objects, types, and methods, as well as implicitly parameterized methods related to the polymorphic functions commonly found in functional languages. By making type declarations optional, Cecil aims to allow mixing of and migration between exploratory and production programming styles. Cecil supports a module mechanism that enables independently-developed subsystems to be encapsulated, allowing them to be type-checked and reasoned about in isolation despite the presence of multi-methods and subclassing. Objects can be extended externally with additional

We describe Vortex, an optimizing compiler intended to produce high-quality code for programs written in a heavily-object-oriented style. To achieve this end, Vortex includes a number of intra- and interprocedural static analyses that can exploit knowledge about the whole program being compiled, including intraprocedural class analysis, class hierarchy analysis, and exhaustive class testing, and profile-guided optimizations such as receiver class prediction and selective specialization. To make whole-program optimization practical, Vortex automatically tracks cross-file optimization dependencies at a fine granularity, triggering selective recompilation of affected compiled files whenever the source program changes. Empirical measurements of five purely object-oriented benchmark programs written in Cecil, ranging in size from several hundred to 75,000 lines of source code, indicate that these optimization techniques improve performance of large programs by more than a factor of three over a system with only intraprocedural static optimizations. Vortex is written in Cecil, and it has been used as its own compiler and optimizer during its development for the past two years. Vortex’s optimizations and implementation techniques should be useful for any language or program where optimizations to reduce the cost of polymorphism are important, including object-oriented languages (we are currently adding front-ends for C++, Modula-3, and Java to Vortex to study its effectiveness on these other language styles) and other highlevel symbolic, functional, and logic languages.

The high performance implementation of Java Virtual Machines (JVM) and Just-In-Time (JIT) compilers is directed toward adaptive compilation optimizations on the basis of online runtime profile in-formation. This paper describes the design and implementation of a dynamic optimization framework in a production-level Java JIT compiler. Our approach is to employ a mixed mode interpreter and a three level optimizing compiler, supporting quick, full, and spe-cial optimization, each of which has a different set of tradeoffs be-tween compilation overhead and execution speed. A lightweight sampling profiler operates continuously during the entire program's execution. When necessary, detailed information on runtime behav-ior is collected by dynamically generating instrumentation code which can be installed to and uninstalled from the specified recom-pilation target code. Value profiling with this instrumentation mechanism allows fully automatic code specialization to be per-formed on the basis of specific parameter values or global data at the highest optimization level. The experimental results show that our approach offers high performance and a low code expansion ra-tio in both program startup and steady state measurements in com-parison to the compile-only approach, and that the code specializa-tion can also contribute modest pertbrmance improvements. 1.

This paper presents a system that provides code generation at load-time and continuous program optimization at run-time. First, the architecture of the system is presented. Then, two optimization techniques are discussed that were developed specifically in the context of continuous optimization. The first of these optimizations continually adjusts the storage layouts of dynamic data structures to maximize data cache locality, while the second performs profile-driven instruction re-scheduling to increase instruction-level parallelism. These two optimizations have very di#erent cost/benefit ratios, presented in a series of benchmarks. The paper concludes with an outlook to future research directions and an enumeration of some remaining research problems. The empirical results presented in this paper make a case in favor of continuous optimization, but indicate that it needs to be applied judiciously. In many situations, the costs of dynamic optimizations outweigh their benefit, so that no break-even point is ever reached. In favorable circumstances, on the other hand, speed-ups of over 120% have been observed. It appears as if the main beneficiaries of continuous optimization are shared libraries, which at di#erent times can be optimized in the context of the currently dominant client application.

Selective dynamic compilation systems, typically driven by annotations that identify run-time constants, can achieve significant program speedups. However, manually inserting annotations is a tedious and time-consuming process that requires careful inspection of a program's static characteristics and run-time behavior and much trial and error in order to select the most beneficial annotations. Calpa is a system that generates annotations automatically for the DyC dynamic compiler. Calpa combines execution frequency and value profile information with a model of dynamic compilation cost and dynamically generated code benefit to choose run-time constants and other dynamic compilation strategies. For the programs tested so far, Calpa generates annotations of the same or better quality as those found by a human, but in a fraction of the time. The result was equal or better program speedups from dynamic compilation, but without the need for programmer intervention. 1. Introduction Dynamic ...