Indirect branch prediction is important to boost instruction-level parallelism in modern processors. Previous indirect branch predictions usually cannot achieve high performance for the ineffectiveness of correlated information. This paper proposes the Compilerassisted value correlation (CVC), a hardware/software cooperative indirect branch prediction scheme. The key is to identify effective va...

In this paper, we investigate the combined use of static techniques and dynamic feedback information to achieve a high level of optimization of both compiler eeciency and code performance. In previous work, we have introduced a new compiler approach, iterative compilation, to select the best tile sizes and unrolling factors. In this approach, many versions of programs are generated and their wo...

Instruction schedulers for superscalar and VLIW processors must expose suucient instruction-level parallelism to the hardware in order to achieve high performance. Traditional compiler instruction scheduling techniques typically take into account the constraints imposed by all execution scenarios in the program. However, there are additional opportunities to increase instruction-level paralleli...

A xed-point C compiler is developed for convenient and efcient programming of TMS320C50 xed-point digital signal processor. This compiler supports the `fix' data type that can have an individual integer word-length according to the range of a variable. It can add or subtract two data having di erent integer word-lengths by automatically inserting shift operations. The accuracy of xed-point mult...

We investigate the problem of code generation for DSP systems on a chip. Such systems devote a limited quantity of silicon to program ROM, so application software must be maximally dense. Additionally, the software must be written so as to meet various high-performance constraints, which may include hard real-time constraints. Unfortunately, current compiler technology is unable to generate den...

Automatic type inference is a popular feature of functional programming languages. Unfortunately, if the compiler cannot infer a correct typing of a program, the produced error message does not always help the programmer to fix the error. The compiler typically reports a single program location in the error message. This location is the point where the type inference failed, but not necessarily...

Our research is concerned with compiler-independent, eecient and convenient garbage collection for C++. Most collectors proposed for C++ have either been implemented in a library or in a compiler. As an intermediate step between those two, this paper proposes using precompilation techniques to augment a C++ source program with code to allow mostly type-accurate garbage collection. There are two...

To eeciently parallelize a scientiic application with a data-parallel compiler requires certain structural properties in the source program, and conversely, the absence of others. A recent parallelization eeort of ours reinforced this observation and motivated this correspondence. Speciically, we have transformed a Fortran 77 version of Gromos, a popular dusty-deck program for molecular dynamic...

While program correctness for compiled languages depends fundamentally on compiler correctness, compiler optimizations are not usually formally verified due to the effort involved, particularly in the presence of concurrency. In this paper, we present a framework for stating and reasoning about compiler optimizations and transformations on programs in the presence of relaxed memory models. The ...

Higher-order functions pose a challenge for both static program analyses and optimizing compilers. To simplify the analysis compilation of languages with higher-order functions, rich body prior work has proposed variety defunctionalization techniques, which can eliminate from by transforming to semantically-equivalent first-order representation. Several modern take this step further, specializi...