基于存储资源迭代重用的低成本寄存器重命名方法

针对超标量深流水线中物理寄存器资源冲突造成的流水线阻塞问题,提出了一种多指令共享同一物理寄存器资源的非阻塞指令发射方法。该方法可在物理寄存器资源冲突下继续分配物理寄存器,利用发射缓冲队列临时缓冲冲突的指令,增加发射流水级实际可分配的物理寄存器数量,释放发射窗口,提高物理寄存器使用的并行性。实验结果表明:相对于传统重命名方法,该方法可减少27.3%的物理寄存器资源实现传统方法相同的性能。     

Denotational Semantics of a Simple Model of Eiffel

A simple abstract model of Eiffel is introduced,and its denotational semantics is defined using VDM style.A static analysis approach is presented to treat multiple inheritance and renaming mechanism.Within the framework of denotational semantics introduced in this paper,the key features of Eiffel,such as identification,classification,multiple inheritance,polymorphism and dynamic binding,can be adequately characterized.     

两级分配多可用重命名寄存器

提出了一种新的寄存器重命名机制——两级分配多可用重命名寄存器,简称2L-MuRR,其特点如下：（1）对重命名寄存器文件（RRF）的占用要经过两级重命名和分配过程,减少了无效占用时间.（2）RRF被划分为多个不同宽度的字段,根据目标值的大小进行按字段分配,这样使每个重命名寄存器都是多可用的.（3）同一重命名寄存器内的多个目标值的写（或读）操作可以合并成一次完成,减少了RRF的访问次数和读/写端口的压力.模拟实验表明,2L-MuRR显著提高了RRF的存储能力（70%以上）,能以较少的重命名寄存器满足SMT高并行度的要求,有效缓解了寄存器文件的设计压力.     

基于PVM的协调检查点设置关键技术

本文论述了基于PVM的并行程序运行回卷恢复系统设计和实现过程中的退出再加入PVM机制、任务号隐式映射机制、任务结束前同步机制、防止PVM库重入机制，信号与消息协同触发机制、应用任务初始化机制以及作为前述各机制实现基础的函数包裹与换名机制等关键技术。这些技术已经成功地应用于我们自主开发的ChaRM系统中，证明了技术的正确性和有效性。     

Java标识符重命名混淆算法及其实现

针对控制流混淆和数据混淆会引入额外开销的问题,提出4种标识符重命名算法,包括滥用标识符算法、重载无关联方法算法、非法标识符代替合法标识符算法以及重写静态方法算法,以这4种算法为核心技术构造Java混淆器JIRO,对JIRO的性能进行初步分析,结果表明,JIRO具有较好的混淆效果,不会引入额外的字节码执行开销。     

Path Analysis and Renaming for Predicated Instruction Scheduling

Increases in instruction level parallelism are needed to exploit the potential parallelism available in future wide issue architectures. Predicated execution is an architectural mechanism that increases instruction level parallelism by removing branches and allowing simultaneous execution of multiple paths of control, only committing instructions from the correct path. In order for the compiler to expose and use such parallelism, traditional compiler data-flow and path analysis needs to be extended to predicated code. In this paper, we motivate the need for renaming and for predicates that reflect path information. We present Predicated Static Single Assignment (PSSA) which uses renaming and introduces Full -Path Predicates to remove false dependences and enable aggressive predicated optimization and instruction scheduling. We demonstrate the usefulness of PSSA for Predicated Speculation and Control Height Reduction. These two predicated code optimizations used during instruction scheduling reduce the dependence length of the critical paths through a predicated region. Our results show that using PSSA to enable speculation and control height reduction reduces execution time from 12 to 68%.     

MAX(1)和MARG(1)中公式改名的复杂性

改名是一个将变元映射到变元本身或它的补的函数,变元改名是公式变元集合上的一个置换,文字改名是一个改名和一个变元改名的组合.研究CNF公式的改名有助于改进DPLL算法.考虑判定问题"对于给定的CNF公式H和F是否存在一个变元(或文字)改名ψ使得ψ(H)=F?"的计算复杂性.MAX(1)和MARG(1)是极小不可满足公式的两个子类,这两个子类中的公式可以用树表示.树同构的判定问题在线性时间内是可解的.证明了对于MAX(1)和MARG(1)中的公式,文字改名问题在线性时间内可解,变元改名问题在平方次时间内可解.     

A generalization of the E-LOTOS renaming operator: a convenience for specification of new forms of process composition

E-LOTOS is one of the standard languages for formal specification of real-time concurrent and reactive systems. As it is process-algebraic, its expressive power lies in its process-composition operators. Of course, not all forms of composition can be directly expressed in the language. In the most difficult cases, one typically resorts to the constraint-oriented specification style. We demonstrate that a slight enhancement of the E-LOTOS renaming operator would make the specification style even more powerful. As an example, we show how to specify choice, parallel composition and disabling enhanced with action priorities.     

Renaming a set of non-Horn clauses

Several extensions of the logic programming language Prolog to nonHorn clauses use case analysis to handle non-Horn clauses.In this paper,analytical and emprirical evidences are presented to show that,by making a set of clauses less “non-Horn“ using predicated renaming.the performance of these case-analysis based procedures can be improved significantly.In addition,the paper also investigated the problem of efficeiently constructing a predicate renaming that reduces the degree of “non-Hornness“ of a clause set maximally.It is shown that this problem of finding a redicate renaming to achieve minimal “non-Hornness“ is NP -complete.     

超标量处理器中寄存器别名表恢复方法的改进研究

寄存器重命名是超标量处理器用于提升指令集并行度的重要方法,其基本实现方式是通过寄存器别名表来记录逻辑寄存器和物理寄存器的映射关系,当发生分支预测错误时需要对寄存器别名表中的内容进行恢复。针对于现有的恢复方法没有利用重命名的局部性特征,在处理器的指令窗口增加时暴露出实现代价过大的问题,提出了一种基于区间进行寄存器别名表恢复的改进型恢复方法,通过对walk方法的改造,使用区间计分板来确定需要扫描的地址范围,并精确地控制每个区间的扫描,可以大大减小资源浪费。通过对逻辑综合结果和性能进行分析,相比于检查点恢复的传统设计,这种方法使用更少的面积和功耗,达到与检查点方式恢复接近的性能,也具有更好的扩展潜力。