Keith Cooper and Taylor Simpson. SCC-based value numbering. Technical Report CRPC-TR95636-S, Center for Research on Parallel Computation, Rice University, October 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....be move to the less frequent point, the edge E(B3, B4) than E(B2, B3) Briggs [3] also introduced live range splitting into Chaitin style register alloca tion. The splitting points are determined using static single assignment (SSA) 15] representations of programs and the loop boundaries [14] [13] [4] Since live range splitting is performed prior to the coloring phase, the decisions regarding split ting points and live range selections for splitting are made prematurely. Live ranges may be split unnecessarily, resulting in execution performance degradation by extra shuffle code. Many ....

Keith Cooper and Taylor Simpson. SCC-based value numbering. Techni- cal Report CRPC-TR95636-S, Center for Research on Parallel Computation, Rice University, October 1995.

....are in the same partition in Step i. The concurrent global value numbering (CGVN) algorithm is an optimistic partitioning algo rithm to find congruent variables. The algorithm is an extension of a global value numbering algorithm for sequential programs that was developed by Cooper and Simpson [23, 83]. The global value numbering is essentially the same as Alpern et al. s value graph partitioning [9] We use value numbers to detect congruence of two variables, i.e. if two variables have the same value number, then they are congruent. The value number of a variable is the value number of its ....

Keith Cooper and Taylor Simpson. SCC-based value numbering. Technical Report CRPC- TR95636-S, Center for Research on Parallel Computation, Rice University, October 1995.

....by Morel and Renvoise [MR79] By targeting partially redundant computations in the program, it automatically removes global common subexpressions and moves invariant computations out of loops. It has since become the most important component in many global optimizers [Cho83, CHKW86, SKL88, BC94, CS95b] In [KRS92, KRS94a] Knoop et al. formulated an alternative placement strategy called lazy code motion that improves on Morel and Renvoise s results by avoiding unnecessary code movements, and by removing the bidirectional nature of the original PRE data flow equations. The result of lazy code ....

....the technique can short circuit propagation through them. Johnson showed how to apply his techniques to the data flow systems in Drechsler and Stadel s variation of Knoop et al. s lazy code motion. Researchers at Rice University have done work aimed at improving the effectiveness of PRE [BC94, CS95b] The work involves the application of some SSA based transformation techniques to prepare the program for optimization by PRE. Their techniques enhance the results of PRE. Their implementation of PRE was based on Drechsler and Stadel s variation of Knoop et al. s lazy code motion, and was ....

K. Cooper and T. Simpson. Value-driven code motion. Technical Report CRPC-TR95637-S, Dept. of Computer Science, Rice University., October 1995.

....11,511 1.8M mpeg2dec 114 11,218 10.2M mpeg2enc 210 16,884 205.9M jpeg 380 36,959 681.2M gs 1,142 68,590 48.2M Figure 13: Program characteristics etc) are given in terms of intermediate code instructions. The compiler s optimizer includes the following passes: global value driven code motion [4], operator strength reduction [5] conditional constant propagation [21] copy propagation, local value numbering, global dead code elimination [10] useless control flow removal, and global register allocation [3] For this study we deliberately avoided using optimization passes that would ....

K. Cooper and T. Simpson. Value-driven code motion. Technical Report CRPC-TR95637-S, Center for Research on Parallel Computation, Rice University, October 1995.

....assembly language [1] The remainder of the compiler operates at the intermediatecode level, and all results (code space, dynamic instruction count, etc) are given in terms of intermediatecode instructions. The compiler s optimizer includes the following passes: global value driven code motion [4], operator strength reduction [5] conditional constant propagation [21] copy propagation, local value numbering, global dead code elimination [10] useless control flow removal, and global register allocation [3] For this study we deliberately avoided using optimization passes that would ....

K. Cooper and T. Simpson. Value-driven code motion. Technical Report CRPC-TR95637-S, Center for Research on Parallel Computation, Rice University, October 1995.

No context found.

Keith Cooper and Taylor Simpson. SCC-based value numbering. Technical Report CRPC-TR95636-S, Center for Research on Parallel Computation, Rice University, October 1995.

No context found.

P. Briggs, K. D. Cooper, and L. T. Simpson. Value numbering. Technical Report CRPC-TR95517-S, Center for Research on Parallel Computation, Rice Univeristy, November 1994. To appear in Software -- Practice and Experience.

No context found.

Cooper K, Simpson LT. SCC-based value numbering. Technical Report CRPC-TR95636-S, Rice University, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC