Many parallel programs require run-time support to implement the communication caused by indirect data references. In previous work, we have developed the inspector-executor paradigm to handle these cases. This paper extends that work by developing a dataflow framework to aid in placing the executor communications calls. Our dataflow analysis determines when it is safe to combine communications statements, move them into less frequently executed code regions, or avoid them altogether in favor of reusing data which are already buffered locally. 1 Introduction We present a dataflow framework that can be employed to systematically use runtime preprocessing methods for loops in which some array references are made through a level of indirecti...
Data-parallel languages, such as H scIGH P scERFORMANCE F scORTRAN or F scORTRAN D, provide a machin...
In this paper, we describe two new ideas by which HPF compiler can deal with irregular computations ...
We present a comprehensive approach to performing data flow analysis in parallel. We identify three ...
We developed a dataflow framework which provides a basis for rigorously defining strategies to make ...
Current parallelizing compilers cannot identify a significant fraction of parallelizable loops becau...
Current parallelizing compilers cannot identify a significant fraction of parallelizable loops becau...
This work was also published as a Rice University thesis/dissertation: http://hdl.handle.net/1911/16...
Loss of precision due to the conservative nature of compile-time dataflow analysis is a general prob...
Data-parallel languages allow programmers to use the familiar machine-independent programming style ...
This paper presents a new analysis for parallelizing compilers called predicated array data-flow ana...
Dataflow analyses are a critical part of many optimizing compilers as well as bug-finding and progra...
Abstract—Message passing is a very popular style of parallel programming, used in a wide variety of ...
Array dataflow information plays an important role for successful automatic parallelization of Fortr...
This work was also published as a Rice University thesis/dissertation: http://hdl.handle.net/1911/19...
Automatic parallelization of real FORTRAN programs does not live up to users expectations yet, and d...
Data-parallel languages, such as H scIGH P scERFORMANCE F scORTRAN or F scORTRAN D, provide a machin...
In this paper, we describe two new ideas by which HPF compiler can deal with irregular computations ...
We present a comprehensive approach to performing data flow analysis in parallel. We identify three ...
We developed a dataflow framework which provides a basis for rigorously defining strategies to make ...
Current parallelizing compilers cannot identify a significant fraction of parallelizable loops becau...
Current parallelizing compilers cannot identify a significant fraction of parallelizable loops becau...
This work was also published as a Rice University thesis/dissertation: http://hdl.handle.net/1911/16...
Loss of precision due to the conservative nature of compile-time dataflow analysis is a general prob...
Data-parallel languages allow programmers to use the familiar machine-independent programming style ...
This paper presents a new analysis for parallelizing compilers called predicated array data-flow ana...
Dataflow analyses are a critical part of many optimizing compilers as well as bug-finding and progra...
Abstract—Message passing is a very popular style of parallel programming, used in a wide variety of ...
Array dataflow information plays an important role for successful automatic parallelization of Fortr...
This work was also published as a Rice University thesis/dissertation: http://hdl.handle.net/1911/19...
Automatic parallelization of real FORTRAN programs does not live up to users expectations yet, and d...
Data-parallel languages, such as H scIGH P scERFORMANCE F scORTRAN or F scORTRAN D, provide a machin...
In this paper, we describe two new ideas by which HPF compiler can deal with irregular computations ...
We present a comprehensive approach to performing data flow analysis in parallel. We identify three ...