Irregular data access/data structures/... in parallelizing compiler design

hzmonte@hotmail.com (hzmonte)
5 May 2003 23:38:59 -0400

          From comp.compilers
Related articles
Irregular data access/data structures/... in parallelizing compiler hzmonte@hotmail.com (2003-05-05)
Re: Irregular data access/data structures/... in parallelizing com matt@peakfive.com (Matt Rosing) (2003-05-16)
| List of all articles for this month |
From: hzmonte@hotmail.com (hzmonte)
Newsgroups: comp.compilers
Date: 5 May 2003 23:38:59 -0400
Organization: http://groups.google.com/
Keywords: parallel
Posted-Date: 05 May 2003 23:38:59 EDT

I am a beginner. When I read literature in parallelizing compiler
design, I often come across terms like "irregular data access",
"irregular data structures", "irregular computations", etc. I tried
to find the meaning in several textbooks to no avail. Do they refer
to the same thing? What do they mean?
I read that in cases involving A[B[i]] and A[B[j]] where A and B are
arrays, it is impossible to determine at compile-time whether they
refer to the same memory location and thus data dependence exists.
Does this have anything to do with the above "irregularity"?
Thanks.

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.