home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.compilers
- Path: sparky!uunet!world!iecc!compilers-sender
- From: bwilson@shasta.stanford.edu (Bob Wilson)
- Subject: Re: control dependence question
- Reply-To: bwilson@shasta.stanford.edu (Bob Wilson)
- Organization: Compilers Central
- Date: Wed, 16 Dec 1992 21:09:45 GMT
- Approved: compilers@iecc.cambridge.ma.us
- Message-ID: <92-12-080@comp.compilers>
- Keywords: optimize, design
- References: <92-12-072@comp.compilers> <92-12-065@comp.compilers>
- Sender: compilers-sender@iecc.cambridge.ma.us
- Lines: 27
-
- The following message which I posted was based on what I remembered from
- reading the PTRAN report some time ago.
-
- Carl McConnell writes:
- > I've come up with a simple algorithm for computing control dependences, ...
-
- I replied:
- > This sounds like the algorithm that was used in the PTRAN compiler.
- > I think it's described in a research report by Vivek Sarkar:
- >
- > "PTRAN - the IBM Parallel Translation System"
- > IBM Research Division Report RC 16194
- > 1990
-
- I guess my memory of their control dependence analysis algorithm was a
- little too fuzzy. That algorithm is the same as the one described in the
- paper "Compact Representations for Control Dependence" by Cytron et al.
- in PLDI '90. Instead of comparing the sets of postdominators as
- McConnell's algorithm does, they build the postdominator tree and
- determine the control dependences based on the immediate postdominators.
- Sorry for the confusion.
-
- Bob Wilson
- bwilson@shasta.stanford.edu
- --
- Send compilers articles to compilers@iecc.cambridge.ma.us or
- {ima | spdcc | world}!iecc!compilers. Meta-mail to compilers-request.
-