Optimised Node Splitting in Irreducible Flow Graphs

From	bilal.a.tariq@gmail.com
Newsgroups	comp.compilers
Subject	Optimised Node Splitting in Irreducible Flow Graphs
Date	2019-01-21 22:09 -0500
Organization	Compilers Central
Message-ID	<19-01-007@comp.compilers> (permalink)

Show all headers | View raw

Hello

I am working on a hobbyist transpiler/optimiser for Java bytecode and am
interested in implementing the controlled node splitting algorithm specified
in "Making graphs reducible with controlled node splitting" by Johan Janssen
and Henk Corporaal. More specifically I am unsure what the best way to
identify the "loop-sets" are, defined as paths (n_1,...,n_k) where n_1 is an
immediate successor of n_k. The loopsets formed here seem similar to strongly
connected components to me but they are not the same. Perhaps there is a
reference implementation I could look at or is there something glaringly
obvious I am missing?

B. Tariq

Back to comp.compilers | Previous | Next | Find similar

Thread

Optimised Node Splitting in Irreducible Flow Graphs bilal.a.tariq@gmail.com - 2019-01-21 22:09 -0500

csiph-web