Optimization phases have no effect #1952
Unanswered
Laschoking
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi,
I want to use soot optimization phase jop.cse to compare the results to an own cse analysis that I wrote in doop.
From doop I have found a number of Common Subexpressions, and I would like to check if soot has similar results.
java -cp soot-4.5.0-20230118.164052-1-jar-with-dependencies.jar soot.Main -cp .:../benchmarks/master/JREs/jre1.11/lib/rt.jar -process-dir ../benchmarks/master/dacapo-2006/antlr -d Out/antlr_jop_cse/ -f J -p jop.cse on -p jb off
What confuses me, is that this phase does not seem to do any changes in the code.
When analyzing some benchmark I once run soot with
-p job.cse on
and once with-p job.cse off
. Withdiff Out/SootCse/ Out/SootNoCse
I do not get any results, meaning that the CSE-phase did not make any changes.I also tried to run this option on a custom example, containing a code snippet like
if (b < a) { g = c * b; }else{ f = c * b; }
which should lead to CSE but also here this phase did not have any impact.
So I am wondering if this phase is actually working / whether I put the correct command line options?
I also noticed for the sop.cpf phase no changes to happened, maybe its a related problem...
In general soot works for me (i.e. enabling the jb -phase results in a lot of changes on the outcoming jimple/shimple code).
For better understanding: is there a difference when enabling
-via-shimple
option in comparison to enable shimple output-f S
?I would appreciate any suggestion & help :)
Beta Was this translation helpful? Give feedback.
All reactions