Hello Ravi,
Thank you for your help. I have tried this yesterday by setting
chord.ssa=false, but it didn't work for we, then by reading the source
code, I figured out that I should set "chord.ssa.kind=nophi", but this
is fine.
The problem for me is that it seems to me that no matter I turn on/off
this switch, SSA is always done.
I have a very simple example, suggested by wiki page of SSA
(
http://en.wikipedia.org/wiki/Static_single_assignment_form)
=================================
package test;
public class HelloWorld {
public static void main(String[] args) {
int y = 1;
y = 2;
int x = y;
}
}
==================================
and I asked chord to print instruction by executing ant
-Dchord.work.dir=./examples/hello_world
-Dchord.print.classes=test.HelloWorld run
The following is two runs' the output of log.txt whether I set
"chord.ssa.kind=nophi" or "chord.ssa.kind=phi". They are the same and
it seems that SSA is always done.
==================================
Control flow graph for main:([Ljava/lang/String;)V...@test.HelloWorld:
BB0 (ENTRY) (in: <none>, out: BB2)
BB2 (in: BB0 (ENTRY), out: BB1 (EXIT))
1: MOVE_I R3, IConst: 1
2: MOVE_I R4, IConst: 2
3: MOVE_I R2, IConst: 2
4: RETURN_V
BB1 (EXIT) (in: BB2, out: <none>)
==================================
Dig a little deeper, I found that Chord calls
"jq_Method.doSSA(true/false);" (program/Program.java) to set SSA flag,
then I was trying to see what did jq_Method does.
The online resource that I can find about API is a little bit
outdated. (
http://joeq.sourceforge.net/apidocs/joeq/Class/jq_Method.html).
No doSSA() function can be found there. Joeq is made as a jar file in
chord release. Followed your online guide
(
http://pag-www.gtisc.gatech.edu/chord/user_guide/start.html#id1) to
download chord-libsrc-2.0.tar.gz, but since it is 2.0 version,
jq_Method.doSSA() function also has a different interface.
public static void doSSA() { doSSA = true; }
So, I am just try to know whether it is the case that SSA is always
going to be done. It seems it is for my tests or I am just miserably
misunderstood.
Thanks,
Lei
> You received this message because you are subscribed to a topic in the
> Google Groups "chord-discuss" group.
> To unsubscribe from this topic, visit
>
https://groups.google.com/d/topic/chord-discuss/aq8d0kNLEYg/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to