[Soot-list] Jedd Question

Ondrej Lhotak olhotak at uwaterloo.ca
Wed Dec 13 10:34:09 EST 2006


On Tue, Dec 12, 2006 at 05:59:15PM -0500, Saswat Anand wrote:
> Hi,
> 
> I am getting "jeddc: SAT solver couldn't assign physical domains." error 
> while compiling my Jedd program. I was wondering if somebody could point 
> out whats wrong.

Does it print an error message with a line number afterwards? If it does
not, the problem is that Jedd cannot execute zchaff and/or
zcore/zverify_df (i.e. they are not installed, or not compiled for the
machine you're using, or not on your path, etc.) The scripts that I use
to execute them can be found at:
http://plg.uwaterloo.ca/~olhotak/build/zcore
http://plg.uwaterloo.ca/~olhotak/build/zchaff

Provide the full path to these scripts in the -s and -sc options to
Jedd.

> I have a relation "paths" which of type  <srcc:C1, from:PB1, tgtc:C2, 
> to:PB2>, where C1, C2 are from paddle and PB1, PB2 are new two physical 
> domains. "from" and "to" are probe statements that are inserted to 
> measure coverage. "paths" is supposed to contain all pairs of probes 
> between which there is a control-flow path.
> 
> For that, I have this code that computes fix point:
> 
> <srcc:C1, from:PB1, tgtc:C2, to:PB2> new_paths = 0B;
> do{
>    paths |= new_paths;
>    new_paths = paths{tgtc,to} <> paths{srcc, from};
>    new_paths -= paths;
> }while(new_paths != 0B);
> 
> After some hit-trials, I found that if I comment out the statement with 
> join operation in the above, the code compiles. But I cannot understand 
> whats wrong with that statement. Or, is the problem somewhere else? Just 
> to be sure, I specified physical domains for every attribute in every 
> relation, but jeddc still fails. Please help!
> 
> Thanks,
> Saswat
> 
> 
> 
> _______________________________________________
> Soot-list mailing list
> Soot-list at sable.mcgill.ca
> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
> 


More information about the Soot-list mailing list