Closed Heizmann closed 9 years ago
I just did. (Sorry for the dead link) Furthermore, I updated the arguments of my Z3 call. (Problem can be reproduced without smt.mbqi=false)
It seems that the problem only occurs if I use the argument SMTLIB2_COMPLIANT\=true
.
got it
yes, it is the issue: the interpolants created mixed integer/real constraints. This is not expressible in SMT-LIB2's AUFLIRA theory. Is the compliant configuration required for your tool?
fixed
Is the compliant configuration required for your tool? Once, we required it. I cannot remember the reason any more. In the last hours I run a few benchmarks without the compliant configuration and I have not noticed any problems.
Right, but with the new fix you can use both compliant and non-compliant mode. Interpolation overwrites the portion that disables coercions in compliant mode.
Executing Z3 ef7915858bca5e759ee0a34df2ae976e7e507bef with the command
on this SMT script, I obtain the following error message.
However, there is no Real in the script at all.