C2po develop#52
Conversation
Adds previous functionality and subsequently removes rate since rate had undefined behavior at i=0 (i.e., there was no i=-1 to accurately calculate rate with)
C2po equivalence
…tl patterns benchmarks, major updates to eqsat.
Co-authored-by: Chris Johannsen <cgjoh@protonmail.com>
Added Functionality to C2PO
C2po fmcad26
Co-authored-by: Alexis Aurandt <lexaurandt@gmail.com>
|
I find it a little annoying that we now have to specify --spec to input the spec into C2PO. Is there ever an option where we won't be providing a .c2po file as input into C2PO? I fixed some problems with check-sat. Should we add a few tests for this? Also, should there be something printed to the console when enabling SAT checking? If I recall correctly, there was previously a message printed indicating the results. I also fixed some other option errors. I think I fixed them all, but you may want to double check. @cgjohannsen Will we release this as 4.2.0? If so, I can update the version numbers in C2PO/R2U2. |
|
The issue with omitting the '--spec' option is that when c2po runs in interactive or script mode it doesn't take as .c2po file as input. So 'python3 c2po.py -i' launches into interactive mode and shouldn't take any file as input. Same for when proving a script as in 'python3 c2po.py -s script.cmd'. As for sat checking messaging, the check_sat command has a '--print' option that will print the output but it's off by default. I figure the best behavior is to only warn the user when their spec is unsat or unknown. And I say we do a minor version bump like you suggest, so to 4.2.0. Thanks for the help on this! |
Major rewrite of c2po, adding REPL capability. Allows for granular control of the compiler and extra analysis.