zaks.anna added a comment.
Update: Adding support for top frame to the CallEvent is difficult, so let's
just use CallEvent API in the checkers for all frames but the top one and have
custom top frame handling in the Recursion checker itself. The top frame
handling does not need to be complete (
NoQ added a comment.
Chatted about this a bit more. Because `CallEvent` is an API designed mostly
for convenience, and `ProgramState` is a core API that needs to stay as clean
and separate from other entities and easy to understand as possible, it would
be the best to
1. construct a `CallEvent
zaks.anna added a comment.
Artem just pointed out that I have "Smalls" instead of "SVals" in my first
comment.
https://reviews.llvm.org/D27091
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cf
zaks.anna added a comment.
> Hmm, i'm thinking of just the opposite - refactor the CallEvent methods to
> use the respective new ProgramState methods.
>
> This way we'd avoid touching the CallEvent allocator for a simple Environment
> lookup, while still avoiding code duplication.
I see what
NoQ added a comment.
Hmm, i'm thinking of just the opposite - refactor the CallEvent methods to use
the respective new ProgramState methods.
This way we'd avoid touching the CallEvent allocator for a simple Environment
lookup, while still avoiding code duplication.
https://reviews.llvm.org/D2
zaks.anna added a comment.
Hi!
Looks like this this is used by the Infinite recursion checker. Specifically,
the checker not only needs to get Smalls for arguments of the current
CallEvent, but it also looks for arguments of other calls on the stack. The
checker walks the LocationContext and u
k-wisniewski created this revision.
k-wisniewski added reviewers: NoQ, zaks.anna, dcoughlin, a.sidorin.
k-wisniewski added subscribers: cfe-commits, NoQ.
his patch adds getArgsSVal method to ProgramState that allows the user to
obtain SVals of argumetns used in a call that created the given Stack