[PATCH] D35450: [analyzer] Support generating and reasoning over more symbolic constraint types

[Dominic Chen via Phabricator via cfe-commits]

ddcc added a comment.

> Is diff 1 the original diff from https://reviews.llvm.org/D28953? It was ok to reopen it, but the new revision is also fine.

No, diff 1 is already different; it contains most of the bugfixes. I couldn't find any way to reopen the previous review, and `arc diff` wouldn't let me update a closed review.

> Regarding 1-bit bools: did you notice https://reviews.llvm.org/D32328 and https://reviews.llvm.org/D35041, do they accidentally help?

I did see those changes, but I think they're a little different. The test failures I ran into were cases where the input is a 1-bit APSInt, and attempting to retrieve the type for that gives a null QualType.


https://reviews.llvm.org/D35450