You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
in this post https://www.npopov.com/2024/01/01/This-year-in-LLVM-2023.html
Nikita says: Undefined behavior in LLVM comes in two flavors: One of those is immediate undefined behavior, which basically works the same as in C, and poison values, which get propagated by most operations and only turn into immediate undefined behavior in certain contexts, for example when you try to branch on them.
I take this as a sign that we can turn off undef inputs in Alive2, basically just flipping the sense of the --disable-undef-inputs flag. This will reduce the number of complaints that we get about timeouts and also, nobody wants to see undef-related bugs. I'm happy to do the pull request for this. We could even add a little warning to the effect of "This result is based on the assumption that inputs are not undef"
The text was updated successfully, but these errors were encountered:
in this post
https://www.npopov.com/2024/01/01/This-year-in-LLVM-2023.html
Nikita says:
Undefined behavior in LLVM comes in two flavors: One of those is immediate undefined behavior, which basically works the same as in C, and poison values, which get propagated by most operations and only turn into immediate undefined behavior in certain contexts, for example when you try to branch on them.
I take this as a sign that we can turn off undef inputs in Alive2, basically just flipping the sense of the
--disable-undef-inputs
flag. This will reduce the number of complaints that we get about timeouts and also, nobody wants to see undef-related bugs. I'm happy to do the pull request for this. We could even add a little warning to the effect of "This result is based on the assumption that inputs are not undef"The text was updated successfully, but these errors were encountered: