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
{{ message }}
This repository has been archived by the owner on May 20, 2018. It is now read-only.
With #158, we will be able to compile well-typed programs to well-typed LLVM. However, the general compilation strategy may not be very efficient, e.g. for datatypes, which are encoded as functions.
Any ad hoc compilation strategy which preserves the semantics is well-formed, and some of them may be vastly more efficient.
Therefore, we should be able to select a different compilation strategy for some (sub-)programs, e.g.:
Boolean → 1-bit integer.
Maybe → pointer w/ 0 representing the nothing case.
Maybe Boolean → 2-bit integer.
List Boolean → bitfield.
The text was updated successfully, but these errors were encountered:
With #158, we will be able to compile well-typed programs to well-typed LLVM. However, the general compilation strategy may not be very efficient, e.g. for datatypes, which are encoded as functions.
Any ad hoc compilation strategy which preserves the semantics is well-formed, and some of them may be vastly more efficient.
Therefore, we should be able to select a different compilation strategy for some (sub-)programs, e.g.:
Boolean
→ 1-bit integer.Maybe
→ pointer w/0
representing thenothing
case.Maybe Boolean
→ 2-bit integer.List Boolean
→ bitfield.The text was updated successfully, but these errors were encountered: