News for hackers. Stuff that matters.
Re:The Cerberus C semantics [pdf] by fanf2 on Saturday May 13, 2023 @09:17AM (#35928076)
These pointer provanance semantics are implemented by LLVM too. They also apply to C++, Rust, Zig, etc.
Re:The Cerberus C semantics [pdf] by tialaramex on Saturday May 13, 2023 @12:45PM (#35930119)
What are "these" semantics? The PDF linked describes all of both PVI and PNVI, so saying LLVM implements "these" semantics in that context would be like if you said "I dated her in college" when somebody is describing the entire US population. Maybe you did date an American in college, but I think you need to be more specific.
Re:The Cerberus C semantics [pdf] by carbotaniuman on Saturday May 13, 2023 @04:17PM (#35932288)
I think the parent was trying to say LLVM implements pointer provenance semantics in some form, as "these" semantics doesn't really make sense. As to what form LLVM implements... a weird mishmash of many semantics, with some optimizations that are contradictory with each other and some that don't make sense.

PNVI-ae-udi looks to be the one that compiler developers and the committee agree on, but we'll see if these efforts gain any more traction.