Gergely =?utf-8?q?Bálint?= <[email protected]>,
Gergely =?utf-8?q?Bálint?= <[email protected]>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/[email protected]>


paschalis-mpeis wrote:

> One piece of doubt that I have since I opened this PR: the term CFI is 
> overloaded, it can mean Call Frame Information and Control Flow Integrity. 
> This is even more problematic than usually in this case, where we talk about 
> the Call Frame Information related to a Control Flow Integrity mechanism, 
> namely Pointer Authentication.
> 
> Another way to name these would be PointerAuthDwarf* (or capital DWARF*), to 
> signal that these passes take care of the DWARF CFIs in the binary.

That is a good point. Generalizing to DWARF may help, but I don't have strong 
opinions.
If the name feels too long, can consider `PAuth<..>`.

It may be worth briefly describing both terms in the doc (as a hint / callout 
near the top) and expanding the CFI acronyms in both `*Analyzer` and `*Fixup` 
headers (ie [here](https://github.com/llvm/llvm-project/pull/164622/files#L10)).

https://github.com/llvm/llvm-project/pull/164622
_______________________________________________
llvm-branch-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits

Reply via email to