PikachuHyA wrote:

> Might be a good moment to check whether this is the best approach (cc 
> @andykaylor @xlauko). Others raised the question if we really need to map 
> them into CIR this early or whether it makes sense to only generate them 
> during lowering time. TBH I don't remember the full discussion we had couple 
> years back in the incubator (cc @Lancern @PikachuHyA), but this has been 
> discussed before. Might be related to the fact that we need to keep some 
> information around to distinguish some of the types, and instead of changing 
> the types themselves it made sense at the time to have TBAA populated as side 
> information, "only pay for what you use" kinda thing.
> 
> My personal take is that this is fine as an initial approach, once the 
> lowering part gets upstreamed we'll have good test coverage to simplify 
> CIRGen if needed.

@bcardosolopes  I’m afraid I don’t really remember the details, and I don’t 
have time to look them up right now.

could you clarify what the CIR merge policy is, especially regarding 
authorship/attribution? I’m a bit concerned whether names from previous work 
might get dropped in the process.

https://github.com/llvm/llvm-project/pull/169226
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to