[llvm] [TableGen] Rework error reporting for duplicate Feature/Processor. (PR #102257)

Sergei Barannikov via llvm-commits llvm-commits at lists.llvm.org
Wed Aug 7 11:25:12 PDT 2024


s-barannikov wrote:

> > What if you use SMLoc from a record with smaller ID (that is use getID() as a tie breaker)?
> 
> I see. So ID seem to be assigned for each Record in the construction order, which I assume would also be the order in which the records were parsed/defined in the TD file, so we can use ID as a tie-breaker between First/Second in the code to determine which one was first seen by TableGen parser. Is that the idea?

Yes, and hopefully it will report the "correct" location for the multiclass example, too.


https://github.com/llvm/llvm-project/pull/102257


More information about the llvm-commits mailing list