[PATCH] D148857: [RISCV] NFC tweaks in extension descriptions, sorting

Craig Topper via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri Apr 21 09:05:04 PDT 2023

craig.topper added inline comments.

Comment at: llvm/lib/Support/RISCVISAInfo.cpp:916
-static const char *ImpliedExtsF[] = {"zicsr"};
 static const char *ImpliedExtsD[] = {"f"};
+static const char *ImpliedExtsF[] = {"zicsr"};
asb wrote:
> There is a canonical order for the single letter extensions (and for ordering those vs Z and X extensions). As such, I'd suggest:
> * Single letter extensions in canonical order (i.e. F then D then V).
> * Then a blank line, then the standard Z* extensions in alphabetical order
> * Then a blank line, then the standard X* extensions in alphabetical order
> I don't feel too strongly about this, so if you have a strongly different view I'd be OK with just plain alphabetic order. But let me know what you think.
Alphabetical does make some sense here. These are all listed again in `ImpliedExts` which must me in alabetical order due to binary search.

  rG LLVM Github Monorepo



More information about the llvm-commits mailing list