[PATCH] D152421: [RISCV] Introduce the concept of DLEN(data path width) into getLMULCost.
Philip Reames via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Mon Jun 12 07:57:02 PDT 2023
reames added inline comments.
================
Comment at: llvm/lib/Target/RISCV/RISCVFeatures.td:798
+def TuneDLenFactor2
+ : SubtargetFeature<"dlen-factor-2", "DLenFactor2", "true",
----------------
Can you add a comment here which defines DLEN?
As an organizing principle, I was a bit surprised by the factor description. I was expecting something more along the lines of VLEN (zvlNb) where we have specific sizes and then computed the factor from VLEN/DLEN if needed. Not objecting, just commenting.
================
Comment at: llvm/lib/Target/RISCV/RISCVFeatures.td:822
+ TuneShortForwardBranchOpt,
+ TuneDLenFactor2]>;
----------------
The default tuning is used for more than the x280, are you sure you want this placement?
================
Comment at: llvm/lib/Target/RISCV/RISCVSubtarget.h:182
+ unsigned getDLenFactor() const {
+ if (DLenFactor2)
----------------
asb wrote:
> I think we need a comment here to explain the purpose of the method.
Needs a comment, and/or a link back to the DLEN definition.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D152421/new/
https://reviews.llvm.org/D152421
More information about the llvm-commits
mailing list