[PATCH] D112406: [Driver][AArch64]Add driver support for neoverse-512tvb target

Mon Oct 25 01:18:06 PDT 2021

dmgreen added a comment.

This is a bit of a shame. I was hoping we wouldn't need the same hacks as GCC. The llvm cost modelling can work quite differently at times to GCC and I didn't think we were close enough to optimal code to need to worry about these kinds of differences. I guess having the option is useful for consistency.

================
Comment at: llvm/lib/Support/Host.cpp:216
         .Case("0xd49", "neoverse-n2")
+        .Case("0xd4a", "neoverse-512tvb")
         .Default("generic");
----------------
This doesn't sound right - for a fake cpu to work with -mcpu=native.

================
Comment at: llvm/lib/Target/AArch64/AArch64Subtarget.cpp:158
+  case Neoverse512TVB:
+    MaxInterleaveFactor = 4;
+    break;
----------------
Should this have Loop Alignment too?

Is the interleave factor higher due to the 512bit vector bandwidth?

================
Comment at: llvm/lib/Target/ARM/ARM.td:1412

+def : ProcNoItin<"neoverse-512tvb",                     [ARMv84a,
+                                                         FeatureHWDivThumb,
----------------
Are we sure gcc has a -mcpu=neoverse-512tvb option for Arm? Or is it AArch64 only?

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D112406/new/

https://reviews.llvm.org/D112406