[PATCH] D44618: [X86] Make the multiply and divide itineraries more consistent.

[Craig Topper via Phabricator via llvm-commits]

craig.topper created this revision.
craig.topper added a reviewer: RKSimon.

Sometimes we used the same itinerary for MEM and REG forms, but that seems inconsistent with our usual usage.

We also used the MUL8 itinerary for MULX32/64 which was also weird.

The test changes are because we were using IIC_IMUL32_RR and IIC_IMUL64_RR instead of IIC_IMUL32_REG/IIC_IMUL64_REG for the 32 and 64 bit multiplies that produce double width result.


Repository:
  rL LLVM

https://reviews.llvm.org/D44618

Files:
  lib/Target/X86/X86InstrArithmetic.td
  lib/Target/X86/X86Schedule.td
  lib/Target/X86/X86ScheduleAtom.td
  test/CodeGen/X86/schedule-x86_64.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D44618.138877.patch
Type: text/x-patch
Size: 16301 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180319/3b38cf79/attachment.bin>