[all-commits] [llvm/llvm-project] 0ee891: [ARM] Fix VMOVSR pattern requirements

[Mikhail Maltsev via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 0ee891fe2e5704ab790b48af49a3e8ca4935cce7
      https://github.com/llvm/llvm-project/commit/0ee891fe2e5704ab790b48af49a3e8ca4935cce7
  Author: Mikhail Maltsev <mikhail.maltsev at arm.com>
  Date:   2023-04-04 (Tue, 04 Apr 2023)

  Changed paths:
    M llvm/lib/Target/ARM/ARMInstrVFP.td
    A llvm/test/CodeGen/Thumb2/float-ops-vmovsr-nofp-xo.ll

  Log Message:
  -----------
  [ARM] Fix VMOVSR pattern requirements

The required target features for VMOVSR instruction (integer -> MVE
FP register copy) are HasFPRegs and UseVMOVSR (see
https://reviews.llvm.org/D60694).

The instruction selection pattern must use the same set of
requirements (i.e., HasFPRegs instead of HasVFP2). The discrepancy
can cause instruction selection failure when compiling code for
execute-only position-independence mode.

Reviewed By: dmgreen

Differential Revision: https://reviews.llvm.org/D147523