[all-commits] [llvm/llvm-project] b7b260: [RISCV] Support strict FP conversion operations.

[Craig Topper via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: b7b260e19a7463464da64501a0991ea17e766a4b
      https://github.com/llvm/llvm-project/commit/b7b260e19a7463464da64501a0991ea17e766a4b
  Author: Craig Topper <craig.topper at sifive.com>
  Date:   2021-12-23 (Thu, 23 Dec 2021)

  Changed paths:
    M llvm/lib/Target/RISCV/RISCVISelLowering.cpp
    M llvm/lib/Target/RISCV/RISCVISelLowering.h
    M llvm/lib/Target/RISCV/RISCVInstrInfoD.td
    M llvm/lib/Target/RISCV/RISCVInstrInfoF.td
    M llvm/lib/Target/RISCV/RISCVInstrInfoZfh.td
    A llvm/test/CodeGen/RISCV/double-convert-strict.ll
    A llvm/test/CodeGen/RISCV/float-convert-strict.ll
    A llvm/test/CodeGen/RISCV/half-convert-strict.ll
    A llvm/test/CodeGen/RISCV/rv64d-double-convert-strict.ll
    A llvm/test/CodeGen/RISCV/rv64f-float-convert-strict.ll
    A llvm/test/CodeGen/RISCV/rv64f-half-convert-strict.ll

  Log Message:
  -----------
  [RISCV] Support strict FP conversion operations.

This adds support for strict conversions between fp types and between
integer and fp.

NOTE: RISCV has static rounding mode instructions, but the constrainted
intrinsic metadata is not used to select static rounding modes. Dynamic
rounding mode is always used.

Differential Revision: https://reviews.llvm.org/D115997