[all-commits] [llvm/llvm-project] 431a5d: [x86] Implement a tagged-globals backend feature.

[Matt Morehouse via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 431a5d8411952b4c9c2185303afd703123ef55cc
      https://github.com/llvm/llvm-project/commit/431a5d8411952b4c9c2185303afd703123ef55cc
  Author: Matt Morehouse <mascasa at google.com>
  Date:   2021-10-18 (Mon, 18 Oct 2021)

  Changed paths:
    M llvm/lib/Target/X86/X86.td
    M llvm/lib/Target/X86/X86Subtarget.cpp
    M llvm/lib/Target/X86/X86Subtarget.h
    A llvm/test/CodeGen/X86/tagged-globals-pic.ll
    A llvm/test/CodeGen/X86/tagged-globals-static.ll

  Log Message:
  -----------
  [x86] Implement a tagged-globals backend feature.

The feature tells the backend to allow tags in the upper bits of global
variable addresses.  These tags will be ignored by upcoming CPUs with
the Intel LAM feature but may be used in instrumentation passes (e.g.,
HWASan).

This patch implements the feature by using @GOTPCREL relocations instead
of direct references to the locally defined global.  Thus the full
tagged address can be loaded by a single instruction:
  movq global at GOTPCREL(%rip), %rax

Reviewed By: eugenis

Differential Revision: https://reviews.llvm.org/D111343