Files
clang-p2996/llvm/test/CodeGen/MIR/X86/load-with-max-alignment.mir
Simon Pilgrim d391e4fe84 [X86] Update RET/LRET instruction to use the same naming convention as IRET (PR36876). NFC
Be more consistent in the naming convention for the various RET instructions to specify in terms of bitwidth.

Helps prevent future scheduler model mismatches like those that were only addressed in D44687.

Differential Revision: https://reviews.llvm.org/D113302
2021-11-07 15:06:54 +00:00

45 lines
1.1 KiB
YAML

# RUN: llc -march=x86-64 -run-pass=none -o - %s | FileCheck %s
# CHECK: name: truly_aligned_load
# CHECK: MOV32rm {{.*}} :: (load (s32), align 4294967296)
# CHECK: name: basealigned_load
# CHECK: MOV32rm {{.*}} :: (load (s32) from unknown-address + 4, basealign 4294967296)
---
name: truly_aligned_load
alignment: 16
tracksRegLiveness: true
liveins:
- { reg: '$rdi' }
frameInfo:
maxAlignment: 1
maxCallFrameSize: 0
machineFunctionInfo: {}
body: |
bb.0:
liveins: $rdi
renamable $eax = MOV32rm killed renamable $rdi, 1, $noreg, 0, $noreg :: (load (s32) from unknown-address, align 4294967296)
RET64 $eax
...
---
name: basealigned_load
alignment: 16
tracksRegLiveness: true
liveins:
- { reg: '$rdi' }
frameInfo:
maxAlignment: 1
maxCallFrameSize: 0
machineFunctionInfo: {}
body: |
bb.0:
liveins: $rdi
renamable $eax = MOV32rm killed renamable $rdi, 1, $noreg, 0, $noreg :: (load (s32) from unknown-address + 4, basealign 4294967296)
RET64 $eax
...