252c42354e
The issue is uncovered by #47698: for assembly files, -triple= specifies the full target triple while -arch= merely sets the architecture part of the default target triple, leaving a target triple which may not make sense, e.g. riscv64-apple-darwin. Therefore, -arch= is error-prone and not recommended for tests. The issue has been benign as we recognize $unknown-apple-darwin as ELF instead of rejecting it outrightly. Due to the nature of the issue, we don't see the issue in tests using architectures that any of Mach-O/COFF/XCOFF supports.
19 lines
683 B
ArmAsm
19 lines
683 B
ArmAsm
// RUN: llvm-mc -triple=amdgcn -mcpu=gfx906 -filetype=obj %s | llvm-objdump -d --mcpu=gfx906 - | FileCheck %s
|
|
|
|
// Check that the offset is correctly calculated.
|
|
|
|
BB0:
|
|
v_nop_e64
|
|
v_nop_e64
|
|
BB1:
|
|
v_nop_e64
|
|
BB2:
|
|
s_add_u32 vcc_lo, vcc_lo, (BB2-BB1)&4294967295
|
|
// CHECK: s_add_u32 vcc_lo, vcc_lo, 8 // 000000000018: 806AFF6A 00000008
|
|
s_addc_u32 vcc_hi, vcc_hi, (BB2-BB1)>>32
|
|
// CHECK: s_addc_u32 vcc_hi, vcc_hi, 0 // 000000000020: 826BFF6B 00000000
|
|
s_add_u32 vcc_lo, vcc_lo, (BB0-BB1)&4294967295
|
|
// CHECK: s_add_u32 vcc_lo, vcc_lo, -16 // 000000000028: 806AFF6A FFFFFFF0
|
|
s_addc_u32 vcc_hi, vcc_hi, (BB0-BB1)>>32
|
|
// CHECK: s_addc_u32 vcc_hi, vcc_hi, -1 // 000000000030: 826BFF6B FFFFFFFF
|