[llvm] [AMDGPU][GlobalISel] Reduce instruction count for store in global addrspace for non-powers-of-2 (PR #206043)
via llvm-commits
llvm-commits at lists.llvm.org
Fri Jun 26 07:19:07 PDT 2026
github-actions[bot] wrote:
<!--PREMERGE ADVISOR COMMENT: Linux-->
# :penguin: Linux x64 Test Results
* 176187 tests passed
* 3547 tests skipped
* 4 tests failed
## Failed Tests
(click on a test name to see its output)
### LLVM
<details>
<summary>LLVM.CodeGen/AMDGPU/GlobalISel/add.vni16.ll</summary>
```
Exit Code: 1
Command Output (stdout):
--
# RUN: at line 2
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel -mtriple=amdgcn-amd-amdpal -mcpu=fiji < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/add.vni16.ll | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefix=GFX8 /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/add.vni16.ll
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel -mtriple=amdgcn-amd-amdpal -mcpu=fiji
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefix=GFX8 /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/add.vni16.ll
# .---command stderr------------
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/add.vni16.ll:24:14: error: GFX8-NEXT: expected string not found in input
# | ; GFX8-NEXT: flat_load_ushort v11, v[2:3]
# | ^
# | <stdin>:19:36: note: scanning from here
# | v_addc_u32_e32 v7, vcc, 0, v3, vcc
# | ^
# | <stdin>:20:2: note: possible intended match here
# | flat_load_ushort v2, v[2:3]
# | ^
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/add.vni16.ll:704:14: error: GFX8-NEXT: expected string not found in input
# | ; GFX8-NEXT: v_add_u32_e32 v6, vcc, 18, v0
# | ^
# | <stdin>:668:98: note: scanning from here
# | v_add_u16_sdwa v10, v6, v10 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
# | ^
# | <stdin>:673:2: note: possible intended match here
# | v_add_u32_e32 v0, vcc, 18, v0
# | ^
# |
# | Input file: <stdin>
# | Check file: /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/add.vni16.ll
# |
# | -dump-input=help explains the following input dump.
# |
# | Input was:
# | <<<<<<
# | .
# | .
# | .
# | 14: flat_load_ushort v9, v[6:7]
# | 15: flat_load_ushort v10, v[0:1]
# | 16: v_add_u32_e32 v0, vcc, 2, v2
# | 17: v_addc_u32_e32 v1, vcc, 0, v3, vcc
# | 18: v_add_u32_e32 v6, vcc, 4, v2
# | 19: v_addc_u32_e32 v7, vcc, 0, v3, vcc
# | next:24'0 X error: no match found
# | 20: flat_load_ushort v2, v[2:3]
# | next:24'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | next:24'1 ? possible intended match
# | 21: flat_load_ushort v3, v[6:7]
# | next:24'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 22: flat_load_ushort v6, v[0:1]
# | next:24'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 23: v_add_u32_e32 v0, vcc, 4, v4
# | next:24'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 24: v_addc_u32_e32 v1, vcc, 0, v5, vcc
# | next:24'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 25: s_waitcnt vmcnt(2)
# | next:24'0 ~~~~~~~~~~~~~~~~~~~~
# | .
# | .
# | .
# | 663: flat_load_ushort v16, v[2:3]
# | 664: v_add_u32_e32 v2, vcc, 16, v0
# | 665: v_addc_u32_e32 v3, vcc, 0, v1, vcc
# | 666: s_waitcnt vmcnt(3)
# | 667: v_add_u16_e32 v17, v6, v10
# | 668: v_add_u16_sdwa v10, v6, v10 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
# | next:704'0 X error: no match found
# | 669: v_add_u32_e32 v6, vcc, 20, v0
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 670: v_add_u16_e32 v18, v7, v11
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 671: v_add_u16_sdwa v11, v7, v11 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 672: v_addc_u32_e32 v7, vcc, 0, v1, vcc
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 673: v_add_u32_e32 v0, vcc, 18, v0
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | next:704'1 ? possible intended match
# | 674: v_addc_u32_e32 v1, vcc, 0, v1, vcc
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 675: flat_load_ushort v22, v[2:3]
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 676: flat_load_ushort v23, v[0:1]
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 677: flat_load_ushort v21, v[6:7]
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 678: v_add_u16_e32 v19, v8, v12
# | next:704'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | .
# | .
# | .
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1
--
```
</details>
<details>
<summary>LLVM.CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir</summary>
```
Exit Code: 1
Command Output (stdout):
--
# RUN: at line 2
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -O0 -run-pass=legalizer /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir -o - | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefix=SI /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -O0 -run-pass=legalizer /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir -o -
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefix=SI /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir
# .---command stderr------------
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir:8446:13: error: SI-NEXT: is not on the line after the previous match
# | ; SI-NEXT: [[DEF:%[0-9]+]]:_(s32) = G_IMPLICIT_DEF
# | ^
# | <stdin>:14835:2: note: 'next' match was here
# | %59:_(s32) = G_IMPLICIT_DEF
# | ^
# | <stdin>:14827:49: note: previous match ended here
# | %15:_(p1) = nuw inbounds G_PTR_ADD %0, %16(s64)
# | ^
# | <stdin>:14828:1: note: non-matching line after previous match is here
# | %72:_(s32) = G_SHL %73, %71(s32)
# | ^
# |
# | Input file: <stdin>
# | Check file: /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-store-global.mir
# |
# | -dump-input=help explains the following input dump.
# |
# | Input was:
# | <<<<<<
# | .
# | .
# | .
# | 14830: %21:_(<2 x s16>) = G_BITCAST %64(s32)
# | 14831: %22:_(s32) = G_BITCAST %21(<2 x s16>)
# | 14832: G_STORE %22(s32), %15(p1) :: (store (s32) into unknown-address + 16, align 16, addrspace 1)
# | 14833: %28:_(s64) = G_CONSTANT i64 4
# | 14834: %26:_(p1) = G_PTR_ADD %15, %28(s64)
# | 14835: %59:_(s32) = G_IMPLICIT_DEF
# | next:8446 !~~~~~~~~~~~~~~~~~~~~~~~~~~ error: match on wrong line
# | 14836: G_STORE %59(s32), %26(p1) :: (store (s16) into unknown-address + 20, align 4, addrspace 1)
# | 14837: ...
# | 14838: ---
# | 14839: name: test_store_global_v12s16_align4
# | 14840: alignment: 1
# | .
# | .
# | .
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1
--
```
</details>
<details>
<summary>LLVM.CodeGen/AMDGPU/GlobalISel/strict_fma.f16.ll</summary>
```
Exit Code: 1
Command Output (stdout):
--
# RUN: at line 2
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/strict_fma.f16.ll | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefixes=GFX8 /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/strict_fma.f16.ll
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefixes=GFX8 /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/strict_fma.f16.ll
# .---command stderr------------
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/strict_fma.f16.ll:259:14: error: GFX8-NEXT: is not on the line after the previous match
# | ; GFX8-NEXT: v_fma_f16 v0, v0, v2, v4
# | ^
# | <stdin>:213:2: note: 'next' match was here
# | v_fma_f16 v0, v0, v2, v4
# | ^
# | <stdin>:210:30: note: previous match ended here
# | v_lshrrev_b32_e32 v8, 16, v0
# | ^
# | <stdin>:211:1: note: non-matching line after previous match is here
# | v_lshrrev_b32_e32 v9, 16, v2
# | ^
# |
# | Input file: <stdin>
# | Check file: /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/strict_fma.f16.ll
# |
# | -dump-input=help explains the following input dump.
# |
# | Input was:
# | <<<<<<
# | .
# | .
# | .
# | 208: ; %bb.0:
# | 209: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
# | 210: v_lshrrev_b32_e32 v8, 16, v0
# | 211: v_lshrrev_b32_e32 v9, 16, v2
# | 212: v_lshrrev_b32_e32 v10, 16, v4
# | 213: v_fma_f16 v0, v0, v2, v4
# | next:259 !~~~~~~~~~~~~~~~~~~~~~~~ error: match on wrong line
# | 214: v_fma_f16 v2, v8, v9, v10
# | 215: v_fma_f16 v3, v1, v3, v5
# | 216: v_lshlrev_b32_e32 v1, 16, v2
# | 217: v_or_b32_e32 v0, v0, v1
# | 218: flat_store_dword v[6:7], v0
# | .
# | .
# | .
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1
--
```
</details>
<details>
<summary>LLVM.CodeGen/AMDGPU/freeze.ll</summary>
```
Exit Code: 1
Command Output (stdout):
--
# RUN: at line 2
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel=0 -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx600 < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefixes=GFX6,GFX6-SDAG /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel=0 -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx600
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefixes=GFX6,GFX6-SDAG /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll
# note: command had no output on stdout or stderr
# RUN: at line 3
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel=1 -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx600 < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefixes=GFX6,GFX6-GISEL /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/llc -global-isel=1 -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx600
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck -check-prefixes=GFX6,GFX6-GISEL /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll
# .---command stderr------------
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll:6811:20: error: GFX6-GISEL-NEXT: expected string not found in input
# | ; GFX6-GISEL-NEXT: v_lshrrev_b32_e32 v4, 16, v0
# | ^
# | <stdin>:1679:20: note: scanning from here
# | s_waitcnt vmcnt(0)
# | ^
# | <stdin>:1691:28: note: possible intended match here
# | .set .Lfreeze_v3i16.private_seg_size, 0
# | ^
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll:7525:20: error: GFX6-GISEL-NEXT: expected string not found in input
# | ; GFX6-GISEL-NEXT: v_lshrrev_b32_e32 v4, 16, v0
# | ^
# | <stdin>:1959:20: note: scanning from here
# | s_waitcnt vmcnt(0)
# | ^
# | <stdin>:1971:28: note: possible intended match here
# | .set .Lfreeze_v3f16.private_seg_size, 0
# | ^
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll:8250:20: error: GFX6-GISEL-NEXT: expected string not found in input
# | ; GFX6-GISEL-NEXT: v_lshrrev_b32_e32 v4, 16, v0
# | ^
# | <stdin>:2239:20: note: scanning from here
# | s_waitcnt vmcnt(0)
# | ^
# | <stdin>:2251:29: note: possible intended match here
# | .set .Lfreeze_v3bf16.private_seg_size, 0
# | ^
# |
# | Input file: <stdin>
# | Check file: /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/CodeGen/AMDGPU/freeze.ll
# |
# | -dump-input=help explains the following input dump.
# |
# | Input was:
# | <<<<<<
# | .
# | .
# | .
# | 1674: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
# | 1675: s_mov_b32 s6, 0
# | 1676: s_mov_b32 s7, 0xf000
# | 1677: s_mov_b64 s[4:5], 0
# | 1678: buffer_load_dwordx2 v[0:1], v[0:1], s[4:7], 0 addr64
# | 1679: s_waitcnt vmcnt(0)
# | next:6811'0 X error: no match found
# | 1680: buffer_store_dword v0, v[2:3], s[4:7], 0 addr64
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1681: buffer_store_short v1, v[2:3], s[4:7], 0 addr64 offset:4
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1682: s_waitcnt vmcnt(0) expcnt(0)
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1683: s_setpc_b64 s[30:31]
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~
# | 1684: .Lfunc_end31:
# | next:6811'0 ~~~~~~~~~~~~~~
# | 1685: .size freeze_v3i16, .Lfunc_end31-freeze_v3i16
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1686: ; -- End function
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~
# | 1687: .set .Lfreeze_v3i16.num_vgpr, 4
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1688: .set .Lfreeze_v3i16.num_agpr, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1689: .set .Lfreeze_v3i16.numbered_sgpr, 32
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1690: .set .Lfreeze_v3i16.num_named_barrier, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1691: .set .Lfreeze_v3i16.private_seg_size, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | next:6811'1 ? possible intended match
# | 1692: .set .Lfreeze_v3i16.uses_vcc, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1693: .set .Lfreeze_v3i16.uses_flat_scratch, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1694: .set .Lfreeze_v3i16.has_dyn_sized_stack, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1695: .set .Lfreeze_v3i16.has_recursion, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1696: .set .Lfreeze_v3i16.has_indirect_call, 0
# | next:6811'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | .
# | .
# | .
# | 1954: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
# | 1955: s_mov_b32 s6, 0
# | 1956: s_mov_b32 s7, 0xf000
# | 1957: s_mov_b64 s[4:5], 0
# | 1958: buffer_load_dwordx2 v[0:1], v[0:1], s[4:7], 0 addr64
# | 1959: s_waitcnt vmcnt(0)
# | next:7525'0 X error: no match found
# | 1960: buffer_store_dword v0, v[2:3], s[4:7], 0 addr64
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1961: buffer_store_short v1, v[2:3], s[4:7], 0 addr64 offset:4
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1962: s_waitcnt vmcnt(0) expcnt(0)
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1963: s_setpc_b64 s[30:31]
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~
# | 1964: .Lfunc_end37:
# | next:7525'0 ~~~~~~~~~~~~~~
# | 1965: .size freeze_v3f16, .Lfunc_end37-freeze_v3f16
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1966: ; -- End function
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~
# | 1967: .set .Lfreeze_v3f16.num_vgpr, 4
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1968: .set .Lfreeze_v3f16.num_agpr, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1969: .set .Lfreeze_v3f16.numbered_sgpr, 32
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1970: .set .Lfreeze_v3f16.num_named_barrier, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1971: .set .Lfreeze_v3f16.private_seg_size, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | next:7525'1 ? possible intended match
# | 1972: .set .Lfreeze_v3f16.uses_vcc, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1973: .set .Lfreeze_v3f16.uses_flat_scratch, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1974: .set .Lfreeze_v3f16.has_dyn_sized_stack, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1975: .set .Lfreeze_v3f16.has_recursion, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 1976: .set .Lfreeze_v3f16.has_indirect_call, 0
# | next:7525'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | .
# | .
# | .
# | 2234: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
# | 2235: s_mov_b32 s6, 0
# | 2236: s_mov_b32 s7, 0xf000
# | 2237: s_mov_b64 s[4:5], 0
# | 2238: buffer_load_dwordx2 v[0:1], v[0:1], s[4:7], 0 addr64
# | 2239: s_waitcnt vmcnt(0)
# | next:8250'0 X error: no match found
# | 2240: buffer_store_dword v0, v[2:3], s[4:7], 0 addr64
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2241: buffer_store_short v1, v[2:3], s[4:7], 0 addr64 offset:4
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2242: s_waitcnt vmcnt(0) expcnt(0)
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2243: s_setpc_b64 s[30:31]
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~
# | 2244: .Lfunc_end43:
# | next:8250'0 ~~~~~~~~~~~~~~
# | 2245: .size freeze_v3bf16, .Lfunc_end43-freeze_v3bf16
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2246: ; -- End function
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~
# | 2247: .set .Lfreeze_v3bf16.num_vgpr, 4
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2248: .set .Lfreeze_v3bf16.num_agpr, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2249: .set .Lfreeze_v3bf16.numbered_sgpr, 32
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2250: .set .Lfreeze_v3bf16.num_named_barrier, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2251: .set .Lfreeze_v3bf16.private_seg_size, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | next:8250'1 ? possible intended match
# | 2252: .set .Lfreeze_v3bf16.uses_vcc, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2253: .set .Lfreeze_v3bf16.uses_flat_scratch, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2254: .set .Lfreeze_v3bf16.has_dyn_sized_stack, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2255: .set .Lfreeze_v3bf16.has_recursion, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | 2256: .set .Lfreeze_v3bf16.has_indirect_call, 0
# | next:8250'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# | .
# | .
# | .
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1
--
```
</details>
If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label.
https://github.com/llvm/llvm-project/pull/206043
More information about the llvm-commits
mailing list