-
Notifications
You must be signed in to change notification settings - Fork 14.4k
[AMDGPU] auto update some tests to prepare for future changes #147256
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Merged
+818
−552
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
@llvm/pr-subscribers-backend-amdgpu Author: Sameer Sahasrabuddhe (ssahasra) ChangesPatch is 141.49 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/147256.diff 2 Files Affected:
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-local.mir b/llvm/test/CodeGen/AMDGPU/memory-legalizer-local.mir
index adaee7ebaddd3..56dd95e373dc6 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-local.mir
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-local.mir
@@ -1,18 +1,23 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5
# RUN: llc -mtriple=amdgcn -mcpu=gfx803 -run-pass si-memory-legalizer %s -o - | FileCheck -check-prefix=GCN %s
---
-# GCN-LABEL: name: load_singlethread_unordered
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_singlethread_unordered
body: |
bb.0:
+ ; GCN-LABEL: name: load_singlethread_unordered
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("singlethread-one-as") unordered (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -26,17 +31,21 @@ body: |
...
---
-# GCN-LABEL: name: load_singlethread_monotonic
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_singlethread_monotonic
body: |
bb.0:
+ ; GCN-LABEL: name: load_singlethread_monotonic
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("singlethread-one-as") monotonic (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -50,17 +59,21 @@ body: |
...
---
-# GCN-LABEL: name: load_singlethread_acquire
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_singlethread_acquire
body: |
bb.0:
+ ; GCN-LABEL: name: load_singlethread_acquire
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("singlethread-one-as") acquire (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -74,17 +87,21 @@ body: |
...
---
-# GCN-LABEL: name: load_singlethread_seq_cst
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_singlethread_seq_cst
body: |
bb.0:
+ ; GCN-LABEL: name: load_singlethread_seq_cst
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("singlethread-one-as") seq_cst (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -98,17 +115,21 @@ body: |
...
---
-# GCN-LABEL: name: load_wavefront_unordered
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_wavefront_unordered
body: |
bb.0:
+ ; GCN-LABEL: name: load_wavefront_unordered
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("wavefront-one-as") unordered (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -122,17 +143,21 @@ body: |
...
---
-# GCN-LABEL: name: load_wavefront_monotonic
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_wavefront_monotonic
body: |
bb.0:
+ ; GCN-LABEL: name: load_wavefront_monotonic
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("wavefront-one-as") monotonic (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -146,17 +171,21 @@ body: |
...
---
-# GCN-LABEL: name: load_wavefront_acquire
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_wavefront_acquire
body: |
bb.0:
+ ; GCN-LABEL: name: load_wavefront_acquire
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("wavefront-one-as") acquire (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -170,17 +199,21 @@ body: |
...
---
-# GCN-LABEL: name: load_wavefront_seq_cst
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_wavefront_seq_cst
body: |
bb.0:
+ ; GCN-LABEL: name: load_wavefront_seq_cst
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("wavefront-one-as") seq_cst (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -194,17 +227,21 @@ body: |
...
---
-# GCN-LABEL: name: load_workgroup_unordered
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_workgroup_unordered
body: |
bb.0:
+ ; GCN-LABEL: name: load_workgroup_unordered
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("workgroup-one-as") unordered (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -218,17 +255,21 @@ body: |
...
---
-# GCN-LABEL: name: load_workgroup_monotonic
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_workgroup_monotonic
body: |
bb.0:
+ ; GCN-LABEL: name: load_workgroup_monotonic
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("workgroup-one-as") monotonic (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -242,17 +283,21 @@ body: |
...
---
-# GCN-LABEL: name: load_workgroup_acquire
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_workgroup_acquire
body: |
bb.0:
+ ; GCN-LABEL: name: load_workgroup_acquire
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 killed renamable $vgpr0, 0, 0, implicit $m0, implicit $exec :: (volatile load syncscope("workgroup-one-as") acquire (s32) from `ptr addrspace(3) poison`, addrspace 3)
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 $sgpr0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $sgpr0_sgpr1
+ ; GCN-NEXT: $vgpr1 = V_MOV_B32_e32 killed $sgpr1, implicit $exec, implicit $sgpr0_sgpr1, implicit $exec
+ ; GCN-NEXT: FLAT_STORE_DWORD killed renamable $vgpr0_vgpr1, killed renamable $vgpr2, 0, 0, implicit $exec, implicit $flat_scr :: (store (s32) into `ptr poison`)
+ ; GCN-NEXT: S_ENDPGM 0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
$sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
$m0 = S_MOV_B32 -1
@@ -266,17 +311,21 @@ body: |
...
---
-# GCN-LABEL: name: load_workgroup_seq_cst
-# GCN-LABEL: bb.0:
-# GCN-NOT: S_WAITCNT
-# GCN: DS_READ_B32
-# GCN-NOT: S_WAITCNT
-# GCN: FLAT_STORE_DWORD
name: load_workgroup_seq_cst
body: |
bb.0:
+ ; GCN-LABEL: name: load_workgroup_seq_cst
+ ; GCN: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 36, 0 :: (dereferenceable invariant load (s32) from `ptr addrspace(4) poison`, addrspace 4)
+ ; GCN-NEXT: $sgpr0_sgpr1 = S_LOAD_DWORDX2_IMM killed $sgpr0_sgpr1, 44, 0 :: (dereferenceable invariant load (s64) from `ptr addrspace(4) poison`, align 4, addrspace 4)
+ ; GCN-NEXT: $m0 = S_MOV_B32 -1
+ ; GCN-NEXT: $vgpr0 = V_MOV_B32_e32 killed $sgpr2, implicit $exec, implicit $exec
+ ; GCN-NEXT: renamable $vgpr2 = DS_READ_B32 k...
[truncated]
|
This was referenced Jul 7, 2025
Pierre-vh
approved these changes
Jul 7, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.