[llvm] 8dfb417 - [AMDGPU] Fix missing waitcnt issue
Piotr Sobczak via llvm-commits
llvm-commits at lists.llvm.org
Wed Jan 19 01:55:15 PST 2022
Author: Piotr Sobczak
Date: 2022-01-19T10:54:44+01:00
New Revision: 8dfb417e67e33055a72980ede92c128763d549ad
URL: https://github.com/llvm/llvm-project/commit/8dfb417e67e33055a72980ede92c128763d549ad
DIFF: https://github.com/llvm/llvm-project/commit/8dfb417e67e33055a72980ede92c128763d549ad.diff
LOG: [AMDGPU] Fix missing waitcnt issue
Ignore out of order counters when merging brackets. The fact that
there was a pending event in the old state does not guarantee that
the waitcnt was generated, so we still need to conservatively re-process
the block.
The patch fixes a correctness issue where the block was not re-processed
and the waitcnt not inserted in consequence.
Differential Revision: https://reviews.llvm.org/D117544
Added:
Modified:
llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
llvm/test/CodeGen/AMDGPU/waitcnt.mir
Removed:
################################################################################
diff --git a/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp b/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
index 6fbe5d45ce0a..45c427c18784 100644
--- a/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
@@ -1382,7 +1382,6 @@ bool WaitcntBrackets::merge(const WaitcntBrackets &Other) {
for (auto T : inst_counter_types()) {
// Merge event flags for this counter
- const bool OldOutOfOrder = counterOutOfOrder(T);
const unsigned OldEvents = PendingEvents & WaitEventMaskForInst[T];
const unsigned OtherEvents = Other.PendingEvents & WaitEventMaskForInst[T];
if (OtherEvents & ~OldEvents)
@@ -1425,7 +1424,7 @@ bool WaitcntBrackets::merge(const WaitcntBrackets &Other) {
}
}
- if (RegStrictDom && !OldOutOfOrder)
+ if (RegStrictDom)
StrictDom = true;
}
diff --git a/llvm/test/CodeGen/AMDGPU/waitcnt.mir b/llvm/test/CodeGen/AMDGPU/waitcnt.mir
index 96bd9b3ff6f2..5f065f9be831 100644
--- a/llvm/test/CodeGen/AMDGPU/waitcnt.mir
+++ b/llvm/test/CodeGen/AMDGPU/waitcnt.mir
@@ -44,6 +44,10 @@
define amdgpu_kernel void @subregs16bit() {
ret void
}
+
+ define amdgpu_kernel void @waitcnt_backedge() {
+ ret void
+ }
...
---
@@ -332,3 +336,35 @@ body: |
$vgpr1 = FLAT_LOAD_USHORT killed $vgpr2_vgpr3, 0, 0, implicit $exec, implicit $flat_scr
V_NOP_e32 implicit $exec, implicit $vgpr0_lo16, implicit $vgpr1_lo16
...
+
+---
+# Waitcnt required before the use of $sgpr10_sgpr11, as the S_LOAD also writes
+# to $sgpr10_sgpr11, and can occur first in the program running order.
+
+# CHECK-LABEL: name: waitcnt_backedge
+# CHECK: S_WAITCNT
+# CHECK: $sgpr10_sgpr11 = S_CSELECT_B64
+# CHECK: $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM
+
+
+name: waitcnt_backedge
+body: |
+ bb.0:
+ renamable $sgpr4_sgpr5_sgpr6_sgpr7 = S_LOAD_DWORDX4_IMM renamable $sgpr2_sgpr3, 32, 0 :: (load (s128) from `i32 addrspace(4)* undef`, addrspace 4)
+
+ bb.4:
+ renamable $sgpr10_sgpr11 = S_CSELECT_B64 -1, 0, implicit killed $scc
+ renamable $vgpr1 = BUFFER_LOAD_DWORD_OFFEN killed renamable $vgpr5, renamable $sgpr4_sgpr5_sgpr6_sgpr7, 0, 0, 1, 0, 0, implicit $exec :: (dereferenceable load (s32), align 1, addrspace 4)
+ renamable $sgpr10_sgpr11 = S_LOAD_DWORDX2_IMM killed renamable $sgpr0_sgpr1, 0, 0 :: (load (s64) from `i32 addrspace(4)* undef`, align 4, addrspace 4)
+ S_CBRANCH_SCC0 %bb.9, implicit killed $scc
+
+ bb.9:
+ renamable $vgpr1 = V_CNDMASK_B32_e64 0, 0, 0, 1, killed $sgpr14_sgpr15, implicit $exec
+ S_CBRANCH_SCC0 %bb.14, implicit killed $scc
+
+ bb.10:
+ S_BRANCH %bb.4
+
+ bb.14:
+ S_ENDPGM 0
+...
More information about the llvm-commits
mailing list