[AMDGPU] Prevent sequences of non-instructions disrupting GCNHazardRecognizer wait state counting

Summary:
This fixes a bug where a large number of implicit def instructions can fill the GCNHazardRecognizer lookahead buffer causing required NOPs to not be inserted.

Reviewers: nhaehnle, arsenm

Reviewed By: arsenm

Subscribers: sheredom, kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits

Differential Revision: https://reviews.llvm.org/D51726

Change-Id: Ie75338f94de704ee5816b05afd0c922c6748a95b
llvm-svn: 341798

1 files changed, 9 insertions, 2 deletions

diff --git a/llvm/lib/Target/AMDGPU/GCNHazardRecognizer.cpp b/llvm/lib/Target/AMDGPU/GCNHazardRecognizer.cpp
index f236f10ba75..c6396de89c4 100644
--- a/llvm/lib/Target/AMDGPU/GCNHazardRecognizer.cpp
+++ b/llvm/lib/Target/AMDGPU/GCNHazardRecognizer.cpp
@@ -215,6 +215,14 @@ void GCNHazardRecognizer::AdvanceCycle() {
   if (!CurrCycleInstr)
     return;
 
+  // Do not track non-instructions which do not affect the wait states.
+  // If included, these instructions can lead to buffer overflow such that
+  // detectable hazards are missed.
+  if (CurrCycleInstr->getOpcode() == AMDGPU::IMPLICIT_DEF)
+    return;
+  else if (CurrCycleInstr->isDebugInstr())
+    return;
+
   unsigned NumWaitStates = TII.getNumWaitStates(*CurrCycleInstr);
 
   // Keep track of emitted instructions
@@ -253,8 +261,7 @@ int GCNHazardRecognizer::getWaitStatesSince(
         return WaitStates;
 
       unsigned Opcode = MI->getOpcode();
-      if (Opcode == AMDGPU::DBG_VALUE || Opcode == AMDGPU::IMPLICIT_DEF ||
-          Opcode == AMDGPU::INLINEASM)
+      if (Opcode == AMDGPU::INLINEASM)
         continue;
     }
     ++WaitStates;