[PATCH] D36193: AMDGPU: IMPLICIT_DEFs do not contribute to wait states

[Nicolai Hähnle via Phabricator via llvm-commits]

nhaehnle created this revision.
Herald added subscribers: t-tye, tpr, dstuttard, yaxunl, wdng, kzhuravl.

This fixes a bug that was exposed on gfx9 in various
GL45-CTS.shaders.loops.*_iterations.select_iteration_count_fragment tests,
e.g. GL45-CTS.shaders.loops.do_while_uniform_iterations.select_iteration_count_fragment


https://reviews.llvm.org/D36193

Files:
  lib/Target/AMDGPU/GCNHazardRecognizer.cpp
  test/CodeGen/AMDGPU/hazard.mir


Index: test/CodeGen/AMDGPU/hazard.mir
===================================================================
--- /dev/null
+++ test/CodeGen/AMDGPU/hazard.mir
@@ -0,0 +1,31 @@
+# RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs -run-pass post-RA-hazard-rec %s -o - | FileCheck -check-prefix=GCN -check-prefix=VI %s
+# RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs -run-pass post-RA-hazard-rec %s -o - | FileCheck -check-prefix=GCN -check-prefix=GFX9 %s
+
+# GCN:    bb.0.entry:
+# GCN:      %m0 = S_MOV_B32
+# GFX9:     S_NOP 0
+# VI-NOT:   S_NOP_0
+# GCN:      V_INTERP_P1_F32
+
+---
+name:            hazard_implicit_def
+alignment:       0
+exposesReturnsTwice: false
+legalized:       false
+regBankSelected: false
+selected:        false
+tracksRegLiveness: true
+registers:
+liveins:
+  - { reg: '%sgpr7', virtual-reg: '' }
+  - { reg: '%vgpr4', virtual-reg: '' }
+body:             |
+  bb.0.entry:
+    liveins: %sgpr7, %vgpr4
+
+    %m0 = S_MOV_B32 killed %sgpr7
+    %vgpr5 = IMPLICIT_DEF
+    %vgpr0 = V_INTERP_P1_F32 killed %vgpr4, 0, 0, implicit %m0, implicit %exec
+    SI_RETURN_TO_EPILOG killed %vgpr5, killed %vgpr0
+
+...
Index: lib/Target/AMDGPU/GCNHazardRecognizer.cpp
===================================================================
--- lib/Target/AMDGPU/GCNHazardRecognizer.cpp
+++ lib/Target/AMDGPU/GCNHazardRecognizer.cpp
@@ -218,12 +218,16 @@
 
 int GCNHazardRecognizer::getWaitStatesSince(
     function_ref<bool(MachineInstr *)> IsHazard) {
-  int WaitStates = -1;
+  int WaitStates = 0;
   for (MachineInstr *MI : EmittedInstrs) {
+    if (MI) {
+      if (IsHazard(MI))
+        return WaitStates;
+
+      if (MI->getOpcode() == AMDGPU::IMPLICIT_DEF)
+        continue;
+    }
     ++WaitStates;
-    if (!MI || !IsHazard(MI))
-      continue;
-    return WaitStates;
   }
   return std::numeric_limits<int>::max();
 }


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D36193.109232.patch
Type: text/x-patch
Size: 1868 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170801/0bde3e42/attachment.bin>