[llvm] 9615d48 - [AMDGPU][Uniformity] SI_IF and SI_ELSE pseudos are always divergent

[Sameer Sahasrabuddhe via llvm-commits]

Author: Sameer Sahasrabuddhe
Date: 2023-05-19T11:49:09+05:30
New Revision: 9615d48540a2d3615df91ae7f053bb5effb72e8e

URL: https://github.com/llvm/llvm-project/commit/9615d48540a2d3615df91ae7f053bb5effb72e8e
DIFF: https://github.com/llvm/llvm-project/commit/9615d48540a2d3615df91ae7f053bb5effb72e8e.diff

LOG: [AMDGPU][Uniformity] SI_IF and SI_ELSE pseudos are always divergent

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D150861

Added: 
    llvm/test/Analysis/UniformityAnalysis/AMDGPU/MIR/control-flow-intrinsics.mir

Modified: 
    llvm/include/llvm/ADT/GenericUniformityImpl.h
    llvm/lib/Target/AMDGPU/SIInstructions.td

Removed: 
    


################################################################################
diff  --git a/llvm/include/llvm/ADT/GenericUniformityImpl.h b/llvm/include/llvm/ADT/GenericUniformityImpl.h
index 71935d12ea03f..b6d4424fbb453 100644
--- a/llvm/include/llvm/ADT/GenericUniformityImpl.h
+++ b/llvm/include/llvm/ADT/GenericUniformityImpl.h
@@ -1147,9 +1147,11 @@ template <typename ContextT>
 void GenericUniformityAnalysisImpl<ContextT>::print(raw_ostream &OS) const {
   bool haveDivergentArgs = false;
 
-  if (DivergentValues.empty()) {
-    assert(DivergentTermBlocks.empty());
-    assert(DivergentExitCycles.empty());
+  // Control flow instructions may be divergent even if their inputs are
+  // uniform. Thus, although exceedingly rare, it is possible to have a program
+  // with no divergent values but with divergent control structures.
+  if (DivergentValues.empty() && DivergentTermBlocks.empty() &&
+      DivergentExitCycles.empty()) {
     OS << "ALL VALUES UNIFORM\n";
     return;
   }

diff  --git a/llvm/lib/Target/AMDGPU/SIInstructions.td b/llvm/lib/Target/AMDGPU/SIInstructions.td
index 0c67de58d602a..31914cc07eeaa 100644
--- a/llvm/lib/Target/AMDGPU/SIInstructions.td
+++ b/llvm/lib/Target/AMDGPU/SIInstructions.td
@@ -397,6 +397,7 @@ def SI_IF: CFPseudoInstSI <
   let Constraints = "";
   let Size = 12;
   let hasSideEffects = 1;
+  let IsNeverUniform = 1;
 }
 
 def SI_ELSE : CFPseudoInstSI <
@@ -404,6 +405,7 @@ def SI_ELSE : CFPseudoInstSI <
   (ins SReg_1:$src, brtarget:$target), [], 1, 1> {
   let Size = 12;
   let hasSideEffects = 1;
+  let IsNeverUniform = 1;
 }
 
 def SI_WATERFALL_LOOP : CFPseudoInstSI <

diff  --git a/llvm/test/Analysis/UniformityAnalysis/AMDGPU/MIR/control-flow-intrinsics.mir b/llvm/test/Analysis/UniformityAnalysis/AMDGPU/MIR/control-flow-intrinsics.mir
new file mode 100644
index 0000000000000..dec55e5662c8c
--- /dev/null
+++ b/llvm/test/Analysis/UniformityAnalysis/AMDGPU/MIR/control-flow-intrinsics.mir
@@ -0,0 +1,39 @@
+# RUN: llc -mtriple=amdgcn-- -mcpu=gfx900 -run-pass=print-machine-uniformity -o - %s 2>&1 | FileCheck %s
+
+---
+name:            f1
+body:             |
+  ; CHECK-LABEL: MachineUniformityInfo for function: f1
+  bb.0:
+    successors: %bb.1, %bb.2
+
+    ; CHECK-NOT: DIVERGENT: %1
+    %1:sreg_64(s64) = G_IMPLICIT_DEF
+    ; CHECK: DIVERGENT: {{.*}} SI_IF
+    %2:sreg_64 = SI_IF %1, %bb.2, implicit-def $exec, implicit-def $scc, implicit $exec
+
+  bb.1:
+    SI_RETURN
+
+  bb.2:
+    G_BR %bb.1
+
+...
+
+---
+name:            f2
+body:             |
+  ; CHECK-LABEL: MachineUniformityInfo for function: f2
+  bb.0:
+    successors: %bb.1, %bb.2
+
+    ; CHECK-NOT: DIVERGENT: %1
+    %1:sreg_64(s64) = G_IMPLICIT_DEF
+    ; CHECK: DIVERGENT: {{.*}} SI_ELSE
+    %2:sreg_64 = SI_ELSE %1, %bb.2, implicit-def $exec, implicit-def $scc, implicit $exec
+
+  bb.1:
+    SI_RETURN
+
+  bb.2:
+    G_BR %bb.1