[SYCL] Turn off part of SimplifyCFG optimizations in SYCL mode.

clang/test/CodeGenSYCL/simplifycfg.cpp

@@ -0,0 +1,30 @@
+// RUN: %clangxx -fsycl -fsycl-device-only -flegacy-pass-manager %s -O3 -S -o - | FileCheck %s
+// RUN: %clangxx -fsycl -fsycl-device-only -fno-legacy-pass-manager %s -O3 -S -o - | FileCheck %s
+//
+// This test checks that foo (which is @_Z3foov) is called twice after O3 optimizations.
+//
+// Usually clang with SimplifyCFG pass optimizes constructs like:
+// if (i % 2 == 0)
+//   func();
+// else
+//   func();
+//
+// into one simple func() invocation.
+// This behaviour might be wrong in cases when func's behaviour depends on
+// a place where it is written.
+// There is a relevant discussion about introducing
+// a reliable tool for such cases: https://reviews.llvm.org/D85603
+
+// CHECK: tail call spir_func void @_Z3foov()
+// CHECK: tail call spir_func void @_Z3foov()
+
+SYCL_EXTERNAL void foo();
+
+SYCL_EXTERNAL void bar(int i) {
+  if (i % 2 == 0) {
+    foo();
+  } else {
+    foo();
+  }
+}
+

llvm/lib/Passes/PassBuilderPipelines.cpp

@@ -223,6 +223,8 @@ extern cl::opt<bool> EnableMatrix;
 
 extern cl::opt<bool> DisablePreInliner;
 extern cl::opt<int> PreInlineThreshold;
+
+extern cl::opt<bool> SYCLOptimizationMode;
 } // namespace llvm
 
 void PassBuilder::invokePeepholeEPCallbacks(FunctionPassManager &FPM,
@@ -575,8 +577,12 @@ PassBuilder::buildFunctionSimplificationPipeline(OptimizationLevel Level,
   for (auto &C : ScalarOptimizerLateEPCallbacks)
     C(FPM, Level);
 
-  FPM.addPass(SimplifyCFGPass(
+  if (SYCLOptimizationMode)
+    FPM.addPass(SimplifyCFGPass());
+  else
+    FPM.addPass(SimplifyCFGPass(
       SimplifyCFGOptions().hoistCommonInsts(true).sinkCommonInsts(true)));
+
   FPM.addPass(InstCombinePass());
   invokePeepholeEPCallbacks(FPM, Level);
 
@@ -1029,7 +1035,8 @@ void PassBuilder::addVectorPasses(OptimizationLevel Level,
   // convert to more optimized IR using more aggressive simplify CFG options.
   // The extra sinking transform can create larger basic blocks, so do this
   // before SLP vectorization.
-  FPM.addPass(SimplifyCFGPass(SimplifyCFGOptions()
+  if (!SYCLOptimizationMode)
+    FPM.addPass(SimplifyCFGPass(SimplifyCFGOptions()
                                   .forwardSwitchCondToPhi(true)
                                   .convertSwitchToLookupTable(true)
                                   .needCanonicalLoops(false)

llvm/lib/Transforms/IPO/PassManagerBuilder.cpp

@@ -72,9 +72,8 @@ static cl::opt<bool>
 RunLoopRerolling("reroll-loops", cl::Hidden,
                  cl::desc("Run the loop rerolling pass"));
 
-static cl::opt<bool>
-    SYCLOptimizationMode("sycl-opt", cl::init(false), cl::Hidden,
-                         cl::desc("Enable SYCL optimization mode."));
+cl::opt<bool> SYCLOptimizationMode("sycl-opt", cl::init(false), cl::Hidden,
+                                   cl::desc("Enable SYCL optimization mode."));
 
 cl::opt<bool> RunNewGVN("enable-newgvn", cl::init(false), cl::Hidden,
                         cl::desc("Run the NewGVN pass"));
@@ -542,8 +541,12 @@ void PassManagerBuilder::addFunctionSimplificationPasses(
     MPM.add(createLoopRerollPass());
 
   // Merge & remove BBs and sink & hoist common instructions.
-  MPM.add(createCFGSimplificationPass(
-      SimplifyCFGOptions().hoistCommonInsts(true).sinkCommonInsts(true)));
+  if (SYCLOptimizationMode)
+    MPM.add(createCFGSimplificationPass());
+  else
+    MPM.add(createCFGSimplificationPass(
+        SimplifyCFGOptions().hoistCommonInsts(true).sinkCommonInsts(true)));
+
   // Clean up after everything.
   MPM.add(createInstructionCombiningPass());
   addExtensionsToPM(EP_Peephole, MPM);
@@ -607,12 +610,13 @@ void PassManagerBuilder::addVectorPasses(legacy::PassManagerBase &PM,
   // convert to more optimized IR using more aggressive simplify CFG options.
   // The extra sinking transform can create larger basic blocks, so do this
   // before SLP vectorization.
-  PM.add(createCFGSimplificationPass(SimplifyCFGOptions()
-                                         .forwardSwitchCondToPhi(true)
-                                         .convertSwitchToLookupTable(true)
-                                         .needCanonicalLoops(false)
-                                         .hoistCommonInsts(true)
-                                         .sinkCommonInsts(true)));
+  if (!SYCLOptimizationMode)
+    PM.add(createCFGSimplificationPass(SimplifyCFGOptions()
+                                           .forwardSwitchCondToPhi(true)
+                                           .convertSwitchToLookupTable(true)
+                                           .needCanonicalLoops(false)
+                                           .hoistCommonInsts(true)
+                                           .sinkCommonInsts(true)));
 
   if (IsFullLTO) {
     PM.add(createSCCPPass());                 // Propagate exposed constants