Direct calls to @CriticalNative methods.

Emit direct calls from compiled managed code to the native
code registered with the method, avoiding the JNI stub.

Golem results:
art-opt-cc                       x86 x86-64    arm  arm64
NativeDowncallStaticCritical  +12.5% +62.5% +75.9% +41.7%
NativeDowncallStaticCritical6 +55.6% +87.5% +72.1% +35.3%
art-opt                          x86 x86-64    arm  arm64
NativeDowncallStaticCritical  +28.6% +85.6% +76.4% +38.4%
NativeDowncallStaticCritical6 +44.6% +44.6% +74.6% +32.2%

Test: Covered by 178-app-image-native-method.
Test: m test-art-host-gtest
Test: testrunner.py --host --debuggable --ndebuggable \
          --optimizing --jit --jit-on-first-use
Test: run-gtests.sh
Test: testrunner.py --target --optimizing
Test: testrunner.py --target --debuggable --ndebuggable \
          --optimizing --jit --jit-on-first-use -t 178
Test: aosp_cf_x86_phone-userdebug boots.
Test: aosp_cf_x86_phone-userdebug/jitzygote boots.
Bug: 112189621
Change-Id: I8b37da51e8fe0b7bc513bb81b127fe0416068866

1 files changed, 32 insertions, 0 deletions

diff --git a/compiler/optimizing/code_generator.h b/compiler/optimizing/code_generator.h
index ff2be4740d..4bfc14a12b 100644
--- a/compiler/optimizing/code_generator.h
+++ b/compiler/optimizing/code_generator.h
@@ -570,6 +570,28 @@ class CodeGenerator : public DeletableArenaObject<kArenaAllocCodeGenerator> {
   static void CreateCommonInvokeLocationSummary(
       HInvoke* invoke, InvokeDexCallingConventionVisitor* visitor);
 
+  template <typename CriticalNativeCallingConventionVisitor,
+            size_t kNativeStackAlignment,
+            size_t GetCriticalNativeDirectCallFrameSize(const char* shorty, uint32_t shorty_len)>
+  static size_t PrepareCriticalNativeCall(HInvokeStaticOrDirect* invoke,
+                                          /*out*/HParallelMove* parallel_move) {
+      DCHECK(!invoke->GetLocations()->Intrinsified());
+      CriticalNativeCallingConventionVisitor calling_convention_visitor(
+          /*for_register_allocation=*/ false);
+      PrepareCriticalNativeArgumentMoves(invoke, &calling_convention_visitor, parallel_move);
+      size_t out_frame_size =
+          RoundUp(calling_convention_visitor.GetStackOffset(), kNativeStackAlignment);
+      if (kIsDebugBuild) {
+        uint32_t shorty_len;
+        const char* shorty = GetCriticalNativeShorty(invoke, &shorty_len);
+        DCHECK_EQ(GetCriticalNativeDirectCallFrameSize(shorty, shorty_len), out_frame_size);
+      }
+      if (out_frame_size != 0u) {
+        AdjustCriticalNativeArgumentMoves(out_frame_size, parallel_move);
+      }
+      return out_frame_size;
+  }
+
   void GenerateInvokeStaticOrDirectRuntimeCall(
       HInvokeStaticOrDirect* invoke, Location temp, SlowPathCode* slow_path);
 
@@ -799,6 +821,16 @@ class CodeGenerator : public DeletableArenaObject<kArenaAllocCodeGenerator> {
                        bool needs_vreg_info = true);
   void EmitVRegInfo(HEnvironment* environment, SlowPathCode* slow_path);
 
+  static void PrepareCriticalNativeArgumentMoves(
+      HInvokeStaticOrDirect* invoke,
+      /*inout*/InvokeDexCallingConventionVisitor* visitor,
+      /*out*/HParallelMove* parallel_move);
+
+  static void AdjustCriticalNativeArgumentMoves(size_t out_frame_size,
+                                                /*inout*/HParallelMove* parallel_move);
+
+  static const char* GetCriticalNativeShorty(HInvokeStaticOrDirect* invoke, uint32_t* shorty_len);
+
   OptimizingCompilerStats* stats_;
 
   HGraph* const graph_;