Make the preemptive compilation threshold configurable

Summary:
This diff allows specifying the preemptive compilation threshold via CompileFlags.
Users get a toggle in CLI and RuntimeConfig that allows choosing between 1. fully eager, 2. fully lazy, 3. use default thresholds ("smart"). The default is #3.

However, since we use `-O` by default, it overrides any lazy compilation when
using the `hermes` binary.

ChangeLog: [Internal] Inspector updated to use new RuntimeConfig laziness control

Reviewed By: tmikov

Differential Revision: D23356463

fbshipit-source-id: 508b7b2e6a218346c69acfec97e7891e388f0e9b

API/hermes/hermes.cpp

@@ -292,6 +292,21 @@ class HermesRuntimeImpl final : public HermesRuntime,
     compileFlags_.debug = true;
 #endif
 
+    switch (runtimeConfig.getCompilationMode()) {
+      case vm::SmartCompilation:
+        compileFlags_.lazy = true;
+        // (Leaves thresholds at default values)
+        break;
+      case vm::ForceEagerCompilation:
+        compileFlags_.lazy = false;
+        break;
+      case vm::ForceLazyCompilation:
+        compileFlags_.lazy = true;
+        compileFlags_.preemptiveFileCompilationThreshold = 0;
+        compileFlags_.preemptiveFunctionCompilationThreshold = 0;
+        break;
+    }
+
 #ifndef HERMESJSI_ON_STACK
     // Register the memory for the runtime if it isn't stored on the stack.
     crashMgr_->registerMemory(&runtime_, sizeof(vm::Runtime));
@@ -1217,10 +1232,8 @@ void HermesRuntime::debugJavaScript(
     const DebugFlags &debugFlags) {
   vm::Runtime &runtime = impl(this)->runtime_;
   vm::GCScope gcScope(&runtime);
-  hbc::CompileFlags flags{};
-  flags.debug = true;
-  flags.lazy = debugFlags.lazy;
-  vm::ExecutionStatus res = runtime.run(src, sourceURL, flags).getStatus();
+  vm::ExecutionStatus res =
+      runtime.run(src, sourceURL, impl(this)->compileFlags_).getStatus();
   impl(this)->checkStatus(res);
 }
 #endif
@@ -1308,9 +1321,6 @@ HermesRuntimeImpl::prepareJavaScript(
     bcErr = hbc::BCProviderFromBuffer::createBCProviderFromBuffer(
         std::move(buffer));
   } else {
-    compileFlags_.lazy =
-        (buffer->size() >=
-         ::hermes::hbc::kDefaultSizeThresholdForLazyCompilation);
 #if defined(HERMESVM_LEAN)
     bcErr.second = "prepareJavaScript source compilation not supported";
 #else

API/hermes/hermes.h

@@ -169,7 +169,9 @@ class HERMES_EXPORT HermesRuntime : public jsi::Runtime {
   debugger::Debugger &getDebugger();
 
   struct DebugFlags {
-    bool lazy{false};
+    // Looking for the .lazy flag? It's no longer necessary.
+    // Source is evaluated lazily by default. See
+    // RuntimeConfig::CompilationMode.
   };
 
   /// Evaluate the given code in an unoptimized form,

include/hermes/AST/Context.h

@@ -164,6 +164,14 @@ class Context {
   /// first time.
   bool lazyCompilation_{false};
 
+  /// Even if lazily compiling, eagerly compile any functions under this size in
+  /// bytes.
+  unsigned preemptiveFunctionCompilationThreshold_{0};
+
+  /// Even if lazily compiling, eagerly compile any files under this size in
+  /// bytes.
+  unsigned preemptiveFileCompilationThreshold_{0};
+
   /// Allows Function.toString() to return original source code. As with lazy
   /// compilation this requires source buffers, and hence this Context instance
   /// to be retained after compilation.
@@ -327,6 +335,22 @@ class Context {
     lazyCompilation_ = lazyCompilation;
   }
 
+  unsigned getPreemptiveFunctionCompilationThreshold() {
+    return preemptiveFunctionCompilationThreshold_;
+  }
+
+  void setPreemptiveFunctionCompilationThreshold(unsigned byteCount) {
+    preemptiveFunctionCompilationThreshold_ = byteCount;
+  };
+
+  unsigned getPreemptiveFileCompilationThreshold() {
+    return preemptiveFileCompilationThreshold_;
+  }
+
+  void setPreemptiveFileCompilationThreshold(unsigned byteCount) {
+    preemptiveFileCompilationThreshold_ = byteCount;
+  };
+
   bool allowFunctionToStringWithRuntimeSource() const {
     return allowFunctionToStringWithRuntimeSource_;
   }

include/hermes/BCGen/HBC/BytecodeProviderFromSrc.h

@@ -21,6 +21,16 @@ struct CompileFlags {
   bool optimize{false};
   bool debug{false};
   bool lazy{false};
+
+  /// Eagerly compile files under this number of bytes, even when lazy.
+  // Lazy compilation has significant per-module overhead, and is best applied
+  // to large bundles with a lot of unused code. Eager compilation is more
+  // efficient when compiling many small bundles with little unused code, such
+  // as when the API user loads smaller chunks of JS code on demand.
+  unsigned preemptiveFileCompilationThreshold{1 << 16};
+  /// Eagerly compile functions under this number of bytes, even when lazy.
+  unsigned preemptiveFunctionCompilationThreshold{160};
+
   bool allowFunctionToStringWithRuntimeSource{false};
   bool strict{false};
   /// The value is optional; when it is set, the optimization setting is based
@@ -41,13 +51,6 @@ struct CompileFlags {
   bool instrumentIR{false};
 };
 
-// The minimum code size in bytes before enabling lazy compilation.
-// Lazy compilation has significant per-module overhead, and is best applied
-// to large bundles with a lot of unused code. Eager compilation is more
-// efficient when compiling many small bundles with little unused code, such as
-// when the API user loads smaller chunks of JS code on demand.
-static constexpr unsigned kDefaultSizeThresholdForLazyCompilation = 1 << 16;
-
 #ifndef HERMESVM_LEAN
 /// BCProviderFromSrc is used when we are construction the bytecode from
 /// source compilation, i.e. we generate BytecodeModule/BytecodeFunction

lib/BCGen/HBC/BytecodeProviderFromSrc.cpp

@@ -162,8 +162,13 @@ BCProviderFromSrc::createBCProviderFromSrc(
 
   context->setStrictMode(compileFlags.strict);
   context->setEnableEval(true);
-  if (!compileFlags.optimize) {
-    context->setLazyCompilation(compileFlags.lazy);
+  context->setPreemptiveFunctionCompilationThreshold(
+      compileFlags.preemptiveFunctionCompilationThreshold);
+  context->setPreemptiveFileCompilationThreshold(
+      compileFlags.preemptiveFileCompilationThreshold);
+
+  if (compileFlags.lazy && !compileFlags.optimize) {
+    context->setLazyCompilation(true);
   }
 
   context->setAllowFunctionToStringWithRuntimeSource(
@@ -186,6 +191,8 @@ BCProviderFromSrc::createBCProviderFromSrc(
     declFileList.push_back(libParsed.getValue());
   }
 
+  bool isLargeFile =
+      buffer->size() >= context->getPreemptiveFileCompilationThreshold();
   int fileBufId = context->getSourceErrorManager().addNewSourceBuffer(
       llvh::make_unique<HermesLLVMMemoryBuffer>(std::move(buffer), sourceURL));
   if (sourceMap != nullptr) {
@@ -197,7 +204,7 @@ BCProviderFromSrc::createBCProviderFromSrc(
 
   auto parserMode = parser::FullParse;
   bool useStaticBuiltinDetected = false;
-  if (context->isLazyCompilation()) {
+  if (context->isLazyCompilation() && isLargeFile) {
     if (!parser::JSParser::preParseBuffer(
             *context, fileBufId, useStaticBuiltinDetected)) {
       return {nullptr, outputManager.getErrorString()};

lib/CompilerDriver/CompilerDriver.cpp

@@ -257,7 +257,13 @@ static opt<bool>
 static opt<bool> LazyCompilation(
     "lazy",
     init(false),
-    desc("Compile source lazily when executing (HBC only)"),
+    desc("Force fully lazy compilation"),
+    cat(CompilerCategory));
+
+static opt<bool> EagerCompilation(
+    "eager",
+    init(false),
+    desc("Force fully eager compilation"),
     cat(CompilerCategory));
 
 /// The following flags are exported so it may be used by the VM driver as well.
@@ -771,6 +777,9 @@ ESTree::NodePtr parseJS(
   // builtin' directive in the source.
   bool useStaticBuiltinDetected = false;
 
+  bool isLargeFile = fileBuf->getBufferSize() >=
+      context->getPreemptiveFileCompilationThreshold();
+
   int fileBufId =
       context->getSourceErrorManager().addNewSourceBuffer(std::move(fileBuf));
   if (sourceMap != nullptr && sourceMapTranslator != nullptr) {
@@ -779,7 +788,7 @@ ESTree::NodePtr parseJS(
 
   auto mode = parser::FullParse;
 
-  if (context->isLazyCompilation()) {
+  if (context->isLazyCompilation() && isLargeFile) {
     if (!parser::JSParser::preParseBuffer(
             *context, fileBufId, useStaticBuiltinDetected)) {
       return nullptr;
@@ -891,6 +900,10 @@ bool validateFlags() {
         "Specify output file with -out filename.");
   }
 
+  if (cl::LazyCompilation && cl::EagerCompilation) {
+    err("Can't specify both -lazy and -eager");
+  }
+
   // Validate lazy compilation flags.
   if (cl::LazyCompilation) {
     if (cl::BytecodeFormat != cl::BytecodeFormatKind::HBC)
@@ -1011,6 +1024,30 @@ std::shared_ptr<Context> createContext(
     context->getSourceErrorManager().disableAllWarnings();
   context->getSourceErrorManager().setErrorLimit(cl::ErrorLimit);
 
+  {
+    // Set default lazy mode using defaults from CompileFlags to keep it in one
+    // place.
+    hermes::hbc::CompileFlags defaultFlags{};
+    context->setPreemptiveFileCompilationThreshold(
+        defaultFlags.preemptiveFileCompilationThreshold);
+    context->setPreemptiveFunctionCompilationThreshold(
+        defaultFlags.preemptiveFunctionCompilationThreshold);
+  }
+
+  if (cl::EagerCompilation || cl::DumpTarget == EmitBundle ||
+      cl::OptimizationLevel > cl::OptLevel::Og) {
+    // Make sure nothing is lazy
+    context->setLazyCompilation(false);
+  } else if (cl::LazyCompilation) {
+    // Make sure everything is lazy
+    context->setLazyCompilation(true);
+    context->setPreemptiveFileCompilationThreshold(0);
+    context->setPreemptiveFunctionCompilationThreshold(0);
+  } else {
+    // By default with no optimization, use lazy compilation for "large" files
+    context->setLazyCompilation(true);
+  }
+
   if (cl::CommonJS) {
     context->setUseCJSModules(true);
   }
@@ -1633,9 +1670,6 @@ CompileResult processSourceFiles(
     }
   }
 
-  // Enable lazy compilation if requested.
-  context->setLazyCompilation(cl::LazyCompilation);
-
   // Allows Function.toString() to return original source code. As with lazy
   // compilation this requires source buffers to be retained after compilation.
   context->setAllowFunctionToStringWithRuntimeSource(cl::AllowFunctionToString);

lib/Parser/JSParserImpl.cpp

@@ -20,10 +20,6 @@ namespace hermes {
 namespace parser {
 namespace detail {
 
-// If a function's source code body is smaller than this number of bytes,
-// compile it immediately instead of creating a lazy stub.
-static const int PreemptiveCompilationThresholdBytes = 160;
-
 /// Declare a RAII recursion tracker. Check whether the recursion limit has
 /// been exceeded, and if so generate an error and return an empty
 /// llvh::Optional<>.
@@ -619,8 +615,8 @@ Optional<ESTree::BlockStatementNode *> JSParserImpl::parseFunctionBody(
     auto startLoc = tok_->getStartLoc();
     assert(preParsed_->bodyStartToEnd.count(startLoc) == 1);
     auto endLoc = preParsed_->bodyStartToEnd[startLoc];
-    if (endLoc.getPointer() - startLoc.getPointer() >
-        PreemptiveCompilationThresholdBytes) {
+    if ((unsigned)(endLoc.getPointer() - startLoc.getPointer()) >=
+        context_.getPreemptiveFunctionCompilationThreshold()) {
       lexer_.seek(endLoc);
       advance();
 

lib/VM/JSLib/eval.cpp

@@ -47,7 +47,7 @@ CallResult<HermesValue> evalInEnvironment(
   compileFlags.includeLibHermes = false;
   compileFlags.optimize = runtime->optimizedEval;
   compileFlags.lazy =
-      utf8code.size() >= hbc::kDefaultSizeThresholdForLazyCompilation;
+      utf8code.size() >= compileFlags.preemptiveFileCompilationThreshold;
   compileFlags.allowFunctionToStringWithRuntimeSource =
       runtime->getAllowFunctionToStringWithRuntimeSource();
 #ifdef HERMES_ENABLE_DEBUGGER

public/hermes/Public/RuntimeConfig.h

@@ -26,6 +26,12 @@ class raw_ostream;
 namespace hermes {
 namespace vm {
 
+enum CompilationMode {
+  SmartCompilation,
+  ForceEagerCompilation,
+  ForceLazyCompilation
+};
+
 class PinnedHermesValue;
 #ifdef HERMESVM_SERIALIZE
 class Serializer;
@@ -105,6 +111,12 @@ class Deserializer;
   /* runtime with CompileFlags::allowFunctionToStringWithRuntimeSource set. */ \
   F(constexpr, bool, AllowFunctionToStringWithRuntimeSource, false)            \
                                                                                \
+  /* Choose lazy/eager compilation mode. */                                    \
+  F(constexpr,                                                                 \
+    CompilationMode,                                                           \
+    CompilationMode,                                                           \
+    CompilationMode::SmartCompilation)                                         \
+                                                                               \
   /* An interface for managing crashes. */                                     \
   F(HERMES_NON_CONSTEXPR,                                                      \
     std::shared_ptr<CrashManager>,                                             \

test/IRGen/lazy-function-in-generator.js

@@ -14,10 +14,6 @@ function f() {
   function* g() {
     var g_var = 32;
     function h() {
-      /* Some text to pad out the function so that it won't be eagerly compiled
-       * for being too short. Lorem ipsum dolor sit amet, consectetur adipiscing
-       * elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.
-       */
       return f_var + g_var;
     }
     yield h();

test/IRGen/lazy-function-in-getter.js

@@ -12,10 +12,6 @@ function f() {
   return {
     get g() {
       function h() {
-        /* Some text to pad out the function so that it won't be eagerly compiled
-         * for being too short. Lorem ipsum dolor sit amet, consectetur adipiscing
-         * elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.
-         */
       }
     }
   };

test/debugger/fuzzy-lazy-oneline-break.js

@@ -9,7 +9,7 @@
 // REQUIRES: debugger
 
 // @nolint
-function foo() { print('foo called'); /** * Some text to pad out the function so that it won't be eagerly compiled * for being too short. Lorem ipsum dolor sit amet, consectetur adipiscing * elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.  */ function bar() { print('bar called'); /** * Some text to pad out the function so that it won't be eagerly compiled * for being too short. Lorem ipsum dolor sit amet, consectetur adipiscing * elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.  */ } function baz() { print('baz called'); /** * Some text to pad out the function so that it won't be eagerly compiled * for being too short. Lorem ipsum dolor sit amet, consectetur adipiscing * elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.  */ }; bar(); baz(); }
+function foo() { print('foo called'); function bar() { print('bar called'); } function baz() { print('baz called'); }; bar(); baz(); }
 
 debugger;
 foo();
@@ -23,15 +23,15 @@ foo();
 // CHECK: Break on 'debugger' statement in global: {{.*}}:14:1
 
 // Breakpoint is in the middle of the 'print' for 'bar called'
-// CHECK-NEXT: Set breakpoint 1 at {{.*}}:12:286
+// CHECK-NEXT: Set breakpoint 1 at {{.*}}:12:56
 // Breakpoint is in the middle of the 'print' for 'baz called'
-// CHECK-NEXT: Set breakpoint 2 at {{.*}}:12:556
+// CHECK-NEXT: Set breakpoint 2 at {{.*}}:12:96
 
 // CHECK-NEXT: Continuing execution
 // CHECK-NEXT: foo called
-// CHECK-NEXT: Break on breakpoint 1 in bar: {{.*}}:12:286
+// CHECK-NEXT: Break on breakpoint 1 in bar: {{.*}}:12:56
 // CHECK-NEXT: Continuing execution
 // CHECK-NEXT: bar called
-// CHECK-NEXT: Break on breakpoint 2 in baz: {{.*}}:12:556
+// CHECK-NEXT: Break on breakpoint 2 in baz: {{.*}}:12:96
 // CHECK-NEXT: Continuing execution
 // CHECK-NEXT: baz called

test/debugger/fuzzy-lazy-oneline-break.js.debug

@@ -1,5 +1,5 @@
-break 12 290
-break 12 560
+break 12 58
+break 12 99
 continue
 continue
 continue

test/debugger/lazy-break-disable.js

@@ -10,16 +10,12 @@
 
 function foo() {
   print('foo called');
-  /* Some text to pad out the function so that it won't be eagerly compiled
-   * for being too short. Lorem ipsum dolor sit amet, consectetur adipiscing
-   * elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.
-   */
 }
 
 debugger;
 foo();
 
-// CHECK: Break on 'debugger' statement in global: {{.*}}:19:1
+// CHECK: Break on 'debugger' statement in global: {{.*}}:15:1
 // CHECK-NEXT: Set breakpoint 1 at {{.*}}:12:3
 // CHECK-NEXT: Disabled breakpoint 1
 // CHECK-NEXT: 1 D {{.*}}:12:3