[builtins] Hash & verify the embedded blob

Embed a hash of the embedded blob at serialization-time, and verify it
when loading the blob.

Drive-by: Keep the size of the blob trimmed, and only page-align for
dynamic page allocation and freeing during mksnapshot.

Bug: v8:6666
Change-Id: I10b6064f4f2847d4cc400f6f1bd6961194985b8f
Reviewed-on: https://chromium-review.googlesource.com/1019144
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#52707}

src/instruction-stream.cc

@@ -69,7 +69,7 @@ void InstructionStream::CreateOffHeapInstructionStream(Isolate* isolate,
                        PageAllocator::kReadExecute));
 
   *data = allocated_bytes;
-  *size = allocated_size;
+  *size = d.size();
 
   d.Dispose();
 }
@@ -77,7 +77,8 @@ void InstructionStream::CreateOffHeapInstructionStream(Isolate* isolate,
 // static
 void InstructionStream::FreeOffHeapInstructionStream(uint8_t* data,
                                                      uint32_t size) {
-  CHECK(FreePages(data, size));
+  const uint32_t page_size = static_cast<uint32_t>(AllocatePageSize());
+  CHECK(FreePages(data, RoundUp(size, page_size)));
 }
 #endif  // V8_EMBEDDED_BUILTINS
 

src/isolate.cc

@@ -94,6 +94,15 @@ void Isolate::SetEmbeddedBlob(const uint8_t* blob, uint32_t blob_size) {
   embedded_blob_size_ = blob_size;
   current_embedded_blob_.store(blob, std::memory_order_relaxed);
   current_embedded_blob_size_.store(blob_size, std::memory_order_relaxed);
+
+#ifdef DEBUG
+  if (blob != nullptr) {
+    // Verify that the contents of the embedded blob are unchanged from
+    // serialization-time, just to ensure the compiler isn't messing with us.
+    EmbeddedData d = EmbeddedData::FromBlob();
+    CHECK_EQ(d.Hash(), d.CreateHash());
+  }
+#endif  // DEBUG
 }
 
 const uint8_t* Isolate::embedded_blob() const { return embedded_blob_; }

src/snapshot/snapshot-common.cc

@@ -393,7 +393,17 @@ EmbeddedData EmbeddedData::FromIsolate(Isolate* isolate) {
                 code->raw_instruction_size());
   }
 
-  return {blob, blob_size};
+  EmbeddedData d(blob, blob_size);
+
+  // Hash the blob and store the result.
+  STATIC_ASSERT(HashSize() == kSizetSize);
+  const size_t hash = d.CreateHash();
+  std::memcpy(blob + HashOffset(), &hash, HashSize());
+
+  DCHECK_EQ(hash, d.CreateHash());
+  DCHECK_EQ(hash, d.Hash());
+
+  return d;
 }
 
 EmbeddedData EmbeddedData::FromBlob() {
@@ -406,10 +416,10 @@ EmbeddedData EmbeddedData::FromBlob() {
 
 Address EmbeddedData::InstructionStartOfBuiltin(int i) const {
   DCHECK(Builtins::IsBuiltinId(i));
-
   const uint32_t* offsets = Offsets();
   const uint8_t* result = RawData() + offsets[i];
-  DCHECK_LT(result, data_ + size_);
+  DCHECK_LE(result, data_ + size_);
+  DCHECK_IMPLIES(result == data_ + size_, InstructionSizeOfBuiltin(i) == 0);
   return reinterpret_cast<Address>(result);
 }
 
@@ -418,6 +428,12 @@ uint32_t EmbeddedData::InstructionSizeOfBuiltin(int i) const {
   const uint32_t* lengths = Lengths();
   return lengths[i];
 }
+
+size_t EmbeddedData::CreateHash() const {
+  STATIC_ASSERT(HashOffset() == 0);
+  STATIC_ASSERT(HashSize() == kSizetSize);
+  return base::hash_range(data_ + HashSize(), data_ + size_);
+}
 #endif
 
 uint32_t Snapshot::ExtractNumContexts(const v8::StartupData* data) {

src/snapshot/snapshot.h

@@ -100,16 +100,26 @@ class EmbeddedData final {
     return RoundUp<kCodeAlignment>(InstructionSizeOfBuiltin(i));
   }
 
+  size_t CreateHash() const;
+  size_t Hash() const {
+    return *reinterpret_cast<const size_t*>(data_ + HashOffset());
+  }
+
   // The layout of the blob is as follows:
   //
-  // [0] offset of instruction stream 0
-  // ... offsets
-  // [N] length of instruction stream 0
-  // ... lengths
-  // ... instruction streams
+  // [0]     hash of the remaining blob
+  // [1]     offset of instruction stream 0
+  // ...     offsets
+  // [N + 1] length of instruction stream 0
+  // ...     lengths
+  // ...     instruction streams
 
   static constexpr uint32_t kTableSize = Builtins::builtin_count;
-  static constexpr uint32_t OffsetsOffset() { return 0; }
+  static constexpr uint32_t HashOffset() { return 0; }
+  static constexpr uint32_t HashSize() { return kSizetSize; }
+  static constexpr uint32_t OffsetsOffset() {
+    return HashOffset() + HashSize();
+  }
   static constexpr uint32_t OffsetsSize() { return kUInt32Size * kTableSize; }
   static constexpr uint32_t LengthsOffset() {
     return OffsetsOffset() + OffsetsSize();