Merge 3.2.x branch into master (#2648)

* Down-integrate internal changes to github.

* Update conformance test failure list.

* Explicitly import used class in nano test to avoid random test fail.

* Update _GNUC_VER to use the correct implementation of atomic operation
on Mac.

* maps_test.js: check whether Symbol is defined before using it (#2524)

Symbol is not yet available on older versions of Node.js and so this
test fails with them. This change just directly checks whether Symbol is
available before we try to use it.

* Added well_known_types_embed.cc to CLEANFILES so that it gets cleaned up

* Updated Makefile.am to fix out-of-tree builds

* Added Bazel genrule for generating well_known_types_embed.cc

In pull request #2517 I made this change for the CMake and autotools
builds but forgot to do it for the Bazel build.

* Update _GNUC_VER to use the correct implementation of atomic operation on Mac.

* Add new js file in extra dist.

* Bump version number to 3.2.0

* Fixed issue with autoloading - Invalid paths (#2538)

* PHP fix int64 decoding (#2516)

* fix int64 decoding

* fix int64 decoding + tests

* Fix int64 decoding on 32-bit machines.

* Fix warning in compiler/js/embed.cc

embed.cc: In function ‘std::string CEscape(const string&)’:
embed.cc:51:32: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
   for (int i = 0; i < str.size(); ++i) {
                                ^

* Fix include in auto-generated well_known_types_embed.cc

Restore include style fix (e3da722) that has been trampled by
auto-generation of well_known_types_embed.cc

* Fixed cross compilations with the Autotools build

Pull request #2517 caused cross compilations to start failing, because
the js_embed binary was being built to run on the target platform
instead of on the build machine. This change updates the Autotools build
to use the AX_PROG_CXX_FOR_BUILD macro to find a suitable compiler for
the build machine and always use that when building js_embed.

* Minor fix for autocreated object repeated fields and maps.

- If setting/clearing a repeated field/map that was objects, check the class
  before checking the autocreator.
- Just to be paranoid, don’t mutate within copy/mutableCopy for the autocreated
  classes to ensure there is less chance of issues if someone does something
  really crazy threading wise.
- Some more tests for the internal AutocreatedArray/AutocreatedDictionary
  classes to ensure things are working as expected.
- Add Xcode 8.2 to the full_mac_build.sh supported list.

* Fix generation of extending nested messages in JavaScript (#2439)

* Fix generation of extending nested messages in JavaScript

* Added missing test8.proto to build

* Fix generated code when there is no namespace but there is enum definition.

* Decoding unknown field should succeed.

* Add embed.cc in src/Makefile.am to fix dist check.

* Fixed "make distcheck" for the Autotools build

To make the test pass I needed to fix out-of-tree builds and update
EXTRA_DIST and CLEANFILES.

* Remove redundent embed.cc from src/Makefile.am

* Update version number to 3.2.0-rc.1 (#2578)

* Change protoc-artifacts version to 3.2.0-rc.1

* Update version number to 3.2.0rc2

* Update change logs for 3.2.0 release.

* Update php README

* Update upb, fixes some bugs (including a hash table problem). (#2611)

* Update upb, fixes some bugs (including a hash table problem).

* Ruby: added a test for the previous hash table corruption.

Verified that this triggers the bug in the currently released
version.

* Ruby: bugfix for SEGV.

* Ruby: removed old code for dup'ing defs.

* Reverting deployment target to 7.0 (#2618)

The Protobuf library doesn’t require the 7.1 deployment target so
reverting it back to 7.0

* Fix typo that breaks builds on big-endian (#2632)

* Bump version number to 3.2.0

1 files changed, 233 insertions, 0 deletions

diff --git a/src/google/protobuf/wire_format_lite.cc b/src/google/protobuf/wire_format_lite.cc
index 5a8857ed..e46ac400 100644
--- a/src/google/protobuf/wire_format_lite.cc
+++ b/src/google/protobuf/wire_format_lite.cc
@@ -34,6 +34,9 @@
 
 #include <google/protobuf/wire_format_lite_inl.h>
 
+#ifdef __SSE_4_1__
+#include <immintrin.h>
+#endif
 #include <stack>
 #include <string>
 #include <vector>
@@ -337,6 +340,90 @@ bool WireFormatLite::ReadPackedEnumPreserveUnknowns(
   return true;
 }
 
+namespace {
+void EncodeFixedSizeValue(float v, uint8* dest) {
+  WireFormatLite::WriteFloatNoTagToArray(v, dest);
+}
+
+void EncodeFixedSizeValue(double v, uint8* dest) {
+  WireFormatLite::WriteDoubleNoTagToArray(v, dest);
+}
+
+void EncodeFixedSizeValue(uint32 v, uint8* dest) {
+  WireFormatLite::WriteFixed32NoTagToArray(v, dest);
+}
+
+void EncodeFixedSizeValue(uint64 v, uint8* dest) {
+  WireFormatLite::WriteFixed64NoTagToArray(v, dest);
+}
+
+void EncodeFixedSizeValue(int32 v, uint8* dest) {
+  WireFormatLite::WriteSFixed32NoTagToArray(v, dest);
+}
+
+void EncodeFixedSizeValue(int64 v, uint8* dest) {
+  WireFormatLite::WriteSFixed64NoTagToArray(v, dest);
+}
+
+void EncodeFixedSizeValue(bool v, uint8* dest) {
+  WireFormatLite::WriteBoolNoTagToArray(v, dest);
+}
+}  // anonymous namespace
+
+template <typename CType>
+static void WriteArray(const CType* a, int n, io::CodedOutputStream* output) {
+#if defined(PROTOBUF_LITTLE_ENDIAN)
+  output->WriteRaw(reinterpret_cast<const char*>(a), n * sizeof(a[0]));
+#else
+  const int kAtATime = 128;
+  uint8 buf[sizeof(CType) * kAtATime];
+  for (int i = 0; i < n; i += kAtATime) {
+    int to_do = min(kAtATime, n - i);
+    uint8* ptr = buf;
+    for (int j = 0; j < to_do; j++) {
+      EncodeFixedSizeValue(a[i+j], ptr);
+      ptr += sizeof(a[0]);
+    }
+    output->WriteRaw(buf, to_do * sizeof(a[0]));
+  }
+#endif
+}
+
+void WireFormatLite::WriteFloatArray(const float* a, int n,
+                                     io::CodedOutputStream* output) {
+  WriteArray<float>(a, n, output);
+}
+
+void WireFormatLite::WriteDoubleArray(const double* a, int n,
+                                     io::CodedOutputStream* output) {
+  WriteArray<double>(a, n, output);
+}
+
+void WireFormatLite::WriteFixed32Array(const uint32* a, int n,
+                                     io::CodedOutputStream* output) {
+  WriteArray<uint32>(a, n, output);
+}
+
+void WireFormatLite::WriteFixed64Array(const uint64* a, int n,
+                                       io::CodedOutputStream* output) {
+  WriteArray<uint64>(a, n, output);
+}
+
+void WireFormatLite::WriteSFixed32Array(const int32* a, int n,
+                                       io::CodedOutputStream* output) {
+  WriteArray<int32>(a, n, output);
+}
+
+void WireFormatLite::WriteSFixed64Array(const int64* a, int n,
+                                       io::CodedOutputStream* output) {
+  WriteArray<int64>(a, n, output);
+}
+
+void WireFormatLite::WriteBoolArray(const bool* a, int n,
+                                    io::CodedOutputStream* output) {
+  WriteArray<bool>(a, n, output);
+}
+
 void WireFormatLite::WriteInt32(int field_number, int32 value,
                                 io::CodedOutputStream* output) {
   WriteTag(field_number, WIRETYPE_VARINT, output);
@@ -540,6 +627,152 @@ bool WireFormatLite::VerifyUtf8String(const char* data,
   return true;
 }
 
+#ifdef __SSE_4_1__
+template<typename T, bool ZigZag, bool SignExtended>
+static size_t VarintSize(
+    const T* data, const int n,
+    const internal::enable_if<sizeof(T) == 4>::type* = NULL) {
+#if __cplusplus >= 201103L
+  // is_unsigned<T> => !ZigZag
+  static_assert((std::is_unsigned<T>::value ^ ZigZag) ||
+                std::is_signed<T>::value,
+                "Cannot ZigZag encode unsigned types");
+  // is_unsigned<T> => !SignExtended
+  static_assert((std::is_unsigned<T>::value ^ SignExtended) ||
+                std::is_signed<T>::value,
+                "Cannot SignExtended unsigned types");
+#endif
+
+  union vus32 {
+    uint32  u[4];
+    int32   s[4];
+    __m128i v;
+  };
+
+  static const vus32 ones = {{1, 1, 1, 1}};
+
+  // CodedOutputStream::VarintSize32SignExtended returns 10 for negative
+  // numbers.  We can apply the UInt32Size algorithm, and simultaneously logical
+  // shift the MSB into the LSB to determine if it is negative.
+  static const vus32 fives = {{5, 5, 5, 5}};
+
+  // sum is the vectorized-output of calling CodedOutputStream::VarintSize32 on
+  // the processed elements.
+  //
+  // msb_sum is the count of set most-significant bits.  When computing the
+  // vectorized CodedOutputStream::VarintSize32SignExtended, negative values
+  // have the most significant bit set.  VarintSize32SignExtended returns 10 and
+  // VarintSize32 returns 5.  msb_sum allows us to compute:
+  //   VarintSize32SignExtended = msb_sum * 5 + VarintSize32
+  vus32 sum, v, msb_sum;
+  sum.v = _mm_setzero_si128();
+  msb_sum.v = _mm_setzero_si128();
+
+  int rounded = n & ~(3);
+  int i;
+  for (i = 0; i < rounded; i += 4) {
+    v.v = _mm_loadu_si128(reinterpret_cast<const __m128i*>(&data[i]));
+
+    if (ZigZag) {
+      // Note:  the right-shift must be arithmetic
+      v.v = _mm_xor_si128(_mm_slli_epi32(v.v, 1), _mm_srai_epi32(v.v, 31));
+    }
+
+    sum.v = _mm_add_epi32(sum.v, ones.v);
+    if (SignExtended) {
+      msb_sum.v = _mm_add_epi32(msb_sum.v, _mm_srli_epi32(v.v, 31));
+    }
+
+    v.v = _mm_srli_epi32(v.v, 7);
+
+    for (int j = 0; j < 4; j++) {
+      __m128i min = _mm_min_epi32(v.v, ones.v);
+
+      sum.v = _mm_add_epi32(sum.v, min);
+      v.v   = _mm_srli_epi32(v.v, 7);
+    }
+  }
+
+  if (SignExtended) {
+    vus32 extensions;
+    extensions.v = _mm_mullo_epi32(msb_sum.v, fives.v);
+
+    sum.v = _mm_add_epi32(sum.v, extensions.v);
+  }
+
+  // TODO(ckennelly): Can we avoid the sign conversion?
+  size_t out = _mm_cvtsi128_si32(
+      _mm_hadd_epi32(_mm_hadd_epi32(sum.v, ones.v), ones.v));
+
+  // Finish tail.
+  for (; i < n; i++) {
+    if (ZigZag) {
+      out += WireFormatLite::SInt32Size(data[i]);
+    } else if (SignExtended) {
+      out += WireFormatLite::Int32Size(data[i]);
+    } else {
+      out += WireFormatLite::UInt32Size(data[i]);
+    }
+  }
+
+  return out;
+}
+
+size_t WireFormatLite::Int32Size(const RepeatedField<int32>& value) {
+  return VarintSize<int32, false, true>(value.data(), value.size());
+}
+
+size_t WireFormatLite::UInt32Size(const RepeatedField<uint32>& value) {
+  return VarintSize<uint32, false, false>(value.data(), value.size());
+}
+
+size_t WireFormatLite::SInt32Size(const RepeatedField<int32>& value) {
+  return VarintSize<int32, true, true>(value.data(), value.size());
+}
+
+size_t WireFormatLite::EnumSize(const RepeatedField<int>& value) {
+  // On ILP64, sizeof(int) == 8, which would require a different template.
+  return VarintSize<int, false, true>(value.data(), value.size());
+}
+
+#else  // !__SSE_4_1__
+size_t WireFormatLite::Int32Size(const RepeatedField<int32>& value) {
+  size_t out = 0;
+  const int n = value.size();
+  for (int i = 0; i < n; i++) {
+    out += Int32Size(value.Get(i));
+  }
+  return out;
+}
+
+size_t WireFormatLite::UInt32Size(const RepeatedField<uint32>& value) {
+  size_t out = 0;
+  const int n = value.size();
+  for (int i = 0; i < n; i++) {
+    out += UInt32Size(value.Get(i));
+  }
+  return out;
+}
+
+size_t WireFormatLite::SInt32Size(const RepeatedField<int32>& value) {
+  size_t out = 0;
+  const int n = value.size();
+  for (int i = 0; i < n; i++) {
+    out += SInt32Size(value.Get(i));
+  }
+  return out;
+}
+
+size_t WireFormatLite::EnumSize(const RepeatedField<int>& value) {
+  size_t out = 0;
+  const int n = value.size();
+  for (int i = 0; i < n; i++) {
+    out += EnumSize(value.Get(i));
+  }
+  return out;
+}
+#endif
+
 }  // namespace internal
 }  // namespace protobuf
 }  // namespace google