1 /* 2 * Copyright (C) 2017 The Android Open Source Project 3 * 4 * Licensed under the Apache License, Version 2.0 (the "License"); 5 * you may not use this file except in compliance with the License. 6 * You may obtain a copy of the License at 7 * 8 * http://www.apache.org/licenses/LICENSE-2.0 9 * 10 * Unless required by applicable law or agreed to in writing, software 11 * distributed under the License is distributed on an "AS IS" BASIS, 12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13 * See the License for the specific language governing permissions and 14 * limitations under the License. 15 */ 16 17 #ifndef INCLUDE_PERFETTO_PROTOZERO_PROTO_UTILS_H_ 18 #define INCLUDE_PERFETTO_PROTOZERO_PROTO_UTILS_H_ 19 20 #include <inttypes.h> 21 #include <stddef.h> 22 23 #include <type_traits> 24 25 namespace protozero { 26 namespace proto_utils { 27 28 // See https://developers.google.com/protocol-buffers/docs/encoding wire types. 29 30 enum FieldType : uint32_t { 31 kFieldTypeVarInt = 0, 32 kFieldTypeFixed64 = 1, 33 kFieldTypeLengthDelimited = 2, 34 kFieldTypeFixed32 = 5, 35 }; 36 37 // Maximum message size supported: 256 MiB (4 x 7-bit due to varint encoding). 38 constexpr size_t kMessageLengthFieldSize = 4; 39 constexpr size_t kMaxMessageLength = (1u << (kMessageLengthFieldSize * 7)) - 1; 40 41 // Field tag is encoded as 32-bit varint (5 bytes at most). 42 // Largest value of simple (not length-delimited) field is 64-bit varint 43 // (10 bytes at most). 15 bytes buffer is enough to store a simple field. 44 constexpr size_t kMaxTagEncodedSize = 5; 45 constexpr size_t kMaxSimpleFieldEncodedSize = kMaxTagEncodedSize + 10; 46 47 // Proto types: (int|uint|sint)(32|64), bool, enum. 48 constexpr uint32_t MakeTagVarInt(uint32_t field_id) { 49 return (field_id << 3) | kFieldTypeVarInt; 50 } 51 52 // Proto types: fixed64, sfixed64, fixed32, sfixed32, double, float. 53 template <typename T> 54 constexpr uint32_t MakeTagFixed(uint32_t field_id) { 55 static_assert(sizeof(T) == 8 || sizeof(T) == 4, "Value must be 4 or 8 bytes"); 56 return (field_id << 3) | 57 (sizeof(T) == 8 ? kFieldTypeFixed64 : kFieldTypeFixed32); 58 } 59 60 // Proto types: string, bytes, embedded messages. 61 constexpr uint32_t MakeTagLengthDelimited(uint32_t field_id) { 62 return (field_id << 3) | kFieldTypeLengthDelimited; 63 } 64 65 // Proto types: sint64, sint32. 66 template <typename T> 67 inline typename std::make_unsigned<T>::type ZigZagEncode(T value) { 68 return static_cast<typename std::make_unsigned<T>::type>( 69 (value << 1) ^ (value >> (sizeof(T) * 8 - 1))); 70 } 71 72 template <typename T> 73 inline uint8_t* WriteVarInt(T value, uint8_t* target) { 74 // Avoid arithmetic (sign expanding) shifts. 75 using UnsignedType = typename std::make_unsigned<T>::type; 76 UnsignedType unsigned_value = static_cast<UnsignedType>(value); 77 78 while (unsigned_value >= 0x80) { 79 *target++ = static_cast<uint8_t>(unsigned_value) | 0x80; 80 unsigned_value >>= 7; 81 } 82 *target = static_cast<uint8_t>(unsigned_value); 83 return target + 1; 84 } 85 86 // Writes a fixed-size redundant encoding of the given |value|. This is 87 // used to backfill fixed-size reservations for the length field using a 88 // non-canonical varint encoding (e.g. \x81\x80\x80\x00 instead of \x01). 89 // See https://github.com/google/protobuf/issues/1530. 90 // In particular, this is used for nested messages. The size of a nested message 91 // is not known until all its field have been written. |kMessageLengthFieldSize| 92 // bytes are reserved to encode the size field and backfilled at the end. 93 inline void WriteRedundantVarInt(uint32_t value, uint8_t* buf) { 94 for (size_t i = 0; i < kMessageLengthFieldSize; ++i) { 95 const uint8_t msb = (i < kMessageLengthFieldSize - 1) ? 0x80 : 0; 96 buf[i] = static_cast<uint8_t>(value) | msb; 97 value >>= 7; 98 } 99 } 100 101 template <uint32_t field_id> 102 void StaticAssertSingleBytePreamble() { 103 static_assert(field_id < 16, 104 "Proto field id too big to fit in a single byte preamble"); 105 }; 106 107 // Parses a VarInt from the encoded buffer [start, end). |end| is STL-style and 108 // points one byte past the end of buffer. 109 // The parsed int value is stored in the output arg |value|. Returns a pointer 110 // to the next unconsumed byte (so start < retval <= end). 111 const uint8_t* ParseVarInt(const uint8_t* start, 112 const uint8_t* end, 113 uint64_t* value); 114 115 // Parses a protobuf field and computes its id, type and value. 116 // Returns a pointer to the next unconsumed byte (|start| < retval <= end) that 117 // is either the beginning of the next field or the end of the parent message. 118 // In the case of a kFieldTypeLengthDelimited field, |field_intvalue| will 119 // store the length of the payload (either a string or a nested message). In 120 // this case, the start of the payload will be at (return value) - 121 // |field_intvalue|. 122 const uint8_t* ParseField(const uint8_t* start, 123 const uint8_t* end, 124 uint32_t* field_id, 125 FieldType* field_type, 126 uint64_t* field_intvalue); 127 128 } // namespace proto_utils 129 } // namespace protozero 130 131 #endif // INCLUDE_PERFETTO_PROTOZERO_PROTO_UTILS_H_ 132