X-Git-Url: http://git.efficios.com/?a=blobdiff_plain;f=gold%2Fstringpool.h;h=b7ac54f5b4e4bcbc3add1d285daa966e4ea1e1a6;hb=8b9a915daf91027050ad7145853ce8874034e60b;hp=bc0149840696a8d08e76aa546fd3e75606cf26a8;hpb=614f30a283ede08a5d2913a0597581edeb7c0905;p=deliverable%2Fbinutils-gdb.git diff --git a/gold/stringpool.h b/gold/stringpool.h index bc01498406..b7ac54f5b4 100644 --- a/gold/stringpool.h +++ b/gold/stringpool.h @@ -1,10 +1,28 @@ // stringpool.h -- a string pool for gold -*- C++ -*- +// Copyright (C) 2006-2016 Free Software Foundation, Inc. +// Written by Ian Lance Taylor . + +// This file is part of gold. + +// This program is free software; you can redistribute it and/or modify +// it under the terms of the GNU General Public License as published by +// the Free Software Foundation; either version 3 of the License, or +// (at your option) any later version. + +// This program is distributed in the hope that it will be useful, +// but WITHOUT ANY WARRANTY; without even the implied warranty of +// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +// GNU General Public License for more details. + +// You should have received a copy of the GNU General Public License +// along with this program; if not, write to the Free Software +// Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, +// MA 02110-1301, USA. + #include #include - -// Stringpool -// Manage a pool of unique strings. +#include #ifndef GOLD_STRINGPOOL_H #define GOLD_STRINGPOOL_H @@ -14,76 +32,275 @@ namespace gold class Output_file; +// Return the length of a string in units of Char_type. + +template +inline size_t +string_length(const Char_type* p) +{ + size_t len = 0; + for (; *p != 0; ++p) + ++len; + return len; +} + +// Specialize string_length for char. Maybe we could just use +// std::char_traits<>::length? + +template<> +inline size_t +string_length(const char* p) +{ + return strlen(p); +} + +// A Stringpool is a pool of unique strings. It provides the +// following features: + +// Every string in the pool is unique. Thus, if you have two strings +// in the Stringpool, you can compare them for equality by using +// pointer comparison rather than string comparison. + +// There is a key associated with every string in the pool. If you +// add strings to the Stringpool in the same order, then the key for +// each string will always be the same for any run of the linker. +// This is not true of the string pointers themselves, as they may +// change due to address space randomization. Some parts of the +// linker (e.g., the symbol table) use the key value instead of the +// string pointer so that repeated runs of the linker will generate +// precisely the same output. + +// When you add a string to a Stringpool, Stringpool will optionally +// make a copy of it. Thus there is no requirement to keep a copy +// elsewhere. + +// A Stringpool can be turned into a string table, a sequential series +// of null terminated strings. The first string may optionally be a +// single zero byte, as required for SHT_STRTAB sections. This +// conversion is only permitted after all strings have been added to +// the Stringpool. After doing this conversion, you can ask for the +// offset of any string (or any key) in the stringpool in the string +// table, and you can write the resulting string table to an output +// file. + +// When a Stringpool is turned into a string table, then as an +// optimization it will reuse string suffixes to avoid duplicating +// strings. That is, given the strings "abc" and "bc", only the +// string "abc" will be stored, and "bc" will be represented by an +// offset into the middle of the string "abc". + + +// A simple chunked vector class--this is a subset of std::vector +// which stores memory in chunks. We don't provide iterators, because +// we don't need them. + +template +class Chunked_vector +{ + public: + Chunked_vector() + : chunks_(), size_(0) + { } + + // Clear the elements. + void + clear() + { + this->chunks_.clear(); + this->size_ = 0; + } + + // Reserve elements. + void + reserve(unsigned int n) + { + if (n > this->chunks_.size() * chunk_size) + { + this->chunks_.resize((n + chunk_size - 1) / chunk_size); + // We need to call reserve() of all chunks since changing + // this->chunks_ casues Element_vectors to be copied. The + // reserved capacity of an Element_vector may be lost in copying. + for (size_t i = 0; i < this->chunks_.size(); ++i) + this->chunks_[i].reserve(chunk_size); + } + } + + // Get the number of elements. + size_t + size() const + { return this->size_; } + + // Push a new element on the back of the vector. + void + push_back(const Element& element) + { + size_t chunk_index = this->size_ / chunk_size; + if (chunk_index >= this->chunks_.size()) + { + this->chunks_.push_back(Element_vector()); + this->chunks_.back().reserve(chunk_size); + gold_assert(chunk_index < this->chunks_.size()); + } + this->chunks_[chunk_index].push_back(element); + this->size_++; + } + + // Return a reference to an entry in the vector. + Element& + operator[](size_t i) + { return this->chunks_[i / chunk_size][i % chunk_size]; } + + const Element& + operator[](size_t i) const + { return this->chunks_[i / chunk_size][i % chunk_size]; } + + private: + static const unsigned int chunk_size = 8192; + + typedef std::vector Element_vector; + typedef std::vector Chunk_vector; + + Chunk_vector chunks_; + size_t size_; +}; + + +// Stringpools are implemented in terms of Stringpool_template, which +// is generalized on the type of character used for the strings. Most +// uses will want the Stringpool type which uses char. Other cases +// are used for merging wide string constants. + template class Stringpool_template { public: - // The type of a key into the stringpool. A key value will always - // be the same during any run of the linker. The string pointers - // may change when using address space randomization. We use key - // values in order to get repeatable runs when the value is inserted - // into an unordered hash table. Zero is never a valid key. + // The type of a key into the stringpool. As described above, a key + // value will always be the same during any run of the linker. Zero + // is never a valid key value. typedef size_t Key; - // Create a Stringpool. ZERO_NULL is true if we should reserve - // offset 0 to hold the empty string. - Stringpool_template(bool zero_null = true); + // Create a Stringpool. + Stringpool_template(uint64_t addralign = 1); ~Stringpool_template(); - // Add a string to the pool. This returns a canonical permanent - // pointer to the string. If PKEY is not NULL, this sets *PKEY to - // the key for the string. + // Clear all the data from the stringpool. + void + clear(); + + // Hint to the stringpool class that you intend to insert n additional + // elements. The stringpool class can use this info however it likes; + // in practice it will resize its internal hashtables to make room. + void + reserve(unsigned int n); + + // Indicate that we should not reserve offset 0 to hold the empty + // string when converting the stringpool to a string table. This + // should not be called for a proper ELF SHT_STRTAB section. + void + set_no_zero_null() + { + gold_assert(this->string_set_.empty() + && this->offset_ == sizeof(Stringpool_char)); + this->zero_null_ = false; + this->offset_ = 0; + } + + // Indicate that this string pool should be optimized, even if not + // running with -O2. + void + set_optimize() + { this->optimize_ = true; } + + // Add the string S to the pool. This returns a canonical permanent + // pointer to the string in the pool. If COPY is true, the string + // is copied into permanent storage. If PKEY is not NULL, this sets + // *PKEY to the key for the string. const Stringpool_char* - add(const Stringpool_char*, Key* pkey); + add(const Stringpool_char* s, bool copy, Key* pkey); + // Add the string S to the pool. const Stringpool_char* - add(const std::basic_string& s, Key* pkey) - { return this->add(s.c_str(), pkey); } + add(const std::basic_string& s, bool copy, Key* pkey) + { return this->add_with_length(s.data(), s.size(), copy, pkey); } - // Add the prefix of a string to the pool. + // Add string S of length LEN characters to the pool. If COPY is + // true, S need not be null terminated. const Stringpool_char* - add(const Stringpool_char*, size_t, Key* pkey); + add_with_length(const Stringpool_char* s, size_t len, bool copy, Key* pkey); - // If a string is present, return the canonical string. Otherwise, - // return NULL. If PKEY is not NULL, set *PKEY to the key. + // If the string S is present in the pool, return the canonical + // string pointer. Otherwise, return NULL. If PKEY is not NULL, + // set *PKEY to the key. const Stringpool_char* - find(const Stringpool_char*, Key* pkey) const; + find(const Stringpool_char* s, Key* pkey) const; - // Turn the stringpool into an ELF strtab: determine the offsets of - // all the strings. + // Turn the stringpool into a string table: determine the offsets of + // all the strings. After this is called, no more strings may be + // added to the stringpool. void set_string_offsets(); - // Get the offset of a string in an ELF strtab. This returns the - // offset in bytes, not characters. - off_t - get_offset(const Stringpool_char*) const; + // Get the offset of the string S in the string table. This returns + // the offset in bytes, not in units of Stringpool_char. This may + // only be called after set_string_offsets has been called. + section_offset_type + get_offset(const Stringpool_char* s) const; - off_t + // Get the offset of the string S in the string table. + section_offset_type get_offset(const std::basic_string& s) const - { return this->get_offset(s.c_str()); } + { return this->get_offset_with_length(s.c_str(), s.size()); } + + // Get the offset of string S, with length LENGTH characters, in the + // string table. + section_offset_type + get_offset_with_length(const Stringpool_char* s, size_t length) const; - // Get the size of the ELF strtab. This returns the number of - // bytes, not characters. - off_t + // Get the offset of the string with key K. + section_offset_type + get_offset_from_key(Key k) const + { + gold_assert(k <= this->key_to_offset_.size()); + return this->key_to_offset_[k - 1]; + } + + // Get the size of the string table. This returns the number of + // bytes, not in units of Stringpool_char. + section_size_type get_strtab_size() const { gold_assert(this->strtab_size_ != 0); return this->strtab_size_; } - // Write the strtab into the output file at the specified offset. + // Write the string table into the output file at the specified + // offset. void write(Output_file*, off_t offset); + // Write the string table into the specified buffer, of the + // specified size. buffer_size should be at least + // get_strtab_size(). + void + write_to_buffer(unsigned char* buffer, section_size_type buffer_size); + + // Dump statistical information to stderr. + void + print_stats(const char*) const; + private: Stringpool_template(const Stringpool_template&); Stringpool_template& operator=(const Stringpool_template&); - // Return the length of a string. + // Return whether two strings are equal. + static bool + string_equal(const Stringpool_char*, const Stringpool_char*); + + // Compute a hash code for a string. LENGTH is the length of the + // string in characters. static size_t - string_length(const Stringpool_char*); + string_hash(const Stringpool_char*, size_t length); // We store the actual data in a list of these buffers. struct Stringdata @@ -92,61 +309,78 @@ class Stringpool_template size_t len; // Allocated size of buffer. size_t alc; - // Buffer index. - unsigned int index; // Buffer. char data[1]; }; + // Add a new key offset entry. + void + new_key_offset(size_t); + // Copy a string into the buffers, returning a canonical string. const Stringpool_char* - add_string(const Stringpool_char*, Key*); + add_string(const Stringpool_char*, size_t); + // Return whether s1 is a suffix of s2. + static bool + is_suffix(const Stringpool_char* s1, size_t len1, + const Stringpool_char* s2, size_t len2); + + // The hash table key includes the string, the length of the string, + // and the hash code for the string. We put the hash code + // explicitly into the key so that we can do a find()/insert() + // sequence without having to recompute the hash. Computing the + // hash code is a significant user of CPU time in the linker. + struct Hashkey + { + const Stringpool_char* string; + // Length is in characters, not bytes. + size_t length; + size_t hash_code; + + // This goes in an STL container, so we need a default + // constructor. + Hashkey() + : string(NULL), length(0), hash_code(0) + { } + + // Note that these constructors are relatively expensive, because + // they compute the hash code. + explicit Hashkey(const Stringpool_char* s) + : string(s), length(string_length(s)), hash_code(string_hash(s, length)) + { } + + Hashkey(const Stringpool_char* s, size_t len) + : string(s), length(len), hash_code(string_hash(s, len)) + { } + }; + + // Hash function. This is trivial, since we have already computed + // the hash. struct Stringpool_hash { size_t - operator()(const Stringpool_char*) const; + operator()(const Hashkey& hk) const + { return hk.hash_code; } }; + // Equality comparison function for hash table. struct Stringpool_eq { bool - operator()(const Stringpool_char* p1, const Stringpool_char* p2) const; + operator()(const Hashkey&, const Hashkey&) const; }; - // Return whether s1 is a suffix of s2. - static bool - is_suffix(const Stringpool_char* s1, size_t len1, - const Stringpool_char* s2, size_t len2); + // The hash table is a map from strings to Keys. - // The hash table is a map from string names to a pair of Key and - // ELF strtab offsets. We only use the offsets if we turn this into - // an ELF strtab section. + typedef Key Hashval; - typedef std::pair Val; - -#ifdef HAVE_TR1_UNORDERED_SET - typedef Unordered_map >, - true> String_set_type; -#else - typedef Unordered_map String_set_type; -#endif - // Comparison routine used when sorting into an ELF strtab. We - // store string-sizes in the sort-vector so we don't have to - // recompute them log(n) times as we sort. - struct Stringpool_sort_info - { - typename String_set_type::iterator it; - size_t string_length; - Stringpool_sort_info(typename String_set_type::iterator i, size_t s) - : it(i), string_length(s) - { } - }; + // Comparison routine used when sorting into a string table. + + typedef typename String_set_type::iterator Stringpool_sort_info; struct Stringpool_sort_comparison { @@ -154,19 +388,29 @@ class Stringpool_template operator()(const Stringpool_sort_info&, const Stringpool_sort_info&) const; }; + // Keys map to offsets via a Chunked_vector. We only use the + // offsets if we turn this into an string table section. + typedef Chunked_vector Key_to_offset; + // List of Stringdata structures. typedef std::list Stringdata_list; // Mapping from const char* to namepool entry. String_set_type string_set_; + // Mapping from Key to string table offset. + Key_to_offset key_to_offset_; // List of buffers. Stringdata_list strings_; - // Size of ELF strtab. - off_t strtab_size_; - // Next Stringdata index. - unsigned int next_index_; + // Size of string table. + section_size_type strtab_size_; // Whether to reserve offset 0 to hold the null string. bool zero_null_; + // Whether to optimize the string table. + bool optimize_; + // offset of the next string. + section_offset_type offset_; + // The alignment of strings in the stringpool. + uint64_t addralign_; }; // The most common type of Stringpool.