naiveproxy/net/tools/dump_cache/dump_files.cc

601 lines
19 KiB
C++
Raw Normal View History

2018-02-02 13:49:39 +03:00
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
// Performs basic inspection of the disk cache files with minimal disruption
// to the actual files (they still may change if an error is detected on the
// files).
#include "net/tools/dump_cache/dump_files.h"
#include <stdio.h>
#include <memory>
#include <set>
#include <string>
#include "base/command_line.h"
#include "base/files/file.h"
#include "base/files/file_enumerator.h"
#include "base/files/file_util.h"
#include "base/format_macros.h"
#include "base/macros.h"
#include "base/message_loop/message_loop.h"
#include "base/strings/string_number_conversions.h"
#include "base/strings/stringprintf.h"
#include "net/disk_cache/blockfile/block_files.h"
#include "net/disk_cache/blockfile/disk_format.h"
#include "net/disk_cache/blockfile/mapped_file.h"
#include "net/disk_cache/blockfile/stats.h"
#include "net/disk_cache/blockfile/storage_block-inl.h"
#include "net/disk_cache/blockfile/storage_block.h"
#include "net/url_request/view_cache_helper.h"
namespace {
const base::FilePath::CharType kIndexName[] = FILE_PATH_LITERAL("index");
// Reads the |header_size| bytes from the beginning of file |name|.
bool ReadHeader(const base::FilePath& name, char* header, int header_size) {
base::File file(name, base::File::FLAG_OPEN | base::File::FLAG_READ);
if (!file.IsValid()) {
printf("Unable to open file %s\n", name.MaybeAsASCII().c_str());
return false;
}
int read = file.Read(0, header, header_size);
if (read != header_size) {
printf("Unable to read file %s\n", name.MaybeAsASCII().c_str());
return false;
}
return true;
}
int GetMajorVersionFromFile(const base::FilePath& name) {
disk_cache::IndexHeader header;
if (!ReadHeader(name, reinterpret_cast<char*>(&header), sizeof(header)))
return 0;
return header.version >> 16;
}
// Dumps the contents of the Stats record.
void DumpStats(const base::FilePath& path, disk_cache::CacheAddr addr) {
// We need a message loop, although we really don't run any task.
base::MessageLoopForIO loop;
disk_cache::BlockFiles block_files(path);
if (!block_files.Init(false)) {
printf("Unable to init block files\n");
return;
}
disk_cache::Addr address(addr);
disk_cache::MappedFile* file = block_files.GetFile(address);
if (!file)
return;
size_t length = (2 + disk_cache::Stats::kDataSizesLength) * sizeof(int32_t) +
disk_cache::Stats::MAX_COUNTER * sizeof(int64_t);
size_t offset = address.start_block() * address.BlockSize() +
disk_cache::kBlockHeaderSize;
std::unique_ptr<int32_t[]> buffer(new int32_t[length]);
if (!file->Read(buffer.get(), length, offset))
return;
printf("Stats:\nSignatrure: 0x%x\n", buffer[0]);
printf("Total size: %d\n", buffer[1]);
for (int i = 0; i < disk_cache::Stats::kDataSizesLength; i++)
printf("Size(%d): %d\n", i, buffer[i + 2]);
int64_t* counters = reinterpret_cast<int64_t*>(
buffer.get() + 2 + disk_cache::Stats::kDataSizesLength);
for (int i = 0; i < disk_cache::Stats::MAX_COUNTER; i++)
printf("Count(%d): %" PRId64 "\n", i, *counters++);
printf("-------------------------\n\n");
}
// Dumps the contents of the Index-file header.
void DumpIndexHeader(const base::FilePath& name,
disk_cache::CacheAddr* stats_addr) {
disk_cache::IndexHeader header;
if (!ReadHeader(name, reinterpret_cast<char*>(&header), sizeof(header)))
return;
printf("Index file:\n");
printf("magic: %x\n", header.magic);
printf("version: %d.%d\n", header.version >> 16, header.version & 0xffff);
printf("entries: %d\n", header.num_entries);
printf("total bytes: %d\n", header.num_bytes);
printf("last file number: %d\n", header.last_file);
printf("current id: %d\n", header.this_id);
printf("table length: %d\n", header.table_len);
printf("last crash: %d\n", header.crash);
printf("experiment: %d\n", header.experiment);
printf("stats: %x\n", header.stats);
for (int i = 0; i < 5; i++) {
printf("head %d: 0x%x\n", i, header.lru.heads[i]);
printf("tail %d: 0x%x\n", i, header.lru.tails[i]);
printf("size %d: 0x%x\n", i, header.lru.sizes[i]);
}
printf("transaction: 0x%x\n", header.lru.transaction);
printf("operation: %d\n", header.lru.operation);
printf("operation list: %d\n", header.lru.operation_list);
printf("-------------------------\n\n");
if (stats_addr)
*stats_addr = header.stats;
}
// Dumps the contents of a block-file header.
void DumpBlockHeader(const base::FilePath& name) {
disk_cache::BlockFileHeader header;
if (!ReadHeader(name, reinterpret_cast<char*>(&header), sizeof(header)))
return;
printf("Block file: %s\n", name.BaseName().MaybeAsASCII().c_str());
printf("magic: %x\n", header.magic);
printf("version: %d.%d\n", header.version >> 16, header.version & 0xffff);
printf("file id: %d\n", header.this_file);
printf("next file id: %d\n", header.next_file);
printf("entry size: %d\n", header.entry_size);
printf("current entries: %d\n", header.num_entries);
printf("max entries: %d\n", header.max_entries);
printf("updating: %d\n", header.updating);
printf("empty sz 1: %d\n", header.empty[0]);
printf("empty sz 2: %d\n", header.empty[1]);
printf("empty sz 3: %d\n", header.empty[2]);
printf("empty sz 4: %d\n", header.empty[3]);
printf("user 0: 0x%x\n", header.user[0]);
printf("user 1: 0x%x\n", header.user[1]);
printf("user 2: 0x%x\n", header.user[2]);
printf("user 3: 0x%x\n", header.user[3]);
printf("-------------------------\n\n");
}
// Simple class that interacts with the set of cache files.
class CacheDumper {
public:
explicit CacheDumper(const base::FilePath& path)
: path_(path),
block_files_(path),
index_(NULL),
current_hash_(0),
next_addr_(0) {
}
bool Init();
// Reads an entry from disk. Return false when all entries have been already
// returned.
bool GetEntry(disk_cache::EntryStore* entry, disk_cache::CacheAddr* addr);
// Loads a specific block from the block files.
bool LoadEntry(disk_cache::CacheAddr addr, disk_cache::EntryStore* entry);
bool LoadRankings(disk_cache::CacheAddr addr,
disk_cache::RankingsNode* rankings);
// Appends the data store at |addr| to |out|.
bool HexDump(disk_cache::CacheAddr addr, std::string* out);
private:
base::FilePath path_;
disk_cache::BlockFiles block_files_;
scoped_refptr<disk_cache::MappedFile> index_file_;
disk_cache::Index* index_;
int current_hash_;
disk_cache::CacheAddr next_addr_;
std::set<disk_cache::CacheAddr> dumped_entries_;
DISALLOW_COPY_AND_ASSIGN(CacheDumper);
};
bool CacheDumper::Init() {
if (!block_files_.Init(false)) {
printf("Unable to init block files\n");
return false;
}
base::FilePath index_name(path_.Append(kIndexName));
index_file_ = new disk_cache::MappedFile;
index_ = reinterpret_cast<disk_cache::Index*>(
index_file_->Init(index_name, 0));
if (!index_) {
printf("Unable to map index\n");
return false;
}
return true;
}
bool CacheDumper::GetEntry(disk_cache::EntryStore* entry,
disk_cache::CacheAddr* addr) {
if (dumped_entries_.find(next_addr_) != dumped_entries_.end()) {
printf("Loop detected\n");
next_addr_ = 0;
current_hash_++;
}
if (next_addr_) {
*addr = next_addr_;
if (LoadEntry(next_addr_, entry))
return true;
printf("Unable to load entry at address 0x%x\n", next_addr_);
next_addr_ = 0;
current_hash_++;
}
for (int i = current_hash_; i < index_->header.table_len; i++) {
// Yes, we'll crash if the table is shorter than expected, but only after
// dumping every entry that we can find.
if (index_->table[i]) {
current_hash_ = i;
*addr = index_->table[i];
if (LoadEntry(index_->table[i], entry))
return true;
printf("Unable to load entry at address 0x%x\n", index_->table[i]);
}
}
return false;
}
bool CacheDumper::LoadEntry(disk_cache::CacheAddr addr,
disk_cache::EntryStore* entry) {
disk_cache::Addr address(addr);
disk_cache::MappedFile* file = block_files_.GetFile(address);
if (!file)
return false;
disk_cache::StorageBlock<disk_cache::EntryStore> entry_block(file, address);
if (!entry_block.Load())
return false;
memcpy(entry, entry_block.Data(), sizeof(*entry));
if (!entry_block.VerifyHash())
printf("Self hash failed at 0x%x\n", addr);
// Prepare for the next entry to load.
next_addr_ = entry->next;
if (next_addr_) {
dumped_entries_.insert(addr);
} else {
current_hash_++;
dumped_entries_.clear();
}
return true;
}
bool CacheDumper::LoadRankings(disk_cache::CacheAddr addr,
disk_cache::RankingsNode* rankings) {
disk_cache::Addr address(addr);
if (address.file_type() != disk_cache::RANKINGS)
return false;
disk_cache::MappedFile* file = block_files_.GetFile(address);
if (!file)
return false;
disk_cache::StorageBlock<disk_cache::RankingsNode> rank_block(file, address);
if (!rank_block.Load())
return false;
if (!rank_block.VerifyHash())
printf("Self hash failed at 0x%x\n", addr);
memcpy(rankings, rank_block.Data(), sizeof(*rankings));
return true;
}
bool CacheDumper::HexDump(disk_cache::CacheAddr addr, std::string* out) {
disk_cache::Addr address(addr);
disk_cache::MappedFile* file = block_files_.GetFile(address);
if (!file)
return false;
size_t size = address.num_blocks() * address.BlockSize();
std::unique_ptr<char> buffer(new char[size]);
size_t offset = address.start_block() * address.BlockSize() +
disk_cache::kBlockHeaderSize;
if (!file->Read(buffer.get(), size, offset))
return false;
base::StringAppendF(out, "0x%x:\n", addr);
net::ViewCacheHelper::HexDump(buffer.get(), size, out);
return true;
}
std::string ToLocalTime(int64_t time_us) {
base::Time time = base::Time::FromInternalValue(time_us);
base::Time::Exploded e;
time.LocalExplode(&e);
return base::StringPrintf("%d/%d/%d %d:%d:%d.%d", e.year, e.month,
e.day_of_month, e.hour, e.minute, e.second,
e.millisecond);
}
void DumpEntry(disk_cache::CacheAddr addr,
const disk_cache::EntryStore& entry,
bool verbose) {
std::string key;
static bool full_key =
base::CommandLine::ForCurrentProcess()->HasSwitch("full-key");
if (!entry.long_key) {
key = std::string(entry.key, std::min(static_cast<size_t>(entry.key_len),
sizeof(entry.key)));
if (entry.key_len > 90 && !full_key)
key.resize(90);
}
printf("Entry at 0x%x\n", addr);
printf("rankings: 0x%x\n", entry.rankings_node);
printf("key length: %d\n", entry.key_len);
printf("key: \"%s\"\n", key.c_str());
if (verbose) {
printf("key addr: 0x%x\n", entry.long_key);
printf("hash: 0x%x\n", entry.hash);
printf("next entry: 0x%x\n", entry.next);
printf("reuse count: %d\n", entry.reuse_count);
printf("refetch count: %d\n", entry.refetch_count);
printf("state: %d\n", entry.state);
printf("creation: %s\n", ToLocalTime(entry.creation_time).c_str());
for (int i = 0; i < 4; i++) {
printf("data size %d: %d\n", i, entry.data_size[i]);
printf("data addr %d: 0x%x\n", i, entry.data_addr[i]);
}
printf("----------\n\n");
}
}
void DumpRankings(disk_cache::CacheAddr addr,
const disk_cache::RankingsNode& rankings,
bool verbose) {
printf("Rankings at 0x%x\n", addr);
printf("next: 0x%x\n", rankings.next);
printf("prev: 0x%x\n", rankings.prev);
printf("entry: 0x%x\n", rankings.contents);
if (verbose) {
printf("dirty: %d\n", rankings.dirty);
if (rankings.last_used != rankings.last_modified)
printf("used: %s\n", ToLocalTime(rankings.last_used).c_str());
printf("modified: %s\n", ToLocalTime(rankings.last_modified).c_str());
printf("hash: 0x%x\n", rankings.self_hash);
printf("----------\n\n");
} else {
printf("\n");
}
}
void PrintCSVHeader() {
printf(
"entry,rankings,next,prev,rank-contents,chain,reuse,key,"
"d0,d1,d2,d3\n");
}
void DumpCSV(disk_cache::CacheAddr addr,
const disk_cache::EntryStore& entry,
const disk_cache::RankingsNode& rankings) {
printf("0x%x,0x%x,0x%x,0x%x,0x%x,0x%x,0x%x,0x%x,0x%x,0x%x,0x%x,0x%x\n", addr,
entry.rankings_node, rankings.next, rankings.prev, rankings.contents,
entry.next, entry.reuse_count, entry.long_key, entry.data_addr[0],
entry.data_addr[1], entry.data_addr[2], entry.data_addr[3]);
if (addr != rankings.contents)
printf("Broken entry\n");
}
bool CanDump(disk_cache::CacheAddr addr) {
disk_cache::Addr address(addr);
return address.is_initialized() && address.is_block_file();
}
} // namespace.
// -----------------------------------------------------------------------
int GetMajorVersion(const base::FilePath& input_path) {
base::FilePath index_name(input_path.Append(kIndexName));
int version = GetMajorVersionFromFile(index_name);
if (!version)
return 0;
base::FilePath data_name(input_path.Append(FILE_PATH_LITERAL("data_0")));
if (version != GetMajorVersionFromFile(data_name))
return 0;
data_name = input_path.Append(FILE_PATH_LITERAL("data_1"));
if (version != GetMajorVersionFromFile(data_name))
return 0;
data_name = input_path.Append(FILE_PATH_LITERAL("data_2"));
if (version != GetMajorVersionFromFile(data_name))
return 0;
data_name = input_path.Append(FILE_PATH_LITERAL("data_3"));
if (version != GetMajorVersionFromFile(data_name))
return 0;
return version;
}
// Dumps the headers of all files.
int DumpHeaders(const base::FilePath& input_path) {
base::FilePath index_name(input_path.Append(kIndexName));
disk_cache::CacheAddr stats_addr = 0;
DumpIndexHeader(index_name, &stats_addr);
base::FileEnumerator iter(input_path, false,
base::FileEnumerator::FILES,
FILE_PATH_LITERAL("data_*"));
for (base::FilePath file = iter.Next(); !file.empty(); file = iter.Next())
DumpBlockHeader(file);
DumpStats(input_path, stats_addr);
return 0;
}
// Dumps all entries from the cache.
int DumpContents(const base::FilePath& input_path) {
bool print_csv = base::CommandLine::ForCurrentProcess()->HasSwitch("csv");
if (!print_csv)
DumpIndexHeader(input_path.Append(kIndexName), nullptr);
// We need a message loop, although we really don't run any task.
base::MessageLoopForIO loop;
CacheDumper dumper(input_path);
if (!dumper.Init())
return -1;
if (print_csv)
PrintCSVHeader();
disk_cache::EntryStore entry;
disk_cache::CacheAddr addr;
bool verbose = base::CommandLine::ForCurrentProcess()->HasSwitch("v");
while (dumper.GetEntry(&entry, &addr)) {
if (!print_csv)
DumpEntry(addr, entry, verbose);
disk_cache::RankingsNode rankings;
if (!dumper.LoadRankings(entry.rankings_node, &rankings))
continue;
if (print_csv)
DumpCSV(addr, entry, rankings);
else
DumpRankings(entry.rankings_node, rankings, verbose);
}
printf("Done.\n");
return 0;
}
int DumpLists(const base::FilePath& input_path) {
base::FilePath index_name(input_path.Append(kIndexName));
disk_cache::IndexHeader header;
if (!ReadHeader(index_name, reinterpret_cast<char*>(&header), sizeof(header)))
return -1;
// We need a message loop, although we really don't run any task.
base::MessageLoopForIO loop;
CacheDumper dumper(input_path);
if (!dumper.Init())
return -1;
printf("list, addr, next, prev, entry\n");
const int kMaxLength = 1 * 1000 * 1000;
for (int i = 0; i < 5; i++) {
int32_t size = header.lru.sizes[i];
if (size < 0 || size > kMaxLength) {
printf("Wrong size %d\n", size);
size = kMaxLength;
}
disk_cache::CacheAddr addr = header.lru.tails[i];
int count = 0;
for (; size && addr; size--) {
count++;
disk_cache::RankingsNode rankings;
if (!dumper.LoadRankings(addr, &rankings)) {
printf("Failed to load node at 0x%x\n", addr);
break;
}
printf("%d, 0x%x, 0x%x, 0x%x, 0x%x\n", i, addr, rankings.next,
rankings.prev, rankings.contents);
if (rankings.prev == addr)
break;
addr = rankings.prev;
}
printf("%d nodes found, %d reported\n", count, header.lru.sizes[i]);
}
printf("Done.\n");
return 0;
}
int DumpEntryAt(const base::FilePath& input_path, const std::string& at) {
disk_cache::CacheAddr addr;
if (!base::HexStringToUInt(at, &addr))
return -1;
if (!CanDump(addr))
return -1;
base::FilePath index_name(input_path.Append(kIndexName));
disk_cache::IndexHeader header;
if (!ReadHeader(index_name, reinterpret_cast<char*>(&header), sizeof(header)))
return -1;
// We need a message loop, although we really don't run any task.
base::MessageLoopForIO loop;
CacheDumper dumper(input_path);
if (!dumper.Init())
return -1;
disk_cache::CacheAddr entry_addr = 0;
disk_cache::CacheAddr rankings_addr = 0;
disk_cache::Addr address(addr);
disk_cache::RankingsNode rankings;
if (address.file_type() == disk_cache::RANKINGS) {
if (dumper.LoadRankings(addr, &rankings)) {
rankings_addr = addr;
addr = rankings.contents;
address = disk_cache::Addr(addr);
}
}
disk_cache::EntryStore entry = {};
if (address.file_type() == disk_cache::BLOCK_256 &&
dumper.LoadEntry(addr, &entry)) {
entry_addr = addr;
DumpEntry(addr, entry, true);
if (!rankings_addr && dumper.LoadRankings(entry.rankings_node, &rankings))
rankings_addr = entry.rankings_node;
}
bool verbose = base::CommandLine::ForCurrentProcess()->HasSwitch("v");
std::string hex_dump;
if (!rankings_addr || verbose)
dumper.HexDump(addr, &hex_dump);
if (rankings_addr)
DumpRankings(rankings_addr, rankings, true);
if (entry_addr && verbose) {
if (entry.long_key && CanDump(entry.long_key))
dumper.HexDump(entry.long_key, &hex_dump);
for (int i = 0; i < 4; i++) {
if (entry.data_addr[i] && CanDump(entry.data_addr[i]))
dumper.HexDump(entry.data_addr[i], &hex_dump);
}
}
printf("%s\n", hex_dump.c_str());
printf("Done.\n");
return 0;
}
int DumpAllocation(const base::FilePath& file) {
disk_cache::BlockFileHeader header;
if (!ReadHeader(file, reinterpret_cast<char*>(&header), sizeof(header)))
return -1;
std::string out;
net::ViewCacheHelper::HexDump(reinterpret_cast<char*>(&header.allocation_map),
sizeof(header.allocation_map), &out);
printf("%s\n", out.c_str());
return 0;
}