chrome/browser/metrics/metrics_log_serializer.cc

   1 // Copyright (c) 2012 The Chromium Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style license that can be
   3 // found in the LICENSE file.
   4
   5 #include "chrome/browser/metrics/metrics_log_serializer.h"
   6
   7 #include <string>
   8
   9 #include "base/base64.h"
  10 #include "base/md5.h"
  11 #include "base/metrics/histogram.h"
  12 #include "base/prefs/pref_service.h"
  13 #include "base/prefs/scoped_user_pref_update.h"
  14 #include "chrome/browser/browser_process.h"
  15 #include "chrome/common/pref_names.h"
  16
  17 namespace {
  18
  19 // The number of "initial" logs to save, and hope to send during a future Chrome
  20 // session.  Initial logs contain crash stats, and are pretty small.
  21 const size_t kInitialLogsPersistLimit = 20;
  22
  23 // The number of ongoing logs to save persistently, and hope to
  24 // send during a this or future sessions.  Note that each log may be pretty
  25 // large, as presumably the related "initial" log wasn't sent (probably nothing
  26 // was, as the user was probably off-line).  As a result, the log probably kept
  27 // accumulating while the "initial" log was stalled, and couldn't be sent.  As a
  28 // result, we don't want to save too many of these mega-logs.
  29 // A "standard shutdown" will create a small log, including just the data that
  30 // was not yet been transmitted, and that is normal (to have exactly one
  31 // ongoing_log_ at startup).
  32 const size_t kOngoingLogsPersistLimit = 8;
  33
  34 // The number of bytes each of initial and ongoing logs that must be stored.
  35 // This ensures that a reasonable amount of history will be stored even if there
  36 // is a long series of very small logs.
  37 const size_t kStorageByteLimitPerLogType = 300000;
  38
  39 // We append (2) more elements to persisted lists: the size of the list and a
  40 // checksum of the elements.
  41 const size_t kChecksumEntryCount = 2;
  42
  43 MetricsLogSerializer::LogReadStatus MakeRecallStatusHistogram(
  44     MetricsLogSerializer::LogReadStatus status) {
  45   UMA_HISTOGRAM_ENUMERATION("PrefService.PersistentLogRecallProtobufs",
  46                             status, MetricsLogSerializer::END_RECALL_STATUS);
  47   return status;
  48 }
  49
  50 }  // namespace
  51
  52
  53 MetricsLogSerializer::MetricsLogSerializer() {}
  54
  55 MetricsLogSerializer::~MetricsLogSerializer() {}
  56
  57 void MetricsLogSerializer::SerializeLogs(
  58     const std::vector<MetricsLogManager::SerializedLog>& logs,
  59     MetricsLogManager::LogType log_type) {
  60   PrefService* local_state = g_browser_process->local_state();
  61   DCHECK(local_state);
  62   const char* pref = NULL;
  63   size_t store_length_limit = 0;
  64   switch (log_type) {
  65     case MetricsLogBase::INITIAL_LOG:
  66       pref = prefs::kMetricsInitialLogs;
  67       store_length_limit = kInitialLogsPersistLimit;
  68       break;
  69     case MetricsLogBase::ONGOING_LOG:
  70       pref = prefs::kMetricsOngoingLogs;
  71       store_length_limit = kOngoingLogsPersistLimit;
  72       break;
  73     case MetricsLogBase::NO_LOG:
  74       NOTREACHED();
  75       return;
  76   };
  77
  78   ListPrefUpdate update(local_state, pref);
  79   WriteLogsToPrefList(logs, store_length_limit, kStorageByteLimitPerLogType,
  80                       update.Get());
  81 }
  82
  83 void MetricsLogSerializer::DeserializeLogs(
  84     MetricsLogManager::LogType log_type,
  85     std::vector<MetricsLogManager::SerializedLog>* logs) {
  86   DCHECK(logs);
  87   PrefService* local_state = g_browser_process->local_state();
  88   DCHECK(local_state);
  89
  90   const char* pref;
  91   if (log_type == MetricsLogBase::INITIAL_LOG)
  92     pref = prefs::kMetricsInitialLogs;
  93   else
  94     pref = prefs::kMetricsOngoingLogs;
  95
  96   const base::ListValue* unsent_logs = local_state->GetList(pref);
  97   ReadLogsFromPrefList(*unsent_logs, logs);
  98 }
  99
 100 // static
 101 void MetricsLogSerializer::WriteLogsToPrefList(
 102     const std::vector<MetricsLogManager::SerializedLog>& local_list,
 103     size_t list_length_limit,
 104     size_t byte_limit,
 105     base::ListValue* list) {
 106   // One of the limit arguments must be non-zero.
 107   DCHECK(list_length_limit > 0 || byte_limit > 0);
 108
 109   list->Clear();
 110   if (local_list.size() == 0)
 111     return;
 112
 113   size_t start = 0;
 114   // If there are too many logs, keep the most recent logs up to the length
 115   // limit, and at least to the minimum number of bytes.
 116   if (local_list.size() > list_length_limit) {
 117     start = local_list.size();
 118     size_t bytes_used = 0;
 119     for (std::vector<MetricsLogManager::SerializedLog>::const_reverse_iterator
 120          it = local_list.rbegin(); it != local_list.rend(); ++it) {
 121       size_t log_size = it->log_text().length();
 122       if (bytes_used >= byte_limit &&
 123           (local_list.size() - start) >= list_length_limit)
 124         break;
 125       bytes_used += log_size;
 126       --start;
 127     }
 128   }
 129   DCHECK_LT(start, local_list.size());
 130   if (start >= local_list.size())
 131     return;
 132
 133   // Store size at the beginning of the list.
 134   list->Append(base::Value::CreateIntegerValue(local_list.size() - start));
 135
 136   base::MD5Context ctx;
 137   base::MD5Init(&ctx);
 138   std::string encoded_log;
 139   for (std::vector<MetricsLogManager::SerializedLog>::const_iterator it =
 140            local_list.begin() + start;
 141        it != local_list.end(); ++it) {
 142     // We encode the compressed log as Value::CreateStringValue() expects to
 143     // take a valid UTF8 string.
 144     base::Base64Encode(it->log_text(), &encoded_log);
 145     base::MD5Update(&ctx, encoded_log);
 146     list->Append(base::Value::CreateStringValue(encoded_log));
 147   }
 148
 149   // Append hash to the end of the list.
 150   base::MD5Digest digest;
 151   base::MD5Final(&digest, &ctx);
 152   list->Append(base::Value::CreateStringValue(base::MD5DigestToBase16(digest)));
 153   DCHECK(list->GetSize() >= 3);  // Minimum of 3 elements (size, data, hash).
 154 }
 155
 156 // static
 157 MetricsLogSerializer::LogReadStatus MetricsLogSerializer::ReadLogsFromPrefList(
 158     const base::ListValue& list,
 159     std::vector<MetricsLogManager::SerializedLog>* local_list) {
 160   if (list.GetSize() == 0)
 161     return MakeRecallStatusHistogram(LIST_EMPTY);
 162   if (list.GetSize() < 3)
 163     return MakeRecallStatusHistogram(LIST_SIZE_TOO_SMALL);
 164
 165   // The size is stored at the beginning of the list.
 166   int size;
 167   bool valid = (*list.begin())->GetAsInteger(&size);
 168   if (!valid)
 169     return MakeRecallStatusHistogram(LIST_SIZE_MISSING);
 170   // Account for checksum and size included in the list.
 171   if (static_cast<unsigned int>(size) !=
 172       list.GetSize() - kChecksumEntryCount) {
 173     return MakeRecallStatusHistogram(LIST_SIZE_CORRUPTION);
 174   }
 175
 176   // Allocate strings for all of the logs we are going to read in.
 177   // Do this ahead of time so that we can decode the string values directly into
 178   // the elements of |local_list|, and thereby avoid making copies of the
 179   // serialized logs, which can be fairly large.
 180   DCHECK(local_list->empty());
 181   local_list->resize(size);
 182
 183   base::MD5Context ctx;
 184   base::MD5Init(&ctx);
 185   std::string encoded_log;
 186   size_t local_index = 0;
 187   for (base::ListValue::const_iterator it = list.begin() + 1;
 188        it != list.end() - 1;  // Last element is the checksum.
 189        ++it, ++local_index) {
 190     bool valid = (*it)->GetAsString(&encoded_log);
 191     if (!valid) {
 192       local_list->clear();
 193       return MakeRecallStatusHistogram(LOG_STRING_CORRUPTION);
 194     }
 195
 196     base::MD5Update(&ctx, encoded_log);
 197
 198     std::string log_text;
 199     if (!base::Base64Decode(encoded_log, &log_text)) {
 200       local_list->clear();
 201       return MakeRecallStatusHistogram(DECODE_FAIL);
 202     }
 203
 204     DCHECK_LT(local_index, local_list->size());
 205     (*local_list)[local_index].SwapLogText(&log_text);
 206   }
 207
 208   // Verify checksum.
 209   base::MD5Digest digest;
 210   base::MD5Final(&digest, &ctx);
 211   std::string recovered_md5;
 212   // We store the hash at the end of the list.
 213   valid = (*(list.end() - 1))->GetAsString(&recovered_md5);
 214   if (!valid) {
 215     local_list->clear();
 216     return MakeRecallStatusHistogram(CHECKSUM_STRING_CORRUPTION);
 217   }
 218   if (recovered_md5 != base::MD5DigestToBase16(digest)) {
 219     local_list->clear();
 220     return MakeRecallStatusHistogram(CHECKSUM_CORRUPTION);
 221   }
 222   return MakeRecallStatusHistogram(RECALL_SUCCESS);
 223 }