| // Copyright 2012 The Chromium Authors |
| // Use of this source code is governed by a BSD-style license that can be |
| // found in the LICENSE file. |
| |
| #ifdef UNSAFE_BUFFERS_BUILD |
| // TODO(crbug.com/390223051): Remove C-library calls to fix the errors. |
| #pragma allow_unsafe_libc_calls |
| #endif |
| |
| // Windows Timer Primer |
| // |
| // A good article: http://www.ddj.com/windows/184416651 |
| // A good mozilla bug: http://bugzilla.mozilla.org/show_bug.cgi?id=363258 |
| // |
| // The default windows timer, GetSystemTimePreciseAsFileTime is quite precise. |
| // However it is not always fast on some hardware and is slower than the |
| // performance counters. |
| // |
| // QueryPerformanceCounter is the logical choice for a high-precision timer. |
| // However, it is known to be buggy on some hardware. Specifically, it can |
| // sometimes "jump". On laptops, QPC can also be very expensive to call. |
| // It's 3-4x slower than timeGetTime() on desktops, but can be 10x slower |
| // on laptops. A unittest exists which will show the relative cost of various |
| // timers on any system. |
| // |
| // The next logical choice is timeGetTime(). timeGetTime has a precision of |
| // 1ms, but only if you call APIs (timeBeginPeriod()) which affect all other |
| // applications on the system. By default, precision is only 15.5ms. |
| // Unfortunately, we don't want to call timeBeginPeriod because we don't |
| // want to affect other applications. Further, on mobile platforms, use of |
| // faster multimedia timers can hurt battery life. See the intel |
| // article about this here: |
| // http://softwarecommunity.intel.com/articles/eng/1086.htm |
| // |
| // To work around all this, we're going to generally use timeGetTime(). We |
| // will only increase the system-wide timer if we're not running on battery |
| // power. |
| |
| #include "base/time/time.h" |
| |
| #include <windows.h> |
| |
| #include <mmsystem.h> |
| #include <stdint.h> |
| #include <windows.foundation.h> |
| |
| #include <atomic> |
| #include <ostream> |
| |
| #include "base/base_switches.h" |
| #include "base/bit_cast.h" |
| #include "base/check_op.h" |
| #include "base/command_line.h" |
| #include "base/cpu.h" |
| #include "base/notreached.h" |
| #include "base/rand_util.h" |
| #include "base/synchronization/lock.h" |
| #include "base/threading/platform_thread.h" |
| #include "base/time/time_override.h" |
| #include "build/build_config.h" |
| |
| namespace base { |
| |
| namespace { |
| |
| // From MSDN, FILETIME "Contains a 64-bit value representing the number of |
| // 100-nanosecond intervals since January 1, 1601 (UTC)." |
| int64_t FileTimeToMicroseconds(const FILETIME& ft) { |
| // Need to bit_cast to fix alignment, then divide by 10 to convert |
| // 100-nanoseconds to microseconds. This only works on little-endian |
| // machines. |
| return bit_cast<int64_t, FILETIME>(ft) / 10; |
| } |
| |
| bool CanConvertToFileTime(int64_t us) { |
| return us >= 0 && us <= (std::numeric_limits<int64_t>::max() / 10); |
| } |
| |
| FILETIME MicrosecondsToFileTime(int64_t us) { |
| DCHECK(CanConvertToFileTime(us)) << "Out-of-range: Cannot convert " << us |
| << " microseconds to FILETIME units."; |
| |
| // Multiply by 10 to convert microseconds to 100-nanoseconds. Bit_cast will |
| // handle alignment problems. This only works on little-endian machines. |
| return bit_cast<FILETIME, int64_t>(us * 10); |
| } |
| |
| int64_t CurrentWallclockMicroseconds() { |
| FILETIME ft; |
| ::GetSystemTimePreciseAsFileTime(&ft); |
| return FileTimeToMicroseconds(ft); |
| } |
| |
| // Time between resampling the un-granular clock for this API. |
| constexpr TimeDelta kMaxTimeToAvoidDrift = Seconds(60); |
| |
| int64_t g_initial_time = 0; |
| TimeTicks g_initial_ticks; |
| |
| void InitializeClock() { |
| g_initial_ticks = subtle::TimeTicksNowIgnoringOverride(); |
| g_initial_time = CurrentWallclockMicroseconds(); |
| } |
| |
| // Track the last value passed to timeBeginPeriod so that we can cancel that |
| // call by calling timeEndPeriod with the same value. A value of zero means that |
| // the timer frequency is not currently raised. |
| UINT g_last_interval_requested_ms = 0; |
| // Track if kMinTimerIntervalHighResMs or kMinTimerIntervalLowResMs is active. |
| // For most purposes this could also be named g_is_on_ac_power. |
| bool g_high_res_timer_enabled = false; |
| // How many times the high resolution timer has been called. |
| uint32_t g_high_res_timer_count = 0; |
| // Start time of the high resolution timer usage monitoring. This is needed |
| // to calculate the usage as percentage of the total elapsed time. |
| TimeTicks g_high_res_timer_usage_start; |
| // The cumulative time the high resolution timer has been in use since |
| // |g_high_res_timer_usage_start| moment. |
| TimeDelta g_high_res_timer_usage; |
| // Timestamp of the last activation change of the high resolution timer. This |
| // is used to calculate the cumulative usage. |
| TimeTicks g_high_res_timer_last_activation; |
| // The lock to control access to the above set of variables. |
| Lock* GetHighResLock() { |
| static auto* lock = new Lock(); |
| return lock; |
| } |
| |
| // The two values that ActivateHighResolutionTimer uses to set the systemwide |
| // timer interrupt frequency on Windows. These control how precise timers are |
| // but also have a big impact on battery life. |
| |
| // Used when a faster timer has been requested (g_high_res_timer_count > 0) and |
| // the computer is running on AC power (plugged in) so that it's okay to go to |
| // the highest frequency. |
| constexpr UINT kMinTimerIntervalHighResMs = 1; |
| |
| // Used when a faster timer has been requested (g_high_res_timer_count > 0) and |
| // the computer is running on DC power (battery) so that we don't want to raise |
| // the timer frequency as much. |
| constexpr UINT kMinTimerIntervalLowResMs = 8; |
| |
| // Calculate the desired timer interrupt interval. Note that zero means that the |
| // system default should be used. |
| UINT GetIntervalMs() { |
| if (!g_high_res_timer_count) { |
| return 0; // Use the default, typically 15.625 |
| } |
| if (g_high_res_timer_enabled) { |
| return kMinTimerIntervalHighResMs; |
| } |
| return kMinTimerIntervalLowResMs; |
| } |
| |
| // Compare the currently requested timer interrupt interval to the last interval |
| // requested and update if necessary (by cancelling the old request and making a |
| // new request). If there is no change then do nothing. |
| void UpdateTimerIntervalLocked() { |
| UINT new_interval = GetIntervalMs(); |
| if (new_interval == g_last_interval_requested_ms) { |
| return; |
| } |
| if (g_last_interval_requested_ms) { |
| // Record how long the timer interrupt frequency was raised. |
| g_high_res_timer_usage += subtle::TimeTicksNowIgnoringOverride() - |
| g_high_res_timer_last_activation; |
| // Reset the timer interrupt back to the default. |
| timeEndPeriod(g_last_interval_requested_ms); |
| } |
| g_last_interval_requested_ms = new_interval; |
| if (g_last_interval_requested_ms) { |
| // Record when the timer interrupt was raised. |
| g_high_res_timer_last_activation = subtle::TimeTicksNowIgnoringOverride(); |
| timeBeginPeriod(g_last_interval_requested_ms); |
| } |
| } |
| |
| // Returns the current value of the performance counter. |
| int64_t QPCNowRaw() { |
| LARGE_INTEGER perf_counter_now = {}; |
| // According to the MSDN documentation for QueryPerformanceCounter(), this |
| // will never fail on systems that run XP or later. |
| // https://msdn.microsoft.com/library/windows/desktop/ms644904.aspx |
| ::QueryPerformanceCounter(&perf_counter_now); |
| return perf_counter_now.QuadPart; |
| } |
| |
| #if !defined(ARCH_CPU_ARM64) |
| // Returns the performance frequency. |
| int64_t QPFRaw() { |
| LARGE_INTEGER perf_counter_frequency = {}; |
| // According to the MSDN documentation for QueryPerformanceFrequency(), this |
| // will never fail on systems that run XP or later. |
| // https://learn.microsoft.com/en-us/windows/win32/api/profileapi/nf-profileapi-queryperformancefrequency |
| ::QueryPerformanceFrequency(&perf_counter_frequency); |
| return perf_counter_frequency.QuadPart; |
| } |
| #endif |
| |
| bool SafeConvertToWord(int in, WORD* out) { |
| CheckedNumeric<WORD> result = in; |
| *out = result.ValueOrDefault(std::numeric_limits<WORD>::max()); |
| return result.IsValid(); |
| } |
| |
| } // namespace |
| |
| // Time ----------------------------------------------------------------------- |
| |
| namespace subtle { |
| Time TimeNowIgnoringOverride() { |
| if (g_initial_time == 0) { |
| InitializeClock(); |
| } |
| |
| // We implement time using the high-resolution timers so that we can get |
| // timeouts which likely are smaller than those if we just used |
| // CurrentWallclockMicroseconds(). |
| // |
| // To make this work, we initialize the clock (g_initial_time) and the |
| // counter (initial_ctr). To compute the initial time, we can check |
| // the number of ticks that have elapsed, and compute the delta. |
| // |
| // To avoid any drift, we periodically resync the counters to the system |
| // clock. |
| while (true) { |
| TimeTicks ticks = TimeTicksNowIgnoringOverride(); |
| |
| // Calculate the time elapsed since we started our timer |
| TimeDelta elapsed = ticks - g_initial_ticks; |
| |
| // Check if enough time has elapsed that we need to resync the clock. |
| if (elapsed > kMaxTimeToAvoidDrift) { |
| InitializeClock(); |
| continue; |
| } |
| |
| return Time() + elapsed + Microseconds(g_initial_time); |
| } |
| } |
| |
| Time TimeNowFromSystemTimeIgnoringOverride() { |
| // Force resync. |
| InitializeClock(); |
| return Time() + Microseconds(g_initial_time); |
| } |
| } // namespace subtle |
| |
| // static |
| Time Time::FromFileTime(FILETIME ft) { |
| if (bit_cast<int64_t, FILETIME>(ft) == 0) { |
| return Time(); |
| } |
| if (ft.dwHighDateTime == std::numeric_limits<DWORD>::max() && |
| ft.dwLowDateTime == std::numeric_limits<DWORD>::max()) { |
| return Max(); |
| } |
| return Time(FileTimeToMicroseconds(ft)); |
| } |
| |
| FILETIME Time::ToFileTime() const { |
| if (is_null()) { |
| return bit_cast<FILETIME, int64_t>(0); |
| } |
| if (is_max()) { |
| FILETIME result; |
| result.dwHighDateTime = std::numeric_limits<DWORD>::max(); |
| result.dwLowDateTime = std::numeric_limits<DWORD>::max(); |
| return result; |
| } |
| return MicrosecondsToFileTime(us_); |
| } |
| |
| // static |
| // Enable raising of the system-global timer interrupt frequency to 1 kHz (when |
| // enable is true, which happens when on AC power) or some lower frequency when |
| // on battery power (when enable is false). If the g_high_res_timer_enabled |
| // setting hasn't actually changed or if if there are no outstanding requests |
| // (if g_high_res_timer_count is zero) then do nothing. |
| // TL;DR - call this when going from AC to DC power or vice-versa. |
| void Time::EnableHighResolutionTimer(bool enable) { |
| AutoLock lock(*GetHighResLock()); |
| g_high_res_timer_enabled = enable; |
| UpdateTimerIntervalLocked(); |
| } |
| |
| // static |
| // Request that the system-global Windows timer interrupt frequency be raised. |
| // How high the frequency is raised depends on the system's power state and |
| // possibly other options. |
| // TL;DR - call this at the beginning and end of a time period where you want |
| // higher frequency timer interrupts. Each call with activating=true must be |
| // paired with a subsequent activating=false call. |
| bool Time::ActivateHighResolutionTimer(bool activating) { |
| // We only do work on the transition from zero to one or one to zero so we |
| // can easily undo the effect (if necessary) when EnableHighResolutionTimer is |
| // called. |
| const uint32_t max = std::numeric_limits<uint32_t>::max(); |
| |
| AutoLock lock(*GetHighResLock()); |
| if (activating) { |
| DCHECK_NE(g_high_res_timer_count, max); |
| ++g_high_res_timer_count; |
| } else { |
| DCHECK_NE(g_high_res_timer_count, 0u); |
| --g_high_res_timer_count; |
| } |
| UpdateTimerIntervalLocked(); |
| return true; |
| } |
| |
| // static |
| // See if the timer interrupt interval has been set to the lowest value. |
| bool Time::IsHighResolutionTimerInUse() { |
| AutoLock lock(*GetHighResLock()); |
| return g_last_interval_requested_ms == kMinTimerIntervalHighResMs; |
| } |
| |
| // static |
| void Time::ResetHighResolutionTimerUsage() { |
| AutoLock lock(*GetHighResLock()); |
| g_high_res_timer_usage = TimeDelta(); |
| g_high_res_timer_usage_start = subtle::TimeTicksNowIgnoringOverride(); |
| if (g_high_res_timer_count > 0) { |
| g_high_res_timer_last_activation = g_high_res_timer_usage_start; |
| } |
| } |
| |
| // static |
| double Time::GetHighResolutionTimerUsage() { |
| AutoLock lock(*GetHighResLock()); |
| TimeTicks now = subtle::TimeTicksNowIgnoringOverride(); |
| TimeDelta elapsed_time = now - g_high_res_timer_usage_start; |
| if (elapsed_time.is_zero()) { |
| // This is unexpected but possible if TimeTicks resolution is low and |
| // GetHighResolutionTimerUsage() is called promptly after |
| // ResetHighResolutionTimerUsage(). |
| return 0.0; |
| } |
| TimeDelta used_time = g_high_res_timer_usage; |
| if (g_high_res_timer_count > 0) { |
| // If currently activated add the remainder of time since the last |
| // activation. |
| used_time += now - g_high_res_timer_last_activation; |
| } |
| return used_time / elapsed_time * 100; |
| } |
| |
| // static |
| bool Time::FromExploded(bool is_local, const Exploded& exploded, Time* time) { |
| // Create the system struct representing our exploded time. It will either be |
| // in local time or UTC.If casting from int to WORD results in overflow, |
| // fail and return Time(0). |
| SYSTEMTIME st; |
| if (!SafeConvertToWord(exploded.year, &st.wYear) || |
| !SafeConvertToWord(exploded.month, &st.wMonth) || |
| !SafeConvertToWord(exploded.day_of_week, &st.wDayOfWeek) || |
| !SafeConvertToWord(exploded.day_of_month, &st.wDay) || |
| !SafeConvertToWord(exploded.hour, &st.wHour) || |
| !SafeConvertToWord(exploded.minute, &st.wMinute) || |
| !SafeConvertToWord(exploded.second, &st.wSecond) || |
| !SafeConvertToWord(exploded.millisecond, &st.wMilliseconds)) { |
| *time = Time(0); |
| return false; |
| } |
| |
| FILETIME ft; |
| bool success = true; |
| // Ensure that it's in UTC. |
| if (is_local) { |
| SYSTEMTIME utc_st; |
| success = TzSpecificLocalTimeToSystemTime(nullptr, &st, &utc_st) && |
| SystemTimeToFileTime(&utc_st, &ft); |
| } else { |
| success = !!SystemTimeToFileTime(&st, &ft); |
| } |
| |
| *time = Time(success ? FileTimeToMicroseconds(ft) : 0); |
| return success; |
| } |
| |
| void Time::Explode(bool is_local, Exploded* exploded) const { |
| if (!CanConvertToFileTime(us_)) { |
| // We are not able to convert it to FILETIME. |
| ZeroMemory(exploded, sizeof(*exploded)); |
| return; |
| } |
| |
| const FILETIME utc_ft = MicrosecondsToFileTime(us_); |
| |
| // FILETIME in local time if necessary. |
| bool success = true; |
| // FILETIME in SYSTEMTIME (exploded). |
| SYSTEMTIME st = {0}; |
| if (is_local) { |
| SYSTEMTIME utc_st; |
| // We don't use FileTimeToLocalFileTime here, since it uses the current |
| // settings for the time zone and daylight saving time. Therefore, if it is |
| // daylight saving time, it will take daylight saving time into account, |
| // even if the time you are converting is in standard time. |
| success = FileTimeToSystemTime(&utc_ft, &utc_st) && |
| SystemTimeToTzSpecificLocalTime(nullptr, &utc_st, &st); |
| } else { |
| success = !!FileTimeToSystemTime(&utc_ft, &st); |
| } |
| |
| if (!success) { |
| ZeroMemory(exploded, sizeof(*exploded)); |
| return; |
| } |
| |
| exploded->year = st.wYear; |
| exploded->month = st.wMonth; |
| exploded->day_of_week = st.wDayOfWeek; |
| exploded->day_of_month = st.wDay; |
| exploded->hour = st.wHour; |
| exploded->minute = st.wMinute; |
| exploded->second = st.wSecond; |
| exploded->millisecond = st.wMilliseconds; |
| } |
| |
| // TimeTicks ------------------------------------------------------------------ |
| |
| namespace { |
| |
| // We define a wrapper to adapt between the __stdcall and __cdecl call of the |
| // mock function, and to avoid a static constructor. Assigning an import to a |
| // function pointer directly would require setup code to fetch from the IAT. |
| DWORD timeGetTimeWrapper() { |
| return timeGetTime(); |
| } |
| |
| DWORD (*g_tick_function)(void) = &timeGetTimeWrapper; |
| |
| // A structure holding the most significant bits of "last seen" and a |
| // "rollover" counter. |
| union LastTimeAndRolloversState { |
| // The state as a single 32-bit opaque value. |
| std::atomic<int32_t> as_opaque_32{0}; |
| |
| // The state as usable values. |
| struct { |
| // The top 8-bits of the "last" time. This is enough to check for rollovers |
| // and the small bit-size means fewer CompareAndSwap operations to store |
| // changes in state, which in turn makes for fewer retries. |
| uint8_t last_8; |
| // A count of the number of detected rollovers. Using this as bits 47-32 |
| // of the upper half of a 64-bit value results in a 48-bit tick counter. |
| // This extends the total rollover period from about 49 days to about 8800 |
| // years while still allowing it to be stored with last_8 in a single |
| // 32-bit value. |
| uint16_t rollovers; |
| } as_values; |
| }; |
| std::atomic<int32_t> g_last_time_and_rollovers = 0; |
| static_assert(sizeof(LastTimeAndRolloversState) <= |
| sizeof(g_last_time_and_rollovers), |
| "LastTimeAndRolloversState does not fit in a single atomic word"); |
| |
| // We use timeGetTime() to implement TimeTicks::Now(). This can be problematic |
| // because it returns the number of milliseconds since Windows has started, |
| // which will roll over the 32-bit value every ~49 days. We try to track |
| // rollover ourselves, which works if TimeTicks::Now() is called at least every |
| // 48.8 days (not 49 days because only changes in the top 8 bits get noticed). |
| TimeTicks RolloverProtectedNow() { |
| LastTimeAndRolloversState state; |
| DWORD now; // DWORD is always unsigned 32 bits. |
| |
| while (true) { |
| // Fetch the "now" and "last" tick values, updating "last" with "now" and |
| // incrementing the "rollovers" counter if the tick-value has wrapped back |
| // around. Atomic operations ensure that both "last" and "rollovers" are |
| // always updated together. |
| int32_t original = |
| g_last_time_and_rollovers.load(std::memory_order_acquire); |
| state.as_opaque_32 = original; |
| now = g_tick_function(); |
| uint8_t now_8 = static_cast<uint8_t>(now >> 24); |
| if (now_8 < state.as_values.last_8) { |
| ++state.as_values.rollovers; |
| } |
| state.as_values.last_8 = now_8; |
| |
| // If the state hasn't changed, exit the loop. |
| if (state.as_opaque_32 == original) { |
| break; |
| } |
| |
| // Save the changed state. If the existing value is unchanged from the |
| // original so that the operation is successful. Exit the loop. |
| bool success = g_last_time_and_rollovers.compare_exchange_strong( |
| original, state.as_opaque_32, std::memory_order_release); |
| if (success) { |
| break; |
| } |
| |
| // Another thread has done something in between so retry from the top. |
| } |
| |
| return TimeTicks() + |
| Milliseconds(now + |
| (static_cast<uint64_t>(state.as_values.rollovers) << 32)); |
| } |
| |
| // Discussion of tick counter options on Windows: |
| // |
| // (1) CPU cycle counter. (Retrieved via RDTSC) |
| // The CPU counter provides the highest resolution time stamp and is the least |
| // expensive to retrieve. However, on older CPUs, two issues can affect its |
| // reliability: First it is maintained per processor and not synchronized |
| // between processors. Also, the counters will change frequency due to thermal |
| // and power changes, and stop in some states. |
| // |
| // (2) QueryPerformanceCounter (QPC). The QPC counter provides a high- |
| // resolution (<1 microsecond) time stamp. On most hardware running today, it |
| // auto-detects and uses the constant-rate RDTSC counter to provide extremely |
| // efficient and reliable time stamps. |
| // |
| // On older CPUs where RDTSC is unreliable, it falls back to using more |
| // expensive (20X to 40X more costly) alternate clocks, such as HPET or the ACPI |
| // PM timer, and can involve system calls; and all this is up to the HAL (with |
| // some help from ACPI). According to |
| // http://blogs.msdn.com/oldnewthing/archive/2005/09/02/459952.aspx, in the |
| // worst case, it gets the counter from the rollover interrupt on the |
| // programmable interrupt timer. In best cases, the HAL may conclude that the |
| // RDTSC counter runs at a constant frequency, then it uses that instead. On |
| // multiprocessor machines, it will try to verify the values returned from |
| // RDTSC on each processor are consistent with each other, and apply a handful |
| // of workarounds for known buggy hardware. In other words, QPC is supposed to |
| // give consistent results on a multiprocessor computer, but for older CPUs it |
| // can be unreliable due bugs in BIOS or HAL. |
| // |
| // (3) System time. The system time provides a low-resolution (from ~1 to ~15.6 |
| // milliseconds) time stamp but is comparatively less expensive to retrieve and |
| // more reliable. Time::EnableHighResolutionTimer() and |
| // Time::ActivateHighResolutionTimer() can be called to alter the resolution of |
| // this timer; and also other Windows applications can alter it, affecting this |
| // one. |
| |
| TimeTicks InitialNowFunction(); |
| |
| // See "threading notes" in InitializeNowFunctionPointer() for details on how |
| // concurrent reads/writes to these globals has been made safe. |
| std::atomic<TimeTicksNowFunction> g_time_ticks_now_ignoring_override_function{ |
| &InitialNowFunction}; |
| int64_t g_qpc_ticks_per_second = 0; |
| |
| TimeDelta QPCValueToTimeDelta(LONGLONG qpc_value) { |
| // Ensure that the assignment to |g_qpc_ticks_per_second|, made in |
| // InitializeNowFunctionPointer(), has happened by this point. |
| std::atomic_thread_fence(std::memory_order_acquire); |
| |
| DCHECK_GT(g_qpc_ticks_per_second, 0); |
| |
| // If the QPC Value is below the overflow threshold, we proceed with |
| // simple multiply and divide. |
| if (qpc_value < Time::kQPCOverflowThreshold) { |
| return Microseconds(qpc_value * Time::kMicrosecondsPerSecond / |
| g_qpc_ticks_per_second); |
| } |
| // Otherwise, calculate microseconds in a round about manner to avoid |
| // overflow and precision issues. |
| int64_t whole_seconds = qpc_value / g_qpc_ticks_per_second; |
| int64_t leftover_ticks = qpc_value - (whole_seconds * g_qpc_ticks_per_second); |
| return Microseconds((whole_seconds * Time::kMicrosecondsPerSecond) + |
| ((leftover_ticks * Time::kMicrosecondsPerSecond) / |
| g_qpc_ticks_per_second)); |
| } |
| |
| TimeTicks QPCNow() { |
| return TimeTicks() + QPCValueToTimeDelta(QPCNowRaw()); |
| } |
| |
| void InitializeNowFunctionPointer() { |
| LARGE_INTEGER ticks_per_sec = {}; |
| // `QueryPerformanceFrequency` always succeeds and sets its out parameter to a |
| // nonzero value on Windows versions more recent than Windows XP: |
| // https://learn.microsoft.com/en-us/windows/win32/api/profileapi/nf-profileapi-queryperformancefrequency |
| // Once these `CHECK`s are shown to not trigger in the wild, this condition |
| // can be changed to a CHECK and `ticks_per_sec.QuadPart <= 0 ` can be removed |
| // from the ternary below that selects the function pointer. |
| if (!QueryPerformanceFrequency(&ticks_per_sec)) { |
| ticks_per_sec.QuadPart = 0; |
| NOTREACHED(base::NotFatalUntil::M138); |
| } else { |
| CHECK(ticks_per_sec.QuadPart > 0, base::NotFatalUntil::M138); |
| } |
| |
| // If the QPC implementation is expensive and/or unreliable, TimeTicks::Now() |
| // will still use the low-resolution clock. A CPU lacking a non-stop time |
| // counter will cause Windows to provide an alternate QPC implementation that |
| // works, but is expensive to use. |
| // |
| // Otherwise, Now uses the high-resolution QPC clock. As of 9 September 2024, |
| // ~97% of users fall within this category. |
| bool eligible_for_high_res_time_ticks = false; |
| |
| // `ticks_per_sec.QuadPart <= 0` shouldn't happen post-WinXP (see CHECKs |
| // above) but if it does, QPC is broken and shouldn't be used for any reason. |
| if (ticks_per_sec.QuadPart > 0) { |
| CPU cpu; |
| // QPC is enabled for all devices with invariant TSCs. |
| // On devices where the CPU doesn't report having an invariant TSC, we would |
| // previously have considered the QPC overhead to be unacceptable. For this |
| // field trial, try enabling the high-res, QPC-based implementation of |
| // TimeTicks on 50% of such devices. |
| bool force_high_res_time_ticks = false; |
| // There is an explicit check for |
| // `base::CommandLine::InitializedForCurrentProcess()` not being null here, |
| // because some targets (like `generate_colors_info`) use `TimeTicks` during |
| // the build without initializing this command line object. In those cases, |
| // it's also not necessary to roll the dice to force high res timer since |
| // we're not running a browser. |
| if (base::CommandLine::InitializedForCurrentProcess()) { |
| if (base::CommandLine::ForCurrentProcess()->HasSwitch( |
| switches::kForceHighResTimeTicks)) { |
| // If `switches::kForceHighResTimeTicks` is present, it's because this |
| // is a child process that is being instructed to use the same clock as |
| // its parent browser process. In this case, force the use of high |
| // resolution TimeTicks iff `switches::kForceHighResTimeTicks` is set to |
| // "enabled". It can also take the value of "disabled" when the browser |
| // is in either the "Control" or "Excluded" groups. |
| auto switch_value = |
| base::CommandLine::ForCurrentProcess()->GetSwitchValueASCII( |
| switches::kForceHighResTimeTicks); |
| if (switch_value == "enabled") { |
| force_high_res_time_ticks = true; |
| } |
| } else { |
| // If `switches::kForceHighResTimeTicks` isn't present, this is either |
| // the browser process so we should roll a dice to determine if we're in |
| // the field trial, or this device already uses high resolution |
| // TimeTicks so the dice roll will not be used. |
| force_high_res_time_ticks = base::RandDouble() < 0.5; |
| } |
| } |
| |
| eligible_for_high_res_time_ticks = |
| cpu.has_non_stop_time_stamp_counter() || force_high_res_time_ticks; |
| } |
| |
| const TimeTicksNowFunction now_function = |
| eligible_for_high_res_time_ticks ? &QPCNow : &RolloverProtectedNow; |
| |
| // Threading note 1: In an unlikely race condition, it's possible for two or |
| // more threads to enter InitializeNowFunctionPointer() in parallel. This is |
| // not a problem since all threads end up writing out the same values |
| // to the global variables, and those variable being atomic are safe to read |
| // from other threads. |
| // |
| // Under the high resolution field trial for low resolution devices, multiple |
| // threads racing could roll a different dice roll and attempt to set the |
| // functions to different values. To avoid having thread A set the "now" |
| // function to something, and thread B set the "now without override" function |
| // to something else, only the thread where the first compare_exchange |
| // succeeds is allowed to proceed with setting the remainder of the global |
| // state. |
| // |
| // Threading note 2: A release fence is placed here to ensure, from the |
| // perspective of other threads using the function pointers, that the |
| // assignment to |g_qpc_ticks_per_second| happens before the function pointers |
| // are changed. |
| g_qpc_ticks_per_second = ticks_per_sec.QuadPart; |
| std::atomic_thread_fence(std::memory_order_release); |
| // memory_order_relaxed is sufficient since an explicit fence was inserted |
| // above. |
| base::TimeTicksNowFunction initial_time_ticks_now_function = |
| &InitialNowFunction; |
| |
| if (g_time_ticks_now_ignoring_override_function.compare_exchange_strong( |
| initial_time_ticks_now_function, now_function, |
| std::memory_order_relaxed)) { |
| // Also set g_time_ticks_now_function to avoid the additional indirection |
| // via TimeTicksNowIgnoringOverride() for future calls to TimeTicks::Now(). |
| internal::g_time_ticks_now_function.store(now_function, |
| std::memory_order_relaxed); |
| } |
| } |
| |
| TimeTicks InitialNowFunction() { |
| InitializeNowFunctionPointer(); |
| return g_time_ticks_now_ignoring_override_function.load( |
| std::memory_order_relaxed)(); |
| } |
| |
| enum class HighResolutionTrialState { |
| kAlreadyHighResolution, |
| kExcludedFromTrial, |
| kDontUseHighResolution, |
| kUseHighResolution, |
| }; |
| |
| HighResolutionTrialState GetHighResolutionTrialState() { |
| // This is a copy of the conditions in `InitializeNowFunctionPointer`, minus |
| // the work around global atomics. The return value of this function shouldn't |
| // vary on the same device. |
| // TODO(crbug.com/410560675): Remove this function once experimentation with |
| // QPC is concluded. |
| |
| // IsHighResolution() initializes the clock if it hasn't been done. |
| bool is_high_res = TimeTicks::IsHighResolution(); |
| if (g_qpc_ticks_per_second == 0) { |
| // QPC is broken and can't be enabled. |
| return HighResolutionTrialState::kExcludedFromTrial; |
| } |
| |
| CPU cpu; |
| if (!cpu.has_non_stop_time_stamp_counter()) { |
| if (is_high_res) { |
| // If the device isn't considered eligible for QPC-based TimeTicks but is |
| // using it regardless, it means that it's part of the experimental QPC |
| // group. |
| return HighResolutionTrialState::kUseHighResolution; |
| } else { |
| // Otherwise, the device is expectedly using low-res TimeTicks, add it to |
| // the control group. |
| return HighResolutionTrialState::kDontUseHighResolution; |
| } |
| } |
| |
| // Don't add clients with ideal QPC implementations to the trial at all. |
| return HighResolutionTrialState::kAlreadyHighResolution; |
| } |
| |
| } // namespace |
| |
| // static |
| TimeTicks::TickFunctionType TimeTicks::SetMockTickFunction( |
| TickFunctionType ticker) { |
| TickFunctionType old = g_tick_function; |
| g_tick_function = ticker; |
| g_last_time_and_rollovers.store(0, std::memory_order_relaxed); |
| return old; |
| } |
| |
| namespace subtle { |
| TimeTicks TimeTicksNowIgnoringOverride() { |
| return g_time_ticks_now_ignoring_override_function.load( |
| std::memory_order_relaxed)(); |
| } |
| |
| TimeTicks TimeTicksLowResolutionNowIgnoringOverride() { |
| return RolloverProtectedNow(); |
| } |
| } // namespace subtle |
| |
| // static |
| bool TimeTicks::IsHighResolution() { |
| if (g_time_ticks_now_ignoring_override_function == &InitialNowFunction) { |
| InitializeNowFunctionPointer(); |
| } |
| return g_time_ticks_now_ignoring_override_function == &QPCNow; |
| } |
| |
| // static |
| bool TimeTicks::IsConsistentAcrossProcesses() { |
| // According to Windows documentation [1] QPC is consistent post-Windows |
| // Vista. So if we are using QPC then we are consistent which is the same as |
| // being high resolution. |
| // |
| // [1] |
| // https://msdn.microsoft.com/en-us/library/windows/desktop/dn553408(v=vs.85).aspx |
| // |
| // "In general, the performance counter results are consistent across all |
| // processors in multi-core and multi-processor systems, even when measured on |
| // different threads or processes. Here are some exceptions to this rule: |
| // - Pre-Windows Vista operating systems that run on certain processors might |
| // violate this consistency because of one of these reasons: |
| // 1. The hardware processors have a non-invariant TSC and the BIOS |
| // doesn't indicate this condition correctly. |
| // 2. The TSC synchronization algorithm that was used wasn't suitable for |
| // systems with large numbers of processors." |
| return IsHighResolution(); |
| } |
| |
| // static |
| TimeTicks::Clock TimeTicks::GetClock() { |
| return IsHighResolution() ? Clock::WIN_QPC |
| : Clock::WIN_ROLLOVER_PROTECTED_TIME_GET_TIME; |
| } |
| |
| // LiveTicks ------------------------------------------------------------------ |
| |
| namespace subtle { |
| LiveTicks LiveTicksNowIgnoringOverride() { |
| ULONGLONG unbiased_interrupt_time; |
| QueryUnbiasedInterruptTimePrecise(&unbiased_interrupt_time); |
| // QueryUnbiasedInterruptTimePrecise gets the interrupt time in system time |
| // units of 100 nanoseconds. |
| return LiveTicks() + Nanoseconds(unbiased_interrupt_time * 100); |
| } |
| } // namespace subtle |
| |
| // ThreadTicks ---------------------------------------------------------------- |
| |
| namespace subtle { |
| ThreadTicks ThreadTicksNowIgnoringOverride() { |
| return ThreadTicks::GetForThread(PlatformThread::CurrentHandle()); |
| } |
| } // namespace subtle |
| |
| // static |
| ThreadTicks ThreadTicks::GetForThread( |
| const PlatformThreadHandle& thread_handle) { |
| DCHECK(IsSupported()); |
| |
| #if defined(ARCH_CPU_ARM64) |
| // QueryThreadCycleTime versus TSCTicksPerSecond doesn't have much relation to |
| // actual elapsed time on Windows on Arm, because QueryThreadCycleTime is |
| // backed by the actual number of CPU cycles executed, rather than a |
| // constant-rate timer like Intel. To work around this, use GetThreadTimes |
| // (which isn't as accurate but is meaningful as a measure of elapsed |
| // per-thread time). |
| FILETIME creation_time, exit_time, kernel_time, user_time; |
| ::GetThreadTimes(thread_handle.platform_handle(), &creation_time, &exit_time, |
| &kernel_time, &user_time); |
| |
| const int64_t us = FileTimeToMicroseconds(user_time); |
| #else |
| // Get the number of TSC ticks used by the current thread. |
| ULONG64 thread_cycle_time = 0; |
| ::QueryThreadCycleTime(thread_handle.platform_handle(), &thread_cycle_time); |
| |
| // Get the frequency of the TSC. |
| const double tsc_ticks_per_second = time_internal::TSCTicksPerSecond(); |
| if (tsc_ticks_per_second == 0) { |
| return ThreadTicks(); |
| } |
| |
| // Return the CPU time of the current thread. |
| const double thread_time_seconds = thread_cycle_time / tsc_ticks_per_second; |
| const int64_t us = |
| static_cast<int64_t>(thread_time_seconds * Time::kMicrosecondsPerSecond); |
| #endif |
| |
| return ThreadTicks(us); |
| } |
| |
| // static |
| bool ThreadTicks::IsSupportedWin() { |
| #if defined(ARCH_CPU_ARM64) |
| // The Arm implementation does not use QueryThreadCycleTime and therefore does |
| // not care about the time stamp counter. |
| return true; |
| #else |
| return time_internal::HasConstantRateTSC(); |
| #endif |
| } |
| |
| // static |
| void ThreadTicks::WaitUntilInitializedWin() { |
| #if !defined(ARCH_CPU_ARM64) |
| while (time_internal::TSCTicksPerSecond() == 0) { |
| ::Sleep(10); |
| } |
| #endif |
| } |
| |
| // static |
| TimeTicks TimeTicks::FromQPCValue(LONGLONG qpc_value) { |
| return TimeTicks() + QPCValueToTimeDelta(qpc_value); |
| } |
| |
| // static |
| bool TimeTicks::GetHighResolutionTimeTicksFieldTrial(std::string* trial_name, |
| std::string* group_name) { |
| auto state = GetHighResolutionTrialState(); |
| |
| switch (state) { |
| case HighResolutionTrialState::kAlreadyHighResolution: |
| return false; |
| case HighResolutionTrialState::kExcludedFromTrial: |
| *group_name = "Excluded"; |
| break; |
| case HighResolutionTrialState::kDontUseHighResolution: |
| *group_name = "Control"; |
| break; |
| case HighResolutionTrialState::kUseHighResolution: |
| *group_name = "Enabled"; |
| break; |
| } |
| |
| *trial_name = "HighResolutionTimeTicks"; |
| return true; |
| } |
| |
| // static |
| void TimeTicks::MaybeAddHighResolutionTimeTicksSwitch( |
| base::CommandLine* command_line) { |
| auto state = GetHighResolutionTrialState(); |
| |
| switch (state) { |
| case HighResolutionTrialState::kAlreadyHighResolution: |
| // If the device is already using an ideal QPC implementation for |
| // TimeTicks, don't pass any command line flag. |
| break; |
| case HighResolutionTrialState::kExcludedFromTrial: |
| // In the cases of "Control" and "Excluded", tell the child process not to |
| // use QPC for TimeTicks to match the browser process. |
| [[fallthrough]]; |
| case HighResolutionTrialState::kDontUseHighResolution: |
| command_line->AppendSwitchASCII(switches::kForceHighResTimeTicks, |
| "disabled"); |
| break; |
| case HighResolutionTrialState::kUseHighResolution: |
| // If the device doesn't report having an invariant TSC, but the browser |
| // process has rolled a dice and is being included in the high-resolution |
| // trial's "enabled" group, pass this information to the child process. |
| command_line->AppendSwitchASCII(switches::kForceHighResTimeTicks, |
| "enabled"); |
| break; |
| } |
| } |
| |
| // TimeDelta ------------------------------------------------------------------ |
| |
| // static |
| TimeDelta TimeDelta::FromQPCValue(LONGLONG qpc_value) { |
| return QPCValueToTimeDelta(qpc_value); |
| } |
| |
| // static |
| TimeDelta TimeDelta::FromFileTime(FILETIME ft) { |
| return Microseconds(FileTimeToMicroseconds(ft)); |
| } |
| |
| // static |
| TimeDelta TimeDelta::FromWinrtDateTime(ABI::Windows::Foundation::DateTime dt) { |
| // UniversalTime is 100 ns intervals since January 1, 1601 (UTC) |
| return Microseconds(dt.UniversalTime / 10); |
| } |
| |
| ABI::Windows::Foundation::DateTime TimeDelta::ToWinrtDateTime() const { |
| ABI::Windows::Foundation::DateTime date_time; |
| date_time.UniversalTime = InMicroseconds() * 10; |
| return date_time; |
| } |
| |
| // static |
| TimeDelta TimeDelta::FromWinrtTimeSpan(ABI::Windows::Foundation::TimeSpan ts) { |
| // Duration is 100 ns intervals |
| return Microseconds(ts.Duration / 10); |
| } |
| |
| ABI::Windows::Foundation::TimeSpan TimeDelta::ToWinrtTimeSpan() const { |
| ABI::Windows::Foundation::TimeSpan time_span; |
| time_span.Duration = InMicroseconds() * 10; |
| return time_span; |
| } |
| |
| #if !defined(ARCH_CPU_ARM64) |
| namespace time_internal { |
| |
| bool HasConstantRateTSC() { |
| static bool is_supported = CPU().has_non_stop_time_stamp_counter(); |
| return is_supported; |
| } |
| |
| double TSCTicksPerSecond() { |
| DCHECK(HasConstantRateTSC()); |
| // The value returned by QueryPerformanceFrequency() cannot be used as the TSC |
| // frequency, because there is no guarantee that the TSC frequency is equal to |
| // the performance counter frequency. |
| // The TSC frequency is cached in a static variable because it takes some time |
| // to compute it. |
| static double tsc_ticks_per_second = 0; |
| if (tsc_ticks_per_second != 0) { |
| return tsc_ticks_per_second; |
| } |
| |
| // Increase the thread priority to reduces the chances of having a context |
| // switch during a reading of the TSC and the performance counter. |
| const int previous_priority = ::GetThreadPriority(::GetCurrentThread()); |
| ::SetThreadPriority(::GetCurrentThread(), THREAD_PRIORITY_HIGHEST); |
| |
| // The first time that this function is called, make an initial reading of the |
| // TSC and the performance counter. |
| |
| static const uint64_t tsc_initial = __rdtsc(); |
| static const int64_t perf_counter_initial = QPCNowRaw(); |
| static const int64_t perf_counter_frequency = QPFRaw(); |
| |
| // Make a another reading of the TSC and the performance counter every time |
| // that this function is called. |
| const uint64_t tsc_now = __rdtsc(); |
| const int64_t perf_counter_now = QPCNowRaw(); |
| |
| // Reset the thread priority. |
| ::SetThreadPriority(::GetCurrentThread(), previous_priority); |
| |
| // Make sure that at least 50 ms elapsed between the 2 readings. The first |
| // time that this function is called, we don't expect this to be the case. |
| // Note: The longer the elapsed time between the 2 readings is, the more |
| // accurate the computed TSC frequency will be. The 50 ms value was |
| // chosen because local benchmarks show that it allows us to get a |
| // stddev of less than 1 tick/us between multiple runs. |
| DCHECK_GE(perf_counter_now, perf_counter_initial); |
| const int64_t perf_counter_ticks = perf_counter_now - perf_counter_initial; |
| const double elapsed_time_seconds = |
| perf_counter_ticks / static_cast<double>(perf_counter_frequency); |
| |
| constexpr double kMinimumEvaluationPeriodSeconds = 0.05; |
| if (elapsed_time_seconds < kMinimumEvaluationPeriodSeconds) { |
| return 0; |
| } |
| |
| // Compute the frequency of the TSC. |
| DCHECK_GE(tsc_now, tsc_initial); |
| const uint64_t tsc_ticks = tsc_now - tsc_initial; |
| tsc_ticks_per_second = tsc_ticks / elapsed_time_seconds; |
| |
| return tsc_ticks_per_second; |
| } |
| |
| } // namespace time_internal |
| #endif // defined(ARCH_CPU_ARM64) |
| |
| } // namespace base |