PPS (Pulse Per Second) generates a hardware pulse every second based on
CLOCK_REALTIME. This works fine when the pulse is generated in software
from a hrtimer callback function.
For hardware which generates the pulse by programming a timer it is
required to convert CLOCK_REALTIME to the underlying hardware clock.
The X86 Timed IO device is based on the Always Running Timer (ART), which
is the base clock of the TSC, which is usually the system clocksource on
X86.
The core code already has functionality to convert base clock timestamps to
system clocksource timestamps, but there is no support for converting the
other way around.
Provide the required functionality to support such devices in a generic
way to avoid code duplication in drivers:
1) ktime_real_to_base_clock() to convert a CLOCK_REALTIME timestamp to a
base clock timestamp
2) timekeeping_clocksource_has_base() to allow drivers to validate that
the system clocksource is based on a particular clocksource ID.
[ tglx: Simplify timekeeping_clocksource_has_base() and add missing READ_ONCE() ]
Co-developed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Co-developed-by: Christopher S. Hall <christopher.s.hall@intel.com>
Signed-off-by: Christopher S. Hall <christopher.s.hall@intel.com>
Signed-off-by: Lakshmi Sowjanya D <lakshmi.sowjanya.d@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240513103813.5666-10-lakshmi.sowjanya.d@intel.com
Hardware time stamps like provided by PTP clock implementations are based
on a clock which feeds both the PCIe device and the system clock. For
further processing the underlying hardwarre clock timestamp must be
converted to the system clock.
Right now this requires drivers to invoke an architecture specific
conversion function, e.g. to convert the ART (Always Running Timer)
timestamp to a TSC timestamp.
As the system clock is aware of the underlying base clock, this can be
moved to the core code by providing a base clock property for the system
clock which contains the conversion factors and assigning a clocksource ID
to the base clock.
Add the required data structures and the conversion infrastructure in the
core code to prepare for converting X86 and the related PTP drivers over.
[ tglx: Added a missing READ_ONCE(). Massaged change log ]
Co-developed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Co-developed-by: Christopher S. Hall <christopher.s.hall@intel.com>
Signed-off-by: Christopher S. Hall <christopher.s.hall@intel.com>
Signed-off-by: Lakshmi Sowjanya D <lakshmi.sowjanya.d@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240513103813.5666-2-lakshmi.sowjanya.d@intel.com
For the case !CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE, forego overflow
protection in the range (mask << 1) < delta <= mask, and interpret it
always as an inconsistency between CPU clock values. That allows
slightly neater code, and it is on a slow path so has no effect on
performance.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-19-adrian.hunter@intel.com
Kernel timekeeping is designed to keep the change in cycles (since the last
timer interrupt) below max_cycles, which prevents multiplication overflow
when converting cycles to nanoseconds. However, if timer interrupts stop,
the calculation will eventually overflow.
Add protection against that. In timekeeping_cycles_to_ns() calculation,
check against max_cycles, falling back to a slower higher precision
calculation. In timekeeping_forward_now(), process delta in chunks of at
most max_cycles.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-18-adrian.hunter@intel.com
Open code clocksource_delta() in timekeeping_cycles_to_ns() so that
overflow safety can be added efficiently.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-17-adrian.hunter@intel.com
timekeeping_delta_to_ns() is now called only from
timekeeping_cycles_to_ns(), and it is not useful otherwise.
Simplify the code by folding it into timekeeping_cycles_to_ns().
No functional change.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-16-adrian.hunter@intel.com
Consolidate timekeeping helpers, making use of timekeeping_cycles_to_ns()
in preference to directly using timekeeping_delta_to_ns().
No functional change.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-15-adrian.hunter@intel.com
Simplify the usage of timekeeping sanity checking, in preparation for
consolidating timekeeping helpers. This works towards eliminating
timekeeping_delta_to_ns() in favour of timekeeping_cycles_to_ns().
No functional change.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-14-adrian.hunter@intel.com
Simplify __timekeeping_get_ns() by reusing timekeeping_cycles_to_ns().
No functional change.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-13-adrian.hunter@intel.com
Put together declaration and initialization of the local variable 'delta'.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-12-adrian.hunter@intel.com
Rename fast_tk_get_delta_ns() to __timekeeping_get_ns() to prepare for its
reuse as a general timekeeping helper function.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-11-adrian.hunter@intel.com
Move timekeeping helper functions to prepare for their reuse.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240325064023.2997-10-adrian.hunter@intel.com
- The hierarchical timer pull model
When timer wheel timers are armed they are placed into the timer wheel
of a CPU which is likely to be busy at the time of expiry. This is done
to avoid wakeups on potentially idle CPUs.
This is wrong in several aspects:
1) The heuristics to select the target CPU are wrong by
definition as the chance to get the prediction right is close
to zero.
2) Due to #1 it is possible that timers are accumulated on a
single target CPU
3) The required computation in the enqueue path is just overhead for
dubious value especially under the consideration that the vast
majority of timer wheel timers are either canceled or rearmed
before they expire.
The timer pull model avoids the above by removing the target
computation on enqueue and queueing timers always on the CPU on which
they get armed.
This is achieved by having separate wheels for CPU pinned timers and
global timers which do not care about where they expire.
As long as a CPU is busy it handles both the pinned and the global
timers which are queued on the CPU local timer wheels.
When a CPU goes idle it evaluates its own timer wheels:
- If the first expiring timer is a pinned timer, then the global
timers can be ignored as the CPU will wake up before they expire.
- If the first expiring timer is a global timer, then the expiry time
is propagated into the timer pull hierarchy and the CPU makes sure
to wake up for the first pinned timer.
The timer pull hierarchy organizes CPUs in groups of eight at the
lowest level and at the next levels groups of eight groups up to the
point where no further aggregation of groups is required, i.e. the
number of levels is log8(NR_CPUS). The magic number of eight has been
established by experimention, but can be adjusted if needed.
In each group one busy CPU acts as the migrator. It's only one CPU to
avoid lock contention on remote timer wheels.
The migrator CPU checks in its own timer wheel handling whether there
are other CPUs in the group which have gone idle and have global timers
to expire. If there are global timers to expire, the migrator locks the
remote CPU timer wheel and handles the expiry.
Depending on the group level in the hierarchy this handling can require
to walk the hierarchy downwards to the CPU level.
Special care is taken when the last CPU goes idle. At this point the
CPU is the systemwide migrator at the top of the hierarchy and it
therefore cannot delegate to the hierarchy. It needs to arm its own
timer device to expire either at the first expiring timer in the
hierarchy or at the first CPU local timer, which ever expires first.
This completely removes the overhead from the enqueue path, which is
e.g. for networking a true hotpath and trades it for a slightly more
complex idle path.
This has been in development for a couple of years and the final series
has been extensively tested by various teams from silicon vendors and
ran through extensive CI.
There have been slight performance improvements observed on network
centric workloads and an Intel team confirmed that this allows them to
power down a die completely on a mult-die socket for the first time in
a mostly idle scenario.
There is only one outstanding ~1.5% regression on a specific overloaded
netperf test which is currently investigated, but the rest is either
positive or neutral performance wise and positive on the power
management side.
- Fixes for the timekeeping interpolation code for cross-timestamps:
cross-timestamps are used for PTP to get snapshots from hardware timers
and interpolated them back to clock MONOTONIC. The changes address a
few corner cases in the interpolation code which got the math and logic
wrong.
- Simplifcation of the clocksource watchdog retry logic to automatically
adjust to handle larger systems correctly instead of having more
incomprehensible command line parameters.
- Treewide consolidation of the VDSO data structures.
- The usual small improvements and cleanups all over the place.
-----BEGIN PGP SIGNATURE-----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=iQlu
-----END PGP SIGNATURE-----
Merge tag 'timers-core-2024-03-10' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer updates from Thomas Gleixner:
"A large set of updates and features for timers and timekeeping:
- The hierarchical timer pull model
When timer wheel timers are armed they are placed into the timer
wheel of a CPU which is likely to be busy at the time of expiry.
This is done to avoid wakeups on potentially idle CPUs.
This is wrong in several aspects:
1) The heuristics to select the target CPU are wrong by
definition as the chance to get the prediction right is
close to zero.
2) Due to #1 it is possible that timers are accumulated on
a single target CPU
3) The required computation in the enqueue path is just overhead
for dubious value especially under the consideration that the
vast majority of timer wheel timers are either canceled or
rearmed before they expire.
The timer pull model avoids the above by removing the target
computation on enqueue and queueing timers always on the CPU on
which they get armed.
This is achieved by having separate wheels for CPU pinned timers
and global timers which do not care about where they expire.
As long as a CPU is busy it handles both the pinned and the global
timers which are queued on the CPU local timer wheels.
When a CPU goes idle it evaluates its own timer wheels:
- If the first expiring timer is a pinned timer, then the global
timers can be ignored as the CPU will wake up before they
expire.
- If the first expiring timer is a global timer, then the expiry
time is propagated into the timer pull hierarchy and the CPU
makes sure to wake up for the first pinned timer.
The timer pull hierarchy organizes CPUs in groups of eight at the
lowest level and at the next levels groups of eight groups up to
the point where no further aggregation of groups is required, i.e.
the number of levels is log8(NR_CPUS). The magic number of eight
has been established by experimention, but can be adjusted if
needed.
In each group one busy CPU acts as the migrator. It's only one CPU
to avoid lock contention on remote timer wheels.
The migrator CPU checks in its own timer wheel handling whether
there are other CPUs in the group which have gone idle and have
global timers to expire. If there are global timers to expire, the
migrator locks the remote CPU timer wheel and handles the expiry.
Depending on the group level in the hierarchy this handling can
require to walk the hierarchy downwards to the CPU level.
Special care is taken when the last CPU goes idle. At this point
the CPU is the systemwide migrator at the top of the hierarchy and
it therefore cannot delegate to the hierarchy. It needs to arm its
own timer device to expire either at the first expiring timer in
the hierarchy or at the first CPU local timer, which ever expires
first.
This completely removes the overhead from the enqueue path, which
is e.g. for networking a true hotpath and trades it for a slightly
more complex idle path.
This has been in development for a couple of years and the final
series has been extensively tested by various teams from silicon
vendors and ran through extensive CI.
There have been slight performance improvements observed on network
centric workloads and an Intel team confirmed that this allows them
to power down a die completely on a mult-die socket for the first
time in a mostly idle scenario.
There is only one outstanding ~1.5% regression on a specific
overloaded netperf test which is currently investigated, but the
rest is either positive or neutral performance wise and positive on
the power management side.
- Fixes for the timekeeping interpolation code for cross-timestamps:
cross-timestamps are used for PTP to get snapshots from hardware
timers and interpolated them back to clock MONOTONIC. The changes
address a few corner cases in the interpolation code which got the
math and logic wrong.
- Simplifcation of the clocksource watchdog retry logic to
automatically adjust to handle larger systems correctly instead of
having more incomprehensible command line parameters.
- Treewide consolidation of the VDSO data structures.
- The usual small improvements and cleanups all over the place"
* tag 'timers-core-2024-03-10' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (62 commits)
timer/migration: Fix quick check reporting late expiry
tick/sched: Fix build failure for CONFIG_NO_HZ_COMMON=n
vdso/datapage: Quick fix - use asm/page-def.h for ARM64
timers: Assert no next dyntick timer look-up while CPU is offline
tick: Assume timekeeping is correctly handed over upon last offline idle call
tick: Shut down low-res tick from dying CPU
tick: Split nohz and highres features from nohz_mode
tick: Move individual bit features to debuggable mask accesses
tick: Move got_idle_tick away from common flags
tick: Assume the tick can't be stopped in NOHZ_MODE_INACTIVE mode
tick: Move broadcast cancellation up to CPUHP_AP_TICK_DYING
tick: Move tick cancellation up to CPUHP_AP_TICK_DYING
tick: Start centralizing tick related CPU hotplug operations
tick/sched: Don't clear ts::next_tick again in can_stop_idle_tick()
tick/sched: Rename tick_nohz_stop_sched_tick() to tick_nohz_full_stop_tick()
tick: Use IS_ENABLED() whenever possible
tick/sched: Remove useless oneshot ifdeffery
tick/nohz: Remove duplicate between lowres and highres handlers
tick/nohz: Remove duplicate between tick_nohz_switch_to_nohz() and tick_setup_sched_timer()
hrtimer: Select housekeeping CPU during migration
...
So far, get_device_system_crosststamp() unconditionally passes
system_counterval.cycles to timekeeping_cycles_to_ns(). But when
interpolating system time (do_interp == true), system_counterval.cycles is
before tkr_mono.cycle_last, contrary to the timekeeping_cycles_to_ns()
expectations.
On x86, CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE will mitigate on
interpolating, setting delta to 0. With delta == 0, xtstamp->sys_monoraw
and xtstamp->sys_realtime are then set to the last update time, as
implicitly expected by adjust_historical_crosststamp(). On other
architectures, the resulting nonsense xtstamp->sys_monoraw and
xtstamp->sys_realtime corrupt the xtstamp (ts) adjustment in
adjust_historical_crosststamp().
Fix this by deriving xtstamp->sys_monoraw and xtstamp->sys_realtime from
the last update time when interpolating, by using the local variable
"cycles". The local variable already has the right value when
interpolating, unlike system_counterval.cycles.
Fixes: 2c756feb18 ("time: Add history to cross timestamp interface supporting slower devices")
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <jstultz@google.com>
Link: https://lore.kernel.org/r/20231218073849.35294-4-peter.hilber@opensynergy.com
The cycle_between() helper checks if parameter test is in the open interval
(before, after). Colloquially speaking, this also applies to the counter
wrap-around special case before > after. get_device_system_crosststamp()
currently uses cycle_between() at the first call site to decide whether to
interpolate for older counter readings.
get_device_system_crosststamp() has the following problem with
cycle_between() testing against an open interval: Assume that, by chance,
cycles == tk->tkr_mono.cycle_last (in the following, "cycle_last" for
brevity). Then, cycle_between() at the first call site, with effective
argument values cycle_between(cycle_last, cycles, now), returns false,
enabling interpolation. During interpolation,
get_device_system_crosststamp() will then call cycle_between() at the
second call site (if a history_begin was supplied). The effective argument
values are cycle_between(history_begin->cycles, cycles, cycles), since
system_counterval.cycles == interval_start == cycles, per the assumption.
Due to the test against the open interval, cycle_between() returns false
again. This causes get_device_system_crosststamp() to return -EINVAL.
This failure should be avoided, since get_device_system_crosststamp() works
both when cycles follows cycle_last (no interpolation), and when cycles
precedes cycle_last (interpolation). For the case cycles == cycle_last,
interpolation is actually unneeded.
Fix this by changing cycle_between() into timestamp_in_interval(), which
now checks against the closed interval, rather than the open interval.
This changes the get_device_system_crosststamp() behavior for three corner
cases:
1. Bypass interpolation in the case cycles == tk->tkr_mono.cycle_last,
fixing the problem described above.
2. At the first timestamp_in_interval() call site, cycles == now no longer
causes failure.
3. At the second timestamp_in_interval() call site, history_begin->cycles
== system_counterval.cycles no longer causes failure.
adjust_historical_crosststamp() also works for this corner case,
where partial_history_cycles == total_history_cycles.
These behavioral changes should not cause any problems.
Fixes: 2c756feb18 ("time: Add history to cross timestamp interface supporting slower devices")
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20231218073849.35294-3-peter.hilber@opensynergy.com
cycle_between() decides whether get_device_system_crosststamp() will
interpolate for older counter readings.
cycle_between() yields wrong results for a counter wrap-around where after
< before < test, and for the case after < test < before.
Fix the comparison logic.
Fixes: 2c756feb18 ("time: Add history to cross timestamp interface supporting slower devices")
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <jstultz@google.com>
Link: https://lore.kernel.org/r/20231218073849.35294-2-peter.hilber@opensynergy.com
Clocksource pointers can be problematic to obtain for drivers which are not
clocksource drivers themselves. In particular, the RFC virtio_rtc driver
[1] would require a new helper function to obtain a pointer to the ARM
Generic Timer clocksource. The ptp_kvm driver also required a similar
workaround.
Address this by evaluating the clocksource ID, rather than the clocksource
pointer, of struct system_counterval_t. By this, setting the clocksource
pointer becomes unneeded, and get_device_system_crosststamp() callers will
no longer need to supply clocksource pointers.
All relevant clocksource drivers provide the ID, so this change is not
changing the behaviour.
[1] https://lore.kernel.org/lkml/20231218073849.35294-1-peter.hilber@opensynergy.com/
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240201010453.2212371-7-peter.hilber@opensynergy.com
The read side of seqcount_latch consists of:
do {
seq = raw_read_seqcount_latch(&latch->seq);
...
} while (read_seqcount_latch_retry(&latch->seq, seq));
which is asymmetric in the raw_ department, and sure enough,
read_seqcount_latch_retry() includes (explicit) instrumentation where
raw_read_seqcount_latch() does not.
This inconsistency becomes a problem when trying to use it from
noinstr code. As such, fix it by renaming and re-implementing
raw_read_seqcount_latch_retry() without the instrumentation.
Specifically the instrumentation in question is kcsan_atomic_next(0)
in do___read_seqcount_retry(). Loosing this annotation is not a
problem because raw_read_seqcount_latch() does not pass through
kcsan_atomic_next(KCSAN_SEQLOCK_REGION_MAX).
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Tested-by: Michael Kelley <mikelley@microsoft.com> # Hyper-V
Link: https://lore.kernel.org/r/20230519102715.233598176@infradead.org
There was never a function named ktime_get_fast_ns().
Presumably these should refer to ktime_get_mono_fast_ns() instead.
Fixes: c1ce406e80 ("timekeeping: Fix up function documentation for the NMI safe accessors")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <jstultz@google.com>
Link: https://lore.kernel.org/r/06df7b3cbd94f016403bbf6cd2b38e4368e7468f.1682516546.git.geert+renesas@glider.be
Clean up kernel-doc complaints about function names and non-kernel-doc
comments in kernel/time/. Fixes these warnings:
kernel/time/time.c:479: warning: expecting prototype for set_normalized_timespec(). Prototype was for set_normalized_timespec64() instead
kernel/time/time.c:553: warning: expecting prototype for msecs_to_jiffies(). Prototype was for __msecs_to_jiffies() instead
kernel/time/timekeeping.c:1595: warning: contents before sections
kernel/time/timekeeping.c:1705: warning: This comment starts with '/**', but isn't a kernel-doc comment.
* We have three kinds of time sources to use for sleep time
kernel/time/timekeeping.c:1726: warning: This comment starts with '/**', but isn't a kernel-doc comment.
* 1) can be determined whether to use or not only when doing
kernel/time/tick-oneshot.c:21: warning: missing initial short description on line:
* tick_program_event
kernel/time/tick-oneshot.c:107: warning: expecting prototype for tick_check_oneshot_mode(). Prototype was for tick_oneshot_mode_active() instead
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20230103032849.12723-1-rdunlap@infradead.org
The rng's random_init() function contributes the real time to the rng at
boot time, so that events can at least start in relation to something
particular in the real world. But this clock might not yet be set that
point in boot, so nothing is contributed. In addition, the relation
between minor clock changes from, say, NTP, and the cycle counter is
potentially useful entropic data.
This commit addresses this by mixing in a time stamp on calls to
settimeofday and adjtimex. No entropy is credited in doing so, so it
doesn't make initialization faster, but it is still useful input to
have.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
-----BEGIN PGP SIGNATURE-----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=SUr3
-----END PGP SIGNATURE-----
Merge tag 'random-5.19-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random
Pull random number generator updates from Jason Donenfeld:
"These updates continue to refine the work began in 5.17 and 5.18 of
modernizing the RNG's crypto and streamlining and documenting its
code.
New for 5.19, the updates aim to improve entropy collection methods
and make some initial decisions regarding the "premature next" problem
and our threat model. The cloc utility now reports that random.c is
931 lines of code and 466 lines of comments, not that basic metrics
like that mean all that much, but at the very least it tells you that
this is very much a manageable driver now.
Here's a summary of the various updates:
- The random_get_entropy() function now always returns something at
least minimally useful. This is the primary entropy source in most
collectors, which in the best case expands to something like RDTSC,
but prior to this change, in the worst case it would just return 0,
contributing nothing. For 5.19, additional architectures are wired
up, and architectures that are entirely missing a cycle counter now
have a generic fallback path, which uses the highest resolution
clock available from the timekeeping subsystem.
Some of those clocks can actually be quite good, despite the CPU
not having a cycle counter of its own, and going off-core for a
stamp is generally thought to increase jitter, something positive
from the perspective of entropy gathering. Done very early on in
the development cycle, this has been sitting in next getting some
testing for a while now and has relevant acks from the archs, so it
should be pretty well tested and fine, but is nonetheless the thing
I'll be keeping my eye on most closely.
- Of particular note with the random_get_entropy() improvements is
MIPS, which, on CPUs that lack the c0 count register, will now
combine the high-speed but short-cycle c0 random register with the
lower-speed but long-cycle generic fallback path.
- With random_get_entropy() now always returning something useful,
the interrupt handler now collects entropy in a consistent
construction.
- Rather than comparing two samples of random_get_entropy() for the
jitter dance, the algorithm now tests many samples, and uses the
amount of differing ones to determine whether or not jitter entropy
is usable and how laborious it must be. The problem with comparing
only two samples was that if the cycle counter was extremely slow,
but just so happened to be on the cusp of a change, the slowness
wouldn't be detected. Taking many samples fixes that to some
degree.
This, combined with the other improvements to random_get_entropy(),
should make future unification of /dev/random and /dev/urandom
maybe more possible. At the very least, were we to attempt it again
today (we're not), it wouldn't break any of Guenter's test rigs
that broke when we tried it with 5.18. So, not today, but perhaps
down the road, that's something we can revisit.
- We attempt to reseed the RNG immediately upon waking up from system
suspend or hibernation, making use of the various timestamps about
suspend time and such available, as well as the usual inputs such
as RDRAND when available.
- Batched randomness now falls back to ordinary randomness before the
RNG is initialized. This provides more consistent guarantees to the
types of random numbers being returned by the various accessors.
- The "pre-init injection" code is now gone for good. I suspect you
in particular will be happy to read that, as I recall you
expressing your distaste for it a few months ago. Instead, to avoid
a "premature first" issue, while still allowing for maximal amount
of entropy availability during system boot, the first 128 bits of
estimated entropy are used immediately as it arrives, with the next
128 bits being buffered. And, as before, after the RNG has been
fully initialized, it winds up reseeding anyway a few seconds later
in most cases. This resulted in a pretty big simplification of the
initialization code and let us remove various ad-hoc mechanisms
like the ugly crng_pre_init_inject().
- The RNG no longer pretends to handle the "premature next" security
model, something that various academics and other RNG designs have
tried to care about in the past. After an interesting mailing list
thread, these issues are thought to be a) mainly academic and not
practical at all, and b) actively harming the real security of the
RNG by delaying new entropy additions after a potential compromise,
making a potentially bad situation even worse. As well, in the
first place, our RNG never even properly handled the premature next
issue, so removing an incomplete solution to a fake problem was
particularly nice.
This allowed for numerous other simplifications in the code, which
is a lot cleaner as a consequence. If you didn't see it before,
https://lore.kernel.org/lkml/YmlMGx6+uigkGiZ0@zx2c4.com/ may be a
thread worth skimming through.
- While the interrupt handler received a separate code path years ago
that avoids locks by using per-cpu data structures and a faster
mixing algorithm, in order to reduce interrupt latency, input and
disk events that are triggered in hardirq handlers were still
hitting locks and more expensive algorithms. Those are now
redirected to use the faster per-cpu data structures.
- Rather than having the fake-crypto almost-siphash-based random32
implementation be used right and left, and in many places where
cryptographically secure randomness is desirable, the batched
entropy code is now fast enough to replace that.
- As usual, numerous code quality and documentation cleanups. For
example, the initialization state machine now uses enum symbolic
constants instead of just hard coding numbers everywhere.
- Since the RNG initializes once, and then is always initialized
thereafter, a pretty heavy amount of code used during that
initialization is never used again. It is now completely cordoned
off using static branches and it winds up in the .text.unlikely
section so that it doesn't reduce cache compactness after the RNG
is ready.
- A variety of functions meant for waiting on the RNG to be
initialized were only used by vsprintf, and in not a particularly
optimal way. Replacing that usage with a more ordinary setup made
it possible to remove those functions.
- A cleanup of how we warn userspace about the use of uninitialized
/dev/urandom and uninitialized get_random_bytes() usage.
Interestingly, with the change you merged for 5.18 that attempts to
use jitter (but does not block if it can't), the majority of users
should never see those warnings for /dev/urandom at all now, and
the one for in-kernel usage is mainly a debug thing.
- The file_operations struct for /dev/[u]random now implements
.read_iter and .write_iter instead of .read and .write, allowing it
to also implement .splice_read and .splice_write, which makes
splice(2) work again after it was broken here (and in many other
places in the tree) during the set_fs() removal. This was a bit of
a last minute arrival from Jens that hasn't had as much time to
bake, so I'll be keeping my eye on this as well, but it seems
fairly ordinary. Unfortunately, read_iter() is around 3% slower
than read() in my tests, which I'm not thrilled about. But Jens and
Al, spurred by this observation, seem to be making progress in
removing the bottlenecks on the iter paths in the VFS layer in
general, which should remove the performance gap for all drivers.
- Assorted other bug fixes, cleanups, and optimizations.
- A small SipHash cleanup"
* tag 'random-5.19-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random: (49 commits)
random: check for signals after page of pool writes
random: wire up fops->splice_{read,write}_iter()
random: convert to using fops->write_iter()
random: convert to using fops->read_iter()
random: unify batched entropy implementations
random: move randomize_page() into mm where it belongs
random: remove mostly unused async readiness notifier
random: remove get_random_bytes_arch() and add rng_has_arch_random()
random: move initialization functions out of hot pages
random: make consistent use of buf and len
random: use proper return types on get_random_{int,long}_wait()
random: remove extern from functions in header
random: use static branch for crng_ready()
random: credit architectural init the exact amount
random: handle latent entropy and command line from random_init()
random: use proper jiffies comparison macro
random: remove ratelimiting for in-kernel unseeded randomness
random: move initialization out of reseeding hot path
random: avoid initializing twice in credit race
random: use symbolic constants for crng_init states
...
- Expose CLOCK_TAI to instrumentation to aid with TSN debugging.
- Ensure that the clockevent is stopped when there is no timer armed to
avoid pointless wakeups.
- Make the sched clock frequency handling and rounding consistent.
- Provide a better debugobject hint for delayed works. The timer callback
is always the same, which makes it difficult to identify the underlying
work. Use the work function as a hint instead.
- Move the timer specific sysctl code into the timer subsystem.
- The usual set of improvements and cleanups
-----BEGIN PGP SIGNATURE-----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=De28
-----END PGP SIGNATURE-----
Merge tag 'timers-core-2022-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer and timekeeping updates from Thomas Gleixner:
- Expose CLOCK_TAI to instrumentation to aid with TSN debugging.
- Ensure that the clockevent is stopped when there is no timer armed to
avoid pointless wakeups.
- Make the sched clock frequency handling and rounding consistent.
- Provide a better debugobject hint for delayed works. The timer
callback is always the same, which makes it difficult to identify the
underlying work. Use the work function as a hint instead.
- Move the timer specific sysctl code into the timer subsystem.
- The usual set of improvements and cleanups
* tag 'timers-core-2022-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timers: Provide a better debugobjects hint for delayed works
time/sched_clock: Fix formatting of frequency reporting code
time/sched_clock: Use Hz as the unit for clock rate reporting below 4kHz
time/sched_clock: Round the frequency reported to nearest rather than down
timekeeping: Consolidate fast timekeeper
timekeeping: Annotate ktime_get_boot_fast_ns() with data_race()
timers/nohz: Switch to ONESHOT_STOPPED in the low-res handler when the tick is stopped
timekeeping: Introduce fast accessor to clock tai
tracing/timer: Add missing argument documentation of trace points
clocksource: Replace cpumask_weight() with cpumask_empty()
timers: Move timer sysctl into the timer code
clockevents: Use dedicated list iterator variable
timers: Simplify calc_index()
timers: Initialize base::next_expiry_recalc in timers_prepare_cpu()
The addition of random_get_entropy_fallback() provides access to
whichever time source has the highest frequency, which is useful for
gathering entropy on platforms without available cycle counters. It's
not necessarily as good as being able to quickly access a cycle counter
that the CPU has, but it's still something, even when it falls back to
being jiffies-based.
In the event that a given arch does not define get_cycles(), falling
back to the get_cycles() default implementation that returns 0 is really
not the best we can do. Instead, at least calling
random_get_entropy_fallback() would be preferable, because that always
needs to return _something_, even falling back to jiffies eventually.
It's not as though random_get_entropy_fallback() is super high precision
or guaranteed to be entropic, but basically anything that's not zero all
the time is better than returning zero all the time.
Finally, since random_get_entropy_fallback() is used during extremely
early boot when randomizing freelists in mm_init(), it can be called
before timekeeping has been initialized. In that case there really is
nothing we can do; jiffies hasn't even started ticking yet. So just give
up and return 0.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Theodore Ts'o <tytso@mit.edu>
Accessing timekeeper::offset_boot in ktime_get_boot_fast_ns() is an
intended data race as the reader side cannot synchronize with a writer and
there is no space in struct tk_read_base of the NMI safe timekeeper.
Mark it so.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20220415091920.956045162@linutronix.de
Introduce fast/NMI safe accessor to clock tai for tracing. The Linux kernel
tracing infrastructure has support for using different clocks to generate
timestamps for trace events. Especially in TSN networks it's useful to have TAI
as trace clock, because the application scheduling is done in accordance to the
network time, which is based on TAI. With a tai trace_clock in place, it becomes
very convenient to correlate network activity with Linux kernel application
traces.
Use the same implementation as ktime_get_boot_fast_ns() does by reading the
monotonic time and adding the TAI offset. The same limitations as for the fast
boot implementation apply. The TAI offset may change at run time e.g., by
setting the time or using adjtimex() with an offset. However, these kind of
offset changes are rare events. Nevertheless, the user has to be aware and deal
with it in post processing.
An alternative approach would be to use the same implementation as
ktime_get_real_fast_ns() does. However, this requires to add an additional u64
member to the tk_read_base struct. This struct together with a seqcount is
designed to fit into a single cache line on 64 bit architectures. Adding a new
member would violate this constraint.
Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20220414091805.89667-2-kurt@linutronix.de
Even after commit e1d7ba8735 ("time: Always make sure wall_to_monotonic
isn't positive") it is still possible to make wall_to_monotonic positive
by running the following code:
int main(void)
{
struct timespec time;
clock_gettime(CLOCK_MONOTONIC, &time);
time.tv_nsec = 0;
clock_settime(CLOCK_REALTIME, &time);
return 0;
}
The reason is that the second parameter of timespec64_compare(), ts_delta,
may be unnormalized because the delta is calculated with an open coded
substraction which causes the comparison of tv_sec to yield the wrong
result:
wall_to_monotonic = { .tv_sec = -10, .tv_nsec = 900000000 }
ts_delta = { .tv_sec = -9, .tv_nsec = -900000000 }
That makes timespec64_compare() claim that wall_to_monotonic < ts_delta,
but actually the result should be wall_to_monotonic > ts_delta.
After normalization, the result of timespec64_compare() is correct because
the tv_sec comparison is not longer misleading:
wall_to_monotonic = { .tv_sec = -10, .tv_nsec = 900000000 }
ts_delta = { .tv_sec = -10, .tv_nsec = 100000000 }
Use timespec64_sub() to ensure that ts_delta is normalized, which fixes the
issue.
Fixes: e1d7ba8735 ("time: Always make sure wall_to_monotonic isn't positive")
Signed-off-by: Yu Liao <liaoyu15@huawei.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211213135727.1656662-1-liaoyu15@huawei.com
clock_was_set() unconditionaly invokes retrigger_next_event() on all online
CPUs. This was necessary because that mechanism was also used for resume
from suspend to idle which is not longer the case.
The bases arguments allows the callers of clock_was_set() to hand in a mask
which tells clock_was_set() which of the hrtimer clock bases are affected
by the clock setting. This mask will be used in the next step to check
whether a CPU base has timers queued on a clock base affected by the event
and avoid the SMP function call if there are none.
Add a @bases argument, provide defines for the active bases masking and
fixup all callsites.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20210713135158.691083465@linutronix.de
do_adjtimex() might end up scheduling a delayed clock_was_set() via
timekeeping_advance() and then invoke clock_was_set() directly which is
pointless.
Make timekeeping_advance() return whether an invocation of clock_was_set()
is required and handle it at the call sites which allows do_adjtimex() to
issue a single direct call if required.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20210713135158.580966888@linutronix.de
Resuming timekeeping is a clock-was-set event and uses the clock-was-set
notification mechanism. This is in the way of making the clock-was-set
update for hrtimers selective so unnecessary IPIs are avoided when a CPU
base does not have timers queued which are affected by the clock setting.
Distangle it by invoking hrtimer_resume() on each unfreezing CPU and invoke
the new timerfd_resume() function from timekeeping_resume() which is the
only place where this is needed.
Rename hrtimer_resume() to hrtimer_resume_local() to reflect the change.
With this the clock_was_set*() functions are not longer required to IPI all
CPUs unconditionally and can get some smarts to avoid them.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20210713135158.488853478@linutronix.de
- Stage-2 isolation for the host kernel when running in protected mode
- Guest SVE support when running in nVHE mode
- Force W^X hypervisor mappings in nVHE mode
- ITS save/restore for guests using direct injection with GICv4.1
- nVHE panics now produce readable backtraces
- Guest support for PTP using the ptp_kvm driver
- Performance improvements in the S2 fault handler
x86:
- Optimizations and cleanup of nested SVM code
- AMD: Support for virtual SPEC_CTRL
- Optimizations of the new MMU code: fast invalidation,
zap under read lock, enable/disably dirty page logging under
read lock
- /dev/kvm API for AMD SEV live migration (guest API coming soon)
- support SEV virtual machines sharing the same encryption context
- support SGX in virtual machines
- add a few more statistics
- improved directed yield heuristics
- Lots and lots of cleanups
Generic:
- Rework of MMU notifier interface, simplifying and optimizing
the architecture-specific code
- Some selftests improvements
-----BEGIN PGP SIGNATURE-----
iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAmCJ13kUHHBib256aW5p
QHJlZGhhdC5jb20ACgkQv/vSX3jHroM1HAgAqzPxEtiTPTFeFJV5cnPPJ3dFoFDK
y/juZJUQ1AOtvuWzzwuf175ewkv9vfmtG6rVohpNSkUlJYeoc6tw7n8BTTzCVC1b
c/4Dnrjeycr6cskYlzaPyV6MSgjSv5gfyj1LA5UEM16LDyekmaynosVWY5wJhju+
Bnyid8l8Utgz+TLLYogfQJQECCrsU0Wm//n+8TWQgLf1uuiwshU5JJe7b43diJrY
+2DX+8p9yWXCTz62sCeDWNahUv8AbXpMeJ8uqZPYcN1P0gSEUGu8xKmLOFf9kR7b
M4U1Gyz8QQbjd2lqnwiWIkvRLX6gyGVbq2zH0QbhUe5gg3qGUX7JjrhdDQ==
=AXUi
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm updates from Paolo Bonzini:
"This is a large update by KVM standards, including AMD PSP (Platform
Security Processor, aka "AMD Secure Technology") and ARM CoreSight
(debug and trace) changes.
ARM:
- CoreSight: Add support for ETE and TRBE
- Stage-2 isolation for the host kernel when running in protected
mode
- Guest SVE support when running in nVHE mode
- Force W^X hypervisor mappings in nVHE mode
- ITS save/restore for guests using direct injection with GICv4.1
- nVHE panics now produce readable backtraces
- Guest support for PTP using the ptp_kvm driver
- Performance improvements in the S2 fault handler
x86:
- AMD PSP driver changes
- Optimizations and cleanup of nested SVM code
- AMD: Support for virtual SPEC_CTRL
- Optimizations of the new MMU code: fast invalidation, zap under
read lock, enable/disably dirty page logging under read lock
- /dev/kvm API for AMD SEV live migration (guest API coming soon)
- support SEV virtual machines sharing the same encryption context
- support SGX in virtual machines
- add a few more statistics
- improved directed yield heuristics
- Lots and lots of cleanups
Generic:
- Rework of MMU notifier interface, simplifying and optimizing the
architecture-specific code
- a handful of "Get rid of oprofile leftovers" patches
- Some selftests improvements"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (379 commits)
KVM: selftests: Speed up set_memory_region_test
selftests: kvm: Fix the check of return value
KVM: x86: Take advantage of kvm_arch_dy_has_pending_interrupt()
KVM: SVM: Skip SEV cache flush if no ASIDs have been used
KVM: SVM: Remove an unnecessary prototype declaration of sev_flush_asids()
KVM: SVM: Drop redundant svm_sev_enabled() helper
KVM: SVM: Move SEV VMCB tracking allocation to sev.c
KVM: SVM: Explicitly check max SEV ASID during sev_hardware_setup()
KVM: SVM: Unconditionally invoke sev_hardware_teardown()
KVM: SVM: Enable SEV/SEV-ES functionality by default (when supported)
KVM: SVM: Condition sev_enabled and sev_es_enabled on CONFIG_KVM_AMD_SEV=y
KVM: SVM: Append "_enabled" to module-scoped SEV/SEV-ES control variables
KVM: SEV: Mask CPUID[0x8000001F].eax according to supported features
KVM: SVM: Move SEV module params/variables to sev.c
KVM: SVM: Disable SEV/SEV-ES if NPT is disabled
KVM: SVM: Free sev_asid_bitmap during init if SEV setup fails
KVM: SVM: Zero out the VMCB array used to track SEV ASID association
x86/sev: Drop redundant and potentially misleading 'sev_enabled'
KVM: x86: Move reverse CPUID helpers to separate header file
KVM: x86: Rename GPR accessors to make mode-aware variants the defaults
...
System time snapshots are not conveying information about the current
clocksource which was used, but callers like the PTP KVM guest
implementation have the requirement to evaluate the clocksource type to
select the appropriate mechanism.
Introduce a clocksource id field in struct clocksource which is by default
set to CSID_GENERIC (0). Clocksource implementations can set that field to
a value which allows to identify the clocksource.
Store the clocksource id of the current clocksource in the
system_time_snapshot so callers can evaluate which clocksource was used to
take the snapshot and act accordingly.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20201209060932.212364-5-jianyong.wu@arm.com
The struct clocksource callbacks enable() and disable() are described as a
way to allow clock sources to enter a power save mode. See commit
4614e6adaf ("clocksource: add enable() and disable() callbacks")
But using runtime PM from these callbacks triggers a cyclic lockdep warning when
switching clock source using change_clocksource().
# echo e60f0000.timer > /sys/devices/system/clocksource/clocksource0/current_clocksource
======================================================
WARNING: possible circular locking dependency detected
------------------------------------------------------
migration/0/11 is trying to acquire lock:
ffff0000403ed220 (&dev->power.lock){-...}-{2:2}, at: __pm_runtime_resume+0x40/0x74
but task is already holding lock:
ffff8000113c8f88 (tk_core.seq.seqcount){----}-{0:0}, at: multi_cpu_stop+0xa4/0x190
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #2 (tk_core.seq.seqcount){----}-{0:0}:
ktime_get+0x28/0xa0
hrtimer_start_range_ns+0x210/0x2dc
generic_sched_clock_init+0x70/0x88
sched_clock_init+0x40/0x64
start_kernel+0x494/0x524
-> #1 (hrtimer_bases.lock){-.-.}-{2:2}:
hrtimer_start_range_ns+0x68/0x2dc
rpm_suspend+0x308/0x5dc
rpm_idle+0xc4/0x2a4
pm_runtime_work+0x98/0xc0
process_one_work+0x294/0x6f0
worker_thread+0x70/0x45c
kthread+0x154/0x160
ret_from_fork+0x10/0x20
-> #0 (&dev->power.lock){-...}-{2:2}:
_raw_spin_lock_irqsave+0x7c/0xc4
__pm_runtime_resume+0x40/0x74
sh_cmt_start+0x1c4/0x260
sh_cmt_clocksource_enable+0x28/0x50
change_clocksource+0x9c/0x160
multi_cpu_stop+0xa4/0x190
cpu_stopper_thread+0x90/0x154
smpboot_thread_fn+0x244/0x270
kthread+0x154/0x160
ret_from_fork+0x10/0x20
other info that might help us debug this:
Chain exists of:
&dev->power.lock --> hrtimer_bases.lock --> tk_core.seq.seqcount
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(tk_core.seq.seqcount);
lock(hrtimer_bases.lock);
lock(tk_core.seq.seqcount);
lock(&dev->power.lock);
*** DEADLOCK ***
2 locks held by migration/0/11:
#0: ffff8000113c9278 (timekeeper_lock){-.-.}-{2:2}, at: change_clocksource+0x2c/0x160
#1: ffff8000113c8f88 (tk_core.seq.seqcount){----}-{0:0}, at: multi_cpu_stop+0xa4/0x190
Rework change_clocksource() so it enables the new clocksource and disables
the old clocksource outside of the timekeeper_lock and seqcount write held
region. There is no requirement that these callbacks are invoked from the
lock held region.
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Link: https://lore.kernel.org/r/20210211134318.323910-1-niklas.soderlund+renesas@ragnatech.se
This cleans up two ancient timer features that were never completed in
the past, CONFIG_GENERIC_CLOCKEVENTS and CONFIG_ARCH_USES_GETTIMEOFFSET.
There was only one user left for the ARCH_USES_GETTIMEOFFSET variant
of clocksource implementations, the ARM EBSA110 platform. Rather than
changing to use modern timekeeping, we remove the platform entirely as
Russell no longer uses his machine and nobody else seems to have one
any more.
The conditional code for using arch_gettimeoffset() is removed as
a result.
For CONFIG_GENERIC_CLOCKEVENTS, there are still a couple of platforms
not using clockevent drivers: parisc, ia64, most of m68k, and one
Arm platform. These all do timer ticks slighly differently, and this
gets cleaned up to the point they at least all call the same helper
function. Instead of most platforms using 'select GENERIC_CLOCKEVENTS'
in Kconfig, the polarity is now reversed, with the few remaining ones
selecting LEGACY_TIMER_TICK instead.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
-----BEGIN PGP SIGNATURE-----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=QpNs
-----END PGP SIGNATURE-----
Merge tag 'asm-generic-timers-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/asm-generic
Pull asm-generic cross-architecture timer cleanup from Arnd Bergmann:
"This cleans up two ancient timer features that were never completed in
the past, CONFIG_GENERIC_CLOCKEVENTS and CONFIG_ARCH_USES_GETTIMEOFFSET.
There was only one user left for the ARCH_USES_GETTIMEOFFSET variant
of clocksource implementations, the ARM EBSA110 platform. Rather than
changing to use modern timekeeping, we remove the platform entirely as
Russell no longer uses his machine and nobody else seems to have one
any more.
The conditional code for using arch_gettimeoffset() is removed as a
result.
For CONFIG_GENERIC_CLOCKEVENTS, there are still a couple of platforms
not using clockevent drivers: parisc, ia64, most of m68k, and one Arm
platform. These all do timer ticks slighly differently, and this gets
cleaned up to the point they at least all call the same helper
function.
Instead of most platforms using 'select GENERIC_CLOCKEVENTS' in
Kconfig, the polarity is now reversed, with the few remaining ones
selecting LEGACY_TIMER_TICK instead"
* tag 'asm-generic-timers-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/asm-generic:
timekeeping: default GENERIC_CLOCKEVENTS to enabled
timekeeping: remove xtime_update
m68k: remove timer_interrupt() function
m68k: change remaining timers to legacy_timer_tick
m68k: m68328: use legacy_timer_tick()
m68k: sun3/sun3c: use legacy_timer_tick
m68k: split heartbeat out of timer function
m68k: coldfire: use legacy_timer_tick()
parisc: use legacy_timer_tick
ARM: rpc: use legacy_timer_tick
ia64: convert to legacy_timer_tick
timekeeping: add CONFIG_LEGACY_TIMER_TICK
timekeeping: remove arch_gettimeoffset
net: remove am79c961a driver
ARM: remove ebsa110 platform
The kernel-doc parser complains:
kernel/time/timekeeping.c:1543: warning: Function parameter or member
'ts' not described in 'read_persistent_clock64'
kernel/time/timekeeping.c:764: warning: Function parameter or member
'tk' not described in 'timekeeping_forward_now'
kernel/time/timekeeping.c:1331: warning: Function parameter or member
'ts' not described in 'timekeeping_inject_offset'
kernel/time/timekeeping.c:1331: warning: Excess function parameter 'tv'
description in 'timekeeping_inject_offset'
Add the missing parameter documentations and rename the 'tv' parameter of
timekeeping_inject_offset() to 'ts' so it matches the implemention.
[ tglx: Reworded a few docs and massaged changelog ]
Signed-off-by: Alex Shi <alex.shi@linux.alibaba.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1605252275-63652-5-git-send-email-alex.shi@linux.alibaba.com
Address the following kernel-doc markup warnings:
kernel/time/timekeeping.c:1563: warning: Function parameter or member
'wall_time' not described in 'read_persistent_wall_and_boot_offset'
kernel/time/timekeeping.c:1563: warning: Function parameter or member
'boot_offset' not described in 'read_persistent_wall_and_boot_offset'
The parameters are described but miss the leading '@' and the colon after
the parameter names.
[ tglx: Massaged changelog ]
Signed-off-by: Alex Shi <alex.shi@linux.alibaba.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1605252275-63652-6-git-send-email-alex.shi@linux.alibaba.com
The kernel-doc parser complains about:
kernel/time/timekeeping.c:651: warning: Function parameter or member
'nb' not described in 'pvclock_gtod_register_notifier'
kernel/time/timekeeping.c:670: warning: Function parameter or member
'nb' not described in 'pvclock_gtod_unregister_notifier'
Add the missing parameter explanations.
[ tglx: Massaged changelog ]
Signed-off-by: Alex Shi <alex.shi@linux.alibaba.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1605252275-63652-3-git-send-email-alex.shi@linux.alibaba.com
Alex reported the following warning:
kernel/time/timekeeping.c:464: warning: Function parameter or member
'tkf' not described in '__ktime_get_fast_ns'
which is not entirely correct because the documented function is
ktime_get_mono_fast_ns() which does not have a parameter, but the
kernel-doc parser looks at the function declaration which follows the
comment and complains about the missing parameter documentation.
Aside of that the documentation for the rest of the NMI safe accessors is
either incomplete or missing.
- Move the function documentation to the right place
- Fixup the references and inconsistencies
- Add the missing documentation for ktime_get_raw_fast_ns()
Reported-by: Alex Shi <alex.shi@linux.alibaba.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Address the following warning:
kernel/time/timekeeping.c:415: warning: Function parameter or member
'tkf' not described in 'update_fast_timekeeper'
[ tglx: Remove the bogus ktime_get_mono_fast_ns() part ]
Signed-off-by: Alex Shi <alex.shi@linux.alibaba.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1605252275-63652-2-git-send-email-alex.shi@linux.alibaba.com
Various static functions in the timekeeping code have function comments
which pretend to be kernel-doc, but are incomplete and trigger parser
warnings.
As these functions are local to the timekeeping core code there is no need
to expose them via kernel-doc.
Remove the double star kernel-doc marker and remove excess newlines.
[ tglx: Massaged changelog and removed excess newlines ]
Signed-off-by: Alex Shi <alex.shi@linux.alibaba.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/1605252275-63652-4-git-send-email-alex.shi@linux.alibaba.com
There are no more users of xtime_update aside from legacy_timer_tick(),
so fold it into that function and remove the declaration.
update_process_times() is now only called inside of the kernel/time/
code, so the declaration can be moved there.
Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org>
Tested-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
With Arm EBSA110 gone, nothing uses it any more, so the corresponding
code and the Kconfig option can be removed.
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
- Add deadlock detection for recursive read-locks. The rationale is outlined
in:
224ec489d3: ("lockdep/Documention: Recursive read lock detection reasoning")
The main deadlock pattern we want to detect is:
TASK A: TASK B:
read_lock(X);
write_lock(X);
read_lock_2(X);
- Add "latch sequence counters" (seqcount_latch_t):
A sequence counter variant where the counter even/odd value is used to
switch between two copies of protected data. This allows the read path,
typically NMIs, to safely interrupt the write side critical section.
We utilize this new variant for sched-clock, and to make x86 TSC handling safer.
- Other seqlock cleanups, fixes and enhancements
- KCSAN updates
- LKMM updates
- Misc updates, cleanups and fixes.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
-----BEGIN PGP SIGNATURE-----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=Iq51
-----END PGP SIGNATURE-----
Merge tag 'locking-core-2020-10-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull locking updates from Ingo Molnar:
"These are the locking updates for v5.10:
- Add deadlock detection for recursive read-locks.
The rationale is outlined in commit 224ec489d3 ("lockdep/
Documention: Recursive read lock detection reasoning")
The main deadlock pattern we want to detect is:
TASK A: TASK B:
read_lock(X);
write_lock(X);
read_lock_2(X);
- Add "latch sequence counters" (seqcount_latch_t):
A sequence counter variant where the counter even/odd value is used
to switch between two copies of protected data. This allows the
read path, typically NMIs, to safely interrupt the write side
critical section.
We utilize this new variant for sched-clock, and to make x86 TSC
handling safer.
- Other seqlock cleanups, fixes and enhancements
- KCSAN updates
- LKMM updates
- Misc updates, cleanups and fixes"
* tag 'locking-core-2020-10-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (67 commits)
lockdep: Revert "lockdep: Use raw_cpu_*() for per-cpu variables"
lockdep: Fix lockdep recursion
lockdep: Fix usage_traceoverflow
locking/atomics: Check atomic-arch-fallback.h too
locking/seqlock: Tweak DEFINE_SEQLOCK() kernel doc
lockdep: Optimize the memory usage of circular queue
seqlock: Unbreak lockdep
seqlock: PREEMPT_RT: Do not starve seqlock_t writers
seqlock: seqcount_LOCKNAME_t: Introduce PREEMPT_RT support
seqlock: seqcount_t: Implement all read APIs as statement expressions
seqlock: Use unique prefix for seqcount_t property accessors
seqlock: seqcount_LOCKNAME_t: Standardize naming convention
seqlock: seqcount latch APIs: Only allow seqcount_latch_t
rbtree_latch: Use seqcount_latch_t
x86/tsc: Use seqcount_latch_t
timekeeping: Use seqcount_latch_t
time/sched_clock: Use seqcount_latch_t
seqlock: Introduce seqcount_latch_t
mm/swap: Do not abuse the seqcount_t latching API
time/sched_clock: Use raw_read_seqcount_latch() during suspend
...
Latch sequence counters are a multiversion concurrency control mechanism
where the seqcount_t counter even/odd value is used to switch between
two data storage copies. This allows the seqcount_t read path to safely
interrupt its write side critical section (e.g. from NMIs).
Initially, latch sequence counters were implemented as a single write
function, raw_write_seqcount_latch(), above plain seqcount_t. The read
path was expected to use plain seqcount_t raw_read_seqcount().
A specialized read function was later added, raw_read_seqcount_latch(),
and became the standardized way for latch read paths. Having unique read
and write APIs meant that latch sequence counters are basically a data
type of their own -- just inappropriately overloading plain seqcount_t.
The seqcount_latch_t data type was thus introduced at seqlock.h.
Use that new data type instead of seqcount_raw_spinlock_t. This ensures
that only latch-safe APIs are to be used with the sequence counter.
Note that the use of seqcount_raw_spinlock_t was not very useful in the
first place. Only the "raw_" subset of seqcount_t APIs were used at
timekeeping.c. This subset was created for contexts where lockdep cannot
be used. seqcount_LOCKTYPE_t's raison d'être -- verifying that the
seqcount_t writer serialization lock is held -- cannot thus be done.
References: 0c3351d451 ("seqlock: Use raw_ prefix instead of _no_lockdep")
References: 55f3560df9 ("seqlock: Extend seqcount API with associated locks")
Signed-off-by: Ahmed S. Darwish <a.darwish@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200827114044.11173-6-a.darwish@linutronix.de
printk wants to store various timestamps (MONOTONIC, REALTIME, BOOTTIME) to
make correlation of dmesg from several systems easier.
Provide an interface to retrieve all three timestamps in one go.
There are some caveats:
1) Boot time and late sleep time injection
Boot time is a racy access on 32bit systems if the sleep time injection
happens late during resume and not in timekeeping_resume(). That could be
avoided by expanding struct tk_read_base with boot offset for 32bit and
adding more overhead to the update. As this is a hard to observe once per
resume event which can be filtered with reasonable effort using the
accurate mono/real timestamps, it's probably not worth the trouble.
Aside of that it might be possible on 32 and 64 bit to observe the
following when the sleep time injection happens late:
CPU 0 CPU 1
timekeeping_resume()
ktime_get_fast_timestamps()
mono, real = __ktime_get_real_fast()
inject_sleep_time()
update boot offset
boot = mono + bootoffset;
That means that boot time already has the sleep time adjustment, but
real time does not. On the next readout both are in sync again.
Preventing this for 64bit is not really feasible without destroying the
careful cache layout of the timekeeper because the sequence count and
struct tk_read_base would then need two cache lines instead of one.
2) Suspend/resume timestamps
Access to the time keeper clock source is disabled accross the innermost
steps of suspend/resume. The accessors still work, but the timestamps
are frozen until time keeping is resumed which happens very early.
For regular suspend/resume there is no observable difference vs. sched
clock, but it might affect some of the nasty low level debug printks.
OTOH, access to sched clock is not guaranteed accross suspend/resume on
all systems either so it depends on the hardware in use.
If that turns out to be a real problem then this could be mitigated by
using sched clock in a similar way as during early boot. But it's not as
trivial as on early boot because it needs some careful protection
against the clock monotonic timestamp jumping backwards on resume.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20200814115512.159981360@linutronix.de
During early boot the NMI safe timekeeper returns 0 until the first
clocksource becomes available.
This prevents it from being used for printk or other facilities which today
use sched clock. sched clock can be available way before timekeeping is
initialized.
The obvious workaround for this is to utilize the early sched clock in the
default dummy clock read function until a clocksource becomes available.
After switching to the clocksource clock MONOTONIC and BOOTTIME will not
jump because the timekeeping_init() bases clock MONOTONIC on sched clock
and the offset between clock MONOTONIC and BOOTTIME is zero during boot.
Clock REALTIME cannot provide useful timestamps during early boot up to
the point where a persistent clock becomes available, which is either in
timekeeping_init() or later when the RTC driver which might depend on I2C
or other subsystems is initialized.
There is a minor difference to sched_clock() vs. suspend/resume. As the
timekeeper clock source might not be accessible during suspend, after
timekeeping_suspend() timestamps freeze up to the point where
timekeeping_resume() is invoked. OTOH this is true for some sched clock
implementations as well.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20200814115512.041422402@linutronix.de
