These changes prevent sysctl(8) from returning proper output,
such as:
1) no output from sysctl(8)
2) erroneously returning ENOMEM with tools like truss(1)
or uname(1)
truss: can not get etype: Cannot allocate memory
there is an environment variable which shall initialize the SYSCTL
during early boot. This works for all SYSCTL types both statically and
dynamically created ones, except for the SYSCTL NODE type and SYSCTLs
which belong to VNETs. A new flag, CTLFLAG_NOFETCH, has been added to
be used in the case a tunable sysctl has a custom initialisation
function allowing the sysctl to still be marked as a tunable. The
kernel SYSCTL API is mostly the same, with a few exceptions for some
special operations like iterating childrens of a static/extern SYSCTL
node. This operation should probably be made into a factored out
common macro, hence some device drivers use this. The reason for
changing the SYSCTL API was the need for a SYSCTL parent OID pointer
and not only the SYSCTL parent OID list pointer in order to quickly
generate the sysctl path. The motivation behind this patch is to avoid
parameter loading cludges inside the OFED driver subsystem. Instead of
adding special code to the OFED driver subsystem to post-load tunables
into dynamically created sysctls, we generalize this in the kernel.
Other changes:
- Corrected a possibly incorrect sysctl name from "hw.cbb.intr_mask"
to "hw.pcic.intr_mask".
- Removed redundant TUNABLE statements throughout the kernel.
- Some minor code rewrites in connection to removing not needed
TUNABLE statements.
- Added a missing SYSCTL_DECL().
- Wrapped two very long lines.
- Avoid malloc()/free() inside sysctl string handling, in case it is
called to initialize a sysctl from a tunable, hence malloc()/free() is
not ready when sysctls from the sysctl dataset are registered.
- Bumped FreeBSD version to indicate SYSCTL API change.
MFC after: 2 weeks
Sponsored by: Mellanox Technologies
Under enough load, the swi's can actually be preempted and migrated
to other currently free cores. When doing RSS experiments, this lead
to the per-CPU TCP timers not lining up any more with the RX CPU said
flows were ending up on, leading to increased lock contention.
Since there was a little pushback on flipping them on by default,
I've left the default at "don't pin."
The other less obvious problem here is that the default swi
is also the same as the destination swi for CPU #0. So if one
pins the swi on CPU #0, there's no default floating swi.
A nice future project would be to create a separate swi for
the "default" floating swi, as well as per-CPU swis that are
(optionally) pinned.
Tested:
* parallel TCP tests (2 x 1g unfortunately for now);
CPU: Intel(R) Xeon(R) CPU E5-2650
Note:
This is based on some initial investigation into RSS/TCP stack lock
contention on FreeBSD-HEAD whilst at Netflix in January 2014.
SBT_MAX, to make it more robust in case internal type representation will
change in the future. All the consumers were migrated to SBT_MAX and
every new consumer (if any) should from now use this interface.
Requested by: bapt, jmg, Ryan Lortie (implictly)
Reviewed by: mav, bde
In its stead use the Solaris / illumos approach of emulating '-' (dash)
in probe names with '__' (two consecutive underscores).
Reviewed by: markj
MFC after: 3 weeks
option, unbreak the lock tracing release semantic by embedding
calls to LOCKSTAT_PROFILE_RELEASE_LOCK() direclty in the inlined
version of the releasing functions for mutex, rwlock and sxlock.
Failing to do so skips the lockstat_probe_func invokation for
unlocking.
- As part of the LOCKSTAT support is inlined in mutex operation, for
kernel compiled without lock debugging options, potentially every
consumer must be compiled including opt_kdtrace.h.
Fix this by moving KDTRACE_HOOKS into opt_global.h and remove the
dependency by opt_kdtrace.h for all files, as now only KDTRACE_FRAMES
is linked there and it is only used as a compile-time stub [0].
[0] immediately shows some new bug as DTRACE-derived support for debug
in sfxge is broken and it was never really tested. As it was not
including correctly opt_kdtrace.h before it was never enabled so it
was kept broken for a while. Fix this by using a protection stub,
leaving sfxge driver authors the responsibility for fixing it
appropriately [1].
Sponsored by: EMC / Isilon storage division
Discussed with: rstone
[0] Reported by: rstone
[1] Discussed with: philip
Without these, if the timeout value passed is "large enough", the
value of the sum of it and other factors (e.g. current time as
returned by sbinuptime() or 'precision' argument) might result in a
negative number. This negative number is then passed to
eventtimers(4), which causes et_start() routine to load et_min_period
into eventtimer, making the CPU where the thread is stuck forever in
timer interrupt handler routine. This is now avoided rounding to
INT64_MAX the timeout period in case of overflow.
Reported by: kib, pho
Discussed with: kib, mav
Tested by: pho (stress2 suite, kevent7.sh scenario)
Approved by: re (kib)
rm_priotracker' directly in the softclock thread. Now consumers can
pass CALLOUT_SHAREDLOCK flag to callout initialization routine safely.
The choice of the already existing flags instead of special casing
shared rmlocks is done to prevent consumer footshooting.
Suggested by: jhb
Reviewed by: jhb
Approved by: re (delphij)
using SDT_PROBE_ARGTYPE(). This will make it easy to extend the SDT(9) API
to allow probes with dynamically-translated types.
There is no functional change.
MFC after: 2 weeks
for migrating callouts to new CPU. This value is passed to
callout_cc_add() in order to update properly precision field in case of
rescheduling/migration.
Reviewed by: mav
Change size requested to malloc(9) now that callwheel buckets are
callout_list and not callout_tailq anymore. This change was already
there but it seems it got lost after code churn in r248032.
Reported by: alc, kib
from being indirectly called via cpu_startup()+vm_ksubmap_init().
The boot order position remains the same at SI_SUB_CPU.
Allocation of the callout array is changed to stardard kernel malloc
from a slightly obscure direct kernel_map allocation.
kern_timeout_callwheel_alloc() is renamed to callout_callwheel_init()
to better describe its purpose.
kern_timeout_callwheel_init() is removed simplifying the per-cpu
initialization.
Reviewed by: davide
kern_timeout_callwheel_alloc() where it is actually used.
This is a mechanical move and no tuning parameters are changed.
The pre-allocated callout array is only used for legacy timeout(9)
calls and is only allocated and active on cpu0. Eventually all
remaining users of timeout(9) should switch to the callout_* API.
Reviewed by: davide
precise time event generation. This greatly improves granularity of
callouts which are not anymore constrained to wait next tick to be
scheduled.
- Extend the callout KPI introducing a set of callout_reset_sbt* functions,
which take a sbintime_t as timeout argument. The new KPI also offers a
way for consumers to specify precision tolerance they allow, so that
callout can coalesce events and reduce number of interrupts as well as
potentially avoid scheduling a SWI thread.
- Introduce support for dispatching callouts directly from hardware
interrupt context, specifying an additional flag. This feature should be
used carefully, as long as interrupt context has some limitations
(e.g. no sleeping locks can be held).
- Enhance mechanisms to gather informations about callwheel, introducing
a new sysctl to obtain stats.
This change breaks the KBI. struct callout fields has been changed, in
particular 'int ticks' (4 bytes) has been replaced with 'sbintime_t'
(8 bytes) and another 'sbintime_t' field was added for precision.
Together with: mav
Reviewed by: attilio, bde, luigi, phk
Sponsored by: Google Summer of Code 2012, iXsystems inc.
Tested by: flo (amd64, sparc64), marius (sparc64), ian (arm),
markj (amd64), mav, Fabian Keil
- As the comment report, CALLOUT_LOCAL_ALLOC cannot be checked
directly from the callout flags but might be checked by a cached
value. Hence, do so before to actually remove the callout, when
needed, in softclock_call_cc().
- In softclock_call_cc() also add a comment in the waiting and deferred
migration case explaining that the dereference should be safe
because of the migration dereference invariants.
Additively:
- In softclock_call_cc(), for the deferred migration case, move all the
accesses to callout structure after the comment stating the callout
must not be destroyed.
- For consistency with this last tweak, use cached c_flags for the
KASSERT() in the deferred migration case. It is not strictly necessary
but this way all the callout accesses happen after the above mentioned
comment, improving consistency.
Pointy hat to: me
Sponsored by: Isilon Systems / EMC Corporation
Reviewed by: kib
MFC after: 2 weeks
X-MFC: 243901
from the callwheel. Calculate the cc->cc_next before removing the
callout, otherwise the code followed the invalid tailq links. After
this, make softclock_call_cc() return void, since it always return
cc->cc_next, which is immediately available to the softclock()
anyway. This also allows to eliminate a label under #ifdef SMP.
Remove the assignment of cc->cc_next from callout_cc_del(), since the
function is called with the callout already removed from callwheel.
If cancelling the migration, also clear the CALLOUT_DFRMIGRATION flag.
Postpone the free of the timeout(9) allocated callouts after the
migration checks are done.
Add some more strict asserts about the state of the callout in
callout_call_cc().
Reviewed by: attilio
Reported and tested by: pho (previous version)
MFC after: 2 weeks
cache line in order to avoid manual frobbing but using
struct mtx_padalign.
The sole exception being nvme and sxfge drivers, where the author
redefined CACHE_LINE_SIZE manually, so they need to be analyzed and
dealt with separately.
Reviwed by: jimharris, alc
sharing especially on the default CPU 0 callout_cpu structure.
This will be followed up by attilio@ with a conversion to the new struct
mtx_padalign but doing this manual conversion first gives an easy MFC
candidate since mtx_padalign is a more extensive system change.
Sponsored by: Intel
Reviewed by: jeff, attilio
MFC after: 1 week
is running on other cpu, the CALLOUT_PENDING flag is temporarily
cleared. Then, callout_stop() on this, in fact active, callout fails
because CALLOUT_PENDING is not set, and callout_stop() returns 0.
Now, in sleepq_check_timeout(), the failed callout_stop() causes the
sleepq code to execute mi_switch() without even setting the wmesg,
since the switch-out is supposed to be transient. In fact, the thread
is put off the CPU for full timeout interval, instead of being put on
runq immediately. Until timeout fires, the process is unkillable for
obvious reasons.
Fix this by marking the migrating callouts with CALLOUT_DFRMIGRATION
flag. The flag is cleared by callout_stop_safe() when the function
detects a migration, besides returning the success. The softclock()
rechecks the flag for migrating callout and cancels its execution if
the flag was cleared meantime.
PR: misc/166340
Reported, debugging traces provided and tested by:
Christian Esken <christian.esken trivago com>
Reviewed by: avg, jhb
MFC after: 1 week
callout cpu lock (and after having dropped it).
If the newly scheduled thread wants to acquire the old queue it will
just spin forever.
Fix this by disabling preemption and interrupts entirely (because fast
interrupt handlers may incur in the same problem too) while switching
locks.
Reported by: hrs, Mike Tancsa <mike AT sentex DOT net>,
Chip Camden <sterling AT camdensoftware DOT com>
Tested by: hrs, Mike Tancsa <mike AT sentex DOT net>,
Chip Camden <sterling AT camdensoftware DOT com>,
Nicholas Esborn <nick AT desert DOT net>
Approved by: re (kib)
MFC after: 10 days
for a detailed explanation of the problems).
The only difference with the previous fix is in Solution2:
CPUBLOCK is no longer set when exiting from callout_reset_*() functions,
which avoid the deadlock (leading to r217161).
There is no need to CPUBLOCK there because the running-and-migrating
assumption is strong enough to avoid problems there.
Furthermore add a better !SMP compliancy (leading to shrinked code and
structures) and facility macros/functions.
Tested by: gianni, pho, dim
MFC after: 3 weeks
That revision is introducing a bug which is more visible than problems
it is trying to fix.
As long as my time is very limited in this period I am going to
commit back this patch just once it is fully fixed.
Reported by: dim, Nicholas Esborn
- Problem1:
Hypothesis: thread1 is doing a callout_reset_on(), within his
callout handler, willing to implicitly or explicitly migrate the
callout. thread2 is draining the callout.
Thesys:
* thread1 calls callout_lock() and locks the old callout cpu
* thread1 performs the checks in the first path of the
callout_reset_on()
* thread1 hits this codepiece:
/*
* If the lock must migrate we have to check the state again as
* we can't hold both the new and old locks simultaneously.
*/
if (c->c_cpu != cpu) {
c->c_cpu = cpu;
CC_UNLOCK(cc);
goto retry;
}
which means it will drop the lock and 'retry'
* thread2 will callout_lock() and locks the new callout cpu.
thread1 spins on the new lock and will not keep going for the
moment.
* thread2 checks that the callout is not pending (as callout is
currently running) and that it is not on cc->cc_curr (because cc
now refers to the new callout and the callout is running on the
old callout cpu) thus it thinks it is done and returns.
* thread1 will now acquire the lock and then adds the callout
to the new callout cpu queue
That seems an obvious race as callout_stop() falsely reports
the callout stopped or worse, callout_drain() falsely returns
while the callout is still in use.
- Solution1:
Fixing this problem would require, in general, to lock both
callout cpus at once while switching the c_cpu field and avoid
cyclic deadlocks between callout cpus locks.
The concept of CPUBLOCK is then introduced (working more or less
like the blocked_lock for thread_lock() function) meaning:
"in callout_lock(), spin until the c->c_cpu is not different from
CPUBLOCK". That way the "original" callout cpu, referred to the
above mentioned code snippet, will remain blocked until the lock
handover is over critical path will remain covered.
- Problem2:
Having the callout currently executed on a specific callout cpu
and contemporary pending on another callout cpu (as it can happen
with current code) breaks, at least, the assumption callout_drain()
returns just once the callout cannot be referenced anymore.
- Solution2:
Callout migration is deferred if the current callout is already
under execution.
The best place to do that is in softclock() and new members are
added to the callout cpu structure in order to specify a pending
migration is requested. That is necessary because the callout
cannot be trusted (not freed) the 100% of times after the execution
of the callout handler.
CPUBLOCK will prevent, in the "deferred migration" case, that the
callout gets freed in this case, stopping any callout_stop() and
callout_drain() possible activity until the migration is
actually performed.
- Problem3:
There is a further race in callout_drain().
In order to avoid a race between sleepqueue lock and callout cpu
spinlock, in _callout_stop_safe(), the callout cpu lock is dropped,
the sleepqueue lock is acquired and a new callout cpu lookup is
performed. Note that the channel used for locking the sleepqueue is
obtained from the "current" callout cpu (&cc->cc_waiting).
If the callout migrated in the meanwhile, callout_drain() will end up
using the wrong wchan for the sleepqueue (the locked one will be the
older, while the new one will not really be locked) leading to a
lock leak and a race access to sleepqueue.
- Solution3:
It is enough to check if a migration happened between the operation
of acquiring the sleepqueue lock and the new callout cpu lock and
eventually unwind all those and try again.
This problems can lead to deathly races on moderate (4-ways) SMP
environment, leading to easy panic or deadlocks.
The 24-ways of the reporter, could easilly panic, with completely
normal workload, almost daily.
gianni@ kindly wrote the following prof-of-concept which can
panic a FreeBSD machine in less than one hour, in smaller SMP:
http://www.freebsd.org/~attilio/callout/test.c
Reported by: Nicholas Esborn <nick at desert dot net>, DesertNet
In collabouration with: gianni, pho, Nicholas Esborn
Reviewed by: jhb
MFC after: 1 week (*)
* Usually, I would aim for a larger MFC timeout, but I really want this
in before 8.2-RELEASE, thus re@ accepted a shorter timeout as a special
case for this patch
to handle current timecounter wraps. Make kern_clocksource.c to honor that
requirement, scheduling sleeps on first CPU for no more then specified
period. Allow other CPUs to sleep up to 1/4 second (for any case).
The main goal of this is to generate timer interrupts only when there is
some work to do. When CPU is busy interrupts are generating at full rate
of hz + stathz to fullfill scheduler and timekeeping requirements. But
when CPU is idle, only minimum set of interrupts (down to 8 interrupts per
second per CPU now), needed to handle scheduled callouts is executed.
This allows significantly increase idle CPU sleep time, increasing effect
of static power-saving technologies. Also it should reduce host CPU load
on virtualized systems, when guest system is idle.
There is set of tunables, also available as writable sysctls, allowing to
control wanted event timer subsystem behavior:
kern.eventtimer.timer - allows to choose event timer hardware to use.
On x86 there is up to 4 different kinds of timers. Depending on whether
chosen timer is per-CPU, behavior of other options slightly differs.
kern.eventtimer.periodic - allows to choose periodic and one-shot
operation mode. In periodic mode, current timer hardware taken as the only
source of time for time events. This mode is quite alike to previous kernel
behavior. One-shot mode instead uses currently selected time counter
hardware to schedule all needed events one by one and program timer to
generate interrupt exactly in specified time. Default value depends of
chosen timer capabilities, but one-shot mode is preferred, until other is
forced by user or hardware.
kern.eventtimer.singlemul - in periodic mode specifies how much times
higher timer frequency should be, to not strictly alias hardclock() and
statclock() events. Default values are 2 and 4, but could be reduced to 1
if extra interrupts are unwanted.
kern.eventtimer.idletick - makes each CPU to receive every timer interrupt
independently of whether they busy or not. By default this options is
disabled. If chosen timer is per-CPU and runs in periodic mode, this option
has no effect - all interrupts are generating.
As soon as this patch modifies cpu_idle() on some platforms, I have also
refactored one on x86. Now it makes use of MONITOR/MWAIT instrunctions
(if supported) under high sleep/wakeup rate, as fast alternative to other
methods. It allows SMP scheduler to wake up sleeping CPUs much faster
without using IPI, significantly increasing performance on some highly
task-switching loads.
Tested by: many (on i386, amd64, sparc64 and powerc)
H/W donated by: Gheorghe Ardelean
Sponsored by: iXsystems, Inc.
use '-' in probe names, matching the probe names in Solaris.[1]
Add userland SDT probes definitions to sys/sdt.h.
Sponsored by: The FreeBSD Foundation
Discussed with: rwaston [1]
struct callout_cpu. From the comment in the file:
+ * There is one struct callout_cpu per cpu, holding all relevant
+ * state for the callout processing thread on the individual CPU.
+ * In particular:
+ * cc_ticks is incremented once per tick in callout_cpu().
+ * It tracks the global 'ticks' but in a way that the individual
+ * threads should not worry about races in the order in which
+ * hardclock() and hardclock_cpu() run on the various CPUs.
+ * cc_softclock is advanced in callout_cpu() to point to the
+ * first entry in cc_callwheel that may need handling. In turn,
+ * a softclock() is scheduled so it can serve the various entries i
+ * such that cc_softclock <= i <= cc_ticks .
Together with a smaller patch committed in september, this fixes a
bug that affects 8.0 with apps that rely on callouts to fire exactly
in the number of ticks specified (qemu among them).
Right now, callouts in 8.0 fire one tick late.
This was discussed in september with JeffR and jhb
MFC after: 3 days
The problem was introduced in SVN 180608/ rev 1.114 and affects
all users of callout_reset() (including select, usleep, setitimer).
A better fix probably involves replicating 'ticks' in the
struct callout_cpu; this commit is just a temporary thing so that
we can MFC it after a suitable test time and RE approval.
MFC after: 3 days
pointers to the callout handler just before and just after the callout
it invoked. I attempted to do this in a manner congruent to tracing in
Solaris's callout mechanism, but couldn't quite use the same names due
to convention and syntax differences.
Example DTrace script to generate a distribution graph of callout
execution times:
callout_execute:::callout_start
{
self->cstart = timestamp;
}
callout_execute:::callout_end
{
@length = quantize(timestamp - self->cstart);
}
Reviewed by: jb
MFC after: 3 days
"ticks" goes negative. This breaks the signed comparison in softclock.
This causes sleep() to never wake up, tcp to stop, etc etc. This is
bad(TM). Use the SEQ_LT() method from tcp's sequence number comparisons.
