15 files changed, 1105 insertions, 1042 deletions

diff --git a/share/man/man9/Makefile b/share/man/man9/Makefile
index e59c14fa9d17..cf64df7c811b 100644
--- a/share/man/man9/Makefile
+++ b/share/man/man9/Makefile
@@ -1570,7 +1570,6 @@ MLINKS+=timeout.9 callout.9 \
 	timeout.9 callout_active.9 \
 	timeout.9 callout_deactivate.9 \
 	timeout.9 callout_drain.9 \
-	timeout.9 callout_drain_async.9 \
 	timeout.9 callout_handle_init.9 \
 	timeout.9 callout_init.9 \
 	timeout.9 callout_init_mtx.9 \
diff --git a/share/man/man9/timeout.9 b/share/man/man9/timeout.9
index 4f52a19cb61f..7202815a7824 100644
--- a/share/man/man9/timeout.9
+++ b/share/man/man9/timeout.9
@@ -29,14 +29,13 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd January 14, 2015
+.Dd October 8, 2014
 .Dt TIMEOUT 9
 .Os
 .Sh NAME
 .Nm callout_active ,
 .Nm callout_deactivate ,
 .Nm callout_drain ,
-.Nm callout_drain_async ,
 .Nm callout_handle_init ,
 .Nm callout_init ,
 .Nm callout_init_mtx ,
@@ -64,232 +63,279 @@
 .In sys/systm.h
 .Bd -literal
 typedef void timeout_t (void *);
-typedef void callout_func_t (void *);
 .Ed
+.Ft int
+.Fn callout_active "struct callout *c"
+.Ft void
+.Fn callout_deactivate "struct callout *c"
+.Ft int
+.Fn callout_drain "struct callout *c"
+.Ft void
+.Fn callout_handle_init "struct callout_handle *handle"
+.Bd -literal
+struct callout_handle handle = CALLOUT_HANDLE_INITIALIZER(&handle);
+.Ed
+.Ft void
+.Fn callout_init "struct callout *c" "int mpsafe"
+.Ft void
+.Fn callout_init_mtx "struct callout *c" "struct mtx *mtx" "int flags"
+.Ft void
+.Fn callout_init_rm "struct callout *c" "struct rmlock *rm" "int flags"
+.Ft void
+.Fn callout_init_rw "struct callout *c" "struct rwlock *rw" "int flags"
+.Ft int
+.Fn callout_pending "struct callout *c"
+.Ft int
+.Fn callout_reset "struct callout *c" "int ticks" "timeout_t *func" "void *arg"
+.Ft int
+.Fn callout_reset_curcpu "struct callout *c" "int ticks" "timeout_t *func" \
+"void *arg"
+.Ft int
+.Fn callout_reset_on "struct callout *c" "int ticks" "timeout_t *func" \
+"void *arg" "int cpu"
+.Ft int
+.Fn callout_reset_sbt "struct callout *c" "sbintime_t sbt" \
+"sbintime_t pr" "timeout_t *func" "void *arg" "int flags"
+.Ft int
+.Fn callout_reset_sbt_curcpu "struct callout *c" "sbintime_t sbt" \
+"sbintime_t pr" "timeout_t *func" "void *arg" "int flags"
+.Ft int
+.Fn callout_reset_sbt_on "struct callout *c" "sbintime_t sbt" \
+"sbintime_t pr" "timeout_t *func" "void *arg" "int cpu" "int flags"
+.Ft int
+.Fn callout_schedule "struct callout *c" "int ticks"
+.Ft int
+.Fn callout_schedule_curcpu "struct callout *c" "int ticks"
+.Ft int
+.Fn callout_schedule_on "struct callout *c" "int ticks" "int cpu"
+.Ft int
+.Fn callout_schedule_sbt "struct callout *c" "sbintime_t sbt" \
+"sbintime_t pr" "int flags"
+.Ft int
+.Fn callout_schedule_sbt_curcpu "struct callout *c" "sbintime_t sbt" \
+"sbintime_t pr" "int flags"
+.Ft int
+.Fn callout_schedule_sbt_on "struct callout *c" "sbintime_t sbt" \
+"sbintime_t pr" "int cpu" "int flags"
+.Ft int
+.Fn callout_stop "struct callout *c"
+.Ft struct callout_handle
+.Fn timeout "timeout_t *func" "void *arg" "int ticks"
+.Ft void
+.Fn untimeout "timeout_t *func" "void *arg" "struct callout_handle handle"
 .Sh DESCRIPTION
 The
 .Nm callout
 API is used to schedule a call to an arbitrary function at a specific
-time in the future in a single-shot fashion.
-Consumers of this API are required to allocate a
-.Pq struct callout
-structure for each pending function invocation.
-The
+time in the future.
+Consumers of this API are required to allocate a callout structure
 .Pq struct callout
-structure stores the full state about any pending function call and
-should be drained by a call to
-.Fn callout_drain
-or
-.Fn callout_drain_async
-before freeing.
-.Sh INITIALISATION
-.Ft void
-.Fn callout_handle_init "struct callout_handle *handle"
-This function is deprecated and is used to prepare a
-.Pq struct callout_handle
-structure before it can be used the first time.
-If this function is called on a pending timeout, the pending timeout
-cannot be cancelled and the
-.Fn untimeout
-function will return as if there was no timeout pending.
+for each pending function invocation.
+This structure stores state about the pending function invocation including
+the function to be called and the time at which the function should be invoked.
+Pending function calls can be cancelled or rescheduled to a different time.
+In addition,
+a callout structure may be reused to schedule a new function call after a
+scheduled call is completed.
 .Pp
-.Fn CALLOUT_HANDLE_INITIALIZER "&handle"
-This macro is deprecated and can be used instead of
-.Fn callout_handle_init
-to assign the default state to the
-.Pq struct callout_handle
-structure when declaring static timeouts.
+Callouts only provide a single-shot mode.
+If a consumer requires a periodic timer,
+it must explicitly reschedule each function call.
+This is normally done by rescheduling the subsequent call within the called
+function.
 .Pp
-.Ft void
-.Fn callout_init "struct callout *c" "int mpsafe"
-This function prepares a
-.Pq struct callout
-structure before it can be used.
-This function should not be used when the callout is pending a timeout.
+Callout functions must not sleep.
+They may not acquire sleepable locks,
+wait on condition variables,
+perform blocking allocation requests,
+or invoke any other action that might sleep.
+.Pp
+Each callout structure must be initialized by
+.Fn callout_init ,
+.Fn callout_init_mtx ,
+.Fn callout_init_rm ,
+or
+.Fn callout_init_rw
+before it is passed to any of the other callout functions.
+The
+.Fn callout_init
+function initializes a callout structure in
+.Fa c
+that is not associated with a specific lock.
 If the
 .Fa mpsafe
-argument is non-zero, the callback function will be running unlocked.
-Else the Giant mutex will be locked before calling the callback function.
+argument is zero,
+the callout structure is not considered to be
+.Dq multi-processor safe ;
+and the Giant lock will be acquired before calling the callout function
+and released when the callout function returns.
 .Pp
-.Ft void
-.Fn callout_init_mtx "struct callout *c" "struct mtx *mtx" "int flags"
-This function prepares a
-.Pq struct callout
-structure before it can be used.
-This function should not be used when the callout is pending a timeout.
 The
-.Fa mtx
-argument should be non-zero and should specify a pointer to a valid
-spinlock type of mutex or a valid regular non-sleepable mutex which
-the callback subsystem should lock before calling the callback
-function.
-Valid
+.Fn callout_init_mtx ,
+.Fn callout_init_rm ,
+and
+.Fn callout_init_rw
+functions initialize a callout structure in
+.Fa c
+that is associated with a specific lock.
+The lock is specified by the
+.Fa mtx ,
+.Fa rm ,
+or
+.Fa rw
+parameter.
+The associated lock must be held while stopping or rescheduling the
+callout.
+The callout subsystem acquires the associated lock before calling the
+callout function and releases it after the function returns.
+If the callout was cancelled while the callout subsystem waited for the
+associated lock,
+the callout function is not called,
+and the associated lock is released.
+This ensures that stopping or rescheduling the callout will abort any
+previously scheduled invocation.
+.Pp
+Only regular mutexes may be used with
+.Fn callout_init_mtx ;
+spin mutexes are not supported.
+A sleepable read-mostly lock
+.Po
+one initialized with the
+.Dv RM_SLEEPABLE
+flag
+.Pc
+may not be used with
+.Fn callout_init_rm .
+Similarly, other sleepable lock types such as
+.Xr sx 9
+and
+.Xr lockmgr 9
+cannot be used with callouts because sleeping is not permitted in
+the callout subsystem.
+.Pp
+These
 .Fa flags
-are:
+may be specified for
+.Fn callout_init_mtx ,
+.Fn callout_init_rm ,
+or
+.Fn callout_init_rw :
 .Bl -tag -width ".Dv CALLOUT_RETURNUNLOCKED"
 .It Dv CALLOUT_RETURNUNLOCKED
-It is assumed that the callout function has released the specified
-mutex before returning.
-Else the callout subsystem will release the specified mutex after the
-callout function has returned.
+The callout function will release the associated lock itself,
+so the callout subsystem should not attempt to unlock it
+after the callout function returns.
+.It Dv CALLOUT_SHAREDLOCK
+The lock is only acquired in read mode when running the callout handler.
+This flag is ignored by
+.Fn callout_init_mtx .
 .El
 .Pp
-.Ft void
-.Fn callout_init_rm "struct callout *c" "struct rmlock *rm" "int flags"
-This function is the same like the
-.Fn callout_init_mtx
-function except it accepts a read-mostly type of lock.
-The read-mostly lock must not be initialised with the
-.Dv RM_SLEEPABLE
-flag.
+The function
+.Fn callout_stop
+cancels a callout
+.Fa c
+if it is currently pending.
+If the callout is pending, then
+.Fn callout_stop
+returns a non-zero value.
+If the callout is not set,
+has already been serviced,
+or is currently being serviced,
+then zero will be returned.
+If the callout has an associated lock,
+then that lock must be held when this function is called.
 .Pp
-.Ft void
-.Fn callout_init_rw "struct callout *c" "struct rwlock *rw" "int flags"
-This function is the same like the
-.Fn callout_init_mtx
-function except it accepts a reader-writer type of lock.
-.Sh SCHEDULING CALLOUTS
-.Ft struct callout_handle
-.Fn timeout "timeout_t *func" "void *arg" "int ticks"
-This function is deprecated and schedules a call to the function given by the argument
-.Fa func
-to take place after
+The function
+.Fn callout_drain
+is identical to
+.Fn callout_stop
+except that it will wait for the callout
+.Fa c
+to complete if it is already in progress.
+This function MUST NOT be called while holding any
+locks on which the callout might block, or deadlock will result.
+Note that if the callout subsystem has already begun processing this
+callout, then the callout function may be invoked before
+.Fn callout_drain
+returns.
+However, the callout subsystem does guarantee that the callout will be
+fully stopped before
+.Fn callout_drain
+returns.
+.Pp
+The
+.Fn callout_reset
+and
+.Fn callout_schedule
+function families schedule a future function invocation for callout
+.Fa c .
+If
+.Fa c
+already has a pending callout,
+it is cancelled before the new invocation is scheduled.
+These functions return a non-zero value if a pending callout was cancelled
+and zero if there was no pending callout.
+If the callout has an associated lock,
+then that lock must be held when any of these functions are called.
+.Pp
+The time at which the callout function will be invoked is determined by
+either the
+.Fa ticks
+argument or the
+.Fa sbt ,
+.Fa pr ,
+and
+.Fa flags
+arguments.
+When
+.Fa ticks
+is used,
+the callout is scheduled to execute after
 .Fa ticks Ns No /hz
 seconds.
 Non-positive values of
 .Fa ticks
 are silently converted to the value
 .Sq 1 .
-The
-.Fa func
-argument should be a valid pointer to a function that takes a single
-.Fa void *
-argument.
-Upon invocation, the
-.Fa func
-function will receive
-.Fa arg
-as its only argument.
-The Giant lock is locked when the
-.Fa arg
-function is invoked and should not be unlocked by this function.
-The returned value from
-.Fn timeout
-is a
-.Ft struct callout_handle
-structure which can be used in conjunction with the
-.Fn untimeout
-function to request that a scheduled timeout be cancelled.
-As handles are recycled by the system, it is possible, although unlikely,
-that a handle from one invocation of
-.Fn timeout
-may match the handle of another invocation of
-.Fn timeout
-if both calls used the same function pointer and argument, and the first
-timeout is expired or canceled before the second call.
-Please ensure that the function and argument pointers are unique when using this function.
-.Pp
-.Ft int
-.Fn callout_reset "struct callout *c" "int ticks" "callout_func_t *func" "void *arg"
-This function is used to schedule or re-schedule a callout.
-This function at first stops the callout given by the
-.Fa c
-argument, if any.
-Then it will start the callout given by the
-.Fa c
-argument.
-The relative time until the timeout callback happens is given by the
-.Fa ticks
-argument.
-The number of ticks in a second is defined by
-.Dv hz
-and can vary from system to system.
-This function returns a non-zero value if the given callout was pending and
-the callback function was prevented from being called.
-Else a value of zero is returned.
-If a lock is associated with the callout given by the
-.Fa c
-argument and it is exclusivly locked when this function is called this
-function will always ensure that previous callback function, if any,
-is never reached.
-In other words the callout will be atomically restarted.
-Else there is no such guarantee.
-The callback function is given by the
-.Fa func
-argument and its function argument is given by the
-.Fa arg
-argument.
-.Pp
-.Ft int
-.Fn callout_reset_curcpu "struct callout *c" "int ticks" "callout_func_t *func" \
-"void *arg"
-This function works the same like the
-.Fn callout_reset
-function except the callback function given by the
-.Fa func
-argument will be executed on the same CPU which called this function.
-A change in the CPU selection can happen if the callout has a lock
-associated with it and is locked when this function is called.
-A change in the CPU selection cannot happen if this function is
-re-scheduled inside a callout function.
-Else the callback function given by the
-.Fa func
-argument will be executed on the same CPU like previously done.
 .Pp
-.Ft int
-.Fn callout_reset_on "struct callout *c" "int ticks" "callout_func_t *func" \
-"void *arg" "int cpu"
-This function works the same like the
-.Fn callout_reset
-function except the callback function given by the
-.Fa func
-argument will be executed on the CPU given by the
-.Fa cpu
-argument.
-A change in the CPU selection can happen if the callout has a lock
-associated with it and is locked when this function is called.
-A change in the CPU selection cannot happen if this function is
-re-scheduled inside a callout function.
-Else the callback function given by the
-.Fa func
-argument will be executed on the same CPU like previously done.
-.Pp
-.Ft int
-.Fn callout_reset_sbt "struct callout *c" "sbintime_t sbt" \
-"sbintime_t pr" "callout_func_t *func" "void *arg" "int flags"
-This function works the same like the
-.Fn callout_reset
-function except the relative or absolute time after which the timeout
-callback should happen is given by the
+The
+.Fa sbt ,
+.Fa pr ,
+and
+.Fa flags
+arguments provide more control over the scheduled time including
+support for higher resolution times,
+specifying the precision of the scheduled time,
+and setting an absolute deadline instead of a relative timeout.
+The callout is scheduled to execute in a time window which begins at
+the time specified in
 .Fa sbt
-argument and extends for the amount of time specified in the
-.Fa pr
-argument.
-This function is used when you need high precision timeouts.
-If the
+and extends for the amount of time specified in
+.Fa pr .
+If
 .Fa sbt
-argument specifies a time in the past,
+specifies a time in the past,
 the window is adjusted to start at the current time.
 A non-zero value for
 .Fa pr
 allows the callout subsystem to coalesce callouts scheduled close to each
 other into fewer timer interrupts,
 reducing processing overhead and power consumption.
-The
+These
 .Fa flags
-argument may be non-zero to adjust the interpretation of the
+may be specified to adjust the interpretation of
 .Fa sbt
-and the
-.Fa pr
-arguments:
+and
+.Fa pr :
 .Bl -tag -width ".Dv C_DIRECT_EXEC"
 .It Dv C_ABSOLUTE
 Handle the
 .Fa sbt
 argument as an absolute time since boot.
-By default, the
+By default,
 .Fa sbt
-argument is treated like a relative amount of time,
+is treated as a relative amount of time,
 similar to
 .Fa ticks .
 .It Dv C_DIRECT_EXEC
@@ -301,7 +347,7 @@ Callout functions run in this context may use only spin mutexes for locking
 and should be as small as possible because they run with absolute priority.
 .It Fn C_PREL
 Specifies relative event time precision as binary logarithm of time interval
-divided by acceptable time deviation: 1 -- 1/2, 2 -- 1/4 and so on.
+divided by acceptable time deviation: 1 -- 1/2, 2 -- 1/4, etc.
 Note that the larger of
 .Fa pr
 or this value is used as the length of the time window.
@@ -314,215 +360,65 @@ Align the timeouts to
 calls if possible.
 .El
 .Pp
-.Ft int
-.Fn callout_reset_sbt_curcpu "struct callout *c" "sbintime_t sbt" \
-"sbintime_t pr" "callout_func_t *func" "void *arg" "int flags"
-This function works the same like the
-.Fn callout_reset_sbt
-function except the callback function given by the
-.Fa func
-argument will be executed on the same CPU which called this function.
-A change in the CPU selection can happen if the callout has a lock
-associated with it and is locked when this function is called.
-A change in the CPU selection cannot happen if this function is
-re-scheduled inside a callout function.
-Else the callback function given by the
-.Fa func
-argument will be executed on the same CPU like previously done.
-.Pp
-.Ft int
-.Fn callout_reset_sbt_on "struct callout *c" "sbintime_t sbt" \
-"sbintime_t pr" "callout_func_t *func" "void *arg" "int cpu" "int flags"
-This function works the same like the
-.Fn callout_reset_sbt
-function except the callback function given by the
+The
+.Fn callout_reset
+functions accept a
 .Fa func
-argument will be executed on the CPU given by the
-.Fa cpu
+argument which identifies the function to be called when the time expires.
+It must be a pointer to a function that takes a single
+.Fa void *
 argument.
-A change in the CPU selection can happen if the callout has a lock
-associated with it and is locked when this function is called.
-A change in the CPU selection cannot happen if this function is
-re-scheduled inside a callout function.
-Else the callback function given by the
+Upon invocation,
 .Fa func
-argument will be executed on the same CPU like previously done.
-.Pp
-.Ft int
-.Fn callout_schedule "struct callout *c" "int ticks"
-This function works the same like the
-.Fn callout_reset
-function except it re-uses the callback function and the callback argument
-already stored in the
-.Pq struct callout
-structure.
-.Pp
-.Ft int
-.Fn callout_schedule_curcpu "struct callout *c" "int ticks"
-This function works the same like the
-.Fn callout_reset_curcpu
-function except it re-uses the callback function and the callback argument
-already stored in the
-.Pq struct callout
-structure.
-.Pp
-.Ft int
-.Fn callout_schedule_on "struct callout *c" "int ticks" "int cpu"
-This function works the same like the
-.Fn callout_reset_on
-function except it re-uses the callback function and the callback argument
-already stored in the
-.Pq struct callout
-structure.
-.Pp
-.Ft int
-.Fn callout_schedule_sbt "struct callout *c" "sbintime_t sbt" \
-"sbintime_t pr" "int flags"
-This function works the same like the
-.Fn callout_reset_sbt
-function except it re-uses the callback function and the callback argument
-already stored in the
-.Pq struct callout
-structure.
-.Pp
-.Ft int
-.Fn callout_schedule_sbt_curcpu "struct callout *c" "sbintime_t sbt" \
-"sbintime_t pr" "int flags"
-This function works the same like the
-.Fn callout_reset_sbt_curcpu
-function except it re-uses the callback function and the callback argument
-already stored in the
-.Pq struct callout
-structure.
-.Pp
-.Ft int
-.Fn callout_schedule_sbt_on "struct callout *c" "sbintime_t sbt" \
-"sbintime_t pr" "int cpu" "int flags"
-This function works the same like the
-.Fn callout_reset_sbt_on
-function except it re-uses the callback function and the callback argument
-already stored in the
-.Pq struct callout
-structure.
-.Sh CHECKING THE STATE OF CALLOUTS
-.Ft int
-.Fn callout_pending "struct callout *c"
-This function returns non-zero if the callout pointed to by the
-.Fa c
-argument is pending for callback.
-Else this function returns zero.
-This function returns zero when inside the callout function if the
-callout is not re-scheduled.
-.Pp
-.Ft int
-.Fn callout_active "struct callout *c"
-This function is deprecated and returns non-zero if the callout
-pointed to by the
-.Fa c
-argument was scheduled in the past.
-Else this function returns zero.
-This function also returns zero after the
-.Fn callout_deactivate
-or the
-.Fn callout_stop
-or the
-.Fn callout_drain
-or the
-.Fn callout_drain_async
-function is called on the same callout as given by the
-.Fa c
-argument.
-.Pp
-.Ft void
-.Fn callout_deactivate "struct callout *c"
-This function is deprecated and ensures that subsequent calls to the
-.Fn callout_activate
-function returns zero until the callout is scheduled again.
-.Sh STOPPING CALLOUTS
-.Ft void
-.Fn untimeout "timeout_t *func" "void *arg" "struct callout_handle handle"
-This function is deprecated and cancels the timeout associated with the
-.Fa handle
-argument using the function pointed to by the
+will receive
+.Fa arg
+as its only argument.
+The
+.Fn callout_schedule
+functions reuse the
 .Fa func
-argument and having the
+and
 .Fa arg
-arguments to validate the handle.
-If the handle does not correspond to a timeout with
-the function
+arguments from the previous callout.
+Note that one of the
+.Fn callout_reset
+functions must always be called to initialize
 .Fa func
-taking the argument
+and
 .Fa arg
-no action is taken. The
-.Fa handle
-must be initialised by a previous call to
-.Fn timeout ,
-.Fn callout_handle_init
-or assigned the value of
-.Fn CALLOUT_HANDLE_INITIALIZER "&handle"
-before being passed to
-.Fn untimeout .
-The behavior of calling
-.Fn untimeout
-with an uninitialised handle
-is undefined.
+before one of the
+.Fn callout_schedule
+functions can be used.
 .Pp
-.Ft int
-.Fn callout_stop "struct callout *c"
-This function is used to stop a timeout function invocation associated with the callout pointed to by the
-.Fa c
-argument, in a non-blocking fashion.
-This function can be called multiple times in a row with no side effects, even if the callout is already stopped. This function however should not be called before the callout has been initialised.
-This function returns a non-zero value if the given callout was pending and
-the callback function was prevented from being called.
-Else a value of zero is returned.
-If a lock is associated with the callout given by the
-.Fa c
-argument and it is exclusivly locked when this function is called, the
-.Fn callout_stop
-function will always ensure that the callback function is never reached.
-In other words the callout will be atomically stopped.
-Else there is no such guarantee.
-.Sh DRAINING CALLOUTS
-.Ft int
-.Fn callout_drain "struct callout *c"
-This function works the same like the
-.Fn callout_stop
-function except it ensures that all callback functions have returned and there are no more references to the callout pointed to by the
-.Fa c
-argument inside the callout subsystem before it returns.
-Also this function ensures that the lock, if any, associated with the
-callout is no longer being used.
-When this function returns, it is safe to free the callout structure pointed to by the
-.Fa c
-argument.
+The callout subsystem provides a softclock thread for each CPU in the system.
+Callouts are assigned to a single CPU and are executed by the softclock thread
+for that CPU.
+Initially,
+callouts are assigned to CPU 0.
+The
+.Fn callout_reset_on ,
+.Fn callout_reset_sbt_on ,
+.Fn callout_schedule_on
+and
+.Fn callout_schedule_sbt_on
+functions assign the callout to CPU
+.Fa cpu .
+The
+.Fn callout_reset_curcpu ,
+.Fn callout_reset_sbt_curpu ,
+.Fn callout_schedule_curcpu
+and
+.Fn callout_schedule_sbt_curcpu
+functions assign the callout to the current CPU.
+The
+.Fn callout_reset ,
+.Fn callout_reset_sbt ,
+.Fn callout_schedule
+and
+.Fn callout_schedule_sbt
+functions schedule the callout to execute in the softclock thread of the CPU
+to which it is currently assigned.
 .Pp
-.Ft int
-.Fn callout_drain_async "struct callout *c" "callout_func_t *fn" "void *arg"
-This function is non-blocking and works the same like the
-.Fn callout_stop
-function except if it returns non-zero it means the callback function pointed to by the
-.Fa fn
-argument will be called back with the
-.Fa arg
-argument when all references to the callout pointed to by the
-.Fa c
-argument are gone.
-If this function returns zero, it is safe to free the callout structure pointed to by the
-.Fa c
-argument right away.
-.Sh CALLOUT FUNCTION RESTRICTIONS
-Callout functions must not sleep.
-They may not acquire sleepable locks, wait on condition variables,
-perform blocking allocation requests, or invoke any other action that
-might sleep.
-.Sh CALLOUT SUBSYSTEM INTERNALS
-The callout subsystem has its own set of spinlocks to protect its internal state.
-The callout subsystem provides a softclock thread for each CPU in the
-system.
-Callouts are assigned to a single CPU and are executed by the
-softclock thread for that CPU.
-Initially, callouts are assigned to CPU 0.
 Softclock threads are not pinned to their respective CPUs by default.
 The softclock thread for CPU 0 can be pinned to CPU 0 by setting the
 .Va kern.pin_default_swi
@@ -531,7 +427,50 @@ Softclock threads for CPUs other than zero can be pinned to their
 respective CPUs by setting the
 .Va kern.pin_pcpu_swi
 loader tunable to a non-zero value.
-.Sh "AVOIDING RACE CONDITIONS"
+.Pp
+The macros
+.Fn callout_pending ,
+.Fn callout_active
+and
+.Fn callout_deactivate
+provide access to the current state of the callout.
+The
+.Fn callout_pending
+macro checks whether a callout is
+.Em pending ;
+a callout is considered
+.Em pending
+when a timeout has been set but the time has not yet arrived.
+Note that once the timeout time arrives and the callout subsystem
+starts to process this callout,
+.Fn callout_pending
+will return
+.Dv FALSE
+even though the callout function may not have finished
+.Pq or even begun
+executing.
+The
+.Fn callout_active
+macro checks whether a callout is marked as
+.Em active ,
+and the
+.Fn callout_deactivate
+macro clears the callout's
+.Em active
+flag.
+The callout subsystem marks a callout as
+.Em active
+when a timeout is set and it clears the
+.Em active
+flag in
+.Fn callout_stop
+and
+.Fn callout_drain ,
+but it
+.Em does not
+clear it when a callout expires normally via the execution of the
+callout function.
+.Ss "Avoiding Race Conditions"
 The callout subsystem invokes callout functions from its own thread
 context.
 Without some kind of synchronization,
@@ -548,7 +487,7 @@ synchronization concerns.
 The first approach is preferred as it is the simplest:
 .Bl -enum -offset indent
 .It
-Callouts can be associated with a specific lock when they are initialised
+Callouts can be associated with a specific lock when they are initialized
 by
 .Fn callout_init_mtx ,
 .Fn callout_init_rm ,
@@ -569,7 +508,7 @@ or
 .Fn callout_schedule
 functions to provide this safety.
 .Pp
-A callout initialised via
+A callout initialized via
 .Fn callout_init
 with
 .Fa mpsafe
@@ -592,8 +531,9 @@ function families
 .Pc
 indicates whether or not the callout was removed.
 If it is known that the callout was set and the callout function has
-not yet executed, then a return value of zero indicates that the
-callout function is about to be called.
+not yet executed, then a return value of
+.Dv FALSE
+indicates that the callout function is about to be called.
 For example:
 .Bd -literal -offset indent
 if (sc->sc_flags & SCFLG_CALLOUT_RUNNING) {
@@ -649,14 +589,16 @@ The callout function should first check the
 .Em pending
 flag and return without action if
 .Fn callout_pending
-returns non-zero.
+returns
+.Dv TRUE .
 This indicates that the callout was rescheduled using
 .Fn callout_reset
 just before the callout function was invoked.
 If
 .Fn callout_active
-returns zero then the callout function should also return without
-action.
+returns
+.Dv FALSE
+then the callout function should also return without action.
 This indicates that the callout has been stopped.
 Finally, the callout function should call
 .Fn callout_deactivate
@@ -726,13 +668,129 @@ a callout should always be drained prior to destroying its associated lock
 or releasing the storage for the callout structure.
 .Sh LEGACY API
 .Bf Sy
-The
+The functions below are a legacy API that will be removed in a future release.
+New code should not use these routines.
+.Ef
+.Pp
+The function
+.Fn timeout
+schedules a call to the function given by the argument
+.Fa func
+to take place after
+.Fa ticks Ns No /hz
+seconds.
+Non-positive values of
+.Fa ticks
+are silently converted to the value
+.Sq 1 .
+.Fa func
+should be a pointer to a function that takes a
+.Fa void *
+argument.
+Upon invocation,
+.Fa func
+will receive
+.Fa arg
+as its only argument.
+The return value from
 .Fn timeout
+is a
+.Ft struct callout_handle
+which can be used in conjunction with the
+.Fn untimeout
+function to request that a scheduled timeout be canceled.
+.Pp
+The function
+.Fn callout_handle_init
+can be used to initialize a handle to a state which will cause
+any calls to
+.Fn untimeout
+with that handle to return with no side
+effects.
+.Pp
+Assigning a callout handle the value of
+.Fn CALLOUT_HANDLE_INITIALIZER
+performs the same function as
+.Fn callout_handle_init
+and is provided for use on statically declared or global callout handles.
+.Pp
+The function
+.Fn untimeout
+cancels the timeout associated with
+.Fa handle
+using the
+.Fa func
 and
+.Fa arg
+arguments to validate the handle.
+If the handle does not correspond to a timeout with
+the function
+.Fa func
+taking the argument
+.Fa arg
+no action is taken.
+.Fa handle
+must be initialized by a previous call to
+.Fn timeout ,
+.Fn callout_handle_init ,
+or assigned the value of
+.Fn CALLOUT_HANDLE_INITIALIZER "&handle"
+before being passed to
+.Fn untimeout .
+The behavior of calling
 .Fn untimeout
-functions are a legacy API that will be removed in a future release.
-New code should not use these routines.
-.Ef
+with an uninitialized handle
+is undefined.
+.Pp
+As handles are recycled by the system, it is possible (although unlikely)
+that a handle from one invocation of
+.Fn timeout
+may match the handle of another invocation of
+.Fn timeout
+if both calls used the same function pointer and argument, and the first
+timeout is expired or canceled before the second call.
+The timeout facility offers O(1) running time for
+.Fn timeout
+and
+.Fn untimeout .
+Timeouts are executed from
+.Fn softclock
+with the
+.Va Giant
+lock held.
+Thus they are protected from re-entrancy.
+.Sh RETURN VALUES
+The
+.Fn callout_active
+macro returns the state of a callout's
+.Em active
+flag.
+.Pp
+The
+.Fn callout_pending
+macro returns the state of a callout's
+.Em pending
+flag.
+.Pp
+The
+.Fn callout_reset
+and
+.Fn callout_schedule
+function families return non-zero if the callout was pending before the new
+function invocation was scheduled.
+.Pp
+The
+.Fn callout_stop
+and
+.Fn callout_drain
+functions return non-zero if the callout was still pending when it was
+called or zero otherwise.
+The
+.Fn timeout
+function returns a
+.Ft struct callout_handle
+that can be passed to
+.Fn untimeout .
 .Sh HISTORY
 The current timeout and untimeout routines are based on the work of
 .An Adam M. Costello
@@ -757,4 +815,4 @@ The current implementation replaces the long standing
 .Bx
 linked list
 callout mechanism which offered O(n) insertion and removal running time
-and did not generate or require handles for untimeout operations.
+but did not generate or require handles for untimeout operations.
diff --git a/sys/kern/init_main.c b/sys/kern/init_main.c
index 922959f8fa57..beb49bc56962 100644
--- a/sys/kern/init_main.c
+++ b/sys/kern/init_main.c
@@ -504,8 +504,7 @@ proc0_init(void *dummy __unused)
 
 	callout_init_mtx(&p->p_itcallout, &p->p_mtx, 0);
 	callout_init_mtx(&p->p_limco, &p->p_mtx, 0);
-	mtx_init(&td->td_slpmutex, "td_slpmutex", NULL, MTX_SPIN);
-	callout_init_mtx(&td->td_slpcallout, &td->td_slpmutex, 0);
+	callout_init(&td->td_slpcallout, CALLOUT_MPSAFE);
 
 	/* Create credentials. */
 	p->p_ucred = crget();
diff --git a/sys/kern/kern_condvar.c b/sys/kern/kern_condvar.c
index 8c2691babb3f..2700a25d477c 100644
--- a/sys/kern/kern_condvar.c
+++ b/sys/kern/kern_condvar.c
@@ -313,13 +313,15 @@ _cv_timedwait_sbt(struct cv *cvp, struct lock_object *lock, sbintime_t sbt,
 	DROP_GIANT();
 
 	sleepq_add(cvp, lock, cvp->cv_description, SLEEPQ_CONDVAR, 0);
-	sleepq_release(cvp);
 	sleepq_set_timeout_sbt(cvp, sbt, pr, flags);
 	if (lock != &Giant.lock_object) {
+		if (class->lc_flags & LC_SLEEPABLE)
+			sleepq_release(cvp);
 		WITNESS_SAVE(lock, lock_witness);
 		lock_state = class->lc_unlock(lock);
+		if (class->lc_flags & LC_SLEEPABLE)
+			sleepq_lock(cvp);
 	}
-	sleepq_lock(cvp);
 	rval = sleepq_timedwait(cvp, 0);
 
 #ifdef KTRACE
@@ -381,13 +383,15 @@ _cv_timedwait_sig_sbt(struct cv *cvp, struct lock_object *lock,
 
 	sleepq_add(cvp, lock, cvp->cv_description, SLEEPQ_CONDVAR |
 	    SLEEPQ_INTERRUPTIBLE, 0);
-	sleepq_release(cvp);
 	sleepq_set_timeout_sbt(cvp, sbt, pr, flags);
 	if (lock != &Giant.lock_object) {
+		if (class->lc_flags & LC_SLEEPABLE)
+			sleepq_release(cvp);
 		WITNESS_SAVE(lock, lock_witness);
 		lock_state = class->lc_unlock(lock);
+		if (class->lc_flags & LC_SLEEPABLE)
+			sleepq_lock(cvp);
 	}
-	sleepq_lock(cvp);
 	rval = sleepq_timedwait_sig(cvp, 0);
 
 #ifdef KTRACE
diff --git a/sys/kern/kern_lock.c b/sys/kern/kern_lock.c
index 38c870709c31..36a8470d2971 100644
--- a/sys/kern/kern_lock.c
+++ b/sys/kern/kern_lock.c
@@ -210,11 +210,9 @@ sleeplk(struct lock *lk, u_int flags, struct lock_object *ilk,
 	GIANT_SAVE();
 	sleepq_add(&lk->lock_object, NULL, wmesg, SLEEPQ_LK | (catch ?
 	    SLEEPQ_INTERRUPTIBLE : 0), queue);
-	if ((flags & LK_TIMELOCK) && timo) {
-		sleepq_release(&lk->lock_object);
+	if ((flags & LK_TIMELOCK) && timo)
 		sleepq_set_timeout(&lk->lock_object, timo);
-		sleepq_lock(&lk->lock_object);
-	}
+
 	/*
 	 * Decisional switch for real sleeping.
 	 */
diff --git a/sys/kern/kern_switch.c b/sys/kern/kern_switch.c
index 61d914944640..d0009b1042f3 100644
--- a/sys/kern/kern_switch.c
+++ b/sys/kern/kern_switch.c
@@ -93,6 +93,8 @@ SCHED_STAT_DEFINE_VAR(turnstile,
     &DPCPU_NAME(sched_switch_stats[SWT_TURNSTILE]), "");
 SCHED_STAT_DEFINE_VAR(sleepq,
     &DPCPU_NAME(sched_switch_stats[SWT_SLEEPQ]), "");
+SCHED_STAT_DEFINE_VAR(sleepqtimo,
+    &DPCPU_NAME(sched_switch_stats[SWT_SLEEPQTIMO]), "");
 SCHED_STAT_DEFINE_VAR(relinquish, 
     &DPCPU_NAME(sched_switch_stats[SWT_RELINQUISH]), "");
 SCHED_STAT_DEFINE_VAR(needresched,
diff --git a/sys/kern/kern_synch.c b/sys/kern/kern_synch.c
index 19bf4e8a094a..9501ba2ba07d 100644
--- a/sys/kern/kern_synch.c
+++ b/sys/kern/kern_synch.c
@@ -236,16 +236,12 @@ _sleep(void *ident, struct lock_object *lock, int priority,
 	 * return from cursig().
 	 */
 	sleepq_add(ident, lock, wmesg, sleepq_flags, 0);
+	if (sbt != 0)
+		sleepq_set_timeout_sbt(ident, sbt, pr, flags);
 	if (lock != NULL && class->lc_flags & LC_SLEEPABLE) {
 		sleepq_release(ident);
 		WITNESS_SAVE(lock, lock_witness);
 		lock_state = class->lc_unlock(lock);
-		if (sbt != 0)
-			sleepq_set_timeout_sbt(ident, sbt, pr, flags);
-		sleepq_lock(ident);
-	} else if (sbt != 0) {
-		sleepq_release(ident);
-		sleepq_set_timeout_sbt(ident, sbt, pr, flags);
 		sleepq_lock(ident);
 	}
 	if (sbt != 0 && catch)
@@ -310,11 +306,8 @@ msleep_spin_sbt(void *ident, struct mtx *mtx, const char *wmesg,
 	 * We put ourselves on the sleep queue and start our timeout.
 	 */
 	sleepq_add(ident, &mtx->lock_object, wmesg, SLEEPQ_SLEEP, 0);
-	if (sbt != 0) {
-		sleepq_release(ident);
+	if (sbt != 0)
 		sleepq_set_timeout_sbt(ident, sbt, pr, flags);
-		sleepq_lock(ident);
-	}
 
 	/*
 	 * Can't call ktrace with any spin locks held so it can lock the
diff --git a/sys/kern/kern_thread.c b/sys/kern/kern_thread.c
index b1e1a1263597..2d0b0d278a56 100644
--- a/sys/kern/kern_thread.c
+++ b/sys/kern/kern_thread.c
@@ -149,9 +149,6 @@ thread_ctor(void *mem, int size, void *arg, int flags)
 	audit_thread_alloc(td);
 #endif
 	umtx_thread_alloc(td);
-
-	mtx_init(&td->td_slpmutex, "td_slpmutex", NULL, MTX_SPIN);
-	callout_init_mtx(&td->td_slpcallout, &td->td_slpmutex, 0);
 	return (0);
 }
 
@@ -165,10 +162,6 @@ thread_dtor(void *mem, int size, void *arg)
 
 	td = (struct thread *)mem;
 
-	/* make sure to drain any use of the "td->td_slpcallout" */
-	callout_drain(&td->td_slpcallout);
-	mtx_destroy(&td->td_slpmutex);
-
 #ifdef INVARIANTS
 	/* Verify that this thread is in a safe state to free. */
 	switch (td->td_state) {
@@ -551,6 +544,7 @@ thread_link(struct thread *td, struct proc *p)
 	LIST_INIT(&td->td_lprof[0]);
 	LIST_INIT(&td->td_lprof[1]);
 	sigqueue_init(&td->td_sigqueue, p);
+	callout_init(&td->td_slpcallout, CALLOUT_MPSAFE);
 	TAILQ_INSERT_TAIL(&p->p_threads, td, td_plist);
 	p->p_numthreads++;
 }
diff --git a/sys/kern/kern_timeout.c b/sys/kern/kern_timeout.c
index 4336faab3bf7..13822fd99ac6 100644
--- a/sys/kern/kern_timeout.c
+++ b/sys/kern/kern_timeout.c
@@ -54,8 +54,6 @@ __FBSDID("$FreeBSD$");
 #include <sys/lock.h>
 #include <sys/malloc.h>
 #include <sys/mutex.h>
-#include <sys/rmlock.h>
-#include <sys/rwlock.h>
 #include <sys/proc.h>
 #include <sys/sdt.h>
 #include <sys/sleepqueue.h>
@@ -126,216 +124,37 @@ SYSCTL_INT(_kern, OID_AUTO, pin_pcpu_swi, CTLFLAG_RDTUN | CTLFLAG_NOFETCH, &pin_
  */
 u_int callwheelsize, callwheelmask;
 
-typedef void callout_mutex_op_t(struct lock_object *);
-typedef int callout_owned_op_t(struct lock_object *);
-
-struct callout_mutex_ops {
-	callout_mutex_op_t *lock;
-	callout_mutex_op_t *unlock;
-	callout_owned_op_t *owned;
-};
-
-enum {
-	CALLOUT_LC_UNUSED_0,
-	CALLOUT_LC_UNUSED_1,
-	CALLOUT_LC_UNUSED_2,
-	CALLOUT_LC_UNUSED_3,
-	CALLOUT_LC_SPIN,
-	CALLOUT_LC_MUTEX,
-	CALLOUT_LC_RW,
-	CALLOUT_LC_RM,
-};
-
-static void
-callout_mutex_op_none(struct lock_object *lock)
-{
-}
-
-static int
-callout_owned_op_none(struct lock_object *lock)
-{
-	return (0);
-}
-
-static void
-callout_mutex_lock(struct lock_object *lock)
-{
-	mtx_lock((struct mtx *)lock);
-}
-
-static void
-callout_mutex_unlock(struct lock_object *lock)
-{
-	mtx_unlock((struct mtx *)lock);
-}
-
-static void
-callout_mutex_lock_spin(struct lock_object *lock)
-{
-	mtx_lock_spin((struct mtx *)lock);
-}
-
-static void
-callout_mutex_unlock_spin(struct lock_object *lock)
-{
-	mtx_unlock_spin((struct mtx *)lock);
-}
-
-static int
-callout_mutex_owned(struct lock_object *lock)
-{
-	return (mtx_owned((struct mtx *)lock));
-}
-
-static void
-callout_rm_wlock(struct lock_object *lock)
-{
-	rm_wlock((struct rmlock *)lock);
-}
-
-static void
-callout_rm_wunlock(struct lock_object *lock)
-{
-	rm_wunlock((struct rmlock *)lock);
-}
-
-static int
-callout_rm_owned(struct lock_object *lock)
-{
-	return (rm_wowned((struct rmlock *)lock));
-}
-
-static void
-callout_rw_wlock(struct lock_object *lock)
-{
-	rw_wlock((struct rwlock *)lock);
-}
-
-static void
-callout_rw_wunlock(struct lock_object *lock)
-{
-	rw_wunlock((struct rwlock *)lock);
-}
-
-static int
-callout_rw_owned(struct lock_object *lock)
-{
-	return (rw_wowned((struct rwlock *)lock));
-}
-
-static const struct callout_mutex_ops callout_mutex_ops[8] = {
-	[CALLOUT_LC_UNUSED_0] = {
-		.lock = callout_mutex_op_none,
-		.unlock = callout_mutex_op_none,
-		.owned = callout_owned_op_none,
-	},
-	[CALLOUT_LC_UNUSED_1] = {
-		.lock = callout_mutex_op_none,
-		.unlock = callout_mutex_op_none,
-		.owned = callout_owned_op_none,
-	},
-	[CALLOUT_LC_UNUSED_2] = {
-		.lock = callout_mutex_op_none,
-		.unlock = callout_mutex_op_none,
-		.owned = callout_owned_op_none,
-	},
-	[CALLOUT_LC_UNUSED_3] = {
-		.lock = callout_mutex_op_none,
-		.unlock = callout_mutex_op_none,
-		.owned = callout_owned_op_none,
-	},
-	[CALLOUT_LC_SPIN] = {
-		.lock = callout_mutex_lock_spin,
-		.unlock = callout_mutex_unlock_spin,
-		.owned = callout_mutex_owned,
-	},
-	[CALLOUT_LC_MUTEX] = {
-		.lock = callout_mutex_lock,
-		.unlock = callout_mutex_unlock,
-		.owned = callout_mutex_owned,
-	},
-	[CALLOUT_LC_RW] = {
-		.lock = callout_rw_wlock,
-		.unlock = callout_rw_wunlock,
-		.owned = callout_rw_owned,
-	},
-	[CALLOUT_LC_RM] = {
-		.lock = callout_rm_wlock,
-		.unlock = callout_rm_wunlock,
-		.owned = callout_rm_owned,
-	},
-};
-
-static void
-callout_lock_client(int c_flags, struct lock_object *c_lock)
-{
-	callout_mutex_ops[CALLOUT_GET_LC(c_flags)].lock(c_lock);
-}
-
-static void
-callout_unlock_client(int c_flags, struct lock_object *c_lock)
-{
-	callout_mutex_ops[CALLOUT_GET_LC(c_flags)].unlock(c_lock);
-}
-
-#ifdef SMP
-static int
-callout_lock_owned_client(int c_flags, struct lock_object *c_lock)
-{
-	return (callout_mutex_ops[CALLOUT_GET_LC(c_flags)].owned(c_lock));
-}
-#endif
-
 /*
- * The callout CPU exec structure represent information necessary for
- * describing the state of callouts currently running on the CPU and
- * for handling deferred callout restarts.
- *
- * In particular, the first entry of the array cc_exec_entity holds
- * information for callouts running from the SWI thread context, while
- * the second one holds information for callouts running directly from
- * the hardware interrupt context.
+ * The callout cpu exec entities represent informations necessary for
+ * describing the state of callouts currently running on the CPU and the ones
+ * necessary for migrating callouts to the new callout cpu. In particular,
+ * the first entry of the array cc_exec_entity holds informations for callout
+ * running in SWI thread context, while the second one holds informations
+ * for callout running directly from hardware interrupt context.
+ * The cached informations are very important for deferring migration when
+ * the migrating callout is already running.
  */
 struct cc_exec {
-	/*
-	 * The "cc_curr" points to the currently executing callout and
-	 * is protected by the "cc_lock" spinlock. If no callback is
-	 * currently executing it is equal to "NULL".
-	 */
+	struct callout		*cc_next;
 	struct callout		*cc_curr;
-	/*
-	 * The "cc_restart_args" structure holds the argument for a
-	 * deferred callback restart and is protected by the "cc_lock"
-	 * spinlock. The structure is only valid if "cc_restart" is
-	 * "true". If "cc_restart" is "false" the information in the
-	 * "cc_restart_args" structure shall be ignored.
-	 */
-	struct callout_args	cc_restart_args;
-	bool			cc_restart;
-	/*
-	 * The "cc_cancel" variable allows the currently pending
-	 * callback to be atomically cancelled. This field is write
-	 * protected by the "cc_lock" spinlock.
-	 */
-	bool cc_cancel;
-	/*
-	 * The "cc_drain_fn" points to a function which shall be
-	 * called with the argument stored in "cc_drain_arg" when an
-	 * asynchronous drain is performed. This field is write
-	 * protected by the "cc_lock" spinlock.
-	 */
-	callout_func_t *cc_drain_fn;
-	void *cc_drain_arg;
+#ifdef SMP
+	void			(*ce_migration_func)(void *);
+	void			*ce_migration_arg;
+	int			ce_migration_cpu;
+	sbintime_t		ce_migration_time;
+	sbintime_t		ce_migration_prec;
+#endif
+	bool			cc_cancel;
+	bool			cc_waiting;
 };
 
 /*
- * There is one "struct callout_cpu" per CPU, holding all relevant
+ * There is one struct callout_cpu per cpu, holding all relevant
  * state for the callout processing thread on the individual CPU.
  */
 struct callout_cpu {
 	struct mtx_padalign	cc_lock;
 	struct cc_exec 		cc_exec_entity[2];
-	struct callout		*cc_exec_next_dir;
 	struct callout		*cc_callout;
 	struct callout_list	*cc_callwheel;
 	struct callout_tailq	cc_expireq;
@@ -347,7 +166,27 @@ struct callout_cpu {
 	char			cc_ktr_event_name[20];
 };
 
+#define	cc_exec_curr		cc_exec_entity[0].cc_curr
+#define	cc_exec_next		cc_exec_entity[0].cc_next
+#define	cc_exec_cancel		cc_exec_entity[0].cc_cancel
+#define	cc_exec_waiting		cc_exec_entity[0].cc_waiting
+#define	cc_exec_curr_dir	cc_exec_entity[1].cc_curr
+#define	cc_exec_next_dir	cc_exec_entity[1].cc_next
+#define	cc_exec_cancel_dir	cc_exec_entity[1].cc_cancel
+#define	cc_exec_waiting_dir	cc_exec_entity[1].cc_waiting
+
 #ifdef SMP
+#define	cc_migration_func	cc_exec_entity[0].ce_migration_func
+#define	cc_migration_arg	cc_exec_entity[0].ce_migration_arg
+#define	cc_migration_cpu	cc_exec_entity[0].ce_migration_cpu
+#define	cc_migration_time	cc_exec_entity[0].ce_migration_time
+#define	cc_migration_prec	cc_exec_entity[0].ce_migration_prec
+#define	cc_migration_func_dir	cc_exec_entity[1].ce_migration_func
+#define	cc_migration_arg_dir	cc_exec_entity[1].ce_migration_arg
+#define	cc_migration_cpu_dir	cc_exec_entity[1].ce_migration_cpu
+#define	cc_migration_time_dir	cc_exec_entity[1].ce_migration_time
+#define	cc_migration_prec_dir	cc_exec_entity[1].ce_migration_prec
+
 struct callout_cpu cc_cpu[MAXCPU];
 #define	CPUBLOCK	MAXCPU
 #define	CC_CPU(cpu)	(&cc_cpu[(cpu)])
@@ -372,9 +211,60 @@ static void	softclock_call_cc(struct callout *c, struct callout_cpu *cc,
 
 static MALLOC_DEFINE(M_CALLOUT, "callout", "Callout datastructures");
 
+/**
+ * Locked by cc_lock:
+ *   cc_curr         - If a callout is in progress, it is cc_curr.
+ *                     If cc_curr is non-NULL, threads waiting in
+ *                     callout_drain() will be woken up as soon as the
+ *                     relevant callout completes.
+ *   cc_cancel       - Changing to 1 with both callout_lock and cc_lock held
+ *                     guarantees that the current callout will not run.
+ *                     The softclock() function sets this to 0 before it
+ *                     drops callout_lock to acquire c_lock, and it calls
+ *                     the handler only if curr_cancelled is still 0 after
+ *                     cc_lock is successfully acquired.
+ *   cc_waiting      - If a thread is waiting in callout_drain(), then
+ *                     callout_wait is nonzero.  Set only when
+ *                     cc_curr is non-NULL.
+ */
+
+/*
+ * Resets the execution entity tied to a specific callout cpu.
+ */
+static void
+cc_cce_cleanup(struct callout_cpu *cc, int direct)
+{
+
+	cc->cc_exec_entity[direct].cc_curr = NULL;
+	cc->cc_exec_entity[direct].cc_next = NULL;
+	cc->cc_exec_entity[direct].cc_cancel = false;
+	cc->cc_exec_entity[direct].cc_waiting = false;
+#ifdef SMP
+	cc->cc_exec_entity[direct].ce_migration_cpu = CPUBLOCK;
+	cc->cc_exec_entity[direct].ce_migration_time = 0;
+	cc->cc_exec_entity[direct].ce_migration_prec = 0;
+	cc->cc_exec_entity[direct].ce_migration_func = NULL;
+	cc->cc_exec_entity[direct].ce_migration_arg = NULL;
+#endif
+}
+
+/*
+ * Checks if migration is requested by a specific callout cpu.
+ */
+static int
+cc_cce_migrating(struct callout_cpu *cc, int direct)
+{
+
+#ifdef SMP
+	return (cc->cc_exec_entity[direct].ce_migration_cpu != CPUBLOCK);
+#else
+	return (0);
+#endif
+}
+
 /*
- * Kernel low level callwheel initialization called from cpu0 during
- * kernel startup:
+ * Kernel low level callwheel initialization
+ * called on cpu0 during kernel startup.
  */
 static void
 callout_callwheel_init(void *dummy)
@@ -434,6 +324,8 @@ callout_cpu_init(struct callout_cpu *cc, int cpu)
 		LIST_INIT(&cc->cc_callwheel[i]);
 	TAILQ_INIT(&cc->cc_expireq);
 	cc->cc_firstevent = SBT_MAX;
+	for (i = 0; i < 2; i++)
+		cc_cce_cleanup(cc, i);
 	snprintf(cc->cc_ktr_event_name, sizeof(cc->cc_ktr_event_name),
 	    "callwheel cpu %d", cpu);
 	if (cc->cc_callout == NULL)	/* Only cpu0 handles timeout(9) */
@@ -441,11 +333,41 @@ callout_cpu_init(struct callout_cpu *cc, int cpu)
 	for (i = 0; i < ncallout; i++) {
 		c = &cc->cc_callout[i];
 		callout_init(c, 0);
-		c->c_flags |= CALLOUT_LOCAL_ALLOC;
+		c->c_flags = CALLOUT_LOCAL_ALLOC;
 		SLIST_INSERT_HEAD(&cc->cc_callfree, c, c_links.sle);
 	}
 }
 
+#ifdef SMP
+/*
+ * Switches the cpu tied to a specific callout.
+ * The function expects a locked incoming callout cpu and returns with
+ * locked outcoming callout cpu.
+ */
+static struct callout_cpu *
+callout_cpu_switch(struct callout *c, struct callout_cpu *cc, int new_cpu)
+{
+	struct callout_cpu *new_cc;
+
+	MPASS(c != NULL && cc != NULL);
+	CC_LOCK_ASSERT(cc);
+
+	/*
+	 * Avoid interrupts and preemption firing after the callout cpu
+	 * is blocked in order to avoid deadlocks as the new thread
+	 * may be willing to acquire the callout cpu lock.
+	 */
+	c->c_cpu = CPUBLOCK;
+	spinlock_enter();
+	CC_UNLOCK(cc);
+	new_cc = CC_CPU(new_cpu);
+	CC_LOCK(new_cc);
+	spinlock_exit();
+	c->c_cpu = new_cpu;
+	return (new_cc);
+}
+#endif
+
 /*
  * Start standard softclock thread.
  */
@@ -522,8 +444,9 @@ callout_process(sbintime_t now)
 #ifdef CALLOUT_PROFILING
 	int depth_dir = 0, mpcalls_dir = 0, lockcalls_dir = 0;
 #endif
+
 	cc = CC_SELF();
-	CC_LOCK(cc);
+	mtx_lock_spin_flags(&cc->cc_lock, MTX_QUIET);
 
 	/* Compute the buckets of the last scan and present times. */
 	firstb = callout_hash(cc->cc_lastscan);
@@ -626,7 +549,7 @@ next:
 	avg_mpcalls_dir += (mpcalls_dir * 1000 - avg_mpcalls_dir) >> 8;
 	avg_lockcalls_dir += (lockcalls_dir * 1000 - avg_lockcalls_dir) >> 8;
 #endif
-	CC_UNLOCK(cc);
+	mtx_unlock_spin_flags(&cc->cc_lock, MTX_QUIET);
 	/*
 	 * swi_sched acquires the thread lock, so we don't want to call it
 	 * with cc_lock held; incorrect locking order.
@@ -639,55 +562,49 @@ static struct callout_cpu *
 callout_lock(struct callout *c)
 {
 	struct callout_cpu *cc;
-	cc = CC_CPU(c->c_cpu);
-	CC_LOCK(cc);
+	int cpu;
+
+	for (;;) {
+		cpu = c->c_cpu;
+#ifdef SMP
+		if (cpu == CPUBLOCK) {
+			while (c->c_cpu == CPUBLOCK)
+				cpu_spinwait();
+			continue;
+		}
+#endif
+		cc = CC_CPU(cpu);
+		CC_LOCK(cc);
+		if (cpu == c->c_cpu)
+			break;
+		CC_UNLOCK(cc);
+	}
 	return (cc);
 }
 
-static struct callout_cpu *
-callout_cc_add_locked(struct callout *c, struct callout_cpu *cc,
-    struct callout_args *coa, bool can_swap_cpu)
+static void
+callout_cc_add(struct callout *c, struct callout_cpu *cc,
+    sbintime_t sbt, sbintime_t precision, void (*func)(void *),
+    void *arg, int cpu, int flags)
 {
-#ifndef NO_EVENTTIMERS
-	sbintime_t sbt;
-#endif
 	int bucket;
 
 	CC_LOCK_ASSERT(cc);
-
-	/* update flags before swapping locks, if any */
-	c->c_flags &= ~(CALLOUT_PROCESSED | CALLOUT_DIRECT | CALLOUT_DEFRESTART);
-	if (coa->flags & C_DIRECT_EXEC)
-		c->c_flags |= (CALLOUT_ACTIVE | CALLOUT_PENDING | CALLOUT_DIRECT);
-	else
-		c->c_flags |= (CALLOUT_ACTIVE | CALLOUT_PENDING);
-
-#ifdef SMP
-	/*
-	 * Check if we are changing the CPU on which the callback
-	 * should be executed and if we have a lock protecting us:
-	 */
-	if (can_swap_cpu != false && coa->cpu != c->c_cpu &&
-	    callout_lock_owned_client(c->c_flags, c->c_lock) != 0) {
-		CC_UNLOCK(cc);
-		c->c_cpu = coa->cpu;
-		cc = callout_lock(c);
-	}
-#endif
-	if (coa->time < cc->cc_lastscan)
-		coa->time = cc->cc_lastscan;
-	c->c_arg = coa->arg;
-	c->c_func = coa->func;
-	c->c_time = coa->time;
-	c->c_precision = coa->precision;
-
+	if (sbt < cc->cc_lastscan)
+		sbt = cc->cc_lastscan;
+	c->c_arg = arg;
+	c->c_flags |= (CALLOUT_ACTIVE | CALLOUT_PENDING);
+	if (flags & C_DIRECT_EXEC)
+		c->c_flags |= CALLOUT_DIRECT;
+	c->c_flags &= ~CALLOUT_PROCESSED;
+	c->c_func = func;
+	c->c_time = sbt;
+	c->c_precision = precision;
 	bucket = callout_get_bucket(c->c_time);
 	CTR3(KTR_CALLOUT, "precision set for %p: %d.%08x",
 	    c, (int)(c->c_precision >> 32),
 	    (u_int)(c->c_precision & 0xffffffff));
 	LIST_INSERT_HEAD(&cc->cc_callwheel[bucket], c, c_links.le);
-
-	/* Ensure we are first to be scanned, if called via a callback */
 	if (cc->cc_bucket == bucket)
 		cc->cc_exec_next_dir = c;
 #ifndef NO_EVENTTIMERS
@@ -700,16 +617,17 @@ callout_cc_add_locked(struct callout *c, struct callout_cpu *cc,
 	sbt = c->c_time + c->c_precision;
 	if (sbt < cc->cc_firstevent) {
 		cc->cc_firstevent = sbt;
-		cpu_new_callout(coa->cpu, sbt, c->c_time);
+		cpu_new_callout(cpu, sbt, c->c_time);
 	}
 #endif
-	return (cc);
 }
 
 static void
 callout_cc_del(struct callout *c, struct callout_cpu *cc)
 {
 
+	if ((c->c_flags & CALLOUT_LOCAL_ALLOC) == 0)
+		return;
 	c->c_func = NULL;
 	SLIST_INSERT_HEAD(&cc->cc_callfree, c, c_links.sle);
 }
@@ -721,10 +639,20 @@ softclock_call_cc(struct callout *c, struct callout_cpu *cc,
 #endif
     int direct)
 {
-	callout_func_t *c_func;
+	struct rm_priotracker tracker;
+	void (*c_func)(void *);
 	void *c_arg;
+	struct lock_class *class;
 	struct lock_object *c_lock;
+	uintptr_t lock_status;
 	int c_flags;
+#ifdef SMP
+	struct callout_cpu *new_cc;
+	void (*new_func)(void *);
+	void *new_arg;
+	int flags, new_cpu;
+	sbintime_t new_prec, new_time;
+#endif
 #if defined(DIAGNOSTIC) || defined(CALLOUT_PROFILING) 
 	sbintime_t sbt1, sbt2;
 	struct timespec ts2;
@@ -735,39 +663,37 @@ softclock_call_cc(struct callout *c, struct callout_cpu *cc,
 	KASSERT((c->c_flags & (CALLOUT_PENDING | CALLOUT_ACTIVE)) ==
 	    (CALLOUT_PENDING | CALLOUT_ACTIVE),
 	    ("softclock_call_cc: pend|act %p %x", c, c->c_flags));
+	class = (c->c_lock != NULL) ? LOCK_CLASS(c->c_lock) : NULL;
+	lock_status = 0;
+	if (c->c_flags & CALLOUT_SHAREDLOCK) {
+		if (class == &lock_class_rm)
+			lock_status = (uintptr_t)&tracker;
+		else
+			lock_status = 1;
+	}
 	c_lock = c->c_lock;
 	c_func = c->c_func;
 	c_arg = c->c_arg;
 	c_flags = c->c_flags;
-
-	/* remove pending bit */
-	c->c_flags &= ~CALLOUT_PENDING;
-
-	/* reset our local state */
+	if (c->c_flags & CALLOUT_LOCAL_ALLOC)
+		c->c_flags = CALLOUT_LOCAL_ALLOC;
+	else
+		c->c_flags &= ~CALLOUT_PENDING;
 	cc->cc_exec_entity[direct].cc_curr = c;
-	cc->cc_exec_entity[direct].cc_restart = false;
-	cc->cc_exec_entity[direct].cc_drain_fn = NULL;
-	cc->cc_exec_entity[direct].cc_drain_arg = NULL;
-
+	cc->cc_exec_entity[direct].cc_cancel = false;
+	CC_UNLOCK(cc);
 	if (c_lock != NULL) {
-		cc->cc_exec_entity[direct].cc_cancel = false;
-		CC_UNLOCK(cc);
-
-		/* unlocked region for switching locks */
-
-		callout_lock_client(c_flags, c_lock);
-
+		class->lc_lock(c_lock, lock_status);
 		/*
-		 * Check if the callout may have been cancelled while
-		 * we were switching locks. Even though the callout is
-		 * specifying a lock, it might not be certain this
-		 * lock is locked when starting and stopping callouts.
+		 * The callout may have been cancelled
+		 * while we switched locks.
 		 */
-		CC_LOCK(cc);
 		if (cc->cc_exec_entity[direct].cc_cancel) {
-			callout_unlock_client(c_flags, c_lock);
-			goto skip_cc_locked;
+			class->lc_unlock(c_lock);
+			goto skip;
 		}
+		/* The callout cannot be stopped now. */
+		cc->cc_exec_entity[direct].cc_cancel = true;
 		if (c_lock == &Giant.lock_object) {
 #ifdef CALLOUT_PROFILING
 			(*gcalls)++;
@@ -788,11 +714,6 @@ softclock_call_cc(struct callout *c, struct callout_cpu *cc,
 		CTR3(KTR_CALLOUT, "callout %p func %p arg %p",
 		    c, c_func, c_arg);
 	}
-	/* The callout cannot be stopped now! */
-	cc->cc_exec_entity[direct].cc_cancel = true;
-	CC_UNLOCK(cc);
-
-	/* unlocked region */
 	KTR_STATE3(KTR_SCHED, "callout", cc->cc_ktr_event_name, "running",
 	    "func:%p", c_func, "arg:%p", c_arg, "direct:%d", direct);
 #if defined(DIAGNOSTIC) || defined(CALLOUT_PROFILING)
@@ -819,40 +740,85 @@ softclock_call_cc(struct callout *c, struct callout_cpu *cc,
 #endif
 	KTR_STATE0(KTR_SCHED, "callout", cc->cc_ktr_event_name, "idle");
 	CTR1(KTR_CALLOUT, "callout %p finished", c);
-
-	/*
-	 * At this point the callback structure might have been freed,
-	 * so we need to check the previously copied value of
-	 * "c->c_flags":
-	 */
 	if ((c_flags & CALLOUT_RETURNUNLOCKED) == 0)
-		callout_unlock_client(c_flags, c_lock);
-
+		class->lc_unlock(c_lock);
+skip:
 	CC_LOCK(cc);
-
-skip_cc_locked:
 	KASSERT(cc->cc_exec_entity[direct].cc_curr == c, ("mishandled cc_curr"));
 	cc->cc_exec_entity[direct].cc_curr = NULL;
-
-	/* Check if there is anything which needs draining */
-	if (cc->cc_exec_entity[direct].cc_drain_fn != NULL) {
+	if (cc->cc_exec_entity[direct].cc_waiting) {
 		/*
-		 * Unlock the CPU callout last, so that any use of
-		 * structures belonging to the callout are complete:
+		 * There is someone waiting for the
+		 * callout to complete.
+		 * If the callout was scheduled for
+		 * migration just cancel it.
 		 */
+		if (cc_cce_migrating(cc, direct)) {
+			cc_cce_cleanup(cc, direct);
+
+			/*
+			 * It should be assert here that the callout is not
+			 * destroyed but that is not easy.
+			 */
+			c->c_flags &= ~CALLOUT_DFRMIGRATION;
+		}
+		cc->cc_exec_entity[direct].cc_waiting = false;
 		CC_UNLOCK(cc);
-		/* call drain function unlocked */
-		cc->cc_exec_entity[direct].cc_drain_fn(
-		    cc->cc_exec_entity[direct].cc_drain_arg);
+		wakeup(&cc->cc_exec_entity[direct].cc_waiting);
 		CC_LOCK(cc);
-	} else if (c_flags & CALLOUT_LOCAL_ALLOC) {
-		/* return callout back to freelist */
-		callout_cc_del(c, cc);
-	} else if (cc->cc_exec_entity[direct].cc_restart) {
-		/* [re-]schedule callout, if any */
-		cc = callout_cc_add_locked(c, cc,
-		    &cc->cc_exec_entity[direct].cc_restart_args, false);
+	} else if (cc_cce_migrating(cc, direct)) {
+		KASSERT((c_flags & CALLOUT_LOCAL_ALLOC) == 0,
+		    ("Migrating legacy callout %p", c));
+#ifdef SMP
+		/*
+		 * If the callout was scheduled for
+		 * migration just perform it now.
+		 */
+		new_cpu = cc->cc_exec_entity[direct].ce_migration_cpu;
+		new_time = cc->cc_exec_entity[direct].ce_migration_time;
+		new_prec = cc->cc_exec_entity[direct].ce_migration_prec;
+		new_func = cc->cc_exec_entity[direct].ce_migration_func;
+		new_arg = cc->cc_exec_entity[direct].ce_migration_arg;
+		cc_cce_cleanup(cc, direct);
+
+		/*
+		 * It should be assert here that the callout is not destroyed
+		 * but that is not easy.
+		 *
+		 * As first thing, handle deferred callout stops.
+		 */
+		if ((c->c_flags & CALLOUT_DFRMIGRATION) == 0) {
+			CTR3(KTR_CALLOUT,
+			     "deferred cancelled %p func %p arg %p",
+			     c, new_func, new_arg);
+			callout_cc_del(c, cc);
+			return;
+		}
+		c->c_flags &= ~CALLOUT_DFRMIGRATION;
+
+		new_cc = callout_cpu_switch(c, cc, new_cpu);
+		flags = (direct) ? C_DIRECT_EXEC : 0;
+		callout_cc_add(c, new_cc, new_time, new_prec, new_func,
+		    new_arg, new_cpu, flags);
+		CC_UNLOCK(new_cc);
+		CC_LOCK(cc);
+#else
+		panic("migration should not happen");
+#endif
 	}
+	/*
+	 * If the current callout is locally allocated (from
+	 * timeout(9)) then put it on the freelist.
+	 *
+	 * Note: we need to check the cached copy of c_flags because
+	 * if it was not local, then it's not safe to deref the
+	 * callout pointer.
+	 */
+	KASSERT((c_flags & CALLOUT_LOCAL_ALLOC) == 0 ||
+	    c->c_flags == CALLOUT_LOCAL_ALLOC,
+	    ("corrupted callout"));
+	if (c_flags & CALLOUT_LOCAL_ALLOC)
+		callout_cc_del(c, cc);
 }
 
 /*
@@ -933,11 +899,10 @@ timeout(timeout_t *ftn, void *arg, int to_ticks)
 		/* XXX Attempt to malloc first */
 		panic("timeout table full");
 	SLIST_REMOVE_HEAD(&cc->cc_callfree, c_links.sle);
+	callout_reset(new, to_ticks, ftn, arg);
 	handle.callout = new;
 	CC_UNLOCK(cc);
 
-	callout_reset(new, to_ticks, ftn, arg);
-
 	return (handle);
 }
 
@@ -945,7 +910,6 @@ void
 untimeout(timeout_t *ftn, void *arg, struct callout_handle handle)
 {
 	struct callout_cpu *cc;
-	bool match;
 
 	/*
 	 * Check for a handle that was initialized
@@ -956,11 +920,9 @@ untimeout(timeout_t *ftn, void *arg, struct callout_handle handle)
 		return;
 
 	cc = callout_lock(handle.callout);
-	match = (handle.callout->c_func == ftn && handle.callout->c_arg == arg);
-	CC_UNLOCK(cc);
-
-	if (match)
+	if (handle.callout->c_func == ftn && handle.callout->c_arg == arg)
 		callout_stop(handle.callout);
+	CC_UNLOCK(cc);
 }
 
 void
@@ -969,119 +931,6 @@ callout_handle_init(struct callout_handle *handle)
 	handle->callout = NULL;
 }
 
-static int
-callout_restart_async(struct callout *c, struct callout_args *coa,
-    callout_func_t *drain_fn, void *drain_arg)
-{
-	struct callout_cpu *cc;
-	int cancelled;
-	int direct;
-
-	cc = callout_lock(c);
-
-	/* Figure out if the callout is direct or not */
-	direct = ((c->c_flags & CALLOUT_DIRECT) != 0);
-
-	/*
-	 * Check if the callback is currently scheduled for
-	 * completion:
-	 */
-	if (cc->cc_exec_entity[direct].cc_curr == c) {
-		/*
-		 * Try to prevent the callback from running by setting
-		 * the "cc_cancel" variable to "true". Also check if
-		 * the callout was previously subject to a deferred
-		 * callout restart:
-		 */
-		if (cc->cc_exec_entity[direct].cc_cancel == false ||
-		    (c->c_flags & CALLOUT_DEFRESTART) != 0) {
-			cc->cc_exec_entity[direct].cc_cancel = true;
-			cancelled = 1;
-		} else {
-			cancelled = 0;
-		}
-
-		/*
-		 * Prevent callback restart if "callout_drain_xxx()"
-		 * is being called or we are stopping the callout or
-		 * the callback was preallocated by us:
-		 */
-		if (cc->cc_exec_entity[direct].cc_drain_fn != NULL ||
-		    coa == NULL || (c->c_flags & CALLOUT_LOCAL_ALLOC) != 0) {
-			CTR4(KTR_CALLOUT, "%s %p func %p arg %p",
-			    cancelled ? "cancelled and draining" : "draining",
-			    c, c->c_func, c->c_arg);
-
-			/* clear old flags, if any */
-			c->c_flags &= ~(CALLOUT_ACTIVE | CALLOUT_PENDING |
-			    CALLOUT_DEFRESTART | CALLOUT_PROCESSED);
-
-			/* clear restart flag, if any */
-			cc->cc_exec_entity[direct].cc_restart = false;
-
-			/* set drain function, if any */
-			if (drain_fn != NULL) {
-				cc->cc_exec_entity[direct].cc_drain_fn = drain_fn;
-				cc->cc_exec_entity[direct].cc_drain_arg = drain_arg;
-				cancelled |= 2;		/* XXX define the value */
-			}
-		} else {
-			CTR4(KTR_CALLOUT, "%s %p func %p arg %p",
-			    cancelled ? "cancelled and restarting" : "restarting",
-			    c, c->c_func, c->c_arg);
-
-			/* get us back into the game */
-			c->c_flags |= (CALLOUT_ACTIVE | CALLOUT_PENDING |
-			    CALLOUT_DEFRESTART);
-			c->c_flags &= ~CALLOUT_PROCESSED;
-
-			/* enable deferred restart */
-			cc->cc_exec_entity[direct].cc_restart = true;
-
-			/* store arguments for the deferred restart, if any */
-			cc->cc_exec_entity[direct].cc_restart_args = *coa;
-		}
-	} else {
-		/* stop callout */
-		if (c->c_flags & CALLOUT_PENDING) {
-			/*
-			 * The callback has not yet been executed, and
-			 * we simply just need to unlink it:
-			 */
-			if ((c->c_flags & CALLOUT_PROCESSED) == 0) {
-				if (cc->cc_exec_next_dir == c)
-					cc->cc_exec_next_dir = LIST_NEXT(c, c_links.le);
-				LIST_REMOVE(c, c_links.le);
-			} else {
-				TAILQ_REMOVE(&cc->cc_expireq, c, c_links.tqe);
-			}
-			cancelled = 1;
-		} else {
-			cancelled = 0;
-		}
-
-		CTR4(KTR_CALLOUT, "%s %p func %p arg %p",
-		    cancelled ? "rescheduled" : "scheduled",
-		    c, c->c_func, c->c_arg);
-
-		/* [re-]schedule callout, if any */
-		if (coa != NULL) {
-			cc = callout_cc_add_locked(c, cc, coa, true);
-		} else {
-			/* clear old flags, if any */
-			c->c_flags &= ~(CALLOUT_ACTIVE | CALLOUT_PENDING |
-			    CALLOUT_DEFRESTART | CALLOUT_PROCESSED);
-
-			/* return callback to pre-allocated list, if any */
-			if ((c->c_flags & CALLOUT_LOCAL_ALLOC) && cancelled != 0) {
-				callout_cc_del(c, cc);
-			}
-		}
-	}
-	CC_UNLOCK(cc);
-	return (cancelled);
-}
-
 /*
  * New interface; clients allocate their own callout structures.
  *
@@ -1100,32 +949,25 @@ callout_restart_async(struct callout *c, struct callout_args *coa,
  */
 int
 callout_reset_sbt_on(struct callout *c, sbintime_t sbt, sbintime_t precision,
-    callout_func_t *ftn, void *arg, int cpu, int flags)
+    void (*ftn)(void *), void *arg, int cpu, int flags)
 {
-	struct callout_args coa;
-
-	/* store arguments for callout add function */
-	coa.func = ftn;
-	coa.arg = arg;
-	coa.precision = precision;
-	coa.flags = flags;
-	coa.cpu = cpu;
-
-	/* compute the rest of the arguments needed */
-	if (coa.flags & C_ABSOLUTE) {
-		coa.time = sbt;
-	} else {
-		sbintime_t pr;
+	sbintime_t to_sbt, pr;
+	struct callout_cpu *cc;
+	int cancelled, direct;
 
-		if ((coa.flags & C_HARDCLOCK) && (sbt < tick_sbt))
+	cancelled = 0;
+	if (flags & C_ABSOLUTE) {
+		to_sbt = sbt;
+	} else {
+		if ((flags & C_HARDCLOCK) && (sbt < tick_sbt))
 			sbt = tick_sbt;
-		if ((coa.flags & C_HARDCLOCK) ||
+		if ((flags & C_HARDCLOCK) ||
 #ifdef NO_EVENTTIMERS
 		    sbt >= sbt_timethreshold) {
-			coa.time = getsbinuptime();
+			to_sbt = getsbinuptime();
 
 			/* Add safety belt for the case of hz > 1000. */
-			coa.time += tc_tick_sbt - tick_sbt;
+			to_sbt += tc_tick_sbt - tick_sbt;
 #else
 		    sbt >= sbt_tickthreshold) {
 			/*
@@ -1135,29 +977,101 @@ callout_reset_sbt_on(struct callout *c, sbintime_t sbt, sbintime_t precision,
 			 * active ones.
 			 */
 #ifdef __LP64__
-			coa.time = DPCPU_GET(hardclocktime);
+			to_sbt = DPCPU_GET(hardclocktime);
 #else
 			spinlock_enter();
-			coa.time = DPCPU_GET(hardclocktime);
+			to_sbt = DPCPU_GET(hardclocktime);
 			spinlock_exit();
 #endif
 #endif
-			if ((coa.flags & C_HARDCLOCK) == 0)
-				coa.time += tick_sbt;
+			if ((flags & C_HARDCLOCK) == 0)
+				to_sbt += tick_sbt;
 		} else
-			coa.time = sbinuptime();
-		if (SBT_MAX - coa.time < sbt)
-			coa.time = SBT_MAX;
+			to_sbt = sbinuptime();
+		if (SBT_MAX - to_sbt < sbt)
+			to_sbt = SBT_MAX;
 		else
-			coa.time += sbt;
-		pr = ((C_PRELGET(coa.flags) < 0) ? sbt >> tc_precexp :
-		    sbt >> C_PRELGET(coa.flags));
-		if (pr > coa.precision)
-			coa.precision = pr;
+			to_sbt += sbt;
+		pr = ((C_PRELGET(flags) < 0) ? sbt >> tc_precexp :
+		    sbt >> C_PRELGET(flags));
+		if (pr > precision)
+			precision = pr;
+	}
+	/*
+	 * Don't allow migration of pre-allocated callouts lest they
+	 * become unbalanced.
+	 */
+	if (c->c_flags & CALLOUT_LOCAL_ALLOC)
+		cpu = c->c_cpu;
+	direct = (c->c_flags & CALLOUT_DIRECT) != 0;
+	KASSERT(!direct || c->c_lock == NULL,
+	    ("%s: direct callout %p has lock", __func__, c));
+	cc = callout_lock(c);
+	if (cc->cc_exec_entity[direct].cc_curr == c) {
+		/*
+		 * We're being asked to reschedule a callout which is
+		 * currently in progress.  If there is a lock then we
+		 * can cancel the callout if it has not really started.
+		 */
+		if (c->c_lock != NULL && !cc->cc_exec_entity[direct].cc_cancel)
+			cancelled = cc->cc_exec_entity[direct].cc_cancel = true;
+		if (cc->cc_exec_entity[direct].cc_waiting) {
+			/*
+			 * Someone has called callout_drain to kill this
+			 * callout.  Don't reschedule.
+			 */
+			CTR4(KTR_CALLOUT, "%s %p func %p arg %p",
+			    cancelled ? "cancelled" : "failed to cancel",
+			    c, c->c_func, c->c_arg);
+			CC_UNLOCK(cc);
+			return (cancelled);
+		}
+	}
+	if (c->c_flags & CALLOUT_PENDING) {
+		if ((c->c_flags & CALLOUT_PROCESSED) == 0) {
+			if (cc->cc_exec_next_dir == c)
+				cc->cc_exec_next_dir = LIST_NEXT(c, c_links.le);
+			LIST_REMOVE(c, c_links.le);
+		} else
+			TAILQ_REMOVE(&cc->cc_expireq, c, c_links.tqe);
+		cancelled = 1;
+		c->c_flags &= ~(CALLOUT_ACTIVE | CALLOUT_PENDING);
+	}
+
+#ifdef SMP
+	/*
+	 * If the callout must migrate try to perform it immediately.
+	 * If the callout is currently running, just defer the migration
+	 * to a more appropriate moment.
+	 */
+	if (c->c_cpu != cpu) {
+		if (cc->cc_exec_entity[direct].cc_curr == c) {
+			cc->cc_exec_entity[direct].ce_migration_cpu = cpu;
+			cc->cc_exec_entity[direct].ce_migration_time
+			    = to_sbt;
+			cc->cc_exec_entity[direct].ce_migration_prec 
+			    = precision;
+			cc->cc_exec_entity[direct].ce_migration_func = ftn;
+			cc->cc_exec_entity[direct].ce_migration_arg = arg;
+			c->c_flags |= CALLOUT_DFRMIGRATION;
+			CTR6(KTR_CALLOUT,
+		    "migration of %p func %p arg %p in %d.%08x to %u deferred",
+			    c, c->c_func, c->c_arg, (int)(to_sbt >> 32),
+			    (u_int)(to_sbt & 0xffffffff), cpu);
+			CC_UNLOCK(cc);
+			return (cancelled);
+		}
+		cc = callout_cpu_switch(c, cc, cpu);
 	}
+#endif
 
-	/* get callback started, if any */
-	return (callout_restart_async(c, &coa, NULL, NULL));
+	callout_cc_add(c, cc, to_sbt, precision, ftn, arg, cpu, flags);
+	CTR6(KTR_CALLOUT, "%sscheduled %p func %p arg %p in %d.%08x",
+	    cancelled ? "re" : "", c, c->c_func, c->c_arg, (int)(to_sbt >> 32),
+	    (u_int)(to_sbt & 0xffffffff));
+	CC_UNLOCK(cc);
+
+	return (cancelled);
 }
 
 /*
@@ -1176,105 +1090,204 @@ callout_schedule(struct callout *c, int to_ticks)
 }
 
 int
-callout_stop(struct callout *c)
+_callout_stop_safe(struct callout *c, int safe)
 {
-	/* get callback stopped, if any */
-	return (callout_restart_async(c, NULL, NULL, NULL));
-}
-
-static void
-callout_drain_function(void *arg)
-{
-	wakeup(arg);
-}
-
-int
-callout_drain_async(struct callout *c, callout_func_t *fn, void *arg)
-{
-	/* get callback stopped, if any */
-	return (callout_restart_async(c, NULL, fn, arg) & 2);
-}
-
-int
-callout_drain(struct callout *c)
-{
-	int cancelled;
-
-	WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL,
-	    "Draining callout");
-
-	callout_lock_client(c->c_flags, c->c_lock);
-
-	/* at this point the "c->c_cpu" field is not changing */
+	struct callout_cpu *cc, *old_cc;
+	struct lock_class *class;
+	int direct, sq_locked, use_lock;
 
-	cancelled = callout_drain_async(c, &callout_drain_function, c);
-
-	if (cancelled != 0) {
-		struct callout_cpu *cc;
-		int direct;
+	/*
+	 * Some old subsystems don't hold Giant while running a callout_stop(),
+	 * so just discard this check for the moment.
+	 */
+	if (!safe && c->c_lock != NULL) {
+		if (c->c_lock == &Giant.lock_object)
+			use_lock = mtx_owned(&Giant);
+		else {
+			use_lock = 1;
+			class = LOCK_CLASS(c->c_lock);
+			class->lc_assert(c->c_lock, LA_XLOCKED);
+		}
+	} else
+		use_lock = 0;
+	direct = (c->c_flags & CALLOUT_DIRECT) != 0;
+	sq_locked = 0;
+	old_cc = NULL;
+again:
+	cc = callout_lock(c);
 
-		CTR3(KTR_CALLOUT, "need to drain %p func %p arg %p",
-		    c, c->c_func, c->c_arg);
+	/*
+	 * If the callout was migrating while the callout cpu lock was
+	 * dropped,  just drop the sleepqueue lock and check the states
+	 * again.
+	 */
+	if (sq_locked != 0 && cc != old_cc) {
+#ifdef SMP
+		CC_UNLOCK(cc);
+		sleepq_release(&old_cc->cc_exec_entity[direct].cc_waiting);
+		sq_locked = 0;
+		old_cc = NULL;
+		goto again;
+#else
+		panic("migration should not happen");
+#endif
+	}
 
-		cc = callout_lock(c);
-		direct = ((c->c_flags & CALLOUT_DIRECT) != 0);
+	/*
+	 * If the callout isn't pending, it's not on the queue, so
+	 * don't attempt to remove it from the queue.  We can try to
+	 * stop it by other means however.
+	 */
+	if (!(c->c_flags & CALLOUT_PENDING)) {
+		c->c_flags &= ~CALLOUT_ACTIVE;
 
 		/*
-		 * We've gotten our callout CPU lock, it is safe to
-		 * drop the initial lock:
+		 * If it wasn't on the queue and it isn't the current
+		 * callout, then we can't stop it, so just bail.
 		 */
-		callout_unlock_client(c->c_flags, c->c_lock);
-
-		/* Wait for drain to complete */
+		if (cc->cc_exec_entity[direct].cc_curr != c) {
+			CTR3(KTR_CALLOUT, "failed to stop %p func %p arg %p",
+			    c, c->c_func, c->c_arg);
+			CC_UNLOCK(cc);
+			if (sq_locked)
+				sleepq_release(
+				    &cc->cc_exec_entity[direct].cc_waiting);
+			return (0);
+		}
 
-		while (cc->cc_exec_entity[direct].cc_curr == c)
-			msleep_spin(c, (struct mtx *)&cc->cc_lock, "codrain", 0);
+		if (safe) {
+			/*
+			 * The current callout is running (or just
+			 * about to run) and blocking is allowed, so
+			 * just wait for the current invocation to
+			 * finish.
+			 */
+			while (cc->cc_exec_entity[direct].cc_curr == c) {
+				/*
+				 * Use direct calls to sleepqueue interface
+				 * instead of cv/msleep in order to avoid
+				 * a LOR between cc_lock and sleepqueue
+				 * chain spinlocks.  This piece of code
+				 * emulates a msleep_spin() call actually.
+				 *
+				 * If we already have the sleepqueue chain
+				 * locked, then we can safely block.  If we
+				 * don't already have it locked, however,
+				 * we have to drop the cc_lock to lock
+				 * it.  This opens several races, so we
+				 * restart at the beginning once we have
+				 * both locks.  If nothing has changed, then
+				 * we will end up back here with sq_locked
+				 * set.
+				 */
+				if (!sq_locked) {
+					CC_UNLOCK(cc);
+					sleepq_lock(
+					&cc->cc_exec_entity[direct].cc_waiting);
+					sq_locked = 1;
+					old_cc = cc;
+					goto again;
+				}
 
+				/*
+				 * Migration could be cancelled here, but
+				 * as long as it is still not sure when it
+				 * will be packed up, just let softclock()
+				 * take care of it.
+				 */
+				cc->cc_exec_entity[direct].cc_waiting = true;
+				DROP_GIANT();
+				CC_UNLOCK(cc);
+				sleepq_add(
+				    &cc->cc_exec_entity[direct].cc_waiting,
+				    &cc->cc_lock.lock_object, "codrain",
+				    SLEEPQ_SLEEP, 0);
+				sleepq_wait(
+				    &cc->cc_exec_entity[direct].cc_waiting,
+					     0);
+				sq_locked = 0;
+				old_cc = NULL;
+
+				/* Reacquire locks previously released. */
+				PICKUP_GIANT();
+				CC_LOCK(cc);
+			}
+		} else if (use_lock &&
+			    !cc->cc_exec_entity[direct].cc_cancel) {
+			/*
+			 * The current callout is waiting for its
+			 * lock which we hold.  Cancel the callout
+			 * and return.  After our caller drops the
+			 * lock, the callout will be skipped in
+			 * softclock().
+			 */
+			cc->cc_exec_entity[direct].cc_cancel = true;
+			CTR3(KTR_CALLOUT, "cancelled %p func %p arg %p",
+			    c, c->c_func, c->c_arg);
+			KASSERT(!cc_cce_migrating(cc, direct),
+			    ("callout wrongly scheduled for migration"));
+			CC_UNLOCK(cc);
+			KASSERT(!sq_locked, ("sleepqueue chain locked"));
+			return (1);
+		} else if ((c->c_flags & CALLOUT_DFRMIGRATION) != 0) {
+			c->c_flags &= ~CALLOUT_DFRMIGRATION;
+			CTR3(KTR_CALLOUT, "postponing stop %p func %p arg %p",
+			    c, c->c_func, c->c_arg);
+			CC_UNLOCK(cc);
+			return (1);
+		}
+		CTR3(KTR_CALLOUT, "failed to stop %p func %p arg %p",
+		    c, c->c_func, c->c_arg);
 		CC_UNLOCK(cc);
-	} else {
-		callout_unlock_client(c->c_flags, c->c_lock);
+		KASSERT(!sq_locked, ("sleepqueue chain still locked"));
+		return (0);
 	}
+	if (sq_locked)
+		sleepq_release(&cc->cc_exec_entity[direct].cc_waiting);
+
+	c->c_flags &= ~(CALLOUT_ACTIVE | CALLOUT_PENDING);
 
 	CTR3(KTR_CALLOUT, "cancelled %p func %p arg %p",
 	    c, c->c_func, c->c_arg);
+	if ((c->c_flags & CALLOUT_PROCESSED) == 0) {
+		if (cc->cc_exec_next_dir == c)
+			cc->cc_exec_next_dir = LIST_NEXT(c, c_links.le);
+		LIST_REMOVE(c, c_links.le);
+	} else
+		TAILQ_REMOVE(&cc->cc_expireq, c, c_links.tqe);
+	callout_cc_del(c, cc);
 
-	return (cancelled & 1);
+	CC_UNLOCK(cc);
+	return (1);
 }
 
 void
 callout_init(struct callout *c, int mpsafe)
 {
+	bzero(c, sizeof *c);
 	if (mpsafe) {
-		_callout_init_lock(c, NULL, CALLOUT_RETURNUNLOCKED);
+		c->c_lock = NULL;
+		c->c_flags = CALLOUT_RETURNUNLOCKED;
 	} else {
-		_callout_init_lock(c, &Giant.lock_object, 0);
+		c->c_lock = &Giant.lock_object;
+		c->c_flags = 0;
 	}
+	c->c_cpu = timeout_cpu;
 }
 
 void
 _callout_init_lock(struct callout *c, struct lock_object *lock, int flags)
 {
 	bzero(c, sizeof *c);
-	KASSERT((flags & ~CALLOUT_RETURNUNLOCKED) == 0,
-	    ("callout_init_lock: bad flags 0x%08x", flags));
-	flags &= CALLOUT_RETURNUNLOCKED;
-	if (lock != NULL) {
-		struct lock_class *class = LOCK_CLASS(lock);
-		if (class == &lock_class_mtx_sleep)
-			flags |= CALLOUT_SET_LC(CALLOUT_LC_MUTEX);
-		else if (class == &lock_class_mtx_spin)
-			flags |= CALLOUT_SET_LC(CALLOUT_LC_SPIN);
-		else if (class == &lock_class_rm)
-			flags |= CALLOUT_SET_LC(CALLOUT_LC_RM);
-		else if (class == &lock_class_rw)
-			flags |= CALLOUT_SET_LC(CALLOUT_LC_RW);
-		else
-			panic("callout_init_lock: Unsupported lock class '%s'\n", class->lc_name);
-	} else {
-		flags |= CALLOUT_SET_LC(CALLOUT_LC_UNUSED_0);
-	}
 	c->c_lock = lock;
-	c->c_flags = flags;
+	KASSERT((flags & ~(CALLOUT_RETURNUNLOCKED | CALLOUT_SHAREDLOCK)) == 0,
+	    ("callout_init_lock: bad flags %d", flags));
+	KASSERT(lock != NULL || (flags & CALLOUT_RETURNUNLOCKED) == 0,
+	    ("callout_init_lock: CALLOUT_RETURNUNLOCKED with no lock"));
+	KASSERT(lock == NULL || !(LOCK_CLASS(lock)->lc_flags &
+	    (LC_SPINLOCK | LC_SLEEPABLE)), ("%s: invalid lock class",
+	    __func__));
+	c->c_flags = flags & (CALLOUT_RETURNUNLOCKED | CALLOUT_SHAREDLOCK);
 	c->c_cpu = timeout_cpu;
 }
 
diff --git a/sys/kern/subr_sleepqueue.c b/sys/kern/subr_sleepqueue.c
index 18dc2a058f50..bbbec920ef0d 100644
--- a/sys/kern/subr_sleepqueue.c
+++ b/sys/kern/subr_sleepqueue.c
@@ -152,8 +152,7 @@ static uma_zone_t sleepq_zone;
  */
 static int	sleepq_catch_signals(void *wchan, int pri);
 static int	sleepq_check_signals(void);
-static int	sleepq_check_timeout(struct thread *);
-static void	sleepq_stop_timeout(struct thread *);
+static int	sleepq_check_timeout(void);
 #ifdef INVARIANTS
 static void	sleepq_dtor(void *mem, int size, void *arg);
 #endif
@@ -374,14 +373,17 @@ void
 sleepq_set_timeout_sbt(void *wchan, sbintime_t sbt, sbintime_t pr,
     int flags)
 {
+	struct sleepqueue_chain *sc;
 	struct thread *td;
 
 	td = curthread;
-
-	mtx_lock_spin(&td->td_slpmutex);
+	sc = SC_LOOKUP(wchan);
+	mtx_assert(&sc->sc_lock, MA_OWNED);
+	MPASS(TD_ON_SLEEPQ(td));
+	MPASS(td->td_sleepqueue == NULL);
+	MPASS(wchan != NULL);
 	callout_reset_sbt_on(&td->td_slpcallout, sbt, pr,
 	    sleepq_timeout, td, PCPU_GET(cpuid), flags | C_DIRECT_EXEC);
-	mtx_unlock_spin(&td->td_slpmutex);
 }
 
 /*
@@ -557,8 +559,11 @@ sleepq_switch(void *wchan, int pri)
  * Check to see if we timed out.
  */
 static int
-sleepq_check_timeout(struct thread *td)
+sleepq_check_timeout(void)
 {
+	struct thread *td;
+
+	td = curthread;
 	THREAD_LOCK_ASSERT(td, MA_OWNED);
 
 	/*
@@ -568,18 +573,25 @@ sleepq_check_timeout(struct thread *td)
 		td->td_flags &= ~TDF_TIMEOUT;
 		return (EWOULDBLOCK);
 	}
-	return (0);
-}
 
-/*
- * Atomically stop the timeout by using a mutex.
- */
-static void
-sleepq_stop_timeout(struct thread *td)
-{
-	mtx_lock_spin(&td->td_slpmutex);
-	callout_stop(&td->td_slpcallout);
-	mtx_unlock_spin(&td->td_slpmutex);
+	/*
+	 * If TDF_TIMOFAIL is set, the timeout ran after we had
+	 * already been woken up.
+	 */
+	if (td->td_flags & TDF_TIMOFAIL)
+		td->td_flags &= ~TDF_TIMOFAIL;
+
+	/*
+	 * If callout_stop() fails, then the timeout is running on
+	 * another CPU, so synchronize with it to avoid having it
+	 * accidentally wake up a subsequent sleep.
+	 */
+	else if (callout_stop(&td->td_slpcallout) == 0) {
+		td->td_flags |= TDF_TIMEOUT;
+		TD_SET_SLEEPING(td);
+		mi_switch(SW_INVOL | SWT_SLEEPQTIMO, NULL);
+	}
+	return (0);
 }
 
 /*
@@ -652,11 +664,9 @@ sleepq_timedwait(void *wchan, int pri)
 	MPASS(!(td->td_flags & TDF_SINTR));
 	thread_lock(td);
 	sleepq_switch(wchan, pri);
-	rval = sleepq_check_timeout(td);
+	rval = sleepq_check_timeout();
 	thread_unlock(td);
 
-	sleepq_stop_timeout(td);
-
 	return (rval);
 }
 
@@ -667,18 +677,12 @@ sleepq_timedwait(void *wchan, int pri)
 int
 sleepq_timedwait_sig(void *wchan, int pri)
 {
-	struct thread *td;
 	int rcatch, rvalt, rvals;
 
-	td = curthread;
-
 	rcatch = sleepq_catch_signals(wchan, pri);
-	rvalt = sleepq_check_timeout(td);
+	rvalt = sleepq_check_timeout();
 	rvals = sleepq_check_signals();
-	thread_unlock(td);
-
-	sleepq_stop_timeout(td);
-
+	thread_unlock(curthread);
 	if (rcatch)
 		return (rcatch);
 	if (rvals)
@@ -885,49 +889,64 @@ sleepq_broadcast(void *wchan, int flags, int pri, int queue)
 static void
 sleepq_timeout(void *arg)
 {
-	struct thread *td = arg;
-	int wakeup_swapper = 0;
+	struct sleepqueue_chain *sc;
+	struct sleepqueue *sq;
+	struct thread *td;
+	void *wchan;
+	int wakeup_swapper;
 
+	td = arg;
+	wakeup_swapper = 0;
 	CTR3(KTR_PROC, "sleepq_timeout: thread %p (pid %ld, %s)",
 	    (void *)td, (long)td->td_proc->p_pid, (void *)td->td_name);
 
-	/* Handle the three cases which can happen */
-
+	/*
+	 * First, see if the thread is asleep and get the wait channel if
+	 * it is.
+	 */
 	thread_lock(td);
-	if (TD_ON_SLEEPQ(td)) {
-		if (TD_IS_SLEEPING(td)) {
-			struct sleepqueue_chain *sc;
-			struct sleepqueue *sq;
-			void *wchan;
+	if (TD_IS_SLEEPING(td) && TD_ON_SLEEPQ(td)) {
+		wchan = td->td_wchan;
+		sc = SC_LOOKUP(wchan);
+		THREAD_LOCKPTR_ASSERT(td, &sc->sc_lock);
+		sq = sleepq_lookup(wchan);
+		MPASS(sq != NULL);
+		td->td_flags |= TDF_TIMEOUT;
+		wakeup_swapper = sleepq_resume_thread(sq, td, 0);
+		thread_unlock(td);
+		if (wakeup_swapper)
+			kick_proc0();
+		return;
+	}
 
-			/*
-			 * Case I - thread is asleep and needs to be
-			 * awoken:
-			 */
-			wchan = td->td_wchan;
-			sc = SC_LOOKUP(wchan);
-			THREAD_LOCKPTR_ASSERT(td, &sc->sc_lock);
-			sq = sleepq_lookup(wchan);
-			MPASS(sq != NULL);
-			td->td_flags |= TDF_TIMEOUT;
-			wakeup_swapper = sleepq_resume_thread(sq, td, 0);
-		} else {
-			/*
-			 * Case II - cancel going to sleep by setting
-			 * the timeout flag because the target thread
-			 * is not asleep yet. It can be on another CPU
-			 * in between sleepq_add() and one of the
-			 * sleepq_*wait*() routines or it can be in
-			 * sleepq_catch_signals().
-			 */
-			td->td_flags |= TDF_TIMEOUT;
-		}
-	} else {
-		/*
-		 * Case III - thread is already woken up by a wakeup
-		 * call and should not timeout. Nothing to do!
-		 */
+	/*
+	 * If the thread is on the SLEEPQ but isn't sleeping yet, it
+	 * can either be on another CPU in between sleepq_add() and
+	 * one of the sleepq_*wait*() routines or it can be in
+	 * sleepq_catch_signals().
+	 */
+	if (TD_ON_SLEEPQ(td)) {
+		td->td_flags |= TDF_TIMEOUT;
+		thread_unlock(td);
+		return;
 	}
+
+	/*
+	 * Now check for the edge cases.  First, if TDF_TIMEOUT is set,
+	 * then the other thread has already yielded to us, so clear
+	 * the flag and resume it.  If TDF_TIMEOUT is not set, then the
+	 * we know that the other thread is not on a sleep queue, but it
+	 * hasn't resumed execution yet.  In that case, set TDF_TIMOFAIL
+	 * to let it know that the timeout has already run and doesn't
+	 * need to be canceled.
+	 */
+	if (td->td_flags & TDF_TIMEOUT) {
+		MPASS(TD_IS_SLEEPING(td));
+		td->td_flags &= ~TDF_TIMEOUT;
+		TD_CLR_SLEEPING(td);
+		wakeup_swapper = setrunnable(td);
+	} else
+		td->td_flags |= TDF_TIMOFAIL;
 	thread_unlock(td);
 	if (wakeup_swapper)
 		kick_proc0();
diff --git a/sys/ofed/include/linux/completion.h b/sys/ofed/include/linux/completion.h
index 088828399c78..df4aec3595e4 100644
--- a/sys/ofed/include/linux/completion.h
+++ b/sys/ofed/include/linux/completion.h
@@ -105,9 +105,7 @@ _wait_for_timeout_common(struct completion *c, long timeout, int flags)
 		if (c->done)
 			break;
 		sleepq_add(c, NULL, "completion", flags, 0);
-		sleepq_release(c);
 		sleepq_set_timeout(c, end - ticks);
-		sleepq_lock(c);
 		if (flags & SLEEPQ_INTERRUPTIBLE) {
 			if (sleepq_timedwait_sig(c, 0) != 0)
 				return (-ERESTARTSYS);
diff --git a/sys/sys/_callout.h b/sys/sys/_callout.h
index f58a3835c973..e186aecb8ddf 100644
--- a/sys/sys/_callout.h
+++ b/sys/sys/_callout.h
@@ -46,30 +46,19 @@ LIST_HEAD(callout_list, callout);
 SLIST_HEAD(callout_slist, callout);
 TAILQ_HEAD(callout_tailq, callout);
 
-typedef void callout_func_t(void *);
-
-struct callout_args {
-	sbintime_t time;		/* absolute time for the event */
-	sbintime_t precision;		/* delta allowed wrt opt */
-	void	*arg;			/* function argument */
-	callout_func_t *func;		/* function to call */
-	int	flags;			/* flags passed to callout_reset() */
-	int	cpu;			/* CPU we're scheduled on */
-};
-
 struct callout {
 	union {
 		LIST_ENTRY(callout) le;
 		SLIST_ENTRY(callout) sle;
 		TAILQ_ENTRY(callout) tqe;
 	} c_links;
-	sbintime_t c_time;			/* absolute time for the event */
+	sbintime_t c_time;			/* ticks to the event */
 	sbintime_t c_precision;			/* delta allowed wrt opt */
 	void	*c_arg;				/* function argument */
-	callout_func_t *c_func;			/* function to call */
-	struct lock_object *c_lock;		/* callback lock */
+	void	(*c_func)(void *);		/* function to call */
+	struct lock_object *c_lock;		/* lock to handle */
 	int	c_flags;			/* state of this entry */
-	int	c_cpu;				/* CPU we're scheduled on */
+	volatile int c_cpu;			/* CPU we're scheduled on */
 };
 
 #endif
diff --git a/sys/sys/callout.h b/sys/sys/callout.h
index 235da99f039a..1096cb26ff92 100644
--- a/sys/sys/callout.h
+++ b/sys/sys/callout.h
@@ -45,12 +45,10 @@
 #define	CALLOUT_PENDING		0x0004 /* callout is waiting for timeout */
 #define	CALLOUT_MPSAFE		0x0008 /* callout handler is mp safe */
 #define	CALLOUT_RETURNUNLOCKED	0x0010 /* handler returns with mtx unlocked */
-#define	CALLOUT_UNUSED_5	0x0020 /* --available-- */
-#define	CALLOUT_DEFRESTART	0x0040 /* callout restart is deferred */
+#define	CALLOUT_SHAREDLOCK	0x0020 /* callout lock held in shared mode */
+#define	CALLOUT_DFRMIGRATION	0x0040 /* callout in deferred migration mode */
 #define	CALLOUT_PROCESSED	0x0080 /* callout in wheel or processing list? */
 #define	CALLOUT_DIRECT 		0x0100 /* allow exec from hw int context */
-#define	CALLOUT_SET_LC(x)	(((x) & 7) << 16) /* set lock class */
-#define	CALLOUT_GET_LC(x)	(((x) >> 16) & 7) /* get lock class */
 
 #define	C_DIRECT_EXEC		0x0001 /* direct execution of callout */
 #define	C_PRELBITS		7
@@ -67,8 +65,7 @@ struct callout_handle {
 #ifdef _KERNEL
 #define	callout_active(c)	((c)->c_flags & CALLOUT_ACTIVE)
 #define	callout_deactivate(c)	((c)->c_flags &= ~CALLOUT_ACTIVE)
-int	callout_drain(struct callout *);
-int	callout_drain_async(struct callout *, callout_func_t *, void *);
+#define	callout_drain(c)	_callout_stop_safe(c, 1)
 void	callout_init(struct callout *, int);
 void	_callout_init_lock(struct callout *, struct lock_object *, int);
 #define	callout_init_mtx(c, mtx, flags)					\
@@ -82,7 +79,7 @@ void	_callout_init_lock(struct callout *, struct lock_object *, int);
 	   NULL, (flags))
 #define	callout_pending(c)	((c)->c_flags & CALLOUT_PENDING)
 int	callout_reset_sbt_on(struct callout *, sbintime_t, sbintime_t,
-	    callout_func_t *, void *, int, int);
+	    void (*)(void *), void *, int, int);
 #define	callout_reset_sbt(c, sbt, pr, fn, arg, flags)			\
     callout_reset_sbt_on((c), (sbt), (pr), (fn), (arg), (c)->c_cpu, (flags))
 #define	callout_reset_sbt_curcpu(c, sbt, pr, fn, arg, flags)		\
@@ -106,7 +103,8 @@ int	callout_schedule(struct callout *, int);
 int	callout_schedule_on(struct callout *, int, int);
 #define	callout_schedule_curcpu(c, on_tick)				\
     callout_schedule_on((c), (on_tick), PCPU_GET(cpuid))
-int	callout_stop(struct callout *);
+#define	callout_stop(c)		_callout_stop_safe(c, 0)
+int	_callout_stop_safe(struct callout *, int);
 void	callout_process(sbintime_t now);
 
 #endif
diff --git a/sys/sys/param.h b/sys/sys/param.h
index 0babf847870c..bf59b0bf7226 100644
--- a/sys/sys/param.h
+++ b/sys/sys/param.h
@@ -58,7 +58,7 @@
  *		in the range 5 to 9.
  */
 #undef __FreeBSD_version
-#define __FreeBSD_version 1100054	/* Master, propagated to newvers */
+#define __FreeBSD_version 1100055	/* Master, propagated to newvers */
 
 /*
  * __FreeBSD_kernel__ indicates that this system uses the kernel of FreeBSD,
diff --git a/sys/sys/proc.h b/sys/sys/proc.h
index e694d918c45e..9b6c695d37fb 100644
--- a/sys/sys/proc.h
+++ b/sys/sys/proc.h
@@ -308,7 +308,6 @@ struct thread {
 	} td_uretoff;			/* (k) Syscall aux returns. */
 #define td_retval	td_uretoff.tdu_retval
 	struct callout	td_slpcallout;	/* (h) Callout for sleep. */
-	struct mtx	td_slpmutex;	/* (h) Mutex for sleep callout */
 	struct trapframe *td_frame;	/* (k) */
 	struct vm_object *td_kstack_obj;/* (a) Kstack object. */
 	vm_offset_t	td_kstack;	/* (a) Kernel VA of kstack. */
@@ -365,7 +364,7 @@ do {									\
 #define	TDF_ALLPROCSUSP	0x00000200 /* suspended by SINGLE_ALLPROC */
 #define	TDF_BOUNDARY	0x00000400 /* Thread suspended at user boundary */
 #define	TDF_ASTPENDING	0x00000800 /* Thread has some asynchronous events. */
-#define	TDF_UNUSED12	0x00001000 /* --available-- */
+#define	TDF_TIMOFAIL	0x00001000 /* Timeout from sleep after we were awake. */
 #define	TDF_SBDRY	0x00002000 /* Stop only on usermode boundary. */
 #define	TDF_UPIBLOCKED	0x00004000 /* Thread blocked on user PI mutex. */
 #define	TDF_NEEDSUSPCHK	0x00008000 /* Thread may need to suspend. */
@@ -707,7 +706,7 @@ struct proc {
 #define	SWT_OWEPREEMPT		2	/* Switching due to opepreempt. */
 #define	SWT_TURNSTILE		3	/* Turnstile contention. */
 #define	SWT_SLEEPQ		4	/* Sleepq wait. */
-#define	SWT_UNUSED5		5	/* --available-- */
+#define	SWT_SLEEPQTIMO		5	/* Sleepq timeout wait. */
 #define	SWT_RELINQUISH		6	/* yield call. */
 #define	SWT_NEEDRESCHED		7	/* NEEDRESCHED was set. */
 #define	SWT_IDLE		8	/* Switching from the idle thread. */