9 files changed, 221 insertions, 48 deletions
diff --git a/share/man/man9/Makefile b/share/man/man9/Makefile
index fa768f5..702fafa 100644
--- a/share/man/man9/Makefile
+++ b/share/man/man9/Makefile
@@ -641,7 +641,11 @@ MLINKS+=counter.9 counter_u64_alloc.9 \
 	counter.9 counter_exit.9 \
 	counter.9 counter_u64_add_protected.9 \
 	counter.9 counter_u64_fetch.9 \
-	counter.9 counter_u64_zero.9
+	counter.9 counter_u64_zero.9 \
+	counter.9 SYSCTL_COUNTER_U64.9 \
+	counter.9 SYSCTL_ADD_COUNTER_U64.9 \
+	counter.9 SYSCTL_COUNTER_U64_ARRAY.9 \
+	counter.9 SYSCTL_ADD_COUNTER_U64_ARRAY.9
 MLINKS+=cpuset.9 CPUSET_T_INITIALIZER.9 \
 	cpuset.9 CPUSET_FSET.9 \
 	cpuset.9 CPU_CLR.9 \
diff --git a/share/man/man9/SYSCALL_MODULE.9 b/share/man/man9/SYSCALL_MODULE.9
index eb90aeb..8fc92aa 100644
--- a/share/man/man9/SYSCALL_MODULE.9
+++ b/share/man/man9/SYSCALL_MODULE.9
@@ -57,6 +57,10 @@ A pointer to an
 which saves the offset in
 .Vt "struct sysent"
 where the syscall is allocated.
+If the location pointed to by
+.Fa offset
+holds a non 0 number it will be used if possible.
+If it holds 0 then one will be assigned.
 .It Fa new_sysent
 is a pointer to a structure that specifies the function implementing
 the syscall and the number of arguments this function needs (see
@@ -72,6 +76,17 @@ The argument passed to the callback functions of the
 .Fa evh
 event handler when it is called.
 .El
+.Pp
+The syscall number assigned to the module can be retrieved using the
+.Xr modstat 3
+and
+.Xr modfind 3
+library functions in libc.
+The MACRO
+.Fn SYSCALL_MODULE_HELPER
+includes 
+.Fn SYSCALL_MODULE
+and much of its boilerplate code.
 .Sh EXAMPLES
 A minimal example for a syscall module can be found in
 .Pa /usr/share/examples/kld/syscall/module/syscall.c .
diff --git a/share/man/man9/counter.9 b/share/man/man9/counter.9
index f703ccf..0fcf2ad 100644
--- a/share/man/man9/counter.9
+++ b/share/man/man9/counter.9
@@ -25,7 +25,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd February 7, 2014
+.Dd March 14, 2016
 .Dt COUNTER 9
 .Os
 .Sh NAME
@@ -54,6 +54,8 @@
 .In sys/sysctl.h
 .Fn SYSCTL_COUNTER_U64 parent nbr name access ptr descr
 .Fn SYSCTL_ADD_COUNTER_U64 ctx parent nbr name access ptr descr
+.Fn SYSCTL_COUNTER_U64_ARRAY parent nbr name access ptr len descr
+.Fn SYSCTL_ADD_COUNTER_U64_ARRAY ctx parent nbr name access ptr len descr
 .Sh DESCRIPTION
 .Nm
 is a generic facility to create counters
@@ -150,6 +152,40 @@ argument should be a pointer to allocated
 A read of the oid returns value obtained through
 .Fn counter_u64_fetch .
 Any write to the oid zeroes it.
+.It Fn SYSCTL_COUNTER_U64_ARRAY parent nbr name access ptr len descr
+Declare a static
+.Xr sysctl  
+oid that would represent an array of
+.Nm .
+The
+.Fa ptr
+argument should be a pointer to allocated array of
+.Vt counter_u64_t's .
+The
+.Fa len
+argument should specify number of elements in the array.
+A read of the oid returns len-sized array of
+.Vt uint64_t
+values  obtained through
+.Fn counter_u64_fetch .
+Any write to the oid zeroes all array elements.
+.It Fn SYSCTL_ADD_COUNTER_U64_ARRAY ctx parent nbr name access ptr len descr
+Create a
+.Xr sysctl
+oid that would represent an array of 
+.Nm .
+The
+.Fa ptr
+argument should be a pointer to allocated array of
+.Vt counter_u64_t's .
+The
+.Fa len
+argument should specify number of elements in the array.
+A read of the oid returns len-sized array of 
+.Vt uint64_t 
+values  obtained through
+.Fn counter_u64_fetch .
+Any write to the oid zeroes all array elements.
 .El
 .Sh IMPLEMENTATION DETAILS
 On all architectures
@@ -179,6 +215,16 @@ amd64 single-instruction implementation.
 On some architectures updating a counter require a
 .Xr critical 9
 section.
+.Sh EXAMPLES
+The following example creates a static counter array exported to
+userspace through a sysctl:
+.Bd -literal -offset indent
+#define MY_SIZE 8
+static counter_u64_t array[MY_SIZE];
+SYSCTL_COUNTER_U64_ARRAY(_debug, OID_AUTO, counter_array, CTLFLAG_RW,
+    &array[0], MY_SIZE, "Test counter array");
+.Ed
+.Pp
 .Sh SEE ALSO
 .Xr atomic 9 ,
 .Xr critical 9 ,
diff --git a/share/man/man9/fail.9 b/share/man/man9/fail.9
index 8dbba1b..cf62365 100644
--- a/share/man/man9/fail.9
+++ b/share/man/man9/fail.9
@@ -26,25 +26,31 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd May 10, 2009
+.Dd March 15, 2016
 .Dt FAIL 9
 .Os
 .Sh NAME
 .Nm KFAIL_POINT_CODE ,
+.Nm KFAIL_POINT_CODE_FLAGS ,
+.Nm KFAIL_POINT_CODE_COND ,
 .Nm KFAIL_POINT_RETURN ,
 .Nm KFAIL_POINT_RETURN_VOID ,
 .Nm KFAIL_POINT_ERROR ,
 .Nm KFAIL_POINT_GOTO ,
+.Nm KFAIL_POINT_SLEEP_CALLBACKS ,
 .Nm fail_point ,
 .Nm DEBUG_FP
 .Nd fail points
 .Sh SYNOPSIS
 .In sys/fail.h
 .Fn KFAIL_POINT_CODE "parent" "name" "code"
+.Fn KFAIL_POINT_CODE_FLAGS "parent" "name" "flags" "code"
+.Fn KFAIL_POINT_CODE_COND "parent" "name" "cond" "flags" "code"
 .Fn KFAIL_POINT_RETURN "parent" "name"
 .Fn KFAIL_POINT_RETURN_VOID "parent" "name"
 .Fn KFAIL_POINT_ERROR "parent" "name" "error_var"
 .Fn KFAIL_POINT_GOTO "parent" "name" "error_var" "label"
+.Fn KFAIL_POINT_SLEEP_CALLBACKS "parent" "name" "pre_func" "pre_arg" "post_func" "post_arg" "code"
 .Sh DESCRIPTION
 Fail points are used to add code points where errors may be injected
 in a user controlled fashion.
@@ -77,6 +83,42 @@ argument, the evaluation of
 is derived from the
 .Fn return
 value set in the sysctl MIB.
+.Pp
+Additionally,
+.Fn KFAIL_POINT_CODE_FLAGS
+provides a
+.Fa flags
+argument which controls the fail point's behaviour.
+This can be used to e.g., mark the fail point's context as non-sleepable,
+which causes the
+.Sy sleep
+action to be coerced to a busy wait.
+The supported flags are:
+.Bl -ohang -offset indent
+.It FAIL_POINT_USE_TIMEOUT_PATH
+Rather than sleeping on a
+.Fn sleep
+call, just fire the post-sleep function after a timeout fires.
+.It FAIL_POINT_NONSLEEPABLE
+Mark the fail point as being in a non-sleepable context, which coerces
+.Fn sleep
+calls to
+.Fn delay
+calls.
+.El
+.Pp
+Likewise,
+.Fn KFAIL_POINT_CODE_COND
+supplies a
+.Fa cond
+argument, which allows you to set the condition under which the fail point's
+code may fire.
+This is equivalent to:
+.Bd -literal
+	if (cond)
+		KFAIL_POINT_CODE_FLAGS(...);
+
+.Ed
 See
 .Sx SYSCTL VARIABLES
 below.
@@ -107,26 +149,12 @@ Many base kernel MIBs can be found in the
 tree (referenced in code by
 .Sy DEBUG_FP ) .
 .Pp
-The sysctl variable may be set using the following grammar:
+The sysctl variable may be set in a number of ways:
 .Bd -literal
-  <fail_point> ::
-      <term> ( "->" <term> )*
-
-  <term> ::
-      ( (<float> "%") | (<integer> "*" ) )*
-      <type>
-      [ "(" <integer> ")" ]
-      [ "[pid " <integer> "]" ]
-
-  <float> ::
-      <integer> [ "." <integer> ] |
-      "." <integer>
-
-  <type> ::
-      "off" | "return" | "sleep" | "panic" | "break" | "print"
+  [<pct>%][<cnt>*]<type>[(args...)][-><more terms>]
 .Ed
 .Pp
-The <type> argument specifies which action to take:
+The <type> argument specifies which action to take; it can be one of:
 .Bl -tag -width ".Dv return"
 .It Sy off
 Take no action (does not trigger fail point code)
@@ -140,13 +168,23 @@ Panic
 Break into the debugger, or trap if there is no debugger support
 .It Sy print
 Print that the fail point executed
+.It Sy pause
+Threads sleep at the fail point until the fail point is set to
+.Sy off
+.It Sy yield
+Thread yields the cpu when the fail point is evaluated
+.It Sy delay
+Similar to sleep, but busy waits the cpu.
+(Useful in non-sleepable contexts.)
 .El
 .Pp
-The <float>% and <integer>* modifiers prior to <type> control when
+The <pct>% and <cnt>* modifiers prior to <type> control when
 <type> is executed.
-The <float>% form (e.g. "1.2%") can be used to specify a
+The <pct>% form (e.g. "1.2%") can be used to specify a
 probability that <type> will execute.
-The <integer>* form (e.g. "5*") can be used to specify the number of
+This is a decimal in the range (0, 100] which can specify up to
+1/10,000% precision.
+The <cnt>* form (e.g. "5*") can be used to specify the number of
 times <type> should be executed before this <term> is disabled.
 Only the last probability and the last count are used if multiple
 are specified, i.e. "1.2%2%" is the same as "2%".
@@ -191,6 +229,10 @@ Return 5 once, when pid 1234 executes the fail point.
 .Sh AUTHORS
 .An -nosplit
 This manual page was written by
+.Pp
+.An Matthew Bryan Aq Mt matthew.bryan@isilon.com
+and
+.Pp
 .An Zach Loafman Aq Mt zml@FreeBSD.org .
 .Sh CAVEATS
 It is easy to shoot yourself in the foot by setting fail points too
@@ -206,3 +248,10 @@ Currently,
 .Fn fail_point_eval
 does not verify whether the context is appropriate for calling
 .Fn msleep .
+You can force it to evaluate a
+.Sy sleep
+action as a
+.Sy delay
+action by specifying the
+.Sy FAIL_POINT_NONSLEEPABLE
+flag at the point the fail point is declared.
diff --git a/share/man/man9/kqueue.9 b/share/man/man9/kqueue.9
index a8137bd..ff379d1 100644
--- a/share/man/man9/kqueue.9
+++ b/share/man/man9/kqueue.9
@@ -372,7 +372,7 @@ The function
 will not be called, and the
 .Vt knote
 will not be returned on the next scan.
-Using this function could leak user land resources if a process uses the
+Using this function could leak userland resources if a process uses the
 .Vt knote
 to track resources.
 .Pp
diff --git a/share/man/man9/netisr.9 b/share/man/man9/netisr.9
index 6d79c57..63112be 100644
--- a/share/man/man9/netisr.9
+++ b/share/man/man9/netisr.9
@@ -120,7 +120,7 @@ whose fields are defined as follows:
 .Bl -tag -width "netisr_handler_t nh_handler"
 .It Vt "const char *" Va nh_name
 Unique character string name of the protocol, which may be included in
-.Xr sysctl 2
+.Xr sysctl 3
 MIB names, so should not contain whitespace.
 .It Vt netisr_handler_t Va nh_handler
 Protocol handler function that will be invoked on each packet received for
diff --git a/share/man/man9/osd.9 b/share/man/man9/osd.9
index e8260fe..c27b5d9 100644
--- a/share/man/man9/osd.9
+++ b/share/man/man9/osd.9
@@ -25,7 +25,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd January 5, 2011
+.Dd March 30, 2016
 .Dt OSD 9
 .Os
 .Sh NAME
@@ -33,6 +33,9 @@
 .Nm osd_register ,
 .Nm osd_deregister ,
 .Nm osd_set ,
+.Nm osd_reserve ,
+.Nm osd_set_reserved ,
+.Nm osd_free_reserved ,
 .Nm osd_get ,
 .Nm osd_del ,
 .Nm osd_call ,
@@ -63,6 +66,22 @@
 .Fa "void *value"
 .Fc
 .Ft void *
+.Fo osd_reserve
+.Fa "u_int slot"
+.Fc
+.Ft int
+.Fo osd_set_reserved
+.Fa "u_int type"
+.Fa "struct osd *osd"
+.Fa "u_int slot"
+.Fa "void *rsv"
+.Fa "void *value"
+.Fc
+.Ft void
+.Fo osd_free_reserved
+.Fa "void *rsv"
+.Fc
+.Ft void *
 .Fo osd_get
 .Fa "u_int type"
 .Fa "struct osd *osd"
@@ -198,6 +217,15 @@ argument points to a data object to associate with
 .Fa osd .
 .Pp
 The
+.Fn osd_set_reserved
+function does the same as
+.Fn osd_set ,
+but with an extra argument
+.Fa rsv
+that is internal-use memory previously allocated via
+.Fn osd_reserve .
+.Pp
+The
 .Fn osd_get
 function returns the data pointer associated with a kernel data structure's
 .Vt struct osd
@@ -324,6 +352,24 @@ will proceed without any
 .Xr realloc 9
 calls.
 .Pp
+It is possible for
+.Fn osd_set
+to fail to allocate this array.  To ensure that such allocation succeeds,
+.Fn osd_reserve
+may be called (in a non-blocking context), and it will pre-allocate the
+memory via
+.Xr malloc 9
+with M_WAITOK.
+Then this pre-allocated memory is passed to
+.Fn osd_set_reserved ,
+which will use it if necessary or otherwise discard it.
+The memory may also be explicitly discarded by calling
+.Fn osd_free_reserved .
+As this method always allocates memory whether or not it is ultimately needed,
+it should be used only rarely, such as in the unlikely event that
+.Fn osd_set
+fails.
+.Pp
 The
 .Nm
 API is geared towards slot identifiers storing pointers to the same underlying
@@ -359,15 +405,27 @@ the kernel including most fast paths.
 returns the slot identifier for the newly registered data type.
 .Pp
 .Fn osd_set
-returns zero on success or ENOMEM if the specified type/slot identifier pair
+and
+.Fn osd_set_reserved
+return zero on success or ENOMEM if the specified type/slot identifier pair
 triggered an internal
 .Xr realloc 9
-which failed.
+which failed
+.Fn ( osd_set_reserved
+will always succeed when
+.Fa rsv
+is non-NULL).
 .Pp
 .Fn osd_get
 returns the data pointer for the specified type/slot identifier pair, or NULL if
 the slot has not been initialised yet.
 .Pp
+.Fn osd_reserve
+returns a pointer suitable for passing to
+.Fn osd_set_reserved
+or
+.Fn osd_free_reserved .
+.Pp
 .Fn osd_call
 returns zero if no method is run or the method for each slot runs successfully.
 If a method for a slot returns non-zero,
diff --git a/share/man/man9/rwlock.9 b/share/man/man9/rwlock.9
index 7598826..9ee16fc 100644
--- a/share/man/man9/rwlock.9
+++ b/share/man/man9/rwlock.9
@@ -24,7 +24,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd December 13, 2014
+.Dd March 28, 2016
 .Dt RWLOCK 9
 .Os
 .Sh NAME
@@ -313,18 +313,19 @@ facility was written by
 This manual page was written by
 .An "Gleb Smirnoff" .
 .Sh BUGS
-If
+A kernel without
 .Dv WITNESS
-is not included in the kernel,
-then it is impossible to assert that the current thread does or does not
-hold a read lock.
-In the
-.Pf non- Dv WITNESS
-case, the
+cannot assert whether the current thread does or does not hold a read lock.
 .Dv RA_LOCKED
 and
 .Dv RA_RLOCKED
-assertions merely check that some thread holds a read lock.
+can only assert that
+.Em any
+thread holds a read lock.
+They cannot ensure that the current thread holds a read lock.
+Further,
+.Dv RA_UNLOCKED
+can only assert that the current thread does not hold a write lock.
 .Pp
 Reader/writer is a bit of an awkward name.
 An
diff --git a/share/man/man9/sx.9 b/share/man/man9/sx.9
index a9931a9..f09584a 100644
--- a/share/man/man9/sx.9
+++ b/share/man/man9/sx.9
@@ -26,7 +26,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd March 13, 2016
+.Dd March 28, 2016
 .Dt SX 9
 .Os
 .Sh NAME
@@ -320,16 +320,16 @@ end up sleeping while holding a mutex, which is not allowed.
 .Xr rwlock 9 ,
 .Xr sema 9
 .Sh BUGS
-Currently there is no way to assert that a lock is not held.
-This is not possible in the
-.No non- Ns Dv WITNESS
-case for asserting that this thread
-does not hold a shared lock.
-In the
-.No non- Ns Dv WITNESS
-case, the
+A kernel without
+.Dv WITNESS
+cannot assert whether the current thread does or does not hold a shared lock.
 .Dv SA_LOCKED
 and
 .Dv SA_SLOCKED
-assertions merely check that some thread holds a shared lock.
-They do not ensure that the current thread holds a shared lock.
+can only assert that
+.Em any
+thread holds a shared lock.
+They cannot ensure that the current thread holds a shared lock.
+Further,
+.Dv SA_UNLOCKED
+can only assert that the current thread does not hold an exclusive lock.