a61bd9573f
multiple processors. In particular, clearly state that the operations are always atomic when they are applied to the default memory type that is used by the kernel (and applications). Reviewed by: kib, jhb (an earlier version) MFC after: 1 week
409 lines
10 KiB
Groff
409 lines
10 KiB
Groff
.\" Copyright (c) 2000-2001 John H. Baldwin <jhb@FreeBSD.org>
|
|
.\" All rights reserved.
|
|
.\"
|
|
.\" Redistribution and use in source and binary forms, with or without
|
|
.\" modification, are permitted provided that the following conditions
|
|
.\" are met:
|
|
.\" 1. Redistributions of source code must retain the above copyright
|
|
.\" notice, this list of conditions and the following disclaimer.
|
|
.\" 2. Redistributions in binary form must reproduce the above copyright
|
|
.\" notice, this list of conditions and the following disclaimer in the
|
|
.\" documentation and/or other materials provided with the distribution.
|
|
.\"
|
|
.\" THIS SOFTWARE IS PROVIDED BY THE DEVELOPERS ``AS IS'' AND ANY EXPRESS OR
|
|
.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
|
|
.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
|
|
.\" IN NO EVENT SHALL THE DEVELOPERS BE LIABLE FOR ANY DIRECT, INDIRECT,
|
|
.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
|
|
.\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
.\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
.\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
.\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
|
|
.\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
.\"
|
|
.\" $FreeBSD$
|
|
.\"
|
|
.Dd August 9, 2015
|
|
.Dt ATOMIC 9
|
|
.Os
|
|
.Sh NAME
|
|
.Nm atomic_add ,
|
|
.Nm atomic_clear ,
|
|
.Nm atomic_cmpset ,
|
|
.Nm atomic_fetchadd ,
|
|
.Nm atomic_load ,
|
|
.Nm atomic_readandclear ,
|
|
.Nm atomic_set ,
|
|
.Nm atomic_subtract ,
|
|
.Nm atomic_store
|
|
.Nd atomic operations
|
|
.Sh SYNOPSIS
|
|
.In sys/types.h
|
|
.In machine/atomic.h
|
|
.Ft void
|
|
.Fn atomic_add_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
|
|
.Ft void
|
|
.Fn atomic_clear_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
|
|
.Ft int
|
|
.Fo atomic_cmpset_[acq_|rel_]<type>
|
|
.Fa "volatile <type> *dst"
|
|
.Fa "<type> old"
|
|
.Fa "<type> new"
|
|
.Fc
|
|
.Ft <type>
|
|
.Fn atomic_fetchadd_<type> "volatile <type> *p" "<type> v"
|
|
.Ft <type>
|
|
.Fn atomic_load_acq_<type> "volatile <type> *p"
|
|
.Ft <type>
|
|
.Fn atomic_readandclear_<type> "volatile <type> *p"
|
|
.Ft void
|
|
.Fn atomic_set_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
|
|
.Ft void
|
|
.Fn atomic_subtract_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
|
|
.Ft void
|
|
.Fn atomic_store_rel_<type> "volatile <type> *p" "<type> v"
|
|
.Ft <type>
|
|
.Fn atomic_swap_<type> "volatile <type> *p" "<type> v"
|
|
.Ft int
|
|
.Fn atomic_testandset_<type> "volatile <type> *p" "u_int v"
|
|
.Sh DESCRIPTION
|
|
Each of the atomic operations is guaranteed to be atomic across multiple
|
|
processors and in the presence of interrupts.
|
|
They can be used to implement reference counts or as building blocks for more
|
|
advanced synchronization primitives such as mutexes.
|
|
.Ss Types
|
|
Each atomic operation operates on a specific
|
|
.Fa type .
|
|
The type to use is indicated in the function name.
|
|
The available types that can be used are:
|
|
.Pp
|
|
.Bl -tag -offset indent -width short -compact
|
|
.It Li int
|
|
unsigned integer
|
|
.It Li long
|
|
unsigned long integer
|
|
.It Li ptr
|
|
unsigned integer the size of a pointer
|
|
.It Li 32
|
|
unsigned 32-bit integer
|
|
.It Li 64
|
|
unsigned 64-bit integer
|
|
.El
|
|
.Pp
|
|
For example, the function to atomically add two integers is called
|
|
.Fn atomic_add_int .
|
|
.Pp
|
|
Certain architectures also provide operations for types smaller than
|
|
.Dq Li int .
|
|
.Pp
|
|
.Bl -tag -offset indent -width short -compact
|
|
.It Li char
|
|
unsigned character
|
|
.It Li short
|
|
unsigned short integer
|
|
.It Li 8
|
|
unsigned 8-bit integer
|
|
.It Li 16
|
|
unsigned 16-bit integer
|
|
.El
|
|
.Pp
|
|
These must not be used in MI code because the instructions to implement them
|
|
efficiently may not be available.
|
|
.Ss Memory Barriers
|
|
Memory barriers are used to guarantee the order of data accesses in
|
|
two ways.
|
|
First, they specify hints to the compiler to not re-order or optimize the
|
|
operations.
|
|
Second, on architectures that do not guarantee ordered data accesses,
|
|
special instructions or special variants of instructions are used to indicate
|
|
to the processor that data accesses need to occur in a certain order.
|
|
As a result, most of the atomic operations have three variants in order to
|
|
include optional memory barriers.
|
|
The first form just performs the operation without any explicit barriers.
|
|
The second form uses a read memory barrier, and the third variant uses a write
|
|
memory barrier.
|
|
.Pp
|
|
The second variant of each operation includes an
|
|
.Em acquire
|
|
memory barrier.
|
|
This barrier ensures that the effects of this operation are completed before the
|
|
effects of any later data accesses.
|
|
As a result, the operation is said to have acquire semantics as it acquires a
|
|
pseudo-lock requiring further operations to wait until it has completed.
|
|
To denote this, the suffix
|
|
.Dq Li _acq
|
|
is inserted into the function name immediately prior to the
|
|
.Dq Li _ Ns Aq Fa type
|
|
suffix.
|
|
For example, to subtract two integers ensuring that any later writes will
|
|
happen after the subtraction is performed, use
|
|
.Fn atomic_subtract_acq_int .
|
|
.Pp
|
|
The third variant of each operation includes a
|
|
.Em release
|
|
memory barrier.
|
|
This ensures that all effects of all previous data accesses are completed
|
|
before this operation takes place.
|
|
As a result, the operation is said to have release semantics as it releases
|
|
any pending data accesses to be completed before its operation is performed.
|
|
To denote this, the suffix
|
|
.Dq Li _rel
|
|
is inserted into the function name immediately prior to the
|
|
.Dq Li _ Ns Aq Fa type
|
|
suffix.
|
|
For example, to add two long integers ensuring that all previous
|
|
writes will happen first, use
|
|
.Fn atomic_add_rel_long .
|
|
.Pp
|
|
A practical example of using memory barriers is to ensure that data accesses
|
|
that are protected by a lock are all performed while the lock is held.
|
|
To achieve this, one would use a read barrier when acquiring the lock to
|
|
guarantee that the lock is held before any protected operations are performed.
|
|
Finally, one would use a write barrier when releasing the lock to ensure that
|
|
all of the protected operations are completed before the lock is released.
|
|
.Ss Multiple Processors
|
|
In multiprocessor systems, the atomicity of the atomic operations on memory
|
|
depends on support for cache coherence in the underlying architecture.
|
|
In general, cache coherence on the default memory type,
|
|
.Dv VM_MEMATTR_DEFAULT ,
|
|
is guaranteed by all architectures that are supported by
|
|
.Fx .
|
|
For example, cache coherence is guaranteed on write-back memory by the
|
|
.Tn amd64
|
|
and
|
|
.Tn i386
|
|
architectures.
|
|
However, on some architectures, cache coherence may not be enabled on all
|
|
memory types.
|
|
To determine if cache coherence is enabled for a non-default memory type,
|
|
consult the architecture's documentation.
|
|
.Ss Semantics
|
|
This section describes the semantics of each operation using a C like notation.
|
|
.Bl -hang
|
|
.It Fn atomic_add p v
|
|
.Bd -literal -compact
|
|
*p += v;
|
|
.Ed
|
|
.It Fn atomic_clear p v
|
|
.Bd -literal -compact
|
|
*p &= ~v;
|
|
.Ed
|
|
.It Fn atomic_cmpset dst old new
|
|
.Bd -literal -compact
|
|
if (*dst == old) {
|
|
*dst = new;
|
|
return (1);
|
|
} else
|
|
return (0);
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_cmpset
|
|
functions are not implemented for the types
|
|
.Dq Li char ,
|
|
.Dq Li short ,
|
|
.Dq Li 8 ,
|
|
and
|
|
.Dq Li 16 .
|
|
.Bl -hang
|
|
.It Fn atomic_fetchadd p v
|
|
.Bd -literal -compact
|
|
tmp = *p;
|
|
*p += v;
|
|
return (tmp);
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_fetchadd
|
|
functions are only implemented for the types
|
|
.Dq Li int ,
|
|
.Dq Li long
|
|
and
|
|
.Dq Li 32
|
|
and do not have any variants with memory barriers at this time.
|
|
.Bl -hang
|
|
.It Fn atomic_load p
|
|
.Bd -literal -compact
|
|
return (*p);
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_load
|
|
functions are only provided with acquire memory barriers.
|
|
.Bl -hang
|
|
.It Fn atomic_readandclear p
|
|
.Bd -literal -compact
|
|
tmp = *p;
|
|
*p = 0;
|
|
return (tmp);
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_readandclear
|
|
functions are not implemented for the types
|
|
.Dq Li char ,
|
|
.Dq Li short ,
|
|
.Dq Li ptr ,
|
|
.Dq Li 8 ,
|
|
and
|
|
.Dq Li 16
|
|
and do not have any variants with memory barriers at this time.
|
|
.Bl -hang
|
|
.It Fn atomic_set p v
|
|
.Bd -literal -compact
|
|
*p |= v;
|
|
.Ed
|
|
.It Fn atomic_subtract p v
|
|
.Bd -literal -compact
|
|
*p -= v;
|
|
.Ed
|
|
.It Fn atomic_store p v
|
|
.Bd -literal -compact
|
|
*p = v;
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_store
|
|
functions are only provided with release memory barriers.
|
|
.Bl -hang
|
|
.It Fn atomic_swap p v
|
|
.Bd -literal -compact
|
|
tmp = *p;
|
|
*p = v;
|
|
return (tmp);
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_swap
|
|
functions are not implemented for the types
|
|
.Dq Li char ,
|
|
.Dq Li short ,
|
|
.Dq Li ptr ,
|
|
.Dq Li 8 ,
|
|
and
|
|
.Dq Li 16
|
|
and do not have any variants with memory barriers at this time.
|
|
.Bl -hang
|
|
.It Fn atomic_testandset p v
|
|
.Bd -literal -compact
|
|
bit = 1 << (v % (sizeof(*p) * NBBY));
|
|
tmp = (*p & bit) != 0;
|
|
*p |= bit;
|
|
return (tmp);
|
|
.Ed
|
|
.El
|
|
.Pp
|
|
The
|
|
.Fn atomic_testandset
|
|
functions are only implemented for the types
|
|
.Dq Li int ,
|
|
.Dq Li long
|
|
and
|
|
.Dq Li 32
|
|
and do not have any variants with memory barriers at this time.
|
|
.Pp
|
|
The type
|
|
.Dq Li 64
|
|
is currently not implemented for any of the atomic operations on the
|
|
.Tn arm ,
|
|
.Tn i386 ,
|
|
and
|
|
.Tn powerpc
|
|
architectures.
|
|
.Sh RETURN VALUES
|
|
The
|
|
.Fn atomic_cmpset
|
|
function returns the result of the compare operation.
|
|
The
|
|
.Fn atomic_fetchadd ,
|
|
.Fn atomic_load ,
|
|
.Fn atomic_readandclear ,
|
|
and
|
|
.Fn atomic_swap
|
|
functions return the value at the specified address.
|
|
The
|
|
.Fn atomic_testandset
|
|
function returns the result of the test operation.
|
|
.Sh EXAMPLES
|
|
This example uses the
|
|
.Fn atomic_cmpset_acq_ptr
|
|
and
|
|
.Fn atomic_set_ptr
|
|
functions to obtain a sleep mutex and handle recursion.
|
|
Since the
|
|
.Va mtx_lock
|
|
member of a
|
|
.Vt "struct mtx"
|
|
is a pointer, the
|
|
.Dq Li ptr
|
|
type is used.
|
|
.Bd -literal
|
|
/* Try to obtain mtx_lock once. */
|
|
#define _obtain_lock(mp, tid) \\
|
|
atomic_cmpset_acq_ptr(&(mp)->mtx_lock, MTX_UNOWNED, (tid))
|
|
|
|
/* Get a sleep lock, deal with recursion inline. */
|
|
#define _get_sleep_lock(mp, tid, opts, file, line) do { \\
|
|
uintptr_t _tid = (uintptr_t)(tid); \\
|
|
\\
|
|
if (!_obtain_lock(mp, tid)) { \\
|
|
if (((mp)->mtx_lock & MTX_FLAGMASK) != _tid) \\
|
|
_mtx_lock_sleep((mp), _tid, (opts), (file), (line));\\
|
|
else { \\
|
|
atomic_set_ptr(&(mp)->mtx_lock, MTX_RECURSE); \\
|
|
(mp)->mtx_recurse++; \\
|
|
} \\
|
|
} \\
|
|
} while (0)
|
|
.Ed
|
|
.Sh HISTORY
|
|
The
|
|
.Fn atomic_add ,
|
|
.Fn atomic_clear ,
|
|
.Fn atomic_set ,
|
|
and
|
|
.Fn atomic_subtract
|
|
operations were first introduced in
|
|
.Fx 3.0 .
|
|
This first set only supported the types
|
|
.Dq Li char ,
|
|
.Dq Li short ,
|
|
.Dq Li int ,
|
|
and
|
|
.Dq Li long .
|
|
The
|
|
.Fn atomic_cmpset ,
|
|
.Fn atomic_load ,
|
|
.Fn atomic_readandclear ,
|
|
and
|
|
.Fn atomic_store
|
|
operations were added in
|
|
.Fx 5.0 .
|
|
The types
|
|
.Dq Li 8 ,
|
|
.Dq Li 16 ,
|
|
.Dq Li 32 ,
|
|
.Dq Li 64 ,
|
|
and
|
|
.Dq Li ptr
|
|
and all of the acquire and release variants
|
|
were added in
|
|
.Fx 5.0
|
|
as well.
|
|
The
|
|
.Fn atomic_fetchadd
|
|
operations were added in
|
|
.Fx 6.0 .
|
|
The
|
|
.Fn atomic_swap
|
|
and
|
|
.Fn atomic_testandset
|
|
operations were added in
|
|
.Fx 10.0 .
|