opnsense-src/lib/libc/sys/cpuset_getaffinity.2
Dmitry Chagin 72bc1e6806 cpuset: Byte swap cpuset for compat32 on big endian architectures
Summary:
BITSET uses long as its basic underlying type, which is dependent on the
compile type, meaning on 32-bit builds the basic type is 32 bits, but on
64-bit builds it's 64 bits.  On little endian architectures this doesn't
matter, because the LSB is always at the low bit, so the words get
effectively concatenated moving between 32-bit and 64-bit, but on
big-endian architectures it throws a wrench in, as setting bit 0 in
32-bit mode is equivalent to setting bit 32 in 64-bit mode.  To
demonstrate:

32-bit mode:

BIT_SET(foo, 0):        0x00000001

64-bit sees: 0x0000000100000000

cpuset is the only system interface that uses bitsets, so solve this
by swapping the integer sub-components at the copyin/copyout points.

Reviewed by:    kib
Sponsored by:   Juniper Networks, Inc.
Differential Revision:  https://reviews.freebsd.org/D35225

(cherry picked from commit 47a57144af)

Fix the build after 47a57144

(cherry picked from commit 89737eb829)

cpuset: Fix the KASAN and KMSAN builds

Rename the "copyin" and "copyout" fields of struct cpuset_copy_cb to
something less generic, since sanitizers define interceptors for
copyin() and copyout() using #define.

Reported by:    syzbot+2db5d644097fc698fb6f@syzkaller.appspotmail.com
Fixes:  47a57144af ("cpuset: Byte swap cpuset for compat32 on big endian architectures")
Sponsored by:   The FreeBSD Foundation

(cherry picked from commit 4a3e51335e)

Use Linux semantics for the thread affinity syscalls.

Linux has more tolerant checks of the user supplied cpuset_t's.

Minimum cpuset_t size that the Linux kernel permits in case of
getaffinity() is the maximum CPU id, present in the system / NBBY,
the maximum size is not limited.
For setaffinity(), Linux does not limit the size of the user-provided
cpuset_t, internally using only the meaningful part of the set, where
the upper bound is the maximum CPU id, present in the system, no larger
than the size of the kernel cpuset_t.
Unlike FreeBSD, Linux ignores high bits if set in the setaffinity(),
so clear it in the sched_setaffinity() and Linuxulator itself.

Reviewed by:            Pau Amma (man pages)
In collaboration with:  jhb
Differential revision:  https://reviews.freebsd.org/D34849
MFC after:              2 weeks

(cherry picked from commit f35093f8d6)
2022-06-17 22:35:14 +03:00

179 lines
4.9 KiB
Groff

.\" Copyright (c) 2008 Christian Brueffer
.\" Copyright (c) 2008 Jeffrey Roberson
.\" 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 AUTHOR AND CONTRIBUTORS ``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 AUTHOR OR CONTRIBUTORS 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 April 27, 2022
.Dt CPUSET_GETAFFINITY 2
.Os
.Sh NAME
.Nm cpuset_getaffinity ,
.Nm cpuset_setaffinity
.Nd manage CPU affinity
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In sys/param.h
.In sys/cpuset.h
.Ft int
.Fn cpuset_getaffinity "cpulevel_t level" "cpuwhich_t which" "id_t id" "size_t setsize" "cpuset_t *mask"
.Ft int
.Fn cpuset_setaffinity "cpulevel_t level" "cpuwhich_t which" "id_t id" "size_t setsize" "const cpuset_t *mask"
.Sh DESCRIPTION
.Fn cpuset_getaffinity
and
.Fn cpuset_setaffinity
allow the manipulation of sets of CPUs available to processes, threads,
interrupts, jails and other resources.
These functions may manipulate sets of CPUs that contain many processes
or per-object anonymous masks that effect only a single object.
.Pp
The valid values for the
.Fa level
and
.Fa which
arguments are documented in
.Xr cpuset 2 .
These arguments specify which object and which set of the object we are
referring to.
Not all possible combinations are valid.
For example, only processes may belong to a numbered set accessed by a
.Fa level
argument of
.Dv CPU_LEVEL_CPUSET .
All resources, however, have a mask which may be manipulated with
.Dv CPU_LEVEL_WHICH .
.Pp
Masks of type
.Ft cpuset_t
are composed using the
.Dv CPU_SET
macros.
If the user-supplied mask is not large enough to fit all of the matching CPUs,
.Fn cpuset_getaffinity
fails with
.Er ERANGE .
Calls to
.Fn cpuset_setaffinity
tolerate masks of any size with no restrictions.
The kernel uses the meaningful part of the mask, where the upper bound is
the maximum CPU id present in the system.
If bits for non-existing CPUs are set, calls to
.Fn cpuset_setaffinity
fails with
.Er EINVAL .
.Pp
The supplied mask should have a size of
.Fa setsize
bytes.
This size is usually provided by calling
.Li sizeof(mask)
which is ultimately determined by the value of
.Dv CPU_SETSIZE
as defined in
.In sys/cpuset.h .
.Pp
.Fn cpuset_getaffinity
retrieves the
mask from the object specified by
.Fa level ,
.Fa which
and
.Fa id
and stores it in the space provided by
.Fa mask .
.Pp
.Fn cpuset_setaffinity
attempts to set the mask for the object specified by
.Fa level ,
.Fa which
and
.Fa id
to the value in
.Fa mask .
.Sh RETURN VALUES
.Rv -std
.Sh ERRORS
The following error codes may be set in
.Va errno :
.Bl -tag -width Er
.\" When changing this list, consider updating share/man/man3/pthread_create.3,
.\" since that function can return any of these errors.
.It Bq Er EINVAL
The
.Fa level
or
.Fa which
argument was not a valid value.
.It Bq Er EINVAL
The
.Fa mask
argument specified when calling
.Fn cpuset_setaffinity
was not a valid value.
.It Bq Er EDEADLK
The
.Fn cpuset_setaffinity
call would leave a thread without a valid CPU to run on because the set
does not overlap with the thread's anonymous mask.
.It Bq Er EFAULT
The mask pointer passed was invalid.
.It Bq Er ESRCH
The object specified by the
.Fa id
and
.Fa which
arguments could not be found.
.It Bq Er ERANGE
The
.Fa cpusetsize
was smaller than needed to fit all of the matching CPUs.
.It Bq Er EPERM
The calling process did not have the credentials required to complete the
operation.
.It Bq Er ECAPMODE
The calling process attempted to act on a process other than itself, while
in capability mode.
See
.Xr capsicum 4 .
.El
.Sh SEE ALSO
.Xr cpuset 1 ,
.Xr cpuset 2 ,
.Xr cpuset_getdomain 2 ,
.Xr cpuset_getid 2 ,
.Xr cpuset_setdomain 2 ,
.Xr cpuset_setid 2 ,
.Xr pthread_affinity_np 3 ,
.Xr pthread_attr_affinity_np 3 ,
.Xr capsicum 4 ,
.Xr cpuset 9
.Sh HISTORY
The
.Nm
family of system calls first appeared in
.Fx 7.1 .
.Sh AUTHORS
.An Jeffrey Roberson Aq Mt jeff@FreeBSD.org