path: root/share/man/man4/mem.4

.\" Copyright (c) 1991 The Regents of the University of California.
.\" 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.
.\" 3. Neither the name of the University nor the names of its contributors
.\"    may be used to endorse or promote products derived from this software
.\"    without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
.\"
.Dd March 24, 2024
.Dt MEM 4
.Os
.Sh NAME
.Nm mem ,
.Nm kmem
.Nd memory files
.Sh SYNOPSIS
.Cd "device mem"
.Sh DESCRIPTION
The special file
.Pa /dev/mem
is an interface to the physical memory of the computer.
Byte offsets in this file are interpreted as physical memory addresses.
Reading and writing this file is equivalent to reading and writing
memory itself.
Only offsets within the bounds of
.Pa /dev/mem
are allowed.
.Pp
Kernel virtual memory is accessed through the interface
.Pa /dev/kmem
in the same manner as
.Pa /dev/mem .
Only kernel virtual addresses that are currently mapped to memory are allowed.
.Pp
On ISA the I/O memory space begins at physical address 0x000a0000
and runs to 0x00100000.
The
per-process data
size
for the current process
is
.Dv UPAGES
long, and ends at virtual
address 0xf0000000.
.Sh IOCTL INTERFACE
.Ss Address Properties
The
.Dv MEM_EXTRACT_PADDR
ioctl can be used to look up the physical address and NUMA domain of a given
virtual address in the calling process' address space.
The request is described by
.Bd -literal
struct mem_extract {
	uint64_t	me_vaddr;	/* input */
	uint64_t	me_paddr;	/* output */
	int		me_domain;	/* output */
	int		me_state;	/* output */
};
.Ed
.Pp
The ioctl returns an error if the address is not valid.
The information returned by
.Dv MEM_EXTRACT_PADDR
may be out of date by the time that the ioctl call returns.
Specifically, concurrent system calls, page faults, or system page reclamation
activity may have unmapped the virtual page or replaced the backing physical
page before the ioctl call returns.
Wired pages, e.g., those locked by
.Xr mlock 2 ,
will not be reclaimed by the system.
.Pp
The
.Fa me_state
field provides information about the state of the virtual page:
.Bl -tag -width indent
.It Dv ME_STATE_INVALID
The virtual address is invalid.
.It Dv ME_STATE_VALID
The virtual address is valid but is not mapped at the time of the ioctl call.
.It Dv ME_STATE_MAPPED
The virtual address corresponds to a physical page mapping, and the
.Fa me_paddr
and
.Fa me_domain
fields are valid.
.El
.Ss Memory Ranges
.Pp
Several architectures allow attributes to be associated with ranges of physical
memory.
These attributes can be manipulated via
.Fn ioctl
calls performed on
.Pa /dev/mem .
Declarations and data types are to be found in
.In sys/memrange.h .
.Pp
The specific attributes, and number of programmable ranges may vary between
architectures.
The full set of supported attributes is:
.Bl -tag -width indent
.It Dv MDF_UNCACHEABLE
The region is not cached.
.It Dv MDF_WRITECOMBINE
Writes to the region may be combined or performed out of order.
.It Dv MDF_WRITETHROUGH
Writes to the region are committed synchronously.
.It Dv MDF_WRITEBACK
Writes to the region are committed asynchronously.
.It Dv MDF_WRITEPROTECT
The region cannot be written to.
.El
.Pp
Memory ranges are described by
.Bd -literal
struct mem_range_desc {
	uint64_t	mr_base;	/* physical base address */
	uint64_t	mr_len;		/* physical length of region */
	int		mr_flags;	/* attributes of region */
	char		mr_owner[8];
};
.Ed
.Pp
In addition to the region attributes listed above, the following flags
may also be set in the
.Fa mr_flags
field:
.Bl -tag -width indent
.It MDF_FIXBASE
The region's base address cannot be changed.
.It MDF_FIXLEN
The region's length cannot be changed.
.It MDF_FIRMWARE
The region is believed to have been established by the system firmware.
.It MDF_ACTIVE
The region is currently active.
.It MDF_BOGUS
We believe the region to be invalid or otherwise erroneous.
.It MDF_FIXACTIVE
The region cannot be disabled.
.It MDF_BUSY
The region is currently owned by another process and may not be
altered.
.El
.Pp
Operations are performed using
.Bd -literal
struct mem_range_op {
	struct mem_range_desc	*mo_desc;
	int			mo_arg[2];
};
.Ed
.Pp
The
.Dv MEMRANGE_GET
ioctl is used to retrieve current memory range attributes.
If
.Va mo_arg[0]
is set to 0, it will be updated with the total number of memory range
descriptors.
If greater than 0, the array at
.Va mo_desc
will be filled with a corresponding number of descriptor structures,
or the maximum, whichever is less.
.Pp
The
.Dv MEMRANGE_SET
ioctl is used to add, alter and remove memory range attributes.
A range
with the
.Dv MDF_FIXACTIVE
flag may not be removed; a range with the
.Dv MDF_BUSY
flag may not be removed or updated.
.Pp
.Va mo_arg[0]
should be set to
.Dv MEMRANGE_SET_UPDATE
to update an existing or establish a new range, or to
.Dv MEMRANGE_SET_REMOVE
to remove a range.
.El
.Ss Live Kernel Dumps
.Pp
The
.Dv MEM_KERNELDUMP
ioctl will initiate a kernel dump against the running system, the contents of
which will be written to a process-owned file descriptor.
The resulting dump output will be in minidump format.
The request is described by
.Bd -literal
struct mem_livedump_arg {
	int	fd;		/* input */
	int	flags		/* input */
	uint8_t	compression	/* input */
};
.Ed
.Pp
The
.Va fd
field is used to pass the file descriptor.
.Pp
The
.Va flags
field is currently unused and must be set to zero.
.Pp
The
.Va compression
field can be used to specify the desired compression to
be applied to the dump output.
The supported values are defined in
.In sys/kerneldump.h ;
that is,
.Dv KERNELDUMP_COMP_NONE ,
.Dv KERNELDUMP_COMP_GZIP ,
or
.Dv KERNELDUMP_COMP_ZSTD .
.Pp
Kernel dumps taken against the running system may have inconsistent kernel data
structures due to allocation, deallocation, or modification of memory
concurrent to the dump procedure.
Thus, the resulting core dump is not guaranteed to be usable.
A system under load is more likely to produce an inconsistent result.
Despite this, live kernel dumps can be useful for offline debugging of certain
types of kernel bugs, such as deadlocks, or in inspecting a particular part of
the system's state.
.Sh RETURN VALUES
.Ss MEM_EXTRACT_PADDR
The
.Dv MEM_EXTRACT_PADDR
ioctl always returns a value of zero.
.Ss MEMRANGE_GET/MEMRANGE_SET
.Bl -tag -width Er
.It Bq Er EOPNOTSUPP
Memory range operations are not supported on this architecture.
.It Bq Er ENXIO
No memory range descriptors are available (e.g., firmware has not enabled
any).
.It Bq Er EINVAL
The memory range supplied as an argument is invalid or overlaps another
range in a fashion not supported by this architecture.
.It Bq Er EBUSY
An attempt to remove or update a range failed because the range is busy.
.It Bq Er ENOSPC
An attempt to create a new range failed due to a shortage of hardware
resources (e.g., descriptor slots).
.It Bq Er ENOENT
An attempt to remove a range failed because no range matches the descriptor
base/length supplied.
.It Bq Er EPERM
An attempt to remove a range failed because the range is permanently
enabled.
.El
.Ss MEM_KERNELDUMP
.Bl -tag -width Er
.It Bq Er EOPNOTSUPP
Kernel minidumps are not supported on this architecture.
.It Bq Er EPERM
An attempt to begin the kernel dump failed because the calling thread lacks the
.It Bq Er EBADF
The supplied file descriptor was invalid, or does not have write permission.
.It Bq Er EBUSY
An attempt to begin the kernel dump failed because one is already in progress.
.It Bq Er EINVAL
An invalid or unsupported value was specified in
.Va flags .
.It Bq Er EINVAL
An invalid or unsupported compression type was specified.
.Dv PRIV_KMEM_READ
privilege.
.El
.Sh FILES
.Bl -tag -width /dev/kmem -compact
.It Pa /dev/mem
.It Pa /dev/kmem
.El
.Sh SEE ALSO
.Xr kvm 3 ,
.Xr memcontrol 8
.Sh HISTORY
The
.Pa /dev/mem
file appeared in
.At v1
and
.Pa /dev/kmem
in
.At v5 .
The ioctl interface for memory range attributes was added in
.Fx 3.2 .
.Sh BUGS
Busy range attributes are not yet managed correctly.
.Pp
This device is required for all users of
.Xr kvm 3
to operate.