.\"	$NetBSD: kmem.9,v 1.23 2017/11/07 18:36:27 christos Exp $
.\"
.\" Copyright (c)2006 YAMAMOTO Takashi,
.\" 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.
.\"
.\" ------------------------------------------------------------
.Dd November 7, 2017
.Dt KMEM 9
.Os
.\" ------------------------------------------------------------
.Sh NAME
.Nm kmem
.Nd kernel wired memory allocator
.\" ------------------------------------------------------------
.Sh SYNOPSIS
.In sys/kmem.h
.\" - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
.Ft void *
.Fn kmem_alloc \
"size_t size" "km_flag_t kmflags"
.Ft void *
.Fn kmem_zalloc \
"size_t size" "km_flag_t kmflags"
.Ft void
.Fn kmem_free \
"void *p" "size_t size"
.\" ---
.Ft void *
.Fn kmem_intr_alloc \
"size_t size" "km_flag_t kmflags"
.Ft void *
.Fn kmem_intr_zalloc \
"size_t size" "km_flag_t kmflags"
.Ft void
.Fn kmem_intr_free \
"void *p" "size_t size"
.\" ---
.Ft char *
.Fn kmem_asprintf \
"const char *fmt" "..."
.\" ---
.Ft char *
.Fn kmem_strdupsize \
"const char *str" "size_t *size" "km_flag_t kmflags"
.Ft void
.Fn kmem_strfree \
"char *str"
.\" ------------------------------------------------------------
.Pp
.Cd "options KMEM_SIZE"
.Cd "options KMEM_REDZONE"
.Cd "options KMEM_GUARD"
.Sh DESCRIPTION
.Fn kmem_alloc
allocates kernel wired memory.
It takes the following arguments.
.Bl -tag -width kmflags
.It Fa size
Specify the size of allocation in bytes.
.It Fa kmflags
Either of the following:
.Bl -tag -width KM_NOSLEEP
.It Dv KM_SLEEP
If the allocation cannot be satisfied immediately, sleep until enough
memory is available.
If
.Dv KM_SLEEP
is specified, then the allocation cannot fail.
.It Dv KM_NOSLEEP
Don't sleep.
Immediately return
.Dv NULL
if there is not enough memory available.
It should only be used when failure to allocate will not have harmful,
user-visible effects.
.Pp
.Bf -symbolic
Use of
.Dv KM_NOSLEEP
is strongly discouraged as it can create transient, hard to debug failures
that occur when the system is under memory pressure.
.Ef
.Pp
In situations where it is not possible to sleep, for example because locks
are held by the caller, the code path should be restructured to allow the
allocation to be made in another place.
.El
.El
.Pp
The contents of allocated memory are uninitialized.
.Pp
Unlike Solaris, kmem_alloc(0, flags) is illegal.
.Pp
.\" ------------------------------------------------------------
.Fn kmem_zalloc
is the equivalent of
.Fn kmem_alloc ,
except that it initializes the memory to zero.
.Pp
.\" ------------------------------------------------------------
.Fn kmem_asprintf
functions as the well known
.Fn asprintf
function, but allocates memory using
.Fn kmem_alloc .
This routine can sleep during allocation.
The size of the allocated area is the length of the returned character string, plus one (for the NUL terminator).
This must be taken into consideration when freeing the returned area with
.Fn kmem_free .
.Pp
.\" ------------------------------------------------------------
.Fn kmem_free
frees kernel wired memory allocated by
.Fn kmem_alloc
or
.Fn kmem_zalloc
so that it can be used for other purposes.
It takes the following arguments.
.Bl -tag -width kmflags
.It Fa p
The pointer to the memory being freed.
It must be the one returned by
.Fn kmem_alloc
or
.Fn kmem_zalloc .
.It Fa size
The size of the memory being freed, in bytes.
It must be the same as the
.Fa size
argument used for
.Fn kmem_alloc
or
.Fn kmem_zalloc
when the memory was allocated.
.El
.Pp
Freeing
.Dv NULL
is illegal.
.Pp
.\" ------------------------------------------------------------
.Fn kmem_intr_alloc ,
.Fn kmem_intr_zalloc
and
.Fn kmem_intr_free
are the equivalents of the above kmem routines which can be called
from the interrupt context.
These routines are for the special cases.
Normally,
.Xr pool_cache 9
should be used for memory allocation from interrupt context.
.Pp
The
.Fn kmem_strdupsize
function is a utility function that can be used to copy the string in the
.Fa str
argument to a new buffer allocated using
.Fn kmem_alloc
and optionally return the size of the allocation (the length of the string
plus the trailing
.Dv NUL )
in the
.Fa size
argument if that is not
.Dv NULL .
.Pp
The
.Fn kmem_strfree
function can be used to free a
.Dv NUL
terminated string computing the length of the string using
.Xr strlen 3
and adding one for the
.Dv NUL
and then using
.Fn kmem_free .
.\" ------------------------------------------------------------
.Sh NOTES
Making
.Dv KM_SLEEP
allocations while holding mutexes or reader/writer locks is discouraged, as the
caller can sleep for an unbounded amount of time in order to satisfy the
allocation.
This can in turn block other threads that wish to acquire locks held by the
caller.
It should be noted that
.Fn kmem_free
may also block.
.Pp
For some locks this is permissible or even unavoidable.
For others, particularly locks that may be taken from soft interrupt context,
it is a serious problem.
As a general rule it is better not to allow this type of situation to develop.
One way to circumvent the problem is to make allocations speculative and part
of a retryable sequence.
For example:
.Bd -literal
  retry:
        /* speculative unlocked check */
        if (need to allocate) {
                new_item = kmem_alloc(sizeof(*new_item), KM_SLEEP);
        } else {
                new_item = NULL;
        }
        mutex_enter(lock);
        /* check while holding lock for true status */
        if (need to allocate) {
                if (new_item == NULL) {
                        mutex_exit(lock);
                        goto retry;
                }
                consume(new_item);
                new_item = NULL;
        }
        mutex_exit(lock);
        if (new_item != NULL) {
                /* did not use it after all */
                kmem_free(new_item, sizeof(*new_item));
        }
.Ed
.\" ------------------------------------------------------------
.Sh OPTIONS
.Ss KMEM_SIZE
Kernels compiled with the
.Dv KMEM_SIZE
option ensure the size given in
.Fn kmem_free
matches the actual allocated size.
On
.Fn kmem_alloc ,
the kernel will allocate an additional contiguous kmem page of eight
bytes in the buffer, will register the allocated size in the first kmem
page of that buffer, and will return a pointer to the second kmem page
in that same buffer.
When freeing, the kernel reads the first page, and compares the
size registered with the one given in
.Fn kmem_free .
Any mismatch triggers a panic.
.Pp
.Dv KMEM_SIZE
is enabled by default on
.Dv DIAGNOSTIC
and
.Dv DEBUG .
.Ss KMEM_REDZONE
Kernels compiled with the
.Dv KMEM_REDZONE
option add a dynamic pattern of two bytes at the end of each allocated
buffer, and check this pattern when freeing to ensure the caller hasn't
written outside the requested area.
This option does not introduce a significant performance impact,
but has two drawbacks: it only catches write overflows, and catches
them only on
.Fn kmem_free .
.Pp
.Dv KMEM_REDZONE
is enabled by default on
.Dv DIAGNOSTIC .
.Ss KMEM_GUARD
Kernels compiled with the
.Dv KMEM_GUARD
option perform CPU intensive sanity checks on kmem operations.
It adds additional, very high overhead runtime verification to kmem
operations.
It must be enabled with
.Dv KMEM_SIZE .
.Pp
.Dv KMEM_GUARD
tries to catch the following types of bugs:
.Bl -bullet
.It
Overflow at time of occurrence, by means of a guard page.
An unmapped guard page sits immediately after the requested area;
a read/write overflow therefore triggers a page fault.
.It
Underflow at
.Fn kmem_free ,
by using
.Dv KMEM_SIZE Ap s
registered size.
If an underflow occurs, the size stored by
.Dv KMEM_SIZE
will be overwritten, which means that when freeing, the kernel will
spot the mismatch.
.It
Use-after-free at time of occurrence.
When freeing, the memory is unmapped, and depending on the value
of kmem_guard_depth, the kernel will more or less delay the recycling
of that memory.
Which means that any ulterior read/write access to the memory will
trigger a page fault, given it hasn't been recycled yet.
.El
.Pp
To enable it, boot the system with the
.Fl d
option, which causes the debugger to be entered early during the kernel
boot process.
Issue commands such as the following:
.Bd -literal
db> w kmem_guard_depth 0t30000
db> c
.Ed
.Pp
This instructs
.Dv kmem_guard
to queue up to 60000 (30000*2) pages of unmapped KVA to catch
use-after-free type errors.
When
.Fn kmem_free
is called, memory backing a freed item is unmapped and the kernel VA
space pushed onto a FIFO.
The VA space will not be reused until another 30k items have been freed.
Until reused the kernel will catch invalid accesses and panic with a page fault.
Limitations:
.Bl -bullet
.It
It has a severe impact on performance.
.It
It is best used on a 64-bit machine with lots of RAM.
.El
.Pp
.Dv KMEM_GUARD
is enabled by default on
.Dv DEBUG .
.Sh RETURN VALUES
On success,
.Fn kmem_alloc ,
.Fn kmem_asprintf ,
.Fn kmem_intr_alloc ,
.Fn kmem_intr_zalloc ,
.Fn kmem_strdupsize ,
and
.Fn kmem_zalloc
return a pointer to allocated memory.
Otherwise,
.Dv NULL
is returned.
.\" ------------------------------------------------------------
.Sh CODE REFERENCES
The
.Nm
subsystem is implemented within the file
.Pa sys/kern/subr_kmem.c .
.\" ------------------------------------------------------------
.Sh SEE ALSO
.Xr intro 9 ,
.Xr memoryallocators 9 ,
.Xr percpu 9 ,
.Xr pool_cache 9 ,
.Xr uvm_km 9
.\" ------------------------------------------------------------
.Sh CAVEATS
The
.Fn kmem_alloc ,
.Fn kmem_asprintf ,
.Fn kmem_free ,
.Fn kmem_strdupsize ,
.Fn kmem_strfree ,
and
.Fn kmem_zalloc
functions cannot be used from interrupt context, from a soft interrupt,
or from a callout.
Use
.Xr pool_cache 9
in these situations.
.\" ------------------------------------------------------------
.Sh SECURITY CONSIDERATIONS
As the memory allocated by
.Fn kmem_alloc
is uninitialized, it can contain security-sensitive data left by its
previous user.
It is the caller's responsibility not to expose it to the world.