xref: /dflybsd-src/share/man/man9/bus_dma.9 (revision 030b0c8c4cf27c560ccec70410c8e21934ae677d)

.\"
.\" Copyright (c) 2002, 2003, 2004 The DragonFly Project.  All rights reserved.
.\"
.\" This code is derived from software contributed to The DragonFly Project
.\" by Hiten Pandya <hmp@backplane.com>.
.\"
.\" 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 DragonFly Project 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 COPYRIGHT HOLDERS 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
.\" COPYRIGHT HOLDERS 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.
.\"
.\" Copyright (c) 1996, 1997, 1998, 2001 The NetBSD Foundation, Inc.
.\" All rights reserved.
.\"
.\" This code is derived from software contributed to The NetBSD Foundation
.\" by Jason R. Thorpe of the Numerical Aerospace Simulation Facility,
.\" NASA Ames Research Center.
.\"
.\" 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. All advertising materials mentioning features or use of this software
.\"    must display the following acknowledgment:
.\" 	This product includes software developed by the NetBSD
.\" 	Foundation, Inc. and its contributors.
.\" 4. Neither the name of The NetBSD Foundation 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 NETBSD FOUNDATION, INC. 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 FOUNDATION 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: /repoman/r/ncvs/src/share/man/man9/bus_dma.9,v 1.7 2003/07/27 14:05:29 mux Exp $
.\" $NetBSD: bus_dma.9,v 1.25 2002/10/14 13:43:16 wiz Exp $
.\"
.Dd December 09, 2023
.Dt BUS_DMA 9
.Os
.Sh NAME
.Nm bus_dma ,
.Nm bus_dma_tag_create ,
.Nm bus_dma_tag_destroy ,
.Nm bus_dmamap_create ,
.Nm bus_dmamap_destroy ,
.Nm bus_dmamap_load ,
.Nm bus_dmamap_load_ccb ,
.Nm bus_dmamap_load_mbuf ,
.Nm bus_dmamap_load_mbuf_segment ,
.Nm bus_dmamap_load_mbuf_defrag ,
.Nm bus_dmamap_load_uio ,
.Nm bus_dmamap_unload ,
.Nm bus_dmamap_sync ,
.Nm bus_dmamem_alloc ,
.Nm bus_dmamem_coherent ,
.Nm bus_dmamem_coherent_any ,
.Nm bus_dmamem_free
.Nd Bus and Machine Independent DMA Mapping Interface
.Sh SYNOPSIS
.In sys/bus.h
.Ft int
.Fn bus_dma_tag_create "bus_dma_tag_t parent" "bus_size_t alignment" \
"bus_size_t boundary" "bus_addr_t lowaddr" "bus_addr_t highaddr" \
"bus_size_t maxsize" "int nsegments" "bus_size_t maxsegsz" \
"int flags" "bus_dma_tag_t *dmat"
.Ft int
.Fn bus_dma_tag_destroy "bus_dma_tag_t dmat"
.Ft int
.Fn bus_dmamap_create "bus_dma_tag_t dmat" "int flags" "bus_dmamap_t *mapp"
.Ft int
.Fn bus_dmamap_destroy "bus_dma_tag_t dmat" "bus_dmamap_t map"
.Ft int
.Fn bus_dmamap_load "bus_dma_tag_t dmat" "bus_dmamap_t map" "void *buf" \
"bus_size_t buflen" "bus_dmamap_callback_t *callback" "void *callback_arg" \
"int flags"
.Ft int
.Fn bus_dmamap_load_ccb "bus_dma_tag_t dmat" "bus_dmamap_t map" \
"union ccb *ccb" "bus_dmamap_callback_t *callback" "void *callback_arg" \
"int flags"
.Ft int
.Fn bus_dmamap_load_mbuf "bus_dma_tag_t dmat" "bus_dmamap_t map" \
"struct mbuf *mbuf" "bus_dmamap_callback2_t *callback" "void *callback_arg" \
"int flags"
.Ft int
.Fn bus_dmamap_load_mbuf_segment "bus_dma_tag_t dmat" "bus_dmamap_t map" \
"struct mbuf *mbuf" "bus_dma_segment_t *segs" "int maxsegs" "int *nsegs" \
"int flags"
.Ft int
.Fn bus_dmamap_load_mbuf_defrag "bus_dma_tag_t dmat" "bus_dmamap_t map" \
"struct mbuf **mbuf" "bus_dma_segment_t *segs" "int maxsegs" "int *nsegs" \
"int flags"
.Ft int
.Fn bus_dmamap_load_uio "bus_dma_tag_t dmat" "bus_dmamap_t map" \
"struct uio *uio" "bus_dmamap_callback2_t *callback" "void *callback_arg" \
"int flags"
.Ft int
.Fn bus_dmamem_alloc "bus_dma_tag_t dmat" "void **vaddr" \
"int flags" "bus_dmamap_t *mapp"
.Ft int
.Fn bus_dmamem_coherent "bus_dma_tag_t parent" "bus_size_t alignment" \
"bus_size_t boundary" "bus_addr_t lowaddr" "bus_addr_t highaddr" \
"bus_size_t maxsize" "int flags" "bus_dmamem_t *dmem"
.Ft void *
.Fn bus_dmamem_coherent_any "bus_dma_tag_t parent" "bus_size_t alignment" \
"bus_size_t maxsize" "int flags" "bus_dma_tag_t *dtag" "bus_dmamap_t *dmap" \
"bus_addr_t *busaddr"
.Ft void
.Fn bus_dmamap_unload "bus_dma_tag_t dmat" "bus_dmamap_t map"
.Ft void
.Fn bus_dmamap_sync "bus_dma_tag_t dmat" "bus_dmamap_t map" \
"bus_dmasync_op_t op"
.Ft void
.Fn bus_dmamem_free "bus_dma_tag_t dmat" "void *vaddr" \
"bus_dmamap_t map"
.Sh DESCRIPTION
Direct Memory Access (DMA) is a method of transferring data
without involving the CPU, thus providing higher performance.
A DMA transaction can be achieved between device to memory,
device to device, or memory to memory.
.Pp
The
.Nm
API is a bus, device, and machine-independent (MI) interface to
DMA mechanisms.
It provides the client with flexibility and simplicity by
abstracting machine dependent issues like setting up
DMA mappings, handling cache issues, bus specific features
and limitations.
.Sh STRUCTURES AND TYPES
.Bl -tag -width compact
.It Vt bus_dma_tag_t
A machine-dependent (MD) opaque type that describes the
characteristics of DMA transactions.
DMA tags are organized into a hierarchy, with each child
tag inheriting the restrictions of its parent.
This allows all devices along the path of DMA transactions
to contribute to the constraints of those transactions.
.It Vt bus_dma_segment_t
A machine-dependent type that describes individual
DMA segments.
.Bd -literal
	bus_addr_t	ds_addr;
	bus_size_t	ds_len;
.Ed
.sp
The
.Fa ds_addr
field contains the device visible address of the DMA segment, and
.Fa ds_len
contains the length of the DMA segment.
Although the DMA segments returned by a mapping call will adhere to
all restrictions necessary for a successful DMA operation, some conversion
(e.g. a conversion from host byte order to the device's byte order) is
almost always required when presenting segment information to the device.
.It Vt bus_dmamap_t
A machine-dependent opaque type describing an individual mapping.
Multiple DMA maps can be associated with one DMA tag.
.It Vt bus_dmamem_t
A machine-dependent type that describes DMA memory created by
.Fn bus_dmamem_coherent .
.Bd -literal
	bus_dma_tag_t	dmem_tag;
	bus_dmamap_t	dmem_map;
	void		*dmem_addr;
	bus_addr_t	dmem_busaddr;
.Ed
.sp
The
.Fa dmem_tag
field contains the DMA tag of the DMA memory and
.Fa dmem_map
field contains the DMA map of the DMA memory.
The
.Fa dmem_addr
field points to the allocated DMA memory in kernel virtual address space.
The
.Fa dmem_busaddr
field contains the device visible address of the DMA memory.
.It Vt bus_dmamap_callback_t
Client specified callback for receiving mapping information resulting from
the load of a
.Vt bus_dmamap_t
via
.Fn bus_dmamap_load
or
.Fn bus_dmamap_load_ccb .
Callbacks are of the format:
.Bl -tag -width compact
.It Ft void
.Fn "client_callback" "void *callback_arg" "bus_dma_segment_t *segs" \
"int nseg" "int error"
.El
.sp
The
.Fa callback_arg
is the callback argument passed to dmamap load functions.
The
.Fa segs
and
.Fa nseg
parameters describe an array of
.Vt bus_dma_segment_t
structures that represent the mapping.
This array is only valid within the scope of the callback function.
The success or failure of the mapping is indicated by the
.Fa error
parameter.
More information on the use of callbacks can be found in the
description of the individual dmamap load functions.
.It Vt bus_dmamap_callback2_t
Client specified callback for receiving mapping information resulting from
the load of a
.Vt bus_dmamap_t
via
.Fn bus_dmamap_load_uio
or
.Fn bus_dmamap_load_mbuf .
.sp
Callback2s are of the format:
.Bl -tag -width compact
.It Ft void
.Fn "client_callback2" "void *callback_arg" "bus_dma_segment_t *segs" \
"int nseg" "bus_size_t mapsize" "int error"
.El
.sp
Callback2's behavior is the same as
.Vt bus_dmamap_callback_t
with the addition that the length of the data mapped is provided via
.Fa mapsize .
.It Vt bus_dmasync_op_t
Memory synchronization operation specifier.
Bus DMA requires explicit synchronization of memory with its device
visible mapping in order to guarantee memory coherency.
The
.Vt bus_dmasync_op_t
allows the type of DMA operation that will be or has been performed
to be communicated to the system so that the correct coherency measures
are taken.
All operations specified below are performed from the DMA engine's
point of view:
.Bl -tag -width BUS_DMASYNC_POSTWRITE
.It Dv BUS_DMASYNC_PREREAD
Perform any synchronization required after an update of memory by the CPU
but prior to DMA read operations.
.It Dv BUS_DMASYNC_PREWRITE
Perform any synchronization required after an update of memory by the CPU
but prior to DMA write operations.
.It Dv BUS_DMASYNC_POSTREAD
Perform any synchronization required after DMA read operations, but prior
to CPU access of the memory.
.It Dv BUS_DMASYNC_POSTWRITE
Perform any synchronization required after DMA write operations, but prior
to CPU access of the memory.
.El
.El
.sp
.Sh FUNCTIONS
.Bl -tag -width compact
.It Fn bus_dma_tag_create "parent" "alignment" "boundary" "lowaddr" \
"highaddr" "maxsize" "nsegments" "maxsegsz" "flags" "*dmat"
Allocates a device specific DMA tag, and initializes it according to
the arguments provided:
.Bl -tag -width alignment -compact
.It Fa parent
Indicates restrictions between the parent bridge, CPU memory, and the
device.
May be NULL, if no DMA restrictions are to be inherited.
.It Fa alignment
Alignment constraint, in bytes, of any mappings created using this tag.
The alignment must be a power of 2.
Hardware that can DMA starting at any address would specify
.Em 1
for byte alignment.
Hardware requiring DMA transfers to start on a multiple of 4K
would specify
.Em 4096 .
.It Fa boundary
Boundary constraint, in bytes, of the target DMA memory region.
The boundary indicates the set of addresses, all multiples of the
boundary argument, that cannot be crossed by a single
.Vt bus_dma_segment_t .
The boundary must be either a power of 2 or 0.
.Ql 0
indicates that there are no boundary restrictions.
.It Fa lowaddr
.It Fa highaddr
Bounds of the window of bus address space that
.Em cannot
be directly accessed by the device.
The window contains all address greater than lowaddr and
less than or equal to highaddr.
For example, a device incapable of DMA above 4GB, would specify
a highaddr of
.Dv BUS_SPACE_MAXADDR
and a lowaddr of
.Dv BUS_SPACE_MAXADDR_32BIT .
Similarly a device that can only dma to addresses below 16MB would
specify a highaddr of
.Dv BUS_SPACE_MAXADDR
and a lowaddr of
.Dv BUS_SPACE_MAXADDR_24BIT .
Some implementations requires that some region of device visible
address space, overlapping available host memory, be outside the
window.
This area of
.Ql safe memory
is used to bounce requests that would otherwise conflict with
the exclusion window.
.It Fa maxsize
Maximum size, in bytes, of the sum of all segment lengths in a given
DMA mapping associated with this tag.
.It Fa nsegments
Number of discontinuities (scatter/gather segments) allowed
in a DMA mapped region.
If there is no restriction,
.Dv BUS_SPACE_UNRESTRICTED
may be specified for the tag intended to be used as the parent.
.Dv BUS_SPACE_UNRESTRICTED
must not be specified for the tags
which will be used to create maps.
For tags which will be used to create maps,
this argument must be less than 16384 on x86_64.
.It Fa maxsegsz
Maximum size, in bytes, of a segment in any DMA mapped region associated
with
.Fa dmat .
.It Fa flags
Are as follows:
.Bl -tag -width ".Dv BUS_DMA_PRIVBZONE" -compact
.It Dv BUS_DMA_ALLOCNOW
Allocate the minimum resources necessary to guarantee that all map load
operations associated with this tag may not block.
If sufficient resources are not available,
.Er ENOMEM
is returned.
.It Dv BUS_DMA_WAITOK
Indicates that it is OK to wait for resources.
However,
unlike
.Xr kmalloc 9 ,
it is not guaranteed that the resource allocation will succeed.
This flag is the default one,
if
.Dv BUS_DMA_NOWAIT
is not supplied.
.It Dv BUS_DMA_NOWAIT
If the resource allocation request cannot be immediately fulfilled,
.Er ENOMEM
is returned.
.It Dv BUS_DMA_ONEBPAGE
Allocte one bounce page at most,
even if the
.Fa maxsize
indicates that multiple bounce pages are needed.
.It Dv BUS_DMA_ALIGNED
Indicates that all memory to be loaded into the DMA maps associated
with this DMA tag is properly aligned according to
.Fa alignment
constraint.
No resources,
e.g. bounce pages,
will be allocated due to the
.Fa alignment
constraint.
If unaligned memory was loaded into the DMA maps associated with this DMA tag,
system will panic.
.It Dv BUS_DMA_PRIVBZONE
Uses a private bounce zone instead of a shared one.
A private bounce zone will vanish if the DMA tag is destroyed.
.It Dv BUS_DMA_ALLOCALL
Allocate all required resources (mainly the bounce buffer).
If any allocation fails,
.Fn bus_dma_tag_create
fails.
.It Dv BUS_DMA_PROTECTED
All of the functions called with the DMA tag are already protected by the
caller, so the
.Nm
code need not protect the internal data structures.
.El
.It Fa dmat
Pointer to a bus_dma_tag_t where the resulting DMA tag will
be stored.
.El
.Pp
Returns
.Er ENOMEM
if sufficient memory is not available for tag creation
or allocating mapping resources.
.It Fn bus_dma_tag_destroy "dmat"
Deallocate the DMA tag
.Fa dmat
that was created by
.Fn bus_dma_tag_create .
.Pp
Returns
.Er EBUSY
if any DMA maps remain associated with
.Fa dmat
or
.Ql 0
on success.
.It Fn bus_dmamap_create "dmat" "flags" "*mapp"
Allocates and initializes a DMA map.
Arguments are as follows:
.Bl -tag -width nsegments -compact
.It Fa dmat
DMA tag.
.It Fa flags
Are as follows:
.Bl -tag -width ".Dv BUS_DMA_ONEBPAGE" -compact
.It Dv BUS_DMA_WAITOK
Indicates that it is OK to wait for resources.
However,
unlike
.Xr kmalloc 9 ,
it is not guaranteed that the resource allocation will succeed.
This flag is the default one,
if
.Dv BUS_DMA_NOWAIT
is not supplied.
.It Dv BUS_DMA_NOWAIT
If the resource allocation request cannot be immediately fulfilled,
.Er ENOMEM
is returned.
.It Dv BUS_DMA_ONEBPAGE
Allocte one bounce page at most,
even if the
.Fa maxsize
used to create the
.Fa dmat
indicates that multiple bounce pages are needed.
.El
.It Fa mapp
Pointer to a
.Vt bus_dmamap_t
where the resulting DMA map will be stored.
.El
.Pp
Returns
.Er ENOMEM
if sufficient memory is not available for creating the
map or allocating mapping resources.
.It Fn bus_dmamap_destroy "dmat" "map"
Frees all resources associated with a given DMA map.
Arguments are as follows:
.Bl -tag -width dmat -compact
.It Fa dmat
DMA tag used to allocate
.Fa map .
.It Fa map
The DMA map to destroy.
.El
.Pp
Returns
.Er EBUSY
if a mapping is still active for
.Fa map .
.It Fn bus_dmamap_load "dmat" "map" "buf" "buflen" "*callback" "..."
Creates a mapping in device visible address space of
.Fa buflen
bytes of
.Fa buf ,
associated with the DMA map
.Fa map .
Arguments are as follows:
.Bl -tag -width buflen -compact
.It Fa dmat
DMA tag used to allocate
.Fa map .
.It Fa map
A DMA map without a currently active mapping.
.It Fa buf
A kernel virtual address pointer to a contiguous (in KVA) buffer, to be
mapped into device visible address space.
.It Fa buflen
The size of the buffer.
.It Fa callback Fa callback_arg
The callback function, and its argument.
.It Fa flags
The value of this argument is currently undefined, and should be
specified as
.Ql 0 .
.El
.Pp
Return values to the caller are as follows:
.Bl -tag -width ".Er EINPROGRESS" -compact
.It 0
The callback has been called and completed.
The status of the mapping has been delivered to the callback.
.It Er EINPROGRESS
The mapping has been deferred for lack of resources.
The callback will be called as soon as resources are available.
Callbacks are serviced in FIFO order.
DMA maps created from DMA tags that are allocated with
the
.Dv BUS_DMA_ALLOCNOW
flag will never return this status for a load operation.
.It Er EINVAL
The load request was invalid.
The callback has not, and will not be called.
This error value may indicate that
.Fa dmat ,
.Fa map ,
.Fa buf ,
or
.Fa callback
were invalid, or
.Fa buslen
was larger than the
.Fa maxsize
argument used to create the dma tag
.Fa dmat .
.El
.Pp
When the callback is called, it is presented with an error value
indicating the disposition of the mapping.
Error may be one of the following:
.Bl -tag -width ".Er EINPROGRESS" -compact
.It 0
The mapping was successful and the
.Fa dm_segs
callback argument contains an array of
.Vt bus_dma_segment_t
elements describing the mapping.
This array is only valid during the scope of the callback function.
.It Er EFBIG
A mapping could not be achieved within the segment constraints provided
in the tag even though the requested allocation size was less than maxsize.
.El
.It Fn bus_dmamap_load_ccb "dmat" "map" "ccb" "callback" "callback_arg" "flags"
This is a variation of
.Fn bus_dmamap_load
which maps data pointed to by
.Fa ccb
for DMA transfers.
.It Fn bus_dmamap_load_mbuf "dmat" "map" "mbuf" "callback2" "callback_arg" \
"flags"
This is a variation of
.Fn bus_dmamap_load
which maps mbuf chains
for DMA transfers.
A
.Vt bus_size_t
argument is also passed to the callback routine, which
contains the mbuf chain's packet header length.
.Pp
Mbuf chains are assumed to be in kernel virtual address space.
.Pp
Returns
.Er EINVAL
if the size of the mbuf chain exceeds the maximum limit of the
DMA tag.
.It Fn bus_dmamap_load_mbuf_segment "dmat" "map" "mbuf" "*segs" "maxsegs" \
"*nsegs" "flags"
It is like
.Fn bus_dmamap_load_mbuf
without callback.
Segmentation information are saved in the
.Fa segs
and
.Fa nsegs
if the loading is successful.
The
.Fa maxsegs ,
which indicates the number of elements in the
.Fa segs ,
must be set by the caller and must be at least 1 and at most equal the
.Fa nsegments
used to create the
.Fa dmat .
The
.Fa flags
must have
.Dv BUS_DMA_NOWAIT
turned on.
.Pp
This function will not block.
When system is short of DMA resources,
this function will return
.Er ENOMEM ,
instead of
.Er EINPROGRESS .
.It Fn bus_dmamap_load_mbuf_defrag "dmat" "map" "*mbuf" "*segs" "maxsegs" \
"*nsegs" "flags"
This function is like
.Fn bus_dmamap_load_mbuf_segment ,
but it will call
.Fn m_defrag
on the
.Fa *mbuf
and try reloading,
if low level code indicates too many fragments in the
.Fa *mbuf ;
the
.Fa mbuf
will be updated under this situation.
However,
.Fa *mbuf
would not be freed by this function,
even if
.Fn m_defrag
failed.
.Pp
Return
.Er ENOBUFS ,
if the calling of
.Fn m_defrag
failed.
.It Fn bus_dmamap_load_uio "dmat" "map" "uio" "callback2" "callback_arg" "flags"
This is a variation of
.Fn bus_dmamap_load
which maps buffers pointed to by
.Fa uio
for DMA transfers.
A
.Vt bus_size_t
argument is also passed to the callback routine, which contains the size of
.Fa uio ,
i.e.
.Fa uio->uio_resid .
.Pp
If
.Fa uio->uio_segflg
is
.Dv UIO_USERSPACE ,
then it is assumed that the buffer,
.Fa uio
is in
.Fa "uio->uio_td->td_proc" Ns 's
address space.
User space memory must be in-core and wired prior to attempting a map
load operation.
.It Fn bus_dmamap_unload "dmat" "map"
Unloads a DMA map.
Arguments are as follows:
.Bl -tag -width dmam -compact
.It Fa dmat
DMA tag used to allocate
.Fa map .
.It Fa map
The DMA map that is to be unloaded.
.El
.Pp
.Fn bus_dmamap_unload
will not perform any implicit synchronization of DMA buffers.
This must be done explicitly by a call to
.Fn bus_dmamap_sync
prior to unloading the map.
.It Fn bus_dmamap_sync "dmat" "map" "op"
Performs synchronization of a device visible mapping with the CPU visible
memory referenced by that mapping.
Arguments are as follows:
.Bl -tag -width dmat -compact
.It Fa dmat
DMA tag used to allocate
.Fa map .
.It Fa map
The DMA mapping to be synchronized.
.It Fa op
Type of synchronization operation to perform.
See the definition of
.Vt bus_dmasync_op_t
for a description of the acceptable values for
.Fa op .
.El
.Pp
.Fn bus_dmamap_sync
is the method used to ensure that CPU and device DMA access to shared
memory is coherent.
For example, the CPU might be used to setup the contents of a buffer
that is to be DMA'ed into a device.
To ensure that the data are visible via the device's mapping of that
memory, the buffer must be loaded and a dma sync operation of
.Dv BUS_DMASYNC_PREREAD
must be performed.
Additional sync operations must be performed after every CPU write
to this memory if additional DMA reads are to be performed.
Conversely, for the DMA write case, the buffer must be loaded,
and a dma sync operation of
.Dv BUS_DMASYNC_PREWRITE
must be performed.
The CPU will only be able to see the results of this DMA write
once the DMA has completed and a
.Dv BUS_DMASYNC_POSTWRITE
operation has been performed.
.Pp
If DMA read and write operations are not preceded and followed by the
appropriate synchronization operations, behavior is undefined.
.It Fn bus_dmamem_alloc "dmat" "**vaddr" "flags" "mapp"
Allocates memory that is mapped into KVA at the address returned
in
.Fa vaddr
that is permanently loaded into the newly created
.Vt bus_dmamap_t
returned via
.Fa mapp .
Arguments are as follows:
.Bl -tag -width alignment -compact
.It Fa dmat
DMA tag describing the constraints of the DMA mapping.
.It Fa vaddr
Pointer to a pointer that will hold the returned KVA mapping of
the allocated region.
.It Fa flags
Flags are defined as follows:
.Bl -tag -width ".Dv BUS_DMA_COHERENT" -compact
.It Dv BUS_DMA_WAITOK
The routine can safely wait (sleep) for resources.
.It Dv BUS_DMA_NOWAIT
The routine is not allowed to wait for resources.
If resources are not available,
.Er ENOMEM
is returned.
.It Dv BUS_DMA_COHERENT
Attempt to map this memory such that cache sync operations are
as cheap as possible.
This flag is typically set on memory that will be accessed by both
a CPU and a DMA engine, frequently.
Use of this flag does not remove the requirement of using
bus_dmamap_sync, but it may reduce the cost of performing
these operations.
.It Dv BUS_DMA_ZERO
Causes the allocated memory to be set to all zeros.
.It Dv BUS_DMA_NOCACHE
The allocated memory will not be cached in the processor caches.
All memory accesses appear on the bus and are executed
without reordering.
On x86_64, the
.Dv BUS_DMA_NOCACHE
flag results in the
Strong Uncacheable PAT to be set for the allocated virtual address range.
.El
.It Fa mapp
Pointer to storage for the returned DMA map.
.El
.Pp
The size of memory to be allocated is
.Fa maxsize
as specified in
.Fa dmat .
.Pp
The current implementation of
.Fn bus_dmamem_alloc
will allocate all requests as a single segment.
.Pp
Although no explicit loading is required to access the memory
referenced by the returned map, the synchronization requirements
as described in the
.Fn bus_dmamap_sync
section still apply.
.Pp
Returns
.Er ENOMEM
if sufficient memory is not available for completing
the operation.
.It Fn bus_dmamem_coherent "parent" "alignment" "boundary" "lowaddr" \
"highaddr" "maxsize" "flags" "*dmem"
This is a convenient function to create one segment of DMA memory.
It combines following
.Nm
function calls:
.Bd -literal
	bus_dma_tag_create(..., dtag);
	bus_dmamem_alloc(*dtag, vaddr, ..., dmap);
	bus_dmamap_load(*dtag, *dmap, *vaddr, ..., \\
			callback, busaddr, ...);
.Ed
.sp
The final results of the above function calls are:
DMA tag,
DMA map,
DMA memory's kernel virtual address and
its device visible address.
.Fn bus_dmamem_coherent
saves the results in
.Fa *dmem .
.Pp
The
.Fa parent ,
.Fa alignment ,
.Fa boundary ,
.Fa lowaddr
and
.Fa highaddr
will be passed to
.Fn bus_dma_tag_create
as they are.
The
.Fa maxsize
will be passed to
.Fn bus_dma_tag_create
as its
.Fa maxsize
and
.Fa maxsegsz
and
.Ql 1
will be passed to
.Fn bus_dma_tag_create
as its
.Fa nsegments .
When
.Fn bus_dmamem_alloc
is called,
.Fa flags
will be first or'ed with
.Dv BUS_DMA_COHERENT
then passed to it.
The final results of the above three functions,
i.e. DMA tag,
DMA map,
DMA memory's kernel virtual address and
its device visible address,
are saved in
.Fa *dmem .
If any of the three functions failed,
this function will return the error code and the
.Fa *dmem
should not be used.
.It Fn bus_dmamem_coherent_any "parent" "alignment" "maxsize" "flags" \
"*dtag" "*dmap" "*busaddr"
This function is a simplified version of
.Fn bus_dmamem_coherent
with
its
.Fa boundary
set to
.Ql 0 ,
.Fa lowaddr
set to
.Dv BUS_SPACE_MAXADDR
and
.Fa highaddr
set to
.Dv BUS_SPACE_MAXADDR .
The
.Fa parent
usually should not be NULL.
.Pp
Return the DMA memory's kernel virtual address.
The DMA tag, DMA map and device visible address are returned in
.Fa *dtag ,
.Fa *dmap ,
and
.Fa *busaddr .
If this function failed,
NULL will be returned;
.Fa *dtag ,
.Fa *dmap ,
and
.Fa *busaddr
are left unchanged.
.It Fn bus_dmamem_free "dmat" "*vaddr" "map"
Frees memory previously allocated by
.Fn bus_dmamem_alloc .
Any mappings
will be invalidated.
Arguments are as follows:
.Bl -tag -width vaddr -compact
.It Fa dmat
DMA tag.
.It Fa vaddr
Kernel virtual address of the memory.
.It Fa map
DMA map to be invalidated.
.El
.El
.Sh RETURN VALUES
Behavior is undefined if invalid arguments are passed to
any of the above functions.
If sufficient resources cannot be allocated for a given
transaction,
.Er ENOMEM
is returned.
All
routines that are not of type,
.Vt void ,
will return 0 on success or an error
code, as discussed above.
.Pp
All
.Vt void
routines will succeed if provided with valid arguments.
.Sh SEE ALSO
.Xr devclass 9 ,
.Xr device 9 ,
.Xr driver 9 ,
.Xr rman 9
.Rs
.%A "Jason R. Thorpe"
.%T "A Machine-Independent DMA Framework for NetBSD"
.%J "Proceedings of the Summer 1998 USENIX Technical Conference"
.%Q "USENIX Association"
.%D "June 1998"
.Re
.Sh HISTORY
The
.Nm
interface first appeared in
.Nx 1.3 .
.Pp
The
.Nm
API was adopted from
.Nx
for use in the CAM SCSI subsystem.
The alterations to the original API were aimed to remove the need for
a
.Vt bus_dma_segment_t
array stored in each
.Vt bus_dmamap_t
while allowing callers to queue up on scarce resources.
.Sh AUTHORS
The
.Nm
interface was designed and implemented by
.An Jason R. Thorpe
of the Numerical Aerospace Simulation Facility, NASA Ames Research Center.
Additional input on the
.Nm
design was provided by
.An -nosplit
.An Chris Demetriou ,
.An Charles Hannum ,
.An Ross Harvey ,
.An Matthew Jacob ,
.An Jonathan Stone ,
and
.An Matt Thomas .
.Pp
This manual page was written by
.An Hiten Pandya
and
.An Justin T. Gibbs .