BUS_DMA(9) BSD Kernel Developer's Manual BUS_DMA(9)
bus_dma, bus_dma_tag_create, bus_dma_tag_destroy, bus_dmamap_create, bus_dmamap_destroy, bus_dmamap_load, bus_dmamap_load_mbuf,
bus_dmamap_load_mbuf_sg, bus_dmamap_load_uio, bus_dmamap_unload, bus_dmamap_sync, bus_dmamem_alloc, bus_dmamem_free -- Bus and Machine Inde-
pendent DMA Mapping Interface
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_dma_filter_t *filtfunc, void *filtfuncarg, bus_size_t maxsize, int nsegments, bus_size_t maxsegsz, int flags,
bus_dma_lock_t *lockfunc, void *lockfuncarg, bus_dma_tag_t *dmat);
bus_dmamap_create(bus_dma_tag_t dmat, int flags, bus_dmamap_t *mapp);
bus_dmamap_destroy(bus_dma_tag_t dmat, bus_dmamap_t map);
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,
bus_dmamap_load_mbuf(bus_dma_tag_t dmat, bus_dmamap_t map, struct mbuf *mbuf, bus_dmamap_callback2_t *callback, void *callback_arg,
bus_dmamap_load_mbuf_sg(bus_dma_tag_t dmat, bus_dmamap_t map, struct mbuf *mbuf, bus_dma_segment_t *segs, int *nsegs, int flags);
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);
bus_dmamap_unload(bus_dma_tag_t dmat, bus_dmamap_t map);
bus_dmamap_sync(bus_dma_tag_t dmat, bus_dmamap_t map, op);
bus_dmamem_alloc(bus_dma_tag_t dmat, void **vaddr, int flags, bus_dmamap_t *mapp);
bus_dmamem_free(bus_dma_tag_t dmat, void *vaddr, bus_dmamap_t map);
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.
The bus_dma API is a bus, device, and machine-independent (MI) interface to DMA mechanisms. It provides the client with flexibility and sim-
plicity by abstracting machine dependent issues like setting up DMA mappings, handling cache issues, bus specific features and limitations.
STRUCTURES AND TYPES
A machine-dependent (MD) opaque type that describes the characteristics of DMA transactions. DMA tags are organized into a hierar-
chy, 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.
Client specified address filter having the format:
int client_filter(void *filtarg, bus_addr_t testaddr)
Address filters can be specified during tag creation to allow for devices whose DMA address restrictions cannot be specified by a
single window. The filtarg argument is specified by the client during tag creation to be passed to all invocations of the callback.
The testaddr argument contains a potential starting address of a DMA mapping. The filter function operates on the set of addresses
from testaddr to 'trunc_page(testaddr) + PAGE_SIZE - 1', inclusive. The filter function should return zero if any mapping in this
range can be accommodated by the device and non-zero otherwise.
A machine-dependent type that describes individual DMA segments. It contains the following fields:
The ds_addr field contains the device visible address of the DMA segment, and 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.
A machine-dependent opaque type describing an individual mapping. One map is used for each memory allocation that will be loaded.
Maps can be reused once they have been unloaded. Multiple maps can be associated with one DMA tag. While the value of the map may
evaluate to NULL on some platforms under certain conditions, it should never be assumed that it will be NULL in all cases.
Client specified callback for receiving mapping information resulting from the load of a bus_dmamap_t via bus_dmamap_load(). Call-
backs are of the format:
void client_callback(void *callback_arg, bus_dma_segment_t *segs, int nseg, int error)
The callback_arg is the callback argument passed to dmamap load functions. The segs and nseg arguments describe an array of
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 error argument. More information on the use of callbacks can be found in the
description of the individual dmamap load functions.
Client specified callback for receiving mapping information resulting from the load of a bus_dmamap_t via bus_dmamap_load_uio() or
Callback2s are of the format:
void client_callback2(void *callback_arg, bus_dma_segment_t *segs, int nseg, bus_size_t mapsize, int error)
Callback2's behavior is the same as bus_dmamap_callback_t with the addition that the length of the data mapped is provided via
Memory synchronization operation specifier. Bus DMA requires explicit synchronization of memory with its device visible mapping in
order to guarantee memory coherency. The 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. The operations are represented as bitfield flags that
can be combined together, though it only makes sense to combine PRE flags or POST flags, not both. See the bus_dmamap_sync()
description below for more details on how to use these operations.
All operations specified below are performed from the host memory point of view, where a read implies data coming from the device to
the host memory, and a write implies data going from the host memory to the device. Alternatively, the operations can be thought of
in terms of driver operations, where reading a network packet or storage sector corresponds to a read operation in bus_dma.
BUS_DMASYNC_PREREAD Perform any synchronization required prior to an update of host memory by the device.
BUS_DMASYNC_PREWRITE Perform any synchronization required after an update of host memory by the CPU and prior to device access to
BUS_DMASYNC_POSTREAD Perform any synchronization required after an update of host memory by the device and prior to CPU access to
BUS_DMASYNC_POSTWRITE Perform any synchronization required after device access to host memory.
Client specified lock/mutex manipulation method. This will be called from within busdma whenever a client lock needs to be manipu-
lated. In its current form, the function will be called immediately before the callback for a DMA load operation that has been
deferred with BUS_DMA_LOCK and immediately after with BUS_DMA_UNLOCK. If the load operation does not need to be deferred, then it
will not be called since the function loading the map should be holding the appropriate locks. This method is of the format:
void lockfunc(void *lockfunc_arg, bus_dma_lock_op_t op)
The lockfuncarg argument is specified by the client during tag creation to be passed to all invocations of the callback. The op
argument specifies the lock operation to perform.
Two lockfunc implementations are provided for convenience. busdma_lock_mutex() performs standard mutex operations on the sleep mutex
provided via lockfuncarg. dflt_lock() will generate a system panic if it is called. It is substituted into the tag when lockfunc is
passed as NULL to bus_dma_tag_create() and is useful for tags that should not be used with deferred load operations.
Operations to be performed by the client-specified lockfunc().
BUS_DMA_LOCK Acquires and/or locks the client locking primitive.
BUS_DMA_UNLOCK Releases and/or unlocks the client locking primitive.
bus_dma_tag_create(parent, alignment, boundary, lowaddr, highaddr, *filtfunc, *filtfuncarg, maxsize, nsegments, maxsegsz, flags, lockfunc,
Allocates a device specific DMA tag, and initializes it according to the arguments provided:
parent Indicates restrictions between the parent bridge, CPU memory, and the device. Each device must use a master parent tag
by calling bus_get_dma_tag().
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 1 for byte alignment. Hardware requiring DMA transfers to start on a
multiple of 4K would specify 4096.
boundary Boundary constraint, in bytes, of the target DMA memory region. The boundary indicates the set of addresses, all multi-
ples of the boundary argument, that cannot be crossed by a single bus_dma_segment_t. The boundary must be a power of 2
and must be no smaller than the maximum segment size. '0' indicates that there are no boundary restrictions.
Bounds of the window of bus address space that cannot be directly accessed by the device. The window contains all
addresses greater than lowaddr and less than or equal to highaddr. For example, a device incapable of DMA above 4GB,
would specify a highaddr of BUS_SPACE_MAXADDR and a lowaddr of BUS_SPACE_MAXADDR_32BIT. Similarly a device that can
only perform DMA to addresses below 16MB would specify a highaddr of BUS_SPACE_MAXADDR and a lowaddr of
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 'safe memory' is used to bounce requests that would other-
wise conflict with the exclusion window.
filtfunc Optional filter function (may be NULL) to be called for any attempt to map memory into the window described by lowaddr
and highaddr. A filter function is only required when the single window described by lowaddr and highaddr cannot ade-
quately describe the constraints of the device. The filter function will be called for every machine page that overlaps
the exclusion window.
filtfuncarg Argument passed to all calls to the filter function for this tag. May be NULL.
maxsize Maximum size, in bytes, of the sum of all segment lengths in a given DMA mapping associated with this tag.
nsegments Number of discontinuities (scatter/gather segments) allowed in a DMA mapped region. If there is no restriction,
BUS_SPACE_UNRESTRICTED may be specified.
maxsegsz Maximum size, in bytes, of a segment in any DMA mapped region associated with dmat.
flags Are as follows:
BUS_DMA_ALLOCNOW Pre-allocate enough resources to handle at least one map load operation on this tag. If sufficient
resources are not available, ENOMEM is returned. This should not be used for tags that only describe
buffers that will be allocated with bus_dmamem_alloc(). Also, due to resource sharing with other
tags, this flag does not guarantee that resources will be allocated or reserved exclusively for this
tag. It should be treated only as a minor optimization.
lockfunc Optional lock manipulation function (may be NULL) to be called when busdma needs to manipulate a lock on behalf of the
client. If NULL is specified, dflt_lock() is used.
lockfuncarg Optional argument to be passed to the function specified by lockfunc.
dmat Pointer to a bus_dma_tag_t where the resulting DMA tag will be stored.
Returns ENOMEM if sufficient memory is not available for tag creation or allocating mapping resources.
Deallocate the DMA tag dmat that was created by bus_dma_tag_create().
Returns EBUSY if any DMA maps remain associated with dmat or '0' on success.
bus_dmamap_create(dmat, flags, *mapp)
Allocates and initializes a DMA map. Arguments are as follows:
dmat DMA tag.
flags Are as follows:
BUS_DMA_COHERENT Attempt to map the memory loaded with this map such that cache sync operations are as cheap as possible.
This flag is typically set on maps when the memory loaded with these will be accessed by both a CPU and
a DMA engine, frequently such as control data and as opposed to streamable data such as receive and
transmit buffers. Use of this flag does not remove the requirement of using bus_dmamap_sync(), but it
may reduce the cost of performing these operations. For bus_dmamap_create(), the BUS_DMA_COHERENT flag
is currently implemented on sparc64.
mapp Pointer to a bus_dmamap_t where the resulting DMA map will be stored.
Returns ENOMEM if sufficient memory is not available for creating the map or allocating mapping resources.
Frees all resources associated with a given DMA map. Arguments are as follows:
dmat DMA tag used to allocate map.
map The DMA map to destroy.
Returns EBUSY if a mapping is still active for map.
bus_dmamap_load(dmat, map, buf, buflen, *callback, callback_arg, flags)
Creates a mapping in device visible address space of buflen bytes of buf, associated with the DMA map map. This call will always
return immediately and will not block for any reason. Arguments are as follows:
dmat DMA tag used to allocate map.
map A DMA map without a currently active mapping.
buf A kernel virtual address pointer to a contiguous (in KVA) buffer, to be mapped into device visible address space.
buflen The size of the buffer.
The callback function, and its argument. This function is called once sufficient mapping resources are available for the DMA
operation. If resources are temporarily unavailable, this function will be deferred until later, but the load operation will
still return immediately to the caller. Thus, callers should not assume that the callback will be called before the load
returns, and code should be structured appropriately to handle this. See below for specific flags and error codes that con-
trol this behavior.
flags Are as follows:
BUS_DMA_NOWAIT The load should not be deferred in case of insufficient mapping resources, and instead should return immedi-
ately with an appropriate error.
The generated transactions to and from the virtual page are non-cacheable. For bus_dmamap_load(), the
BUS_DMA_NOCACHE flag is currently implemented on sparc64.
Return values to the caller are as follows:
0 The callback has been called and completed. The status of the mapping has been delivered to the callback.
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. To ensure that ordering is guaranteed, all subsequent load requests will also be
deferred until all callbacks have been processed.
ENOMEM The load request has failed due to insufficient resources, and the caller specifically used the BUS_DMA_NOWAIT flag.
EINVAL The load request was invalid. The callback has been called and has been provided the same error. This error value may
indicate that dmat, map, buf, or callback were invalid, or buflen was larger than the maxsize argument used to create
the dma tag dmat.
When the callback is called, it is presented with an error value indicating the disposition of the mapping. Error may be one of the
0 The mapping was successful and the dm_segs callback argument contains an array of bus_dma_segment_t elements describing
the mapping. This array is only valid during the scope of the callback function.
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.
bus_dmamap_load_mbuf(dmat, map, mbuf, callback2, callback_arg, flags)
This is a variation of bus_dmamap_load() which maps mbuf chains for DMA transfers. A bus_size_t argument is also passed to the call-
back routine, which contains the mbuf chain's packet header length. The BUS_DMA_NOWAIT flag is implied, thus no callback deferral
Mbuf chains are assumed to be in kernel virtual address space.
Beside the error values listed for bus_dmamap_load(), EINVAL will be returned if the size of the mbuf chain exceeds the maximum limit
of the DMA tag.
bus_dmamap_load_mbuf_sg(dmat, map, mbuf, segs, nsegs, flags)
This is just like bus_dmamap_load_mbuf() except that it returns immediately without calling a callback function. It is provided for
efficiency. The scatter/gather segment array segs is provided by the caller and filled in directly by the function. The nsegs argu-
ment is returned with the number of segments filled in. Returns the same errors as bus_dmamap_load_mbuf().
bus_dmamap_load_uio(dmat, map, uio, callback2, callback_arg, flags)
This is a variation of bus_dmamap_load() which maps buffers pointed to by uio for DMA transfers. A bus_size_t argument is also
passed to the callback routine, which contains the size of uio, i.e. uio->uio_resid. The BUS_DMA_NOWAIT flag is implied, thus no
callback deferral will happen. Returns the same errors as bus_dmamap_load().
If uio->uio_segflg is UIO_USERSPACE, then it is assumed that the buffer, uio is in uio->uio_td->td_proc's address space. User space
memory must be in-core and wired prior to attempting a map load operation. Pages may be locked using vslock(9).
Unloads a DMA map. Arguments are as follows:
dmat DMA tag used to allocate map.
map The DMA map that is to be unloaded.
bus_dmamap_unload() will not perform any implicit synchronization of DMA buffers. This must be done explicitly by a call to
bus_dmamap_sync() prior to unloading the map.
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 fol-
dmat DMA tag used to allocate map.
map The DMA mapping to be synchronized.
op Type of synchronization operation to perform. See the definition of bus_dmasync_op_t for a description of the acceptable val-
ues for op.
The bus_dmamap_sync() function is the method used to ensure that CPU's and device's direct memory access (DMA) to shared memory is
coherent. For example, the CPU might be used to set up the contents of a buffer that is to be made available to 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
BUS_DMASYNC_PREWRITE must be performed after the CPU has updated the buffer and before the device access is initiated. If the CPU
modifies this buffer again later, another BUS_DMASYNC_PREWRITE sync operation must be performed before an additional device access.
Conversely, suppose a device updates memory that is to be read by a CPU. In this case, the buffer must be loaded, and a DMA sync
operation of BUS_DMASYNC_PREREAD must be performed before the device access is initiated. The CPU will only be able to see the
results of this memory update once the DMA operation has completed and a BUS_DMASYNC_POSTREAD sync operation has been performed.
If read and write operations are not preceded and followed by the appropriate synchronization operations, behavior is undefined.
bus_dmamem_alloc(dmat, **vaddr, flags, *mapp)
Allocates memory that is mapped into KVA at the address returned in vaddr and that is permanently loaded into the newly created
bus_dmamap_t returned via mapp. Arguments are as follows:
dmat DMA tag describing the constraints of the DMA mapping.
vaddr Pointer to a pointer that will hold the returned KVA mapping of the allocated region.
flags Flags are defined as follows:
BUS_DMA_WAITOK The routine can safely wait (sleep) for resources.
BUS_DMA_NOWAIT The routine is not allowed to wait for resources. If resources are not available, ENOMEM is returned.
Attempt to map this memory in a coherent fashion. See bus_dmamap_create() above for a description of this
flag. For bus_dmamem_alloc(), the BUS_DMA_COHERENT flag is currently implemented on arm and sparc64.
BUS_DMA_ZERO Causes the allocated memory to be set to all zeros.
The allocated memory will not be cached in the processor caches. All memory accesses appear on the bus
and are executed without reordering. For bus_dmamem_alloc(), the BUS_DMA_NOCACHE flag is currently imple-
mented on amd64 and i386 where it results in the Strong Uncacheable PAT to be set for the allocated vir-
tual address range.
mapp Pointer to a bus_dmamap_t where the resulting DMA map will be stored.
The size of memory to be allocated is maxsize as specified in the call to bus_dma_tag_create() for dmat.
The current implementation of bus_dmamem_alloc() will allocate all requests as a single segment.
An initial load operation is required to obtain the bus address of the allocated memory, and an unload operation is required before
freeing the memory, as described below in bus_dmamem_free(). Maps are automatically handled by this function and should not be
explicitly allocated or destroyed.
Although an explicit load is not required for each access to the memory referenced by the returned map, the synchronization require-
ments as described in the bus_dmamap_sync() section still apply and should be used to achieve portability on architectures without
Returns ENOMEM if sufficient memory is not available for completing the operation.
bus_dmamem_free(dmat, *vaddr, map)
Frees memory previously allocated by bus_dmamem_alloc(). Any mappings will be invalidated. Arguments are as follows:
dmat DMA tag.
vaddr Kernel virtual address of the memory.
map DMA map to be invalidated.
Behavior is undefined if invalid arguments are passed to any of the above functions. If sufficient resources cannot be allocated for a given
transaction, ENOMEM is returned. All routines that are not of type void will return 0 on success or an error code on failure as discussed
All void routines will succeed if provided with valid arguments.
Two locking protocols are used by bus_dma. The first is a private global lock that is used to synchronize access to the bounce buffer pool
on the architectures that make use of them. This lock is strictly a leaf lock that is only used internally to bus_dma and is not exposed to
clients of the API.
The second protocol involves protecting various resources stored in the tag. Since almost all bus_dma operations are done through requests
from the driver that created the tag, the most efficient way to protect the tag resources is through the lock that the driver uses. In cases
where bus_dma acts on its own without being called by the driver, the lock primitive specified in the tag is acquired and released automati-
cally. An example of this is when the bus_dmamap_load() callback function is called from a deferred context instead of the driver context.
This means that certain bus_dma functions must always be called with the same lock held that is specified in the tag. These functions
There is one exception to this rule. It is common practice to call some of these functions during driver start-up without any locks held.
So long as there is a guarantee of no possible concurrent use of the tag by different threads during this operation, it is safe to not hold a
lock for these functions.
Certain bus_dma operations should not be called with the driver lock held, either because they are already protected by an internal lock, or
because they might sleep due to memory or resource allocation. The following functions must not be called with any non-sleepable locks held:
All other functions do not have a locking protocol and can thus be called with or without any system or driver locks held.
devclass(9), device(9), driver(9), rman(9), vslock(9)
Jason R. Thorpe, "A Machine-Independent DMA Framework for NetBSD", Proceedings of the Summer 1998 USENIX Technical Conference, USENIX
Association, June 1998.
The bus_dma interface first appeared in NetBSD 1.3.
The bus_dma API was adopted from NetBSD for use in the CAM SCSI subsystem. The alterations to the original API were aimed to remove the need
for a bus_dma_segment_t array stored in each bus_dmamap_t while allowing callers to queue up on scarce resources.
The bus_dma interface was designed and implemented by Jason R. Thorpe of the Numerical Aerospace Simulation Facility, NASA Ames Research Cen-
ter. Additional input on the bus_dma design was provided by Chris Demetriou, Charles Hannum, Ross Harvey, Matthew Jacob, Jonathan Stone, and
The bus_dma interface in FreeBSD benefits from the contributions of Justin T. Gibbs, Peter Wemm, Doug Rabson, Matthew N. Dodd, Sam Leffler,
Maxime Henrion, Jake Burkholder, Takahashi Yoshihiro, Scott Long and many others.
This manual page was written by Hiten M. Pandya and Justin T. Gibbs.
May 12, 2009 BSD