p71924506
/
linux_old1
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

DMA-API.txt: standardize document format 

Each text file under Documentation follows a different
format. Some doesn't even have titles!

Change its representation to follow the adopted standard,
using ReST markups for it to be parseable by Sphinx:

- Fix some title marks to match ReST;
- use :Author: for author name;
- foo_ is an hyperlink. Get rid of it;
- Mark literal blocks as such;
- Use tables on some places that are almost using the
  table format.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
This commit is contained in:
Mauro Carvalho Chehab 2017-05-14 07:27:52 -03:00 committed by Jonathan Corbet
parent a2fbbcea7b
commit 03158a70ad
1 changed files with 320 additions and 242 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

562
Documentation/DMA-API.txt
View File

@ -1,7 +1,8 @@
Dynamic DMA mapping using the generic device ============================================
============================================ Dynamic DMA mapping using the generic device
============================================
James E.J. Bottomley <James.Bottomley@HansenPartnership.com> :Author: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
This document describes the DMA API. For a more gentle introduction This document describes the DMA API. For a more gentle introduction
of the API (and actual examples), see Documentation/DMA-API-HOWTO.txt. of the API (and actual examples), see Documentation/DMA-API-HOWTO.txt.
@ -12,10 +13,10 @@ machines. Unless you know that your driver absolutely has to support
non-consistent platforms (this is usually only legacy platforms) you non-consistent platforms (this is usually only legacy platforms) you
should only use the API described in part I. should only use the API described in part I.
Part I - dma_ API Part I - dma_API
------------------------------------- ----------------
To get the dma_ API, you must #include <linux/dma-mapping.h>. This To get the dma_API, you must #include <linux/dma-mapping.h>. This
provides dma_addr_t and the interfaces described below. provides dma_addr_t and the interfaces described below.
A dma_addr_t can hold any valid DMA address for the platform. It can be A dma_addr_t can hold any valid DMA address for the platform. It can be
@ -26,9 +27,11 @@ address space and the DMA address space.
Part Ia - Using large DMA-coherent buffers Part Ia - Using large DMA-coherent buffers
------------------------------------------ ------------------------------------------
void * ::
dma_alloc_coherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag) void *
dma_alloc_coherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag)
Consistent memory is memory for which a write by either the device or Consistent memory is memory for which a write by either the device or
the processor can immediately be read by the processor or device the processor can immediately be read by the processor or device
@ -51,20 +54,24 @@ consolidate your requests for consistent memory as much as possible.
The simplest way to do that is to use the dma_pool calls (see below). The simplest way to do that is to use the dma_pool calls (see below).
The flag parameter (dma_alloc_coherent() only) allows the caller to The flag parameter (dma_alloc_coherent() only) allows the caller to
specify the GFP_ flags (see kmalloc()) for the allocation (the specify the ``GFP_`` flags (see kmalloc()) for the allocation (the
implementation may choose to ignore flags that affect the location of implementation may choose to ignore flags that affect the location of
the returned memory, like GFP_DMA). the returned memory, like GFP_DMA).
void * ::
dma_zalloc_coherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag) void *
dma_zalloc_coherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag)
Wraps dma_alloc_coherent() and also zeroes the returned memory if the Wraps dma_alloc_coherent() and also zeroes the returned memory if the
allocation attempt succeeded. allocation attempt succeeded.
void ::
dma_free_coherent(struct device *dev, size_t size, void *cpu_addr,
dma_addr_t dma_handle) void
dma_free_coherent(struct device *dev, size_t size, void *cpu_addr,
dma_addr_t dma_handle)
Free a region of consistent memory you previously allocated. dev, Free a region of consistent memory you previously allocated. dev,
size and dma_handle must all be the same as those passed into size and dma_handle must all be the same as those passed into
@ -78,7 +85,7 @@ may only be called with IRQs enabled.
Part Ib - Using small DMA-coherent buffers Part Ib - Using small DMA-coherent buffers
------------------------------------------ ------------------------------------------
To get this part of the dma_ API, you must #include <linux/dmapool.h> To get this part of the dma_API, you must #include <linux/dmapool.h>
Many drivers need lots of small DMA-coherent memory regions for DMA Many drivers need lots of small DMA-coherent memory regions for DMA
descriptors or I/O buffers. Rather than allocating in units of a page descriptors or I/O buffers. Rather than allocating in units of a page
@ -88,6 +95,8 @@ not __get_free_pages(). Also, they understand common hardware constraints
for alignment, like queue heads needing to be aligned on N-byte boundaries. for alignment, like queue heads needing to be aligned on N-byte boundaries.
::
struct dma_pool * struct dma_pool *
dma_pool_create(const char *name, struct device *dev, dma_pool_create(const char *name, struct device *dev,
size_t size, size_t align, size_t alloc); size_t size, size_t align, size_t alloc);
@ -103,16 +112,21 @@ in bytes, and must be a power of two). If your device has no boundary
crossing restrictions, pass 0 for alloc; passing 4096 says memory allocated crossing restrictions, pass 0 for alloc; passing 4096 says memory allocated
from this pool must not cross 4KByte boundaries. from this pool must not cross 4KByte boundaries.
::
void *dma_pool_zalloc(struct dma_pool *pool, gfp_t mem_flags, void *
dma_addr_t *handle) dma_pool_zalloc(struct dma_pool *pool, gfp_t mem_flags,
dma_addr_t *handle)
Wraps dma_pool_alloc() and also zeroes the returned memory if the Wraps dma_pool_alloc() and also zeroes the returned memory if the
allocation attempt succeeded. allocation attempt succeeded.
void *dma_pool_alloc(struct dma_pool *pool, gfp_t gfp_flags, ::
dma_addr_t *dma_handle);
void *
dma_pool_alloc(struct dma_pool *pool, gfp_t gfp_flags,
dma_addr_t *dma_handle);
This allocates memory from the pool; the returned memory will meet the This allocates memory from the pool; the returned memory will meet the
size and alignment requirements specified at creation time. Pass size and alignment requirements specified at creation time. Pass
@ -122,16 +136,20 @@ blocking. Like dma_alloc_coherent(), this returns two values: an
address usable by the CPU, and the DMA address usable by the pool's address usable by the CPU, and the DMA address usable by the pool's
device. device.
::
void dma_pool_free(struct dma_pool *pool, void *vaddr, void
dma_addr_t addr); dma_pool_free(struct dma_pool *pool, void *vaddr,
dma_addr_t addr);
This puts memory back into the pool. The pool is what was passed to This puts memory back into the pool. The pool is what was passed to
dma_pool_alloc(); the CPU (vaddr) and DMA addresses are what dma_pool_alloc(); the CPU (vaddr) and DMA addresses are what
were returned when that routine allocated the memory being freed. were returned when that routine allocated the memory being freed.
::
void dma_pool_destroy(struct dma_pool *pool); void
dma_pool_destroy(struct dma_pool *pool);
dma_pool_destroy() frees the resources of the pool. It must be dma_pool_destroy() frees the resources of the pool. It must be
called in a context which can sleep. Make sure you've freed all allocated called in a context which can sleep. Make sure you've freed all allocated
@ -141,32 +159,40 @@ memory back to the pool before you destroy it.
Part Ic - DMA addressing limitations Part Ic - DMA addressing limitations
------------------------------------ ------------------------------------
int ::
dma_set_mask_and_coherent(struct device *dev, u64 mask)
int
dma_set_mask_and_coherent(struct device *dev, u64 mask)
Checks to see if the mask is possible and updates the device Checks to see if the mask is possible and updates the device
streaming and coherent DMA mask parameters if it is. streaming and coherent DMA mask parameters if it is.
Returns: 0 if successful and a negative error if not. Returns: 0 if successful and a negative error if not.
int ::
dma_set_mask(struct device *dev, u64 mask)
int
dma_set_mask(struct device *dev, u64 mask)
Checks to see if the mask is possible and updates the device Checks to see if the mask is possible and updates the device
parameters if it is. parameters if it is.
Returns: 0 if successful and a negative error if not. Returns: 0 if successful and a negative error if not.
int ::
dma_set_coherent_mask(struct device *dev, u64 mask)
int
dma_set_coherent_mask(struct device *dev, u64 mask)
Checks to see if the mask is possible and updates the device Checks to see if the mask is possible and updates the device
parameters if it is. parameters if it is.
Returns: 0 if successful and a negative error if not. Returns: 0 if successful and a negative error if not.
u64 ::
dma_get_required_mask(struct device *dev)
u64
dma_get_required_mask(struct device *dev)
This API returns the mask that the platform requires to This API returns the mask that the platform requires to
operate efficiently. Usually this means the returned mask operate efficiently. Usually this means the returned mask
@ -182,94 +208,107 @@ call to set the mask to the value returned.
Part Id - Streaming DMA mappings Part Id - Streaming DMA mappings
-------------------------------- --------------------------------
dma_addr_t ::
dma_map_single(struct device *dev, void *cpu_addr, size_t size,
enum dma_data_direction direction) dma_addr_t
dma_map_single(struct device *dev, void *cpu_addr, size_t size,
enum dma_data_direction direction)
Maps a piece of processor virtual memory so it can be accessed by the Maps a piece of processor virtual memory so it can be accessed by the
device and returns the DMA address of the memory. device and returns the DMA address of the memory.
The direction for both APIs may be converted freely by casting. The direction for both APIs may be converted freely by casting.
However the dma_ API uses a strongly typed enumerator for its However the dma_API uses a strongly typed enumerator for its
direction: direction:
======================= =============================================
DMA_NONE no direction (used for debugging) DMA_NONE no direction (used for debugging)
DMA_TO_DEVICE data is going from the memory to the device DMA_TO_DEVICE data is going from the memory to the device
DMA_FROM_DEVICE data is coming from the device to the memory DMA_FROM_DEVICE data is coming from the device to the memory
DMA_BIDIRECTIONAL direction isn't known DMA_BIDIRECTIONAL direction isn't known
======================= =============================================
Notes: Not all memory regions in a machine can be mapped by this API. .. note::
Further, contiguous kernel virtual space may not be contiguous as
physical memory. Since this API does not provide any scatter/gather
capability, it will fail if the user tries to map a non-physically
contiguous piece of memory. For this reason, memory to be mapped by
this API should be obtained from sources which guarantee it to be
physically contiguous (like kmalloc).
Further, the DMA address of the memory must be within the Not all memory regions in a machine can be mapped by this API.
dma_mask of the device (the dma_mask is a bit mask of the Further, contiguous kernel virtual space may not be contiguous as
addressable region for the device, i.e., if the DMA address of physical memory. Since this API does not provide any scatter/gather
the memory ANDed with the dma_mask is still equal to the DMA capability, it will fail if the user tries to map a non-physically
address, then the device can perform DMA to the memory). To contiguous piece of memory. For this reason, memory to be mapped by
ensure that the memory allocated by kmalloc is within the dma_mask, this API should be obtained from sources which guarantee it to be
the driver may specify various platform-dependent flags to restrict physically contiguous (like kmalloc).
the DMA address range of the allocation (e.g., on x86, GFP_DMA
guarantees to be within the first 16MB of available DMA addresses,
as required by ISA devices).
Note also that the above constraints on physical contiguity and Further, the DMA address of the memory must be within the
dma_mask may not apply if the platform has an IOMMU (a device which dma_mask of the device (the dma_mask is a bit mask of the
maps an I/O DMA address to a physical memory address). However, to be addressable region for the device, i.e., if the DMA address of
portable, device driver writers may *not* assume that such an IOMMU the memory ANDed with the dma_mask is still equal to the DMA
exists. address, then the device can perform DMA to the memory). To
ensure that the memory allocated by kmalloc is within the dma_mask,
the driver may specify various platform-dependent flags to restrict
the DMA address range of the allocation (e.g., on x86, GFP_DMA
guarantees to be within the first 16MB of available DMA addresses,
as required by ISA devices).
Warnings: Memory coherency operates at a granularity called the cache Note also that the above constraints on physical contiguity and
line width. In order for memory mapped by this API to operate dma_mask may not apply if the platform has an IOMMU (a device which
correctly, the mapped region must begin exactly on a cache line maps an I/O DMA address to a physical memory address). However, to be
boundary and end exactly on one (to prevent two separately mapped portable, device driver writers may *not* assume that such an IOMMU
regions from sharing a single cache line). Since the cache line size exists.
may not be known at compile time, the API will not enforce this
requirement. Therefore, it is recommended that driver writers who
don't take special care to determine the cache line size at run time
only map virtual regions that begin and end on page boundaries (which
are guaranteed also to be cache line boundaries).
DMA_TO_DEVICE synchronisation must be done after the last modification .. warning::
of the memory region by the software and before it is handed off to
the device. Once this primitive is used, memory covered by this
primitive should be treated as read-only by the device. If the device
may write to it at any point, it should be DMA_BIDIRECTIONAL (see
below).
DMA_FROM_DEVICE synchronisation must be done before the driver Memory coherency operates at a granularity called the cache
accesses data that may be changed by the device. This memory should line width. In order for memory mapped by this API to operate
be treated as read-only by the driver. If the driver needs to write correctly, the mapped region must begin exactly on a cache line
to it at any point, it should be DMA_BIDIRECTIONAL (see below). boundary and end exactly on one (to prevent two separately mapped
regions from sharing a single cache line). Since the cache line size
may not be known at compile time, the API will not enforce this
requirement. Therefore, it is recommended that driver writers who
don't take special care to determine the cache line size at run time
only map virtual regions that begin and end on page boundaries (which
are guaranteed also to be cache line boundaries).
DMA_BIDIRECTIONAL requires special handling: it means that the driver DMA_TO_DEVICE synchronisation must be done after the last modification
isn't sure if the memory was modified before being handed off to the of the memory region by the software and before it is handed off to
device and also isn't sure if the device will also modify it. Thus, the device. Once this primitive is used, memory covered by this
you must always sync bidirectional memory twice: once before the primitive should be treated as read-only by the device. If the device
memory is handed off to the device (to make sure all memory changes may write to it at any point, it should be DMA_BIDIRECTIONAL (see
are flushed from the processor) and once before the data may be below).
accessed after being used by the device (to make sure any processor
cache lines are updated with data that the device may have changed).
void DMA_FROM_DEVICE synchronisation must be done before the driver
dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size, accesses data that may be changed by the device. This memory should
enum dma_data_direction direction) be treated as read-only by the driver. If the driver needs to write
to it at any point, it should be DMA_BIDIRECTIONAL (see below).
DMA_BIDIRECTIONAL requires special handling: it means that the driver
isn't sure if the memory was modified before being handed off to the
device and also isn't sure if the device will also modify it. Thus,
you must always sync bidirectional memory twice: once before the
memory is handed off to the device (to make sure all memory changes
are flushed from the processor) and once before the data may be
accessed after being used by the device (to make sure any processor
cache lines are updated with data that the device may have changed).
::
void
dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
enum dma_data_direction direction)
Unmaps the region previously mapped. All the parameters passed in Unmaps the region previously mapped. All the parameters passed in
must be identical to those passed in (and returned) by the mapping must be identical to those passed in (and returned) by the mapping
API. API.
dma_addr_t ::
dma_map_page(struct device *dev, struct page *page,
unsigned long offset, size_t size, dma_addr_t
enum dma_data_direction direction) dma_map_page(struct device *dev, struct page *page,
void unsigned long offset, size_t size,
dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size, enum dma_data_direction direction)
enum dma_data_direction direction)
void
dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size,
enum dma_data_direction direction)
API for mapping and unmapping for pages. All the notes and warnings API for mapping and unmapping for pages. All the notes and warnings
for the other mapping APIs apply here. Also, although the <offset> for the other mapping APIs apply here. Also, although the <offset>
@ -277,20 +316,24 @@ and <size> parameters are provided to do partial page mapping, it is
recommended that you never use these unless you really know what the recommended that you never use these unless you really know what the
cache width is. cache width is.
dma_addr_t ::
dma_map_resource(struct device *dev, phys_addr_t phys_addr, size_t size,
enum dma_data_direction dir, unsigned long attrs)
void dma_addr_t
dma_unmap_resource(struct device *dev, dma_addr_t addr, size_t size, dma_map_resource(struct device *dev, phys_addr_t phys_addr, size_t size,
enum dma_data_direction dir, unsigned long attrs) enum dma_data_direction dir, unsigned long attrs)
void
dma_unmap_resource(struct device *dev, dma_addr_t addr, size_t size,
enum dma_data_direction dir, unsigned long attrs)
API for mapping and unmapping for MMIO resources. All the notes and API for mapping and unmapping for MMIO resources. All the notes and
warnings for the other mapping APIs apply here. The API should only be warnings for the other mapping APIs apply here. The API should only be
used to map device MMIO resources, mapping of RAM is not permitted. used to map device MMIO resources, mapping of RAM is not permitted.
int ::
dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
int
dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
In some circumstances dma_map_single(), dma_map_page() and dma_map_resource() In some circumstances dma_map_single(), dma_map_page() and dma_map_resource()
will fail to create a mapping. A driver can check for these errors by testing will fail to create a mapping. A driver can check for these errors by testing
@ -298,9 +341,11 @@ the returned DMA address with dma_mapping_error(). A non-zero return value
means the mapping could not be created and the driver should take appropriate means the mapping could not be created and the driver should take appropriate
action (e.g. reduce current DMA mapping usage or delay and try again later). action (e.g. reduce current DMA mapping usage or delay and try again later).
::
int int
dma_map_sg(struct device *dev, struct scatterlist *sg, dma_map_sg(struct device *dev, struct scatterlist *sg,
int nents, enum dma_data_direction direction) int nents, enum dma_data_direction direction)
Returns: the number of DMA address segments mapped (this may be shorter Returns: the number of DMA address segments mapped (this may be shorter
than <nents> passed in if some elements of the scatter/gather list are than <nents> passed in if some elements of the scatter/gather list are
@ -316,7 +361,7 @@ critical that the driver do something, in the case of a block driver
aborting the request or even oopsing is better than doing nothing and aborting the request or even oopsing is better than doing nothing and
corrupting the filesystem. corrupting the filesystem.
With scatterlists, you use the resulting mapping like this: With scatterlists, you use the resulting mapping like this::
int i, count = dma_map_sg(dev, sglist, nents, direction); int i, count = dma_map_sg(dev, sglist, nents, direction);
struct scatterlist *sg; struct scatterlist *sg;
@ -337,9 +382,11 @@ Then you should loop count times (note: this can be less than nents times)
and use sg_dma_address() and sg_dma_len() macros where you previously and use sg_dma_address() and sg_dma_len() macros where you previously
accessed sg->address and sg->length as shown above. accessed sg->address and sg->length as shown above.
::
void void
dma_unmap_sg(struct device *dev, struct scatterlist *sg, dma_unmap_sg(struct device *dev, struct scatterlist *sg,
int nents, enum dma_data_direction direction) int nents, enum dma_data_direction direction)
Unmap the previously mapped scatter/gather list. All the parameters Unmap the previously mapped scatter/gather list. All the parameters
must be the same as those and passed in to the scatter/gather mapping must be the same as those and passed in to the scatter/gather mapping
@ -348,18 +395,27 @@ API.
Note: <nents> must be the number you passed in, *not* the number of Note: <nents> must be the number you passed in, *not* the number of
DMA address entries returned. DMA address entries returned.
void ::
dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
enum dma_data_direction direction) void
void dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle,
dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size, size_t size,
enum dma_data_direction direction) enum dma_data_direction direction)
void
dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int nents, void
enum dma_data_direction direction) dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle,
void size_t size,
dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int nents, enum dma_data_direction direction)
enum dma_data_direction direction)
void
dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg,
int nents,
enum dma_data_direction direction)
void
dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg,
int nents,
enum dma_data_direction direction)
Synchronise a single contiguous or scatter/gather mapping for the CPU Synchronise a single contiguous or scatter/gather mapping for the CPU
and device. With the sync_sg API, all the parameters must be the same and device. With the sync_sg API, all the parameters must be the same
@ -367,36 +423,41 @@ as those passed into the single mapping API. With the sync_single API,
you can use dma_handle and size parameters that aren't identical to you can use dma_handle and size parameters that aren't identical to
those passed into the single mapping API to do a partial sync. those passed into the single mapping API to do a partial sync.
Notes: You must do this:
- Before reading values that have been written by DMA from the device .. note::
(use the DMA_FROM_DEVICE direction)
- After writing values that will be written to the device using DMA You must do this:
(use the DMA_TO_DEVICE) direction
- before *and* after handing memory to the device if the memory is - Before reading values that have been written by DMA from the device
DMA_BIDIRECTIONAL (use the DMA_FROM_DEVICE direction)
- After writing values that will be written to the device using DMA
(use the DMA_TO_DEVICE) direction
- before *and* after handing memory to the device if the memory is
DMA_BIDIRECTIONAL
See also dma_map_single(). See also dma_map_single().
dma_addr_t ::
dma_map_single_attrs(struct device *dev, void *cpu_addr, size_t size,
enum dma_data_direction dir,
unsigned long attrs)
void dma_addr_t
dma_unmap_single_attrs(struct device *dev, dma_addr_t dma_addr, dma_map_single_attrs(struct device *dev, void *cpu_addr, size_t size,
size_t size, enum dma_data_direction dir, enum dma_data_direction dir,
unsigned long attrs) unsigned long attrs)
int void
dma_map_sg_attrs(struct device *dev, struct scatterlist *sgl, dma_unmap_single_attrs(struct device *dev, dma_addr_t dma_addr,
int nents, enum dma_data_direction dir, size_t size, enum dma_data_direction dir,
unsigned long attrs) unsigned long attrs)
void int
dma_unmap_sg_attrs(struct device *dev, struct scatterlist *sgl, dma_map_sg_attrs(struct device *dev, struct scatterlist *sgl,
int nents, enum dma_data_direction dir, int nents, enum dma_data_direction dir,
unsigned long attrs) unsigned long attrs)
void
dma_unmap_sg_attrs(struct device *dev, struct scatterlist *sgl,
int nents, enum dma_data_direction dir,
unsigned long attrs)
The four functions above are just like the counterpart functions The four functions above are just like the counterpart functions
without the _attrs suffixes, except that they pass an optional without the _attrs suffixes, except that they pass an optional
@ -410,37 +471,38 @@ is identical to those of the corresponding function
without the _attrs suffix. As a result dma_map_single_attrs() without the _attrs suffix. As a result dma_map_single_attrs()
can generally replace dma_map_single(), etc. can generally replace dma_map_single(), etc.
As an example of the use of the *_attrs functions, here's how As an example of the use of the ``*_attrs`` functions, here's how
you could pass an attribute DMA_ATTR_FOO when mapping memory you could pass an attribute DMA_ATTR_FOO when mapping memory
for DMA: for DMA::
#include <linux/dma-mapping.h> #include <linux/dma-mapping.h>
/* DMA_ATTR_FOO should be defined in linux/dma-mapping.h and /* DMA_ATTR_FOO should be defined in linux/dma-mapping.h and
* documented in Documentation/DMA-attributes.txt */ * documented in Documentation/DMA-attributes.txt */
... ...
unsigned long attr; unsigned long attr;
attr |= DMA_ATTR_FOO; attr |= DMA_ATTR_FOO;
.... ....
n = dma_map_sg_attrs(dev, sg, nents, DMA_TO_DEVICE, attr); n = dma_map_sg_attrs(dev, sg, nents, DMA_TO_DEVICE, attr);
.... ....
Architectures that care about DMA_ATTR_FOO would check for its Architectures that care about DMA_ATTR_FOO would check for its
presence in their implementations of the mapping and unmapping presence in their implementations of the mapping and unmapping
routines, e.g.: routines, e.g.:::
void whizco_dma_map_sg_attrs(struct device *dev, dma_addr_t dma_addr, void whizco_dma_map_sg_attrs(struct device *dev, dma_addr_t dma_addr,
size_t size, enum dma_data_direction dir, size_t size, enum dma_data_direction dir,
unsigned long attrs) unsigned long attrs)
{ {
.... ....
if (attrs & DMA_ATTR_FOO) if (attrs & DMA_ATTR_FOO)
/* twizzle the frobnozzle */ /* twizzle the frobnozzle */
.... ....
}
Part II - Advanced dma_ usage Part II - Advanced dma usage
----------------------------- ----------------------------
Warning: These pieces of the DMA API should not be used in the Warning: These pieces of the DMA API should not be used in the
majority of cases, since they cater for unlikely corner cases that majority of cases, since they cater for unlikely corner cases that
@ -450,9 +512,11 @@ If you don't understand how cache line coherency works between a
processor and an I/O device, you should not be using this part of the processor and an I/O device, you should not be using this part of the
API at all. API at all.
void * ::
dma_alloc_noncoherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag) void *
dma_alloc_noncoherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag)
Identical to dma_alloc_coherent() except that the platform will Identical to dma_alloc_coherent() except that the platform will
choose to return either consistent or non-consistent memory as it sees choose to return either consistent or non-consistent memory as it sees
@ -468,39 +532,49 @@ only use this API if you positively know your driver will be
required to work on one of the rare (usually non-PCI) architectures required to work on one of the rare (usually non-PCI) architectures
that simply cannot make consistent memory. that simply cannot make consistent memory.
void ::
dma_free_noncoherent(struct device *dev, size_t size, void *cpu_addr,
dma_addr_t dma_handle) void
dma_free_noncoherent(struct device *dev, size_t size, void *cpu_addr,
dma_addr_t dma_handle)
Free memory allocated by the nonconsistent API. All parameters must Free memory allocated by the nonconsistent API. All parameters must
be identical to those passed in (and returned by be identical to those passed in (and returned by
dma_alloc_noncoherent()). dma_alloc_noncoherent()).
int ::
dma_get_cache_alignment(void)
int
dma_get_cache_alignment(void)
Returns the processor cache alignment. This is the absolute minimum Returns the processor cache alignment. This is the absolute minimum
alignment *and* width that you must observe when either mapping alignment *and* width that you must observe when either mapping
memory or doing partial flushes. memory or doing partial flushes.
Notes: This API may return a number *larger* than the actual cache .. note::
line, but it will guarantee that one or more cache lines fit exactly
into the width returned by this call. It will also always be a power
of two for easy alignment.
void This API may return a number *larger* than the actual cache
dma_cache_sync(struct device *dev, void *vaddr, size_t size, line, but it will guarantee that one or more cache lines fit exactly
enum dma_data_direction direction) into the width returned by this call. It will also always be a power
of two for easy alignment.
::
void
dma_cache_sync(struct device *dev, void *vaddr, size_t size,
enum dma_data_direction direction)
Do a partial sync of memory that was allocated by Do a partial sync of memory that was allocated by
dma_alloc_noncoherent(), starting at virtual address vaddr and dma_alloc_noncoherent(), starting at virtual address vaddr and
continuing on for size. Again, you *must* observe the cache line continuing on for size. Again, you *must* observe the cache line
boundaries when doing this. boundaries when doing this.
int ::
dma_declare_coherent_memory(struct device *dev, phys_addr_t phys_addr,
dma_addr_t device_addr, size_t size, int int
flags) dma_declare_coherent_memory(struct device *dev, phys_addr_t phys_addr,
dma_addr_t device_addr, size_t size, int
flags)
Declare region of memory to be handed out by dma_alloc_coherent() when Declare region of memory to be handed out by dma_alloc_coherent() when
it's asked for coherent memory for this device. it's asked for coherent memory for this device.
@ -516,21 +590,21 @@ size is the size of the area (must be multiples of PAGE_SIZE).
flags can be ORed together and are: flags can be ORed together and are:
DMA_MEMORY_MAP - request that the memory returned from - DMA_MEMORY_MAP - request that the memory returned from
dma_alloc_coherent() be directly writable. dma_alloc_coherent() be directly writable.
DMA_MEMORY_IO - request that the memory returned from - DMA_MEMORY_IO - request that the memory returned from
dma_alloc_coherent() be addressable using read()/write()/memcpy_toio() etc. dma_alloc_coherent() be addressable using read()/write()/memcpy_toio() etc.
One or both of these flags must be present. One or both of these flags must be present.
DMA_MEMORY_INCLUDES_CHILDREN - make the declared memory be allocated by - DMA_MEMORY_INCLUDES_CHILDREN - make the declared memory be allocated by
dma_alloc_coherent of any child devices of this one (for memory residing dma_alloc_coherent of any child devices of this one (for memory residing
on a bridge). on a bridge).
DMA_MEMORY_EXCLUSIVE - only allocate memory from the declared regions. - DMA_MEMORY_EXCLUSIVE - only allocate memory from the declared regions.
Do not allow dma_alloc_coherent() to fall back to system memory when Do not allow dma_alloc_coherent() to fall back to system memory when
it's out of memory in the declared region. it's out of memory in the declared region.
The return value will be either DMA_MEMORY_MAP or DMA_MEMORY_IO and The return value will be either DMA_MEMORY_MAP or DMA_MEMORY_IO and
must correspond to a passed in flag (i.e. no returning DMA_MEMORY_IO must correspond to a passed in flag (i.e. no returning DMA_MEMORY_IO
@ -543,15 +617,17 @@ must be accessed using the correct bus functions. If your driver
isn't prepared to handle this contingency, it should not specify isn't prepared to handle this contingency, it should not specify
DMA_MEMORY_IO in the input flags. DMA_MEMORY_IO in the input flags.
As a simplification for the platforms, only *one* such region of As a simplification for the platforms, only **one** such region of
memory may be declared per device. memory may be declared per device.
For reasons of efficiency, most platforms choose to track the declared For reasons of efficiency, most platforms choose to track the declared
region only at the granularity of a page. For smaller allocations, region only at the granularity of a page. For smaller allocations,
you should use the dma_pool() API. you should use the dma_pool() API.
void ::
dma_release_declared_memory(struct device *dev)
void
dma_release_declared_memory(struct device *dev)
Remove the memory region previously declared from the system. This Remove the memory region previously declared from the system. This
API performs *no* in-use checking for this region and will return API performs *no* in-use checking for this region and will return
@ -559,9 +635,11 @@ unconditionally having removed all the required structures. It is the
driver's job to ensure that no parts of this memory region are driver's job to ensure that no parts of this memory region are
currently in use. currently in use.
void * ::
dma_mark_declared_memory_occupied(struct device *dev,
dma_addr_t device_addr, size_t size) void *
dma_mark_declared_memory_occupied(struct device *dev,
dma_addr_t device_addr, size_t size)
This is used to occupy specific regions of the declared space This is used to occupy specific regions of the declared space
(dma_alloc_coherent() will hand out the first free region it finds). (dma_alloc_coherent() will hand out the first free region it finds).
@ -592,38 +670,37 @@ option has a performance impact. Do not enable it in production kernels.
If you boot the resulting kernel will contain code which does some bookkeeping If you boot the resulting kernel will contain code which does some bookkeeping
about what DMA memory was allocated for which device. If this code detects an about what DMA memory was allocated for which device. If this code detects an
error it prints a warning message with some details into your kernel log. An error it prints a warning message with some details into your kernel log. An
example warning message may look like this: example warning message may look like this::
------------[ cut here ]------------ WARNING: at /data2/repos/linux-2.6-iommu/lib/dma-debug.c:448
WARNING: at /data2/repos/linux-2.6-iommu/lib/dma-debug.c:448 check_unmap+0x203/0x490()
check_unmap+0x203/0x490() Hardware name:
Hardware name: forcedeth 0000:00:08.0: DMA-API: device driver frees DMA memory with wrong
forcedeth 0000:00:08.0: DMA-API: device driver frees DMA memory with wrong function [device address=0x00000000640444be] [size=66 bytes] [mapped as
function [device address=0x00000000640444be] [size=66 bytes] [mapped as single] [unmapped as page]
single] [unmapped as page] Modules linked in: nfsd exportfs bridge stp llc r8169
Modules linked in: nfsd exportfs bridge stp llc r8169 Pid: 0, comm: swapper Tainted: G W 2.6.28-dmatest-09289-g8bb99c0 #1
Pid: 0, comm: swapper Tainted: G W 2.6.28-dmatest-09289-g8bb99c0 #1 Call Trace:
Call Trace: <IRQ> [<ffffffff80240b22>] warn_slowpath+0xf2/0x130
<IRQ> [<ffffffff80240b22>] warn_slowpath+0xf2/0x130 [<ffffffff80647b70>] _spin_unlock+0x10/0x30
[<ffffffff80647b70>] _spin_unlock+0x10/0x30 [<ffffffff80537e75>] usb_hcd_link_urb_to_ep+0x75/0xc0
[<ffffffff80537e75>] usb_hcd_link_urb_to_ep+0x75/0xc0 [<ffffffff80647c22>] _spin_unlock_irqrestore+0x12/0x40
[<ffffffff80647c22>] _spin_unlock_irqrestore+0x12/0x40 [<ffffffff8055347f>] ohci_urb_enqueue+0x19f/0x7c0
[<ffffffff8055347f>] ohci_urb_enqueue+0x19f/0x7c0 [<ffffffff80252f96>] queue_work+0x56/0x60
[<ffffffff80252f96>] queue_work+0x56/0x60 [<ffffffff80237e10>] enqueue_task_fair+0x20/0x50
[<ffffffff80237e10>] enqueue_task_fair+0x20/0x50 [<ffffffff80539279>] usb_hcd_submit_urb+0x379/0xbc0
[<ffffffff80539279>] usb_hcd_submit_urb+0x379/0xbc0 [<ffffffff803b78c3>] cpumask_next_and+0x23/0x40
[<ffffffff803b78c3>] cpumask_next_and+0x23/0x40 [<ffffffff80235177>] find_busiest_group+0x207/0x8a0
[<ffffffff80235177>] find_busiest_group+0x207/0x8a0 [<ffffffff8064784f>] _spin_lock_irqsave+0x1f/0x50
[<ffffffff8064784f>] _spin_lock_irqsave+0x1f/0x50 [<ffffffff803c7ea3>] check_unmap+0x203/0x490
[<ffffffff803c7ea3>] check_unmap+0x203/0x490 [<ffffffff803c8259>] debug_dma_unmap_page+0x49/0x50
[<ffffffff803c8259>] debug_dma_unmap_page+0x49/0x50 [<ffffffff80485f26>] nv_tx_done_optimized+0xc6/0x2c0
[<ffffffff80485f26>] nv_tx_done_optimized+0xc6/0x2c0 [<ffffffff80486c13>] nv_nic_irq_optimized+0x73/0x2b0
[<ffffffff80486c13>] nv_nic_irq_optimized+0x73/0x2b0 [<ffffffff8026df84>] handle_IRQ_event+0x34/0x70
[<ffffffff8026df84>] handle_IRQ_event+0x34/0x70 [<ffffffff8026ffe9>] handle_edge_irq+0xc9/0x150
[<ffffffff8026ffe9>] handle_edge_irq+0xc9/0x150 [<ffffffff8020e3ab>] do_IRQ+0xcb/0x1c0
[<ffffffff8020e3ab>] do_IRQ+0xcb/0x1c0 [<ffffffff8020c093>] ret_from_intr+0x0/0xa
[<ffffffff8020c093>] ret_from_intr+0x0/0xa <EOI> <4>---[ end trace f6435a98e2a38c0e ]---
<EOI> <4>---[ end trace f6435a98e2a38c0e ]---
The driver developer can find the driver and the device including a stacktrace The driver developer can find the driver and the device including a stacktrace
of the DMA-API call which caused this warning. of the DMA-API call which caused this warning.
@ -637,43 +714,42 @@ details.
The debugfs directory for the DMA-API debugging code is called dma-api/. In The debugfs directory for the DMA-API debugging code is called dma-api/. In
this directory the following files can currently be found: this directory the following files can currently be found:
dma-api/all_errors This file contains a numeric value. If this =============================== ===============================================
dma-api/all_errors This file contains a numeric value. If this
value is not equal to zero the debugging code value is not equal to zero the debugging code
will print a warning for every error it finds will print a warning for every error it finds
into the kernel log. Be careful with this into the kernel log. Be careful with this
option, as it can easily flood your logs. option, as it can easily flood your logs.
dma-api/disabled This read-only file contains the character 'Y' dma-api/disabled This read-only file contains the character 'Y'
if the debugging code is disabled. This can if the debugging code is disabled. This can
happen when it runs out of memory or if it was happen when it runs out of memory or if it was
disabled at boot time disabled at boot time
dma-api/error_count This file is read-only and shows the total dma-api/error_count This file is read-only and shows the total
numbers of errors found. numbers of errors found.
dma-api/num_errors The number in this file shows how many dma-api/num_errors The number in this file shows how many
warnings will be printed to the kernel log warnings will be printed to the kernel log
before it stops. This number is initialized to before it stops. This number is initialized to
one at system boot and be set by writing into one at system boot and be set by writing into
this file this file
dma-api/min_free_entries dma-api/min_free_entries This read-only file can be read to get the
This read-only file can be read to get the
minimum number of free dma_debug_entries the minimum number of free dma_debug_entries the
allocator has ever seen. If this value goes allocator has ever seen. If this value goes
down to zero the code will disable itself down to zero the code will disable itself
because it is not longer reliable. because it is not longer reliable.
dma-api/num_free_entries dma-api/num_free_entries The current number of free dma_debug_entries
The current number of free dma_debug_entries
in the allocator. in the allocator.
dma-api/driver-filter dma-api/driver-filter You can write a name of a driver into this file
You can write a name of a driver into this file
to limit the debug output to requests from that to limit the debug output to requests from that
particular driver. Write an empty string to particular driver. Write an empty string to
that file to disable the filter and see that file to disable the filter and see
all errors again. all errors again.
=============================== ===============================================
If you have this code compiled into your kernel it will be enabled by default. If you have this code compiled into your kernel it will be enabled by default.
If you want to boot without the bookkeeping anyway you can provide If you want to boot without the bookkeeping anyway you can provide
@ -692,7 +768,10 @@ of preallocated entries is defined per architecture. If it is too low for you
boot with 'dma_debug_entries=<your_desired_number>' to overwrite the boot with 'dma_debug_entries=<your_desired_number>' to overwrite the
architectural default. architectural default.
void debug_dma_mapping_error(struct device *dev, dma_addr_t dma_addr); ::
void
debug_dma_mapping_error(struct device *dev, dma_addr_t dma_addr);
dma-debug interface debug_dma_mapping_error() to debug drivers that fail dma-debug interface debug_dma_mapping_error() to debug drivers that fail
to check DMA mapping errors on addresses returned by dma_map_single() and to check DMA mapping errors on addresses returned by dma_map_single() and
@ -702,4 +781,3 @@ the driver. When driver does unmap, debug_dma_unmap() checks the flag and if
this flag is still set, prints warning message that includes call trace that this flag is still set, prints warning message that includes call trace that
leads up to the unmap. This interface can be called from dma_mapping_error() leads up to the unmap. This interface can be called from dma_mapping_error()
routines to enable DMA mapping error check debugging. routines to enable DMA mapping error check debugging.