Blame view
Documentation/DMA-API-HOWTO.txt
31.6 KB
216bf58f4 Documentation: co... |
1 2 |
Dynamic DMA mapping Guide ========================= |
1da177e4c Linux-2.6.12-rc2 |
3 4 5 6 |
David S. Miller <davem@redhat.com> Richard Henderson <rth@cygnus.com> Jakub Jelinek <jakub@redhat.com> |
216bf58f4 Documentation: co... |
7 8 |
This is a guide to device driver writers on how to use the DMA API with example pseudo-code. For a concise description of the API, see |
1da177e4c Linux-2.6.12-rc2 |
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 |
DMA-API.txt. Most of the 64bit platforms have special hardware that translates bus addresses (DMA addresses) into physical addresses. This is similar to how page tables and/or a TLB translates virtual addresses to physical addresses on a CPU. This is needed so that e.g. PCI devices can access with a Single Address Cycle (32bit DMA address) any page in the 64bit physical address space. Previously in Linux those 64bit platforms had to set artificial limits on the maximum RAM size in the system, so that the virt_to_bus() static scheme works (the DMA address translation tables were simply filled on bootup to map each bus address to the physical page __pa(bus_to_virt())). So that Linux can use the dynamic DMA mapping, it needs some help from the drivers, namely it has to take into account that DMA addresses should be mapped only for the time they are actually used and unmapped after the DMA transfer. The following API will work of course even on platforms where no such |
216bf58f4 Documentation: co... |
28 29 30 31 32 |
hardware exists. Note that the DMA API works with any bus independent of the underlying microprocessor architecture. You should use the DMA API rather than the bus specific DMA API (e.g. pci_dma_*). |
1da177e4c Linux-2.6.12-rc2 |
33 34 |
First of all, you should make sure |
216bf58f4 Documentation: co... |
35 |
#include <linux/dma-mapping.h> |
1da177e4c Linux-2.6.12-rc2 |
36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 |
is in your driver. This file will obtain for you the definition of the dma_addr_t (which can hold any valid DMA address for the platform) type which should be used everywhere you hold a DMA (bus) address returned from the DMA mapping functions. What memory is DMA'able? The first piece of information you must know is what kernel memory can be used with the DMA mapping facilities. There has been an unwritten set of rules regarding this, and this text is an attempt to finally write them down. If you acquired your memory via the page allocator (i.e. __get_free_page*()) or the generic memory allocators (i.e. kmalloc() or kmem_cache_alloc()) then you may DMA to/from that memory using the addresses returned from those routines. This means specifically that you may _not_ use the memory/addresses returned from vmalloc() for DMA. It is possible to DMA to the _underlying_ memory mapped into a vmalloc() area, but this requires walking page tables to get the physical addresses, and then translating each of those pages back to a kernel address using something like __va(). [ EDIT: Update this when we integrate Gerd Knorr's generic code which does this. ] |
21440d313 [PATCH] dma doc u... |
61 62 63 64 65 66 67 68 69 |
This rule also means that you may use neither kernel image addresses (items in data/text/bss segments), nor module image addresses, nor stack addresses for DMA. These could all be mapped somewhere entirely different than the rest of physical memory. Even if those classes of memory could physically work with DMA, you'd need to ensure the I/O buffers were cacheline-aligned. Without that, you'd see cacheline sharing problems (data corruption) on CPUs with DMA-incoherent caches. (The CPU could write to one word, DMA would write to a different one in the same cache line, and one of them could be overwritten.) |
1da177e4c Linux-2.6.12-rc2 |
70 71 72 73 74 75 76 77 78 79 80 |
Also, this means that you cannot take the return of a kmap() call and DMA to/from that. This is similar to vmalloc(). What about block I/O and networking buffers? The block I/O and networking subsystems make sure that the buffers they use are valid for you to DMA from/to. DMA addressing limitations Does your device have any DMA addressing limitations? For example, is |
216bf58f4 Documentation: co... |
81 82 |
your device only capable of driving the low order 24-bits of address? If so, you need to inform the kernel of this fact. |
1da177e4c Linux-2.6.12-rc2 |
83 84 |
By default, the kernel assumes that your device can address the full |
216bf58f4 Documentation: co... |
85 86 87 88 89 90 91 92 93 94 95 96 97 |
32-bits. For a 64-bit capable device, this needs to be increased. And for a device with limitations, as discussed in the previous paragraph, it needs to be decreased. Special note about PCI: PCI-X specification requires PCI-X devices to support 64-bit addressing (DAC) for all transactions. And at least one platform (SGI SN2) requires 64-bit consistent allocations to operate correctly when the IO bus is in PCI-X mode. For correct operation, you must interrogate the kernel in your device probe routine to see if the DMA controller on the machine can properly support the DMA addressing limitation your device has. It is good style to do this even if your device holds the default setting, |
1da177e4c Linux-2.6.12-rc2 |
98 99 |
because this shows that you did think about these issues wrt. your device. |
4aa806b77 DMA-API: provide ... |
100 |
The query is performed via a call to dma_set_mask_and_coherent(): |
1da177e4c Linux-2.6.12-rc2 |
101 |
|
4aa806b77 DMA-API: provide ... |
102 |
int dma_set_mask_and_coherent(struct device *dev, u64 mask); |
1da177e4c Linux-2.6.12-rc2 |
103 |
|
4aa806b77 DMA-API: provide ... |
104 105 106 |
which will query the mask for both streaming and coherent APIs together. If you have some special requirements, then the following two separate queries can be used instead: |
1da177e4c Linux-2.6.12-rc2 |
107 |
|
4aa806b77 DMA-API: provide ... |
108 109 110 111 112 113 114 115 116 |
The query for streaming mappings is performed via a call to dma_set_mask(): int dma_set_mask(struct device *dev, u64 mask); The query for consistent allocations is performed via a call to dma_set_coherent_mask(): int dma_set_coherent_mask(struct device *dev, u64 mask); |
1da177e4c Linux-2.6.12-rc2 |
117 |
|
216bf58f4 Documentation: co... |
118 119 120 121 122 123 124 125 |
Here, dev is a pointer to the device struct of your device, and mask is a bit mask describing which bits of an address your device supports. It returns zero if your card can perform DMA properly on the machine given the address mask you provided. In general, the device struct of your device is embedded in the bus specific device struct of your device. For example, a pointer to the device struct of your PCI device is pdev->dev (pdev is a pointer to the PCI device struct of your device). |
1da177e4c Linux-2.6.12-rc2 |
126 |
|
84eb8d060 Fix "can not" in ... |
127 |
If it returns non-zero, your device cannot perform DMA properly on |
1da177e4c Linux-2.6.12-rc2 |
128 129 130 131 132 133 134 135 136 137 138 139 140 141 |
this platform, and attempting to do so will result in undefined behavior. You must either use a different mask, or not use DMA. This means that in the failure case, you have three options: 1) Use another DMA mask, if possible (see below). 2) Use some non-DMA mode for data transfer, if possible. 3) Ignore this device and do not initialize it. It is recommended that your driver print a kernel KERN_WARNING message when you end up performing either #2 or #3. In this manner, if a user of your driver reports that performance is bad or that the device is not even detected, you can ask them for the kernel messages to find out exactly why. |
216bf58f4 Documentation: co... |
142 |
The standard 32-bit addressing device would do something like this: |
1da177e4c Linux-2.6.12-rc2 |
143 |
|
4aa806b77 DMA-API: provide ... |
144 |
if (dma_set_mask_and_coherent(dev, DMA_BIT_MASK(32))) { |
1da177e4c Linux-2.6.12-rc2 |
145 146 147 148 149 |
printk(KERN_WARNING "mydev: No suitable DMA available. "); goto ignore_this_device; } |
216bf58f4 Documentation: co... |
150 151 152 153 154 155 156 |
Another common scenario is a 64-bit capable device. The approach here is to try for 64-bit addressing, but back down to a 32-bit mask that should not fail. The kernel may fail the 64-bit mask not because the platform is not capable of 64-bit addressing. Rather, it may fail in this case simply because 32-bit addressing is done more efficiently than 64-bit addressing. For example, Sparc64 PCI SAC addressing is more efficient than DAC addressing. |
1da177e4c Linux-2.6.12-rc2 |
157 158 159 160 161 |
Here is how you would handle a 64-bit capable device which can drive all 64-bits when accessing streaming DMA: int using_dac; |
216bf58f4 Documentation: co... |
162 |
if (!dma_set_mask(dev, DMA_BIT_MASK(64))) { |
1da177e4c Linux-2.6.12-rc2 |
163 |
using_dac = 1; |
216bf58f4 Documentation: co... |
164 |
} else if (!dma_set_mask(dev, DMA_BIT_MASK(32))) { |
1da177e4c Linux-2.6.12-rc2 |
165 166 167 168 169 170 171 172 173 174 175 176 |
using_dac = 0; } else { printk(KERN_WARNING "mydev: No suitable DMA available. "); goto ignore_this_device; } If a card is capable of using 64-bit consistent allocations as well, the case would look like this: int using_dac, consistent_using_dac; |
4aa806b77 DMA-API: provide ... |
177 |
if (!dma_set_mask_and_coherent(dev, DMA_BIT_MASK(64))) { |
1da177e4c Linux-2.6.12-rc2 |
178 179 |
using_dac = 1; consistent_using_dac = 1; |
4aa806b77 DMA-API: provide ... |
180 |
} else if (!dma_set_mask_and_coherent(dev, DMA_BIT_MASK(32))) { |
1da177e4c Linux-2.6.12-rc2 |
181 182 |
using_dac = 0; consistent_using_dac = 0; |
1da177e4c Linux-2.6.12-rc2 |
183 184 185 186 187 188 |
} else { printk(KERN_WARNING "mydev: No suitable DMA available. "); goto ignore_this_device; } |
4aa806b77 DMA-API: provide ... |
189 190 |
The coherent coherent mask will always be able to set the same or a smaller mask as the streaming mask. However for the rare case that a |
1da177e4c Linux-2.6.12-rc2 |
191 |
device driver only uses consistent allocations, one would have to |
216bf58f4 Documentation: co... |
192 |
check the return value from dma_set_coherent_mask(). |
1da177e4c Linux-2.6.12-rc2 |
193 |
|
1da177e4c Linux-2.6.12-rc2 |
194 |
Finally, if your device can only drive the low 24-bits of |
216bf58f4 Documentation: co... |
195 |
address you might do something like: |
1da177e4c Linux-2.6.12-rc2 |
196 |
|
216bf58f4 Documentation: co... |
197 |
if (dma_set_mask(dev, DMA_BIT_MASK(24))) { |
1da177e4c Linux-2.6.12-rc2 |
198 199 200 201 202 |
printk(KERN_WARNING "mydev: 24-bit DMA addressing not available. "); goto ignore_this_device; } |
4aa806b77 DMA-API: provide ... |
203 204 205 |
When dma_set_mask() or dma_set_mask_and_coherent() is successful, and returns zero, the kernel saves away this mask you have provided. The kernel will use this information later when you make DMA mappings. |
1da177e4c Linux-2.6.12-rc2 |
206 207 208 209 210 211 212 |
There is a case which we are aware of at this time, which is worth mentioning in this documentation. If your device supports multiple functions (for example a sound card provides playback and record functions) and the various different functions have _different_ DMA addressing limitations, you may wish to probe each mask and only provide the functionality which the machine can handle. It |
216bf58f4 Documentation: co... |
213 |
is important that the last call to dma_set_mask() be for the |
1da177e4c Linux-2.6.12-rc2 |
214 215 216 |
most specific mask. Here is pseudo-code showing how this might be done: |
2c5510d4e dma-mapping: upda... |
217 |
#define PLAYBACK_ADDRESS_BITS DMA_BIT_MASK(32) |
038f7d002 use DMA_BIT_MASK ... |
218 |
#define RECORD_ADDRESS_BITS DMA_BIT_MASK(24) |
1da177e4c Linux-2.6.12-rc2 |
219 220 |
struct my_sound_card *card; |
216bf58f4 Documentation: co... |
221 |
struct device *dev; |
1da177e4c Linux-2.6.12-rc2 |
222 223 |
... |
216bf58f4 Documentation: co... |
224 |
if (!dma_set_mask(dev, PLAYBACK_ADDRESS_BITS)) { |
1da177e4c Linux-2.6.12-rc2 |
225 226 227 |
card->playback_enabled = 1; } else { card->playback_enabled = 0; |
472c06441 Use KERN_WARNING ... |
228 229 |
printk(KERN_WARNING "%s: Playback disabled due to DMA limitations. ", |
1da177e4c Linux-2.6.12-rc2 |
230 231 |
card->name); } |
216bf58f4 Documentation: co... |
232 |
if (!dma_set_mask(dev, RECORD_ADDRESS_BITS)) { |
1da177e4c Linux-2.6.12-rc2 |
233 234 235 |
card->record_enabled = 1; } else { card->record_enabled = 0; |
472c06441 Use KERN_WARNING ... |
236 237 |
printk(KERN_WARNING "%s: Record disabled due to DMA limitations. ", |
1da177e4c Linux-2.6.12-rc2 |
238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 |
card->name); } A sound card was used as an example here because this genre of PCI devices seems to be littered with ISA chips given a PCI front end, and thus retaining the 16MB DMA addressing limitations of ISA. Types of DMA mappings There are two types of DMA mappings: - Consistent DMA mappings which are usually mapped at driver initialization, unmapped at the end and for which the hardware should guarantee that the device and the CPU can access the data in parallel and will see updates made by each other without any explicit software flushing. Think of "consistent" as "synchronous" or "coherent". The current default is to return consistent memory in the low 32 |
216bf58f4 Documentation: co... |
258 259 |
bits of the bus space. However, for future compatibility you should set the consistent mask even if this default is fine for your |
1da177e4c Linux-2.6.12-rc2 |
260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 |
driver. Good examples of what to use consistent mappings for are: - Network card DMA ring descriptors. - SCSI adapter mailbox command data structures. - Device firmware microcode executed out of main memory. The invariant these examples all require is that any CPU store to memory is immediately visible to the device, and vice versa. Consistent mappings guarantee this. IMPORTANT: Consistent DMA memory does not preclude the usage of proper memory barriers. The CPU may reorder stores to consistent memory just as it may normal memory. Example: if it is important for the device to see the first word of a descriptor updated before the second, you must do something like: desc->word0 = address; wmb(); desc->word1 = DESC_VALID; in order to get correct behavior on all platforms. |
21440d313 [PATCH] dma doc u... |
285 286 287 288 |
Also, on some platforms your driver may need to flush CPU write buffers in much the same way as it needs to flush write buffers found in PCI bridges (such as by reading a register's value after writing it). |
216bf58f4 Documentation: co... |
289 290 291 |
- Streaming DMA mappings which are usually mapped for one DMA transfer, unmapped right after it (unless you use dma_sync_* below) and for which hardware can optimize for sequential accesses. |
1da177e4c Linux-2.6.12-rc2 |
292 293 294 295 296 297 298 299 300 301 302 303 304 |
This of "streaming" as "asynchronous" or "outside the coherency domain". Good examples of what to use streaming mappings for are: - Networking buffers transmitted/received by a device. - Filesystem buffers written/read by a SCSI device. The interfaces for using this type of mapping were designed in such a way that an implementation can make whatever performance optimizations the hardware allows. To this end, when using such mappings you must be explicit about what you want to happen. |
216bf58f4 Documentation: co... |
305 306 |
Neither type of DMA mapping has alignment restrictions that come from the underlying bus, although some devices may have such restrictions. |
21440d313 [PATCH] dma doc u... |
307 308 |
Also, systems with caches that aren't DMA-coherent will work better when the underlying buffers don't share cache lines with other data. |
1da177e4c Linux-2.6.12-rc2 |
309 310 311 312 313 314 315 |
Using Consistent DMA mappings. To allocate and map large (PAGE_SIZE or so) consistent DMA regions, you should do: dma_addr_t dma_handle; |
216bf58f4 Documentation: co... |
316 |
cpu_addr = dma_alloc_coherent(dev, size, &dma_handle, gfp); |
1da177e4c Linux-2.6.12-rc2 |
317 |
|
216bf58f4 Documentation: co... |
318 319 |
where device is a struct device *. This may be called in interrupt context with the GFP_ATOMIC flag. |
1da177e4c Linux-2.6.12-rc2 |
320 321 322 323 324 325 |
Size is the length of the region you want to allocate, in bytes. This routine will allocate RAM for that region, so it acts similarly to __get_free_pages (but takes size instead of a page order). If your driver needs regions sized smaller than a page, you may prefer using |
216bf58f4 Documentation: co... |
326 327 328 329 330 331 332 333 334 335 336 |
the dma_pool interface, described below. The consistent DMA mapping interfaces, for non-NULL dev, will by default return a DMA address which is 32-bit addressable. Even if the device indicates (via DMA mask) that it may address the upper 32-bits, consistent allocation will only return > 32-bit addresses for DMA if the consistent DMA mask has been explicitly changed via dma_set_coherent_mask(). This is true of the dma_pool interface as well. dma_alloc_coherent returns two values: the virtual address which you |
1da177e4c Linux-2.6.12-rc2 |
337 338 339 340 341 342 343 344 345 346 347 |
can use to access it from the CPU and dma_handle which you pass to the card. The cpu return address and the DMA bus master address are both guaranteed to be aligned to the smallest PAGE_SIZE order which is greater than or equal to the requested size. This invariant exists (for example) to guarantee that if you allocate a chunk which is smaller than or equal to 64 kilobytes, the extent of the buffer you receive will not cross a 64K boundary. To unmap and free such a DMA region, you call: |
216bf58f4 Documentation: co... |
348 |
dma_free_coherent(dev, size, cpu_addr, dma_handle); |
1da177e4c Linux-2.6.12-rc2 |
349 |
|
216bf58f4 Documentation: co... |
350 351 |
where dev, size are the same as in the above call and cpu_addr and dma_handle are the values dma_alloc_coherent returned to you. |
1da177e4c Linux-2.6.12-rc2 |
352 353 354 |
This function may not be called in interrupt context. If your driver needs lots of smaller memory regions, you can write |
216bf58f4 Documentation: co... |
355 356 357 |
custom code to subdivide pages returned by dma_alloc_coherent, or you can use the dma_pool API to do that. A dma_pool is like a kmem_cache, but it uses dma_alloc_coherent not __get_free_pages. |
1da177e4c Linux-2.6.12-rc2 |
358 359 |
Also, it understands common hardware constraints for alignment, like queue heads needing to be aligned on N byte boundaries. |
216bf58f4 Documentation: co... |
360 |
Create a dma_pool like this: |
1da177e4c Linux-2.6.12-rc2 |
361 |
|
216bf58f4 Documentation: co... |
362 |
struct dma_pool *pool; |
1da177e4c Linux-2.6.12-rc2 |
363 |
|
216bf58f4 Documentation: co... |
364 |
pool = dma_pool_create(name, dev, size, align, alloc); |
1da177e4c Linux-2.6.12-rc2 |
365 |
|
216bf58f4 Documentation: co... |
366 |
The "name" is for diagnostics (like a kmem_cache name); dev and size |
1da177e4c Linux-2.6.12-rc2 |
367 368 369 370 371 |
are as above. The device's hardware alignment requirement for this type of data is "align" (which is expressed 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 from this pool must not cross 4KByte boundaries (but at that time it may be better to |
216bf58f4 Documentation: co... |
372 |
go for dma_alloc_coherent directly instead). |
1da177e4c Linux-2.6.12-rc2 |
373 |
|
216bf58f4 Documentation: co... |
374 |
Allocate memory from a dma pool like this: |
1da177e4c Linux-2.6.12-rc2 |
375 |
|
216bf58f4 Documentation: co... |
376 |
cpu_addr = dma_pool_alloc(pool, flags, &dma_handle); |
1da177e4c Linux-2.6.12-rc2 |
377 378 |
flags are SLAB_KERNEL if blocking is permitted (not in_interrupt nor |
216bf58f4 Documentation: co... |
379 |
holding SMP locks), SLAB_ATOMIC otherwise. Like dma_alloc_coherent, |
1da177e4c Linux-2.6.12-rc2 |
380 |
this returns two values, cpu_addr and dma_handle. |
216bf58f4 Documentation: co... |
381 |
Free memory that was allocated from a dma_pool like this: |
1da177e4c Linux-2.6.12-rc2 |
382 |
|
216bf58f4 Documentation: co... |
383 |
dma_pool_free(pool, cpu_addr, dma_handle); |
1da177e4c Linux-2.6.12-rc2 |
384 |
|
216bf58f4 Documentation: co... |
385 386 |
where pool is what you passed to dma_pool_alloc, and cpu_addr and dma_handle are the values dma_pool_alloc returned. This function |
1da177e4c Linux-2.6.12-rc2 |
387 |
may be called in interrupt context. |
216bf58f4 Documentation: co... |
388 |
Destroy a dma_pool by calling: |
1da177e4c Linux-2.6.12-rc2 |
389 |
|
216bf58f4 Documentation: co... |
390 |
dma_pool_destroy(pool); |
1da177e4c Linux-2.6.12-rc2 |
391 |
|
216bf58f4 Documentation: co... |
392 |
Make sure you've called dma_pool_free for all memory allocated |
1da177e4c Linux-2.6.12-rc2 |
393 394 395 396 397 398 399 400 |
from a pool before you destroy the pool. This function may not be called in interrupt context. DMA Direction The interfaces described in subsequent portions of this document take a DMA direction argument, which is an integer and takes on one of the following values: |
216bf58f4 Documentation: co... |
401 402 403 404 |
DMA_BIDIRECTIONAL DMA_TO_DEVICE DMA_FROM_DEVICE DMA_NONE |
1da177e4c Linux-2.6.12-rc2 |
405 406 |
One should provide the exact DMA direction if you know it. |
216bf58f4 Documentation: co... |
407 408 |
DMA_TO_DEVICE means "from main memory to the device" DMA_FROM_DEVICE means "from the device to main memory" |
1da177e4c Linux-2.6.12-rc2 |
409 410 411 412 413 414 415 |
It is the direction in which the data moves during the DMA transfer. You are _strongly_ encouraged to specify this as precisely as you possibly can. If you absolutely cannot know the direction of the DMA transfer, |
216bf58f4 Documentation: co... |
416 |
specify DMA_BIDIRECTIONAL. It means that the DMA can go in |
1da177e4c Linux-2.6.12-rc2 |
417 418 419 |
either direction. The platform guarantees that you may legally specify this, and that it will work, but this may be at the cost of performance for example. |
216bf58f4 Documentation: co... |
420 |
The value DMA_NONE is to be used for debugging. One can |
1da177e4c Linux-2.6.12-rc2 |
421 422 423 424 425 426 427 428 429 |
hold this in a data structure before you come to know the precise direction, and this will help catch cases where your direction tracking logic has failed to set things up properly. Another advantage of specifying this value precisely (outside of potential platform-specific optimizations of such) is for debugging. Some platforms actually have a write permission boolean which DMA mappings can be marked with, much like page protections in the user program address space. Such platforms can and do report errors in the |
216bf58f4 Documentation: co... |
430 |
kernel logs when the DMA controller hardware detects violation of the |
1da177e4c Linux-2.6.12-rc2 |
431 432 433 434 |
permission setting. Only streaming mappings specify a direction, consistent mappings implicitly have a direction attribute setting of |
216bf58f4 Documentation: co... |
435 |
DMA_BIDIRECTIONAL. |
1da177e4c Linux-2.6.12-rc2 |
436 |
|
be7db055d [PATCH] remove ol... |
437 438 439 |
The SCSI subsystem tells you the direction to use in the 'sc_data_direction' member of the SCSI command your driver is working on. |
1da177e4c Linux-2.6.12-rc2 |
440 441 |
For Networking drivers, it's a rather simple affair. For transmit |
216bf58f4 Documentation: co... |
442 |
packets, map/unmap them with the DMA_TO_DEVICE direction |
1da177e4c Linux-2.6.12-rc2 |
443 |
specifier. For receive packets, just the opposite, map/unmap them |
216bf58f4 Documentation: co... |
444 |
with the DMA_FROM_DEVICE direction specifier. |
1da177e4c Linux-2.6.12-rc2 |
445 446 447 448 449 450 451 452 453 |
Using Streaming DMA mappings The streaming DMA mapping routines can be called from interrupt context. There are two versions of each map/unmap, one which will map/unmap a single memory region, and one which will map/unmap a scatterlist. To map a single region, you do: |
216bf58f4 Documentation: co... |
454 |
struct device *dev = &my_dev->dev; |
1da177e4c Linux-2.6.12-rc2 |
455 456 457 |
dma_addr_t dma_handle; void *addr = buffer->ptr; size_t size = buffer->len; |
216bf58f4 Documentation: co... |
458 |
dma_handle = dma_map_single(dev, addr, size, direction); |
8d7f62e6a Documentation DMA... |
459 460 461 462 463 464 465 466 |
if (dma_mapping_error(dma_handle)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ goto map_error_handling; } |
1da177e4c Linux-2.6.12-rc2 |
467 468 |
and to unmap it: |
216bf58f4 Documentation: co... |
469 |
dma_unmap_single(dev, dma_handle, size, direction); |
1da177e4c Linux-2.6.12-rc2 |
470 |
|
8d7f62e6a Documentation DMA... |
471 472 473 474 475 476 477 |
You should call dma_mapping_error() as dma_map_single() could fail and return error. Not all dma implementations support dma_mapping_error() interface. However, it is a good practice to call dma_mapping_error() interface, which will invoke the generic mapping error check interface. Doing so will ensure that the mapping code will work correctly on all dma implementations without any dependency on the specifics of the underlying implementation. Using the returned address without checking for errors could result in failures ranging |
be62bc410 Documentation/DMA... |
478 479 480 481 |
from panics to silent data corruption. A couple of examples of incorrect ways to check for errors that make assumptions about the underlying dma implementation are as follows and these are applicable to dma_map_page() as well. |
8d7f62e6a Documentation DMA... |
482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 |
Incorrect example 1: dma_addr_t dma_handle; dma_handle = dma_map_single(dev, addr, size, direction); if ((dma_handle & 0xffff != 0) || (dma_handle >= 0x1000000)) { goto map_error; } Incorrect example 2: dma_addr_t dma_handle; dma_handle = dma_map_single(dev, addr, size, direction); if (dma_handle == DMA_ERROR_CODE) { goto map_error; } |
216bf58f4 Documentation: co... |
498 |
You should call dma_unmap_single when the DMA activity is finished, e.g. |
1da177e4c Linux-2.6.12-rc2 |
499 500 501 502 |
from the interrupt which told you that the DMA transfer is done. Using cpu pointers like this for single mappings has a disadvantage, you cannot reference HIGHMEM memory in this way. Thus, there is a |
216bf58f4 Documentation: co... |
503 |
map/unmap interface pair akin to dma_{map,unmap}_single. These |
1da177e4c Linux-2.6.12-rc2 |
504 505 |
interfaces deal with page/offset pairs instead of cpu pointers. Specifically: |
216bf58f4 Documentation: co... |
506 |
struct device *dev = &my_dev->dev; |
1da177e4c Linux-2.6.12-rc2 |
507 508 509 510 |
dma_addr_t dma_handle; struct page *page = buffer->page; unsigned long offset = buffer->offset; size_t size = buffer->len; |
216bf58f4 Documentation: co... |
511 |
dma_handle = dma_map_page(dev, page, offset, size, direction); |
8d7f62e6a Documentation DMA... |
512 513 514 515 516 517 518 519 |
if (dma_mapping_error(dma_handle)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ goto map_error_handling; } |
1da177e4c Linux-2.6.12-rc2 |
520 521 |
... |
216bf58f4 Documentation: co... |
522 |
dma_unmap_page(dev, dma_handle, size, direction); |
1da177e4c Linux-2.6.12-rc2 |
523 524 |
Here, "offset" means byte offset within the given page. |
8d7f62e6a Documentation DMA... |
525 526 527 528 529 |
You should call dma_mapping_error() as dma_map_page() could fail and return error as outlined under the dma_map_single() discussion. You should call dma_unmap_page when the DMA activity is finished, e.g. from the interrupt which told you that the DMA transfer is done. |
1da177e4c Linux-2.6.12-rc2 |
530 |
With scatterlists, you map a region gathered from several regions by: |
216bf58f4 Documentation: co... |
531 |
int i, count = dma_map_sg(dev, sglist, nents, direction); |
1da177e4c Linux-2.6.12-rc2 |
532 |
struct scatterlist *sg; |
4c2f6d4c2 use sg helper fun... |
533 |
for_each_sg(sglist, sg, count, i) { |
1da177e4c Linux-2.6.12-rc2 |
534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 |
hw_address[i] = sg_dma_address(sg); hw_len[i] = sg_dma_len(sg); } where nents is the number of entries in the sglist. The implementation is free to merge several consecutive sglist entries into one (e.g. if DMA mapping is done with PAGE_SIZE granularity, any consecutive sglist entries can be merged into one provided the first one ends and the second one starts on a page boundary - in fact this is a huge advantage for cards which either cannot do scatter-gather or have very limited number of scatter-gather entries) and returns the actual number of sg entries it mapped them to. On failure 0 is returned. 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 accessed sg->address and sg->length as shown above. To unmap a scatterlist, just call: |
216bf58f4 Documentation: co... |
553 |
dma_unmap_sg(dev, sglist, nents, direction); |
1da177e4c Linux-2.6.12-rc2 |
554 555 |
Again, make sure DMA activity has already finished. |
216bf58f4 Documentation: co... |
556 557 |
PLEASE NOTE: The 'nents' argument to the dma_unmap_sg call must be the _same_ one you passed into the dma_map_sg call, |
1da177e4c Linux-2.6.12-rc2 |
558 |
it should _NOT_ be the 'count' value _returned_ from the |
216bf58f4 Documentation: co... |
559 |
dma_map_sg call. |
1da177e4c Linux-2.6.12-rc2 |
560 |
|
216bf58f4 Documentation: co... |
561 |
Every dma_map_{single,sg} call should have its dma_unmap_{single,sg} |
1da177e4c Linux-2.6.12-rc2 |
562 563 564 565 566 567 568 569 570 |
counterpart, because the bus address space is a shared resource (although in some ports the mapping is per each BUS so less devices contend for the same bus address space) and you could render the machine unusable by eating all bus addresses. If you need to use the same streaming DMA region multiple times and touch the data in between the DMA transfers, the buffer needs to be synced properly in order for the cpu and device to see the most uptodate and correct copy of the DMA buffer. |
216bf58f4 Documentation: co... |
571 |
So, firstly, just map it with dma_map_{single,sg}, and after each DMA |
1da177e4c Linux-2.6.12-rc2 |
572 |
transfer call either: |
216bf58f4 Documentation: co... |
573 |
dma_sync_single_for_cpu(dev, dma_handle, size, direction); |
1da177e4c Linux-2.6.12-rc2 |
574 575 |
or: |
216bf58f4 Documentation: co... |
576 |
dma_sync_sg_for_cpu(dev, sglist, nents, direction); |
1da177e4c Linux-2.6.12-rc2 |
577 578 579 580 581 582 |
as appropriate. Then, if you wish to let the device get at the DMA area again, finish accessing the data with the cpu, and then before actually giving the buffer to the hardware call either: |
216bf58f4 Documentation: co... |
583 |
dma_sync_single_for_device(dev, dma_handle, size, direction); |
1da177e4c Linux-2.6.12-rc2 |
584 585 |
or: |
216bf58f4 Documentation: co... |
586 |
dma_sync_sg_for_device(dev, sglist, nents, direction); |
1da177e4c Linux-2.6.12-rc2 |
587 588 589 590 |
as appropriate. After the last DMA transfer call one of the DMA unmap routines |
216bf58f4 Documentation: co... |
591 592 |
dma_unmap_{single,sg}. If you don't touch the data from the first dma_map_* call till dma_unmap_*, then you don't have to call the dma_sync_* |
1da177e4c Linux-2.6.12-rc2 |
593 594 595 |
routines at all. Here is pseudo code which shows a situation in which you would need |
216bf58f4 Documentation: co... |
596 |
to use the dma_sync_*() interfaces. |
1da177e4c Linux-2.6.12-rc2 |
597 598 599 600 |
my_card_setup_receive_buffer(struct my_card *cp, char *buffer, int len) { dma_addr_t mapping; |
216bf58f4 Documentation: co... |
601 |
mapping = dma_map_single(cp->dev, buffer, len, DMA_FROM_DEVICE); |
8d7f62e6a Documentation DMA... |
602 603 604 605 606 607 608 609 |
if (dma_mapping_error(dma_handle)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ goto map_error_handling; } |
1da177e4c Linux-2.6.12-rc2 |
610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 |
cp->rx_buf = buffer; cp->rx_len = len; cp->rx_dma = mapping; give_rx_buf_to_card(cp); } ... my_card_interrupt_handler(int irq, void *devid, struct pt_regs *regs) { struct my_card *cp = devid; ... if (read_card_status(cp) == RX_BUF_TRANSFERRED) { struct my_card_header *hp; /* Examine the header to see if we wish * to accept the data. But synchronize * the DMA transfer with the CPU first * so that we see updated contents. */ |
216bf58f4 Documentation: co... |
633 634 635 |
dma_sync_single_for_cpu(&cp->dev, cp->rx_dma, cp->rx_len, DMA_FROM_DEVICE); |
1da177e4c Linux-2.6.12-rc2 |
636 637 638 639 |
/* Now it is safe to examine the buffer. */ hp = (struct my_card_header *) cp->rx_buf; if (header_is_ok(hp)) { |
216bf58f4 Documentation: co... |
640 641 |
dma_unmap_single(&cp->dev, cp->rx_dma, cp->rx_len, DMA_FROM_DEVICE); |
1da177e4c Linux-2.6.12-rc2 |
642 643 644 |
pass_to_upper_layers(cp->rx_buf); make_and_setup_new_rx_buf(cp); } else { |
3f0fb4e85 Documentation/DMA... |
645 646 647 648 649 650 |
/* CPU should not write to * DMA_FROM_DEVICE-mapped area, * so dma_sync_single_for_device() is * not needed here. It would be required * for DMA_BIDIRECTIONAL mapping if * the memory was modified. |
1da177e4c Linux-2.6.12-rc2 |
651 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
652 653 654 655 656 657 658 659 660 |
give_rx_buf_to_card(cp); } } } Drivers converted fully to this interface should not use virt_to_bus any longer, nor should they use bus_to_virt. Some drivers have to be changed a little bit, because there is no longer an equivalent to bus_to_virt in the dynamic DMA mapping scheme - you have to always store the DMA addresses |
216bf58f4 Documentation: co... |
661 662 |
returned by the dma_alloc_coherent, dma_pool_alloc, and dma_map_single calls (dma_map_sg stores them in the scatterlist itself if the platform |
1da177e4c Linux-2.6.12-rc2 |
663 664 |
supports dynamic DMA mapping in hardware) in your driver structures and/or in the card registers. |
216bf58f4 Documentation: co... |
665 666 |
All drivers should be using these interfaces with no exceptions. It is planned to completely remove virt_to_bus() and bus_to_virt() as |
1da177e4c Linux-2.6.12-rc2 |
667 668 |
they are entirely deprecated. Some ports already do not provide these as it is impossible to correctly support them. |
4ae9ca825 Documentation: mo... |
669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 |
Handling Errors DMA address space is limited on some architectures and an allocation failure can be determined by: - checking if dma_alloc_coherent returns NULL or dma_map_sg returns 0 - checking the returned dma_addr_t of dma_map_single and dma_map_page by using dma_mapping_error(): dma_addr_t dma_handle; dma_handle = dma_map_single(dev, addr, size, direction); if (dma_mapping_error(dev, dma_handle)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ |
8d7f62e6a Documentation DMA... |
688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 |
goto map_error_handling; } - unmap pages that are already mapped, when mapping error occurs in the middle of a multiple page mapping attempt. These example are applicable to dma_map_page() as well. Example 1: dma_addr_t dma_handle1; dma_addr_t dma_handle2; dma_handle1 = dma_map_single(dev, addr, size, direction); if (dma_mapping_error(dev, dma_handle1)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ goto map_error_handling1; } dma_handle2 = dma_map_single(dev, addr, size, direction); if (dma_mapping_error(dev, dma_handle2)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ goto map_error_handling2; } ... map_error_handling2: dma_unmap_single(dma_handle1); map_error_handling1: |
11cd3db01 Documentation/DMA... |
723 |
Example 2: (if buffers are allocated in a loop, unmap all mapped buffers when |
8d7f62e6a Documentation DMA... |
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 |
mapping error is detected in the middle) dma_addr_t dma_addr; dma_addr_t array[DMA_BUFFERS]; int save_index = 0; for (i = 0; i < DMA_BUFFERS; i++) { ... dma_addr = dma_map_single(dev, addr, size, direction); if (dma_mapping_error(dev, dma_addr)) { /* * reduce current DMA mapping usage, * delay and try again later or * reset driver. */ goto map_error_handling; } array[i].dma_addr = dma_addr; save_index++; } ... map_error_handling: for (i = 0; i < save_index; i++) { ... dma_unmap_single(array[i].dma_addr); |
4ae9ca825 Documentation: mo... |
756 757 758 759 760 761 762 763 764 765 |
} Networking drivers must call dev_kfree_skb to free the socket buffer and return NETDEV_TX_OK if the DMA mapping fails on the transmit hook (ndo_start_xmit). This means that the socket buffer is just dropped in the failure case. SCSI drivers must return SCSI_MLQUEUE_HOST_BUSY if the DMA mapping fails in the queuecommand hook. This means that the SCSI subsystem passes the command to the driver again later. |
1da177e4c Linux-2.6.12-rc2 |
766 |
Optimizing Unmap State Space Consumption |
216bf58f4 Documentation: co... |
767 |
On many platforms, dma_unmap_{single,page}() is simply a nop. |
1da177e4c Linux-2.6.12-rc2 |
768 769 770 771 772 773 774 |
Therefore, keeping track of the mapping address and length is a waste of space. Instead of filling your drivers up with ifdefs and the like to "work around" this (which would defeat the whole purpose of a portable API) the following facilities are provided. Actually, instead of describing the macros one by one, we'll transform some example code. |
216bf58f4 Documentation: co... |
775 |
1) Use DEFINE_DMA_UNMAP_{ADDR,LEN} in state saving structures. |
1da177e4c Linux-2.6.12-rc2 |
776 777 778 779 780 781 782 783 784 785 786 787 |
Example, before: struct ring_state { struct sk_buff *skb; dma_addr_t mapping; __u32 len; }; after: struct ring_state { struct sk_buff *skb; |
216bf58f4 Documentation: co... |
788 789 |
DEFINE_DMA_UNMAP_ADDR(mapping); DEFINE_DMA_UNMAP_LEN(len); |
1da177e4c Linux-2.6.12-rc2 |
790 |
}; |
216bf58f4 Documentation: co... |
791 |
2) Use dma_unmap_{addr,len}_set to set these values. |
1da177e4c Linux-2.6.12-rc2 |
792 793 794 795 796 797 |
Example, before: ringp->mapping = FOO; ringp->len = BAR; after: |
216bf58f4 Documentation: co... |
798 799 |
dma_unmap_addr_set(ringp, mapping, FOO); dma_unmap_len_set(ringp, len, BAR); |
1da177e4c Linux-2.6.12-rc2 |
800 |
|
216bf58f4 Documentation: co... |
801 |
3) Use dma_unmap_{addr,len} to access these values. |
1da177e4c Linux-2.6.12-rc2 |
802 |
Example, before: |
216bf58f4 Documentation: co... |
803 804 |
dma_unmap_single(dev, ringp->mapping, ringp->len, DMA_FROM_DEVICE); |
1da177e4c Linux-2.6.12-rc2 |
805 806 |
after: |
216bf58f4 Documentation: co... |
807 808 809 810 |
dma_unmap_single(dev, dma_unmap_addr(ringp, mapping), dma_unmap_len(ringp, len), DMA_FROM_DEVICE); |
1da177e4c Linux-2.6.12-rc2 |
811 812 813 814 815 816 817 818 819 820 821 822 |
It really should be self-explanatory. We treat the ADDR and LEN separately, because it is possible for an implementation to only need the address in order to perform the unmap operation. Platform Issues If you are just writing drivers for Linux and do not maintain an architecture port for the kernel, you can safely skip down to "Closing". 1) Struct scatterlist requirements. |
b02de8716 Documentation: up... |
823 824 825 826 |
Don't invent the architecture specific struct scatterlist; just use <asm-generic/scatterlist.h>. You need to enable CONFIG_NEED_SG_DMA_LENGTH if the architecture supports IOMMUs (including software IOMMU). |
1da177e4c Linux-2.6.12-rc2 |
827 |
|
ce00f7feb Documentation: DM... |
828 |
2) ARCH_DMA_MINALIGN |
2fd74e25d Documentation/DMA... |
829 830 831 832 833 |
Architectures must ensure that kmalloc'ed buffer is DMA-safe. Drivers and subsystems depend on it. If an architecture isn't fully DMA-coherent (i.e. hardware doesn't ensure that data in the CPU cache is identical to data in main memory), |
ce00f7feb Documentation: DM... |
834 |
ARCH_DMA_MINALIGN must be set so that the memory allocator |
2fd74e25d Documentation/DMA... |
835 836 |
makes sure that kmalloc'ed buffer doesn't share a cache line with the others. See arch/arm/include/asm/cache.h as an example. |
ce00f7feb Documentation: DM... |
837 |
Note that ARCH_DMA_MINALIGN is about DMA memory alignment |
2fd74e25d Documentation/DMA... |
838 839 840 |
constraints. You don't need to worry about the architecture data alignment constraints (e.g. the alignment constraints about 64-bit objects). |
1da177e4c Linux-2.6.12-rc2 |
841 |
|
c31e74c4c Documentation: DM... |
842 843 844 845 846 847 848 849 850 |
3) Supporting multiple types of IOMMUs If your architecture needs to support multiple types of IOMMUs, you can use include/linux/asm-generic/dma-mapping-common.h. It's a library to support the DMA API with multiple types of IOMMUs. Lots of architectures (x86, powerpc, sh, alpha, ia64, microblaze and sparc) use it. Choose one to see how it can be used. If you need to support multiple types of IOMMUs in a single system, the example of x86 or powerpc helps. |
1da177e4c Linux-2.6.12-rc2 |
851 |
Closing |
a33f32244 Documentation/: i... |
852 |
This document, and the API itself, would not be in its current |
1da177e4c Linux-2.6.12-rc2 |
853 854 855 856 857 858 859 860 861 862 863 |
form without the feedback and suggestions from numerous individuals. We would like to specifically mention, in no particular order, the following people: Russell King <rmk@arm.linux.org.uk> Leo Dagum <dagum@barrel.engr.sgi.com> Ralf Baechle <ralf@oss.sgi.com> Grant Grundler <grundler@cup.hp.com> Jay Estabrook <Jay.Estabrook@compaq.com> Thomas Sailer <sailer@ife.ee.ethz.ch> Andrea Arcangeli <andrea@suse.de> |
26bbb29a2 Update Jens Axboe... |
864 |
Jens Axboe <jens.axboe@oracle.com> |
1da177e4c Linux-2.6.12-rc2 |
865 |
David Mosberger-Tang <davidm@hpl.hp.com> |