[patch 2.6.16-git] dma doc updates

From: David Brownell
Date: Sat Apr 01 2006 - 13:46:18 EST


This updates docs to clarify some points:

- dma_*map_sg() usage, mostly cloning text from the pci-specific writeup

- mention cpu (and pci bridge) write buffers as cases where having memory
mapped as dma-coherent isn't sufficient for correctness;

- mention cacheline coherence issues for systems with dma-unaware caches.

Please merge ...
This updates the DMA API documentation to address a few issues:

- The dma_map_sg() call results are used like pci_map_sg() results:
using sg_dma_address() and sg_dma_len(). That's not wholly obvious
to folk reading _only_ the "new" DMA-API.txt writeup.

- Buffers allocated by dma_alloc_coherent() may not be completely
free of coherency concerns ... some CPUs also have write buffers
that may need to be flushed.

- Cacheline coherence issues are now mentioned as being among issues
which affect dma buffers, and complicate/prevent using of static and
(especially) stack based buffers with the DMA calls.

I don't think many drivers currently need to worry about flushing write
buffers, but I did hit it with one SOC using external SDRAM for DMA
descriptors: without explicit writebuffer flushing, the on-chip DMA
controller accessed descriptors before the CPU completed the writes.

Signed-off-by: David Brownell <dbrownell@xxxxxxxxxxxxxxxxxxxxx>

Index: g26/Documentation/DMA-API.txt
===================================================================
--- g26.orig/Documentation/DMA-API.txt 2006-03-31 22:24:55.000000000 -0800
+++ g26/Documentation/DMA-API.txt 2006-03-31 22:42:37.000000000 -0800
@@ -33,7 +33,9 @@ pci_alloc_consistent(struct pci_dev *dev

Consistent memory is memory for which a write by either the device or
the processor can immediately be read by the processor or device
-without having to worry about caching effects.
+without having to worry about caching effects. (You may however need
+to make sure to flush the processor's write buffers before telling
+devices to read that memory.)

This routine allocates a region of <size> bytes of consistent memory.
it also returns a <dma_handle> which may be cast to an unsigned
@@ -304,12 +306,12 @@ dma address with dma_mapping_error(). A
could not be created and the driver should take appropriate action (eg
reduce current DMA mapping usage or delay and try again later).

-int
-dma_map_sg(struct device *dev, struct scatterlist *sg, int nents,
- enum dma_data_direction direction)
-int
-pci_map_sg(struct pci_dev *hwdev, struct scatterlist *sg,
- int nents, int direction)
+ int
+ dma_map_sg(struct device *dev, struct scatterlist *sg,
+ int nents, enum dma_data_direction direction)
+ int
+ pci_map_sg(struct pci_dev *hwdev, struct scatterlist *sg,
+ int nents, int direction)

Maps a scatter gather list from the block layer.

@@ -327,12 +329,33 @@ critical that the driver do something, i
aborting the request or even oopsing is better than doing nothing and
corrupting the filesystem.

-void
-dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nhwentries,
- enum dma_data_direction direction)
-void
-pci_unmap_sg(struct pci_dev *hwdev, struct scatterlist *sg,
- int nents, int direction)
+With scatterlists, you use the resulting mapping like this:
+
+ int i, count = dma_map_sg(dev, sglist, nents, direction);
+ struct scatterlist *sg;
+
+ for (i = 0, sg = sglist; i < count; i++, sg++) {
+ 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. with an IOMMU, or if several pages just happen to be
+physically contiguous) 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.
+
+ void
+ dma_unmap_sg(struct device *dev, struct scatterlist *sg,
+ int nhwentries, enum dma_data_direction direction)
+ void
+ pci_unmap_sg(struct pci_dev *hwdev, struct scatterlist *sg,
+ int nents, int direction)

unmap the previously mapped scatter/gather list. All the parameters
must be the same as those and passed in to the scatter/gather mapping
Index: g26/Documentation/DMA-mapping.txt
===================================================================
--- g26.orig/Documentation/DMA-mapping.txt 2006-03-31 22:35:29.000000000 -0800
+++ g26/Documentation/DMA-mapping.txt 2006-03-31 22:44:51.000000000 -0800
@@ -58,11 +58,15 @@ translating each of those pages back to
something like __va(). [ EDIT: Update this when we integrate
Gerd Knorr's generic code which does this. ]

-This rule also means that you may not use kernel image addresses
-(ie. items in the kernel's data/text/bss segment, or your driver's)
-nor may you use kernel stack addresses for DMA. Both of these items
-might be mapped somewhere entirely different than the rest of physical
-memory.
+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.)

Also, this means that you cannot take the return of a kmap()
call and DMA to/from that. This is similar to vmalloc().
@@ -284,6 +288,11 @@ There are two types of DMA mappings:

in order to get correct behavior on all platforms.

+ 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).
+
- Streaming DMA mappings which are usually mapped for one DMA transfer,
unmapped right after it (unless you use pci_dma_sync_* below) and for which
hardware can optimize for sequential accesses.
@@ -303,6 +312,9 @@ There are two types of DMA mappings:

Neither type of DMA mapping has alignment restrictions that come
from PCI, although some devices may have such restrictions.
+Also, systems with caches that aren't DMA-coherent will work better
+when the underlying buffers don't share cache lines with other data.
+

Using Consistent DMA mappings.