linux-toradex.git/lib/swiotlb.c, branch v2.6.28

swiotlb: use coherent_dma_mask in alloc_coherent

2008-11-17T08:12:02+00:00

Impact: fix DMA buffer allocation coherency bug in certain configs

This patch fixes swiotlb to use dev->coherent_dma_mask in
swiotlb_alloc_coherent().

coherent_dma_mask is a subset of dma_mask (equal to it most of
the time), enumerating the address range that a given device
is able to DMA to/from in a cache-coherent way.

But currently, swiotlb uses dev->dma_mask in alloc_coherent()
implicitly via address_needs_mapping(), but alloc_coherent is really
supposed to use coherent_dma_mask.

This bug could break drivers that uses smaller coherent_dma_mask than
dma_mask (though the current code works for the majority that use the
same mask for coherent_dma_mask and dma_mask).

Signed-off-by: FUJITA Tomonori 
Cc: tony.luck@intel.com
Signed-off-by: Ingo Molnar

swiotlb: remove panic for alloc_coherent failure

2008-10-23T19:54:38+00:00

swiotlb_alloc_coherent calls panic() when allocated swiotlb pages is
not fit for a device's dma mask. However, alloc_coherent failure is
not a disaster at all. AFAIK, none of other x86 and IA64 IOMMU
implementations don't crash in case of alloc_coherent failure.

There are some drivers that don't check alloc_coherent failure but not
many (about ten and I've already started to fix some of
them). alloc_coherent returns NULL in case of failure so it's likely
that these guilty drivers crash immediately. So swiotlb doesn't need
to call panic() just for them.

Reported-by: Takashi Iwai 
Signed-off-by: FUJITA Tomonori 
Tested-by: Takashi Iwai 
Signed-off-by: Ingo Molnar

convert swiotlb to use dma_get_mask

2008-09-19T08:20:41+00:00

swiotlb can use dma_get_mask() instead of the homegrown function.

Signed-off-by: FUJITA Tomonori 
Cc: tony.luck@intel.com
Signed-off-by: Ingo Molnar

swiotlb: convert swiotlb to use is_buffer_dma_capable helper function

2008-09-10T09:33:45+00:00

Signed-off-by: FUJITA Tomonori 
Acked-by: Joerg Roedel 
Signed-off-by: Ingo Molnar

swiotlb: add is_swiotlb_buffer helper function

2008-09-08T13:53:31+00:00

This adds is_swiotlb_buffer() helper function to see whether a buffer
belongs to the swiotlb buffer or not.

Signed-off-by: FUJITA Tomonori 
Signed-off-by: Ingo Molnar

swiotlb: use unmap_single instead of swiotlb_unmap_single in swiotlb_free_coherent

2008-09-08T13:53:31+00:00

We don't need any check in swiotlb_unmap_single here. unmap_single is
appropriate.

Signed-off-by: FUJITA Tomonori 
Signed-off-by: Ingo Molnar

swiotlb: use map_single instead of swiotlb_map_single in swiotlb_alloc_coherent

2008-09-08T13:53:30+00:00

We always need swiotlb memory here so address_needs_mapping and
swiotlb_force testings are irrelevant. map_single should be used here
instead of swiotlb_map_single.

Signed-off-by: FUJITA Tomonori 
Signed-off-by: Ingo Molnar

swiotlb: remove GFP_DMA hack in swiotlb_alloc_coherent

2008-09-08T13:50:07+00:00

The callers are supposed to set up the gfp flags appropriately.

Signed-off-by: FUJITA Tomonori 
Acked-by: Joerg Roedel 
Signed-off-by: Ingo Molnar

dma-mapping: add the device argument to dma_mapping_error()

2008-07-26T19:00:03+00:00

Add per-device dma_mapping_ops support for CONFIG_X86_64 as POWER
architecture does:

This enables us to cleanly fix the Calgary IOMMU issue that some devices
are not behind the IOMMU (http://lkml.org/lkml/2008/5/8/423).

I think that per-device dma_mapping_ops support would be also helpful for
KVM people to support PCI passthrough but Andi thinks that this makes it
difficult to support the PCI passthrough (see the above thread).  So I
CC'ed this to KVM camp.  Comments are appreciated.

A pointer to dma_mapping_ops to struct dev_archdata is added.  If the
pointer is non NULL, DMA operations in asm/dma-mapping.h use it.  If it's
NULL, the system-wide dma_ops pointer is used as before.

If it's useful for KVM people, I plan to implement a mechanism to register
a hook called when a new pci (or dma capable) device is created (it works
with hot plugging).  It enables IOMMUs to set up an appropriate
dma_mapping_ops per device.

The major obstacle is that dma_mapping_error doesn't take a pointer to the
device unlike other DMA operations.  So x86 can't have dma_mapping_ops per
device.  Note all the POWER IOMMUs use the same dma_mapping_error function
so this is not a problem for POWER but x86 IOMMUs use different
dma_mapping_error functions.

The first patch adds the device argument to dma_mapping_error.  The patch
is trivial but large since it touches lots of drivers and dma-mapping.h in
all the architecture.

This patch:

dma_mapping_error() doesn't take a pointer to the device unlike other DMA
operations.  So we can't have dma_mapping_ops per device.

Note that POWER already has dma_mapping_ops per device but all the POWER
IOMMUs use the same dma_mapping_error function.  x86 IOMMUs use device
argument.

[akpm@linux-foundation.org: fix sge]
[akpm@linux-foundation.org: fix svc_rdma]
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: fix bnx2x]
[akpm@linux-foundation.org: fix s2io]
[akpm@linux-foundation.org: fix pasemi_mac]
[akpm@linux-foundation.org: fix sdhci]
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: fix sparc]
[akpm@linux-foundation.org: fix ibmvscsi]
Signed-off-by: FUJITA Tomonori 
Cc: Muli Ben-Yehuda 
Cc: Andi Kleen 
Cc: Thomas Gleixner 
Cc: Ingo Molnar 
Cc: Avi Kivity 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

dma/ia64: update ia64 machvecs, swiotlb.c

2008-04-29T15:06:12+00:00

Change all ia64 machvecs to use the new dma_*map*_attrs() interfaces.
Implement the old dma_*map_*() interfaces in terms of the corresponding new
interfaces.  For ia64/sn, make use of one dma attribute,
DMA_ATTR_WRITE_BARRIER.  Introduce swiotlb_*map*_attrs() functions.

Signed-off-by: Arthur Kepner 
Cc: Tony Luck 
Cc: Jesse Barnes 
Cc: Jes Sorensen 
Cc: Randy Dunlap 
Cc: Roland Dreier 
Cc: James Bottomley 
Cc: David Miller 
Cc: Benjamin Herrenschmidt 
Cc: Grant Grundler 
Cc: Michael Ellerman 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds