[1/3] dma-mapping: remove the default map_resource implementation

Message ID 20190111181731.11782-2-hch@lst.de
State Not Applicable
Headers show
Series
  • [1/3] dma-mapping: remove the default map_resource implementation
Related show

Checks

Context Check Description
snowpatch_ozlabs/checkpatch warning total: 0 errors, 0 warnings, 2 checks, 87 lines checked
snowpatch_ozlabs/apply_patch success next/apply_patch Successfully applied

Commit Message

Christoph Hellwig Jan. 11, 2019, 6:17 p.m.
Just returning the physical address when not map_resource method is
present is highly dangerous as it doesn't take any offset in the
direct mapping into account and does the completely wrong thing for
IOMMUs.  Instead provide a proper implementation in the direct mapping
code, and also wire it up for arm and powerpc.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 arch/arm/mm/dma-mapping.c         |  2 ++
 arch/powerpc/kernel/dma-swiotlb.c |  1 +
 arch/powerpc/kernel/dma.c         |  1 +
 include/linux/dma-mapping.h       | 12 +++++++-----
 kernel/dma/direct.c               | 14 ++++++++++++++
 5 files changed, 25 insertions(+), 5 deletions(-)

Comments

Robin Murphy Jan. 14, 2019, 1:12 p.m. | #1
On 11/01/2019 18:17, Christoph Hellwig wrote:
> Just returning the physical address when not map_resource method is
> present is highly dangerous as it doesn't take any offset in the
> direct mapping into account and does the completely wrong thing for
> IOMMUs.  Instead provide a proper implementation in the direct mapping
> code, and also wire it up for arm and powerpc.

Ignoring the offset was kind of intentional there, because at the time I 
was primarily thinking about it in terms of the Keystone 2 platform 
where the peripherals are all in the same place (0-2GB) in both the bus 
and CPU physical address maps, and only the view of RAM differs between 
the two (2-4GB vs. 32-34GB). However, on something like BCM283x, the 
peripherals region is also offset from its bus address in the CPU view, 
but at a *different* offset relative to that of RAM.

Fortunately, I'm not aware of any platform which has a DMA engine behind 
an IOMMU (and thus *needs* to use dma_map_resource() to avoid said IOMMU 
blocking the slave device register reads/writes) and also has any 
nonzero offsets, and AFAIK the IOMMU-less platforms above aren't using 
dma_map_resource() at all, so this change shouldn't actually break 
anything, but I guess we have a bit of a problem making it truly generic 
and robust :(

Is this perhaps another shove in the direction of overhauling 
dma_pfn_offset into an arbitrary "DMA ranges" lookup table?

> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>   arch/arm/mm/dma-mapping.c         |  2 ++
>   arch/powerpc/kernel/dma-swiotlb.c |  1 +
>   arch/powerpc/kernel/dma.c         |  1 +
>   include/linux/dma-mapping.h       | 12 +++++++-----
>   kernel/dma/direct.c               | 14 ++++++++++++++
>   5 files changed, 25 insertions(+), 5 deletions(-)
> 
> diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
> index f1e2922e447c..3c8534904209 100644
> --- a/arch/arm/mm/dma-mapping.c
> +++ b/arch/arm/mm/dma-mapping.c
> @@ -188,6 +188,7 @@ const struct dma_map_ops arm_dma_ops = {
>   	.unmap_page		= arm_dma_unmap_page,
>   	.map_sg			= arm_dma_map_sg,
>   	.unmap_sg		= arm_dma_unmap_sg,
> +	.map_resource		= dma_direct_map_resource,
>   	.sync_single_for_cpu	= arm_dma_sync_single_for_cpu,
>   	.sync_single_for_device	= arm_dma_sync_single_for_device,
>   	.sync_sg_for_cpu	= arm_dma_sync_sg_for_cpu,
> @@ -211,6 +212,7 @@ const struct dma_map_ops arm_coherent_dma_ops = {
>   	.get_sgtable		= arm_dma_get_sgtable,
>   	.map_page		= arm_coherent_dma_map_page,
>   	.map_sg			= arm_dma_map_sg,
> +	.map_resource		= dma_direct_map_resource,
>   	.dma_supported		= arm_dma_supported,
>   };
>   EXPORT_SYMBOL(arm_coherent_dma_ops);
> diff --git a/arch/powerpc/kernel/dma-swiotlb.c b/arch/powerpc/kernel/dma-swiotlb.c
> index 7d5fc9751622..fbb2506a414e 100644
> --- a/arch/powerpc/kernel/dma-swiotlb.c
> +++ b/arch/powerpc/kernel/dma-swiotlb.c
> @@ -55,6 +55,7 @@ const struct dma_map_ops powerpc_swiotlb_dma_ops = {
>   	.dma_supported = swiotlb_dma_supported,
>   	.map_page = dma_direct_map_page,
>   	.unmap_page = dma_direct_unmap_page,
> +	.map_resource = dma_direct_map_resource,
>   	.sync_single_for_cpu = dma_direct_sync_single_for_cpu,
>   	.sync_single_for_device = dma_direct_sync_single_for_device,
>   	.sync_sg_for_cpu = dma_direct_sync_sg_for_cpu,
> diff --git a/arch/powerpc/kernel/dma.c b/arch/powerpc/kernel/dma.c
> index b1903ebb2e9c..258b9e8ebb99 100644
> --- a/arch/powerpc/kernel/dma.c
> +++ b/arch/powerpc/kernel/dma.c
> @@ -273,6 +273,7 @@ const struct dma_map_ops dma_nommu_ops = {
>   	.dma_supported			= dma_nommu_dma_supported,
>   	.map_page			= dma_nommu_map_page,
>   	.unmap_page			= dma_nommu_unmap_page,
> +	.map_resource			= dma_direct_map_resource,
>   	.get_required_mask		= dma_nommu_get_required_mask,
>   #ifdef CONFIG_NOT_COHERENT_CACHE
>   	.sync_single_for_cpu 		= dma_nommu_sync_single,
> diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
> index cef2127e1d70..d3087829a6df 100644
> --- a/include/linux/dma-mapping.h
> +++ b/include/linux/dma-mapping.h
> @@ -208,6 +208,8 @@ dma_addr_t dma_direct_map_page(struct device *dev, struct page *page,
>   		unsigned long attrs);
>   int dma_direct_map_sg(struct device *dev, struct scatterlist *sgl, int nents,
>   		enum dma_data_direction dir, unsigned long attrs);
> +dma_addr_t dma_direct_map_resource(struct device *dev, phys_addr_t paddr,
> +		size_t size, enum dma_data_direction dir, unsigned long attrs);
>   
>   #if defined(CONFIG_ARCH_HAS_SYNC_DMA_FOR_DEVICE) || \
>       defined(CONFIG_SWIOTLB)
> @@ -346,19 +348,19 @@ static inline dma_addr_t dma_map_resource(struct device *dev,
>   					  unsigned long attrs)
>   {
>   	const struct dma_map_ops *ops = get_dma_ops(dev);
> -	dma_addr_t addr;
> +	dma_addr_t addr = DMA_MAPPING_ERROR;
>   
>   	BUG_ON(!valid_dma_direction(dir));
>   
>   	/* Don't allow RAM to be mapped */
>   	BUG_ON(pfn_valid(PHYS_PFN(phys_addr)));
>   
> -	addr = phys_addr;
> -	if (ops && ops->map_resource)
> +	if (dma_is_direct(ops))
> +		addr = dma_direct_map_resource(dev, phys_addr, size, dir, attrs);
> +	else if (ops->map_resource)
>   		addr = ops->map_resource(dev, phys_addr, size, dir, attrs);

Might it be reasonable to do:

	if (!dma_is_direct(ops) && ops->map_resource)
		addr = ops->map_resource(...);
	else
		addr = dma_direct_map_resource(...);

and avoid having to explicitly wire up the dma_direct callback elsewhere?

Robin.

>   
>   	debug_dma_map_resource(dev, phys_addr, size, dir, addr);
> -
>   	return addr;
>   }
>   
> @@ -369,7 +371,7 @@ static inline void dma_unmap_resource(struct device *dev, dma_addr_t addr,
>   	const struct dma_map_ops *ops = get_dma_ops(dev);
>   
>   	BUG_ON(!valid_dma_direction(dir));
> -	if (ops && ops->unmap_resource)
> +	if (!dma_is_direct(ops) && ops->unmap_resource)
>   		ops->unmap_resource(dev, addr, size, dir, attrs);
>   	debug_dma_unmap_resource(dev, addr, size, dir);
>   }
> diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c
> index 355d16acee6d..8e0359b04957 100644
> --- a/kernel/dma/direct.c
> +++ b/kernel/dma/direct.c
> @@ -356,6 +356,20 @@ int dma_direct_map_sg(struct device *dev, struct scatterlist *sgl, int nents,
>   }
>   EXPORT_SYMBOL(dma_direct_map_sg);
>   
> +dma_addr_t dma_direct_map_resource(struct device *dev, phys_addr_t paddr,
> +		size_t size, enum dma_data_direction dir, unsigned long attrs)
> +{
> +	dma_addr_t dma_addr = phys_to_dma(dev, paddr);
> +
> +	if (unlikely(!dma_direct_possible(dev, dma_addr, size))) {
> +		report_addr(dev, dma_addr, size);
> +		return DMA_MAPPING_ERROR;
> +	}
> +
> +	return dma_addr;
> +}
> +EXPORT_SYMBOL(dma_direct_map_resource);
> +
>   /*
>    * Because 32-bit DMA masks are so common we expect every architecture to be
>    * able to satisfy them - either by not supporting more physical memory, or by
>
Christoph Hellwig Jan. 14, 2019, 5:08 p.m. | #2
On Mon, Jan 14, 2019 at 01:12:33PM +0000, Robin Murphy wrote:
> Ignoring the offset was kind of intentional there, because at the time I 
> was primarily thinking about it in terms of the Keystone 2 platform where 
> the peripherals are all in the same place (0-2GB) in both the bus and CPU 
> physical address maps, and only the view of RAM differs between the two 
> (2-4GB vs. 32-34GB). However, on something like BCM283x, the peripherals 
> region is also offset from its bus address in the CPU view, but at a 
> *different* offset relative to that of RAM.

I was more thinking of the PCIe P2P case, where we need to apply a
consistent offset to translate between the CPU and the bus view.

But this isn't really used for PCIe P2P, so I guess keeping the original
sematics might be a better idea.  That being said the videobuf code seems
to rely on these offsets, so we might be between a rock and a hard place.

> Fortunately, I'm not aware of any platform which has a DMA engine behind an 
> IOMMU (and thus *needs* to use dma_map_resource() to avoid said IOMMU 
> blocking the slave device register reads/writes) and also has any nonzero 
> offsets, and AFAIK the IOMMU-less platforms above aren't using 
> dma_map_resource() at all, so this change shouldn't actually break 
> anything, but I guess we have a bit of a problem making it truly generic 
> and robust :(

Note that we don't actually use the code in this patch for ARM/ARM64
platforms with IOMMUs, as both the ARM and the ARM64 iommu code have
their own implementations of ->map_resource that actually program
the iommu (which at least for the PCIe P2P case would be wrong).

> Is this perhaps another shove in the direction of overhauling 
> dma_pfn_offset into an arbitrary "DMA ranges" lookup table?

It is long overdue anyway.

>>   		addr = ops->map_resource(dev, phys_addr, size, dir, attrs);
>
> Might it be reasonable to do:
>
> 	if (!dma_is_direct(ops) && ops->map_resource)
> 		addr = ops->map_resource(...);
> 	else
> 		addr = dma_direct_map_resource(...);
>
> and avoid having to explicitly wire up the dma_direct callback elsewhere?

No, I absolutely _want_ the callback explicitly wired up.  That is the
only way to ensure we actually intentionally support it and don't just
get a default that often won't work.  Same issue for ->mmap and
->get_sgtable.

Patch

diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
index f1e2922e447c..3c8534904209 100644
--- a/arch/arm/mm/dma-mapping.c
+++ b/arch/arm/mm/dma-mapping.c
@@ -188,6 +188,7 @@  const struct dma_map_ops arm_dma_ops = {
 	.unmap_page		= arm_dma_unmap_page,
 	.map_sg			= arm_dma_map_sg,
 	.unmap_sg		= arm_dma_unmap_sg,
+	.map_resource		= dma_direct_map_resource,
 	.sync_single_for_cpu	= arm_dma_sync_single_for_cpu,
 	.sync_single_for_device	= arm_dma_sync_single_for_device,
 	.sync_sg_for_cpu	= arm_dma_sync_sg_for_cpu,
@@ -211,6 +212,7 @@  const struct dma_map_ops arm_coherent_dma_ops = {
 	.get_sgtable		= arm_dma_get_sgtable,
 	.map_page		= arm_coherent_dma_map_page,
 	.map_sg			= arm_dma_map_sg,
+	.map_resource		= dma_direct_map_resource,
 	.dma_supported		= arm_dma_supported,
 };
 EXPORT_SYMBOL(arm_coherent_dma_ops);
diff --git a/arch/powerpc/kernel/dma-swiotlb.c b/arch/powerpc/kernel/dma-swiotlb.c
index 7d5fc9751622..fbb2506a414e 100644
--- a/arch/powerpc/kernel/dma-swiotlb.c
+++ b/arch/powerpc/kernel/dma-swiotlb.c
@@ -55,6 +55,7 @@  const struct dma_map_ops powerpc_swiotlb_dma_ops = {
 	.dma_supported = swiotlb_dma_supported,
 	.map_page = dma_direct_map_page,
 	.unmap_page = dma_direct_unmap_page,
+	.map_resource = dma_direct_map_resource,
 	.sync_single_for_cpu = dma_direct_sync_single_for_cpu,
 	.sync_single_for_device = dma_direct_sync_single_for_device,
 	.sync_sg_for_cpu = dma_direct_sync_sg_for_cpu,
diff --git a/arch/powerpc/kernel/dma.c b/arch/powerpc/kernel/dma.c
index b1903ebb2e9c..258b9e8ebb99 100644
--- a/arch/powerpc/kernel/dma.c
+++ b/arch/powerpc/kernel/dma.c
@@ -273,6 +273,7 @@  const struct dma_map_ops dma_nommu_ops = {
 	.dma_supported			= dma_nommu_dma_supported,
 	.map_page			= dma_nommu_map_page,
 	.unmap_page			= dma_nommu_unmap_page,
+	.map_resource			= dma_direct_map_resource,
 	.get_required_mask		= dma_nommu_get_required_mask,
 #ifdef CONFIG_NOT_COHERENT_CACHE
 	.sync_single_for_cpu 		= dma_nommu_sync_single,
diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index cef2127e1d70..d3087829a6df 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -208,6 +208,8 @@  dma_addr_t dma_direct_map_page(struct device *dev, struct page *page,
 		unsigned long attrs);
 int dma_direct_map_sg(struct device *dev, struct scatterlist *sgl, int nents,
 		enum dma_data_direction dir, unsigned long attrs);
+dma_addr_t dma_direct_map_resource(struct device *dev, phys_addr_t paddr,
+		size_t size, enum dma_data_direction dir, unsigned long attrs);
 
 #if defined(CONFIG_ARCH_HAS_SYNC_DMA_FOR_DEVICE) || \
     defined(CONFIG_SWIOTLB)
@@ -346,19 +348,19 @@  static inline dma_addr_t dma_map_resource(struct device *dev,
 					  unsigned long attrs)
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
-	dma_addr_t addr;
+	dma_addr_t addr = DMA_MAPPING_ERROR;
 
 	BUG_ON(!valid_dma_direction(dir));
 
 	/* Don't allow RAM to be mapped */
 	BUG_ON(pfn_valid(PHYS_PFN(phys_addr)));
 
-	addr = phys_addr;
-	if (ops && ops->map_resource)
+	if (dma_is_direct(ops))
+		addr = dma_direct_map_resource(dev, phys_addr, size, dir, attrs);
+	else if (ops->map_resource)
 		addr = ops->map_resource(dev, phys_addr, size, dir, attrs);
 
 	debug_dma_map_resource(dev, phys_addr, size, dir, addr);
-
 	return addr;
 }
 
@@ -369,7 +371,7 @@  static inline void dma_unmap_resource(struct device *dev, dma_addr_t addr,
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
 	BUG_ON(!valid_dma_direction(dir));
-	if (ops && ops->unmap_resource)
+	if (!dma_is_direct(ops) && ops->unmap_resource)
 		ops->unmap_resource(dev, addr, size, dir, attrs);
 	debug_dma_unmap_resource(dev, addr, size, dir);
 }
diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c
index 355d16acee6d..8e0359b04957 100644
--- a/kernel/dma/direct.c
+++ b/kernel/dma/direct.c
@@ -356,6 +356,20 @@  int dma_direct_map_sg(struct device *dev, struct scatterlist *sgl, int nents,
 }
 EXPORT_SYMBOL(dma_direct_map_sg);
 
+dma_addr_t dma_direct_map_resource(struct device *dev, phys_addr_t paddr,
+		size_t size, enum dma_data_direction dir, unsigned long attrs)
+{
+	dma_addr_t dma_addr = phys_to_dma(dev, paddr);
+
+	if (unlikely(!dma_direct_possible(dev, dma_addr, size))) {
+		report_addr(dev, dma_addr, size);
+		return DMA_MAPPING_ERROR;
+	}
+
+	return dma_addr;
+}
+EXPORT_SYMBOL(dma_direct_map_resource);
+
 /*
  * Because 32-bit DMA masks are so common we expect every architecture to be
  * able to satisfy them - either by not supporting more physical memory, or by