* [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address
@ 2014-05-12 10:15 David Vrabel
2014-05-12 10:15 ` [PATCH 1/2] dma: add dma_get_required_mask_from_max_pfn() David Vrabel
` (5 more replies)
0 siblings, 6 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Thomas Gleixner, Ingo Molnar, H. Peter Anvin, x86,
Greg Kroah-Hartman, xen-devel, Konrad Rzeszutek Wilk,
Boris Ostrovsky, David Vrabel
x86 uses the generic dma_get_required_mask() which assumes that DMA
address == physical address. For Xen PV guests this is not the case
and certain drivers will cause excessive use of the SWIOTLB.
Patch #1 is a straight forward refactoring of a common function.
Patch #2 addes the required arch-specific implementation of
dma_get_required_mask() to x86.
David
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 1/2] dma: add dma_get_required_mask_from_max_pfn()
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
@ 2014-05-12 10:15 ` David Vrabel
2014-05-12 10:15 ` David Vrabel
` (4 subsequent siblings)
5 siblings, 0 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Thomas Gleixner, Ingo Molnar, H. Peter Anvin, x86,
Greg Kroah-Hartman, xen-devel, Konrad Rzeszutek Wilk,
Boris Ostrovsky, David Vrabel, Tony Luck, Fenghua Yu, linux-ia64
ia64 provides a duplicate of the generic dma_get_required_mask()
because it has ARCH_HAS_GET_REQUIRED_MASK. Provide a common
dma_get_require_mask_max_pfn() instead.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: linux-ia64@vger.kernel.org
---
arch/ia64/include/asm/machvec.h | 2 +-
arch/ia64/include/asm/machvec_init.h | 1 -
arch/ia64/pci/pci.c | 20 --------------------
drivers/base/platform.c | 10 ++++++++--
include/linux/dma-mapping.h | 1 +
5 files changed, 10 insertions(+), 24 deletions(-)
diff --git a/arch/ia64/include/asm/machvec.h b/arch/ia64/include/asm/machvec.h
index 9c39bdf..beaa47d 100644
--- a/arch/ia64/include/asm/machvec.h
+++ b/arch/ia64/include/asm/machvec.h
@@ -287,7 +287,7 @@ extern struct dma_map_ops *dma_get_ops(struct device *);
# define platform_dma_get_ops dma_get_ops
#endif
#ifndef platform_dma_get_required_mask
-# define platform_dma_get_required_mask ia64_dma_get_required_mask
+# define platform_dma_get_required_mask dma_get_required_mask_from_max_pfn
#endif
#ifndef platform_irq_to_vector
# define platform_irq_to_vector __ia64_irq_to_vector
diff --git a/arch/ia64/include/asm/machvec_init.h b/arch/ia64/include/asm/machvec_init.h
index 37a4698..ef964b2 100644
--- a/arch/ia64/include/asm/machvec_init.h
+++ b/arch/ia64/include/asm/machvec_init.h
@@ -3,7 +3,6 @@
extern ia64_mv_send_ipi_t ia64_send_ipi;
extern ia64_mv_global_tlb_purge_t ia64_global_tlb_purge;
-extern ia64_mv_dma_get_required_mask ia64_dma_get_required_mask;
extern ia64_mv_irq_to_vector __ia64_irq_to_vector;
extern ia64_mv_local_vector_to_irq __ia64_local_vector_to_irq;
extern ia64_mv_pci_get_legacy_mem_t ia64_pci_get_legacy_mem;
diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c
index 291a582..79da21b 100644
--- a/arch/ia64/pci/pci.c
+++ b/arch/ia64/pci/pci.c
@@ -791,26 +791,6 @@ static void __init set_pci_dfl_cacheline_size(void)
pci_dfl_cache_line_size = (1 << cci.pcci_line_size) / 4;
}
-u64 ia64_dma_get_required_mask(struct device *dev)
-{
- u32 low_totalram = ((max_pfn - 1) << PAGE_SHIFT);
- u32 high_totalram = ((max_pfn - 1) >> (32 - PAGE_SHIFT));
- u64 mask;
-
- if (!high_totalram) {
- /* convert to mask just covering totalram */
- low_totalram = (1 << (fls(low_totalram) - 1));
- low_totalram += low_totalram - 1;
- mask = low_totalram;
- } else {
- high_totalram = (1 << (fls(high_totalram) - 1));
- high_totalram += high_totalram - 1;
- mask = (((u64)high_totalram) << 32) + 0xffffffff;
- }
- return mask;
-}
-EXPORT_SYMBOL_GPL(ia64_dma_get_required_mask);
-
u64 dma_get_required_mask(struct device *dev)
{
return platform_dma_get_required_mask(dev);
diff --git a/drivers/base/platform.c b/drivers/base/platform.c
index 5b47210..a53ef9d 100644
--- a/drivers/base/platform.c
+++ b/drivers/base/platform.c
@@ -942,8 +942,7 @@ int __init platform_bus_init(void)
return error;
}
-#ifndef ARCH_HAS_DMA_GET_REQUIRED_MASK
-u64 dma_get_required_mask(struct device *dev)
+u64 dma_get_required_mask_from_max_pfn(struct device *dev)
{
u32 low_totalram = ((max_pfn - 1) << PAGE_SHIFT);
u32 high_totalram = ((max_pfn - 1) >> (32 - PAGE_SHIFT));
@@ -961,6 +960,13 @@ u64 dma_get_required_mask(struct device *dev)
}
return mask;
}
+EXPORT_SYMBOL_GPL(dma_get_required_mask_from_max_pfn);
+
+#ifndef ARCH_HAS_DMA_GET_REQUIRED_MASK
+u64 dma_get_required_mask(struct device *dev)
+{
+ return dma_get_required_mask_from_max_pfn(dev);
+}
EXPORT_SYMBOL_GPL(dma_get_required_mask);
#endif
diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index fd4aee2..bc6e2b6 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -121,6 +121,7 @@ static inline int dma_coerce_mask_and_coherent(struct device *dev, u64 mask)
return dma_set_mask_and_coherent(dev, mask);
}
+extern u64 dma_get_required_mask_from_max_pfn(struct device *dev);
extern u64 dma_get_required_mask(struct device *dev);
static inline unsigned int dma_get_max_seg_size(struct device *dev)
--
1.7.10.4
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH 1/2] dma: add dma_get_required_mask_from_max_pfn()
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
@ 2014-05-12 10:15 ` David Vrabel
2014-05-12 10:15 ` David Vrabel
` (4 subsequent siblings)
5 siblings, 0 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Fenghua Yu, Tony Luck, linux-ia64, Greg Kroah-Hartman, x86,
Ingo Molnar, David Vrabel, H. Peter Anvin, xen-devel,
Thomas Gleixner, Boris Ostrovsky
ia64 provides a duplicate of the generic dma_get_required_mask()
because it has ARCH_HAS_GET_REQUIRED_MASK. Provide a common
dma_get_require_mask_max_pfn() instead.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: linux-ia64@vger.kernel.org
---
arch/ia64/include/asm/machvec.h | 2 +-
arch/ia64/include/asm/machvec_init.h | 1 -
arch/ia64/pci/pci.c | 20 --------------------
drivers/base/platform.c | 10 ++++++++--
include/linux/dma-mapping.h | 1 +
5 files changed, 10 insertions(+), 24 deletions(-)
diff --git a/arch/ia64/include/asm/machvec.h b/arch/ia64/include/asm/machvec.h
index 9c39bdf..beaa47d 100644
--- a/arch/ia64/include/asm/machvec.h
+++ b/arch/ia64/include/asm/machvec.h
@@ -287,7 +287,7 @@ extern struct dma_map_ops *dma_get_ops(struct device *);
# define platform_dma_get_ops dma_get_ops
#endif
#ifndef platform_dma_get_required_mask
-# define platform_dma_get_required_mask ia64_dma_get_required_mask
+# define platform_dma_get_required_mask dma_get_required_mask_from_max_pfn
#endif
#ifndef platform_irq_to_vector
# define platform_irq_to_vector __ia64_irq_to_vector
diff --git a/arch/ia64/include/asm/machvec_init.h b/arch/ia64/include/asm/machvec_init.h
index 37a4698..ef964b2 100644
--- a/arch/ia64/include/asm/machvec_init.h
+++ b/arch/ia64/include/asm/machvec_init.h
@@ -3,7 +3,6 @@
extern ia64_mv_send_ipi_t ia64_send_ipi;
extern ia64_mv_global_tlb_purge_t ia64_global_tlb_purge;
-extern ia64_mv_dma_get_required_mask ia64_dma_get_required_mask;
extern ia64_mv_irq_to_vector __ia64_irq_to_vector;
extern ia64_mv_local_vector_to_irq __ia64_local_vector_to_irq;
extern ia64_mv_pci_get_legacy_mem_t ia64_pci_get_legacy_mem;
diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c
index 291a582..79da21b 100644
--- a/arch/ia64/pci/pci.c
+++ b/arch/ia64/pci/pci.c
@@ -791,26 +791,6 @@ static void __init set_pci_dfl_cacheline_size(void)
pci_dfl_cache_line_size = (1 << cci.pcci_line_size) / 4;
}
-u64 ia64_dma_get_required_mask(struct device *dev)
-{
- u32 low_totalram = ((max_pfn - 1) << PAGE_SHIFT);
- u32 high_totalram = ((max_pfn - 1) >> (32 - PAGE_SHIFT));
- u64 mask;
-
- if (!high_totalram) {
- /* convert to mask just covering totalram */
- low_totalram = (1 << (fls(low_totalram) - 1));
- low_totalram += low_totalram - 1;
- mask = low_totalram;
- } else {
- high_totalram = (1 << (fls(high_totalram) - 1));
- high_totalram += high_totalram - 1;
- mask = (((u64)high_totalram) << 32) + 0xffffffff;
- }
- return mask;
-}
-EXPORT_SYMBOL_GPL(ia64_dma_get_required_mask);
-
u64 dma_get_required_mask(struct device *dev)
{
return platform_dma_get_required_mask(dev);
diff --git a/drivers/base/platform.c b/drivers/base/platform.c
index 5b47210..a53ef9d 100644
--- a/drivers/base/platform.c
+++ b/drivers/base/platform.c
@@ -942,8 +942,7 @@ int __init platform_bus_init(void)
return error;
}
-#ifndef ARCH_HAS_DMA_GET_REQUIRED_MASK
-u64 dma_get_required_mask(struct device *dev)
+u64 dma_get_required_mask_from_max_pfn(struct device *dev)
{
u32 low_totalram = ((max_pfn - 1) << PAGE_SHIFT);
u32 high_totalram = ((max_pfn - 1) >> (32 - PAGE_SHIFT));
@@ -961,6 +960,13 @@ u64 dma_get_required_mask(struct device *dev)
}
return mask;
}
+EXPORT_SYMBOL_GPL(dma_get_required_mask_from_max_pfn);
+
+#ifndef ARCH_HAS_DMA_GET_REQUIRED_MASK
+u64 dma_get_required_mask(struct device *dev)
+{
+ return dma_get_required_mask_from_max_pfn(dev);
+}
EXPORT_SYMBOL_GPL(dma_get_required_mask);
#endif
diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index fd4aee2..bc6e2b6 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -121,6 +121,7 @@ static inline int dma_coerce_mask_and_coherent(struct device *dev, u64 mask)
return dma_set_mask_and_coherent(dev, mask);
}
+extern u64 dma_get_required_mask_from_max_pfn(struct device *dev);
extern u64 dma_get_required_mask(struct device *dev);
static inline unsigned int dma_get_max_seg_size(struct device *dev)
--
1.7.10.4
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH 1/2] dma: add dma_get_required_mask_from_max_pfn()
@ 2014-05-12 10:15 ` David Vrabel
0 siblings, 0 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Thomas Gleixner, Ingo Molnar, H. Peter Anvin, x86,
Greg Kroah-Hartman, xen-devel, Konrad Rzeszutek Wilk,
Boris Ostrovsky, David Vrabel, Tony Luck, Fenghua Yu, linux-ia64
ia64 provides a duplicate of the generic dma_get_required_mask()
because it has ARCH_HAS_GET_REQUIRED_MASK. Provide a common
dma_get_require_mask_max_pfn() instead.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: linux-ia64@vger.kernel.org
---
arch/ia64/include/asm/machvec.h | 2 +-
arch/ia64/include/asm/machvec_init.h | 1 -
arch/ia64/pci/pci.c | 20 --------------------
drivers/base/platform.c | 10 ++++++++--
include/linux/dma-mapping.h | 1 +
5 files changed, 10 insertions(+), 24 deletions(-)
diff --git a/arch/ia64/include/asm/machvec.h b/arch/ia64/include/asm/machvec.h
index 9c39bdf..beaa47d 100644
--- a/arch/ia64/include/asm/machvec.h
+++ b/arch/ia64/include/asm/machvec.h
@@ -287,7 +287,7 @@ extern struct dma_map_ops *dma_get_ops(struct device *);
# define platform_dma_get_ops dma_get_ops
#endif
#ifndef platform_dma_get_required_mask
-# define platform_dma_get_required_mask ia64_dma_get_required_mask
+# define platform_dma_get_required_mask dma_get_required_mask_from_max_pfn
#endif
#ifndef platform_irq_to_vector
# define platform_irq_to_vector __ia64_irq_to_vector
diff --git a/arch/ia64/include/asm/machvec_init.h b/arch/ia64/include/asm/machvec_init.h
index 37a4698..ef964b2 100644
--- a/arch/ia64/include/asm/machvec_init.h
+++ b/arch/ia64/include/asm/machvec_init.h
@@ -3,7 +3,6 @@
extern ia64_mv_send_ipi_t ia64_send_ipi;
extern ia64_mv_global_tlb_purge_t ia64_global_tlb_purge;
-extern ia64_mv_dma_get_required_mask ia64_dma_get_required_mask;
extern ia64_mv_irq_to_vector __ia64_irq_to_vector;
extern ia64_mv_local_vector_to_irq __ia64_local_vector_to_irq;
extern ia64_mv_pci_get_legacy_mem_t ia64_pci_get_legacy_mem;
diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c
index 291a582..79da21b 100644
--- a/arch/ia64/pci/pci.c
+++ b/arch/ia64/pci/pci.c
@@ -791,26 +791,6 @@ static void __init set_pci_dfl_cacheline_size(void)
pci_dfl_cache_line_size = (1 << cci.pcci_line_size) / 4;
}
-u64 ia64_dma_get_required_mask(struct device *dev)
-{
- u32 low_totalram = ((max_pfn - 1) << PAGE_SHIFT);
- u32 high_totalram = ((max_pfn - 1) >> (32 - PAGE_SHIFT));
- u64 mask;
-
- if (!high_totalram) {
- /* convert to mask just covering totalram */
- low_totalram = (1 << (fls(low_totalram) - 1));
- low_totalram += low_totalram - 1;
- mask = low_totalram;
- } else {
- high_totalram = (1 << (fls(high_totalram) - 1));
- high_totalram += high_totalram - 1;
- mask = (((u64)high_totalram) << 32) + 0xffffffff;
- }
- return mask;
-}
-EXPORT_SYMBOL_GPL(ia64_dma_get_required_mask);
-
u64 dma_get_required_mask(struct device *dev)
{
return platform_dma_get_required_mask(dev);
diff --git a/drivers/base/platform.c b/drivers/base/platform.c
index 5b47210..a53ef9d 100644
--- a/drivers/base/platform.c
+++ b/drivers/base/platform.c
@@ -942,8 +942,7 @@ int __init platform_bus_init(void)
return error;
}
-#ifndef ARCH_HAS_DMA_GET_REQUIRED_MASK
-u64 dma_get_required_mask(struct device *dev)
+u64 dma_get_required_mask_from_max_pfn(struct device *dev)
{
u32 low_totalram = ((max_pfn - 1) << PAGE_SHIFT);
u32 high_totalram = ((max_pfn - 1) >> (32 - PAGE_SHIFT));
@@ -961,6 +960,13 @@ u64 dma_get_required_mask(struct device *dev)
}
return mask;
}
+EXPORT_SYMBOL_GPL(dma_get_required_mask_from_max_pfn);
+
+#ifndef ARCH_HAS_DMA_GET_REQUIRED_MASK
+u64 dma_get_required_mask(struct device *dev)
+{
+ return dma_get_required_mask_from_max_pfn(dev);
+}
EXPORT_SYMBOL_GPL(dma_get_required_mask);
#endif
diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index fd4aee2..bc6e2b6 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -121,6 +121,7 @@ static inline int dma_coerce_mask_and_coherent(struct device *dev, u64 mask)
return dma_set_mask_and_coherent(dev, mask);
}
+extern u64 dma_get_required_mask_from_max_pfn(struct device *dev);
extern u64 dma_get_required_mask(struct device *dev);
static inline unsigned int dma_get_max_seg_size(struct device *dev)
--
1.7.10.4
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH 2/2] x86,xen: correct dma_get_required_mask() for Xen PV guests
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
2014-05-12 10:15 ` [PATCH 1/2] dma: add dma_get_required_mask_from_max_pfn() David Vrabel
2014-05-12 10:15 ` David Vrabel
@ 2014-05-12 10:15 ` David Vrabel
2014-05-12 10:15 ` [PATCH 2/2] x86, xen: " David Vrabel
` (2 subsequent siblings)
5 siblings, 0 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Thomas Gleixner, Ingo Molnar, H. Peter Anvin, x86,
Greg Kroah-Hartman, xen-devel, Konrad Rzeszutek Wilk,
Boris Ostrovsky, David Vrabel
On systems where DMA addresses and physical addresses are not 1:1
(such as Xen PV guests), the generic dma_get_required_mask() will not
return the correct mask (since it uses max_pfn).
Some device drivers (such as mptsas, mpt2sas) use
dma_get_required_mask() to set device DMA masks to allow them to use
only 32-bit DMA addresses in hardware structures. This results in
unnecessary use of the SWIOTLB if DMA addresses are more than 32-bits,
impacting performance significantly.
Provide an arch-specific dma_get_required_mask() that defaults to the
generic dma_get_required_mask_from_pfn().
Under Xen, the required DMA mask can then be set to always 64-bits.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
arch/x86/include/asm/device.h | 2 ++
arch/x86/kernel/pci-dma.c | 7 +++++++
arch/x86/xen/pci-swiotlb-xen.c | 1 +
drivers/xen/swiotlb-xen.c | 7 +++++++
include/xen/swiotlb-xen.h | 4 ++++
5 files changed, 21 insertions(+)
diff --git a/arch/x86/include/asm/device.h b/arch/x86/include/asm/device.h
index 03dd729..10bc628 100644
--- a/arch/x86/include/asm/device.h
+++ b/arch/x86/include/asm/device.h
@@ -13,4 +13,6 @@ struct dev_archdata {
struct pdev_archdata {
};
+#define ARCH_HAS_DMA_GET_REQUIRED_MASK
+
#endif /* _ASM_X86_DEVICE_H */
diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
index f7d0672..ad2c3e2 100644
--- a/arch/x86/kernel/pci-dma.c
+++ b/arch/x86/kernel/pci-dma.c
@@ -135,6 +135,13 @@ void dma_generic_free_coherent(struct device *dev, size_t size, void *vaddr,
free_pages((unsigned long)vaddr, get_order(size));
}
+u64 dma_get_required_mask(struct device *dev)
+{
+ if (dma_ops->get_required_mask)
+ return dma_ops->get_required_mask(dev);
+ return dma_get_required_mask_from_max_pfn(dev);
+}
+
/*
* See <Documentation/x86/x86_64/boot-options.txt> for the iommu kernel
* parameter documentation.
diff --git a/arch/x86/xen/pci-swiotlb-xen.c b/arch/x86/xen/pci-swiotlb-xen.c
index 0e98e5d..a5d180a 100644
--- a/arch/x86/xen/pci-swiotlb-xen.c
+++ b/arch/x86/xen/pci-swiotlb-xen.c
@@ -31,6 +31,7 @@ static struct dma_map_ops xen_swiotlb_dma_ops = {
.map_page = xen_swiotlb_map_page,
.unmap_page = xen_swiotlb_unmap_page,
.dma_supported = xen_swiotlb_dma_supported,
+ .get_required_mask = xen_swiotlb_get_required_mask,
};
/*
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index ebd8f21..798f62b 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -683,3 +683,10 @@ xen_swiotlb_set_dma_mask(struct device *dev, u64 dma_mask)
return 0;
}
EXPORT_SYMBOL_GPL(xen_swiotlb_set_dma_mask);
+
+u64
+xen_swiotlb_get_required_mask(struct device *dev)
+{
+ return DMA_BIT_MASK(64);
+}
+EXPORT_SYMBOL_GPL(xen_swiotlb_get_required_mask);
diff --git a/include/xen/swiotlb-xen.h b/include/xen/swiotlb-xen.h
index 8b2eb93..6408888 100644
--- a/include/xen/swiotlb-xen.h
+++ b/include/xen/swiotlb-xen.h
@@ -58,4 +58,8 @@ xen_swiotlb_dma_supported(struct device *hwdev, u64 mask);
extern int
xen_swiotlb_set_dma_mask(struct device *dev, u64 dma_mask);
+
+extern u64
+xen_swiotlb_get_required_mask(struct device *dev);
+
#endif /* __LINUX_SWIOTLB_XEN_H */
--
1.7.10.4
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH 2/2] x86, xen: correct dma_get_required_mask() for Xen PV guests
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
` (2 preceding siblings ...)
2014-05-12 10:15 ` [PATCH 2/2] x86,xen: correct dma_get_required_mask() for Xen PV guests David Vrabel
@ 2014-05-12 10:15 ` David Vrabel
2014-05-12 15:28 ` [PATCH 0/2] dma, x86, xen: fix dma_get_required_mask() when dma != phys address Konrad Rzeszutek Wilk
2014-05-12 15:28 ` [PATCH 0/2] dma,x86,xen: " Konrad Rzeszutek Wilk
5 siblings, 0 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Greg Kroah-Hartman, x86, Ingo Molnar, David Vrabel,
H. Peter Anvin, xen-devel, Thomas Gleixner, Boris Ostrovsky
On systems where DMA addresses and physical addresses are not 1:1
(such as Xen PV guests), the generic dma_get_required_mask() will not
return the correct mask (since it uses max_pfn).
Some device drivers (such as mptsas, mpt2sas) use
dma_get_required_mask() to set device DMA masks to allow them to use
only 32-bit DMA addresses in hardware structures. This results in
unnecessary use of the SWIOTLB if DMA addresses are more than 32-bits,
impacting performance significantly.
Provide an arch-specific dma_get_required_mask() that defaults to the
generic dma_get_required_mask_from_pfn().
Under Xen, the required DMA mask can then be set to always 64-bits.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
arch/x86/include/asm/device.h | 2 ++
arch/x86/kernel/pci-dma.c | 7 +++++++
arch/x86/xen/pci-swiotlb-xen.c | 1 +
drivers/xen/swiotlb-xen.c | 7 +++++++
include/xen/swiotlb-xen.h | 4 ++++
5 files changed, 21 insertions(+)
diff --git a/arch/x86/include/asm/device.h b/arch/x86/include/asm/device.h
index 03dd729..10bc628 100644
--- a/arch/x86/include/asm/device.h
+++ b/arch/x86/include/asm/device.h
@@ -13,4 +13,6 @@ struct dev_archdata {
struct pdev_archdata {
};
+#define ARCH_HAS_DMA_GET_REQUIRED_MASK
+
#endif /* _ASM_X86_DEVICE_H */
diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
index f7d0672..ad2c3e2 100644
--- a/arch/x86/kernel/pci-dma.c
+++ b/arch/x86/kernel/pci-dma.c
@@ -135,6 +135,13 @@ void dma_generic_free_coherent(struct device *dev, size_t size, void *vaddr,
free_pages((unsigned long)vaddr, get_order(size));
}
+u64 dma_get_required_mask(struct device *dev)
+{
+ if (dma_ops->get_required_mask)
+ return dma_ops->get_required_mask(dev);
+ return dma_get_required_mask_from_max_pfn(dev);
+}
+
/*
* See <Documentation/x86/x86_64/boot-options.txt> for the iommu kernel
* parameter documentation.
diff --git a/arch/x86/xen/pci-swiotlb-xen.c b/arch/x86/xen/pci-swiotlb-xen.c
index 0e98e5d..a5d180a 100644
--- a/arch/x86/xen/pci-swiotlb-xen.c
+++ b/arch/x86/xen/pci-swiotlb-xen.c
@@ -31,6 +31,7 @@ static struct dma_map_ops xen_swiotlb_dma_ops = {
.map_page = xen_swiotlb_map_page,
.unmap_page = xen_swiotlb_unmap_page,
.dma_supported = xen_swiotlb_dma_supported,
+ .get_required_mask = xen_swiotlb_get_required_mask,
};
/*
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index ebd8f21..798f62b 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -683,3 +683,10 @@ xen_swiotlb_set_dma_mask(struct device *dev, u64 dma_mask)
return 0;
}
EXPORT_SYMBOL_GPL(xen_swiotlb_set_dma_mask);
+
+u64
+xen_swiotlb_get_required_mask(struct device *dev)
+{
+ return DMA_BIT_MASK(64);
+}
+EXPORT_SYMBOL_GPL(xen_swiotlb_get_required_mask);
diff --git a/include/xen/swiotlb-xen.h b/include/xen/swiotlb-xen.h
index 8b2eb93..6408888 100644
--- a/include/xen/swiotlb-xen.h
+++ b/include/xen/swiotlb-xen.h
@@ -58,4 +58,8 @@ xen_swiotlb_dma_supported(struct device *hwdev, u64 mask);
extern int
xen_swiotlb_set_dma_mask(struct device *dev, u64 dma_mask);
+
+extern u64
+xen_swiotlb_get_required_mask(struct device *dev);
+
#endif /* __LINUX_SWIOTLB_XEN_H */
--
1.7.10.4
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
` (4 preceding siblings ...)
2014-05-12 15:28 ` [PATCH 0/2] dma, x86, xen: fix dma_get_required_mask() when dma != phys address Konrad Rzeszutek Wilk
@ 2014-05-12 15:28 ` Konrad Rzeszutek Wilk
5 siblings, 0 replies; 9+ messages in thread
From: Konrad Rzeszutek Wilk @ 2014-05-12 15:28 UTC (permalink / raw)
To: David Vrabel
Cc: linux-kernel, Thomas Gleixner, Ingo Molnar, H. Peter Anvin, x86,
Greg Kroah-Hartman, xen-devel, Boris Ostrovsky
On Mon, May 12, 2014 at 11:15:40AM +0100, David Vrabel wrote:
> x86 uses the generic dma_get_required_mask() which assumes that DMA
> address == physical address. For Xen PV guests this is not the case
> and certain drivers will cause excessive use of the SWIOTLB.
>
> Patch #1 is a straight forward refactoring of a common function.
>
> Patch #2 addes the required arch-specific implementation of
> dma_get_required_mask() to x86.
David,
Thank you for taking a look at that. If you would like, you can
tack on Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
And with the *SWIOTLB maintainer hat on - you can add that tag
as well.
Thanks!
>
> David
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 0/2] dma, x86, xen: fix dma_get_required_mask() when dma != phys address
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
` (3 preceding siblings ...)
2014-05-12 10:15 ` [PATCH 2/2] x86, xen: " David Vrabel
@ 2014-05-12 15:28 ` Konrad Rzeszutek Wilk
2014-05-12 15:28 ` [PATCH 0/2] dma,x86,xen: " Konrad Rzeszutek Wilk
5 siblings, 0 replies; 9+ messages in thread
From: Konrad Rzeszutek Wilk @ 2014-05-12 15:28 UTC (permalink / raw)
To: David Vrabel
Cc: Greg Kroah-Hartman, x86, linux-kernel, Ingo Molnar,
H. Peter Anvin, xen-devel, Thomas Gleixner, Boris Ostrovsky
On Mon, May 12, 2014 at 11:15:40AM +0100, David Vrabel wrote:
> x86 uses the generic dma_get_required_mask() which assumes that DMA
> address == physical address. For Xen PV guests this is not the case
> and certain drivers will cause excessive use of the SWIOTLB.
>
> Patch #1 is a straight forward refactoring of a common function.
>
> Patch #2 addes the required arch-specific implementation of
> dma_get_required_mask() to x86.
David,
Thank you for taking a look at that. If you would like, you can
tack on Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
And with the *SWIOTLB maintainer hat on - you can add that tag
as well.
Thanks!
>
> David
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 0/2] dma, x86, xen: fix dma_get_required_mask() when dma != phys address
@ 2014-05-12 10:15 David Vrabel
0 siblings, 0 replies; 9+ messages in thread
From: David Vrabel @ 2014-05-12 10:15 UTC (permalink / raw)
To: linux-kernel
Cc: Greg Kroah-Hartman, x86, Ingo Molnar, David Vrabel,
H. Peter Anvin, xen-devel, Thomas Gleixner, Boris Ostrovsky
x86 uses the generic dma_get_required_mask() which assumes that DMA
address == physical address. For Xen PV guests this is not the case
and certain drivers will cause excessive use of the SWIOTLB.
Patch #1 is a straight forward refactoring of a common function.
Patch #2 addes the required arch-specific implementation of
dma_get_required_mask() to x86.
David
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2014-05-12 15:29 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-05-12 10:15 [PATCH 0/2] dma,x86,xen: fix dma_get_required_mask() when dma != phys address David Vrabel
2014-05-12 10:15 ` [PATCH 1/2] dma: add dma_get_required_mask_from_max_pfn() David Vrabel
2014-05-12 10:15 ` David Vrabel
2014-05-12 10:15 ` David Vrabel
2014-05-12 10:15 ` [PATCH 2/2] x86,xen: correct dma_get_required_mask() for Xen PV guests David Vrabel
2014-05-12 10:15 ` [PATCH 2/2] x86, xen: " David Vrabel
2014-05-12 15:28 ` [PATCH 0/2] dma, x86, xen: fix dma_get_required_mask() when dma != phys address Konrad Rzeszutek Wilk
2014-05-12 15:28 ` [PATCH 0/2] dma,x86,xen: " Konrad Rzeszutek Wilk
-- strict thread matches above, loose matches on Subject: below --
2014-05-12 10:15 [PATCH 0/2] dma, x86, xen: " David Vrabel
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.