Re: [PATCH 2/2] arm/xen: Don't use xen DMA ops when the device is protected by an IOMMU

From: Julien Grall
Date: Fri Mar 14 2014 - 12:50:36 EST


On 02/24/2014 08:49 PM, Stefano Stabellini wrote:
> On Mon, 24 Feb 2014, gregkh@xxxxxxxxxxxxxxxxxxx wrote:
> Julien is proposing to store the list of "safe" devices on an hash table
> in the Xen specific code (in arch/arm/xen/enlighten.c, see
> http://marc.info/?l=linux-kernel&m=139291370526082&w=2).
> Whenever Linux is about to do DMA, we would check in the hashtable to
> figure out whether we need to go through the swiotlb or we can simply
> use the native dma_ops.
>
> Ian and I were thinking that it would be much easier and faster to have
> a "xen_safe_device" parameter in struct device and just check for that.
> It doesn't actually need to be in struct device, it could simply be a
> flag in struct device_dma_parameters as Ian was suggesting.
>
> Julien, could you please come up with a simple patch to demonstrate the
> concept?

Hello Stefano and Greg,

Sorry for the late answer. I wrote a simple patch which depend on patch #1.
Let me know if it's the right direction.

Regards,

commit ca55e82bc191678b284792d2f0d200fa1ce08e16
Author: Julien Grall <julien.grall@xxxxxxxxxx>
Date: Fri Mar 14 16:27:01 2014 +0000

ARM: platform_device: dev_archdata: Add xen specific boolean

Until now, every DMA-capable devices are using specific Xen DMA ops when Linux
is running as DOM0. These DMA ops call swiotlb-xen to bounce buffer.

With the support of IOMMU drivers in Xen, every device protected by IOMMU
must not use swiotlb DMA ops.

This patch introduces a boolean in dev_archdata to indicate if the device
can safely use its own DMA ops or swiotlb ops.

Signed-off-by: Julien Grall <julien.grall@xxxxxxxxxx>

diff --git a/arch/arm/include/asm/device.h b/arch/arm/include/asm/device.h
index dc662fc..345a96e 100644
--- a/arch/arm/include/asm/device.h
+++ b/arch/arm/include/asm/device.h
@@ -17,6 +17,9 @@ struct dev_archdata {
#ifdef CONFIG_ARM_DMA_USE_IOMMU
struct dma_iommu_mapping *mapping;
#endif
+#ifdef CONFIG_XEN
+ bool is_protected;
+#endif
};

struct omap_device;
diff --git a/arch/arm/include/asm/xen/dma-mapping.h b/arch/arm/include/asm/xen/dma-mapping.h
index 002fc57..d6cc012 100644
--- a/arch/arm/include/asm/xen/dma-mapping.h
+++ b/arch/arm/include/asm/xen/dma-mapping.h
@@ -5,9 +5,21 @@

extern struct dma_map_ops *xen_dma_ops;

+#ifdef CONFIG_XEN
+static inline bool xen_is_protected_device(const struct device *dev)
+{
+ return dev->archdata.is_protected;
+}
+#else
+static inline bool xen_is_protected_device(const struct device *dev)
+{
+ return 0;
+}
+#endif
+
static inline bool need_xen_dma_ops(struct device *dev)
{
- return xen_initial_domain();
+ return xen_initial_domain() && !xen_is_protected_device(dev);
}

#endif /* _ASM_ARM_XEN_DMA_MAPPING_H */

--
Julien Grall
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/