[2/7] Initial skeleton of VFIO support for Device Tree based devices

Commit Message

Antonios Motakis Sept. 30, 2013, 3:28 p.m. UTC

Platform devices in the Linux kernel are usually managed by the DT
interface. This patch forms the base to support these kind of devices
with VFIO.

Signed-off-by: Antonios Motakis <[e-mail]>
---
 drivers/vfio/Kconfig         |  11 +++
 drivers/vfio/Makefile        |   1 +
 drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++
 include/uapi/linux/vfio.h    |   1 +
 4 files changed, 200 insertions(+)
 create mode 100644 drivers/vfio/vfio_platform.c

Comments

Bharat Bhushan Sept. 30, 2013, 3:37 p.m. UTC | #1

> -----Original Message-----
> From: [e-mail] [mailto:iommu-
> [e-mail]] On Behalf Of Antonios Motakis
> Sent: Monday, September 30, 2013 8:59 PM
> To: [e-mail]; [e-mail]
> Cc: [e-mail]; [e-mail]; [e-mail]; Yoder
> Stuart-B08248; [e-mail]; Antonios Motakis;
> [e-mail]
> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree based
> devices
> 
> Platform devices in the Linux kernel are usually managed by the DT interface.
> This patch forms the base to support these kind of devices with VFIO.
> 
> Signed-off-by: Antonios Motakis <[e-mail]>
> ---
>  drivers/vfio/Kconfig         |  11 +++
>  drivers/vfio/Makefile        |   1 +
>  drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++
>  include/uapi/linux/vfio.h    |   1 +
>  4 files changed, 200 insertions(+)
>  create mode 100644 drivers/vfio/vfio_platform.c
> 
> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index 1f84eda..35254b7
> 100644
> --- a/drivers/vfio/Kconfig
> +++ b/drivers/vfio/Kconfig
> @@ -13,4 +13,15 @@ menuconfig VFIO
> 
>  	  If you don't know what to do here, say N.
> 
> +config VFIO_PLATFORM
> +	tristate "VFIO support for device tree based platform devices"
> +	depends on VFIO && EVENTFD && OF
> +	help
> +	  Support for platform devices with VFIO. This is required to make
> +	  use of platform devices present on device tree nodes using the VFIO
> +	  framework. Devices that are not described in the device tree cannot
> +	  be used by this driver.
> +
> +	  If you don't know what to do here, say N.
> +
>  source "drivers/vfio/pci/Kconfig"
> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index
> 2398d4a..575c8dd 100644
> --- a/drivers/vfio/Makefile
> +++ b/drivers/vfio/Makefile
> @@ -1,3 +1,4 @@
>  obj-$(CONFIG_VFIO) += vfio.o
>  obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
>  obj-$(CONFIG_VFIO_PCI) += pci/
> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
> diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c new

We can make this parallel to PCI, something like drivers/vfio/platform/platform.c

-Bharat

> file mode 100644 index 0000000..b9686b0
> --- /dev/null
> +++ b/drivers/vfio/vfio_platform.c
> @@ -0,0 +1,187 @@
> +/*
> + * Copyright (C) 2013 - Virtual Open Systems
> + * Author: Antonios Motakis <[e-mail]>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License, version 2, as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
> + */
> +
> +#include <linux/device.h>
> +#include <linux/eventfd.h>
> +#include <linux/interrupt.h>
> +#include <linux/iommu.h>
> +#include <linux/module.h>
> +#include <linux/mutex.h>
> +#include <linux/notifier.h>
> +#include <linux/pm_runtime.h>
> +#include <linux/slab.h>
> +#include <linux/types.h>
> +#include <linux/uaccess.h>
> +#include <linux/vfio.h>
> +
> +#define DRIVER_VERSION  "0.1"
> +#define DRIVER_AUTHOR   "Antonios Motakis <[e-mail]>"
> +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
> +
> +struct vfio_platform_device {
> +	struct platform_device	*pdev;
> +};
> +
> +static void vfio_platform_release(void *device_data) {
> +	module_put(THIS_MODULE);
> +}
> +
> +static int vfio_platform_open(void *device_data) {
> +	if (!try_module_get(THIS_MODULE))
> +		return -ENODEV;
> +
> +	return 0;
> +}
> +
> +static long vfio_platform_ioctl(void *device_data,
> +			   unsigned int cmd, unsigned long arg) {
> +	struct vfio_platform_device *vdev = device_data;
> +	unsigned long minsz;
> +
> +	if (cmd == VFIO_DEVICE_GET_INFO) {
> +		struct vfio_device_info info;
> +
> +		minsz = offsetofend(struct vfio_device_info, num_irqs);
> +
> +		if (copy_from_user(&info, (void __user *)arg, minsz))
> +			return -EFAULT;
> +
> +		if (info.argsz < minsz)
> +			return -EINVAL;
> +
> +		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
> +		info.num_regions = 0;
> +		info.num_irqs = 0;
> +
> +		return copy_to_user((void __user *)arg, &info, minsz);
> +
> +	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
> +		return -EINVAL;
> +
> +	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
> +		return -EINVAL;
> +
> +	else if (cmd == VFIO_DEVICE_SET_IRQS)
> +		return -EINVAL;
> +
> +	else if (cmd == VFIO_DEVICE_RESET)
> +		return -EINVAL;
> +
> +	return -ENOTTY;
> +}
> +
> +static ssize_t vfio_platform_read(void *device_data, char __user *buf,
> +			     size_t count, loff_t *ppos)
> +{
> +	return 0;
> +}
> +
> +static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
> +			      size_t count, loff_t *ppos)
> +{
> +	return 0;
> +}
> +
> +static int vfio_platform_mmap(void *device_data, struct vm_area_struct
> +*vma) {
> +	return -EINVAL;
> +}
> +
> +static const struct vfio_device_ops vfio_platform_ops = {
> +	.name		= "vfio-platform",
> +	.open		= vfio_platform_open,
> +	.release	= vfio_platform_release,
> +	.ioctl		= vfio_platform_ioctl,
> +	.read		= vfio_platform_read,
> +	.write		= vfio_platform_write,
> +	.mmap		= vfio_platform_mmap,
> +};
> +
> +static int vfio_platform_probe(struct platform_device *pdev) {
> +	struct vfio_platform_device *vdev;
> +	struct iommu_group *group;
> +	int ret;
> +
> +	group = iommu_group_get(&pdev->dev);
> +	if (!group) {
> +		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
> +		return -EINVAL;
> +	}
> +
> +	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
> +	if (!vdev) {
> +		iommu_group_put(group);
> +		return -ENOMEM;
> +	}
> +
> +	vdev->pdev = pdev;
> +
> +	ret = vfio_add_group_dev(&pdev->dev, &vfio_platform_ops, vdev);
> +	if (ret) {
> +		iommu_group_put(group);
> +		kfree(vdev);
> +	}
> +
> +	return ret;
> +}
> +
> +static int vfio_platform_remove(struct platform_device *pdev) {
> +	struct vfio_platform_device *vdev;
> +
> +	vdev = vfio_del_group_dev(&pdev->dev);
> +	if (!vdev)
> +		return -EINVAL;
> +
> +	iommu_group_put(pdev->dev.iommu_group);
> +	kfree(vdev);
> +
> +	return 0;
> +}
> +
> +static const struct of_device_id vfio_platform_match[] = {
> +	/* Before this series can be applied, we shall implement a better
> +	 * mechanism to bind the module to any device. For now add the
> +	 * compatible property to the dtb of the devices we want to use. */
> +	{
> +		.compatible = "vfio",
> +	},
> +	{},
> +};
> +MODULE_DEVICE_TABLE(of, vfio_platform_match);
> +
> +static struct platform_driver vfio_platform_driver = {
> +	.probe		= vfio_platform_probe,
> +	.remove		= vfio_platform_remove,
> +	.driver	= {
> +		.name	= "vfio-platform",
> +		.owner	= THIS_MODULE,
> +		.of_match_table = vfio_platform_match,
> +	},
> +};
> +
> +module_platform_driver(vfio_platform_driver);
> +
> +MODULE_VERSION(DRIVER_VERSION);
> +MODULE_LICENSE("GPL v2");
> +MODULE_AUTHOR(DRIVER_AUTHOR);
> +MODULE_DESCRIPTION(DRIVER_DESC);
> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h index
> 284ff24..8d7434d 100644
> --- a/include/uapi/linux/vfio.h
> +++ b/include/uapi/linux/vfio.h
> @@ -147,6 +147,7 @@ struct vfio_device_info {
>  	__u32	flags;
>  #define VFIO_DEVICE_FLAGS_RESET	(1 << 0)	/* Device supports reset */
>  #define VFIO_DEVICE_FLAGS_PCI	(1 << 1)	/* vfio-pci device */
> +#define VFIO_DEVICE_FLAGS_PLATFORM (1 << 2)	/* vfio-platform device */
>  	__u32	num_regions;	/* Max region index + 1 */
>  	__u32	num_irqs;	/* Max IRQ index + 1 */
>  };
> --
> 1.8.1.2
> 
> _______________________________________________
> iommu mailing list
> [e-mail]
> https://lists.linuxfoundation.org/mailman/listinfo/iommu


--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Alex Williamson Sept. 30, 2013, 7:41 p.m. UTC | #2

On Mon, 2013-09-30 at 17:28 +0200, Antonios Motakis wrote:
> Platform devices in the Linux kernel are usually managed by the DT
> interface. This patch forms the base to support these kind of devices
> with VFIO.
> 
> Signed-off-by: Antonios Motakis <[e-mail]>
> ---
>  drivers/vfio/Kconfig         |  11 +++
>  drivers/vfio/Makefile        |   1 +
>  drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++

I agree with Bharat, I'd probably make a platform sub-directory.  Like
pci, you'll probably have several support files by the end and it seems
like platform devices may need device specific handling in places, which
would also be cleaner to have in the directory.

>  include/uapi/linux/vfio.h    |   1 +
>  4 files changed, 200 insertions(+)
>  create mode 100644 drivers/vfio/vfio_platform.c
> 
> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig
> index 1f84eda..35254b7 100644
> --- a/drivers/vfio/Kconfig
> +++ b/drivers/vfio/Kconfig
> @@ -13,4 +13,15 @@ menuconfig VFIO
>  
>  	  If you don't know what to do here, say N.
>  
> +config VFIO_PLATFORM
> +	tristate "VFIO support for device tree based platform devices"
> +	depends on VFIO && EVENTFD && OF
> +	help
> +	  Support for platform devices with VFIO. This is required to make
> +	  use of platform devices present on device tree nodes using the VFIO
> +	  framework. Devices that are not described in the device tree cannot
> +	  be used by this driver.
> +
> +	  If you don't know what to do here, say N.
> +
>  source "drivers/vfio/pci/Kconfig"
> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile
> index 2398d4a..575c8dd 100644
> --- a/drivers/vfio/Makefile
> +++ b/drivers/vfio/Makefile
> @@ -1,3 +1,4 @@
>  obj-$(CONFIG_VFIO) += vfio.o
>  obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
>  obj-$(CONFIG_VFIO_PCI) += pci/
> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
> diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c
> new file mode 100644
> index 0000000..b9686b0
> --- /dev/null
> +++ b/drivers/vfio/vfio_platform.c
> @@ -0,0 +1,187 @@
> +/*
> + * Copyright (C) 2013 - Virtual Open Systems
> + * Author: Antonios Motakis <[e-mail]>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License, version 2, as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
> + */
> +
> +#include <linux/device.h>
> +#include <linux/eventfd.h>
> +#include <linux/interrupt.h>
> +#include <linux/iommu.h>
> +#include <linux/module.h>
> +#include <linux/mutex.h>
> +#include <linux/notifier.h>
> +#include <linux/pm_runtime.h>
> +#include <linux/slab.h>
> +#include <linux/types.h>
> +#include <linux/uaccess.h>
> +#include <linux/vfio.h>
> +
> +#define DRIVER_VERSION  "0.1"
> +#define DRIVER_AUTHOR   "Antonios Motakis <[e-mail]>"
> +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
> +
> +struct vfio_platform_device {
> +	struct platform_device	*pdev;
> +};
> +
> +static void vfio_platform_release(void *device_data)
> +{
> +	module_put(THIS_MODULE);
> +}
> +
> +static int vfio_platform_open(void *device_data)
> +{
> +	if (!try_module_get(THIS_MODULE))
> +		return -ENODEV;
> +
> +	return 0;
> +}
> +
> +static long vfio_platform_ioctl(void *device_data,
> +			   unsigned int cmd, unsigned long arg)
> +{
> +	struct vfio_platform_device *vdev = device_data;
> +	unsigned long minsz;
> +
> +	if (cmd == VFIO_DEVICE_GET_INFO) {
> +		struct vfio_device_info info;
> +
> +		minsz = offsetofend(struct vfio_device_info, num_irqs);
> +
> +		if (copy_from_user(&info, (void __user *)arg, minsz))
> +			return -EFAULT;
> +
> +		if (info.argsz < minsz)
> +			return -EINVAL;
> +
> +		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
> +		info.num_regions = 0;
> +		info.num_irqs = 0;
> +
> +		return copy_to_user((void __user *)arg, &info, minsz);
> +
> +	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
> +		return -EINVAL;
> +
> +	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
> +		return -EINVAL;
> +
> +	else if (cmd == VFIO_DEVICE_SET_IRQS)
> +		return -EINVAL;
> +
> +	else if (cmd == VFIO_DEVICE_RESET)
> +		return -EINVAL;
> +
> +	return -ENOTTY;
> +}
> +
> +static ssize_t vfio_platform_read(void *device_data, char __user *buf,
> +			     size_t count, loff_t *ppos)
> +{
> +	return 0;
> +}
> +
> +static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
> +			      size_t count, loff_t *ppos)
> +{
> +	return 0;
> +}
> +
> +static int vfio_platform_mmap(void *device_data, struct vm_area_struct *vma)
> +{
> +	return -EINVAL;
> +}
> +
> +static const struct vfio_device_ops vfio_platform_ops = {
> +	.name		= "vfio-platform",
> +	.open		= vfio_platform_open,
> +	.release	= vfio_platform_release,
> +	.ioctl		= vfio_platform_ioctl,
> +	.read		= vfio_platform_read,
> +	.write		= vfio_platform_write,
> +	.mmap		= vfio_platform_mmap,
> +};
> +
> +static int vfio_platform_probe(struct platform_device *pdev)
> +{
> +	struct vfio_platform_device *vdev;
> +	struct iommu_group *group;
> +	int ret;
> +
> +	group = iommu_group_get(&pdev->dev);
> +	if (!group) {
> +		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
> +		return -EINVAL;
> +	}
> +
> +	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
> +	if (!vdev) {
> +		iommu_group_put(group);
> +		return -ENOMEM;
> +	}
> +
> +	vdev->pdev = pdev;
> +
> +	ret = vfio_add_group_dev(&pdev->dev, &vfio_platform_ops, vdev);
> +	if (ret) {
> +		iommu_group_put(group);
> +		kfree(vdev);
> +	}
> +
> +	return ret;
> +}
> +
> +static int vfio_platform_remove(struct platform_device *pdev)
> +{
> +	struct vfio_platform_device *vdev;
> +
> +	vdev = vfio_del_group_dev(&pdev->dev);
> +	if (!vdev)
> +		return -EINVAL;
> +
> +	iommu_group_put(pdev->dev.iommu_group);
> +	kfree(vdev);
> +
> +	return 0;
> +}
> +
> +static const struct of_device_id vfio_platform_match[] = {
> +	/* Before this series can be applied, we shall implement a better
> +	 * mechanism to bind the module to any device. For now add the
> +	 * compatible property to the dtb of the devices we want to use. */
> +	{
> +		.compatible = "vfio",
> +	},
> +	{},
> +};
> +MODULE_DEVICE_TABLE(of, vfio_platform_match);
> +
> +static struct platform_driver vfio_platform_driver = {
> +	.probe		= vfio_platform_probe,
> +	.remove		= vfio_platform_remove,
> +	.driver	= {
> +		.name	= "vfio-platform",
> +		.owner	= THIS_MODULE,
> +		.of_match_table = vfio_platform_match,
> +	},
> +};
> +
> +module_platform_driver(vfio_platform_driver);
> +
> +MODULE_VERSION(DRIVER_VERSION);
> +MODULE_LICENSE("GPL v2");
> +MODULE_AUTHOR(DRIVER_AUTHOR);
> +MODULE_DESCRIPTION(DRIVER_DESC);
> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
> index 284ff24..8d7434d 100644
> --- a/include/uapi/linux/vfio.h
> +++ b/include/uapi/linux/vfio.h
> @@ -147,6 +147,7 @@ struct vfio_device_info {
>  	__u32	flags;
>  #define VFIO_DEVICE_FLAGS_RESET	(1 << 0)	/* Device supports reset */
>  #define VFIO_DEVICE_FLAGS_PCI	(1 << 1)	/* vfio-pci device */
> +#define VFIO_DEVICE_FLAGS_PLATFORM (1 << 2)	/* vfio-platform device */
>  	__u32	num_regions;	/* Max region index + 1 */
>  	__u32	num_irqs;	/* Max IRQ index + 1 */
>  };



--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Antonios Motakis Oct. 1, 2013, 9:36 a.m. UTC | #3

On Mon, Sep 30, 2013 at 9:41 PM, Alex Williamson
<[e-mail]> wrote:
>
> On Mon, 2013-09-30 at 17:28 +0200, Antonios Motakis wrote:
> > Platform devices in the Linux kernel are usually managed by the DT
> > interface. This patch forms the base to support these kind of devices
> > with VFIO.
> >
> > Signed-off-by: Antonios Motakis <[e-mail]>
> > ---
> >  drivers/vfio/Kconfig         |  11 +++
> >  drivers/vfio/Makefile        |   1 +
> >  drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++
>
> I agree with Bharat, I'd probably make a platform sub-directory.  Like
> pci, you'll probably have several support files by the end and it seems
> like platform devices may need device specific handling in places, which
> would also be cleaner to have in the directory.
>
> >  include/uapi/linux/vfio.h    |   1 +
> >  4 files changed, 200 insertions(+)
> >  create mode 100644 drivers/vfio/vfio_platform.c
> >
> > diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig
> > index 1f84eda..35254b7 100644
> > --- a/drivers/vfio/Kconfig
> > +++ b/drivers/vfio/Kconfig
> > @@ -13,4 +13,15 @@ menuconfig VFIO
> >
> >         If you don't know what to do here, say N.
> >
> > +config VFIO_PLATFORM
> > +     tristate "VFIO support for device tree based platform devices"
> > +     depends on VFIO && EVENTFD && OF
> > +     help
> > +       Support for platform devices with VFIO. This is required to make
> > +       use of platform devices present on device tree nodes using the VFIO
> > +       framework. Devices that are not described in the device tree cannot
> > +       be used by this driver.
> > +
> > +       If you don't know what to do here, say N.
> > +
> >  source "drivers/vfio/pci/Kconfig"
> > diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile
> > index 2398d4a..575c8dd 100644
> > --- a/drivers/vfio/Makefile
> > +++ b/drivers/vfio/Makefile
> > @@ -1,3 +1,4 @@
> >  obj-$(CONFIG_VFIO) += vfio.o
> >  obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
> >  obj-$(CONFIG_VFIO_PCI) += pci/
> > +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
> > diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c
> > new file mode 100644
> > index 0000000..b9686b0
> > --- /dev/null
> > +++ b/drivers/vfio/vfio_platform.c
> > @@ -0,0 +1,187 @@
> > +/*
> > + * Copyright (C) 2013 - Virtual Open Systems
> > + * Author: Antonios Motakis <[e-mail]>
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License, version 2, as
> > + * published by the Free Software Foundation.
> > + *
> > + * This program is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > + * GNU General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU General Public License
> > + * along with this program; if not, write to the Free Software
> > + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
> > + */
> > +
> > +#include <linux/device.h>
> > +#include <linux/eventfd.h>
> > +#include <linux/interrupt.h>
> > +#include <linux/iommu.h>
> > +#include <linux/module.h>
> > +#include <linux/mutex.h>
> > +#include <linux/notifier.h>
> > +#include <linux/pm_runtime.h>
> > +#include <linux/slab.h>
> > +#include <linux/types.h>
> > +#include <linux/uaccess.h>
> > +#include <linux/vfio.h>
> > +
> > +#define DRIVER_VERSION  "0.1"
> > +#define DRIVER_AUTHOR   "Antonios Motakis <[e-mail]>"
> > +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
> > +
> > +struct vfio_platform_device {
> > +     struct platform_device  *pdev;
> > +};
> > +
> > +static void vfio_platform_release(void *device_data)
> > +{
> > +     module_put(THIS_MODULE);
> > +}
> > +
> > +static int vfio_platform_open(void *device_data)
> > +{
> > +     if (!try_module_get(THIS_MODULE))
> > +             return -ENODEV;
> > +
> > +     return 0;
> > +}
> > +
> > +static long vfio_platform_ioctl(void *device_data,
> > +                        unsigned int cmd, unsigned long arg)
> > +{
> > +     struct vfio_platform_device *vdev = device_data;
> > +     unsigned long minsz;
> > +
> > +     if (cmd == VFIO_DEVICE_GET_INFO) {
> > +             struct vfio_device_info info;
> > +
> > +             minsz = offsetofend(struct vfio_device_info, num_irqs);
> > +
> > +             if (copy_from_user(&info, (void __user *)arg, minsz))
> > +                     return -EFAULT;
> > +
> > +             if (info.argsz < minsz)
> > +                     return -EINVAL;
> > +
> > +             info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
> > +             info.num_regions = 0;
> > +             info.num_irqs = 0;
> > +
> > +             return copy_to_user((void __user *)arg, &info, minsz);
> > +
> > +     } else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
> > +             return -EINVAL;
> > +
> > +     else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
> > +             return -EINVAL;
> > +
> > +     else if (cmd == VFIO_DEVICE_SET_IRQS)
> > +             return -EINVAL;
> > +
> > +     else if (cmd == VFIO_DEVICE_RESET)
> > +             return -EINVAL;
> > +
> > +     return -ENOTTY;
> > +}
> > +
> > +static ssize_t vfio_platform_read(void *device_data, char __user *buf,
> > +                          size_t count, loff_t *ppos)
> > +{
> > +     return 0;
> > +}
> > +
> > +static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
> > +                           size_t count, loff_t *ppos)
> > +{
> > +     return 0;
> > +}
> > +
> > +static int vfio_platform_mmap(void *device_data, struct vm_area_struct *vma)
> > +{
> > +     return -EINVAL;
> > +}
> > +
> > +static const struct vfio_device_ops vfio_platform_ops = {
> > +     .name           = "vfio-platform",
> > +     .open           = vfio_platform_open,
> > +     .release        = vfio_platform_release,
> > +     .ioctl          = vfio_platform_ioctl,
> > +     .read           = vfio_platform_read,
> > +     .write          = vfio_platform_write,
> > +     .mmap           = vfio_platform_mmap,
> > +};
> > +
> > +static int vfio_platform_probe(struct platform_device *pdev)
> > +{
> > +     struct vfio_platform_device *vdev;
> > +     struct iommu_group *group;
> > +     int ret;
> > +
> > +     group = iommu_group_get(&pdev->dev);
> > +     if (!group) {
> > +             pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
> > +             return -EINVAL;
> > +     }
> > +
> > +     vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
> > +     if (!vdev) {
> > +             iommu_group_put(group);
> > +             return -ENOMEM;
> > +     }
> > +
> > +     vdev->pdev = pdev;
> > +
> > +     ret = vfio_add_group_dev(&pdev->dev, &vfio_platform_ops, vdev);
> > +     if (ret) {
> > +             iommu_group_put(group);
> > +             kfree(vdev);
> > +     }
> > +
> > +     return ret;
> > +}
> > +
> > +static int vfio_platform_remove(struct platform_device *pdev)
> > +{
> > +     struct vfio_platform_device *vdev;
> > +
> > +     vdev = vfio_del_group_dev(&pdev->dev);
> > +     if (!vdev)
> > +             return -EINVAL;
> > +
> > +     iommu_group_put(pdev->dev.iommu_group);
> > +     kfree(vdev);
> > +
> > +     return 0;
> > +}
> > +
> > +static const struct of_device_id vfio_platform_match[] = {
> > +     /* Before this series can be applied, we shall implement a better
> > +      * mechanism to bind the module to any device. For now add the
> > +      * compatible property to the dtb of the devices we want to use. */
> > +     {
> > +             .compatible = "vfio",
> > +     },
> > +     {},
> > +};
> > +MODULE_DEVICE_TABLE(of, vfio_platform_match);
> > +
> > +static struct platform_driver vfio_platform_driver = {
> > +     .probe          = vfio_platform_probe,
> > +     .remove         = vfio_platform_remove,
> > +     .driver = {
> > +             .name   = "vfio-platform",
> > +             .owner  = THIS_MODULE,
> > +             .of_match_table = vfio_platform_match,
> > +     },
> > +};
> > +
> > +module_platform_driver(vfio_platform_driver);
> > +
> > +MODULE_VERSION(DRIVER_VERSION);
> > +MODULE_LICENSE("GPL v2");
> > +MODULE_AUTHOR(DRIVER_AUTHOR);
> > +MODULE_DESCRIPTION(DRIVER_DESC);
> > diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
> > index 284ff24..8d7434d 100644
> > --- a/include/uapi/linux/vfio.h
> > +++ b/include/uapi/linux/vfio.h
> > @@ -147,6 +147,7 @@ struct vfio_device_info {
> >       __u32   flags;
> >  #define VFIO_DEVICE_FLAGS_RESET      (1 << 0)        /* Device supports reset */
> >  #define VFIO_DEVICE_FLAGS_PCI        (1 << 1)        /* vfio-pci device */
> > +#define VFIO_DEVICE_FLAGS_PLATFORM (1 << 2)  /* vfio-platform device */
> >       __u32   num_regions;    /* Max region index + 1 */
> >       __u32   num_irqs;       /* Max IRQ index + 1 */
> >  };
>
>
>

Ack, thanks.
--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Donald Dutile Oct. 29, 2013, 1:29 a.m. UTC | #4

On 09/30/2013 11:37 AM, Bhushan Bharat-R65777 wrote:
>
>
>> -----Original Message-----
>> From: [e-mail] [mailto:iommu-
>> [e-mail]] On Behalf Of Antonios Motakis
>> Sent: Monday, September 30, 2013 8:59 PM
>> To: [e-mail]; [e-mail]
>> Cc: [e-mail]; [e-mail]; [e-mail]; Yoder
>> Stuart-B08248; [e-mail]; Antonios Motakis;
>> [e-mail]
>> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree based
>> devices
>>
>> Platform devices in the Linux kernel are usually managed by the DT interface.
>> This patch forms the base to support these kind of devices with VFIO.
>>
>> Signed-off-by: Antonios Motakis<[e-mail]>
>> ---
>>   drivers/vfio/Kconfig         |  11 +++
>>   drivers/vfio/Makefile        |   1 +
>>   drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++
>>   include/uapi/linux/vfio.h    |   1 +
>>   4 files changed, 200 insertions(+)
>>   create mode 100644 drivers/vfio/vfio_platform.c
>>
>> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index 1f84eda..35254b7
>> 100644
>> --- a/drivers/vfio/Kconfig
>> +++ b/drivers/vfio/Kconfig
>> @@ -13,4 +13,15 @@ menuconfig VFIO
>>
>>   	  If you don't know what to do here, say N.
>>
>> +config VFIO_PLATFORM
>> +	tristate "VFIO support for device tree based platform devices"
>> +	depends on VFIO&&  EVENTFD&&  OF
>> +	help
>> +	  Support for platform devices with VFIO. This is required to make
>> +	  use of platform devices present on device tree nodes using the VFIO
>> +	  framework. Devices that are not described in the device tree cannot
>> +	  be used by this driver.
>> +
>> +	  If you don't know what to do here, say N.
>> +
>>   source "drivers/vfio/pci/Kconfig"
>> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index
>> 2398d4a..575c8dd 100644
>> --- a/drivers/vfio/Makefile
>> +++ b/drivers/vfio/Makefile
>> @@ -1,3 +1,4 @@
>>   obj-$(CONFIG_VFIO) += vfio.o
>>   obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
>>   obj-$(CONFIG_VFIO_PCI) += pci/
>> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
>> diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c new
>
> We can make this parallel to PCI, something like drivers/vfio/platform/platform.c
>
pls, no.  'platform' is too generic, and it really means 'arm-dt' ... so can
move it to the arch/arm space, and have it's kconfig conditional on ARM&&VFIO.
if kept under drivers/vfio, then use a better directory name that ties it to arm-dt.
thanks.
- Don


> -Bharat
>
>> file mode 100644 index 0000000..b9686b0
>> --- /dev/null
>> +++ b/drivers/vfio/vfio_platform.c
>> @@ -0,0 +1,187 @@
>> +/*
>> + * Copyright (C) 2013 - Virtual Open Systems
>> + * Author: Antonios Motakis<[e-mail]>
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License, version 2, as
>> + * published by the Free Software Foundation.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program; if not, write to the Free Software
>> + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
>> + */
>> +
>> +#include<linux/device.h>
>> +#include<linux/eventfd.h>
>> +#include<linux/interrupt.h>
>> +#include<linux/iommu.h>
>> +#include<linux/module.h>
>> +#include<linux/mutex.h>
>> +#include<linux/notifier.h>
>> +#include<linux/pm_runtime.h>
>> +#include<linux/slab.h>
>> +#include<linux/types.h>
>> +#include<linux/uaccess.h>
>> +#include<linux/vfio.h>
>> +
>> +#define DRIVER_VERSION  "0.1"
>> +#define DRIVER_AUTHOR   "Antonios Motakis<[e-mail]>"
>> +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
>> +
>> +struct vfio_platform_device {
>> +	struct platform_device	*pdev;
>> +};
>> +
>> +static void vfio_platform_release(void *device_data) {
>> +	module_put(THIS_MODULE);
>> +}
>> +
>> +static int vfio_platform_open(void *device_data) {
>> +	if (!try_module_get(THIS_MODULE))
>> +		return -ENODEV;
>> +
>> +	return 0;
>> +}
>> +
>> +static long vfio_platform_ioctl(void *device_data,
>> +			   unsigned int cmd, unsigned long arg) {
>> +	struct vfio_platform_device *vdev = device_data;
>> +	unsigned long minsz;
>> +
>> +	if (cmd == VFIO_DEVICE_GET_INFO) {
>> +		struct vfio_device_info info;
>> +
>> +		minsz = offsetofend(struct vfio_device_info, num_irqs);
>> +
>> +		if (copy_from_user(&info, (void __user *)arg, minsz))
>> +			return -EFAULT;
>> +
>> +		if (info.argsz<  minsz)
>> +			return -EINVAL;
>> +
>> +		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
>> +		info.num_regions = 0;
>> +		info.num_irqs = 0;
>> +
>> +		return copy_to_user((void __user *)arg,&info, minsz);
>> +
>> +	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
>> +		return -EINVAL;
>> +
>> +	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
>> +		return -EINVAL;
>> +
>> +	else if (cmd == VFIO_DEVICE_SET_IRQS)
>> +		return -EINVAL;
>> +
>> +	else if (cmd == VFIO_DEVICE_RESET)
>> +		return -EINVAL;
>> +
>> +	return -ENOTTY;
>> +}
>> +
>> +static ssize_t vfio_platform_read(void *device_data, char __user *buf,
>> +			     size_t count, loff_t *ppos)
>> +{
>> +	return 0;
>> +}
>> +
>> +static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
>> +			      size_t count, loff_t *ppos)
>> +{
>> +	return 0;
>> +}
>> +
>> +static int vfio_platform_mmap(void *device_data, struct vm_area_struct
>> +*vma) {
>> +	return -EINVAL;
>> +}
>> +
>> +static const struct vfio_device_ops vfio_platform_ops = {
>> +	.name		= "vfio-platform",
>> +	.open		= vfio_platform_open,
>> +	.release	= vfio_platform_release,
>> +	.ioctl		= vfio_platform_ioctl,
>> +	.read		= vfio_platform_read,
>> +	.write		= vfio_platform_write,
>> +	.mmap		= vfio_platform_mmap,
>> +};
>> +
>> +static int vfio_platform_probe(struct platform_device *pdev) {
>> +	struct vfio_platform_device *vdev;
>> +	struct iommu_group *group;
>> +	int ret;
>> +
>> +	group = iommu_group_get(&pdev->dev);
>> +	if (!group) {
>> +		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
>> +		return -EINVAL;
>> +	}
>> +
>> +	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
>> +	if (!vdev) {
>> +		iommu_group_put(group);
>> +		return -ENOMEM;
>> +	}
>> +
>> +	vdev->pdev = pdev;
>> +
>> +	ret = vfio_add_group_dev(&pdev->dev,&vfio_platform_ops, vdev);
>> +	if (ret) {
>> +		iommu_group_put(group);
>> +		kfree(vdev);
>> +	}
>> +
>> +	return ret;
>> +}
>> +
>> +static int vfio_platform_remove(struct platform_device *pdev) {
>> +	struct vfio_platform_device *vdev;
>> +
>> +	vdev = vfio_del_group_dev(&pdev->dev);
>> +	if (!vdev)
>> +		return -EINVAL;
>> +
>> +	iommu_group_put(pdev->dev.iommu_group);
>> +	kfree(vdev);
>> +
>> +	return 0;
>> +}
>> +
>> +static const struct of_device_id vfio_platform_match[] = {
>> +	/* Before this series can be applied, we shall implement a better
>> +	 * mechanism to bind the module to any device. For now add the
>> +	 * compatible property to the dtb of the devices we want to use. */
>> +	{
>> +		.compatible = "vfio",
>> +	},
>> +	{},
>> +};
>> +MODULE_DEVICE_TABLE(of, vfio_platform_match);
>> +
>> +static struct platform_driver vfio_platform_driver = {
>> +	.probe		= vfio_platform_probe,
>> +	.remove		= vfio_platform_remove,
>> +	.driver	= {
>> +		.name	= "vfio-platform",
>> +		.owner	= THIS_MODULE,
>> +		.of_match_table = vfio_platform_match,
>> +	},
>> +};
>> +
>> +module_platform_driver(vfio_platform_driver);
>> +
>> +MODULE_VERSION(DRIVER_VERSION);
>> +MODULE_LICENSE("GPL v2");
>> +MODULE_AUTHOR(DRIVER_AUTHOR);
>> +MODULE_DESCRIPTION(DRIVER_DESC);
>> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h index
>> 284ff24..8d7434d 100644
>> --- a/include/uapi/linux/vfio.h
>> +++ b/include/uapi/linux/vfio.h
>> @@ -147,6 +147,7 @@ struct vfio_device_info {
>>   	__u32	flags;
>>   #define VFIO_DEVICE_FLAGS_RESET	(1<<  0)	/* Device supports reset */
>>   #define VFIO_DEVICE_FLAGS_PCI	(1<<  1)	/* vfio-pci device */
>> +#define VFIO_DEVICE_FLAGS_PLATFORM (1<<  2)	/* vfio-platform device */
>>   	__u32	num_regions;	/* Max region index + 1 */
>>   	__u32	num_irqs;	/* Max IRQ index + 1 */
>>   };
>> --
>> 1.8.1.2
>>
>> _______________________________________________
>> iommu mailing list
>> [e-mail]
>> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>
>
> _______________________________________________
> iommu mailing list
> [e-mail]
> https://lists.linuxfoundation.org/mailman/listinfo/iommu

--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Alex Williamson Oct. 29, 2013, 11:47 a.m. UTC | #5

On Mon, 2013-10-28 at 21:29 -0400, Don Dutile wrote:
> On 09/30/2013 11:37 AM, Bhushan Bharat-R65777 wrote:
> >
> >
> >> -----Original Message-----
> >> From: [e-mail] [mailto:iommu-
> >> [e-mail]] On Behalf Of Antonios Motakis
> >> Sent: Monday, September 30, 2013 8:59 PM
> >> To: [e-mail]; [e-mail]
> >> Cc: [e-mail]; [e-mail]; [e-mail]; Yoder
> >> Stuart-B08248; [e-mail]; Antonios Motakis;
> >> [e-mail]
> >> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree based
> >> devices
> >>
> >> Platform devices in the Linux kernel are usually managed by the DT interface.
> >> This patch forms the base to support these kind of devices with VFIO.
> >>
> >> Signed-off-by: Antonios Motakis<[e-mail]>
> >> ---
> >>   drivers/vfio/Kconfig         |  11 +++
> >>   drivers/vfio/Makefile        |   1 +
> >>   drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++
> >>   include/uapi/linux/vfio.h    |   1 +
> >>   4 files changed, 200 insertions(+)
> >>   create mode 100644 drivers/vfio/vfio_platform.c
> >>
> >> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index 1f84eda..35254b7
> >> 100644
> >> --- a/drivers/vfio/Kconfig
> >> +++ b/drivers/vfio/Kconfig
> >> @@ -13,4 +13,15 @@ menuconfig VFIO
> >>
> >>   	  If you don't know what to do here, say N.
> >>
> >> +config VFIO_PLATFORM
> >> +	tristate "VFIO support for device tree based platform devices"
> >> +	depends on VFIO&&  EVENTFD&&  OF
> >> +	help
> >> +	  Support for platform devices with VFIO. This is required to make
> >> +	  use of platform devices present on device tree nodes using the VFIO
> >> +	  framework. Devices that are not described in the device tree cannot
> >> +	  be used by this driver.
> >> +
> >> +	  If you don't know what to do here, say N.
> >> +
> >>   source "drivers/vfio/pci/Kconfig"
> >> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index
> >> 2398d4a..575c8dd 100644
> >> --- a/drivers/vfio/Makefile
> >> +++ b/drivers/vfio/Makefile
> >> @@ -1,3 +1,4 @@
> >>   obj-$(CONFIG_VFIO) += vfio.o
> >>   obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
> >>   obj-$(CONFIG_VFIO_PCI) += pci/
> >> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
> >> diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c new
> >
> > We can make this parallel to PCI, something like drivers/vfio/platform/platform.c
> >
> pls, no.  'platform' is too generic, and it really means 'arm-dt' ... so can
> move it to the arch/arm space, and have it's kconfig conditional on ARM&&VFIO.
> if kept under drivers/vfio, then use a better directory name that ties it to arm-dt.
> thanks.

The intention is that vfio platform device support is not arm-dt
specific.  This is to be used by both arm and embedded ppc.  The devices
we intend to support with them are known as platform drivers in the
kernel, thus the name.  I suppose the question remains whether the
interface here is really generic for any "platform" device or whether
we're making whether we're making an interface specifically for device
tree platform devices, or if those are one in the same.  In any case,
arm-dt is certainly not the answer.

Alex

> >> file mode 100644 index 0000000..b9686b0
> >> --- /dev/null
> >> +++ b/drivers/vfio/vfio_platform.c
> >> @@ -0,0 +1,187 @@
> >> +/*
> >> + * Copyright (C) 2013 - Virtual Open Systems
> >> + * Author: Antonios Motakis<[e-mail]>
> >> + *
> >> + * This program is free software; you can redistribute it and/or modify
> >> + * it under the terms of the GNU General Public License, version 2, as
> >> + * published by the Free Software Foundation.
> >> + *
> >> + * This program is distributed in the hope that it will be useful,
> >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >> + * GNU General Public License for more details.
> >> + *
> >> + * You should have received a copy of the GNU General Public License
> >> + * along with this program; if not, write to the Free Software
> >> + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
> >> + */
> >> +
> >> +#include<linux/device.h>
> >> +#include<linux/eventfd.h>
> >> +#include<linux/interrupt.h>
> >> +#include<linux/iommu.h>
> >> +#include<linux/module.h>
> >> +#include<linux/mutex.h>
> >> +#include<linux/notifier.h>
> >> +#include<linux/pm_runtime.h>
> >> +#include<linux/slab.h>
> >> +#include<linux/types.h>
> >> +#include<linux/uaccess.h>
> >> +#include<linux/vfio.h>
> >> +
> >> +#define DRIVER_VERSION  "0.1"
> >> +#define DRIVER_AUTHOR   "Antonios Motakis<[e-mail]>"
> >> +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
> >> +
> >> +struct vfio_platform_device {
> >> +	struct platform_device	*pdev;
> >> +};
> >> +
> >> +static void vfio_platform_release(void *device_data) {
> >> +	module_put(THIS_MODULE);
> >> +}
> >> +
> >> +static int vfio_platform_open(void *device_data) {
> >> +	if (!try_module_get(THIS_MODULE))
> >> +		return -ENODEV;
> >> +
> >> +	return 0;
> >> +}
> >> +
> >> +static long vfio_platform_ioctl(void *device_data,
> >> +			   unsigned int cmd, unsigned long arg) {
> >> +	struct vfio_platform_device *vdev = device_data;
> >> +	unsigned long minsz;
> >> +
> >> +	if (cmd == VFIO_DEVICE_GET_INFO) {
> >> +		struct vfio_device_info info;
> >> +
> >> +		minsz = offsetofend(struct vfio_device_info, num_irqs);
> >> +
> >> +		if (copy_from_user(&info, (void __user *)arg, minsz))
> >> +			return -EFAULT;
> >> +
> >> +		if (info.argsz<  minsz)
> >> +			return -EINVAL;
> >> +
> >> +		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
> >> +		info.num_regions = 0;
> >> +		info.num_irqs = 0;
> >> +
> >> +		return copy_to_user((void __user *)arg,&info, minsz);
> >> +
> >> +	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
> >> +		return -EINVAL;
> >> +
> >> +	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
> >> +		return -EINVAL;
> >> +
> >> +	else if (cmd == VFIO_DEVICE_SET_IRQS)
> >> +		return -EINVAL;
> >> +
> >> +	else if (cmd == VFIO_DEVICE_RESET)
> >> +		return -EINVAL;
> >> +
> >> +	return -ENOTTY;
> >> +}
> >> +
> >> +static ssize_t vfio_platform_read(void *device_data, char __user *buf,
> >> +			     size_t count, loff_t *ppos)
> >> +{
> >> +	return 0;
> >> +}
> >> +
> >> +static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
> >> +			      size_t count, loff_t *ppos)
> >> +{
> >> +	return 0;
> >> +}
> >> +
> >> +static int vfio_platform_mmap(void *device_data, struct vm_area_struct
> >> +*vma) {
> >> +	return -EINVAL;
> >> +}
> >> +
> >> +static const struct vfio_device_ops vfio_platform_ops = {
> >> +	.name		= "vfio-platform",
> >> +	.open		= vfio_platform_open,
> >> +	.release	= vfio_platform_release,
> >> +	.ioctl		= vfio_platform_ioctl,
> >> +	.read		= vfio_platform_read,
> >> +	.write		= vfio_platform_write,
> >> +	.mmap		= vfio_platform_mmap,
> >> +};
> >> +
> >> +static int vfio_platform_probe(struct platform_device *pdev) {
> >> +	struct vfio_platform_device *vdev;
> >> +	struct iommu_group *group;
> >> +	int ret;
> >> +
> >> +	group = iommu_group_get(&pdev->dev);
> >> +	if (!group) {
> >> +		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
> >> +		return -EINVAL;
> >> +	}
> >> +
> >> +	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
> >> +	if (!vdev) {
> >> +		iommu_group_put(group);
> >> +		return -ENOMEM;
> >> +	}
> >> +
> >> +	vdev->pdev = pdev;
> >> +
> >> +	ret = vfio_add_group_dev(&pdev->dev,&vfio_platform_ops, vdev);
> >> +	if (ret) {
> >> +		iommu_group_put(group);
> >> +		kfree(vdev);
> >> +	}
> >> +
> >> +	return ret;
> >> +}
> >> +
> >> +static int vfio_platform_remove(struct platform_device *pdev) {
> >> +	struct vfio_platform_device *vdev;
> >> +
> >> +	vdev = vfio_del_group_dev(&pdev->dev);
> >> +	if (!vdev)
> >> +		return -EINVAL;
> >> +
> >> +	iommu_group_put(pdev->dev.iommu_group);
> >> +	kfree(vdev);
> >> +
> >> +	return 0;
> >> +}
> >> +
> >> +static const struct of_device_id vfio_platform_match[] = {
> >> +	/* Before this series can be applied, we shall implement a better
> >> +	 * mechanism to bind the module to any device. For now add the
> >> +	 * compatible property to the dtb of the devices we want to use. */
> >> +	{
> >> +		.compatible = "vfio",
> >> +	},
> >> +	{},
> >> +};
> >> +MODULE_DEVICE_TABLE(of, vfio_platform_match);
> >> +
> >> +static struct platform_driver vfio_platform_driver = {
> >> +	.probe		= vfio_platform_probe,
> >> +	.remove		= vfio_platform_remove,
> >> +	.driver	= {
> >> +		.name	= "vfio-platform",
> >> +		.owner	= THIS_MODULE,
> >> +		.of_match_table = vfio_platform_match,
> >> +	},
> >> +};
> >> +
> >> +module_platform_driver(vfio_platform_driver);
> >> +
> >> +MODULE_VERSION(DRIVER_VERSION);
> >> +MODULE_LICENSE("GPL v2");
> >> +MODULE_AUTHOR(DRIVER_AUTHOR);
> >> +MODULE_DESCRIPTION(DRIVER_DESC);
> >> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h index
> >> 284ff24..8d7434d 100644
> >> --- a/include/uapi/linux/vfio.h
> >> +++ b/include/uapi/linux/vfio.h
> >> @@ -147,6 +147,7 @@ struct vfio_device_info {
> >>   	__u32	flags;
> >>   #define VFIO_DEVICE_FLAGS_RESET	(1<<  0)	/* Device supports reset */
> >>   #define VFIO_DEVICE_FLAGS_PCI	(1<<  1)	/* vfio-pci device */
> >> +#define VFIO_DEVICE_FLAGS_PLATFORM (1<<  2)	/* vfio-platform device */
> >>   	__u32	num_regions;	/* Max region index + 1 */
> >>   	__u32	num_irqs;	/* Max IRQ index + 1 */
> >>   };
> >> --
> >> 1.8.1.2
> >>
> >> _______________________________________________
> >> iommu mailing list
> >> [e-mail]
> >> https://lists.linuxfoundation.org/mailman/listinfo/iommu
> >
> >
> > _______________________________________________
> > iommu mailing list
> > [e-mail]
> > https://lists.linuxfoundation.org/mailman/listinfo/iommu
> 
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to [e-mail]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html



--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Yoder Stuart-B08248 Oct. 29, 2013, 1:07 p.m. UTC | #6

> -----Original Message-----
> From: Don Dutile [mailto:[e-mail]]
> Sent: Monday, October 28, 2013 8:30 PM
> To: Bhushan Bharat-R65777
> Cc: Antonios Motakis; [e-mail];
> [e-mail]; [e-mail];
> [e-mail]; [e-mail]; Yoder Stuart-B08248;
> [e-mail]; [e-mail]
> Subject: Re: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree
> based devices
> 
> On 09/30/2013 11:37 AM, Bhushan Bharat-R65777 wrote:
> >
> >
> >> -----Original Message-----
> >> From: [e-mail] [mailto:iommu-
> >> [e-mail]] On Behalf Of Antonios Motakis
> >> Sent: Monday, September 30, 2013 8:59 PM
> >> To: [e-mail]; [e-mail]
> >> Cc: [e-mail]; [e-mail];
> [e-mail]; Yoder
> >> Stuart-B08248; [e-mail]; Antonios Motakis;
> >> [e-mail]
> >> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree
> based
> >> devices
> >>
> >> Platform devices in the Linux kernel are usually managed by the DT
> interface.
> >> This patch forms the base to support these kind of devices with VFIO.
> >>
> >> Signed-off-by: Antonios Motakis<[e-mail]>
> >> ---
> >>   drivers/vfio/Kconfig         |  11 +++
> >>   drivers/vfio/Makefile        |   1 +
> >>   drivers/vfio/vfio_platform.c | 187
> +++++++++++++++++++++++++++++++++++++++++++
> >>   include/uapi/linux/vfio.h    |   1 +
> >>   4 files changed, 200 insertions(+)
> >>   create mode 100644 drivers/vfio/vfio_platform.c
> >>
> >> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index
> 1f84eda..35254b7
> >> 100644
> >> --- a/drivers/vfio/Kconfig
> >> +++ b/drivers/vfio/Kconfig
> >> @@ -13,4 +13,15 @@ menuconfig VFIO
> >>
> >>   	  If you don't know what to do here, say N.
> >>
> >> +config VFIO_PLATFORM
> >> +	tristate "VFIO support for device tree based platform devices"
> >> +	depends on VFIO&&  EVENTFD&&  OF
> >> +	help
> >> +	  Support for platform devices with VFIO. This is required to make
> >> +	  use of platform devices present on device tree nodes using the
> VFIO
> >> +	  framework. Devices that are not described in the device tree
> cannot
> >> +	  be used by this driver.
> >> +
> >> +	  If you don't know what to do here, say N.
> >> +
> >>   source "drivers/vfio/pci/Kconfig"
> >> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index
> >> 2398d4a..575c8dd 100644
> >> --- a/drivers/vfio/Makefile
> >> +++ b/drivers/vfio/Makefile
> >> @@ -1,3 +1,4 @@
> >>   obj-$(CONFIG_VFIO) += vfio.o
> >>   obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
> >>   obj-$(CONFIG_VFIO_PCI) += pci/
> >> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
> >> diff --git a/drivers/vfio/vfio_platform.c
> b/drivers/vfio/vfio_platform.c new
> >
> > We can make this parallel to PCI, something like
> drivers/vfio/platform/platform.c
> >
> pls, no.  'platform' is too generic, and it really means 'arm-dt' ... so
> can
> move it to the arch/arm space, and have it's kconfig conditional on
> ARM&&VFIO.
> if kept under drivers/vfio, then use a better directory name that ties it
> to arm-dt.
> thanks.

This has nothing to do with ARM specifically.  We need this for Power architecture
as well.

And the name should be 'platform' as that _is_ the name of the bus
that these devices sit on.

This has nothing to do with device tree's per se.  It so happens that
the device tree is parsed and devices discovered there are added
to the platform bus.

The intent is to allow devices on the platform bus to be passed through
to user space using vfio.

Stuart

--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Bharat Bhushan Oct. 29, 2013, 1:37 p.m. UTC | #7

> -----Original Message-----

> From: [e-mail] [mailto:[e-mail]] On Behalf Of

> Alex Williamson

> Sent: Tuesday, October 29, 2013 5:17 PM

> To: Don Dutile

> Cc: Bhushan Bharat-R65777; Antonios Motakis; [e-mail];

> [e-mail]; [e-mail]; [e-mail]; Yoder

> Stuart-B08248; [e-mail]; [e-mail]

> Subject: Re: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree based

> devices

> 

> On Mon, 2013-10-28 at 21:29 -0400, Don Dutile wrote:

> > On 09/30/2013 11:37 AM, Bhushan Bharat-R65777 wrote:

> > >

> > >

> > >> -----Original Message-----

> > >> From: [e-mail] [mailto:iommu-

> > >> [e-mail]] On Behalf Of Antonios Motakis

> > >> Sent: Monday, September 30, 2013 8:59 PM

> > >> To: [e-mail]; [e-mail]

> > >> Cc: [e-mail]; [e-mail];

> > >> [e-mail]; Yoder Stuart-B08248;

> > >> [e-mail]; Antonios Motakis;

> > >> [e-mail]

> > >> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device

> > >> Tree based devices

> > >>

> > >> Platform devices in the Linux kernel are usually managed by the DT

> interface.

> > >> This patch forms the base to support these kind of devices with VFIO.

> > >>

> > >> Signed-off-by: Antonios Motakis<[e-mail]>

> > >> ---

> > >>   drivers/vfio/Kconfig         |  11 +++

> > >>   drivers/vfio/Makefile        |   1 +

> > >>   drivers/vfio/vfio_platform.c | 187

> +++++++++++++++++++++++++++++++++++++++++++

> > >>   include/uapi/linux/vfio.h    |   1 +

> > >>   4 files changed, 200 insertions(+)

> > >>   create mode 100644 drivers/vfio/vfio_platform.c

> > >>

> > >> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index

> > >> 1f84eda..35254b7

> > >> 100644

> > >> --- a/drivers/vfio/Kconfig

> > >> +++ b/drivers/vfio/Kconfig

> > >> @@ -13,4 +13,15 @@ menuconfig VFIO

> > >>

> > >>   	  If you don't know what to do here, say N.

> > >>

> > >> +config VFIO_PLATFORM

> > >> +	tristate "VFIO support for device tree based platform devices"

> > >> +	depends on VFIO&&  EVENTFD&&  OF

> > >> +	help

> > >> +	  Support for platform devices with VFIO. This is required to make

> > >> +	  use of platform devices present on device tree nodes using the VFIO

> > >> +	  framework. Devices that are not described in the device tree cannot

> > >> +	  be used by this driver.

> > >> +

> > >> +	  If you don't know what to do here, say N.

> > >> +

> > >>   source "drivers/vfio/pci/Kconfig"

> > >> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index

> > >> 2398d4a..575c8dd 100644

> > >> --- a/drivers/vfio/Makefile

> > >> +++ b/drivers/vfio/Makefile

> > >> @@ -1,3 +1,4 @@

> > >>   obj-$(CONFIG_VFIO) += vfio.o

> > >>   obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o

> > >>   obj-$(CONFIG_VFIO_PCI) += pci/

> > >> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o

> > >> diff --git a/drivers/vfio/vfio_platform.c

> > >> b/drivers/vfio/vfio_platform.c new

> > >

> > > We can make this parallel to PCI, something like

> > > drivers/vfio/platform/platform.c

> > >

> > pls, no.  'platform' is too generic, and it really means 'arm-dt' ...

> > so can move it to the arch/arm space, and have it's kconfig conditional on

> ARM&&VFIO.

> > if kept under drivers/vfio, then use a better directory name that ties it to

> arm-dt.

> > thanks.

> 

> The intention is that vfio platform device support is not arm-dt specific.  This

> is to be used by both arm and embedded ppc.  The devices we intend to support

> with them are known as platform drivers in the kernel, thus the name.  I suppose

> the question remains whether the interface here is really generic for any

> "platform" device or whether we're making whether we're making an interface

> specifically for device tree platform devices, or if those are one in the same.

> In any case, arm-dt is certainly not the answer.


I think we should stick to "platform".

-Bharat

> 

> Alex

> 

> > >> file mode 100644 index 0000000..b9686b0

> > >> --- /dev/null

> > >> +++ b/drivers/vfio/vfio_platform.c

> > >> @@ -0,0 +1,187 @@

> > >> +/*

> > >> + * Copyright (C) 2013 - Virtual Open Systems

> > >> + * Author: Antonios Motakis<[e-mail]>

> > >> + *

> > >> + * This program is free software; you can redistribute it and/or

> > >> +modify

> > >> + * it under the terms of the GNU General Public License, version

> > >> +2, as

> > >> + * published by the Free Software Foundation.

> > >> + *

> > >> + * This program is distributed in the hope that it will be useful,

> > >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of

> > >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

> > >> + * GNU General Public License for more details.

> > >> + *

> > >> + * You should have received a copy of the GNU General Public

> > >> +License

> > >> + * along with this program; if not, write to the Free Software

> > >> + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301,

> USA.

> > >> + */

> > >> +

> > >> +#include<linux/device.h>

> > >> +#include<linux/eventfd.h>

> > >> +#include<linux/interrupt.h>

> > >> +#include<linux/iommu.h>

> > >> +#include<linux/module.h>

> > >> +#include<linux/mutex.h>

> > >> +#include<linux/notifier.h>

> > >> +#include<linux/pm_runtime.h>

> > >> +#include<linux/slab.h>

> > >> +#include<linux/types.h>

> > >> +#include<linux/uaccess.h>

> > >> +#include<linux/vfio.h>

> > >> +

> > >> +#define DRIVER_VERSION  "0.1"

> > >> +#define DRIVER_AUTHOR   "Antonios

> Motakis<[e-mail]>"

> > >> +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-

> driver"

> > >> +

> > >> +struct vfio_platform_device {

> > >> +	struct platform_device	*pdev;

> > >> +};

> > >> +

> > >> +static void vfio_platform_release(void *device_data) {

> > >> +	module_put(THIS_MODULE);

> > >> +}

> > >> +

> > >> +static int vfio_platform_open(void *device_data) {

> > >> +	if (!try_module_get(THIS_MODULE))

> > >> +		return -ENODEV;

> > >> +

> > >> +	return 0;

> > >> +}

> > >> +

> > >> +static long vfio_platform_ioctl(void *device_data,

> > >> +			   unsigned int cmd, unsigned long arg) {

> > >> +	struct vfio_platform_device *vdev = device_data;

> > >> +	unsigned long minsz;

> > >> +

> > >> +	if (cmd == VFIO_DEVICE_GET_INFO) {

> > >> +		struct vfio_device_info info;

> > >> +

> > >> +		minsz = offsetofend(struct vfio_device_info, num_irqs);

> > >> +

> > >> +		if (copy_from_user(&info, (void __user *)arg, minsz))

> > >> +			return -EFAULT;

> > >> +

> > >> +		if (info.argsz<  minsz)

> > >> +			return -EINVAL;

> > >> +

> > >> +		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;

> > >> +		info.num_regions = 0;

> > >> +		info.num_irqs = 0;

> > >> +

> > >> +		return copy_to_user((void __user *)arg,&info, minsz);

> > >> +

> > >> +	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)

> > >> +		return -EINVAL;

> > >> +

> > >> +	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)

> > >> +		return -EINVAL;

> > >> +

> > >> +	else if (cmd == VFIO_DEVICE_SET_IRQS)

> > >> +		return -EINVAL;

> > >> +

> > >> +	else if (cmd == VFIO_DEVICE_RESET)

> > >> +		return -EINVAL;

> > >> +

> > >> +	return -ENOTTY;

> > >> +}

> > >> +

> > >> +static ssize_t vfio_platform_read(void *device_data, char __user *buf,

> > >> +			     size_t count, loff_t *ppos) {

> > >> +	return 0;

> > >> +}

> > >> +

> > >> +static ssize_t vfio_platform_write(void *device_data, const char __user

> *buf,

> > >> +			      size_t count, loff_t *ppos) {

> > >> +	return 0;

> > >> +}

> > >> +

> > >> +static int vfio_platform_mmap(void *device_data, struct

> > >> +vm_area_struct

> > >> +*vma) {

> > >> +	return -EINVAL;

> > >> +}

> > >> +

> > >> +static const struct vfio_device_ops vfio_platform_ops = {

> > >> +	.name		= "vfio-platform",

> > >> +	.open		= vfio_platform_open,

> > >> +	.release	= vfio_platform_release,

> > >> +	.ioctl		= vfio_platform_ioctl,

> > >> +	.read		= vfio_platform_read,

> > >> +	.write		= vfio_platform_write,

> > >> +	.mmap		= vfio_platform_mmap,

> > >> +};

> > >> +

> > >> +static int vfio_platform_probe(struct platform_device *pdev) {

> > >> +	struct vfio_platform_device *vdev;

> > >> +	struct iommu_group *group;

> > >> +	int ret;

> > >> +

> > >> +	group = iommu_group_get(&pdev->dev);

> > >> +	if (!group) {

> > >> +		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);

> > >> +		return -EINVAL;

> > >> +	}

> > >> +

> > >> +	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);

> > >> +	if (!vdev) {

> > >> +		iommu_group_put(group);

> > >> +		return -ENOMEM;

> > >> +	}

> > >> +

> > >> +	vdev->pdev = pdev;

> > >> +

> > >> +	ret = vfio_add_group_dev(&pdev->dev,&vfio_platform_ops, vdev);

> > >> +	if (ret) {

> > >> +		iommu_group_put(group);

> > >> +		kfree(vdev);

> > >> +	}

> > >> +

> > >> +	return ret;

> > >> +}

> > >> +

> > >> +static int vfio_platform_remove(struct platform_device *pdev) {

> > >> +	struct vfio_platform_device *vdev;

> > >> +

> > >> +	vdev = vfio_del_group_dev(&pdev->dev);

> > >> +	if (!vdev)

> > >> +		return -EINVAL;

> > >> +

> > >> +	iommu_group_put(pdev->dev.iommu_group);

> > >> +	kfree(vdev);

> > >> +

> > >> +	return 0;

> > >> +}

> > >> +

> > >> +static const struct of_device_id vfio_platform_match[] = {

> > >> +	/* Before this series can be applied, we shall implement a better

> > >> +	 * mechanism to bind the module to any device. For now add the

> > >> +	 * compatible property to the dtb of the devices we want to use. */

> > >> +	{

> > >> +		.compatible = "vfio",

> > >> +	},

> > >> +	{},

> > >> +};

> > >> +MODULE_DEVICE_TABLE(of, vfio_platform_match);

> > >> +

> > >> +static struct platform_driver vfio_platform_driver = {

> > >> +	.probe		= vfio_platform_probe,

> > >> +	.remove		= vfio_platform_remove,

> > >> +	.driver	= {

> > >> +		.name	= "vfio-platform",

> > >> +		.owner	= THIS_MODULE,

> > >> +		.of_match_table = vfio_platform_match,

> > >> +	},

> > >> +};

> > >> +

> > >> +module_platform_driver(vfio_platform_driver);

> > >> +

> > >> +MODULE_VERSION(DRIVER_VERSION);

> > >> +MODULE_LICENSE("GPL v2");

> > >> +MODULE_AUTHOR(DRIVER_AUTHOR);

> > >> +MODULE_DESCRIPTION(DRIVER_DESC);

> > >> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h

> > >> index 284ff24..8d7434d 100644

> > >> --- a/include/uapi/linux/vfio.h

> > >> +++ b/include/uapi/linux/vfio.h

> > >> @@ -147,6 +147,7 @@ struct vfio_device_info {

> > >>   	__u32	flags;

> > >>   #define VFIO_DEVICE_FLAGS_RESET	(1<<  0)	/* Device supports reset */

> > >>   #define VFIO_DEVICE_FLAGS_PCI	(1<<  1)	/* vfio-pci device */

> > >> +#define VFIO_DEVICE_FLAGS_PLATFORM (1<<  2)	/* vfio-platform device */

> > >>   	__u32	num_regions;	/* Max region index + 1 */

> > >>   	__u32	num_irqs;	/* Max IRQ index + 1 */

> > >>   };

> > >> --

> > >> 1.8.1.2

> > >>

> > >> _______________________________________________

> > >> iommu mailing list

> > >> [e-mail]

> > >> https://lists.linuxfoundation.org/mailman/listinfo/iommu

> > >

> > >

> > > _______________________________________________

> > > iommu mailing list

> > > [e-mail]

> > > https://lists.linuxfoundation.org/mailman/listinfo/iommu

> >

> > --

> > To unsubscribe from this list: send the line "unsubscribe kvm" in the

> > body of a message to [e-mail] More majordomo info at

> > http://vger.kernel.org/majordomo-info.html

> 

> 

> 

> --

> To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a

> message to [e-mail] More majordomo info at

> http://vger.kernel.org/majordomo-info.html

Donald Dutile Oct. 29, 2013, 5:13 p.m. UTC | #8

On 10/29/2013 07:47 AM, Alex Williamson wrote:
> On Mon, 2013-10-28 at 21:29 -0400, Don Dutile wrote:
>> On 09/30/2013 11:37 AM, Bhushan Bharat-R65777 wrote:
>>>
>>>
>>>> -----Original Message-----
>>>> From: [e-mail] [mailto:iommu-
>>>> [e-mail]] On Behalf Of Antonios Motakis
>>>> Sent: Monday, September 30, 2013 8:59 PM
>>>> To: [e-mail]; [e-mail]
>>>> Cc: [e-mail]; [e-mail]; [e-mail]; Yoder
>>>> Stuart-B08248; [e-mail]; Antonios Motakis;
>>>> [e-mail]
>>>> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree based
>>>> devices
>>>>
>>>> Platform devices in the Linux kernel are usually managed by the DT interface.
>>>> This patch forms the base to support these kind of devices with VFIO.
>>>>
>>>> Signed-off-by: Antonios Motakis<[e-mail]>
>>>> ---
>>>>    drivers/vfio/Kconfig         |  11 +++
>>>>    drivers/vfio/Makefile        |   1 +
>>>>    drivers/vfio/vfio_platform.c | 187 +++++++++++++++++++++++++++++++++++++++++++
>>>>    include/uapi/linux/vfio.h    |   1 +
>>>>    4 files changed, 200 insertions(+)
>>>>    create mode 100644 drivers/vfio/vfio_platform.c
>>>>
>>>> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index 1f84eda..35254b7
>>>> 100644
>>>> --- a/drivers/vfio/Kconfig
>>>> +++ b/drivers/vfio/Kconfig
>>>> @@ -13,4 +13,15 @@ menuconfig VFIO
>>>>
>>>>    	  If you don't know what to do here, say N.
>>>>
>>>> +config VFIO_PLATFORM
>>>> +	tristate "VFIO support for device tree based platform devices"
>>>> +	depends on VFIO&&   EVENTFD&&   OF
>>>> +	help
>>>> +	  Support for platform devices with VFIO. This is required to make
>>>> +	  use of platform devices present on device tree nodes using the VFIO
>>>> +	  framework. Devices that are not described in the device tree cannot
>>>> +	  be used by this driver.
>>>> +
>>>> +	  If you don't know what to do here, say N.
>>>> +
>>>>    source "drivers/vfio/pci/Kconfig"
>>>> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index
>>>> 2398d4a..575c8dd 100644
>>>> --- a/drivers/vfio/Makefile
>>>> +++ b/drivers/vfio/Makefile
>>>> @@ -1,3 +1,4 @@
>>>>    obj-$(CONFIG_VFIO) += vfio.o
>>>>    obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
>>>>    obj-$(CONFIG_VFIO_PCI) += pci/
>>>> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
>>>> diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c new
>>>
>>> We can make this parallel to PCI, something like drivers/vfio/platform/platform.c
>>>
>> pls, no.  'platform' is too generic, and it really means 'arm-dt' ... so can
>> move it to the arch/arm space, and have it's kconfig conditional on ARM&&VFIO.
>> if kept under drivers/vfio, then use a better directory name that ties it to arm-dt.
>> thanks.
>
> The intention is that vfio platform device support is not arm-dt
> specific.  This is to be used by both arm and embedded ppc.  The devices
> we intend to support with them are known as platform drivers in the
> kernel, thus the name.  I suppose the question remains whether the
> interface here is really generic for any "platform" device or whether
> we're making whether we're making an interface specifically for device
> tree platform devices, or if those are one in the same.  In any case,
> arm-dt is certainly not the answer.
>
> Alex
>
I thought that was the intention until I saw this use in platform.c:
	static const struct of_device_id vfio_platform_match[] = {

So, of_device_id hit me as DT-specific, and thus, the file should have
a name that implies its not as platform-generic as one may want/expect.
I agree that the 'arm' part can be dropped, but it's not of-agnostic atm.


>>>> file mode 100644 index 0000000..b9686b0
>>>> --- /dev/null
>>>> +++ b/drivers/vfio/vfio_platform.c
>>>> @@ -0,0 +1,187 @@
>>>> +/*
>>>> + * Copyright (C) 2013 - Virtual Open Systems
>>>> + * Author: Antonios Motakis<[e-mail]>
>>>> + *
>>>> + * This program is free software; you can redistribute it and/or modify
>>>> + * it under the terms of the GNU General Public License, version 2, as
>>>> + * published by the Free Software Foundation.
>>>> + *
>>>> + * This program is distributed in the hope that it will be useful,
>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>> + * GNU General Public License for more details.
>>>> + *
>>>> + * You should have received a copy of the GNU General Public License
>>>> + * along with this program; if not, write to the Free Software
>>>> + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
>>>> + */
>>>> +
>>>> +#include<linux/device.h>
>>>> +#include<linux/eventfd.h>
>>>> +#include<linux/interrupt.h>
>>>> +#include<linux/iommu.h>
>>>> +#include<linux/module.h>
>>>> +#include<linux/mutex.h>
>>>> +#include<linux/notifier.h>
>>>> +#include<linux/pm_runtime.h>
>>>> +#include<linux/slab.h>
>>>> +#include<linux/types.h>
>>>> +#include<linux/uaccess.h>
>>>> +#include<linux/vfio.h>
>>>> +
>>>> +#define DRIVER_VERSION  "0.1"
>>>> +#define DRIVER_AUTHOR   "Antonios Motakis<[e-mail]>"
>>>> +#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
>>>> +
>>>> +struct vfio_platform_device {
>>>> +	struct platform_device	*pdev;
>>>> +};
>>>> +
>>>> +static void vfio_platform_release(void *device_data) {
>>>> +	module_put(THIS_MODULE);
>>>> +}
>>>> +
>>>> +static int vfio_platform_open(void *device_data) {
>>>> +	if (!try_module_get(THIS_MODULE))
>>>> +		return -ENODEV;
>>>> +
>>>> +	return 0;
>>>> +}
>>>> +
>>>> +static long vfio_platform_ioctl(void *device_data,
>>>> +			   unsigned int cmd, unsigned long arg) {
>>>> +	struct vfio_platform_device *vdev = device_data;
>>>> +	unsigned long minsz;
>>>> +
>>>> +	if (cmd == VFIO_DEVICE_GET_INFO) {
>>>> +		struct vfio_device_info info;
>>>> +
>>>> +		minsz = offsetofend(struct vfio_device_info, num_irqs);
>>>> +
>>>> +		if (copy_from_user(&info, (void __user *)arg, minsz))
>>>> +			return -EFAULT;
>>>> +
>>>> +		if (info.argsz<   minsz)
>>>> +			return -EINVAL;
>>>> +
>>>> +		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
>>>> +		info.num_regions = 0;
>>>> +		info.num_irqs = 0;
>>>> +
>>>> +		return copy_to_user((void __user *)arg,&info, minsz);
>>>> +
>>>> +	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
>>>> +		return -EINVAL;
>>>> +
>>>> +	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
>>>> +		return -EINVAL;
>>>> +
>>>> +	else if (cmd == VFIO_DEVICE_SET_IRQS)
>>>> +		return -EINVAL;
>>>> +
>>>> +	else if (cmd == VFIO_DEVICE_RESET)
>>>> +		return -EINVAL;
>>>> +
>>>> +	return -ENOTTY;
>>>> +}
>>>> +
>>>> +static ssize_t vfio_platform_read(void *device_data, char __user *buf,
>>>> +			     size_t count, loff_t *ppos)
>>>> +{
>>>> +	return 0;
>>>> +}
>>>> +
>>>> +static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
>>>> +			      size_t count, loff_t *ppos)
>>>> +{
>>>> +	return 0;
>>>> +}
>>>> +
>>>> +static int vfio_platform_mmap(void *device_data, struct vm_area_struct
>>>> +*vma) {
>>>> +	return -EINVAL;
>>>> +}
>>>> +
>>>> +static const struct vfio_device_ops vfio_platform_ops = {
>>>> +	.name		= "vfio-platform",
>>>> +	.open		= vfio_platform_open,
>>>> +	.release	= vfio_platform_release,
>>>> +	.ioctl		= vfio_platform_ioctl,
>>>> +	.read		= vfio_platform_read,
>>>> +	.write		= vfio_platform_write,
>>>> +	.mmap		= vfio_platform_mmap,
>>>> +};
>>>> +
>>>> +static int vfio_platform_probe(struct platform_device *pdev) {
>>>> +	struct vfio_platform_device *vdev;
>>>> +	struct iommu_group *group;
>>>> +	int ret;
>>>> +
>>>> +	group = iommu_group_get(&pdev->dev);
>>>> +	if (!group) {
>>>> +		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
>>>> +		return -EINVAL;
>>>> +	}
>>>> +
>>>> +	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
>>>> +	if (!vdev) {
>>>> +		iommu_group_put(group);
>>>> +		return -ENOMEM;
>>>> +	}
>>>> +
>>>> +	vdev->pdev = pdev;
>>>> +
>>>> +	ret = vfio_add_group_dev(&pdev->dev,&vfio_platform_ops, vdev);
>>>> +	if (ret) {
>>>> +		iommu_group_put(group);
>>>> +		kfree(vdev);
>>>> +	}
>>>> +
>>>> +	return ret;
>>>> +}
>>>> +
>>>> +static int vfio_platform_remove(struct platform_device *pdev) {
>>>> +	struct vfio_platform_device *vdev;
>>>> +
>>>> +	vdev = vfio_del_group_dev(&pdev->dev);
>>>> +	if (!vdev)
>>>> +		return -EINVAL;
>>>> +
>>>> +	iommu_group_put(pdev->dev.iommu_group);
>>>> +	kfree(vdev);
>>>> +
>>>> +	return 0;
>>>> +}
>>>> +
>>>> +static const struct of_device_id vfio_platform_match[] = {
>>>> +	/* Before this series can be applied, we shall implement a better
>>>> +	 * mechanism to bind the module to any device. For now add the
>>>> +	 * compatible property to the dtb of the devices we want to use. */
>>>> +	{
>>>> +		.compatible = "vfio",
>>>> +	},
>>>> +	{},
>>>> +};
>>>> +MODULE_DEVICE_TABLE(of, vfio_platform_match);
>>>> +
>>>> +static struct platform_driver vfio_platform_driver = {
>>>> +	.probe		= vfio_platform_probe,
>>>> +	.remove		= vfio_platform_remove,
>>>> +	.driver	= {
>>>> +		.name	= "vfio-platform",
>>>> +		.owner	= THIS_MODULE,
>>>> +		.of_match_table = vfio_platform_match,
>>>> +	},
>>>> +};
>>>> +
>>>> +module_platform_driver(vfio_platform_driver);
>>>> +
>>>> +MODULE_VERSION(DRIVER_VERSION);
>>>> +MODULE_LICENSE("GPL v2");
>>>> +MODULE_AUTHOR(DRIVER_AUTHOR);
>>>> +MODULE_DESCRIPTION(DRIVER_DESC);
>>>> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h index
>>>> 284ff24..8d7434d 100644
>>>> --- a/include/uapi/linux/vfio.h
>>>> +++ b/include/uapi/linux/vfio.h
>>>> @@ -147,6 +147,7 @@ struct vfio_device_info {
>>>>    	__u32	flags;
>>>>    #define VFIO_DEVICE_FLAGS_RESET	(1<<   0)	/* Device supports reset */
>>>>    #define VFIO_DEVICE_FLAGS_PCI	(1<<   1)	/* vfio-pci device */
>>>> +#define VFIO_DEVICE_FLAGS_PLATFORM (1<<   2)	/* vfio-platform device */
>>>>    	__u32	num_regions;	/* Max region index + 1 */
>>>>    	__u32	num_irqs;	/* Max IRQ index + 1 */
>>>>    };
>>>> --
>>>> 1.8.1.2
>>>>
>>>> _______________________________________________
>>>> iommu mailing list
>>>> [e-mail]
>>>> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>>>
>>>
>>> _______________________________________________
>>> iommu mailing list
>>> [e-mail]
>>> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe kvm" in
>> the body of a message to [e-mail]
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>
>

--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Yoder Stuart-B08248 Oct. 29, 2013, 9:58 p.m. UTC | #9

> -----Original Message-----
> From: Don Dutile [mailto:[e-mail]]
> Sent: Tuesday, October 29, 2013 12:13 PM
> To: Alex Williamson
> Cc: Bhushan Bharat-R65777; Antonios Motakis;
> [e-mail]; [e-mail];
> [e-mail]; [e-mail]; Yoder Stuart-B08248;
> [e-mail]; [e-mail]
> Subject: Re: [PATCH 2/7] Initial skeleton of VFIO support for Device Tree
> based devices
> 
> On 10/29/2013 07:47 AM, Alex Williamson wrote:
> > On Mon, 2013-10-28 at 21:29 -0400, Don Dutile wrote:
> >> On 09/30/2013 11:37 AM, Bhushan Bharat-R65777 wrote:
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: [e-mail] [mailto:iommu-
> >>>> [e-mail]] On Behalf Of Antonios Motakis
> >>>> Sent: Monday, September 30, 2013 8:59 PM
> >>>> To: [e-mail]; [e-mail]
> >>>> Cc: [e-mail]; [e-mail];
> [e-mail]; Yoder
> >>>> Stuart-B08248; [e-mail]; Antonios Motakis;
> >>>> [e-mail]
> >>>> Subject: [PATCH 2/7] Initial skeleton of VFIO support for Device
> Tree based
> >>>> devices
> >>>>
> >>>> Platform devices in the Linux kernel are usually managed by the DT
> interface.
> >>>> This patch forms the base to support these kind of devices with
> VFIO.
> >>>>
> >>>> Signed-off-by: Antonios Motakis<[e-mail]>
> >>>> ---
> >>>>    drivers/vfio/Kconfig         |  11 +++
> >>>>    drivers/vfio/Makefile        |   1 +
> >>>>    drivers/vfio/vfio_platform.c | 187
> +++++++++++++++++++++++++++++++++++++++++++
> >>>>    include/uapi/linux/vfio.h    |   1 +
> >>>>    4 files changed, 200 insertions(+)
> >>>>    create mode 100644 drivers/vfio/vfio_platform.c
> >>>>
> >>>> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index
> 1f84eda..35254b7
> >>>> 100644
> >>>> --- a/drivers/vfio/Kconfig
> >>>> +++ b/drivers/vfio/Kconfig
> >>>> @@ -13,4 +13,15 @@ menuconfig VFIO
> >>>>
> >>>>    	  If you don't know what to do here, say N.
> >>>>
> >>>> +config VFIO_PLATFORM
> >>>> +	tristate "VFIO support for device tree based platform
> devices"
> >>>> +	depends on VFIO&&   EVENTFD&&   OF
> >>>> +	help
> >>>> +	  Support for platform devices with VFIO. This is required to
> make
> >>>> +	  use of platform devices present on device tree nodes using
> the VFIO
> >>>> +	  framework. Devices that are not described in the device
> tree cannot
> >>>> +	  be used by this driver.
> >>>> +
> >>>> +	  If you don't know what to do here, say N.
> >>>> +
> >>>>    source "drivers/vfio/pci/Kconfig"
> >>>> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index
> >>>> 2398d4a..575c8dd 100644
> >>>> --- a/drivers/vfio/Makefile
> >>>> +++ b/drivers/vfio/Makefile
> >>>> @@ -1,3 +1,4 @@
> >>>>    obj-$(CONFIG_VFIO) += vfio.o
> >>>>    obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
> >>>>    obj-$(CONFIG_VFIO_PCI) += pci/
> >>>> +obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
> >>>> diff --git a/drivers/vfio/vfio_platform.c
> b/drivers/vfio/vfio_platform.c new
> >>>
> >>> We can make this parallel to PCI, something like
> drivers/vfio/platform/platform.c
> >>>
> >> pls, no.  'platform' is too generic, and it really means 'arm-dt' ...
> so can
> >> move it to the arch/arm space, and have it's kconfig conditional on
> ARM&&VFIO.
> >> if kept under drivers/vfio, then use a better directory name that ties
> it to arm-dt.
> >> thanks.
> >
> > The intention is that vfio platform device support is not arm-dt
> > specific.  This is to be used by both arm and embedded ppc.  The
> devices
> > we intend to support with them are known as platform drivers in the
> > kernel, thus the name.  I suppose the question remains whether the
> > interface here is really generic for any "platform" device or whether
> > we're making whether we're making an interface specifically for device
> > tree platform devices, or if those are one in the same.  In any case,
> > arm-dt is certainly not the answer.
> >
> > Alex
> >
> I thought that was the intention until I saw this use in platform.c:
> 	static const struct of_device_id vfio_platform_match[] = {
> 
> So, of_device_id hit me as DT-specific, and thus, the file should have
> a name that implies its not as platform-generic as one may want/expect.
> I agree that the 'arm' part can be dropped, but it's not of-agnostic atm.

That part of the patch needs to be re-worked...and the comment
indicates it is not final.  The vfio-platform driver should not register
to handle specific compatible strings.  Kim Phillips has a 
patch posted that adds 'wildcard' support to platform drivers
so that a driver can bind to any device.

Stuart

--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig
index 1f84eda..35254b7 100644
--- a/drivers/vfio/Kconfig
+++ b/drivers/vfio/Kconfig
@@ -13,4 +13,15 @@  menuconfig VFIO
 
 	  If you don't know what to do here, say N.
 
+config VFIO_PLATFORM
+	tristate "VFIO support for device tree based platform devices"
+	depends on VFIO && EVENTFD && OF
+	help
+	  Support for platform devices with VFIO. This is required to make
+	  use of platform devices present on device tree nodes using the VFIO
+	  framework. Devices that are not described in the device tree cannot
+	  be used by this driver.
+
+	  If you don't know what to do here, say N.
+
 source "drivers/vfio/pci/Kconfig"
diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile
index 2398d4a..575c8dd 100644
--- a/drivers/vfio/Makefile
+++ b/drivers/vfio/Makefile
@@ -1,3 +1,4 @@ 
 obj-$(CONFIG_VFIO) += vfio.o
 obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o
 obj-$(CONFIG_VFIO_PCI) += pci/
+obj-$(CONFIG_VFIO_PLATFORM) += vfio_platform.o
diff --git a/drivers/vfio/vfio_platform.c b/drivers/vfio/vfio_platform.c
new file mode 100644
index 0000000..b9686b0
--- /dev/null
+++ b/drivers/vfio/vfio_platform.c
@@ -0,0 +1,187 @@ 
+/*
+ * Copyright (C) 2013 - Virtual Open Systems
+ * Author: Antonios Motakis <[e-mail]>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+
+#include <linux/device.h>
+#include <linux/eventfd.h>
+#include <linux/interrupt.h>
+#include <linux/iommu.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/notifier.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <linux/uaccess.h>
+#include <linux/vfio.h>
+
+#define DRIVER_VERSION  "0.1"
+#define DRIVER_AUTHOR   "Antonios Motakis <[e-mail]>"
+#define DRIVER_DESC     "VFIO Device Tree devices - User Level meta-driver"
+
+struct vfio_platform_device {
+	struct platform_device	*pdev;
+};
+
+static void vfio_platform_release(void *device_data)
+{
+	module_put(THIS_MODULE);
+}
+
+static int vfio_platform_open(void *device_data)
+{
+	if (!try_module_get(THIS_MODULE))
+		return -ENODEV;
+
+	return 0;
+}
+
+static long vfio_platform_ioctl(void *device_data,
+			   unsigned int cmd, unsigned long arg)
+{
+	struct vfio_platform_device *vdev = device_data;
+	unsigned long minsz;
+
+	if (cmd == VFIO_DEVICE_GET_INFO) {
+		struct vfio_device_info info;
+
+		minsz = offsetofend(struct vfio_device_info, num_irqs);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if (info.argsz < minsz)
+			return -EINVAL;
+
+		info.flags = VFIO_DEVICE_FLAGS_PLATFORM;
+		info.num_regions = 0;
+		info.num_irqs = 0;
+
+		return copy_to_user((void __user *)arg, &info, minsz);
+
+	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO)
+		return -EINVAL;
+
+	else if (cmd == VFIO_DEVICE_GET_IRQ_INFO)
+		return -EINVAL;
+
+	else if (cmd == VFIO_DEVICE_SET_IRQS)
+		return -EINVAL;
+
+	else if (cmd == VFIO_DEVICE_RESET)
+		return -EINVAL;
+
+	return -ENOTTY;
+}
+
+static ssize_t vfio_platform_read(void *device_data, char __user *buf,
+			     size_t count, loff_t *ppos)
+{
+	return 0;
+}
+
+static ssize_t vfio_platform_write(void *device_data, const char __user *buf,
+			      size_t count, loff_t *ppos)
+{
+	return 0;
+}
+
+static int vfio_platform_mmap(void *device_data, struct vm_area_struct *vma)
+{
+	return -EINVAL;
+}
+
+static const struct vfio_device_ops vfio_platform_ops = {
+	.name		= "vfio-platform",
+	.open		= vfio_platform_open,
+	.release	= vfio_platform_release,
+	.ioctl		= vfio_platform_ioctl,
+	.read		= vfio_platform_read,
+	.write		= vfio_platform_write,
+	.mmap		= vfio_platform_mmap,
+};
+
+static int vfio_platform_probe(struct platform_device *pdev)
+{
+	struct vfio_platform_device *vdev;
+	struct iommu_group *group;
+	int ret;
+
+	group = iommu_group_get(&pdev->dev);
+	if (!group) {
+		pr_err("VFIO: No IOMMU group for device %s\n", pdev->name);
+		return -EINVAL;
+	}
+
+	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
+	if (!vdev) {
+		iommu_group_put(group);
+		return -ENOMEM;
+	}
+
+	vdev->pdev = pdev;
+
+	ret = vfio_add_group_dev(&pdev->dev, &vfio_platform_ops, vdev);
+	if (ret) {
+		iommu_group_put(group);
+		kfree(vdev);
+	}
+
+	return ret;
+}
+
+static int vfio_platform_remove(struct platform_device *pdev)
+{
+	struct vfio_platform_device *vdev;
+
+	vdev = vfio_del_group_dev(&pdev->dev);
+	if (!vdev)
+		return -EINVAL;
+
+	iommu_group_put(pdev->dev.iommu_group);
+	kfree(vdev);
+
+	return 0;
+}
+
+static const struct of_device_id vfio_platform_match[] = {
+	/* Before this series can be applied, we shall implement a better
+	 * mechanism to bind the module to any device. For now add the
+	 * compatible property to the dtb of the devices we want to use. */
+	{
+		.compatible = "vfio",
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, vfio_platform_match);
+
+static struct platform_driver vfio_platform_driver = {
+	.probe		= vfio_platform_probe,
+	.remove		= vfio_platform_remove,
+	.driver	= {
+		.name	= "vfio-platform",
+		.owner	= THIS_MODULE,
+		.of_match_table = vfio_platform_match,
+	},
+};
+
+module_platform_driver(vfio_platform_driver);
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index 284ff24..8d7434d 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -147,6 +147,7 @@  struct vfio_device_info {
 	__u32	flags;
 #define VFIO_DEVICE_FLAGS_RESET	(1 << 0)	/* Device supports reset */
 #define VFIO_DEVICE_FLAGS_PCI	(1 << 1)	/* vfio-pci device */
+#define VFIO_DEVICE_FLAGS_PLATFORM (1 << 2)	/* vfio-platform device */
 	__u32	num_regions;	/* Max region index + 1 */
 	__u32	num_irqs;	/* Max IRQ index + 1 */
 };