[Patch V4 0/3] xen, usb: support pvUSB frontend driver

[Patch V4 0/3] xen, usb: support pvUSB frontend driver

[Juergen Gross]

This series adds XEN guest pvUSB support. With pvUSB it is possible to
use physical USB devices from a XEN domain.

The support consists of a frontend in form of a virtual hcd driver in
the unprivileged domU passing I/O-requests to the backend in a driver
domain (usually Dom0). The backend is not part of this patch series,
as it will be supported via qemu.

The code is taken (and adapted) from the original pvUSB implementation
done for Linux 2.6 in 2008 by Fujitsu.

Normal operation of USB devices by adding and removing them dynamically
to/from a domain has been tested using various USB devices (USB 1.1,
2.0 and 3.0). The pvUSB backend for these tests was a SUSE SLES Dom0
with a kernel based backend driver.

Changes in V4:
- remove sysfs file from frontend, as it violated the "one value per file"
  rule and didn't serve any real purpose.

Changes in V3:
- move frontend to drivers/usb/host and rename it to xen-hcd.
- changed name prefixes in patch 1 to "xenusb" as requested by Greg
- use __u<n> types rather than uint<n>_t as requested by Greg

Changes in V2:
- removed backend, as it can be implemented in user land
- added some access macros and definitions to the pvUSB interface
  description to make it independant from linux kernel USB internals
- adapted frontend to newer kernel version and use new pvUSB
  interface macros
- set port status in one chunk as suggested by Oliver Neukum


Juergen Gross (3):
  usb: Add Xen pvUSB protocol description
  usb: Introduce Xen pvUSB frontend (xen hcd)
  xen: add Xen pvUSB maintainer

 MAINTAINERS                      |    8 +
 drivers/usb/host/Kconfig         |   11 +
 drivers/usb/host/Makefile        |    1 +
 drivers/usb/host/xen-hcd.c       | 1594 ++++++++++++++++++++++++++++++++++++++
 include/xen/interface/io/usbif.h |  252 ++++++
 5 files changed, 1866 insertions(+)
 create mode 100644 drivers/usb/host/xen-hcd.c
 create mode 100644 include/xen/interface/io/usbif.h

-- 
2.1.4

[Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Juergen Gross]

Add the definition of pvUSB protocol used between the pvUSB frontend in
a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).

This header was originally provided by Fujitsu for Xen based on Linux
2.6.18.

Changes are:
- adapt to Linux style guide

Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
---
 include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 252 insertions(+)
 create mode 100644 include/xen/interface/io/usbif.h

diff --git a/include/xen/interface/io/usbif.h b/include/xen/interface/io/usbif.h
new file mode 100644
index 0000000..4ce0b40
--- /dev/null
+++ b/include/xen/interface/io/usbif.h
@@ -0,0 +1,252 @@
+/*
+ * usbif.h
+ *
+ * USB I/O interface for Xen guest OSes.
+ *
+ * Copyright (C) 2009, FUJITSU LABORATORIES LTD.
+ * Author: Noboru Iwamatsu <n_iwamatsu-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef __XEN_PUBLIC_IO_USBIF_H__
+#define __XEN_PUBLIC_IO_USBIF_H__
+
+#include "ring.h"
+#include "../grant_table.h"
+
+/*
+ * Feature and Parameter Negotiation
+ * =================================
+ * The two halves of a Xen pvUSB driver utilize nodes within the XenStore to
+ * communicate capabilities and to negotiate operating parameters. This
+ * section enumerates these nodes which reside in the respective front and
+ * backend portions of the XenStore, following the XenBus convention.
+ *
+ * Any specified default value is in effect if the corresponding XenBus node
+ * is not present in the XenStore.
+ *
+ * XenStore nodes in sections marked "PRIVATE" are solely for use by the
+ * driver side whose XenBus tree contains them.
+ *
+ *****************************************************************************
+ *                            Backend XenBus Nodes
+ *****************************************************************************
+ *
+ *------------------ Backend Device Identification (PRIVATE) ------------------
+ *
+ * num-ports
+ *      Values:         unsigned [1...31]
+ *
+ *      Number of ports for this (virtual) USB host connector.
+ *
+ * usb-ver
+ *      Values:         unsigned [1...2]
+ *
+ *      USB version of this host connector: 1 = USB 1.1, 2 = USB 2.0.
+ *
+ * port/[1...31]
+ *      Values:         string
+ *
+ *      Physical USB device connected to the given port, e.g. "3-1.5".
+ *
+ *****************************************************************************
+ *                            Frontend XenBus Nodes
+ *****************************************************************************
+ *
+ *----------------------- Request Transport Parameters -----------------------
+ *
+ * event-channel
+ *      Values:         unsigned
+ *
+ *      The identifier of the Xen event channel used to signal activity
+ *      in the ring buffer.
+ *
+ * urb-ring-ref
+ *      Values:         unsigned
+ *
+ *      The Xen grant reference granting permission for the backend to map
+ *      the sole page in a single page sized ring buffer. This is the ring
+ *      buffer for urb requests.
+ *
+ * conn-ring-ref
+ *      Values:         unsigned
+ *
+ *      The Xen grant reference granting permission for the backend to map
+ *      the sole page in a single page sized ring buffer. This is the ring
+ *      buffer for connection/disconnection requests.
+ *
+ * protocol
+ *      Values:         string (XEN_IO_PROTO_ABI_*)
+ *      Default Value:  XEN_IO_PROTO_ABI_NATIVE
+ *
+ *      The machine ABI rules governing the format of all ring request and
+ *      response structures.
+ *
+ */
+
+enum xenusb_spec_version {
+	XENUSB_VER_UNKNOWN = 0,
+	XENUSB_VER_USB11,
+	XENUSB_VER_USB20,
+	XENUSB_VER_USB30,	/* not supported yet */
+};
+
+/*
+ *  USB pipe in xenusb_request
+ *
+ *  - port number:	bits 0-4
+ *				(USB_MAXCHILDREN is 31)
+ *
+ *  - operation flag:	bit 5
+ *				(0 = submit urb,
+ *				 1 = unlink urb)
+ *
+ *  - direction:	bit 7
+ *				(0 = Host-to-Device [Out]
+ *				 1 = Device-to-Host [In])
+ *
+ *  - device address:	bits 8-14
+ *
+ *  - endpoint:		bits 15-18
+ *
+ *  - pipe type:	bits 30-31
+ *				(00 = isochronous, 01 = interrupt,
+ *				 10 = control, 11 = bulk)
+ */
+
+#define XENUSB_PIPE_PORT_MASK	0x0000001f
+#define XENUSB_PIPE_UNLINK	0x00000020
+#define XENUSB_PIPE_DIR		0x00000080
+#define XENUSB_PIPE_DEV_MASK	0x0000007f
+#define XENUSB_PIPE_DEV_SHIFT	8
+#define XENUSB_PIPE_EP_MASK	0x0000000f
+#define XENUSB_PIPE_EP_SHIFT	15
+#define XENUSB_PIPE_TYPE_MASK	0x00000003
+#define XENUSB_PIPE_TYPE_SHIFT	30
+#define XENUSB_PIPE_TYPE_ISOC	0
+#define XENUSB_PIPE_TYPE_INT	1
+#define XENUSB_PIPE_TYPE_CTRL	2
+#define XENUSB_PIPE_TYPE_BULK	3
+
+#define xenusb_pipeportnum(pipe)		((pipe) & XENUSB_PIPE_PORT_MASK)
+#define xenusb_setportnum_pipe(pipe, portnum)	((pipe) | (portnum))
+
+#define xenusb_pipeunlink(pipe)			((pipe) & XENUSB_PIPE_UNLINK)
+#define xenusb_pipesubmit(pipe)			(!xenusb_pipeunlink(pipe))
+#define xenusb_setunlink_pipe(pipe)		((pipe) | XENUSB_PIPE_UNLINK)
+
+#define xenusb_pipein(pipe)			((pipe) & XENUSB_PIPE_DIR)
+#define xenusb_pipeout(pipe)			(!xenusb_pipein(pipe))
+
+#define xenusb_pipedevice(pipe)			\
+		(((pipe) >> XENUSB_PIPE_DEV_SHIFT) & XENUSB_PIPE_DEV_MASK)
+
+#define xenusb_pipeendpoint(pipe)		\
+		(((pipe) >> XENUSB_PIPE_EP_SHIFT) & XENUSB_PIPE_EP_MASK)
+
+#define xenusb_pipetype(pipe)			\
+		(((pipe) >> XENUSB_PIPE_TYPE_SHIFT) & XENUSB_PIPE_TYPE_MASK)
+#define xenusb_pipeisoc(pipe)	(xenusb_pipetype(pipe) == XENUSB_PIPE_TYPE_ISOC)
+#define xenusb_pipeint(pipe)	(xenusb_pipetype(pipe) == XENUSB_PIPE_TYPE_INT)
+#define xenusb_pipectrl(pipe)	(xenusb_pipetype(pipe) == XENUSB_PIPE_TYPE_CTRL)
+#define xenusb_pipebulk(pipe)	(xenusb_pipetype(pipe) == XENUSB_PIPE_TYPE_BULK)
+
+#define XENUSB_MAX_SEGMENTS_PER_REQUEST (16)
+#define XENUSB_MAX_PORTNR	31
+
+/*
+ * RING for transferring urbs.
+ */
+struct xenusb_request_segment {
+	grant_ref_t gref;
+	__u16 offset;
+	__u16 length;
+};
+
+struct xenusb_urb_request {
+	__u16 id;			/* request id */
+	__u16 nr_buffer_segs;	/* # of urb->transfer_buffer segments */
+
+	/* basic urb parameter */
+	__u32 pipe;
+	__u16 transfer_flags;
+	__u16 buffer_length;
+	union {
+		__u8 ctrl[8];			/* pipe type control */
+						/* setup packet */
+
+		struct {
+			__u16 interval;		/* max (1024*8) in usb core */
+			__u16 start_frame;		/* start frame */
+			__u16 number_of_packets;	/* # of ISO packets */
+			__u16 nr_frame_desc_segs;
+					/* # of iso_frame_desc segments */
+		} isoc;				/* pipe type isochronous */
+
+		struct {
+			__u16 interval;		/* max (1024*8) in usb core */
+			__u16 pad[3];
+		} intr;				/* pipe type interrupt */
+
+		struct {
+			__u16 unlink_id;	/* unlink request id */
+			__u16 pad[3];
+		} unlink;			/* pipe unlink */
+
+	} u;
+
+	/* urb data segments */
+	struct xenusb_request_segment seg[XENUSB_MAX_SEGMENTS_PER_REQUEST];
+};
+
+struct xenusb_urb_response {
+	__u16 id;		/* request id */
+	__u16 start_frame;	/* start frame (ISO) */
+	__s32 status;		/* status (non-ISO) */
+	__s32 actual_length;	/* actual transfer length */
+	__s32 error_count;	/* number of ISO errors */
+};
+
+DEFINE_RING_TYPES(xenusb_urb, struct xenusb_urb_request,
+		  struct xenusb_urb_response);
+#define XENUSB_URB_RING_SIZE __CONST_RING_SIZE(xenusb_urb, PAGE_SIZE)
+
+/*
+ * RING for notifying connect/disconnect events to frontend
+ */
+struct xenusb_conn_request {
+	__u16 id;
+};
+
+struct xenusb_conn_response {
+	__u16 id;		/* request id */
+	__u8 portnum;		/* port number */
+	__u8 speed;		/* usb device speed */
+#define XENUSB_SPEED_NONE	0
+#define XENUSB_SPEED_LOW	1
+#define XENUSB_SPEED_FULL	2
+#define XENUSB_SPEED_HIGH	3
+};
+
+DEFINE_RING_TYPES(xenusb_conn, struct xenusb_conn_request,
+		  struct xenusb_conn_response);
+#define XENUSB_CONN_RING_SIZE __CONST_RING_SIZE(xenusb_conn, PAGE_SIZE)
+
+#endif /* __XEN_PUBLIC_IO_USBIF_H__ */
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in

[Patch V4 2/3] usb: Introduce Xen pvUSB frontend (xen hcd)

[Juergen Gross]

Introduces the Xen pvUSB frontend. With pvUSB it is possible for a Xen
domU to communicate with a USB device assigned to that domU. The
communication is all done via the pvUSB backend in a driver domain
(usually Dom0) which is owner of the physical device.

The pvUSB frontend is a USB hcd for a virtual USB host connector.

The code is taken from the pvUSB implementation in Xen done by Fujitsu
based on Linux kernel 2.6.18.

Changes from the original version are:
- port to upstream kernel
- put all code in just one source file
- move module to appropriate location in kernel tree
- adapt to Linux style guide
- minor code modifications to increase readability

Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
---
 drivers/usb/host/Kconfig   |   11 +
 drivers/usb/host/Makefile  |    1 +
 drivers/usb/host/xen-hcd.c | 1594 ++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 1606 insertions(+)
 create mode 100644 drivers/usb/host/xen-hcd.c

diff --git a/drivers/usb/host/Kconfig b/drivers/usb/host/Kconfig
index 197a6a3..3361b4b 100644
--- a/drivers/usb/host/Kconfig
+++ b/drivers/usb/host/Kconfig
@@ -788,3 +788,14 @@ config USB_HCD_TEST_MODE
 	  This option is of interest only to developers who need to validate
 	  their USB hardware designs.  It is not needed for normal use.  If
 	  unsure, say N.
+
+config USB_XEN_HCD
+	tristate "Xen usb virtual host driver"
+	depends on XEN
+	select XEN_XENBUS_FRONTEND
+	help
+	  The Xen usb virtual host driver serves as a frontend driver enabling
+	  a Xen guest system to access USB Devices passed through to the guest
+	  by the Xen host (usually Dom0).
+	  Only needed if the kernel is running in a Xen guest and generic
+	  access to a USB device is needed.
diff --git a/drivers/usb/host/Makefile b/drivers/usb/host/Makefile
index 65b0b6a..3779696 100644
--- a/drivers/usb/host/Makefile
+++ b/drivers/usb/host/Makefile
@@ -75,3 +75,4 @@ obj-$(CONFIG_USB_HCD_SSB)	+= ssb-hcd.o
 obj-$(CONFIG_USB_FUSBH200_HCD)	+= fusbh200-hcd.o
 obj-$(CONFIG_USB_FOTG210_HCD)	+= fotg210-hcd.o
 obj-$(CONFIG_USB_MAX3421_HCD)	+= max3421-hcd.o
+obj-$(CONFIG_USB_XEN_HCD)	+= xen-hcd.o
diff --git a/drivers/usb/host/xen-hcd.c b/drivers/usb/host/xen-hcd.c
new file mode 100644
index 0000000..982d986
--- /dev/null
+++ b/drivers/usb/host/xen-hcd.c
@@ -0,0 +1,1594 @@
+/*
+ * xen-hcd.c
+ *
+ * Xen USB Virtual Host Controller driver
+ *
+ * Copyright (C) 2009, FUJITSU LABORATORIES LTD.
+ * Author: Noboru Iwamatsu <n_iwamatsu-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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, see <http://www.gnu.org/licenses/>.
+ *
+ * Or, by your choice:
+ *
+ * When distributed separately from the Linux kernel or incorporated into
+ * other software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/module.h>
+#include <linux/usb.h>
+#include <linux/list.h>
+#include <linux/usb/hcd.h>
+#include <linux/io.h>
+
+#include <xen/xen.h>
+#include <xen/xenbus.h>
+#include <xen/grant_table.h>
+#include <xen/events.h>
+#include <xen/page.h>
+
+#include <xen/interface/io/usbif.h>
+
+/* Private per-URB data */
+struct urb_priv {
+	struct list_head list;
+	struct urb *urb;
+	int req_id;		/* RING_REQUEST id for submitting */
+	int unlink_req_id;	/* RING_REQUEST id for unlinking */
+	int status;
+	unsigned unlinked:1;	/* dequeued marker */
+};
+
+/* virtual roothub port status */
+struct rhport_status {
+	__u32 status;
+	unsigned resuming:1;		/* in resuming */
+	unsigned c_connection:1;	/* connection changed */
+	unsigned long timeout;
+};
+
+/* status of attached device */
+struct vdevice_status {
+	int devnum;
+	enum usb_device_state status;
+	enum usb_device_speed speed;
+};
+
+/* RING request shadow */
+struct usb_shadow {
+	struct xenusb_urb_request req;
+	struct urb *urb;
+};
+
+struct xenhcd_info {
+	/* Virtual Host Controller has 4 urb queues */
+	struct list_head pending_submit_list;
+	struct list_head pending_unlink_list;
+	struct list_head in_progress_list;
+	struct list_head giveback_waiting_list;
+
+	spinlock_t lock;
+
+	/* timer that kick pending and giveback waiting urbs */
+	struct timer_list watchdog;
+	unsigned long actions;
+
+	/* virtual root hub */
+	int rh_numports;
+	struct rhport_status ports[XENUSB_MAX_PORTNR];
+	struct vdevice_status devices[XENUSB_MAX_PORTNR];
+
+	/* Xen related staff */
+	struct xenbus_device *xbdev;
+	int urb_ring_ref;
+	int conn_ring_ref;
+	struct xenusb_urb_front_ring urb_ring;
+	struct xenusb_conn_front_ring conn_ring;
+
+	unsigned int evtchn;
+	unsigned int irq;
+	struct usb_shadow shadow[XENUSB_URB_RING_SIZE];
+	unsigned shadow_free;
+};
+
+#define GRANT_INVALID_REF 0
+
+#define XENHCD_RING_JIFFIES (HZ/200)
+#define XENHCD_SCAN_JIFFIES 1
+
+enum xenhcd_timer_action {
+	TIMER_RING_WATCHDOG,
+	TIMER_SCAN_PENDING_URBS,
+};
+
+static struct kmem_cache *xenhcd_urbp_cachep;
+
+static inline struct xenhcd_info *xenhcd_hcd_to_info(struct usb_hcd *hcd)
+{
+	return (struct xenhcd_info *)hcd->hcd_priv;
+}
+
+static inline struct usb_hcd *xenhcd_info_to_hcd(struct xenhcd_info *info)
+{
+	return container_of((void *)info, struct usb_hcd, hcd_priv);
+}
+
+static inline void xenhcd_timer_action_done(struct xenhcd_info *info,
+					    enum xenhcd_timer_action action)
+{
+	clear_bit(action, &info->actions);
+}
+
+static void xenhcd_timer_action(struct xenhcd_info *info,
+				enum xenhcd_timer_action action)
+{
+	if (timer_pending(&info->watchdog) &&
+	    test_bit(TIMER_SCAN_PENDING_URBS, &info->actions))
+		return;
+
+	if (!test_and_set_bit(action, &info->actions)) {
+		unsigned long t;
+
+		switch (action) {
+		case TIMER_RING_WATCHDOG:
+			t = XENHCD_RING_JIFFIES;
+			break;
+		default:
+			t = XENHCD_SCAN_JIFFIES;
+			break;
+		}
+		mod_timer(&info->watchdog, t + jiffies);
+	}
+}
+
+/*
+ * set virtual port connection status
+ */
+static void xenhcd_set_connect_state(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	if (info->ports[port].status & USB_PORT_STAT_POWER) {
+		switch (info->devices[port].speed) {
+		case XENUSB_SPEED_NONE:
+			info->ports[port].status &=
+				~(USB_PORT_STAT_CONNECTION |
+				  USB_PORT_STAT_ENABLE |
+				  USB_PORT_STAT_LOW_SPEED |
+				  USB_PORT_STAT_HIGH_SPEED |
+				  USB_PORT_STAT_SUSPEND);
+			break;
+		case XENUSB_SPEED_LOW:
+			info->ports[port].status |= USB_PORT_STAT_CONNECTION;
+			info->ports[port].status |= USB_PORT_STAT_LOW_SPEED;
+			break;
+		case XENUSB_SPEED_FULL:
+			info->ports[port].status |= USB_PORT_STAT_CONNECTION;
+			break;
+		case XENUSB_SPEED_HIGH:
+			info->ports[port].status |= USB_PORT_STAT_CONNECTION;
+			info->ports[port].status |= USB_PORT_STAT_HIGH_SPEED;
+			break;
+		default: /* error */
+			return;
+		}
+		info->ports[port].status |= (USB_PORT_STAT_C_CONNECTION << 16);
+	}
+}
+
+/*
+ * set virtual device connection status
+ */
+static void xenhcd_rhport_connect(struct xenhcd_info *info, __u8 portnum,
+				  __u8 speed)
+{
+	int port;
+
+	if (portnum < 1 || portnum > info->rh_numports)
+		return; /* invalid port number */
+
+	port = portnum - 1;
+	if (info->devices[port].speed != speed) {
+		switch (speed) {
+		case XENUSB_SPEED_NONE: /* disconnect */
+			info->devices[port].status = USB_STATE_NOTATTACHED;
+			break;
+		case XENUSB_SPEED_LOW:
+		case XENUSB_SPEED_FULL:
+		case XENUSB_SPEED_HIGH:
+			info->devices[port].status = USB_STATE_ATTACHED;
+			break;
+		default: /* error */
+			return;
+		}
+		info->devices[port].speed = speed;
+		info->ports[port].c_connection = 1;
+
+		xenhcd_set_connect_state(info, portnum);
+	}
+}
+
+/*
+ * SetPortFeature(PORT_SUSPENDED)
+ */
+static void xenhcd_rhport_suspend(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	info->ports[port].status |= USB_PORT_STAT_SUSPEND;
+	info->devices[port].status = USB_STATE_SUSPENDED;
+}
+
+/*
+ * ClearPortFeature(PORT_SUSPENDED)
+ */
+static void xenhcd_rhport_resume(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	if (info->ports[port].status & USB_PORT_STAT_SUSPEND) {
+		info->ports[port].resuming = 1;
+		info->ports[port].timeout = jiffies + msecs_to_jiffies(20);
+	}
+}
+
+/*
+ * SetPortFeature(PORT_POWER)
+ */
+static void xenhcd_rhport_power_on(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	if ((info->ports[port].status & USB_PORT_STAT_POWER) == 0) {
+		info->ports[port].status |= USB_PORT_STAT_POWER;
+		if (info->devices[port].status != USB_STATE_NOTATTACHED)
+			info->devices[port].status = USB_STATE_POWERED;
+		if (info->ports[port].c_connection)
+			xenhcd_set_connect_state(info, portnum);
+	}
+}
+
+/*
+ * ClearPortFeature(PORT_POWER)
+ * SetConfiguration(non-zero)
+ * Power_Source_Off
+ * Over-current
+ */
+static void xenhcd_rhport_power_off(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	if (info->ports[port].status & USB_PORT_STAT_POWER) {
+		info->ports[port].status = 0;
+		if (info->devices[port].status != USB_STATE_NOTATTACHED)
+			info->devices[port].status = USB_STATE_ATTACHED;
+	}
+}
+
+/*
+ * ClearPortFeature(PORT_ENABLE)
+ */
+static void xenhcd_rhport_disable(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	info->ports[port].status &= ~USB_PORT_STAT_ENABLE;
+	info->ports[port].status &= ~USB_PORT_STAT_SUSPEND;
+	info->ports[port].resuming = 0;
+	if (info->devices[port].status != USB_STATE_NOTATTACHED)
+		info->devices[port].status = USB_STATE_POWERED;
+}
+
+/*
+ * SetPortFeature(PORT_RESET)
+ */
+static void xenhcd_rhport_reset(struct xenhcd_info *info, int portnum)
+{
+	int port;
+
+	port = portnum - 1;
+	info->ports[port].status &= ~(USB_PORT_STAT_ENABLE |
+				      USB_PORT_STAT_LOW_SPEED |
+				      USB_PORT_STAT_HIGH_SPEED);
+	info->ports[port].status |= USB_PORT_STAT_RESET;
+
+	if (info->devices[port].status != USB_STATE_NOTATTACHED)
+		info->devices[port].status = USB_STATE_ATTACHED;
+
+	/* 10msec reset signaling */
+	info->ports[port].timeout = jiffies + msecs_to_jiffies(10);
+}
+
+#ifdef CONFIG_PM
+static int xenhcd_bus_suspend(struct usb_hcd *hcd)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+	int ret = 0;
+	int i, ports;
+
+	ports = info->rh_numports;
+
+	spin_lock_irq(&info->lock);
+	if (!test_bit(HCD_FLAG_HW_ACCESSIBLE, &hcd->flags)) {
+		ret = -ESHUTDOWN;
+	} else {
+		/* suspend any active ports*/
+		for (i = 1; i <= ports; i++)
+			xenhcd_rhport_suspend(info, i);
+	}
+	spin_unlock_irq(&info->lock);
+
+	del_timer_sync(&info->watchdog);
+
+	return ret;
+}
+
+static int xenhcd_bus_resume(struct usb_hcd *hcd)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+	int ret = 0;
+	int i, ports;
+
+	ports = info->rh_numports;
+
+	spin_lock_irq(&info->lock);
+	if (!test_bit(HCD_FLAG_HW_ACCESSIBLE, &hcd->flags)) {
+		ret = -ESHUTDOWN;
+	} else {
+		/* resume any suspended ports*/
+		for (i = 1; i <= ports; i++)
+			xenhcd_rhport_resume(info, i);
+	}
+	spin_unlock_irq(&info->lock);
+
+	return ret;
+}
+#endif
+
+static void xenhcd_hub_descriptor(struct xenhcd_info *info,
+				  struct usb_hub_descriptor *desc)
+{
+	__u16 temp;
+	int ports = info->rh_numports;
+
+	desc->bDescriptorType = 0x29;
+	desc->bPwrOn2PwrGood = 10; /* EHCI says 20ms max */
+	desc->bHubContrCurrent = 0;
+	desc->bNbrPorts = ports;
+
+	/* size of DeviceRemovable and PortPwrCtrlMask fields */
+	temp = 1 + (ports / 8);
+	desc->bDescLength = 7 + 2 * temp;
+
+	/* bitmaps for DeviceRemovable and PortPwrCtrlMask */
+	memset(&desc->u.hs.DeviceRemovable[0], 0, temp);
+	memset(&desc->u.hs.DeviceRemovable[temp], 0xff, temp);
+
+	/* per-port over current reporting and no power switching */
+	temp = 0x000a;
+	desc->wHubCharacteristics = cpu_to_le16(temp);
+}
+
+/* port status change mask for hub_status_data */
+#define PORT_C_MASK	((USB_PORT_STAT_C_CONNECTION |		\
+			  USB_PORT_STAT_C_ENABLE |		\
+			  USB_PORT_STAT_C_SUSPEND |		\
+			  USB_PORT_STAT_C_OVERCURRENT |		\
+			  USB_PORT_STAT_C_RESET) << 16)
+
+/*
+ * See USB 2.0 Spec, 11.12.4 Hub and Port Status Change Bitmap.
+ * If port status changed, writes the bitmap to buf and return
+ * that length(number of bytes).
+ * If Nothing changed, return 0.
+ */
+static int xenhcd_hub_status_data(struct usb_hcd *hcd, char *buf)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+	int ports;
+	int i;
+	unsigned long flags;
+	int ret;
+	int changed = 0;
+
+	/* initialize the status to no-changes */
+	ports = info->rh_numports;
+	ret = 1 + (ports / 8);
+	memset(buf, 0, ret);
+
+	spin_lock_irqsave(&info->lock, flags);
+
+	for (i = 0; i < ports; i++) {
+		/* check status for each port */
+		if (info->ports[i].status & PORT_C_MASK) {
+			buf[(i + 1) / 8] |= 1 << (i + 1) % 8;
+			changed = 1;
+		}
+	}
+
+	if ((hcd->state == HC_STATE_SUSPENDED) && (changed == 1))
+		usb_hcd_resume_root_hub(hcd);
+
+	spin_unlock_irqrestore(&info->lock, flags);
+
+	return changed ? ret : 0;
+}
+
+static int xenhcd_hub_control(struct usb_hcd *hcd, __u16 typeReq, __u16 wValue,
+			      __u16 wIndex, char *buf, __u16 wLength)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+	int ports = info->rh_numports;
+	unsigned long flags;
+	int ret = 0;
+	int i;
+	int changed = 0;
+
+	spin_lock_irqsave(&info->lock, flags);
+	switch (typeReq) {
+	case ClearHubFeature:
+		/* ignore this request */
+		break;
+	case ClearPortFeature:
+		if (!wIndex || wIndex > ports)
+			goto error;
+
+		switch (wValue) {
+		case USB_PORT_FEAT_SUSPEND:
+			xenhcd_rhport_resume(info, wIndex);
+			break;
+		case USB_PORT_FEAT_POWER:
+			xenhcd_rhport_power_off(info, wIndex);
+			break;
+		case USB_PORT_FEAT_ENABLE:
+			xenhcd_rhport_disable(info, wIndex);
+			break;
+		case USB_PORT_FEAT_C_CONNECTION:
+			info->ports[wIndex-1].c_connection = 0;
+			/* falling through */
+		default:
+			info->ports[wIndex-1].status &= ~(1 << wValue);
+			break;
+		}
+		break;
+	case GetHubDescriptor:
+		xenhcd_hub_descriptor(info, (struct usb_hub_descriptor *)buf);
+		break;
+	case GetHubStatus:
+		/* always local power supply good and no over-current exists. */
+		*(__le32 *)buf = cpu_to_le32(0);
+		break;
+	case GetPortStatus:
+		if (!wIndex || wIndex > ports)
+			goto error;
+
+		wIndex--;
+
+		/* resume completion */
+		if (info->ports[wIndex].resuming &&
+		    time_after_eq(jiffies, info->ports[wIndex].timeout)) {
+			info->ports[wIndex].status |=
+				USB_PORT_STAT_C_SUSPEND << 16;
+			info->ports[wIndex].status &= ~USB_PORT_STAT_SUSPEND;
+		}
+
+		/* reset completion */
+		if ((info->ports[wIndex].status & USB_PORT_STAT_RESET) != 0 &&
+		    time_after_eq(jiffies, info->ports[wIndex].timeout)) {
+			info->ports[wIndex].status |=
+				USB_PORT_STAT_C_RESET << 16;
+			info->ports[wIndex].status &= ~USB_PORT_STAT_RESET;
+
+			if (info->devices[wIndex].status !=
+			    USB_STATE_NOTATTACHED) {
+				info->ports[wIndex].status |=
+					USB_PORT_STAT_ENABLE;
+				info->devices[wIndex].status =
+					USB_STATE_DEFAULT;
+			}
+
+			switch (info->devices[wIndex].speed) {
+			case XENUSB_SPEED_LOW:
+				info->ports[wIndex].status |=
+					USB_PORT_STAT_LOW_SPEED;
+				break;
+			case XENUSB_SPEED_HIGH:
+				info->ports[wIndex].status |=
+					USB_PORT_STAT_HIGH_SPEED;
+				break;
+			default:
+				break;
+			}
+		}
+
+		*(__le32 *)buf = cpu_to_le32(info->ports[wIndex].status);
+		break;
+	case SetPortFeature:
+		if (!wIndex || wIndex > ports)
+			goto error;
+
+		switch (wValue) {
+		case USB_PORT_FEAT_POWER:
+			xenhcd_rhport_power_on(info, wIndex);
+			break;
+		case USB_PORT_FEAT_RESET:
+			xenhcd_rhport_reset(info, wIndex);
+			break;
+		case USB_PORT_FEAT_SUSPEND:
+			xenhcd_rhport_suspend(info, wIndex);
+			break;
+		default:
+			if (info->ports[wIndex-1].status & USB_PORT_STAT_POWER)
+				info->ports[wIndex-1].status |= (1 << wValue);
+		}
+		break;
+
+	case SetHubFeature:
+		/* not supported */
+	default:
+error:
+		ret = -EPIPE;
+	}
+	spin_unlock_irqrestore(&info->lock, flags);
+
+	/* check status for each port */
+	for (i = 0; i < ports; i++) {
+		if (info->ports[i].status & PORT_C_MASK)
+			changed = 1;
+	}
+	if (changed)
+		usb_hcd_poll_rh_status(hcd);
+
+	return ret;
+}
+
+static void xenhcd_free_urb_priv(struct urb_priv *urbp)
+{
+	urbp->urb->hcpriv = NULL;
+	kmem_cache_free(xenhcd_urbp_cachep, urbp);
+}
+
+static inline unsigned xenhcd_get_id_from_freelist(struct xenhcd_info *info)
+{
+	unsigned free;
+
+	free = info->shadow_free;
+	BUG_ON(free >= XENUSB_URB_RING_SIZE);
+	info->shadow_free = info->shadow[free].req.id;
+	info->shadow[free].req.id = 0x0fff; /* debug */
+	return free;
+}
+
+static inline void xenhcd_add_id_to_freelist(struct xenhcd_info *info,
+					     unsigned id)
+{
+	info->shadow[id].req.id	= info->shadow_free;
+	info->shadow[id].urb = NULL;
+	info->shadow_free = id;
+}
+
+static inline int xenhcd_count_pages(void *addr, int length)
+{
+	unsigned long vaddr = (unsigned long)addr;
+
+	return PFN_UP(vaddr + length) - PFN_DOWN(vaddr);
+}
+
+static void xenhcd_gnttab_map(struct xenhcd_info *info, void *addr, int length,
+			      grant_ref_t *gref_head,
+			      struct xenusb_request_segment *seg,
+			      int nr_pages, int flags)
+{
+	grant_ref_t ref;
+	unsigned long buffer_mfn;
+	unsigned int offset;
+	unsigned int len = length;
+	unsigned int bytes;
+	int i;
+
+	for (i = 0; i < nr_pages; i++) {
+		BUG_ON(!len);
+
+		buffer_mfn = PFN_DOWN(arbitrary_virt_to_machine(addr).maddr);
+		offset = offset_in_page(addr);
+
+		bytes = PAGE_SIZE - offset;
+		if (bytes > len)
+			bytes = len;
+
+		ref = gnttab_claim_grant_reference(gref_head);
+		BUG_ON(ref == -ENOSPC);
+		gnttab_grant_foreign_access_ref(ref, info->xbdev->otherend_id,
+						buffer_mfn, flags);
+		seg[i].gref = ref;
+		seg[i].offset = (__u16)offset;
+		seg[i].length = (__u16)bytes;
+
+		addr += bytes;
+		len -= bytes;
+	}
+}
+
+static __u32 xenhcd_pipe_urb_to_xenusb(__u32 urb_pipe, __u8 port)
+{
+	static __u32 pipe;
+
+	pipe = usb_pipedevice(urb_pipe) << XENUSB_PIPE_DEV_SHIFT;
+	pipe |= usb_pipeendpoint(urb_pipe) << XENUSB_PIPE_EP_SHIFT;
+	if (usb_pipein(urb_pipe))
+		pipe |= XENUSB_PIPE_DIR;
+	switch (usb_pipetype(urb_pipe)) {
+	case PIPE_ISOCHRONOUS:
+		pipe |= XENUSB_PIPE_TYPE_ISOC << XENUSB_PIPE_TYPE_SHIFT;
+		break;
+	case PIPE_INTERRUPT:
+		pipe |= XENUSB_PIPE_TYPE_INT << XENUSB_PIPE_TYPE_SHIFT;
+		break;
+	case PIPE_CONTROL:
+		pipe |= XENUSB_PIPE_TYPE_CTRL << XENUSB_PIPE_TYPE_SHIFT;
+		break;
+	case PIPE_BULK:
+		pipe |= XENUSB_PIPE_TYPE_BULK << XENUSB_PIPE_TYPE_SHIFT;
+		break;
+	}
+	pipe = xenusb_setportnum_pipe(pipe, port);
+
+	return pipe;
+}
+
+static int xenhcd_map_urb_for_request(struct xenhcd_info *info, struct urb *urb,
+				      struct xenusb_urb_request *req)
+{
+	grant_ref_t gref_head;
+	int nr_buff_pages = 0;
+	int nr_isodesc_pages = 0;
+	int nr_grants = 0;
+
+	if (urb->transfer_buffer_length) {
+		nr_buff_pages = xenhcd_count_pages(urb->transfer_buffer,
+						urb->transfer_buffer_length);
+
+		if (usb_pipeisoc(urb->pipe))
+			nr_isodesc_pages = xenhcd_count_pages(
+				&urb->iso_frame_desc[0],
+				sizeof(struct usb_iso_packet_descriptor) *
+				urb->number_of_packets);
+
+		nr_grants = nr_buff_pages + nr_isodesc_pages;
+		if (nr_grants > XENUSB_MAX_SEGMENTS_PER_REQUEST) {
+			pr_err("xenhcd: error: %d grants\n", nr_grants);
+			return -E2BIG;
+		}
+
+		if (gnttab_alloc_grant_references(nr_grants, &gref_head)) {
+			pr_err("xenhcd: gnttab_alloc_grant_references() error\n");
+			return -ENOMEM;
+		}
+
+		xenhcd_gnttab_map(info, urb->transfer_buffer,
+				  urb->transfer_buffer_length, &gref_head,
+				  &req->seg[0], nr_buff_pages,
+				  usb_pipein(urb->pipe) ? 0 : GTF_readonly);
+	}
+
+	req->pipe = xenhcd_pipe_urb_to_xenusb(urb->pipe, urb->dev->portnum);
+	req->transfer_flags = urb->transfer_flags;
+	req->buffer_length = urb->transfer_buffer_length;
+	req->nr_buffer_segs = nr_buff_pages;
+
+	switch (usb_pipetype(urb->pipe)) {
+	case PIPE_ISOCHRONOUS:
+		req->u.isoc.interval = urb->interval;
+		req->u.isoc.start_frame = urb->start_frame;
+		req->u.isoc.number_of_packets = urb->number_of_packets;
+		req->u.isoc.nr_frame_desc_segs = nr_isodesc_pages;
+
+		/* urb->number_of_packets must be > 0 */
+		BUG_ON(urb->number_of_packets <= 0);
+
+		xenhcd_gnttab_map(info, &urb->iso_frame_desc[0],
+				  sizeof(struct usb_iso_packet_descriptor) *
+				  urb->number_of_packets,
+				  &gref_head, &req->seg[nr_buff_pages],
+				  nr_isodesc_pages, 0);
+		break;
+	case PIPE_INTERRUPT:
+		req->u.intr.interval = urb->interval;
+		break;
+	case PIPE_CONTROL:
+		if (urb->setup_packet)
+			memcpy(req->u.ctrl, urb->setup_packet, 8);
+		break;
+	case PIPE_BULK:
+		break;
+	default:
+		BUG();
+	}
+
+	if (nr_grants)
+		gnttab_free_grant_references(gref_head);
+
+	return 0;
+}
+
+static void xenhcd_gnttab_done(struct usb_shadow *shadow)
+{
+	int nr_segs = 0;
+	int i;
+
+	nr_segs = shadow->req.nr_buffer_segs;
+
+	if (xenusb_pipeisoc(shadow->req.pipe))
+		nr_segs += shadow->req.u.isoc.nr_frame_desc_segs;
+
+	for (i = 0; i < nr_segs; i++)
+		gnttab_end_foreign_access(shadow->req.seg[i].gref, 0, 0UL);
+
+	shadow->req.nr_buffer_segs = 0;
+	shadow->req.u.isoc.nr_frame_desc_segs = 0;
+}
+
+static void xenhcd_giveback_urb(struct xenhcd_info *info, struct urb *urb,
+				int status)
+{
+	struct urb_priv *urbp = (struct urb_priv *)urb->hcpriv;
+	int priv_status = urbp->status;
+
+	list_del_init(&urbp->list);
+	xenhcd_free_urb_priv(urbp);
+
+	if (urb->status == -EINPROGRESS)
+		urb->status = status;
+
+	spin_unlock(&info->lock);
+	usb_hcd_giveback_urb(xenhcd_info_to_hcd(info), urb,
+			     priv_status <= 0 ? priv_status : urb->status);
+	spin_lock(&info->lock);
+}
+
+static int xenhcd_do_request(struct xenhcd_info *info, struct urb_priv *urbp)
+{
+	struct xenusb_urb_request *req;
+	struct urb *urb = urbp->urb;
+	unsigned id;
+	int notify;
+	int ret;
+
+	req = RING_GET_REQUEST(&info->urb_ring, info->urb_ring.req_prod_pvt);
+	id = xenhcd_get_id_from_freelist(info);
+	req->id = id;
+
+	if (unlikely(urbp->unlinked)) {
+		req->u.unlink.unlink_id = urbp->req_id;
+		req->pipe = xenusb_setunlink_pipe(xenhcd_pipe_urb_to_xenusb(
+						 urb->pipe, urb->dev->portnum));
+		urbp->unlink_req_id = id;
+	} else {
+		ret = xenhcd_map_urb_for_request(info, urb, req);
+		if (ret) {
+			xenhcd_add_id_to_freelist(info, id);
+			return ret;
+		}
+		urbp->req_id = id;
+	}
+
+	info->urb_ring.req_prod_pvt++;
+	info->shadow[id].urb = urb;
+	info->shadow[id].req = *req;
+
+	RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->urb_ring, notify);
+	if (notify)
+		notify_remote_via_irq(info->irq);
+
+	return 0;
+}
+
+static void xenhcd_kick_pending_urbs(struct xenhcd_info *info)
+{
+	struct urb_priv *urbp;
+
+	while (!list_empty(&info->pending_submit_list)) {
+		if (RING_FULL(&info->urb_ring)) {
+			xenhcd_timer_action(info, TIMER_RING_WATCHDOG);
+			return;
+		}
+
+		urbp = list_entry(info->pending_submit_list.next,
+				  struct urb_priv, list);
+		if (!xenhcd_do_request(info, urbp))
+			list_move_tail(&urbp->list, &info->in_progress_list);
+		else
+			xenhcd_giveback_urb(info, urbp->urb, -ESHUTDOWN);
+	}
+	xenhcd_timer_action_done(info, TIMER_SCAN_PENDING_URBS);
+}
+
+/*
+ * caller must lock info->lock
+ */
+static void xenhcd_cancel_all_enqueued_urbs(struct xenhcd_info *info)
+{
+	struct urb_priv *urbp, *tmp;
+	int req_id;
+
+	list_for_each_entry_safe(urbp, tmp, &info->in_progress_list, list) {
+		req_id = urbp->req_id;
+		if (!urbp->unlinked) {
+			xenhcd_gnttab_done(&info->shadow[req_id]);
+			if (urbp->urb->status == -EINPROGRESS)
+				/* not dequeued */
+				xenhcd_giveback_urb(info, urbp->urb,
+						    -ESHUTDOWN);
+			else	/* dequeued */
+				xenhcd_giveback_urb(info, urbp->urb,
+						    urbp->urb->status);
+		}
+		info->shadow[req_id].urb = NULL;
+	}
+
+	list_for_each_entry_safe(urbp, tmp, &info->pending_submit_list, list)
+		xenhcd_giveback_urb(info, urbp->urb, -ESHUTDOWN);
+}
+
+/*
+ * caller must lock info->lock
+ */
+static void xenhcd_giveback_unlinked_urbs(struct xenhcd_info *info)
+{
+	struct urb_priv *urbp, *tmp;
+
+	list_for_each_entry_safe(urbp, tmp, &info->giveback_waiting_list, list)
+		xenhcd_giveback_urb(info, urbp->urb, urbp->urb->status);
+}
+
+static int xenhcd_submit_urb(struct xenhcd_info *info, struct urb_priv *urbp)
+{
+	int ret;
+
+	if (RING_FULL(&info->urb_ring)) {
+		list_add_tail(&urbp->list, &info->pending_submit_list);
+		xenhcd_timer_action(info, TIMER_RING_WATCHDOG);
+		return 0;
+	}
+
+	if (!list_empty(&info->pending_submit_list)) {
+		list_add_tail(&urbp->list, &info->pending_submit_list);
+		xenhcd_timer_action(info, TIMER_SCAN_PENDING_URBS);
+		return 0;
+	}
+
+	ret = xenhcd_do_request(info, urbp);
+	if (ret == 0)
+		list_add_tail(&urbp->list, &info->in_progress_list);
+
+	return ret;
+}
+
+static int xenhcd_unlink_urb(struct xenhcd_info *info, struct urb_priv *urbp)
+{
+	int ret;
+
+	/* already unlinked? */
+	if (urbp->unlinked)
+		return -EBUSY;
+
+	urbp->unlinked = 1;
+
+	/* the urb is still in pending_submit queue */
+	if (urbp->req_id == ~0) {
+		list_move_tail(&urbp->list, &info->giveback_waiting_list);
+		xenhcd_timer_action(info, TIMER_SCAN_PENDING_URBS);
+		return 0;
+	}
+
+	/* send unlink request to backend */
+	if (RING_FULL(&info->urb_ring)) {
+		list_move_tail(&urbp->list, &info->pending_unlink_list);
+		xenhcd_timer_action(info, TIMER_RING_WATCHDOG);
+		return 0;
+	}
+
+	if (!list_empty(&info->pending_unlink_list)) {
+		list_move_tail(&urbp->list, &info->pending_unlink_list);
+		xenhcd_timer_action(info, TIMER_SCAN_PENDING_URBS);
+		return 0;
+	}
+
+	ret = xenhcd_do_request(info, urbp);
+	if (ret == 0)
+		list_move_tail(&urbp->list, &info->in_progress_list);
+
+	return ret;
+}
+
+static int xenhcd_urb_request_done(struct xenhcd_info *info)
+{
+	struct xenusb_urb_response *res;
+	struct urb *urb;
+	RING_IDX i, rp;
+	__u16 id;
+	int more_to_do = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&info->lock, flags);
+
+	rp = info->urb_ring.sring->rsp_prod;
+	rmb(); /* ensure we see queued responses up to "rp" */
+
+	for (i = info->urb_ring.rsp_cons; i != rp; i++) {
+		res = RING_GET_RESPONSE(&info->urb_ring, i);
+		id = res->id;
+
+		if (likely(xenusb_pipesubmit(info->shadow[id].req.pipe))) {
+			xenhcd_gnttab_done(&info->shadow[id]);
+			urb = info->shadow[id].urb;
+			if (likely(urb)) {
+				urb->actual_length = res->actual_length;
+				urb->error_count = res->error_count;
+				urb->start_frame = res->start_frame;
+				xenhcd_giveback_urb(info, urb, res->status);
+			}
+		}
+
+		xenhcd_add_id_to_freelist(info, id);
+	}
+	info->urb_ring.rsp_cons = i;
+
+	if (i != info->urb_ring.req_prod_pvt)
+		RING_FINAL_CHECK_FOR_RESPONSES(&info->urb_ring, more_to_do);
+	else
+		info->urb_ring.sring->rsp_event = i + 1;
+
+	spin_unlock_irqrestore(&info->lock, flags);
+
+	return more_to_do;
+}
+
+static int xenhcd_conn_notify(struct xenhcd_info *info)
+{
+	struct xenusb_conn_response *res;
+	struct xenusb_conn_request *req;
+	RING_IDX rc, rp;
+	__u16 id;
+	__u8 portnum, speed;
+	int more_to_do = 0;
+	int notify;
+	int port_changed = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&info->lock, flags);
+
+	rc = info->conn_ring.rsp_cons;
+	rp = info->conn_ring.sring->rsp_prod;
+	rmb(); /* ensure we see queued responses up to "rp" */
+
+	while (rc != rp) {
+		res = RING_GET_RESPONSE(&info->conn_ring, rc);
+		id = res->id;
+		portnum = res->portnum;
+		speed = res->speed;
+		info->conn_ring.rsp_cons = ++rc;
+
+		xenhcd_rhport_connect(info, portnum, speed);
+		if (info->ports[portnum - 1].c_connection)
+			port_changed = 1;
+
+		barrier();
+
+		req = RING_GET_REQUEST(&info->conn_ring,
+				       info->conn_ring.req_prod_pvt);
+		req->id = id;
+		info->conn_ring.req_prod_pvt++;
+	}
+
+	if (rc != info->conn_ring.req_prod_pvt)
+		RING_FINAL_CHECK_FOR_RESPONSES(&info->conn_ring, more_to_do);
+	else
+		info->conn_ring.sring->rsp_event = rc + 1;
+
+	RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->conn_ring, notify);
+	if (notify)
+		notify_remote_via_irq(info->irq);
+
+	spin_unlock_irqrestore(&info->lock, flags);
+
+	if (port_changed)
+		usb_hcd_poll_rh_status(xenhcd_info_to_hcd(info));
+
+	return more_to_do;
+}
+
+static irqreturn_t xenhcd_int(int irq, void *dev_id)
+{
+	struct xenhcd_info *info = (struct xenhcd_info *)dev_id;
+
+	while (xenhcd_urb_request_done(info) | xenhcd_conn_notify(info))
+		/* Yield point for this unbounded loop. */
+		cond_resched();
+
+	return IRQ_HANDLED;
+}
+
+static void xenhcd_destroy_rings(struct xenhcd_info *info)
+{
+	if (info->irq)
+		unbind_from_irqhandler(info->irq, info);
+	info->irq = 0;
+
+	if (info->urb_ring_ref != GRANT_INVALID_REF) {
+		gnttab_end_foreign_access(info->urb_ring_ref, 0,
+					  (unsigned long)info->urb_ring.sring);
+		info->urb_ring_ref = GRANT_INVALID_REF;
+	}
+	info->urb_ring.sring = NULL;
+
+	if (info->conn_ring_ref != GRANT_INVALID_REF) {
+		gnttab_end_foreign_access(info->conn_ring_ref, 0,
+					  (unsigned long)info->conn_ring.sring);
+		info->conn_ring_ref = GRANT_INVALID_REF;
+	}
+	info->conn_ring.sring = NULL;
+}
+
+static int xenhcd_setup_rings(struct xenbus_device *dev,
+			      struct xenhcd_info *info)
+{
+	struct xenusb_urb_sring *urb_sring;
+	struct xenusb_conn_sring *conn_sring;
+	grant_ref_t gref;
+	int err;
+
+	info->urb_ring_ref = GRANT_INVALID_REF;
+	info->conn_ring_ref = GRANT_INVALID_REF;
+
+	urb_sring = (struct xenusb_urb_sring *)get_zeroed_page(
+							GFP_NOIO | __GFP_HIGH);
+	if (!urb_sring) {
+		xenbus_dev_fatal(dev, -ENOMEM, "allocating urb ring");
+		return -ENOMEM;
+	}
+	SHARED_RING_INIT(urb_sring);
+	FRONT_RING_INIT(&info->urb_ring, urb_sring, PAGE_SIZE);
+
+	err = xenbus_grant_ring(dev, urb_sring, 1, &gref);
+	if (err < 0) {
+		free_page((unsigned long)urb_sring);
+		info->urb_ring.sring = NULL;
+		goto fail;
+	}
+	info->urb_ring_ref = gref;
+
+	conn_sring = (struct xenusb_conn_sring *)get_zeroed_page(
+							GFP_NOIO | __GFP_HIGH);
+	if (!conn_sring) {
+		xenbus_dev_fatal(dev, -ENOMEM, "allocating conn ring");
+		err = -ENOMEM;
+		goto fail;
+	}
+	SHARED_RING_INIT(conn_sring);
+	FRONT_RING_INIT(&info->conn_ring, conn_sring, PAGE_SIZE);
+
+	err = xenbus_grant_ring(dev, conn_sring, 1, &gref);
+	if (err < 0) {
+		free_page((unsigned long)conn_sring);
+		info->conn_ring.sring = NULL;
+		goto fail;
+	}
+	info->conn_ring_ref = gref;
+
+	err = xenbus_alloc_evtchn(dev, &info->evtchn);
+	if (err) {
+		xenbus_dev_fatal(dev, err, "xenbus_alloc_evtchn");
+		goto fail;
+	}
+
+	err = bind_evtchn_to_irq(info->evtchn);
+	if (err <= 0) {
+		xenbus_dev_fatal(dev, err, "bind_evtchn_to_irq");
+		goto fail;
+	}
+
+	info->irq = err;
+
+	err = request_threaded_irq(info->irq, NULL, xenhcd_int,
+				   IRQF_ONESHOT, "xenhcd", info);
+	if (err) {
+		xenbus_dev_fatal(dev, err, "request_threaded_irq");
+		goto free_irq;
+	}
+
+	return 0;
+
+free_irq:
+	unbind_from_irqhandler(info->irq, info);
+fail:
+	xenhcd_destroy_rings(info);
+	return err;
+}
+
+static int xenhcd_talk_to_backend(struct xenbus_device *dev,
+				  struct xenhcd_info *info)
+{
+	const char *message;
+	struct xenbus_transaction xbt;
+	int err;
+
+	err = xenhcd_setup_rings(dev, info);
+	if (err)
+		return err;
+
+again:
+	err = xenbus_transaction_start(&xbt);
+	if (err) {
+		xenbus_dev_fatal(dev, err, "starting transaction");
+		goto destroy_ring;
+	}
+
+	err = xenbus_printf(xbt, dev->nodename, "urb-ring-ref", "%u",
+			    info->urb_ring_ref);
+	if (err) {
+		message = "writing urb-ring-ref";
+		goto abort_transaction;
+	}
+
+	err = xenbus_printf(xbt, dev->nodename, "conn-ring-ref", "%u",
+			    info->conn_ring_ref);
+	if (err) {
+		message = "writing conn-ring-ref";
+		goto abort_transaction;
+	}
+
+	err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
+			    info->evtchn);
+	if (err) {
+		message = "writing event-channel";
+		goto abort_transaction;
+	}
+
+	err = xenbus_transaction_end(xbt, 0);
+	if (err) {
+		if (err == -EAGAIN)
+			goto again;
+		xenbus_dev_fatal(dev, err, "completing transaction");
+		goto destroy_ring;
+	}
+
+	return 0;
+
+abort_transaction:
+	xenbus_transaction_end(xbt, 1);
+	xenbus_dev_fatal(dev, err, "%s", message);
+
+destroy_ring:
+	xenhcd_destroy_rings(info);
+
+	return err;
+}
+
+static int xenhcd_connect(struct xenbus_device *dev)
+{
+	struct xenhcd_info *info = dev_get_drvdata(&dev->dev);
+	struct xenusb_conn_request *req;
+	int idx, err;
+	int notify;
+	char name[TASK_COMM_LEN];
+	struct usb_hcd *hcd;
+
+	hcd = xenhcd_info_to_hcd(info);
+	snprintf(name, TASK_COMM_LEN, "xenhcd.%d", hcd->self.busnum);
+
+	err = xenhcd_talk_to_backend(dev, info);
+	if (err)
+		return err;
+
+	/* prepare ring for hotplug notification */
+	for (idx = 0; idx < XENUSB_CONN_RING_SIZE; idx++) {
+		req = RING_GET_REQUEST(&info->conn_ring, idx);
+		req->id = idx;
+	}
+	info->conn_ring.req_prod_pvt = idx;
+
+	RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->conn_ring, notify);
+	if (notify)
+		notify_remote_via_irq(info->irq);
+
+	return 0;
+}
+
+static void xenhcd_disconnect(struct xenbus_device *dev)
+{
+	struct xenhcd_info *info = dev_get_drvdata(&dev->dev);
+	struct usb_hcd *hcd = xenhcd_info_to_hcd(info);
+
+	usb_remove_hcd(hcd);
+	xenbus_frontend_closed(dev);
+}
+
+static void xenhcd_watchdog(unsigned long param)
+{
+	struct xenhcd_info *info = (struct xenhcd_info *)param;
+	unsigned long flags;
+
+	spin_lock_irqsave(&info->lock, flags);
+	if (likely(HC_IS_RUNNING(xenhcd_info_to_hcd(info)->state))) {
+		xenhcd_timer_action_done(info, TIMER_RING_WATCHDOG);
+		xenhcd_giveback_unlinked_urbs(info);
+		xenhcd_kick_pending_urbs(info);
+	}
+	spin_unlock_irqrestore(&info->lock, flags);
+}
+
+/*
+ * one-time HC init
+ */
+static int xenhcd_setup(struct usb_hcd *hcd)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+
+	spin_lock_init(&info->lock);
+	INIT_LIST_HEAD(&info->pending_submit_list);
+	INIT_LIST_HEAD(&info->pending_unlink_list);
+	INIT_LIST_HEAD(&info->in_progress_list);
+	INIT_LIST_HEAD(&info->giveback_waiting_list);
+	init_timer(&info->watchdog);
+	info->watchdog.function = xenhcd_watchdog;
+	info->watchdog.data = (unsigned long)info;
+
+	hcd->has_tt = (hcd->driver->flags & HCD_MASK) != HCD_USB11;
+
+	return 0;
+}
+
+/*
+ * start HC running
+ */
+static int xenhcd_run(struct usb_hcd *hcd)
+{
+	hcd->uses_new_polling = 1;
+	clear_bit(HCD_FLAG_POLL_RH, &hcd->flags);
+	hcd->state = HC_STATE_RUNNING;
+	return 0;
+}
+
+/*
+ * stop running HC
+ */
+static void xenhcd_stop(struct usb_hcd *hcd)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+
+	del_timer_sync(&info->watchdog);
+	spin_lock_irq(&info->lock);
+	/* cancel all urbs */
+	hcd->state = HC_STATE_HALT;
+	xenhcd_cancel_all_enqueued_urbs(info);
+	xenhcd_giveback_unlinked_urbs(info);
+	spin_unlock_irq(&info->lock);
+}
+
+/*
+ * called as .urb_enqueue()
+ * non-error returns are promise to giveback the urb later
+ */
+static int xenhcd_urb_enqueue(struct usb_hcd *hcd, struct urb *urb,
+			      gfp_t mem_flags)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+	struct urb_priv *urbp;
+	unsigned long flags;
+	int ret;
+
+
+	urbp = kmem_cache_zalloc(xenhcd_urbp_cachep, mem_flags);
+	if (!urbp)
+		return -ENOMEM;
+
+	spin_lock_irqsave(&info->lock, flags);
+
+	urbp->urb = urb;
+	urb->hcpriv = urbp;
+	urbp->req_id = ~0;
+	urbp->unlink_req_id = ~0;
+	INIT_LIST_HEAD(&urbp->list);
+	urbp->status = 1;
+	urb->unlinked = 0;
+
+	ret = xenhcd_submit_urb(info, urbp);
+
+	if (ret)
+		xenhcd_free_urb_priv(urbp);
+
+	spin_unlock_irqrestore(&info->lock, flags);
+
+	return ret;
+}
+
+/*
+ * called as .urb_dequeue()
+ */
+static int xenhcd_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status)
+{
+	struct xenhcd_info *info = xenhcd_hcd_to_info(hcd);
+	struct urb_priv *urbp;
+	unsigned long flags;
+	int ret = 0;
+
+	spin_lock_irqsave(&info->lock, flags);
+
+	urbp = urb->hcpriv;
+	if (urbp) {
+		urbp->status = status;
+		ret = xenhcd_unlink_urb(info, urbp);
+	}
+
+	spin_unlock_irqrestore(&info->lock, flags);
+
+	return ret;
+}
+
+/*
+ * called from usb_get_current_frame_number(),
+ * but, almost all drivers not use such function.
+ */
+static int xenhcd_get_frame(struct usb_hcd *hcd)
+{
+	/* it means error, but probably no problem :-) */
+	return 0;
+}
+
+static struct hc_driver xenhcd_usb20_hc_driver = {
+	.description = "xen-hcd",
+	.product_desc = "Xen USB2.0 Virtual Host Controller",
+	.hcd_priv_size = sizeof(struct xenhcd_info),
+	.flags = HCD_USB2,
+
+	/* basic HC lifecycle operations */
+	.reset = xenhcd_setup,
+	.start = xenhcd_run,
+	.stop = xenhcd_stop,
+
+	/* managing urb I/O */
+	.urb_enqueue = xenhcd_urb_enqueue,
+	.urb_dequeue = xenhcd_urb_dequeue,
+	.get_frame_number = xenhcd_get_frame,
+
+	/* root hub operations */
+	.hub_status_data = xenhcd_hub_status_data,
+	.hub_control = xenhcd_hub_control,
+#ifdef CONFIG_PM
+	.bus_suspend = xenhcd_bus_suspend,
+	.bus_resume = xenhcd_bus_resume,
+#endif
+};
+
+static struct hc_driver xenhcd_usb11_hc_driver = {
+	.description = "xen-hcd",
+	.product_desc = "Xen USB1.1 Virtual Host Controller",
+	.hcd_priv_size = sizeof(struct xenhcd_info),
+	.flags = HCD_USB11,
+
+	/* basic HC lifecycle operations */
+	.reset = xenhcd_setup,
+	.start = xenhcd_run,
+	.stop = xenhcd_stop,
+
+	/* managing urb I/O */
+	.urb_enqueue = xenhcd_urb_enqueue,
+	.urb_dequeue = xenhcd_urb_dequeue,
+	.get_frame_number = xenhcd_get_frame,
+
+	/* root hub operations */
+	.hub_status_data = xenhcd_hub_status_data,
+	.hub_control = xenhcd_hub_control,
+#ifdef CONFIG_PM
+	.bus_suspend = xenhcd_bus_suspend,
+	.bus_resume = xenhcd_bus_resume,
+#endif
+};
+
+static struct usb_hcd *xenhcd_create_hcd(struct xenbus_device *dev)
+{
+	int i;
+	int err = 0;
+	int num_ports;
+	int usb_ver;
+	struct usb_hcd *hcd = NULL;
+	struct xenhcd_info *info;
+
+	err = xenbus_scanf(XBT_NIL, dev->otherend, "num-ports", "%d",
+			   &num_ports);
+	if (err != 1) {
+		xenbus_dev_fatal(dev, err, "reading num-ports");
+		return ERR_PTR(-EINVAL);
+	}
+	if (num_ports < 1 || num_ports > XENUSB_MAX_PORTNR) {
+		xenbus_dev_fatal(dev, err, "invalid num-ports");
+		return ERR_PTR(-EINVAL);
+	}
+
+	err = xenbus_scanf(XBT_NIL, dev->otherend, "usb-ver", "%d", &usb_ver);
+	if (err != 1) {
+		xenbus_dev_fatal(dev, err, "reading usb-ver");
+		return ERR_PTR(-EINVAL);
+	}
+	switch (usb_ver) {
+	case XENUSB_VER_USB11:
+		hcd = usb_create_hcd(&xenhcd_usb11_hc_driver, &dev->dev,
+				     dev_name(&dev->dev));
+		break;
+	case XENUSB_VER_USB20:
+		hcd = usb_create_hcd(&xenhcd_usb20_hc_driver, &dev->dev,
+				     dev_name(&dev->dev));
+		break;
+	default:
+		xenbus_dev_fatal(dev, err, "invalid usb-ver");
+		return ERR_PTR(-EINVAL);
+	}
+	if (!hcd) {
+		xenbus_dev_fatal(dev, err,
+				 "fail to allocate USB host controller");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	info = xenhcd_hcd_to_info(hcd);
+	info->xbdev = dev;
+	info->rh_numports = num_ports;
+
+	for (i = 0; i < XENUSB_URB_RING_SIZE; i++) {
+		info->shadow[i].req.id = i + 1;
+		info->shadow[i].urb = NULL;
+	}
+	info->shadow[XENUSB_URB_RING_SIZE - 1].req.id = 0x0fff;
+
+	return hcd;
+}
+
+static void xenhcd_backend_changed(struct xenbus_device *dev,
+				   enum xenbus_state backend_state)
+{
+	switch (backend_state) {
+	case XenbusStateInitialising:
+	case XenbusStateReconfiguring:
+	case XenbusStateReconfigured:
+	case XenbusStateUnknown:
+		break;
+
+	case XenbusStateInitWait:
+	case XenbusStateInitialised:
+	case XenbusStateConnected:
+		if (dev->state != XenbusStateInitialising)
+			break;
+		if (!xenhcd_connect(dev))
+			xenbus_switch_state(dev, XenbusStateConnected);
+		break;
+
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's Closing state -- fallthrough */
+	case XenbusStateClosing:
+		xenhcd_disconnect(dev);
+		break;
+
+	default:
+		xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
+				 backend_state);
+		break;
+	}
+}
+
+static int xenhcd_remove(struct xenbus_device *dev)
+{
+	struct xenhcd_info *info = dev_get_drvdata(&dev->dev);
+	struct usb_hcd *hcd = xenhcd_info_to_hcd(info);
+
+	xenhcd_destroy_rings(info);
+	usb_put_hcd(hcd);
+
+	return 0;
+}
+
+static int xenhcd_probe(struct xenbus_device *dev,
+			const struct xenbus_device_id *id)
+{
+	int err;
+	struct usb_hcd *hcd;
+	struct xenhcd_info *info;
+
+	if (usb_disabled())
+		return -ENODEV;
+
+	hcd = xenhcd_create_hcd(dev);
+	if (IS_ERR(hcd)) {
+		err = PTR_ERR(hcd);
+		xenbus_dev_fatal(dev, err,
+				 "fail to create usb host controller");
+		return err;
+	}
+
+	info = xenhcd_hcd_to_info(hcd);
+	dev_set_drvdata(&dev->dev, info);
+
+	err = usb_add_hcd(hcd, 0, 0);
+	if (err) {
+		xenbus_dev_fatal(dev, err, "fail to add USB host controller");
+		usb_put_hcd(hcd);
+		dev_set_drvdata(&dev->dev, NULL);
+	}
+
+	return err;
+}
+
+static const struct xenbus_device_id xenhcd_ids[] = {
+	{ "vusb" },
+	{ "" },
+};
+
+static struct xenbus_driver xenhcd_driver = {
+	.ids			= xenhcd_ids,
+	.probe			= xenhcd_probe,
+	.otherend_changed	= xenhcd_backend_changed,
+	.remove			= xenhcd_remove,
+};
+
+static int __init xenhcd_init(void)
+{
+	if (!xen_domain())
+		return -ENODEV;
+
+	xenhcd_urbp_cachep = kmem_cache_create("xenhcd_urb_priv",
+					sizeof(struct urb_priv), 0, 0, NULL);
+	if (!xenhcd_urbp_cachep) {
+		pr_err("xenhcd failed to create kmem cache\n");
+		return -ENOMEM;
+	}
+
+	return xenbus_register_frontend(&xenhcd_driver);
+}
+module_init(xenhcd_init);
+
+static void __exit xenhcd_exit(void)
+{
+	kmem_cache_destroy(xenhcd_urbp_cachep);
+	xenbus_unregister_driver(&xenhcd_driver);
+}
+module_exit(xenhcd_exit);
+
+MODULE_ALIAS("xen:vusb");
+MODULE_AUTHOR("Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>");
+MODULE_DESCRIPTION("Xen USB Virtual Host Controller driver (xen-hcd)");
+MODULE_LICENSE("Dual BSD/GPL");
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in

[Patch V4 3/3] xen: add Xen pvUSB maintainer

[Juergen Gross]

Add myself as maintainer for the Xen pvUSB stuff.

Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
Acked-by: Konrad Rzeszutek Wilk <Konrad.wilk-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
---
 MAINTAINERS | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index d8afd29..8ef4e83 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -10965,6 +10965,14 @@ F:	drivers/scsi/xen-scsifront.c
 F:	drivers/xen/xen-scsiback.c
 F:	include/xen/interface/io/vscsiif.h
 
+XEN PVUSB DRIVER
+M:	Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
+L:	xen-devel-GuqFBffKawtpuQazS67q72D2FQJk+8+b@public.gmane.org (moderated for non-subscribers)
+L:	linux-usb-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
+S:	Supported
+F:	divers/usb/host/xen*
+F:	include/xen/interface/io/usbif.h
+
 XEN SWIOTLB SUBSYSTEM
 M:	Konrad Rzeszutek Wilk <konrad.wilk-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
 L:	xen-devel-GuqFBffKawtpuQazS67q72D2FQJk+8+b@public.gmane.org (moderated for non-subscribers)
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in

Re: [Patch V4 0/3] xen, usb: support pvUSB frontend driver

[Juergen Gross]

Ping?

On 06/23/2015 08:53 AM, Juergen Gross wrote:
> This series adds XEN guest pvUSB support. With pvUSB it is possible to
> use physical USB devices from a XEN domain.
>
> The support consists of a frontend in form of a virtual hcd driver in
> the unprivileged domU passing I/O-requests to the backend in a driver
> domain (usually Dom0). The backend is not part of this patch series,
> as it will be supported via qemu.
>
> The code is taken (and adapted) from the original pvUSB implementation
> done for Linux 2.6 in 2008 by Fujitsu.
>
> Normal operation of USB devices by adding and removing them dynamically
> to/from a domain has been tested using various USB devices (USB 1.1,
> 2.0 and 3.0). The pvUSB backend for these tests was a SUSE SLES Dom0
> with a kernel based backend driver.
>
> Changes in V4:
> - remove sysfs file from frontend, as it violated the "one value per file"
>    rule and didn't serve any real purpose.
>
> Changes in V3:
> - move frontend to drivers/usb/host and rename it to xen-hcd.
> - changed name prefixes in patch 1 to "xenusb" as requested by Greg
> - use __u<n> types rather than uint<n>_t as requested by Greg
>
> Changes in V2:
> - removed backend, as it can be implemented in user land
> - added some access macros and definitions to the pvUSB interface
>    description to make it independant from linux kernel USB internals
> - adapted frontend to newer kernel version and use new pvUSB
>    interface macros
> - set port status in one chunk as suggested by Oliver Neukum
>
>
> Juergen Gross (3):
>    usb: Add Xen pvUSB protocol description
>    usb: Introduce Xen pvUSB frontend (xen hcd)
>    xen: add Xen pvUSB maintainer
>
>   MAINTAINERS                      |    8 +
>   drivers/usb/host/Kconfig         |   11 +
>   drivers/usb/host/Makefile        |    1 +
>   drivers/usb/host/xen-hcd.c       | 1594 ++++++++++++++++++++++++++++++++++++++
>   include/xen/interface/io/usbif.h |  252 ++++++
>   5 files changed, 1866 insertions(+)
>   create mode 100644 drivers/usb/host/xen-hcd.c
>   create mode 100644 include/xen/interface/io/usbif.h
>

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Greg KH]

On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
> Add the definition of pvUSB protocol used between the pvUSB frontend in
> a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
> 
> This header was originally provided by Fujitsu for Xen based on Linux
> 2.6.18.
> 
> Changes are:
> - adapt to Linux style guide
> 
> Signed-off-by: Juergen Gross <jgross@suse.com>
> ---
>  include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++

Why is this a different interface than the existing ones we have today
(i.e. usbip?)  Where is it documented?  Do the Xen developers /
maintainers agree with this interface and code?  I need their sign-off
before I can accept such a thing.

thanks,

greg k-h

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Juergen Gross]

On 07/23/2015 01:46 AM, Greg KH wrote:
> On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
>> Add the definition of pvUSB protocol used between the pvUSB frontend in
>> a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
>>
>> This header was originally provided by Fujitsu for Xen based on Linux
>> 2.6.18.
>>
>> Changes are:
>> - adapt to Linux style guide
>>
>> Signed-off-by: Juergen Gross <jgross@suse.com>
>> ---
>>   include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
>
> Why is this a different interface than the existing ones we have today
> (i.e. usbip?)  Where is it documented?  Do the Xen developers /

The interface definition is living in the Xen git repository for several
years now:

git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h

It is used e.g. in SUSE's xen kernel since 2.6.18.

The differences between the Xen version and the one I've posted here are
only style and name space related.

> maintainers agree with this interface and code?  I need their sign-off
> before I can accept such a thing.

Sure.

David, Konrad, Boris, could one of you please comment on the patches?


Juergen

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Greg KH]

On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
> On 07/23/2015 01:46 AM, Greg KH wrote:
> >On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
> >>Add the definition of pvUSB protocol used between the pvUSB frontend in
> >>a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
> >>
> >>This header was originally provided by Fujitsu for Xen based on Linux
> >>2.6.18.
> >>
> >>Changes are:
> >>- adapt to Linux style guide
> >>
> >>Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
> >>---
> >>  include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
> >
> >Why is this a different interface than the existing ones we have today
> >(i.e. usbip?)  Where is it documented?  Do the Xen developers /
> 
> The interface definition is living in the Xen git repository for several
> years now:
> 
> git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h

That's header file, not a document describing the api here.

> It is used e.g. in SUSE's xen kernel since 2.6.18.

I am very aware of the amount of Xen crap in SuSE's kernel, don't use
that as an excuse for me to merge it to mainline :)

thanks,

greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Juergen Gross]

On 07/23/2015 06:36 AM, Greg KH wrote:
> On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
>> On 07/23/2015 01:46 AM, Greg KH wrote:
>>> On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
>>>> Add the definition of pvUSB protocol used between the pvUSB frontend in
>>>> a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
>>>>
>>>> This header was originally provided by Fujitsu for Xen based on Linux
>>>> 2.6.18.
>>>>
>>>> Changes are:
>>>> - adapt to Linux style guide
>>>>
>>>> Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
>>>> ---
>>>>   include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
>>>
>>> Why is this a different interface than the existing ones we have today
>>> (i.e. usbip?)  Where is it documented?  Do the Xen developers /
>>
>> The interface definition is living in the Xen git repository for several
>> years now:
>>
>> git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h
>
> That's header file, not a document describing the api here.

I suppose you want to tell me I should add something like:

Documentation/DocBook/usb/API-struct-urb.html

I can do this, of course.

>> It is used e.g. in SUSE's xen kernel since 2.6.18.
>
> I am very aware of the amount of Xen crap in SuSE's kernel, don't use
> that as an excuse for me to merge it to mainline :)

:-)

Wasn't meant as an excuse, just a hint why the interface can't be the
same as for usbip. We have to ensure compatibility with those kernels
and possibly other operating systems (BSD?, Windows?) which already
might be using pvUSB with a Dom0 based on the SUSE xen kernel.


Juergen

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Xen-devel] [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[David Vrabel]

On 23/07/15 07:46, Juergen Gross wrote:
> On 07/23/2015 06:36 AM, Greg KH wrote:
>> On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
>>> On 07/23/2015 01:46 AM, Greg KH wrote:
>>>> On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
>>>>> Add the definition of pvUSB protocol used between the pvUSB
>>>>> frontend in
>>>>> a Xen domU and the pvUSB backend in a Xen driver domain (usually
>>>>> Dom0).
>>>>>
>>>>> This header was originally provided by Fujitsu for Xen based on Linux
>>>>> 2.6.18.
>>>>>
>>>>> Changes are:
>>>>> - adapt to Linux style guide
>>>>>
>>>>> Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
>>>>> ---
>>>>>   include/xen/interface/io/usbif.h | 252
>>>>> +++++++++++++++++++++++++++++++++++++++
>>>>
>>>> Why is this a different interface than the existing ones we have today
>>>> (i.e. usbip?)  Where is it documented?  Do the Xen developers /
>>>
>>> The interface definition is living in the Xen git repository for several
>>> years now:
>>>
>>> git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h
>>
>> That's header file, not a document describing the api here.

This is the format that Xen uses to document the interface to frontend
drivers.  This is also the style used for netfront, blkfront etc.

The documentation in the header could be expanded if necessary but...

> I suppose you want to tell me I should add something like:
> 
> Documentation/DocBook/usb/API-struct-urb.html

... we certainly don't want a second location for the interface
description in the Linux source.

David
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Xen-devel] [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[David Vrabel]

On 23/06/15 07:53, Juergen Gross wrote:
> Add the definition of pvUSB protocol used between the pvUSB frontend in
> a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
> 
> This header was originally provided by Fujitsu for Xen based on Linux
> 2.6.18.
> 
> Changes are:
> - adapt to Linux style guide

Acked-by: David Vrabel <david.vrabel-Sxgqhf6Nn4DQT0dZR+AlfA@public.gmane.org>

David
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch V4 2/3] usb: Introduce Xen pvUSB frontend (xen hcd)

[David Vrabel]

On 23/06/15 07:53, Juergen Gross wrote:
> Introduces the Xen pvUSB frontend. With pvUSB it is possible for a Xen
> domU to communicate with a USB device assigned to that domU. The
> communication is all done via the pvUSB backend in a driver domain
> (usually Dom0) which is owner of the physical device.
> 
> The pvUSB frontend is a USB hcd for a virtual USB host connector.

For the Xen-related parts:

Reveiwed-by: David Vrabel <david.vrabel-Sxgqhf6Nn4DQT0dZR+AlfA@public.gmane.org>

David
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Greg KH]

On Thu, Jul 23, 2015 at 08:46:17AM +0200, Juergen Gross wrote:
> On 07/23/2015 06:36 AM, Greg KH wrote:
> >On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
> >>On 07/23/2015 01:46 AM, Greg KH wrote:
> >>>On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
> >>>>Add the definition of pvUSB protocol used between the pvUSB frontend in
> >>>>a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
> >>>>
> >>>>This header was originally provided by Fujitsu for Xen based on Linux
> >>>>2.6.18.
> >>>>
> >>>>Changes are:
> >>>>- adapt to Linux style guide
> >>>>
> >>>>Signed-off-by: Juergen Gross <jgross@suse.com>
> >>>>---
> >>>>  include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
> >>>
> >>>Why is this a different interface than the existing ones we have today
> >>>(i.e. usbip?)  Where is it documented?  Do the Xen developers /
> >>
> >>The interface definition is living in the Xen git repository for several
> >>years now:
> >>
> >>git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h
> >
> >That's header file, not a document describing the api here.
> 
> I suppose you want to tell me I should add something like:
> 
> Documentation/DocBook/usb/API-struct-urb.html

Somewhere that people can refer to that describes this public-facing API
that "must not ever be broken or changed".  If you want to put it in a
documentation file, or a .h file, I don't care.

> >>It is used e.g. in SUSE's xen kernel since 2.6.18.
> >
> >I am very aware of the amount of Xen crap in SuSE's kernel, don't use
> >that as an excuse for me to merge it to mainline :)
> 
> :-)
> 
> Wasn't meant as an excuse, just a hint why the interface can't be the
> same as for usbip. We have to ensure compatibility with those kernels

This shouldn't be a kernel/kernel compability issue, as the api talks
between Xen and the OS, not between different OSs, right?

> and possibly other operating systems (BSD?, Windows?) which already
> might be using pvUSB with a Dom0 based on the SUSE xen kernel.

Are there other operating system drivers today that use this API?  Is
this an API in the Xen core today that we have to support?

Some more background / descriptions would be nice to have.

thanks,

greg k-h

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Pasi Kärkkäinen]

On Thu, Jul 23, 2015 at 12:08:01PM -0700, Greg KH wrote:
> 
> Somewhere that people can refer to that describes this public-facing API
> that "must not ever be broken or changed".  If you want to put it in a
> documentation file, or a .h file, I don't care.
> 
> > >>It is used e.g. in SUSE's xen kernel since 2.6.18.
> > >
> > >I am very aware of the amount of Xen crap in SuSE's kernel, don't use
> > >that as an excuse for me to merge it to mainline :)
> > 
> > :-)
> > 
> > Wasn't meant as an excuse, just a hint why the interface can't be the
> > same as for usbip. We have to ensure compatibility with those kernels
> 
> This shouldn't be a kernel/kernel compability issue, as the api talks
> between Xen and the OS, not between different OSs, right?
> 
> > and possibly other operating systems (BSD?, Windows?) which already
> > might be using pvUSB with a Dom0 based on the SUSE xen kernel.
> 
> Are there other operating system drivers today that use this API?  Is
> this an API in the Xen core today that we have to support?
> 
> Some more background / descriptions would be nice to have.
>

For example Xen "GPLPV" drivers for Windows do have PVUSB frontend driver..


-- Pasi

 
> thanks,
> 
> greg k-h
> 

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Juergen Gross]

On 07/23/2015 09:08 PM, Greg KH wrote:
> On Thu, Jul 23, 2015 at 08:46:17AM +0200, Juergen Gross wrote:
>> On 07/23/2015 06:36 AM, Greg KH wrote:
>>> On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
>>>> On 07/23/2015 01:46 AM, Greg KH wrote:
>>>>> On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
>>>>>> Add the definition of pvUSB protocol used between the pvUSB frontend in
>>>>>> a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
>>>>>>
>>>>>> This header was originally provided by Fujitsu for Xen based on Linux
>>>>>> 2.6.18.
>>>>>>
>>>>>> Changes are:
>>>>>> - adapt to Linux style guide
>>>>>>
>>>>>> Signed-off-by: Juergen Gross <jgross@suse.com>
>>>>>> ---
>>>>>>   include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
>>>>>
>>>>> Why is this a different interface than the existing ones we have today
>>>>> (i.e. usbip?)  Where is it documented?  Do the Xen developers /
>>>>
>>>> The interface definition is living in the Xen git repository for several
>>>> years now:
>>>>
>>>> git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h
>>>
>>> That's header file, not a document describing the api here.
>>
>> I suppose you want to tell me I should add something like:
>>
>> Documentation/DocBook/usb/API-struct-urb.html
>
> Somewhere that people can refer to that describes this public-facing API
> that "must not ever be broken or changed".  If you want to put it in a
> documentation file, or a .h file, I don't care.
 >
>>>> It is used e.g. in SUSE's xen kernel since 2.6.18.
>>>
>>> I am very aware of the amount of Xen crap in SuSE's kernel, don't use
>>> that as an excuse for me to merge it to mainline :)
>>
>> :-)
>>
>> Wasn't meant as an excuse, just a hint why the interface can't be the
>> same as for usbip. We have to ensure compatibility with those kernels
>
> This shouldn't be a kernel/kernel compability issue, as the api talks
> between Xen and the OS, not between different OSs, right?

Depends on where the backend is living. It's the backend the frontend is
talking to.

There is a backend in SUSE's kernels up to SLE12. So compatibility is
to be maintained to those kernels.

Looks as if in future there will be one in qemu.

>> and possibly other operating systems (BSD?, Windows?) which already
>> might be using pvUSB with a Dom0 based on the SUSE xen kernel.
>
> Are there other operating system drivers today that use this API?  Is
> this an API in the Xen core today that we have to support?

Yes.

> Some more background / descriptions would be nice to have.

I guess a documentation file giving a brief explanation about the
interfaces of Xen wouldn't be a bad idea. This could avoid discussions
like this.

It shouldn't define each interface, but the classes of interfaces which
are existing (between kernel and hypervisor, frontends and backends)
and the stability requirements. Headers like the one we are discussing
here could then refer to this document.


Juergen

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Greg KH]

On Fri, Jul 24, 2015 at 05:51:04AM +0200, Juergen Gross wrote:
> On 07/23/2015 09:08 PM, Greg KH wrote:
> >On Thu, Jul 23, 2015 at 08:46:17AM +0200, Juergen Gross wrote:
> >>On 07/23/2015 06:36 AM, Greg KH wrote:
> >>>On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
> >>>>On 07/23/2015 01:46 AM, Greg KH wrote:
> >>>>>On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
> >>>>>>Add the definition of pvUSB protocol used between the pvUSB frontend in
> >>>>>>a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
> >>>>>>
> >>>>>>This header was originally provided by Fujitsu for Xen based on Linux
> >>>>>>2.6.18.
> >>>>>>
> >>>>>>Changes are:
> >>>>>>- adapt to Linux style guide
> >>>>>>
> >>>>>>Signed-off-by: Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org>
> >>>>>>---
> >>>>>>  include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
> >>>>>
> >>>>>Why is this a different interface than the existing ones we have today
> >>>>>(i.e. usbip?)  Where is it documented?  Do the Xen developers /
> >>>>
> >>>>The interface definition is living in the Xen git repository for several
> >>>>years now:
> >>>>
> >>>>git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h
> >>>
> >>>That's header file, not a document describing the api here.
> >>
> >>I suppose you want to tell me I should add something like:
> >>
> >>Documentation/DocBook/usb/API-struct-urb.html
> >
> >Somewhere that people can refer to that describes this public-facing API
> >that "must not ever be broken or changed".  If you want to put it in a
> >documentation file, or a .h file, I don't care.
> >
> >>>>It is used e.g. in SUSE's xen kernel since 2.6.18.
> >>>
> >>>I am very aware of the amount of Xen crap in SuSE's kernel, don't use
> >>>that as an excuse for me to merge it to mainline :)
> >>
> >>:-)
> >>
> >>Wasn't meant as an excuse, just a hint why the interface can't be the
> >>same as for usbip. We have to ensure compatibility with those kernels
> >
> >This shouldn't be a kernel/kernel compability issue, as the api talks
> >between Xen and the OS, not between different OSs, right?
> 
> Depends on where the backend is living. It's the backend the frontend is
> talking to.
> 
> There is a backend in SUSE's kernels up to SLE12. So compatibility is
> to be maintained to those kernels.

Note, just because a distro merged an out of tree patch, does not mean
that mainline has to accept the same api as-is :)

> Looks as if in future there will be one in qemu.

So there's only one other backend talking to this, in one distro?

> >>and possibly other operating systems (BSD?, Windows?) which already
> >>might be using pvUSB with a Dom0 based on the SUSE xen kernel.
> >
> >Are there other operating system drivers today that use this API?  Is
> >this an API in the Xen core today that we have to support?
> 
> Yes.

Yes to both?  Which other operating systems have such a driver?

> >Some more background / descriptions would be nice to have.
> 
> I guess a documentation file giving a brief explanation about the
> interfaces of Xen wouldn't be a bad idea. This could avoid discussions
> like this.

Yes it would.

> It shouldn't define each interface, but the classes of interfaces which
> are existing (between kernel and hypervisor, frontends and backends)
> and the stability requirements. Headers like the one we are discussing
> here could then refer to this document.

Why shouldn't the protocol be documented?

thanks,

greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch V4 1/3] usb: Add Xen pvUSB protocol description

[Konrad Rzeszutek Wilk]

On Fri, Jul 24, 2015 at 11:57:47AM -0700, Greg KH wrote:
> On Fri, Jul 24, 2015 at 05:51:04AM +0200, Juergen Gross wrote:
> > On 07/23/2015 09:08 PM, Greg KH wrote:
> > >On Thu, Jul 23, 2015 at 08:46:17AM +0200, Juergen Gross wrote:
> > >>On 07/23/2015 06:36 AM, Greg KH wrote:
> > >>>On Thu, Jul 23, 2015 at 06:04:39AM +0200, Juergen Gross wrote:
> > >>>>On 07/23/2015 01:46 AM, Greg KH wrote:
> > >>>>>On Tue, Jun 23, 2015 at 08:53:23AM +0200, Juergen Gross wrote:
> > >>>>>>Add the definition of pvUSB protocol used between the pvUSB frontend in
> > >>>>>>a Xen domU and the pvUSB backend in a Xen driver domain (usually Dom0).
> > >>>>>>
> > >>>>>>This header was originally provided by Fujitsu for Xen based on Linux
> > >>>>>>2.6.18.
> > >>>>>>
> > >>>>>>Changes are:
> > >>>>>>- adapt to Linux style guide
> > >>>>>>
> > >>>>>>Signed-off-by: Juergen Gross <jgross@suse.com>
> > >>>>>>---
> > >>>>>>  include/xen/interface/io/usbif.h | 252 +++++++++++++++++++++++++++++++++++++++
> > >>>>>
> > >>>>>Why is this a different interface than the existing ones we have today
> > >>>>>(i.e. usbip?)  Where is it documented?  Do the Xen developers /
> > >>>>
> > >>>>The interface definition is living in the Xen git repository for several
> > >>>>years now:
> > >>>>
> > >>>>git://xenbits.xen.org/xen.git -> xen/include/public/io/usbif.h
> > >>>
> > >>>That's header file, not a document describing the api here.
> > >>
> > >>I suppose you want to tell me I should add something like:
> > >>
> > >>Documentation/DocBook/usb/API-struct-urb.html
> > >
> > >Somewhere that people can refer to that describes this public-facing API
> > >that "must not ever be broken or changed".  If you want to put it in a
> > >documentation file, or a .h file, I don't care.
> > >
> > >>>>It is used e.g. in SUSE's xen kernel since 2.6.18.
> > >>>
> > >>>I am very aware of the amount of Xen crap in SuSE's kernel, don't use
> > >>>that as an excuse for me to merge it to mainline :)
> > >>
> > >>:-)
> > >>
> > >>Wasn't meant as an excuse, just a hint why the interface can't be the
> > >>same as for usbip. We have to ensure compatibility with those kernels
> > >
> > >This shouldn't be a kernel/kernel compability issue, as the api talks
> > >between Xen and the OS, not between different OSs, right?
> > 
> > Depends on where the backend is living. It's the backend the frontend is
> > talking to.
> > 
> > There is a backend in SUSE's kernels up to SLE12. So compatibility is
> > to be maintained to those kernels.
> 
> Note, just because a distro merged an out of tree patch, does not mean
> that mainline has to accept the same api as-is :)
> 
> > Looks as if in future there will be one in qemu.
> 
> So there's only one other backend talking to this, in one distro?

No. All distros which use QEMU will have it.

> 
> > >>and possibly other operating systems (BSD?, Windows?) which already
> > >>might be using pvUSB with a Dom0 based on the SUSE xen kernel.
> > >
> > >Are there other operating system drivers today that use this API?  Is
> > >this an API in the Xen core today that we have to support?
> > 
> > Yes.
> 
> Yes to both?  Which other operating systems have such a driver?

Windows XP, Windows 7, Windows 8, and so on using one of the Xen
PV drivers.

The old RHEL5 PV kernels, SLES kernels, and NetBSD [edit: Only
the header file, odd]


>
> > >Some more background / descriptions would be nice to have.
> > 
> > I guess a documentation file giving a brief explanation about the
> > interfaces of Xen wouldn't be a bad idea. This could avoid discussions
> > like this.
> 
> Yes it would.
> 
> > It shouldn't define each interface, but the classes of interfaces which
> > are existing (between kernel and hypervisor, frontends and backends)
> > and the stability requirements. Headers like the one we are discussing
> > here could then refer to this document.
> 
> Why shouldn't the protocol be documented?

I presume you mean the wire spec? Like what is in the memory and
exchanged between the guests? Classes sounds like an C++ thing :-)

There is also the XenBus which does the negotiation of parameters
(like the VirtioPCI spec) - that could be expanded to enumerate
the USB drivers ones ?

> 
> thanks,
> 
> greg k-h

Re: [Patch V4 0/3] xen, usb: support pvUSB frontend driver

[Oleksandr Tyshchenko]

[-- Attachment #1.1: Type: text/plain, Size: 3002 bytes --]

Hi Juergen, All.

I would like to apply PVUSB drivers for using in our platform while PVUSB
stuff doesn't reach upstream.

Unfortunately, I couldn't find more recent version of "kernel based"
backend driver in mailing list.
I have found it in v0 only.

[Xen-devel] [PATCH 0/4] xen, usb: support pvUSB drivers
<http://lists.xenproject.org/archives/html/xen-devel/2015-02/msg03245.html>

Can I use this frontend driver (v4) together with kernel based backend
driver (v0)?
Are they fully compatible with each other?

Thank you.

On Mon, Jul 20, 2015 at 8:13 AM, Juergen Gross <jgross@suse.com> wrote:

> Ping?
>
>
> On 06/23/2015 08:53 AM, Juergen Gross wrote:
>
>> This series adds XEN guest pvUSB support. With pvUSB it is possible to
>> use physical USB devices from a XEN domain.
>>
>> The support consists of a frontend in form of a virtual hcd driver in
>> the unprivileged domU passing I/O-requests to the backend in a driver
>> domain (usually Dom0). The backend is not part of this patch series,
>> as it will be supported via qemu.
>>
>> The code is taken (and adapted) from the original pvUSB implementation
>> done for Linux 2.6 in 2008 by Fujitsu.
>>
>> Normal operation of USB devices by adding and removing them dynamically
>> to/from a domain has been tested using various USB devices (USB 1.1,
>> 2.0 and 3.0). The pvUSB backend for these tests was a SUSE SLES Dom0
>> with a kernel based backend driver.
>>
>> Changes in V4:
>> - remove sysfs file from frontend, as it violated the "one value per file"
>>    rule and didn't serve any real purpose.
>>
>> Changes in V3:
>> - move frontend to drivers/usb/host and rename it to xen-hcd.
>> - changed name prefixes in patch 1 to "xenusb" as requested by Greg
>> - use __u<n> types rather than uint<n>_t as requested by Greg
>>
>> Changes in V2:
>> - removed backend, as it can be implemented in user land
>> - added some access macros and definitions to the pvUSB interface
>>    description to make it independant from linux kernel USB internals
>> - adapted frontend to newer kernel version and use new pvUSB
>>    interface macros
>> - set port status in one chunk as suggested by Oliver Neukum
>>
>>
>> Juergen Gross (3):
>>    usb: Add Xen pvUSB protocol description
>>    usb: Introduce Xen pvUSB frontend (xen hcd)
>>    xen: add Xen pvUSB maintainer
>>
>>   MAINTAINERS                      |    8 +
>>   drivers/usb/host/Kconfig         |   11 +
>>   drivers/usb/host/Makefile        |    1 +
>>   drivers/usb/host/xen-hcd.c       | 1594
>> ++++++++++++++++++++++++++++++++++++++
>>   include/xen/interface/io/usbif.h |  252 ++++++
>>   5 files changed, 1866 insertions(+)
>>   create mode 100644 drivers/usb/host/xen-hcd.c
>>   create mode 100644 include/xen/interface/io/usbif.h
>>
>>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel
>



-- 

Oleksandr Tyshchenko | Embedded Dev
GlobalLogic
www.globallogic.com
<http://www.globallogic.com/>

[-- Attachment #1.2: Type: text/html, Size: 5546 bytes --]

[-- Attachment #2: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [Xen-devel] [Patch V4 0/3] xen, usb: support pvUSB frontend driver

[Juergen Gross]

Hi Oleksandr,

On 17/03/16 20:43, Oleksandr Tyshchenko wrote:
> Hi Juergen, All.
> 
> I would like to apply PVUSB drivers for using in our platform while
> PVUSB stuff doesn't reach upstream.
> 
> Unfortunately, I couldn't find more recent version of "kernel based"
> backend driver in mailing list.
> I have found it in v0 only.
> 
> [Xen-devel] [PATCH 0/4] xen, usb: support pvUSB drivers
> <http://lists.xenproject.org/archives/html/xen-devel/2015-02/msg03245.html>
> 
> Can I use this frontend driver (v4) together with kernel based backend
> driver (v0)?
> Are they fully compatible with each other?

I didn't send any further version as it was regarded to be better to
provide such a backend via qemu (patches are being reviewed right now,
I sent them last week).

Regarding compatibility: should work. I've tested all possible
combinations of frontends/backends with both my upstream variants
you are referring to and the SUSE SLE variants which are kernel-xen
specific and SUSE-only.

You should be aware, however, that there might be some bugs in the
v0 backend driver. It was tested by me only.

> Thank you.

You're welcome. :-)


Juergen

> 
> On Mon, Jul 20, 2015 at 8:13 AM, Juergen Gross <jgross-IBi9RG/b67k@public.gmane.org
> <mailto:jgross-IBi9RG/b67k@public.gmane.org>> wrote:
> 
>     Ping?
> 
> 
>     On 06/23/2015 08:53 AM, Juergen Gross wrote:
> 
>         This series adds XEN guest pvUSB support. With pvUSB it is
>         possible to
>         use physical USB devices from a XEN domain.
> 
>         The support consists of a frontend in form of a virtual hcd
>         driver in
>         the unprivileged domU passing I/O-requests to the backend in a
>         driver
>         domain (usually Dom0). The backend is not part of this patch series,
>         as it will be supported via qemu.
> 
>         The code is taken (and adapted) from the original pvUSB
>         implementation
>         done for Linux 2.6 in 2008 by Fujitsu.
> 
>         Normal operation of USB devices by adding and removing them
>         dynamically
>         to/from a domain has been tested using various USB devices (USB 1.1,
>         2.0 and 3.0). The pvUSB backend for these tests was a SUSE SLES Dom0
>         with a kernel based backend driver.
> 
>         Changes in V4:
>         - remove sysfs file from frontend, as it violated the "one value
>         per file"
>            rule and didn't serve any real purpose.
> 
>         Changes in V3:
>         - move frontend to drivers/usb/host and rename it to xen-hcd.
>         - changed name prefixes in patch 1 to "xenusb" as requested by Greg
>         - use __u<n> types rather than uint<n>_t as requested by Greg
> 
>         Changes in V2:
>         - removed backend, as it can be implemented in user land
>         - added some access macros and definitions to the pvUSB interface
>            description to make it independant from linux kernel USB
>         internals
>         - adapted frontend to newer kernel version and use new pvUSB
>            interface macros
>         - set port status in one chunk as suggested by Oliver Neukum
> 
> 
>         Juergen Gross (3):
>            usb: Add Xen pvUSB protocol description
>            usb: Introduce Xen pvUSB frontend (xen hcd)
>            xen: add Xen pvUSB maintainer
> 
>           MAINTAINERS                      |    8 +
>           drivers/usb/host/Kconfig         |   11 +
>           drivers/usb/host/Makefile        |    1 +
>           drivers/usb/host/xen-hcd.c       | 1594
>         ++++++++++++++++++++++++++++++++++++++
>           include/xen/interface/io/usbif.h |  252 ++++++
>           5 files changed, 1866 insertions(+)
>           create mode 100644 drivers/usb/host/xen-hcd.c
>           create mode 100644 include/xen/interface/io/usbif.h
> 
> 
> 
>     _______________________________________________
>     Xen-devel mailing list
>     Xen-devel-GuqFBffKawuEi8DpZVb4nw@public.gmane.org <mailto:Xen-devel-GuqFBffKawuEi8DpZVb4nw@public.gmane.org>
>     http://lists.xen.org/xen-devel
> 
> 
> 
> 
> -- 
> 
> Oleksandr Tyshchenko | Embedded Dev
> GlobalLogic
> www.globallogic.com <http://www.globallogic.com/>
> <http://www.globallogic.com/>

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch V4 0/3] xen, usb: support pvUSB frontend driver

[Oleksandr Tyshchenko]

[-- Attachment #1.1: Type: text/plain, Size: 4777 bytes --]

On Fri, Mar 18, 2016 at 10:32 AM, Juergen Gross <jgross@suse.com> wrote:

> Hi Oleksandr,

Hi Juergen.
Thank you for the quick answer.


>
> On 17/03/16 20:43, Oleksandr Tyshchenko wrote:
> > Hi Juergen, All.
> >
> > I would like to apply PVUSB drivers for using in our platform while
> > PVUSB stuff doesn't reach upstream.
> >
> > Unfortunately, I couldn't find more recent version of "kernel based"
> > backend driver in mailing list.
> > I have found it in v0 only.
> >
> > [Xen-devel] [PATCH 0/4] xen, usb: support pvUSB drivers
> > <
> http://lists.xenproject.org/archives/html/xen-devel/2015-02/msg03245.html>
> >
> > Can I use this frontend driver (v4) together with kernel based backend
> > driver (v0)?
> > Are they fully compatible with each other?
>
> I didn't send any further version as it was regarded to be better to
> provide such a backend via qemu (patches are being reviewed right now,
> I sent them last week).
>
I got it. I have found patch series for "qemu based" backed driver but our
target is a kernel based backend driver like blkback or netback.


> Regarding compatibility: should work. I've tested all possible
> combinations of frontends/backends with both my upstream variants
> you are referring to and the SUSE SLE variants which are kernel-xen
> specific and SUSE-only.
>
> You should be aware, however, that there might be some bugs in the
> v0 backend driver. It was tested by me only.
>
OK, I am going to test it. I will let you know if something doesn't work.


>
> > Thank you.
>
> You're welcome. :-)
>
>
> Juergen
>
> >
> > On Mon, Jul 20, 2015 at 8:13 AM, Juergen Gross <jgross@suse.com
> > <mailto:jgross@suse.com>> wrote:
> >
> >     Ping?
> >
> >
> >     On 06/23/2015 08:53 AM, Juergen Gross wrote:
> >
> >         This series adds XEN guest pvUSB support. With pvUSB it is
> >         possible to
> >         use physical USB devices from a XEN domain.
> >
> >         The support consists of a frontend in form of a virtual hcd
> >         driver in
> >         the unprivileged domU passing I/O-requests to the backend in a
> >         driver
> >         domain (usually Dom0). The backend is not part of this patch
> series,
> >         as it will be supported via qemu.
> >
> >         The code is taken (and adapted) from the original pvUSB
> >         implementation
> >         done for Linux 2.6 in 2008 by Fujitsu.
> >
> >         Normal operation of USB devices by adding and removing them
> >         dynamically
> >         to/from a domain has been tested using various USB devices (USB
> 1.1,
> >         2.0 and 3.0). The pvUSB backend for these tests was a SUSE SLES
> Dom0
> >         with a kernel based backend driver.
> >
> >         Changes in V4:
> >         - remove sysfs file from frontend, as it violated the "one value
> >         per file"
> >            rule and didn't serve any real purpose.
> >
> >         Changes in V3:
> >         - move frontend to drivers/usb/host and rename it to xen-hcd.
> >         - changed name prefixes in patch 1 to "xenusb" as requested by
> Greg
> >         - use __u<n> types rather than uint<n>_t as requested by Greg
> >
> >         Changes in V2:
> >         - removed backend, as it can be implemented in user land
> >         - added some access macros and definitions to the pvUSB interface
> >            description to make it independant from linux kernel USB
> >         internals
> >         - adapted frontend to newer kernel version and use new pvUSB
> >            interface macros
> >         - set port status in one chunk as suggested by Oliver Neukum
> >
> >
> >         Juergen Gross (3):
> >            usb: Add Xen pvUSB protocol description
> >            usb: Introduce Xen pvUSB frontend (xen hcd)
> >            xen: add Xen pvUSB maintainer
> >
> >           MAINTAINERS                      |    8 +
> >           drivers/usb/host/Kconfig         |   11 +
> >           drivers/usb/host/Makefile        |    1 +
> >           drivers/usb/host/xen-hcd.c       | 1594
> >         ++++++++++++++++++++++++++++++++++++++
> >           include/xen/interface/io/usbif.h |  252 ++++++
> >           5 files changed, 1866 insertions(+)
> >           create mode 100644 drivers/usb/host/xen-hcd.c
> >           create mode 100644 include/xen/interface/io/usbif.h
> >
> >
> >
> >     _______________________________________________
> >     Xen-devel mailing list
> >     Xen-devel@lists.xen.org <mailto:Xen-devel@lists.xen.org>
> >     http://lists.xen.org/xen-devel
> >
> >
> >
> >
> > --
> >
> > Oleksandr Tyshchenko | Embedded Dev
> > GlobalLogic
> > www.globallogic.com <http://www.globallogic.com/>
> > <http://www.globallogic.com/>
>
>


-- 

Oleksandr Tyshchenko | Embedded Dev
GlobalLogic
www.globallogic.com
<http://www.globallogic.com/>

[-- Attachment #1.2: Type: text/html, Size: 9118 bytes --]

[-- Attachment #2: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel