Re: [RFC] 9p: add KVM/QEMU pci transport

Re: [RFC] 9p: add KVM/QEMU pci transport

[Anthony Liguori]

I think that it would be nicer to implement the p9 transport on top of
virtio instead of directly on top of PCI.  I think your PCI transport
would make a pretty nice start of a PCI virtio transport though.

Regards,

Anthony Liguori

On Tue, 2007-08-28 at 13:52 -0500, Eric Van Hensbergen wrote:
> From: Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>
> 
> This adds a shared memory transport for a synthetic 9p device for
> paravirtualized file system support under KVM/QEMU.
> 
> Signed-off-by: Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>
> Signed-off-by: Eric Van Hensbergen <ericvh-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> ---
>  Documentation/filesystems/9p.txt |    2 +
>  net/9p/Kconfig                   |   10 ++-
>  net/9p/Makefile                  |    4 +
>  net/9p/trans_pci.c               |  295 ++++++++++++++++++++++++++++++++++++++
>  4 files changed, 310 insertions(+), 1 deletions(-)
>  create mode 100644 net/9p/trans_pci.c
> 
> diff --git a/Documentation/filesystems/9p.txt b/Documentation/filesystems/9p.txt
> index 1a5f50d..e1879bd 100644
> --- a/Documentation/filesystems/9p.txt
> +++ b/Documentation/filesystems/9p.txt
> @@ -46,6 +46,8 @@ OPTIONS
>   			tcp  - specifying a normal TCP/IP connection
>   			fd   - used passed file descriptors for connection
>                                  (see rfdno and wfdno)
> +			pci  - use a PCI pseudo device for 9p communication
> +				over shared memory between a guest and host
>  
>    uname=name	user name to attempt mount as on the remote server.  The
>    		server may override or ignore this value.  Certain user
> diff --git a/net/9p/Kconfig b/net/9p/Kconfig
> index 09566ae..8517560 100644
> --- a/net/9p/Kconfig
> +++ b/net/9p/Kconfig
> @@ -16,13 +16,21 @@ menuconfig NET_9P
>  config NET_9P_FD
>  	depends on NET_9P
>  	default y if NET_9P
> -	tristate "9P File Descriptor Transports (Experimental)"
> +	tristate "9p File Descriptor Transports (Experimental)"
>  	help
>  	  This builds support for file descriptor transports for 9p
>  	  which includes support for TCP/IP, named pipes, or passed
>  	  file descriptors.  TCP/IP is the default transport for 9p,
>  	  so if you are going to use 9p, you'll likely want this.  
>  
> +config NET_9P_PCI
> +	depends on NET_9P
> +	tristate "9p PCI Shared Memory Transport (Experimental)"
> +	help
> +	  This builds support for a PCI psuedo-device currently available
> +	  under KVM/QEMU which allows for 9p transactions over shared
> +	  memory between the guest and the host.
> +
>  config NET_9P_DEBUG
>  	bool "Debug information"
>  	depends on NET_9P
> diff --git a/net/9p/Makefile b/net/9p/Makefile
> index 7b2a67a..26ce89d 100644
> --- a/net/9p/Makefile
> +++ b/net/9p/Makefile
> @@ -1,5 +1,6 @@
>  obj-$(CONFIG_NET_9P) := 9pnet.o
>  obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o
> +obj-$(CONFIG_NET_9P_PCI) += 9pnet_pci.o
>  
>  9pnet-objs := \
>  	mod.o \
> @@ -14,3 +15,6 @@ obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o
>  
>  9pnet_fd-objs := \
>  	trans_fd.o \
> +
> +9pnet_pci-objs := \
> +	trans_pci.o \
> diff --git a/net/9p/trans_pci.c b/net/9p/trans_pci.c
> new file mode 100644
> index 0000000..36ddc5f
> --- /dev/null
> +++ b/net/9p/trans_pci.c
> @@ -0,0 +1,295 @@
> +/*
> + * net/9p/trans_pci.c
> + *
> + * 9P over PCI transport layer. For use with KVM/QEMU.
> + *
> + *  Copyright (C) 2007 by Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@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 version 2
> + *  as published by the Free Software Foundation.
> + *
> + *  This program is distributed in the hope that it will be useful,
> + *  but WITHOUT ANY WARRANTY; without even the implied warranty of
> + *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + *  GNU General Public License for more details.
> + *
> + *  You should have received a copy of the GNU General Public License
> + *  along with this program; if not, write to:
> + *  Free Software Foundation
> + *  51 Franklin Street, Fifth Floor
> + *  Boston, MA  02111-1301  USA
> + *
> + */
> +
> +#include <linux/module.h>
> +#include <linux/kernel.h>
> +#include <linux/compiler.h>
> +#include <linux/pci.h>
> +#include <linux/init.h>
> +#include <linux/ioport.h>
> +#include <linux/completion.h>
> +#include <linux/interrupt.h>
> +#include <linux/io.h>
> +#include <linux/uaccess.h>
> +#include <linux/irq.h>
> +#include <linux/poll.h>
> +#include <net/9p/9p.h>
> +#include <net/9p/transport.h>
> +
> +#define P9PCI_DRIVER_NAME "9P PCI Device"
> +#define P9PCI_DRIVER_VERSION "1"
> +
> +#define PCI_VENDOR_ID_9P	0x5002
> +#define PCI_DEVICE_ID_9P	0x000D
> +
> +#define MAX_PCI_BUF		(4*1024) /* TODO: Get a number from lucho */
> +
> +struct p9pci_trans {
> +	struct pci_dev		*pdev;
> +	void __iomem		*ioaddr;
> +	void __iomem		*tx;
> +	void __iomem		*rx;
> +	int			irq;
> +	int			pos;
> +	int			len;
> +	wait_queue_head_t	wait;
> +};
> +static struct p9pci_trans *p9pci_trans; /* single channel for now */
> +
> +static struct pci_device_id p9pci_tbl[] = {
> +	{PCI_VENDOR_ID_9P, PCI_DEVICE_ID_9P, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 },
> +	{0,}
> +};
> +
> +static irqreturn_t p9pci_interrupt(int irq, void *dev)
> +{
> +	p9pci_trans = dev;
> +	p9pci_trans->len = le32_to_cpu(readl(p9pci_trans->rx));
> +	P9_DPRINTK(P9_DEBUG_TRANS, "%p len %d\n", p9pci_trans->pdev,
> +							p9pci_trans->len);
> +	iowrite32(0, p9pci_trans->ioaddr + 4);
> +	wake_up_interruptible(&p9pci_trans->wait);
> +	return IRQ_HANDLED;
> +}
> +
> +static int p9pci_read(struct p9_trans *trans, void *v, int len)
> +{
> +	struct p9pci_trans *ts;
> +
> +	if (!trans || trans->status == Disconnected || !trans->priv)
> +		return -EREMOTEIO;
> +
> +	ts = trans->priv;
> +
> +	P9_DPRINTK(P9_DEBUG_TRANS, "trans %p rx %p tx %p buf %p len %d\n",
> +						trans, ts->rx, ts->tx, v, len);
> +	if (len > ts->len)
> +		len = ts->len;
> +
> +	if (len) {
> +		memcpy_fromio(v, ts->rx, len);
> +		ts->len = 0;
> +		/* let the host knows the message is consumed */
> +		writel(0, ts->rx);
> +		iowrite32(0, p9pci_trans->ioaddr + 4);
> +		P9_DPRINTK(P9_DEBUG_TRANS, "zero rxlen %d txlen %d\n",
> +						readl(ts->rx), readl(ts->tx));
> +	}
> +
> +	return len;
> +}
> +
> +static int p9pci_write(struct p9_trans *trans, void *v, int len)
> +{
> +	struct p9pci_trans *ts;
> +
> +	if (!trans || trans->status == Disconnected || !trans->priv)
> +		return -EREMOTEIO;
> +
> +	ts = trans->priv;
> +	P9_DPRINTK(P9_DEBUG_TRANS, "trans %p rx %p tx %p buf %p len %d\n",
> +						trans, ts->rx, ts->tx, v, len);
> +	P9_DPRINTK(P9_DEBUG_TRANS, "rxlen %d\n", readl(ts->rx));
> +	if (readb(ts->tx) != 0)
> +		return 0;
> +
> +	P9_DPRINTK(P9_DEBUG_TRANS, "tx addr %p io addr %p\n", ts->tx,
> +								ts->ioaddr);
> +	memcpy_toio(ts->tx, v, len);
> +	iowrite32(len, ts->ioaddr);
> +	return len;
> +}
> +
> +static unsigned int
> +p9pci_poll(struct p9_trans *trans, struct poll_table_struct *pt)
> +{
> +	int ret;
> +	struct p9pci_trans *ts;
> +
> +	P9_DPRINTK(P9_DEBUG_TRANS, "trans %p\n", trans);
> +	if (!trans || trans->status != Connected || !trans->priv)
> +		return -EREMOTEIO;
> +
> +	ts = trans->priv;
> +	poll_wait(NULL, &ts->wait, pt);
> +	ret = 0;
> +	if (!readl(ts->tx))
> +		ret |= POLLOUT;
> +	if (readl(ts->rx))
> +		ret |= POLLIN;
> +
> +	P9_DPRINTK(P9_DEBUG_TRANS, "txlen %d rxlen %d\n", readl(ts->tx),
> +								readl(ts->rx));
> +	return ret;
> +}
> +
> +/**
> + * p9_sock_close - shutdown socket
> + * @trans: private socket structure
> + *
> + */
> +static void p9pci_close(struct p9_trans *trans)
> +{
> +	P9_DPRINTK(P9_DEBUG_TRANS, "trans %p\n", trans);
> +}
> +
> +static struct p9_trans *p9pci_trans_create(const char *name, char *arg)
> +{
> +	struct p9_trans *trans;
> +
> +	P9_DPRINTK(P9_DEBUG_TRANS, "\n");
> +	trans = kmalloc(sizeof(struct p9_trans), GFP_KERNEL);
> +	if (!trans)
> +		return ERR_PTR(-ENOMEM);
> +
> +	trans->status = Connected;
> +	trans->write = p9pci_write;
> +	trans->read = p9pci_read;
> +	trans->close = p9pci_close;
> +	trans->poll = p9pci_poll;
> +	trans->priv = p9pci_trans;
> +	writel(0, p9pci_trans->tx);
> +	writel(0, p9pci_trans->rx);
> +
> +	return trans;
> +}
> +
> +static int __devinit p9pci_probe(struct pci_dev *pdev,
> +		const struct pci_device_id *ent)
> +{
> +	int err;
> +	u8 pci_rev;
> +
> +	if (p9pci_trans)
> +		return -1;
> +
> +	pci_read_config_byte(pdev, PCI_REVISION_ID, &pci_rev);
> +
> +	if (pdev->vendor == PCI_VENDOR_ID_9P &&
> +	    pdev->device == PCI_DEVICE_ID_9P)
> +		printk(KERN_INFO "pci dev %s (id %04x:%04x rev %02x) is a 9P\n",
> +		       pci_name(pdev), pdev->vendor, pdev->device, pci_rev);
> +
> +	P9_DPRINTK(P9_DEBUG_TRANS, "%p\n", pdev);
> +	p9pci_trans = kzalloc(sizeof(*p9pci_trans), GFP_KERNEL);
> +	p9pci_trans->irq = -1;
> +	init_waitqueue_head(&p9pci_trans->wait);
> +	err = pci_enable_device(pdev);
> +	if (err)
> +		goto error;
> +
> +	p9pci_trans->pdev = pdev;
> +	err = pci_request_regions(pdev, "9p");
> +	if (err)
> +		goto error;
> +
> +	p9pci_trans->ioaddr = pci_iomap(pdev, 0, 8);
> +	if (!p9pci_trans->ioaddr) {
> +		P9_DPRINTK(P9_DEBUG_ERROR, "Cannot remap MMIO, aborting\n");
> +		err = -EIO;
> +		goto error;
> +	}
> +
> +	p9pci_trans->tx = pci_iomap(pdev, 1, 0x20000);
> +	p9pci_trans->rx = pci_iomap(pdev, 2, 0x20000);
> +	pci_set_drvdata(pdev, p9pci_trans);
> +	err = request_irq(pdev->irq, &p9pci_interrupt, 0, "p9pci", p9pci_trans);
> +	if (err)
> +		goto error;
> +
> +	p9pci_trans->irq = pdev->irq;
> +	return 0;
> +
> +error:
> +	P9_DPRINTK(P9_DEBUG_ERROR, "error %d\n", err);
> +	if (p9pci_trans->irq >= 0) {
> +		synchronize_irq(p9pci_trans->irq);
> +		free_irq(p9pci_trans->irq, p9pci_trans);
> +	}
> +
> +	if (p9pci_trans->pdev) {
> +		pci_release_regions(pdev);
> +		pci_iounmap(pdev, p9pci_trans->ioaddr);
> +		pci_set_drvdata(pdev, NULL);
> +		pci_disable_device(pdev);
> +	}
> +
> +	kfree(p9pci_trans);
> +	return -1;
> +}
> +
> +static void __devexit p9pci_remove(struct pci_dev *pdev)
> +{
> +	P9_DPRINTK(P9_DEBUG_TRANS, "%p\n", pdev);
> +	p9pci_trans = pci_get_drvdata(pdev);
> +	if (!p9pci_trans)
> +		return;
> +
> +	if (p9pci_trans->irq) {
> +		synchronize_irq(p9pci_trans->irq);
> +		free_irq(p9pci_trans->irq, p9pci_trans);
> +	}
> +
> +	pci_release_regions(pdev);
> +	pci_iounmap(pdev, p9pci_trans->ioaddr);
> +	pci_set_drvdata(pdev, NULL);
> +	kfree(p9pci_trans);
> +	pci_disable_device(pdev);
> +}
> +
> +static struct pci_driver p9pci_driver = {
> +	.name		= P9PCI_DRIVER_NAME,
> +	.id_table	= p9pci_tbl,
> +	.probe		= p9pci_probe,
> +	.remove		= __devexit_p(p9pci_remove),
> +};
> +
> +static struct p9_trans_module p9_pci_trans = {
> +	.name = "pci",
> +	.maxsize = MAX_PCI_BUF,
> +	.def = 0,
> +	.create = p9pci_trans_create,
> +};
> +
> +static int __init p9pci_init_module(void)
> +{
> +	v9fs_register_trans(&p9_pci_trans);
> +	return pci_register_driver(&p9pci_driver);
> +}
> +
> +static void __exit p9pci_cleanup_module(void)
> +{
> +	pci_unregister_driver(&p9pci_driver);
> +	printk(KERN_ERR "Removal of 9p transports not implemented\n");
> +	BUG();
> +}
> +
> +module_init(p9pci_init_module);
> +module_exit(p9pci_cleanup_module);
> +
> +MODULE_DEVICE_TABLE(pci, p9pci_tbl);
> +MODULE_AUTHOR("Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>");
> +MODULE_DESCRIPTION(P9PCI_DRIVER_NAME);
> +MODULE_LICENSE("GPL");
> +MODULE_VERSION(P9PCI_DRIVER_VERSION);


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/

Re: [V9fs-developer] [RFC] 9p: add KVM/QEMU pci transport

[Latchesar Ionkov]

That's also in our plans. There was no virtio support in KVM when I
started working in the transport.

Thanks,
    Lucho

On 8/29/07, Anthony Liguori <anthony-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org> wrote:
> I think that it would be nicer to implement the p9 transport on top of
> virtio instead of directly on top of PCI.  I think your PCI transport
> would make a pretty nice start of a PCI virtio transport though.
>
> Regards,
>
> Anthony Liguori
>
> On Tue, 2007-08-28 at 13:52 -0500, Eric Van Hensbergen wrote:
> > From: Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>
> >
> > This adds a shared memory transport for a synthetic 9p device for
> > paravirtualized file system support under KVM/QEMU.
> >
> > Signed-off-by: Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>
> > Signed-off-by: Eric Van Hensbergen <ericvh-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> > ---
> >  Documentation/filesystems/9p.txt |    2 +
> >  net/9p/Kconfig                   |   10 ++-
> >  net/9p/Makefile                  |    4 +
> >  net/9p/trans_pci.c               |  295 ++++++++++++++++++++++++++++++++++++++
> >  4 files changed, 310 insertions(+), 1 deletions(-)
> >  create mode 100644 net/9p/trans_pci.c
> >
> > diff --git a/Documentation/filesystems/9p.txt b/Documentation/filesystems/9p.txt
> > index 1a5f50d..e1879bd 100644
> > --- a/Documentation/filesystems/9p.txt
> > +++ b/Documentation/filesystems/9p.txt
> > @@ -46,6 +46,8 @@ OPTIONS
> >                       tcp  - specifying a normal TCP/IP connection
> >                       fd   - used passed file descriptors for connection
> >                                  (see rfdno and wfdno)
> > +                     pci  - use a PCI pseudo device for 9p communication
> > +                             over shared memory between a guest and host
> >
> >    uname=name user name to attempt mount as on the remote server.  The
> >               server may override or ignore this value.  Certain user
> > diff --git a/net/9p/Kconfig b/net/9p/Kconfig
> > index 09566ae..8517560 100644
> > --- a/net/9p/Kconfig
> > +++ b/net/9p/Kconfig
> > @@ -16,13 +16,21 @@ menuconfig NET_9P
> >  config NET_9P_FD
> >       depends on NET_9P
> >       default y if NET_9P
> > -     tristate "9P File Descriptor Transports (Experimental)"
> > +     tristate "9p File Descriptor Transports (Experimental)"
> >       help
> >         This builds support for file descriptor transports for 9p
> >         which includes support for TCP/IP, named pipes, or passed
> >         file descriptors.  TCP/IP is the default transport for 9p,
> >         so if you are going to use 9p, you'll likely want this.
> >
> > +config NET_9P_PCI
> > +     depends on NET_9P
> > +     tristate "9p PCI Shared Memory Transport (Experimental)"
> > +     help
> > +       This builds support for a PCI psuedo-device currently available
> > +       under KVM/QEMU which allows for 9p transactions over shared
> > +       memory between the guest and the host.
> > +
> >  config NET_9P_DEBUG
> >       bool "Debug information"
> >       depends on NET_9P
> > diff --git a/net/9p/Makefile b/net/9p/Makefile
> > index 7b2a67a..26ce89d 100644
> > --- a/net/9p/Makefile
> > +++ b/net/9p/Makefile
> > @@ -1,5 +1,6 @@
> >  obj-$(CONFIG_NET_9P) := 9pnet.o
> >  obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o
> > +obj-$(CONFIG_NET_9P_PCI) += 9pnet_pci.o
> >
> >  9pnet-objs := \
> >       mod.o \
> > @@ -14,3 +15,6 @@ obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o
> >
> >  9pnet_fd-objs := \
> >       trans_fd.o \
> > +
> > +9pnet_pci-objs := \
> > +     trans_pci.o \
> > diff --git a/net/9p/trans_pci.c b/net/9p/trans_pci.c
> > new file mode 100644
> > index 0000000..36ddc5f
> > --- /dev/null
> > +++ b/net/9p/trans_pci.c
> > @@ -0,0 +1,295 @@
> > +/*
> > + * net/9p/trans_pci.c
> > + *
> > + * 9P over PCI transport layer. For use with KVM/QEMU.
> > + *
> > + *  Copyright (C) 2007 by Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@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 version 2
> > + *  as published by the Free Software Foundation.
> > + *
> > + *  This program is distributed in the hope that it will be useful,
> > + *  but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > + *  GNU General Public License for more details.
> > + *
> > + *  You should have received a copy of the GNU General Public License
> > + *  along with this program; if not, write to:
> > + *  Free Software Foundation
> > + *  51 Franklin Street, Fifth Floor
> > + *  Boston, MA  02111-1301  USA
> > + *
> > + */
> > +
> > +#include <linux/module.h>
> > +#include <linux/kernel.h>
> > +#include <linux/compiler.h>
> > +#include <linux/pci.h>
> > +#include <linux/init.h>
> > +#include <linux/ioport.h>
> > +#include <linux/completion.h>
> > +#include <linux/interrupt.h>
> > +#include <linux/io.h>
> > +#include <linux/uaccess.h>
> > +#include <linux/irq.h>
> > +#include <linux/poll.h>
> > +#include <net/9p/9p.h>
> > +#include <net/9p/transport.h>
> > +
> > +#define P9PCI_DRIVER_NAME "9P PCI Device"
> > +#define P9PCI_DRIVER_VERSION "1"
> > +
> > +#define PCI_VENDOR_ID_9P     0x5002
> > +#define PCI_DEVICE_ID_9P     0x000D
> > +
> > +#define MAX_PCI_BUF          (4*1024) /* TODO: Get a number from lucho */
> > +
> > +struct p9pci_trans {
> > +     struct pci_dev          *pdev;
> > +     void __iomem            *ioaddr;
> > +     void __iomem            *tx;
> > +     void __iomem            *rx;
> > +     int                     irq;
> > +     int                     pos;
> > +     int                     len;
> > +     wait_queue_head_t       wait;
> > +};
> > +static struct p9pci_trans *p9pci_trans; /* single channel for now */
> > +
> > +static struct pci_device_id p9pci_tbl[] = {
> > +     {PCI_VENDOR_ID_9P, PCI_DEVICE_ID_9P, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 },
> > +     {0,}
> > +};
> > +
> > +static irqreturn_t p9pci_interrupt(int irq, void *dev)
> > +{
> > +     p9pci_trans = dev;
> > +     p9pci_trans->len = le32_to_cpu(readl(p9pci_trans->rx));
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "%p len %d\n", p9pci_trans->pdev,
> > +                                                     p9pci_trans->len);
> > +     iowrite32(0, p9pci_trans->ioaddr + 4);
> > +     wake_up_interruptible(&p9pci_trans->wait);
> > +     return IRQ_HANDLED;
> > +}
> > +
> > +static int p9pci_read(struct p9_trans *trans, void *v, int len)
> > +{
> > +     struct p9pci_trans *ts;
> > +
> > +     if (!trans || trans->status == Disconnected || !trans->priv)
> > +             return -EREMOTEIO;
> > +
> > +     ts = trans->priv;
> > +
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p rx %p tx %p buf %p len %d\n",
> > +                                             trans, ts->rx, ts->tx, v, len);
> > +     if (len > ts->len)
> > +             len = ts->len;
> > +
> > +     if (len) {
> > +             memcpy_fromio(v, ts->rx, len);
> > +             ts->len = 0;
> > +             /* let the host knows the message is consumed */
> > +             writel(0, ts->rx);
> > +             iowrite32(0, p9pci_trans->ioaddr + 4);
> > +             P9_DPRINTK(P9_DEBUG_TRANS, "zero rxlen %d txlen %d\n",
> > +                                             readl(ts->rx), readl(ts->tx));
> > +     }
> > +
> > +     return len;
> > +}
> > +
> > +static int p9pci_write(struct p9_trans *trans, void *v, int len)
> > +{
> > +     struct p9pci_trans *ts;
> > +
> > +     if (!trans || trans->status == Disconnected || !trans->priv)
> > +             return -EREMOTEIO;
> > +
> > +     ts = trans->priv;
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p rx %p tx %p buf %p len %d\n",
> > +                                             trans, ts->rx, ts->tx, v, len);
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "rxlen %d\n", readl(ts->rx));
> > +     if (readb(ts->tx) != 0)
> > +             return 0;
> > +
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "tx addr %p io addr %p\n", ts->tx,
> > +                                                             ts->ioaddr);
> > +     memcpy_toio(ts->tx, v, len);
> > +     iowrite32(len, ts->ioaddr);
> > +     return len;
> > +}
> > +
> > +static unsigned int
> > +p9pci_poll(struct p9_trans *trans, struct poll_table_struct *pt)
> > +{
> > +     int ret;
> > +     struct p9pci_trans *ts;
> > +
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p\n", trans);
> > +     if (!trans || trans->status != Connected || !trans->priv)
> > +             return -EREMOTEIO;
> > +
> > +     ts = trans->priv;
> > +     poll_wait(NULL, &ts->wait, pt);
> > +     ret = 0;
> > +     if (!readl(ts->tx))
> > +             ret |= POLLOUT;
> > +     if (readl(ts->rx))
> > +             ret |= POLLIN;
> > +
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "txlen %d rxlen %d\n", readl(ts->tx),
> > +                                                             readl(ts->rx));
> > +     return ret;
> > +}
> > +
> > +/**
> > + * p9_sock_close - shutdown socket
> > + * @trans: private socket structure
> > + *
> > + */
> > +static void p9pci_close(struct p9_trans *trans)
> > +{
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p\n", trans);
> > +}
> > +
> > +static struct p9_trans *p9pci_trans_create(const char *name, char *arg)
> > +{
> > +     struct p9_trans *trans;
> > +
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "\n");
> > +     trans = kmalloc(sizeof(struct p9_trans), GFP_KERNEL);
> > +     if (!trans)
> > +             return ERR_PTR(-ENOMEM);
> > +
> > +     trans->status = Connected;
> > +     trans->write = p9pci_write;
> > +     trans->read = p9pci_read;
> > +     trans->close = p9pci_close;
> > +     trans->poll = p9pci_poll;
> > +     trans->priv = p9pci_trans;
> > +     writel(0, p9pci_trans->tx);
> > +     writel(0, p9pci_trans->rx);
> > +
> > +     return trans;
> > +}
> > +
> > +static int __devinit p9pci_probe(struct pci_dev *pdev,
> > +             const struct pci_device_id *ent)
> > +{
> > +     int err;
> > +     u8 pci_rev;
> > +
> > +     if (p9pci_trans)
> > +             return -1;
> > +
> > +     pci_read_config_byte(pdev, PCI_REVISION_ID, &pci_rev);
> > +
> > +     if (pdev->vendor == PCI_VENDOR_ID_9P &&
> > +         pdev->device == PCI_DEVICE_ID_9P)
> > +             printk(KERN_INFO "pci dev %s (id %04x:%04x rev %02x) is a 9P\n",
> > +                    pci_name(pdev), pdev->vendor, pdev->device, pci_rev);
> > +
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "%p\n", pdev);
> > +     p9pci_trans = kzalloc(sizeof(*p9pci_trans), GFP_KERNEL);
> > +     p9pci_trans->irq = -1;
> > +     init_waitqueue_head(&p9pci_trans->wait);
> > +     err = pci_enable_device(pdev);
> > +     if (err)
> > +             goto error;
> > +
> > +     p9pci_trans->pdev = pdev;
> > +     err = pci_request_regions(pdev, "9p");
> > +     if (err)
> > +             goto error;
> > +
> > +     p9pci_trans->ioaddr = pci_iomap(pdev, 0, 8);
> > +     if (!p9pci_trans->ioaddr) {
> > +             P9_DPRINTK(P9_DEBUG_ERROR, "Cannot remap MMIO, aborting\n");
> > +             err = -EIO;
> > +             goto error;
> > +     }
> > +
> > +     p9pci_trans->tx = pci_iomap(pdev, 1, 0x20000);
> > +     p9pci_trans->rx = pci_iomap(pdev, 2, 0x20000);
> > +     pci_set_drvdata(pdev, p9pci_trans);
> > +     err = request_irq(pdev->irq, &p9pci_interrupt, 0, "p9pci", p9pci_trans);
> > +     if (err)
> > +             goto error;
> > +
> > +     p9pci_trans->irq = pdev->irq;
> > +     return 0;
> > +
> > +error:
> > +     P9_DPRINTK(P9_DEBUG_ERROR, "error %d\n", err);
> > +     if (p9pci_trans->irq >= 0) {
> > +             synchronize_irq(p9pci_trans->irq);
> > +             free_irq(p9pci_trans->irq, p9pci_trans);
> > +     }
> > +
> > +     if (p9pci_trans->pdev) {
> > +             pci_release_regions(pdev);
> > +             pci_iounmap(pdev, p9pci_trans->ioaddr);
> > +             pci_set_drvdata(pdev, NULL);
> > +             pci_disable_device(pdev);
> > +     }
> > +
> > +     kfree(p9pci_trans);
> > +     return -1;
> > +}
> > +
> > +static void __devexit p9pci_remove(struct pci_dev *pdev)
> > +{
> > +     P9_DPRINTK(P9_DEBUG_TRANS, "%p\n", pdev);
> > +     p9pci_trans = pci_get_drvdata(pdev);
> > +     if (!p9pci_trans)
> > +             return;
> > +
> > +     if (p9pci_trans->irq) {
> > +             synchronize_irq(p9pci_trans->irq);
> > +             free_irq(p9pci_trans->irq, p9pci_trans);
> > +     }
> > +
> > +     pci_release_regions(pdev);
> > +     pci_iounmap(pdev, p9pci_trans->ioaddr);
> > +     pci_set_drvdata(pdev, NULL);
> > +     kfree(p9pci_trans);
> > +     pci_disable_device(pdev);
> > +}
> > +
> > +static struct pci_driver p9pci_driver = {
> > +     .name           = P9PCI_DRIVER_NAME,
> > +     .id_table       = p9pci_tbl,
> > +     .probe          = p9pci_probe,
> > +     .remove         = __devexit_p(p9pci_remove),
> > +};
> > +
> > +static struct p9_trans_module p9_pci_trans = {
> > +     .name = "pci",
> > +     .maxsize = MAX_PCI_BUF,
> > +     .def = 0,
> > +     .create = p9pci_trans_create,
> > +};
> > +
> > +static int __init p9pci_init_module(void)
> > +{
> > +     v9fs_register_trans(&p9_pci_trans);
> > +     return pci_register_driver(&p9pci_driver);
> > +}
> > +
> > +static void __exit p9pci_cleanup_module(void)
> > +{
> > +     pci_unregister_driver(&p9pci_driver);
> > +     printk(KERN_ERR "Removal of 9p transports not implemented\n");
> > +     BUG();
> > +}
> > +
> > +module_init(p9pci_init_module);
> > +module_exit(p9pci_cleanup_module);
> > +
> > +MODULE_DEVICE_TABLE(pci, p9pci_tbl);
> > +MODULE_AUTHOR("Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>");
> > +MODULE_DESCRIPTION(P9PCI_DRIVER_NAME);
> > +MODULE_LICENSE("GPL");
> > +MODULE_VERSION(P9PCI_DRIVER_VERSION);
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> V9fs-developer mailing list
> V9fs-developer-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/v9fs-developer
>

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/

Re: [Lguest] [V9fs-developer] [RFC] 9p: add KVM/QEMUpci transport

[Dor Laor]

My current view of the IO stack is the following:

        -------------- --------------  ------------    ----------
----------   ---------
        |NET_PCI_BACK| |BLK_PCI_BACK|  |9P_PCI_BACK|   |NET_FRONT|
|BLK_FRONT| |9P_FRONT|
        -------------- --------------  ------------    ----------
----------   ---------

               -------------      ---------------
-------------------
               |KVM_PCI_BUS|      |hypercall_ops|
|shared_mem_virtio|
               -------------      ---------------
-------------------

So the 9P implementation should add the front end logic and the
p9_pci_backend that glues the shared_memory, pci_bus and hypercalls
together.                    



>That's also in our plans. There was no virtio support in KVM when I
>started working in the transport.
>
>Thanks,
>    Lucho
>
>On 8/29/07, Anthony Liguori <anthony-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org> wrote:
>> I think that it would be nicer to implement the p9 transport on top
of
>> virtio instead of directly on top of PCI.  I think your PCI transport
>> would make a pretty nice start of a PCI virtio transport though.
>>
>> Regards,
>>
>> Anthony Liguori
>>
>> On Tue, 2007-08-28 at 13:52 -0500, Eric Van Hensbergen wrote:
>> > From: Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>
>> >
>> > This adds a shared memory transport for a synthetic 9p device for
>> > paravirtualized file system support under KVM/QEMU.
>> >
>> > Signed-off-by: Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>
>> > Signed-off-by: Eric Van Hensbergen <ericvh-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> > ---
>> >  Documentation/filesystems/9p.txt |    2 +
>> >  net/9p/Kconfig                   |   10 ++-
>> >  net/9p/Makefile                  |    4 +
>> >  net/9p/trans_pci.c               |  295
>++++++++++++++++++++++++++++++++++++++
>> >  4 files changed, 310 insertions(+), 1 deletions(-)
>> >  create mode 100644 net/9p/trans_pci.c
>> >
>> > diff --git a/Documentation/filesystems/9p.txt
>b/Documentation/filesystems/9p.txt
>> > index 1a5f50d..e1879bd 100644
>> > --- a/Documentation/filesystems/9p.txt
>> > +++ b/Documentation/filesystems/9p.txt
>> > @@ -46,6 +46,8 @@ OPTIONS
>> >                       tcp  - specifying a normal TCP/IP connection
>> >                       fd   - used passed file descriptors for
>connection
>> >                                  (see rfdno and wfdno)
>> > +                     pci  - use a PCI pseudo device for 9p
>communication
>> > +                             over shared memory between a guest
and
>host
>> >
>> >    uname=name user name to attempt mount as on the remote server.
>The
>> >               server may override or ignore this value.  Certain
>user
>> > diff --git a/net/9p/Kconfig b/net/9p/Kconfig
>> > index 09566ae..8517560 100644
>> > --- a/net/9p/Kconfig
>> > +++ b/net/9p/Kconfig
>> > @@ -16,13 +16,21 @@ menuconfig NET_9P
>> >  config NET_9P_FD
>> >       depends on NET_9P
>> >       default y if NET_9P
>> > -     tristate "9P File Descriptor Transports (Experimental)"
>> > +     tristate "9p File Descriptor Transports (Experimental)"
>> >       help
>> >         This builds support for file descriptor transports for 9p
>> >         which includes support for TCP/IP, named pipes, or passed
>> >         file descriptors.  TCP/IP is the default transport for 9p,
>> >         so if you are going to use 9p, you'll likely want this.
>> >
>> > +config NET_9P_PCI
>> > +     depends on NET_9P
>> > +     tristate "9p PCI Shared Memory Transport (Experimental)"
>> > +     help
>> > +       This builds support for a PCI psuedo-device currently
>available
>> > +       under KVM/QEMU which allows for 9p transactions over shared
>> > +       memory between the guest and the host.
>> > +
>> >  config NET_9P_DEBUG
>> >       bool "Debug information"
>> >       depends on NET_9P
>> > diff --git a/net/9p/Makefile b/net/9p/Makefile
>> > index 7b2a67a..26ce89d 100644
>> > --- a/net/9p/Makefile
>> > +++ b/net/9p/Makefile
>> > @@ -1,5 +1,6 @@
>> >  obj-$(CONFIG_NET_9P) := 9pnet.o
>> >  obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o
>> > +obj-$(CONFIG_NET_9P_PCI) += 9pnet_pci.o
>> >
>> >  9pnet-objs := \
>> >       mod.o \
>> > @@ -14,3 +15,6 @@ obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o
>> >
>> >  9pnet_fd-objs := \
>> >       trans_fd.o \
>> > +
>> > +9pnet_pci-objs := \
>> > +     trans_pci.o \
>> > diff --git a/net/9p/trans_pci.c b/net/9p/trans_pci.c
>> > new file mode 100644
>> > index 0000000..36ddc5f
>> > --- /dev/null
>> > +++ b/net/9p/trans_pci.c
>> > @@ -0,0 +1,295 @@
>> > +/*
>> > + * net/9p/trans_pci.c
>> > + *
>> > + * 9P over PCI transport layer. For use with KVM/QEMU.
>> > + *
>> > + *  Copyright (C) 2007 by Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@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 version 2
>> > + *  as published by the Free Software Foundation.
>> > + *
>> > + *  This program is distributed in the hope that it will be
useful,
>> > + *  but WITHOUT ANY WARRANTY; without even the implied warranty of
>> > + *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> > + *  GNU General Public License for more details.
>> > + *
>> > + *  You should have received a copy of the GNU General Public
>License
>> > + *  along with this program; if not, write to:
>> > + *  Free Software Foundation
>> > + *  51 Franklin Street, Fifth Floor
>> > + *  Boston, MA  02111-1301  USA
>> > + *
>> > + */
>> > +
>> > +#include <linux/module.h>
>> > +#include <linux/kernel.h>
>> > +#include <linux/compiler.h>
>> > +#include <linux/pci.h>
>> > +#include <linux/init.h>
>> > +#include <linux/ioport.h>
>> > +#include <linux/completion.h>
>> > +#include <linux/interrupt.h>
>> > +#include <linux/io.h>
>> > +#include <linux/uaccess.h>
>> > +#include <linux/irq.h>
>> > +#include <linux/poll.h>
>> > +#include <net/9p/9p.h>
>> > +#include <net/9p/transport.h>
>> > +
>> > +#define P9PCI_DRIVER_NAME "9P PCI Device"
>> > +#define P9PCI_DRIVER_VERSION "1"
>> > +
>> > +#define PCI_VENDOR_ID_9P     0x5002
>> > +#define PCI_DEVICE_ID_9P     0x000D
>> > +
>> > +#define MAX_PCI_BUF          (4*1024) /* TODO: Get a number from
>lucho */
>> > +
>> > +struct p9pci_trans {
>> > +     struct pci_dev          *pdev;
>> > +     void __iomem            *ioaddr;
>> > +     void __iomem            *tx;
>> > +     void __iomem            *rx;
>> > +     int                     irq;
>> > +     int                     pos;
>> > +     int                     len;
>> > +     wait_queue_head_t       wait;
>> > +};
>> > +static struct p9pci_trans *p9pci_trans; /* single channel for now
>*/
>> > +
>> > +static struct pci_device_id p9pci_tbl[] = {
>> > +     {PCI_VENDOR_ID_9P, PCI_DEVICE_ID_9P, PCI_ANY_ID, PCI_ANY_ID,
>0, 0, 0 },
>> > +     {0,}
>> > +};
>> > +
>> > +static irqreturn_t p9pci_interrupt(int irq, void *dev)
>> > +{
>> > +     p9pci_trans = dev;
>> > +     p9pci_trans->len = le32_to_cpu(readl(p9pci_trans->rx));
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "%p len %d\n", p9pci_trans->pdev,
>> > +                                                     p9pci_trans-
>>len);
>> > +     iowrite32(0, p9pci_trans->ioaddr + 4);
>> > +     wake_up_interruptible(&p9pci_trans->wait);
>> > +     return IRQ_HANDLED;
>> > +}
>> > +
>> > +static int p9pci_read(struct p9_trans *trans, void *v, int len)
>> > +{
>> > +     struct p9pci_trans *ts;
>> > +
>> > +     if (!trans || trans->status == Disconnected || !trans->priv)
>> > +             return -EREMOTEIO;
>> > +
>> > +     ts = trans->priv;
>> > +
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p rx %p tx %p buf %p len
>%d\n",
>> > +                                             trans, ts->rx,
ts->tx,
>v, len);
>> > +     if (len > ts->len)
>> > +             len = ts->len;
>> > +
>> > +     if (len) {
>> > +             memcpy_fromio(v, ts->rx, len);
>> > +             ts->len = 0;
>> > +             /* let the host knows the message is consumed */
>> > +             writel(0, ts->rx);
>> > +             iowrite32(0, p9pci_trans->ioaddr + 4);
>> > +             P9_DPRINTK(P9_DEBUG_TRANS, "zero rxlen %d txlen
%d\n",
>> > +                                             readl(ts->rx),
>readl(ts->tx));
>> > +     }
>> > +
>> > +     return len;
>> > +}
>> > +
>> > +static int p9pci_write(struct p9_trans *trans, void *v, int len)
>> > +{
>> > +     struct p9pci_trans *ts;
>> > +
>> > +     if (!trans || trans->status == Disconnected || !trans->priv)
>> > +             return -EREMOTEIO;
>> > +
>> > +     ts = trans->priv;
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p rx %p tx %p buf %p len
>%d\n",
>> > +                                             trans, ts->rx,
ts->tx,
>v, len);
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "rxlen %d\n", readl(ts->rx));
>> > +     if (readb(ts->tx) != 0)
>> > +             return 0;
>> > +
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "tx addr %p io addr %p\n", ts->tx,
>> > +                                                             ts-
>>ioaddr);
>> > +     memcpy_toio(ts->tx, v, len);
>> > +     iowrite32(len, ts->ioaddr);
>> > +     return len;
>> > +}
>> > +
>> > +static unsigned int
>> > +p9pci_poll(struct p9_trans *trans, struct poll_table_struct *pt)
>> > +{
>> > +     int ret;
>> > +     struct p9pci_trans *ts;
>> > +
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p\n", trans);
>> > +     if (!trans || trans->status != Connected || !trans->priv)
>> > +             return -EREMOTEIO;
>> > +
>> > +     ts = trans->priv;
>> > +     poll_wait(NULL, &ts->wait, pt);
>> > +     ret = 0;
>> > +     if (!readl(ts->tx))
>> > +             ret |= POLLOUT;
>> > +     if (readl(ts->rx))
>> > +             ret |= POLLIN;
>> > +
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "txlen %d rxlen %d\n", readl(ts-
>>tx),
>> > +
>readl(ts->rx));
>> > +     return ret;
>> > +}
>> > +
>> > +/**
>> > + * p9_sock_close - shutdown socket
>> > + * @trans: private socket structure
>> > + *
>> > + */
>> > +static void p9pci_close(struct p9_trans *trans)
>> > +{
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "trans %p\n", trans);
>> > +}
>> > +
>> > +static struct p9_trans *p9pci_trans_create(const char *name, char
>*arg)
>> > +{
>> > +     struct p9_trans *trans;
>> > +
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "\n");
>> > +     trans = kmalloc(sizeof(struct p9_trans), GFP_KERNEL);
>> > +     if (!trans)
>> > +             return ERR_PTR(-ENOMEM);
>> > +
>> > +     trans->status = Connected;
>> > +     trans->write = p9pci_write;
>> > +     trans->read = p9pci_read;
>> > +     trans->close = p9pci_close;
>> > +     trans->poll = p9pci_poll;
>> > +     trans->priv = p9pci_trans;
>> > +     writel(0, p9pci_trans->tx);
>> > +     writel(0, p9pci_trans->rx);
>> > +
>> > +     return trans;
>> > +}
>> > +
>> > +static int __devinit p9pci_probe(struct pci_dev *pdev,
>> > +             const struct pci_device_id *ent)
>> > +{
>> > +     int err;
>> > +     u8 pci_rev;
>> > +
>> > +     if (p9pci_trans)
>> > +             return -1;
>> > +
>> > +     pci_read_config_byte(pdev, PCI_REVISION_ID, &pci_rev);
>> > +
>> > +     if (pdev->vendor == PCI_VENDOR_ID_9P &&
>> > +         pdev->device == PCI_DEVICE_ID_9P)
>> > +             printk(KERN_INFO "pci dev %s (id %04x:%04x rev %02x)
>is a 9P\n",
>> > +                    pci_name(pdev), pdev->vendor, pdev->device,
>pci_rev);
>> > +
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "%p\n", pdev);
>> > +     p9pci_trans = kzalloc(sizeof(*p9pci_trans), GFP_KERNEL);
>> > +     p9pci_trans->irq = -1;
>> > +     init_waitqueue_head(&p9pci_trans->wait);
>> > +     err = pci_enable_device(pdev);
>> > +     if (err)
>> > +             goto error;
>> > +
>> > +     p9pci_trans->pdev = pdev;
>> > +     err = pci_request_regions(pdev, "9p");
>> > +     if (err)
>> > +             goto error;
>> > +
>> > +     p9pci_trans->ioaddr = pci_iomap(pdev, 0, 8);
>> > +     if (!p9pci_trans->ioaddr) {
>> > +             P9_DPRINTK(P9_DEBUG_ERROR, "Cannot remap MMIO,
>aborting\n");
>> > +             err = -EIO;
>> > +             goto error;
>> > +     }
>> > +
>> > +     p9pci_trans->tx = pci_iomap(pdev, 1, 0x20000);
>> > +     p9pci_trans->rx = pci_iomap(pdev, 2, 0x20000);
>> > +     pci_set_drvdata(pdev, p9pci_trans);
>> > +     err = request_irq(pdev->irq, &p9pci_interrupt, 0, "p9pci",
>p9pci_trans);
>> > +     if (err)
>> > +             goto error;
>> > +
>> > +     p9pci_trans->irq = pdev->irq;
>> > +     return 0;
>> > +
>> > +error:
>> > +     P9_DPRINTK(P9_DEBUG_ERROR, "error %d\n", err);
>> > +     if (p9pci_trans->irq >= 0) {
>> > +             synchronize_irq(p9pci_trans->irq);
>> > +             free_irq(p9pci_trans->irq, p9pci_trans);
>> > +     }
>> > +
>> > +     if (p9pci_trans->pdev) {
>> > +             pci_release_regions(pdev);
>> > +             pci_iounmap(pdev, p9pci_trans->ioaddr);
>> > +             pci_set_drvdata(pdev, NULL);
>> > +             pci_disable_device(pdev);
>> > +     }
>> > +
>> > +     kfree(p9pci_trans);
>> > +     return -1;
>> > +}
>> > +
>> > +static void __devexit p9pci_remove(struct pci_dev *pdev)
>> > +{
>> > +     P9_DPRINTK(P9_DEBUG_TRANS, "%p\n", pdev);
>> > +     p9pci_trans = pci_get_drvdata(pdev);
>> > +     if (!p9pci_trans)
>> > +             return;
>> > +
>> > +     if (p9pci_trans->irq) {
>> > +             synchronize_irq(p9pci_trans->irq);
>> > +             free_irq(p9pci_trans->irq, p9pci_trans);
>> > +     }
>> > +
>> > +     pci_release_regions(pdev);
>> > +     pci_iounmap(pdev, p9pci_trans->ioaddr);
>> > +     pci_set_drvdata(pdev, NULL);
>> > +     kfree(p9pci_trans);
>> > +     pci_disable_device(pdev);
>> > +}
>> > +
>> > +static struct pci_driver p9pci_driver = {
>> > +     .name           = P9PCI_DRIVER_NAME,
>> > +     .id_table       = p9pci_tbl,
>> > +     .probe          = p9pci_probe,
>> > +     .remove         = __devexit_p(p9pci_remove),
>> > +};
>> > +
>> > +static struct p9_trans_module p9_pci_trans = {
>> > +     .name = "pci",
>> > +     .maxsize = MAX_PCI_BUF,
>> > +     .def = 0,
>> > +     .create = p9pci_trans_create,
>> > +};
>> > +
>> > +static int __init p9pci_init_module(void)
>> > +{
>> > +     v9fs_register_trans(&p9_pci_trans);
>> > +     return pci_register_driver(&p9pci_driver);
>> > +}
>> > +
>> > +static void __exit p9pci_cleanup_module(void)
>> > +{
>> > +     pci_unregister_driver(&p9pci_driver);
>> > +     printk(KERN_ERR "Removal of 9p transports not
implemented\n");
>> > +     BUG();
>> > +}
>> > +
>> > +module_init(p9pci_init_module);
>> > +module_exit(p9pci_cleanup_module);
>> > +
>> > +MODULE_DEVICE_TABLE(pci, p9pci_tbl);
>> > +MODULE_AUTHOR("Latchesar Ionkov <lucho-OnYtXJJ0/fesTnJN9+BGXg@public.gmane.org>");
>> > +MODULE_DESCRIPTION(P9PCI_DRIVER_NAME);
>> > +MODULE_LICENSE("GPL");
>> > +MODULE_VERSION(P9PCI_DRIVER_VERSION);
>>
>>
>>
----------------------------------------------------------------------
>---
>> This SF.net email is sponsored by: Splunk Inc.
>> Still grepping through log files to find problems?  Stop.
>> Now Search log events and configuration files using AJAX and a
>browser.
>> Download your FREE copy of Splunk now >>  http://get.splunk.com/
>> _______________________________________________
>> V9fs-developer mailing list
>> V9fs-developer-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
>> https://lists.sourceforge.net/lists/listinfo/v9fs-developer
>>
>_______________________________________________
>Lguest mailing list
>Lguest-mnsaURCQ41sdnm+yROfE0A@public.gmane.org
>https://ozlabs.org/mailman/listinfo/lguest

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/

Re: [RFC] 9p Virtualization Transports

[Andi Kleen]

On Tuesday 28 August 2007 20:52:36 Eric Van Hensbergen wrote:
> This patch set contains a set of virtualization transports for the 9p file
> system intended to provide a mechanism for guests to access a portion of the
> hosts name space without having to go through a virtualized network.  

It might be useful to convert UML's hostfs to something like this. It's essentially
the same.

-Andi

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/

Re: [Lguest] [RFC] 9p Virtualization Transports

[Rusty Russell]

On Tue, 2007-08-28 at 13:52 -0500, Eric Van Hensbergen wrote:
> The lguest and kvm transports are functional, but we are still working out
> remaining bugs and need to spend some time focusing on performance issues.
> I wanted to send out this "preview" patch set to the community to solicit
> ideas on things we can do differently/better.

Hi Eric,

	Patches look reasonable, but just a heads-up: lguest will be moving to
virtio, as will kvm.  That means a single implementation for both
(yay!), but it does complicate your life in the short term 8(

	Dor has published a kvm virtio implementation, and we've already
discussed a couple of modifications.  I expect that to be nailed in the
next 2 weeks tho, and lguest will follow.

Thanks!
Rusty.


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/

Re: [Lguest] [RFC] 9p Virtualization Transports

[Eric Van Hensbergen]

On 9/1/07, Rusty Russell <rusty-8n+1lVoiYb80n/F98K4Iww@public.gmane.org> wrote:
> On Tue, 2007-08-28 at 13:52 -0500, Eric Van Hensbergen wrote:
> > The lguest and kvm transports are functional, but we are still working out
> > remaining bugs and need to spend some time focusing on performance issues.
> > I wanted to send out this "preview" patch set to the community to solicit
> > ideas on things we can do differently/better.
>
>         Patches look reasonable, but just a heads-up: lguest will be moving to
> virtio, as will kvm.  That means a single implementation for both
> (yay!), but it does complicate your life in the short term 8(
>
>         Dor has published a kvm virtio implementation, and we've already
> discussed a couple of modifications.  I expect that to be nailed in the
> next 2 weeks tho, and lguest will follow.
>

yeah, I've been emailing Dor -- it sounds like he'll have stuff ready
for the 2.6.24 merge window -- that being the case, I'll write a
virtio transport and mothball the PCI and lguest transports.  They
were straightforward to write (a couple hours for the lguest
transport) and the lguest transport was a good learning experience --
so I'm not shedding tears over wasted effort.

            -eric

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/