Re: [Xen-devel] [PATCH 2/4] Qemu-Xen-vTPM: Register Xen stubdom vTPM frontend driver

["Xu, Quan" <quan.xu@xxxxxxxxx>]

> -----Original Message-----
> From: Stefano Stabellini [mailto:stefano.stabellini@xxxxxxxxxxxxx]
> Sent: Monday, November 03, 2014 7:54 PM
> To: Xu, Quan
> Cc: qemu-devel@xxxxxxxxxx; xen-devel@xxxxxxxxxxxxx;
> stefano.stabellini@xxxxxxxxxxxxx
> Subject: Re: [PATCH 2/4] Qemu-Xen-vTPM: Register Xen stubdom vTPM
> frontend driver
> 
> On Sun, 2 Nov 2014, Quan Xu wrote:
> > This drvier transfers any request/repond between TPM xenstubdoms
> > driver and Xen vTPM stubdom, and facilitates communications between
> > Xen vTPM stubdom domain and vTPM xenstubdoms driver
> >
> > Signed-off-by: Quan Xu <quan.xu@xxxxxxxxx>
> 
> Please describe what changes did make to xen_backend.c and why.
> The commit message should contains info on all the changes made by the
> patch below.
> 

The following is code process when Qemu is running with Xen. 
##code process
[...]
 xen_hvm_init()
    -->xen_be_register()
        -->xenstore_scan()
            -->xen_be_check_state()

    -->xen_vtpm_register()

ideally, I can register 'vtpm' via xen_vtpm_register() as

+ xen_be_register("console", &xen_console_ops);
+ xen_be_register("vkbd", &xen_kbdmouse_ops);
+ xen_be_register("qdisk", &xen_blkdev_ops);

but there are 2 reasons why I add xen_vtpm_register(), instead
of xen_be_register().

1. The backend of TPM is runing in a Xen stubDom, not Domain 0.
some functions are not working, for example 'setup watch' and
'look for backend' in xenstore_scan()

2. there is a thread runing in Xen stubDom [event_thread()], it will
handle backend status when the frontend is initialized. It is not
compatible with xen_be_check_state(). xen_be_check_state() always tries 
to modify the status of backend. 

as there is always a tradeoff, if I force to integrate this case into
xen_be_register(), there are maybe a lot of 'if ... else.. '. It will
break the code architecture. Also I should leverage existing source
code with minimum modifcation. i add 'DEVOPS_FLAG_STUBDOM_BE' flag in
include/hw/xen/xen_backend.h to indicate that device backend is Xen
stubDom.


> Please also describe what is the "Xen vTPM stubdom", what is the
> "vTPM xenstubdoms driver" and how the communicate with each others.
> 
In previous email, I have explained what is "Xen vTPM stubdom", what is the 
" vTPM xenstubdoms driver ". 

Let me describe how the communicate with each others.

            |        |  ^       |
            |        v  |       |
            |        vTPM      |
            | XenStubdoms driver |  (new ..)
            +-----------------------------+
                     |  ^
                     v  |
            +-----------------------------+
            |  xen_vtpmdev_ops |  (new ..)
            +-----------------------------+

xen_vtpmdev_ops is initialized with the following process:
 xen_hvm_init()
    [...]
    -->xen_vtpm_register()
      [...]
        --> vtpm_alloc()
        --> vtpm_initialise()

## 
vTPM XenStubdoms driver is initialized by Qemu command line options,
        "-tpmdev xenstubdoms,id=xenvtpm0 -device tpm-tis,tpmdev=xenvtpm0"


(the communicate with each others in following function.
.. hw/tpm/tpm_xenstubdoms.c
static int tpm_xenstubdoms_unix_transfer(const TPMLocality *locty_data)
{
[...]
    xen_vtpm_send(locty_data->w_buffer.buffer, locty_data->w_offset);
    xen_vtpm_recv(locty_data->r_buffer.buffer, &rlen);
[...]
} 


> Where does the vTPM backend lives?

Xen stubDom. 


> 
> 
> >  hw/xen/Makefile.objs         |   1 +
> >  hw/xen/xen_backend.c         | 182 ++++++++++++++++++++++-
> >  hw/xen/xen_stubdom_vtpm.c    | 333
> +++++++++++++++++++++++++++++++++++++++++++
> >  include/hw/xen/xen_backend.h |  11 ++
> >  include/hw/xen/xen_common.h  |   6 +
> >  xen-hvm.c                    |  13 ++
> >  6 files changed, 544 insertions(+), 2 deletions(-)
> >  create mode 100644 hw/xen/xen_stubdom_vtpm.c
> >
> > diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
> > index a0ca0aa..724df8d 100644
> > --- a/hw/xen/Makefile.objs
> > +++ b/hw/xen/Makefile.objs
> > @@ -1,5 +1,6 @@
> >  # xen backend driver support
> >  common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o
> xen_devconfig.o
> > +common-obj-$(CONFIG_TPM_XENSTUBDOMS) += xen_stubdom_vtpm.o
> >
> >  obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
> >  obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o
> xen_pt_config_init.o xen_pt_msi.o
> > diff --git a/hw/xen/xen_backend.c b/hw/xen/xen_backend.c
> > index b2cb22b..45a5778 100644
> > --- a/hw/xen/xen_backend.c
> > +++ b/hw/xen/xen_backend.c
> > @@ -194,6 +194,32 @@ int xen_be_set_state(struct XenDevice *xendev,
> enum xenbus_state state)
> >      return 0;
> >  }
> >
> > +/*get stubdom backend*/
> > +static char *xen_stubdom_be(const char *type, int dom, int dev)
> > +{
> > +    char *val, *domu;
> > +    char path[XEN_BUFSIZE];
> > +    unsigned int len, ival;
> > +
> > +    /*front domu*/
> > +    domu = xs_get_domain_path(xenstore, dom);
> > +    snprintf(path, sizeof(path), "%s/device/%s/%d/backend-id",
> > +             domu, type, dev);
> > +    g_free(domu);
> > +
> > +    val = xs_read(xenstore, 0, path, &len);
> > +    if (!val || 1 != sscanf(val, "%d", &ival)) {
> > +        g_free(val);
> > +        return NULL;
> > +    }
> > +    g_free(val);
> > +
> > +    /*backend domu*/
> > +    domu = xs_get_domain_path(xenstore, ival);
> > +
> > +    return domu;
> > +}
> > +
> >  /* ------------------------------------------------------------- */
> >
> >  struct XenDevice *xen_be_find_xendev(const char *type, int dom, int
> dev)
> > @@ -222,6 +248,7 @@ static struct XenDevice *xen_be_get_xendev(const
> char *type, int dom, int dev,
> >                                             struct XenDevOps *ops)
> >  {
> >      struct XenDevice *xendev;
> > +    char *stub;
> >
> >      xendev = xen_be_find_xendev(type, dom, dev);
> >      if (xendev) {
> > @@ -235,8 +262,15 @@ static struct XenDevice
> *xen_be_get_xendev(const char *type, int dom, int dev,
> >      xendev->dev   = dev;
> >      xendev->ops   = ops;
> >
> > -    snprintf(xendev->be, sizeof(xendev->be), "backend/%s/%d/%d",
> > -             xendev->type, xendev->dom, xendev->dev);
> > +    if (ops->flags & DEVOPS_FLAG_STUBDOM_BE) {
> > +        stub = xen_stubdom_be(xendev->type, xendev->dom,
> xendev->dev);
> > +        snprintf(xendev->be, sizeof(xendev->be),
> "%s/backend/%s/%d/%d",
> > +                 stub, xendev->type, xendev->dom, xendev->dev);
> > +        g_free(stub);
> > +    } else {
> > +        snprintf(xendev->be, sizeof(xendev->be), "backend/%s/%d/%d",
> > +                 xendev->type, xendev->dom, xendev->dev);
> > +    }
> >      snprintf(xendev->name, sizeof(xendev->name), "%s-%d",
> >               xendev->type, xendev->dev);
> >
> > @@ -611,6 +645,47 @@ static int xenstore_scan(const char *type, int
> dom, struct XenDevOps *ops)
> >      return 0;
> >  }
> >
> > +static void stubdom_update_be(char *watch, char *type, int dom,
> > +                              struct XenDevOps *ops)
> > +{
> > +    struct XenDevice *xendev;
> > +    char path[XEN_BUFSIZE];
> > +    char *ptr, *bepath;
> > +    unsigned int len, dev;
> > +
> > +    if (!(ops->flags & DEVOPS_FLAG_STUBDOM_BE)) {
> > +        return;
> > +    }
> > +
> > +    len = snprintf(path, sizeof(path), "backend/%s/%d", type, dom);
> > +    ptr = strstr(watch, path);
> > +    if (ptr == NULL) {
> > +        return;
> > +    }
> > +
> > +    if (sscanf(ptr+len, "/%u/%255s", &dev, path) != 2) {
> > +        strcpy(path, "");
> > +        if (sscanf(ptr+len, "/%u", &dev) != 1) {
> > +            dev = -1;
> > +        }
> > +    }
> > +
> > +    if (dev == -1) {
> > +        return;
> > +    }
> > +
> > +    xendev = xen_be_get_xendev(type, dom, dev, ops);
> > +    if (xendev != NULL) {
> > +        bepath = xs_read(xenstore, 0, xendev->be, &len);
> > +        if (bepath == NULL) {
> > +            xen_be_del_xendev(dom, dev);
> > +        } else {
> > +            free(bepath);
> > +            xen_be_backend_changed(xendev, path);
> > +        }
> > +    }
> > +}
> > +
> >  static void xenstore_update_be(char *watch, char *type, int dom,
> >                                 struct XenDevOps *ops)
> >  {
> > @@ -681,6 +756,10 @@ static void xenstore_update(void *unused)
> >      if (sscanf(vec[XS_WATCH_TOKEN], "fe:%" PRIxPTR, &ptr) == 1) {
> >          xenstore_update_fe(vec[XS_WATCH_PATH], (void*)ptr);
> >      }
> > +    if (sscanf(vec[XS_WATCH_TOKEN], "stub:%" PRIxPTR ":%d:%"
> PRIxPTR,
> > +               &type, &dom, &ops) == 3) {
> > +        stubdom_update_be(vec[XS_WATCH_PATH], (void *)type, dom,
> (void *)ops);
> > +    }
> >
> >  cleanup:
> >      free(vec);
> > @@ -732,11 +811,74 @@ err:
> >      return -1;
> >  }
> >
> > +int xen_vtpm_register(struct XenDevOps *ops)
> > +{
> > +    struct XenDevice *xendev;
> > +    char path[XEN_BUFSIZE], token[XEN_BUFSIZE];
> > +    char *domu;
> > +    unsigned int cdev, j, rc;
> > +    const char *type = "vtpm";
> > +    char **dev = NULL;
> > +
> > +    /*front domu*/
> > +    domu = xs_get_domain_path(xenstore, xen_domid);
> > +    snprintf(path, sizeof(path), "%s/device/%s",
> > +             domu, type);
> > +    free(domu);
> > +    dev = xs_directory(xenstore, 0, path, &cdev);
> > +    if (dev == NULL) {
> > +        return 0;
> > +    }
> > +
> > +    for (j = 0; j < cdev; j++) {
> > +        xendev = xen_be_get_xendev(type, xen_domid, atoi(dev[j]),
> ops);
> > +        if (xendev == NULL) {
> > +            xen_be_printf(xendev, 0, "xen_vtpm_register xendev is
> NULL.\n");
> > +            continue;
> > +        }
> > +
> > +        if (xendev->ops->initialise) {
> > +            rc = xendev->ops->initialise(xendev);
> > +
> > +            /*if initialise failed, delete it*/
> > +            if (rc != 0) {
> > +                xen_be_del_xendev(xen_domid, atoi(dev[j]));
> > +                continue;
> > +            }
> > +        }
> > +
> > +        /*setup watch*/
> > +        snprintf(token, sizeof(token), "stub:%p:%d:%p",
> > +                 type, xen_domid, xendev->ops);
> > +        if (!xs_watch(xenstore, xendev->be, token)) {
> > +            xen_be_printf(xendev, 0, "xen_vtpm_register xs_watch
> failed.\n");
> > +            return -1;
> > +        }
> > +    }
> > +
> > +    free(dev);
> > +    return 0;
> > +}
> 
> What does this function do? I sholdn't need  to guess from the code, I
> should be able to tell from the patch description.
> 

1. setup watch
2. look for backends
3. initialize the frontend

> 
> >  int xen_be_register(const char *type, struct XenDevOps *ops)
> >  {
> >      return xenstore_scan(type, xen_domid, ops);
> >  }
> >
> > +int xen_be_alloc_unbound(struct XenDevice *xendev, int dom, int
> remote_dom)
> > +{
> > +    xendev->local_port =
> xc_evtchn_bind_unbound_port(xendev->evtchndev,
> > +
> remote_dom);
> > +    if (xendev->local_port == -1) {
> > +        xen_be_printf(xendev, 0, "xc_evtchn_alloc_unbound failed\n");
> > +        return -1;
> > +    }
> > +    xen_be_printf(xendev, 2, "bind evtchn port %d\n",
> xendev->local_port);
> > +    qemu_set_fd_handler(xc_evtchn_fd(xendev->evtchndev),
> > +                        xen_be_evtchn_event, NULL, xendev);
> > +    return 0;
> > +}
> > +
> >  int xen_be_bind_evtchn(struct XenDevice *xendev)
> >  {
> >      if (xendev->local_port != -1) {
> > @@ -770,6 +912,42 @@ int xen_be_send_notify(struct XenDevice
> *xendev)
> >      return xc_evtchn_notify(xendev->evtchndev, xendev->local_port);
> >  }
> >
> > +int xen_vtpm_send(unsigned char *buf, size_t count)
> > +{
> > +    struct XenDevice *xendev;
> > +    int rc = -1;
> > +
> > +    xendev = xen_be_find_xendev("vtpm", xen_domid, 0);
> > +    if (xendev == NULL) {
> > +        xen_be_printf(xendev, 0, "Con not find vtpm device\n");
> > +        return -1;
> > +    }
> > +
> > +    if (xendev->ops->send) {
> > +        rc = xendev->ops->send(xendev, buf, count);
> > +    }
> > +
> > +    return rc;
> > +}
> > +
> > +int xen_vtpm_recv(unsigned char *buf, size_t *count)
> > +{
> > +    struct XenDevice *xendev;
> > +    int rc = -1;
> > +
> > +    xendev = xen_be_find_xendev("vtpm", xen_domid, 0);
> > +    if (xendev == NULL) {
> > +        xen_be_printf(xendev, 0, "Con not find vtpm device\n");
> > +        return -1;
> > +    }
> > +
> > +    if (xendev->ops->recv) {
> > +        xendev->ops->recv(xendev, buf, count);
> > +    }
> > +
> > +    return rc;
> > +}
> 
> xen_backend.c is supposed to be generic, so stubdom functions might be
> OK but vtpm specific functions should not be here.
> 


OK, I will move it to hw/tpm. 

> 
> >  /*
> >   * msg_level:
> >   *  0 == errors (stderr + logfile).
> > diff --git a/hw/xen/xen_stubdom_vtpm.c b/hw/xen/xen_stubdom_vtpm.c
> > new file mode 100644
> > index 0000000..0d740c1
> > --- /dev/null
> > +++ b/hw/xen/xen_stubdom_vtpm.c
> > @@ -0,0 +1,333 @@
> > +/*
> > + * Connect to Xen vTPM stubdom domain
> > + *
> > + *  Copyright (c) 2014 Intel Corporation
> > + *  Authors:
> > + *    Quan Xu <quan.xu@xxxxxxxxx>
> > + *
> > + * This library is free software; you can redistribute it and/or
> > + * modify it under the terms of the GNU Lesser General Public
> > + * License as published by the Free Software Foundation; either
> > + * version 2 of the License, or (at your option) any later version.
> > + *
> > + * This library 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
> > + * Lesser General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU Lesser General Public
> > + * License along with this library; if not, see
> <http://www.gnu.org/licenses/>
> > + */
> > +
> > +#include <stdio.h>
> > +#include <stdlib.h>
> > +#include <stdarg.h>
> > +#include <string.h>
> > +#include <unistd.h>
> > +#include <signal.h>
> > +#include <inttypes.h>
> > +#include <time.h>
> > +#include <fcntl.h>
> > +#include <errno.h>
> > +#include <sys/ioctl.h>
> > +#include <sys/types.h>
> > +#include <sys/stat.h>
> > +#include <sys/mman.h>
> > +#include <sys/uio.h>
> > +
> > +#include "hw/hw.h"
> > +#include "block/aio.h"
> > +#include "hw/xen/xen_backend.h"
> > +
> > +enum tpmif_state {
> > +    TPMIF_STATE_IDLE,        /* no contents / vTPM idle / cancel
> complete */
> > +    TPMIF_STATE_SUBMIT,      /* request ready / vTPM working */
> > +    TPMIF_STATE_FINISH,      /* response ready / vTPM idle */
> > +    TPMIF_STATE_CANCEL,      /* cancel requested / vTPM working */
> > +};
> > +
> > +static AioContext *vtpm_aio_ctx;
> > +
> > +enum status_bits {
> > +    VTPM_STATUS_RUNNING  = 0x1,
> > +    VTPM_STATUS_IDLE     = 0x2,
> > +    VTPM_STATUS_RESULT   = 0x4,
> > +    VTPM_STATUS_CANCELED = 0x8,
> > +};
> > +
> > +struct tpmif_shared_page {
> > +    uint32_t length;         /* request/response length in bytes */
> > +
> > +    uint8_t  state;           /* enum tpmif_state */
> > +    uint8_t  locality;        /* for the current request */
> > +    uint8_t  pad;             /* should be zero */
> > +
> > +    uint8_t  nr_extra_pages;  /* extra pages for long packets; may be
> zero */
> > +    uint32_t extra_pages[0]; /* grant IDs; length is actually
> nr_extra_pages */
> > +};
> > +
> > +struct QEMUBH {
> > +    AioContext *ctx;
> > +    QEMUBHFunc *cb;
> > +    void *opaque;
> > +    QEMUBH *next;
> > +    bool scheduled;
> > +    bool idle;
> > +    bool deleted;
> > +};
> > +
> > +struct XenVtpmDev {
> > +    struct XenDevice xendev;  /* must be first */
> > +    struct           tpmif_shared_page *shr;
> > +    xc_gntshr        *xen_xcs;
> > +    int              ring_ref;
> > +    int              bedomid;
> > +    QEMUBH           *sr_bh;
> > +};
> > +
> > +static uint8_t vtpm_status(struct XenVtpmDev *vtpmdev)
> > +{
> > +    switch (vtpmdev->shr->state) {
> > +    case TPMIF_STATE_IDLE:
> > +    case TPMIF_STATE_FINISH:
> > +        return VTPM_STATUS_IDLE;
> > +    case TPMIF_STATE_SUBMIT:
> > +    case TPMIF_STATE_CANCEL:
> > +        return VTPM_STATUS_RUNNING;
> > +    default:
> > +        return 0;
> > +    }
> > +}
> > +
> > +static int xenbus_switch_state(struct XenDevice *xendev, enum
> xenbus_state xbus)
> > +{
> > +    xs_transaction_t xbt = XBT_NULL;
> > +
> > +    if (xendev->fe_state == xbus) {
> > +        return 0;
> > +    }
> > +
> > +    xendev->fe_state = xbus;
> > +
> > +retry_transaction:
> > +    xbt = xs_transaction_start(xenstore);
> > +    if (xbt == XBT_NULL) {
> > +        goto abort_transaction;
> > +    }
> > +
> > +    if (xenstore_write_int(xendev->fe, "state", xbus)) {
> > +        goto abort_transaction;
> > +    }
> > +
> > +    if (!xs_transaction_end(xenstore, xbt, 0)) {
> > +        if (errno == EAGAIN) {
> > +            goto retry_transaction;
> > +        }
> > +    }
> > +
> > +    return 0;
> > +
> > +abort_transaction:
> > +    xs_transaction_end(xenstore, xbt, 1);
> > +    return -1;
> > +}
> > +
> > +static int vtpm_aio_wait(QEMUBH *qbh)
> > +{
> > +    return aio_poll(qbh->ctx, true);
> > +}
> > +
> > +static void sr_bh_handler(void *opaque)
> > +{
> > +}
> > +
> > +static int vtpm_recv(struct XenDevice *xendev, uint8_t* buf, size_t
> *count)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +    struct tpmif_shared_page *shr = vtpmdev->shr;
> > +    unsigned int offset;
> > +
> > +    if (shr->state == TPMIF_STATE_IDLE) {
> > +        return -ECANCELED;
> > +    }
> > +
> > +    while (vtpm_status(vtpmdev) != VTPM_STATUS_IDLE) {
> > +        vtpm_aio_wait(vtpmdev->sr_bh);
> > +    }
> > +
> > +    offset = sizeof(*shr) + 4*shr->nr_extra_pages;
> > +    memcpy(buf, offset + (uint8_t *)shr, shr->length);
> > +    *count = shr->length;
> > +
> > +    return 0;
> > +}
> > +
> > +static int vtpm_send(struct XenDevice *xendev, uint8_t* buf, size_t
> count)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +    struct tpmif_shared_page *shr = vtpmdev->shr;
> > +    unsigned int offset = sizeof(*shr) + 4*shr->nr_extra_pages;
> > +
> > +    while (vtpm_status(vtpmdev) != VTPM_STATUS_IDLE) {
> > +        vtpm_aio_wait(vtpmdev->sr_bh);
> > +    }
> > +
> > +    memcpy(offset + (uint8_t *)shr, buf, count);
> > +    shr->length = count;
> > +    barrier();
> > +    shr->state = TPMIF_STATE_SUBMIT;
> > +    xen_wmb();
> > +    xen_be_send_notify(&vtpmdev->xendev);
> > +
> > +    while (vtpm_status(vtpmdev) != VTPM_STATUS_IDLE) {
> > +        vtpm_aio_wait(vtpmdev->sr_bh);
> > +    }
> > +
> > +    return count;
> > +}
> > +
> > +static int vtpm_initialise(struct XenDevice *xendev)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +    xs_transaction_t xbt = XBT_NULL;
> > +    unsigned int ring_ref;
> > +
> > +    vtpmdev->xendev.fe = xenstore_read_be_str(&vtpmdev->xendev,
> "frontend");
> > +    if (vtpmdev->xendev.fe == NULL) {
> > +        return -1;
> > +    }
> > +
> > +    /* Get backend domid */
> > +    if (xenstore_read_fe_int(&vtpmdev->xendev, "backend-id",
> > +                             &vtpmdev->bedomid)) {
> > +        return -1;
> > +    }
> > +
> > +    /*alloc share page*/
> > +    vtpmdev->shr = xc_gntshr_share_pages(vtpmdev->xen_xcs,
> vtpmdev->bedomid, 1,
> > +                                         &ring_ref,
> PROT_READ|PROT_WRITE);
> > +    vtpmdev->ring_ref = ring_ref;
> > +    if (vtpmdev->shr == NULL) {
> > +        return -1;
> > +    }
> > +
> > +    /*Create event channel */
> > +    if (xen_be_alloc_unbound(&vtpmdev->xendev, 0,
> vtpmdev->bedomid)) {
> > +        xc_gntshr_munmap(vtpmdev->xen_xcs, vtpmdev->shr, 1);
> > +        return -1;
> > +    }
> > +
> > +    xc_evtchn_unmask(vtpmdev->xendev.evtchndev,
> > +                     vtpmdev->xendev.local_port);
> > +
> > +again:
> > +    xbt = xs_transaction_start(xenstore);
> > +    if (xbt == XBT_NULL) {
> > +        goto abort_transaction;
> > +    }
> > +
> > +    if (xenstore_write_int(vtpmdev->xendev.fe, "ring-ref",
> > +                           vtpmdev->ring_ref)) {
> > +        goto abort_transaction;
> > +    }
> > +
> > +    if (xenstore_write_int(vtpmdev->xendev.fe, "event-channel",
> > +                           vtpmdev->xendev.local_port)) {
> > +        goto abort_transaction;
> > +    }
> > +
> > +    /* Publish protocol v2 feature */
> > +    if (xenstore_write_int(vtpmdev->xendev.fe, "feature-protocol-v2", 1))
> {
> > +        goto abort_transaction;
> > +    }
> > +
> > +    if (!xs_transaction_end(xenstore, xbt, 0)) {
> > +        if (errno == EAGAIN) {
> > +            goto again;
> > +        }
> > +    }
> > +    /* Tell vtpm backend that we are ready */
> > +    xenbus_switch_state(&vtpmdev->xendev, XenbusStateInitialised);
> > +
> > +    return 0;
> > +
> > +abort_transaction:
> > +    xc_gntshr_munmap(vtpmdev->xen_xcs, vtpmdev->shr, 1);
> > +    xs_transaction_end(xenstore, xbt, 1);
> > +    return -1;
> > +}
> > +
> > +static void vtpm_backend_changed(struct XenDevice *xendev, const char
> *node)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +    int be_state;
> > +
> > +    if (strcmp(node, "state") == 0) {
> > +        xenstore_read_be_int(&vtpmdev->xendev, node, &be_state);
> > +        switch (be_state) {
> > +        case XenbusStateConnected:
> > +            /*TODO*/
> > +            break;
> > +        case XenbusStateClosing:
> > +        case XenbusStateClosed:
> > +            xenbus_switch_state(&vtpmdev->xendev,
> XenbusStateClosing);
> > +            break;
> > +        default:
> > +            break;
> > +        }
> > +    }
> > +}
> > +
> > +static int vtpm_free(struct XenDevice *xendev)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +    QEMUBH *qbh = vtpmdev->sr_bh;
> > +
> > +    aio_poll(qbh->ctx, false);
> > +    qemu_bh_delete(vtpmdev->sr_bh);
> > +    if (vtpmdev->shr) {
> > +        xc_gntshr_munmap(vtpmdev->xen_xcs, vtpmdev->shr, 1);
> > +    }
> > +    xc_interface_close(vtpmdev->xen_xcs);
> > +    return 0;
> > +}
> > +
> > +static void vtpm_alloc(struct XenDevice *xendev)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +
> > +    vtpm_aio_ctx = aio_context_new(NULL);
> > +    if (vtpm_aio_ctx == NULL) {
> > +        return;
> > +    }
> > +    vtpmdev->sr_bh = aio_bh_new(vtpm_aio_ctx, sr_bh_handler,
> vtpmdev);
> > +    qemu_bh_schedule(vtpmdev->sr_bh);
> > +    vtpmdev->xen_xcs = xen_xc_gntshr_open(0, 0);
> > +}
> > +
> > +static void vtpm_event(struct XenDevice *xendev)
> > +{
> > +    struct XenVtpmDev *vtpmdev = container_of(xendev, struct
> XenVtpmDev,
> > +                                              xendev);
> > +
> > +    qemu_bh_schedule(vtpmdev->sr_bh);
> > +}
> > +
> > +struct XenDevOps xen_vtpmdev_ops = {
> > +    .size             = sizeof(struct XenVtpmDev),
> > +    .flags            = DEVOPS_FLAG_IGNORE_STATE |
> > +                        DEVOPS_FLAG_STUBDOM_BE,
> > +    .event            = vtpm_event,
> > +    .free             = vtpm_free,
> > +    .alloc            = vtpm_alloc,
> > +    .initialise       = vtpm_initialise,
> > +    .backend_changed  = vtpm_backend_changed,
> > +    .recv             = vtpm_recv,
> > +    .send             = vtpm_send,
> > +};
> 
> Is this the frontend, like the subject line would seem to imply?
> If so, XenDevOps are made for backends, while this is a frontend. In
> fact this is the first PV frontend in QEMU. We need to introduce
> something generic and similar to struct XenDevOps and xen_backend.c but
> for frontends.
> 
> 
> > diff --git a/include/hw/xen/xen_backend.h
> b/include/hw/xen/xen_backend.h
> > index 3b4125e..45fd6d3 100644
> > --- a/include/hw/xen/xen_backend.h
> > +++ b/include/hw/xen/xen_backend.h
> > @@ -15,6 +15,8 @@ struct XenDevice;
> >  #define DEVOPS_FLAG_NEED_GNTDEV   1
> >  /* don't expect frontend doing correct state transitions (aka console
> quirk) */
> >  #define DEVOPS_FLAG_IGNORE_STATE  2
> > +/*dev backend is stubdom*/
> > +#define DEVOPS_FLAG_STUBDOM_BE    4
> >
> >  struct XenDevOps {
> >      size_t    size;
> > @@ -26,6 +28,8 @@ struct XenDevOps {
> >      void      (*event)(struct XenDevice *xendev);
> >      void      (*disconnect)(struct XenDevice *xendev);
> >      int       (*free)(struct XenDevice *xendev);
> > +    int       (*send)(struct XenDevice *xendev, uint8_t* buf, size_t
> count);
> > +    int       (*recv)(struct XenDevice *xendev, uint8_t* buf, size_t
> *count);
> >      void      (*backend_changed)(struct XenDevice *xendev, const
> char *node);
> >      void      (*frontend_changed)(struct XenDevice *xendev, const
> char *node);
> >  };
> > @@ -91,12 +95,19 @@ int xen_be_send_notify(struct XenDevice
> *xendev);
> >  void xen_be_printf(struct XenDevice *xendev, int msg_level, const char
> *fmt, ...)
> >      GCC_FMT_ATTR(3, 4);
> >
> > +/*Xen stubdom vtpm*/
> > +int xen_vtpm_register(struct XenDevOps *ops);
> > +int xen_be_alloc_unbound(struct XenDevice *xendev, int dom, int
> remote_dom);
> > +int xen_vtpm_send(unsigned char *buf, size_t count);
> > +int xen_vtpm_recv(unsigned char *buf, size_t *count);
> > +
> >  /* actual backend drivers */
> >  extern struct XenDevOps xen_console_ops;      /* xen_console.c
> */
> >  extern struct XenDevOps xen_kbdmouse_ops;     /* xen_framebuffer.c
> */
> >  extern struct XenDevOps xen_framebuffer_ops;  /* xen_framebuffer.c */
> >  extern struct XenDevOps xen_blkdev_ops;       /* xen_disk.c
> */
> >  extern struct XenDevOps xen_netdev_ops;       /* xen_nic.c
> */
> > +extern struct XenDevOps xen_vtpmdev_ops;      /*
> xen_stubdom_vtpm.c*/
> >
> >  void xen_init_display(int domid);
> >
> > diff --git a/include/hw/xen/xen_common.h
> b/include/hw/xen/xen_common.h
> > index 95612a4..fb43084 100644
> > --- a/include/hw/xen/xen_common.h
> > +++ b/include/hw/xen/xen_common.h
> > @@ -130,6 +130,12 @@ static inline XenXC xen_xc_interface_open(void
> *logger, void *dombuild_logger,
> >      return xc_interface_open(logger, dombuild_logger, open_flags);
> >  }
> >
> > +static inline xc_gntshr *xen_xc_gntshr_open(void *logger,
> > +                                           unsigned int open_flags)
> > +{
> > +    return xc_gntshr_open(logger, open_flags);
> > +}
> > +
> >  /* FIXME There is now way to have the xen fd */
> >  static inline int xc_fd(xc_interface *xen_xc)
> >  {
> > diff --git a/xen-hvm.c b/xen-hvm.c
> > index 21f1cbb..c99ace8 100644
> > --- a/xen-hvm.c
> > +++ b/xen-hvm.c
> > @@ -1067,6 +1067,11 @@ int xen_hvm_init(ram_addr_t
> *below_4g_mem_size, ram_addr_t *above_4g_mem_size,
> >      int i, rc;
> >      unsigned long ioreq_pfn;
> >      unsigned long bufioreq_evtchn;
> > +
> > +#ifdef CONFIG_TPM_XENSTUBDOMS
> > +    unsigned long stubdom_vtpm = 0;
> > +#endif
> > +
> >      XenIOState *state;
> >
> >      state = g_malloc0(sizeof (XenIOState));
> > @@ -1169,6 +1174,14 @@ int xen_hvm_init(ram_addr_t
> *below_4g_mem_size, ram_addr_t *above_4g_mem_size,
> >          fprintf(stderr, "%s: xen backend core setup failed\n",
> __FUNCTION__);
> >          return -1;
> >      }
> > +
> > +#ifdef CONFIG_TPM_XENSTUBDOMS
> > +    xc_get_hvm_param(xen_xc, xen_domid,
> HVM_PARAM_STUBDOM_VTPM, &stubdom_vtpm);
> > +    if (stubdom_vtpm) {
> > +        xen_vtpm_register(&xen_vtpmdev_ops);
> > +    }
> > +#endif
> 
> Given that vtpm is just a PV frontend, can't you just detect whether is
> present on xenstore and initialize it based on that? Like all the
> backend below?

Refer to above explanation. 

Thanks 
Quan 

> 
> 
> >      xen_be_register("console", &xen_console_ops);
> >      xen_be_register("vkbd", &xen_kbdmouse_ops);
> >      xen_be_register("qdisk", &xen_blkdev_ops);
> > --
> > 1.8.3.2
> >

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