how to read value from xenstore inside kernel

how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 222 bytes --]

Can someone please point me to a code( or help me to find some code)where a
value corresponding to a key is read from Xenstore inside the kernel . My
requirement is to read a xenstore value from netback.c .

Thanks
Mathew

[-- Attachment #1.2: Type: text/html, Size: 240 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: how to read value from xenstore inside kernel

[Konrad Rzeszutek Wilk]

On Mon, Aug 17, 2009 at 12:10:21PM +0530, Kuriakose Mathew wrote:
> Can someone please point me to a code( or help me to find some code)where a
> value corresponding to a key is read from Xenstore inside the kernel . My
> requirement is to read a xenstore value from netback.c .

In drivers/xen/blkback/xenbus.c, the function backend_changed exclusivy job is
to read those values. Here is an example:

278         err = xenbus_scanf(XBT_NIL, dev->nodename, "physical-device", "%x:%x",
279                            &major, &minor);

I would recommend you get a copy of "Understanding Xen" to understand how
XenStore "transfers" the data from one domain to another.

> 
> Thanks
> Mathew

> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 1368 bytes --]

Thanks for the response.
But I want the value to be read in netback.c . I checked the
drivers/xen/netback/xenbus.c
and found some functions using xenbus_scanf . But couldn't find where these
functions are getting called.
Whether it will be possible to directly call xenbus_scanf directly in
netback.c ? I want a value in xenstore to be read in some function in
netback.c.

Please help.
-Mathew


On Mon, Aug 17, 2009 at 8:08 PM, Konrad Rzeszutek Wilk <
konrad.wilk@oracle.com> wrote:

> On Mon, Aug 17, 2009 at 12:10:21PM +0530, Kuriakose Mathew wrote:
> > Can someone please point me to a code( or help me to find some code)where
> a
> > value corresponding to a key is read from Xenstore inside the kernel . My
> > requirement is to read a xenstore value from netback.c .
>
> In drivers/xen/blkback/xenbus.c, the function backend_changed exclusivy job
> is
> to read those values. Here is an example:
>
> 278         err = xenbus_scanf(XBT_NIL, dev->nodename, "physical-device",
> "%x:%x",
> 279                            &major, &minor);
>
> I would recommend you get a copy of "Understanding Xen" to understand how
> XenStore "transfers" the data from one domain to another.
>
> >
> > Thanks
> > Mathew
>
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@lists.xensource.com
> > http://lists.xensource.com/xen-devel
>
>

[-- Attachment #1.2: Type: text/html, Size: 1968 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: how to read value from xenstore inside kernel

[Konrad Rzeszutek Wilk]

On Mon, Aug 17, 2009 at 11:25:36PM +0530, Kuriakose Mathew wrote:
> Thanks for the response.
> But I want the value to be read in netback.c . I checked the
> drivers/xen/netback/xenbus.c
> and found some functions using xenbus_scanf . But couldn't find where these
> functions are getting called.

I recommend you get acquinted with 'cscope' and run under your linux directory this
command: 'make cscope' to create the cscope files. Then look for xenbus_scanf.

> Whether it will be possible to directly call xenbus_scanf directly in
> netback.c ? I want a value in xenstore to be read in some function in

Yes, if you have all of the #include files set right.

> netback.c.

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 1287 bytes --]

Hi ,

I have been trying to use to xenbus_read.

I added the below code in netback.c , static void net_tx_action(unsigned
long unused) function.

char * test;
const char *nodename = "console";
test = xenbus_read(XBT_NIL, nodename, "limit", NULL);

It is compiling fine. But when the kernel is crashing at this point during
runtime.
I expect to read the /local/domain/0/console/limit , which I am able to read
from shell using
xenstore-read . Am I missing something ?

Thanks
Mathew




On Mon, Aug 17, 2009 at 11:50 PM, Konrad Rzeszutek Wilk <
konrad.wilk@oracle.com> wrote:

> On Mon, Aug 17, 2009 at 11:25:36PM +0530, Kuriakose Mathew wrote:
> > Thanks for the response.
> > But I want the value to be read in netback.c . I checked the
> > drivers/xen/netback/xenbus.c
> > and found some functions using xenbus_scanf . But couldn't find where
> these
> > functions are getting called.
>
> I recommend you get acquinted with 'cscope' and run under your linux
> directory this
> command: 'make cscope' to create the cscope files. Then look for
> xenbus_scanf.
>
> > Whether it will be possible to directly call xenbus_scanf directly in
> > netback.c ? I want a value in xenstore to be read in some function in
>
> Yes, if you have all of the #include files set right.
>
> > netback.c.
>

[-- Attachment #1.2: Type: text/html, Size: 1764 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

RE: how to read value from xenstore inside kernel

[James Harper]

> Hi ,
> 
> I have been trying to use to xenbus_read.
> 
> I added the below code in netback.c , static void
net_tx_action(unsigned long
> unused) function.
> 
> char * test;
> const char *nodename = "console";
> test = xenbus_read(XBT_NIL, nodename, "limit", NULL);
> 
> It is compiling fine. But when the kernel is crashing at this point
during
> runtime.
> I expect to read the /local/domain/0/console/limit , which I am able
to read
> from shell using
> xenstore-read . Am I missing something ?

Not sure how much this helps, but you should not be re-reading from
xenbus every time a packet is sent. That would involve huge per-packet
overheads. You can read it initially and then set a 'watch' on the value
to get notified every time it changes.

Have you got a copy of the crash message?

James

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 1383 bytes --]

Thanks James.
I will try to get the crash message but not how to get it in a file. When I
restart the system after making the change, it gets rebooted at this point.
So I have to find a way to get those message displayed while a system is
booting. I ll try to get it.

Also regarding 'watch' , it is called when the value of key changes but if I
want to read some value which is static then will this help?
Not sure if I am right.

-Mathew


On Tue, Aug 25, 2009 at 4:37 PM, James Harper <james.harper@bendigoit.com.au
> wrote:

> > Hi ,
> >
> > I have been trying to use to xenbus_read.
> >
> > I added the below code in netback.c , static void
> net_tx_action(unsigned long
> > unused) function.
> >
> > char * test;
> > const char *nodename = "console";
> > test = xenbus_read(XBT_NIL, nodename, "limit", NULL);
> >
> > It is compiling fine. But when the kernel is crashing at this point
> during
> > runtime.
> > I expect to read the /local/domain/0/console/limit , which I am able
> to read
> > from shell using
> > xenstore-read . Am I missing something ?
>
> Not sure how much this helps, but you should not be re-reading from
> xenbus every time a packet is sent. That would involve huge per-packet
> overheads. You can read it initially and then set a 'watch' on the value
> to get notified every time it changes.
>
> Have you got a copy of the crash message?
>
> James
>
>

[-- Attachment #1.2: Type: text/html, Size: 2012 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

RE: how to read value from xenstore inside kernel

[James Harper]

> 
> Thanks James.
> I will try to get the crash message but not how to get it in a file.
When I
> restart the system after making the change, it gets rebooted at this
point. So
> I have to find a way to get those message displayed while a system is
booting.
> I ll try to get it.
> 
> Also regarding 'watch' , it is called when the value of key changes
but if I
> want to read some value which is static then will this help?
> Not sure if I am right.
> 

If it's static then just read it when the other values are read.

James

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 3183 bytes --]

When I am using
  char * test;
  const char *nodename = "vif/2/0";
  test = xenbus_read(XBT_NIL, nodename, "mac", NULL);

in netback .c ,I am getting this error when the kernel is rebooted.

guest1 BUG: scheduling while atomic: xenwatch/0x00000100/13

and for a different nodename ( console , mac )

crash msg is as follows

In netback.c::net_tx_action called
BUG: scheduling while atomic: syslogd/0x00000100/4532
 [<c02f73c6>] schedule+0x5c6/0x990
 [<c0138654>] autoremove_wake_function+0x14/0x40
 [<c011aeb9>] __wake_up_common+0x39/0x60
 [<c011b468>] __wake_up+0x38/0x50
 [<c02f84f3>] __mutex_lock_slowpath+0x53/0x8d
 [<c02f853c>] .text.lock.mutex+0xf/0x23
 [<c0248c19>] xs_talkv+0x49/0x180
 [<c0248e96>] xs_single+0x46/0x60
 [<c024913a>] xenbus_read+0x3a/0x60
 [<c025410e>] net_tx_action+0x2e/0xe90
 [<c015470e>] __generic_file_aio_write_nolock+0x34e/0x510
 [<c0154a3b>] __generic_file_write_nolock+0x8b/0xb0
 [<c0138640>] autoremove_wake_function+0x0/0x40
 [<c0188354>] do_path_lookup+0xc4/0x2e0
 [<c0186e9a>] getname+0xaa/0xe0
 [<c02f8398>] mutex_lock+0x8/0x20
 [<c017746b>] do_readv_writev+0x1fb/0x300
 [<c012905a>] tasklet_action+0x7a/0x130
 [<c0128a24>] __do_softirq+0x94/0x130
 [<c0128b4b>] do_softirq+0x8b/0xa0
 [<c0107444>] do_IRQ+0x44/0xa0
 [<c0243620>] evtchn_do_upcall+0xc0/0x190
 [<c010591d>] hypervisor_callback+0x3d/0x45
..
..
BUG: scheduling while atomic: syslogd/0x00000100/4532
 [<c02f73c6>] schedule+0x5c6/0x990
 [<c024861a>] xb_write+0x14a/0x1e0
 [<c0248ab5>] read_reply+0x75/0x100
 [<c0138640>] autoremove_wake_function+0x0/0x40
 [<c0248c71>] xs_talkv+0xa1/0x180
 [<c0248e96>] xs_single+0x46/0x60
 [<c024913a>] xenbus_read+0x3a/0x60
 [<c025410e>] net_tx_action+0x2e/0xe90
 [<c015470e>] __generic_file_aio_write_nolock+0x34e/0x510
 [<c0154a3b>] __generic_file_write_nolock+0x8b/0xb0
 [<c0138640>] autoremove_wake_function+0x0/0x40
 [<c0188354>] do_path_lookup+0xc4/0x2e0
 [<c0186e9a>] getname+0xaa/0xe0
 [<c02f8398>] mutex_lock+0x8/0x20
 [<c017746b>] do_readv_writev+0x1fb/0x300
 [<c012905a>] tasklet_action+0x7a/0x130
 [<c0128a24>] __do_softirq+0x94/0x130
 [<c0128b4b>] do_softirq+0x8b/0xa0
 [<c0107444>] do_IRQ+0x44/0xa0
 [<c0243620>] evtchn_do_upcall+0xc0/0x190
 [<c010591d>] hypervisor_callback+0x3d/0x45

-Mathew


On Tue, Aug 25, 2009 at 4:37 PM, James Harper <james.harper@bendigoit.com.au
> wrote:

> > Hi ,
> >
> > I have been trying to use to xenbus_read.
> >
> > I added the below code in netback.c , static void
> net_tx_action(unsigned long
> > unused) function.
> >
> > char * test;
> > const char *nodename = "console";
> > test = xenbus_read(XBT_NIL, nodename, "limit", NULL);
> >
> > It is compiling fine. But when the kernel is crashing at this point
> during
> > runtime.
> > I expect to read the /local/domain/0/console/limit , which I am able
> to read
> > from shell using
> > xenstore-read . Am I missing something ?
>
> Not sure how much this helps, but you should not be re-reading from
> xenbus every time a packet is sent. That would involve huge per-packet
> overheads. You can read it initially and then set a 'watch' on the value
> to get notified every time it changes.
>
> Have you got a copy of the crash message?
>
> James
>
>

[-- Attachment #1.2: Type: text/html, Size: 4173 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

RE: how to read value from xenstore inside kernel

[James Harper]

> When I am using
>   char * test;
>   const char *nodename = "vif/2/0";
>   test = xenbus_read(XBT_NIL, nodename, "mac", NULL);
> 
> in netback .c ,I am getting this error when the kernel is rebooted.
> 
> guest1 BUG: scheduling while atomic: xenwatch/0x00000100/13
> 

That will be because the xenbus_read involves a wait of some sort, and
you can't do that in that sort of function.

Basically the rx and tx and other hardware event handlers need to
execute as fast as possible, which precludes any waiting for xenbus
responses to arrive.

Read your value at startup, or put a watch in at startup and read it
when it changes.

James

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 1475 bytes --]

In netback.c , I intialized a target_watch struct and a callback function
target_watch.
Now for registering it , I used the below function in " static int __init
netback_init(void)  "

spin_lock_init(&net_schedule_list_lock);
INIT_LIST_HEAD(&net_schedule_list);

netif_xenbus_init();
//ADDED CODE
err = register_xenbus_watch(&target_watch);
    if(IS_ERR(err)) {
              printk(KERN_NOTICE "In netback.c::init Failed to initialize
watcher\n");
        } else {
              printk(KERN_NOTICE " In netback.c::init xenbus watcher
initialized\n");
      }
///END OF ADD

When I reboot , it is hung at the point above the
'register_xenbus_watch(&target_watch)' .
Neither of the printk messages are dispayed.



On Fri, Aug 28, 2009 at 2:37 PM, James Harper <james.harper@bendigoit.com.au
> wrote:

> > When I am using
> >   char * test;
> >   const char *nodename = "vif/2/0";
> >   test = xenbus_read(XBT_NIL, nodename, "mac", NULL);
> >
> > in netback .c ,I am getting this error when the kernel is rebooted.
> >
> > guest1 BUG: scheduling while atomic: xenwatch/0x00000100/13
> >
>
> That will be because the xenbus_read involves a wait of some sort, and
> you can't do that in that sort of function.
>
> Basically the rx and tx and other hardware event handlers need to
> execute as fast as possible, which precludes any waiting for xenbus
> responses to arrive.
>
> Read your value at startup, or put a watch in at startup and read it
> when it changes.
>
> James
>

[-- Attachment #1.2: Type: text/html, Size: 2029 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

RE: how to read value from xenstore inside kernel

[James Harper]

> When I reboot , it is hung at the point above the
> 'register_xenbus_watch(&target_watch)' .
> Neither of the printk messages are dispayed.
> 

Please show the code where you initialize target_watch

James

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 732 bytes --]

The .node  member I initialized while declaring  and .callback I initialized
in netbacl init function (just above where I use register_xenbus_watch )as
shown below. (used balloon.c as guideline)


static struct xenbus_watch target_watch =
{
    .node = "memory/target"

};

static int __init netback_init(void)
{
......
target_watch.callback = watch_target;
... register_xenbus_watch(&target_watch);
...
}

Thanks
Mathew

On Fri, Aug 28, 2009 at 5:14 PM, James Harper <james.harper@bendigoit.com.au
> wrote:

> > When I reboot , it is hung at the point above the
> > 'register_xenbus_watch(&target_watch)' .
> > Neither of the printk messages are dispayed.
> >
>
> Please show the code where you initialize target_watch
>
> James
>

[-- Attachment #1.2: Type: text/html, Size: 1176 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

RE: how to read value from xenstore inside kernel

[James Harper]

> 
> The .node  member I initialized while declaring  and .callback I
initialized
> in netbacl init function (just above where I use register_xenbus_watch
)as
> shown below. (used balloon.c as guideline)
> 

Try adding the watch (or just a read as a test) in netback/xenbus.c
inside the connect() function where a lot of other xenbus activity is
happening

James

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 4406 bytes --]

I tried watch in xenbus.c ( connect function)

code is as shown below.

//////////ADDED CODE////////////////////


static struct xenbus_watch target_watch =
{
    .node = "memory/target"

};

////////////////////////////////////////////////

/* React to a change in the target key */
static void mywatch_target(struct xenbus_watch *watch,
             const char **vec, unsigned int len)
{
    unsigned long long new_target;
    int err;
    printk(KERN_NOTICE "In netback.c::watch_target called \n");
    err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
    if (err != 1) {
        /* This is ok (for domain0 at least) - so just return */
        return;
    }

    /* The given memory/target value is in KiB, so it needs converting to
     * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
     */
    //balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
}


////////////////////END OF ADD ////////////////////////////


static void connect(struct backend_info *be)
{
    int err;
    struct xenbus_device *dev = be->dev;

    err = connect_rings(be);
    if (err)
        return;

    err = xen_net_read_mac(dev, be->netif->fe_dev_addr);
    if (err) {
        xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
        return;
    }

    xen_net_read_rate(dev, &be->netif->credit_bytes,
              &be->netif->credit_usec);
    be->netif->remaining_credit = be->netif->credit_bytes;

    xenbus_switch_state(dev, XenbusStateConnected);

    netif_wake_queue(be->netif->dev);

   /////////////////ADDED CODE////////////
    target_watch.callback = mywatch_target;

    printk(KERN_NOTICE "In xenbus.c::connect b4 register \n");
    err = register_xenbus_watch(&target_watch);
    if(IS_ERR(err)) {
              printk(KERN_NOTICE "In xenbus.c::connect Failed to initialize
watcher\n");
        } else {
              printk(KERN_NOTICE " In xenbus.c::connect xenbus watcher
initialized\n");
      }
    //////////////END OF ADD//////////////
}

The crash message is

In xenbus.c::connect b4 register
------------[ cut here ]------------
kernel BUG at
/home/saransh/xen-3.1.3/linux-2.6.18.8-xen/drivers/xen/xenbus/xenbus_xs.c:607!
invalid opcode: 0000 [#1]
SMP
Modules linked in: xt_tcpudp xt_physdev iptable_filter ip_tables x_tables
ppdev ac video container button sbs i2c_ec i2c_core asus_acpi battery bridge
lp serio_raw parport_pc parport serial_core rtc pcspkr shpchp pci_hotplug
ipv6 tsdev evdev ext3 jbd sg sd_mod amd74xx usbhid ehci_hcd ohci_hcd usbcore
sata_nv libata scsi_mod forcedeth thermal processor fan fuse
CPU:    1
EIP:    0061:[<c0249306>]    Not tainted VLI
EFLAGS: 00010286   (2.6.18.8-xen #1)
EIP is at register_xenbus_watch+0xc6/0xe0
eax: c0364bd4   ebx: c0364bd4   ecx: c0f2da10   edx: c0364bd4
esi: fffffffe   edi: c1511f03   ebp: ed00cb40   esp: c1511ef4
ds: 007b   es: 007b   ss: 0069
Process xenwatch (pid: 13, ti=c1510000 task=c0446620 task.ti=c1510000)
Stack: c1511f03 c032d562 c0364bd4 43fe0400 34363330 00344442 c0fe0000
fffffffe
       c0fe0400 c02564db c03218b0 ed00c540 c032db45 c030ec5b c1511f5c
c032db12
       c031610b c1511f68 c032dc1d c031ac96 c1511f64 00000000 c0fe0408
ed00c540
Call Trace:
 [<c02564db>] frontend_changed+0x4db/0x650
 [<c024a8d8>] otherend_changed+0xa8/0xb0
 [<c0249710>] xenwatch_thread+0x0/0x140
 [<c0248b5a>] xenwatch_handle_callback+0x1a/0x60
 [<c0249840>] xenwatch_thread+0x130/0x140
 [<c0248b40>] xenwatch_handle_callback+0x0/0x60
 [<c0138640>] autoremove_wake_function+0x0/0x40
 [<c0138486>] kthread+0xf6/0x100
 [<c0138390>] kthread+0x0/0x100
 [<c0103005>] kernel_thread_helper+0x5/0x10
Code: 10 00 c7 43 04 00 02 20 00 86 05 e0 3d 36 c0 b8 a8 3e 41 c0 e8 0c 2a
ef ff 89 f8 8b 5c 24 18 8b 74 24 1c 8b 7c 24 20 83 c4 24 c3 <0f> 0b 5f 02 6c
03 32 c0 e9 75 ff ff ff 8d b6 00 00 00 00 8d bc
EIP: [<c0249306>] register_xenbus_watch+0xc6/0xe0 SS:ESP 0069:c1511ef4
 <6>Bluetooth: Core ver 2.10


Thanks
Mathew


On Fri, Aug 28, 2009 at 5:27 PM, James Harper <james.harper@bendigoit.com.au
> wrote:

> >
> > The .node  member I initialized while declaring  and .callback I
> initialized
> > in netbacl init function (just above where I use register_xenbus_watch
> )as
> > shown below. (used balloon.c as guideline)
> >
>
> Try adding the watch (or just a read as a test) in netback/xenbus.c
> inside the connect() function where a lot of other xenbus activity is
> happening
>
> James
>

[-- Attachment #1.2: Type: text/html, Size: 5402 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 5069 bytes --]

Hi ,

Thanks. I think the registration is happening. I mistook the call trace to
be a crash. But now when I write a new value to xenstore memory/target , it
is showing some odd behavior like the shell is scrolling up .

I think the callback function is getting called again and again as though
some kind of infinite loop.

-Mathew

On Fri, Aug 28, 2009 at 8:09 PM, Kuriakose Mathew <kmathew123@gmail.com>wrote:

> I tried watch in xenbus.c ( connect function)
>
> code is as shown below.
>
> //////////ADDED CODE////////////////////
>
>
> static struct xenbus_watch target_watch =
> {
>     .node = "memory/target"
>
> };
>
> ////////////////////////////////////////////////
>
> /* React to a change in the target key */
> static void mywatch_target(struct xenbus_watch *watch,
>              const char **vec, unsigned int len)
> {
>     unsigned long long new_target;
>     int err;
>     printk(KERN_NOTICE "In netback.c::watch_target called \n");
>     err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
>     if (err != 1) {
>         /* This is ok (for domain0 at least) - so just return */
>         return;
>     }
>
>     /* The given memory/target value is in KiB, so it needs converting to
>      * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
>      */
>     //balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
> }
>
>
> ////////////////////END OF ADD ////////////////////////////
>
>
> static void connect(struct backend_info *be)
> {
>     int err;
>     struct xenbus_device *dev = be->dev;
>
>     err = connect_rings(be);
>     if (err)
>         return;
>
>     err = xen_net_read_mac(dev, be->netif->fe_dev_addr);
>     if (err) {
>         xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
>         return;
>     }
>
>     xen_net_read_rate(dev, &be->netif->credit_bytes,
>               &be->netif->credit_usec);
>     be->netif->remaining_credit = be->netif->credit_bytes;
>
>     xenbus_switch_state(dev, XenbusStateConnected);
>
>     netif_wake_queue(be->netif->dev);
>
>    /////////////////ADDED CODE////////////
>     target_watch.callback = mywatch_target;
>
>     printk(KERN_NOTICE "In xenbus.c::connect b4 register \n");
>     err = register_xenbus_watch(&target_watch);
>     if(IS_ERR(err)) {
>               printk(KERN_NOTICE "In xenbus.c::connect Failed to
> initialize  watcher\n");
>         } else {
>               printk(KERN_NOTICE " In xenbus.c::connect xenbus watcher
> initialized\n");
>       }
>     //////////////END OF ADD//////////////
> }
>
> The crash message is
>
> In xenbus.c::connect b4 register
> ------------[ cut here ]------------
> kernel BUG at
> /home/saransh/xen-3.1.3/linux-2.6.18.8-xen/drivers/xen/xenbus/xenbus_xs.c:607!
> invalid opcode: 0000 [#1]
> SMP
> Modules linked in: xt_tcpudp xt_physdev iptable_filter ip_tables x_tables
> ppdev ac video container button sbs i2c_ec i2c_core asus_acpi battery bridge
> lp serio_raw parport_pc parport serial_core rtc pcspkr shpchp pci_hotplug
> ipv6 tsdev evdev ext3 jbd sg sd_mod amd74xx usbhid ehci_hcd ohci_hcd usbcore
> sata_nv libata scsi_mod forcedeth thermal processor fan fuse
> CPU:    1
> EIP:    0061:[<c0249306>]    Not tainted VLI
> EFLAGS: 00010286   (2.6.18.8-xen #1)
> EIP is at register_xenbus_watch+0xc6/0xe0
> eax: c0364bd4   ebx: c0364bd4   ecx: c0f2da10   edx: c0364bd4
> esi: fffffffe   edi: c1511f03   ebp: ed00cb40   esp: c1511ef4
> ds: 007b   es: 007b   ss: 0069
> Process xenwatch (pid: 13, ti=c1510000 task=c0446620 task.ti=c1510000)
> Stack: c1511f03 c032d562 c0364bd4 43fe0400 34363330 00344442 c0fe0000
> fffffffe
>        c0fe0400 c02564db c03218b0 ed00c540 c032db45 c030ec5b c1511f5c
> c032db12
>        c031610b c1511f68 c032dc1d c031ac96 c1511f64 00000000 c0fe0408
> ed00c540
> Call Trace:
>  [<c02564db>] frontend_changed+0x4db/0x650
>  [<c024a8d8>] otherend_changed+0xa8/0xb0
>  [<c0249710>] xenwatch_thread+0x0/0x140
>  [<c0248b5a>] xenwatch_handle_callback+0x1a/0x60
>  [<c0249840>] xenwatch_thread+0x130/0x140
>  [<c0248b40>] xenwatch_handle_callback+0x0/0x60
>  [<c0138640>] autoremove_wake_function+0x0/0x40
>  [<c0138486>] kthread+0xf6/0x100
>  [<c0138390>] kthread+0x0/0x100
>  [<c0103005>] kernel_thread_helper+0x5/0x10
> Code: 10 00 c7 43 04 00 02 20 00 86 05 e0 3d 36 c0 b8 a8 3e 41 c0 e8 0c 2a
> ef ff 89 f8 8b 5c 24 18 8b 74 24 1c 8b 7c 24 20 83 c4 24 c3 <0f> 0b 5f 02 6c
> 03 32 c0 e9 75 ff ff ff 8d b6 00 00 00 00 8d bc
> EIP: [<c0249306>] register_xenbus_watch+0xc6/0xe0 SS:ESP 0069:c1511ef4
>  <6>Bluetooth: Core ver 2.10
>
>
> Thanks
> Mathew
>
>
>
> On Fri, Aug 28, 2009 at 5:27 PM, James Harper <
> james.harper@bendigoit.com.au> wrote:
>
>> >
>> > The .node  member I initialized while declaring  and .callback I
>> initialized
>> > in netbacl init function (just above where I use register_xenbus_watch
>> )as
>> > shown below. (used balloon.c as guideline)
>> >
>>
>> Try adding the watch (or just a read as a test) in netback/xenbus.c
>> inside the connect() function where a lot of other xenbus activity is
>> happening
>>
>> James
>>
>
>

[-- Attachment #1.2: Type: text/html, Size: 6274 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: how to read value from xenstore inside kernel

[Kuriakose Mathew]

[-- Attachment #1.1: Type: text/plain, Size: 5435 bytes --]

How to avoid the callback function  getting called again and again after
registration of xenbus_watch?

-Mathew



On Sat, Aug 29, 2009 at 8:31 PM, Kuriakose Mathew <kmathew123@gmail.com>wrote:

> Hi ,
>
> Thanks. I think the registration is happening. I mistook the call trace to
> be a crash. But now when I write a new value to xenstore memory/target , it
> is showing some odd behavior like the shell is scrolling up .
>
> I think the callback function is getting called again and again as though
> some kind of infinite loop.
>
> -Mathew
>
>
> On Fri, Aug 28, 2009 at 8:09 PM, Kuriakose Mathew <kmathew123@gmail.com>wrote:
>
>> I tried watch in xenbus.c ( connect function)
>>
>> code is as shown below.
>>
>> //////////ADDED CODE////////////////////
>>
>>
>> static struct xenbus_watch target_watch =
>> {
>>     .node = "memory/target"
>>
>> };
>>
>> ////////////////////////////////////////////////
>>
>> /* React to a change in the target key */
>> static void mywatch_target(struct xenbus_watch *watch,
>>              const char **vec, unsigned int len)
>> {
>>     unsigned long long new_target;
>>     int err;
>>     printk(KERN_NOTICE "In netback.c::watch_target called \n");
>>     err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
>>     if (err != 1) {
>>         /* This is ok (for domain0 at least) - so just return */
>>         return;
>>     }
>>
>>     /* The given memory/target value is in KiB, so it needs converting to
>>      * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
>>      */
>>     //balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
>> }
>>
>>
>> ////////////////////END OF ADD ////////////////////////////
>>
>>
>> static void connect(struct backend_info *be)
>> {
>>     int err;
>>     struct xenbus_device *dev = be->dev;
>>
>>     err = connect_rings(be);
>>     if (err)
>>         return;
>>
>>     err = xen_net_read_mac(dev, be->netif->fe_dev_addr);
>>     if (err) {
>>         xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
>>         return;
>>     }
>>
>>     xen_net_read_rate(dev, &be->netif->credit_bytes,
>>               &be->netif->credit_usec);
>>     be->netif->remaining_credit = be->netif->credit_bytes;
>>
>>     xenbus_switch_state(dev, XenbusStateConnected);
>>
>>     netif_wake_queue(be->netif->dev);
>>
>>    /////////////////ADDED CODE////////////
>>     target_watch.callback = mywatch_target;
>>
>>     printk(KERN_NOTICE "In xenbus.c::connect b4 register \n");
>>     err = register_xenbus_watch(&target_watch);
>>     if(IS_ERR(err)) {
>>               printk(KERN_NOTICE "In xenbus.c::connect Failed to
>> initialize  watcher\n");
>>         } else {
>>               printk(KERN_NOTICE " In xenbus.c::connect xenbus watcher
>> initialized\n");
>>       }
>>     //////////////END OF ADD//////////////
>> }
>>
>> The crash message is
>>
>> In xenbus.c::connect b4 register
>> ------------[ cut here ]------------
>> kernel BUG at
>> /home/saransh/xen-3.1.3/linux-2.6.18.8-xen/drivers/xen/xenbus/xenbus_xs.c:607!
>> invalid opcode: 0000 [#1]
>> SMP
>> Modules linked in: xt_tcpudp xt_physdev iptable_filter ip_tables x_tables
>> ppdev ac video container button sbs i2c_ec i2c_core asus_acpi battery bridge
>> lp serio_raw parport_pc parport serial_core rtc pcspkr shpchp pci_hotplug
>> ipv6 tsdev evdev ext3 jbd sg sd_mod amd74xx usbhid ehci_hcd ohci_hcd usbcore
>> sata_nv libata scsi_mod forcedeth thermal processor fan fuse
>> CPU:    1
>> EIP:    0061:[<c0249306>]    Not tainted VLI
>> EFLAGS: 00010286   (2.6.18.8-xen #1)
>> EIP is at register_xenbus_watch+0xc6/0xe0
>> eax: c0364bd4   ebx: c0364bd4   ecx: c0f2da10   edx: c0364bd4
>> esi: fffffffe   edi: c1511f03   ebp: ed00cb40   esp: c1511ef4
>> ds: 007b   es: 007b   ss: 0069
>> Process xenwatch (pid: 13, ti=c1510000 task=c0446620 task.ti=c1510000)
>> Stack: c1511f03 c032d562 c0364bd4 43fe0400 34363330 00344442 c0fe0000
>> fffffffe
>>        c0fe0400 c02564db c03218b0 ed00c540 c032db45 c030ec5b c1511f5c
>> c032db12
>>        c031610b c1511f68 c032dc1d c031ac96 c1511f64 00000000 c0fe0408
>> ed00c540
>> Call Trace:
>>  [<c02564db>] frontend_changed+0x4db/0x650
>>  [<c024a8d8>] otherend_changed+0xa8/0xb0
>>  [<c0249710>] xenwatch_thread+0x0/0x140
>>  [<c0248b5a>] xenwatch_handle_callback+0x1a/0x60
>>  [<c0249840>] xenwatch_thread+0x130/0x140
>>  [<c0248b40>] xenwatch_handle_callback+0x0/0x60
>>  [<c0138640>] autoremove_wake_function+0x0/0x40
>>   [<c0138486>] kthread+0xf6/0x100
>>  [<c0138390>] kthread+0x0/0x100
>>  [<c0103005>] kernel_thread_helper+0x5/0x10
>> Code: 10 00 c7 43 04 00 02 20 00 86 05 e0 3d 36 c0 b8 a8 3e 41 c0 e8 0c 2a
>> ef ff 89 f8 8b 5c 24 18 8b 74 24 1c 8b 7c 24 20 83 c4 24 c3 <0f> 0b 5f 02 6c
>> 03 32 c0 e9 75 ff ff ff 8d b6 00 00 00 00 8d bc
>> EIP: [<c0249306>] register_xenbus_watch+0xc6/0xe0 SS:ESP 0069:c1511ef4
>>  <6>Bluetooth: Core ver 2.10
>>
>>
>> Thanks
>> Mathew
>>
>>
>>
>> On Fri, Aug 28, 2009 at 5:27 PM, James Harper <
>> james.harper@bendigoit.com.au> wrote:
>>
>>> >
>>> > The .node  member I initialized while declaring  and .callback I
>>> initialized
>>> > in netbacl init function (just above where I use register_xenbus_watch
>>> )as
>>> > shown below. (used balloon.c as guideline)
>>> >
>>>
>>> Try adding the watch (or just a read as a test) in netback/xenbus.c
>>> inside the connect() function where a lot of other xenbus activity is
>>> happening
>>>
>>> James
>>>
>>
>>
>

[-- Attachment #1.2: Type: text/html, Size: 6854 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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