[PATCH 1/2] Xen: Use native_pte_flags instead of native_pte_val for .pte_flags

[PATCH 1/2] Xen: Use native_pte_flags instead of native_pte_val for .pte_flags

[Eduardo Habkost]

Using native_pte_val triggers the BUG_ON() in the paravirt_ops
version of pte_flags().

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
---
 arch/x86/xen/enlighten.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 3fc0c15..6b169b7 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -1378,7 +1378,7 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
 	.ptep_modify_prot_commit = __ptep_modify_prot_commit,
 
 	.pte_val = xen_pte_val,
-	.pte_flags = native_pte_val,
+	.pte_flags = native_pte_flags,
 	.pgd_val = xen_pgd_val,
 
 	.make_pte = xen_make_pte,
-- 
1.5.5.GIT

[PATCH 2/2] xen-netfront: call netif_start_queue() on xennet_open()

[Eduardo Habkost]

xen-netfront was waking the queue before calling netif_start_queue(),
triggering the BUG_ON() on __netif_schedule().

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
---
 drivers/net/xen-netfront.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
index 902bbe7..c749bdb 100644
--- a/drivers/net/xen-netfront.c
+++ b/drivers/net/xen-netfront.c
@@ -329,7 +329,7 @@ static int xennet_open(struct net_device *dev)
 	}
 	spin_unlock_bh(&np->rx_lock);
 
-	xennet_maybe_wake_tx(dev);
+	netif_start_queue(dev);
 
 	return 0;
 }
-- 
1.5.5.GIT

Re: [PATCH 2/2] xen-netfront: call netif_start_queue() on xennet_open()

[Jeremy Fitzhardinge]

Eduardo Habkost wrote:
> xen-netfront was waking the queue before calling netif_start_queue(),
> triggering the BUG_ON() on __netif_schedule().
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>   
Ok,
Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>

but Ingo, do you want to pick this up?  You're gathering a collection of 
these patches anyway, or should we kick it over to Dave/Jeff?

    J

> ---
>  drivers/net/xen-netfront.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
> index 902bbe7..c749bdb 100644
> --- a/drivers/net/xen-netfront.c
> +++ b/drivers/net/xen-netfront.c
> @@ -329,7 +329,7 @@ static int xennet_open(struct net_device *dev)
>  	}
>  	spin_unlock_bh(&np->rx_lock);
>  
> -	xennet_maybe_wake_tx(dev);
> +	netif_start_queue(dev);
>  
>  	return 0;
>  }
>   

Re: [PATCH 2/2] xen-netfront: call netif_start_queue() on xennet_open()

[Jeff Garzik]

Jeremy Fitzhardinge wrote:
> Eduardo Habkost wrote:
>> xen-netfront was waking the queue before calling netif_start_queue(),
>> triggering the BUG_ON() on __netif_schedule().
>>
>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>>   
> Ok,
> Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
> 
> but Ingo, do you want to pick this up?  You're gathering a collection of 
> these patches anyway, or should we kick it over to Dave/Jeff?
> 
>    J
> 
>> ---
>>  drivers/net/xen-netfront.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
>> index 902bbe7..c749bdb 100644
>> --- a/drivers/net/xen-netfront.c
>> +++ b/drivers/net/xen-netfront.c
>> @@ -329,7 +329,7 @@ static int xennet_open(struct net_device *dev)
>>      }
>>      spin_unlock_bh(&np->rx_lock);
>>  
>> -    xennet_maybe_wake_tx(dev);
>> +    netif_start_queue(dev);

This is clearly a net driver change, it should go this way...

	Jeff

Re: [PATCH 2/2] xen-netfront: call netif_start_queue() on xennet_open()

[Jeremy Fitzhardinge]

Jeff Garzik wrote:
> Jeremy Fitzhardinge wrote:
>> Eduardo Habkost wrote:
>>> xen-netfront was waking the queue before calling netif_start_queue(),
>>> triggering the BUG_ON() on __netif_schedule().
>>>
>>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>>>   
>> Ok,
>> Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
>>
>> but Ingo, do you want to pick this up?  You're gathering a collection 
>> of these patches anyway, or should we kick it over to Dave/Jeff?
>>
>>    J
>>
>>> ---
>>>  drivers/net/xen-netfront.c |    2 +-
>>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
>>> index 902bbe7..c749bdb 100644
>>> --- a/drivers/net/xen-netfront.c
>>> +++ b/drivers/net/xen-netfront.c
>>> @@ -329,7 +329,7 @@ static int xennet_open(struct net_device *dev)
>>>      }
>>>      spin_unlock_bh(&np->rx_lock);
>>>  
>>> -    xennet_maybe_wake_tx(dev);
>>> +    netif_start_queue(dev);
>
> This is clearly a net driver change, it should go this way... 

OK, all your's.

    J

Re: [PATCH 2/2] xen-netfront: call netif_start_queue() on xennet_open()

[Eduardo Habkost]

On Thu, Jul 24, 2008 at 11:39:25AM -0700, Jeremy Fitzhardinge wrote:
> Jeff Garzik wrote:
>> Jeremy Fitzhardinge wrote:
>>> Eduardo Habkost wrote:
>>>> xen-netfront was waking the queue before calling netif_start_queue(),
>>>> triggering the BUG_ON() on __netif_schedule().
>>>>
>>>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>>>>   
>>> Ok,
>>> Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
>>>
>>> but Ingo, do you want to pick this up?  You're gathering a collection of 
>>> these patches anyway, or should we kick it over to Dave/Jeff?
>>>
>>>    J
>>>
>>>> ---
>>>>  drivers/net/xen-netfront.c |    2 +-
>>>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>>>
>>>> diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
>>>> index 902bbe7..c749bdb 100644
>>>> --- a/drivers/net/xen-netfront.c
>>>> +++ b/drivers/net/xen-netfront.c
>>>> @@ -329,7 +329,7 @@ static int xennet_open(struct net_device *dev)
>>>>      }
>>>>      spin_unlock_bh(&np->rx_lock);
>>>>  -    xennet_maybe_wake_tx(dev);
>>>> +    netif_start_queue(dev);
>>
>> This is clearly a net driver change, it should go this way... 
>
> OK, all your's.

I didn't find it on any git repository. Is it queued somewhere? If not,
should I resend?

-- 
Eduardo

Re: [PATCH 2/2] xen-netfront: call netif_start_queue() on xennet_open()

[Jeremy Fitzhardinge]

Eduardo Habkost wrote:
> I didn't find it on any git repository. Is it queued somewhere? If not,
> should I resend?
>   

Yes, I think it fell through the cracks.  Resend to Jeff.

    J

Re: [PATCH 1/2] Xen: Use native_pte_flags instead of native_pte_val for .pte_flags

[Jeremy Fitzhardinge]

Eduardo Habkost wrote:
> Using native_pte_val triggers the BUG_ON() in the paravirt_ops
> version of pte_flags().
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>   

Good point.

Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>

    J
> ---
>  arch/x86/xen/enlighten.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
> index 3fc0c15..6b169b7 100644
> --- a/arch/x86/xen/enlighten.c
> +++ b/arch/x86/xen/enlighten.c
> @@ -1378,7 +1378,7 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
>  	.ptep_modify_prot_commit = __ptep_modify_prot_commit,
>  
>  	.pte_val = xen_pte_val,
> -	.pte_flags = native_pte_val,
> +	.pte_flags = native_pte_flags,
>  	.pgd_val = xen_pgd_val,
>  
>  	.make_pte = xen_make_pte,
>   

Re: [PATCH 1/2] Xen: Use native_pte_flags instead of native_pte_val for .pte_flags

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Eduardo Habkost wrote:
>> Using native_pte_val triggers the BUG_ON() in the paravirt_ops
>> version of pte_flags().
>>
>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>>   
>
> Good point.
>
> Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>

applied to tip/x86/xen - thanks.

	Ingo