[PATCH 0/1] KVM: s390: One small fix for post 4.1

[PATCH 0/1] KVM: s390: One small fix for post 4.1

[Christian Borntraeger]

Paolo,

here is a small fixup for KVM on s390. It is also necessary for
4.1 which I am a bit late for - so cc stable.
No pull request as it is only one patch.


Jens Freimann (1):
  KVM: s390: clear floating interrupt bitmap and parameters

 arch/s390/kvm/interrupt.c | 3 +++
 1 file changed, 3 insertions(+)

-- 
2.3.0

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

[PATCH 1/1] KVM: s390: clear floating interrupt bitmap and parameters

[Christian Borntraeger]

From: Jens Freimann <jfrei@linux.vnet.ibm.com>

commit 6d3da24141 ("KVM: s390: deliver floating interrupts in order
of priority") introduced a regression for the reset handling.

We don't clear the bitmap of pending floating interrupts
and interrupt parameters. This could result in stale interrupts
even after a reset. Let's fix this by clearing the pending bitmap
and the parameters for service and machine check interrupts.

Cc: stable@vger.kernel.org # 4.1
Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/kvm/interrupt.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 322ef9c..480b975 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -1621,6 +1621,9 @@ void kvm_s390_clear_float_irqs(struct kvm *kvm)
 	int i;
 
 	spin_lock(&fi->lock);
+	fi->pending_irqs = 0;
+	memset(&fi->srv_signal, 0, sizeof(fi->srv_signal));
+	memset(&fi->mchk, 0, sizeof(fi->mchk));
 	for (i = 0; i < FIRQ_LIST_COUNT; i++)
 		clear_irq_list(&fi->lists[i]);
 	for (i = 0; i < FIRQ_MAX_COUNT; i++)
-- 
2.3.0

Re: [PATCH 1/1] KVM: s390: clear floating interrupt bitmap and parameters

[Paolo Bonzini]

On 22/06/2015 13:20, Christian Borntraeger wrote:
> From: Jens Freimann <jfrei@linux.vnet.ibm.com>
> 
> commit 6d3da24141 ("KVM: s390: deliver floating interrupts in order
> of priority") introduced a regression for the reset handling.
> 
> We don't clear the bitmap of pending floating interrupts
> and interrupt parameters. This could result in stale interrupts
> even after a reset. Let's fix this by clearing the pending bitmap
> and the parameters for service and machine check interrupts.
> 
> Cc: stable@vger.kernel.org # 4.1
> Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>
> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
> ---
>  arch/s390/kvm/interrupt.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
> index 322ef9c..480b975 100644
> --- a/arch/s390/kvm/interrupt.c
> +++ b/arch/s390/kvm/interrupt.c
> @@ -1621,6 +1621,9 @@ void kvm_s390_clear_float_irqs(struct kvm *kvm)
>  	int i;
>  
>  	spin_lock(&fi->lock);
> +	fi->pending_irqs = 0;
> +	memset(&fi->srv_signal, 0, sizeof(fi->srv_signal));
> +	memset(&fi->mchk, 0, sizeof(fi->mchk));
>  	for (i = 0; i < FIRQ_LIST_COUNT; i++)
>  		clear_irq_list(&fi->lists[i]);
>  	for (i = 0; i < FIRQ_MAX_COUNT; i++)
> 

Applied, thanks.

Paolo