Re: [PATCH 2/4] xen/arm: do not use is_running to decide whether we can write directly to the LR registers

[Stefano Stabellini <stefano.stabellini@eu.citrix.com>]

On Fri, 15 Feb 2013, Stefano Stabellini wrote:
> During context switch is_running is set for the next vcpu before the
> gic state is actually saved.
> This leads to possible nasty races when interrupts need to be injected
> after is_running is set to the next vcpu but before the currently
> running gic state has been saved from the previous vcpu.
> Introduce a new gic_running internal variable to precisely determine
> which one is the vcpu currently using the gic.
> 
> Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>

Although the description of the problem is accurate, the fix doesn't
take SMP into consideration. Probably gic_running needs to be a per_cpu
variable.
In any case, I'll rework and resent the patch.


>  xen/arch/arm/gic.c |    8 +++++++-
>  1 files changed, 7 insertions(+), 1 deletions(-)
> 
> diff --git a/xen/arch/arm/gic.c b/xen/arch/arm/gic.c
> index 0ecc0f1..88f2d3a 100644
> --- a/xen/arch/arm/gic.c
> +++ b/xen/arch/arm/gic.c
> @@ -53,6 +53,7 @@ static irq_desc_t irq_desc[NR_IRQS];
>  static DEFINE_PER_CPU(irq_desc_t[NR_LOCAL_IRQS], local_irq_desc);
>  static DEFINE_PER_CPU(uint64_t, lr_mask);
>  static gic_callback_fn_t gic_callbacks[NR_IRQS];
> +static struct vcpu *gic_running;
>  
>  unsigned nr_lrs;
>  
> @@ -70,6 +71,7 @@ void gic_save_state(struct vcpu *v)
>      for ( i=0; i<nr_lrs; i++)
>          v->arch.gic_lr[i] = GICH[GICH_LR + i];
>      v->arch.lr_mask = this_cpu(lr_mask);
> +    gic_running = NULL;
>      spin_unlock_irq(&gic.lock);
>      /* Disable until next VCPU scheduled */
>      GICH[GICH_HCR] = 0;
> @@ -81,12 +83,16 @@ void gic_restore_state(struct vcpu *v)
>      int i;
>  
>      if ( is_idle_vcpu(v) )
> +    {
> +        gic_running = v;
>          return;
> +    }
>  
>      spin_lock_irq(&gic.lock);
>      this_cpu(lr_mask) = v->arch.lr_mask;
>      for ( i=0; i<nr_lrs; i++)
>          GICH[GICH_LR + i] = v->arch.gic_lr[i];
> +    gic_running = v;
>      spin_unlock_irq(&gic.lock);
>      GICH[GICH_HCR] = GICH_HCR_EN;
>      isb();
> @@ -481,7 +487,7 @@ void gic_set_guest_irq(struct vcpu *v, unsigned int virtual_irq,
>  
>      spin_lock_irqsave(&gic.lock, flags);
>  
> -    if ( v->is_running && list_empty(&v->arch.vgic.lr_pending) )
> +    if ( v == gic_running && list_empty(&v->arch.vgic.lr_pending) )
>      {
>          i = find_first_zero_bit(&this_cpu(lr_mask), nr_lrs);
>          if (i < nr_lrs) {
> -- 
> 1.7.2.5
>