Re: [PATCH v1 02/11] KVM: x86: add kvm_apic_map_get_dest_lapic

From: Radim KrÄmÃÅ
Date: Fri Jul 01 2016 - 08:46:11 EST


2016-07-01 09:57+0200, Paolo Bonzini:
> On 30/06/2016 22:54, Radim KrÄmÃÅ wrote:
>> kvm_irq_delivery_to_apic_fast and kvm_intr_is_single_vcpu_fast both
>> compute the interrupt destination. Factor the code.
>>
>> 'struct kvm_lapic **dst = NULL' had to be added to silence GCC.
>> GCC might complain about potential NULL access in the future, because it
>> missed conditions that avoided uninitialized uses of dst.
>>
>> Signed-off-by: Radim KrÄmÃÅ <rkrcmar@xxxxxxxxxx>
>> ---
>> v1: improved comment for kvm_apic_map_get_dest_lapic() [Peter]
>>
>> arch/x86/kvm/lapic.c | 241 ++++++++++++++++++++++-----------------------------
>> 1 file changed, 103 insertions(+), 138 deletions(-)
>>
>> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
>> index 22a6474af220..238b87b068db 100644
>> --- a/arch/x86/kvm/lapic.c
>> +++ b/arch/x86/kvm/lapic.c
>> @@ -671,14 +671,98 @@ static void kvm_apic_disabled_lapic_found(struct kvm *kvm)
>> }
>> }
>>
>> +/* Return true if the interrupt can be handled by using *bitmap as index mask
>> + * for valid destinations in *dst array.
>> + * Return false if kvm_apic_map_get_dest_lapic did nothing useful.
>> + * Note: we may have zero kvm_lapic destinations when we return true, which
>> + * means that the interrupt should be dropped. In this case, *bitmap would be
>> + * zero and *dst undefined.
>> + */
>> +static inline bool kvm_apic_map_get_dest_lapic(struct kvm *kvm, struct kvm_lapic *src,
>> + struct kvm_lapic_irq *irq, struct kvm_apic_map *map,
>> + struct kvm_lapic ***dst, unsigned long *bitmap)
>> +{
>> + int i, lowest;
>> + bool x2apic_ipi;
>> + u16 cid;
>> +
>> + if (irq->shorthand == APIC_DEST_SELF) {
>> + *dst = &src;
>
> This is not valid, &src dies as soon as you leave the function. You
> need to pass &src into kvm_apic_map_get_dest_lapic and here do something
> like

Indeed, that would have been bad.

> if (irq->shorthand) {
> if (irq->shorthand == APIC_DEST_SELF && p_src) {
> *dst = p_src;
> *bitmap = 1;
> return true;
> }
> return false;
> }
>
> If it's not too hard, I'd like to have patch 4 before this one.

Sure, thanks.