public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] kvm: x86: lapic: remove one redundant judging condition
@ 2014-11-05  9:03 Tiejun Chen
  2014-11-05 10:22 ` Paolo Bonzini
  0 siblings, 1 reply; 3+ messages in thread
From: Tiejun Chen @ 2014-11-05  9:03 UTC (permalink / raw)
  To: pbonzini; +Cc: kvm

Finally we always return highest_irr so its unnecessary to return -1
after check if highest_irr == -1.

Signed-off-by: Tiejun Chen <tiejun.chen@intel.com>
---
 arch/x86/kvm/lapic.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 5f574b4..e6a7eb6 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1638,8 +1638,7 @@ int kvm_apic_has_interrupt(struct kvm_vcpu *vcpu)
 
 	apic_update_ppr(apic);
 	highest_irr = apic_find_highest_irr(apic);
-	if ((highest_irr == -1) ||
-	    ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI)))
+	if ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI))
 		return -1;
 	return highest_irr;
 }
-- 
1.9.1


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] kvm: x86: lapic: remove one redundant judging condition
  2014-11-05  9:03 [PATCH] kvm: x86: lapic: remove one redundant judging condition Tiejun Chen
@ 2014-11-05 10:22 ` Paolo Bonzini
  2014-11-06  1:29   ` Chen, Tiejun
  0 siblings, 1 reply; 3+ messages in thread
From: Paolo Bonzini @ 2014-11-05 10:22 UTC (permalink / raw)
  To: Tiejun Chen; +Cc: kvm

On 05/11/2014 10:03, Tiejun Chen wrote:
> Finally we always return highest_irr so its unnecessary to return -1
> after check if highest_irr == -1.
> 
> Signed-off-by: Tiejun Chen <tiejun.chen@intel.com>
> ---
>  arch/x86/kvm/lapic.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index 5f574b4..e6a7eb6 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -1638,8 +1638,7 @@ int kvm_apic_has_interrupt(struct kvm_vcpu *vcpu)
>  
>  	apic_update_ppr(apic);
>  	highest_irr = apic_find_highest_irr(apic);
> -	if ((highest_irr == -1) ||
> -	    ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI)))
> +	if ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI))
>  		return -1;
>  	return highest_irr;
>  }

I think the code is clearer without this change.

The two returns mean:

- return -1: no interrupt to inject

- return highest_irr: inject this interrupt

With IRR equal to all zeroes (highest_irr = -1), your patch would make 
the "if" always false ("current PPR is low, can inject the interrupt"), 
but computing highest_irr & 0xF0 would make no sense if highest_irr == 
-1.

To put it another way, imagine the code looked like this:

static inline int int_prio(int vector)
{
	WARN_ON(vector == -1);
	return vector & 0xF0;
}
...

 	apic_update_ppr(apic);
 	highest_irr = apic_find_highest_irr(apic);
	if (highest_irr == -1 ||
	    int_prio(highest_irr) <= kvm_apic_get_reg(apic, APIC_PROCPRI))
 		return -1;
 	return highest_irr;

Then removing the check on highest_irr == -1 would trigger a warning.

Paolo

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] kvm: x86: lapic: remove one redundant judging condition
  2014-11-05 10:22 ` Paolo Bonzini
@ 2014-11-06  1:29   ` Chen, Tiejun
  0 siblings, 0 replies; 3+ messages in thread
From: Chen, Tiejun @ 2014-11-06  1:29 UTC (permalink / raw)
  To: Paolo Bonzini; +Cc: kvm

On 2014/11/5 18:22, Paolo Bonzini wrote:
> On 05/11/2014 10:03, Tiejun Chen wrote:
>> Finally we always return highest_irr so its unnecessary to return -1
>> after check if highest_irr == -1.
>>
>> Signed-off-by: Tiejun Chen <tiejun.chen@intel.com>
>> ---
>>   arch/x86/kvm/lapic.c | 3 +--
>>   1 file changed, 1 insertion(+), 2 deletions(-)
>>
>> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
>> index 5f574b4..e6a7eb6 100644
>> --- a/arch/x86/kvm/lapic.c
>> +++ b/arch/x86/kvm/lapic.c
>> @@ -1638,8 +1638,7 @@ int kvm_apic_has_interrupt(struct kvm_vcpu *vcpu)
>>
>>   	apic_update_ppr(apic);
>>   	highest_irr = apic_find_highest_irr(apic);
>> -	if ((highest_irr == -1) ||
>> -	    ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI)))
>> +	if ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI))
>>   		return -1;
>>   	return highest_irr;
>>   }
>
> I think the code is clearer without this change.
>
> The two returns mean:
>
> - return -1: no interrupt to inject
>
> - return highest_irr: inject this interrupt
>
> With IRR equal to all zeroes (highest_irr = -1), your patch would make
> the "if" always false ("current PPR is low, can inject the interrupt"),
> but computing highest_irr & 0xF0 would make no sense if highest_irr ==
> -1.

Yeah, you're right so here is just a little confusion to read.

Actually what this code is doing looks like,

@@ -1638,7 +1638,7 @@ int kvm_apic_has_interrupt(struct kvm_vcpu *vcpu)

         apic_update_ppr(apic);
         highest_irr = apic_find_highest_irr(apic);
-       if ((highest_irr == -1) ||
+       if ((highest_irr != -1) &&
             ((highest_irr & 0xF0) <= kvm_apic_get_reg(apic, APIC_PROCPRI)))
                 return -1;
         return highest_irr;

But it's really no big deal so we can keep the original alive.

Thanks
Tiejun

>
> To put it another way, imagine the code looked like this:
>
> static inline int int_prio(int vector)
> {
> 	WARN_ON(vector == -1);
> 	return vector & 0xF0;
> }
> ...
>
>   	apic_update_ppr(apic);
>   	highest_irr = apic_find_highest_irr(apic);
> 	if (highest_irr == -1 ||
> 	    int_prio(highest_irr) <= kvm_apic_get_reg(apic, APIC_PROCPRI))
>   		return -1;
>   	return highest_irr;
>
> Then removing the check on highest_irr == -1 would trigger a warning.
>
> Paolo
>

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2014-11-06  1:29 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-11-05  9:03 [PATCH] kvm: x86: lapic: remove one redundant judging condition Tiejun Chen
2014-11-05 10:22 ` Paolo Bonzini
2014-11-06  1:29   ` Chen, Tiejun

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox