merge kvm_flush_remote_tlbs when modifing shadow pte

merge kvm_flush_remote_tlbs when modifing shadow pte

[Dong, Eddie]

[-- Attachment #1: Type: text/plain, Size: 685 bytes --]


host side TLB flush can be merged together if multiple
spte need to be write-protected.

Signed-off-by: Yaozu (Eddie) Dong <eddie.dong-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
index 92ac0d1..749db46 100644
--- a/drivers/kvm/mmu.c
+++ b/drivers/kvm/mmu.c
@@ -524,9 +524,9 @@ static void rmap_write_protect(struct kvm *kvm, u64
gfn)
 		rmap_printk("rmap_write_protect: spte %p %llx\n", spte,
*spte);
 		if (is_writeble_pte(*spte))
 			set_shadow_pte(spte, *spte & ~PT_WRITABLE_MASK);
-		kvm_flush_remote_tlbs(kvm);
 		spte = rmap_next(kvm, rmapp, spte);
 	}
+	kvm_flush_remote_tlbs(kvm);
 }
 
 #ifdef MMU_DEBUG

[-- Attachment #2: tlb-clean1.patch --]
[-- Type: application/octet-stream, Size: 632 bytes --]


host side TLB flush can be merged together if multiple
spte need to be write-protected.

Signed-off-by: Yaozu (Eddie) Dong <eddie.dong@intel.com>

diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
index 92ac0d1..749db46 100644
--- a/drivers/kvm/mmu.c
+++ b/drivers/kvm/mmu.c
@@ -524,9 +524,9 @@ static void rmap_write_protect(struct kvm *kvm, u64 gfn)
 		rmap_printk("rmap_write_protect: spte %p %llx\n", spte, *spte);
 		if (is_writeble_pte(*spte))
 			set_shadow_pte(spte, *spte & ~PT_WRITABLE_MASK);
-		kvm_flush_remote_tlbs(kvm);
 		spte = rmap_next(kvm, rmapp, spte);
 	}
+	kvm_flush_remote_tlbs(kvm);
 }
 
 #ifdef MMU_DEBUG

[-- Attachment #3: Type: text/plain, Size: 308 bytes --]

-------------------------------------------------------------------------
SF.Net email is sponsored by:
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services
for just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace

[-- Attachment #4: Type: text/plain, Size: 186 bytes --]

_______________________________________________
kvm-devel mailing list
kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/kvm-devel

Re: merge kvm_flush_remote_tlbs when modifing shadow pte

[Avi Kivity]

Dong, Eddie wrote:
> host side TLB flush can be merged together if multiple
> spte need to be write-protected.
>
> Signed-off-by: Yaozu (Eddie) Dong <eddie.dong-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>
> diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
> index 92ac0d1..749db46 100644
> --- a/drivers/kvm/mmu.c
> +++ b/drivers/kvm/mmu.c
> @@ -524,9 +524,9 @@ static void rmap_write_protect(struct kvm *kvm, u64
> gfn)
>  		rmap_printk("rmap_write_protect: spte %p %llx\n", spte,
> *spte);
>  		if (is_writeble_pte(*spte))
>  			set_shadow_pte(spte, *spte & ~PT_WRITABLE_MASK);
> -		kvm_flush_remote_tlbs(kvm);
>  		spte = rmap_next(kvm, rmapp, spte);
>  	}
> +	kvm_flush_remote_tlbs(kvm);
>  }
>   


I think this can be improved; in case we don't touch the shadow pte at 
all, there is no need to flush the tlbs.

Also, please pull from kvm.git; things have moved around (mmu.c is now 
in arch/x86/kvm/).


-- 
Any sufficiently difficult bug is indistinguishable from a feature.


-------------------------------------------------------------------------
SF.Net email is sponsored by:
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services
for just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace

Re: merge kvm_flush_remote_tlbs when modifing shadow pte

[Dong, Eddie]

[-- Attachment #1: Type: text/plain, Size: 1004 bytes --]

>Dong, Eddie wrote:
>> host side TLB flush can be merged together if multiple
>> spte need to be write-protected.
>>
>> Signed-off-by: Yaozu (Eddie) Dong <eddie.dong-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>>
>> diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
>> index 92ac0d1..749db46 100644
>> --- a/drivers/kvm/mmu.c
>> +++ b/drivers/kvm/mmu.c
>> @@ -524,9 +524,9 @@ static void rmap_write_protect(struct 
>kvm *kvm, u64
>> gfn)
>>  		rmap_printk("rmap_write_protect: spte %p %llx\n", spte,
>> *spte);
>>  		if (is_writeble_pte(*spte))
>>  			set_shadow_pte(spte, *spte & ~PT_WRITABLE_MASK);
>> -		kvm_flush_remote_tlbs(kvm);
>>  		spte = rmap_next(kvm, rmapp, spte);
>>  	}
>> +	kvm_flush_remote_tlbs(kvm);
>>  }
>>   
>
>
>I think this can be improved; in case we don't touch the shadow pte at 
>all, there is no need to flush the tlbs.
>
>Also, please pull from kvm.git; things have moved around (mmu.c is now 
>in arch/x86/kvm/).
>
>
Sure, rebased.
thx,eddie

[-- Attachment #2: tlb-clean1.patch2 --]
[-- Type: application/octet-stream, Size: 1131 bytes --]

commit 799271f149506598750e312ae83e719618ea247e
Author: root <root@vt32-pae.(none)>
Date:   Tue Dec 18 06:08:27 2007 +0800

    host side TLB flush can be merged together if multiple
    spte need to be write-protected.
    
    Signed-off-by: Yaozu (Eddie) Dong <eddie.dong@intel.com>

diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 401eb7c..9a57e1a 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -513,6 +513,7 @@ static void rmap_write_protect(struct kvm *kvm, u64 gfn)
 {
 	unsigned long *rmapp;
 	u64 *spte;
+	int write_protected = 0;
 
 	gfn = unalias_gfn(kvm, gfn);
 	rmapp = gfn_to_rmap(kvm, gfn);
@@ -522,11 +523,14 @@ static void rmap_write_protect(struct kvm *kvm, u64 gfn)
 		BUG_ON(!spte);
 		BUG_ON(!(*spte & PT_PRESENT_MASK));
 		rmap_printk("rmap_write_protect: spte %p %llx\n", spte, *spte);
-		if (is_writeble_pte(*spte))
+		if (is_writeble_pte(*spte)) {
 			set_shadow_pte(spte, *spte & ~PT_WRITABLE_MASK);
-		kvm_flush_remote_tlbs(kvm);
+			write_protected = 1;
+		}
 		spte = rmap_next(kvm, rmapp, spte);
 	}
+	if (write_protected)
+		kvm_flush_remote_tlbs(kvm);
 }
 
 #ifdef MMU_DEBUG

[-- Attachment #3: Type: text/plain, Size: 308 bytes --]

-------------------------------------------------------------------------
SF.Net email is sponsored by:
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services
for just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace

[-- Attachment #4: Type: text/plain, Size: 186 bytes --]

_______________________________________________
kvm-devel mailing list
kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/kvm-devel

Re: merge kvm_flush_remote_tlbs when modifing shadow pte

[Avi Kivity]

Dong, Eddie wrote:
>> Dong, Eddie wrote:
>>     
>>> host side TLB flush can be merged together if multiple
>>> spte need to be write-protected.
>>>
>>> Signed-off-by: Yaozu (Eddie) Dong <eddie.dong-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>>>
>>> diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
>>> index 92ac0d1..749db46 100644
>>> --- a/drivers/kvm/mmu.c
>>> +++ b/drivers/kvm/mmu.c
>>> @@ -524,9 +524,9 @@ static void rmap_write_protect(struct 
>>>       
>> kvm *kvm, u64
>>     
>>> gfn)
>>>  		rmap_printk("rmap_write_protect: spte %p %llx\n", spte,
>>> *spte);
>>>  		if (is_writeble_pte(*spte))
>>>  			set_shadow_pte(spte, *spte & ~PT_WRITABLE_MASK);
>>> -		kvm_flush_remote_tlbs(kvm);
>>>  		spte = rmap_next(kvm, rmapp, spte);
>>>  	}
>>> +	kvm_flush_remote_tlbs(kvm);
>>>  }
>>>   
>>>       
>> I think this can be improved; in case we don't touch the shadow pte at 
>> all, there is no need to flush the tlbs.
>>
>> Also, please pull from kvm.git; things have moved around (mmu.c is now 
>> in arch/x86/kvm/).
>>
>>
>>     
> Sure, rebased.
> thx,eddie
>   

Applied, thanks.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
SF.Net email is sponsored by:
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services
for just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace