[v19,060/130] KVM: x86/tdp_mmu: Apply mmu notifier callback to only shared GPA

Commit Message

Isaku Yamahata Feb. 26, 2024, 8:26 a.m. UTC

From: Isaku Yamahata <[e-mail]>

The private GPAs that typically guest memfd backs aren't subject to MMU
notifier because it isn't mapped into virtual address of user process.
kvm_tdp_mmu_handle_gfn() handles the callback of the MMU notifier,
clear_flush_young(), clear_young(), test_young()() and change_pte().  Make
kvm_tdp_mmu_handle_gfn() aware of private mapping and skip private mapping.

Even with AS_UNMOVABLE set, those mmu notifier are called.  For example,
ksmd triggers change_pte().

Signed-off-by: Isaku Yamahata <[e-mail]>
Reviewed-by: Binbin Wu <[e-mail]>
---
v19:
- type: test_gfn() => test_young()

v18:
- newly added

Signed-off-by: Isaku Yamahata <[e-mail]>
---
 arch/x86/kvm/mmu/tdp_mmu.c | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

Comments

Binbin Wu March 28, 2024, 8:29 a.m. UTC | #1

On 2/26/2024 4:26 PM, [e-mail] wrote:
> From: Isaku Yamahata <[e-mail]>
>
> The private GPAs that typically guest memfd backs aren't subject to MMU
> notifier because it isn't mapped into virtual address of user process.
> kvm_tdp_mmu_handle_gfn() handles the callback of the MMU notifier,
> clear_flush_young(), clear_young(), test_young()() and change_pte().  Make
                                                    ^
                                                    an extra "()"
> kvm_tdp_mmu_handle_gfn() aware of private mapping and skip private mapping.
>
> Even with AS_UNMOVABLE set, those mmu notifier are called.  For example,
> ksmd triggers change_pte().

The description about the "AS_UNMOVABLE", you are refering to shared 
memory, right?
Then, it seems not related to the change of this patch.

>
> Signed-off-by: Isaku Yamahata <[e-mail]>
> Reviewed-by: Binbin Wu <[e-mail]>
> ---
> v19:
> - type: test_gfn() => test_young()
>
> v18:
> - newly added
>
> Signed-off-by: Isaku Yamahata <[e-mail]>
> ---
>   arch/x86/kvm/mmu/tdp_mmu.c | 22 +++++++++++++++++++++-
>   1 file changed, 21 insertions(+), 1 deletion(-)
>
> diff --git a/arch/x86/kvm/mmu/tdp_mmu.c b/arch/x86/kvm/mmu/tdp_mmu.c
> index e7514a807134..10507920f36b 100644
> --- a/arch/x86/kvm/mmu/tdp_mmu.c
> +++ b/arch/x86/kvm/mmu/tdp_mmu.c
> @@ -1157,9 +1157,29 @@ static __always_inline bool kvm_tdp_mmu_handle_gfn(struct kvm *kvm,
>   	 * into this helper allow blocking; it'd be dead, wasteful code.
>   	 */
>   	for_each_tdp_mmu_root(kvm, root, range->slot->as_id) {
> +		gfn_t start, end;
> +
> +		/*
> +		 * This function is called on behalf of mmu_notifier of
> +		 * clear_flush_young(), clear_young(), test_young()(), and

^
                                                                an extra 
"()""

> +		 * change_pte().  They apply to only shared GPAs.
> +		 */
> +		WARN_ON_ONCE(range->only_private);
> +		WARN_ON_ONCE(!range->only_shared);
> +		if (is_private_sp(root))
> +			continue;
> +
> +		/*
> +		 * For TDX shared mapping, set GFN shared bit to the range,
> +		 * so the handler() doesn't need to set it, to avoid duplicated
> +		 * code in multiple handler()s.
> +		 */
> +		start = kvm_gfn_to_shared(kvm, range->start);
> +		end = kvm_gfn_to_shared(kvm, range->end);
> +
>   		rcu_read_lock();
>   
> -		tdp_root_for_each_leaf_pte(iter, root, range->start, range->end)
> +		tdp_root_for_each_leaf_pte(iter, root, start, end)
>   			ret |= handler(kvm, &iter, range);
>   
>   		rcu_read_unlock();

Isaku Yamahata March 28, 2024, 8:55 p.m. UTC | #2

On Thu, Mar 28, 2024 at 04:29:50PM +0800,
Binbin Wu <[e-mail]> wrote:

> 
> 
> On 2/26/2024 4:26 PM, [e-mail] wrote:
> > From: Isaku Yamahata <[e-mail]>
> > 
> > The private GPAs that typically guest memfd backs aren't subject to MMU
> > notifier because it isn't mapped into virtual address of user process.
> > kvm_tdp_mmu_handle_gfn() handles the callback of the MMU notifier,
> > clear_flush_young(), clear_young(), test_young()() and change_pte().  Make
>                                                    ^
>                                                    an extra "()"

Will fix it. Thanks.

> > kvm_tdp_mmu_handle_gfn() aware of private mapping and skip private mapping.
> > 
> > Even with AS_UNMOVABLE set, those mmu notifier are called.  For example,
> > ksmd triggers change_pte().
> 
> The description about the "AS_UNMOVABLE", you are refering to shared memory,
> right?
> Then, it seems not related to the change of this patch.

Ok, will remove this sentence.

Patch

diff --git a/arch/x86/kvm/mmu/tdp_mmu.c b/arch/x86/kvm/mmu/tdp_mmu.c
index e7514a807134..10507920f36b 100644
--- a/arch/x86/kvm/mmu/tdp_mmu.c
+++ b/arch/x86/kvm/mmu/tdp_mmu.c
@@ -1157,9 +1157,29 @@  static __always_inline bool kvm_tdp_mmu_handle_gfn(struct kvm *kvm,
 	 * into this helper allow blocking; it'd be dead, wasteful code.
 	 */
 	for_each_tdp_mmu_root(kvm, root, range->slot->as_id) {
+		gfn_t start, end;
+
+		/*
+		 * This function is called on behalf of mmu_notifier of
+		 * clear_flush_young(), clear_young(), test_young()(), and
+		 * change_pte().  They apply to only shared GPAs.
+		 */
+		WARN_ON_ONCE(range->only_private);
+		WARN_ON_ONCE(!range->only_shared);
+		if (is_private_sp(root))
+			continue;
+
+		/*
+		 * For TDX shared mapping, set GFN shared bit to the range,
+		 * so the handler() doesn't need to set it, to avoid duplicated
+		 * code in multiple handler()s.
+		 */
+		start = kvm_gfn_to_shared(kvm, range->start);
+		end = kvm_gfn_to_shared(kvm, range->end);
+
 		rcu_read_lock();
 
-		tdp_root_for_each_leaf_pte(iter, root, range->start, range->end)
+		tdp_root_for_each_leaf_pte(iter, root, start, end)
 			ret |= handler(kvm, &iter, range);
 
 		rcu_read_unlock();