[PATCH] Fix unsynchronized access to sev members through svm_register_enc_region

[PATCH] Fix unsynchronized access to sev members through svm_register_enc_region

[Peter Gonda]

sev_pin_memory assumes that callers hold the kvm->lock. This was true for
all callers except svm_register_enc_region since it does not originate
from svm_mem_enc_op. Also added lockdep annotation to help prevent
future regressions.

Tested: Booted SEV enabled VM on host.

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Cc: Sean Christopherson <seanjc@google.com>
Cc: x86@kernel.org
Cc: kvm@vger.kernel.org
Cc: stable@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Fixes: 116a2214c5173 (KVM: SVM: Pin guest memory when SEV is active)
Signed-off-by: Peter Gonda <pgonda@google.com>

---
 arch/x86/kvm/svm.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index afdc5b44fe9f..9884e57f3d0f 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -1699,6 +1699,8 @@ static struct page **sev_pin_memory(struct kvm *kvm, unsigned long uaddr,
 	struct page **pages;
 	unsigned long first, last;
 
+	lockdep_assert_held(&kvm->lock);
+
 	if (ulen == 0 || uaddr + ulen < uaddr)
 		return NULL;
 
@@ -7228,12 +7230,19 @@ static int svm_register_enc_region(struct kvm *kvm,
 	if (!region)
 		return -ENOMEM;
 
+	mutex_lock(&kvm->lock);
 	region->pages = sev_pin_memory(kvm, range->addr, range->size, &region->npages, 1);
 	if (!region->pages) {
 		ret = -ENOMEM;
 		goto e_free;
 	}
 
+	region->uaddr = range->addr;
+	region->size = range->size;
+
+	list_add_tail(&region->list, &sev->regions_list);
+	mutex_unlock(&kvm->lock);
+
 	/*
 	 * The guest may change the memory encryption attribute from C=0 -> C=1
 	 * or vice versa for this memory range. Lets make sure caches are
@@ -7242,13 +7251,6 @@ static int svm_register_enc_region(struct kvm *kvm,
 	 */
 	sev_clflush_pages(region->pages, region->npages);
 
-	region->uaddr = range->addr;
-	region->size = range->size;
-
-	mutex_lock(&kvm->lock);
-	list_add_tail(&region->list, &sev->regions_list);
-	mutex_unlock(&kvm->lock);
-
 	return ret;
 
 e_free:
-- 
2.30.0.280.ga3ce27912f-goog

Re: [PATCH] Fix unsynchronized access to sev members through svm_register_enc_region

[Sean Christopherson]

On Tue, Jan 26, 2021, Peter Gonda wrote:
> sev_pin_memory assumes that callers hold the kvm->lock. This was true for
> all callers except svm_register_enc_region since it does not originate

This doesn't actually state what change it being made, is only describes the
problem. I'd also reword these sentences to avoid talking about assumptions, and
instead use stronger language.
 
> from svm_mem_enc_op. Also added lockdep annotation to help prevent

s/Also added/Add, i.e. describe what the patch is doing, not what you did in the
past.

E.g.

  Grab kvm->lock before pinning memory when registering an encrypted
  region; sev_pin_memory() relies on kvm->lock being held to ensure
  correctness when checking and updating the number of pinned pages.

  Add a lockdep assertion to help prevent future regressions.

> future regressions.
> 
> Tested: Booted SEV enabled VM on host.

Personally I'd just leave this out.  Unless stated otherwise, it's implied that
you've tested the patch.

> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@redhat.com>
> Cc: "H. Peter Anvin" <hpa@zytor.com>
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Joerg Roedel <joro@8bytes.org>
> Cc: Tom Lendacky <thomas.lendacky@amd.com>
> Cc: Brijesh Singh <brijesh.singh@amd.com>
> Cc: Sean Christopherson <seanjc@google.com>
> Cc: x86@kernel.org
> Cc: kvm@vger.kernel.org
> Cc: stable@vger.kernel.org
> Cc: linux-kernel@vger.kernel.org
> Fixes: 116a2214c5173 (KVM: SVM: Pin guest memory when SEV is active)
> Signed-off-by: Peter Gonda <pgonda@google.com>
> 
> ---
>  arch/x86/kvm/svm.c | 16 +++++++++-------
>  1 file changed, 9 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
> index afdc5b44fe9f..9884e57f3d0f 100644
> --- a/arch/x86/kvm/svm.c
> +++ b/arch/x86/kvm/svm.c
> @@ -1699,6 +1699,8 @@ static struct page **sev_pin_memory(struct kvm *kvm, unsigned long uaddr,
>  	struct page **pages;
>  	unsigned long first, last;
>  
> +	lockdep_assert_held(&kvm->lock);
> +
>  	if (ulen == 0 || uaddr + ulen < uaddr)
>  		return NULL;
>  
> @@ -7228,12 +7230,19 @@ static int svm_register_enc_region(struct kvm *kvm,
>  	if (!region)
>  		return -ENOMEM;
>  
> +	mutex_lock(&kvm->lock);
>  	region->pages = sev_pin_memory(kvm, range->addr, range->size, &region->npages, 1);
>  	if (!region->pages) {
>  		ret = -ENOMEM;
>  		goto e_free;

This error path needs to do mutex_unlock().

>  	}
>  
> +	region->uaddr = range->addr;
> +	region->size = range->size;
> +
> +	list_add_tail(&region->list, &sev->regions_list);
> +	mutex_unlock(&kvm->lock);
> +
>  	/*
>  	 * The guest may change the memory encryption attribute from C=0 -> C=1
>  	 * or vice versa for this memory range. Lets make sure caches are
> @@ -7242,13 +7251,6 @@ static int svm_register_enc_region(struct kvm *kvm,
>  	 */
>  	sev_clflush_pages(region->pages, region->npages);
>  
> -	region->uaddr = range->addr;
> -	region->size = range->size;
> -
> -	mutex_lock(&kvm->lock);
> -	list_add_tail(&region->list, &sev->regions_list);
> -	mutex_unlock(&kvm->lock);
> -
>  	return ret;
>  
>  e_free:
> -- 
> 2.30.0.280.ga3ce27912f-goog

Re: [PATCH] Fix unsynchronized access to sev members through svm_register_enc_region

[Tom Lendacky]

On 1/26/21 12:54 PM, Peter Gonda wrote:
> sev_pin_memory assumes that callers hold the kvm->lock. This was true for
> all callers except svm_register_enc_region since it does not originate
> from svm_mem_enc_op. Also added lockdep annotation to help prevent
> future regressions.

I'm not exactly sure what the problem is that your fixing? What is the
symptom that you're seeing?

> 
> Tested: Booted SEV enabled VM on host.
> 
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@redhat.com>
> Cc: "H. Peter Anvin" <hpa@zytor.com>
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Joerg Roedel <joro@8bytes.org>
> Cc: Tom Lendacky <thomas.lendacky@amd.com>
> Cc: Brijesh Singh <brijesh.singh@amd.com>
> Cc: Sean Christopherson <seanjc@google.com>
> Cc: x86@kernel.org
> Cc: kvm@vger.kernel.org
> Cc: stable@vger.kernel.org
> Cc: linux-kernel@vger.kernel.org
> Fixes: 116a2214c5173 (KVM: SVM: Pin guest memory when SEV is active)

I can't find this commit. The Linux and KVM trees have it as:

1e80fdc09d12 ("KVM: SVM: Pin guest memory when SEV is active")

> Signed-off-by: Peter Gonda <pgonda@google.com>
> 
> ---
>  arch/x86/kvm/svm.c | 16 +++++++++-------

This patch won't apply, as it has already been a few releases since svm.c
was moved to the arch/x86/kvm/svm directory and this function now lives in
sev.c.

Thanks,
Tom

>  1 file changed, 9 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
> index afdc5b44fe9f..9884e57f3d0f 100644
> --- a/arch/x86/kvm/svm.c
> +++ b/arch/x86/kvm/svm.c
> @@ -1699,6 +1699,8 @@ static struct page **sev_pin_memory(struct kvm *kvm, unsigned long uaddr,
>  	struct page **pages;
>  	unsigned long first, last;
>  
> +	lockdep_assert_held(&kvm->lock);
> +
>  	if (ulen == 0 || uaddr + ulen < uaddr)
>  		return NULL;
>  
> @@ -7228,12 +7230,19 @@ static int svm_register_enc_region(struct kvm *kvm,
>  	if (!region)
>  		return -ENOMEM;
>  
> +	mutex_lock(&kvm->lock);
>  	region->pages = sev_pin_memory(kvm, range->addr, range->size, &region->npages, 1);
>  	if (!region->pages) {
>  		ret = -ENOMEM;
>  		goto e_free;
>  	}
>  
> +	region->uaddr = range->addr;
> +	region->size = range->size;
> +
> +	list_add_tail(&region->list, &sev->regions_list);
> +	mutex_unlock(&kvm->lock);
> +
>  	/*
>  	 * The guest may change the memory encryption attribute from C=0 -> C=1
>  	 * or vice versa for this memory range. Lets make sure caches are
> @@ -7242,13 +7251,6 @@ static int svm_register_enc_region(struct kvm *kvm,
>  	 */
>  	sev_clflush_pages(region->pages, region->npages);
>  
> -	region->uaddr = range->addr;
> -	region->size = range->size;
> -
> -	mutex_lock(&kvm->lock);
> -	list_add_tail(&region->list, &sev->regions_list);
> -	mutex_unlock(&kvm->lock);
> -
>  	return ret;
>  
>  e_free:
> 

Re: [PATCH] Fix unsynchronized access to sev members through svm_register_enc_region

[Sean Christopherson]

On Tue, Jan 26, 2021, Tom Lendacky wrote:
> On 1/26/21 12:54 PM, Peter Gonda wrote:
> > sev_pin_memory assumes that callers hold the kvm->lock. This was true for
> > all callers except svm_register_enc_region since it does not originate
> > from svm_mem_enc_op. Also added lockdep annotation to help prevent
> > future regressions.
> 
> I'm not exactly sure what the problem is that your fixing? What is the
> symptom that you're seeing?

svm_register_enc_region() calls sev_pin_memory() without holding kvm->lock.  If
userspace does multiple KVM_MEMORY_ENCRYPT_REG_REGION in parallel, it could
circumvent the rlimit(RLIMIT_MEMLOCK) check.