Skip to contentSkip to conversation

KVM: x86: Async #PF MSR fix and cleanups

PATCH Applied Sean Christopherson 4 participants 11 messages Apr 6 – May 19

Subsystems (3)
lore
Share & tools
lore

Sean Christopherson

author maintainer
#0 lore raw

v3 cover KVM: x86: Async #PF MSR fix and cleanups

Fix a user-triggerable WARN found syzkaller, then clean up related code to eliminate potential stale data usage by removing what are effectively duplicate fields.

Ethan Yang (1):
KVM: x86: Don't leave APF half-enabled on bad APF data GPA

Sean Christopherson (2):
KVM: x86: Drop superfluous caching of
KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT
KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_SEND_ALWAYS

 arch/x86/include/asm/kvm_host.h |  2 --
 arch/x86/kvm/x86.c              | 33 +++++++++++++++++----------------
 2 files changed, 17 insertions(+), 18 deletions(-)

base-commit: b89df297a47e641581ee67793592e5c6ae0428f4

Signature · 1 line
2.53.0.1213.gd9a14994de-goog
author maintainer
#1 lore raw

PATCH v3 1/3 KVM: x86: Don't leave APF half-enabled on bad APF data GPA

From: Ethan Yang <[email protected]>

kvm_pv_enable_async_pf() updates vcpu->arch.apf.msr_en_val before initializing the APF data gfn_to_hva cache. If userspace provides an invalid GPA, kvm_gfn_to_hva_cache_init() fails, but msr_en_val stays enabled and leaves APF state half-initialized.

Later APF paths can then try to use the empty cache and trigger WARN_ON() in kvm_read_guest_offset_cached().

Determine the new APF enabled state from the incoming MSR value, do cache initialization first on the enable path, and commit msr_en_val only after successful initialization. Keep the disable path behavior unchanged.

Reported-by [email protected]
Fixes 344d9588a9df ("KVM: Add PV MSR to enable asynchronous page faults delivery.")
[sean: don't bother with a local "enable" variable]
---
 arch/x86/kvm/x86.c | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)
arch/x86/kvm/x86.c +13 -7
index 0a1b63c63d1a..c35d359b56dd 100644
@@ -1043,11 +1043,16 @@ bool kvm_require_dr(struct kvm_vcpu *vcpu, int dr)
 }
 EXPORT_SYMBOL_FOR_KVM_INTERNAL(kvm_require_dr);
 
-static bool kvm_pv_async_pf_enabled(struct kvm_vcpu *vcpu)
+static bool __kvm_pv_async_pf_enabled(u64 data)
 {
 	u64 mask = KVM_ASYNC_PF_ENABLED | KVM_ASYNC_PF_DELIVERY_AS_INT;
 
-	return (vcpu->arch.apf.msr_en_val & mask) == mask;
+	return (data & mask) == mask;
+}
+
+static bool kvm_pv_async_pf_enabled(struct kvm_vcpu *vcpu)
+{
+	return __kvm_pv_async_pf_enabled(vcpu->arch.apf.msr_en_val);
 }
 
 static inline u64 pdptr_rsvd_bits(struct kvm_vcpu *vcpu)
@@ -3647,18 +3652,19 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 	if (!lapic_in_kernel(vcpu))
 		return data ? 1 : 0;
 
+	if (__kvm_pv_async_pf_enabled(data) &&
+	    kvm_gfn_to_hva_cache_init(vcpu->kvm, &vcpu->arch.apf.data, gpa,
+				      sizeof(u64)))
+		return 1;
+
 	vcpu->arch.apf.msr_en_val = data;
 
-	if (!kvm_pv_async_pf_enabled(vcpu)) {
+	if (!__kvm_pv_async_pf_enabled(data)) {
 		kvm_clear_async_pf_completion_queue(vcpu);
 		kvm_async_pf_hash_reset(vcpu);
 		return 0;
 	}
 
-	if (kvm_gfn_to_hva_cache_init(vcpu->kvm, &vcpu->arch.apf.data, gpa,
-					sizeof(u64)))
-		return 1;
-
 	vcpu->arch.apf.send_always = (data & KVM_ASYNC_PF_SEND_ALWAYS);
 	vcpu->arch.apf.delivery_as_pf_vmexit = data & KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT;
 
Signature · 1 line
2.53.0.1213.gd9a14994de-goog
Reviewed-by on v3 1/3
#2 lore raw

Re: [PATCH v3 1/3] KVM: x86: Don't leave APF half-enabled on bad APF data GPA

Quoted 14 lines · Sean Christopherson

From: Ethan Yang <[email protected]>

kvm_pv_enable_async_pf() updates vcpu->arch.apf.msr_en_val before initializing the APF data gfn_to_hva cache. If userspace provides an invalid GPA, kvm_gfn_to_hva_cache_init() fails, but msr_en_val stays enabled and leaves APF state half-initialized.

Later APF paths can then try to use the empty cache and trigger WARN_ON() in kvm_read_guest_offset_cached().

Determine the new APF enabled state from the incoming MSR value, do cache initialization first on the enable path, and commit msr_en_val only after successful initialization. Keep the disable path behavior unchanged.

Quoted 9 lines
Reported-by [email protected]
Fixes 344d9588a9df ("KVM: Add PV MSR to enable asynchronous page faults delivery.")
[sean: don't bother with a local "enable" variable]
author maintainer
#3 lore raw

PATCH v3 2/3 KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT

In reply to Sean Christopherson

Drop kvm_vcpu_arch.delivery_as_pf_vmexit and instead use msr_en_val as the source of truth to reduce the probability of operating on stale data. This fixes flaws where KVM fails to update delivery_as_pf_vmexit when APF is explicitly disabled by the guest or implicitly disabled by KVM on INIT. Absent other bugs, the flaws are benign as KVM *shouldn't* consume delivery_as_pf_vmexit when PV APF support is disabled.

Simply delete the field, as there's zero benefit to maintaining a separate "cache" of the state.

Fixes 52a5c155cf79 ("KVM: async_pf: Let guest support delivery of async_pf from guest mode")
---
 arch/x86/include/asm/kvm_host.h | 1 -
 arch/x86/kvm/x86.c              | 3 +--
 2 files changed, 1 insertion(+), 3 deletions(-)
arch/x86/include/asm/kvm_host.h +0 -1
index c470e40a00aa..fae1f4aeca5a 100644
@@ -1039,7 +1039,6 @@ struct kvm_vcpu_arch {
 		u32 id;
 		u32 host_apf_flags;
 		bool send_always;
-		bool delivery_as_pf_vmexit;
 		bool pageready_pending;
 	} apf;
 
arch/x86/kvm/x86.c +1 -2
index c35d359b56dd..4632222a5d1c 100644
@@ -3666,7 +3666,6 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 	}
 
 	vcpu->arch.apf.send_always = (data & KVM_ASYNC_PF_SEND_ALWAYS);
-	vcpu->arch.apf.delivery_as_pf_vmexit = data & KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT;
 
 	kvm_async_pf_wakeup_all(vcpu);
 
@@ -14035,7 +14034,7 @@ static bool kvm_can_deliver_async_pf(struct kvm_vcpu *vcpu)
 		 * L1 needs to opt into the special #PF vmexits that are
 		 * used to deliver async page faults.
 		 */
-		return vcpu->arch.apf.delivery_as_pf_vmexit;
+		return vcpu->arch.apf.msr_en_val & KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT;
 	} else {
 		/*
 		 * Play it safe in case the guest temporarily disables paging.
Signature · 1 line
2.53.0.1213.gd9a14994de-goog
Reviewed-by on v3 2/3
#4 lore raw

Re: [PATCH v3 2/3] KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT

Quoted 7 lines · Sean Christopherson

Drop kvm_vcpu_arch.delivery_as_pf_vmexit and instead use msr_en_val as the source of truth to reduce the probability of operating on stale data. This fixes flaws where KVM fails to update delivery_as_pf_vmexit when APF is explicitly disabled by the guest or implicitly disabled by KVM on INIT. Absent other bugs, the flaws are benign as KVM *shouldn't* consume delivery_as_pf_vmexit when PV APF support is disabled.

Quoted 5 lines

Simply delete the field, as there's zero benefit to maintaining a separate "cache" of the state.

Fixes 52a5c155cf79 ("KVM: async_pf: Let guest support delivery of async_pf from guest mode")
Reviewed-by on v3 2/3
#5 lore raw

Re: [PATCH v3 2/3] KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT

In reply to Sean Christopherson
Quoted 1 line · Sean Christopherson

Drop kvm_vcpu_arch.delivery_as_pf_vmexit and instead use msr_en_val as the

Nit:
kvm_vcpu_arch.apf.elivery_as_pf_vmexit

Quoted 11 lines

source of truth to reduce the probability of operating on stale data. This fixes flaws where KVM fails to update delivery_as_pf_vmexit when APF is explicitly disabled by the guest or implicitly disabled by KVM on INIT. Absent other bugs, the flaws are benign as KVM *shouldn't* consume delivery_as_pf_vmexit when PV APF support is disabled.

Simply delete the field, as there's zero benefit to maintaining a separate "cache" of the state.

Fixes 52a5c155cf79 ("KVM: async_pf: Let guest support delivery of async_pf from guest mode")
Quoted 38 lines
---
 arch/x86/include/asm/kvm_host.h | 1 -
 arch/x86/kvm/x86.c              | 3 +--
 2 files changed, 1 insertion(+), 3 deletions(-)
arch/x86/include/asm/kvm_host.h +0 -1
index c470e40a00aa..fae1f4aeca5a 100644
@@ -1039,7 +1039,6 @@ struct kvm_vcpu_arch {
 		u32 id;
 		u32 host_apf_flags;
 		bool send_always;
-		bool delivery_as_pf_vmexit;
 		bool pageready_pending;
 	} apf;
 
arch/x86/kvm/x86.c +1 -2
index c35d359b56dd..4632222a5d1c 100644
@@ -3666,7 +3666,6 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 	}
 
 	vcpu->arch.apf.send_always = (data & KVM_ASYNC_PF_SEND_ALWAYS);
-	vcpu->arch.apf.delivery_as_pf_vmexit = data & KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT;
 
 	kvm_async_pf_wakeup_all(vcpu);
 
@@ -14035,7 +14034,7 @@ static bool kvm_can_deliver_async_pf(struct kvm_vcpu *vcpu)
 		 * L1 needs to opt into the special #PF vmexits that are
 		 * used to deliver async page faults.
 		 */
-		return vcpu->arch.apf.delivery_as_pf_vmexit;
+		return vcpu->arch.apf.msr_en_val & KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT;
 	} else {
 		/*
 		 * Play it safe in case the guest temporarily disables paging.
author maintainer
#6 lore raw

PATCH v3 3/3 KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_SEND_ALWAYS

In reply to Sean Christopherson

Drop kvm_vcpu_arch.send_always and instead use msr_en_val as the source of truth to reduce the probability of operating on stale data. This fixes flaws where KVM fails to update send_always when APF is explicitly disabled by the guest or implicitly disabled by KVM on INIT. Absent other bugs, the flaws are benign as KVM *shouldn't* consume send_always when PV APF support is disabled.

Simply delete the field, as there's zero benefit to maintaining a separate "cache" of the state.

Opportunistically turn the enabled vs. disabled logic at the end of kvm_pv_enable_async_pf() into an if-else instead of using an early return, e.g. so that it's more obvious that both paths are "success" paths.

Fixes 6adba5274206 ("KVM: Let host know whether the guest can handle async PF in non-userspace context.")
---
 arch/x86/include/asm/kvm_host.h |  1 -
 arch/x86/kvm/x86.c              | 12 ++++--------
 2 files changed, 4 insertions(+), 9 deletions(-)
arch/x86/include/asm/kvm_host.h +0 -1
index fae1f4aeca5a..2a6906597637 100644
@@ -1038,7 +1038,6 @@ struct kvm_vcpu_arch {
 		u16 vec;
 		u32 id;
 		u32 host_apf_flags;
-		bool send_always;
 		bool pageready_pending;
 	} apf;
 
arch/x86/kvm/x86.c +4 -8
index 4632222a5d1c..e24877353f17 100644
@@ -3659,16 +3659,12 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 
 	vcpu->arch.apf.msr_en_val = data;
 
-	if (!__kvm_pv_async_pf_enabled(data)) {
+	if (__kvm_pv_async_pf_enabled(data)) {
+		kvm_async_pf_wakeup_all(vcpu);
+	} else {
 		kvm_clear_async_pf_completion_queue(vcpu);
 		kvm_async_pf_hash_reset(vcpu);
-		return 0;
 	}
-
-	vcpu->arch.apf.send_always = (data & KVM_ASYNC_PF_SEND_ALWAYS);
-
-	kvm_async_pf_wakeup_all(vcpu);
-
 	return 0;
 }
 
@@ -14025,7 +14021,7 @@ static bool kvm_can_deliver_async_pf(struct kvm_vcpu *vcpu)
 	if (!kvm_pv_async_pf_enabled(vcpu))
 		return false;
 
-	if (!vcpu->arch.apf.send_always &&
+	if (!(vcpu->arch.apf.msr_en_val & KVM_ASYNC_PF_SEND_ALWAYS) &&
 	    (vcpu->arch.guest_state_protected || !kvm_x86_call(get_cpl)(vcpu)))
 		return false;
 
Signature · 1 line
2.53.0.1213.gd9a14994de-goog
Reviewed-by on v3 3/3
#7 lore raw

Re: [PATCH v3 3/3] KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_SEND_ALWAYS

Quoted 1 line · Sean Christopherson

Drop kvm_vcpu_arch.send_always and instead use msr_en_val as the source of

Nit: kvm_vcpu_arch.apf.send_always

Quoted 9 lines

truth to reduce the probability of operating on stale data. This fixes flaws where KVM fails to update send_always when APF is explicitly disabled by the guest or implicitly disabled by KVM on INIT. Absent other bugs, the flaws are benign as KVM *shouldn't* consume send_always when PV APF support is disabled.

Simply delete the field, as there's zero benefit to maintaining a separate "cache" of the state.

Quoted 6 lines

Opportunistically turn the enabled vs. disabled logic at the end of kvm_pv_enable_async_pf() into an if-else instead of using an early return, e.g. so that it's more obvious that both paths are "success" paths.

Fixes 6adba5274206 ("KVM: Let host know whether the guest can handle async PF in non-userspace context.")
Quoted 50 lines
---
  arch/x86/include/asm/kvm_host.h |  1 -
  arch/x86/kvm/x86.c              | 12 ++++--------
  2 files changed, 4 insertions(+), 9 deletions(-)
arch/x86/include/asm/kvm_host.h +0 -1
index fae1f4aeca5a..2a6906597637 100644
@@ -1038,7 +1038,6 @@ struct kvm_vcpu_arch {
  		u16 vec;
  		u32 id;
  		u32 host_apf_flags;
-		bool send_always;
  		bool pageready_pending;
  	} apf;
 
arch/x86/kvm/x86.c +4 -8
index 4632222a5d1c..e24877353f17 100644
@@ -3659,16 +3659,12 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 
  	vcpu->arch.apf.msr_en_val = data;
 
-	if (!__kvm_pv_async_pf_enabled(data)) {
+	if (__kvm_pv_async_pf_enabled(data)) {
+		kvm_async_pf_wakeup_all(vcpu);
+	} else {
  		kvm_clear_async_pf_completion_queue(vcpu);
  		kvm_async_pf_hash_reset(vcpu);
-		return 0;
  	}
-
-	vcpu->arch.apf.send_always = (data & KVM_ASYNC_PF_SEND_ALWAYS);
-
-	kvm_async_pf_wakeup_all(vcpu);
-
  	return 0;
  }
 
@@ -14025,7 +14021,7 @@ static bool kvm_can_deliver_async_pf(struct kvm_vcpu *vcpu)
  	if (!kvm_pv_async_pf_enabled(vcpu))
  		return false;
 
-	if (!vcpu->arch.apf.send_always &&
+	if (!(vcpu->arch.apf.msr_en_val & KVM_ASYNC_PF_SEND_ALWAYS) &&
  	    (vcpu->arch.guest_state_protected || !kvm_x86_call(get_cpl)(vcpu)))
  		return false;
 
Reviewed-by on v3 3/3
#8 lore raw

Re: [PATCH v3 3/3] KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_SEND_ALWAYS

In reply to Sean Christopherson
Quoted 7 lines · Sean Christopherson

Drop kvm_vcpu_arch.send_always and instead use msr_en_val as the source of truth to reduce the probability of operating on stale data. This fixes flaws where KVM fails to update send_always when APF is explicitly disabled by the guest or implicitly disabled by KVM on INIT. Absent other bugs, the flaws are benign as KVM *shouldn't* consume send_always when PV APF support is disabled.

Quoted 6 lines

Simply delete the field, as there's zero benefit to maintaining a separate "cache" of the state.

Opportunistically turn the enabled vs. disabled logic at the end of kvm_pv_enable_async_pf() into an if-else instead of using an early return, e.g. so that it's more obvious that both paths are "success" paths.

Nit:
Drop "e.g." ?

Quoted 3 lines
Fixes 6adba5274206 ("KVM: Let host know whether the guest can handle async PF in non-userspace context.")
Quoted 50 lines
---
 arch/x86/include/asm/kvm_host.h |  1 -
 arch/x86/kvm/x86.c              | 12 ++++--------
 2 files changed, 4 insertions(+), 9 deletions(-)
arch/x86/include/asm/kvm_host.h +0 -1
index fae1f4aeca5a..2a6906597637 100644
@@ -1038,7 +1038,6 @@ struct kvm_vcpu_arch {
 		u16 vec;
 		u32 id;
 		u32 host_apf_flags;
-		bool send_always;
 		bool pageready_pending;
 	} apf;
 
arch/x86/kvm/x86.c +4 -8
index 4632222a5d1c..e24877353f17 100644
@@ -3659,16 +3659,12 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 
 	vcpu->arch.apf.msr_en_val = data;
 
-	if (!__kvm_pv_async_pf_enabled(data)) {
+	if (__kvm_pv_async_pf_enabled(data)) {
+		kvm_async_pf_wakeup_all(vcpu);
+	} else {
 		kvm_clear_async_pf_completion_queue(vcpu);
 		kvm_async_pf_hash_reset(vcpu);
-		return 0;
 	}
-
-	vcpu->arch.apf.send_always = (data & KVM_ASYNC_PF_SEND_ALWAYS);
-
-	kvm_async_pf_wakeup_all(vcpu);
-
 	return 0;
 }
 
@@ -14025,7 +14021,7 @@ static bool kvm_can_deliver_async_pf(struct kvm_vcpu *vcpu)
 	if (!kvm_pv_async_pf_enabled(vcpu))
 		return false;
 
-	if (!vcpu->arch.apf.send_always &&
+	if (!(vcpu->arch.apf.msr_en_val & KVM_ASYNC_PF_SEND_ALWAYS) &&
 	    (vcpu->arch.guest_state_protected || !kvm_x86_call(get_cpl)(vcpu)))
 		return false;
 
author maintainer Applied
#10 lore raw
In reply to Sean Christopherson
Quoted 8 lines · Sean Christopherson

Fix a user-triggerable WARN found syzkaller, then clean up related code to eliminate potential stale data usage by removing what are effectively duplicate fields.

Ethan Yang (1):
KVM: x86: Don't leave APF half-enabled on bad APF data GPA

[...]

Applied to kvm-x86 misc, thanks!

[1/3] KVM: x86: Don't leave APF half-enabled on bad APF data GPA

      https://github.com/kvm-x86/linux/commit/923ca078f08c
[2/3] KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT
      https://github.com/kvm-x86/linux/commit/3c0bf11dd41f
[3/3] KVM: x86: Drop superfluous caching of KVM_ASYNC_PF_SEND_ALWAYS
      https://github.com/kvm-x86/linux/commit/7e985021ef2f

--
https://github.com/kvm-x86/linux/tree/next