Skip to contentSkip to conversation

KVM: Avoid illegal stage2 mapping on invalid memory slot

PATCH In review Gavin Shan 6 participants 6 messages Jun 15, 2023 – Jun 16, 2023

Subsystems (2)
lore
Share & tools
lore

Gavin Shan

author
#0 lore raw

PATCH v3 1/1 KVM: Avoid illegal stage2 mapping on invalid memory slot

We run into guest hang in edk2 firmware when KSM is kept as running on the host. The edk2 firmware is waiting for status 0x80 from QEMU's pflash device (TYPE_PFLASH_CFI01) during the operation of sector erasing or buffered write. The status is returned by reading the memory region of the pflash device and the read request should have been forwarded to QEMU and emulated by it. Unfortunately, the read request is covered by an illegal stage2 mapping when the guest hang issue occurs. The read request is completed with QEMU bypassed and wrong status is fetched. The edk2 firmware runs into an infinite loop with the wrong status.

The illegal stage2 mapping is populated due to same page sharing by KSM at (C) even the associated memory slot has been marked as invalid at (B) when the memory slot is requested to be deleted. It's notable that the active and inactive memory slots can't be swapped when we're in the middle of kvm_mmu_notifier_change_pte() because kvm->mn_active_invalidate_count is elevated, and kvm_swap_active_memslots() will busy loop until it reaches to zero again. Besides, the swapping from the active to the inactive memory slots is also avoided by holding &kvm->srcu in __kvm_handle_hva_range(), corresponding to synchronize_srcu_expedited() in kvm_swap_active_memslots().

CPU-A CPU-B
----- -----
ioctl(kvm_fd, KVM_SET_USER_MEMORY_REGION) kvm_vm_ioctl_set_memory_region kvm_set_memory_region __kvm_set_memory_region kvm_set_memslot(kvm, old, NULL, KVM_MR_DELETE)
kvm_invalidate_memslot
kvm_copy_memslot kvm_replace_memslot
kvm_swap_active_memslots (A)
kvm_arch_flush_shadow_memslot (B)
same page sharing by KSM kvm_mmu_notifier_invalidate_range_start

        :
  kvm_mmu_notifier_change_pte
    kvm_handle_hva_range
    __kvm_handle_hva_range
    kvm_set_spte_gfn            (C)
        :

kvm_mmu_notifier_invalidate_range_end

Fix the issue by skipping the invalid memory slot at (C) to avoid the illegal stage2 mapping so that the read request for the pflash's status is forwarded to QEMU and emulated by it. In this way, the correct pflash's status can be returned from QEMU to break the infinite loop in the edk2 firmware.

We tried a git-bisect and the first problematic commit is cd4c71835228 ("
KVM: arm64: Convert to the gfn-based MMU notifier callbacks"). With this, clean_dcache_guest_page() is called after the memory slots are iterated in kvm_mmu_notifier_change_pte(). clean_dcache_guest_page() is called before the iteration on the memory slots before this commit. This change literally enlarges the racy window between kvm_mmu_notifier_change_pte() and memory slot removal so that we're able to reproduce the issue in a practical test case. However, the issue exists since commit d5d8184d35c9 ("KVM: ARM: Memory virtualization setup").

Fixes d5d8184d35c9 ("KVM: ARM: Memory virtualization setup")

---
v3: Skip the invalid memory slots in change_pte() MMU notifier only,
suggested by Sean. Improved changelog to describe how the fixes tag is given.

---
 virt/kvm/kvm_main.c | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)
virt/kvm/kvm_main.c +19 -1
index 479802a892d4..65f94f592ff8 100644
@@ -686,6 +686,24 @@ static __always_inline int kvm_handle_hva_range_no_flush(struct mmu_notifier *mn
 
 	return __kvm_handle_hva_range(kvm, &range);
 }
+
+static bool kvm_change_spte_gfn(struct kvm *kvm, struct kvm_gfn_range *range)
+{
+	/*
+	 * Skipping invalid memslots is correct if and only change_pte() is
+	 * surrounded by invalidate_range_{start,end}(), which is currently
+	 * guaranteed by the primary MMU.  If that ever changes, KVM needs to
+	 * unmap the memslot instead of skipping the memslot to ensure that KVM
+	 * doesn't hold references to the old PFN.
+	 */
+	WARN_ON_ONCE(!READ_ONCE(kvm->mn_active_invalidate_count));
+
+	if (range->slot->flags & KVM_MEMSLOT_INVALID)
+		return false;
+
+	return kvm_set_spte_gfn(kvm, range);
+}
+
 static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn,
 					struct mm_struct *mm,
 					unsigned long address,
@@ -707,7 +725,7 @@ static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn,
 	if (!READ_ONCE(kvm->mmu_invalidate_in_progress))
 		return;
 
-	kvm_handle_hva_range(mn, address, address + 1, pte, kvm_set_spte_gfn);
+	kvm_handle_hva_range(mn, address, address + 1, pte, kvm_change_spte_gfn);
 }
 
 void kvm_mmu_invalidate_begin(struct kvm *kvm, unsigned long start,
Signature · 1 line
2.40.1
Reviewed-by on v3 1/1
#1 lore raw
Quoted 101 lines · Gavin Shan

We run into guest hang in edk2 firmware when KSM is kept as running on the host. The edk2 firmware is waiting for status 0x80 from QEMU's pflash device (TYPE_PFLASH_CFI01) during the operation of sector erasing or buffered write. The status is returned by reading the memory region of the pflash device and the read request should have been forwarded to QEMU and emulated by it. Unfortunately, the read request is covered by an illegal stage2 mapping when the guest hang issue occurs. The read request is completed with QEMU bypassed and wrong status is fetched. The edk2 firmware runs into an infinite loop with the wrong status.

The illegal stage2 mapping is populated due to same page sharing by KSM at (C) even the associated memory slot has been marked as invalid at (B) when the memory slot is requested to be deleted. It's notable that the active and inactive memory slots can't be swapped when we're in the middle of kvm_mmu_notifier_change_pte() because kvm->mn_active_invalidate_count is elevated, and kvm_swap_active_memslots() will busy loop until it reaches to zero again. Besides, the swapping from the active to the inactive memory slots is also avoided by holding &kvm->srcu in __kvm_handle_hva_range(), corresponding to synchronize_srcu_expedited() in kvm_swap_active_memslots().

   CPU-A                    CPU-B
   -----                    -----
                            ioctl(kvm_fd, KVM_SET_USER_MEMORY_REGION)
                            kvm_vm_ioctl_set_memory_region
                            kvm_set_memory_region
                            __kvm_set_memory_region
                            kvm_set_memslot(kvm, old, NULL, KVM_MR_DELETE)
                              kvm_invalidate_memslot
                                kvm_copy_memslot
                                kvm_replace_memslot
                                kvm_swap_active_memslots        (A)
                                kvm_arch_flush_shadow_memslot   (B)
   same page sharing by KSM
   kvm_mmu_notifier_invalidate_range_start
         :
   kvm_mmu_notifier_change_pte
     kvm_handle_hva_range
     __kvm_handle_hva_range
     kvm_set_spte_gfn            (C)
         :
   kvm_mmu_notifier_invalidate_range_end

Fix the issue by skipping the invalid memory slot at (C) to avoid the illegal stage2 mapping so that the read request for the pflash's status is forwarded to QEMU and emulated by it. In this way, the correct pflash's status can be returned from QEMU to break the infinite loop in the edk2 firmware.

We tried a git-bisect and the first problematic commit is cd4c71835228 ("
KVM: arm64: Convert to the gfn-based MMU notifier callbacks"). With this, clean_dcache_guest_page() is called after the memory slots are iterated in kvm_mmu_notifier_change_pte(). clean_dcache_guest_page() is called before the iteration on the memory slots before this commit. This change literally enlarges the racy window between kvm_mmu_notifier_change_pte() and memory slot removal so that we're able to reproduce the issue in a practical test case. However, the issue exists since commit d5d8184d35c9 ("KVM: ARM: Memory virtualization setup").

Fixes d5d8184d35c9 ("KVM: ARM: Memory virtualization setup")

---
v3: Skip the invalid memory slots in change_pte() MMU notifier only,
suggested by Sean. Improved changelog to describe how the fixes tag is given.

---
  virt/kvm/kvm_main.c | 20 +++++++++++++++++++-
  1 file changed, 19 insertions(+), 1 deletion(-)
virt/kvm/kvm_main.c +18 -0
index 479802a892d4..65f94f592ff8 100644
@@ -686,6 +686,24 @@ static __always_inline int kvm_handle_hva_range_no_flush(struct mmu_notifier *mn
 
  	return __kvm_handle_hva_range(kvm, &range);
  }
+
+static bool kvm_change_spte_gfn(struct kvm *kvm, struct kvm_gfn_range *range)
+{
+	/*
+	 * Skipping invalid memslots is correct if and only change_pte() is
+	 * surrounded by invalidate_range_{start,end}(), which is currently
+	 * guaranteed by the primary MMU.  If that ever changes, KVM needs to
+	 * unmap the memslot instead of skipping the memslot to ensure that KVM
+	 * doesn't hold references to the old PFN.
+	 */
+	WARN_ON_ONCE(!READ_ONCE(kvm->mn_active_invalidate_count));
+
+	if (range->slot->flags & KVM_MEMSLOT_INVALID)
+		return false;
+
+	return kvm_set_spte_gfn(kvm, range);
+}
+
  static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn,
  					struct mm_struct *mm,
  					unsigned long address,

virt/kvm/kvm_main.c static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn, · line 731

Quoted 8 lines
virt/kvm/kvm_main.c +1 -1
@@ -707,7 +725,7 @@ static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn,
  	if (!READ_ONCE(kvm->mmu_invalidate_in_progress))
  		return;
 
-	kvm_handle_hva_range(mn, address, address + 1, pte, kvm_set_spte_gfn);
+	kvm_handle_hva_range(mn, address, address + 1, pte, kvm_change_spte_gfn);
  }
 
  void kvm_mmu_invalidate_begin(struct kvm *kvm, unsigned long start,

Makes perfect sense to me as well

Signature · 3 lines
Cheers,

David / dhildenb
Reviewed-by on v3 1/1
#2 lore raw
In reply to Gavin Shan
Quoted 9 lines · Gavin Shan

We run into guest hang in edk2 firmware when KSM is kept as running on the host. The edk2 firmware is waiting for status 0x80 from QEMU's pflash device (TYPE_PFLASH_CFI01) during the operation of sector erasing or buffered write. The status is returned by reading the memory region of the pflash device and the read request should have been forwarded to QEMU and emulated by it. Unfortunately, the read request is covered by an illegal stage2 mapping when the guest hang issue occurs. The read request is completed with QEMU bypassed and wrong status is fetched. The edk2 firmware runs into an infinite loop with the wrong status.

[...]

Quoted 16 lines

Fix the issue by skipping the invalid memory slot at (C) to avoid the illegal stage2 mapping so that the read request for the pflash's status is forwarded to QEMU and emulated by it. In this way, the correct pflash's status can be returned from QEMU to break the infinite loop in the edk2 firmware.

We tried a git-bisect and the first problematic commit is cd4c71835228 ("
KVM: arm64: Convert to the gfn-based MMU notifier callbacks"). With this, clean_dcache_guest_page() is called after the memory slots are iterated in kvm_mmu_notifier_change_pte(). clean_dcache_guest_page() is called before the iteration on the memory slots before this commit. This change literally enlarges the racy window between kvm_mmu_notifier_change_pte() and memory slot removal so that we're able to reproduce the issue in a practical test case. However, the issue exists since commit d5d8184d35c9 ("KVM: ARM: Memory virtualization setup").

Quoted 5 lines
Fixes d5d8184d35c9 ("KVM: ARM: Memory virtualization setup")

Thanks for fixing this Gavin. I'm more than happy to take this through the kvmarm tree since we got burned, but it'd probably be best if Paolo got it.

Paolo, if you wind up grabbing it:

Quoted 13 lines

---
v3: Skip the invalid memory slots in change_pte() MMU notifier only,
suggested by Sean. Improved changelog to describe how the fixes tag is given.

---
 virt/kvm/kvm_main.c | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)
virt/kvm/kvm_main.c +0 -0
index 479802a892d4..65f94f592ff8 100644

virt/kvm/kvm_main.c static __always_inline int kvm_handle_hva_range_no_flush(struct mmu_notifier *mn · line 693

Quoted 8 lines
virt/kvm/kvm_main.c +5 -0
@@ -686,6 +686,24 @@ static __always_inline int kvm_handle_hva_range_no_flush(struct mmu_notifier *mn
 
 	return __kvm_handle_hva_range(kvm, &range);
 }
+
+static bool kvm_change_spte_gfn(struct kvm *kvm, struct kvm_gfn_range *range)
+{
+	/*
+	 * Skipping invalid memslots is correct if and only change_pte() is

nit: I think there's a typo here: s/if and only/if and only if/

This is tiny, and can be fixed up when the patch is applied.

Signature · 2 lines
Thanks,
Oliver
Reviewed-by on v3 1/1
#3 lore raw
In reply to Gavin Shan
Quoted 63 lines · Gavin Shan

We run into guest hang in edk2 firmware when KSM is kept as running on the host. The edk2 firmware is waiting for status 0x80 from QEMU's pflash device (TYPE_PFLASH_CFI01) during the operation of sector erasing or buffered write. The status is returned by reading the memory region of the pflash device and the read request should have been forwarded to QEMU and emulated by it. Unfortunately, the read request is covered by an illegal stage2 mapping when the guest hang issue occurs. The read request is completed with QEMU bypassed and wrong status is fetched. The edk2 firmware runs into an infinite loop with the wrong status.

The illegal stage2 mapping is populated due to same page sharing by KSM at (C) even the associated memory slot has been marked as invalid at (B) when the memory slot is requested to be deleted. It's notable that the active and inactive memory slots can't be swapped when we're in the middle of kvm_mmu_notifier_change_pte() because kvm->mn_active_invalidate_count is elevated, and kvm_swap_active_memslots() will busy loop until it reaches to zero again. Besides, the swapping from the active to the inactive memory slots is also avoided by holding &kvm->srcu in __kvm_handle_hva_range(), corresponding to synchronize_srcu_expedited() in kvm_swap_active_memslots().

CPU-A CPU-B
----- -----
ioctl(kvm_fd, KVM_SET_USER_MEMORY_REGION) kvm_vm_ioctl_set_memory_region kvm_set_memory_region __kvm_set_memory_region kvm_set_memslot(kvm, old, NULL, KVM_MR_DELETE)
kvm_invalidate_memslot
kvm_copy_memslot kvm_replace_memslot
kvm_swap_active_memslots (A)
kvm_arch_flush_shadow_memslot (B)
same page sharing by KSM kvm_mmu_notifier_invalidate_range_start

        :
  kvm_mmu_notifier_change_pte
    kvm_handle_hva_range
    __kvm_handle_hva_range
    kvm_set_spte_gfn            (C)
        :

kvm_mmu_notifier_invalidate_range_end

Fix the issue by skipping the invalid memory slot at (C) to avoid the illegal stage2 mapping so that the read request for the pflash's status is forwarded to QEMU and emulated by it. In this way, the correct pflash's status can be returned from QEMU to break the infinite loop in the edk2 firmware.

We tried a git-bisect and the first problematic commit is cd4c71835228 ("
KVM: arm64: Convert to the gfn-based MMU notifier callbacks"). With this, clean_dcache_guest_page() is called after the memory slots are iterated in kvm_mmu_notifier_change_pte(). clean_dcache_guest_page() is called before the iteration on the memory slots before this commit. This change literally enlarges the racy window between kvm_mmu_notifier_change_pte() and memory slot removal so that we're able to reproduce the issue in a practical test case. However, the issue exists since commit d5d8184d35c9 ("KVM: ARM: Memory virtualization setup").

Fixes d5d8184d35c9 ("KVM: ARM: Memory virtualization setup")
Quoted 4 lines

---
v3: Skip the invalid memory slots in change_pte() MMU notifier only,
suggested by Sean. Improved changelog to describe how the fixes tag is given.

True.. FWIW:

Signature · 1 line
Peter Xu
Reviewed-by on v3 1/1
#5 lore raw
In reply to Gavin Shan
Quoted 58 lines · Gavin Shan

We run into guest hang in edk2 firmware when KSM is kept as running on the host. The edk2 firmware is waiting for status 0x80 from QEMU's pflash device (TYPE_PFLASH_CFI01) during the operation of sector erasing or buffered write. The status is returned by reading the memory region of the pflash device and the read request should have been forwarded to QEMU and emulated by it. Unfortunately, the read request is covered by an illegal stage2 mapping when the guest hang issue occurs. The read request is completed with QEMU bypassed and wrong status is fetched. The edk2 firmware runs into an infinite loop with the wrong status.

The illegal stage2 mapping is populated due to same page sharing by KSM at (C) even the associated memory slot has been marked as invalid at (B) when the memory slot is requested to be deleted. It's notable that the active and inactive memory slots can't be swapped when we're in the middle of kvm_mmu_notifier_change_pte() because kvm->mn_active_invalidate_count is elevated, and kvm_swap_active_memslots() will busy loop until it reaches to zero again. Besides, the swapping from the active to the inactive memory slots is also avoided by holding &kvm->srcu in __kvm_handle_hva_range(), corresponding to synchronize_srcu_expedited() in kvm_swap_active_memslots().

   CPU-A                    CPU-B
   -----                    -----
                            ioctl(kvm_fd, KVM_SET_USER_MEMORY_REGION)
                            kvm_vm_ioctl_set_memory_region
                            kvm_set_memory_region
                            __kvm_set_memory_region
                            kvm_set_memslot(kvm, old, NULL, KVM_MR_DELETE)
                              kvm_invalidate_memslot
                                kvm_copy_memslot
                                kvm_replace_memslot
                                kvm_swap_active_memslots        (A)
                                kvm_arch_flush_shadow_memslot   (B)
   same page sharing by KSM
   kvm_mmu_notifier_invalidate_range_start
         :
   kvm_mmu_notifier_change_pte
     kvm_handle_hva_range
     __kvm_handle_hva_range
     kvm_set_spte_gfn            (C)
         :
   kvm_mmu_notifier_invalidate_range_end

Fix the issue by skipping the invalid memory slot at (C) to avoid the illegal stage2 mapping so that the read request for the pflash's status is forwarded to QEMU and emulated by it. In this way, the correct pflash's status can be returned from QEMU to break the infinite loop in the edk2 firmware.

We tried a git-bisect and the first problematic commit is cd4c71835228 ("
KVM: arm64: Convert to the gfn-based MMU notifier callbacks"). With this, clean_dcache_guest_page() is called after the memory slots are iterated in kvm_mmu_notifier_change_pte(). clean_dcache_guest_page() is called before the iteration on the memory slots before this commit. This change literally enlarges the racy window between kvm_mmu_notifier_change_pte() and memory slot removal so that we're able to reproduce the issue in a practical test case. However, the issue exists since commit d5d8184d35c9 ("KVM: ARM: Memory virtualization setup").

Quoted 5 lines
Fixes d5d8184d35c9 ("KVM: ARM: Memory virtualization setup")
Quoted 46 lines

---
v3: Skip the invalid memory slots in change_pte() MMU notifier only,
suggested by Sean. Improved changelog to describe how the fixes tag is given.

---
  virt/kvm/kvm_main.c | 20 +++++++++++++++++++-
  1 file changed, 19 insertions(+), 1 deletion(-)
virt/kvm/kvm_main.c +19 -1
index 479802a892d4..65f94f592ff8 100644
@@ -686,6 +686,24 @@ static __always_inline int kvm_handle_hva_range_no_flush(struct mmu_notifier *mn
 
  	return __kvm_handle_hva_range(kvm, &range);
  }
+
+static bool kvm_change_spte_gfn(struct kvm *kvm, struct kvm_gfn_range *range)
+{
+	/*
+	 * Skipping invalid memslots is correct if and only change_pte() is
+	 * surrounded by invalidate_range_{start,end}(), which is currently
+	 * guaranteed by the primary MMU.  If that ever changes, KVM needs to
+	 * unmap the memslot instead of skipping the memslot to ensure that KVM
+	 * doesn't hold references to the old PFN.
+	 */
+	WARN_ON_ONCE(!READ_ONCE(kvm->mn_active_invalidate_count));
+
+	if (range->slot->flags & KVM_MEMSLOT_INVALID)
+		return false;
+
+	return kvm_set_spte_gfn(kvm, range);
+}
+
  static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn,
  					struct mm_struct *mm,
  					unsigned long address,
@@ -707,7 +725,7 @@ static void kvm_mmu_notifier_change_pte(struct mmu_notifier *mn,
  	if (!READ_ONCE(kvm->mmu_invalidate_in_progress))
  		return;
 
-	kvm_handle_hva_range(mn, address, address + 1, pte, kvm_set_spte_gfn);
+	kvm_handle_hva_range(mn, address, address + 1, pte, kvm_change_spte_gfn);
  }
 
  void kvm_mmu_invalidate_begin(struct kvm *kvm, unsigned long start,
Signature · 1 line
Shaoqin