s390/vfio-ap: fix stale pqap_hook pointer on error in vfio_ap_mdev_set_kvm()
Subsystems (5)
Anthony Krowiak
maintainerPATCH v2 1/1 s390/vfio-ap: fix stale pqap_hook pointer on error in vfio_ap_mdev_set_kvm()
In vfio_ap_mdev_set_kvm(), kvm->arch.crypto.pqap_hook is set to &matrix_mdev->pqap_hook before the update locks are acquired and the mdev list is checked for a conflicting assignment. If another mdev is already attached to the same KVM instance, the function returns -EPERM without restoring the hook pointer, leaving kvm->arch.crypto.pqap_hook pointing at the failing matrix_mdev instead of the mdev that legitimately owns the KVM.
Since matrix_mdev->kvm is never set on this error path, vfio_ap_mdev_unset_kvm() will not clean up the hook when matrix_mdev is later closed. If matrix_mdev is subsequently freed, any PQAP instruction executed by the guest will dereference the stale pointer through pqap_hook_rwsem, resulting in a use-after-free.
Since kvm->arch.crypto.pqap_hook is only set in the vfio_ap_mdev_set_kvm() function and is cleared in the vfio_ap_mdev_unset_kvm() function, a check for 'kvm->arch.crypto.pqap_hook != NULL' is all that is needed to determine whether it belongs to another mdev. This will alleviate the need to iterate the matrix_dev->mdev_list list to see if the kvm object is assigned to another mdev. This likewise eliminates the need to hold the matrix_dev->mdevs_lock while iterating the list.
Note that the 'matrix_mdev' pointer retrieved via the 'container_of' macro in the 'vfio_ap_mdev_open_device()' callback will be available for the duration of that function because 'matrix_mdev' cannot be freed while 'open_device' (which calls the 'vfio_ap_mdev_open_device()' function) is running.
--- drivers/s390/crypto/vfio_ap_ops.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-)
drivers/s390/crypto/vfio_ap_ops.c +10 -11
@@ -1821,26 +1821,25 @@ static const struct attribute_group *vfio_ap_mdev_attr_groups[] = {static int vfio_ap_mdev_set_kvm (struct ap_matrix_mdev *matrix_mdev,struct kvm *kvm) { -struct ap_matrix_mdev *m; -if (kvm->arch.crypto.crycbd) { +/* +* Taking the kvm->arch.crypto.pqap_hook_rwsem lock while +* holding the update locks (below) could result in a +* deadlock condition, so let's hold the rwsem here while we +* check and update the pqap_hook. +*/ down_write (&kvm->arch.crypto.pqap_hook_rwsem); +if (kvm->arch.crypto.pqap_hook) { +up_write (&kvm->arch.crypto.pqap_hook_rwsem); +return -EPERM ; + } kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook;up_write (&kvm->arch.crypto.pqap_hook_rwsem);get_update_locks_for_kvm (kvm); - -list_for_each_entry (m, &matrix_dev->mdev_list, node) { -if (m != matrix_mdev && m->kvm == kvm) { -release_update_locks_for_kvm (kvm); -return -EPERM ; - } - } -kvm_get_kvm (kvm); matrix_mdev->kvm = kvm;vfio_ap_mdev_update_guest_apcb (matrix_mdev); -release_update_locks_for_kvm (kvm); }
Signature · 1 line
2.53.0
[email protected]
AI reviewThank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] Removing the `mdev_list` iteration from within the `update_locks` critical section introduces a…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Removing the mdev_list iteration from within the update_locks critical section introduces a Time-Of-Check to Time-Of-Use (TOCTOU) race condition between vfio_ap_mdev_set_kvm() and vfio_ap_mdev_unset_kvm() that silently clobbers KVM APCB masks for a newly attached mdev. --
--- Patch [1]: [PATCH v2] s390/vfio-ap: fix stale pqap_hook pointer on error in vfio_ap_mdev_set_kvm() --- commit 0834fa20c3fb715cac9e1ff09376e43784d44814
Author: Anthony Krowiak <[email protected]>
s390/vfio-ap: fix stale pqap_hook pointer on error in vfio_ap_mdev_set_kvm()
This commit changes vfio_ap_mdev_set_kvm() to use pqap_hook to detect conflicting assignments instead of iterating mdev_list under the update locks, resolving a potential stale pointer dereference.
Quoted 4 lines
drivers/s390/crypto/vfio_ap_ops.c +0
-0
[ ... ]
Quoted 12 lines
drivers/s390/crypto/vfio_ap_ops.c +5
-2@@ -1821,26 +1821,25 @@ static const struct attribute_group *vfio_ap_mdev_attr_groups[] = {static int vfio_ap_mdev_set_kvm (struct ap_matrix_mdev *matrix_mdev,struct kvm *kvm) { -struct ap_matrix_mdev *m; -if (kvm->arch.crypto.crycbd) { +/* +* Taking the kvm->arch.crypto.pqap_hook_rwsem lock while +* holding the update locks (below) could result in a +* deadlock condition, so let's hold the rwsem here while we +* check and update the pqap_hook.
drivers/s390/crypto/vfio_ap_ops.c static const struct attribute_group *vfio_ap_mdev_attr_groups[] = { · line 1839
Quoted 10 lines
drivers/s390/crypto/vfio_ap_ops.c +5
-0+*/ down_write (&kvm->arch.crypto.pqap_hook_rwsem); +if (kvm->arch.crypto.pqap_hook) { +up_write (&kvm->arch.crypto.pqap_hook_rwsem); +return -EPERM ; + } kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook;up_write (&kvm->arch.crypto.pqap_hook_rwsem);get_update_locks_for_kvm (kvm);
[Severity: High]
Can this introduce a race with vfio_ap_mdev_unset_kvm()?
By moving the conflict check outside of the update_locks critical section, is it possible for the following sequence to occur during concurrent hot-plug/unplug operations?
Thread 1 (in vfio_ap_mdev_unset_kvm):
down_write(&kvm->arch.crypto.pqap_hook_rwsem);
kvm->arch.crypto.pqap_hook = NULL;
up_write(&kvm->arch.crypto.pqap_hook_rwsem);Thread 2 (in vfio_ap_mdev_set_kvm):
down_write(&kvm->arch.crypto.pqap_hook_rwsem);
// Sees pqap_hook is NULL
kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook;
up_write(&kvm->arch.crypto.pqap_hook_rwsem);
get_update_locks_for_kvm(kvm);
vfio_ap_mdev_update_guest_apcb(matrix_mdev);
release_update_locks_for_kvm(kvm);Thread 1 (continuing in vfio_ap_mdev_unset_kvm):
get_update_locks_for_kvm(kvm);
kvm_arch_crypto_clear_masks(kvm);If Thread 1 interleaves after Thread 2 sets the guest masks but before Thread 1 clears them, would this result in Thread 1 unconditionally clearing the KVM APCB masks that Thread 2 just successfully configured, leaving the newly attached mdev with inaccessible queues?
Quoted 15 lines
drivers/s390/crypto/vfio_ap_ops.c +0
-9- -list_for_each_entry (m, &matrix_dev->mdev_list, node) { -if (m != matrix_mdev && m->kvm == kvm) { -release_update_locks_for_kvm (kvm); -return -EPERM ; - } - } -kvm_get_kvm (kvm); matrix_mdev->kvm = kvm;vfio_ap_mdev_update_guest_apcb (matrix_mdev); -release_update_locks_for_kvm (kvm); }
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
Anthony Krowiak
maintainerQuoted 64 lines · On 7/28/26 12:41 PM, [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Removing themdev_listiteration from within theupdate_lockscritical section introduces a Time-Of-Check to Time-Of-Use (TOCTOU) race condition betweenvfio_ap_mdev_set_kvm()andvfio_ap_mdev_unset_kvm()that silently clobbers KVM APCB masks for a newly attached mdev. ----- Patch [1]: [PATCH v2] s390/vfio-ap: fix stale
pqap_hookpointer on error invfio_ap_mdev_set_kvm()--- commit0834fa20c3fb715cac9e1ff09376e43784d44814
Author: Anthony Krowiak <[email protected]>
s390/vfio-ap: fix stalepqap_hookpointer on error invfio_ap_mdev_set_kvm()This commit changes
vfio_ap_mdev_set_kvm()to usepqap_hookto detect conflicting assignments instead of iteratingmdev_listunder the update locks, resolving a potential stale pointer dereference.Fixes 86956e70761b3 ("s390/vfio-ap: replace open coded locks for VFIO_GROUP_NOTIFY_SET_KVM notification")Quoted 4 lines
drivers/s390/crypto/vfio_ap_ops.c +0
-0[ ... ]
Quoted 12 lines
drivers/s390/crypto/vfio_ap_ops.c +5
-2@@ -1821,26 +1821,25 @@ static const struct attribute_group *vfio_ap_mdev_attr_groups[] = {static int vfio_ap_mdev_set_kvm (struct ap_matrix_mdev *matrix_mdev,struct kvm *kvm) { -struct ap_matrix_mdev *m; -if (kvm->arch.crypto.crycbd) { +/* +* Taking the kvm->arch.crypto.pqap_hook_rwsem lock while +* holding the update locks (below) could result in a +* deadlock condition, so let's hold the rwsem here while we +* check and update the pqap_hook. drivers/s390/crypto/vfio_ap_ops.c static const struct attribute_group *vfio_ap_mdev_attr_groups[] = { · line 1839
Quoted 10 lines
drivers/s390/crypto/vfio_ap_ops.c +5
-0+*/ down_write (&kvm->arch.crypto.pqap_hook_rwsem); +if (kvm->arch.crypto.pqap_hook) { +up_write (&kvm->arch.crypto.pqap_hook_rwsem); +return -EPERM ; + } kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook;up_write (&kvm->arch.crypto.pqap_hook_rwsem);get_update_locks_for_kvm (kvm);[Severity: High]
Can this introduce a race withvfio_ap_mdev_unset_kvm()?By moving the conflict check outside of the
update_lockscritical section, is it possible for the following sequence to occur during concurrent hot-plug/unplug operations?Thread 1 (in
vfio_ap_mdev_unset_kvm):down_write(&kvm->arch.crypto.pqap_hook_rwsem); kvm->arch.crypto.pqap_hook = NULL; up_write(&kvm->arch.crypto.pqap_hook_rwsem);Thread 2 (in
vfio_ap_mdev_set_kvm):down_write(&kvm->arch.crypto.pqap_hook_rwsem); // Sees pqap_hook is NULL kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook; up_write(&kvm->arch.crypto.pqap_hook_rwsem); get_update_locks_for_kvm(kvm); vfio_ap_mdev_update_guest_apcb(matrix_mdev); release_update_locks_for_kvm(kvm);
Quoted 8 lines
Thread 1 (continuing in
vfio_ap_mdev_unset_kvm):get_update_locks_for_kvm(kvm); kvm_arch_crypto_clear_masks(kvm);If Thread 1 interleaves after Thread 2 sets the guest masks but before Thread 1 clears them, would this result in Thread 1 unconditionally clearing the KVM APCB masks that Thread 2 just successfully configured, leaving the newly attached mdev with inaccessible queues?
This scenario requires Thread 1 (unset_kvm) to be tearing down the same matrix_mdev that Thread 2 (set_kvm) is attaching. But look at the entry conditions:
* vfio_ap_mdev_unset_kvm is called from vfio_ap_mdev_close_device
* vfio_ap_mdev_set_kvm is called from vfio_ap_mdev_open_device
open_device and close_device on the same vfio_device are serialized by the vfio core under dev_set->lock; you cannot have both executing concurrently for the same mdev.
Furthermore, vfio_ap_mdev_unset_kvm reads matrix_mdev->kvm which is only set to a non-NULL value inside vfio_ap_mdev_set_kvm. For Thread 1's unset_kvm to operate on the same kvm as Thread 2's set_kvm, Thread 1's matrix_mdev would have to have already gone through a complete set_kvm cycle and have ->kvm set; in other words, it is a different matrix_mdev that previously attached to the same KVM.
This fix prevents that: the pqap_hook != NULL check blocks Thread 2 from attaching to a KVM that already has an owner. So Thread 1 and Thread 2 can never legitimately be racing over the same kvm in the scenario described.
Quoted 16 lines
Quoted 16 lines
drivers/s390/crypto/vfio_ap_ops.c +0
-9- -list_for_each_entry (m, &matrix_dev->mdev_list, node) { -if (m != matrix_mdev && m->kvm == kvm) { -release_update_locks_for_kvm (kvm); -return -EPERM ; - } - } -kvm_get_kvm (kvm); matrix_mdev->kvm = kvm;vfio_ap_mdev_update_guest_apcb (matrix_mdev); -release_update_locks_for_kvm (kvm); }
Anthony Krowiak
maintainerQuoted 42 lines · On 7/28/26 12:41 PM, [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Removing themdev_listiteration from within theupdate_lockscritical section introduces a Time-Of-Check to Time-Of-Use (TOCTOU) race condition betweenvfio_ap_mdev_set_kvm()andvfio_ap_mdev_unset_kvm()that silently clobbers KVM APCB masks for a newly attached mdev. ----- Patch [1]: [PATCH v2] s390/vfio-ap: fix stale
pqap_hookpointer on error invfio_ap_mdev_set_kvm()--- commit0834fa20c3fb715cac9e1ff09376e43784d44814
Author: Anthony Krowiak <[email protected]>
s390/vfio-ap: fix stalepqap_hookpointer on error invfio_ap_mdev_set_kvm()This commit changes
vfio_ap_mdev_set_kvm()to usepqap_hookto detect conflicting assignments instead of iteratingmdev_listunder the update locks, resolving a potential stale pointer dereference.Fixes 86956e70761b3 ("s390/vfio-ap: replace open coded locks for VFIO_GROUP_NOTIFY_SET_KVM notification")Quoted 4 lines
drivers/s390/crypto/vfio_ap_ops.c +0
-0[ ... ]
Quoted 21 lines
drivers/s390/crypto/vfio_ap_ops.c +10
-2@@ -1821,26 +1821,25 @@ static const struct attribute_group *vfio_ap_mdev_attr_groups[] = {static int vfio_ap_mdev_set_kvm (struct ap_matrix_mdev *matrix_mdev,struct kvm *kvm) { -struct ap_matrix_mdev *m; -if (kvm->arch.crypto.crycbd) { +/* +* Taking the kvm->arch.crypto.pqap_hook_rwsem lock while +* holding the update locks (below) could result in a +* deadlock condition, so let's hold the rwsem here while we +* check and update the pqap_hook. +*/ down_write (&kvm->arch.crypto.pqap_hook_rwsem); +if (kvm->arch.crypto.pqap_hook) { +up_write (&kvm->arch.crypto.pqap_hook_rwsem); +return -EPERM ; + } kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook;up_write (&kvm->arch.crypto.pqap_hook_rwsem);
Quoted 3 lines
drivers/s390/crypto/vfio_ap_ops.c static const struct attribute_group *vfio_ap_mdev_attr_groups[] = { · line 1839
Quoted 1 line
drivers/s390/crypto/vfio_ap_ops.c +0
-0get_update_locks_for_kvm (kvm);[Severity: High]
Can this introduce a race withvfio_ap_mdev_unset_kvm()?
After further review, it looks like this is a legitimate concern. As the comment preceding the the code inside of the 'if' block stated, we needed to take the pqap_hook_rwsem lock outside of the update_locks to avoid a deadlock situation when the pqap_hook is executed. There is no reason, however, that we can't take the update_locks while holding the pqap_hook_rwsem. The same thing will have to also be done in the vfio_ap_mdev_unset_kvm function also.
Quoted 43 lines
By moving the conflict check outside of the
update_lockscritical section, is it possible for the following sequence to occur during concurrent hot-plug/unplug operations?Thread 1 (in
vfio_ap_mdev_unset_kvm):down_write(&kvm->arch.crypto.pqap_hook_rwsem); kvm->arch.crypto.pqap_hook = NULL; up_write(&kvm->arch.crypto.pqap_hook_rwsem);Thread 2 (in
vfio_ap_mdev_set_kvm):down_write(&kvm->arch.crypto.pqap_hook_rwsem); // Sees pqap_hook is NULL kvm->arch.crypto.pqap_hook = &matrix_mdev->pqap_hook; up_write(&kvm->arch.crypto.pqap_hook_rwsem); get_update_locks_for_kvm(kvm); vfio_ap_mdev_update_guest_apcb(matrix_mdev); release_update_locks_for_kvm(kvm);Thread 1 (continuing in
vfio_ap_mdev_unset_kvm):get_update_locks_for_kvm(kvm); kvm_arch_crypto_clear_masks(kvm);If Thread 1 interleaves after Thread 2 sets the guest masks but before Thread 1 clears them, would this result in Thread 1 unconditionally clearing the KVM APCB masks that Thread 2 just successfully configured, leaving the newly attached mdev with inaccessible queues?
Quoted 15 lines
drivers/s390/crypto/vfio_ap_ops.c +0
-9- -list_for_each_entry (m, &matrix_dev->mdev_list, node) { -if (m != matrix_mdev && m->kvm == kvm) { -release_update_locks_for_kvm (kvm); -return -EPERM ; - } - } -kvm_get_kvm (kvm); matrix_mdev->kvm = kvm;vfio_ap_mdev_update_guest_apcb (matrix_mdev); -release_update_locks_for_kvm (kvm); }