KVM: arm64: Introduce hyp_assert_lock_held()

Introduce a poor man's lockdep implementation at EL2 which allows to
BUG() whenever a hyp spinlock is not held when it should. Hide this
feature behind a new Kconfig option that targets the EL2 object
specifically, instead of piggy backing on the existing CONFIG_LOCKDEP.
EL2 cannot WARN() cleanly to report locking issues, hence BUG() is the
only option and it is not clear whether we want this widely enabled.
This is most likely going to be useful for local testing until the EL2
WARN() situation has improved.

Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210809152448.1810400-3-qperret@google.com
This commit is contained in:
Quentin Perret 2021-08-09 16:24:29 +01:00 committed by Marc Zyngier
parent d21292f13f
commit 8e049e0daf
2 changed files with 26 additions and 0 deletions

View file

@ -46,6 +46,15 @@ if KVM
source "virt/kvm/Kconfig"
config NVHE_EL2_DEBUG
bool "Debug mode for non-VHE EL2 object"
help
Say Y here to enable the debug mode for the non-VHE KVM EL2 object.
Failure reports will BUG() in the hypervisor. This is intended for
local EL2 hypervisor development.
If unsure, say N.
endif # KVM
endif # VIRTUALIZATION

View file

@ -97,4 +97,21 @@ static inline bool hyp_spin_is_locked(hyp_spinlock_t *lock)
return lockval.owner != lockval.next;
}
#ifdef CONFIG_NVHE_EL2_DEBUG
static inline void hyp_assert_lock_held(hyp_spinlock_t *lock)
{
/*
* The __pkvm_init() path accesses protected data-structures without
* holding locks as the other CPUs are guaranteed to not enter EL2
* concurrently at this point in time. The point by which EL2 is
* initialized on all CPUs is reflected in the pkvm static key, so
* wait until it is set before checking the lock state.
*/
if (static_branch_likely(&kvm_protected_mode_initialized))
BUG_ON(!hyp_spin_is_locked(lock));
}
#else
static inline void hyp_assert_lock_held(hyp_spinlock_t *lock) { }
#endif
#endif /* __ARM64_KVM_NVHE_SPINLOCK_H__ */