On Mon, 2023-12-04 at 14:43 +0000, Paul Durrant wrote:
From: Paul Durrant pdurrant@amazon.com
Taking a write lock on a pfncache will be disruptive if the cache is heavily used (which only requires a read lock). Hence, in the MMU notifier callback, take read locks on caches to check for a match; only taking a write lock to actually perform an invalidation (after a another check).
Signed-off-by: Paul Durrant pdurrant@amazon.com
Reviewed-by: David Woodhouse dwmw@amazon.co.uk
In particular, the previous 'don't block on pfncache locks in kvm_xen_set_evtchn_fast()' patch in this series is easy to justify on the basis that it only falls back to the slow path if it can't take a read lock immediately. And surely it should *always* be able to take a read lock immediately unless there's an actual *writer* — which should be a rare event, and means the cache was probably going to be invalidates anyway.
But then we realised the MMU notifier was going to disrupt that.
Cc: Sean Christopherson seanjc@google.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: David Woodhouse dwmw2@infradead.org
v10: - New in this version.
virt/kvm/pfncache.c | 22 +++++++++++++++++++--- 1 file changed, 19 insertions(+), 3 deletions(-)
diff --git a/virt/kvm/pfncache.c b/virt/kvm/pfncache.c index c2a2d1e145b6..4da16d494f4b 100644 --- a/virt/kvm/pfncache.c +++ b/virt/kvm/pfncache.c @@ -29,14 +29,30 @@ void gfn_to_pfn_cache_invalidate_start(struct kvm *kvm, unsigned long start, spin_lock(&kvm->gpc_lock); list_for_each_entry(gpc, &kvm->gpc_list, list) { - write_lock_irq(&gpc->lock); + read_lock_irq(&gpc->lock); /* Only a single page so no need to care about length */ if (gpc->valid && !is_error_noslot_pfn(gpc->pfn) && gpc->uhva >= start && gpc->uhva < end) { - gpc->valid = false; + read_unlock_irq(&gpc->lock);
+ /* + * There is a small window here where the cache could + * be modified, and invalidation would no longer be + * necessary. Hence check again whether invalidation + * is still necessary once the write lock has been + * acquired. + */
+ write_lock_irq(&gpc->lock); + if (gpc->valid && !is_error_noslot_pfn(gpc->pfn) && + gpc->uhva >= start && gpc->uhva < end) + gpc->valid = false; + write_unlock_irq(&gpc->lock); + continue; } - write_unlock_irq(&gpc->lock);
+ read_unlock_irq(&gpc->lock); } spin_unlock(&kvm->gpc_lock); }