@@ -5771,7 +5771,7 @@ will be initialized to 1 when created. This also improves performance because
dirty logging can be enabled gradually in small chunks on the first call
to KVM_CLEAR_DIRTY_LOG. KVM_DIRTY_LOG_INITIALLY_SET depends on
KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (it is also only available on
-x86 for now).
+x86 and arm64 for now).
KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name
KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make
@@ -358,8 +358,6 @@ static inline bool kvm_arm_is_pvtime_enabled(struct kvm_vcpu_arch *vcpu_arch)
return false;
}
-void kvm_mmu_wp_memory_region(struct kvm *kvm, int slot);
-
struct kvm_vcpu *kvm_mpidr_to_vcpu(struct kvm *kvm, unsigned long mpidr);
static inline bool kvm_arch_requires_vhe(void) { return false; }
@@ -16,6 +16,7 @@
#include <linux/jump_label.h>
#include <linux/kvm_types.h>
#include <linux/percpu.h>
+#include <linux/kvm.h>
#include <asm/arch_gicv3.h>
#include <asm/barrier.h>
#include <asm/cpufeature.h>
@@ -45,6 +46,9 @@
#define KVM_REQ_VCPU_RESET KVM_ARCH_REQ(2)
#define KVM_REQ_RECORD_STEAL KVM_ARCH_REQ(3)
+#define KVM_DIRTY_LOG_MANUAL_CAPS (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | \
+ KVM_DIRTY_LOG_INITIALLY_SET)
+
DECLARE_STATIC_KEY_FALSE(userspace_irqchip_in_use);
extern unsigned int kvm_sve_max_vl;
@@ -478,7 +482,6 @@ u64 __kvm_call_hyp(void *hypfn, ...);
})
void force_vm_exit(const cpumask_t *mask);
-void kvm_mmu_wp_memory_region(struct kvm *kvm, int slot);
int handle_exit(struct kvm_vcpu *vcpu, struct kvm_run *run,
int exception_index);
@@ -1530,7 +1530,7 @@ static void stage2_wp_range(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
* Acquires kvm_mmu_lock. Called with kvm->slots_lock mutex acquired,
* serializing operations for VM memory regions.
*/
-void kvm_mmu_wp_memory_region(struct kvm *kvm, int slot)
+static void kvm_mmu_wp_memory_region(struct kvm *kvm, int slot)
{
struct kvm_memslots *slots = kvm_memslots(kvm);
struct kvm_memory_slot *memslot = id_to_memslot(slots, slot);
@@ -2265,8 +2265,18 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
* allocated dirty_bitmap[], dirty pages will be be tracked while the
* memory slot is write protected.
*/
- if (change != KVM_MR_DELETE && mem->flags & KVM_MEM_LOG_DIRTY_PAGES)
- kvm_mmu_wp_memory_region(kvm, mem->slot);
+ if (change != KVM_MR_DELETE && mem->flags & KVM_MEM_LOG_DIRTY_PAGES) {
+ if (!kvm_dirty_log_manual_protect_and_init_set(kvm)) {
+ kvm_mmu_wp_memory_region(kvm, mem->slot);
+ } else {
+ /*
+ * If we're with initial-all-set, we don't need to
+ * write protect any pages because they're reported
+ * as dirty here.
+ */
+ bitmap_set(new->dirty_bitmap, 0, new->npages);
+ }
+ }
}
int kvm_arch_prepare_memory_region(struct kvm *kvm,
There is already support of enabling dirty log gradually in small chunks for x86. This adds support for arm64. However, unlike x86, all pages, including huge pages and normal pages, don't need to be write protected when enabling dirty log if KVM_DIRTY_LOG_INITIALLY_SET is eanbled. For that both huge pages and normal pages can be write protected later by userspace manually. Under the Huawei Kunpeng 920 2.6GHz platform, I did some tests on 128G Linux VMs with different page size. The memory pressure is 127G in each case. The time taken of memory_global_dirty_log_start in QEMU is listed below: Page Size Before After Optimization 4K 650ms 1.8ms 2M 4ms 1.8ms 1G 2ms 1.8ms Signed-off-by: Keqian Zhu <zhukeqian1@huawei.com> --- Documentation/virt/kvm/api.rst | 2 +- arch/arm/include/asm/kvm_host.h | 2 -- arch/arm64/include/asm/kvm_host.h | 5 ++++- virt/kvm/arm/mmu.c | 16 +++++++++++++--- 4 files changed, 18 insertions(+), 7 deletions(-)