[PATCH 03/28] KVM: x86/mmu: Remove spurious TLB flushes in TDP MMU zap collapsible path

From: Sean Christopherson
Date: Fri Nov 19 2021 - 23:51:51 EST


Drop the "flush" param and return values to/from the TDP MMU's helper for
zapping collapsible SPTEs. Because the helper runs with mmu_lock held
for read, not write, it uses tdp_mmu_zap_spte_atomic(), and the atomic
zap handles the necessary remote TLB flush.

Similarly, because mmu_lock is dropped and re-acquired between zapping
legacy MMUs and zapping TDP MMUs, kvm_mmu_zap_collapsible_sptes() must
handle remote TLB flushes from the legacy MMU before calling into the TDP
MMU.

Opportunistically drop the local "flush" variable in the common helper.

Fixes: e2209710ccc5d ("KVM: x86/mmu: Skip rmap operations if rmaps not allocated")
Signed-off-by: Sean Christopherson <seanjc@xxxxxxxxxx>
---
arch/x86/kvm/mmu/mmu.c | 9 ++-------
arch/x86/kvm/mmu/tdp_mmu.c | 22 +++++++---------------
arch/x86/kvm/mmu/tdp_mmu.h | 5 ++---
3 files changed, 11 insertions(+), 25 deletions(-)

diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 457336f67636..d2ad12a4d66e 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -5861,8 +5861,6 @@ static bool kvm_mmu_zap_collapsible_spte(struct kvm *kvm,
void kvm_mmu_zap_collapsible_sptes(struct kvm *kvm,
const struct kvm_memory_slot *slot)
{
- bool flush = false;
-
if (kvm_memslots_have_rmaps(kvm)) {
write_lock(&kvm->mmu_lock);
/*
@@ -5870,17 +5868,14 @@ void kvm_mmu_zap_collapsible_sptes(struct kvm *kvm,
* logging at a 4k granularity and never creates collapsible
* 2m SPTEs during dirty logging.
*/
- flush = slot_handle_level_4k(kvm, slot, kvm_mmu_zap_collapsible_spte, true);
- if (flush)
+ if (slot_handle_level_4k(kvm, slot, kvm_mmu_zap_collapsible_spte, true))
kvm_arch_flush_remote_tlbs_memslot(kvm, slot);
write_unlock(&kvm->mmu_lock);
}

if (is_tdp_mmu_enabled(kvm)) {
read_lock(&kvm->mmu_lock);
- flush = kvm_tdp_mmu_zap_collapsible_sptes(kvm, slot, flush);
- if (flush)
- kvm_arch_flush_remote_tlbs_memslot(kvm, slot);
+ kvm_tdp_mmu_zap_collapsible_sptes(kvm, slot);
read_unlock(&kvm->mmu_lock);
}
}
diff --git a/arch/x86/kvm/mmu/tdp_mmu.c b/arch/x86/kvm/mmu/tdp_mmu.c
index 4cd6bf7e73f0..1db8496259ad 100644
--- a/arch/x86/kvm/mmu/tdp_mmu.c
+++ b/arch/x86/kvm/mmu/tdp_mmu.c
@@ -1362,10 +1362,9 @@ void kvm_tdp_mmu_clear_dirty_pt_masked(struct kvm *kvm,
* Clear leaf entries which could be replaced by large mappings, for
* GFNs within the slot.
*/
-static bool zap_collapsible_spte_range(struct kvm *kvm,
+static void zap_collapsible_spte_range(struct kvm *kvm,
struct kvm_mmu_page *root,
- const struct kvm_memory_slot *slot,
- bool flush)
+ const struct kvm_memory_slot *slot)
{
gfn_t start = slot->base_gfn;
gfn_t end = start + slot->npages;
@@ -1376,10 +1375,8 @@ static bool zap_collapsible_spte_range(struct kvm *kvm,

tdp_root_for_each_pte(iter, root, start, end) {
retry:
- if (tdp_mmu_iter_cond_resched(kvm, &iter, flush, true)) {
- flush = false;
+ if (tdp_mmu_iter_cond_resched(kvm, &iter, false, true))
continue;
- }

if (!is_shadow_present_pte(iter.old_spte) ||
!is_last_spte(iter.old_spte, iter.level))
@@ -1391,6 +1388,7 @@ static bool zap_collapsible_spte_range(struct kvm *kvm,
pfn, PG_LEVEL_NUM))
continue;

+ /* Note, a successful atomic zap also does a remote TLB flush. */
if (!tdp_mmu_zap_spte_atomic(kvm, &iter)) {
/*
* The iter must explicitly re-read the SPTE because
@@ -1399,30 +1397,24 @@ static bool zap_collapsible_spte_range(struct kvm *kvm,
iter.old_spte = READ_ONCE(*rcu_dereference(iter.sptep));
goto retry;
}
- flush = true;
}

rcu_read_unlock();
-
- return flush;
}

/*
* Clear non-leaf entries (and free associated page tables) which could
* be replaced by large mappings, for GFNs within the slot.
*/
-bool kvm_tdp_mmu_zap_collapsible_sptes(struct kvm *kvm,
- const struct kvm_memory_slot *slot,
- bool flush)
+void kvm_tdp_mmu_zap_collapsible_sptes(struct kvm *kvm,
+ const struct kvm_memory_slot *slot)
{
struct kvm_mmu_page *root;

lockdep_assert_held_read(&kvm->mmu_lock);

for_each_tdp_mmu_root_yield_safe(kvm, root, slot->as_id, true)
- flush = zap_collapsible_spte_range(kvm, root, slot, flush);
-
- return flush;
+ zap_collapsible_spte_range(kvm, root, slot);
}

/*
diff --git a/arch/x86/kvm/mmu/tdp_mmu.h b/arch/x86/kvm/mmu/tdp_mmu.h
index 476b133544dd..3899004a5d91 100644
--- a/arch/x86/kvm/mmu/tdp_mmu.h
+++ b/arch/x86/kvm/mmu/tdp_mmu.h
@@ -64,9 +64,8 @@ void kvm_tdp_mmu_clear_dirty_pt_masked(struct kvm *kvm,
struct kvm_memory_slot *slot,
gfn_t gfn, unsigned long mask,
bool wrprot);
-bool kvm_tdp_mmu_zap_collapsible_sptes(struct kvm *kvm,
- const struct kvm_memory_slot *slot,
- bool flush);
+void kvm_tdp_mmu_zap_collapsible_sptes(struct kvm *kvm,
+ const struct kvm_memory_slot *slot);

bool kvm_tdp_mmu_write_protect_gfn(struct kvm *kvm,
struct kvm_memory_slot *slot, gfn_t gfn,
--
2.34.0.rc2.393.gf8c9666880-goog