KVM: x86/mmu: Always drop mmu_lock to allocate TDP MMU SPs for eager splitting
PATCH v3 1/1 KVM: x86/mmu: Always drop mmu_lock to allocate TDP MMU SPs for eager splitting
Always drop mmu_lock to allocate shadow pages in the TDP MMU when doing eager page splitting. Dropping mmu_lock during eager page splitting is cheap since KVM does not have to flush remote TLBs, and avoids stalling vCPU threads that are taking page faults in parallel.
This change reduces 20%+ dips in MySQL throughput during live migration in a 160 vCPU VM while userspace is issuing CLEAR_DIRTY_LOG ioctls (tested with 1GiB and 8GiB CLEARs). Userspace could issue finer-grained CLEARs, which would also reduce contention on mmu_lock, but doing so will increase the rate of remote TLB flushing (KVM must flush TLBs before returning from CLEAR_DITY_LOG).
When there isn't contention on mmu_lock[1], this change does not regress the time it takes to perform eager page splitting.
[1] Tested with dirty_log_perf_test, which does not run vCPUs during eager page splitting, and with a 16 vCPU VM Live Migration with manual-protect disabled (where mmu_lock is held in read-mode).
---
v3:
- Only drop mmu_lock during TDP MMU eager page splitting. This fixes the perfomance regression without regressing CLEAR_DIRTY_LOG performance on other architectures from frequent lock dropping.
v2: https://lore.kernel.org/kvm/[email protected]/
- Rebase onto kvm/queue [Marc]
v1: https://lore.kernel.org/kvm/[email protected]/
arch/x86/kvm/mmu/tdp_mmu.c | 22 ++++------------------ 1 file changed, 4 insertions(+), 18 deletions(-)
arch/x86/kvm/mmu/tdp_mmu.c +4 -18
@@ -1385,11 +1385,11 @@ bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm,return spte_set; } -static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (gfp_t gfp) +static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (void ) { +gfp_t gfp =GFP_KERNEL_ACCOUNT | __GFP_ZERO;struct kvm_mmu_page *sp; - gfp |= __GFP_ZERO; sp =kmem_cache_alloc (mmu_page_header_cache, gfp);if (!sp)
@@ -1412,19 +1412,6 @@ static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm,kvm_lockdep_assert_mmu_lock_held (kvm, shared); -/* -* Since we are allocating while under the MMU lock we have to be -* careful about GFP flags. Use GFP_NOWAIT to avoid blocking on direct -* reclaim and to avoid making any filesystem callbacks (which can end -* up invoking KVM MMU notifiers, resulting in a deadlock). -* -* If this allocation fails we drop the lock and retry with reclaim -* allowed. -*/ - sp =__tdp_mmu_alloc_sp_for_split (GFP_NOWAIT | __GFP_ACCOUNT); -if (sp) -return sp; -rcu_read_unlock ();if (shared)
@@ -1433,7 +1420,7 @@ static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm,write_unlock (&kvm->mmu_lock); iter->yielded = true; - sp =__tdp_mmu_alloc_sp_for_split (GFP_KERNEL_ACCOUNT ); + sp =__tdp_mmu_alloc_sp_for_split ();if (shared)read_lock (&kvm->mmu_lock);
@@ -1524,8 +1511,7 @@ static int tdp_mmu_split_huge_pages_root(struct kvm *kvm,break ; } -if (iter.yielded) -continue ; +continue ; }tdp_mmu_init_child_sp (sp, &iter);
base-commit: 2d181d84af38146748042a6974c577fc46c3f1c3
Signature · 1 line
2.45.0.118.g7fe29c98d7-goog
Sean Christopherson
maintainerQuoted 43 lines · David Matlack
Always drop
mmu_lockto allocate shadow pages in the TDP MMU when doing eager page splitting. Droppingmmu_lockduring eager page splitting is cheap since KVM does not have to flush remote TLBs, and avoids stalling vCPU threads that are taking page faults in parallel.This change reduces 20%+ dips in MySQL throughput during live migration in a 160 vCPU VM while userspace is issuing
CLEAR_DIRTY_LOGioctls (tested with 1GiB and 8GiB CLEARs). Userspace could issue finer-grained CLEARs, which would also reduce contention onmmu_lock, but doing so will increase the rate of remote TLB flushing (KVM must flush TLBs before returning fromCLEAR_DITY_LOG).When there isn't contention on mmu_lock[1], this change does not regress the time it takes to perform eager page splitting.
[1] Tested with
dirty_log_perf_test, which does not run vCPUs during eager page splitting, and with a 16 vCPU VM Live Migration with manual-protect disabled (wheremmu_lockis held in read-mode).Cc Bibo Mao <[email protected]>Cc Sean Christopherson <[email protected]>Signed-off-by David Matlack <[email protected]>---
v3:
- Only dropmmu_lockduring TDP MMU eager page splitting. This fixes the perfomance regression without regressingCLEAR_DIRTY_LOGperformance on other architectures from frequent lock dropping.v2: https://lore.kernel.org/kvm/[email protected]/
- Rebase onto kvm/queue [Marc]v1: https://lore.kernel.org/kvm/[email protected]/
arch/x86/kvm/mmu/tdp_mmu.c | 22 ++++------------------ 1 file changed, 4 insertions(+), 18 deletions(-)arch/x86/kvm/mmu/tdp_mmu.c +0
-0@@ -1385,11 +1385,11 @@ bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm,return spte_set; }
arch/x86/kvm/mmu/tdp_mmu.c bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm, · line 1394
Quoted 10 lines
arch/x86/kvm/mmu/tdp_mmu.c +2
-2-static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (gfp_t gfp) +static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (void ) { +gfp_t gfp =GFP_KERNEL_ACCOUNT | __GFP_ZERO;struct kvm_mmu_page *sp; - gfp |= __GFP_ZERO; sp =kmem_cache_alloc (mmu_page_header_cache, gfp);
This can more simply and cleary be:
sp = kmem_cache_zalloc(mmu_page_header_cache, GFP_KERNEL_ACCOUNT);
Quoted 22 lines
arch/x86/kvm/mmu/tdp_mmu.c +0
-13if (!sp)@@ -1412,19 +1412,6 @@ static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm,kvm_lockdep_assert_mmu_lock_held (kvm, shared); -/* -* Since we are allocating while under the MMU lock we have to be -* careful about GFP flags. Use GFP_NOWAIT to avoid blocking on direct -* reclaim and to avoid making any filesystem callbacks (which can end -* up invoking KVM MMU notifiers, resulting in a deadlock). -* -* If this allocation fails we drop the lock and retry with reclaim -* allowed. -*/ - sp =__tdp_mmu_alloc_sp_for_split (GFP_NOWAIT | __GFP_ACCOUNT); -if (sp) -return sp; -rcu_read_unlock ();if (shared)
arch/x86/kvm/mmu/tdp_mmu.c static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm, · line 1422
Quoted 3 lines
arch/x86/kvm/mmu/tdp_mmu.c +0
-0@@ -1433,7 +1420,7 @@ static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm,write_unlock (&kvm->mmu_lock); iter->yielded = true;
Now that yielding is unconditional, this really should be in the loop itself. The bare continue looks weird, and it's unnecessarily hard to see that "yielded" is being set.
And there's definitely no reason to have two helpers.
Not sure how many patches it'll take, but I think we should end up with:
static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(void)
{
struct kvm_mmu_page *sp;
sp = kmem_cache_zalloc(mmu_page_header_cache, GFP_KERNEL_ACCOUNT);
if (!sp)
return NULL;
sp->spt = (void *)__get_free_page(gfp);
if (!sp->spt) {
kmem_cache_free(mmu_page_header_cache, sp);
return NULL;
}
return sp;
}static int tdp_mmu_split_huge_pages_root(struct kvm *kvm,
struct kvm_mmu_page *root, gfn_t start, gfn_t end, int target_level, bool shared)
{
struct kvm_mmu_page *sp = NULL;
struct tdp_iter iter;
rcu_read_lock();
/*
* Traverse the page table splitting all huge pages above the target
* level into one lower level. For example, if we encounter a 1GB page
* we split it into 512 2MB pages.
*
* Since the TDP iterator uses a pre-order traversal, we are guaranteed
* to visit an SPTE before ever visiting its children, which means we
* will correctly recursively split huge pages that are more than one
* level above the target level (e.g. splitting a 1GB to 512 2MB pages,
* and then splitting each of those to 512 4KB pages).
*/
for_each_tdp_pte_min_level(iter, root, target_level + 1, start, end) {
retry:
if (tdp_mmu_iter_cond_resched(kvm, &iter, false, shared))
continue;
if (!is_shadow_present_pte(iter.old_spte) || !is_large_pte(iter.old_spte))
continue;
if (!sp) {
rcu_read_unlock();
if (shared)
read_unlock(&kvm->mmu_lock);
else
write_unlock(&kvm->mmu_lock);
sp = tdp_mmu_alloc_sp_for_split(kvm, &iter, shared);
if (shared)
read_lock(&kvm->mmu_lock);
else
write_lock(&kvm->mmu_lock);
if (!sp) {
trace_kvm_mmu_split_huge_page(iter.gfn,
iter.old_spte,
iter.level, -ENOMEM);
return -ENOMEM;
}
rcu_read_lock();
iter->yielded = true;
continue;
}
tdp_mmu_init_child_sp(sp, &iter);
if (tdp_mmu_split_huge_page(kvm, &iter, sp, shared))
goto retry;
sp = NULL;
}
rcu_read_unlock();
/*
* It's possible to exit the loop having never used the last sp if, for
* example, a vCPU doing HugePage NX splitting wins the race and
* installs its own sp in place of the last sp we tried to split.
*/
if (sp)
tdp_mmu_free_sp(sp);
return 0;
}Quoted 15 lines · Sean Christopherson
Quoted 14 lines · David Matlack
arch/x86/kvm/mmu/tdp_mmu.c +2
-1@@ -1385,11 +1385,11 @@ bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm,return spte_set; } -static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (gfp_t gfp) +static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (void ) { +gfp_t gfp =GFP_KERNEL_ACCOUNT | __GFP_ZERO;struct kvm_mmu_page *sp;
Quoted 7 lines
arch/x86/kvm/mmu/tdp_mmu.c bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm, · line 1394
Quoted 3 lines
arch/x86/kvm/mmu/tdp_mmu.c +0
-1- gfp |= __GFP_ZERO; sp =kmem_cache_alloc (mmu_page_header_cache, gfp);This can more simply and cleary be:
sp = kmem_cache_zalloc(mmu_page_header_cache, GFP_KERNEL_ACCOUNT);
Will do. And I assume you'd prefer get_zeroed_page(GFP_KERNEL_ACCOUNT) as well below?
Quoted 121 lines
Quoted 23 lines
arch/x86/kvm/mmu/tdp_mmu.c +0
-13if (!sp)@@ -1412,19 +1412,6 @@ static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm,kvm_lockdep_assert_mmu_lock_held (kvm, shared); -/* -* Since we are allocating while under the MMU lock we have to be -* careful about GFP flags. Use GFP_NOWAIT to avoid blocking on direct -* reclaim and to avoid making any filesystem callbacks (which can end -* up invoking KVM MMU notifiers, resulting in a deadlock). -* -* If this allocation fails we drop the lock and retry with reclaim -* allowed. -*/ - sp =__tdp_mmu_alloc_sp_for_split (GFP_NOWAIT | __GFP_ACCOUNT); -if (sp) -return sp; -rcu_read_unlock ();if (shared)arch/x86/kvm/mmu/tdp_mmu.c static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm, · line 1422
Quoted 3 lines
arch/x86/kvm/mmu/tdp_mmu.c +0
-0@@ -1433,7 +1420,7 @@ static struct kvm_mmu_page *tdp_mmu_alloc_sp_for_split(struct kvm *kvm,write_unlock (&kvm->mmu_lock); iter->yielded = true;Now that yielding is unconditional, this really should be in the loop itself. The bare continue looks weird, and it's unnecessarily hard to see that "yielded" is being set.
And there's definitely no reason to have two helpers.
Not sure how many patches it'll take, but I think we should end up with:
static struct
kvm_mmu_page*tdp_mmu_alloc_sp_for_split(void){ struct kvm_mmu_page *sp; sp = kmem_cache_zalloc(mmu_page_header_cache, GFP_KERNEL_ACCOUNT); if (!sp) return NULL; sp->spt = (void *)__get_free_page(gfp); if (!sp->spt) { kmem_cache_free(mmu_page_header_cache, sp); return NULL; } return sp; }static int tdp_mmu_split_huge_pages_root(struct kvm *kvm,
structkvm_mmu_page*root,gfn_tstart,gfn_tend, inttarget_level, bool shared)
{
structkvm_mmu_page*sp = NULL;
structtdp_iteriter;
rcu_read_lock();/*
* Traverse the page table splitting all huge pages above the target
* level into one lower level. For example, if we encounter a 1GB page
* we split it into 512 2MB pages.
*
* Since the TDP iterator uses a pre-order traversal, we are guaranteed
* to visit an SPTE before ever visiting its children, which means we
* will correctly recursively split huge pages that are more than one
* level above the target level (e.g. splitting a 1GB to 512 2MB pages,
* and then splitting each of those to 512 4KB pages).
*/
for_each_tdp_pte_min_level(iter, root, target_level + 1, start, end){
retry:if (tdp_mmu_iter_cond_resched(kvm, &iter, false, shared)) continue; if (!is_shadow_present_pte(iter.old_spte) || !is_large_pte(iter.old_spte)) continue; if (!sp) { rcu_read_unlock(); if (shared) read_unlock(&kvm->mmu_lock); else write_unlock(&kvm->mmu_lock); sp = tdp_mmu_alloc_sp_for_split(kvm, &iter, shared); if (shared) read_lock(&kvm->mmu_lock); else write_lock(&kvm->mmu_lock); if (!sp) { trace_kvm_mmu_split_huge_page(iter.gfn, iter.old_spte, iter.level, -ENOMEM); return -ENOMEM; } rcu_read_lock(); iter->yielded = true; continue; } tdp_mmu_init_child_sp(sp, &iter); if (tdp_mmu_split_huge_page(kvm, &iter, sp, shared)) goto retry; sp = NULL; } rcu_read_unlock(); /* * It's possible to exit the loop having never used the last sp if, for
Quoted 8 lines
* example, a vCPU doing HugePage NX splitting wins the race and * installs its own sp in place of the last sp we tried to split. */ if (sp) tdp_mmu_free_sp(sp); return 0; }
Ack, will do.
Sean Christopherson
maintainerQuoted 18 lines · David Matlack
Quoted 17 lines · Sean Christopherson
Quoted 16 lines · David Matlack
arch/x86/kvm/mmu/tdp_mmu.c +2
-2@@ -1385,11 +1385,11 @@ bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm,return spte_set; } -static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (gfp_t gfp) +static struct kvm_mmu_page *__tdp_mmu_alloc_sp_for_split (void ) { +gfp_t gfp =GFP_KERNEL_ACCOUNT | __GFP_ZERO;struct kvm_mmu_page *sp; - gfp |= __GFP_ZERO;
Quoted 8 lines
Quoted 5 lines
arch/x86/kvm/mmu/tdp_mmu.c bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm, · line 1394
Quoted 1 line
arch/x86/kvm/mmu/tdp_mmu.c +0
-0sp =kmem_cache_alloc (mmu_page_header_cache, gfp);This can more simply and cleary be:
sp = kmem_cache_zalloc(mmu_page_header_cache, GFP_KERNEL_ACCOUNT);Will do. And I assume you'd prefer
get_zeroed_page(GFP_KERNEL_ACCOUNT)as well below?
Ah, yeah, good catch!