Re: [RFC PATCH] kvm: nv: Optimize the unmapping of shadow S2-MMU tables.

From: Ganapatrao Kulkarni
Date: Tue Mar 05 2024 - 08:29:39 EST




On 05-03-2024 04:43 pm, Marc Zyngier wrote:
[re-sending with kvmarm@ fixed]

On Tue, 05 Mar 2024 05:46:06 +0000,
Ganapatrao Kulkarni <gankulkarni@xxxxxxxxxxxxxxxxxxxxxx> wrote:

As per 'commit 178a6915434c ("KVM: arm64: nv: Unmap/flush shadow stage 2

$ git describe --contains 178a6915434c --match=v\*
fatal: cannot describe '178a6915434c141edefd116b8da3d55555ea3e63'


My bad(I would have been more verbose), I missed to mention that this patch is on top of NV-V11 patch series.

This commit simply doesn't exist upstream. It only lives in a
now deprecated branch that will never be merged.

page tables")', when ever there is unmap of pages that
are mapped to L1, they are invalidated from both L1 S2-MMU and from
all the active shadow/L2 S2-MMU tables. Since there is no mapping
to invalidate the IPAs of Shadow S2 to a page, there is a complete
S2-MMU page table walk and invalidation is done covering complete
address space allocated to a L2. This has performance impacts and
even soft lockup for NV(L1 and L2) boots with higher number of
CPUs and large Memory.

Adding a lookup table of mapping of Shadow IPA to Canonical IPA
whenever a page is mapped to any of the L2. While any page is
unmaped, this lookup is helpful to unmap only if it is mapped in
any of the shadow S2-MMU tables. Hence avoids unnecessary long
iterations of S2-MMU table walk-through and invalidation for the
complete address space.

All of this falls in the "premature optimisation" bucket. Why should
we bother with any of this when not even 'AT S1' works correctly,

Hmm, I am not aware of this, is this something new issue of V11?

making it trivial to prevent a guest from making forward progress? You
also show no numbers that would hint at a measurable improvement under
any particular workload.

This patch is avoiding long iterations of unmap which was resulting in soft-lockup, when tried L1 and L2 with 192 cores.
Fixing soft lockup isn't a required fix for feature enablement?


I am genuinely puzzled that you are wasting valuable engineering time
on *this*.


Signed-off-by: Ganapatrao Kulkarni <gankulkarni@xxxxxxxxxxxxxxxxxxxxxx>
---
arch/arm64/include/asm/kvm_emulate.h | 5 ++
arch/arm64/include/asm/kvm_host.h | 14 ++++
arch/arm64/include/asm/kvm_nested.h | 4 +
arch/arm64/kvm/mmu.c | 19 ++++-
arch/arm64/kvm/nested.c | 113 +++++++++++++++++++++++++++
5 files changed, 152 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
index 5173f8cf2904..f503b2eaedc4 100644
--- a/arch/arm64/include/asm/kvm_emulate.h
+++ b/arch/arm64/include/asm/kvm_emulate.h
@@ -656,4 +656,9 @@ static inline bool kvm_is_shadow_s2_fault(struct kvm_vcpu *vcpu)
vcpu->arch.hw_mmu->nested_stage2_enabled);
}
+static inline bool kvm_is_l1_using_shadow_s2(struct kvm_vcpu *vcpu)
+{
+ return (vcpu->arch.hw_mmu != &vcpu->kvm->arch.mmu);
+}

Isn't that the very definition of "!in_hyp_ctxt()"? You are abusing

"!in_hyp_ctxt()" isn't true for non-NV case also?
This function added to know that L1 is NV enabled and using shadow S2.

the hw_mmu pointer to derive something, but the source of truth is the
translation regime, as defined by HCR_EL2.{E2H,TGE} and PSTATE.M.


OK, I can try HCR_EL2.{E2H,TGE} and PSTATE.M instead of hw_mmu in next version.

+
#endif /* __ARM64_KVM_EMULATE_H__ */
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 8da3c9a81ae3..f61c674c300a 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -144,6 +144,13 @@ struct kvm_vmid {
atomic64_t id;
};
+struct mapipa_node {
+ struct rb_node node;
+ phys_addr_t ipa;
+ phys_addr_t shadow_ipa;
+ long size;
+};
+
struct kvm_s2_mmu {
struct kvm_vmid vmid;
@@ -216,6 +223,13 @@ struct kvm_s2_mmu {
* >0: Somebody is actively using this.
*/
atomic_t refcnt;
+
+ /*
+ * For a Canonical IPA to Shadow IPA mapping.
+ */
+ struct rb_root nested_mapipa_root;

Why isn't this a maple tree? If there is no overlap between mappings
(and it really shouldn't be any), why should we use a bare-bone rb-tree?

+ rwlock_t mmu_lock;

Hell no. We have plenty of locking already, and there is no reason why
this should gain its own locking. I can't see a case where you would
take this lock outside of holding the *real* mmu_lock -- extra bonus
point for the ill-chosen name.

OK, this should be avoided with maple tree.

+
};
static inline bool kvm_s2_mmu_valid(struct kvm_s2_mmu *mmu)
diff --git a/arch/arm64/include/asm/kvm_nested.h b/arch/arm64/include/asm/kvm_nested.h
index da7ebd2f6e24..c31a59a1fdc6 100644
--- a/arch/arm64/include/asm/kvm_nested.h
+++ b/arch/arm64/include/asm/kvm_nested.h
@@ -65,6 +65,9 @@ extern void kvm_init_nested(struct kvm *kvm);
extern int kvm_vcpu_init_nested(struct kvm_vcpu *vcpu);
extern void kvm_init_nested_s2_mmu(struct kvm_s2_mmu *mmu);
extern struct kvm_s2_mmu *lookup_s2_mmu(struct kvm_vcpu *vcpu);
+extern void add_shadow_ipa_map_node(
+ struct kvm_s2_mmu *mmu,
+ phys_addr_t ipa, phys_addr_t shadow_ipa, long size);
union tlbi_info;
@@ -123,6 +126,7 @@ extern int kvm_s2_handle_perm_fault(struct kvm_vcpu *vcpu,
extern int kvm_inject_s2_fault(struct kvm_vcpu *vcpu, u64 esr_el2);
extern void kvm_nested_s2_wp(struct kvm *kvm);
extern void kvm_nested_s2_unmap(struct kvm *kvm);
+extern void kvm_nested_s2_unmap_range(struct kvm *kvm, struct kvm_gfn_range *range);
extern void kvm_nested_s2_flush(struct kvm *kvm);
int handle_wfx_nested(struct kvm_vcpu *vcpu, bool is_wfe);
diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c
index 61bdd8798f83..3948681426a0 100644
--- a/arch/arm64/kvm/mmu.c
+++ b/arch/arm64/kvm/mmu.c
@@ -1695,6 +1695,13 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
memcache,
KVM_PGTABLE_WALK_HANDLE_FAULT |
KVM_PGTABLE_WALK_SHARED);
+ if ((nested || kvm_is_l1_using_shadow_s2(vcpu)) && !ret) {

I don't understand this condition. If nested is non-NULL, it's because
we're using a shadow S2. So why the additional condition?

No, nested is set only for L2, for L1 it is not.
To handle L1 shadow S2 case, I have added this condition.


+ struct kvm_s2_mmu *shadow_s2_mmu;
+
+ ipa &= ~(vma_pagesize - 1);
+ shadow_s2_mmu = lookup_s2_mmu(vcpu);
+ add_shadow_ipa_map_node(shadow_s2_mmu, ipa, fault_ipa, vma_pagesize);
+ }
}
/* Mark the page dirty only if the fault is handled successfully */
@@ -1918,7 +1925,7 @@ bool kvm_unmap_gfn_range(struct kvm *kvm, struct kvm_gfn_range *range)
(range->end - range->start) << PAGE_SHIFT,
range->may_block);
- kvm_nested_s2_unmap(kvm);
+ kvm_nested_s2_unmap_range(kvm, range);
return false;
}
@@ -1953,7 +1960,7 @@ bool kvm_set_spte_gfn(struct kvm *kvm, struct kvm_gfn_range *range)
PAGE_SIZE, __pfn_to_phys(pfn),
KVM_PGTABLE_PROT_R, NULL, 0);
- kvm_nested_s2_unmap(kvm);
+ kvm_nested_s2_unmap_range(kvm, range);
return false;
}
@@ -2223,12 +2230,18 @@ void kvm_arch_memslots_updated(struct kvm *kvm, u64 gen)
void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
struct kvm_memory_slot *slot)
{
+ struct kvm_gfn_range range;
+
gpa_t gpa = slot->base_gfn << PAGE_SHIFT;
phys_addr_t size = slot->npages << PAGE_SHIFT;
+ range.start = gpa;
+ range.end = gpa + size;
+ range.may_block = true;
+
write_lock(&kvm->mmu_lock);
kvm_unmap_stage2_range(&kvm->arch.mmu, gpa, size);
- kvm_nested_s2_unmap(kvm);
+ kvm_nested_s2_unmap_range(kvm, &range);
write_unlock(&kvm->mmu_lock);
}
diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c
index f88d9213c6b3..888ec9fba4a0 100644
--- a/arch/arm64/kvm/nested.c
+++ b/arch/arm64/kvm/nested.c
@@ -565,6 +565,88 @@ void kvm_s2_mmu_iterate_by_vmid(struct kvm *kvm, u16 vmid,
write_unlock(&kvm->mmu_lock);
}
+/*
+ * Create a node and add to lookup table, when a page is mapped to
+ * Canonical IPA and also mapped to Shadow IPA.
+ */
+void add_shadow_ipa_map_node(struct kvm_s2_mmu *mmu,
+ phys_addr_t ipa,
+ phys_addr_t shadow_ipa, long size)
+{
+ struct rb_root *ipa_root = &(mmu->nested_mapipa_root);
+ struct rb_node **node = &(ipa_root->rb_node), *parent = NULL;
+ struct mapipa_node *new;
+
+ new = kzalloc(sizeof(struct mapipa_node), GFP_KERNEL);
+ if (!new)
+ return;
+
+ new->shadow_ipa = shadow_ipa;
+ new->ipa = ipa;
+ new->size = size;
+
+ write_lock(&mmu->mmu_lock);
+
+ while (*node) {
+ struct mapipa_node *tmp;
+
+ tmp = container_of(*node, struct mapipa_node, node);
+ parent = *node;
+ if (new->ipa < tmp->ipa) {
+ node = &(*node)->rb_left;
+ } else if (new->ipa > tmp->ipa) {
+ node = &(*node)->rb_right;
+ } else {
+ write_unlock(&mmu->mmu_lock);
+ kfree(new);
+ return;
+ }
+ }
+
+ rb_link_node(&new->node, parent, node);
+ rb_insert_color(&new->node, ipa_root);
+ write_unlock(&mmu->mmu_lock);

All this should be removed in favour of simply using a maple tree.


Thanks for the suggestion to use maple tree. I will use it in next version, which help to avoid the locks.

+}
+
+/*
+ * Iterate over the lookup table of Canonical IPA to Shadow IPA.
+ * Return Shadow IPA, if the page mapped to Canonical IPA is
+ * also mapped to a Shadow IPA.
+ *
+ */
+bool get_shadow_ipa(struct kvm_s2_mmu *mmu, phys_addr_t ipa, phys_addr_t *shadow_ipa, long *size)

static?

It should be, thanks.

+{
+ struct rb_node *node;
+ struct mapipa_node *tmp = NULL;
+
+ read_lock(&mmu->mmu_lock);
+ node = mmu->nested_mapipa_root.rb_node;
+
+ while (node) {
+ tmp = container_of(node, struct mapipa_node, node);
+
+ if (tmp->ipa == ipa)

What guarantees that the mapping you have for L1 has the same starting
address as the one you have for L2? L1 could have a 2MB mapping and L2
only 4kB *in the middle*.

IIUC, when a page is mapped to 2MB in L1, it won't be
mapped to L2 and we iterate with the step of PAGE_SIZE and we should be hitting the L2's IPA in lookup table, provided the L2 page falls in unmap range.


+ break;
+ else if (ipa > tmp->ipa)
+ node = node->rb_right;
+ else
+ node = node->rb_left;
+ }
+
+ read_unlock(&mmu->mmu_lock);

Why would you drop the lock here....

+
+ if (tmp && tmp->ipa == ipa) {
+ *shadow_ipa = tmp->shadow_ipa;
+ *size = tmp->size;
+ write_lock(&mmu->mmu_lock);

... if taking it again here? What could have changed in between?

+ rb_erase(&tmp->node, &mmu->nested_mapipa_root);
+ write_unlock(&mmu->mmu_lock);
+ kfree(tmp);
+ return true;
+ }
+ return false;
+}

So simply hitting in the reverse mapping structure *frees* it? Meaning
that you cannot use it as a way to update a mapping?

Freeing it since this page already unmapped/migrated on host and will be done on shadow S2 after this lookup. I should have considered other cases as well, as Oliver mentioned.


+
/* Must be called with kvm->mmu_lock held */
struct kvm_s2_mmu *lookup_s2_mmu(struct kvm_vcpu *vcpu)
{
@@ -674,6 +756,7 @@ void kvm_init_nested_s2_mmu(struct kvm_s2_mmu *mmu)
mmu->tlb_vttbr = 1;
mmu->nested_stage2_enabled = false;
atomic_set(&mmu->refcnt, 0);
+ mmu->nested_mapipa_root = RB_ROOT;
}
void kvm_vcpu_load_hw_mmu(struct kvm_vcpu *vcpu)
@@ -760,6 +843,36 @@ void kvm_nested_s2_unmap(struct kvm *kvm)
}
}
+void kvm_nested_s2_unmap_range(struct kvm *kvm, struct kvm_gfn_range *range)
+{
+ int i;
+ long size;
+ bool ret;
+
+ for (i = 0; i < kvm->arch.nested_mmus_size; i++) {
+ struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i];
+
+ if (kvm_s2_mmu_valid(mmu)) {
+ phys_addr_t shadow_ipa, start, end;
+
+ start = range->start << PAGE_SHIFT;
+ end = range->end << PAGE_SHIFT;
+
+ while (start < end) {
+ size = PAGE_SIZE;
+ /*
+ * get the Shadow IPA if the page is mapped
+ * to L1 and also mapped to any of active L2.
+ */

Why is L1 relevant here?

We do map while L1 boots(early stage) in shadow S2, at that moment
if the L1 mapped page is unmapped/migrated we do need to unmap from L1's S2 table also.


+ ret = get_shadow_ipa(mmu, start, &shadow_ipa, &size);
+ if (ret)
+ kvm_unmap_stage2_range(mmu, shadow_ipa, size);
+ start += size;
+ }
+ }
+ }
+}
+
/* expects kvm->mmu_lock to be held */
void kvm_nested_s2_flush(struct kvm *kvm)
{

There are a bunch of worrying issues with this patch. But more
importantly, this looks like a waste of effort until the core issues
that NV still has are solved, and I will not consider anything of the
sort until then.

OK thanks for letting us know, I will pause the work on V2 of this patch until then.


I get the ugly feeling that you are trying to make it look as if it
was "production ready", which it won't be for another few years,
specially if the few interested people (such as you) are ignoring the
core issues in favour of marketing driven features ("make it fast").


What are the core issues (please forgive me if you mentioned already)? certainly we will prioritise them than this.

Thanks,

M.


Thanks,
Ganapat