KVM: x86/mmu: Add comment on try_cmpxchg64 usage in tdp_mmu_set_spte_atomic
PATCH 1/1 KVM: x86/mmu: Add comment on try_cmpxchg64 usage in tdp_mmu_set_spte_atomic
Commit aee98a6838d5 ("KVM: x86/mmu: Use try_cmpxchg64 in tdp_mmu_set_spte_atomic") removed the comment that iter->old_spte is updated when different logical CPU modifies the page table entry. Although this is what try_cmpxchg does implicitly, it won't hurt if this fact is explicitly mentioned in a restored comment.
--- arch/x86/kvm/mmu/tdp_mmu.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
arch/x86/kvm/mmu/tdp_mmu.c +9 -1
@@ -655,8 +655,16 @@ static inline int tdp_mmu_set_spte_atomic(struct kvm *kvm,* Note, fast_pf_fix_direct_spte() can also modify TDP MMU SPTEs and * does not hold the mmu_lock. */ -if (!try_cmpxchg64 (sptep, &iter->old_spte, new_spte)) +if (!try_cmpxchg64 (sptep, &iter->old_spte, new_spte)) { +/* +* The page table entry was modified by a different logical +* CPU. In this case the above try_cmpxchg updates +* iter->old_spte with the current value, so the caller +* operates on fresh data, e.g. if it retries +* tdp_mmu_set_spte_atomic(). +*/ return -EBUSY ; + }__handle_changed_spte (kvm, iter->as_id, iter->gfn, iter->old_spte, new_spte, iter->level, true);
Signature · 1 line
2.40.0
Sean Christopherson
maintainerQuoted 21 lines · Uros Bizjak
Commit
aee98a6838d5("KVM: x86/mmu: Usetry_cmpxchg64intdp_mmu_set_spte_atomic") removed the comment thatiter->old_spteis updated when different logical CPU modifies the page table entry. Although this is whattry_cmpxchgdoes implicitly, it won't hurt if this fact is explicitly mentioned in a restored comment.Cc Paolo Bonzini <[email protected]>Cc Sean Christopherson <[email protected]>Cc David Matlack <[email protected]>Signed-off-by Uros Bizjak <[email protected]>--- arch/x86/kvm/mmu/tdp_mmu.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)arch/x86/kvm/mmu/tdp_mmu.c +0
-0@@ -655,8 +655,16 @@ static inline int tdp_mmu_set_spte_atomic(struct kvm *kvm,* Note, fast_pf_fix_direct_spte() can also modify TDP MMU SPTEs and * does not hold the mmu_lock.
arch/x86/kvm/mmu/tdp_mmu.c static inline int tdp_mmu_set_spte_atomic(struct kvm *kvm, · line 665
Quoted 10 lines
arch/x86/kvm/mmu/tdp_mmu.c +8
-1*/ -if (!try_cmpxchg64 (sptep, &iter->old_spte, new_spte)) +if (!try_cmpxchg64 (sptep, &iter->old_spte, new_spte)) { +/* +* The page table entry was modified by a different logical +* CPU. In this case the above try_cmpxchg updates +* iter->old_spte with the current value, so the caller +* operates on fresh data, e.g. if it retries +* tdp_mmu_set_spte_atomic(). +*/
If there's no objection, when applying I'll massage this to extend the comment above the try_cmpxchg64(), e.g.
/*
* Note, fast_pf_fix_direct_spte() can also modify TDP MMU SPTEs and
* does not hold the mmu_lock. On failure, i.e. if a different logical
* CPU modified the SPTE, try_cmpxchg64() updates iter->old_spte with
* the current value, so the caller operates on fresh data, e.g. if it
* retries tdp_mmu_set_spte_atomic()
*/
if (!try_cmpxchg64(sptep, &iter->old_spte, new_spte))
return -EBUSY;
Quoted 8 lines · Uros Bizjak
Commit
aee98a6838d5("KVM: x86/mmu: Usetry_cmpxchg64intdp_mmu_set_spte_atomic") removed the comment thatiter->old_spteis updated when different logical CPU modifies the page table entry. Although this is whattry_cmpxchgdoes implicitly, it won't hurt if this fact is explicitly mentioned in a restored comment.[...]
Applied to kvm-x86 mmu, thanks!
[1/1] KVM: x86/mmu: Add comment on try_cmpxchg64 usage in tdp_mmu_set_spte_atomic
https://github.com/kvm-x86/linux/commit/12ced095956a
--
https://github.com/kvm-x86/linux/tree/next
https://github.com/kvm-x86/linux/tree/fixes