1KVM Lock Overview 2================= 3 41. Acquisition Orders 5--------------------- 6 7The acquisition orders for mutexes are as follows: 8 9- kvm->lock is taken outside vcpu->mutex 10 11- kvm->lock is taken outside kvm->slots_lock and kvm->irq_lock 12 13- kvm->slots_lock is taken outside kvm->irq_lock, though acquiring 14 them together is quite rare. 15 16On x86, vcpu->mutex is taken outside kvm->arch.hyperv.hv_lock. 17 18For spinlocks, kvm_lock is taken outside kvm->mmu_lock. 19 20Everything else is a leaf: no other lock is taken inside the critical 21sections. 22 232: Exception 24------------ 25 26Fast page fault: 27 28Fast page fault is the fast path which fixes the guest page fault out of 29the mmu-lock on x86. Currently, the page fault can be fast in one of the 30following two cases: 31 321. Access Tracking: The SPTE is not present, but it is marked for access 33tracking i.e. the SPTE_SPECIAL_MASK is set. That means we need to 34restore the saved R/X bits. This is described in more detail later below. 35 362. Write-Protection: The SPTE is present and the fault is 37caused by write-protect. That means we just need to change the W bit of the 38spte. 39 40What we use to avoid all the race is the SPTE_HOST_WRITEABLE bit and 41SPTE_MMU_WRITEABLE bit on the spte: 42- SPTE_HOST_WRITEABLE means the gfn is writable on host. 43- SPTE_MMU_WRITEABLE means the gfn is writable on mmu. The bit is set when 44 the gfn is writable on guest mmu and it is not write-protected by shadow 45 page write-protection. 46 47On fast page fault path, we will use cmpxchg to atomically set the spte W 48bit if spte.SPTE_HOST_WRITEABLE = 1 and spte.SPTE_WRITE_PROTECT = 1, or 49restore the saved R/X bits if VMX_EPT_TRACK_ACCESS mask is set, or both. This 50is safe because whenever changing these bits can be detected by cmpxchg. 51 52But we need carefully check these cases: 531): The mapping from gfn to pfn 54The mapping from gfn to pfn may be changed since we can only ensure the pfn 55is not changed during cmpxchg. This is a ABA problem, for example, below case 56will happen: 57 58At the beginning: 59gpte = gfn1 60gfn1 is mapped to pfn1 on host 61spte is the shadow page table entry corresponding with gpte and 62spte = pfn1 63 64 VCPU 0 VCPU0 65on fast page fault path: 66 67 old_spte = *spte; 68 pfn1 is swapped out: 69 spte = 0; 70 71 pfn1 is re-alloced for gfn2. 72 73 gpte is changed to point to 74 gfn2 by the guest: 75 spte = pfn1; 76 77 if (cmpxchg(spte, old_spte, old_spte+W) 78 mark_page_dirty(vcpu->kvm, gfn1) 79 OOPS!!! 80 81We dirty-log for gfn1, that means gfn2 is lost in dirty-bitmap. 82 83For direct sp, we can easily avoid it since the spte of direct sp is fixed 84to gfn. For indirect sp, before we do cmpxchg, we call gfn_to_pfn_atomic() 85to pin gfn to pfn, because after gfn_to_pfn_atomic(): 86- We have held the refcount of pfn that means the pfn can not be freed and 87 be reused for another gfn. 88- The pfn is writable that means it can not be shared between different gfns 89 by KSM. 90 91Then, we can ensure the dirty bitmaps is correctly set for a gfn. 92 93Currently, to simplify the whole things, we disable fast page fault for 94indirect shadow page. 95 962): Dirty bit tracking 97In the origin code, the spte can be fast updated (non-atomically) if the 98spte is read-only and the Accessed bit has already been set since the 99Accessed bit and Dirty bit can not be lost. 100 101But it is not true after fast page fault since the spte can be marked 102writable between reading spte and updating spte. Like below case: 103 104At the beginning: 105spte.W = 0 106spte.Accessed = 1 107 108 VCPU 0 VCPU0 109In mmu_spte_clear_track_bits(): 110 111 old_spte = *spte; 112 113 /* 'if' condition is satisfied. */ 114 if (old_spte.Accessed == 1 && 115 old_spte.W == 0) 116 spte = 0ull; 117 on fast page fault path: 118 spte.W = 1 119 memory write on the spte: 120 spte.Dirty = 1 121 122 123 else 124 old_spte = xchg(spte, 0ull) 125 126 127 if (old_spte.Accessed == 1) 128 kvm_set_pfn_accessed(spte.pfn); 129 if (old_spte.Dirty == 1) 130 kvm_set_pfn_dirty(spte.pfn); 131 OOPS!!! 132 133The Dirty bit is lost in this case. 134 135In order to avoid this kind of issue, we always treat the spte as "volatile" 136if it can be updated out of mmu-lock, see spte_has_volatile_bits(), it means, 137the spte is always atomically updated in this case. 138 1393): flush tlbs due to spte updated 140If the spte is updated from writable to readonly, we should flush all TLBs, 141otherwise rmap_write_protect will find a read-only spte, even though the 142writable spte might be cached on a CPU's TLB. 143 144As mentioned before, the spte can be updated to writable out of mmu-lock on 145fast page fault path, in order to easily audit the path, we see if TLBs need 146be flushed caused by this reason in mmu_spte_update() since this is a common 147function to update spte (present -> present). 148 149Since the spte is "volatile" if it can be updated out of mmu-lock, we always 150atomically update the spte, the race caused by fast page fault can be avoided, 151See the comments in spte_has_volatile_bits() and mmu_spte_update(). 152 153Lockless Access Tracking: 154 155This is used for Intel CPUs that are using EPT but do not support the EPT A/D 156bits. In this case, when the KVM MMU notifier is called to track accesses to a 157page (via kvm_mmu_notifier_clear_flush_young), it marks the PTE as not-present 158by clearing the RWX bits in the PTE and storing the original R & X bits in 159some unused/ignored bits. In addition, the SPTE_SPECIAL_MASK is also set on the 160PTE (using the ignored bit 62). When the VM tries to access the page later on, 161a fault is generated and the fast page fault mechanism described above is used 162to atomically restore the PTE to a Present state. The W bit is not saved when 163the PTE is marked for access tracking and during restoration to the Present 164state, the W bit is set depending on whether or not it was a write access. If 165it wasn't, then the W bit will remain clear until a write access happens, at 166which time it will be set using the Dirty tracking mechanism described above. 167 1683. Reference 169------------ 170 171Name: kvm_lock 172Type: spinlock_t 173Arch: any 174Protects: - vm_list 175 176Name: kvm_count_lock 177Type: raw_spinlock_t 178Arch: any 179Protects: - hardware virtualization enable/disable 180Comment: 'raw' because hardware enabling/disabling must be atomic /wrt 181 migration. 182 183Name: kvm_arch::tsc_write_lock 184Type: raw_spinlock 185Arch: x86 186Protects: - kvm_arch::{last_tsc_write,last_tsc_nsec,last_tsc_offset} 187 - tsc offset in vmcb 188Comment: 'raw' because updating the tsc offsets must not be preempted. 189 190Name: kvm->mmu_lock 191Type: spinlock_t 192Arch: any 193Protects: -shadow page/shadow tlb entry 194Comment: it is a spinlock since it is used in mmu notifier. 195 196Name: kvm->srcu 197Type: srcu lock 198Arch: any 199Protects: - kvm->memslots 200 - kvm->buses 201Comment: The srcu read lock must be held while accessing memslots (e.g. 202 when using gfn_to_* functions) and while accessing in-kernel 203 MMIO/PIO address->device structure mapping (kvm->buses). 204 The srcu index can be stored in kvm_vcpu->srcu_idx per vcpu 205 if it is needed by multiple functions. 206 207Name: blocked_vcpu_on_cpu_lock 208Type: spinlock_t 209Arch: x86 210Protects: blocked_vcpu_on_cpu 211Comment: This is a per-CPU lock and it is used for VT-d posted-interrupts. 212 When VT-d posted-interrupts is supported and the VM has assigned 213 devices, we put the blocked vCPU on the list blocked_vcpu_on_cpu 214 protected by blocked_vcpu_on_cpu_lock, when VT-d hardware issues 215 wakeup notification event since external interrupts from the 216 assigned devices happens, we will find the vCPU on the list to 217 wakeup. 218