• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
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
16For spinlocks, kvm_lock is taken outside kvm->mmu_lock.  Everything
17else is a leaf: no other lock is taken inside the critical sections.
18
192: Exception
20------------
21
22Fast page fault:
23
24Fast page fault is the fast path which fixes the guest page fault out of
25the mmu-lock on x86. Currently, the page fault can be fast only if the
26shadow page table is present and it is caused by write-protect, that means
27we just need change the W bit of the spte.
28
29What we use to avoid all the race is the SPTE_HOST_WRITEABLE bit and
30SPTE_MMU_WRITEABLE bit on the spte:
31- SPTE_HOST_WRITEABLE means the gfn is writable on host.
32- SPTE_MMU_WRITEABLE means the gfn is writable on mmu. The bit is set when
33  the gfn is writable on guest mmu and it is not write-protected by shadow
34  page write-protection.
35
36On fast page fault path, we will use cmpxchg to atomically set the spte W
37bit if spte.SPTE_HOST_WRITEABLE = 1 and spte.SPTE_WRITE_PROTECT = 1, this
38is safe because whenever changing these bits can be detected by cmpxchg.
39
40But we need carefully check these cases:
411): The mapping from gfn to pfn
42The mapping from gfn to pfn may be changed since we can only ensure the pfn
43is not changed during cmpxchg. This is a ABA problem, for example, below case
44will happen:
45
46At the beginning:
47gpte = gfn1
48gfn1 is mapped to pfn1 on host
49spte is the shadow page table entry corresponding with gpte and
50spte = pfn1
51
52   VCPU 0                           VCPU0
53on fast page fault path:
54
55   old_spte = *spte;
56                                 pfn1 is swapped out:
57                                    spte = 0;
58
59                                 pfn1 is re-alloced for gfn2.
60
61                                 gpte is changed to point to
62                                 gfn2 by the guest:
63                                    spte = pfn1;
64
65   if (cmpxchg(spte, old_spte, old_spte+W)
66	mark_page_dirty(vcpu->kvm, gfn1)
67             OOPS!!!
68
69We dirty-log for gfn1, that means gfn2 is lost in dirty-bitmap.
70
71For direct sp, we can easily avoid it since the spte of direct sp is fixed
72to gfn. For indirect sp, before we do cmpxchg, we call gfn_to_pfn_atomic()
73to pin gfn to pfn, because after gfn_to_pfn_atomic():
74- We have held the refcount of pfn that means the pfn can not be freed and
75  be reused for another gfn.
76- The pfn is writable that means it can not be shared between different gfns
77  by KSM.
78
79Then, we can ensure the dirty bitmaps is correctly set for a gfn.
80
81Currently, to simplify the whole things, we disable fast page fault for
82indirect shadow page.
83
842): Dirty bit tracking
85In the origin code, the spte can be fast updated (non-atomically) if the
86spte is read-only and the Accessed bit has already been set since the
87Accessed bit and Dirty bit can not be lost.
88
89But it is not true after fast page fault since the spte can be marked
90writable between reading spte and updating spte. Like below case:
91
92At the beginning:
93spte.W = 0
94spte.Accessed = 1
95
96   VCPU 0                                       VCPU0
97In mmu_spte_clear_track_bits():
98
99   old_spte = *spte;
100
101   /* 'if' condition is satisfied. */
102   if (old_spte.Accessed == 1 &&
103        old_spte.W == 0)
104      spte = 0ull;
105                                         on fast page fault path:
106                                             spte.W = 1
107                                         memory write on the spte:
108                                             spte.Dirty = 1
109
110
111   else
112      old_spte = xchg(spte, 0ull)
113
114
115   if (old_spte.Accessed == 1)
116      kvm_set_pfn_accessed(spte.pfn);
117   if (old_spte.Dirty == 1)
118      kvm_set_pfn_dirty(spte.pfn);
119      OOPS!!!
120
121The Dirty bit is lost in this case.
122
123In order to avoid this kind of issue, we always treat the spte as "volatile"
124if it can be updated out of mmu-lock, see spte_has_volatile_bits(), it means,
125the spte is always atomically updated in this case.
126
1273): flush tlbs due to spte updated
128If the spte is updated from writable to readonly, we should flush all TLBs,
129otherwise rmap_write_protect will find a read-only spte, even though the
130writable spte might be cached on a CPU's TLB.
131
132As mentioned before, the spte can be updated to writable out of mmu-lock on
133fast page fault path, in order to easily audit the path, we see if TLBs need
134be flushed caused by this reason in mmu_spte_update() since this is a common
135function to update spte (present -> present).
136
137Since the spte is "volatile" if it can be updated out of mmu-lock, we always
138atomically update the spte, the race caused by fast page fault can be avoided,
139See the comments in spte_has_volatile_bits() and mmu_spte_update().
140
1413. Reference
142------------
143
144Name:		kvm_lock
145Type:		spinlock_t
146Arch:		any
147Protects:	- vm_list
148
149Name:		kvm_count_lock
150Type:		raw_spinlock_t
151Arch:		any
152Protects:	- hardware virtualization enable/disable
153Comment:	'raw' because hardware enabling/disabling must be atomic /wrt
154		migration.
155
156Name:		kvm_arch::tsc_write_lock
157Type:		raw_spinlock
158Arch:		x86
159Protects:	- kvm_arch::{last_tsc_write,last_tsc_nsec,last_tsc_offset}
160		- tsc offset in vmcb
161Comment:	'raw' because updating the tsc offsets must not be preempted.
162
163Name:		kvm->mmu_lock
164Type:		spinlock_t
165Arch:		any
166Protects:	-shadow page/shadow tlb entry
167Comment:	it is a spinlock since it is used in mmu notifier.
168
169Name:		kvm->srcu
170Type:		srcu lock
171Arch:		any
172Protects:	- kvm->memslots
173		- kvm->buses
174Comment:	The srcu read lock must be held while accessing memslots (e.g.
175		when using gfn_to_* functions) and while accessing in-kernel
176		MMIO/PIO address->device structure mapping (kvm->buses).
177		The srcu index can be stored in kvm_vcpu->srcu_idx per vcpu
178		if it is needed by multiple functions.
179
180Name:		blocked_vcpu_on_cpu_lock
181Type:		spinlock_t
182Arch:		x86
183Protects:	blocked_vcpu_on_cpu
184Comment:	This is a per-CPU lock and it is used for VT-d posted-interrupts.
185		When VT-d posted-interrupts is supported and the VM has assigned
186		devices, we put the blocked vCPU on the list blocked_vcpu_on_cpu
187		protected by blocked_vcpu_on_cpu_lock, when VT-d hardware issues
188		wakeup notification event since external interrupts from the
189		assigned devices happens, we will find the vCPU on the list to
190		wakeup.
191