summaryrefslogtreecommitdiffstats
path: root/Documentation/virtual/kvm/vcpu-requests.rst
blob: 5feb3706a7ae58278bc9c747123d042f8b6f8958 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
=================
KVM VCPU Requests
=================

Overview
========

KVM supports an internal API enabling threads to request a VCPU thread to
perform some activity.  For example, a thread may request a VCPU to flush
its TLB with a VCPU request.  The API consists of the following functions::

  /* Check if any requests are pending for VCPU @vcpu. */
  bool kvm_request_pending(struct kvm_vcpu *vcpu);

  /* Check if VCPU @vcpu has request @req pending. */
  bool kvm_test_request(int req, struct kvm_vcpu *vcpu);

  /* Clear request @req for VCPU @vcpu. */
  void kvm_clear_request(int req, struct kvm_vcpu *vcpu);

  /*
   * Check if VCPU @vcpu has request @req pending. When the request is
   * pending it will be cleared and a memory barrier, which pairs with
   * another in kvm_make_request(), will be issued.
   */
  bool kvm_check_request(int req, struct kvm_vcpu *vcpu);

  /*
   * Make request @req of VCPU @vcpu. Issues a memory barrier, which pairs
   * with another in kvm_check_request(), prior to setting the request.
   */
  void kvm_make_request(int req, struct kvm_vcpu *vcpu);

  /* Make request @req of all VCPUs of the VM with struct kvm @kvm. */
  bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req);

Typically a requester wants the VCPU to perform the activity as soon
as possible after making the request.  This means most requests
(kvm_make_request() calls) are followed by a call to kvm_vcpu_kick(),
and kvm_make_all_cpus_request() has the kicking of all VCPUs built
into it.

VCPU Kicks
----------

The goal of a VCPU kick is to bring a VCPU thread out of guest mode in
order to perform some KVM maintenance.  To do so, an IPI is sent, forcing
a guest mode exit.  However, a VCPU thread may not be in guest mode at the
time of the kick.  Therefore, depending on the mode and state of the VCPU
thread, there are two other actions a kick may take.  All three actions
are listed below:

1) Send an IPI.  This forces a guest mode exit.
2) Waking a sleeping VCPU.  Sleeping VCPUs are VCPU threads outside guest
   mode that wait on waitqueues.  Waking them removes the threads from
   the waitqueues, allowing the threads to run again.  This behavior
   may be suppressed, see KVM_REQUEST_NO_WAKEUP below.
3) Nothing.  When the VCPU is not in guest mode and the VCPU thread is not
   sleeping, then there is nothing to do.

VCPU Mode
---------

VCPUs have a mode state, ``vcpu->mode``, that is used to track whether the
guest is running in guest mode or not, as well as some specific
outside guest mode states.  The architecture may use ``vcpu->mode`` to
ensure VCPU requests are seen by VCPUs (see "Ensuring Requests Are Seen"),
as well as to avoid sending unnecessary IPIs (see "IPI Reduction"), and
even to ensure IPI acknowledgements are waited upon (see "Waiting for
Acknowledgements").  The following modes are defined:

OUTSIDE_GUEST_MODE

  The VCPU thread is outside guest mode.

IN_GUEST_MODE

  The VCPU thread is in guest mode.

EXITING_GUEST_MODE

  The VCPU thread is transitioning from IN_GUEST_MODE to
  OUTSIDE_GUEST_MODE.

READING_SHADOW_PAGE_TABLES

  The VCPU thread is outside guest mode, but it wants the sender of
  certain VCPU requests, namely KVM_REQ_TLB_FLUSH, to wait until the VCPU
  thread is done reading the page tables.

VCPU Request Internals
======================

VCPU requests are simply bit indices of the ``vcpu->requests`` bitmap.
This means general bitops, like those documented in [atomic-ops]_ could
also be used, e.g. ::

  clear_bit(KVM_REQ_UNHALT & KVM_REQUEST_MASK, &vcpu->requests);

However, VCPU request users should refrain from doing so, as it would
break the abstraction.  The first 8 bits are reserved for architecture
independent requests, all additional bits are available for architecture
dependent requests.

Architecture Independent Requests
---------------------------------

KVM_REQ_TLB_FLUSH

  KVM's common MMU notifier may need to flush all of a guest's TLB
  entries, calling kvm_flush_remote_tlbs() to do so.  Architectures that
  choose to use the common kvm_flush_remote_tlbs() implementation will
  need to handle this VCPU request.

KVM_REQ_MMU_RELOAD

  When shadow page tables are used and memory slots are removed it's
  necessary to inform each VCPU to completely refresh the tables.  This
  request is used for that.

KVM_REQ_PENDING_TIMER

  This request may be made from a timer handler run on the host on behalf
  of a VCPU.  It informs the VCPU thread to inject a timer interrupt.

KVM_REQ_UNHALT

  This request may be made from the KVM common function kvm_vcpu_block(),
  which is used to emulate an instruction that causes a CPU to halt until
  one of an architectural specific set of events and/or interrupts is
  received (determined by checking kvm_arch_vcpu_runnable()).  When that
  event or interrupt arrives kvm_vcpu_block() makes the request.  This is
  in contrast to when kvm_vcpu_block() returns due to any other reason,
  such as a pending signal, which does not indicate the VCPU's halt
  emulation should stop, and therefore does not make the request.

KVM_REQUEST_MASK
----------------

VCPU requests should be masked by KVM_REQUEST_MASK before using them with
bitops.  This is because only the lower 8 bits are used to represent the
request's number.  The upper bits are used as flags.  Currently only two
flags are defined.

VCPU Request Flags
------------------

KVM_REQUEST_NO_WAKEUP

  This flag is applied to requests that only need immediate attention
  from VCPUs running in guest mode.  That is, sleeping VCPUs do not need
  to be awaken for these requests.  Sleeping VCPUs will handle the
  requests when they are awaken later for some other reason.

KVM_REQUEST_WAIT

  When requests with this flag are made with kvm_make_all_cpus_request(),
  then the caller will wait for each VCPU to acknowledge its IPI before
  proceeding.  This flag only applies to VCPUs that would receive IPIs.
  If, for example, the VCPU is sleeping, so no IPI is necessary, then
  the requesting thread does not wait.  This means that this flag may be
  safely combined with KVM_REQUEST_NO_WAKEUP.  See "Waiting for
  Acknowledgements" for more information about requests with
  KVM_REQUEST_WAIT.

VCPU Requests with Associated State
===================================

Requesters that want the receiving VCPU to handle new state need to ensure
the newly written state is observable to the receiving VCPU thread's CPU
by the time it observes the request.  This means a write memory barrier
must be inserted after writing the new state and before setting the VCPU
request bit.  Additionally, on the receiving VCPU thread's side, a
corresponding read barrier must be inserted after reading the request bit
and before proceeding to read the new state associated with it.  See
scenario 3, Message and Flag, of [lwn-mb]_ and the kernel documentation
[memory-barriers]_.

The pair of functions, kvm_check_request() and kvm_make_request(), provide
the memory barriers, allowing this requirement to be handled internally by
the API.

Ensuring Requests Are Seen
==========================

When making requests to VCPUs, we want to avoid the receiving VCPU
executing in guest mode for an arbitrary long time without handling the
request.  We can be sure this won't happen as long as we ensure the VCPU
thread checks kvm_request_pending() before entering guest mode and that a
kick will send an IPI to force an exit from guest mode when necessary.
Extra care must be taken to cover the period after the VCPU thread's last
kvm_request_pending() check and before it has entered guest mode, as kick
IPIs will only trigger guest mode exits for VCPU threads that are in guest
mode or at least have already disabled interrupts in order to prepare to
enter guest mode.  This means that an optimized implementation (see "IPI
Reduction") must be certain when it's safe to not send the IPI.  One
solution, which all architectures except s390 apply, is to:

- set ``vcpu->mode`` to IN_GUEST_MODE between disabling the interrupts and
  the last kvm_request_pending() check;
- enable interrupts atomically when entering the guest.

This solution also requires memory barriers to be placed carefully in both
the requesting thread and the receiving VCPU.  With the memory barriers we
can exclude the possibility of a VCPU thread observing
!kvm_request_pending() on its last check and then not receiving an IPI for
the next request made of it, even if the request is made immediately after
the check.  This is done by way of the Dekker memory barrier pattern
(scenario 10 of [lwn-mb]_).  As the Dekker pattern requires two variables,
this solution pairs ``vcpu->mode`` with ``vcpu->requests``.  Substituting
them into the pattern gives::

  CPU1                                    CPU2
  =================                       =================
  local_irq_disable();
  WRITE_ONCE(vcpu->mode, IN_GUEST_MODE);  kvm_make_request(REQ, vcpu);
  smp_mb();                               smp_mb();
  if (kvm_request_pending(vcpu)) {        if (READ_ONCE(vcpu->mode) ==
                                              IN_GUEST_MODE) {
      ...abort guest entry...                 ...send IPI...
  }                                       }

As stated above, the IPI is only useful for VCPU threads in guest mode or
that have already disabled interrupts.  This is why this specific case of
the Dekker pattern has been extended to disable interrupts before setting
``vcpu->mode`` to IN_GUEST_MODE.  WRITE_ONCE() and READ_ONCE() are used to
pedantically implement the memory barrier pattern, guaranteeing the
compiler doesn't interfere with ``vcpu->mode``'s carefully planned
accesses.

IPI Reduction
-------------

As only one IPI is needed to get a VCPU to check for any/all requests,
then they may be coalesced.  This is easily done by having the first IPI
sending kick also change the VCPU mode to something !IN_GUEST_MODE.  The
transitional state, EXITING_GUEST_MODE, is used for this purpose.

Waiting for Acknowledgements
----------------------------

Some requests, those with the KVM_REQUEST_WAIT flag set, require IPIs to
be sent, and the acknowledgements to be waited upon, even when the target
VCPU threads are in modes other than IN_GUEST_MODE.  For example, one case
is when a target VCPU thread is in READING_SHADOW_PAGE_TABLES mode, which
is set after disabling interrupts.  To support these cases, the
KVM_REQUEST_WAIT flag changes the condition for sending an IPI from
checking that the VCPU is IN_GUEST_MODE to checking that it is not
OUTSIDE_GUEST_MODE.

Request-less VCPU Kicks
-----------------------

As the determination of whether or not to send an IPI depends on the
two-variable Dekker memory barrier pattern, then it's clear that
request-less VCPU kicks are almost never correct.  Without the assurance
that a non-IPI generating kick will still result in an action by the
receiving VCPU, as the final kvm_request_pending() check does for
request-accompanying kicks, then the kick may not do anything useful at
all.  If, for instance, a request-less kick was made to a VCPU that was
just about to set its mode to IN_GUEST_MODE, meaning no IPI is sent, then
the VCPU thread may continue its entry without actually having done
whatever it was the kick was meant to initiate.

One exception is x86's posted interrupt mechanism.  In this case, however,
even the request-less VCPU kick is coupled with the same
local_irq_disable() + smp_mb() pattern described above; the ON bit
(Outstanding Notification) in the posted interrupt descriptor takes the
role of ``vcpu->requests``.  When sending a posted interrupt, PIR.ON is
set before reading ``vcpu->mode``; dually, in the VCPU thread,
vmx_sync_pir_to_irr() reads PIR after setting ``vcpu->mode`` to
IN_GUEST_MODE.

Additional Considerations
=========================

Sleeping VCPUs
--------------

VCPU threads may need to consider requests before and/or after calling
functions that may put them to sleep, e.g. kvm_vcpu_block().  Whether they
do or not, and, if they do, which requests need consideration, is
architecture dependent.  kvm_vcpu_block() calls kvm_arch_vcpu_runnable()
to check if it should awaken.  One reason to do so is to provide
architectures a function where requests may be checked if necessary.

Clearing Requests
-----------------

Generally it only makes sense for the receiving VCPU thread to clear a
request.  However, in some circumstances, such as when the requesting
thread and the receiving VCPU thread are executed serially, such as when
they are the same thread, or when they are using some form of concurrency
control to temporarily execute synchronously, then it's possible to know
that the request may be cleared immediately, rather than waiting for the
receiving VCPU thread to handle the request in VCPU RUN.  The only current
examples of this are kvm_vcpu_block() calls made by VCPUs to block
themselves.  A possible side-effect of that call is to make the
KVM_REQ_UNHALT request, which may then be cleared immediately when the
VCPU returns from the call.

References
==========

.. [atomic-ops] Documentation/core-api/atomic_ops.rst
.. [memory-barriers] Documentation/memory-barriers.txt
.. [lwn-mb] https://lwn.net/Articles/573436/