>> 1 .. _userfaultfd:
>> 2
1 =========== 3 ===========
2 Userfaultfd 4 Userfaultfd
3 =========== 5 ===========
4 6
5 Objective 7 Objective
6 ========= 8 =========
7 9
8 Userfaults allow the implementation of on-dema 10 Userfaults allow the implementation of on-demand paging from userland
9 and more generally they allow userland to take 11 and more generally they allow userland to take control of various
10 memory page faults, something otherwise only t 12 memory page faults, something otherwise only the kernel code could do.
11 13
12 For example userfaults allows a proper and mor 14 For example userfaults allows a proper and more optimal implementation
13 of the ``PROT_NONE+SIGSEGV`` trick. 15 of the ``PROT_NONE+SIGSEGV`` trick.
14 16
15 Design 17 Design
16 ====== 18 ======
17 19
18 Userspace creates a new userfaultfd, initializ 20 Userspace creates a new userfaultfd, initializes it, and registers one or more
19 regions of virtual memory with it. Then, any p 21 regions of virtual memory with it. Then, any page faults which occur within the
20 region(s) result in a message being delivered 22 region(s) result in a message being delivered to the userfaultfd, notifying
21 userspace of the fault. 23 userspace of the fault.
22 24
23 The ``userfaultfd`` (aside from registering an 25 The ``userfaultfd`` (aside from registering and unregistering virtual
24 memory ranges) provides two primary functional 26 memory ranges) provides two primary functionalities:
25 27
26 1) ``read/POLLIN`` protocol to notify a userla 28 1) ``read/POLLIN`` protocol to notify a userland thread of the faults
27 happening 29 happening
28 30
29 2) various ``UFFDIO_*`` ioctls that can manage 31 2) various ``UFFDIO_*`` ioctls that can manage the virtual memory regions
30 registered in the ``userfaultfd`` that allo 32 registered in the ``userfaultfd`` that allows userland to efficiently
31 resolve the userfaults it receives via 1) o 33 resolve the userfaults it receives via 1) or to manage the virtual
32 memory in the background 34 memory in the background
33 35
34 The real advantage of userfaults if compared t 36 The real advantage of userfaults if compared to regular virtual memory
35 management of mremap/mprotect is that the user 37 management of mremap/mprotect is that the userfaults in all their
36 operations never involve heavyweight structure 38 operations never involve heavyweight structures like vmas (in fact the
37 ``userfaultfd`` runtime load never takes the m 39 ``userfaultfd`` runtime load never takes the mmap_lock for writing).
38 Vmas are not suitable for page- (or hugepage) 40 Vmas are not suitable for page- (or hugepage) granular fault tracking
39 when dealing with virtual address spaces that 41 when dealing with virtual address spaces that could span
40 Terabytes. Too many vmas would be needed for t 42 Terabytes. Too many vmas would be needed for that.
41 43
42 The ``userfaultfd``, once created, can also be 44 The ``userfaultfd``, once created, can also be
43 passed using unix domain sockets to a manager 45 passed using unix domain sockets to a manager process, so the same
44 manager process could handle the userfaults of 46 manager process could handle the userfaults of a multitude of
45 different processes without them being aware a 47 different processes without them being aware about what is going on
46 (well of course unless they later try to use t 48 (well of course unless they later try to use the ``userfaultfd``
47 themselves on the same region the manager is a 49 themselves on the same region the manager is already tracking, which
48 is a corner case that would currently return ` 50 is a corner case that would currently return ``-EBUSY``).
49 51
50 API 52 API
51 === 53 ===
52 54
53 Creating a userfaultfd 55 Creating a userfaultfd
54 ---------------------- 56 ----------------------
55 57
56 There are two ways to create a new userfaultfd 58 There are two ways to create a new userfaultfd, each of which provide ways to
57 restrict access to this functionality (since h 59 restrict access to this functionality (since historically userfaultfds which
58 handle kernel page faults have been a useful t 60 handle kernel page faults have been a useful tool for exploiting the kernel).
59 61
60 The first way, supported since userfaultfd was 62 The first way, supported since userfaultfd was introduced, is the
61 userfaultfd(2) syscall. Access to this is cont 63 userfaultfd(2) syscall. Access to this is controlled in several ways:
62 64
63 - Any user can always create a userfaultfd whi 65 - Any user can always create a userfaultfd which traps userspace page faults
64 only. Such a userfaultfd can be created usin 66 only. Such a userfaultfd can be created using the userfaultfd(2) syscall
65 with the flag UFFD_USER_MODE_ONLY. 67 with the flag UFFD_USER_MODE_ONLY.
66 68
67 - In order to also trap kernel page faults for 69 - In order to also trap kernel page faults for the address space, either the
68 process needs the CAP_SYS_PTRACE capability, 70 process needs the CAP_SYS_PTRACE capability, or the system must have
69 vm.unprivileged_userfaultfd set to 1. By def 71 vm.unprivileged_userfaultfd set to 1. By default, vm.unprivileged_userfaultfd
70 is set to 0. 72 is set to 0.
71 73
72 The second way, added to the kernel more recen 74 The second way, added to the kernel more recently, is by opening
73 /dev/userfaultfd and issuing a USERFAULTFD_IOC 75 /dev/userfaultfd and issuing a USERFAULTFD_IOC_NEW ioctl to it. This method
74 yields equivalent userfaultfds to the userfaul 76 yields equivalent userfaultfds to the userfaultfd(2) syscall.
75 77
76 Unlike userfaultfd(2), access to /dev/userfaul 78 Unlike userfaultfd(2), access to /dev/userfaultfd is controlled via normal
77 filesystem permissions (user/group/mode), whic 79 filesystem permissions (user/group/mode), which gives fine grained access to
78 userfaultfd specifically, without also grantin 80 userfaultfd specifically, without also granting other unrelated privileges at
79 the same time (as e.g. granting CAP_SYS_PTRACE 81 the same time (as e.g. granting CAP_SYS_PTRACE would do). Users who have access
80 to /dev/userfaultfd can always create userfaul 82 to /dev/userfaultfd can always create userfaultfds that trap kernel page faults;
81 vm.unprivileged_userfaultfd is not considered. 83 vm.unprivileged_userfaultfd is not considered.
82 84
83 Initializing a userfaultfd 85 Initializing a userfaultfd
84 -------------------------- 86 --------------------------
85 87
86 When first opened the ``userfaultfd`` must be 88 When first opened the ``userfaultfd`` must be enabled invoking the
87 ``UFFDIO_API`` ioctl specifying a ``uffdio_api 89 ``UFFDIO_API`` ioctl specifying a ``uffdio_api.api`` value set to ``UFFD_API`` (or
88 a later API version) which will specify the `` 90 a later API version) which will specify the ``read/POLLIN`` protocol
89 userland intends to speak on the ``UFFD`` and 91 userland intends to speak on the ``UFFD`` and the ``uffdio_api.features``
90 userland requires. The ``UFFDIO_API`` ioctl if 92 userland requires. The ``UFFDIO_API`` ioctl if successful (i.e. if the
91 requested ``uffdio_api.api`` is spoken also by 93 requested ``uffdio_api.api`` is spoken also by the running kernel and the
92 requested features are going to be enabled) wi 94 requested features are going to be enabled) will return into
93 ``uffdio_api.features`` and ``uffdio_api.ioctl 95 ``uffdio_api.features`` and ``uffdio_api.ioctls`` two 64bit bitmasks of
94 respectively all the available features of the 96 respectively all the available features of the read(2) protocol and
95 the generic ioctl available. 97 the generic ioctl available.
96 98
97 The ``uffdio_api.features`` bitmask returned b 99 The ``uffdio_api.features`` bitmask returned by the ``UFFDIO_API`` ioctl
98 defines what memory types are supported by the 100 defines what memory types are supported by the ``userfaultfd`` and what
99 events, except page fault notifications, may b 101 events, except page fault notifications, may be generated:
100 102
101 - The ``UFFD_FEATURE_EVENT_*`` flags indicate 103 - The ``UFFD_FEATURE_EVENT_*`` flags indicate that various other events
102 other than page faults are supported. These 104 other than page faults are supported. These events are described in more
103 detail below in the `Non-cooperative userfau 105 detail below in the `Non-cooperative userfaultfd`_ section.
104 106
105 - ``UFFD_FEATURE_MISSING_HUGETLBFS`` and ``UFF 107 - ``UFFD_FEATURE_MISSING_HUGETLBFS`` and ``UFFD_FEATURE_MISSING_SHMEM``
106 indicate that the kernel supports ``UFFDIO_R 108 indicate that the kernel supports ``UFFDIO_REGISTER_MODE_MISSING``
107 registrations for hugetlbfs and shared memor 109 registrations for hugetlbfs and shared memory (covering all shmem APIs,
108 i.e. tmpfs, ``IPCSHM``, ``/dev/zero``, ``MAP 110 i.e. tmpfs, ``IPCSHM``, ``/dev/zero``, ``MAP_SHARED``, ``memfd_create``,
109 etc) virtual memory areas, respectively. 111 etc) virtual memory areas, respectively.
110 112
111 - ``UFFD_FEATURE_MINOR_HUGETLBFS`` indicates t 113 - ``UFFD_FEATURE_MINOR_HUGETLBFS`` indicates that the kernel supports
112 ``UFFDIO_REGISTER_MODE_MINOR`` registration 114 ``UFFDIO_REGISTER_MODE_MINOR`` registration for hugetlbfs virtual memory
113 areas. ``UFFD_FEATURE_MINOR_SHMEM`` is the a 115 areas. ``UFFD_FEATURE_MINOR_SHMEM`` is the analogous feature indicating
114 support for shmem virtual memory areas. 116 support for shmem virtual memory areas.
115 117
116 - ``UFFD_FEATURE_MOVE`` indicates that the ker <<
117 existing page contents from userspace. <<
118 <<
119 The userland application should set the featur 118 The userland application should set the feature flags it intends to use
120 when invoking the ``UFFDIO_API`` ioctl, to req 119 when invoking the ``UFFDIO_API`` ioctl, to request that those features be
121 enabled if supported. 120 enabled if supported.
122 121
123 Once the ``userfaultfd`` API has been enabled 122 Once the ``userfaultfd`` API has been enabled the ``UFFDIO_REGISTER``
124 ioctl should be invoked (if present in the ret 123 ioctl should be invoked (if present in the returned ``uffdio_api.ioctls``
125 bitmask) to register a memory range in the ``u 124 bitmask) to register a memory range in the ``userfaultfd`` by setting the
126 uffdio_register structure accordingly. The ``u 125 uffdio_register structure accordingly. The ``uffdio_register.mode``
127 bitmask will specify to the kernel which kind 126 bitmask will specify to the kernel which kind of faults to track for
128 the range. The ``UFFDIO_REGISTER`` ioctl will 127 the range. The ``UFFDIO_REGISTER`` ioctl will return the
129 ``uffdio_register.ioctls`` bitmask of ioctls t 128 ``uffdio_register.ioctls`` bitmask of ioctls that are suitable to resolve
130 userfaults on the range registered. Not all io 129 userfaults on the range registered. Not all ioctls will necessarily be
131 supported for all memory types (e.g. anonymous 130 supported for all memory types (e.g. anonymous memory vs. shmem vs.
132 hugetlbfs), or all types of intercepted faults 131 hugetlbfs), or all types of intercepted faults.
133 132
134 Userland can use the ``uffdio_register.ioctls` 133 Userland can use the ``uffdio_register.ioctls`` to manage the virtual
135 address space in the background (to add or pot 134 address space in the background (to add or potentially also remove
136 memory from the ``userfaultfd`` registered ran 135 memory from the ``userfaultfd`` registered range). This means a userfault
137 could be triggering just before userland maps 136 could be triggering just before userland maps in the background the
138 user-faulted page. 137 user-faulted page.
139 138
140 Resolving Userfaults 139 Resolving Userfaults
141 -------------------- 140 --------------------
142 141
143 There are three basic ways to resolve userfaul 142 There are three basic ways to resolve userfaults:
144 143
145 - ``UFFDIO_COPY`` atomically copies some exist 144 - ``UFFDIO_COPY`` atomically copies some existing page contents from
146 userspace. 145 userspace.
147 146
148 - ``UFFDIO_ZEROPAGE`` atomically zeros the new 147 - ``UFFDIO_ZEROPAGE`` atomically zeros the new page.
149 148
150 - ``UFFDIO_CONTINUE`` maps an existing, previo 149 - ``UFFDIO_CONTINUE`` maps an existing, previously-populated page.
151 150
152 These operations are atomic in the sense that 151 These operations are atomic in the sense that they guarantee nothing can
153 see a half-populated page, since readers will 152 see a half-populated page, since readers will keep userfaulting until the
154 operation has finished. 153 operation has finished.
155 154
156 By default, these wake up userfaults blocked o 155 By default, these wake up userfaults blocked on the range in question.
157 They support a ``UFFDIO_*_MODE_DONTWAKE`` ``mo 156 They support a ``UFFDIO_*_MODE_DONTWAKE`` ``mode`` flag, which indicates
158 that waking will be done separately at some la 157 that waking will be done separately at some later time.
159 158
160 Which ioctl to choose depends on the kind of p 159 Which ioctl to choose depends on the kind of page fault, and what we'd
161 like to do to resolve it: 160 like to do to resolve it:
162 161
163 - For ``UFFDIO_REGISTER_MODE_MISSING`` faults, 162 - For ``UFFDIO_REGISTER_MODE_MISSING`` faults, the fault needs to be
164 resolved by either providing a new page (``U 163 resolved by either providing a new page (``UFFDIO_COPY``), or mapping
165 the zero page (``UFFDIO_ZEROPAGE``). By defa 164 the zero page (``UFFDIO_ZEROPAGE``). By default, the kernel would map
166 the zero page for a missing fault. With user 165 the zero page for a missing fault. With userfaultfd, userspace can
167 decide what content to provide before the fa 166 decide what content to provide before the faulting thread continues.
168 167
169 - For ``UFFDIO_REGISTER_MODE_MINOR`` faults, t 168 - For ``UFFDIO_REGISTER_MODE_MINOR`` faults, there is an existing page (in
170 the page cache). Userspace has the option of 169 the page cache). Userspace has the option of modifying the page's
171 contents before resolving the fault. Once th 170 contents before resolving the fault. Once the contents are correct
172 (modified or not), userspace asks the kernel 171 (modified or not), userspace asks the kernel to map the page and let the
173 faulting thread continue with ``UFFDIO_CONTI 172 faulting thread continue with ``UFFDIO_CONTINUE``.
174 173
175 Notes: 174 Notes:
176 175
177 - You can tell which kind of fault occurred by 176 - You can tell which kind of fault occurred by examining
178 ``pagefault.flags`` within the ``uffd_msg``, 177 ``pagefault.flags`` within the ``uffd_msg``, checking for the
179 ``UFFD_PAGEFAULT_FLAG_*`` flags. 178 ``UFFD_PAGEFAULT_FLAG_*`` flags.
180 179
181 - None of the page-delivering ioctls default t 180 - None of the page-delivering ioctls default to the range that you
182 registered with. You must fill in all field 181 registered with. You must fill in all fields for the appropriate
183 ioctl struct including the range. 182 ioctl struct including the range.
184 183
185 - You get the address of the access that trigg 184 - You get the address of the access that triggered the missing page
186 event out of a struct uffd_msg that you read 185 event out of a struct uffd_msg that you read in the thread from the
187 uffd. You can supply as many pages as you w 186 uffd. You can supply as many pages as you want with these IOCTLs.
188 Keep in mind that unless you used DONTWAKE t 187 Keep in mind that unless you used DONTWAKE then the first of any of
189 those IOCTLs wakes up the faulting thread. 188 those IOCTLs wakes up the faulting thread.
190 189
191 - Be sure to test for all errors including 190 - Be sure to test for all errors including
192 (``pollfd[0].revents & POLLERR``). This can 191 (``pollfd[0].revents & POLLERR``). This can happen, e.g. when ranges
193 supplied were incorrect. 192 supplied were incorrect.
194 193
195 Write Protect Notifications 194 Write Protect Notifications
196 --------------------------- 195 ---------------------------
197 196
198 This is equivalent to (but faster than) using 197 This is equivalent to (but faster than) using mprotect and a SIGSEGV
199 signal handler. 198 signal handler.
200 199
201 Firstly you need to register a range with ``UF 200 Firstly you need to register a range with ``UFFDIO_REGISTER_MODE_WP``.
202 Instead of using mprotect(2) you use 201 Instead of using mprotect(2) you use
203 ``ioctl(uffd, UFFDIO_WRITEPROTECT, struct *uff 202 ``ioctl(uffd, UFFDIO_WRITEPROTECT, struct *uffdio_writeprotect)``
204 while ``mode = UFFDIO_WRITEPROTECT_MODE_WP`` 203 while ``mode = UFFDIO_WRITEPROTECT_MODE_WP``
205 in the struct passed in. The range does not d 204 in the struct passed in. The range does not default to and does not
206 have to be identical to the range you register 205 have to be identical to the range you registered with. You can write
207 protect as many ranges as you like (inside the 206 protect as many ranges as you like (inside the registered range).
208 Then, in the thread reading from uffd the stru 207 Then, in the thread reading from uffd the struct will have
209 ``msg.arg.pagefault.flags & UFFD_PAGEFAULT_FLA 208 ``msg.arg.pagefault.flags & UFFD_PAGEFAULT_FLAG_WP`` set. Now you send
210 ``ioctl(uffd, UFFDIO_WRITEPROTECT, struct *uff 209 ``ioctl(uffd, UFFDIO_WRITEPROTECT, struct *uffdio_writeprotect)``
211 again while ``pagefault.mode`` does not have ` 210 again while ``pagefault.mode`` does not have ``UFFDIO_WRITEPROTECT_MODE_WP``
212 set. This wakes up the thread which will conti 211 set. This wakes up the thread which will continue to run with writes. This
213 allows you to do the bookkeeping about the wri 212 allows you to do the bookkeeping about the write in the uffd reading
214 thread before the ioctl. 213 thread before the ioctl.
215 214
216 If you registered with both ``UFFDIO_REGISTER_ 215 If you registered with both ``UFFDIO_REGISTER_MODE_MISSING`` and
217 ``UFFDIO_REGISTER_MODE_WP`` then you need to t 216 ``UFFDIO_REGISTER_MODE_WP`` then you need to think about the sequence in
218 which you supply a page and undo write protect 217 which you supply a page and undo write protect. Note that there is a
219 difference between writes into a WP area and i 218 difference between writes into a WP area and into a !WP area. The
220 former will have ``UFFD_PAGEFAULT_FLAG_WP`` se 219 former will have ``UFFD_PAGEFAULT_FLAG_WP`` set, the latter
221 ``UFFD_PAGEFAULT_FLAG_WRITE``. The latter did 220 ``UFFD_PAGEFAULT_FLAG_WRITE``. The latter did not fail on protection but
222 you still need to supply a page when ``UFFDIO_ 221 you still need to supply a page when ``UFFDIO_REGISTER_MODE_MISSING`` was
223 used. 222 used.
224 <<
225 Userfaultfd write-protect mode currently behav <<
226 (when e.g. page is missing) over different typ <<
227 <<
228 For anonymous memory, ``ioctl(UFFDIO_WRITEPROT <<
229 (e.g. when pages are missing and not populated <<
230 like shmem and hugetlbfs, none ptes will be wr <<
231 present pte. In other words, there will be a <<
232 message generated when writing to a missing pa <<
233 as long as the page range was write-protected <<
234 not be generated on anonymous memories by defa <<
235 <<
236 If the application wants to be able to write p <<
237 memory, one can pre-populate the memory with e <<
238 newer kernels, one can also detect the feature <<
239 and set the feature bit in advance to make sur <<
240 write protected even upon anonymous memory. <<
241 <<
242 When using ``UFFDIO_REGISTER_MODE_WP`` in comb <<
243 ``UFFDIO_REGISTER_MODE_MISSING`` or ``UFFDIO_R <<
244 resolving missing / minor faults with ``UFFDIO <<
245 respectively, it may be desirable for the new <<
246 write-protected (so future writes will also re <<
247 support a mode flag (``UFFDIO_COPY_MODE_WP`` o <<
248 respectively) to configure the mapping this wa <<
249 <<
250 If the userfaultfd context has ``UFFD_FEATURE_ <<
251 any vma registered with write-protection will <<
252 than the default sync mode. <<
253 <<
254 In async mode, there will be no message genera <<
255 happens, meanwhile the write-protection will b <<
256 the kernel. It can be seen as a more accurate <<
257 tracking and it can be different in a few ways <<
258 <<
259 - The dirty result will not be affected by v <<
260 merging) because the dirty is only tracked <<
261 <<
262 - It supports range operations by default, s <<
263 any range of memory as long as page aligne <<
264 <<
265 - Dirty information will not get lost if the <<
266 various reasons (e.g. during split of a sh <<
267 <<
268 - Due to a reverted meaning of soft-dirty (p <<
269 set; dirty when uffd-wp bit cleared), it h <<
270 some of the memory operations. For exampl <<
271 anonymous (or ``MADV_REMOVE`` on a file ma <<
272 dirtying of memory by dropping uffd-wp bit <<
273 <<
274 The user app can collect the "written/dirty" s <<
275 uffd-wp bit for the pages being interested in <<
276 <<
277 The page will not be under track of uffd-wp as <<
278 explicitly write-protected by ``ioctl(UFFDIO_W <<
279 flag ``UFFDIO_WRITEPROTECT_MODE_WP`` set. Try <<
280 that was tracked by async mode userfaultfd-wp <<
281 <<
282 When userfaultfd-wp async mode is used alone, <<
283 kinds of memory. <<
284 <<
285 Memory Poisioning Emulation <<
286 --------------------------- <<
287 <<
288 In response to a fault (either missing or mino <<
289 take to "resolve" it is to issue a ``UFFDIO_PO <<
290 future faulters to either get a SIGBUS, or in <<
291 receive an MCE as if there were hardware memor <<
292 <<
293 This is used to emulate hardware memory poison <<
294 machine which experiences a real hardware memo <<
295 the VM to another physical machine. Since we w <<
296 transparent to the guest, we want that same ad <<
297 still poisoned, even though it's on a new phys <<
298 doesn't have a memory error in the exact same <<
299 223
300 QEMU/KVM 224 QEMU/KVM
301 ======== 225 ========
302 226
303 QEMU/KVM is using the ``userfaultfd`` syscall 227 QEMU/KVM is using the ``userfaultfd`` syscall to implement postcopy live
304 migration. Postcopy live migration is one form 228 migration. Postcopy live migration is one form of memory
305 externalization consisting of a virtual machin 229 externalization consisting of a virtual machine running with part or
306 all of its memory residing on a different node 230 all of its memory residing on a different node in the cloud. The
307 ``userfaultfd`` abstraction is generic enough 231 ``userfaultfd`` abstraction is generic enough that not a single line of
308 KVM kernel code had to be modified in order to 232 KVM kernel code had to be modified in order to add postcopy live
309 migration to QEMU. 233 migration to QEMU.
310 234
311 Guest async page faults, ``FOLL_NOWAIT`` and a 235 Guest async page faults, ``FOLL_NOWAIT`` and all other ``GUP*`` features work
312 just fine in combination with userfaults. User 236 just fine in combination with userfaults. Userfaults trigger async
313 page faults in the guest scheduler so those gu 237 page faults in the guest scheduler so those guest processes that
314 aren't waiting for userfaults (i.e. network bo 238 aren't waiting for userfaults (i.e. network bound) can keep running in
315 the guest vcpus. 239 the guest vcpus.
316 240
317 It is generally beneficial to run one pass of 241 It is generally beneficial to run one pass of precopy live migration
318 just before starting postcopy live migration, 242 just before starting postcopy live migration, in order to avoid
319 generating userfaults for readonly guest regio 243 generating userfaults for readonly guest regions.
320 244
321 The implementation of postcopy live migration 245 The implementation of postcopy live migration currently uses one
322 single bidirectional socket but in the future 246 single bidirectional socket but in the future two different sockets
323 will be used (to reduce the latency of the use 247 will be used (to reduce the latency of the userfaults to the minimum
324 possible without having to decrease ``/proc/sy 248 possible without having to decrease ``/proc/sys/net/ipv4/tcp_wmem``).
325 249
326 The QEMU in the source node writes all pages t 250 The QEMU in the source node writes all pages that it knows are missing
327 in the destination node, into the socket, and 251 in the destination node, into the socket, and the migration thread of
328 the QEMU running in the destination node runs 252 the QEMU running in the destination node runs ``UFFDIO_COPY|ZEROPAGE``
329 ioctls on the ``userfaultfd`` in order to map 253 ioctls on the ``userfaultfd`` in order to map the received pages into the
330 guest (``UFFDIO_ZEROCOPY`` is used if the sour 254 guest (``UFFDIO_ZEROCOPY`` is used if the source page was a zero page).
331 255
332 A different postcopy thread in the destination 256 A different postcopy thread in the destination node listens with
333 poll() to the ``userfaultfd`` in parallel. Whe 257 poll() to the ``userfaultfd`` in parallel. When a ``POLLIN`` event is
334 generated after a userfault triggers, the post 258 generated after a userfault triggers, the postcopy thread read() from
335 the ``userfaultfd`` and receives the fault add 259 the ``userfaultfd`` and receives the fault address (or ``-EAGAIN`` in case the
336 userfault was already resolved and waken by a 260 userfault was already resolved and waken by a ``UFFDIO_COPY|ZEROPAGE`` run
337 by the parallel QEMU migration thread). 261 by the parallel QEMU migration thread).
338 262
339 After the QEMU postcopy thread (running in the 263 After the QEMU postcopy thread (running in the destination node) gets
340 the userfault address it writes the informatio 264 the userfault address it writes the information about the missing page
341 into the socket. The QEMU source node receives 265 into the socket. The QEMU source node receives the information and
342 roughly "seeks" to that page address and conti 266 roughly "seeks" to that page address and continues sending all
343 remaining missing pages from that new page off 267 remaining missing pages from that new page offset. Soon after that
344 (just the time to flush the tcp_wmem queue thr 268 (just the time to flush the tcp_wmem queue through the network) the
345 migration thread in the QEMU running in the de 269 migration thread in the QEMU running in the destination node will
346 receive the page that triggered the userfault 270 receive the page that triggered the userfault and it'll map it as
347 usual with the ``UFFDIO_COPY|ZEROPAGE`` (witho 271 usual with the ``UFFDIO_COPY|ZEROPAGE`` (without actually knowing if it
348 was spontaneously sent by the source or if it 272 was spontaneously sent by the source or if it was an urgent page
349 requested through a userfault). 273 requested through a userfault).
350 274
351 By the time the userfaults start, the QEMU in 275 By the time the userfaults start, the QEMU in the destination node
352 doesn't need to keep any per-page state bitmap 276 doesn't need to keep any per-page state bitmap relative to the live
353 migration around and a single per-page bitmap 277 migration around and a single per-page bitmap has to be maintained in
354 the QEMU running in the source node to know wh 278 the QEMU running in the source node to know which pages are still
355 missing in the destination node. The bitmap in 279 missing in the destination node. The bitmap in the source node is
356 checked to find which missing pages to send in 280 checked to find which missing pages to send in round robin and we seek
357 over it when receiving incoming userfaults. Af 281 over it when receiving incoming userfaults. After sending each page of
358 course the bitmap is updated accordingly. It's 282 course the bitmap is updated accordingly. It's also useful to avoid
359 sending the same page twice (in case the userf 283 sending the same page twice (in case the userfault is read by the
360 postcopy thread just before ``UFFDIO_COPY|ZERO 284 postcopy thread just before ``UFFDIO_COPY|ZEROPAGE`` runs in the migration
361 thread). 285 thread).
362 286
363 Non-cooperative userfaultfd 287 Non-cooperative userfaultfd
364 =========================== 288 ===========================
365 289
366 When the ``userfaultfd`` is monitored by an ex 290 When the ``userfaultfd`` is monitored by an external manager, the manager
367 must be able to track changes in the process v 291 must be able to track changes in the process virtual memory
368 layout. Userfaultfd can notify the manager abo 292 layout. Userfaultfd can notify the manager about such changes using
369 the same read(2) protocol as for the page faul 293 the same read(2) protocol as for the page fault notifications. The
370 manager has to explicitly enable these events 294 manager has to explicitly enable these events by setting appropriate
371 bits in ``uffdio_api.features`` passed to ``UF 295 bits in ``uffdio_api.features`` passed to ``UFFDIO_API`` ioctl:
372 296
373 ``UFFD_FEATURE_EVENT_FORK`` 297 ``UFFD_FEATURE_EVENT_FORK``
374 enable ``userfaultfd`` hooks for fork( 298 enable ``userfaultfd`` hooks for fork(). When this feature is
375 enabled, the ``userfaultfd`` context o 299 enabled, the ``userfaultfd`` context of the parent process is
376 duplicated into the newly created proc 300 duplicated into the newly created process. The manager
377 receives ``UFFD_EVENT_FORK`` with file 301 receives ``UFFD_EVENT_FORK`` with file descriptor of the new
378 ``userfaultfd`` context in the ``uffd_ 302 ``userfaultfd`` context in the ``uffd_msg.fork``.
379 303
380 ``UFFD_FEATURE_EVENT_REMAP`` 304 ``UFFD_FEATURE_EVENT_REMAP``
381 enable notifications about mremap() ca 305 enable notifications about mremap() calls. When the
382 non-cooperative process moves a virtua 306 non-cooperative process moves a virtual memory area to a
383 different location, the manager will r 307 different location, the manager will receive
384 ``UFFD_EVENT_REMAP``. The ``uffd_msg.r 308 ``UFFD_EVENT_REMAP``. The ``uffd_msg.remap`` will contain the old and
385 new addresses of the area and its orig 309 new addresses of the area and its original length.
386 310
387 ``UFFD_FEATURE_EVENT_REMOVE`` 311 ``UFFD_FEATURE_EVENT_REMOVE``
388 enable notifications about madvise(MAD 312 enable notifications about madvise(MADV_REMOVE) and
389 madvise(MADV_DONTNEED) calls. The even 313 madvise(MADV_DONTNEED) calls. The event ``UFFD_EVENT_REMOVE`` will
390 be generated upon these calls to madvi 314 be generated upon these calls to madvise(). The ``uffd_msg.remove``
391 will contain start and end addresses o 315 will contain start and end addresses of the removed area.
392 316
393 ``UFFD_FEATURE_EVENT_UNMAP`` 317 ``UFFD_FEATURE_EVENT_UNMAP``
394 enable notifications about memory unma 318 enable notifications about memory unmapping. The manager will
395 get ``UFFD_EVENT_UNMAP`` with ``uffd_m 319 get ``UFFD_EVENT_UNMAP`` with ``uffd_msg.remove`` containing start and
396 end addresses of the unmapped area. 320 end addresses of the unmapped area.
397 321
398 Although the ``UFFD_FEATURE_EVENT_REMOVE`` and 322 Although the ``UFFD_FEATURE_EVENT_REMOVE`` and ``UFFD_FEATURE_EVENT_UNMAP``
399 are pretty similar, they quite differ in the a 323 are pretty similar, they quite differ in the action expected from the
400 ``userfaultfd`` manager. In the former case, t 324 ``userfaultfd`` manager. In the former case, the virtual memory is
401 removed, but the area is not, the area remains 325 removed, but the area is not, the area remains monitored by the
402 ``userfaultfd``, and if a page fault occurs in 326 ``userfaultfd``, and if a page fault occurs in that area it will be
403 delivered to the manager. The proper resolutio 327 delivered to the manager. The proper resolution for such page fault is
404 to zeromap the faulting address. However, in t 328 to zeromap the faulting address. However, in the latter case, when an
405 area is unmapped, either explicitly (with munm 329 area is unmapped, either explicitly (with munmap() system call), or
406 implicitly (e.g. during mremap()), the area is 330 implicitly (e.g. during mremap()), the area is removed and in turn the
407 ``userfaultfd`` context for such area disappea 331 ``userfaultfd`` context for such area disappears too and the manager will
408 not get further userland page faults from the 332 not get further userland page faults from the removed area. Still, the
409 notification is required in order to prevent m 333 notification is required in order to prevent manager from using
410 ``UFFDIO_COPY`` on the unmapped area. 334 ``UFFDIO_COPY`` on the unmapped area.
411 335
412 Unlike userland page faults which have to be s 336 Unlike userland page faults which have to be synchronous and require
413 explicit or implicit wakeup, all the events ar 337 explicit or implicit wakeup, all the events are delivered
414 asynchronously and the non-cooperative process 338 asynchronously and the non-cooperative process resumes execution as
415 soon as manager executes read(). The ``userfau 339 soon as manager executes read(). The ``userfaultfd`` manager should
416 carefully synchronize calls to ``UFFDIO_COPY`` 340 carefully synchronize calls to ``UFFDIO_COPY`` with the events
417 processing. To aid the synchronization, the `` 341 processing. To aid the synchronization, the ``UFFDIO_COPY`` ioctl will
418 return ``-ENOSPC`` when the monitored process 342 return ``-ENOSPC`` when the monitored process exits at the time of
419 ``UFFDIO_COPY``, and ``-ENOENT``, when the non 343 ``UFFDIO_COPY``, and ``-ENOENT``, when the non-cooperative process has changed
420 its virtual memory layout simultaneously with 344 its virtual memory layout simultaneously with outstanding ``UFFDIO_COPY``
421 operation. 345 operation.
422 346
423 The current asynchronous model of the event de 347 The current asynchronous model of the event delivery is optimal for
424 single threaded non-cooperative ``userfaultfd` 348 single threaded non-cooperative ``userfaultfd`` manager implementations. A
425 synchronous event delivery model can be added 349 synchronous event delivery model can be added later as a new
426 ``userfaultfd`` feature to facilitate multithr 350 ``userfaultfd`` feature to facilitate multithreading enhancements of the
427 non cooperative manager, for example to allow 351 non cooperative manager, for example to allow ``UFFDIO_COPY`` ioctls to
428 run in parallel to the event reception. Single 352 run in parallel to the event reception. Single threaded
429 implementations should continue to use the cur 353 implementations should continue to use the current async event
430 delivery model instead. 354 delivery model instead.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.