6 Add description of notifier of memory hotplug Oct 11 2007
8 This document is about memory hotplug including how-to-use and current status.
9 Because Memory Hotplug is still under development, contents of this text will
13 1.1 purpose of memory hotplug
14 1.2. Phases of memory hotplug
15 1.3. Unit of Memory online/offline operation
16 2. Kernel Configuration
17 3. sysfs files for memory hotplug
18 4. Physical memory hot-add phase
19 4.1 Hardware(Firmware) Support
20 4.2 Notify memory hot-add event by hand
21 5. Logical Memory hot-add phase
23 5.2. How to online memory
24 6. Logical memory remove
25 6.1 Memory offline and ZONE_MOVABLE
26 6.2. How to offline memory
27 7. Physical memory remove
28 8. Memory hotplug event notifier
31 Note(1): x86_64's has special implementation for memory hotplug.
32 This text does not describe it.
33 Note(2): This text assumes that sysfs is mounted at /sys.
40 1.1 purpose of memory hotplug
42 Memory Hotplug allows users to increase/decrease the amount of memory.
43 Generally, there are two purposes.
45 (A) For changing the amount of memory.
46 This is to allow a feature like capacity on demand.
47 (B) For installing/removing DIMMs or NUMA-nodes physically.
48 This is to exchange DIMMs/NUMA-nodes, reduce power consumption, etc.
50 (A) is required by highly virtualized environments and (B) is required by
51 hardware which supports memory power management.
53 Linux memory hotplug is designed for both purpose.
56 1.2. Phases of memory hotplug
58 There are 2 phases in Memory Hotplug.
59 1) Physical Memory Hotplug phase
60 2) Logical Memory Hotplug phase.
62 The First phase is to communicate hardware/firmware and make/erase
63 environment for hotplugged memory. Basically, this phase is necessary
64 for the purpose (B), but this is good phase for communication between
65 highly virtualized environments too.
67 When memory is hotplugged, the kernel recognizes new memory, makes new memory
68 management tables, and makes sysfs files for new memory's operation.
70 If firmware supports notification of connection of new memory to OS,
71 this phase is triggered automatically. ACPI can notify this event. If not,
72 "probe" operation by system administration is used instead.
75 Logical Memory Hotplug phase is to change memory state into
76 available/unavailable for users. Amount of memory from user's view is
77 changed by this phase. The kernel makes all memory in it as free pages
78 when a memory range is available.
80 In this document, this phase is described as online/offline.
82 Logical Memory Hotplug phase is triggered by write of sysfs file by system
83 administrator. For the hot-add case, it must be executed after Physical Hotplug
85 (However, if you writes udev's hotplug scripts for memory hotplug, these
86 phases can be execute in seamless way.)
89 1.3. Unit of Memory online/offline operation
91 Memory hotplug uses SPARSEMEM memory model which allows memory to be divided
92 into chunks of the same size. These chunks are called "sections". The size of
93 a memory section is architecture dependent. For example, power uses 16MiB, ia64
96 Memory sections are combined into chunks referred to as "memory blocks". The
97 size of a memory block is architecture dependent and represents the logical
98 unit upon which memory online/offline operations are to be performed. The
99 default size of a memory block is the same as memory section size unless an
100 architecture specifies otherwise. (see Section 3.)
102 To determine the size (in bytes) of a memory block please read this file:
104 /sys/devices/system/memory/block_size_bytes
107 -----------------------
108 2. Kernel Configuration
109 -----------------------
110 To use memory hotplug feature, kernel must be compiled with following
113 - For all memory hotplug
114 Memory model -> Sparse Memory (CONFIG_SPARSEMEM)
115 Allow for memory hot-add (CONFIG_MEMORY_HOTPLUG)
117 - To enable memory removal, the followings are also necessary
118 Allow for memory hot remove (CONFIG_MEMORY_HOTREMOVE)
119 Page Migration (CONFIG_MIGRATION)
121 - For ACPI memory hotplug, the followings are also necessary
122 Memory hotplug (under ACPI Support menu) (CONFIG_ACPI_HOTPLUG_MEMORY)
123 This option can be kernel module.
125 - As a related configuration, if your box has a feature of NUMA-node hotplug
126 via ACPI, then this option is necessary too.
127 ACPI0004,PNP0A05 and PNP0A06 Container Driver (under ACPI Support menu)
128 (CONFIG_ACPI_CONTAINER).
129 This option can be kernel module too.
132 --------------------------------
133 3 sysfs files for memory hotplug
134 --------------------------------
135 All memory blocks have their device information in sysfs. Each memory block
136 is described under /sys/devices/system/memory as
138 /sys/devices/system/memory/memoryXXX
139 (XXX is the memory block id.)
141 For the memory block covered by the sysfs directory. It is expected that all
142 memory sections in this range are present and no memory holes exist in the
143 range. Currently there is no way to determine if there is a memory hole, but
144 the existence of one should not affect the hotplug capabilities of the memory
147 For example, assume 1GiB memory block size. A device for a memory starting at
148 0x100000000 is /sys/device/system/memory/memory4
149 (0x100000000 / 1Gib = 4)
150 This device covers address range [0x100000000 ... 0x140000000)
152 Under each memory block, you can see 5 files:
154 /sys/devices/system/memory/memoryXXX/phys_index
155 /sys/devices/system/memory/memoryXXX/phys_device
156 /sys/devices/system/memory/memoryXXX/state
157 /sys/devices/system/memory/memoryXXX/removable
158 /sys/devices/system/memory/memoryXXX/valid_zones
160 'phys_index' : read-only and contains memory block id, same as XXX.
162 at read: contains online/offline state of memory.
163 at write: user can specify "online_kernel",
164 "online_movable", "online", "offline" command
165 which will be performed on all sections in the block.
166 'phys_device' : read-only: designed to show the name of physical memory
167 device. This is not well implemented now.
168 'removable' : read-only: contains an integer value indicating
169 whether the memory block is removable or not
170 removable. A value of 1 indicates that the memory
171 block is removable and a value of 0 indicates that
172 it is not removable. A memory block is removable only if
173 every section in the block is removable.
174 'valid_zones' : read-only: designed to show which zones this memory block
176 The first column shows it's default zone.
177 "memory6/valid_zones: Normal Movable" shows this memoryblock
178 can be onlined to ZONE_NORMAL by default and to ZONE_MOVABLE
180 "memory7/valid_zones: Movable Normal" shows this memoryblock
181 can be onlined to ZONE_MOVABLE by default and to ZONE_NORMAL
185 These directories/files appear after physical memory hotplug phase.
187 If CONFIG_NUMA is enabled the memoryXXX/ directories can also be accessed
188 via symbolic links located in the /sys/devices/system/node/node* directories.
191 /sys/devices/system/node/node0/memory9 -> ../../memory/memory9
193 A backlink will also be created:
194 /sys/devices/system/memory/memory9/node0 -> ../../node/node0
197 --------------------------------
198 4. Physical memory hot-add phase
199 --------------------------------
201 4.1 Hardware(Firmware) Support
203 On x86_64/ia64 platform, memory hotplug by ACPI is supported.
205 In general, the firmware (ACPI) which supports memory hotplug defines
206 memory class object of _HID "PNP0C80". When a notify is asserted to PNP0C80,
207 Linux's ACPI handler does hot-add memory to the system and calls a hotplug udev
208 script. This will be done automatically.
210 But scripts for memory hotplug are not contained in generic udev package(now).
211 You may have to write it by yourself or online/offline memory by hand.
212 Please see "How to online memory", "How to offline memory" in this text.
214 If firmware supports NUMA-node hotplug, and defines an object _HID "ACPI0004",
215 "PNP0A05", or "PNP0A06", notification is asserted to it, and ACPI handler
216 calls hotplug code for all of objects which are defined in it.
217 If memory device is found, memory hotplug code will be called.
220 4.2 Notify memory hot-add event by hand
222 On some architectures, the firmware may not notify the kernel of a memory
223 hotplug event. Therefore, the memory "probe" interface is supported to
224 explicitly notify the kernel. This interface depends on
225 CONFIG_ARCH_MEMORY_PROBE and can be configured on powerpc, sh, and x86
226 if hotplug is supported, although for x86 this should be handled by ACPI
229 Probe interface is located at
230 /sys/devices/system/memory/probe
232 You can tell the physical address of new memory to the kernel by
234 % echo start_address_of_new_memory > /sys/devices/system/memory/probe
236 Then, [start_address_of_new_memory, start_address_of_new_memory +
237 memory_block_size] memory range is hot-added. In this case, hotplug script is
238 not called (in current implementation). You'll have to online memory by
239 yourself. Please see "How to online memory" in this text.
242 ------------------------------
243 5. Logical Memory hot-add phase
244 ------------------------------
248 To see (online/offline) state of a memory block, read 'state' file.
250 % cat /sys/device/system/memory/memoryXXX/state
253 If the memory block is online, you'll read "online".
254 If the memory block is offline, you'll read "offline".
257 5.2. How to online memory
259 When the memory is hot-added, the kernel decides whether or not to "online"
260 it according to the policy which can be read from "auto_online_blocks" file:
262 % cat /sys/devices/system/memory/auto_online_blocks
264 The default is "offline" which means the newly added memory is not in a
265 ready-to-use state and you have to "online" the newly added memory blocks
266 manually. Automatic onlining can be requested by writing "online" to
267 "auto_online_blocks" file:
269 % echo online > /sys/devices/system/memory/auto_online_blocks
271 This sets a global policy and impacts all memory blocks that will subsequently
272 be hotplugged. Currently offline blocks keep their state. It is possible, under
273 certain circumstances, that some memory blocks will be added but will fail to
274 online. User space tools can check their "state" files
275 (/sys/devices/system/memory/memoryXXX/state) and try to online them manually.
277 If the automatic onlining wasn't requested, failed, or some memory block was
278 offlined it is possible to change the individual block's state by writing to the
281 % echo online > /sys/devices/system/memory/memoryXXX/state
283 This onlining will not change the ZONE type of the target memory block,
284 If the memory block is in ZONE_NORMAL, you can change it to ZONE_MOVABLE:
286 % echo online_movable > /sys/devices/system/memory/memoryXXX/state
287 (NOTE: current limit: this memory block must be adjacent to ZONE_MOVABLE)
289 And if the memory block is in ZONE_MOVABLE, you can change it to ZONE_NORMAL:
291 % echo online_kernel > /sys/devices/system/memory/memoryXXX/state
292 (NOTE: current limit: this memory block must be adjacent to ZONE_NORMAL)
294 After this, memory block XXX's state will be 'online' and the amount of
295 available memory will be increased.
297 Currently, newly added memory is added as ZONE_NORMAL (for powerpc, ZONE_DMA).
298 This may be changed in future.
302 ------------------------
303 6. Logical memory remove
304 ------------------------
306 6.1 Memory offline and ZONE_MOVABLE
308 Memory offlining is more complicated than memory online. Because memory offline
309 has to make the whole memory block be unused, memory offline can fail if
310 the memory block includes memory which cannot be freed.
312 In general, memory offline can use 2 techniques.
314 (1) reclaim and free all memory in the memory block.
315 (2) migrate all pages in the memory block.
317 In the current implementation, Linux's memory offline uses method (2), freeing
318 all pages in the memory block by page migration. But not all pages are
319 migratable. Under current Linux, migratable pages are anonymous pages and
320 page caches. For offlining a memory block by migration, the kernel has to
321 guarantee that the memory block contains only migratable pages.
323 Now, a boot option for making a memory block which consists of migratable pages
324 is supported. By specifying "kernelcore=" or "movablecore=" boot option, you can
325 create ZONE_MOVABLE...a zone which is just used for movable pages.
326 (See also Documentation/kernel-parameters.txt)
328 Assume the system has "TOTAL" amount of memory at boot time, this boot option
329 creates ZONE_MOVABLE as following.
331 1) When kernelcore=YYYY boot option is used,
332 Size of memory not for movable pages (not for offline) is YYYY.
333 Size of memory for movable pages (for offline) is TOTAL-YYYY.
335 2) When movablecore=ZZZZ boot option is used,
336 Size of memory not for movable pages (not for offline) is TOTAL - ZZZZ.
337 Size of memory for movable pages (for offline) is ZZZZ.
340 Note: Unfortunately, there is no information to show which memory block belongs
341 to ZONE_MOVABLE. This is TBD.
344 6.2. How to offline memory
346 You can offline a memory block by using the same sysfs interface that was used
349 % echo offline > /sys/devices/system/memory/memoryXXX/state
351 If offline succeeds, the state of the memory block is changed to be "offline".
352 If it fails, some error core (like -EBUSY) will be returned by the kernel.
353 Even if a memory block does not belong to ZONE_MOVABLE, you can try to offline
354 it. If it doesn't contain 'unmovable' memory, you'll get success.
356 A memory block under ZONE_MOVABLE is considered to be able to be offlined
357 easily. But under some busy state, it may return -EBUSY. Even if a memory
358 block cannot be offlined due to -EBUSY, you can retry offlining it and may be
359 able to offline it (or not). (For example, a page is referred to by some kernel
360 internal call and released soon.)
363 Memory hotplug's design direction is to make the possibility of memory offlining
364 higher and to guarantee unplugging memory under any situation. But it needs
365 more work. Returning -EBUSY under some situation may be good because the user
366 can decide to retry more or not by himself. Currently, memory offlining code
367 does some amount of retry with 120 seconds timeout.
369 -------------------------
370 7. Physical memory remove
371 -------------------------
372 Need more implementation yet....
373 - Notification completion of remove works by OS to firmware.
374 - Guard from remove if not yet.
376 --------------------------------
377 8. Memory hotplug event notifier
378 --------------------------------
379 Hotplugging events are sent to a notification queue.
381 There are six types of notification defined in include/linux/memory.h:
384 Generated before new memory becomes available in order to be able to
385 prepare subsystems to handle memory. The page allocator is still unable
386 to allocate from the new memory.
389 Generated if MEMORY_GOING_ONLINE fails.
392 Generated when memory has successfully brought online. The callback may
393 allocate pages from the new memory.
396 Generated to begin the process of offlining memory. Allocations are no
397 longer possible from the memory but some of the memory to be offlined
398 is still in use. The callback can be used to free memory known to a
399 subsystem from the indicated memory block.
402 Generated if MEMORY_GOING_OFFLINE fails. Memory is available again from
403 the memory block that we attempted to offline.
406 Generated after offlining memory is complete.
408 A callback routine can be registered by calling
410 hotplug_memory_notifier(callback_func, priority)
412 Callback functions with higher values of priority are called before callback
413 functions with lower values.
415 A callback function must have the following prototype:
418 struct notifier_block *self, unsigned long action, void *arg);
420 The first argument of the callback function (self) is a pointer to the block
421 of the notifier chain that points to the callback function itself.
422 The second argument (action) is one of the event types described above.
423 The third argument (arg) passes a pointer of struct memory_notify.
425 struct memory_notify {
426 unsigned long start_pfn;
427 unsigned long nr_pages;
428 int status_change_nid_normal;
429 int status_change_nid_high;
430 int status_change_nid;
433 start_pfn is start_pfn of online/offline memory.
434 nr_pages is # of pages of online/offline memory.
435 status_change_nid_normal is set node id when N_NORMAL_MEMORY of nodemask
436 is (will be) set/clear, if this is -1, then nodemask status is not changed.
437 status_change_nid_high is set node id when N_HIGH_MEMORY of nodemask
438 is (will be) set/clear, if this is -1, then nodemask status is not changed.
439 status_change_nid is set node id when N_MEMORY of nodemask is (will be)
440 set/clear. It means a new(memoryless) node gets new memory by online and a
441 node loses all memory. If this is -1, then nodemask status is not changed.
442 If status_changed_nid* >= 0, callback should create/discard structures for the
445 The callback routine shall return one of the values
446 NOTIFY_DONE, NOTIFY_OK, NOTIFY_BAD, NOTIFY_STOP
447 defined in include/linux/notifier.h
449 NOTIFY_DONE and NOTIFY_OK have no effect on the further processing.
451 NOTIFY_BAD is used as response to the MEM_GOING_ONLINE, MEM_GOING_OFFLINE,
452 MEM_ONLINE, or MEM_OFFLINE action to cancel hotplugging. It stops
453 further processing of the notification queue.
455 NOTIFY_STOP stops further processing of the notification queue.
460 - allowing memory hot-add to ZONE_MOVABLE. maybe we need some switch like
461 sysctl or new control file.
462 - showing memory block and physical device relationship.
463 - test and make it better memory offlining.
464 - support HugeTLB page migration and offlining.
465 - memmap removing at memory offline.
466 - physical remove memory.