Root/Documentation/memory-hotplug.txt

1==============
2Memory Hotplug
3==============
4
5Created: Jul 28 2007
6Add description of notifier of memory hotplug Oct 11 2007
7
8This document is about memory hotplug including how-to-use and current status.
9Because Memory Hotplug is still under development, contents of this text will
10be changed often.
11
121. Introduction
13  1.1 purpose of memory hotplug
14  1.2. Phases of memory hotplug
15  1.3. Unit of Memory online/offline operation
162. Kernel Configuration
173. sysfs files for memory hotplug
184. Physical memory hot-add phase
19  4.1 Hardware(Firmware) Support
20  4.2 Notify memory hot-add event by hand
215. Logical Memory hot-add phase
22  5.1. State of memory
23  5.2. How to online memory
246. Logical memory remove
25  6.1 Memory offline and ZONE_MOVABLE
26  6.2. How to offline memory
277. Physical memory remove
288. Memory hotplug event notifier
299. Future Work List
30
31Note(1): x86_64's has special implementation for memory hotplug.
32         This text does not describe it.
33Note(2): This text assumes that sysfs is mounted at /sys.
34
35
36---------------
371. Introduction
38---------------
39
401.1 purpose of memory hotplug
41------------
42Memory Hotplug allows users to increase/decrease the amount of memory.
43Generally, there are two purposes.
44
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.
49
50(A) is required by highly virtualized environments and (B) is required by
51hardware which supports memory power management.
52
53Linux memory hotplug is designed for both purpose.
54
55
561.2. Phases of memory hotplug
57---------------
58There are 2 phases in Memory Hotplug.
59  1) Physical Memory Hotplug phase
60  2) Logical Memory Hotplug phase.
61
62The First phase is to communicate hardware/firmware and make/erase
63environment for hotplugged memory. Basically, this phase is necessary
64for the purpose (B), but this is good phase for communication between
65highly virtualized environments too.
66
67When memory is hotplugged, the kernel recognizes new memory, makes new memory
68management tables, and makes sysfs files for new memory's operation.
69
70If firmware supports notification of connection of new memory to OS,
71this phase is triggered automatically. ACPI can notify this event. If not,
72"probe" operation by system administration is used instead.
73(see Section 4.).
74
75Logical Memory Hotplug phase is to change memory state into
76available/unavailable for users. Amount of memory from user's view is
77changed by this phase. The kernel makes all memory in it as free pages
78when a memory range is available.
79
80In this document, this phase is described as online/offline.
81
82Logical Memory Hotplug phase is triggered by write of sysfs file by system
83administrator. For the hot-add case, it must be executed after Physical Hotplug
84phase by hand.
85(However, if you writes udev's hotplug scripts for memory hotplug, these
86 phases can be execute in seamless way.)
87
88
891.3. Unit of Memory online/offline operation
90------------
91Memory hotplug uses SPARSEMEM memory model. SPARSEMEM divides the whole memory
92into chunks of the same size. The chunk is called a "section". The size of
93a section is architecture dependent. For example, power uses 16MiB, ia64 uses
941GiB. The unit of online/offline operation is "one section". (see Section 3.)
95
96To determine the size of sections, please read this file:
97
98/sys/devices/system/memory/block_size_bytes
99
100This file shows the size of sections in byte.
101
102-----------------------
1032. Kernel Configuration
104-----------------------
105To use memory hotplug feature, kernel must be compiled with following
106config options.
107
108- For all memory hotplug
109    Memory model -> Sparse Memory (CONFIG_SPARSEMEM)
110    Allow for memory hot-add (CONFIG_MEMORY_HOTPLUG)
111
112- To enable memory removal, the followings are also necessary
113    Allow for memory hot remove (CONFIG_MEMORY_HOTREMOVE)
114    Page Migration (CONFIG_MIGRATION)
115
116- For ACPI memory hotplug, the followings are also necessary
117    Memory hotplug (under ACPI Support menu) (CONFIG_ACPI_HOTPLUG_MEMORY)
118    This option can be kernel module.
119
120- As a related configuration, if your box has a feature of NUMA-node hotplug
121  via ACPI, then this option is necessary too.
122    ACPI0004,PNP0A05 and PNP0A06 Container Driver (under ACPI Support menu)
123    (CONFIG_ACPI_CONTAINER).
124    This option can be kernel module too.
125
126--------------------------------
1274 sysfs files for memory hotplug
128--------------------------------
129All sections have their device information in sysfs. Each section is part of
130a memory block under /sys/devices/system/memory as
131
132/sys/devices/system/memory/memoryXXX
133(XXX is the section id.)
134
135Now, XXX is defined as (start_address_of_section / section_size) of the first
136section contained in the memory block. The files 'phys_index' and
137'end_phys_index' under each directory report the beginning and end section id's
138for the memory block covered by the sysfs directory. It is expected that all
139memory sections in this range are present and no memory holes exist in the
140range. Currently there is no way to determine if there is a memory hole, but
141the existence of one should not affect the hotplug capabilities of the memory
142block.
143
144For example, assume 1GiB section size. A device for a memory starting at
1450x100000000 is /sys/device/system/memory/memory4
146(0x100000000 / 1Gib = 4)
147This device covers address range [0x100000000 ... 0x140000000)
148
149Under each section, you can see 4 or 5 files, the end_phys_index file being
150a recent addition and not present on older kernels.
151
152/sys/devices/system/memory/memoryXXX/start_phys_index
153/sys/devices/system/memory/memoryXXX/end_phys_index
154/sys/devices/system/memory/memoryXXX/phys_device
155/sys/devices/system/memory/memoryXXX/state
156/sys/devices/system/memory/memoryXXX/removable
157
158'phys_index' : read-only and contains section id of the first section
159            in the memory block, same as XXX.
160'end_phys_index' : read-only and contains section id of the last section
161            in the memory block.
162'state' : read-write
163                    at read: contains online/offline state of memory.
164                    at write: user can specify "online_kernel",
165                    "online_movable", "online", "offline" command
166                    which will be performed on all sections in the block.
167'phys_device' : read-only: designed to show the name of physical memory
168                    device. This is not well implemented now.
169'removable' : read-only: contains an integer value indicating
170                    whether the memory block is removable or not
171                    removable. A value of 1 indicates that the memory
172                    block is removable and a value of 0 indicates that
173                    it is not removable. A memory block is removable only if
174                    every section in the block is removable.
175
176NOTE:
177  These directories/files appear after physical memory hotplug phase.
178
179If CONFIG_NUMA is enabled the memoryXXX/ directories can also be accessed
180via symbolic links located in the /sys/devices/system/node/node* directories.
181
182For example:
183/sys/devices/system/node/node0/memory9 -> ../../memory/memory9
184
185A backlink will also be created:
186/sys/devices/system/memory/memory9/node0 -> ../../node/node0
187
188--------------------------------
1894. Physical memory hot-add phase
190--------------------------------
191
1924.1 Hardware(Firmware) Support
193------------
194On x86_64/ia64 platform, memory hotplug by ACPI is supported.
195
196In general, the firmware (ACPI) which supports memory hotplug defines
197memory class object of _HID "PNP0C80". When a notify is asserted to PNP0C80,
198Linux's ACPI handler does hot-add memory to the system and calls a hotplug udev
199script. This will be done automatically.
200
201But scripts for memory hotplug are not contained in generic udev package(now).
202You may have to write it by yourself or online/offline memory by hand.
203Please see "How to online memory", "How to offline memory" in this text.
204
205If firmware supports NUMA-node hotplug, and defines an object _HID "ACPI0004",
206"PNP0A05", or "PNP0A06", notification is asserted to it, and ACPI handler
207calls hotplug code for all of objects which are defined in it.
208If memory device is found, memory hotplug code will be called.
209
210
2114.2 Notify memory hot-add event by hand
212------------
213On powerpc, the firmware does not notify a memory hotplug event to the kernel.
214Therefore, "probe" interface is supported to notify the event to the kernel.
215This interface depends on CONFIG_ARCH_MEMORY_PROBE.
216
217CONFIG_ARCH_MEMORY_PROBE is supported on powerpc only. On x86, this config
218option is disabled by default since ACPI notifies a memory hotplug event to
219the kernel, which performs its hotplug operation as the result. Please
220enable this option if you need the "probe" interface for testing purposes
221on x86.
222
223Probe interface is located at
224/sys/devices/system/memory/probe
225
226You can tell the physical address of new memory to the kernel by
227
228% echo start_address_of_new_memory > /sys/devices/system/memory/probe
229
230Then, [start_address_of_new_memory, start_address_of_new_memory + section_size)
231memory range is hot-added. In this case, hotplug script is not called (in
232current implementation). You'll have to online memory by yourself.
233Please see "How to online memory" in this text.
234
235
236
237------------------------------
2385. Logical Memory hot-add phase
239------------------------------
240
2415.1. State of memory
242------------
243To see (online/offline) state of memory section, read 'state' file.
244
245% cat /sys/device/system/memory/memoryXXX/state
246
247
248If the memory section is online, you'll read "online".
249If the memory section is offline, you'll read "offline".
250
251
2525.2. How to online memory
253------------
254Even if the memory is hot-added, it is not at ready-to-use state.
255For using newly added memory, you have to "online" the memory section.
256
257For onlining, you have to write "online" to the section's state file as:
258
259% echo online > /sys/devices/system/memory/memoryXXX/state
260
261This onlining will not change the ZONE type of the target memory section,
262If the memory section is in ZONE_NORMAL, you can change it to ZONE_MOVABLE:
263
264% echo online_movable > /sys/devices/system/memory/memoryXXX/state
265(NOTE: current limit: this memory section must be adjacent to ZONE_MOVABLE)
266
267And if the memory section is in ZONE_MOVABLE, you can change it to ZONE_NORMAL:
268
269% echo online_kernel > /sys/devices/system/memory/memoryXXX/state
270(NOTE: current limit: this memory section must be adjacent to ZONE_NORMAL)
271
272After this, section memoryXXX's state will be 'online' and the amount of
273available memory will be increased.
274
275Currently, newly added memory is added as ZONE_NORMAL (for powerpc, ZONE_DMA).
276This may be changed in future.
277
278
279
280------------------------
2816. Logical memory remove
282------------------------
283
2846.1 Memory offline and ZONE_MOVABLE
285------------
286Memory offlining is more complicated than memory online. Because memory offline
287has to make the whole memory section be unused, memory offline can fail if
288the section includes memory which cannot be freed.
289
290In general, memory offline can use 2 techniques.
291
292(1) reclaim and free all memory in the section.
293(2) migrate all pages in the section.
294
295In the current implementation, Linux's memory offline uses method (2), freeing
296all pages in the section by page migration. But not all pages are
297migratable. Under current Linux, migratable pages are anonymous pages and
298page caches. For offlining a section by migration, the kernel has to guarantee
299that the section contains only migratable pages.
300
301Now, a boot option for making a section which consists of migratable pages is
302supported. By specifying "kernelcore=" or "movablecore=" boot option, you can
303create ZONE_MOVABLE...a zone which is just used for movable pages.
304(See also Documentation/kernel-parameters.txt)
305
306Assume the system has "TOTAL" amount of memory at boot time, this boot option
307creates ZONE_MOVABLE as following.
308
3091) When kernelcore=YYYY boot option is used,
310  Size of memory not for movable pages (not for offline) is YYYY.
311  Size of memory for movable pages (for offline) is TOTAL-YYYY.
312
3132) When movablecore=ZZZZ boot option is used,
314  Size of memory not for movable pages (not for offline) is TOTAL - ZZZZ.
315  Size of memory for movable pages (for offline) is ZZZZ.
316
317
318Note) Unfortunately, there is no information to show which section belongs
319to ZONE_MOVABLE. This is TBD.
320
321
3226.2. How to offline memory
323------------
324You can offline a section by using the same sysfs interface that was used in
325memory onlining.
326
327% echo offline > /sys/devices/system/memory/memoryXXX/state
328
329If offline succeeds, the state of the memory section is changed to be "offline".
330If it fails, some error core (like -EBUSY) will be returned by the kernel.
331Even if a section does not belong to ZONE_MOVABLE, you can try to offline it.
332If it doesn't contain 'unmovable' memory, you'll get success.
333
334A section under ZONE_MOVABLE is considered to be able to be offlined easily.
335But under some busy state, it may return -EBUSY. Even if a memory section
336cannot be offlined due to -EBUSY, you can retry offlining it and may be able to
337offline it (or not).
338(For example, a page is referred to by some kernel internal call and released
339 soon.)
340
341Consideration:
342Memory hotplug's design direction is to make the possibility of memory offlining
343higher and to guarantee unplugging memory under any situation. But it needs
344more work. Returning -EBUSY under some situation may be good because the user
345can decide to retry more or not by himself. Currently, memory offlining code
346does some amount of retry with 120 seconds timeout.
347
348-------------------------
3497. Physical memory remove
350-------------------------
351Need more implementation yet....
352 - Notification completion of remove works by OS to firmware.
353 - Guard from remove if not yet.
354
355--------------------------------
3568. Memory hotplug event notifier
357--------------------------------
358Memory hotplug has event notifier. There are 6 types of notification.
359
360MEMORY_GOING_ONLINE
361  Generated before new memory becomes available in order to be able to
362  prepare subsystems to handle memory. The page allocator is still unable
363  to allocate from the new memory.
364
365MEMORY_CANCEL_ONLINE
366  Generated if MEMORY_GOING_ONLINE fails.
367
368MEMORY_ONLINE
369  Generated when memory has successfully brought online. The callback may
370  allocate pages from the new memory.
371
372MEMORY_GOING_OFFLINE
373  Generated to begin the process of offlining memory. Allocations are no
374  longer possible from the memory but some of the memory to be offlined
375  is still in use. The callback can be used to free memory known to a
376  subsystem from the indicated memory section.
377
378MEMORY_CANCEL_OFFLINE
379  Generated if MEMORY_GOING_OFFLINE fails. Memory is available again from
380  the section that we attempted to offline.
381
382MEMORY_OFFLINE
383  Generated after offlining memory is complete.
384
385A callback routine can be registered by
386  hotplug_memory_notifier(callback_func, priority)
387
388The second argument of callback function (action) is event types of above.
389The third argument is passed by pointer of struct memory_notify.
390
391struct memory_notify {
392       unsigned long start_pfn;
393       unsigned long nr_pages;
394       int status_change_nid_normal;
395       int status_change_nid_high;
396       int status_change_nid;
397}
398
399start_pfn is start_pfn of online/offline memory.
400nr_pages is # of pages of online/offline memory.
401status_change_nid_normal is set node id when N_NORMAL_MEMORY of nodemask
402is (will be) set/clear, if this is -1, then nodemask status is not changed.
403status_change_nid_high is set node id when N_HIGH_MEMORY of nodemask
404is (will be) set/clear, if this is -1, then nodemask status is not changed.
405status_change_nid is set node id when N_MEMORY of nodemask is (will be)
406set/clear. It means a new(memoryless) node gets new memory by online and a
407node loses all memory. If this is -1, then nodemask status is not changed.
408If status_changed_nid* >= 0, callback should create/discard structures for the
409node if necessary.
410
411--------------
4129. Future Work
413--------------
414  - allowing memory hot-add to ZONE_MOVABLE. maybe we need some switch like
415    sysctl or new control file.
416  - showing memory section and physical device relationship.
417  - showing memory section is under ZONE_MOVABLE or not
418  - test and make it better memory offlining.
419  - support HugeTLB page migration and offlining.
420  - memmap removing at memory offline.
421  - physical remove memory.
422
423

Archive Download this file



interactive