blob: 168117bd6ee8b8d7eafbb72c626a2434556f4cc9 [file] [log] [blame]
Yasunori Goto6867c932007-08-10 13:00:59 -07001==============
2Memory Hotplug
3==============
4
Yasunori Goto10020ca2007-10-21 16:41:36 -07005Created: Jul 28 2007
6Add description of notifier of memory hotplug Oct 11 2007
Yasunori Goto6867c932007-08-10 13:00:59 -07007
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
Yasunori Goto10020ca2007-10-21 16:41:36 -0700288. Memory hotplug event notifier
299. Future Work List
Yasunori Goto6867c932007-08-10 13:00:59 -070030
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
76avaiable/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 triggred 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--------------------------------
1273 sysfs files for memory hotplug
128--------------------------------
129All sections have their device information under /sys/devices/system/memory as
130
131/sys/devices/system/memory/memoryXXX
132(XXX is section id.)
133
134Now, XXX is defined as start_address_of_section / section_size.
135
136For example, assume 1GiB section size. A device for a memory starting at
1370x100000000 is /sys/device/system/memory/memory4
138(0x100000000 / 1Gib = 4)
139This device covers address range [0x100000000 ... 0x140000000)
140
141Under each section, you can see 3 files.
142
143/sys/devices/system/memory/memoryXXX/phys_index
144/sys/devices/system/memory/memoryXXX/phys_device
145/sys/devices/system/memory/memoryXXX/state
146
147'phys_index' : read-only and contains section id, same as XXX.
148'state' : read-write
149 at read: contains online/offline state of memory.
150 at write: user can specify "online", "offline" command
151'phys_device': read-only: designed to show the name of physical memory device.
152 This is not well implemented now.
153
154NOTE:
155 These directories/files appear after physical memory hotplug phase.
156
157
158--------------------------------
1594. Physical memory hot-add phase
160--------------------------------
161
1624.1 Hardware(Firmware) Support
163------------
164On x86_64/ia64 platform, memory hotplug by ACPI is supported.
165
166In general, the firmware (ACPI) which supports memory hotplug defines
167memory class object of _HID "PNP0C80". When a notify is asserted to PNP0C80,
168Linux's ACPI handler does hot-add memory to the system and calls a hotplug udev
169script. This will be done automatically.
170
171But scripts for memory hotplug are not contained in generic udev package(now).
172You may have to write it by yourself or online/offline memory by hand.
173Please see "How to online memory", "How to offline memory" in this text.
174
175If firmware supports NUMA-node hotplug, and defines an object _HID "ACPI0004",
176"PNP0A05", or "PNP0A06", notification is asserted to it, and ACPI handler
177calls hotplug code for all of objects which are defined in it.
178If memory device is found, memory hotplug code will be called.
179
180
1814.2 Notify memory hot-add event by hand
182------------
183In some environments, especially virtualized environment, firmware will not
184notify memory hotplug event to the kernel. For such environment, "probe"
185interface is supported. This interface depends on CONFIG_ARCH_MEMORY_PROBE.
186
187Now, CONFIG_ARCH_MEMORY_PROBE is supported only by powerpc but it does not
188contain highly architecture codes. Please add config if you need "probe"
189interface.
190
191Probe interface is located at
192/sys/devices/system/memory/probe
193
194You can tell the physical address of new memory to the kernel by
195
196% echo start_address_of_new_memory > /sys/devices/system/memory/probe
197
198Then, [start_address_of_new_memory, start_address_of_new_memory + section_size)
199memory range is hot-added. In this case, hotplug script is not called (in
200current implementation). You'll have to online memory by yourself.
201Please see "How to online memory" in this text.
202
203
204
205------------------------------
2065. Logical Memory hot-add phase
207------------------------------
208
2095.1. State of memory
210------------
211To see (online/offline) state of memory section, read 'state' file.
212
213% cat /sys/device/system/memory/memoryXXX/state
214
215
216If the memory section is online, you'll read "online".
217If the memory section is offline, you'll read "offline".
218
219
2205.2. How to online memory
221------------
222Even if the memory is hot-added, it is not at ready-to-use state.
223For using newly added memory, you have to "online" the memory section.
224
225For onlining, you have to write "online" to the section's state file as:
226
227% echo online > /sys/devices/system/memory/memoryXXX/state
228
229After this, section memoryXXX's state will be 'online' and the amount of
230available memory will be increased.
231
232Currently, newly added memory is added as ZONE_NORMAL (for powerpc, ZONE_DMA).
233This may be changed in future.
234
235
236
237------------------------
2386. Logical memory remove
239------------------------
240
2416.1 Memory offline and ZONE_MOVABLE
242------------
243Memory offlining is more complicated than memory online. Because memory offline
244has to make the whole memory section be unused, memory offline can fail if
245the section includes memory which cannot be freed.
246
247In general, memory offline can use 2 techniques.
248
249(1) reclaim and free all memory in the section.
250(2) migrate all pages in the section.
251
252In the current implementation, Linux's memory offline uses method (2), freeing
253all pages in the section by page migration. But not all pages are
254migratable. Under current Linux, migratable pages are anonymous pages and
255page caches. For offlining a section by migration, the kernel has to guarantee
256that the section contains only migratable pages.
257
258Now, a boot option for making a section which consists of migratable pages is
259supported. By specifying "kernelcore=" or "movablecore=" boot option, you can
260create ZONE_MOVABLE...a zone which is just used for movable pages.
261(See also Documentation/kernel-parameters.txt)
262
263Assume the system has "TOTAL" amount of memory at boot time, this boot option
264creates ZONE_MOVABLE as following.
265
2661) When kernelcore=YYYY boot option is used,
267 Size of memory not for movable pages (not for offline) is YYYY.
268 Size of memory for movable pages (for offline) is TOTAL-YYYY.
269
2702) When movablecore=ZZZZ boot option is used,
271 Size of memory not for movable pages (not for offline) is TOTAL - ZZZZ.
272 Size of memory for movable pages (for offline) is ZZZZ.
273
274
275Note) Unfortunately, there is no information to show which section belongs
276to ZONE_MOVABLE. This is TBD.
277
278
2796.2. How to offline memory
280------------
281You can offline a section by using the same sysfs interface that was used in
282memory onlining.
283
284% echo offline > /sys/devices/system/memory/memoryXXX/state
285
286If offline succeeds, the state of the memory section is changed to be "offline".
287If it fails, some error core (like -EBUSY) will be returned by the kernel.
288Even if a section does not belong to ZONE_MOVABLE, you can try to offline it.
289If it doesn't contain 'unmovable' memory, you'll get success.
290
291A section under ZONE_MOVABLE is considered to be able to be offlined easily.
292But under some busy state, it may return -EBUSY. Even if a memory section
293cannot be offlined due to -EBUSY, you can retry offlining it and may be able to
294offline it (or not).
295(For example, a page is referred to by some kernel internal call and released
296 soon.)
297
298Consideration:
299Memory hotplug's design direction is to make the possibility of memory offlining
300higher and to guarantee unplugging memory under any situation. But it needs
301more work. Returning -EBUSY under some situation may be good because the user
302can decide to retry more or not by himself. Currently, memory offlining code
303does some amount of retry with 120 seconds timeout.
304
305-------------------------
3067. Physical memory remove
307-------------------------
308Need more implementation yet....
309 - Notification completion of remove works by OS to firmware.
310 - Guard from remove if not yet.
311
Yasunori Goto10020ca2007-10-21 16:41:36 -0700312--------------------------------
3138. Memory hotplug event notifier
314--------------------------------
315Memory hotplug has event notifer. There are 6 types of notification.
316
317MEMORY_GOING_ONLINE
318 Generated before new memory becomes available in order to be able to
319 prepare subsystems to handle memory. The page allocator is still unable
320 to allocate from the new memory.
321
322MEMORY_CANCEL_ONLINE
323 Generated if MEMORY_GOING_ONLINE fails.
324
325MEMORY_ONLINE
326 Generated when memory has succesfully brought online. The callback may
327 allocate pages from the new memory.
328
329MEMORY_GOING_OFFLINE
330 Generated to begin the process of offlining memory. Allocations are no
331 longer possible from the memory but some of the memory to be offlined
332 is still in use. The callback can be used to free memory known to a
333 subsystem from the indicated memory section.
334
335MEMORY_CANCEL_OFFLINE
336 Generated if MEMORY_GOING_OFFLINE fails. Memory is available again from
337 the section that we attempted to offline.
338
339MEMORY_OFFLINE
340 Generated after offlining memory is complete.
341
342A callback routine can be registered by
343 hotplug_memory_notifier(callback_func, priority)
344
345The second argument of callback function (action) is event types of above.
346The third argument is passed by pointer of struct memory_notify.
347
348struct memory_notify {
349 unsigned long start_pfn;
350 unsigned long nr_pages;
351 int status_cahnge_nid;
352}
353
354start_pfn is start_pfn of online/offline memory.
355nr_pages is # of pages of online/offline memory.
356status_change_nid is set node id when N_HIGH_MEMORY of nodemask is (will be)
357set/clear. It means a new(memoryless) node gets new memory by online and a
358node loses all memory. If this is -1, then nodemask status is not changed.
359If status_changed_nid >= 0, callback should create/discard structures for the
360node if necessary.
361
Yasunori Goto6867c932007-08-10 13:00:59 -0700362--------------
Yasunori Goto10020ca2007-10-21 16:41:36 -07003639. Future Work
Yasunori Goto6867c932007-08-10 13:00:59 -0700364--------------
365 - allowing memory hot-add to ZONE_MOVABLE. maybe we need some switch like
366 sysctl or new control file.
367 - showing memory section and physical device relationship.
368 - showing memory section and node relationship (maybe good for NUMA)
369 - showing memory section is under ZONE_MOVABLE or not
370 - test and make it better memory offlining.
371 - support HugeTLB page migration and offlining.
372 - memmap removing at memory offline.
373 - physical remove memory.
374