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 4 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
159 'phys_index' : read-only and contains memory block id, same as XXX.
161 at read: contains online/offline state of memory.
162 at write: user can specify "online_kernel",
163 "online_movable", "online", "offline" command
164 which will be performed on all sections in the block.
165 'phys_device' : read-only: designed to show the name of physical memory
166 device. This is not well implemented now.
167 'removable' : read-only: contains an integer value indicating
168 whether the memory block is removable or not
169 removable. A value of 1 indicates that the memory
170 block is removable and a value of 0 indicates that
171 it is not removable. A memory block is removable only if
172 every section in the block is removable.
175 These directories/files appear after physical memory hotplug phase.
177 If CONFIG_NUMA is enabled the memoryXXX/ directories can also be accessed
178 via symbolic links located in the /sys/devices/system/node/node* directories.
181 /sys/devices/system/node/node0/memory9 -> ../../memory/memory9
183 A backlink will also be created:
184 /sys/devices/system/memory/memory9/node0 -> ../../node/node0
187 --------------------------------
188 4. Physical memory hot-add phase
189 --------------------------------
191 4.1 Hardware(Firmware) Support
193 On x86_64/ia64 platform, memory hotplug by ACPI is supported.
195 In general, the firmware (ACPI) which supports memory hotplug defines
196 memory class object of _HID "PNP0C80". When a notify is asserted to PNP0C80,
197 Linux's ACPI handler does hot-add memory to the system and calls a hotplug udev
198 script. This will be done automatically.
200 But scripts for memory hotplug are not contained in generic udev package(now).
201 You may have to write it by yourself or online/offline memory by hand.
202 Please see "How to online memory", "How to offline memory" in this text.
204 If firmware supports NUMA-node hotplug, and defines an object _HID "ACPI0004",
205 "PNP0A05", or "PNP0A06", notification is asserted to it, and ACPI handler
206 calls hotplug code for all of objects which are defined in it.
207 If memory device is found, memory hotplug code will be called.
210 4.2 Notify memory hot-add event by hand
212 On some architectures, the firmware may not notify the kernel of a memory
213 hotplug event. Therefore, the memory "probe" interface is supported to
214 explicitly notify the kernel. This interface depends on
215 CONFIG_ARCH_MEMORY_PROBE and can be configured on powerpc, sh, and x86
216 if hotplug is supported, although for x86 this should be handled by ACPI
219 Probe interface is located at
220 /sys/devices/system/memory/probe
222 You can tell the physical address of new memory to the kernel by
224 % echo start_address_of_new_memory > /sys/devices/system/memory/probe
226 Then, [start_address_of_new_memory, start_address_of_new_memory +
227 memory_block_size] memory range is hot-added. In this case, hotplug script is
228 not called (in current implementation). You'll have to online memory by
229 yourself. Please see "How to online memory" in this text.
232 ------------------------------
233 5. Logical Memory hot-add phase
234 ------------------------------
238 To see (online/offline) state of a memory block, read 'state' file.
240 % cat /sys/device/system/memory/memoryXXX/state
243 If the memory block is online, you'll read "online".
244 If the memory block is offline, you'll read "offline".
247 5.2. How to online memory
249 Even if the memory is hot-added, it is not at ready-to-use state.
250 For using newly added memory, you have to "online" the memory block.
252 For onlining, you have to write "online" to the memory block's state file as:
254 % echo online > /sys/devices/system/memory/memoryXXX/state
256 This onlining will not change the ZONE type of the target memory block,
257 If the memory block is in ZONE_NORMAL, you can change it to ZONE_MOVABLE:
259 % echo online_movable > /sys/devices/system/memory/memoryXXX/state
260 (NOTE: current limit: this memory block must be adjacent to ZONE_MOVABLE)
262 And if the memory block is in ZONE_MOVABLE, you can change it to ZONE_NORMAL:
264 % echo online_kernel > /sys/devices/system/memory/memoryXXX/state
265 (NOTE: current limit: this memory block must be adjacent to ZONE_NORMAL)
267 After this, memory block XXX's state will be 'online' and the amount of
268 available memory will be increased.
270 Currently, newly added memory is added as ZONE_NORMAL (for powerpc, ZONE_DMA).
271 This may be changed in future.
275 ------------------------
276 6. Logical memory remove
277 ------------------------
279 6.1 Memory offline and ZONE_MOVABLE
281 Memory offlining is more complicated than memory online. Because memory offline
282 has to make the whole memory block be unused, memory offline can fail if
283 the memory block includes memory which cannot be freed.
285 In general, memory offline can use 2 techniques.
287 (1) reclaim and free all memory in the memory block.
288 (2) migrate all pages in the memory block.
290 In the current implementation, Linux's memory offline uses method (2), freeing
291 all pages in the memory block by page migration. But not all pages are
292 migratable. Under current Linux, migratable pages are anonymous pages and
293 page caches. For offlining a memory block by migration, the kernel has to
294 guarantee that the memory block contains only migratable pages.
296 Now, a boot option for making a memory block which consists of migratable pages
297 is supported. By specifying "kernelcore=" or "movablecore=" boot option, you can
298 create ZONE_MOVABLE...a zone which is just used for movable pages.
299 (See also Documentation/kernel-parameters.txt)
301 Assume the system has "TOTAL" amount of memory at boot time, this boot option
302 creates ZONE_MOVABLE as following.
304 1) When kernelcore=YYYY boot option is used,
305 Size of memory not for movable pages (not for offline) is YYYY.
306 Size of memory for movable pages (for offline) is TOTAL-YYYY.
308 2) When movablecore=ZZZZ boot option is used,
309 Size of memory not for movable pages (not for offline) is TOTAL - ZZZZ.
310 Size of memory for movable pages (for offline) is ZZZZ.
313 Note: Unfortunately, there is no information to show which memory block belongs
314 to ZONE_MOVABLE. This is TBD.
317 6.2. How to offline memory
319 You can offline a memory block by using the same sysfs interface that was used
322 % echo offline > /sys/devices/system/memory/memoryXXX/state
324 If offline succeeds, the state of the memory block is changed to be "offline".
325 If it fails, some error core (like -EBUSY) will be returned by the kernel.
326 Even if a memory block does not belong to ZONE_MOVABLE, you can try to offline
327 it. If it doesn't contain 'unmovable' memory, you'll get success.
329 A memory block under ZONE_MOVABLE is considered to be able to be offlined
330 easily. But under some busy state, it may return -EBUSY. Even if a memory
331 block cannot be offlined due to -EBUSY, you can retry offlining it and may be
332 able to offline it (or not). (For example, a page is referred to by some kernel
333 internal call and released soon.)
336 Memory hotplug's design direction is to make the possibility of memory offlining
337 higher and to guarantee unplugging memory under any situation. But it needs
338 more work. Returning -EBUSY under some situation may be good because the user
339 can decide to retry more or not by himself. Currently, memory offlining code
340 does some amount of retry with 120 seconds timeout.
342 -------------------------
343 7. Physical memory remove
344 -------------------------
345 Need more implementation yet....
346 - Notification completion of remove works by OS to firmware.
347 - Guard from remove if not yet.
349 --------------------------------
350 8. Memory hotplug event notifier
351 --------------------------------
352 Memory hotplug has event notifier. There are 6 types of notification.
355 Generated before new memory becomes available in order to be able to
356 prepare subsystems to handle memory. The page allocator is still unable
357 to allocate from the new memory.
360 Generated if MEMORY_GOING_ONLINE fails.
363 Generated when memory has successfully brought online. The callback may
364 allocate pages from the new memory.
367 Generated to begin the process of offlining memory. Allocations are no
368 longer possible from the memory but some of the memory to be offlined
369 is still in use. The callback can be used to free memory known to a
370 subsystem from the indicated memory block.
372 MEMORY_CANCEL_OFFLINE
373 Generated if MEMORY_GOING_OFFLINE fails. Memory is available again from
374 the memory block that we attempted to offline.
377 Generated after offlining memory is complete.
379 A callback routine can be registered by
380 hotplug_memory_notifier(callback_func, priority)
382 The second argument of callback function (action) is event types of above.
383 The third argument is passed by pointer of struct memory_notify.
385 struct memory_notify {
386 unsigned long start_pfn;
387 unsigned long nr_pages;
388 int status_change_nid_normal;
389 int status_change_nid_high;
390 int status_change_nid;
393 start_pfn is start_pfn of online/offline memory.
394 nr_pages is # of pages of online/offline memory.
395 status_change_nid_normal is set node id when N_NORMAL_MEMORY of nodemask
396 is (will be) set/clear, if this is -1, then nodemask status is not changed.
397 status_change_nid_high is set node id when N_HIGH_MEMORY of nodemask
398 is (will be) set/clear, if this is -1, then nodemask status is not changed.
399 status_change_nid is set node id when N_MEMORY of nodemask is (will be)
400 set/clear. It means a new(memoryless) node gets new memory by online and a
401 node loses all memory. If this is -1, then nodemask status is not changed.
402 If status_changed_nid* >= 0, callback should create/discard structures for the
408 - allowing memory hot-add to ZONE_MOVABLE. maybe we need some switch like
409 sysctl or new control file.
410 - showing memory block and physical device relationship.
411 - showing memory block is under ZONE_MOVABLE or not
412 - test and make it better memory offlining.
413 - support HugeTLB page migration and offlining.
414 - memmap removing at memory offline.
415 - physical remove memory.