1 =============================== 1 ===============================
2 Documentation for /proc/sys/vm/ 2 Documentation for /proc/sys/vm/
3 =============================== 3 ===============================
4 4
5 kernel version 2.6.29 5 kernel version 2.6.29
6 6
7 Copyright (c) 1998, 1999, Rik van Riel <riel@n 7 Copyright (c) 1998, 1999, Rik van Riel <riel@nl.linux.org>
8 8
9 Copyright (c) 2008 Peter W. Morreale <p 9 Copyright (c) 2008 Peter W. Morreale <pmorreale@novell.com>
10 10
11 For general info and legal blurb, please look 11 For general info and legal blurb, please look in index.rst.
12 12
13 ---------------------------------------------- 13 ------------------------------------------------------------------------------
14 14
15 This file contains the documentation for the s 15 This file contains the documentation for the sysctl files in
16 /proc/sys/vm and is valid for Linux kernel ver 16 /proc/sys/vm and is valid for Linux kernel version 2.6.29.
17 17
18 The files in this directory can be used to tun 18 The files in this directory can be used to tune the operation
19 of the virtual memory (VM) subsystem of the Li 19 of the virtual memory (VM) subsystem of the Linux kernel and
20 the writeout of dirty data to disk. 20 the writeout of dirty data to disk.
21 21
22 Default values and initialization routines for 22 Default values and initialization routines for most of these
23 files can be found in mm/swap.c. 23 files can be found in mm/swap.c.
24 24
25 Currently, these files are in /proc/sys/vm: 25 Currently, these files are in /proc/sys/vm:
26 26
27 - admin_reserve_kbytes 27 - admin_reserve_kbytes
28 - compact_memory 28 - compact_memory
29 - compaction_proactiveness 29 - compaction_proactiveness
30 - compact_unevictable_allowed 30 - compact_unevictable_allowed
31 - dirty_background_bytes 31 - dirty_background_bytes
32 - dirty_background_ratio 32 - dirty_background_ratio
33 - dirty_bytes 33 - dirty_bytes
34 - dirty_expire_centisecs 34 - dirty_expire_centisecs
35 - dirty_ratio 35 - dirty_ratio
36 - dirtytime_expire_seconds 36 - dirtytime_expire_seconds
37 - dirty_writeback_centisecs 37 - dirty_writeback_centisecs
38 - drop_caches 38 - drop_caches
39 - enable_soft_offline <<
40 - extfrag_threshold 39 - extfrag_threshold
41 - highmem_is_dirtyable 40 - highmem_is_dirtyable
42 - hugetlb_shm_group 41 - hugetlb_shm_group
43 - laptop_mode 42 - laptop_mode
44 - legacy_va_layout 43 - legacy_va_layout
45 - lowmem_reserve_ratio 44 - lowmem_reserve_ratio
46 - max_map_count 45 - max_map_count
47 - mem_profiling (only if CONFIG_MEM_AL <<
48 - memory_failure_early_kill 46 - memory_failure_early_kill
49 - memory_failure_recovery 47 - memory_failure_recovery
50 - min_free_kbytes 48 - min_free_kbytes
51 - min_slab_ratio 49 - min_slab_ratio
52 - min_unmapped_ratio 50 - min_unmapped_ratio
53 - mmap_min_addr 51 - mmap_min_addr
54 - mmap_rnd_bits 52 - mmap_rnd_bits
55 - mmap_rnd_compat_bits 53 - mmap_rnd_compat_bits
56 - nr_hugepages 54 - nr_hugepages
57 - nr_hugepages_mempolicy 55 - nr_hugepages_mempolicy
58 - nr_overcommit_hugepages 56 - nr_overcommit_hugepages
59 - nr_trim_pages (only if CONFIG_MMU=n) 57 - nr_trim_pages (only if CONFIG_MMU=n)
60 - numa_zonelist_order 58 - numa_zonelist_order
61 - oom_dump_tasks 59 - oom_dump_tasks
62 - oom_kill_allocating_task 60 - oom_kill_allocating_task
63 - overcommit_kbytes 61 - overcommit_kbytes
64 - overcommit_memory 62 - overcommit_memory
65 - overcommit_ratio 63 - overcommit_ratio
66 - page-cluster 64 - page-cluster
67 - page_lock_unfairness 65 - page_lock_unfairness
68 - panic_on_oom 66 - panic_on_oom
69 - percpu_pagelist_high_fraction 67 - percpu_pagelist_high_fraction
70 - stat_interval 68 - stat_interval
71 - stat_refresh 69 - stat_refresh
72 - numa_stat 70 - numa_stat
73 - swappiness 71 - swappiness
74 - unprivileged_userfaultfd 72 - unprivileged_userfaultfd
75 - user_reserve_kbytes 73 - user_reserve_kbytes
76 - vfs_cache_pressure 74 - vfs_cache_pressure
77 - watermark_boost_factor 75 - watermark_boost_factor
78 - watermark_scale_factor 76 - watermark_scale_factor
79 - zone_reclaim_mode 77 - zone_reclaim_mode
80 78
81 79
82 admin_reserve_kbytes 80 admin_reserve_kbytes
83 ==================== 81 ====================
84 82
85 The amount of free memory in the system that s 83 The amount of free memory in the system that should be reserved for users
86 with the capability cap_sys_admin. 84 with the capability cap_sys_admin.
87 85
88 admin_reserve_kbytes defaults to min(3% of fre 86 admin_reserve_kbytes defaults to min(3% of free pages, 8MB)
89 87
90 That should provide enough for the admin to lo 88 That should provide enough for the admin to log in and kill a process,
91 if necessary, under the default overcommit 'gu 89 if necessary, under the default overcommit 'guess' mode.
92 90
93 Systems running under overcommit 'never' shoul 91 Systems running under overcommit 'never' should increase this to account
94 for the full Virtual Memory Size of programs u 92 for the full Virtual Memory Size of programs used to recover. Otherwise,
95 root may not be able to log in to recover the 93 root may not be able to log in to recover the system.
96 94
97 How do you calculate a minimum useful reserve? 95 How do you calculate a minimum useful reserve?
98 96
99 sshd or login + bash (or some other shell) + t 97 sshd or login + bash (or some other shell) + top (or ps, kill, etc.)
100 98
101 For overcommit 'guess', we can sum resident se 99 For overcommit 'guess', we can sum resident set sizes (RSS).
102 On x86_64 this is about 8MB. 100 On x86_64 this is about 8MB.
103 101
104 For overcommit 'never', we can take the max of 102 For overcommit 'never', we can take the max of their virtual sizes (VSZ)
105 and add the sum of their RSS. 103 and add the sum of their RSS.
106 On x86_64 this is about 128MB. 104 On x86_64 this is about 128MB.
107 105
108 Changing this takes effect whenever an applica 106 Changing this takes effect whenever an application requests memory.
109 107
110 108
111 compact_memory 109 compact_memory
112 ============== 110 ==============
113 111
114 Available only when CONFIG_COMPACTION is set. 112 Available only when CONFIG_COMPACTION is set. When 1 is written to the file,
115 all zones are compacted such that free memory 113 all zones are compacted such that free memory is available in contiguous
116 blocks where possible. This can be important f 114 blocks where possible. This can be important for example in the allocation of
117 huge pages although processes will also direct 115 huge pages although processes will also directly compact memory as required.
118 116
119 compaction_proactiveness 117 compaction_proactiveness
120 ======================== 118 ========================
121 119
122 This tunable takes a value in the range [0, 10 120 This tunable takes a value in the range [0, 100] with a default value of
123 20. This tunable determines how aggressively c 121 20. This tunable determines how aggressively compaction is done in the
124 background. Write of a non zero value to this 122 background. Write of a non zero value to this tunable will immediately
125 trigger the proactive compaction. Setting it t 123 trigger the proactive compaction. Setting it to 0 disables proactive compaction.
126 124
127 Note that compaction has a non-trivial system- 125 Note that compaction has a non-trivial system-wide impact as pages
128 belonging to different processes are moved aro 126 belonging to different processes are moved around, which could also lead
129 to latency spikes in unsuspecting applications 127 to latency spikes in unsuspecting applications. The kernel employs
130 various heuristics to avoid wasting CPU cycles 128 various heuristics to avoid wasting CPU cycles if it detects that
131 proactive compaction is not being effective. 129 proactive compaction is not being effective.
132 130
133 Be careful when setting it to extreme values l 131 Be careful when setting it to extreme values like 100, as that may
134 cause excessive background compaction activity 132 cause excessive background compaction activity.
135 133
136 compact_unevictable_allowed 134 compact_unevictable_allowed
137 =========================== 135 ===========================
138 136
139 Available only when CONFIG_COMPACTION is set. 137 Available only when CONFIG_COMPACTION is set. When set to 1, compaction is
140 allowed to examine the unevictable lru (mlocke 138 allowed to examine the unevictable lru (mlocked pages) for pages to compact.
141 This should be used on systems where stalls fo 139 This should be used on systems where stalls for minor page faults are an
142 acceptable trade for large contiguous free mem 140 acceptable trade for large contiguous free memory. Set to 0 to prevent
143 compaction from moving pages that are unevicta 141 compaction from moving pages that are unevictable. Default value is 1.
144 On CONFIG_PREEMPT_RT the default value is 0 in 142 On CONFIG_PREEMPT_RT the default value is 0 in order to avoid a page fault, due
145 to compaction, which would block the task from 143 to compaction, which would block the task from becoming active until the fault
146 is resolved. 144 is resolved.
147 145
148 146
149 dirty_background_bytes 147 dirty_background_bytes
150 ====================== 148 ======================
151 149
152 Contains the amount of dirty memory at which t 150 Contains the amount of dirty memory at which the background kernel
153 flusher threads will start writeback. 151 flusher threads will start writeback.
154 152
155 Note: 153 Note:
156 dirty_background_bytes is the counterpart of 154 dirty_background_bytes is the counterpart of dirty_background_ratio. Only
157 one of them may be specified at a time. When 155 one of them may be specified at a time. When one sysctl is written it is
158 immediately taken into account to evaluate t 156 immediately taken into account to evaluate the dirty memory limits and the
159 other appears as 0 when read. 157 other appears as 0 when read.
160 158
161 159
162 dirty_background_ratio 160 dirty_background_ratio
163 ====================== 161 ======================
164 162
165 Contains, as a percentage of total available m 163 Contains, as a percentage of total available memory that contains free pages
166 and reclaimable pages, the number of pages at 164 and reclaimable pages, the number of pages at which the background kernel
167 flusher threads will start writing out dirty d 165 flusher threads will start writing out dirty data.
168 166
169 The total available memory is not equal to tot 167 The total available memory is not equal to total system memory.
170 168
171 169
172 dirty_bytes 170 dirty_bytes
173 =========== 171 ===========
174 172
175 Contains the amount of dirty memory at which a 173 Contains the amount of dirty memory at which a process generating disk writes
176 will itself start writeback. 174 will itself start writeback.
177 175
178 Note: dirty_bytes is the counterpart of dirty_ 176 Note: dirty_bytes is the counterpart of dirty_ratio. Only one of them may be
179 specified at a time. When one sysctl is writte 177 specified at a time. When one sysctl is written it is immediately taken into
180 account to evaluate the dirty memory limits an 178 account to evaluate the dirty memory limits and the other appears as 0 when
181 read. 179 read.
182 180
183 Note: the minimum value allowed for dirty_byte 181 Note: the minimum value allowed for dirty_bytes is two pages (in bytes); any
184 value lower than this limit will be ignored an 182 value lower than this limit will be ignored and the old configuration will be
185 retained. 183 retained.
186 184
187 185
188 dirty_expire_centisecs 186 dirty_expire_centisecs
189 ====================== 187 ======================
190 188
191 This tunable is used to define when dirty data 189 This tunable is used to define when dirty data is old enough to be eligible
192 for writeout by the kernel flusher threads. I 190 for writeout by the kernel flusher threads. It is expressed in 100'ths
193 of a second. Data which has been dirty in-mem 191 of a second. Data which has been dirty in-memory for longer than this
194 interval will be written out next time a flush 192 interval will be written out next time a flusher thread wakes up.
195 193
196 194
197 dirty_ratio 195 dirty_ratio
198 =========== 196 ===========
199 197
200 Contains, as a percentage of total available m 198 Contains, as a percentage of total available memory that contains free pages
201 and reclaimable pages, the number of pages at 199 and reclaimable pages, the number of pages at which a process which is
202 generating disk writes will itself start writi 200 generating disk writes will itself start writing out dirty data.
203 201
204 The total available memory is not equal to tot 202 The total available memory is not equal to total system memory.
205 203
206 204
207 dirtytime_expire_seconds 205 dirtytime_expire_seconds
208 ======================== 206 ========================
209 207
210 When a lazytime inode is constantly having its 208 When a lazytime inode is constantly having its pages dirtied, the inode with
211 an updated timestamp will never get chance to 209 an updated timestamp will never get chance to be written out. And, if the
212 only thing that has happened on the file syste 210 only thing that has happened on the file system is a dirtytime inode caused
213 by an atime update, a worker will be scheduled 211 by an atime update, a worker will be scheduled to make sure that inode
214 eventually gets pushed out to disk. This tuna 212 eventually gets pushed out to disk. This tunable is used to define when dirty
215 inode is old enough to be eligible for writeba 213 inode is old enough to be eligible for writeback by the kernel flusher threads.
216 And, it is also used as the interval to wakeup 214 And, it is also used as the interval to wakeup dirtytime_writeback thread.
217 215
218 216
219 dirty_writeback_centisecs 217 dirty_writeback_centisecs
220 ========================= 218 =========================
221 219
222 The kernel flusher threads will periodically w 220 The kernel flusher threads will periodically wake up and write `old` data
223 out to disk. This tunable expresses the inter 221 out to disk. This tunable expresses the interval between those wakeups, in
224 100'ths of a second. 222 100'ths of a second.
225 223
226 Setting this to zero disables periodic writeba 224 Setting this to zero disables periodic writeback altogether.
227 225
228 226
229 drop_caches 227 drop_caches
230 =========== 228 ===========
231 229
232 Writing to this will cause the kernel to drop 230 Writing to this will cause the kernel to drop clean caches, as well as
233 reclaimable slab objects like dentries and ino 231 reclaimable slab objects like dentries and inodes. Once dropped, their
234 memory becomes free. 232 memory becomes free.
235 233
236 To free pagecache:: 234 To free pagecache::
237 235
238 echo 1 > /proc/sys/vm/drop_caches 236 echo 1 > /proc/sys/vm/drop_caches
239 237
240 To free reclaimable slab objects (includes den 238 To free reclaimable slab objects (includes dentries and inodes)::
241 239
242 echo 2 > /proc/sys/vm/drop_caches 240 echo 2 > /proc/sys/vm/drop_caches
243 241
244 To free slab objects and pagecache:: 242 To free slab objects and pagecache::
245 243
246 echo 3 > /proc/sys/vm/drop_caches 244 echo 3 > /proc/sys/vm/drop_caches
247 245
248 This is a non-destructive operation and will n 246 This is a non-destructive operation and will not free any dirty objects.
249 To increase the number of objects freed by thi 247 To increase the number of objects freed by this operation, the user may run
250 `sync` prior to writing to /proc/sys/vm/drop_c 248 `sync` prior to writing to /proc/sys/vm/drop_caches. This will minimize the
251 number of dirty objects on the system and crea 249 number of dirty objects on the system and create more candidates to be
252 dropped. 250 dropped.
253 251
254 This file is not a means to control the growth 252 This file is not a means to control the growth of the various kernel caches
255 (inodes, dentries, pagecache, etc...) These o 253 (inodes, dentries, pagecache, etc...) These objects are automatically
256 reclaimed by the kernel when memory is needed 254 reclaimed by the kernel when memory is needed elsewhere on the system.
257 255
258 Use of this file can cause performance problem 256 Use of this file can cause performance problems. Since it discards cached
259 objects, it may cost a significant amount of I 257 objects, it may cost a significant amount of I/O and CPU to recreate the
260 dropped objects, especially if they were under 258 dropped objects, especially if they were under heavy use. Because of this,
261 use outside of a testing or debugging environm 259 use outside of a testing or debugging environment is not recommended.
262 260
263 You may see informational messages in your ker 261 You may see informational messages in your kernel log when this file is
264 used:: 262 used::
265 263
266 cat (1234): drop_caches: 3 264 cat (1234): drop_caches: 3
267 265
268 These are informational only. They do not mea 266 These are informational only. They do not mean that anything is wrong
269 with your system. To disable them, echo 4 (bi 267 with your system. To disable them, echo 4 (bit 2) into drop_caches.
270 268
271 enable_soft_offline <<
272 =================== <<
273 Correctable memory errors are very common on s <<
274 solution for memory pages having (excessive) c <<
275 <<
276 For different types of page, soft-offline has <<
277 <<
278 - For a raw error page, soft-offline migrates <<
279 a new raw page. <<
280 <<
281 - For a page that is part of a transparent hug <<
282 transparent hugepage into raw pages, then mi <<
283 As a result, user is transparently backed by <<
284 memory access performance. <<
285 <<
286 - For a page that is part of a HugeTLB hugepag <<
287 the entire HugeTLB hugepage, during which a <<
288 as migration target. Then the original huge <<
289 pages without compensation, reducing the cap <<
290 <<
291 It is user's call to choose between reliabilit <<
292 physical memory) vs performance / capacity imp <<
293 HugeTLB cases. <<
294 <<
295 For all architectures, enable_soft_offline con <<
296 memory pages. When set to 1, kernel attempts <<
297 whenever it thinks needed. When set to 0, ker <<
298 the request to soft offline the pages. Its de <<
299 <<
300 It is worth mentioning that after setting enab <<
301 following requests to soft offline pages will <<
302 <<
303 - Request to soft offline pages from RAS Corre <<
304 <<
305 - On ARM, the request to soft offline pages fr <<
306 <<
307 - On PARISC, the request to soft offline pages <<
308 269
309 extfrag_threshold 270 extfrag_threshold
310 ================= 271 =================
311 272
312 This parameter affects whether the kernel will 273 This parameter affects whether the kernel will compact memory or direct
313 reclaim to satisfy a high-order allocation. Th 274 reclaim to satisfy a high-order allocation. The extfrag/extfrag_index file in
314 debugfs shows what the fragmentation index for 275 debugfs shows what the fragmentation index for each order is in each zone in
315 the system. Values tending towards 0 imply all 276 the system. Values tending towards 0 imply allocations would fail due to lack
316 of memory, values towards 1000 imply failures 277 of memory, values towards 1000 imply failures are due to fragmentation and -1
317 implies that the allocation will succeed as lo 278 implies that the allocation will succeed as long as watermarks are met.
318 279
319 The kernel will not compact memory in a zone i 280 The kernel will not compact memory in a zone if the
320 fragmentation index is <= extfrag_threshold. T 281 fragmentation index is <= extfrag_threshold. The default value is 500.
321 282
322 283
323 highmem_is_dirtyable 284 highmem_is_dirtyable
324 ==================== 285 ====================
325 286
326 Available only for systems with CONFIG_HIGHMEM 287 Available only for systems with CONFIG_HIGHMEM enabled (32b systems).
327 288
328 This parameter controls whether the high memor 289 This parameter controls whether the high memory is considered for dirty
329 writers throttling. This is not the case by d 290 writers throttling. This is not the case by default which means that
330 only the amount of memory directly visible/usa 291 only the amount of memory directly visible/usable by the kernel can
331 be dirtied. As a result, on systems with a lar 292 be dirtied. As a result, on systems with a large amount of memory and
332 lowmem basically depleted writers might be thr 293 lowmem basically depleted writers might be throttled too early and
333 streaming writes can get very slow. 294 streaming writes can get very slow.
334 295
335 Changing the value to non zero would allow mor 296 Changing the value to non zero would allow more memory to be dirtied
336 and thus allow writers to write more data whic 297 and thus allow writers to write more data which can be flushed to the
337 storage more effectively. Note this also comes 298 storage more effectively. Note this also comes with a risk of pre-mature
338 OOM killer because some writers (e.g. direct b 299 OOM killer because some writers (e.g. direct block device writes) can
339 only use the low memory and they can fill it u 300 only use the low memory and they can fill it up with dirty data without
340 any throttling. 301 any throttling.
341 302
342 303
343 hugetlb_shm_group 304 hugetlb_shm_group
344 ================= 305 =================
345 306
346 hugetlb_shm_group contains group id that is al 307 hugetlb_shm_group contains group id that is allowed to create SysV
347 shared memory segment using hugetlb page. 308 shared memory segment using hugetlb page.
348 309
349 310
350 laptop_mode 311 laptop_mode
351 =========== 312 ===========
352 313
353 laptop_mode is a knob that controls "laptop mo 314 laptop_mode is a knob that controls "laptop mode". All the things that are
354 controlled by this knob are discussed in Docum 315 controlled by this knob are discussed in Documentation/admin-guide/laptops/laptop-mode.rst.
355 316
356 317
357 legacy_va_layout 318 legacy_va_layout
358 ================ 319 ================
359 320
360 If non-zero, this sysctl disables the new 32-b 321 If non-zero, this sysctl disables the new 32-bit mmap layout - the kernel
361 will use the legacy (2.4) layout for all proce 322 will use the legacy (2.4) layout for all processes.
362 323
363 324
364 lowmem_reserve_ratio 325 lowmem_reserve_ratio
365 ==================== 326 ====================
366 327
367 For some specialised workloads on highmem mach 328 For some specialised workloads on highmem machines it is dangerous for
368 the kernel to allow process memory to be alloc 329 the kernel to allow process memory to be allocated from the "lowmem"
369 zone. This is because that memory could then 330 zone. This is because that memory could then be pinned via the mlock()
370 system call, or by unavailability of swapspace 331 system call, or by unavailability of swapspace.
371 332
372 And on large highmem machines this lack of rec 333 And on large highmem machines this lack of reclaimable lowmem memory
373 can be fatal. 334 can be fatal.
374 335
375 So the Linux page allocator has a mechanism wh 336 So the Linux page allocator has a mechanism which prevents allocations
376 which *could* use highmem from using too much 337 which *could* use highmem from using too much lowmem. This means that
377 a certain amount of lowmem is defended from th 338 a certain amount of lowmem is defended from the possibility of being
378 captured into pinned user memory. 339 captured into pinned user memory.
379 340
380 (The same argument applies to the old 16 megab 341 (The same argument applies to the old 16 megabyte ISA DMA region. This
381 mechanism will also defend that region from al 342 mechanism will also defend that region from allocations which could use
382 highmem or lowmem). 343 highmem or lowmem).
383 344
384 The `lowmem_reserve_ratio` tunable determines 345 The `lowmem_reserve_ratio` tunable determines how aggressive the kernel is
385 in defending these lower zones. 346 in defending these lower zones.
386 347
387 If you have a machine which uses highmem or IS 348 If you have a machine which uses highmem or ISA DMA and your
388 applications are using mlock(), or if you are 349 applications are using mlock(), or if you are running with no swap then
389 you probably should change the lowmem_reserve_ 350 you probably should change the lowmem_reserve_ratio setting.
390 351
391 The lowmem_reserve_ratio is an array. You can 352 The lowmem_reserve_ratio is an array. You can see them by reading this file::
392 353
393 % cat /proc/sys/vm/lowmem_reserve_rati 354 % cat /proc/sys/vm/lowmem_reserve_ratio
394 256 256 32 355 256 256 32
395 356
396 But, these values are not used directly. The k 357 But, these values are not used directly. The kernel calculates # of protection
397 pages for each zones from them. These are show 358 pages for each zones from them. These are shown as array of protection pages
398 in /proc/zoneinfo like the following. (This is !! 359 in /proc/zoneinfo like followings. (This is an example of x86-64 box).
399 Each zone has an array of protection pages lik 360 Each zone has an array of protection pages like this::
400 361
401 Node 0, zone DMA 362 Node 0, zone DMA
402 pages free 1355 363 pages free 1355
403 min 3 364 min 3
404 low 3 365 low 3
405 high 4 366 high 4
406 : 367 :
407 : 368 :
408 numa_other 0 369 numa_other 0
409 protection: (0, 2004, 2004, 2004) 370 protection: (0, 2004, 2004, 2004)
410 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 371 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
411 pagesets 372 pagesets
412 cpu: 0 pcp: 0 373 cpu: 0 pcp: 0
413 : 374 :
414 375
415 These protections are added to score to judge 376 These protections are added to score to judge whether this zone should be used
416 for page allocation or should be reclaimed. 377 for page allocation or should be reclaimed.
417 378
418 In this example, if normal pages (index=2) are 379 In this example, if normal pages (index=2) are required to this DMA zone and
419 watermark[WMARK_HIGH] is used for watermark, t 380 watermark[WMARK_HIGH] is used for watermark, the kernel judges this zone should
420 not be used because pages_free(1355) is smalle 381 not be used because pages_free(1355) is smaller than watermark + protection[2]
421 (4 + 2004 = 2008). If this protection value is 382 (4 + 2004 = 2008). If this protection value is 0, this zone would be used for
422 normal page requirement. If requirement is DMA 383 normal page requirement. If requirement is DMA zone(index=0), protection[0]
423 (=0) is used. 384 (=0) is used.
424 385
425 zone[i]'s protection[j] is calculated by follo 386 zone[i]'s protection[j] is calculated by following expression::
426 387
427 (i < j): 388 (i < j):
428 zone[i]->protection[j] 389 zone[i]->protection[j]
429 = (total sums of managed_pages from zone[i 390 = (total sums of managed_pages from zone[i+1] to zone[j] on the node)
430 / lowmem_reserve_ratio[i]; 391 / lowmem_reserve_ratio[i];
431 (i = j): 392 (i = j):
432 (should not be protected. = 0; 393 (should not be protected. = 0;
433 (i > j): 394 (i > j):
434 (not necessary, but looks 0) 395 (not necessary, but looks 0)
435 396
436 The default values of lowmem_reserve_ratio[i] 397 The default values of lowmem_reserve_ratio[i] are
437 398
438 === ==================================== 399 === ====================================
439 256 (if zone[i] means DMA or DMA32 zone) 400 256 (if zone[i] means DMA or DMA32 zone)
440 32 (others) 401 32 (others)
441 === ==================================== 402 === ====================================
442 403
443 As above expression, they are reciprocal numbe 404 As above expression, they are reciprocal number of ratio.
444 256 means 1/256. # of protection pages becomes 405 256 means 1/256. # of protection pages becomes about "0.39%" of total managed
445 pages of higher zones on the node. 406 pages of higher zones on the node.
446 407
447 If you would like to protect more pages, small 408 If you would like to protect more pages, smaller values are effective.
448 The minimum value is 1 (1/1 -> 100%). The valu 409 The minimum value is 1 (1/1 -> 100%). The value less than 1 completely
449 disables protection of the pages. 410 disables protection of the pages.
450 411
451 412
452 max_map_count: 413 max_map_count:
453 ============== 414 ==============
454 415
455 This file contains the maximum number of memor 416 This file contains the maximum number of memory map areas a process
456 may have. Memory map areas are used as a side- 417 may have. Memory map areas are used as a side-effect of calling
457 malloc, directly by mmap, mprotect, and madvis 418 malloc, directly by mmap, mprotect, and madvise, and also when loading
458 shared libraries. 419 shared libraries.
459 420
460 While most applications need less than a thous 421 While most applications need less than a thousand maps, certain
461 programs, particularly malloc debuggers, may c 422 programs, particularly malloc debuggers, may consume lots of them,
462 e.g., up to one or two maps per allocation. 423 e.g., up to one or two maps per allocation.
463 424
464 The default value is 65530. 425 The default value is 65530.
465 426
466 427
467 mem_profiling <<
468 ============== <<
469 <<
470 Enable memory profiling (when CONFIG_MEM_ALLOC <<
471 <<
472 1: Enable memory profiling. <<
473 <<
474 0: Disable memory profiling. <<
475 <<
476 Enabling memory profiling introduces a small p <<
477 memory allocations. <<
478 <<
479 The default value depends on CONFIG_MEM_ALLOC_ <<
480 <<
481 <<
482 memory_failure_early_kill: 428 memory_failure_early_kill:
483 ========================== 429 ==========================
484 430
485 Control how to kill processes when uncorrected 431 Control how to kill processes when uncorrected memory error (typically
486 a 2bit error in a memory module) is detected i 432 a 2bit error in a memory module) is detected in the background by hardware
487 that cannot be handled by the kernel. In some 433 that cannot be handled by the kernel. In some cases (like the page
488 still having a valid copy on disk) the kernel 434 still having a valid copy on disk) the kernel will handle the failure
489 transparently without affecting any applicatio 435 transparently without affecting any applications. But if there is
490 no other up-to-date copy of the data it will k !! 436 no other uptodate copy of the data it will kill to prevent any data
491 corruptions from propagating. 437 corruptions from propagating.
492 438
493 1: Kill all processes that have the corrupted 439 1: Kill all processes that have the corrupted and not reloadable page mapped
494 as soon as the corruption is detected. Note t 440 as soon as the corruption is detected. Note this is not supported
495 for a few types of pages, like kernel internal 441 for a few types of pages, like kernel internally allocated data or
496 the swap cache, but works for the majority of 442 the swap cache, but works for the majority of user pages.
497 443
498 0: Only unmap the corrupted page from all proc 444 0: Only unmap the corrupted page from all processes and only kill a process
499 who tries to access it. 445 who tries to access it.
500 446
501 The kill is done using a catchable SIGBUS with 447 The kill is done using a catchable SIGBUS with BUS_MCEERR_AO, so processes can
502 handle this if they want to. 448 handle this if they want to.
503 449
504 This is only active on architectures/platforms 450 This is only active on architectures/platforms with advanced machine
505 check handling and depends on the hardware cap 451 check handling and depends on the hardware capabilities.
506 452
507 Applications can override this setting individ 453 Applications can override this setting individually with the PR_MCE_KILL prctl
508 454
509 455
510 memory_failure_recovery 456 memory_failure_recovery
511 ======================= 457 =======================
512 458
513 Enable memory failure recovery (when supported 459 Enable memory failure recovery (when supported by the platform)
514 460
515 1: Attempt recovery. 461 1: Attempt recovery.
516 462
517 0: Always panic on a memory failure. 463 0: Always panic on a memory failure.
518 464
519 465
520 min_free_kbytes 466 min_free_kbytes
521 =============== 467 ===============
522 468
523 This is used to force the Linux VM to keep a m 469 This is used to force the Linux VM to keep a minimum number
524 of kilobytes free. The VM uses this number to 470 of kilobytes free. The VM uses this number to compute a
525 watermark[WMARK_MIN] value for each lowmem zon 471 watermark[WMARK_MIN] value for each lowmem zone in the system.
526 Each lowmem zone gets a number of reserved fre 472 Each lowmem zone gets a number of reserved free pages based
527 proportionally on its size. 473 proportionally on its size.
528 474
529 Some minimal amount of memory is needed to sat 475 Some minimal amount of memory is needed to satisfy PF_MEMALLOC
530 allocations; if you set this to lower than 102 476 allocations; if you set this to lower than 1024KB, your system will
531 become subtly broken, and prone to deadlock un 477 become subtly broken, and prone to deadlock under high loads.
532 478
533 Setting this too high will OOM your machine in 479 Setting this too high will OOM your machine instantly.
534 480
535 481
536 min_slab_ratio 482 min_slab_ratio
537 ============== 483 ==============
538 484
539 This is available only on NUMA kernels. 485 This is available only on NUMA kernels.
540 486
541 A percentage of the total pages in each zone. 487 A percentage of the total pages in each zone. On Zone reclaim
542 (fallback from the local zone occurs) slabs wi 488 (fallback from the local zone occurs) slabs will be reclaimed if more
543 than this percentage of pages in a zone are re 489 than this percentage of pages in a zone are reclaimable slab pages.
544 This insures that the slab growth stays under 490 This insures that the slab growth stays under control even in NUMA
545 systems that rarely perform global reclaim. 491 systems that rarely perform global reclaim.
546 492
547 The default is 5 percent. 493 The default is 5 percent.
548 494
549 Note that slab reclaim is triggered in a per z 495 Note that slab reclaim is triggered in a per zone / node fashion.
550 The process of reclaiming slab memory is curre 496 The process of reclaiming slab memory is currently not node specific
551 and may not be fast. 497 and may not be fast.
552 498
553 499
554 min_unmapped_ratio 500 min_unmapped_ratio
555 ================== 501 ==================
556 502
557 This is available only on NUMA kernels. 503 This is available only on NUMA kernels.
558 504
559 This is a percentage of the total pages in eac 505 This is a percentage of the total pages in each zone. Zone reclaim will
560 only occur if more than this percentage of pag 506 only occur if more than this percentage of pages are in a state that
561 zone_reclaim_mode allows to be reclaimed. 507 zone_reclaim_mode allows to be reclaimed.
562 508
563 If zone_reclaim_mode has the value 4 OR'd, the 509 If zone_reclaim_mode has the value 4 OR'd, then the percentage is compared
564 against all file-backed unmapped pages includi 510 against all file-backed unmapped pages including swapcache pages and tmpfs
565 files. Otherwise, only unmapped pages backed b 511 files. Otherwise, only unmapped pages backed by normal files but not tmpfs
566 files and similar are considered. 512 files and similar are considered.
567 513
568 The default is 1 percent. 514 The default is 1 percent.
569 515
570 516
571 mmap_min_addr 517 mmap_min_addr
572 ============= 518 =============
573 519
574 This file indicates the amount of address spac 520 This file indicates the amount of address space which a user process will
575 be restricted from mmapping. Since kernel nul 521 be restricted from mmapping. Since kernel null dereference bugs could
576 accidentally operate based on the information 522 accidentally operate based on the information in the first couple of pages
577 of memory userspace processes should not be al 523 of memory userspace processes should not be allowed to write to them. By
578 default this value is set to 0 and no protecti 524 default this value is set to 0 and no protections will be enforced by the
579 security module. Setting this value to someth 525 security module. Setting this value to something like 64k will allow the
580 vast majority of applications to work correctl 526 vast majority of applications to work correctly and provide defense in depth
581 against future potential kernel bugs. 527 against future potential kernel bugs.
582 528
583 529
584 mmap_rnd_bits 530 mmap_rnd_bits
585 ============= 531 =============
586 532
587 This value can be used to select the number of 533 This value can be used to select the number of bits to use to
588 determine the random offset to the base addres 534 determine the random offset to the base address of vma regions
589 resulting from mmap allocations on architectur 535 resulting from mmap allocations on architectures which support
590 tuning address space randomization. This valu 536 tuning address space randomization. This value will be bounded
591 by the architecture's minimum and maximum supp 537 by the architecture's minimum and maximum supported values.
592 538
593 This value can be changed after boot using the 539 This value can be changed after boot using the
594 /proc/sys/vm/mmap_rnd_bits tunable 540 /proc/sys/vm/mmap_rnd_bits tunable
595 541
596 542
597 mmap_rnd_compat_bits 543 mmap_rnd_compat_bits
598 ==================== 544 ====================
599 545
600 This value can be used to select the number of 546 This value can be used to select the number of bits to use to
601 determine the random offset to the base addres 547 determine the random offset to the base address of vma regions
602 resulting from mmap allocations for applicatio 548 resulting from mmap allocations for applications run in
603 compatibility mode on architectures which supp 549 compatibility mode on architectures which support tuning address
604 space randomization. This value will be bound 550 space randomization. This value will be bounded by the
605 architecture's minimum and maximum supported v 551 architecture's minimum and maximum supported values.
606 552
607 This value can be changed after boot using the 553 This value can be changed after boot using the
608 /proc/sys/vm/mmap_rnd_compat_bits tunable 554 /proc/sys/vm/mmap_rnd_compat_bits tunable
609 555
610 556
611 nr_hugepages 557 nr_hugepages
612 ============ 558 ============
613 559
614 Change the minimum size of the hugepage pool. 560 Change the minimum size of the hugepage pool.
615 561
616 See Documentation/admin-guide/mm/hugetlbpage.r 562 See Documentation/admin-guide/mm/hugetlbpage.rst
617 563
618 564
619 hugetlb_optimize_vmemmap 565 hugetlb_optimize_vmemmap
620 ======================== 566 ========================
621 567
622 This knob is not available when the size of 's 568 This knob is not available when the size of 'struct page' (a structure defined
623 in include/linux/mm_types.h) is not power of t 569 in include/linux/mm_types.h) is not power of two (an unusual system config could
624 result in this). 570 result in this).
625 571
626 Enable (set to 1) or disable (set to 0) HugeTL 572 Enable (set to 1) or disable (set to 0) HugeTLB Vmemmap Optimization (HVO).
627 573
628 Once enabled, the vmemmap pages of subsequent 574 Once enabled, the vmemmap pages of subsequent allocation of HugeTLB pages from
629 buddy allocator will be optimized (7 pages per 575 buddy allocator will be optimized (7 pages per 2MB HugeTLB page and 4095 pages
630 per 1GB HugeTLB page), whereas already allocat 576 per 1GB HugeTLB page), whereas already allocated HugeTLB pages will not be
631 optimized. When those optimized HugeTLB pages 577 optimized. When those optimized HugeTLB pages are freed from the HugeTLB pool
632 to the buddy allocator, the vmemmap pages repr 578 to the buddy allocator, the vmemmap pages representing that range needs to be
633 remapped again and the vmemmap pages discarded 579 remapped again and the vmemmap pages discarded earlier need to be rellocated
634 again. If your use case is that HugeTLB pages 580 again. If your use case is that HugeTLB pages are allocated 'on the fly' (e.g.
635 never explicitly allocating HugeTLB pages with 581 never explicitly allocating HugeTLB pages with 'nr_hugepages' but only set
636 'nr_overcommit_hugepages', those overcommitted 582 'nr_overcommit_hugepages', those overcommitted HugeTLB pages are allocated 'on
637 the fly') instead of being pulled from the Hug 583 the fly') instead of being pulled from the HugeTLB pool, you should weigh the
638 benefits of memory savings against the more ov 584 benefits of memory savings against the more overhead (~2x slower than before)
639 of allocation or freeing HugeTLB pages between 585 of allocation or freeing HugeTLB pages between the HugeTLB pool and the buddy
640 allocator. Another behavior to note is that i 586 allocator. Another behavior to note is that if the system is under heavy memory
641 pressure, it could prevent the user from freei 587 pressure, it could prevent the user from freeing HugeTLB pages from the HugeTLB
642 pool to the buddy allocator since the allocati 588 pool to the buddy allocator since the allocation of vmemmap pages could be
643 failed, you have to retry later if your system 589 failed, you have to retry later if your system encounter this situation.
644 590
645 Once disabled, the vmemmap pages of subsequent 591 Once disabled, the vmemmap pages of subsequent allocation of HugeTLB pages from
646 buddy allocator will not be optimized meaning 592 buddy allocator will not be optimized meaning the extra overhead at allocation
647 time from buddy allocator disappears, whereas 593 time from buddy allocator disappears, whereas already optimized HugeTLB pages
648 will not be affected. If you want to make sur 594 will not be affected. If you want to make sure there are no optimized HugeTLB
649 pages, you can set "nr_hugepages" to 0 first a 595 pages, you can set "nr_hugepages" to 0 first and then disable this. Note that
650 writing 0 to nr_hugepages will make any "in us 596 writing 0 to nr_hugepages will make any "in use" HugeTLB pages become surplus
651 pages. So, those surplus pages are still opti 597 pages. So, those surplus pages are still optimized until they are no longer
652 in use. You would need to wait for those surp 598 in use. You would need to wait for those surplus pages to be released before
653 there are no optimized pages in the system. 599 there are no optimized pages in the system.
654 600
655 601
656 nr_hugepages_mempolicy 602 nr_hugepages_mempolicy
657 ====================== 603 ======================
658 604
659 Change the size of the hugepage pool at run-ti 605 Change the size of the hugepage pool at run-time on a specific
660 set of NUMA nodes. 606 set of NUMA nodes.
661 607
662 See Documentation/admin-guide/mm/hugetlbpage.r 608 See Documentation/admin-guide/mm/hugetlbpage.rst
663 609
664 610
665 nr_overcommit_hugepages 611 nr_overcommit_hugepages
666 ======================= 612 =======================
667 613
668 Change the maximum size of the hugepage pool. 614 Change the maximum size of the hugepage pool. The maximum is
669 nr_hugepages + nr_overcommit_hugepages. 615 nr_hugepages + nr_overcommit_hugepages.
670 616
671 See Documentation/admin-guide/mm/hugetlbpage.r 617 See Documentation/admin-guide/mm/hugetlbpage.rst
672 618
673 619
674 nr_trim_pages 620 nr_trim_pages
675 ============= 621 =============
676 622
677 This is available only on NOMMU kernels. 623 This is available only on NOMMU kernels.
678 624
679 This value adjusts the excess page trimming be 625 This value adjusts the excess page trimming behaviour of power-of-2 aligned
680 NOMMU mmap allocations. 626 NOMMU mmap allocations.
681 627
682 A value of 0 disables trimming of allocations 628 A value of 0 disables trimming of allocations entirely, while a value of 1
683 trims excess pages aggressively. Any value >= 629 trims excess pages aggressively. Any value >= 1 acts as the watermark where
684 trimming of allocations is initiated. 630 trimming of allocations is initiated.
685 631
686 The default value is 1. 632 The default value is 1.
687 633
688 See Documentation/admin-guide/mm/nommu-mmap.rs 634 See Documentation/admin-guide/mm/nommu-mmap.rst for more information.
689 635
690 636
691 numa_zonelist_order 637 numa_zonelist_order
692 =================== 638 ===================
693 639
694 This sysctl is only for NUMA and it is depreca 640 This sysctl is only for NUMA and it is deprecated. Anything but
695 Node order will fail! 641 Node order will fail!
696 642
697 'where the memory is allocated from' is contro 643 'where the memory is allocated from' is controlled by zonelists.
698 644
699 (This documentation ignores ZONE_HIGHMEM/ZONE_ 645 (This documentation ignores ZONE_HIGHMEM/ZONE_DMA32 for simple explanation.
700 you may be able to read ZONE_DMA as ZONE_DMA32 646 you may be able to read ZONE_DMA as ZONE_DMA32...)
701 647
702 In non-NUMA case, a zonelist for GFP_KERNEL is 648 In non-NUMA case, a zonelist for GFP_KERNEL is ordered as following.
703 ZONE_NORMAL -> ZONE_DMA 649 ZONE_NORMAL -> ZONE_DMA
704 This means that a memory allocation request fo 650 This means that a memory allocation request for GFP_KERNEL will
705 get memory from ZONE_DMA only when ZONE_NORMAL 651 get memory from ZONE_DMA only when ZONE_NORMAL is not available.
706 652
707 In NUMA case, you can think of following 2 typ 653 In NUMA case, you can think of following 2 types of order.
708 Assume 2 node NUMA and below is zonelist of No 654 Assume 2 node NUMA and below is zonelist of Node(0)'s GFP_KERNEL::
709 655
710 (A) Node(0) ZONE_NORMAL -> Node(0) ZONE_DMA 656 (A) Node(0) ZONE_NORMAL -> Node(0) ZONE_DMA -> Node(1) ZONE_NORMAL
711 (B) Node(0) ZONE_NORMAL -> Node(1) ZONE_NORM 657 (B) Node(0) ZONE_NORMAL -> Node(1) ZONE_NORMAL -> Node(0) ZONE_DMA.
712 658
713 Type(A) offers the best locality for processes 659 Type(A) offers the best locality for processes on Node(0), but ZONE_DMA
714 will be used before ZONE_NORMAL exhaustion. Th 660 will be used before ZONE_NORMAL exhaustion. This increases possibility of
715 out-of-memory(OOM) of ZONE_DMA because ZONE_DM 661 out-of-memory(OOM) of ZONE_DMA because ZONE_DMA is tend to be small.
716 662
717 Type(B) cannot offer the best locality but is 663 Type(B) cannot offer the best locality but is more robust against OOM of
718 the DMA zone. 664 the DMA zone.
719 665
720 Type(A) is called as "Node" order. Type (B) is 666 Type(A) is called as "Node" order. Type (B) is "Zone" order.
721 667
722 "Node order" orders the zonelists by node, the 668 "Node order" orders the zonelists by node, then by zone within each node.
723 Specify "[Nn]ode" for node order 669 Specify "[Nn]ode" for node order
724 670
725 "Zone Order" orders the zonelists by zone type 671 "Zone Order" orders the zonelists by zone type, then by node within each
726 zone. Specify "[Zz]one" for zone order. 672 zone. Specify "[Zz]one" for zone order.
727 673
728 Specify "[Dd]efault" to request automatic conf 674 Specify "[Dd]efault" to request automatic configuration.
729 675
730 On 32-bit, the Normal zone needs to be preserv 676 On 32-bit, the Normal zone needs to be preserved for allocations accessible
731 by the kernel, so "zone" order will be selecte 677 by the kernel, so "zone" order will be selected.
732 678
733 On 64-bit, devices that require DMA32/DMA are 679 On 64-bit, devices that require DMA32/DMA are relatively rare, so "node"
734 order will be selected. 680 order will be selected.
735 681
736 Default order is recommended unless this is ca 682 Default order is recommended unless this is causing problems for your
737 system/application. 683 system/application.
738 684
739 685
740 oom_dump_tasks 686 oom_dump_tasks
741 ============== 687 ==============
742 688
743 Enables a system-wide task dump (excluding ker 689 Enables a system-wide task dump (excluding kernel threads) to be produced
744 when the kernel performs an OOM-killing and in 690 when the kernel performs an OOM-killing and includes such information as
745 pid, uid, tgid, vm size, rss, pgtables_bytes, 691 pid, uid, tgid, vm size, rss, pgtables_bytes, swapents, oom_score_adj
746 score, and name. This is helpful to determine 692 score, and name. This is helpful to determine why the OOM killer was
747 invoked, to identify the rogue task that cause 693 invoked, to identify the rogue task that caused it, and to determine why
748 the OOM killer chose the task it did to kill. 694 the OOM killer chose the task it did to kill.
749 695
750 If this is set to zero, this information is su 696 If this is set to zero, this information is suppressed. On very
751 large systems with thousands of tasks it may n 697 large systems with thousands of tasks it may not be feasible to dump
752 the memory state information for each one. Su 698 the memory state information for each one. Such systems should not
753 be forced to incur a performance penalty in OO 699 be forced to incur a performance penalty in OOM conditions when the
754 information may not be desired. 700 information may not be desired.
755 701
756 If this is set to non-zero, this information i 702 If this is set to non-zero, this information is shown whenever the
757 OOM killer actually kills a memory-hogging tas 703 OOM killer actually kills a memory-hogging task.
758 704
759 The default value is 1 (enabled). 705 The default value is 1 (enabled).
760 706
761 707
762 oom_kill_allocating_task 708 oom_kill_allocating_task
763 ======================== 709 ========================
764 710
765 This enables or disables killing the OOM-trigg 711 This enables or disables killing the OOM-triggering task in
766 out-of-memory situations. 712 out-of-memory situations.
767 713
768 If this is set to zero, the OOM killer will sc 714 If this is set to zero, the OOM killer will scan through the entire
769 tasklist and select a task based on heuristics 715 tasklist and select a task based on heuristics to kill. This normally
770 selects a rogue memory-hogging task that frees 716 selects a rogue memory-hogging task that frees up a large amount of
771 memory when killed. 717 memory when killed.
772 718
773 If this is set to non-zero, the OOM killer sim 719 If this is set to non-zero, the OOM killer simply kills the task that
774 triggered the out-of-memory condition. This a 720 triggered the out-of-memory condition. This avoids the expensive
775 tasklist scan. 721 tasklist scan.
776 722
777 If panic_on_oom is selected, it takes preceden 723 If panic_on_oom is selected, it takes precedence over whatever value
778 is used in oom_kill_allocating_task. 724 is used in oom_kill_allocating_task.
779 725
780 The default value is 0. 726 The default value is 0.
781 727
782 728
783 overcommit_kbytes 729 overcommit_kbytes
784 ================= 730 =================
785 731
786 When overcommit_memory is set to 2, the commit 732 When overcommit_memory is set to 2, the committed address space is not
787 permitted to exceed swap plus this amount of p 733 permitted to exceed swap plus this amount of physical RAM. See below.
788 734
789 Note: overcommit_kbytes is the counterpart of 735 Note: overcommit_kbytes is the counterpart of overcommit_ratio. Only one
790 of them may be specified at a time. Setting on 736 of them may be specified at a time. Setting one disables the other (which
791 then appears as 0 when read). 737 then appears as 0 when read).
792 738
793 739
794 overcommit_memory 740 overcommit_memory
795 ================= 741 =================
796 742
797 This value contains a flag that enables memory 743 This value contains a flag that enables memory overcommitment.
798 744
799 When this flag is 0, the kernel compares the u !! 745 When this flag is 0, the kernel attempts to estimate the amount
800 size against total memory plus swap and reject !! 746 of free memory left when userspace requests more memory.
801 747
802 When this flag is 1, the kernel pretends there 748 When this flag is 1, the kernel pretends there is always enough
803 memory until it actually runs out. 749 memory until it actually runs out.
804 750
805 When this flag is 2, the kernel uses a "never 751 When this flag is 2, the kernel uses a "never overcommit"
806 policy that attempts to prevent any overcommit 752 policy that attempts to prevent any overcommit of memory.
807 Note that user_reserve_kbytes affects this pol 753 Note that user_reserve_kbytes affects this policy.
808 754
809 This feature can be very useful because there 755 This feature can be very useful because there are a lot of
810 programs that malloc() huge amounts of memory 756 programs that malloc() huge amounts of memory "just-in-case"
811 and don't use much of it. 757 and don't use much of it.
812 758
813 The default value is 0. 759 The default value is 0.
814 760
815 See Documentation/mm/overcommit-accounting.rst 761 See Documentation/mm/overcommit-accounting.rst and
816 mm/util.c::__vm_enough_memory() for more infor 762 mm/util.c::__vm_enough_memory() for more information.
817 763
818 764
819 overcommit_ratio 765 overcommit_ratio
820 ================ 766 ================
821 767
822 When overcommit_memory is set to 2, the commit 768 When overcommit_memory is set to 2, the committed address
823 space is not permitted to exceed swap plus thi 769 space is not permitted to exceed swap plus this percentage
824 of physical RAM. See above. 770 of physical RAM. See above.
825 771
826 772
827 page-cluster 773 page-cluster
828 ============ 774 ============
829 775
830 page-cluster controls the number of pages up t 776 page-cluster controls the number of pages up to which consecutive pages
831 are read in from swap in a single attempt. Thi 777 are read in from swap in a single attempt. This is the swap counterpart
832 to page cache readahead. 778 to page cache readahead.
833 The mentioned consecutivity is not in terms of 779 The mentioned consecutivity is not in terms of virtual/physical addresses,
834 but consecutive on swap space - that means the 780 but consecutive on swap space - that means they were swapped out together.
835 781
836 It is a logarithmic value - setting it to zero 782 It is a logarithmic value - setting it to zero means "1 page", setting
837 it to 1 means "2 pages", setting it to 2 means 783 it to 1 means "2 pages", setting it to 2 means "4 pages", etc.
838 Zero disables swap readahead completely. 784 Zero disables swap readahead completely.
839 785
840 The default value is three (eight pages at a t 786 The default value is three (eight pages at a time). There may be some
841 small benefits in tuning this to a different v 787 small benefits in tuning this to a different value if your workload is
842 swap-intensive. 788 swap-intensive.
843 789
844 Lower values mean lower latencies for initial 790 Lower values mean lower latencies for initial faults, but at the same time
845 extra faults and I/O delays for following faul 791 extra faults and I/O delays for following faults if they would have been part of
846 that consecutive pages readahead would have br 792 that consecutive pages readahead would have brought in.
847 793
848 794
849 page_lock_unfairness 795 page_lock_unfairness
850 ==================== 796 ====================
851 797
852 This value determines the number of times that 798 This value determines the number of times that the page lock can be
853 stolen from under a waiter. After the lock is 799 stolen from under a waiter. After the lock is stolen the number of times
854 specified in this file (default is 5), the "fa 800 specified in this file (default is 5), the "fair lock handoff" semantics
855 will apply, and the waiter will only be awaken 801 will apply, and the waiter will only be awakened if the lock can be taken.
856 802
857 panic_on_oom 803 panic_on_oom
858 ============ 804 ============
859 805
860 This enables or disables panic on out-of-memor 806 This enables or disables panic on out-of-memory feature.
861 807
862 If this is set to 0, the kernel will kill some 808 If this is set to 0, the kernel will kill some rogue process,
863 called oom_killer. Usually, oom_killer can ki 809 called oom_killer. Usually, oom_killer can kill rogue processes and
864 system will survive. 810 system will survive.
865 811
866 If this is set to 1, the kernel panics when ou 812 If this is set to 1, the kernel panics when out-of-memory happens.
867 However, if a process limits using nodes by me 813 However, if a process limits using nodes by mempolicy/cpusets,
868 and those nodes become memory exhaustion statu 814 and those nodes become memory exhaustion status, one process
869 may be killed by oom-killer. No panic occurs i 815 may be killed by oom-killer. No panic occurs in this case.
870 Because other nodes' memory may be free. This 816 Because other nodes' memory may be free. This means system total status
871 may be not fatal yet. 817 may be not fatal yet.
872 818
873 If this is set to 2, the kernel panics compuls 819 If this is set to 2, the kernel panics compulsorily even on the
874 above-mentioned. Even oom happens under memory 820 above-mentioned. Even oom happens under memory cgroup, the whole
875 system panics. 821 system panics.
876 822
877 The default value is 0. 823 The default value is 0.
878 824
879 1 and 2 are for failover of clustering. Please 825 1 and 2 are for failover of clustering. Please select either
880 according to your policy of failover. 826 according to your policy of failover.
881 827
882 panic_on_oom=2+kdump gives you very strong too 828 panic_on_oom=2+kdump gives you very strong tool to investigate
883 why oom happens. You can get snapshot. 829 why oom happens. You can get snapshot.
884 830
885 831
886 percpu_pagelist_high_fraction 832 percpu_pagelist_high_fraction
887 ============================= 833 =============================
888 834
889 This is the fraction of pages in each zone tha 835 This is the fraction of pages in each zone that are can be stored to
890 per-cpu page lists. It is an upper boundary th 836 per-cpu page lists. It is an upper boundary that is divided depending
891 on the number of online CPUs. The min value fo 837 on the number of online CPUs. The min value for this is 8 which means
892 that we do not allow more than 1/8th of pages 838 that we do not allow more than 1/8th of pages in each zone to be stored
893 on per-cpu page lists. This entry only changes 839 on per-cpu page lists. This entry only changes the value of hot per-cpu
894 page lists. A user can specify a number like 1 840 page lists. A user can specify a number like 100 to allocate 1/100th of
895 each zone between per-cpu lists. 841 each zone between per-cpu lists.
896 842
897 The batch value of each per-cpu page list rema 843 The batch value of each per-cpu page list remains the same regardless of
898 the value of the high fraction so allocation l 844 the value of the high fraction so allocation latencies are unaffected.
899 845
900 The initial value is zero. Kernel uses this va 846 The initial value is zero. Kernel uses this value to set the high pcp->high
901 mark based on the low watermark for the zone a 847 mark based on the low watermark for the zone and the number of local
902 online CPUs. If the user writes '0' to this s 848 online CPUs. If the user writes '0' to this sysctl, it will revert to
903 this default behavior. 849 this default behavior.
904 850
905 851
906 stat_interval 852 stat_interval
907 ============= 853 =============
908 854
909 The time interval between which vm statistics 855 The time interval between which vm statistics are updated. The default
910 is 1 second. 856 is 1 second.
911 857
912 858
913 stat_refresh 859 stat_refresh
914 ============ 860 ============
915 861
916 Any read or write (by root only) flushes all t 862 Any read or write (by root only) flushes all the per-cpu vm statistics
917 into their global totals, for more accurate re 863 into their global totals, for more accurate reports when testing
918 e.g. cat /proc/sys/vm/stat_refresh /proc/memin 864 e.g. cat /proc/sys/vm/stat_refresh /proc/meminfo
919 865
920 As a side-effect, it also checks for negative 866 As a side-effect, it also checks for negative totals (elsewhere reported
921 as 0) and "fails" with EINVAL if any are found 867 as 0) and "fails" with EINVAL if any are found, with a warning in dmesg.
922 (At time of writing, a few stats are known som 868 (At time of writing, a few stats are known sometimes to be found negative,
923 with no ill effects: errors and warnings on th 869 with no ill effects: errors and warnings on these stats are suppressed.)
924 870
925 871
926 numa_stat 872 numa_stat
927 ========= 873 =========
928 874
929 This interface allows runtime configuration of 875 This interface allows runtime configuration of numa statistics.
930 876
931 When page allocation performance becomes a bot 877 When page allocation performance becomes a bottleneck and you can tolerate
932 some possible tool breakage and decreased numa 878 some possible tool breakage and decreased numa counter precision, you can
933 do:: 879 do::
934 880
935 echo 0 > /proc/sys/vm/numa_stat 881 echo 0 > /proc/sys/vm/numa_stat
936 882
937 When page allocation performance is not a bott 883 When page allocation performance is not a bottleneck and you want all
938 tooling to work, you can do:: 884 tooling to work, you can do::
939 885
940 echo 1 > /proc/sys/vm/numa_stat 886 echo 1 > /proc/sys/vm/numa_stat
941 887
942 888
943 swappiness 889 swappiness
944 ========== 890 ==========
945 891
946 This control is used to define the rough relat 892 This control is used to define the rough relative IO cost of swapping
947 and filesystem paging, as a value between 0 an 893 and filesystem paging, as a value between 0 and 200. At 100, the VM
948 assumes equal IO cost and will thus apply memo 894 assumes equal IO cost and will thus apply memory pressure to the page
949 cache and swap-backed pages equally; lower val 895 cache and swap-backed pages equally; lower values signify more
950 expensive swap IO, higher values indicates che 896 expensive swap IO, higher values indicates cheaper.
951 897
952 Keep in mind that filesystem IO patterns under 898 Keep in mind that filesystem IO patterns under memory pressure tend to
953 be more efficient than swap's random IO. An op 899 be more efficient than swap's random IO. An optimal value will require
954 experimentation and will also be workload-depe 900 experimentation and will also be workload-dependent.
955 901
956 The default value is 60. 902 The default value is 60.
957 903
958 For in-memory swap, like zram or zswap, as wel 904 For in-memory swap, like zram or zswap, as well as hybrid setups that
959 have swap on faster devices than the filesyste 905 have swap on faster devices than the filesystem, values beyond 100 can
960 be considered. For example, if the random IO a 906 be considered. For example, if the random IO against the swap device
961 is on average 2x faster than IO from the files 907 is on average 2x faster than IO from the filesystem, swappiness should
962 be 133 (x + 2x = 200, 2x = 133.33). 908 be 133 (x + 2x = 200, 2x = 133.33).
963 909
964 At 0, the kernel will not initiate swap until 910 At 0, the kernel will not initiate swap until the amount of free and
965 file-backed pages is less than the high waterm 911 file-backed pages is less than the high watermark in a zone.
966 912
967 913
968 unprivileged_userfaultfd 914 unprivileged_userfaultfd
969 ======================== 915 ========================
970 916
971 This flag controls the mode in which unprivile 917 This flag controls the mode in which unprivileged users can use the
972 userfaultfd system calls. Set this to 0 to res 918 userfaultfd system calls. Set this to 0 to restrict unprivileged users
973 to handle page faults in user mode only. In th 919 to handle page faults in user mode only. In this case, users without
974 SYS_CAP_PTRACE must pass UFFD_USER_MODE_ONLY i 920 SYS_CAP_PTRACE must pass UFFD_USER_MODE_ONLY in order for userfaultfd to
975 succeed. Prohibiting use of userfaultfd for ha 921 succeed. Prohibiting use of userfaultfd for handling faults from kernel
976 mode may make certain vulnerabilities more dif 922 mode may make certain vulnerabilities more difficult to exploit.
977 923
978 Set this to 1 to allow unprivileged users to u 924 Set this to 1 to allow unprivileged users to use the userfaultfd system
979 calls without any restrictions. 925 calls without any restrictions.
980 926
981 The default value is 0. 927 The default value is 0.
982 928
983 Another way to control permissions for userfau <<
984 /dev/userfaultfd instead of userfaultfd(2). Se <<
985 Documentation/admin-guide/mm/userfaultfd.rst. <<
986 929
987 user_reserve_kbytes 930 user_reserve_kbytes
988 =================== 931 ===================
989 932
990 When overcommit_memory is set to 2, "never ove 933 When overcommit_memory is set to 2, "never overcommit" mode, reserve
991 min(3% of current process size, user_reserve_k 934 min(3% of current process size, user_reserve_kbytes) of free memory.
992 This is intended to prevent a user from starti 935 This is intended to prevent a user from starting a single memory hogging
993 process, such that they cannot recover (kill t 936 process, such that they cannot recover (kill the hog).
994 937
995 user_reserve_kbytes defaults to min(3% of the 938 user_reserve_kbytes defaults to min(3% of the current process size, 128MB).
996 939
997 If this is reduced to zero, then the user will 940 If this is reduced to zero, then the user will be allowed to allocate
998 all free memory with a single process, minus a 941 all free memory with a single process, minus admin_reserve_kbytes.
999 Any subsequent attempts to execute a command w 942 Any subsequent attempts to execute a command will result in
1000 "fork: Cannot allocate memory". 943 "fork: Cannot allocate memory".
1001 944
1002 Changing this takes effect whenever an applic 945 Changing this takes effect whenever an application requests memory.
1003 946
1004 947
1005 vfs_cache_pressure 948 vfs_cache_pressure
1006 ================== 949 ==================
1007 950
1008 This percentage value controls the tendency o 951 This percentage value controls the tendency of the kernel to reclaim
1009 the memory which is used for caching of direc 952 the memory which is used for caching of directory and inode objects.
1010 953
1011 At the default value of vfs_cache_pressure=10 954 At the default value of vfs_cache_pressure=100 the kernel will attempt to
1012 reclaim dentries and inodes at a "fair" rate 955 reclaim dentries and inodes at a "fair" rate with respect to pagecache and
1013 swapcache reclaim. Decreasing vfs_cache_pres 956 swapcache reclaim. Decreasing vfs_cache_pressure causes the kernel to prefer
1014 to retain dentry and inode caches. When vfs_c 957 to retain dentry and inode caches. When vfs_cache_pressure=0, the kernel will
1015 never reclaim dentries and inodes due to memo 958 never reclaim dentries and inodes due to memory pressure and this can easily
1016 lead to out-of-memory conditions. Increasing 959 lead to out-of-memory conditions. Increasing vfs_cache_pressure beyond 100
1017 causes the kernel to prefer to reclaim dentri 960 causes the kernel to prefer to reclaim dentries and inodes.
1018 961
1019 Increasing vfs_cache_pressure significantly b 962 Increasing vfs_cache_pressure significantly beyond 100 may have negative
1020 performance impact. Reclaim code needs to tak 963 performance impact. Reclaim code needs to take various locks to find freeable
1021 directory and inode objects. With vfs_cache_p 964 directory and inode objects. With vfs_cache_pressure=1000, it will look for
1022 ten times more freeable objects than there ar 965 ten times more freeable objects than there are.
1023 966
1024 967
1025 watermark_boost_factor 968 watermark_boost_factor
1026 ====================== 969 ======================
1027 970
1028 This factor controls the level of reclaim whe 971 This factor controls the level of reclaim when memory is being fragmented.
1029 It defines the percentage of the high waterma 972 It defines the percentage of the high watermark of a zone that will be
1030 reclaimed if pages of different mobility are 973 reclaimed if pages of different mobility are being mixed within pageblocks.
1031 The intent is that compaction has less work t 974 The intent is that compaction has less work to do in the future and to
1032 increase the success rate of future high-orde 975 increase the success rate of future high-order allocations such as SLUB
1033 allocations, THP and hugetlbfs pages. 976 allocations, THP and hugetlbfs pages.
1034 977
1035 To make it sensible with respect to the water 978 To make it sensible with respect to the watermark_scale_factor
1036 parameter, the unit is in fractions of 10,000 979 parameter, the unit is in fractions of 10,000. The default value of
1037 15,000 means that up to 150% of the high wate 980 15,000 means that up to 150% of the high watermark will be reclaimed in the
1038 event of a pageblock being mixed due to fragm 981 event of a pageblock being mixed due to fragmentation. The level of reclaim
1039 is determined by the number of fragmentation 982 is determined by the number of fragmentation events that occurred in the
1040 recent past. If this value is smaller than a 983 recent past. If this value is smaller than a pageblock then a pageblocks
1041 worth of pages will be reclaimed (e.g. 2MB o 984 worth of pages will be reclaimed (e.g. 2MB on 64-bit x86). A boost factor
1042 of 0 will disable the feature. 985 of 0 will disable the feature.
1043 986
1044 987
1045 watermark_scale_factor 988 watermark_scale_factor
1046 ====================== 989 ======================
1047 990
1048 This factor controls the aggressiveness of ks 991 This factor controls the aggressiveness of kswapd. It defines the
1049 amount of memory left in a node/system before 992 amount of memory left in a node/system before kswapd is woken up and
1050 how much memory needs to be free before kswap 993 how much memory needs to be free before kswapd goes back to sleep.
1051 994
1052 The unit is in fractions of 10,000. The defau 995 The unit is in fractions of 10,000. The default value of 10 means the
1053 distances between watermarks are 0.1% of the 996 distances between watermarks are 0.1% of the available memory in the
1054 node/system. The maximum value is 3000, or 30 997 node/system. The maximum value is 3000, or 30% of memory.
1055 998
1056 A high rate of threads entering direct reclai 999 A high rate of threads entering direct reclaim (allocstall) or kswapd
1057 going to sleep prematurely (kswapd_low_wmark_ 1000 going to sleep prematurely (kswapd_low_wmark_hit_quickly) can indicate
1058 that the number of free pages kswapd maintain 1001 that the number of free pages kswapd maintains for latency reasons is
1059 too small for the allocation bursts occurring 1002 too small for the allocation bursts occurring in the system. This knob
1060 can then be used to tune kswapd aggressivenes 1003 can then be used to tune kswapd aggressiveness accordingly.
1061 1004
1062 1005
1063 zone_reclaim_mode 1006 zone_reclaim_mode
1064 ================= 1007 =================
1065 1008
1066 Zone_reclaim_mode allows someone to set more 1009 Zone_reclaim_mode allows someone to set more or less aggressive approaches to
1067 reclaim memory when a zone runs out of memory 1010 reclaim memory when a zone runs out of memory. If it is set to zero then no
1068 zone reclaim occurs. Allocations will be sati 1011 zone reclaim occurs. Allocations will be satisfied from other zones / nodes
1069 in the system. 1012 in the system.
1070 1013
1071 This is value OR'ed together of 1014 This is value OR'ed together of
1072 1015
1073 = =================================== 1016 = ===================================
1074 1 Zone reclaim on 1017 1 Zone reclaim on
1075 2 Zone reclaim writes dirty pages out 1018 2 Zone reclaim writes dirty pages out
1076 4 Zone reclaim swaps pages 1019 4 Zone reclaim swaps pages
1077 = =================================== 1020 = ===================================
1078 1021
1079 zone_reclaim_mode is disabled by default. Fo 1022 zone_reclaim_mode is disabled by default. For file servers or workloads
1080 that benefit from having their data cached, z 1023 that benefit from having their data cached, zone_reclaim_mode should be
1081 left disabled as the caching effect is likely 1024 left disabled as the caching effect is likely to be more important than
1082 data locality. 1025 data locality.
1083 1026
1084 Consider enabling one or more zone_reclaim mo 1027 Consider enabling one or more zone_reclaim mode bits if it's known that the
1085 workload is partitioned such that each partit 1028 workload is partitioned such that each partition fits within a NUMA node
1086 and that accessing remote memory would cause 1029 and that accessing remote memory would cause a measurable performance
1087 reduction. The page allocator will take addi 1030 reduction. The page allocator will take additional actions before
1088 allocating off node pages. 1031 allocating off node pages.
1089 1032
1090 Allowing zone reclaim to write out pages stop 1033 Allowing zone reclaim to write out pages stops processes that are
1091 writing large amounts of data from dirtying p 1034 writing large amounts of data from dirtying pages on other nodes. Zone
1092 reclaim will write out dirty pages if a zone 1035 reclaim will write out dirty pages if a zone fills up and so effectively
1093 throttle the process. This may decrease the p 1036 throttle the process. This may decrease the performance of a single process
1094 since it cannot use all of system memory to b 1037 since it cannot use all of system memory to buffer the outgoing writes
1095 anymore but it preserve the memory on other n 1038 anymore but it preserve the memory on other nodes so that the performance
1096 of other processes running on other nodes wil 1039 of other processes running on other nodes will not be affected.
1097 1040
1098 Allowing regular swap effectively restricts a 1041 Allowing regular swap effectively restricts allocations to the local
1099 node unless explicitly overridden by memory p 1042 node unless explicitly overridden by memory policies or cpuset
1100 configurations. 1043 configurations.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.