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