1 config SELECT_MEMORY_MODEL
3 depends on EXPERIMENTAL || ARCH_SELECT_MEMORY_MODEL
7 depends on SELECT_MEMORY_MODEL
8 default DISCONTIGMEM_MANUAL if ARCH_DISCONTIGMEM_DEFAULT
9 default SPARSEMEM_MANUAL if ARCH_SPARSEMEM_DEFAULT
10 default FLATMEM_MANUAL
14 depends on !(ARCH_DISCONTIGMEM_ENABLE || ARCH_SPARSEMEM_ENABLE) || ARCH_FLATMEM_ENABLE
16 This option allows you to change some of the ways that
17 Linux manages its memory internally. Most users will
18 only have one option here: FLATMEM. This is normal
21 Some users of more advanced features like NUMA and
22 memory hotplug may have different options here.
23 DISCONTIGMEM is an more mature, better tested system,
24 but is incompatible with memory hotplug and may suffer
25 decreased performance over SPARSEMEM. If unsure between
26 "Sparse Memory" and "Discontiguous Memory", choose
27 "Discontiguous Memory".
29 If unsure, choose this option (Flat Memory) over any other.
31 config DISCONTIGMEM_MANUAL
32 bool "Discontiguous Memory"
33 depends on ARCH_DISCONTIGMEM_ENABLE
35 This option provides enhanced support for discontiguous
36 memory systems, over FLATMEM. These systems have holes
37 in their physical address spaces, and this option provides
38 more efficient handling of these holes. However, the vast
39 majority of hardware has quite flat address spaces, and
40 can have degraded performance from the extra overhead that
43 Many NUMA configurations will have this as the only option.
45 If unsure, choose "Flat Memory" over this option.
47 config SPARSEMEM_MANUAL
49 depends on ARCH_SPARSEMEM_ENABLE
51 This will be the only option for some systems, including
52 memory hotplug systems. This is normal.
54 For many other systems, this will be an alternative to
55 "Discontiguous Memory". This option provides some potential
56 performance benefits, along with decreased code complexity,
57 but it is newer, and more experimental.
59 If unsure, choose "Discontiguous Memory" or "Flat Memory"
66 depends on (!SELECT_MEMORY_MODEL && ARCH_DISCONTIGMEM_ENABLE) || DISCONTIGMEM_MANUAL
70 depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL
74 depends on (!DISCONTIGMEM && !SPARSEMEM) || FLATMEM_MANUAL
76 config FLAT_NODE_MEM_MAP
81 # Both the NUMA code and DISCONTIGMEM use arrays of pg_data_t's
82 # to represent different areas of memory. This variable allows
83 # those dependencies to exist individually.
85 config NEED_MULTIPLE_NODES
87 depends on DISCONTIGMEM || NUMA
89 config HAVE_MEMORY_PRESENT
91 depends on ARCH_HAVE_MEMORY_PRESENT || SPARSEMEM
94 # SPARSEMEM_EXTREME (which is the default) does some bootmem
95 # allocations when memory_present() is called. If this cannot
96 # be done on your architecture, select this option. However,
97 # statically allocating the mem_section[] array can potentially
98 # consume vast quantities of .bss, so be careful.
100 # This option will also potentially produce smaller runtime code
101 # with gcc 3.4 and later.
103 config SPARSEMEM_STATIC
107 # Architecture platforms which require a two level mem_section in SPARSEMEM
108 # must select this option. This is usually for architecture platforms with
109 # an extremely sparse physical address space.
111 config SPARSEMEM_EXTREME
113 depends on SPARSEMEM && !SPARSEMEM_STATIC
115 config SPARSEMEM_VMEMMAP_ENABLE
118 config SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
120 depends on SPARSEMEM && X86_64
122 config SPARSEMEM_VMEMMAP
123 bool "Sparse Memory virtual memmap"
124 depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE
127 SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise
128 pfn_to_page and page_to_pfn operations. This is the most
129 efficient option when sufficient kernel resources are available.
134 config HAVE_MEMBLOCK_NODE_MAP
137 config ARCH_DISCARD_MEMBLOCK
143 config MEMORY_ISOLATION
147 boolean "Enable to assign a node which has only movable memory"
148 depends on HAVE_MEMBLOCK
149 depends on NO_BOOTMEM
154 # eventually, we can have this option just 'select SPARSEMEM'
155 config MEMORY_HOTPLUG
156 bool "Allow for memory hot-add"
157 select MEMORY_ISOLATION
158 depends on SPARSEMEM || X86_64_ACPI_NUMA
159 depends on HOTPLUG && ARCH_ENABLE_MEMORY_HOTPLUG
160 depends on (IA64 || X86 || PPC_BOOK3S_64 || SUPERH || S390)
162 config MEMORY_HOTPLUG_SPARSE
164 depends on SPARSEMEM && MEMORY_HOTPLUG
166 config MEMORY_HOTREMOVE
167 bool "Allow for memory hot remove"
168 depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE
172 # If we have space for more page flags then we can enable additional
173 # optimizations and functionality.
175 # Regular Sparsemem takes page flag bits for the sectionid if it does not
176 # use a virtual memmap. Disable extended page flags for 32 bit platforms
177 # that require the use of a sectionid in the page flags.
179 config PAGEFLAGS_EXTENDED
181 depends on 64BIT || SPARSEMEM_VMEMMAP || !SPARSEMEM
183 # Heavily threaded applications may benefit from splitting the mm-wide
184 # page_table_lock, so that faults on different parts of the user address
185 # space can be handled with less contention: split it at this NR_CPUS.
186 # Default to 4 for wider testing, though 8 might be more appropriate.
187 # ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
188 # PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
189 # DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page.
191 config SPLIT_PTLOCK_CPUS
193 default "999999" if ARM && !CPU_CACHE_VIPT
194 default "999999" if PARISC && !PA20
195 default "999999" if DEBUG_SPINLOCK || DEBUG_LOCK_ALLOC
199 # support for memory balloon compaction
200 config BALLOON_COMPACTION
201 bool "Allow for balloon memory compaction/migration"
203 depends on COMPACTION && VIRTIO_BALLOON
205 Memory fragmentation introduced by ballooning might reduce
206 significantly the number of 2MB contiguous memory blocks that can be
207 used within a guest, thus imposing performance penalties associated
208 with the reduced number of transparent huge pages that could be used
209 by the guest workload. Allowing the compaction & migration for memory
210 pages enlisted as being part of memory balloon devices avoids the
211 scenario aforementioned and helps improving memory defragmentation.
214 # support for memory compaction
216 bool "Allow for memory compaction"
221 Allows the compaction of memory for the allocation of huge pages.
224 # support for page migration
227 bool "Page migration"
229 depends on NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA
231 Allows the migration of the physical location of pages of processes
232 while the virtual addresses are not changed. This is useful in
233 two situations. The first is on NUMA systems to put pages nearer
234 to the processors accessing. The second is when allocating huge
235 pages as migration can relocate pages to satisfy a huge page
236 allocation instead of reclaiming.
238 config PHYS_ADDR_T_64BIT
239 def_bool 64BIT || ARCH_PHYS_ADDR_T_64BIT
243 default "0" if !ZONE_DMA
248 depends on BLOCK && MMU && (ZONE_DMA || HIGHMEM)
258 depends on !ARCH_NO_VIRT_TO_BUS
264 bool "Enable KSM for page merging"
267 Enable Kernel Samepage Merging: KSM periodically scans those areas
268 of an application's address space that an app has advised may be
269 mergeable. When it finds pages of identical content, it replaces
270 the many instances by a single page with that content, so
271 saving memory until one or another app needs to modify the content.
272 Recommended for use with KVM, or with other duplicative applications.
273 See Documentation/vm/ksm.txt for more information: KSM is inactive
274 until a program has madvised that an area is MADV_MERGEABLE, and
275 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
277 config DEFAULT_MMAP_MIN_ADDR
278 int "Low address space to protect from user allocation"
282 This is the portion of low virtual memory which should be protected
283 from userspace allocation. Keeping a user from writing to low pages
284 can help reduce the impact of kernel NULL pointer bugs.
286 For most ia64, ppc64 and x86 users with lots of address space
287 a value of 65536 is reasonable and should cause no problems.
288 On arm and other archs it should not be higher than 32768.
289 Programs which use vm86 functionality or have some need to map
290 this low address space will need CAP_SYS_RAWIO or disable this
291 protection by setting the value to 0.
293 This value can be changed after boot using the
294 /proc/sys/vm/mmap_min_addr tunable.
296 config ARCH_SUPPORTS_MEMORY_FAILURE
299 config MEMORY_FAILURE
301 depends on ARCH_SUPPORTS_MEMORY_FAILURE
302 bool "Enable recovery from hardware memory errors"
303 select MEMORY_ISOLATION
305 Enables code to recover from some memory failures on systems
306 with MCA recovery. This allows a system to continue running
307 even when some of its memory has uncorrected errors. This requires
308 special hardware support and typically ECC memory.
310 config HWPOISON_INJECT
311 tristate "HWPoison pages injector"
312 depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS
313 select PROC_PAGE_MONITOR
315 config NOMMU_INITIAL_TRIM_EXCESS
316 int "Turn on mmap() excess space trimming before booting"
320 The NOMMU mmap() frequently needs to allocate large contiguous chunks
321 of memory on which to store mappings, but it can only ask the system
322 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently
323 more than it requires. To deal with this, mmap() is able to trim off
324 the excess and return it to the allocator.
326 If trimming is enabled, the excess is trimmed off and returned to the
327 system allocator, which can cause extra fragmentation, particularly
328 if there are a lot of transient processes.
330 If trimming is disabled, the excess is kept, but not used, which for
331 long-term mappings means that the space is wasted.
333 Trimming can be dynamically controlled through a sysctl option
334 (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of
335 excess pages there must be before trimming should occur, or zero if
336 no trimming is to occur.
338 This option specifies the initial value of this option. The default
339 of 1 says that all excess pages should be trimmed.
341 See Documentation/nommu-mmap.txt for more information.
343 config TRANSPARENT_HUGEPAGE
344 bool "Transparent Hugepage Support"
345 depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE
348 Transparent Hugepages allows the kernel to use huge pages and
349 huge tlb transparently to the applications whenever possible.
350 This feature can improve computing performance to certain
351 applications by speeding up page faults during memory
352 allocation, by reducing the number of tlb misses and by speeding
353 up the pagetable walking.
355 If memory constrained on embedded, you may want to say N.
358 prompt "Transparent Hugepage Support sysfs defaults"
359 depends on TRANSPARENT_HUGEPAGE
360 default TRANSPARENT_HUGEPAGE_ALWAYS
362 Selects the sysfs defaults for Transparent Hugepage Support.
364 config TRANSPARENT_HUGEPAGE_ALWAYS
367 Enabling Transparent Hugepage always, can increase the
368 memory footprint of applications without a guaranteed
369 benefit but it will work automatically for all applications.
371 config TRANSPARENT_HUGEPAGE_MADVISE
374 Enabling Transparent Hugepage madvise, will only provide a
375 performance improvement benefit to the applications using
376 madvise(MADV_HUGEPAGE) but it won't risk to increase the
377 memory footprint of applications without a guaranteed
381 config CROSS_MEMORY_ATTACH
382 bool "Cross Memory Support"
386 Enabling this option adds the system calls process_vm_readv and
387 process_vm_writev which allow a process with the correct privileges
388 to directly read from or write to to another process's address space.
389 See the man page for more details.
392 # UP and nommu archs use km based percpu allocator
394 config NEED_PER_CPU_KM
400 bool "Enable cleancache driver to cache clean pages if tmem is present"
403 Cleancache can be thought of as a page-granularity victim cache
404 for clean pages that the kernel's pageframe replacement algorithm
405 (PFRA) would like to keep around, but can't since there isn't enough
406 memory. So when the PFRA "evicts" a page, it first attempts to use
407 cleancache code to put the data contained in that page into
408 "transcendent memory", memory that is not directly accessible or
409 addressable by the kernel and is of unknown and possibly
410 time-varying size. And when a cleancache-enabled
411 filesystem wishes to access a page in a file on disk, it first
412 checks cleancache to see if it already contains it; if it does,
413 the page is copied into the kernel and a disk access is avoided.
414 When a transcendent memory driver is available (such as zcache or
415 Xen transcendent memory), a significant I/O reduction
416 may be achieved. When none is available, all cleancache calls
417 are reduced to a single pointer-compare-against-NULL resulting
418 in a negligible performance hit.
420 If unsure, say Y to enable cleancache
423 bool "Enable frontswap to cache swap pages if tmem is present"
427 Frontswap is so named because it can be thought of as the opposite
428 of a "backing" store for a swap device. The data is stored into
429 "transcendent memory", memory that is not directly accessible or
430 addressable by the kernel and is of unknown and possibly
431 time-varying size. When space in transcendent memory is available,
432 a significant swap I/O reduction may be achieved. When none is
433 available, all frontswap calls are reduced to a single pointer-
434 compare-against-NULL resulting in a negligible performance hit
435 and swap data is stored as normal on the matching swap device.
437 If unsure, say Y to enable frontswap.