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 # eventually, we can have this option just 'select SPARSEMEM'
135 config MEMORY_HOTPLUG
136 bool "Allow for memory hot-add"
137 depends on SPARSEMEM || X86_64_ACPI_NUMA
138 depends on HOTPLUG && ARCH_ENABLE_MEMORY_HOTPLUG
139 depends on (IA64 || X86 || PPC_BOOK3S_64 || SUPERH || S390)
141 config MEMORY_HOTPLUG_SPARSE
143 depends on SPARSEMEM && MEMORY_HOTPLUG
145 config MEMORY_HOTREMOVE
146 bool "Allow for memory hot remove"
147 depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE
151 # If we have space for more page flags then we can enable additional
152 # optimizations and functionality.
154 # Regular Sparsemem takes page flag bits for the sectionid if it does not
155 # use a virtual memmap. Disable extended page flags for 32 bit platforms
156 # that require the use of a sectionid in the page flags.
158 config PAGEFLAGS_EXTENDED
160 depends on 64BIT || SPARSEMEM_VMEMMAP || !SPARSEMEM
162 # Heavily threaded applications may benefit from splitting the mm-wide
163 # page_table_lock, so that faults on different parts of the user address
164 # space can be handled with less contention: split it at this NR_CPUS.
165 # Default to 4 for wider testing, though 8 might be more appropriate.
166 # ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
167 # PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
168 # DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page.
170 config SPLIT_PTLOCK_CPUS
172 default "999999" if ARM && !CPU_CACHE_VIPT
173 default "999999" if PARISC && !PA20
174 default "999999" if DEBUG_SPINLOCK || DEBUG_LOCK_ALLOC
178 # support for memory compaction
180 bool "Allow for memory compaction"
184 Allows the compaction of memory for the allocation of huge pages.
187 # support for page migration
190 bool "Page migration"
192 depends on NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION
194 Allows the migration of the physical location of pages of processes
195 while the virtual addresses are not changed. This is useful in
196 two situations. The first is on NUMA systems to put pages nearer
197 to the processors accessing. The second is when allocating huge
198 pages as migration can relocate pages to satisfy a huge page
199 allocation instead of reclaiming.
201 config PHYS_ADDR_T_64BIT
202 def_bool 64BIT || ARCH_PHYS_ADDR_T_64BIT
206 default "0" if !ZONE_DMA
211 depends on BLOCK && MMU && (ZONE_DMA || HIGHMEM)
221 depends on !ARCH_NO_VIRT_TO_BUS
227 bool "Enable KSM for page merging"
230 Enable Kernel Samepage Merging: KSM periodically scans those areas
231 of an application's address space that an app has advised may be
232 mergeable. When it finds pages of identical content, it replaces
233 the many instances by a single page with that content, so
234 saving memory until one or another app needs to modify the content.
235 Recommended for use with KVM, or with other duplicative applications.
236 See Documentation/vm/ksm.txt for more information: KSM is inactive
237 until a program has madvised that an area is MADV_MERGEABLE, and
238 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
240 config DEFAULT_MMAP_MIN_ADDR
241 int "Low address space to protect from user allocation"
245 This is the portion of low virtual memory which should be protected
246 from userspace allocation. Keeping a user from writing to low pages
247 can help reduce the impact of kernel NULL pointer bugs.
249 For most ia64, ppc64 and x86 users with lots of address space
250 a value of 65536 is reasonable and should cause no problems.
251 On arm and other archs it should not be higher than 32768.
252 Programs which use vm86 functionality or have some need to map
253 this low address space will need CAP_SYS_RAWIO or disable this
254 protection by setting the value to 0.
256 This value can be changed after boot using the
257 /proc/sys/vm/mmap_min_addr tunable.
259 config ARCH_SUPPORTS_MEMORY_FAILURE
262 config MEMORY_FAILURE
264 depends on ARCH_SUPPORTS_MEMORY_FAILURE
265 bool "Enable recovery from hardware memory errors"
267 Enables code to recover from some memory failures on systems
268 with MCA recovery. This allows a system to continue running
269 even when some of its memory has uncorrected errors. This requires
270 special hardware support and typically ECC memory.
272 config HWPOISON_INJECT
273 tristate "HWPoison pages injector"
274 depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS
275 select PROC_PAGE_MONITOR
277 config NOMMU_INITIAL_TRIM_EXCESS
278 int "Turn on mmap() excess space trimming before booting"
282 The NOMMU mmap() frequently needs to allocate large contiguous chunks
283 of memory on which to store mappings, but it can only ask the system
284 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently
285 more than it requires. To deal with this, mmap() is able to trim off
286 the excess and return it to the allocator.
288 If trimming is enabled, the excess is trimmed off and returned to the
289 system allocator, which can cause extra fragmentation, particularly
290 if there are a lot of transient processes.
292 If trimming is disabled, the excess is kept, but not used, which for
293 long-term mappings means that the space is wasted.
295 Trimming can be dynamically controlled through a sysctl option
296 (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of
297 excess pages there must be before trimming should occur, or zero if
298 no trimming is to occur.
300 This option specifies the initial value of this option. The default
301 of 1 says that all excess pages should be trimmed.
303 See Documentation/nommu-mmap.txt for more information.
305 config TRANSPARENT_HUGEPAGE
306 bool "Transparent Hugepage Support"
307 depends on X86 && MMU
310 Transparent Hugepages allows the kernel to use huge pages and
311 huge tlb transparently to the applications whenever possible.
312 This feature can improve computing performance to certain
313 applications by speeding up page faults during memory
314 allocation, by reducing the number of tlb misses and by speeding
315 up the pagetable walking.
317 If memory constrained on embedded, you may want to say N.
320 prompt "Transparent Hugepage Support sysfs defaults"
321 depends on TRANSPARENT_HUGEPAGE
322 default TRANSPARENT_HUGEPAGE_ALWAYS
324 Selects the sysfs defaults for Transparent Hugepage Support.
326 config TRANSPARENT_HUGEPAGE_ALWAYS
329 Enabling Transparent Hugepage always, can increase the
330 memory footprint of applications without a guaranteed
331 benefit but it will work automatically for all applications.
333 config TRANSPARENT_HUGEPAGE_MADVISE
336 Enabling Transparent Hugepage madvise, will only provide a
337 performance improvement benefit to the applications using
338 madvise(MADV_HUGEPAGE) but it won't risk to increase the
339 memory footprint of applications without a guaranteed
344 # UP and nommu archs use km based percpu allocator
346 config NEED_PER_CPU_KM
352 bool "Enable cleancache driver to cache clean pages if tmem is present"
355 Cleancache can be thought of as a page-granularity victim cache
356 for clean pages that the kernel's pageframe replacement algorithm
357 (PFRA) would like to keep around, but can't since there isn't enough
358 memory. So when the PFRA "evicts" a page, it first attempts to use
359 cleancache code to put the data contained in that page into
360 "transcendent memory", memory that is not directly accessible or
361 addressable by the kernel and is of unknown and possibly
362 time-varying size. And when a cleancache-enabled
363 filesystem wishes to access a page in a file on disk, it first
364 checks cleancache to see if it already contains it; if it does,
365 the page is copied into the kernel and a disk access is avoided.
366 When a transcendent memory driver is available (such as zcache or
367 Xen transcendent memory), a significant I/O reduction
368 may be achieved. When none is available, all cleancache calls
369 are reduced to a single pointer-compare-against-NULL resulting
370 in a negligible performance hit.
372 If unsure, say Y to enable cleancache
375 bool "Enable frontswap to cache swap pages if tmem is present"
379 Frontswap is so named because it can be thought of as the opposite
380 of a "backing" store for a swap device. The data is stored into
381 "transcendent memory", memory that is not directly accessible or
382 addressable by the kernel and is of unknown and possibly
383 time-varying size. When space in transcendent memory is available,
384 a significant swap I/O reduction may be achieved. When none is
385 available, all frontswap calls are reduced to a single pointer-
386 compare-against-NULL resulting in a negligible performance hit
387 and swap data is stored as normal on the matching swap device.
389 If unsure, say Y to enable frontswap.