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