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
146 # eventually, we can have this option just 'select SPARSEMEM'
147 config MEMORY_HOTPLUG
148 bool "Allow for memory hot-add"
149 select MEMORY_ISOLATION
150 depends on SPARSEMEM || X86_64_ACPI_NUMA
151 depends on HOTPLUG && ARCH_ENABLE_MEMORY_HOTPLUG
152 depends on (IA64 || X86 || PPC_BOOK3S_64 || SUPERH || S390)
154 config MEMORY_HOTPLUG_SPARSE
156 depends on SPARSEMEM && MEMORY_HOTPLUG
158 config MEMORY_HOTREMOVE
159 bool "Allow for memory hot remove"
160 depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE
164 # If we have space for more page flags then we can enable additional
165 # optimizations and functionality.
167 # Regular Sparsemem takes page flag bits for the sectionid if it does not
168 # use a virtual memmap. Disable extended page flags for 32 bit platforms
169 # that require the use of a sectionid in the page flags.
171 config PAGEFLAGS_EXTENDED
173 depends on 64BIT || SPARSEMEM_VMEMMAP || !SPARSEMEM
175 # Heavily threaded applications may benefit from splitting the mm-wide
176 # page_table_lock, so that faults on different parts of the user address
177 # space can be handled with less contention: split it at this NR_CPUS.
178 # Default to 4 for wider testing, though 8 might be more appropriate.
179 # ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
180 # PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
181 # DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page.
183 config SPLIT_PTLOCK_CPUS
185 default "999999" if ARM && !CPU_CACHE_VIPT
186 default "999999" if PARISC && !PA20
187 default "999999" if DEBUG_SPINLOCK || DEBUG_LOCK_ALLOC
191 # support for memory compaction
193 bool "Allow for memory compaction"
197 Allows the compaction of memory for the allocation of huge pages.
200 # support for page migration
203 bool "Page migration"
205 depends on NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA
207 Allows the migration of the physical location of pages of processes
208 while the virtual addresses are not changed. This is useful in
209 two situations. The first is on NUMA systems to put pages nearer
210 to the processors accessing. The second is when allocating huge
211 pages as migration can relocate pages to satisfy a huge page
212 allocation instead of reclaiming.
214 config PHYS_ADDR_T_64BIT
215 def_bool 64BIT || ARCH_PHYS_ADDR_T_64BIT
219 default "0" if !ZONE_DMA
224 depends on BLOCK && MMU && (ZONE_DMA || HIGHMEM)
234 depends on !ARCH_NO_VIRT_TO_BUS
240 bool "Enable KSM for page merging"
243 Enable Kernel Samepage Merging: KSM periodically scans those areas
244 of an application's address space that an app has advised may be
245 mergeable. When it finds pages of identical content, it replaces
246 the many instances by a single page with that content, so
247 saving memory until one or another app needs to modify the content.
248 Recommended for use with KVM, or with other duplicative applications.
249 See Documentation/vm/ksm.txt for more information: KSM is inactive
250 until a program has madvised that an area is MADV_MERGEABLE, and
251 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
253 config DEFAULT_MMAP_MIN_ADDR
254 int "Low address space to protect from user allocation"
258 This is the portion of low virtual memory which should be protected
259 from userspace allocation. Keeping a user from writing to low pages
260 can help reduce the impact of kernel NULL pointer bugs.
262 For most ia64, ppc64 and x86 users with lots of address space
263 a value of 65536 is reasonable and should cause no problems.
264 On arm and other archs it should not be higher than 32768.
265 Programs which use vm86 functionality or have some need to map
266 this low address space will need CAP_SYS_RAWIO or disable this
267 protection by setting the value to 0.
269 This value can be changed after boot using the
270 /proc/sys/vm/mmap_min_addr tunable.
272 config ARCH_SUPPORTS_MEMORY_FAILURE
275 config MEMORY_FAILURE
277 depends on ARCH_SUPPORTS_MEMORY_FAILURE
278 bool "Enable recovery from hardware memory errors"
279 select MEMORY_ISOLATION
281 Enables code to recover from some memory failures on systems
282 with MCA recovery. This allows a system to continue running
283 even when some of its memory has uncorrected errors. This requires
284 special hardware support and typically ECC memory.
286 config HWPOISON_INJECT
287 tristate "HWPoison pages injector"
288 depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS
289 select PROC_PAGE_MONITOR
291 config NOMMU_INITIAL_TRIM_EXCESS
292 int "Turn on mmap() excess space trimming before booting"
296 The NOMMU mmap() frequently needs to allocate large contiguous chunks
297 of memory on which to store mappings, but it can only ask the system
298 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently
299 more than it requires. To deal with this, mmap() is able to trim off
300 the excess and return it to the allocator.
302 If trimming is enabled, the excess is trimmed off and returned to the
303 system allocator, which can cause extra fragmentation, particularly
304 if there are a lot of transient processes.
306 If trimming is disabled, the excess is kept, but not used, which for
307 long-term mappings means that the space is wasted.
309 Trimming can be dynamically controlled through a sysctl option
310 (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of
311 excess pages there must be before trimming should occur, or zero if
312 no trimming is to occur.
314 This option specifies the initial value of this option. The default
315 of 1 says that all excess pages should be trimmed.
317 See Documentation/nommu-mmap.txt for more information.
319 config TRANSPARENT_HUGEPAGE
320 bool "Transparent Hugepage Support"
321 depends on X86 && MMU
324 Transparent Hugepages allows the kernel to use huge pages and
325 huge tlb transparently to the applications whenever possible.
326 This feature can improve computing performance to certain
327 applications by speeding up page faults during memory
328 allocation, by reducing the number of tlb misses and by speeding
329 up the pagetable walking.
331 If memory constrained on embedded, you may want to say N.
334 prompt "Transparent Hugepage Support sysfs defaults"
335 depends on TRANSPARENT_HUGEPAGE
336 default TRANSPARENT_HUGEPAGE_ALWAYS
338 Selects the sysfs defaults for Transparent Hugepage Support.
340 config TRANSPARENT_HUGEPAGE_ALWAYS
343 Enabling Transparent Hugepage always, can increase the
344 memory footprint of applications without a guaranteed
345 benefit but it will work automatically for all applications.
347 config TRANSPARENT_HUGEPAGE_MADVISE
350 Enabling Transparent Hugepage madvise, will only provide a
351 performance improvement benefit to the applications using
352 madvise(MADV_HUGEPAGE) but it won't risk to increase the
353 memory footprint of applications without a guaranteed
357 config CROSS_MEMORY_ATTACH
358 bool "Cross Memory Support"
362 Enabling this option adds the system calls process_vm_readv and
363 process_vm_writev which allow a process with the correct privileges
364 to directly read from or write to to another process's address space.
365 See the man page for more details.
368 # UP and nommu archs use km based percpu allocator
370 config NEED_PER_CPU_KM
376 bool "Enable cleancache driver to cache clean pages if tmem is present"
379 Cleancache can be thought of as a page-granularity victim cache
380 for clean pages that the kernel's pageframe replacement algorithm
381 (PFRA) would like to keep around, but can't since there isn't enough
382 memory. So when the PFRA "evicts" a page, it first attempts to use
383 cleancache code to put the data contained in that page into
384 "transcendent memory", memory that is not directly accessible or
385 addressable by the kernel and is of unknown and possibly
386 time-varying size. And when a cleancache-enabled
387 filesystem wishes to access a page in a file on disk, it first
388 checks cleancache to see if it already contains it; if it does,
389 the page is copied into the kernel and a disk access is avoided.
390 When a transcendent memory driver is available (such as zcache or
391 Xen transcendent memory), a significant I/O reduction
392 may be achieved. When none is available, all cleancache calls
393 are reduced to a single pointer-compare-against-NULL resulting
394 in a negligible performance hit.
396 If unsure, say Y to enable cleancache
399 bool "Enable frontswap to cache swap pages if tmem is present"
403 Frontswap is so named because it can be thought of as the opposite
404 of a "backing" store for a swap device. The data is stored into
405 "transcendent memory", memory that is not directly accessible or
406 addressable by the kernel and is of unknown and possibly
407 time-varying size. When space in transcendent memory is available,
408 a significant swap I/O reduction may be achieved. When none is
409 available, all frontswap calls are reduced to a single pointer-
410 compare-against-NULL resulting in a negligible performance hit
411 and swap data is stored as normal on the matching swap device.
413 If unsure, say Y to enable frontswap.