1ec8f24b7SThomas Gleixner# SPDX-License-Identifier: GPL-2.0-only 259e0b520SChristoph Hellwig 359e0b520SChristoph Hellwigmenu "Memory Management options" 459e0b520SChristoph Hellwig 5e1785e85SDave Hansenconfig SELECT_MEMORY_MODEL 6e1785e85SDave Hansen def_bool y 7a8826eebSKees Cook depends on ARCH_SELECT_MEMORY_MODEL 8e1785e85SDave Hansen 93a9da765SDave Hansenchoice 103a9da765SDave Hansen prompt "Memory model" 11e1785e85SDave Hansen depends on SELECT_MEMORY_MODEL 12e1785e85SDave Hansen default DISCONTIGMEM_MANUAL if ARCH_DISCONTIGMEM_DEFAULT 13d41dee36SAndy Whitcroft default SPARSEMEM_MANUAL if ARCH_SPARSEMEM_DEFAULT 14e1785e85SDave Hansen default FLATMEM_MANUAL 15d66d109dSMike Rapoport help 16d66d109dSMike Rapoport This option allows you to change some of the ways that 17d66d109dSMike Rapoport Linux manages its memory internally. Most users will 18d66d109dSMike Rapoport only have one option here selected by the architecture 19d66d109dSMike Rapoport configuration. This is normal. 203a9da765SDave Hansen 21e1785e85SDave Hansenconfig FLATMEM_MANUAL 223a9da765SDave Hansen bool "Flat Memory" 23c898ec16SAnton Blanchard depends on !(ARCH_DISCONTIGMEM_ENABLE || ARCH_SPARSEMEM_ENABLE) || ARCH_FLATMEM_ENABLE 243a9da765SDave Hansen help 25d66d109dSMike Rapoport This option is best suited for non-NUMA systems with 26d66d109dSMike Rapoport flat address space. The FLATMEM is the most efficient 27d66d109dSMike Rapoport system in terms of performance and resource consumption 28d66d109dSMike Rapoport and it is the best option for smaller systems. 293a9da765SDave Hansen 30d66d109dSMike Rapoport For systems that have holes in their physical address 31d66d109dSMike Rapoport spaces and for features like NUMA and memory hotplug, 32dd33d29aSRandy Dunlap choose "Sparse Memory". 33d41dee36SAndy Whitcroft 34d41dee36SAndy Whitcroft If unsure, choose this option (Flat Memory) over any other. 353a9da765SDave Hansen 36e1785e85SDave Hansenconfig DISCONTIGMEM_MANUAL 37f3519f91SDave Hansen bool "Discontiguous Memory" 383a9da765SDave Hansen depends on ARCH_DISCONTIGMEM_ENABLE 393a9da765SDave Hansen help 40785dcd44SDave Hansen This option provides enhanced support for discontiguous 41785dcd44SDave Hansen memory systems, over FLATMEM. These systems have holes 42785dcd44SDave Hansen in their physical address spaces, and this option provides 43d66d109dSMike Rapoport more efficient handling of these holes. 44785dcd44SDave Hansen 45d66d109dSMike Rapoport Although "Discontiguous Memory" is still used by several 46d66d109dSMike Rapoport architectures, it is considered deprecated in favor of 47d66d109dSMike Rapoport "Sparse Memory". 48785dcd44SDave Hansen 49d66d109dSMike Rapoport If unsure, choose "Sparse Memory" over this option. 503a9da765SDave Hansen 51d41dee36SAndy Whitcroftconfig SPARSEMEM_MANUAL 52d41dee36SAndy Whitcroft bool "Sparse Memory" 53d41dee36SAndy Whitcroft depends on ARCH_SPARSEMEM_ENABLE 54d41dee36SAndy Whitcroft help 55d41dee36SAndy Whitcroft This will be the only option for some systems, including 56d66d109dSMike Rapoport memory hot-plug systems. This is normal. 57d41dee36SAndy Whitcroft 58d66d109dSMike Rapoport This option provides efficient support for systems with 59d66d109dSMike Rapoport holes is their physical address space and allows memory 60d66d109dSMike Rapoport hot-plug and hot-remove. 61d41dee36SAndy Whitcroft 62d66d109dSMike Rapoport If unsure, choose "Flat Memory" over this option. 63d41dee36SAndy Whitcroft 643a9da765SDave Hansenendchoice 653a9da765SDave Hansen 66e1785e85SDave Hansenconfig DISCONTIGMEM 67e1785e85SDave Hansen def_bool y 68e1785e85SDave Hansen depends on (!SELECT_MEMORY_MODEL && ARCH_DISCONTIGMEM_ENABLE) || DISCONTIGMEM_MANUAL 69e1785e85SDave Hansen 70d41dee36SAndy Whitcroftconfig SPARSEMEM 71d41dee36SAndy Whitcroft def_bool y 721a83e175SRussell King depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL 73d41dee36SAndy Whitcroft 74e1785e85SDave Hansenconfig FLATMEM 75e1785e85SDave Hansen def_bool y 76d41dee36SAndy Whitcroft depends on (!DISCONTIGMEM && !SPARSEMEM) || FLATMEM_MANUAL 77d41dee36SAndy Whitcroft 78d41dee36SAndy Whitcroftconfig FLAT_NODE_MEM_MAP 79d41dee36SAndy Whitcroft def_bool y 80d41dee36SAndy Whitcroft depends on !SPARSEMEM 81e1785e85SDave Hansen 8293b7504eSDave Hansen# 8393b7504eSDave Hansen# Both the NUMA code and DISCONTIGMEM use arrays of pg_data_t's 8493b7504eSDave Hansen# to represent different areas of memory. This variable allows 8593b7504eSDave Hansen# those dependencies to exist individually. 8693b7504eSDave Hansen# 8793b7504eSDave Hansenconfig NEED_MULTIPLE_NODES 8893b7504eSDave Hansen def_bool y 8993b7504eSDave Hansen depends on DISCONTIGMEM || NUMA 90af705362SAndy Whitcroft 91af705362SAndy Whitcroftconfig HAVE_MEMORY_PRESENT 92af705362SAndy Whitcroft def_bool y 93d41dee36SAndy Whitcroft depends on ARCH_HAVE_MEMORY_PRESENT || SPARSEMEM 94802f192eSBob Picco 95802f192eSBob Picco# 963e347261SBob Picco# SPARSEMEM_EXTREME (which is the default) does some bootmem 973e347261SBob Picco# allocations when memory_present() is called. If this cannot 983e347261SBob Picco# be done on your architecture, select this option. However, 993e347261SBob Picco# statically allocating the mem_section[] array can potentially 1003e347261SBob Picco# consume vast quantities of .bss, so be careful. 1013e347261SBob Picco# 1023e347261SBob Picco# This option will also potentially produce smaller runtime code 1033e347261SBob Picco# with gcc 3.4 and later. 1043e347261SBob Picco# 1053e347261SBob Piccoconfig SPARSEMEM_STATIC 1069ba16087SJan Beulich bool 1073e347261SBob Picco 1083e347261SBob Picco# 10944c09201SMatt LaPlante# Architecture platforms which require a two level mem_section in SPARSEMEM 110802f192eSBob Picco# must select this option. This is usually for architecture platforms with 111802f192eSBob Picco# an extremely sparse physical address space. 112802f192eSBob Picco# 1133e347261SBob Piccoconfig SPARSEMEM_EXTREME 1143e347261SBob Picco def_bool y 1153e347261SBob Picco depends on SPARSEMEM && !SPARSEMEM_STATIC 1164c21e2f2SHugh Dickins 11729c71111SAndy Whitcroftconfig SPARSEMEM_VMEMMAP_ENABLE 1189ba16087SJan Beulich bool 11929c71111SAndy Whitcroft 12029c71111SAndy Whitcroftconfig SPARSEMEM_VMEMMAP 121a5ee6daaSGeoff Levand bool "Sparse Memory virtual memmap" 122a5ee6daaSGeoff Levand depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE 123a5ee6daaSGeoff Levand default y 124a5ee6daaSGeoff Levand help 125a5ee6daaSGeoff Levand SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise 126a5ee6daaSGeoff Levand pfn_to_page and page_to_pfn operations. This is the most 127a5ee6daaSGeoff Levand efficient option when sufficient kernel resources are available. 12829c71111SAndy Whitcroft 1297c0caeb8STejun Heoconfig HAVE_MEMBLOCK_NODE_MAP 1306341e62bSChristoph Jaeger bool 1317c0caeb8STejun Heo 13270210ed9SPhilipp Hachtmannconfig HAVE_MEMBLOCK_PHYS_MAP 1336341e62bSChristoph Jaeger bool 13470210ed9SPhilipp Hachtmann 13567a929e0SChristoph Hellwigconfig HAVE_FAST_GUP 136050a9adcSChristoph Hellwig depends on MMU 1376341e62bSChristoph Jaeger bool 1382667f50eSSteve Capper 139350e88baSMike Rapoportconfig ARCH_KEEP_MEMBLOCK 1406341e62bSChristoph Jaeger bool 141c378ddd5STejun Heo 142*1e5d8e1eSDan Williams# Keep arch NUMA mapping infrastructure post-init. 143*1e5d8e1eSDan Williamsconfig NUMA_KEEP_MEMINFO 144*1e5d8e1eSDan Williams bool 145*1e5d8e1eSDan Williams 146ee6f509cSMinchan Kimconfig MEMORY_ISOLATION 1476341e62bSChristoph Jaeger bool 148ee6f509cSMinchan Kim 14946723bfaSYasuaki Ishimatsu# 15046723bfaSYasuaki Ishimatsu# Only be set on architectures that have completely implemented memory hotplug 15146723bfaSYasuaki Ishimatsu# feature. If you are not sure, don't touch it. 15246723bfaSYasuaki Ishimatsu# 15346723bfaSYasuaki Ishimatsuconfig HAVE_BOOTMEM_INFO_NODE 15446723bfaSYasuaki Ishimatsu def_bool n 15546723bfaSYasuaki Ishimatsu 1563947be19SDave Hansen# eventually, we can have this option just 'select SPARSEMEM' 1573947be19SDave Hansenconfig MEMORY_HOTPLUG 1583947be19SDave Hansen bool "Allow for memory hot-add" 159ec69acbbSKeith Mannthey depends on SPARSEMEM || X86_64_ACPI_NUMA 16040b31360SStephen Rothwell depends on ARCH_ENABLE_MEMORY_HOTPLUG 161*1e5d8e1eSDan Williams select NUMA_KEEP_MEMINFO if NUMA 1623947be19SDave Hansen 163ec69acbbSKeith Manntheyconfig MEMORY_HOTPLUG_SPARSE 164ec69acbbSKeith Mannthey def_bool y 165ec69acbbSKeith Mannthey depends on SPARSEMEM && MEMORY_HOTPLUG 166ec69acbbSKeith Mannthey 1678604d9e5SVitaly Kuznetsovconfig MEMORY_HOTPLUG_DEFAULT_ONLINE 1688604d9e5SVitaly Kuznetsov bool "Online the newly added memory blocks by default" 1698604d9e5SVitaly Kuznetsov depends on MEMORY_HOTPLUG 1708604d9e5SVitaly Kuznetsov help 1718604d9e5SVitaly Kuznetsov This option sets the default policy setting for memory hotplug 1728604d9e5SVitaly Kuznetsov onlining policy (/sys/devices/system/memory/auto_online_blocks) which 1738604d9e5SVitaly Kuznetsov determines what happens to newly added memory regions. Policy setting 1748604d9e5SVitaly Kuznetsov can always be changed at runtime. 175cb1aaebeSMauro Carvalho Chehab See Documentation/admin-guide/mm/memory-hotplug.rst for more information. 1768604d9e5SVitaly Kuznetsov 1778604d9e5SVitaly Kuznetsov Say Y here if you want all hot-plugged memory blocks to appear in 1788604d9e5SVitaly Kuznetsov 'online' state by default. 1798604d9e5SVitaly Kuznetsov Say N here if you want the default policy to keep all hot-plugged 1808604d9e5SVitaly Kuznetsov memory blocks in 'offline' state. 1818604d9e5SVitaly Kuznetsov 1820c0e6195SKAMEZAWA Hiroyukiconfig MEMORY_HOTREMOVE 1830c0e6195SKAMEZAWA Hiroyuki bool "Allow for memory hot remove" 18446723bfaSYasuaki Ishimatsu select MEMORY_ISOLATION 185f7e3334aSNathan Fontenot select HAVE_BOOTMEM_INFO_NODE if (X86_64 || PPC64) 1860c0e6195SKAMEZAWA Hiroyuki depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE 1870c0e6195SKAMEZAWA Hiroyuki depends on MIGRATION 1880c0e6195SKAMEZAWA Hiroyuki 1894c21e2f2SHugh Dickins# Heavily threaded applications may benefit from splitting the mm-wide 1904c21e2f2SHugh Dickins# page_table_lock, so that faults on different parts of the user address 1914c21e2f2SHugh Dickins# space can be handled with less contention: split it at this NR_CPUS. 1924c21e2f2SHugh Dickins# Default to 4 for wider testing, though 8 might be more appropriate. 1934c21e2f2SHugh Dickins# ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock. 1947b6ac9dfSHugh Dickins# PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes. 195a70caa8bSHugh Dickins# DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page. 1964c21e2f2SHugh Dickins# 1974c21e2f2SHugh Dickinsconfig SPLIT_PTLOCK_CPUS 1984c21e2f2SHugh Dickins int 1999164550eSKirill A. Shutemov default "999999" if !MMU 200a70caa8bSHugh Dickins default "999999" if ARM && !CPU_CACHE_VIPT 201a70caa8bSHugh Dickins default "999999" if PARISC && !PA20 2024c21e2f2SHugh Dickins default "4" 2037cbe34cfSChristoph Lameter 204e009bb30SKirill A. Shutemovconfig ARCH_ENABLE_SPLIT_PMD_PTLOCK 2056341e62bSChristoph Jaeger bool 206e009bb30SKirill A. Shutemov 2077cbe34cfSChristoph Lameter# 20809316c09SKonstantin Khlebnikov# support for memory balloon 20909316c09SKonstantin Khlebnikovconfig MEMORY_BALLOON 2106341e62bSChristoph Jaeger bool 21109316c09SKonstantin Khlebnikov 21209316c09SKonstantin Khlebnikov# 21318468d93SRafael Aquini# support for memory balloon compaction 21418468d93SRafael Aquiniconfig BALLOON_COMPACTION 21518468d93SRafael Aquini bool "Allow for balloon memory compaction/migration" 21618468d93SRafael Aquini def_bool y 21709316c09SKonstantin Khlebnikov depends on COMPACTION && MEMORY_BALLOON 21818468d93SRafael Aquini help 21918468d93SRafael Aquini Memory fragmentation introduced by ballooning might reduce 22018468d93SRafael Aquini significantly the number of 2MB contiguous memory blocks that can be 22118468d93SRafael Aquini used within a guest, thus imposing performance penalties associated 22218468d93SRafael Aquini with the reduced number of transparent huge pages that could be used 22318468d93SRafael Aquini by the guest workload. Allowing the compaction & migration for memory 22418468d93SRafael Aquini pages enlisted as being part of memory balloon devices avoids the 22518468d93SRafael Aquini scenario aforementioned and helps improving memory defragmentation. 22618468d93SRafael Aquini 22718468d93SRafael Aquini# 228e9e96b39SMel Gorman# support for memory compaction 229e9e96b39SMel Gormanconfig COMPACTION 230e9e96b39SMel Gorman bool "Allow for memory compaction" 23105106e6aSRik van Riel def_bool y 232e9e96b39SMel Gorman select MIGRATION 23333a93877SAndrea Arcangeli depends on MMU 234e9e96b39SMel Gorman help 235b32eaf71SMichal Hocko Compaction is the only memory management component to form 236b32eaf71SMichal Hocko high order (larger physically contiguous) memory blocks 237b32eaf71SMichal Hocko reliably. The page allocator relies on compaction heavily and 238b32eaf71SMichal Hocko the lack of the feature can lead to unexpected OOM killer 239b32eaf71SMichal Hocko invocations for high order memory requests. You shouldn't 240b32eaf71SMichal Hocko disable this option unless there really is a strong reason for 241b32eaf71SMichal Hocko it and then we would be really interested to hear about that at 242b32eaf71SMichal Hocko linux-mm@kvack.org. 243e9e96b39SMel Gorman 244e9e96b39SMel Gorman# 2457cbe34cfSChristoph Lameter# support for page migration 2467cbe34cfSChristoph Lameter# 2477cbe34cfSChristoph Lameterconfig MIGRATION 248b20a3503SChristoph Lameter bool "Page migration" 2496c5240aeSChristoph Lameter def_bool y 250de32a817SChen Gang depends on (NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA) && MMU 251b20a3503SChristoph Lameter help 252b20a3503SChristoph Lameter Allows the migration of the physical location of pages of processes 253e9e96b39SMel Gorman while the virtual addresses are not changed. This is useful in 254e9e96b39SMel Gorman two situations. The first is on NUMA systems to put pages nearer 255e9e96b39SMel Gorman to the processors accessing. The second is when allocating huge 256e9e96b39SMel Gorman pages as migration can relocate pages to satisfy a huge page 257e9e96b39SMel Gorman allocation instead of reclaiming. 2586550e07fSGreg Kroah-Hartman 259c177c81eSNaoya Horiguchiconfig ARCH_ENABLE_HUGEPAGE_MIGRATION 2606341e62bSChristoph Jaeger bool 261c177c81eSNaoya Horiguchi 2629c670ea3SNaoya Horiguchiconfig ARCH_ENABLE_THP_MIGRATION 2639c670ea3SNaoya Horiguchi bool 2649c670ea3SNaoya Horiguchi 2658df995f6SAlexandre Ghiticonfig CONTIG_ALLOC 2668df995f6SAlexandre Ghiti def_bool (MEMORY_ISOLATION && COMPACTION) || CMA 2678df995f6SAlexandre Ghiti 268600715dcSJeremy Fitzhardingeconfig PHYS_ADDR_T_64BIT 269d4a451d5SChristoph Hellwig def_bool 64BIT 270600715dcSJeremy Fitzhardinge 2712a7326b5SChristoph Lameterconfig BOUNCE 2729ca24e2eSVinayak Menon bool "Enable bounce buffers" 2739ca24e2eSVinayak Menon default y 2742a7326b5SChristoph Lameter depends on BLOCK && MMU && (ZONE_DMA || HIGHMEM) 2759ca24e2eSVinayak Menon help 2769ca24e2eSVinayak Menon Enable bounce buffers for devices that cannot access 2779ca24e2eSVinayak Menon the full range of memory available to the CPU. Enabled 2789ca24e2eSVinayak Menon by default when ZONE_DMA or HIGHMEM is selected, but you 2799ca24e2eSVinayak Menon may say n to override this. 2802a7326b5SChristoph Lameter 281f057eac0SStephen Rothwellconfig VIRT_TO_BUS 2824febd95aSStephen Rothwell bool 2834febd95aSStephen Rothwell help 2844febd95aSStephen Rothwell An architecture should select this if it implements the 2854febd95aSStephen Rothwell deprecated interface virt_to_bus(). All new architectures 2864febd95aSStephen Rothwell should probably not select this. 2874febd95aSStephen Rothwell 288cddb8a5cSAndrea Arcangeli 289cddb8a5cSAndrea Arcangeliconfig MMU_NOTIFIER 290cddb8a5cSAndrea Arcangeli bool 29183fe27eaSPranith Kumar select SRCU 29299cb252fSJason Gunthorpe select INTERVAL_TREE 293fc4d5c29SDavid Howells 294f8af4da3SHugh Dickinsconfig KSM 295f8af4da3SHugh Dickins bool "Enable KSM for page merging" 296f8af4da3SHugh Dickins depends on MMU 29759e1a2f4STimofey Titovets select XXHASH 298f8af4da3SHugh Dickins help 299f8af4da3SHugh Dickins Enable Kernel Samepage Merging: KSM periodically scans those areas 300f8af4da3SHugh Dickins of an application's address space that an app has advised may be 301f8af4da3SHugh Dickins mergeable. When it finds pages of identical content, it replaces 302d0f209f6SHugh Dickins the many instances by a single page with that content, so 303f8af4da3SHugh Dickins saving memory until one or another app needs to modify the content. 304f8af4da3SHugh Dickins Recommended for use with KVM, or with other duplicative applications. 305ad56b738SMike Rapoport See Documentation/vm/ksm.rst for more information: KSM is inactive 306c73602adSHugh Dickins until a program has madvised that an area is MADV_MERGEABLE, and 307c73602adSHugh Dickins root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set). 308f8af4da3SHugh Dickins 309e0a94c2aSChristoph Lameterconfig DEFAULT_MMAP_MIN_ADDR 310e0a94c2aSChristoph Lameter int "Low address space to protect from user allocation" 3116e141546SDavid Howells depends on MMU 312e0a94c2aSChristoph Lameter default 4096 313e0a94c2aSChristoph Lameter help 314e0a94c2aSChristoph Lameter This is the portion of low virtual memory which should be protected 315e0a94c2aSChristoph Lameter from userspace allocation. Keeping a user from writing to low pages 316e0a94c2aSChristoph Lameter can help reduce the impact of kernel NULL pointer bugs. 317e0a94c2aSChristoph Lameter 318e0a94c2aSChristoph Lameter For most ia64, ppc64 and x86 users with lots of address space 319e0a94c2aSChristoph Lameter a value of 65536 is reasonable and should cause no problems. 320e0a94c2aSChristoph Lameter On arm and other archs it should not be higher than 32768. 321788084abSEric Paris Programs which use vm86 functionality or have some need to map 322788084abSEric Paris this low address space will need CAP_SYS_RAWIO or disable this 323788084abSEric Paris protection by setting the value to 0. 324e0a94c2aSChristoph Lameter 325e0a94c2aSChristoph Lameter This value can be changed after boot using the 326e0a94c2aSChristoph Lameter /proc/sys/vm/mmap_min_addr tunable. 327e0a94c2aSChristoph Lameter 328d949f36fSLinus Torvaldsconfig ARCH_SUPPORTS_MEMORY_FAILURE 329d949f36fSLinus Torvalds bool 330e0a94c2aSChristoph Lameter 3316a46079cSAndi Kleenconfig MEMORY_FAILURE 3326a46079cSAndi Kleen depends on MMU 333d949f36fSLinus Torvalds depends on ARCH_SUPPORTS_MEMORY_FAILURE 3346a46079cSAndi Kleen bool "Enable recovery from hardware memory errors" 335ee6f509cSMinchan Kim select MEMORY_ISOLATION 33697f0b134SXie XiuQi select RAS 3376a46079cSAndi Kleen help 3386a46079cSAndi Kleen Enables code to recover from some memory failures on systems 3396a46079cSAndi Kleen with MCA recovery. This allows a system to continue running 3406a46079cSAndi Kleen even when some of its memory has uncorrected errors. This requires 3416a46079cSAndi Kleen special hardware support and typically ECC memory. 3426a46079cSAndi Kleen 343cae681fcSAndi Kleenconfig HWPOISON_INJECT 344413f9efbSAndi Kleen tristate "HWPoison pages injector" 34527df5068SAndi Kleen depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS 346478c5ffcSWu Fengguang select PROC_PAGE_MONITOR 347cae681fcSAndi Kleen 348fc4d5c29SDavid Howellsconfig NOMMU_INITIAL_TRIM_EXCESS 349fc4d5c29SDavid Howells int "Turn on mmap() excess space trimming before booting" 350fc4d5c29SDavid Howells depends on !MMU 351fc4d5c29SDavid Howells default 1 352fc4d5c29SDavid Howells help 353fc4d5c29SDavid Howells The NOMMU mmap() frequently needs to allocate large contiguous chunks 354fc4d5c29SDavid Howells of memory on which to store mappings, but it can only ask the system 355fc4d5c29SDavid Howells allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently 356fc4d5c29SDavid Howells more than it requires. To deal with this, mmap() is able to trim off 357fc4d5c29SDavid Howells the excess and return it to the allocator. 358fc4d5c29SDavid Howells 359fc4d5c29SDavid Howells If trimming is enabled, the excess is trimmed off and returned to the 360fc4d5c29SDavid Howells system allocator, which can cause extra fragmentation, particularly 361fc4d5c29SDavid Howells if there are a lot of transient processes. 362fc4d5c29SDavid Howells 363fc4d5c29SDavid Howells If trimming is disabled, the excess is kept, but not used, which for 364fc4d5c29SDavid Howells long-term mappings means that the space is wasted. 365fc4d5c29SDavid Howells 366fc4d5c29SDavid Howells Trimming can be dynamically controlled through a sysctl option 367fc4d5c29SDavid Howells (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of 368fc4d5c29SDavid Howells excess pages there must be before trimming should occur, or zero if 369fc4d5c29SDavid Howells no trimming is to occur. 370fc4d5c29SDavid Howells 371fc4d5c29SDavid Howells This option specifies the initial value of this option. The default 372fc4d5c29SDavid Howells of 1 says that all excess pages should be trimmed. 373fc4d5c29SDavid Howells 374fc4d5c29SDavid Howells See Documentation/nommu-mmap.txt for more information. 375bbddff05STejun Heo 3764c76d9d1SAndrea Arcangeliconfig TRANSPARENT_HUGEPAGE 37713ece886SAndrea Arcangeli bool "Transparent Hugepage Support" 37815626062SGerald Schaefer depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE 3795d689240SAndrea Arcangeli select COMPACTION 3803a08cd52SMatthew Wilcox select XARRAY_MULTI 3814c76d9d1SAndrea Arcangeli help 3824c76d9d1SAndrea Arcangeli Transparent Hugepages allows the kernel to use huge pages and 3834c76d9d1SAndrea Arcangeli huge tlb transparently to the applications whenever possible. 3844c76d9d1SAndrea Arcangeli This feature can improve computing performance to certain 3854c76d9d1SAndrea Arcangeli applications by speeding up page faults during memory 3864c76d9d1SAndrea Arcangeli allocation, by reducing the number of tlb misses and by speeding 3874c76d9d1SAndrea Arcangeli up the pagetable walking. 3884c76d9d1SAndrea Arcangeli 3894c76d9d1SAndrea Arcangeli If memory constrained on embedded, you may want to say N. 3904c76d9d1SAndrea Arcangeli 39113ece886SAndrea Arcangelichoice 39213ece886SAndrea Arcangeli prompt "Transparent Hugepage Support sysfs defaults" 39313ece886SAndrea Arcangeli depends on TRANSPARENT_HUGEPAGE 39413ece886SAndrea Arcangeli default TRANSPARENT_HUGEPAGE_ALWAYS 39513ece886SAndrea Arcangeli help 39613ece886SAndrea Arcangeli Selects the sysfs defaults for Transparent Hugepage Support. 39713ece886SAndrea Arcangeli 39813ece886SAndrea Arcangeli config TRANSPARENT_HUGEPAGE_ALWAYS 39913ece886SAndrea Arcangeli bool "always" 40013ece886SAndrea Arcangeli help 40113ece886SAndrea Arcangeli Enabling Transparent Hugepage always, can increase the 40213ece886SAndrea Arcangeli memory footprint of applications without a guaranteed 40313ece886SAndrea Arcangeli benefit but it will work automatically for all applications. 40413ece886SAndrea Arcangeli 40513ece886SAndrea Arcangeli config TRANSPARENT_HUGEPAGE_MADVISE 40613ece886SAndrea Arcangeli bool "madvise" 40713ece886SAndrea Arcangeli help 40813ece886SAndrea Arcangeli Enabling Transparent Hugepage madvise, will only provide a 40913ece886SAndrea Arcangeli performance improvement benefit to the applications using 41013ece886SAndrea Arcangeli madvise(MADV_HUGEPAGE) but it won't risk to increase the 41113ece886SAndrea Arcangeli memory footprint of applications without a guaranteed 41213ece886SAndrea Arcangeli benefit. 41313ece886SAndrea Arcangeliendchoice 41413ece886SAndrea Arcangeli 41538d8b4e6SHuang Yingconfig ARCH_WANTS_THP_SWAP 41638d8b4e6SHuang Ying def_bool n 41738d8b4e6SHuang Ying 41838d8b4e6SHuang Yingconfig THP_SWAP 41938d8b4e6SHuang Ying def_bool y 42014fef284SHuang Ying depends on TRANSPARENT_HUGEPAGE && ARCH_WANTS_THP_SWAP && SWAP 42138d8b4e6SHuang Ying help 42238d8b4e6SHuang Ying Swap transparent huge pages in one piece, without splitting. 42314fef284SHuang Ying XXX: For now, swap cluster backing transparent huge page 42414fef284SHuang Ying will be split after swapout. 42538d8b4e6SHuang Ying 42638d8b4e6SHuang Ying For selection by architectures with reasonable THP sizes. 42738d8b4e6SHuang Ying 428e496cf3dSKirill A. Shutemovconfig TRANSPARENT_HUGE_PAGECACHE 429e496cf3dSKirill A. Shutemov def_bool y 430953c66c2SAneesh Kumar K.V depends on TRANSPARENT_HUGEPAGE 431e496cf3dSKirill A. Shutemov 432e496cf3dSKirill A. Shutemov# 433bbddff05STejun Heo# UP and nommu archs use km based percpu allocator 434bbddff05STejun Heo# 435bbddff05STejun Heoconfig NEED_PER_CPU_KM 436bbddff05STejun Heo depends on !SMP 437bbddff05STejun Heo bool 438bbddff05STejun Heo default y 439077b1f83SDan Magenheimer 440077b1f83SDan Magenheimerconfig CLEANCACHE 441077b1f83SDan Magenheimer bool "Enable cleancache driver to cache clean pages if tmem is present" 442077b1f83SDan Magenheimer help 443077b1f83SDan Magenheimer Cleancache can be thought of as a page-granularity victim cache 444077b1f83SDan Magenheimer for clean pages that the kernel's pageframe replacement algorithm 445077b1f83SDan Magenheimer (PFRA) would like to keep around, but can't since there isn't enough 446077b1f83SDan Magenheimer memory. So when the PFRA "evicts" a page, it first attempts to use 447140a1ef2SMichael Witten cleancache code to put the data contained in that page into 448077b1f83SDan Magenheimer "transcendent memory", memory that is not directly accessible or 449077b1f83SDan Magenheimer addressable by the kernel and is of unknown and possibly 450077b1f83SDan Magenheimer time-varying size. And when a cleancache-enabled 451077b1f83SDan Magenheimer filesystem wishes to access a page in a file on disk, it first 452077b1f83SDan Magenheimer checks cleancache to see if it already contains it; if it does, 453077b1f83SDan Magenheimer the page is copied into the kernel and a disk access is avoided. 454077b1f83SDan Magenheimer When a transcendent memory driver is available (such as zcache or 455077b1f83SDan Magenheimer Xen transcendent memory), a significant I/O reduction 456077b1f83SDan Magenheimer may be achieved. When none is available, all cleancache calls 457077b1f83SDan Magenheimer are reduced to a single pointer-compare-against-NULL resulting 458077b1f83SDan Magenheimer in a negligible performance hit. 459077b1f83SDan Magenheimer 460077b1f83SDan Magenheimer If unsure, say Y to enable cleancache 46127c6aec2SDan Magenheimer 46227c6aec2SDan Magenheimerconfig FRONTSWAP 46327c6aec2SDan Magenheimer bool "Enable frontswap to cache swap pages if tmem is present" 46427c6aec2SDan Magenheimer depends on SWAP 46527c6aec2SDan Magenheimer help 46627c6aec2SDan Magenheimer Frontswap is so named because it can be thought of as the opposite 46727c6aec2SDan Magenheimer of a "backing" store for a swap device. The data is stored into 46827c6aec2SDan Magenheimer "transcendent memory", memory that is not directly accessible or 46927c6aec2SDan Magenheimer addressable by the kernel and is of unknown and possibly 47027c6aec2SDan Magenheimer time-varying size. When space in transcendent memory is available, 47127c6aec2SDan Magenheimer a significant swap I/O reduction may be achieved. When none is 47227c6aec2SDan Magenheimer available, all frontswap calls are reduced to a single pointer- 47327c6aec2SDan Magenheimer compare-against-NULL resulting in a negligible performance hit 47427c6aec2SDan Magenheimer and swap data is stored as normal on the matching swap device. 47527c6aec2SDan Magenheimer 47627c6aec2SDan Magenheimer If unsure, say Y to enable frontswap. 477f825c736SAneesh Kumar K.V 478f825c736SAneesh Kumar K.Vconfig CMA 479f825c736SAneesh Kumar K.V bool "Contiguous Memory Allocator" 480aca52c39SMike Rapoport depends on MMU 481f825c736SAneesh Kumar K.V select MIGRATION 482f825c736SAneesh Kumar K.V select MEMORY_ISOLATION 483f825c736SAneesh Kumar K.V help 484f825c736SAneesh Kumar K.V This enables the Contiguous Memory Allocator which allows other 485f825c736SAneesh Kumar K.V subsystems to allocate big physically-contiguous blocks of memory. 486f825c736SAneesh Kumar K.V CMA reserves a region of memory and allows only movable pages to 487f825c736SAneesh Kumar K.V be allocated from it. This way, the kernel can use the memory for 488f825c736SAneesh Kumar K.V pagecache and when a subsystem requests for contiguous area, the 489f825c736SAneesh Kumar K.V allocated pages are migrated away to serve the contiguous request. 490f825c736SAneesh Kumar K.V 491f825c736SAneesh Kumar K.V If unsure, say "n". 492f825c736SAneesh Kumar K.V 493f825c736SAneesh Kumar K.Vconfig CMA_DEBUG 494f825c736SAneesh Kumar K.V bool "CMA debug messages (DEVELOPMENT)" 495f825c736SAneesh Kumar K.V depends on DEBUG_KERNEL && CMA 496f825c736SAneesh Kumar K.V help 497f825c736SAneesh Kumar K.V Turns on debug messages in CMA. This produces KERN_DEBUG 498f825c736SAneesh Kumar K.V messages for every CMA call as well as various messages while 499f825c736SAneesh Kumar K.V processing calls such as dma_alloc_from_contiguous(). 500f825c736SAneesh Kumar K.V This option does not affect warning and error messages. 501bf550fc9SAlexander Graf 50228b24c1fSSasha Levinconfig CMA_DEBUGFS 50328b24c1fSSasha Levin bool "CMA debugfs interface" 50428b24c1fSSasha Levin depends on CMA && DEBUG_FS 50528b24c1fSSasha Levin help 50628b24c1fSSasha Levin Turns on the DebugFS interface for CMA. 50728b24c1fSSasha Levin 508a254129eSJoonsoo Kimconfig CMA_AREAS 509a254129eSJoonsoo Kim int "Maximum count of the CMA areas" 510a254129eSJoonsoo Kim depends on CMA 511a254129eSJoonsoo Kim default 7 512a254129eSJoonsoo Kim help 513a254129eSJoonsoo Kim CMA allows to create CMA areas for particular purpose, mainly, 514a254129eSJoonsoo Kim used as device private area. This parameter sets the maximum 515a254129eSJoonsoo Kim number of CMA area in the system. 516a254129eSJoonsoo Kim 517a254129eSJoonsoo Kim If unsure, leave the default value "7". 518a254129eSJoonsoo Kim 519af8d417aSDan Streetmanconfig MEM_SOFT_DIRTY 520af8d417aSDan Streetman bool "Track memory changes" 521af8d417aSDan Streetman depends on CHECKPOINT_RESTORE && HAVE_ARCH_SOFT_DIRTY && PROC_FS 522af8d417aSDan Streetman select PROC_PAGE_MONITOR 5234e2e2770SSeth Jennings help 524af8d417aSDan Streetman This option enables memory changes tracking by introducing a 525af8d417aSDan Streetman soft-dirty bit on pte-s. This bit it set when someone writes 526af8d417aSDan Streetman into a page just as regular dirty bit, but unlike the latter 527af8d417aSDan Streetman it can be cleared by hands. 528af8d417aSDan Streetman 5291ad1335dSMike Rapoport See Documentation/admin-guide/mm/soft-dirty.rst for more details. 5304e2e2770SSeth Jennings 5312b281117SSeth Jenningsconfig ZSWAP 5322b281117SSeth Jennings bool "Compressed cache for swap pages (EXPERIMENTAL)" 5332b281117SSeth Jennings depends on FRONTSWAP && CRYPTO=y 5342b281117SSeth Jennings select CRYPTO_LZO 53512d79d64SDan Streetman select ZPOOL 5362b281117SSeth Jennings help 5372b281117SSeth Jennings A lightweight compressed cache for swap pages. It takes 5382b281117SSeth Jennings pages that are in the process of being swapped out and attempts to 5392b281117SSeth Jennings compress them into a dynamically allocated RAM-based memory pool. 5402b281117SSeth Jennings This can result in a significant I/O reduction on swap device and, 5412b281117SSeth Jennings in the case where decompressing from RAM is faster that swap device 5422b281117SSeth Jennings reads, can also improve workload performance. 5432b281117SSeth Jennings 5442b281117SSeth Jennings This is marked experimental because it is a new feature (as of 5452b281117SSeth Jennings v3.11) that interacts heavily with memory reclaim. While these 5462b281117SSeth Jennings interactions don't cause any known issues on simple memory setups, 5472b281117SSeth Jennings they have not be fully explored on the large set of potential 5482b281117SSeth Jennings configurations and workloads that exist. 5492b281117SSeth Jennings 550af8d417aSDan Streetmanconfig ZPOOL 551af8d417aSDan Streetman tristate "Common API for compressed memory storage" 5520f8975ecSPavel Emelyanov help 553af8d417aSDan Streetman Compressed memory storage API. This allows using either zbud or 554af8d417aSDan Streetman zsmalloc. 5550f8975ecSPavel Emelyanov 556af8d417aSDan Streetmanconfig ZBUD 5579a001fc1SVitaly Wool tristate "Low (Up to 2x) density storage for compressed pages" 558af8d417aSDan Streetman help 559af8d417aSDan Streetman A special purpose allocator for storing compressed pages. 560af8d417aSDan Streetman It is designed to store up to two compressed pages per physical 561af8d417aSDan Streetman page. While this design limits storage density, it has simple and 562af8d417aSDan Streetman deterministic reclaim properties that make it preferable to a higher 563af8d417aSDan Streetman density approach when reclaim will be used. 564bcf1647dSMinchan Kim 5659a001fc1SVitaly Woolconfig Z3FOLD 5669a001fc1SVitaly Wool tristate "Up to 3x density storage for compressed pages" 5679a001fc1SVitaly Wool depends on ZPOOL 5689a001fc1SVitaly Wool help 5699a001fc1SVitaly Wool A special purpose allocator for storing compressed pages. 5709a001fc1SVitaly Wool It is designed to store up to three compressed pages per physical 5719a001fc1SVitaly Wool page. It is a ZBUD derivative so the simplicity and determinism are 5729a001fc1SVitaly Wool still there. 5739a001fc1SVitaly Wool 574bcf1647dSMinchan Kimconfig ZSMALLOC 575d867f203SMinchan Kim tristate "Memory allocator for compressed pages" 576bcf1647dSMinchan Kim depends on MMU 577bcf1647dSMinchan Kim help 578bcf1647dSMinchan Kim zsmalloc is a slab-based memory allocator designed to store 579bcf1647dSMinchan Kim compressed RAM pages. zsmalloc uses virtual memory mapping 580bcf1647dSMinchan Kim in order to reduce fragmentation. However, this results in a 581bcf1647dSMinchan Kim non-standard allocator interface where a handle, not a pointer, is 582bcf1647dSMinchan Kim returned by an alloc(). This handle must be mapped in order to 583bcf1647dSMinchan Kim access the allocated space. 584bcf1647dSMinchan Kim 585bcf1647dSMinchan Kimconfig PGTABLE_MAPPING 586bcf1647dSMinchan Kim bool "Use page table mapping to access object in zsmalloc" 587bcf1647dSMinchan Kim depends on ZSMALLOC 588bcf1647dSMinchan Kim help 589bcf1647dSMinchan Kim By default, zsmalloc uses a copy-based object mapping method to 590bcf1647dSMinchan Kim access allocations that span two pages. However, if a particular 591bcf1647dSMinchan Kim architecture (ex, ARM) performs VM mapping faster than copying, 592bcf1647dSMinchan Kim then you should select this. This causes zsmalloc to use page table 593bcf1647dSMinchan Kim mapping rather than copying for object mapping. 594bcf1647dSMinchan Kim 5952216ee85SBen Hutchings You can check speed with zsmalloc benchmark: 5962216ee85SBen Hutchings https://github.com/spartacus06/zsmapbench 5979e5c33d7SMark Salter 5980f050d99SGanesh Mahendranconfig ZSMALLOC_STAT 5990f050d99SGanesh Mahendran bool "Export zsmalloc statistics" 6000f050d99SGanesh Mahendran depends on ZSMALLOC 6010f050d99SGanesh Mahendran select DEBUG_FS 6020f050d99SGanesh Mahendran help 6030f050d99SGanesh Mahendran This option enables code in the zsmalloc to collect various 6040f050d99SGanesh Mahendran statistics about whats happening in zsmalloc and exports that 6050f050d99SGanesh Mahendran information to userspace via debugfs. 6060f050d99SGanesh Mahendran If unsure, say N. 6070f050d99SGanesh Mahendran 6089e5c33d7SMark Salterconfig GENERIC_EARLY_IOREMAP 6099e5c33d7SMark Salter bool 610042d27acSHelge Deller 611042d27acSHelge Dellerconfig MAX_STACK_SIZE_MB 612042d27acSHelge Deller int "Maximum user stack size for 32-bit processes (MB)" 613042d27acSHelge Deller default 80 614042d27acSHelge Deller range 8 2048 615042d27acSHelge Deller depends on STACK_GROWSUP && (!64BIT || COMPAT) 616042d27acSHelge Deller help 617042d27acSHelge Deller This is the maximum stack size in Megabytes in the VM layout of 32-bit 618042d27acSHelge Deller user processes when the stack grows upwards (currently only on parisc 6195f171577SJames Hogan arch). The stack will be located at the highest memory address minus 6205f171577SJames Hogan the given value, unless the RLIMIT_STACK hard limit is changed to a 6215f171577SJames Hogan smaller value in which case that is used. 622042d27acSHelge Deller 623042d27acSHelge Deller A sane initial value is 80 MB. 6243a80a7faSMel Gorman 6253a80a7faSMel Gormanconfig DEFERRED_STRUCT_PAGE_INIT 6261ce22103SVlastimil Babka bool "Defer initialisation of struct pages to kthreads" 627d39f8fb4SMike Rapoport depends on SPARSEMEM 628ab1e8d89SPavel Tatashin depends on !NEED_PER_CPU_KM 629889c695dSPasha Tatashin depends on 64BIT 6303a80a7faSMel Gorman help 6313a80a7faSMel Gorman Ordinarily all struct pages are initialised during early boot in a 6323a80a7faSMel Gorman single thread. On very large machines this can take a considerable 6333a80a7faSMel Gorman amount of time. If this option is set, large machines will bring up 6343a80a7faSMel Gorman a subset of memmap at boot and then initialise the rest in parallel 6351ce22103SVlastimil Babka by starting one-off "pgdatinitX" kernel thread for each node X. This 6361ce22103SVlastimil Babka has a potential performance impact on processes running early in the 6371ce22103SVlastimil Babka lifetime of the system until these kthreads finish the 6381ce22103SVlastimil Babka initialisation. 639033fbae9SDan Williams 64033c3fc71SVladimir Davydovconfig IDLE_PAGE_TRACKING 64133c3fc71SVladimir Davydov bool "Enable idle page tracking" 64233c3fc71SVladimir Davydov depends on SYSFS && MMU 64333c3fc71SVladimir Davydov select PAGE_EXTENSION if !64BIT 64433c3fc71SVladimir Davydov help 64533c3fc71SVladimir Davydov This feature allows to estimate the amount of user pages that have 64633c3fc71SVladimir Davydov not been touched during a given period of time. This information can 64733c3fc71SVladimir Davydov be useful to tune memory cgroup limits and/or for job placement 64833c3fc71SVladimir Davydov within a compute cluster. 64933c3fc71SVladimir Davydov 6501ad1335dSMike Rapoport See Documentation/admin-guide/mm/idle_page_tracking.rst for 6511ad1335dSMike Rapoport more details. 65233c3fc71SVladimir Davydov 65317596731SRobin Murphyconfig ARCH_HAS_PTE_DEVMAP 65465f7d049SOliver O'Halloran bool 65565f7d049SOliver O'Halloran 656033fbae9SDan Williamsconfig ZONE_DEVICE 6575042db43SJérôme Glisse bool "Device memory (pmem, HMM, etc...) hotplug support" 658033fbae9SDan Williams depends on MEMORY_HOTPLUG 659033fbae9SDan Williams depends on MEMORY_HOTREMOVE 66099490f16SDan Williams depends on SPARSEMEM_VMEMMAP 66117596731SRobin Murphy depends on ARCH_HAS_PTE_DEVMAP 6623a08cd52SMatthew Wilcox select XARRAY_MULTI 663033fbae9SDan Williams 664033fbae9SDan Williams help 665033fbae9SDan Williams Device memory hotplug support allows for establishing pmem, 666033fbae9SDan Williams or other device driver discovered memory regions, in the 667033fbae9SDan Williams memmap. This allows pfn_to_page() lookups of otherwise 668033fbae9SDan Williams "device-physical" addresses which is needed for using a DAX 669033fbae9SDan Williams mapping in an O_DIRECT operation, among other things. 670033fbae9SDan Williams 671033fbae9SDan Williams If FS_DAX is enabled, then say Y. 67206a660adSLinus Torvalds 673e7638488SDan Williamsconfig DEV_PAGEMAP_OPS 674e7638488SDan Williams bool 675e7638488SDan Williams 6769c240a7bSChristoph Hellwig# 6779c240a7bSChristoph Hellwig# Helpers to mirror range of the CPU page tables of a process into device page 6789c240a7bSChristoph Hellwig# tables. 6799c240a7bSChristoph Hellwig# 680c0b12405SJérôme Glisseconfig HMM_MIRROR 6819c240a7bSChristoph Hellwig bool 682f442c283SChristoph Hellwig depends on MMU 683c0b12405SJérôme Glisse 6845042db43SJérôme Glisseconfig DEVICE_PRIVATE 6855042db43SJérôme Glisse bool "Unaddressable device memory (GPU memory, ...)" 6867328d9ccSChristoph Hellwig depends on ZONE_DEVICE 687e7638488SDan Williams select DEV_PAGEMAP_OPS 6885042db43SJérôme Glisse 6895042db43SJérôme Glisse help 6905042db43SJérôme Glisse Allows creation of struct pages to represent unaddressable device 6915042db43SJérôme Glisse memory; i.e., memory that is only accessible from the device (or 6925042db43SJérôme Glisse group of devices). You likely also want to select HMM_MIRROR. 6935042db43SJérôme Glisse 6948025e5ddSJan Karaconfig FRAME_VECTOR 6958025e5ddSJan Kara bool 69663c17fb8SDave Hansen 69763c17fb8SDave Hansenconfig ARCH_USES_HIGH_VMA_FLAGS 69863c17fb8SDave Hansen bool 69966d37570SDave Hansenconfig ARCH_HAS_PKEYS 70066d37570SDave Hansen bool 70130a5b536SDennis Zhou 70230a5b536SDennis Zhouconfig PERCPU_STATS 70330a5b536SDennis Zhou bool "Collect percpu memory statistics" 70430a5b536SDennis Zhou help 70530a5b536SDennis Zhou This feature collects and exposes statistics via debugfs. The 70630a5b536SDennis Zhou information includes global and per chunk statistics, which can 70730a5b536SDennis Zhou be used to help understand percpu memory usage. 70864c349f4SKirill A. Shutemov 70964c349f4SKirill A. Shutemovconfig GUP_BENCHMARK 71064c349f4SKirill A. Shutemov bool "Enable infrastructure for get_user_pages_fast() benchmarking" 71164c349f4SKirill A. Shutemov help 71264c349f4SKirill A. Shutemov Provides /sys/kernel/debug/gup_benchmark that helps with testing 71364c349f4SKirill A. Shutemov performance of get_user_pages_fast(). 71464c349f4SKirill A. Shutemov 71564c349f4SKirill A. Shutemov See tools/testing/selftests/vm/gup_benchmark.c 7163010a5eaSLaurent Dufour 71739656e83SChristoph Hellwigconfig GUP_GET_PTE_LOW_HIGH 71839656e83SChristoph Hellwig bool 71939656e83SChristoph Hellwig 72099cb0dbdSSong Liuconfig READ_ONLY_THP_FOR_FS 72199cb0dbdSSong Liu bool "Read-only THP for filesystems (EXPERIMENTAL)" 72299cb0dbdSSong Liu depends on TRANSPARENT_HUGE_PAGECACHE && SHMEM 72399cb0dbdSSong Liu 72499cb0dbdSSong Liu help 72599cb0dbdSSong Liu Allow khugepaged to put read-only file-backed pages in THP. 72699cb0dbdSSong Liu 72799cb0dbdSSong Liu This is marked experimental because it is a new feature. Write 72899cb0dbdSSong Liu support of file THPs will be developed in the next few release 72999cb0dbdSSong Liu cycles. 73099cb0dbdSSong Liu 7313010a5eaSLaurent Dufourconfig ARCH_HAS_PTE_SPECIAL 7323010a5eaSLaurent Dufour bool 73359e0b520SChristoph Hellwig 734cbd34da7SChristoph Hellwig# 735cbd34da7SChristoph Hellwig# Some architectures require a special hugepage directory format that is 736cbd34da7SChristoph Hellwig# required to support multiple hugepage sizes. For example a4fe3ce76 737cbd34da7SChristoph Hellwig# "powerpc/mm: Allow more flexible layouts for hugepage pagetables" 738cbd34da7SChristoph Hellwig# introduced it on powerpc. This allows for a more flexible hugepage 739cbd34da7SChristoph Hellwig# pagetable layouts. 740cbd34da7SChristoph Hellwig# 741cbd34da7SChristoph Hellwigconfig ARCH_HAS_HUGEPD 742cbd34da7SChristoph Hellwig bool 743cbd34da7SChristoph Hellwig 744c5acad84SThomas Hellstromconfig MAPPING_DIRTY_HELPERS 745c5acad84SThomas Hellstrom bool 746c5acad84SThomas Hellstrom 74759e0b520SChristoph Hellwigendmenu 748