1ec8f24b7SThomas Gleixner# SPDX-License-Identifier: GPL-2.0-only 259e0b520SChristoph Hellwig 359e0b520SChristoph Hellwigmenu "Memory Management options" 459e0b520SChristoph Hellwig 57b42f104SJohannes Weiner# 67b42f104SJohannes Weiner# For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can 77b42f104SJohannes Weiner# add proper SWAP support to them, in which case this can be remove. 87b42f104SJohannes Weiner# 97b42f104SJohannes Weinerconfig ARCH_NO_SWAP 107b42f104SJohannes Weiner bool 117b42f104SJohannes Weiner 12b3fbd58fSJohannes Weinerconfig ZPOOL 13b3fbd58fSJohannes Weiner bool 14b3fbd58fSJohannes Weiner 15519bcb79SJohannes Weinermenuconfig SWAP 167b42f104SJohannes Weiner bool "Support for paging of anonymous memory (swap)" 177b42f104SJohannes Weiner depends on MMU && BLOCK && !ARCH_NO_SWAP 187b42f104SJohannes Weiner default y 197b42f104SJohannes Weiner help 207b42f104SJohannes Weiner This option allows you to choose whether you want to have support 217b42f104SJohannes Weiner for so called swap devices or swap files in your kernel that are 227b42f104SJohannes Weiner used to provide more virtual memory than the actual RAM present 237b42f104SJohannes Weiner in your computer. If unsure say Y. 247b42f104SJohannes Weiner 25519bcb79SJohannes Weinerconfig ZSWAP 26fcab9b44SDavid Heidelberg bool "Compressed cache for swap pages" 27b3fbd58fSJohannes Weiner depends on SWAP 28519bcb79SJohannes Weiner select FRONTSWAP 29b3fbd58fSJohannes Weiner select CRYPTO 30519bcb79SJohannes Weiner select ZPOOL 31519bcb79SJohannes Weiner help 32519bcb79SJohannes Weiner A lightweight compressed cache for swap pages. It takes 33519bcb79SJohannes Weiner pages that are in the process of being swapped out and attempts to 34519bcb79SJohannes Weiner compress them into a dynamically allocated RAM-based memory pool. 35519bcb79SJohannes Weiner This can result in a significant I/O reduction on swap device and, 361a44131dSSophia Gabriella in the case where decompressing from RAM is faster than swap device 37519bcb79SJohannes Weiner reads, can also improve workload performance. 38519bcb79SJohannes Weiner 39b3fbd58fSJohannes Weinerconfig ZSWAP_DEFAULT_ON 40b3fbd58fSJohannes Weiner bool "Enable the compressed cache for swap pages by default" 41b3fbd58fSJohannes Weiner depends on ZSWAP 42b3fbd58fSJohannes Weiner help 43b3fbd58fSJohannes Weiner If selected, the compressed cache for swap pages will be enabled 44b3fbd58fSJohannes Weiner at boot, otherwise it will be disabled. 45b3fbd58fSJohannes Weiner 46b3fbd58fSJohannes Weiner The selection made here can be overridden by using the kernel 47b3fbd58fSJohannes Weiner command line 'zswap.enabled=' option. 48b3fbd58fSJohannes Weiner 49519bcb79SJohannes Weinerchoice 50b3fbd58fSJohannes Weiner prompt "Default compressor" 51519bcb79SJohannes Weiner depends on ZSWAP 52519bcb79SJohannes Weiner default ZSWAP_COMPRESSOR_DEFAULT_LZO 53519bcb79SJohannes Weiner help 54519bcb79SJohannes Weiner Selects the default compression algorithm for the compressed cache 55519bcb79SJohannes Weiner for swap pages. 56519bcb79SJohannes Weiner 57519bcb79SJohannes Weiner For an overview what kind of performance can be expected from 58519bcb79SJohannes Weiner a particular compression algorithm please refer to the benchmarks 59519bcb79SJohannes Weiner available at the following LWN page: 60519bcb79SJohannes Weiner https://lwn.net/Articles/751795/ 61519bcb79SJohannes Weiner 62519bcb79SJohannes Weiner If in doubt, select 'LZO'. 63519bcb79SJohannes Weiner 64519bcb79SJohannes Weiner The selection made here can be overridden by using the kernel 65519bcb79SJohannes Weiner command line 'zswap.compressor=' option. 66519bcb79SJohannes Weiner 67519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT_DEFLATE 68519bcb79SJohannes Weiner bool "Deflate" 69519bcb79SJohannes Weiner select CRYPTO_DEFLATE 70519bcb79SJohannes Weiner help 71519bcb79SJohannes Weiner Use the Deflate algorithm as the default compression algorithm. 72519bcb79SJohannes Weiner 73519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT_LZO 74519bcb79SJohannes Weiner bool "LZO" 75519bcb79SJohannes Weiner select CRYPTO_LZO 76519bcb79SJohannes Weiner help 77519bcb79SJohannes Weiner Use the LZO algorithm as the default compression algorithm. 78519bcb79SJohannes Weiner 79519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT_842 80519bcb79SJohannes Weiner bool "842" 81519bcb79SJohannes Weiner select CRYPTO_842 82519bcb79SJohannes Weiner help 83519bcb79SJohannes Weiner Use the 842 algorithm as the default compression algorithm. 84519bcb79SJohannes Weiner 85519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT_LZ4 86519bcb79SJohannes Weiner bool "LZ4" 87519bcb79SJohannes Weiner select CRYPTO_LZ4 88519bcb79SJohannes Weiner help 89519bcb79SJohannes Weiner Use the LZ4 algorithm as the default compression algorithm. 90519bcb79SJohannes Weiner 91519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT_LZ4HC 92519bcb79SJohannes Weiner bool "LZ4HC" 93519bcb79SJohannes Weiner select CRYPTO_LZ4HC 94519bcb79SJohannes Weiner help 95519bcb79SJohannes Weiner Use the LZ4HC algorithm as the default compression algorithm. 96519bcb79SJohannes Weiner 97519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT_ZSTD 98519bcb79SJohannes Weiner bool "zstd" 99519bcb79SJohannes Weiner select CRYPTO_ZSTD 100519bcb79SJohannes Weiner help 101519bcb79SJohannes Weiner Use the zstd algorithm as the default compression algorithm. 102519bcb79SJohannes Weinerendchoice 103519bcb79SJohannes Weiner 104519bcb79SJohannes Weinerconfig ZSWAP_COMPRESSOR_DEFAULT 105519bcb79SJohannes Weiner string 106519bcb79SJohannes Weiner depends on ZSWAP 107519bcb79SJohannes Weiner default "deflate" if ZSWAP_COMPRESSOR_DEFAULT_DEFLATE 108519bcb79SJohannes Weiner default "lzo" if ZSWAP_COMPRESSOR_DEFAULT_LZO 109519bcb79SJohannes Weiner default "842" if ZSWAP_COMPRESSOR_DEFAULT_842 110519bcb79SJohannes Weiner default "lz4" if ZSWAP_COMPRESSOR_DEFAULT_LZ4 111519bcb79SJohannes Weiner default "lz4hc" if ZSWAP_COMPRESSOR_DEFAULT_LZ4HC 112519bcb79SJohannes Weiner default "zstd" if ZSWAP_COMPRESSOR_DEFAULT_ZSTD 113519bcb79SJohannes Weiner default "" 114519bcb79SJohannes Weiner 115519bcb79SJohannes Weinerchoice 116b3fbd58fSJohannes Weiner prompt "Default allocator" 117519bcb79SJohannes Weiner depends on ZSWAP 118519bcb79SJohannes Weiner default ZSWAP_ZPOOL_DEFAULT_ZBUD 119519bcb79SJohannes Weiner help 120519bcb79SJohannes Weiner Selects the default allocator for the compressed cache for 121519bcb79SJohannes Weiner swap pages. 122519bcb79SJohannes Weiner The default is 'zbud' for compatibility, however please do 123519bcb79SJohannes Weiner read the description of each of the allocators below before 124519bcb79SJohannes Weiner making a right choice. 125519bcb79SJohannes Weiner 126519bcb79SJohannes Weiner The selection made here can be overridden by using the kernel 127519bcb79SJohannes Weiner command line 'zswap.zpool=' option. 128519bcb79SJohannes Weiner 129519bcb79SJohannes Weinerconfig ZSWAP_ZPOOL_DEFAULT_ZBUD 130519bcb79SJohannes Weiner bool "zbud" 131519bcb79SJohannes Weiner select ZBUD 132519bcb79SJohannes Weiner help 133519bcb79SJohannes Weiner Use the zbud allocator as the default allocator. 134519bcb79SJohannes Weiner 135519bcb79SJohannes Weinerconfig ZSWAP_ZPOOL_DEFAULT_Z3FOLD 136519bcb79SJohannes Weiner bool "z3fold" 137519bcb79SJohannes Weiner select Z3FOLD 138519bcb79SJohannes Weiner help 139519bcb79SJohannes Weiner Use the z3fold allocator as the default allocator. 140519bcb79SJohannes Weiner 141519bcb79SJohannes Weinerconfig ZSWAP_ZPOOL_DEFAULT_ZSMALLOC 142519bcb79SJohannes Weiner bool "zsmalloc" 143519bcb79SJohannes Weiner select ZSMALLOC 144519bcb79SJohannes Weiner help 145519bcb79SJohannes Weiner Use the zsmalloc allocator as the default allocator. 146519bcb79SJohannes Weinerendchoice 147519bcb79SJohannes Weiner 148519bcb79SJohannes Weinerconfig ZSWAP_ZPOOL_DEFAULT 149519bcb79SJohannes Weiner string 150519bcb79SJohannes Weiner depends on ZSWAP 151519bcb79SJohannes Weiner default "zbud" if ZSWAP_ZPOOL_DEFAULT_ZBUD 152519bcb79SJohannes Weiner default "z3fold" if ZSWAP_ZPOOL_DEFAULT_Z3FOLD 153519bcb79SJohannes Weiner default "zsmalloc" if ZSWAP_ZPOOL_DEFAULT_ZSMALLOC 154519bcb79SJohannes Weiner default "" 155519bcb79SJohannes Weiner 156519bcb79SJohannes Weinerconfig ZBUD 157b3fbd58fSJohannes Weiner tristate "2:1 compression allocator (zbud)" 158b3fbd58fSJohannes Weiner depends on ZSWAP 159519bcb79SJohannes Weiner help 160519bcb79SJohannes Weiner A special purpose allocator for storing compressed pages. 161519bcb79SJohannes Weiner It is designed to store up to two compressed pages per physical 162519bcb79SJohannes Weiner page. While this design limits storage density, it has simple and 163519bcb79SJohannes Weiner deterministic reclaim properties that make it preferable to a higher 164519bcb79SJohannes Weiner density approach when reclaim will be used. 165519bcb79SJohannes Weiner 166519bcb79SJohannes Weinerconfig Z3FOLD 167b3fbd58fSJohannes Weiner tristate "3:1 compression allocator (z3fold)" 168b3fbd58fSJohannes Weiner depends on ZSWAP 169519bcb79SJohannes Weiner help 170519bcb79SJohannes Weiner A special purpose allocator for storing compressed pages. 171519bcb79SJohannes Weiner It is designed to store up to three compressed pages per physical 172519bcb79SJohannes Weiner page. It is a ZBUD derivative so the simplicity and determinism are 173519bcb79SJohannes Weiner still there. 174519bcb79SJohannes Weiner 175519bcb79SJohannes Weinerconfig ZSMALLOC 176b3fbd58fSJohannes Weiner tristate 177b3fbd58fSJohannes Weiner prompt "N:1 compression allocator (zsmalloc)" if ZSWAP 178519bcb79SJohannes Weiner depends on MMU 179519bcb79SJohannes Weiner help 180519bcb79SJohannes Weiner zsmalloc is a slab-based memory allocator designed to store 181b3fbd58fSJohannes Weiner pages of various compression levels efficiently. It achieves 182b3fbd58fSJohannes Weiner the highest storage density with the least amount of fragmentation. 183519bcb79SJohannes Weiner 184519bcb79SJohannes Weinerconfig ZSMALLOC_STAT 185519bcb79SJohannes Weiner bool "Export zsmalloc statistics" 186519bcb79SJohannes Weiner depends on ZSMALLOC 187519bcb79SJohannes Weiner select DEBUG_FS 188519bcb79SJohannes Weiner help 189519bcb79SJohannes Weiner This option enables code in the zsmalloc to collect various 190519bcb79SJohannes Weiner statistics about what's happening in zsmalloc and exports that 191519bcb79SJohannes Weiner information to userspace via debugfs. 192519bcb79SJohannes Weiner If unsure, say N. 193519bcb79SJohannes Weiner 194519bcb79SJohannes Weinermenu "SLAB allocator options" 195519bcb79SJohannes Weiner 1967b42f104SJohannes Weinerchoice 1977b42f104SJohannes Weiner prompt "Choose SLAB allocator" 1987b42f104SJohannes Weiner default SLUB 1997b42f104SJohannes Weiner help 2007b42f104SJohannes Weiner This option allows to select a slab allocator. 2017b42f104SJohannes Weiner 2027b42f104SJohannes Weinerconfig SLAB 2037b42f104SJohannes Weiner bool "SLAB" 2047b42f104SJohannes Weiner depends on !PREEMPT_RT 2057b42f104SJohannes Weiner select HAVE_HARDENED_USERCOPY_ALLOCATOR 2067b42f104SJohannes Weiner help 2077b42f104SJohannes Weiner The regular slab allocator that is established and known to work 2087b42f104SJohannes Weiner well in all environments. It organizes cache hot objects in 2097b42f104SJohannes Weiner per cpu and per node queues. 2107b42f104SJohannes Weiner 2117b42f104SJohannes Weinerconfig SLUB 2127b42f104SJohannes Weiner bool "SLUB (Unqueued Allocator)" 2137b42f104SJohannes Weiner select HAVE_HARDENED_USERCOPY_ALLOCATOR 2147b42f104SJohannes Weiner help 2157b42f104SJohannes Weiner SLUB is a slab allocator that minimizes cache line usage 2167b42f104SJohannes Weiner instead of managing queues of cached objects (SLAB approach). 2177b42f104SJohannes Weiner Per cpu caching is realized using slabs of objects instead 2187b42f104SJohannes Weiner of queues of objects. SLUB can use memory efficiently 2197b42f104SJohannes Weiner and has enhanced diagnostics. SLUB is the default choice for 2207b42f104SJohannes Weiner a slab allocator. 2217b42f104SJohannes Weiner 2227b42f104SJohannes Weinerconfig SLOB 2237b42f104SJohannes Weiner depends on EXPERT 2247b42f104SJohannes Weiner bool "SLOB (Simple Allocator)" 2257b42f104SJohannes Weiner depends on !PREEMPT_RT 2267b42f104SJohannes Weiner help 2277b42f104SJohannes Weiner SLOB replaces the stock allocator with a drastically simpler 2287b42f104SJohannes Weiner allocator. SLOB is generally more space efficient but 2297b42f104SJohannes Weiner does not perform as well on large systems. 2307b42f104SJohannes Weiner 2317b42f104SJohannes Weinerendchoice 2327b42f104SJohannes Weiner 2337b42f104SJohannes Weinerconfig SLAB_MERGE_DEFAULT 2347b42f104SJohannes Weiner bool "Allow slab caches to be merged" 2357b42f104SJohannes Weiner default y 2367b42f104SJohannes Weiner depends on SLAB || SLUB 2377b42f104SJohannes Weiner help 2387b42f104SJohannes Weiner For reduced kernel memory fragmentation, slab caches can be 2397b42f104SJohannes Weiner merged when they share the same size and other characteristics. 2407b42f104SJohannes Weiner This carries a risk of kernel heap overflows being able to 2417b42f104SJohannes Weiner overwrite objects from merged caches (and more easily control 2427b42f104SJohannes Weiner cache layout), which makes such heap attacks easier to exploit 2437b42f104SJohannes Weiner by attackers. By keeping caches unmerged, these kinds of exploits 2447b42f104SJohannes Weiner can usually only damage objects in the same cache. To disable 2457b42f104SJohannes Weiner merging at runtime, "slab_nomerge" can be passed on the kernel 2467b42f104SJohannes Weiner command line. 2477b42f104SJohannes Weiner 2487b42f104SJohannes Weinerconfig SLAB_FREELIST_RANDOM 2497b42f104SJohannes Weiner bool "Randomize slab freelist" 2507b42f104SJohannes Weiner depends on SLAB || SLUB 2517b42f104SJohannes Weiner help 2527b42f104SJohannes Weiner Randomizes the freelist order used on creating new pages. This 2537b42f104SJohannes Weiner security feature reduces the predictability of the kernel slab 2547b42f104SJohannes Weiner allocator against heap overflows. 2557b42f104SJohannes Weiner 2567b42f104SJohannes Weinerconfig SLAB_FREELIST_HARDENED 2577b42f104SJohannes Weiner bool "Harden slab freelist metadata" 2587b42f104SJohannes Weiner depends on SLAB || SLUB 2597b42f104SJohannes Weiner help 2607b42f104SJohannes Weiner Many kernel heap attacks try to target slab cache metadata and 2617b42f104SJohannes Weiner other infrastructure. This options makes minor performance 2627b42f104SJohannes Weiner sacrifices to harden the kernel slab allocator against common 2637b42f104SJohannes Weiner freelist exploit methods. Some slab implementations have more 2647b42f104SJohannes Weiner sanity-checking than others. This option is most effective with 2657b42f104SJohannes Weiner CONFIG_SLUB. 2667b42f104SJohannes Weiner 2670710d012SVlastimil Babkaconfig SLUB_STATS 2680710d012SVlastimil Babka default n 2690710d012SVlastimil Babka bool "Enable SLUB performance statistics" 2700710d012SVlastimil Babka depends on SLUB && SYSFS 2710710d012SVlastimil Babka help 2720710d012SVlastimil Babka SLUB statistics are useful to debug SLUBs allocation behavior in 2730710d012SVlastimil Babka order find ways to optimize the allocator. This should never be 2740710d012SVlastimil Babka enabled for production use since keeping statistics slows down 2750710d012SVlastimil Babka the allocator by a few percentage points. The slabinfo command 2760710d012SVlastimil Babka supports the determination of the most active slabs to figure 2770710d012SVlastimil Babka out which slabs are relevant to a particular load. 2780710d012SVlastimil Babka Try running: slabinfo -DA 2790710d012SVlastimil Babka 280519bcb79SJohannes Weinerconfig SLUB_CPU_PARTIAL 281519bcb79SJohannes Weiner default y 282519bcb79SJohannes Weiner depends on SLUB && SMP 283519bcb79SJohannes Weiner bool "SLUB per cpu partial cache" 284519bcb79SJohannes Weiner help 285519bcb79SJohannes Weiner Per cpu partial caches accelerate objects allocation and freeing 286519bcb79SJohannes Weiner that is local to a processor at the price of more indeterminism 287519bcb79SJohannes Weiner in the latency of the free. On overflow these caches will be cleared 288519bcb79SJohannes Weiner which requires the taking of locks that may cause latency spikes. 289519bcb79SJohannes Weiner Typically one would choose no for a realtime system. 290519bcb79SJohannes Weiner 291519bcb79SJohannes Weinerendmenu # SLAB allocator options 292519bcb79SJohannes Weiner 2937b42f104SJohannes Weinerconfig SHUFFLE_PAGE_ALLOCATOR 2947b42f104SJohannes Weiner bool "Page allocator randomization" 2957b42f104SJohannes Weiner default SLAB_FREELIST_RANDOM && ACPI_NUMA 2967b42f104SJohannes Weiner help 2977b42f104SJohannes Weiner Randomization of the page allocator improves the average 2987b42f104SJohannes Weiner utilization of a direct-mapped memory-side-cache. See section 2997b42f104SJohannes Weiner 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI 3007b42f104SJohannes Weiner 6.2a specification for an example of how a platform advertises 3017b42f104SJohannes Weiner the presence of a memory-side-cache. There are also incidental 3027b42f104SJohannes Weiner security benefits as it reduces the predictability of page 3037b42f104SJohannes Weiner allocations to compliment SLAB_FREELIST_RANDOM, but the 3047b42f104SJohannes Weiner default granularity of shuffling on the "MAX_ORDER - 1" i.e, 3057b42f104SJohannes Weiner 10th order of pages is selected based on cache utilization 3067b42f104SJohannes Weiner benefits on x86. 3077b42f104SJohannes Weiner 3087b42f104SJohannes Weiner While the randomization improves cache utilization it may 3097b42f104SJohannes Weiner negatively impact workloads on platforms without a cache. For 3107b42f104SJohannes Weiner this reason, by default, the randomization is enabled only 3117b42f104SJohannes Weiner after runtime detection of a direct-mapped memory-side-cache. 3127b42f104SJohannes Weiner Otherwise, the randomization may be force enabled with the 3137b42f104SJohannes Weiner 'page_alloc.shuffle' kernel command line parameter. 3147b42f104SJohannes Weiner 3157b42f104SJohannes Weiner Say Y if unsure. 3167b42f104SJohannes Weiner 3170710d012SVlastimil Babkaconfig COMPAT_BRK 3180710d012SVlastimil Babka bool "Disable heap randomization" 3190710d012SVlastimil Babka default y 3200710d012SVlastimil Babka help 3210710d012SVlastimil Babka Randomizing heap placement makes heap exploits harder, but it 3220710d012SVlastimil Babka also breaks ancient binaries (including anything libc5 based). 3230710d012SVlastimil Babka This option changes the bootup default to heap randomization 3240710d012SVlastimil Babka disabled, and can be overridden at runtime by setting 3250710d012SVlastimil Babka /proc/sys/kernel/randomize_va_space to 2. 3260710d012SVlastimil Babka 3270710d012SVlastimil Babka On non-ancient distros (post-2000 ones) N is usually a safe choice. 3280710d012SVlastimil Babka 3290710d012SVlastimil Babkaconfig MMAP_ALLOW_UNINITIALIZED 3300710d012SVlastimil Babka bool "Allow mmapped anonymous memory to be uninitialized" 3310710d012SVlastimil Babka depends on EXPERT && !MMU 3320710d012SVlastimil Babka default n 3330710d012SVlastimil Babka help 3340710d012SVlastimil Babka Normally, and according to the Linux spec, anonymous memory obtained 3350710d012SVlastimil Babka from mmap() has its contents cleared before it is passed to 3360710d012SVlastimil Babka userspace. Enabling this config option allows you to request that 3370710d012SVlastimil Babka mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 3380710d012SVlastimil Babka providing a huge performance boost. If this option is not enabled, 3390710d012SVlastimil Babka then the flag will be ignored. 3400710d012SVlastimil Babka 3410710d012SVlastimil Babka This is taken advantage of by uClibc's malloc(), and also by 3420710d012SVlastimil Babka ELF-FDPIC binfmt's brk and stack allocator. 3430710d012SVlastimil Babka 3440710d012SVlastimil Babka Because of the obvious security issues, this option should only be 3450710d012SVlastimil Babka enabled on embedded devices where you control what is run in 3460710d012SVlastimil Babka userspace. Since that isn't generally a problem on no-MMU systems, 3470710d012SVlastimil Babka it is normally safe to say Y here. 3480710d012SVlastimil Babka 3490710d012SVlastimil Babka See Documentation/admin-guide/mm/nommu-mmap.rst for more information. 3500710d012SVlastimil Babka 351e1785e85SDave Hansenconfig SELECT_MEMORY_MODEL 352e1785e85SDave Hansen def_bool y 353a8826eebSKees Cook depends on ARCH_SELECT_MEMORY_MODEL 354e1785e85SDave Hansen 3553a9da765SDave Hansenchoice 3563a9da765SDave Hansen prompt "Memory model" 357e1785e85SDave Hansen depends on SELECT_MEMORY_MODEL 358d41dee36SAndy Whitcroft default SPARSEMEM_MANUAL if ARCH_SPARSEMEM_DEFAULT 359e1785e85SDave Hansen default FLATMEM_MANUAL 360d66d109dSMike Rapoport help 361d66d109dSMike Rapoport This option allows you to change some of the ways that 362d66d109dSMike Rapoport Linux manages its memory internally. Most users will 363d66d109dSMike Rapoport only have one option here selected by the architecture 364d66d109dSMike Rapoport configuration. This is normal. 3653a9da765SDave Hansen 366e1785e85SDave Hansenconfig FLATMEM_MANUAL 3673a9da765SDave Hansen bool "Flat Memory" 368bb1c50d3SMike Rapoport depends on !ARCH_SPARSEMEM_ENABLE || ARCH_FLATMEM_ENABLE 3693a9da765SDave Hansen help 370d66d109dSMike Rapoport This option is best suited for non-NUMA systems with 371d66d109dSMike Rapoport flat address space. The FLATMEM is the most efficient 372d66d109dSMike Rapoport system in terms of performance and resource consumption 373d66d109dSMike Rapoport and it is the best option for smaller systems. 3743a9da765SDave Hansen 375d66d109dSMike Rapoport For systems that have holes in their physical address 376d66d109dSMike Rapoport spaces and for features like NUMA and memory hotplug, 377dd33d29aSRandy Dunlap choose "Sparse Memory". 378d41dee36SAndy Whitcroft 379d41dee36SAndy Whitcroft If unsure, choose this option (Flat Memory) over any other. 3803a9da765SDave Hansen 381d41dee36SAndy Whitcroftconfig SPARSEMEM_MANUAL 382d41dee36SAndy Whitcroft bool "Sparse Memory" 383d41dee36SAndy Whitcroft depends on ARCH_SPARSEMEM_ENABLE 384d41dee36SAndy Whitcroft help 385d41dee36SAndy Whitcroft This will be the only option for some systems, including 386d66d109dSMike Rapoport memory hot-plug systems. This is normal. 387d41dee36SAndy Whitcroft 388d66d109dSMike Rapoport This option provides efficient support for systems with 389d66d109dSMike Rapoport holes is their physical address space and allows memory 390d66d109dSMike Rapoport hot-plug and hot-remove. 391d41dee36SAndy Whitcroft 392d66d109dSMike Rapoport If unsure, choose "Flat Memory" over this option. 393d41dee36SAndy Whitcroft 3943a9da765SDave Hansenendchoice 3953a9da765SDave Hansen 396d41dee36SAndy Whitcroftconfig SPARSEMEM 397d41dee36SAndy Whitcroft def_bool y 3981a83e175SRussell King depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL 399d41dee36SAndy Whitcroft 400e1785e85SDave Hansenconfig FLATMEM 401e1785e85SDave Hansen def_bool y 402bb1c50d3SMike Rapoport depends on !SPARSEMEM || FLATMEM_MANUAL 403d41dee36SAndy Whitcroft 40493b7504eSDave Hansen# 4053e347261SBob Picco# SPARSEMEM_EXTREME (which is the default) does some bootmem 406c89ab04fSMike Rapoport# allocations when sparse_init() is called. If this cannot 4073e347261SBob Picco# be done on your architecture, select this option. However, 4083e347261SBob Picco# statically allocating the mem_section[] array can potentially 4093e347261SBob Picco# consume vast quantities of .bss, so be careful. 4103e347261SBob Picco# 4113e347261SBob Picco# This option will also potentially produce smaller runtime code 4123e347261SBob Picco# with gcc 3.4 and later. 4133e347261SBob Picco# 4143e347261SBob Piccoconfig SPARSEMEM_STATIC 4159ba16087SJan Beulich bool 4163e347261SBob Picco 4173e347261SBob Picco# 41844c09201SMatt LaPlante# Architecture platforms which require a two level mem_section in SPARSEMEM 419802f192eSBob Picco# must select this option. This is usually for architecture platforms with 420802f192eSBob Picco# an extremely sparse physical address space. 421802f192eSBob Picco# 4223e347261SBob Piccoconfig SPARSEMEM_EXTREME 4233e347261SBob Picco def_bool y 4243e347261SBob Picco depends on SPARSEMEM && !SPARSEMEM_STATIC 4254c21e2f2SHugh Dickins 42629c71111SAndy Whitcroftconfig SPARSEMEM_VMEMMAP_ENABLE 4279ba16087SJan Beulich bool 42829c71111SAndy Whitcroft 42929c71111SAndy Whitcroftconfig SPARSEMEM_VMEMMAP 430a5ee6daaSGeoff Levand bool "Sparse Memory virtual memmap" 431a5ee6daaSGeoff Levand depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE 432a5ee6daaSGeoff Levand default y 433a5ee6daaSGeoff Levand help 434a5ee6daaSGeoff Levand SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise 435a5ee6daaSGeoff Levand pfn_to_page and page_to_pfn operations. This is the most 436a5ee6daaSGeoff Levand efficient option when sufficient kernel resources are available. 43729c71111SAndy Whitcroft 43870210ed9SPhilipp Hachtmannconfig HAVE_MEMBLOCK_PHYS_MAP 4396341e62bSChristoph Jaeger bool 44070210ed9SPhilipp Hachtmann 44167a929e0SChristoph Hellwigconfig HAVE_FAST_GUP 442050a9adcSChristoph Hellwig depends on MMU 4436341e62bSChristoph Jaeger bool 4442667f50eSSteve Capper 44552219aeaSDavid Hildenbrand# Don't discard allocated memory used to track "memory" and "reserved" memblocks 44652219aeaSDavid Hildenbrand# after early boot, so it can still be used to test for validity of memory. 44752219aeaSDavid Hildenbrand# Also, memblocks are updated with memory hot(un)plug. 448350e88baSMike Rapoportconfig ARCH_KEEP_MEMBLOCK 4496341e62bSChristoph Jaeger bool 450c378ddd5STejun Heo 4511e5d8e1eSDan Williams# Keep arch NUMA mapping infrastructure post-init. 4521e5d8e1eSDan Williamsconfig NUMA_KEEP_MEMINFO 4531e5d8e1eSDan Williams bool 4541e5d8e1eSDan Williams 455ee6f509cSMinchan Kimconfig MEMORY_ISOLATION 4566341e62bSChristoph Jaeger bool 457ee6f509cSMinchan Kim 458a9e7b8d4SDavid Hildenbrand# IORESOURCE_SYSTEM_RAM regions in the kernel resource tree that are marked 459a9e7b8d4SDavid Hildenbrand# IORESOURCE_EXCLUSIVE cannot be mapped to user space, for example, via 460a9e7b8d4SDavid Hildenbrand# /dev/mem. 461a9e7b8d4SDavid Hildenbrandconfig EXCLUSIVE_SYSTEM_RAM 462a9e7b8d4SDavid Hildenbrand def_bool y 463a9e7b8d4SDavid Hildenbrand depends on !DEVMEM || STRICT_DEVMEM 464a9e7b8d4SDavid Hildenbrand 46546723bfaSYasuaki Ishimatsu# 46646723bfaSYasuaki Ishimatsu# Only be set on architectures that have completely implemented memory hotplug 46746723bfaSYasuaki Ishimatsu# feature. If you are not sure, don't touch it. 46846723bfaSYasuaki Ishimatsu# 46946723bfaSYasuaki Ishimatsuconfig HAVE_BOOTMEM_INFO_NODE 47046723bfaSYasuaki Ishimatsu def_bool n 47146723bfaSYasuaki Ishimatsu 47291024b3cSAnshuman Khandualconfig ARCH_ENABLE_MEMORY_HOTPLUG 47391024b3cSAnshuman Khandual bool 47491024b3cSAnshuman Khandual 475519bcb79SJohannes Weinerconfig ARCH_ENABLE_MEMORY_HOTREMOVE 476519bcb79SJohannes Weiner bool 477519bcb79SJohannes Weiner 4783947be19SDave Hansen# eventually, we can have this option just 'select SPARSEMEM' 479519bcb79SJohannes Weinermenuconfig MEMORY_HOTPLUG 480519bcb79SJohannes Weiner bool "Memory hotplug" 481b30c5927SDavid Hildenbrand select MEMORY_ISOLATION 48271b6f2ddSDavid Hildenbrand depends on SPARSEMEM 48340b31360SStephen Rothwell depends on ARCH_ENABLE_MEMORY_HOTPLUG 4847ec58a2bSDavid Hildenbrand depends on 64BIT 4851e5d8e1eSDan Williams select NUMA_KEEP_MEMINFO if NUMA 4863947be19SDave Hansen 487519bcb79SJohannes Weinerif MEMORY_HOTPLUG 488519bcb79SJohannes Weiner 4898604d9e5SVitaly Kuznetsovconfig MEMORY_HOTPLUG_DEFAULT_ONLINE 4908604d9e5SVitaly Kuznetsov bool "Online the newly added memory blocks by default" 4918604d9e5SVitaly Kuznetsov depends on MEMORY_HOTPLUG 4928604d9e5SVitaly Kuznetsov help 4938604d9e5SVitaly Kuznetsov This option sets the default policy setting for memory hotplug 4948604d9e5SVitaly Kuznetsov onlining policy (/sys/devices/system/memory/auto_online_blocks) which 4958604d9e5SVitaly Kuznetsov determines what happens to newly added memory regions. Policy setting 4968604d9e5SVitaly Kuznetsov can always be changed at runtime. 497cb1aaebeSMauro Carvalho Chehab See Documentation/admin-guide/mm/memory-hotplug.rst for more information. 4988604d9e5SVitaly Kuznetsov 4998604d9e5SVitaly Kuznetsov Say Y here if you want all hot-plugged memory blocks to appear in 5008604d9e5SVitaly Kuznetsov 'online' state by default. 5018604d9e5SVitaly Kuznetsov Say N here if you want the default policy to keep all hot-plugged 5028604d9e5SVitaly Kuznetsov memory blocks in 'offline' state. 5038604d9e5SVitaly Kuznetsov 5040c0e6195SKAMEZAWA Hiroyukiconfig MEMORY_HOTREMOVE 5050c0e6195SKAMEZAWA Hiroyuki bool "Allow for memory hot remove" 506f7e3334aSNathan Fontenot select HAVE_BOOTMEM_INFO_NODE if (X86_64 || PPC64) 5070c0e6195SKAMEZAWA Hiroyuki depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE 5080c0e6195SKAMEZAWA Hiroyuki depends on MIGRATION 5090c0e6195SKAMEZAWA Hiroyuki 510a08a2ae3SOscar Salvadorconfig MHP_MEMMAP_ON_MEMORY 511a08a2ae3SOscar Salvador def_bool y 512a08a2ae3SOscar Salvador depends on MEMORY_HOTPLUG && SPARSEMEM_VMEMMAP 513a08a2ae3SOscar Salvador depends on ARCH_MHP_MEMMAP_ON_MEMORY_ENABLE 514a08a2ae3SOscar Salvador 515519bcb79SJohannes Weinerendif # MEMORY_HOTPLUG 516519bcb79SJohannes Weiner 5174c21e2f2SHugh Dickins# Heavily threaded applications may benefit from splitting the mm-wide 5184c21e2f2SHugh Dickins# page_table_lock, so that faults on different parts of the user address 5194c21e2f2SHugh Dickins# space can be handled with less contention: split it at this NR_CPUS. 5204c21e2f2SHugh Dickins# Default to 4 for wider testing, though 8 might be more appropriate. 5214c21e2f2SHugh Dickins# ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock. 5227b6ac9dfSHugh Dickins# PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes. 52360bccaa6SWill Deacon# SPARC32 allocates multiple pte tables within a single page, and therefore 52460bccaa6SWill Deacon# a per-page lock leads to problems when multiple tables need to be locked 52560bccaa6SWill Deacon# at the same time (e.g. copy_page_range()). 526a70caa8bSHugh Dickins# DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page. 5274c21e2f2SHugh Dickins# 5284c21e2f2SHugh Dickinsconfig SPLIT_PTLOCK_CPUS 5294c21e2f2SHugh Dickins int 5309164550eSKirill A. Shutemov default "999999" if !MMU 531a70caa8bSHugh Dickins default "999999" if ARM && !CPU_CACHE_VIPT 532a70caa8bSHugh Dickins default "999999" if PARISC && !PA20 53360bccaa6SWill Deacon default "999999" if SPARC32 5344c21e2f2SHugh Dickins default "4" 5357cbe34cfSChristoph Lameter 536e009bb30SKirill A. Shutemovconfig ARCH_ENABLE_SPLIT_PMD_PTLOCK 5376341e62bSChristoph Jaeger bool 538e009bb30SKirill A. Shutemov 5397cbe34cfSChristoph Lameter# 54009316c09SKonstantin Khlebnikov# support for memory balloon 54109316c09SKonstantin Khlebnikovconfig MEMORY_BALLOON 5426341e62bSChristoph Jaeger bool 54309316c09SKonstantin Khlebnikov 54409316c09SKonstantin Khlebnikov# 54518468d93SRafael Aquini# support for memory balloon compaction 54618468d93SRafael Aquiniconfig BALLOON_COMPACTION 54718468d93SRafael Aquini bool "Allow for balloon memory compaction/migration" 54818468d93SRafael Aquini def_bool y 54909316c09SKonstantin Khlebnikov depends on COMPACTION && MEMORY_BALLOON 55018468d93SRafael Aquini help 55118468d93SRafael Aquini Memory fragmentation introduced by ballooning might reduce 55218468d93SRafael Aquini significantly the number of 2MB contiguous memory blocks that can be 55318468d93SRafael Aquini used within a guest, thus imposing performance penalties associated 55418468d93SRafael Aquini with the reduced number of transparent huge pages that could be used 55518468d93SRafael Aquini by the guest workload. Allowing the compaction & migration for memory 55618468d93SRafael Aquini pages enlisted as being part of memory balloon devices avoids the 55718468d93SRafael Aquini scenario aforementioned and helps improving memory defragmentation. 55818468d93SRafael Aquini 55918468d93SRafael Aquini# 560e9e96b39SMel Gorman# support for memory compaction 561e9e96b39SMel Gormanconfig COMPACTION 562e9e96b39SMel Gorman bool "Allow for memory compaction" 56305106e6aSRik van Riel def_bool y 564e9e96b39SMel Gorman select MIGRATION 56533a93877SAndrea Arcangeli depends on MMU 566e9e96b39SMel Gorman help 567b32eaf71SMichal Hocko Compaction is the only memory management component to form 568b32eaf71SMichal Hocko high order (larger physically contiguous) memory blocks 569b32eaf71SMichal Hocko reliably. The page allocator relies on compaction heavily and 570b32eaf71SMichal Hocko the lack of the feature can lead to unexpected OOM killer 571b32eaf71SMichal Hocko invocations for high order memory requests. You shouldn't 572b32eaf71SMichal Hocko disable this option unless there really is a strong reason for 573b32eaf71SMichal Hocko it and then we would be really interested to hear about that at 574b32eaf71SMichal Hocko linux-mm@kvack.org. 575e9e96b39SMel Gorman 576c7e0b3d0SThomas Gleixnerconfig COMPACT_UNEVICTABLE_DEFAULT 577c7e0b3d0SThomas Gleixner int 578c7e0b3d0SThomas Gleixner depends on COMPACTION 579c7e0b3d0SThomas Gleixner default 0 if PREEMPT_RT 580c7e0b3d0SThomas Gleixner default 1 581c7e0b3d0SThomas Gleixner 582e9e96b39SMel Gorman# 58336e66c55SAlexander Duyck# support for free page reporting 58436e66c55SAlexander Duyckconfig PAGE_REPORTING 58536e66c55SAlexander Duyck bool "Free page reporting" 58636e66c55SAlexander Duyck def_bool n 58736e66c55SAlexander Duyck help 58836e66c55SAlexander Duyck Free page reporting allows for the incremental acquisition of 58936e66c55SAlexander Duyck free pages from the buddy allocator for the purpose of reporting 59036e66c55SAlexander Duyck those pages to another entity, such as a hypervisor, so that the 59136e66c55SAlexander Duyck memory can be freed within the host for other uses. 59236e66c55SAlexander Duyck 59336e66c55SAlexander Duyck# 5947cbe34cfSChristoph Lameter# support for page migration 5957cbe34cfSChristoph Lameter# 5967cbe34cfSChristoph Lameterconfig MIGRATION 597b20a3503SChristoph Lameter bool "Page migration" 5986c5240aeSChristoph Lameter def_bool y 599de32a817SChen Gang depends on (NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA) && MMU 600b20a3503SChristoph Lameter help 601b20a3503SChristoph Lameter Allows the migration of the physical location of pages of processes 602e9e96b39SMel Gorman while the virtual addresses are not changed. This is useful in 603e9e96b39SMel Gorman two situations. The first is on NUMA systems to put pages nearer 604e9e96b39SMel Gorman to the processors accessing. The second is when allocating huge 605e9e96b39SMel Gorman pages as migration can relocate pages to satisfy a huge page 606e9e96b39SMel Gorman allocation instead of reclaiming. 6076550e07fSGreg Kroah-Hartman 60876cbbeadSChristoph Hellwigconfig DEVICE_MIGRATION 609d90a25f8SChristoph Hellwig def_bool MIGRATION && ZONE_DEVICE 61076cbbeadSChristoph Hellwig 611c177c81eSNaoya Horiguchiconfig ARCH_ENABLE_HUGEPAGE_MIGRATION 6126341e62bSChristoph Jaeger bool 613c177c81eSNaoya Horiguchi 6149c670ea3SNaoya Horiguchiconfig ARCH_ENABLE_THP_MIGRATION 6159c670ea3SNaoya Horiguchi bool 6169c670ea3SNaoya Horiguchi 6174bfb68a0SAnshuman Khandualconfig HUGETLB_PAGE_SIZE_VARIABLE 6184bfb68a0SAnshuman Khandual def_bool n 6194bfb68a0SAnshuman Khandual help 6204bfb68a0SAnshuman Khandual Allows the pageblock_order value to be dynamic instead of just standard 6214bfb68a0SAnshuman Khandual HUGETLB_PAGE_ORDER when there are multiple HugeTLB page sizes available 6224bfb68a0SAnshuman Khandual on a platform. 6234bfb68a0SAnshuman Khandual 624b3d40a2bSDavid Hildenbrand Note that the pageblock_order cannot exceed MAX_ORDER - 1 and will be 625b3d40a2bSDavid Hildenbrand clamped down to MAX_ORDER - 1. 626b3d40a2bSDavid Hildenbrand 6278df995f6SAlexandre Ghiticonfig CONTIG_ALLOC 6288df995f6SAlexandre Ghiti def_bool (MEMORY_ISOLATION && COMPACTION) || CMA 6298df995f6SAlexandre Ghiti 630600715dcSJeremy Fitzhardingeconfig PHYS_ADDR_T_64BIT 631d4a451d5SChristoph Hellwig def_bool 64BIT 632600715dcSJeremy Fitzhardinge 6332a7326b5SChristoph Lameterconfig BOUNCE 6349ca24e2eSVinayak Menon bool "Enable bounce buffers" 6359ca24e2eSVinayak Menon default y 636ce288e05SChristoph Hellwig depends on BLOCK && MMU && HIGHMEM 6379ca24e2eSVinayak Menon help 638ce288e05SChristoph Hellwig Enable bounce buffers for devices that cannot access the full range of 639ce288e05SChristoph Hellwig memory available to the CPU. Enabled by default when HIGHMEM is 640ce288e05SChristoph Hellwig selected, but you may say n to override this. 6412a7326b5SChristoph Lameter 642cddb8a5cSAndrea Arcangeliconfig MMU_NOTIFIER 643cddb8a5cSAndrea Arcangeli bool 64483fe27eaSPranith Kumar select SRCU 64599cb252fSJason Gunthorpe select INTERVAL_TREE 646fc4d5c29SDavid Howells 647f8af4da3SHugh Dickinsconfig KSM 648f8af4da3SHugh Dickins bool "Enable KSM for page merging" 649f8af4da3SHugh Dickins depends on MMU 65059e1a2f4STimofey Titovets select XXHASH 651f8af4da3SHugh Dickins help 652f8af4da3SHugh Dickins Enable Kernel Samepage Merging: KSM periodically scans those areas 653f8af4da3SHugh Dickins of an application's address space that an app has advised may be 654f8af4da3SHugh Dickins mergeable. When it finds pages of identical content, it replaces 655d0f209f6SHugh Dickins the many instances by a single page with that content, so 656f8af4da3SHugh Dickins saving memory until one or another app needs to modify the content. 657f8af4da3SHugh Dickins Recommended for use with KVM, or with other duplicative applications. 658ee65728eSMike Rapoport See Documentation/mm/ksm.rst for more information: KSM is inactive 659c73602adSHugh Dickins until a program has madvised that an area is MADV_MERGEABLE, and 660c73602adSHugh Dickins root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set). 661f8af4da3SHugh Dickins 662e0a94c2aSChristoph Lameterconfig DEFAULT_MMAP_MIN_ADDR 663e0a94c2aSChristoph Lameter int "Low address space to protect from user allocation" 6646e141546SDavid Howells depends on MMU 665e0a94c2aSChristoph Lameter default 4096 666e0a94c2aSChristoph Lameter help 667e0a94c2aSChristoph Lameter This is the portion of low virtual memory which should be protected 668e0a94c2aSChristoph Lameter from userspace allocation. Keeping a user from writing to low pages 669e0a94c2aSChristoph Lameter can help reduce the impact of kernel NULL pointer bugs. 670e0a94c2aSChristoph Lameter 671e0a94c2aSChristoph Lameter For most ia64, ppc64 and x86 users with lots of address space 672e0a94c2aSChristoph Lameter a value of 65536 is reasonable and should cause no problems. 673e0a94c2aSChristoph Lameter On arm and other archs it should not be higher than 32768. 674788084abSEric Paris Programs which use vm86 functionality or have some need to map 675788084abSEric Paris this low address space will need CAP_SYS_RAWIO or disable this 676788084abSEric Paris protection by setting the value to 0. 677e0a94c2aSChristoph Lameter 678e0a94c2aSChristoph Lameter This value can be changed after boot using the 679e0a94c2aSChristoph Lameter /proc/sys/vm/mmap_min_addr tunable. 680e0a94c2aSChristoph Lameter 681d949f36fSLinus Torvaldsconfig ARCH_SUPPORTS_MEMORY_FAILURE 682d949f36fSLinus Torvalds bool 683e0a94c2aSChristoph Lameter 6846a46079cSAndi Kleenconfig MEMORY_FAILURE 6856a46079cSAndi Kleen depends on MMU 686d949f36fSLinus Torvalds depends on ARCH_SUPPORTS_MEMORY_FAILURE 6876a46079cSAndi Kleen bool "Enable recovery from hardware memory errors" 688ee6f509cSMinchan Kim select MEMORY_ISOLATION 68997f0b134SXie XiuQi select RAS 6906a46079cSAndi Kleen help 6916a46079cSAndi Kleen Enables code to recover from some memory failures on systems 6926a46079cSAndi Kleen with MCA recovery. This allows a system to continue running 6936a46079cSAndi Kleen even when some of its memory has uncorrected errors. This requires 6946a46079cSAndi Kleen special hardware support and typically ECC memory. 6956a46079cSAndi Kleen 696cae681fcSAndi Kleenconfig HWPOISON_INJECT 697413f9efbSAndi Kleen tristate "HWPoison pages injector" 69827df5068SAndi Kleen depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS 699478c5ffcSWu Fengguang select PROC_PAGE_MONITOR 700cae681fcSAndi Kleen 701fc4d5c29SDavid Howellsconfig NOMMU_INITIAL_TRIM_EXCESS 702fc4d5c29SDavid Howells int "Turn on mmap() excess space trimming before booting" 703fc4d5c29SDavid Howells depends on !MMU 704fc4d5c29SDavid Howells default 1 705fc4d5c29SDavid Howells help 706fc4d5c29SDavid Howells The NOMMU mmap() frequently needs to allocate large contiguous chunks 707fc4d5c29SDavid Howells of memory on which to store mappings, but it can only ask the system 708fc4d5c29SDavid Howells allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently 709fc4d5c29SDavid Howells more than it requires. To deal with this, mmap() is able to trim off 710fc4d5c29SDavid Howells the excess and return it to the allocator. 711fc4d5c29SDavid Howells 712fc4d5c29SDavid Howells If trimming is enabled, the excess is trimmed off and returned to the 713fc4d5c29SDavid Howells system allocator, which can cause extra fragmentation, particularly 714fc4d5c29SDavid Howells if there are a lot of transient processes. 715fc4d5c29SDavid Howells 716fc4d5c29SDavid Howells If trimming is disabled, the excess is kept, but not used, which for 717fc4d5c29SDavid Howells long-term mappings means that the space is wasted. 718fc4d5c29SDavid Howells 719fc4d5c29SDavid Howells Trimming can be dynamically controlled through a sysctl option 720fc4d5c29SDavid Howells (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of 721fc4d5c29SDavid Howells excess pages there must be before trimming should occur, or zero if 722fc4d5c29SDavid Howells no trimming is to occur. 723fc4d5c29SDavid Howells 724fc4d5c29SDavid Howells This option specifies the initial value of this option. The default 725fc4d5c29SDavid Howells of 1 says that all excess pages should be trimmed. 726fc4d5c29SDavid Howells 727dd19d293SStephen Kitt See Documentation/admin-guide/mm/nommu-mmap.rst for more information. 728bbddff05STejun Heo 729519bcb79SJohannes Weinerconfig ARCH_WANT_GENERAL_HUGETLB 730519bcb79SJohannes Weiner bool 731519bcb79SJohannes Weiner 732519bcb79SJohannes Weinerconfig ARCH_WANTS_THP_SWAP 733519bcb79SJohannes Weiner def_bool n 734519bcb79SJohannes Weiner 735519bcb79SJohannes Weinermenuconfig TRANSPARENT_HUGEPAGE 73613ece886SAndrea Arcangeli bool "Transparent Hugepage Support" 737554b0f3cSSebastian Andrzej Siewior depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT 7385d689240SAndrea Arcangeli select COMPACTION 7393a08cd52SMatthew Wilcox select XARRAY_MULTI 7404c76d9d1SAndrea Arcangeli help 7414c76d9d1SAndrea Arcangeli Transparent Hugepages allows the kernel to use huge pages and 7424c76d9d1SAndrea Arcangeli huge tlb transparently to the applications whenever possible. 7434c76d9d1SAndrea Arcangeli This feature can improve computing performance to certain 7444c76d9d1SAndrea Arcangeli applications by speeding up page faults during memory 7454c76d9d1SAndrea Arcangeli allocation, by reducing the number of tlb misses and by speeding 7464c76d9d1SAndrea Arcangeli up the pagetable walking. 7474c76d9d1SAndrea Arcangeli 7484c76d9d1SAndrea Arcangeli If memory constrained on embedded, you may want to say N. 7494c76d9d1SAndrea Arcangeli 750519bcb79SJohannes Weinerif TRANSPARENT_HUGEPAGE 751519bcb79SJohannes Weiner 75213ece886SAndrea Arcangelichoice 75313ece886SAndrea Arcangeli prompt "Transparent Hugepage Support sysfs defaults" 75413ece886SAndrea Arcangeli depends on TRANSPARENT_HUGEPAGE 75513ece886SAndrea Arcangeli default TRANSPARENT_HUGEPAGE_ALWAYS 75613ece886SAndrea Arcangeli help 75713ece886SAndrea Arcangeli Selects the sysfs defaults for Transparent Hugepage Support. 75813ece886SAndrea Arcangeli 75913ece886SAndrea Arcangeli config TRANSPARENT_HUGEPAGE_ALWAYS 76013ece886SAndrea Arcangeli bool "always" 76113ece886SAndrea Arcangeli help 76213ece886SAndrea Arcangeli Enabling Transparent Hugepage always, can increase the 76313ece886SAndrea Arcangeli memory footprint of applications without a guaranteed 76413ece886SAndrea Arcangeli benefit but it will work automatically for all applications. 76513ece886SAndrea Arcangeli 76613ece886SAndrea Arcangeli config TRANSPARENT_HUGEPAGE_MADVISE 76713ece886SAndrea Arcangeli bool "madvise" 76813ece886SAndrea Arcangeli help 76913ece886SAndrea Arcangeli Enabling Transparent Hugepage madvise, will only provide a 77013ece886SAndrea Arcangeli performance improvement benefit to the applications using 77113ece886SAndrea Arcangeli madvise(MADV_HUGEPAGE) but it won't risk to increase the 77213ece886SAndrea Arcangeli memory footprint of applications without a guaranteed 77313ece886SAndrea Arcangeli benefit. 77413ece886SAndrea Arcangeliendchoice 77513ece886SAndrea Arcangeli 77638d8b4e6SHuang Yingconfig THP_SWAP 77738d8b4e6SHuang Ying def_bool y 778*dad6a5ebSHugh Dickins depends on TRANSPARENT_HUGEPAGE && ARCH_WANTS_THP_SWAP && SWAP && 64BIT 77938d8b4e6SHuang Ying help 78038d8b4e6SHuang Ying Swap transparent huge pages in one piece, without splitting. 78114fef284SHuang Ying XXX: For now, swap cluster backing transparent huge page 78214fef284SHuang Ying will be split after swapout. 78338d8b4e6SHuang Ying 78438d8b4e6SHuang Ying For selection by architectures with reasonable THP sizes. 78538d8b4e6SHuang Ying 786519bcb79SJohannes Weinerconfig READ_ONLY_THP_FOR_FS 787519bcb79SJohannes Weiner bool "Read-only THP for filesystems (EXPERIMENTAL)" 788519bcb79SJohannes Weiner depends on TRANSPARENT_HUGEPAGE && SHMEM 789519bcb79SJohannes Weiner 790519bcb79SJohannes Weiner help 791519bcb79SJohannes Weiner Allow khugepaged to put read-only file-backed pages in THP. 792519bcb79SJohannes Weiner 793519bcb79SJohannes Weiner This is marked experimental because it is a new feature. Write 794519bcb79SJohannes Weiner support of file THPs will be developed in the next few release 795519bcb79SJohannes Weiner cycles. 796519bcb79SJohannes Weiner 797519bcb79SJohannes Weinerendif # TRANSPARENT_HUGEPAGE 798519bcb79SJohannes Weiner 799e496cf3dSKirill A. Shutemov# 800bbddff05STejun Heo# UP and nommu archs use km based percpu allocator 801bbddff05STejun Heo# 802bbddff05STejun Heoconfig NEED_PER_CPU_KM 8033583521aSVladimir Murzin depends on !SMP || !MMU 804bbddff05STejun Heo bool 805bbddff05STejun Heo default y 806077b1f83SDan Magenheimer 8077ecd19cfSKefeng Wangconfig NEED_PER_CPU_EMBED_FIRST_CHUNK 8087ecd19cfSKefeng Wang bool 8097ecd19cfSKefeng Wang 8107ecd19cfSKefeng Wangconfig NEED_PER_CPU_PAGE_FIRST_CHUNK 8117ecd19cfSKefeng Wang bool 8127ecd19cfSKefeng Wang 8137ecd19cfSKefeng Wangconfig USE_PERCPU_NUMA_NODE_ID 8147ecd19cfSKefeng Wang bool 8157ecd19cfSKefeng Wang 8167ecd19cfSKefeng Wangconfig HAVE_SETUP_PER_CPU_AREA 8177ecd19cfSKefeng Wang bool 8187ecd19cfSKefeng Wang 81927c6aec2SDan Magenheimerconfig FRONTSWAP 8206e61dde8SChristoph Hellwig bool 821f825c736SAneesh Kumar K.V 822f825c736SAneesh Kumar K.Vconfig CMA 823f825c736SAneesh Kumar K.V bool "Contiguous Memory Allocator" 824aca52c39SMike Rapoport depends on MMU 825f825c736SAneesh Kumar K.V select MIGRATION 826f825c736SAneesh Kumar K.V select MEMORY_ISOLATION 827f825c736SAneesh Kumar K.V help 828f825c736SAneesh Kumar K.V This enables the Contiguous Memory Allocator which allows other 829f825c736SAneesh Kumar K.V subsystems to allocate big physically-contiguous blocks of memory. 830f825c736SAneesh Kumar K.V CMA reserves a region of memory and allows only movable pages to 831f825c736SAneesh Kumar K.V be allocated from it. This way, the kernel can use the memory for 832f825c736SAneesh Kumar K.V pagecache and when a subsystem requests for contiguous area, the 833f825c736SAneesh Kumar K.V allocated pages are migrated away to serve the contiguous request. 834f825c736SAneesh Kumar K.V 835f825c736SAneesh Kumar K.V If unsure, say "n". 836f825c736SAneesh Kumar K.V 837f825c736SAneesh Kumar K.Vconfig CMA_DEBUG 838f825c736SAneesh Kumar K.V bool "CMA debug messages (DEVELOPMENT)" 839f825c736SAneesh Kumar K.V depends on DEBUG_KERNEL && CMA 840f825c736SAneesh Kumar K.V help 841f825c736SAneesh Kumar K.V Turns on debug messages in CMA. This produces KERN_DEBUG 842f825c736SAneesh Kumar K.V messages for every CMA call as well as various messages while 843f825c736SAneesh Kumar K.V processing calls such as dma_alloc_from_contiguous(). 844f825c736SAneesh Kumar K.V This option does not affect warning and error messages. 845bf550fc9SAlexander Graf 84628b24c1fSSasha Levinconfig CMA_DEBUGFS 84728b24c1fSSasha Levin bool "CMA debugfs interface" 84828b24c1fSSasha Levin depends on CMA && DEBUG_FS 84928b24c1fSSasha Levin help 85028b24c1fSSasha Levin Turns on the DebugFS interface for CMA. 85128b24c1fSSasha Levin 85243ca106fSMinchan Kimconfig CMA_SYSFS 85343ca106fSMinchan Kim bool "CMA information through sysfs interface" 85443ca106fSMinchan Kim depends on CMA && SYSFS 85543ca106fSMinchan Kim help 85643ca106fSMinchan Kim This option exposes some sysfs attributes to get information 85743ca106fSMinchan Kim from CMA. 85843ca106fSMinchan Kim 859a254129eSJoonsoo Kimconfig CMA_AREAS 860a254129eSJoonsoo Kim int "Maximum count of the CMA areas" 861a254129eSJoonsoo Kim depends on CMA 862b7176c26SBarry Song default 19 if NUMA 863a254129eSJoonsoo Kim default 7 864a254129eSJoonsoo Kim help 865a254129eSJoonsoo Kim CMA allows to create CMA areas for particular purpose, mainly, 866a254129eSJoonsoo Kim used as device private area. This parameter sets the maximum 867a254129eSJoonsoo Kim number of CMA area in the system. 868a254129eSJoonsoo Kim 869b7176c26SBarry Song If unsure, leave the default value "7" in UMA and "19" in NUMA. 870a254129eSJoonsoo Kim 871af8d417aSDan Streetmanconfig MEM_SOFT_DIRTY 872af8d417aSDan Streetman bool "Track memory changes" 873af8d417aSDan Streetman depends on CHECKPOINT_RESTORE && HAVE_ARCH_SOFT_DIRTY && PROC_FS 874af8d417aSDan Streetman select PROC_PAGE_MONITOR 8754e2e2770SSeth Jennings help 876af8d417aSDan Streetman This option enables memory changes tracking by introducing a 877af8d417aSDan Streetman soft-dirty bit on pte-s. This bit it set when someone writes 878af8d417aSDan Streetman into a page just as regular dirty bit, but unlike the latter 879af8d417aSDan Streetman it can be cleared by hands. 880af8d417aSDan Streetman 8811ad1335dSMike Rapoport See Documentation/admin-guide/mm/soft-dirty.rst for more details. 8824e2e2770SSeth Jennings 8839e5c33d7SMark Salterconfig GENERIC_EARLY_IOREMAP 8849e5c33d7SMark Salter bool 885042d27acSHelge Deller 88622ee3ea5SHelge Dellerconfig STACK_MAX_DEFAULT_SIZE_MB 88722ee3ea5SHelge Deller int "Default maximum user stack size for 32-bit processes (MB)" 88822ee3ea5SHelge Deller default 100 889042d27acSHelge Deller range 8 2048 890042d27acSHelge Deller depends on STACK_GROWSUP && (!64BIT || COMPAT) 891042d27acSHelge Deller help 892042d27acSHelge Deller This is the maximum stack size in Megabytes in the VM layout of 32-bit 893042d27acSHelge Deller user processes when the stack grows upwards (currently only on parisc 89422ee3ea5SHelge Deller arch) when the RLIMIT_STACK hard limit is unlimited. 895042d27acSHelge Deller 89622ee3ea5SHelge Deller A sane initial value is 100 MB. 8973a80a7faSMel Gorman 8983a80a7faSMel Gormanconfig DEFERRED_STRUCT_PAGE_INIT 8991ce22103SVlastimil Babka bool "Defer initialisation of struct pages to kthreads" 900d39f8fb4SMike Rapoport depends on SPARSEMEM 901ab1e8d89SPavel Tatashin depends on !NEED_PER_CPU_KM 902889c695dSPasha Tatashin depends on 64BIT 903e4443149SDaniel Jordan select PADATA 9043a80a7faSMel Gorman help 9053a80a7faSMel Gorman Ordinarily all struct pages are initialised during early boot in a 9063a80a7faSMel Gorman single thread. On very large machines this can take a considerable 9073a80a7faSMel Gorman amount of time. If this option is set, large machines will bring up 908e4443149SDaniel Jordan a subset of memmap at boot and then initialise the rest in parallel. 909e4443149SDaniel Jordan This has a potential performance impact on tasks running early in the 9101ce22103SVlastimil Babka lifetime of the system until these kthreads finish the 9111ce22103SVlastimil Babka initialisation. 912033fbae9SDan Williams 9131c676e0dSSeongJae Parkconfig PAGE_IDLE_FLAG 9141c676e0dSSeongJae Park bool 9151c676e0dSSeongJae Park select PAGE_EXTENSION if !64BIT 9161c676e0dSSeongJae Park help 9171c676e0dSSeongJae Park This adds PG_idle and PG_young flags to 'struct page'. PTE Accessed 9181c676e0dSSeongJae Park bit writers can set the state of the bit in the flags so that PTE 9191c676e0dSSeongJae Park Accessed bit readers may avoid disturbance. 9201c676e0dSSeongJae Park 92133c3fc71SVladimir Davydovconfig IDLE_PAGE_TRACKING 92233c3fc71SVladimir Davydov bool "Enable idle page tracking" 92333c3fc71SVladimir Davydov depends on SYSFS && MMU 9241c676e0dSSeongJae Park select PAGE_IDLE_FLAG 92533c3fc71SVladimir Davydov help 92633c3fc71SVladimir Davydov This feature allows to estimate the amount of user pages that have 92733c3fc71SVladimir Davydov not been touched during a given period of time. This information can 92833c3fc71SVladimir Davydov be useful to tune memory cgroup limits and/or for job placement 92933c3fc71SVladimir Davydov within a compute cluster. 93033c3fc71SVladimir Davydov 9311ad1335dSMike Rapoport See Documentation/admin-guide/mm/idle_page_tracking.rst for 9321ad1335dSMike Rapoport more details. 93333c3fc71SVladimir Davydov 934c2280be8SAnshuman Khandualconfig ARCH_HAS_CACHE_LINE_SIZE 935c2280be8SAnshuman Khandual bool 936c2280be8SAnshuman Khandual 9372792d84eSKees Cookconfig ARCH_HAS_CURRENT_STACK_POINTER 9382792d84eSKees Cook bool 9392792d84eSKees Cook help 9402792d84eSKees Cook In support of HARDENED_USERCOPY performing stack variable lifetime 9412792d84eSKees Cook checking, an architecture-agnostic way to find the stack pointer 9422792d84eSKees Cook is needed. Once an architecture defines an unsigned long global 9432792d84eSKees Cook register alias named "current_stack_pointer", this config can be 9442792d84eSKees Cook selected. 9452792d84eSKees Cook 94617596731SRobin Murphyconfig ARCH_HAS_PTE_DEVMAP 94765f7d049SOliver O'Halloran bool 94865f7d049SOliver O'Halloran 94963703f37SKefeng Wangconfig ARCH_HAS_ZONE_DMA_SET 95063703f37SKefeng Wang bool 95163703f37SKefeng Wang 95263703f37SKefeng Wangconfig ZONE_DMA 95363703f37SKefeng Wang bool "Support DMA zone" if ARCH_HAS_ZONE_DMA_SET 95463703f37SKefeng Wang default y if ARM64 || X86 95563703f37SKefeng Wang 95663703f37SKefeng Wangconfig ZONE_DMA32 95763703f37SKefeng Wang bool "Support DMA32 zone" if ARCH_HAS_ZONE_DMA_SET 95863703f37SKefeng Wang depends on !X86_32 95963703f37SKefeng Wang default y if ARM64 96063703f37SKefeng Wang 961033fbae9SDan Williamsconfig ZONE_DEVICE 9625042db43SJérôme Glisse bool "Device memory (pmem, HMM, etc...) hotplug support" 963033fbae9SDan Williams depends on MEMORY_HOTPLUG 964033fbae9SDan Williams depends on MEMORY_HOTREMOVE 96599490f16SDan Williams depends on SPARSEMEM_VMEMMAP 96617596731SRobin Murphy depends on ARCH_HAS_PTE_DEVMAP 9673a08cd52SMatthew Wilcox select XARRAY_MULTI 968033fbae9SDan Williams 969033fbae9SDan Williams help 970033fbae9SDan Williams Device memory hotplug support allows for establishing pmem, 971033fbae9SDan Williams or other device driver discovered memory regions, in the 972033fbae9SDan Williams memmap. This allows pfn_to_page() lookups of otherwise 973033fbae9SDan Williams "device-physical" addresses which is needed for using a DAX 974033fbae9SDan Williams mapping in an O_DIRECT operation, among other things. 975033fbae9SDan Williams 976033fbae9SDan Williams If FS_DAX is enabled, then say Y. 97706a660adSLinus Torvalds 9789c240a7bSChristoph Hellwig# 9799c240a7bSChristoph Hellwig# Helpers to mirror range of the CPU page tables of a process into device page 9809c240a7bSChristoph Hellwig# tables. 9819c240a7bSChristoph Hellwig# 982c0b12405SJérôme Glisseconfig HMM_MIRROR 9839c240a7bSChristoph Hellwig bool 984f442c283SChristoph Hellwig depends on MMU 985c0b12405SJérôme Glisse 98614b80582SDan Williamsconfig GET_FREE_REGION 98714b80582SDan Williams depends on SPARSEMEM 98814b80582SDan Williams bool 98914b80582SDan Williams 9905042db43SJérôme Glisseconfig DEVICE_PRIVATE 9915042db43SJérôme Glisse bool "Unaddressable device memory (GPU memory, ...)" 9927328d9ccSChristoph Hellwig depends on ZONE_DEVICE 99314b80582SDan Williams select GET_FREE_REGION 9945042db43SJérôme Glisse 9955042db43SJérôme Glisse help 9965042db43SJérôme Glisse Allows creation of struct pages to represent unaddressable device 9975042db43SJérôme Glisse memory; i.e., memory that is only accessible from the device (or 9985042db43SJérôme Glisse group of devices). You likely also want to select HMM_MIRROR. 9995042db43SJérôme Glisse 10003e9a9e25SChristoph Hellwigconfig VMAP_PFN 10013e9a9e25SChristoph Hellwig bool 10023e9a9e25SChristoph Hellwig 100363c17fb8SDave Hansenconfig ARCH_USES_HIGH_VMA_FLAGS 100463c17fb8SDave Hansen bool 100566d37570SDave Hansenconfig ARCH_HAS_PKEYS 100666d37570SDave Hansen bool 100730a5b536SDennis Zhou 10080710d012SVlastimil Babkaconfig VM_EVENT_COUNTERS 10090710d012SVlastimil Babka default y 10100710d012SVlastimil Babka bool "Enable VM event counters for /proc/vmstat" if EXPERT 10110710d012SVlastimil Babka help 10120710d012SVlastimil Babka VM event counters are needed for event counts to be shown. 10130710d012SVlastimil Babka This option allows the disabling of the VM event counters 10140710d012SVlastimil Babka on EXPERT systems. /proc/vmstat will only show page counts 10150710d012SVlastimil Babka if VM event counters are disabled. 10160710d012SVlastimil Babka 101730a5b536SDennis Zhouconfig PERCPU_STATS 101830a5b536SDennis Zhou bool "Collect percpu memory statistics" 101930a5b536SDennis Zhou help 102030a5b536SDennis Zhou This feature collects and exposes statistics via debugfs. The 102130a5b536SDennis Zhou information includes global and per chunk statistics, which can 102230a5b536SDennis Zhou be used to help understand percpu memory usage. 102364c349f4SKirill A. Shutemov 10249c84f229SJohn Hubbardconfig GUP_TEST 10259c84f229SJohn Hubbard bool "Enable infrastructure for get_user_pages()-related unit tests" 1026d0de8241SBarry Song depends on DEBUG_FS 102764c349f4SKirill A. Shutemov help 10289c84f229SJohn Hubbard Provides /sys/kernel/debug/gup_test, which in turn provides a way 10299c84f229SJohn Hubbard to make ioctl calls that can launch kernel-based unit tests for 10309c84f229SJohn Hubbard the get_user_pages*() and pin_user_pages*() family of API calls. 103164c349f4SKirill A. Shutemov 10329c84f229SJohn Hubbard These tests include benchmark testing of the _fast variants of 10339c84f229SJohn Hubbard get_user_pages*() and pin_user_pages*(), as well as smoke tests of 10349c84f229SJohn Hubbard the non-_fast variants. 10359c84f229SJohn Hubbard 1036f4f9bda4SJohn Hubbard There is also a sub-test that allows running dump_page() on any 1037f4f9bda4SJohn Hubbard of up to eight pages (selected by command line args) within the 1038f4f9bda4SJohn Hubbard range of user-space addresses. These pages are either pinned via 1039f4f9bda4SJohn Hubbard pin_user_pages*(), or pinned via get_user_pages*(), as specified 1040f4f9bda4SJohn Hubbard by other command line arguments. 1041f4f9bda4SJohn Hubbard 10429c84f229SJohn Hubbard See tools/testing/selftests/vm/gup_test.c 10433010a5eaSLaurent Dufour 1044d0de8241SBarry Songcomment "GUP_TEST needs to have DEBUG_FS enabled" 1045d0de8241SBarry Song depends on !GUP_TEST && !DEBUG_FS 10463010a5eaSLaurent Dufour 104739656e83SChristoph Hellwigconfig GUP_GET_PTE_LOW_HIGH 104839656e83SChristoph Hellwig bool 104939656e83SChristoph Hellwig 10503010a5eaSLaurent Dufourconfig ARCH_HAS_PTE_SPECIAL 10513010a5eaSLaurent Dufour bool 105259e0b520SChristoph Hellwig 1053cbd34da7SChristoph Hellwig# 1054cbd34da7SChristoph Hellwig# Some architectures require a special hugepage directory format that is 1055cbd34da7SChristoph Hellwig# required to support multiple hugepage sizes. For example a4fe3ce76 1056cbd34da7SChristoph Hellwig# "powerpc/mm: Allow more flexible layouts for hugepage pagetables" 1057cbd34da7SChristoph Hellwig# introduced it on powerpc. This allows for a more flexible hugepage 1058cbd34da7SChristoph Hellwig# pagetable layouts. 1059cbd34da7SChristoph Hellwig# 1060cbd34da7SChristoph Hellwigconfig ARCH_HAS_HUGEPD 1061cbd34da7SChristoph Hellwig bool 1062cbd34da7SChristoph Hellwig 1063c5acad84SThomas Hellstromconfig MAPPING_DIRTY_HELPERS 1064c5acad84SThomas Hellstrom bool 1065c5acad84SThomas Hellstrom 1066298fa1adSThomas Gleixnerconfig KMAP_LOCAL 1067298fa1adSThomas Gleixner bool 1068298fa1adSThomas Gleixner 1069825c43f5SArd Biesheuvelconfig KMAP_LOCAL_NON_LINEAR_PTE_ARRAY 1070825c43f5SArd Biesheuvel bool 1071825c43f5SArd Biesheuvel 10721fbaf8fcSChristoph Hellwig# struct io_mapping based helper. Selected by drivers that need them 10731fbaf8fcSChristoph Hellwigconfig IO_MAPPING 10741fbaf8fcSChristoph Hellwig bool 10751507f512SMike Rapoport 10761507f512SMike Rapoportconfig SECRETMEM 10771507f512SMike Rapoport def_bool ARCH_HAS_SET_DIRECT_MAP && !EMBEDDED 10781507f512SMike Rapoport 10799a10064fSColin Crossconfig ANON_VMA_NAME 10809a10064fSColin Cross bool "Anonymous VMA name support" 10819a10064fSColin Cross depends on PROC_FS && ADVISE_SYSCALLS && MMU 10829a10064fSColin Cross 10839a10064fSColin Cross help 10849a10064fSColin Cross Allow naming anonymous virtual memory areas. 10859a10064fSColin Cross 10869a10064fSColin Cross This feature allows assigning names to virtual memory areas. Assigned 10879a10064fSColin Cross names can be later retrieved from /proc/pid/maps and /proc/pid/smaps 10889a10064fSColin Cross and help identifying individual anonymous memory areas. 10899a10064fSColin Cross Assigning a name to anonymous virtual memory area might prevent that 10909a10064fSColin Cross area from being merged with adjacent virtual memory areas due to the 10919a10064fSColin Cross difference in their name. 10929a10064fSColin Cross 1093430529b5SPeter Xuconfig USERFAULTFD 1094430529b5SPeter Xu bool "Enable userfaultfd() system call" 1095430529b5SPeter Xu depends on MMU 1096430529b5SPeter Xu help 1097430529b5SPeter Xu Enable the userfaultfd() system call that allows to intercept and 1098430529b5SPeter Xu handle page faults in userland. 1099430529b5SPeter Xu 1100430529b5SPeter Xuconfig HAVE_ARCH_USERFAULTFD_WP 1101430529b5SPeter Xu bool 1102430529b5SPeter Xu help 1103430529b5SPeter Xu Arch has userfaultfd write protection support 1104430529b5SPeter Xu 1105430529b5SPeter Xuconfig HAVE_ARCH_USERFAULTFD_MINOR 1106430529b5SPeter Xu bool 1107430529b5SPeter Xu help 1108430529b5SPeter Xu Arch has userfaultfd minor fault support 1109430529b5SPeter Xu 11101db9dbc2SPeter Xuconfig PTE_MARKER_UFFD_WP 111181e0f15fSPeter Xu bool "Userfaultfd write protection support for shmem/hugetlbfs" 111281e0f15fSPeter Xu default y 111381e0f15fSPeter Xu depends on HAVE_ARCH_USERFAULTFD_WP 11141db9dbc2SPeter Xu 11151db9dbc2SPeter Xu help 11161db9dbc2SPeter Xu Allows to create marker PTEs for userfaultfd write protection 11171db9dbc2SPeter Xu purposes. It is required to enable userfaultfd write protection on 11181db9dbc2SPeter Xu file-backed memory types like shmem and hugetlbfs. 11191db9dbc2SPeter Xu 1120ac35a490SYu Zhao# multi-gen LRU { 1121ec1c86b2SYu Zhaoconfig LRU_GEN 1122ec1c86b2SYu Zhao bool "Multi-Gen LRU" 1123ec1c86b2SYu Zhao depends on MMU 1124ec1c86b2SYu Zhao # make sure folio->flags has enough spare bits 1125ec1c86b2SYu Zhao depends on 64BIT || !SPARSEMEM || SPARSEMEM_VMEMMAP 1126ec1c86b2SYu Zhao help 112707017acbSYu Zhao A high performance LRU implementation to overcommit memory. See 112807017acbSYu Zhao Documentation/admin-guide/mm/multigen_lru.rst for details. 1129ec1c86b2SYu Zhao 1130354ed597SYu Zhaoconfig LRU_GEN_ENABLED 1131354ed597SYu Zhao bool "Enable by default" 1132354ed597SYu Zhao depends on LRU_GEN 1133354ed597SYu Zhao help 1134354ed597SYu Zhao This option enables the multi-gen LRU by default. 1135354ed597SYu Zhao 1136ac35a490SYu Zhaoconfig LRU_GEN_STATS 1137ac35a490SYu Zhao bool "Full stats for debugging" 1138ac35a490SYu Zhao depends on LRU_GEN 1139ac35a490SYu Zhao help 1140ac35a490SYu Zhao Do not enable this option unless you plan to look at historical stats 1141ac35a490SYu Zhao from evicted generations for debugging purpose. 1142ac35a490SYu Zhao 1143ac35a490SYu Zhao This option has a per-memcg and per-node memory overhead. 1144ac35a490SYu Zhao# } 1145ac35a490SYu Zhao 11462224d848SSeongJae Parksource "mm/damon/Kconfig" 11472224d848SSeongJae Park 114859e0b520SChristoph Hellwigendmenu 1149