€•VŒsphinx.addnodes”Œdocument”“”)”}”(Œ rawsource”Œ”Œchildren”]”(Œtranslations”Œ LanguagesNode”“”)”}”(hhh]”(hŒpending_xref”“”)”}”(hhh]”Œdocutils.nodes”ŒText”“”ŒChinese (Simplified)”…””}”Œparent”hsbaŒ attributes”}”(Œids”]”Œclasses”]”Œnames”]”Œdupnames”]”Œbackrefs”]”Œ refdomain”Œstd”Œreftype”Œdoc”Œ reftarget”Œ&/translations/zh_CN/mm/unevictable-lru”Œmodname”NŒ classname”NŒrefexplicit”ˆuŒtagname”hhhubh)”}”(hhh]”hŒChinese (Traditional)”…””}”hh2sbah}”(h]”h ]”h"]”h$]”h&]”Œ refdomain”h)Œreftype”h+Œ reftarget”Œ&/translations/zh_TW/mm/unevictable-lru”Œmodname”NŒ classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒItalian”…””}”hhFsbah}”(h]”h ]”h"]”h$]”h&]”Œ refdomain”h)Œreftype”h+Œ reftarget”Œ&/translations/it_IT/mm/unevictable-lru”Œmodname”NŒ classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒJapanese”…””}”hhZsbah}”(h]”h ]”h"]”h$]”h&]”Œ refdomain”h)Œreftype”h+Œ reftarget”Œ&/translations/ja_JP/mm/unevictable-lru”Œmodname”NŒ classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒKorean”…””}”hhnsbah}”(h]”h ]”h"]”h$]”h&]”Œ refdomain”h)Œreftype”h+Œ reftarget”Œ&/translations/ko_KR/mm/unevictable-lru”Œmodname”NŒ classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒSpanish”…””}”hh‚sbah}”(h]”h ]”h"]”h$]”h&]”Œ refdomain”h)Œreftype”h+Œ reftarget”Œ&/translations/sp_SP/mm/unevictable-lru”Œmodname”NŒ classname”NŒrefexplicit”ˆuh1hhhubeh}”(h]”h ]”h"]”h$]”h&]”Œcurrent_language”ŒEnglish”uh1h hhŒ _document”hŒsource”NŒline”NubhŒsection”“”)”}”(hhh]”(hŒtitle”“”)”}”(hŒUnevictable LRU Infrastructure”h]”hŒUnevictable LRU Infrastructure”…””}”(hh¨hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hh£hžhhŸŒ@/var/lib/git/docbuild/linux/Documentation/mm/unevictable-lru.rst”h KubhŒtopic”“”)”}”(hhh]”hŒbullet_list”“”)”}”(hhh]”(hŒ list_item”“”)”}”(hhh]”hŒ paragraph”“”)”}”(hhh]”hŒ reference”“”)”}”(hhh]”hŒIntroduction”…””}”(hhÍhžhhŸNh Nubah}”(h]”Œid1”ah ]”h"]”h$]”h&]”Œrefid”Œintroduction”uh1hËhhÈubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhhÃubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhh¾ubhÂ)”}”(hhh]”(hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒThe Unevictable LRU”…””}”(hhïhžhhŸNh Nubah}”(h]”Œid2”ah ]”h"]”h$]”h&]”Œrefid”Œthe-unevictable-lru”uh1hËhhìubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhhéubh½)”}”(hhh]”(hÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒThe Unevictable LRU Folio List”…””}”(hjhžhhŸNh Nubah}”(h]”Œid3”ah ]”h"]”h$]”h&]”Œrefid”Œthe-unevictable-lru-folio-list”uh1hËhjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ Memory Control Group Interaction”…””}”(hj0hžhhŸNh Nubah}”(h]”Œid4”ah ]”h"]”h$]”h&]”Œrefid”Œ memory-control-group-interaction”uh1hËhj-ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhj*ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ"Marking Address Spaces Unevictable”…””}”(hjRhžhhŸNh Nubah}”(h]”Œid5”ah ]”h"]”h$]”h&]”Œrefid”Œ"marking-address-spaces-unevictable”uh1hËhjOubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjLubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒDetecting Unevictable Pages”…””}”(hjthžhhŸNh Nubah}”(h]”Œid6”ah ]”h"]”h$]”h&]”Œrefid”Œdetecting-unevictable-pages”uh1hËhjqubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjnubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ)Vmscanâ€™s Handling of Unevictable Folios”…””}”(hj–hžhhŸNh Nubah}”(h]”Œid7”ah ]”h"]”h$]”h&]”Œrefid”Œ'vmscan-s-handling-of-unevictable-folios”uh1hËhj“ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubeh}”(h]”h ]”h"]”h$]”h&]”uh1h¼hhéubeh}”(h]”h ]”h"]”h$]”h&]”uh1hÁhh¾ubhÂ)”}”(hhh]”(hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ MLOCKED Pages”…””}”(hjÄhžhhŸNh Nubah}”(h]”Œid8”ah ]”h"]”h$]”h&]”Œrefid”Œ mlocked-pages”uh1hËhjÁubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhj¾ubh½)”}”(hhh]”(hÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒHistory”…””}”(hjãhžhhŸNh Nubah}”(h]”Œid9”ah ]”h"]”h$]”h&]”Œrefid”Œhistory”uh1hËhjàubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjÝubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒBasic Management”…””}”(hjhžhhŸNh Nubah}”(h]”Œid10”ah ]”h"]”h$]”h&]”Œrefid”Œbasic-management”uh1hËhjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjÿubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ0mlock()/mlock2()/mlockall() System Call Handling”…””}”(hj'hžhhŸNh Nubah}”(h]”Œid11”ah ]”h"]”h$]”h&]”Œrefid”Œ*mlock-mlock2-mlockall-system-call-handling”uh1hËhj$ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhj!ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒFiltering Special VMAs”…””}”(hjIhžhhŸNh Nubah}”(h]”Œid12”ah ]”h"]”h$]”h&]”Œrefid”Œfiltering-special-vmas”uh1hËhjFubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjCubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ+munlock()/munlockall() System Call Handling”…””}”(hjkhžhhŸNh Nubah}”(h]”Œid13”ah ]”h"]”h$]”h&]”Œrefid”Œ'munlock-munlockall-system-call-handling”uh1hËhjhubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjeubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒMigrating MLOCKED Pages”…””}”(hjhžhhŸNh Nubah}”(h]”Œid14”ah ]”h"]”h$]”h&]”Œrefid”Œmigrating-mlocked-pages”uh1hËhjŠubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhj‡ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒCompacting MLOCKED Pages”…””}”(hj¯hžhhŸNh Nubah}”(h]”Œid15”ah ]”h"]”h$]”h&]”Œrefid”Œcompacting-mlocked-pages”uh1hËhj¬ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhj©ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒMLOCKING Transparent Huge Pages”…””}”(hjÑhžhhŸNh Nubah}”(h]”Œid16”ah ]”h"]”h$]”h&]”Œrefid”Œmlocking-transparent-huge-pages”uh1hËhjÎubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjËubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ%mmap(MAP_LOCKED) System Call Handling”…””}”(hjóhžhhŸNh Nubah}”(h]”Œid17”ah ]”h"]”h$]”h&]”Œrefid”Œ$mmap-map-locked-system-call-handling”uh1hËhjðubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjíubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒ+munmap()/exit()/exec() System Call Handling”…””}”(hjhžhhŸNh Nubah}”(h]”Œid18”ah ]”h"]”h$]”h&]”Œrefid”Œ%munmap-exit-exec-system-call-handling”uh1hËhjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒTruncating MLOCKED Pages”…””}”(hj7hžhhŸNh Nubah}”(h]”Œid19”ah ]”h"]”h$]”h&]”Œrefid”Œtruncating-mlocked-pages”uh1hËhj4ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhj1ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubhÂ)”}”(hhh]”hÇ)”}”(hhh]”hÌ)”}”(hhh]”hŒPage Reclaim in shrink_*_list()”…””}”(hjYhžhhŸNh Nubah}”(h]”Œid20”ah ]”h"]”h$]”h&]”Œrefid”Œpage-reclaim-in-shrink-list”uh1hËhjVubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjSubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÚubeh}”(h]”h ]”h"]”h$]”h&]”uh1h¼hj¾ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hÁhh¾ubeh}”(h]”h ]”h"]”h$]”h&]”uh1h¼hh¹hžhhŸNh Nubah}”(h]”Œcontents”ah ]”(Œcontents”Œlocal”eh"]”Œcontents”ah$]”h&]”uh1h·hŸh¶h Khh£hžhubh¢)”}”(hhh]”(h§)”}”(hŒIntroduction”h]”hŒIntroduction”…””}”(hj”hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”Œrefid”hÖuh1h¦hj‘hžhhŸh¶h K ubhÇ)”}”(hŒ˜This document describes the Linux memory manager's "Unevictable LRU" infrastructure and the use of this to manage several types of "unevictable" folios.”h]”hŒ¢This document describes the Linux memory managerâ€™s â€œUnevictable LRUâ€ infrastructure and the use of this to manage several types of â€œunevictableâ€ folios.”…””}”(hj£hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Khj‘hžhubhÇ)”}”(hXÂThe document attempts to provide the overall rationale behind this mechanism and the rationale for some of the design decisions that drove the implementation. The latter design rationale is discussed in the context of an implementation description. Admittedly, one can obtain the implementation details - the "what does it do?" - by reading the code. One hopes that the descriptions below add value by provide the answer to "why does it do that?".”h]”hXÊThe document attempts to provide the overall rationale behind this mechanism and the rationale for some of the design decisions that drove the implementation. The latter design rationale is discussed in the context of an implementation description. Admittedly, one can obtain the implementation details - the â€œwhat does it do?â€ - by reading the code. One hopes that the descriptions below add value by provide the answer to â€œwhy does it do that?â€.”…””}”(hj±hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Khj‘hžhubeh}”(h]”hÜah ]”h"]”Œintroduction”ah$]”h&]”uh1h¡hh£hžhhŸh¶h K ubh¢)”}”(hhh]”(h§)”}”(hŒThe Unevictable LRU”h]”hŒThe Unevictable LRU”…””}”(hjÉhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢høuh1h¦hjÆhžhhŸh¶h KubhÇ)”}”(hXRThe Unevictable LRU facility adds an additional LRU list to track unevictable folios and to hide these folios from vmscan. This mechanism is based on a patch by Larry Woodman of Red Hat to address several scalability problems with folio reclaim in Linux. The problems have been observed at customer sites on large memory x86_64 systems.”h]”hXRThe Unevictable LRU facility adds an additional LRU list to track unevictable folios and to hide these folios from vmscan. This mechanism is based on a patch by Larry Woodman of Red Hat to address several scalability problems with folio reclaim in Linux. The problems have been observed at customer sites on large memory x86_64 systems.”…””}”(hj×hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KhjÆhžhubhÇ)”}”(hXðTo illustrate this with an example, a non-NUMA x86_64 platform with 128GB of main memory will have over 32 million 4k pages in a single node. When a large fraction of these pages are not evictable for any reason [see below], vmscan will spend a lot of time scanning the LRU lists looking for the small fraction of pages that are evictable. This can result in a situation where all CPUs are spending 100% of their time in vmscan for hours or days on end, with the system completely unresponsive.”h]”hXðTo illustrate this with an example, a non-NUMA x86_64 platform with 128GB of main memory will have over 32 million 4k pages in a single node. When a large fraction of these pages are not evictable for any reason [see below], vmscan will spend a lot of time scanning the LRU lists looking for the small fraction of pages that are evictable. This can result in a situation where all CPUs are spending 100% of their time in vmscan for hours or days on end, with the system completely unresponsive.”…””}”(hjåhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K!hjÆhžhubhÇ)”}”(hŒJThe unevictable list addresses the following classes of unevictable pages:”h]”hŒJThe unevictable list addresses the following classes of unevictable pages:”…””}”(hjóhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K)hjÆhžhubhŒblock_quote”“”)”}”(hŒ¶* Those owned by ramfs. * Those owned by tmpfs with the noswap mount option. * Those mapped into SHM_LOCK'd shared memory regions. * Those mapped into VM_LOCKED [mlock()ed] VMAs. ”h]”h½)”}”(hhh]”(hÂ)”}”(hŒThose owned by ramfs. ”h]”hÇ)”}”(hŒThose owned by ramfs.”h]”hŒThose owned by ramfs.”…””}”(hjhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K+hj ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hŒ3Those owned by tmpfs with the noswap mount option. ”h]”hÇ)”}”(hŒ2Those owned by tmpfs with the noswap mount option.”h]”hŒ2Those owned by tmpfs with the noswap mount option.”…””}”(hj&hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K-hj"ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hŒ4Those mapped into SHM_LOCK'd shared memory regions. ”h]”hÇ)”}”(hŒ3Those mapped into SHM_LOCK'd shared memory regions.”h]”hŒ5Those mapped into SHM_LOCKâ€™d shared memory regions.”…””}”(hj>hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K/hj:ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubhÂ)”}”(hŒ.Those mapped into VM_LOCKED [mlock()ed] VMAs. ”h]”hÇ)”}”(hŒ-Those mapped into VM_LOCKED [mlock()ed] VMAs.”h]”hŒ-Those mapped into VM_LOCKED [mlock()ed] VMAs.”…””}”(hjVhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K1hjRubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjubeh}”(h]”h ]”h"]”h$]”h&]”Œbullet”Œ*”uh1h¼hŸh¶h K+hjubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h K+hjÆhžhubhÇ)”}”(hŒ“The infrastructure may also be able to handle other conditions that make pages unevictable, either by definition or by circumstance, in the future.”h]”hŒ“The infrastructure may also be able to handle other conditions that make pages unevictable, either by definition or by circumstance, in the future.”…””}”(hjxhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K3hjÆhžhubh¢)”}”(hhh]”(h§)”}”(hŒThe Unevictable LRU Folio List”h]”hŒThe Unevictable LRU Folio List”…””}”(hj‰hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢juh1h¦hj†hžhhŸh¶h K8ubhÇ)”}”(hXDThe Unevictable LRU folio list is a lie. It was never an LRU-ordered list, but a companion to the LRU-ordered anonymous and file, active and inactive folio lists; and now it is not even a folio list. But following familiar convention, here in this document and in the source, we often imagine it as a fifth LRU folio list.”h]”hXDThe Unevictable LRU folio list is a lie. It was never an LRU-ordered list, but a companion to the LRU-ordered anonymous and file, active and inactive folio lists; and now it is not even a folio list. But following familiar convention, here in this document and in the source, we often imagine it as a fifth LRU folio list.”…””}”(hj—hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K:hj†hžhubhÇ)”}”(hŒáThe Unevictable LRU infrastructure consists of an additional, per-node, LRU list called the "unevictable" list and an associated folio flag, PG_unevictable, to indicate that the folio is being managed on the unevictable list.”h]”hŒåThe Unevictable LRU infrastructure consists of an additional, per-node, LRU list called the â€œunevictableâ€ list and an associated folio flag, PG_unevictable, to indicate that the folio is being managed on the unevictable list.”…””}”(hj¥hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K@hj†hžhubhÇ)”}”(hŒ£The PG_unevictable flag is analogous to, and mutually exclusive with, the PG_active flag in that it indicates on which LRU list a folio resides when PG_lru is set.”h]”hŒ£The PG_unevictable flag is analogous to, and mutually exclusive with, the PG_active flag in that it indicates on which LRU list a folio resides when PG_lru is set.”…””}”(hj³hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KDhj†hžhubhÇ)”}”(hŒ|The Unevictable LRU infrastructure maintains unevictable folios as if they were on an additional LRU list for a few reasons:”h]”hŒ|The Unevictable LRU infrastructure maintains unevictable folios as if they were on an additional LRU list for a few reasons:”…””}”(hjÁhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KHhj†hžhubj)”}”(hXá(1) We get to "treat unevictable folios just like we treat other folios in the system - which means we get to use the same code to manipulate them, the same code to isolate them (for migrate, etc.), the same code to keep track of the statistics, etc..." [Rik van Riel] (2) We want to be able to migrate unevictable folios between nodes for memory defragmentation, workload management and memory hotplug. The Linux kernel can only migrate folios that it can successfully isolate from the LRU lists (or "Movable" folios: outside of consideration here). If we were to maintain folios elsewhere than on an LRU-like list, where they can be detected by folio_isolate_lru(), we would prevent their migration. ”h]”hŒenumerated_list”“”)”}”(hhh]”(hÂ)”}”(hX We get to "treat unevictable folios just like we treat other folios in the system - which means we get to use the same code to manipulate them, the same code to isolate them (for migrate, etc.), the same code to keep track of the statistics, etc..." [Rik van Riel] ”h]”hÇ)”}”(hXWe get to "treat unevictable folios just like we treat other folios in the system - which means we get to use the same code to manipulate them, the same code to isolate them (for migrate, etc.), the same code to keep track of the statistics, etc..." [Rik van Riel]”h]”hXWe get to â€œtreat unevictable folios just like we treat other folios in the system - which means we get to use the same code to manipulate them, the same code to isolate them (for migrate, etc.), the same code to keep track of the statistics, etc...â€ [Rik van Riel]”…””}”(hjÜhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KKhjØubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÕubhÂ)”}”(hX¯We want to be able to migrate unevictable folios between nodes for memory defragmentation, workload management and memory hotplug. The Linux kernel can only migrate folios that it can successfully isolate from the LRU lists (or "Movable" folios: outside of consideration here). If we were to maintain folios elsewhere than on an LRU-like list, where they can be detected by folio_isolate_lru(), we would prevent their migration. ”h]”hÇ)”}”(hX®We want to be able to migrate unevictable folios between nodes for memory defragmentation, workload management and memory hotplug. The Linux kernel can only migrate folios that it can successfully isolate from the LRU lists (or "Movable" folios: outside of consideration here). If we were to maintain folios elsewhere than on an LRU-like list, where they can be detected by folio_isolate_lru(), we would prevent their migration.”h]”hX²We want to be able to migrate unevictable folios between nodes for memory defragmentation, workload management and memory hotplug. The Linux kernel can only migrate folios that it can successfully isolate from the LRU lists (or â€œMovableâ€ folios: outside of consideration here). If we were to maintain folios elsewhere than on an LRU-like list, where they can be detected by folio_isolate_lru(), we would prevent their migration.”…””}”(hjôhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KPhjðubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÕubeh}”(h]”h ]”h"]”h$]”h&]”Œenumtype”Œarabic”Œprefix”Œ(”Œsuffix”Œ)”uh1jÓhjÏubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h KKhj†hžhubhÇ)”}”(hŒ´The unevictable list does not differentiate between file-backed and anonymous, swap-backed folios. This differentiation is only important while the folios are, in fact, evictable.”h]”hŒ´The unevictable list does not differentiate between file-backed and anonymous, swap-backed folios. This differentiation is only important while the folios are, in fact, evictable.”…””}”(hjhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KWhj†hžhubhÇ)”}”(hŒ•The unevictable list benefits from the "arrayification" of the per-node LRU lists and statistics originally proposed and posted by Christoph Lameter.”h]”hŒ™The unevictable list benefits from the â€œarrayificationâ€ of the per-node LRU lists and statistics originally proposed and posted by Christoph Lameter.”…””}”(hj(hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K[hj†hžhubeh}”(h]”jah ]”h"]”Œthe unevictable lru folio list”ah$]”h&]”uh1h¡hjÆhžhhŸh¶h K8ubh¢)”}”(hhh]”(h§)”}”(hŒ Memory Control Group Interaction”h]”hŒ Memory Control Group Interaction”…””}”(hj@hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢j9uh1h¦hj=hžhhŸh¶h K`ubhÇ)”}”(hŒ°The unevictable LRU facility interacts with the memory control group [aka memory controller; see Documentation/admin-guide/cgroup-v1/memory.rst] by extending the lru_list enum.”h]”hŒ°The unevictable LRU facility interacts with the memory control group [aka memory controller; see Documentation/admin-guide/cgroup-v1/memory.rst] by extending the lru_list enum.”…””}”(hjNhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kbhj=hžhubhÇ)”}”(hXThe memory controller data structure automatically gets a per-node unevictable list as a result of the "arrayification" of the per-node LRU lists (one per lru_list enum element). The memory controller tracks the movement of pages to and from the unevictable list.”h]”hXThe memory controller data structure automatically gets a per-node unevictable list as a result of the â€œarrayificationâ€ of the per-node LRU lists (one per lru_list enum element). The memory controller tracks the movement of pages to and from the unevictable list.”…””}”(hj\hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kfhj=hžhubhÇ)”}”(hŒ¡When a memory control group comes under memory pressure, the controller will not attempt to reclaim pages on the unevictable list. This has a couple of effects:”h]”hŒ¡When a memory control group comes under memory pressure, the controller will not attempt to reclaim pages on the unevictable list. This has a couple of effects:”…””}”(hjjhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kkhj=hžhubj)”}”(hXÛ(1) Because the pages are "hidden" from reclaim on the unevictable list, the reclaim process can be more efficient, dealing only with pages that have a chance of being reclaimed. (2) On the other hand, if too many of the pages charged to the control group are unevictable, the evictable portion of the working set of the tasks in the control group may not fit into the available memory. This can cause the control group to thrash or to OOM-kill tasks. ”h]”jÔ)”}”(hhh]”(hÂ)”}”(hŒ¯Because the pages are "hidden" from reclaim on the unevictable list, the reclaim process can be more efficient, dealing only with pages that have a chance of being reclaimed. ”h]”hÇ)”}”(hŒ®Because the pages are "hidden" from reclaim on the unevictable list, the reclaim process can be more efficient, dealing only with pages that have a chance of being reclaimed.”h]”hŒ²Because the pages are â€œhiddenâ€ from reclaim on the unevictable list, the reclaim process can be more efficient, dealing only with pages that have a chance of being reclaimed.”…””}”(hjƒhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kohjubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj|ubhÂ)”}”(hXOn the other hand, if too many of the pages charged to the control group are unevictable, the evictable portion of the working set of the tasks in the control group may not fit into the available memory. This can cause the control group to thrash or to OOM-kill tasks. ”h]”hÇ)”}”(hX On the other hand, if too many of the pages charged to the control group are unevictable, the evictable portion of the working set of the tasks in the control group may not fit into the available memory. This can cause the control group to thrash or to OOM-kill tasks.”h]”hX On the other hand, if too many of the pages charged to the control group are unevictable, the evictable portion of the working set of the tasks in the control group may not fit into the available memory. This can cause the control group to thrash or to OOM-kill tasks.”…””}”(hj›hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kshj—ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj|ubeh}”(h]”h ]”h"]”h$]”h&]”jjjjjjuh1jÓhjxubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h Kohj=hžhubhŒtarget”“”)”}”(hŒ.. _mark_addr_space_unevict:”h]”h}”(h]”h ]”h"]”h$]”h&]”j¢Œmark-addr-space-unevict”uh1j»h Kyhj=hžhhŸh¶ubeh}”(h]”j?ah ]”h"]”Œ memory control group interaction”ah$]”h&]”uh1h¡hjÆhžhhŸh¶h K`ubh¢)”}”(hhh]”(h§)”}”(hŒ"Marking Address Spaces Unevictable”h]”hŒ"Marking Address Spaces Unevictable”…””}”(hjÒhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢j[uh1h¦hjÏhžhhŸh¶h K|ubhÇ)”}”(hXFor facilities such as ramfs none of the pages attached to the address space may be evicted. To prevent eviction of any such pages, the AS_UNEVICTABLE address space flag is provided, and this can be manipulated by a filesystem using a number of wrapper functions:”h]”hXFor facilities such as ramfs none of the pages attached to the address space may be evicted. To prevent eviction of any such pages, the AS_UNEVICTABLE address space flag is provided, and this can be manipulated by a filesystem using a number of wrapper functions:”…””}”(hjàhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K~hjÏhžhubj)”}”(hX”* ``void mapping_set_unevictable(struct address_space *mapping);`` Mark the address space as being completely unevictable. * ``void mapping_clear_unevictable(struct address_space *mapping);`` Mark the address space as being evictable. * ``int mapping_unevictable(struct address_space *mapping);`` Query the address space, and return true if it is completely unevictable. ”h]”h½)”}”(hhh]”(hÂ)”}”(hŒ``void mapping_set_unevictable(struct address_space *mapping);`` Mark the address space as being completely unevictable. ”h]”(hÇ)”}”(hŒ@``void mapping_set_unevictable(struct address_space *mapping);``”h]”hŒliteral”“”)”}”(hjûh]”hŒvoid mapping_clear_unevictable(struct address_space *mapping);”…””}”(hj8hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1jýhj4ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K‡hj0ubj)”}”(hŒ+Mark the address space as being evictable. ”h]”hÇ)”}”(hŒ*Mark the address space as being evictable.”h]”hŒ*Mark the address space as being evictable.”…””}”(hjOhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K‰hjKubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h K‰hj0ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjòubhÂ)”}”(hŒ‘``int mapping_unevictable(struct address_space *mapping);`` Query the address space, and return true if it is completely unevictable. ”h]”(hÇ)”}”(hŒ;``int mapping_unevictable(struct address_space *mapping);``”h]”jþ)”}”(hjoh]”hŒ7int mapping_unevictable(struct address_space *mapping);”…””}”(hjqhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1jýhjmubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K‹hjiubj)”}”(hŒJQuery the address space, and return true if it is completely unevictable. ”h]”hÇ)”}”(hŒIQuery the address space, and return true if it is completely unevictable.”h]”hŒIQuery the address space, and return true if it is completely unevictable.”…””}”(hjˆhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Khj„ubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h Khjiubeh}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjòubeh}”(h]”h ]”h"]”h$]”h&]”jpjquh1h¼hŸh¶h Kƒhjîubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h KƒhjÏhžhubhÇ)”}”(hŒ7These are currently used in three places in the kernel:”h]”hŒ7These are currently used in three places in the kernel:”…””}”(hj®hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KhjÏhžhubj)”}”(hX[(1) By ramfs to mark the address spaces of its inodes when they are created, and this mark remains for the life of the inode. (2) By SYSV SHM to mark SHM_LOCK'd address spaces until SHM_UNLOCK is called. Note that SHM_LOCK is not required to page in the locked pages if they're swapped out; the application must touch the pages manually if it wants to ensure they're in memory. (3) By the i915 driver to mark pinned address space until it's unpinned. The amount of unevictable memory marked by i915 driver is roughly the bounded object size in debugfs/dri/0/i915_gem_objects. ”h]”jÔ)”}”(hhh]”(hÂ)”}”(hŒzBy ramfs to mark the address spaces of its inodes when they are created, and this mark remains for the life of the inode. ”h]”hÇ)”}”(hŒyBy ramfs to mark the address spaces of its inodes when they are created, and this mark remains for the life of the inode.”h]”hŒyBy ramfs to mark the address spaces of its inodes when they are created, and this mark remains for the life of the inode.”…””}”(hjÇhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K’hjÃubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÀubhÂ)”}”(hŒøBy SYSV SHM to mark SHM_LOCK'd address spaces until SHM_UNLOCK is called. Note that SHM_LOCK is not required to page in the locked pages if they're swapped out; the application must touch the pages manually if it wants to ensure they're in memory. ”h]”hÇ)”}”(hŒ÷By SYSV SHM to mark SHM_LOCK'd address spaces until SHM_UNLOCK is called. Note that SHM_LOCK is not required to page in the locked pages if they're swapped out; the application must touch the pages manually if it wants to ensure they're in memory.”h]”hŒýBy SYSV SHM to mark SHM_LOCKâ€™d address spaces until SHM_UNLOCK is called. Note that SHM_LOCK is not required to page in the locked pages if theyâ€™re swapped out; the application must touch the pages manually if it wants to ensure theyâ€™re in memory.”…””}”(hjßhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K•hjÛubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÀubhÂ)”}”(hŒÃBy the i915 driver to mark pinned address space until it's unpinned. The amount of unevictable memory marked by i915 driver is roughly the bounded object size in debugfs/dri/0/i915_gem_objects. ”h]”hÇ)”}”(hŒÁBy the i915 driver to mark pinned address space until it's unpinned. The amount of unevictable memory marked by i915 driver is roughly the bounded object size in debugfs/dri/0/i915_gem_objects.”h]”hŒÃBy the i915 driver to mark pinned address space until itâ€™s unpinned. The amount of unevictable memory marked by i915 driver is roughly the bounded object size in debugfs/dri/0/i915_gem_objects.”…””}”(hj÷hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kšhjóubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjÀubeh}”(h]”h ]”h"]”h$]”h&]”jjjjjjuh1jÓhj¼ubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h K’hjÏhžhubeh}”(h]”(jajÇeh ]”h"]”(Œ"marking address spaces unevictable”Œmark_addr_space_unevict”eh$]”h&]”uh1h¡hjÆhžhhŸh¶h K|Œexpect_referenced_by_name”}”jj½sŒexpect_referenced_by_id”}”jÇj½subh¢)”}”(hhh]”(h§)”}”(hŒDetecting Unevictable Pages”h]”hŒDetecting Unevictable Pages”…””}”(hj&hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢j}uh1h¦hj#hžhhŸh¶h K ubhÇ)”}”(hŒúThe function folio_evictable() in mm/internal.h determines whether a folio is evictable or not using the query function outlined above [see section :ref:`Marking address spaces unevictable `] to check the AS_UNEVICTABLE flag.”h]”(hŒ”The function folio_evictable() in mm/internal.h determines whether a folio is evictable or not using the query function outlined above [see section ”…””}”(hj4hžhhŸNh Nubh)”}”(hŒC:ref:`Marking address spaces unevictable `”h]”hŒinline”“”)”}”(hj>h]”hŒ"Marking address spaces unevictable”…””}”(hjBhžhhŸNh Nubah}”(h]”h ]”(Œxref”Œstd”Œstd-ref”eh"]”h$]”h&]”uh1j@hj<ubah}”(h]”h ]”h"]”h$]”h&]”Œrefdoc”Œmm/unevictable-lru”Œ refdomain”jMŒreftype”Œref”Œrefexplicit”ˆŒrefwarn”ˆŒ reftarget”Œmark_addr_space_unevict”uh1hhŸh¶h K¢hj4ubhŒ#] to check the AS_UNEVICTABLE flag.”…””}”(hj4hžhhŸNh Nubeh}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K¢hj#hžhubhÇ)”}”(hX›For address spaces that are so marked after being populated (as SHM regions might be), the lock action (e.g. SHM_LOCK) can be lazy, and need not populate the page tables for the region as does, for example, mlock(), nor need it make any special effort to push any pages in the SHM_LOCK'd area to the unevictable list. Instead, vmscan will do this if and when it encounters the folios during a reclamation scan.”h]”hXFor address spaces that are so marked after being populated (as SHM regions might be), the lock action (e.g. SHM_LOCK) can be lazy, and need not populate the page tables for the region as does, for example, mlock(), nor need it make any special effort to push any pages in the SHM_LOCKâ€™d area to the unevictable list. Instead, vmscan will do this if and when it encounters the folios during a reclamation scan.”…””}”(hjkhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K§hj#hžhubhÇ)”}”(hXFOn an unlock action (such as SHM_UNLOCK), the unlocker (e.g. shmctl()) must scan the pages in the region and "rescue" them from the unevictable list if no other condition is keeping them unevictable. If an unevictable region is destroyed, the pages are also "rescued" from the unevictable list in the process of freeing them.”h]”hXNOn an unlock action (such as SHM_UNLOCK), the unlocker (e.g. shmctl()) must scan the pages in the region and â€œrescueâ€ them from the unevictable list if no other condition is keeping them unevictable. If an unevictable region is destroyed, the pages are also â€œrescuedâ€ from the unevictable list in the process of freeing them.”…””}”(hjyhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K®hj#hžhubhÇ)”}”(hŒ¯folio_evictable() also checks for mlocked folios by calling folio_test_mlocked(), which is set when a folio is faulted into a VM_LOCKED VMA, or found in a VMA being VM_LOCKED.”h]”hŒ¯folio_evictable() also checks for mlocked folios by calling folio_test_mlocked(), which is set when a folio is faulted into a VM_LOCKED VMA, or found in a VMA being VM_LOCKED.”…””}”(hj‡hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K´hj#hžhubeh}”(h]”jƒah ]”h"]”Œdetecting unevictable pages”ah$]”h&]”uh1h¡hjÆhžhhŸh¶h K ubh¢)”}”(hhh]”(h§)”}”(hŒ'Vmscan's Handling of Unevictable Folios”h]”hŒ)Vmscanâ€™s Handling of Unevictable Folios”…””}”(hjŸhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢jŸuh1h¦hjœhžhhŸh¶h KºubhÇ)”}”(hX´If unevictable folios are culled in the fault path, or moved to the unevictable list at mlock() or mmap() time, vmscan will not encounter the folios until they have become evictable again (via munlock() for example) and have been "rescued" from the unevictable list. However, there may be situations where we decide, for the sake of expediency, to leave an unevictable folio on one of the regular active/inactive LRU lists for vmscan to deal with. vmscan checks for such folios in all of the shrink_{active|inactive|folio}_list() functions and will "cull" such folios that it encounters: that is, it diverts those folios to the unevictable list for the memory cgroup and node being scanned.”h]”hX¼If unevictable folios are culled in the fault path, or moved to the unevictable list at mlock() or mmap() time, vmscan will not encounter the folios until they have become evictable again (via munlock() for example) and have been â€œrescuedâ€ from the unevictable list. However, there may be situations where we decide, for the sake of expediency, to leave an unevictable folio on one of the regular active/inactive LRU lists for vmscan to deal with. vmscan checks for such folios in all of the shrink_{active|inactive|folio}_list() functions and will â€œcullâ€ such folios that it encounters: that is, it diverts those folios to the unevictable list for the memory cgroup and node being scanned.”…””}”(hjhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h K¼hjœhžhubhÇ)”}”(hX…There may be situations where a folio is mapped into a VM_LOCKED VMA, but the folio does not have the mlocked flag set. Such folios will make it all the way to shrink_active_list() or shrink_folio_list() where they will be detected when vmscan walks the reverse map in folio_referenced() or try_to_unmap(). The folio is culled to the unevictable list when it is released by the shrinker.”h]”hX…There may be situations where a folio is mapped into a VM_LOCKED VMA, but the folio does not have the mlocked flag set. Such folios will make it all the way to shrink_active_list() or shrink_folio_list() where they will be detected when vmscan walks the reverse map in folio_referenced() or try_to_unmap(). The folio is culled to the unevictable list when it is released by the shrinker.”…””}”(hj»hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KÆhjœhžhubhÇ)”}”(hX‹To "cull" an unevictable folio, vmscan simply puts the folio back on the LRU list using folio_putback_lru() - the inverse operation to folio_isolate_lru() - after dropping the folio lock. Because the condition which makes the folio unevictable may change once the folio is unlocked, __pagevec_lru_add_fn() will recheck the unevictable state of a folio before placing it on the unevictable list.”h]”hXTo â€œcullâ€ an unevictable folio, vmscan simply puts the folio back on the LRU list using folio_putback_lru() - the inverse operation to folio_isolate_lru() - after dropping the folio lock. Because the condition which makes the folio unevictable may change once the folio is unlocked, __pagevec_lru_add_fn() will recheck the unevictable state of a folio before placing it on the unevictable list.”…””}”(hjÉhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KÍhjœhžhubeh}”(h]”j¥ah ]”h"]”Œ'vmscan's handling of unevictable folios”ah$]”h&]”uh1h¡hjÆhžhhŸh¶h Kºubeh}”(h]”hþah ]”h"]”Œthe unevictable lru”ah$]”h&]”uh1h¡hh£hžhhŸh¶h Kubh¢)”}”(hhh]”(h§)”}”(hŒ MLOCKED Pages”h]”hŒ MLOCKED Pages”…””}”(hjèhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢jÍuh1h¦hjåhžhhŸh¶h KÖubhÇ)”}”(hŒÕThe unevictable folio list is also useful for mlock(), in addition to ramfs and SYSV SHM. Note that mlock() is only available in CONFIG_MMU=y situations; in NOMMU situations, all mappings are effectively mlocked.”h]”hŒÕThe unevictable folio list is also useful for mlock(), in addition to ramfs and SYSV SHM. Note that mlock() is only available in CONFIG_MMU=y situations; in NOMMU situations, all mappings are effectively mlocked.”…””}”(hjöhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KØhjåhžhubh¢)”}”(hhh]”(h§)”}”(hŒHistory”h]”hŒHistory”…””}”(hjhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢jìuh1h¦hjhžhhŸh¶h KÞubhÇ)”}”(hX&The "Unevictable mlocked Pages" infrastructure is based on work originally posted by Nick Piggin in an RFC patch entitled "mm: mlocked pages off LRU". Nick posted his patch as an alternative to a patch posted by Christoph Lameter to achieve the same objective: hiding mlocked pages from vmscan.”h]”hX.The â€œUnevictable mlocked Pagesâ€ infrastructure is based on work originally posted by Nick Piggin in an RFC patch entitled â€œmm: mlocked pages off LRUâ€. Nick posted his patch as an alternative to a patch posted by Christoph Lameter to achieve the same objective: hiding mlocked pages from vmscan.”…””}”(hjhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KàhjhžhubhÇ)”}”(hX®In Nick's patch, he used one of the struct page LRU list link fields as a count of VM_LOCKED VMAs that map the page (Rik van Riel had the same idea three years earlier). But this use of the link field for a count prevented the management of the pages on an LRU list, and thus mlocked pages were not migratable as folio_isolate_lru() could not detect them, and the LRU list link field was not available to the migration subsystem.”h]”hX°In Nickâ€™s patch, he used one of the struct page LRU list link fields as a count of VM_LOCKED VMAs that map the page (Rik van Riel had the same idea three years earlier). But this use of the link field for a count prevented the management of the pages on an LRU list, and thus mlocked pages were not migratable as folio_isolate_lru() could not detect them, and the LRU list link field was not available to the migration subsystem.”…””}”(hj#hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KåhjhžhubhÇ)”}”(hXYNick resolved this by putting mlocked pages back on the LRU list before attempting to isolate them, thus abandoning the count of VM_LOCKED VMAs. When Nick's patch was integrated with the Unevictable LRU work, the count was replaced by walking the reverse map when munlocking, to determine whether any other VM_LOCKED VMAs still mapped the page.”h]”hX[Nick resolved this by putting mlocked pages back on the LRU list before attempting to isolate them, thus abandoning the count of VM_LOCKED VMAs. When Nickâ€™s patch was integrated with the Unevictable LRU work, the count was replaced by walking the reverse map when munlocking, to determine whether any other VM_LOCKED VMAs still mapped the page.”…””}”(hj1hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KìhjhžhubhÇ)”}”(hX However, walking the reverse map for each page when munlocking was ugly and inefficient, and could lead to catastrophic contention on a file's rmap lock, when many processes which had it mlocked were trying to exit. In 5.18, the idea of keeping mlock_count in Unevictable LRU list link field was revived and put to work, without preventing the migration of mlocked pages. This is why the "Unevictable LRU list" cannot be a linked list of pages now; but there was no use for that linked list anyway - though its size is maintained for meminfo.”h]”hX&However, walking the reverse map for each page when munlocking was ugly and inefficient, and could lead to catastrophic contention on a fileâ€™s rmap lock, when many processes which had it mlocked were trying to exit. In 5.18, the idea of keeping mlock_count in Unevictable LRU list link field was revived and put to work, without preventing the migration of mlocked pages. This is why the â€œUnevictable LRU listâ€ cannot be a linked list of pages now; but there was no use for that linked list anyway - though its size is maintained for meminfo.”…””}”(hj?hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Kòhjhžhubeh}”(h]”jòah ]”h"]”Œhistory”ah$]”h&]”uh1h¡hjåhžhhŸh¶h KÞubh¢)”}”(hhh]”(h§)”}”(hŒBasic Management”h]”hŒBasic Management”…””}”(hjWhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢juh1h¦hjThžhhŸh¶h KüubhÇ)”}”(hX$mlocked pages - pages mapped into a VM_LOCKED VMA - are a class of unevictable pages. When such a page has been "noticed" by the memory management subsystem, the folio is marked with the PG_mlocked flag. This can be manipulated using folio_set_mlocked() and folio_clear_mlocked() functions.”h]”hX(mlocked pages - pages mapped into a VM_LOCKED VMA - are a class of unevictable pages. When such a page has been â€œnoticedâ€ by the memory management subsystem, the folio is marked with the PG_mlocked flag. This can be manipulated using folio_set_mlocked() and folio_clear_mlocked() functions.”…””}”(hjehžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h KþhjThžhubhÇ)”}”(hŒšA PG_mlocked page will be placed on the unevictable list when it is added to the LRU. Such pages can be "noticed" by memory management in several places:”h]”hŒžA PG_mlocked page will be placed on the unevictable list when it is added to the LRU. Such pages can be â€œnoticedâ€ by memory management in several places:”…””}”(hjshžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MhjThžhubj)”}”(hXÖ(1) in the mlock()/mlock2()/mlockall() system call handlers; (2) in the mmap() system call handler when mmapping a region with the MAP_LOCKED flag; (3) mmapping a region in a task that has called mlockall() with the MCL_FUTURE flag; (4) in the fault path and when a VM_LOCKED stack segment is expanded; or (5) as mentioned above, in vmscan:shrink_folio_list() when attempting to reclaim a page in a VM_LOCKED VMA by folio_referenced() or try_to_unmap(). ”h]”jÔ)”}”(hhh]”(hÂ)”}”(hŒ9in the mlock()/mlock2()/mlockall() system call handlers; ”h]”hÇ)”}”(hŒ8in the mlock()/mlock2()/mlockall() system call handlers;”h]”hŒ8in the mlock()/mlock2()/mlockall() system call handlers;”…””}”(hjŒhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhjˆubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj…ubhÂ)”}”(hŒSin the mmap() system call handler when mmapping a region with the MAP_LOCKED flag; ”h]”hÇ)”}”(hŒRin the mmap() system call handler when mmapping a region with the MAP_LOCKED flag;”h]”hŒRin the mmap() system call handler when mmapping a region with the MAP_LOCKED flag;”…””}”(hj¤hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhj ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj…ubhÂ)”}”(hŒQmmapping a region in a task that has called mlockall() with the MCL_FUTURE flag; ”h]”hÇ)”}”(hŒPmmapping a region in a task that has called mlockall() with the MCL_FUTURE flag;”h]”hŒPmmapping a region in a task that has called mlockall() with the MCL_FUTURE flag;”…””}”(hj¼hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhj¸ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj…ubhÂ)”}”(hŒEin the fault path and when a VM_LOCKED stack segment is expanded; or ”h]”hÇ)”}”(hŒDin the fault path and when a VM_LOCKED stack segment is expanded; or”h]”hŒDin the fault path and when a VM_LOCKED stack segment is expanded; or”…””}”(hjÔhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MhjÐubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj…ubhÂ)”}”(hŒas mentioned above, in vmscan:shrink_folio_list() when attempting to reclaim a page in a VM_LOCKED VMA by folio_referenced() or try_to_unmap(). ”h]”hÇ)”}”(hŒas mentioned above, in vmscan:shrink_folio_list() when attempting to reclaim a page in a VM_LOCKED VMA by folio_referenced() or try_to_unmap().”h]”hŒas mentioned above, in vmscan:shrink_folio_list() when attempting to reclaim a page in a VM_LOCKED VMA by folio_referenced() or try_to_unmap().”…””}”(hjìhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhjèubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj…ubeh}”(h]”h ]”h"]”h$]”h&]”jjjjjjuh1jÓhjubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h MhjThžhubhÇ)”}”(hŒImlocked pages become unlocked and rescued from the unevictable list when:”h]”hŒImlocked pages become unlocked and rescued from the unevictable list when:”…””}”(hj hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MhjThžhubj)”}”(hX<(1) mapped in a range unlocked via the munlock()/munlockall() system calls; (2) munmap()'d out of the last VM_LOCKED VMA that maps the page, including unmapping at task exit; (3) when the page is truncated from the last VM_LOCKED VMA of an mmapped file; or (4) before a page is COW'd in a VM_LOCKED VMA. ”h]”jÔ)”}”(hhh]”(hÂ)”}”(hŒHmapped in a range unlocked via the munlock()/munlockall() system calls; ”h]”hÇ)”}”(hŒGmapped in a range unlocked via the munlock()/munlockall() system calls;”h]”hŒGmapped in a range unlocked via the munlock()/munlockall() system calls;”…””}”(hj% hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhj! ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj ubhÂ)”}”(hŒ_munmap()'d out of the last VM_LOCKED VMA that maps the page, including unmapping at task exit; ”h]”hÇ)”}”(hŒ^munmap()'d out of the last VM_LOCKED VMA that maps the page, including unmapping at task exit;”h]”hŒ`munmap()â€™d out of the last VM_LOCKED VMA that maps the page, including unmapping at task exit;”…””}”(hj= hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhj9 ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj ubhÂ)”}”(hŒNwhen the page is truncated from the last VM_LOCKED VMA of an mmapped file; or ”h]”hÇ)”}”(hŒMwhen the page is truncated from the last VM_LOCKED VMA of an mmapped file; or”h]”hŒMwhen the page is truncated from the last VM_LOCKED VMA of an mmapped file; or”…””}”(hjU hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MhjQ ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj ubhÂ)”}”(hŒ,before a page is COW'd in a VM_LOCKED VMA. ”h]”hÇ)”}”(hŒ*before a page is COW'd in a VM_LOCKED VMA.”h]”hŒ,before a page is COWâ€™d in a VM_LOCKED VMA.”…””}”(hjm hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhji ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhj ubeh}”(h]”h ]”h"]”h$]”h&]”jjjjjjuh1jÓhj ubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h MhjThžhubeh}”(h]”jah ]”h"]”Œbasic management”ah$]”h&]”uh1h¡hjåhžhhŸh¶h Küubh¢)”}”(hhh]”(h§)”}”(hŒ0mlock()/mlock2()/mlockall() System Call Handling”h]”hŒ0mlock()/mlock2()/mlockall() System Call Handling”…””}”(hj— hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢j0uh1h¦hj” hžhhŸh¶h M!ubhÇ)”}”(hX·mlock(), mlock2() and mlockall() system call handlers proceed to mlock_fixup() for each VMA in the range specified by the call. In the case of mlockall(), this is the entire active address space of the task. Note that mlock_fixup() is used for both mlocking and munlocking a range of memory. A call to mlock() an already VM_LOCKED VMA, or to munlock() a VMA that is not VM_LOCKED, is treated as a no-op and mlock_fixup() simply returns.”h]”hX·mlock(), mlock2() and mlockall() system call handlers proceed to mlock_fixup() for each VMA in the range specified by the call. In the case of mlockall(), this is the entire active address space of the task. Note that mlock_fixup() is used for both mlocking and munlocking a range of memory. A call to mlock() an already VM_LOCKED VMA, or to munlock() a VMA that is not VM_LOCKED, is treated as a no-op and mlock_fixup() simply returns.”…””}”(hj¥ hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M#hj” hžhubhÇ)”}”(hXwIf the VMA passes some filtering as described in "Filtering Special VMAs" below, mlock_fixup() will attempt to merge the VMA with its neighbors or split off a subset of the VMA if the range does not cover the entire VMA. Any pages already present in the VMA are then marked as mlocked by mlock_folio() via mlock_pte_range() via walk_page_range() via mlock_vma_pages_range().”h]”hX{If the VMA passes some filtering as described in â€œFiltering Special VMAsâ€ below, mlock_fixup() will attempt to merge the VMA with its neighbors or split off a subset of the VMA if the range does not cover the entire VMA. Any pages already present in the VMA are then marked as mlocked by mlock_folio() via mlock_pte_range() via walk_page_range() via mlock_vma_pages_range().”…””}”(hj³ hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M*hj” hžhubhÇ)”}”(hŒÆBefore returning from the system call, do_mlock() or mlockall() will call __mm_populate() to fault in the remaining pages via get_user_pages() and to mark those pages as mlocked as they are faulted.”h]”hŒÆBefore returning from the system call, do_mlock() or mlockall() will call __mm_populate() to fault in the remaining pages via get_user_pages() and to mark those pages as mlocked as they are faulted.”…””}”(hjÁ hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M0hj” hžhubhÇ)”}”(hX7Note that the VMA being mlocked might be mapped with PROT_NONE. In this case, get_user_pages() will be unable to fault in the pages. That's okay. If pages do end up getting faulted into this VM_LOCKED VMA, they will be handled in the fault path - which is also how mlock2()'s MLOCK_ONFAULT areas are handled.”h]”hX;Note that the VMA being mlocked might be mapped with PROT_NONE. In this case, get_user_pages() will be unable to fault in the pages. Thatâ€™s okay. If pages do end up getting faulted into this VM_LOCKED VMA, they will be handled in the fault path - which is also how mlock2()â€™s MLOCK_ONFAULT areas are handled.”…””}”(hjÏ hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M4hj” hžhubhÇ)”}”(hX½For each PTE (or PMD) being faulted into a VMA, the page add rmap function calls mlock_vma_folio(), which calls mlock_folio() when the VMA is VM_LOCKED (unless it is a PTE mapping of a part of a transparent huge page). Or when it is a newly allocated anonymous page, folio_add_lru_vma() calls mlock_new_folio() instead: similar to mlock_folio(), but can make better judgments, since this page is held exclusively and known not to be on LRU yet.”h]”hX½For each PTE (or PMD) being faulted into a VMA, the page add rmap function calls mlock_vma_folio(), which calls mlock_folio() when the VMA is VM_LOCKED (unless it is a PTE mapping of a part of a transparent huge page). Or when it is a newly allocated anonymous page, folio_add_lru_vma() calls mlock_new_folio() instead: similar to mlock_folio(), but can make better judgments, since this page is held exclusively and known not to be on LRU yet.”…””}”(hjÝ hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M9hj” hžhubhÇ)”}”(hXÉmlock_folio() sets PG_mlocked immediately, then places the page on the CPU's mlock folio batch, to batch up the rest of the work to be done under lru_lock by __mlock_folio(). __mlock_folio() sets PG_unevictable, initializes mlock_count and moves the page to unevictable state ("the unevictable LRU", but with mlock_count in place of LRU threading). Or if the page was already PG_lru and PG_unevictable and PG_mlocked, it simply increments the mlock_count.”h]”hXÏmlock_folio() sets PG_mlocked immediately, then places the page on the CPUâ€™s mlock folio batch, to batch up the rest of the work to be done under lru_lock by __mlock_folio(). __mlock_folio() sets PG_unevictable, initializes mlock_count and moves the page to unevictable state (â€œthe unevictable LRUâ€, but with mlock_count in place of LRU threading). Or if the page was already PG_lru and PG_unevictable and PG_mlocked, it simply increments the mlock_count.”…””}”(hjë hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M@hj” hžhubhÇ)”}”(hX7But in practice that may not work ideally: the page may not yet be on an LRU, or it may have been temporarily isolated from LRU. In such cases the mlock_count field cannot be touched, but will be set to 0 later when __munlock_folio() returns the page to "LRU". Races prohibit mlock_count from being set to 1 then: rather than risk stranding a page indefinitely as unevictable, always err with mlock_count on the low side, so that when munlocked the page will be rescued to an evictable LRU, then perhaps be mlocked again later if vmscan finds it in a VM_LOCKED VMA.”h]”hX;But in practice that may not work ideally: the page may not yet be on an LRU, or it may have been temporarily isolated from LRU. In such cases the mlock_count field cannot be touched, but will be set to 0 later when __munlock_folio() returns the page to â€œLRUâ€. Races prohibit mlock_count from being set to 1 then: rather than risk stranding a page indefinitely as unevictable, always err with mlock_count on the low side, so that when munlocked the page will be rescued to an evictable LRU, then perhaps be mlocked again later if vmscan finds it in a VM_LOCKED VMA.”…””}”(hjù hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MGhj” hžhubeh}”(h]”j6ah ]”h"]”Œ0mlock()/mlock2()/mlockall() system call handling”ah$]”h&]”uh1h¡hjåhžhhŸh¶h M!ubh¢)”}”(hhh]”(h§)”}”(hŒFiltering Special VMAs”h]”hŒFiltering Special VMAs”…””}”(hj hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢jRuh1h¦hj hžhhŸh¶h MRubhÇ)”}”(hŒ8mlock_fixup() filters several classes of "special" VMAs:”h]”hŒhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MéhjhžhubhÇ)”}”(hX0munlock_folio() uses the mlock pagevec to batch up work to be done under lru_lock by __munlock_folio(). __munlock_folio() decrements the folio's mlock_count, and when that reaches 0 it clears the mlocked flag and clears the unevictable flag, moving the folio from unevictable state to the inactive LRU.”h]”hX2munlock_folio() uses the mlock pagevec to batch up work to be done under lru_lock by __munlock_folio(). __munlock_folio() decrements the folioâ€™s mlock_count, and when that reaches 0 it clears the mlocked flag and clears the unevictable flag, moving the folio from unevictable state to the inactive LRU.”…””}”(hjLhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MíhjhžhubhÇ)”}”(hXgBut in practice that may not work ideally: the folio may not yet have reached "the unevictable LRU", or it may have been temporarily isolated from it. In those cases its mlock_count field is unusable and must be assumed to be 0: so that the folio will be rescued to an evictable LRU, then perhaps be mlocked again later if vmscan finds it in a VM_LOCKED VMA.”h]”hXkBut in practice that may not work ideally: the folio may not yet have reached â€œthe unevictable LRUâ€, or it may have been temporarily isolated from it. In those cases its mlock_count field is unusable and must be assumed to be 0: so that the folio will be rescued to an evictable LRU, then perhaps be mlocked again later if vmscan finds it in a VM_LOCKED VMA.”…””}”(hjZhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Móhjhžhubeh}”(h]”j$ah ]”h"]”Œ+munmap()/exit()/exec() system call handling”ah$]”h&]”uh1h¡hjåhžhhŸh¶h Máubh¢)”}”(hhh]”(h§)”}”(hŒTruncating MLOCKED Pages”h]”hŒTruncating MLOCKED Pages”…””}”(hjrhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢j@uh1h¦hjohžhhŸh¶h MûubhÇ)”}”(hŒÙFile truncation or hole punching forcibly unmaps the deleted pages from userspace; truncation even unmaps and deletes any private anonymous pages which had been Copied-On-Write from the file pages now being truncated.”h]”hŒÙFile truncation or hole punching forcibly unmaps the deleted pages from userspace; truncation even unmaps and deletes any private anonymous pages which had been Copied-On-Write from the file pages now being truncated.”…””}”(hj€hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MýhjohžhubhÇ)”}”(hX'Mlocked pages can be munlocked and deleted in this way: like with munmap(), for each PTE (or PMD) being unmapped from a VMA, folio_remove_rmap_*() calls munlock_vma_folio(), which calls munlock_folio() when the VMA is VM_LOCKED (unless it was a PTE mapping of a part of a transparent huge page).”h]”hX'Mlocked pages can be munlocked and deleted in this way: like with munmap(), for each PTE (or PMD) being unmapped from a VMA, folio_remove_rmap_*() calls munlock_vma_folio(), which calls munlock_folio() when the VMA is VM_LOCKED (unless it was a PTE mapping of a part of a transparent huge page).”…””}”(hjŽhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h MhjohžhubhÇ)”}”(hXˆHowever, if there is a racing munlock(), since mlock_vma_pages_range() starts munlocking by clearing VM_LOCKED from a VMA, before munlocking all the pages present, if one of those pages were unmapped by truncation or hole punch before mlock_pte_range() reached it, it would not be recognized as mlocked by this VMA, and would not be counted out of mlock_count. In this rare case, a page may still appear as PG_mlocked after it has been fully unmapped: and it is left to release_pages() (or __page_cache_release()) to clear it and update statistics before freeing (this event is counted in /proc/vmstat unevictable_pgs_cleared, which is usually 0).”h]”hXˆHowever, if there is a racing munlock(), since mlock_vma_pages_range() starts munlocking by clearing VM_LOCKED from a VMA, before munlocking all the pages present, if one of those pages were unmapped by truncation or hole punch before mlock_pte_range() reached it, it would not be recognized as mlocked by this VMA, and would not be counted out of mlock_count. In this rare case, a page may still appear as PG_mlocked after it has been fully unmapped: and it is left to release_pages() (or __page_cache_release()) to clear it and update statistics before freeing (this event is counted in /proc/vmstat unevictable_pgs_cleared, which is usually 0).”…””}”(hjœhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhjohžhubeh}”(h]”jFah ]”h"]”Œtruncating mlocked pages”ah$]”h&]”uh1h¡hjåhžhhŸh¶h Mûubh¢)”}”(hhh]”(h§)”}”(hŒPage Reclaim in shrink_*_list()”h]”hŒPage Reclaim in shrink_*_list()”…””}”(hj´hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”j¢jbuh1h¦hj±hžhhŸh¶h MubhÇ)”}”(hX“vmscan's shrink_active_list() culls any obviously unevictable pages - i.e. !page_evictable(page) pages - diverting those to the unevictable list. However, shrink_active_list() only sees unevictable pages that made it onto the active/inactive LRU lists. Note that these pages do not have PG_unevictable set - otherwise they would be on the unevictable list and shrink_active_list() would never see them.”h]”hX•vmscanâ€™s shrink_active_list() culls any obviously unevictable pages - i.e. !page_evictable(page) pages - diverting those to the unevictable list. However, shrink_active_list() only sees unevictable pages that made it onto the active/inactive LRU lists. Note that these pages do not have PG_unevictable set - otherwise they would be on the unevictable list and shrink_active_list() would never see them.”…””}”(hjÂhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhj±hžhubhÇ)”}”(hŒ>Some examples of these unevictable pages on the LRU lists are:”h]”hŒ>Some examples of these unevictable pages on the LRU lists are:”…””}”(hjÐhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhj±hžhubj)”}”(hXÛ(1) ramfs pages that have been placed on the LRU lists when first allocated. (2) SHM_LOCK'd shared memory pages. shmctl(SHM_LOCK) does not attempt to allocate or fault in the pages in the shared memory region. This happens when an application accesses the page the first time after SHM_LOCK'ing the segment. (3) pages still mapped into VM_LOCKED VMAs, which should be marked mlocked, but events left mlock_count too low, so they were munlocked too early. ”h]”jÔ)”}”(hhh]”(hÂ)”}”(hŒIramfs pages that have been placed on the LRU lists when first allocated. ”h]”hÇ)”}”(hŒHramfs pages that have been placed on the LRU lists when first allocated.”h]”hŒHramfs pages that have been placed on the LRU lists when first allocated.”…””}”(hjéhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhjåubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjâubhÂ)”}”(hŒåSHM_LOCK'd shared memory pages. shmctl(SHM_LOCK) does not attempt to allocate or fault in the pages in the shared memory region. This happens when an application accesses the page the first time after SHM_LOCK'ing the segment. ”h]”hÇ)”}”(hŒäSHM_LOCK'd shared memory pages. shmctl(SHM_LOCK) does not attempt to allocate or fault in the pages in the shared memory region. This happens when an application accesses the page the first time after SHM_LOCK'ing the segment.”h]”hŒèSHM_LOCKâ€™d shared memory pages. shmctl(SHM_LOCK) does not attempt to allocate or fault in the pages in the shared memory region. This happens when an application accesses the page the first time after SHM_LOCKâ€™ing the segment.”…””}”(hj hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h Mhjýubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjâubhÂ)”}”(hŒpages still mapped into VM_LOCKED VMAs, which should be marked mlocked, but events left mlock_count too low, so they were munlocked too early. ”h]”hÇ)”}”(hŒŽpages still mapped into VM_LOCKED VMAs, which should be marked mlocked, but events left mlock_count too low, so they were munlocked too early.”h]”hŒŽpages still mapped into VM_LOCKED VMAs, which should be marked mlocked, but events left mlock_count too low, so they were munlocked too early.”…””}”(hj hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M$hj ubah}”(h]”h ]”h"]”h$]”h&]”uh1hÁhjâubeh}”(h]”h ]”h"]”h$]”h&]”jjjjjjuh1jÓhjÞubah}”(h]”h ]”h"]”h$]”h&]”uh1jhŸh¶h Mhj±hžhubhÇ)”}”(hŒ·vmscan's shrink_inactive_list() and shrink_folio_list() also divert obviously unevictable pages found on the inactive lists to the appropriate memory cgroup and node unevictable list.”h]”hŒ¹vmscanâ€™s shrink_inactive_list() and shrink_folio_list() also divert obviously unevictable pages found on the inactive lists to the appropriate memory cgroup and node unevictable list.”…””}”(hj9 hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M'hj±hžhubhÇ)”}”(hXMrmap's folio_referenced_one(), called via vmscan's shrink_active_list() or shrink_folio_list(), and rmap's try_to_unmap_one() called via shrink_folio_list(), check for (3) pages still mapped into VM_LOCKED VMAs, and call mlock_vma_folio() to correct them. Such pages are culled to the unevictable list when released by the shrinker.”h]”hXSrmapâ€™s folio_referenced_one(), called via vmscanâ€™s shrink_active_list() or shrink_folio_list(), and rmapâ€™s try_to_unmap_one() called via shrink_folio_list(), check for (3) pages still mapped into VM_LOCKED VMAs, and call mlock_vma_folio() to correct them. Such pages are culled to the unevictable list when released by the shrinker.”…””}”(hjG hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhŸh¶h M+hj±hžhubeh}”(h]”jhah ]”h"]”Œpage reclaim in shrink_*_list()”ah$]”h&]”uh1h¡hjåhžhhŸh¶h Mubeh}”(h]”jÓah ]”h"]”Œ mlocked pages”ah$]”h&]”uh1h¡hh£hžhhŸh¶h KÖubeh}”(h]”Œunevictable-lru-infrastructure”ah ]”h"]”Œunevictable lru infrastructure”ah$]”h&]”uh1h¡hhhžhhŸh¶h Kubeh}”(h]”h ]”h"]”h$]”h&]”Œsource”h¶uh1hŒcurrent_source”NŒcurrent_line”NŒsettings”Œdocutils.frontend”ŒValues”“”)”}”(h¦NŒ generator”NŒ datestamp”NŒsource_link”NŒ source_url”NŒ toc_backlinks”Œentry”Œfootnote_backlinks”KŒ sectnum_xform”KŒstrip_comments”NŒstrip_elements_with_classes”NŒ strip_classes”NŒreport_level”KŒ halt_level”KŒexit_status_level”KŒdebug”NŒwarning_stream”NŒ traceback”ˆŒinput_encoding”Œ utf-8-sig”Œinput_encoding_error_handler”Œstrict”Œoutput_encoding”Œutf-8”Œoutput_encoding_error_handler”jŽ Œerror_encoding”Œutf-8”Œerror_encoding_error_handler”Œbackslashreplace”Œ language_code”Œen”Œrecord_dependencies”NŒconfig”NŒ id_prefix”hŒauto_id_prefix”Œid”Œ dump_settings”NŒdump_internals”NŒdump_transforms”NŒdump_pseudo_xml”NŒexpose_internals”NŒstrict_visitor”NŒ_disable_config”NŒ_source”h¶Œ_destination”NŒ _config_files”]”Œ7/var/lib/git/docbuild/linux/Documentation/docutils.conf”aŒfile_insertion_enabled”ˆŒraw_enabled”KŒline_length_limit”M'Œpep_references”NŒpep_base_url”Œhttps://peps.python.org/”Œpep_file_url_template”Œpep-%04d”Œrfc_references”NŒrfc_base_url”Œ&https://datatracker.ietf.org/doc/html/”Œ tab_width”KŒtrim_footnote_reference_space”‰Œsyntax_highlight”Œlong”Œsmart_quotes”ˆŒsmartquotes_locales”]”Œcharacter_level_inline_markup”‰Œdoctitle_xform”‰Œ docinfo_xform”KŒsectsubtitle_xform”‰Œ image_loading”Œlink”Œembed_stylesheet”‰Œcloak_email_addresses”ˆŒsection_self_link”‰Œenv”NubŒreporter”NŒindirect_targets”]”Œsubstitution_defs”}”Œsubstitution_names”}”Œrefnames”}”Œrefids”}”jÇ]”j½asŒnameids”}”(jh je jŽj‰jÃhÜjâhþj:jjÌj?jjÇjjaj™jƒjÛj¥j` jÓjQjòj‘ jj j6j¨ jXjø jzjHjœjnj¾jÚjàjjjlj$j®jFjY jhuŒ nametypes”}”(jh ‰jŽ‰jÃ‰jâ‰j:‰jÌ‰jˆj‰j™‰jÛ‰j` ‰jQ‰j‘ ‰j ‰j¨ ‰jø ‰jH‰jn‰jÚ‰j‰jl‰j®‰jY ‰uh}”(je h£j‰h¹hÜj‘hþjÆjj†j?j=jÇjÏjajÏjƒj#j¥jœjÓjåjòjjjTj6j” jXj jzj« jœjû j¾jKjàjqjjÝj$jjFjojhj±hÖhÍhøhïjjj9j0j[jRj}jtjŸj–jÍjÄjìjãjjj0j'jRjIjtjkj–jj¸j¯jÚjÑjüjójjj@j7jbjYuŒ footnote_refs”}”Œ citation_refs”}”Œ autofootnotes”]”Œautofootnote_refs”]”Œsymbol_footnotes”]”Œsymbol_footnote_refs”]”Œ footnotes”]”Œ citations”]”Œautofootnote_start”KŒsymbol_footnote_start”KŒ id_counter”Œcollections”ŒCounter”“”}”jœ Ks…”R”Œparse_messages”]”Œtransform_messages”]”hŒsystem_message”“”)”}”(hhh]”hÇ)”}”(hhh]”hŒ=Hyperlink target "mark-addr-space-unevict" is not referenced.”…””}”hjø sbah}”(h]”h ]”h"]”h$]”h&]”uh1hÆhjõ ubah}”(h]”h ]”h"]”h$]”h&]”Œlevel”KŒtype”ŒINFO”Œsource”h¶Œline”Kyuh1jó ubaŒtransformer”NŒinclude_log”]”Œ decoration”Nhžhub.