1============================
2Transparent Hugepage Support
3============================
4
5This document describes design principles for Transparent Hugepage (THP)
6support and its interaction with other parts of the memory management
7system.
8
9Design principles
10=================
11
12- "graceful fallback": mm components which don't have transparent hugepage
13  knowledge fall back to breaking huge pmd mapping into table of ptes and,
14  if necessary, split a transparent hugepage. Therefore these components
15  can continue working on the regular pages or regular pte mappings.
16
17- if a hugepage allocation fails because of memory fragmentation,
18  regular pages should be gracefully allocated instead and mixed in
19  the same vma without any failure or significant delay and without
20  userland noticing
21
22- if some task quits and more hugepages become available (either
23  immediately in the buddy or through the VM), guest physical memory
24  backed by regular pages should be relocated on hugepages
25  automatically (with khugepaged)
26
27- it doesn't require memory reservation and in turn it uses hugepages
28  whenever possible (the only possible reservation here is kernelcore=
29  to avoid unmovable pages to fragment all the memory but such a tweak
30  is not specific to transparent hugepage support and it's a generic
31  feature that applies to all dynamic high order allocations in the
32  kernel)
33
34get_user_pages and follow_page
35==============================
36
37get_user_pages and follow_page if run on a hugepage, will return the
38head or tail pages as usual (exactly as they would do on
39hugetlbfs). Most GUP users will only care about the actual physical
40address of the page and its temporary pinning to release after the I/O
41is complete, so they won't ever notice the fact the page is huge. But
42if any driver is going to mangle over the page structure of the tail
43page (like for checking page->mapping or other bits that are relevant
44for the head page and not the tail page), it should be updated to jump
45to check head page instead. Taking a reference on any head/tail page would
46prevent the page from being split by anyone.
47
48.. note::
49   these aren't new constraints to the GUP API, and they match the
50   same constraints that apply to hugetlbfs too, so any driver capable
51   of handling GUP on hugetlbfs will also work fine on transparent
52   hugepage backed mappings.
53
54Graceful fallback
55=================
56
57Code walking pagetables but unaware about huge pmds can simply call
58split_huge_pmd(vma, pmd, addr) where the pmd is the one returned by
59pmd_offset. It's trivial to make the code transparent hugepage aware
60by just grepping for "pmd_offset" and adding split_huge_pmd where
61missing after pmd_offset returns the pmd. Thanks to the graceful
62fallback design, with a one liner change, you can avoid to write
63hundreds if not thousands of lines of complex code to make your code
64hugepage aware.
65
66If you're not walking pagetables but you run into a physical hugepage
67that you can't handle natively in your code, you can split it by
68calling split_huge_page(page). This is what the Linux VM does before
69it tries to swapout the hugepage for example. split_huge_page() can fail
70if the page is pinned and you must handle this correctly.
71
72Example to make mremap.c transparent hugepage aware with a one liner
73change::
74
75	diff --git a/mm/mremap.c b/mm/mremap.c
76	--- a/mm/mremap.c
77	+++ b/mm/mremap.c
78	@@ -41,6 +41,7 @@ static pmd_t *get_old_pmd(struct mm_stru
79			return NULL;
80
81		pmd = pmd_offset(pud, addr);
82	+	split_huge_pmd(vma, pmd, addr);
83		if (pmd_none_or_clear_bad(pmd))
84			return NULL;
85
86Locking in hugepage aware code
87==============================
88
89We want as much code as possible hugepage aware, as calling
90split_huge_page() or split_huge_pmd() has a cost.
91
92To make pagetable walks huge pmd aware, all you need to do is to call
93pmd_trans_huge() on the pmd returned by pmd_offset. You must hold the
94mmap_lock in read (or write) mode to be sure a huge pmd cannot be
95created from under you by khugepaged (khugepaged collapse_huge_page
96takes the mmap_lock in write mode in addition to the anon_vma lock). If
97pmd_trans_huge returns false, you just fallback in the old code
98paths. If instead pmd_trans_huge returns true, you have to take the
99page table lock (pmd_lock()) and re-run pmd_trans_huge. Taking the
100page table lock will prevent the huge pmd being converted into a
101regular pmd from under you (split_huge_pmd can run in parallel to the
102pagetable walk). If the second pmd_trans_huge returns false, you
103should just drop the page table lock and fallback to the old code as
104before. Otherwise, you can proceed to process the huge pmd and the
105hugepage natively. Once finished, you can drop the page table lock.
106
107Refcounts and transparent huge pages
108====================================
109
110Refcounting on THP is mostly consistent with refcounting on other compound
111pages:
112
113  - get_page()/put_page() and GUP operate on the folio->_refcount.
114
115  - ->_refcount in tail pages is always zero: get_page_unless_zero() never
116    succeeds on tail pages.
117
118  - map/unmap of a PMD entry for the whole THP increment/decrement
119    folio->_entire_mapcount and also increment/decrement
120    folio->_nr_pages_mapped by COMPOUND_MAPPED when _entire_mapcount
121    goes from -1 to 0 or 0 to -1.
122
123  - map/unmap of individual pages with PTE entry increment/decrement
124    page->_mapcount and also increment/decrement folio->_nr_pages_mapped
125    when page->_mapcount goes from -1 to 0 or 0 to -1 as this counts
126    the number of pages mapped by PTE.
127
128split_huge_page internally has to distribute the refcounts in the head
129page to the tail pages before clearing all PG_head/tail bits from the page
130structures. It can be done easily for refcounts taken by page table
131entries, but we don't have enough information on how to distribute any
132additional pins (i.e. from get_user_pages). split_huge_page() fails any
133requests to split pinned huge pages: it expects page count to be equal to
134the sum of mapcount of all sub-pages plus one (split_huge_page caller must
135have a reference to the head page).
136
137split_huge_page uses migration entries to stabilize page->_refcount and
138page->_mapcount of anonymous pages. File pages just get unmapped.
139
140We are safe against physical memory scanners too: the only legitimate way
141a scanner can get a reference to a page is get_page_unless_zero().
142
143All tail pages have zero ->_refcount until atomic_add(). This prevents the
144scanner from getting a reference to the tail page up to that point. After the
145atomic_add() we don't care about the ->_refcount value. We already know how
146many references should be uncharged from the head page.
147
148For head page get_page_unless_zero() will succeed and we don't mind. It's
149clear where references should go after split: it will stay on the head page.
150
151Note that split_huge_pmd() doesn't have any limitations on refcounting:
152pmd can be split at any point and never fails.
153
154Partial unmap and deferred_split_folio()
155========================================
156
157Unmapping part of THP (with munmap() or other way) is not going to free
158memory immediately. Instead, we detect that a subpage of THP is not in use
159in page_remove_rmap() and queue the THP for splitting if memory pressure
160comes. Splitting will free up unused subpages.
161
162Splitting the page right away is not an option due to locking context in
163the place where we can detect partial unmap. It also might be
164counterproductive since in many cases partial unmap happens during exit(2) if
165a THP crosses a VMA boundary.
166
167The function deferred_split_folio() is used to queue a folio for splitting.
168The splitting itself will happen when we get memory pressure via shrinker
169interface.
170