1.. _pagemap:
2
3=============================
4Examining Process Page Tables
5=============================
6
7pagemap is a new (as of 2.6.25) set of interfaces in the kernel that allow
8userspace programs to examine the page tables and related information by
9reading files in ``/proc``.
10
11There are four components to pagemap:
12
13 * ``/proc/pid/pagemap``.  This file lets a userspace process find out which
14   physical frame each virtual page is mapped to.  It contains one 64-bit
15   value for each virtual page, containing the following data (from
16   ``fs/proc/task_mmu.c``, above pagemap_read):
17
18    * Bits 0-54  page frame number (PFN) if present
19    * Bits 0-4   swap type if swapped
20    * Bits 5-54  swap offset if swapped
21    * Bit  55    pte is soft-dirty (see
22      :ref:`Documentation/admin-guide/mm/soft-dirty.rst <soft_dirty>`)
23    * Bit  56    page exclusively mapped (since 4.2)
24    * Bits 57-60 zero
25    * Bit  61    page is file-page or shared-anon (since 3.5)
26    * Bit  62    page swapped
27    * Bit  63    page present
28
29   Since Linux 4.0 only users with the CAP_SYS_ADMIN capability can get PFNs.
30   In 4.0 and 4.1 opens by unprivileged fail with -EPERM.  Starting from
31   4.2 the PFN field is zeroed if the user does not have CAP_SYS_ADMIN.
32   Reason: information about PFNs helps in exploiting Rowhammer vulnerability.
33
34   If the page is not present but in swap, then the PFN contains an
35   encoding of the swap file number and the page's offset into the
36   swap. Unmapped pages return a null PFN. This allows determining
37   precisely which pages are mapped (or in swap) and comparing mapped
38   pages between processes.
39
40   Efficient users of this interface will use ``/proc/pid/maps`` to
41   determine which areas of memory are actually mapped and llseek to
42   skip over unmapped regions.
43
44 * ``/proc/kpagecount``.  This file contains a 64-bit count of the number of
45   times each page is mapped, indexed by PFN.
46
47The page-types tool in the tools/vm directory can be used to query the
48number of times a page is mapped.
49
50 * ``/proc/kpageflags``.  This file contains a 64-bit set of flags for each
51   page, indexed by PFN.
52
53   The flags are (from ``fs/proc/page.c``, above kpageflags_read):
54
55    0. LOCKED
56    1. ERROR
57    2. REFERENCED
58    3. UPTODATE
59    4. DIRTY
60    5. LRU
61    6. ACTIVE
62    7. SLAB
63    8. WRITEBACK
64    9. RECLAIM
65    10. BUDDY
66    11. MMAP
67    12. ANON
68    13. SWAPCACHE
69    14. SWAPBACKED
70    15. COMPOUND_HEAD
71    16. COMPOUND_TAIL
72    17. HUGE
73    18. UNEVICTABLE
74    19. HWPOISON
75    20. NOPAGE
76    21. KSM
77    22. THP
78    23. BALLOON
79    24. ZERO_PAGE
80    25. IDLE
81
82 * ``/proc/kpagecgroup``.  This file contains a 64-bit inode number of the
83   memory cgroup each page is charged to, indexed by PFN. Only available when
84   CONFIG_MEMCG is set.
85
86Short descriptions to the page flags
87====================================
88
890 - LOCKED
90   page is being locked for exclusive access, e.g. by undergoing read/write IO
917 - SLAB
92   page is managed by the SLAB/SLOB/SLUB/SLQB kernel memory allocator
93   When compound page is used, SLUB/SLQB will only set this flag on the head
94   page; SLOB will not flag it at all.
9510 - BUDDY
96    a free memory block managed by the buddy system allocator
97    The buddy system organizes free memory in blocks of various orders.
98    An order N block has 2^N physically contiguous pages, with the BUDDY flag
99    set for and _only_ for the first page.
10015 - COMPOUND_HEAD
101    A compound page with order N consists of 2^N physically contiguous pages.
102    A compound page with order 2 takes the form of "HTTT", where H donates its
103    head page and T donates its tail page(s).  The major consumers of compound
104    pages are hugeTLB pages
105    (:ref:`Documentation/admin-guide/mm/hugetlbpage.rst <hugetlbpage>`),
106    the SLUB etc.  memory allocators and various device drivers.
107    However in this interface, only huge/giga pages are made visible
108    to end users.
10916 - COMPOUND_TAIL
110    A compound page tail (see description above).
11117 - HUGE
112    this is an integral part of a HugeTLB page
11319 - HWPOISON
114    hardware detected memory corruption on this page: don't touch the data!
11520 - NOPAGE
116    no page frame exists at the requested address
11721 - KSM
118    identical memory pages dynamically shared between one or more processes
11922 - THP
120    contiguous pages which construct transparent hugepages
12123 - BALLOON
122    balloon compaction page
12324 - ZERO_PAGE
124    zero page for pfn_zero or huge_zero page
12525 - IDLE
126    page has not been accessed since it was marked idle (see
127    :ref:`Documentation/admin-guide/mm/idle_page_tracking.rst <idle_page_tracking>`).
128    Note that this flag may be stale in case the page was accessed via
129    a PTE. To make sure the flag is up-to-date one has to read
130    ``/sys/kernel/mm/page_idle/bitmap`` first.
131
132IO related page flags
133---------------------
134
1351 - ERROR
136   IO error occurred
1373 - UPTODATE
138   page has up-to-date data
139   ie. for file backed page: (in-memory data revision >= on-disk one)
1404 - DIRTY
141   page has been written to, hence contains new data
142   i.e. for file backed page: (in-memory data revision >  on-disk one)
1438 - WRITEBACK
144   page is being synced to disk
145
146LRU related page flags
147----------------------
148
1495 - LRU
150   page is in one of the LRU lists
1516 - ACTIVE
152   page is in the active LRU list
15318 - UNEVICTABLE
154   page is in the unevictable (non-)LRU list It is somehow pinned and
155   not a candidate for LRU page reclaims, e.g. ramfs pages,
156   shmctl(SHM_LOCK) and mlock() memory segments
1572 - REFERENCED
158   page has been referenced since last LRU list enqueue/requeue
1599 - RECLAIM
160   page will be reclaimed soon after its pageout IO completed
16111 - MMAP
162   a memory mapped page
16312 - ANON
164   a memory mapped page that is not part of a file
16513 - SWAPCACHE
166   page is mapped to swap space, i.e. has an associated swap entry
16714 - SWAPBACKED
168   page is backed by swap/RAM
169
170The page-types tool in the tools/vm directory can be used to query the
171above flags.
172
173Using pagemap to do something useful
174====================================
175
176The general procedure for using pagemap to find out about a process' memory
177usage goes like this:
178
179 1. Read ``/proc/pid/maps`` to determine which parts of the memory space are
180    mapped to what.
181 2. Select the maps you are interested in -- all of them, or a particular
182    library, or the stack or the heap, etc.
183 3. Open ``/proc/pid/pagemap`` and seek to the pages you would like to examine.
184 4. Read a u64 for each page from pagemap.
185 5. Open ``/proc/kpagecount`` and/or ``/proc/kpageflags``.  For each PFN you
186    just read, seek to that entry in the file, and read the data you want.
187
188For example, to find the "unique set size" (USS), which is the amount of
189memory that a process is using that is not shared with any other process,
190you can go through every map in the process, find the PFNs, look those up
191in kpagecount, and tally up the number of pages that are only referenced
192once.
193
194Other notes
195===========
196
197Reading from any of the files will return -EINVAL if you are not starting
198the read on an 8-byte boundary (e.g., if you sought an odd number of bytes
199into the file), or if the size of the read is not a multiple of 8 bytes.
200
201Before Linux 3.11 pagemap bits 55-60 were used for "page-shift" (which is
202always 12 at most architectures). Since Linux 3.11 their meaning changes
203after first clear of soft-dirty bits. Since Linux 4.2 they are used for
204flags unconditionally.
205