1MOVE_PAGES(2) Linux Programmer's Manual MOVE_PAGES(2)
2
6 move_pages - move individual pages of a process to another node
7
9 #include <numaif.h>
10 long move_pages(int pid, unsigned long count, void **pages,
12 const int *nodes, int *status, int flags);
13 Link with -lnuma.
15
17 move_pages() moves the specified pages of the process pid to the memory
18 nodes specified by nodes. The result of the move is reflected in sta‐
19 tus. The flags indicate constraints on the pages to be moved.
20 pid is the ID of the process in which pages are to be moved. To move
22 pages in another process, the caller must be privileged (CAP_SYS_NICE)
23 or the real or effective user ID of the calling process must match the
24 real or saved-set user ID of the target process. If pid is 0, then
25 move_pages() moves pages of the calling process.
26 count is the number of pages to move. It defines the size of the three
28 arrays pages, nodes, and status.
29 pages is an array of pointers to the pages that should be moved. These
31 are pointers that should be aligned to page boundaries. Addresses are
32 specified as seen by the process specified by pid.
33 nodes is an array of integers that specify the desired location for
35 each page. Each element in the array is a node number. nodes can also
36 be NULL, in which case move_pages() does not move any pages but instead
37 will return the node where each page currently resides, in the status
38 array. Obtaining the status of each page may be necessary to determine
39 pages that need to be moved.
40 status is an array of integers that return the status of each page.
42 The array only contains valid values if move_pages() did not return an
43 error.
44 flags specify what types of pages to move. MPOL_MF_MOVE means that
46 only pages that are in exclusive use by the process are to be moved.
47 MPOL_MF_MOVE_ALL means that pages shared between multiple processes can
48 also be moved. The process must be privileged (CAP_SYS_NICE) to use
49 MPOL_MF_MOVE_ALL.
50
52 The following values can be returned in each element of the status
53 array.
54 0..MAX_NUMNODES
56 Identifies the node on which the page resides.
57 -EACCES
59 The page is mapped by multiple processes and can only be moved
60 if MPOL_MF_MOVE_ALL is specified.
61 -EBUSY The page is currently busy and cannot be moved. Try again
63 later. This occurs if a page is undergoing I/O or another ker‐
64 nel subsystem is holding a reference to the page.
65 -EFAULT
67 This is a zero page or the memory area is not mapped by the
68 process.
69 -EIO Unable to write back a page. The page has to be written back in
71 order to move it since the page is dirty and the file system
72 does not provide a migration function that would allow the move
73 of dirty pages.
74 -EINVAL
76 A dirty page cannot be moved. The file system does not provide
77 a migration function and has no ability to write back pages.
78 -ENOENT
80 The page is not present.
81 -ENOMEM
83 Unable to allocate memory on target node.
84
86 On success move_pages() returns zero. On error, it returns -1, and
87 sets errno to indicate the error.
88
90 E2BIG Too many pages to move.
91 EACCES One of the target nodes is not allowed by the current cpuset.
93 EFAULT Parameter array could not be accessed.
95 EINVAL Flags other than MPOL_MF_MOVE and MPOL_MF_MOVE_ALL was specified
97 or an attempt was made to migrate pages of a kernel thread.
98 ENODEV One of the target nodes is not online.
100 ENOENT No pages were found that require moving. All pages are either
102 already on the target node, not present, had an invalid address
103 or could not be moved because they were mapped by multiple pro‐
104 cesses.
105 EPERM The caller specified MPOL_MF_MOVE_ALL without sufficient privi‐
107 leges (CAP_SYS_NICE). Or, the caller attempted to move pages of
108 a process belonging to another user but did not have privilege
109 to do so (CAP_SYS_NICE).
110 ESRCH Process does not exist.
112
114 move_pages() first appeared on Linux in version 2.6.18.
115
117 This system call is Linux-specific.
118
120 For information on library support, see numa(7).
121 Use get_mempolicy(2) with the MPOL_F_MEMS_ALLOWED flag to obtain the
123 set of nodes that are allowed by the current cpuset. Note that this
124 information is subject to change at any time by manual or automatic
125 reconfiguration of the cpuset.
126 Use of this function may result in pages whose location (node) violates
128 the memory policy established for the specified addresses (See
129 mbind(2)) and/or the specified process (See set_mempolicy(2)). That
130 is, memory policy does not constrain the destination nodes used by
131 move_pages().
132 The <numaif.h> header is not included with glibc, but requires
134 installing libnuma-devel or a similar package.
135
137 get_mempolicy(2), mbind(2), set_mempolicy(2), numa(3), numa_maps(5),
138 cpuset(7), numa(7), migratepages(8), numa_stat(8)
139
141 This page is part of release 3.25 of the Linux man-pages project. A
142 description of the project, and information about reporting bugs, can
143 be found at http://www.kernel.org/doc/man-pages/.
144Linux 2010-06-11 MOVE_PAGES(2)