tjh.dev
Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1Page migration
2--------------
3
4Page migration allows the moving of the physical location of pages between
5nodes in a numa system while the process is running. This means that the
6virtual addresses that the process sees do not change. However, the
7system rearranges the physical location of those pages.
8
9The main intend of page migration is to reduce the latency of memory access
10by moving pages near to the processor where the process accessing that memory
11is running.
12
13Page migration allows a process to manually relocate the node on which its
14pages are located through the MF_MOVE and MF_MOVE_ALL options while setting
15a new memory policy via mbind(). The pages of process can also be relocated
16from another process using the sys_migrate_pages() function call. The
17migrate_pages function call takes two sets of nodes and moves pages of a
18process that are located on the from nodes to the destination nodes.
19Page migration functions are provided by the numactl package by Andi Kleen
20(a version later than 0.9.3 is required. Get it from
21ftp://ftp.suse.com/pub/people/ak). numactl provided libnuma which
22provides an interface similar to other numa functionality for page migration.
23cat /proc/<pid>/numa_maps allows an easy review of where the pages of
24a process are located. See also the numa_maps manpage in the numactl package.
25
26Manual migration is useful if for example the scheduler has relocated
27a process to a processor on a distant node. A batch scheduler or an
28administrator may detect the situation and move the pages of the process
29nearer to the new processor. The kernel itself does only provide
30manual page migration support. Automatic page migration may be implemented
31through user space processes that move pages. A special function call
32"move_pages" allows the moving of individual pages within a process.
33A NUMA profiler may f.e. obtain a log showing frequent off node
34accesses and may use the result to move pages to more advantageous
35locations.
36
37Larger installations usually partition the system using cpusets into
38sections of nodes. Paul Jackson has equipped cpusets with the ability to
39move pages when a task is moved to another cpuset (See ../cpusets.txt).
40Cpusets allows the automation of process locality. If a task is moved to
41a new cpuset then also all its pages are moved with it so that the
42performance of the process does not sink dramatically. Also the pages
43of processes in a cpuset are moved if the allowed memory nodes of a
44cpuset are changed.
45
46Page migration allows the preservation of the relative location of pages
47within a group of nodes for all migration techniques which will preserve a
48particular memory allocation pattern generated even after migrating a
49process. This is necessary in order to preserve the memory latencies.
50Processes will run with similar performance after migration.
51
52Page migration occurs in several steps. First a high level
53description for those trying to use migrate_pages() from the kernel
54(for userspace usage see the Andi Kleen's numactl package mentioned above)
55and then a low level description of how the low level details work.
56
57A. In kernel use of migrate_pages()
58-----------------------------------
59
601. Remove pages from the LRU.
61
62 Lists of pages to be migrated are generated by scanning over
63 pages and moving them into lists. This is done by
64 calling isolate_lru_page().
65 Calling isolate_lru_page increases the references to the page
66 so that it cannot vanish while the page migration occurs.
67 It also prevents the swapper or other scans to encounter
68 the page.
69
702. We need to have a function of type new_page_t that can be
71 passed to migrate_pages(). This function should figure out
72 how to allocate the correct new page given the old page.
73
743. The migrate_pages() function is called which attempts
75 to do the migration. It will call the function to allocate
76 the new page for each page that is considered for
77 moving.
78
79B. How migrate_pages() works
80----------------------------
81
82migrate_pages() does several passes over its list of pages. A page is moved
83if all references to a page are removable at the time. The page has
84already been removed from the LRU via isolate_lru_page() and the refcount
85is increased so that the page cannot be freed while page migration occurs.
86
87Steps:
88
891. Lock the page to be migrated
90
912. Insure that writeback is complete.
92
933. Prep the new page that we want to move to. It is locked
94 and set to not being uptodate so that all accesses to the new
95 page immediately lock while the move is in progress.
96
974. The new page is prepped with some settings from the old page so that
98 accesses to the new page will discover a page with the correct settings.
99
1005. All the page table references to the page are converted
101 to migration entries or dropped (nonlinear vmas).
102 This decrease the mapcount of a page. If the resulting
103 mapcount is not zero then we do not migrate the page.
104 All user space processes that attempt to access the page
105 will now wait on the page lock.
106
1076. The radix tree lock is taken. This will cause all processes trying
108 to access the page via the mapping to block on the radix tree spinlock.
109
1107. The refcount of the page is examined and we back out if references remain
111 otherwise we know that we are the only one referencing this page.
112
1138. The radix tree is checked and if it does not contain the pointer to this
114 page then we back out because someone else modified the radix tree.
115
1169. The radix tree is changed to point to the new page.
117
11810. The reference count of the old page is dropped because the radix tree
119 reference is gone. A reference to the new page is established because
120 the new page is referenced to by the radix tree.
121
12211. The radix tree lock is dropped. With that lookups in the mapping
123 become possible again. Processes will move from spinning on the tree_lock
124 to sleeping on the locked new page.
125
12612. The page contents are copied to the new page.
127
12813. The remaining page flags are copied to the new page.
129
13014. The old page flags are cleared to indicate that the page does
131 not provide any information anymore.
132
13315. Queued up writeback on the new page is triggered.
134
13516. If migration entries were page then replace them with real ptes. Doing
136 so will enable access for user space processes not already waiting for
137 the page lock.
138
13919. The page locks are dropped from the old and new page.
140 Processes waiting on the page lock will redo their page faults
141 and will reach the new page.
142
14320. The new page is moved to the LRU and can be scanned by the swapper
144 etc again.
145
146Christoph Lameter, May 8, 2006.
147