aboutsummaryrefslogtreecommitdiffstats
path: root/Documentation/vm
diff options
context:
space:
mode:
authorChristoph Lameter <clameter@sgi.com>2006-06-23 05:03:39 -0400
committerLinus Torvalds <torvalds@g5.osdl.org>2006-06-23 10:42:51 -0400
commit8d3c138b77f195ca0eee6fb639ae73f5ea9edb6b (patch)
tree832ac3c87484f5875b3b4afeab1de5eca1115770 /Documentation/vm
parent04e62a29bf157ce1edd168f2b71b533c80d13628 (diff)
[PATCH] page migration: Update documentation
Signed-off-by: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'Documentation/vm')
-rw-r--r--Documentation/vm/page_migration91
1 files changed, 33 insertions, 58 deletions
diff --git a/Documentation/vm/page_migration b/Documentation/vm/page_migration
index 0dd4ef30c361..0a5d5fb18854 100644
--- a/Documentation/vm/page_migration
+++ b/Documentation/vm/page_migration
@@ -62,15 +62,15 @@ A. In kernel use of migrate_pages()
62 It also prevents the swapper or other scans to encounter 62 It also prevents the swapper or other scans to encounter
63 the page. 63 the page.
64 64
652. Generate a list of newly allocates page. These pages will contain the 652. Generate a list of newly allocates pages. These pages will contain the
66 contents of the pages from the first list after page migration is 66 contents of the pages from the first list after page migration is
67 complete. 67 complete.
68 68
693. The migrate_pages() function is called which attempts 693. The migrate_pages() function is called which attempts
70 to do the migration. It returns the moved pages in the 70 to do the migration. It returns the moved pages in the
71 list specified as the third parameter and the failed 71 list specified as the third parameter and the failed
72 migrations in the fourth parameter. The first parameter 72 migrations in the fourth parameter. When the function
73 will contain the pages that could still be retried. 73 returns the first list will contain the pages that could still be retried.
74 74
754. The leftover pages of various types are returned 754. The leftover pages of various types are returned
76 to the LRU using putback_to_lru_pages() or otherwise 76 to the LRU using putback_to_lru_pages() or otherwise
@@ -93,83 +93,58 @@ Steps:
93 93
942. Insure that writeback is complete. 942. Insure that writeback is complete.
95 95
963. Make sure that the page has assigned swap cache entry if 963. Prep the new page that we want to move to. It is locked
97 it is an anonyous page. The swap cache reference is necessary
98 to preserve the information contain in the page table maps while
99 page migration occurs.
100
1014. Prep the new page that we want to move to. It is locked
102 and set to not being uptodate so that all accesses to the new 97 and set to not being uptodate so that all accesses to the new
103 page immediately lock while the move is in progress. 98 page immediately lock while the move is in progress.
104 99
1055. All the page table references to the page are either dropped (file 1004. The new page is prepped with some settings from the old page so that
106 backed pages) or converted to swap references (anonymous pages). 101 accesses to the new page will discover a page with the correct settings.
107 This should decrease the reference count. 102
1035. All the page table references to the page are converted
104 to migration entries or dropped (nonlinear vmas).
105 This decrease the mapcount of a page. If the resulting
106 mapcount is not zero then we do not migrate the page.
107 All user space processes that attempt to access the page
108 will now wait on the page lock.
108 109
1096. The radix tree lock is taken. This will cause all processes trying 1106. The radix tree lock is taken. This will cause all processes trying
110 to reestablish a pte to block on the radix tree spinlock. 111 to access the page via the mapping to block on the radix tree spinlock.
111 112
1127. The refcount of the page is examined and we back out if references remain 1137. The refcount of the page is examined and we back out if references remain
113 otherwise we know that we are the only one referencing this page. 114 otherwise we know that we are the only one referencing this page.
114 115
1158. The radix tree is checked and if it does not contain the pointer to this 1168. The radix tree is checked and if it does not contain the pointer to this
116 page then we back out because someone else modified the mapping first. 117 page then we back out because someone else modified the radix tree.
117
1189. The mapping is checked. If the mapping is gone then a truncate action may
119 be in progress and we back out.
120
12110. The new page is prepped with some settings from the old page so that
122 accesses to the new page will be discovered to have the correct settings.
123 118
12411. The radix tree is changed to point to the new page. 1199. The radix tree is changed to point to the new page.
125 120
12612. The reference count of the old page is dropped because the radix tree 12110. The reference count of the old page is dropped because the radix tree
127 reference is gone. 122 reference is gone. A reference to the new page is established because
123 the new page is referenced to by the radix tree.
128 124
12913. The radix tree lock is dropped. With that lookups become possible again 12511. The radix tree lock is dropped. With that lookups in the mapping
130 and other processes will move from spinning on the tree lock to sleeping on 126 become possible again. Processes will move from spinning on the tree_lock
131 the locked new page. 127 to sleeping on the locked new page.
132 128
13314. The page contents are copied to the new page. 12912. The page contents are copied to the new page.
134 130
13515. The remaining page flags are copied to the new page. 13113. The remaining page flags are copied to the new page.
136 132
13716. The old page flags are cleared to indicate that the page does 13314. The old page flags are cleared to indicate that the page does
138 not use any information anymore. 134 not provide any information anymore.
139 135
14017. Queued up writeback on the new page is triggered. 13615. Queued up writeback on the new page is triggered.
141 137
14218. If swap pte's were generated for the page then replace them with real 13816. If migration entries were page then replace them with real ptes. Doing
143 ptes. This will reenable access for processes not blocked by the page lock. 139 so will enable access for user space processes not already waiting for
140 the page lock.
144 141
14519. The page locks are dropped from the old and new page. 14219. The page locks are dropped from the old and new page.
146 Processes waiting on the page lock can continue. 143 Processes waiting on the page lock will redo their page faults
144 and will reach the new page.
147 145
14820. The new page is moved to the LRU and can be scanned by the swapper 14620. The new page is moved to the LRU and can be scanned by the swapper
149 etc again. 147 etc again.
150 148
151TODO list 149Christoph Lameter, May 8, 2006.
152---------
153
154- Page migration requires the use of swap handles to preserve the
155 information of the anonymous page table entries. This means that swap
156 space is reserved but never used. The maximum number of swap handles used
157 is determined by CHUNK_SIZE (see mm/mempolicy.c) per ongoing migration.
158 Reservation of pages could be avoided by having a special type of swap
159 handle that does not require swap space and that would only track the page
160 references. Something like that was proposed by Marcelo Tosatti in the
161 past (search for migration cache on lkml or linux-mm@kvack.org).
162
163- Page migration unmaps ptes for file backed pages and requires page
164 faults to reestablish these ptes. This could be optimized by somehow
165 recording the references before migration and then reestablish them later.
166 However, there are several locking challenges that have to be overcome
167 before this is possible.
168
169- Page migration generates read ptes for anonymous pages. Dirty page
170 faults are required to make the pages writable again. It may be possible
171 to generate a pte marked dirty if it is known that the page is dirty and
172 that this process has the only reference to that page.
173
174Christoph Lameter, March 8, 2006.
175 150