Blame view
mm/page_alloc.c
159 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
/* * linux/mm/page_alloc.c * * Manages the free list, the system allocates free pages here. * Note that kmalloc() lives in slab.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds * Swap reorganised 29.12.95, Stephen Tweedie * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999 * Reshaped it to be a zoned allocator, Ingo Molnar, Red Hat, 1999 * Discontiguous memory support, Kanoj Sarcar, SGI, Nov 1999 * Zone balancing, Kanoj Sarcar, SGI, Jan 2000 * Per cpu hot/cold page lists, bulk allocation, Martin J. Bligh, Sept 2002 * (lots of bits borrowed from Ingo Molnar & Andrew Morton) */ |
1da177e4c Linux-2.6.12-rc2 |
16 17 18 19 20 |
#include <linux/stddef.h> #include <linux/mm.h> #include <linux/swap.h> #include <linux/interrupt.h> #include <linux/pagemap.h> |
10ed273f5 zlc_setup(): hand... |
21 |
#include <linux/jiffies.h> |
1da177e4c Linux-2.6.12-rc2 |
22 |
#include <linux/bootmem.h> |
edbe7d23b memblock: Add fin... |
23 |
#include <linux/memblock.h> |
1da177e4c Linux-2.6.12-rc2 |
24 |
#include <linux/compiler.h> |
9f1583339 [PATCH] use add_t... |
25 |
#include <linux/kernel.h> |
b1eeab676 kmemcheck: add ho... |
26 |
#include <linux/kmemcheck.h> |
1da177e4c Linux-2.6.12-rc2 |
27 28 29 30 31 |
#include <linux/module.h> #include <linux/suspend.h> #include <linux/pagevec.h> #include <linux/blkdev.h> #include <linux/slab.h> |
a238ab5b0 mm: break out pag... |
32 |
#include <linux/ratelimit.h> |
5a3135c2e oom: move prototy... |
33 |
#include <linux/oom.h> |
1da177e4c Linux-2.6.12-rc2 |
34 35 36 37 38 |
#include <linux/notifier.h> #include <linux/topology.h> #include <linux/sysctl.h> #include <linux/cpu.h> #include <linux/cpuset.h> |
bdc8cb984 [PATCH] memory ho... |
39 |
#include <linux/memory_hotplug.h> |
1da177e4c Linux-2.6.12-rc2 |
40 41 |
#include <linux/nodemask.h> #include <linux/vmalloc.h> |
a6cccdc36 mm, mem-hotplug: ... |
42 |
#include <linux/vmstat.h> |
4be38e351 [PATCH] mm: move ... |
43 |
#include <linux/mempolicy.h> |
6811378e7 [PATCH] wait_tabl... |
44 |
#include <linux/stop_machine.h> |
c713216de [PATCH] Introduce... |
45 46 |
#include <linux/sort.h> #include <linux/pfn.h> |
3fcfab16c [PATCH] separate ... |
47 |
#include <linux/backing-dev.h> |
933e312e7 [PATCH] fault-inj... |
48 |
#include <linux/fault-inject.h> |
a5d76b54a memory unplug: pa... |
49 |
#include <linux/page-isolation.h> |
52d4b9ac0 memcg: allocate a... |
50 |
#include <linux/page_cgroup.h> |
3ac7fe5a4 infrastructure to... |
51 |
#include <linux/debugobjects.h> |
dbb1f81ca kmemleak: Add kme... |
52 |
#include <linux/kmemleak.h> |
925cc71e5 mm: Add notifier ... |
53 |
#include <linux/memory.h> |
56de7263f mm: compaction: d... |
54 |
#include <linux/compaction.h> |
0d3d062a6 tracing, page-all... |
55 |
#include <trace/events/kmem.h> |
718a38211 mm: introduce dum... |
56 |
#include <linux/ftrace_event.h> |
f212ad7cf memcg: add memcg ... |
57 |
#include <linux/memcontrol.h> |
268bb0ce3 sanitize <linux/p... |
58 |
#include <linux/prefetch.h> |
1da177e4c Linux-2.6.12-rc2 |
59 60 |
#include <asm/tlbflush.h> |
ac924c603 [PATCH] setup_per... |
61 |
#include <asm/div64.h> |
1da177e4c Linux-2.6.12-rc2 |
62 |
#include "internal.h" |
728120192 numa: add generic... |
63 64 65 66 |
#ifdef CONFIG_USE_PERCPU_NUMA_NODE_ID DEFINE_PER_CPU(int, numa_node); EXPORT_PER_CPU_SYMBOL(numa_node); #endif |
7aac78988 numa: introduce n... |
67 68 69 70 71 72 73 74 75 76 |
#ifdef CONFIG_HAVE_MEMORYLESS_NODES /* * N.B., Do NOT reference the '_numa_mem_' per cpu variable directly. * It will not be defined when CONFIG_HAVE_MEMORYLESS_NODES is not defined. * Use the accessor functions set_numa_mem(), numa_mem_id() and cpu_to_mem() * defined in <linux/topology.h>. */ DEFINE_PER_CPU(int, _numa_mem_); /* Kernel "local memory" node */ EXPORT_PER_CPU_SYMBOL(_numa_mem_); #endif |
1da177e4c Linux-2.6.12-rc2 |
77 |
/* |
138089107 Memoryless nodes:... |
78 |
* Array of node states. |
1da177e4c Linux-2.6.12-rc2 |
79 |
*/ |
138089107 Memoryless nodes:... |
80 81 82 83 84 85 86 87 88 89 90 91 |
nodemask_t node_states[NR_NODE_STATES] __read_mostly = { [N_POSSIBLE] = NODE_MASK_ALL, [N_ONLINE] = { { [0] = 1UL } }, #ifndef CONFIG_NUMA [N_NORMAL_MEMORY] = { { [0] = 1UL } }, #ifdef CONFIG_HIGHMEM [N_HIGH_MEMORY] = { { [0] = 1UL } }, #endif [N_CPU] = { { [0] = 1UL } }, #endif /* NUMA */ }; EXPORT_SYMBOL(node_states); |
6c231b7ba [PATCH] Additions... |
92 |
unsigned long totalram_pages __read_mostly; |
cb45b0e96 [PATCH] overcommi... |
93 |
unsigned long totalreserve_pages __read_mostly; |
8ad4b1fb8 [PATCH] Make high... |
94 |
int percpu_pagelist_fraction; |
dcce284a2 mm: Extend gfp ma... |
95 |
gfp_t gfp_allowed_mask __read_mostly = GFP_BOOT_MASK; |
1da177e4c Linux-2.6.12-rc2 |
96 |
|
452aa6999 mm/pm: force GFP_... |
97 98 99 100 101 102 103 104 105 |
#ifdef CONFIG_PM_SLEEP /* * The following functions are used by the suspend/hibernate code to temporarily * change gfp_allowed_mask in order to avoid using I/O during memory allocations * while devices are suspended. To avoid races with the suspend/hibernate code, * they should always be called with pm_mutex held (gfp_allowed_mask also should * only be modified with pm_mutex held, unless the suspend/hibernate code is * guaranteed not to run in parallel with that modification). */ |
c9e664f1f PM / Hibernate: F... |
106 107 108 109 |
static gfp_t saved_gfp_mask; void pm_restore_gfp_mask(void) |
452aa6999 mm/pm: force GFP_... |
110 111 |
{ WARN_ON(!mutex_is_locked(&pm_mutex)); |
c9e664f1f PM / Hibernate: F... |
112 113 114 115 |
if (saved_gfp_mask) { gfp_allowed_mask = saved_gfp_mask; saved_gfp_mask = 0; } |
452aa6999 mm/pm: force GFP_... |
116 |
} |
c9e664f1f PM / Hibernate: F... |
117 |
void pm_restrict_gfp_mask(void) |
452aa6999 mm/pm: force GFP_... |
118 |
{ |
452aa6999 mm/pm: force GFP_... |
119 |
WARN_ON(!mutex_is_locked(&pm_mutex)); |
c9e664f1f PM / Hibernate: F... |
120 121 122 |
WARN_ON(saved_gfp_mask); saved_gfp_mask = gfp_allowed_mask; gfp_allowed_mask &= ~GFP_IOFS; |
452aa6999 mm/pm: force GFP_... |
123 124 |
} #endif /* CONFIG_PM_SLEEP */ |
d9c234005 Do not depend on ... |
125 126 127 |
#ifdef CONFIG_HUGETLB_PAGE_SIZE_VARIABLE int pageblock_order __read_mostly; #endif |
d98c7a098 [PATCH] compound ... |
128 |
static void __free_pages_ok(struct page *page, unsigned int order); |
a226f6c89 [PATCH] FRV: Clea... |
129 |
|
1da177e4c Linux-2.6.12-rc2 |
130 131 132 133 134 135 136 |
/* * results with 256, 32 in the lowmem_reserve sysctl: * 1G machine -> (16M dma, 800M-16M normal, 1G-800M high) * 1G machine -> (16M dma, 784M normal, 224M high) * NORMAL allocation will leave 784M/256 of ram reserved in the ZONE_DMA * HIGHMEM allocation will leave 224M/32 of ram reserved in ZONE_NORMAL * HIGHMEM allocation will (224M+784M)/256 of ram reserved in ZONE_DMA |
a2f1b4249 [PATCH] x86_64: A... |
137 138 139 |
* * TBD: should special case ZONE_DMA32 machines here - in those we normally * don't need any ZONE_NORMAL reservation |
1da177e4c Linux-2.6.12-rc2 |
140 |
*/ |
2f1b62486 [PATCH] reduce MA... |
141 |
int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES-1] = { |
4b51d6698 [PATCH] optional ... |
142 |
#ifdef CONFIG_ZONE_DMA |
2f1b62486 [PATCH] reduce MA... |
143 |
256, |
4b51d6698 [PATCH] optional ... |
144 |
#endif |
fb0e7942b [PATCH] reduce MA... |
145 |
#ifdef CONFIG_ZONE_DMA32 |
2f1b62486 [PATCH] reduce MA... |
146 |
256, |
fb0e7942b [PATCH] reduce MA... |
147 |
#endif |
e53ef38d0 [PATCH] reduce MA... |
148 |
#ifdef CONFIG_HIGHMEM |
2a1e274ac Create the ZONE_M... |
149 |
32, |
e53ef38d0 [PATCH] reduce MA... |
150 |
#endif |
2a1e274ac Create the ZONE_M... |
151 |
32, |
2f1b62486 [PATCH] reduce MA... |
152 |
}; |
1da177e4c Linux-2.6.12-rc2 |
153 154 |
EXPORT_SYMBOL(totalram_pages); |
1da177e4c Linux-2.6.12-rc2 |
155 |
|
15ad7cdcf [PATCH] struct se... |
156 |
static char * const zone_names[MAX_NR_ZONES] = { |
4b51d6698 [PATCH] optional ... |
157 |
#ifdef CONFIG_ZONE_DMA |
2f1b62486 [PATCH] reduce MA... |
158 |
"DMA", |
4b51d6698 [PATCH] optional ... |
159 |
#endif |
fb0e7942b [PATCH] reduce MA... |
160 |
#ifdef CONFIG_ZONE_DMA32 |
2f1b62486 [PATCH] reduce MA... |
161 |
"DMA32", |
fb0e7942b [PATCH] reduce MA... |
162 |
#endif |
2f1b62486 [PATCH] reduce MA... |
163 |
"Normal", |
e53ef38d0 [PATCH] reduce MA... |
164 |
#ifdef CONFIG_HIGHMEM |
2a1e274ac Create the ZONE_M... |
165 |
"HighMem", |
e53ef38d0 [PATCH] reduce MA... |
166 |
#endif |
2a1e274ac Create the ZONE_M... |
167 |
"Movable", |
2f1b62486 [PATCH] reduce MA... |
168 |
}; |
1da177e4c Linux-2.6.12-rc2 |
169 |
int min_free_kbytes = 1024; |
2c85f51d2 mm: also use allo... |
170 171 |
static unsigned long __meminitdata nr_kernel_pages; static unsigned long __meminitdata nr_all_pages; |
a3142c8e1 Fix section misma... |
172 |
static unsigned long __meminitdata dma_reserve; |
1da177e4c Linux-2.6.12-rc2 |
173 |
|
c713216de [PATCH] Introduce... |
174 175 |
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP /* |
183ff22bb spelling fixes: mm/ |
176 |
* MAX_ACTIVE_REGIONS determines the maximum number of distinct |
c713216de [PATCH] Introduce... |
177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 |
* ranges of memory (RAM) that may be registered with add_active_range(). * Ranges passed to add_active_range() will be merged if possible * so the number of times add_active_range() can be called is * related to the number of nodes and the number of holes */ #ifdef CONFIG_MAX_ACTIVE_REGIONS /* Allow an architecture to set MAX_ACTIVE_REGIONS to save memory */ #define MAX_ACTIVE_REGIONS CONFIG_MAX_ACTIVE_REGIONS #else #if MAX_NUMNODES >= 32 /* If there can be many nodes, allow up to 50 holes per node */ #define MAX_ACTIVE_REGIONS (MAX_NUMNODES*50) #else /* By default, allow up to 256 distinct regions */ #define MAX_ACTIVE_REGIONS 256 #endif #endif |
98011f569 mm: fix improper ... |
194 195 196 197 |
static struct node_active_region __meminitdata early_node_map[MAX_ACTIVE_REGIONS]; static int __meminitdata nr_nodemap_entries; static unsigned long __meminitdata arch_zone_lowest_possible_pfn[MAX_NR_ZONES]; static unsigned long __meminitdata arch_zone_highest_possible_pfn[MAX_NR_ZONES]; |
b69a7288e mm/page_alloc.c: ... |
198 |
static unsigned long __initdata required_kernelcore; |
484f51f82 mm/page_alloc.c: ... |
199 |
static unsigned long __initdata required_movablecore; |
b69a7288e mm/page_alloc.c: ... |
200 |
static unsigned long __meminitdata zone_movable_pfn[MAX_NUMNODES]; |
2a1e274ac Create the ZONE_M... |
201 202 203 204 |
/* movable_zone is the "real" zone pages in ZONE_MOVABLE are taken from */ int movable_zone; EXPORT_SYMBOL(movable_zone); |
c713216de [PATCH] Introduce... |
205 |
#endif /* CONFIG_ARCH_POPULATES_NODE_MAP */ |
418508c13 fix unused setup_... |
206 207 |
#if MAX_NUMNODES > 1 int nr_node_ids __read_mostly = MAX_NUMNODES; |
62bc62a87 page allocator: u... |
208 |
int nr_online_nodes __read_mostly = 1; |
418508c13 fix unused setup_... |
209 |
EXPORT_SYMBOL(nr_node_ids); |
62bc62a87 page allocator: u... |
210 |
EXPORT_SYMBOL(nr_online_nodes); |
418508c13 fix unused setup_... |
211 |
#endif |
9ef9acb05 Do not group page... |
212 |
int page_group_by_mobility_disabled __read_mostly; |
b2a0ac887 Split the free li... |
213 214 |
static void set_pageblock_migratetype(struct page *page, int migratetype) { |
49255c619 page allocator: m... |
215 216 217 |
if (unlikely(page_group_by_mobility_disabled)) migratetype = MIGRATE_UNMOVABLE; |
b2a0ac887 Split the free li... |
218 219 220 |
set_pageblock_flags_group(page, (unsigned long)migratetype, PB_migrate, PB_migrate_end); } |
7f33d49a2 mm, PM/Freezer: D... |
221 |
bool oom_killer_disabled __read_mostly; |
13e7444b0 [PATCH] mm: remov... |
222 |
#ifdef CONFIG_DEBUG_VM |
c6a57e19e [PATCH] memory ho... |
223 |
static int page_outside_zone_boundaries(struct zone *zone, struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
224 |
{ |
bdc8cb984 [PATCH] memory ho... |
225 226 227 |
int ret = 0; unsigned seq; unsigned long pfn = page_to_pfn(page); |
c6a57e19e [PATCH] memory ho... |
228 |
|
bdc8cb984 [PATCH] memory ho... |
229 230 231 232 233 234 235 236 237 |
do { seq = zone_span_seqbegin(zone); if (pfn >= zone->zone_start_pfn + zone->spanned_pages) ret = 1; else if (pfn < zone->zone_start_pfn) ret = 1; } while (zone_span_seqretry(zone, seq)); return ret; |
c6a57e19e [PATCH] memory ho... |
238 239 240 241 |
} static int page_is_consistent(struct zone *zone, struct page *page) { |
14e072984 add pfn_valid_wit... |
242 |
if (!pfn_valid_within(page_to_pfn(page))) |
c6a57e19e [PATCH] memory ho... |
243 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
244 |
if (zone != page_zone(page)) |
c6a57e19e [PATCH] memory ho... |
245 246 247 248 249 250 251 252 253 254 |
return 0; return 1; } /* * Temporary debugging check for pages not lying within a given zone. */ static int bad_range(struct zone *zone, struct page *page) { if (page_outside_zone_boundaries(zone, page)) |
1da177e4c Linux-2.6.12-rc2 |
255 |
return 1; |
c6a57e19e [PATCH] memory ho... |
256 257 |
if (!page_is_consistent(zone, page)) return 1; |
1da177e4c Linux-2.6.12-rc2 |
258 259 |
return 0; } |
13e7444b0 [PATCH] mm: remov... |
260 261 262 263 264 265 |
#else static inline int bad_range(struct zone *zone, struct page *page) { return 0; } #endif |
224abf92b [PATCH] mm: bad_p... |
266 |
static void bad_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
267 |
{ |
d936cf9b3 badpage: ratelimi... |
268 269 270 |
static unsigned long resume; static unsigned long nr_shown; static unsigned long nr_unshown; |
2a7684a23 HWPOISON: check a... |
271 272 |
/* Don't complain about poisoned pages */ if (PageHWPoison(page)) { |
ef2b4b95a mm: PageBuddy and... |
273 |
reset_page_mapcount(page); /* remove PageBuddy */ |
2a7684a23 HWPOISON: check a... |
274 275 |
return; } |
d936cf9b3 badpage: ratelimi... |
276 277 278 279 280 281 282 283 284 285 |
/* * Allow a burst of 60 reports, then keep quiet for that minute; * or allow a steady drip of one report per second. */ if (nr_shown == 60) { if (time_before(jiffies, resume)) { nr_unshown++; goto out; } if (nr_unshown) { |
1e9e63650 badpage: KERN_ALE... |
286 287 288 |
printk(KERN_ALERT "BUG: Bad page state: %lu messages suppressed ", |
d936cf9b3 badpage: ratelimi... |
289 290 291 292 293 294 295 |
nr_unshown); nr_unshown = 0; } nr_shown = 0; } if (nr_shown++ == 0) resume = jiffies + 60 * HZ; |
1e9e63650 badpage: KERN_ALE... |
296 297 |
printk(KERN_ALERT "BUG: Bad page state in process %s pfn:%05lx ", |
3dc147414 badpage: replace ... |
298 |
current->comm, page_to_pfn(page)); |
718a38211 mm: introduce dum... |
299 |
dump_page(page); |
3dc147414 badpage: replace ... |
300 |
|
1da177e4c Linux-2.6.12-rc2 |
301 |
dump_stack(); |
d936cf9b3 badpage: ratelimi... |
302 |
out: |
8cc3b3922 badpage: keep any... |
303 |
/* Leave bad fields for debug, except PageBuddy could make trouble */ |
ef2b4b95a mm: PageBuddy and... |
304 |
reset_page_mapcount(page); /* remove PageBuddy */ |
9f1583339 [PATCH] use add_t... |
305 |
add_taint(TAINT_BAD_PAGE); |
1da177e4c Linux-2.6.12-rc2 |
306 |
} |
1da177e4c Linux-2.6.12-rc2 |
307 308 309 310 311 312 313 314 315 316 |
/* * Higher-order pages are called "compound pages". They are structured thusly: * * The first PAGE_SIZE page is called the "head page". * * The remaining PAGE_SIZE pages are called "tail pages". * * All pages have PG_compound set. All pages have their ->private pointing at * the head page (even the head page has this). * |
41d78ba55 [PATCH] compound ... |
317 318 319 |
* The first tail page's ->lru.next holds the address of the compound page's * put_page() function. Its ->lru.prev holds the order of allocation. * This usage means that zero-order pages may not be compound. |
1da177e4c Linux-2.6.12-rc2 |
320 |
*/ |
d98c7a098 [PATCH] compound ... |
321 322 323 |
static void free_compound_page(struct page *page) { |
d85f33855 Make page->privat... |
324 |
__free_pages_ok(page, compound_order(page)); |
d98c7a098 [PATCH] compound ... |
325 |
} |
01ad1c082 mm: export prep_c... |
326 |
void prep_compound_page(struct page *page, unsigned long order) |
1da177e4c Linux-2.6.12-rc2 |
327 328 329 |
{ int i; int nr_pages = 1 << order; |
18229df5b hugetlb: pull gig... |
330 331 332 333 334 335 336 337 338 339 340 |
set_compound_page_dtor(page, free_compound_page); set_compound_order(page, order); __SetPageHead(page); for (i = 1; i < nr_pages; i++) { struct page *p = page + i; __SetPageTail(p); p->first_page = page; } } |
59ff42163 thp: comment remi... |
341 |
/* update __split_huge_page_refcount if you change this function */ |
8cc3b3922 badpage: keep any... |
342 |
static int destroy_compound_page(struct page *page, unsigned long order) |
1da177e4c Linux-2.6.12-rc2 |
343 344 345 |
{ int i; int nr_pages = 1 << order; |
8cc3b3922 badpage: keep any... |
346 |
int bad = 0; |
1da177e4c Linux-2.6.12-rc2 |
347 |
|
8cc3b3922 badpage: keep any... |
348 349 |
if (unlikely(compound_order(page) != order) || unlikely(!PageHead(page))) { |
224abf92b [PATCH] mm: bad_p... |
350 |
bad_page(page); |
8cc3b3922 badpage: keep any... |
351 352 |
bad++; } |
1da177e4c Linux-2.6.12-rc2 |
353 |
|
6d7779538 mm: optimize comp... |
354 |
__ClearPageHead(page); |
8cc3b3922 badpage: keep any... |
355 |
|
18229df5b hugetlb: pull gig... |
356 357 |
for (i = 1; i < nr_pages; i++) { struct page *p = page + i; |
1da177e4c Linux-2.6.12-rc2 |
358 |
|
e713a21d8 trivial: Fix dubi... |
359 |
if (unlikely(!PageTail(p) || (p->first_page != page))) { |
224abf92b [PATCH] mm: bad_p... |
360 |
bad_page(page); |
8cc3b3922 badpage: keep any... |
361 362 |
bad++; } |
d85f33855 Make page->privat... |
363 |
__ClearPageTail(p); |
1da177e4c Linux-2.6.12-rc2 |
364 |
} |
8cc3b3922 badpage: keep any... |
365 366 |
return bad; |
1da177e4c Linux-2.6.12-rc2 |
367 |
} |
1da177e4c Linux-2.6.12-rc2 |
368 |
|
17cf44064 [PATCH] mm: clean... |
369 370 371 |
static inline void prep_zero_page(struct page *page, int order, gfp_t gfp_flags) { int i; |
6626c5d53 [PATCH] mm: prep_... |
372 373 374 375 |
/* * clear_highpage() will use KM_USER0, so it's a bug to use __GFP_ZERO * and __GFP_HIGHMEM from hard or soft interrupt context. */ |
725d704ec [PATCH] mm: VM_BU... |
376 |
VM_BUG_ON((gfp_flags & __GFP_HIGHMEM) && in_interrupt()); |
17cf44064 [PATCH] mm: clean... |
377 378 379 |
for (i = 0; i < (1 << order); i++) clear_highpage(page + i); } |
6aa3001b2 [PATCH] page_allo... |
380 381 |
static inline void set_page_order(struct page *page, int order) { |
4c21e2f24 [PATCH] mm: split... |
382 |
set_page_private(page, order); |
676165a8a [PATCH] Fix buddy... |
383 |
__SetPageBuddy(page); |
1da177e4c Linux-2.6.12-rc2 |
384 385 386 387 |
} static inline void rmv_page_order(struct page *page) { |
676165a8a [PATCH] Fix buddy... |
388 |
__ClearPageBuddy(page); |
4c21e2f24 [PATCH] mm: split... |
389 |
set_page_private(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 |
} /* * Locate the struct page for both the matching buddy in our * pair (buddy1) and the combined O(n+1) page they form (page). * * 1) Any buddy B1 will have an order O twin B2 which satisfies * the following equation: * B2 = B1 ^ (1 << O) * For example, if the starting buddy (buddy2) is #8 its order * 1 buddy is #10: * B2 = 8 ^ (1 << 1) = 8 ^ 2 = 10 * * 2) Any buddy B will have an order O+1 parent P which * satisfies the following equation: * P = B & ~(1 << O) * |
d6e05edc5 spelling fixes |
407 |
* Assumption: *_mem_map is contiguous at least up to MAX_ORDER |
1da177e4c Linux-2.6.12-rc2 |
408 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
409 |
static inline unsigned long |
43506fad2 mm/page_alloc.c: ... |
410 |
__find_buddy_index(unsigned long page_idx, unsigned int order) |
1da177e4c Linux-2.6.12-rc2 |
411 |
{ |
43506fad2 mm/page_alloc.c: ... |
412 |
return page_idx ^ (1 << order); |
1da177e4c Linux-2.6.12-rc2 |
413 414 415 416 417 |
} /* * This function checks whether a page is free && is the buddy * we can do coalesce a page and its buddy if |
13e7444b0 [PATCH] mm: remov... |
418 |
* (a) the buddy is not in a hole && |
676165a8a [PATCH] Fix buddy... |
419 |
* (b) the buddy is in the buddy system && |
cb2b95e1c [PATCH] zone hand... |
420 421 |
* (c) a page and its buddy have the same order && * (d) a page and its buddy are in the same zone. |
676165a8a [PATCH] Fix buddy... |
422 |
* |
5f24ce5fd thp: remove PG_buddy |
423 424 |
* For recording whether a page is in the buddy system, we set ->_mapcount -2. * Setting, clearing, and testing _mapcount -2 is serialized by zone->lock. |
1da177e4c Linux-2.6.12-rc2 |
425 |
* |
676165a8a [PATCH] Fix buddy... |
426 |
* For recording page's order, we use page_private(page). |
1da177e4c Linux-2.6.12-rc2 |
427 |
*/ |
cb2b95e1c [PATCH] zone hand... |
428 429 |
static inline int page_is_buddy(struct page *page, struct page *buddy, int order) |
1da177e4c Linux-2.6.12-rc2 |
430 |
{ |
14e072984 add pfn_valid_wit... |
431 |
if (!pfn_valid_within(page_to_pfn(buddy))) |
13e7444b0 [PATCH] mm: remov... |
432 |
return 0; |
13e7444b0 [PATCH] mm: remov... |
433 |
|
cb2b95e1c [PATCH] zone hand... |
434 435 436 437 |
if (page_zone_id(page) != page_zone_id(buddy)) return 0; if (PageBuddy(buddy) && page_order(buddy) == order) { |
a3af9c389 page allocator: d... |
438 |
VM_BUG_ON(page_count(buddy) != 0); |
6aa3001b2 [PATCH] page_allo... |
439 |
return 1; |
676165a8a [PATCH] Fix buddy... |
440 |
} |
6aa3001b2 [PATCH] page_allo... |
441 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 |
} /* * Freeing function for a buddy system allocator. * * The concept of a buddy system is to maintain direct-mapped table * (containing bit values) for memory blocks of various "orders". * The bottom level table contains the map for the smallest allocatable * units of memory (here, pages), and each level above it describes * pairs of units from the levels below, hence, "buddies". * At a high level, all that happens here is marking the table entry * at the bottom level available, and propagating the changes upward * as necessary, plus some accounting needed to play nicely with other * parts of the VM system. * At each level, we keep a list of pages, which are heads of continuous |
5f24ce5fd thp: remove PG_buddy |
457 |
* free pages of length of (1 << order) and marked with _mapcount -2. Page's |
4c21e2f24 [PATCH] mm: split... |
458 |
* order is recorded in page_private(page) field. |
1da177e4c Linux-2.6.12-rc2 |
459 460 461 462 463 464 465 466 |
* So when we are allocating or freeing one, we can derive the state of the * other. That is, if we allocate a small block, and both were * free, the remainder of the region must be split into blocks. * If a block is freed, and its buddy is also free, then this * triggers coalescing into a block of larger size. * * -- wli */ |
48db57f8f [PATCH] mm: free_... |
467 |
static inline void __free_one_page(struct page *page, |
ed0ae21dc page allocator: d... |
468 469 |
struct zone *zone, unsigned int order, int migratetype) |
1da177e4c Linux-2.6.12-rc2 |
470 471 |
{ unsigned long page_idx; |
6dda9d55b page allocator: r... |
472 |
unsigned long combined_idx; |
43506fad2 mm/page_alloc.c: ... |
473 |
unsigned long uninitialized_var(buddy_idx); |
6dda9d55b page allocator: r... |
474 |
struct page *buddy; |
1da177e4c Linux-2.6.12-rc2 |
475 |
|
224abf92b [PATCH] mm: bad_p... |
476 |
if (unlikely(PageCompound(page))) |
8cc3b3922 badpage: keep any... |
477 478 |
if (unlikely(destroy_compound_page(page, order))) return; |
1da177e4c Linux-2.6.12-rc2 |
479 |
|
ed0ae21dc page allocator: d... |
480 |
VM_BUG_ON(migratetype == -1); |
1da177e4c Linux-2.6.12-rc2 |
481 |
page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1); |
f2260e6b1 page allocator: u... |
482 |
VM_BUG_ON(page_idx & ((1 << order) - 1)); |
725d704ec [PATCH] mm: VM_BU... |
483 |
VM_BUG_ON(bad_range(zone, page)); |
1da177e4c Linux-2.6.12-rc2 |
484 |
|
1da177e4c Linux-2.6.12-rc2 |
485 |
while (order < MAX_ORDER-1) { |
43506fad2 mm/page_alloc.c: ... |
486 487 |
buddy_idx = __find_buddy_index(page_idx, order); buddy = page + (buddy_idx - page_idx); |
cb2b95e1c [PATCH] zone hand... |
488 |
if (!page_is_buddy(page, buddy, order)) |
3c82d0ce2 buddy: clarify co... |
489 |
break; |
13e7444b0 [PATCH] mm: remov... |
490 |
|
3c82d0ce2 buddy: clarify co... |
491 |
/* Our buddy is free, merge with it and move up one order. */ |
1da177e4c Linux-2.6.12-rc2 |
492 |
list_del(&buddy->lru); |
b2a0ac887 Split the free li... |
493 |
zone->free_area[order].nr_free--; |
1da177e4c Linux-2.6.12-rc2 |
494 |
rmv_page_order(buddy); |
43506fad2 mm/page_alloc.c: ... |
495 |
combined_idx = buddy_idx & page_idx; |
1da177e4c Linux-2.6.12-rc2 |
496 497 498 499 500 |
page = page + (combined_idx - page_idx); page_idx = combined_idx; order++; } set_page_order(page, order); |
6dda9d55b page allocator: r... |
501 502 503 504 505 506 507 508 509 |
/* * If this is not the largest possible page, check if the buddy * of the next-highest order is free. If it is, it's possible * that pages are being freed that will coalesce soon. In case, * that is happening, add the free page to the tail of the list * so it's less likely to be used soon and more likely to be merged * as a higher order page */ |
b7f50cfa3 mm, page-allocato... |
510 |
if ((order < MAX_ORDER-2) && pfn_valid_within(page_to_pfn(buddy))) { |
6dda9d55b page allocator: r... |
511 |
struct page *higher_page, *higher_buddy; |
43506fad2 mm/page_alloc.c: ... |
512 513 514 515 |
combined_idx = buddy_idx & page_idx; higher_page = page + (combined_idx - page_idx); buddy_idx = __find_buddy_index(combined_idx, order + 1); higher_buddy = page + (buddy_idx - combined_idx); |
6dda9d55b page allocator: r... |
516 517 518 519 520 521 522 523 524 |
if (page_is_buddy(higher_page, higher_buddy, order + 1)) { list_add_tail(&page->lru, &zone->free_area[order].free_list[migratetype]); goto out; } } list_add(&page->lru, &zone->free_area[order].free_list[migratetype]); out: |
1da177e4c Linux-2.6.12-rc2 |
525 526 |
zone->free_area[order].nr_free++; } |
092cead61 page allocator: m... |
527 528 529 530 531 532 533 |
/* * free_page_mlock() -- clean up attempts to free and mlocked() page. * Page should not be on lru, so no need to fix that up. * free_pages_check() will verify... */ static inline void free_page_mlock(struct page *page) { |
092cead61 page allocator: m... |
534 535 536 |
__dec_zone_page_state(page, NR_MLOCK); __count_vm_event(UNEVICTABLE_MLOCKFREED); } |
092cead61 page allocator: m... |
537 |
|
224abf92b [PATCH] mm: bad_p... |
538 |
static inline int free_pages_check(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
539 |
{ |
92be2e33b [PATCH] mm: micro... |
540 541 |
if (unlikely(page_mapcount(page) | (page->mapping != NULL) | |
a3af9c389 page allocator: d... |
542 |
(atomic_read(&page->_count) != 0) | |
f212ad7cf memcg: add memcg ... |
543 544 |
(page->flags & PAGE_FLAGS_CHECK_AT_FREE) | (mem_cgroup_bad_page_check(page)))) { |
224abf92b [PATCH] mm: bad_p... |
545 |
bad_page(page); |
79f4b7bf3 badpage: simplify... |
546 |
return 1; |
8cc3b3922 badpage: keep any... |
547 |
} |
79f4b7bf3 badpage: simplify... |
548 549 550 |
if (page->flags & PAGE_FLAGS_CHECK_AT_PREP) page->flags &= ~PAGE_FLAGS_CHECK_AT_PREP; return 0; |
1da177e4c Linux-2.6.12-rc2 |
551 552 553 |
} /* |
5f8dcc212 page-allocator: s... |
554 |
* Frees a number of pages from the PCP lists |
1da177e4c Linux-2.6.12-rc2 |
555 |
* Assumes all pages on list are in same zone, and of same order. |
207f36eec [PATCH] remove in... |
556 |
* count is the number of pages to free. |
1da177e4c Linux-2.6.12-rc2 |
557 558 559 560 561 562 563 |
* * If the zone was previously in an "all pages pinned" state then look to * see if this freeing clears that state. * * And clear the zone's pages_scanned counter, to hold off the "all pages are * pinned" detection logic. */ |
5f8dcc212 page-allocator: s... |
564 565 |
static void free_pcppages_bulk(struct zone *zone, int count, struct per_cpu_pages *pcp) |
1da177e4c Linux-2.6.12-rc2 |
566 |
{ |
5f8dcc212 page-allocator: s... |
567 |
int migratetype = 0; |
a6f9edd65 page-allocator: m... |
568 |
int batch_free = 0; |
72853e299 mm: page allocato... |
569 |
int to_free = count; |
5f8dcc212 page-allocator: s... |
570 |
|
c54ad30c7 [PATCH] mm: pagea... |
571 |
spin_lock(&zone->lock); |
93e4a89a8 mm: restore zone-... |
572 |
zone->all_unreclaimable = 0; |
1da177e4c Linux-2.6.12-rc2 |
573 |
zone->pages_scanned = 0; |
f2260e6b1 page allocator: u... |
574 |
|
72853e299 mm: page allocato... |
575 |
while (to_free) { |
48db57f8f [PATCH] mm: free_... |
576 |
struct page *page; |
5f8dcc212 page-allocator: s... |
577 578 579 |
struct list_head *list; /* |
a6f9edd65 page-allocator: m... |
580 581 582 583 584 |
* Remove pages from lists in a round-robin fashion. A * batch_free count is maintained that is incremented when an * empty list is encountered. This is so more pages are freed * off fuller lists instead of spinning excessively around empty * lists |
5f8dcc212 page-allocator: s... |
585 586 |
*/ do { |
a6f9edd65 page-allocator: m... |
587 |
batch_free++; |
5f8dcc212 page-allocator: s... |
588 589 590 591 |
if (++migratetype == MIGRATE_PCPTYPES) migratetype = 0; list = &pcp->lists[migratetype]; } while (list_empty(list)); |
48db57f8f [PATCH] mm: free_... |
592 |
|
1d16871d8 mm: batch-free pc... |
593 594 595 |
/* This is the only non-empty list. Free them all. */ if (batch_free == MIGRATE_PCPTYPES) batch_free = to_free; |
a6f9edd65 page-allocator: m... |
596 597 598 599 |
do { page = list_entry(list->prev, struct page, lru); /* must delete as __free_one_page list manipulates */ list_del(&page->lru); |
a7016235a mm: fix migratety... |
600 601 602 |
/* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */ __free_one_page(page, zone, 0, page_private(page)); trace_mm_page_pcpu_drain(page, 0, page_private(page)); |
72853e299 mm: page allocato... |
603 |
} while (--to_free && --batch_free && !list_empty(list)); |
1da177e4c Linux-2.6.12-rc2 |
604 |
} |
72853e299 mm: page allocato... |
605 |
__mod_zone_page_state(zone, NR_FREE_PAGES, count); |
c54ad30c7 [PATCH] mm: pagea... |
606 |
spin_unlock(&zone->lock); |
1da177e4c Linux-2.6.12-rc2 |
607 |
} |
ed0ae21dc page allocator: d... |
608 609 |
static void free_one_page(struct zone *zone, struct page *page, int order, int migratetype) |
1da177e4c Linux-2.6.12-rc2 |
610 |
{ |
006d22d9b [PATCH] Optimize ... |
611 |
spin_lock(&zone->lock); |
93e4a89a8 mm: restore zone-... |
612 |
zone->all_unreclaimable = 0; |
006d22d9b [PATCH] Optimize ... |
613 |
zone->pages_scanned = 0; |
f2260e6b1 page allocator: u... |
614 |
|
ed0ae21dc page allocator: d... |
615 |
__free_one_page(page, zone, order, migratetype); |
72853e299 mm: page allocato... |
616 |
__mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order); |
006d22d9b [PATCH] Optimize ... |
617 |
spin_unlock(&zone->lock); |
48db57f8f [PATCH] mm: free_... |
618 |
} |
ec95f53aa mm: introduce fre... |
619 |
static bool free_pages_prepare(struct page *page, unsigned int order) |
48db57f8f [PATCH] mm: free_... |
620 |
{ |
1da177e4c Linux-2.6.12-rc2 |
621 |
int i; |
8cc3b3922 badpage: keep any... |
622 |
int bad = 0; |
1da177e4c Linux-2.6.12-rc2 |
623 |
|
f650316c8 mm/page_alloc.c: ... |
624 |
trace_mm_page_free_direct(page, order); |
b1eeab676 kmemcheck: add ho... |
625 |
kmemcheck_free_shadow(page, order); |
8dd60a3a6 thp: clear compou... |
626 627 628 629 |
if (PageAnon(page)) page->mapping = NULL; for (i = 0; i < (1 << order); i++) bad += free_pages_check(page + i); |
8cc3b3922 badpage: keep any... |
630 |
if (bad) |
ec95f53aa mm: introduce fre... |
631 |
return false; |
689bcebfd [PATCH] unpaged: ... |
632 |
|
3ac7fe5a4 infrastructure to... |
633 |
if (!PageHighMem(page)) { |
9858db504 [PATCH] mm: locks... |
634 |
debug_check_no_locks_freed(page_address(page),PAGE_SIZE<<order); |
3ac7fe5a4 infrastructure to... |
635 636 637 |
debug_check_no_obj_freed(page_address(page), PAGE_SIZE << order); } |
dafb13673 [PATCH] mm: arch_... |
638 |
arch_free_page(page, order); |
48db57f8f [PATCH] mm: free_... |
639 |
kernel_map_pages(page, 1 << order, 0); |
dafb13673 [PATCH] mm: arch_... |
640 |
|
ec95f53aa mm: introduce fre... |
641 642 643 644 645 646 647 648 649 650 |
return true; } static void __free_pages_ok(struct page *page, unsigned int order) { unsigned long flags; int wasMlocked = __TestClearPageMlocked(page); if (!free_pages_prepare(page, order)) return; |
c54ad30c7 [PATCH] mm: pagea... |
651 |
local_irq_save(flags); |
c277331d5 mm: page_alloc: c... |
652 |
if (unlikely(wasMlocked)) |
da456f14d page allocator: d... |
653 |
free_page_mlock(page); |
f8891e5e1 [PATCH] Light wei... |
654 |
__count_vm_events(PGFREE, 1 << order); |
ed0ae21dc page allocator: d... |
655 656 |
free_one_page(page_zone(page), page, order, get_pageblock_migratetype(page)); |
c54ad30c7 [PATCH] mm: pagea... |
657 |
local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
658 |
} |
a226f6c89 [PATCH] FRV: Clea... |
659 660 661 |
/* * permit the bootmem allocator to evade page validation on high-order frees */ |
af370fb8c memory hotplug: s... |
662 |
void __meminit __free_pages_bootmem(struct page *page, unsigned int order) |
a226f6c89 [PATCH] FRV: Clea... |
663 664 665 666 |
{ if (order == 0) { __ClearPageReserved(page); set_page_count(page, 0); |
7835e98b2 [PATCH] remove se... |
667 |
set_page_refcounted(page); |
545b1ea9b [PATCH] mm: clean... |
668 |
__free_page(page); |
a226f6c89 [PATCH] FRV: Clea... |
669 |
} else { |
a226f6c89 [PATCH] FRV: Clea... |
670 |
int loop; |
545b1ea9b [PATCH] mm: clean... |
671 |
prefetchw(page); |
a226f6c89 [PATCH] FRV: Clea... |
672 673 |
for (loop = 0; loop < BITS_PER_LONG; loop++) { struct page *p = &page[loop]; |
545b1ea9b [PATCH] mm: clean... |
674 675 |
if (loop + 1 < BITS_PER_LONG) prefetchw(p + 1); |
a226f6c89 [PATCH] FRV: Clea... |
676 677 678 |
__ClearPageReserved(p); set_page_count(p, 0); } |
7835e98b2 [PATCH] remove se... |
679 |
set_page_refcounted(page); |
545b1ea9b [PATCH] mm: clean... |
680 |
__free_pages(page, order); |
a226f6c89 [PATCH] FRV: Clea... |
681 682 |
} } |
1da177e4c Linux-2.6.12-rc2 |
683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 |
/* * The order of subdivision here is critical for the IO subsystem. * Please do not alter this order without good reasons and regression * testing. Specifically, as large blocks of memory are subdivided, * the order in which smaller blocks are delivered depends on the order * they're subdivided in this function. This is the primary factor * influencing the order in which pages are delivered to the IO * subsystem according to empirical testing, and this is also justified * by considering the behavior of a buddy system containing a single * large block of memory acted on by a series of small allocations. * This behavior is a critical factor in sglist merging's success. * * -- wli */ |
085cc7d5d [PATCH] mm: page_... |
698 |
static inline void expand(struct zone *zone, struct page *page, |
b2a0ac887 Split the free li... |
699 700 |
int low, int high, struct free_area *area, int migratetype) |
1da177e4c Linux-2.6.12-rc2 |
701 702 703 704 705 706 707 |
{ unsigned long size = 1 << high; while (high > low) { area--; high--; size >>= 1; |
725d704ec [PATCH] mm: VM_BU... |
708 |
VM_BUG_ON(bad_range(zone, &page[size])); |
b2a0ac887 Split the free li... |
709 |
list_add(&page[size].lru, &area->free_list[migratetype]); |
1da177e4c Linux-2.6.12-rc2 |
710 711 712 |
area->nr_free++; set_page_order(&page[size], high); } |
1da177e4c Linux-2.6.12-rc2 |
713 |
} |
1da177e4c Linux-2.6.12-rc2 |
714 715 716 |
/* * This page is about to be returned from the page allocator */ |
2a7684a23 HWPOISON: check a... |
717 |
static inline int check_new_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
718 |
{ |
92be2e33b [PATCH] mm: micro... |
719 720 |
if (unlikely(page_mapcount(page) | (page->mapping != NULL) | |
a3af9c389 page allocator: d... |
721 |
(atomic_read(&page->_count) != 0) | |
f212ad7cf memcg: add memcg ... |
722 723 |
(page->flags & PAGE_FLAGS_CHECK_AT_PREP) | (mem_cgroup_bad_page_check(page)))) { |
224abf92b [PATCH] mm: bad_p... |
724 |
bad_page(page); |
689bcebfd [PATCH] unpaged: ... |
725 |
return 1; |
8cc3b3922 badpage: keep any... |
726 |
} |
2a7684a23 HWPOISON: check a... |
727 728 729 730 731 732 733 734 735 736 737 738 |
return 0; } static int prep_new_page(struct page *page, int order, gfp_t gfp_flags) { int i; for (i = 0; i < (1 << order); i++) { struct page *p = page + i; if (unlikely(check_new_page(p))) return 1; } |
689bcebfd [PATCH] unpaged: ... |
739 |
|
4c21e2f24 [PATCH] mm: split... |
740 |
set_page_private(page, 0); |
7835e98b2 [PATCH] remove se... |
741 |
set_page_refcounted(page); |
cc1025090 [PATCH] mm: add a... |
742 743 |
arch_alloc_page(page, order); |
1da177e4c Linux-2.6.12-rc2 |
744 |
kernel_map_pages(page, 1 << order, 1); |
17cf44064 [PATCH] mm: clean... |
745 746 747 748 749 750 |
if (gfp_flags & __GFP_ZERO) prep_zero_page(page, order, gfp_flags); if (order && (gfp_flags & __GFP_COMP)) prep_compound_page(page, order); |
689bcebfd [PATCH] unpaged: ... |
751 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
752 |
} |
56fd56b86 Bias the location... |
753 754 755 756 |
/* * Go through the free lists for the given migratetype and remove * the smallest available page from the freelists */ |
728ec980f page allocator: i... |
757 758 |
static inline struct page *__rmqueue_smallest(struct zone *zone, unsigned int order, |
56fd56b86 Bias the location... |
759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 |
int migratetype) { unsigned int current_order; struct free_area * area; struct page *page; /* Find a page of the appropriate size in the preferred list */ for (current_order = order; current_order < MAX_ORDER; ++current_order) { area = &(zone->free_area[current_order]); if (list_empty(&area->free_list[migratetype])) continue; page = list_entry(area->free_list[migratetype].next, struct page, lru); list_del(&page->lru); rmv_page_order(page); area->nr_free--; |
56fd56b86 Bias the location... |
776 777 778 779 780 781 |
expand(zone, page, order, current_order, area, migratetype); return page; } return NULL; } |
b2a0ac887 Split the free li... |
782 783 784 785 786 |
/* * This array describes the order lists are fallen back to when * the free lists for the desirable migrate type are depleted */ static int fallbacks[MIGRATE_TYPES][MIGRATE_TYPES-1] = { |
64c5e135b don't group high ... |
787 788 789 790 |
[MIGRATE_UNMOVABLE] = { MIGRATE_RECLAIMABLE, MIGRATE_MOVABLE, MIGRATE_RESERVE }, [MIGRATE_RECLAIMABLE] = { MIGRATE_UNMOVABLE, MIGRATE_MOVABLE, MIGRATE_RESERVE }, [MIGRATE_MOVABLE] = { MIGRATE_RECLAIMABLE, MIGRATE_UNMOVABLE, MIGRATE_RESERVE }, [MIGRATE_RESERVE] = { MIGRATE_RESERVE, MIGRATE_RESERVE, MIGRATE_RESERVE }, /* Never used */ |
b2a0ac887 Split the free li... |
791 |
}; |
c361be55b Move free pages b... |
792 793 |
/* * Move the free pages in a range to the free lists of the requested type. |
d9c234005 Do not depend on ... |
794 |
* Note that start_page and end_pages are not aligned on a pageblock |
c361be55b Move free pages b... |
795 796 |
* boundary. If alignment is required, use move_freepages_block() */ |
b69a7288e mm/page_alloc.c: ... |
797 798 799 |
static int move_freepages(struct zone *zone, struct page *start_page, struct page *end_page, int migratetype) |
c361be55b Move free pages b... |
800 801 802 |
{ struct page *page; unsigned long order; |
d100313fd Fix calculation i... |
803 |
int pages_moved = 0; |
c361be55b Move free pages b... |
804 805 806 807 808 809 810 |
#ifndef CONFIG_HOLES_IN_ZONE /* * page_zone is not safe to call in this context when * CONFIG_HOLES_IN_ZONE is set. This bug check is probably redundant * anyway as we check zone boundaries in move_freepages_block(). * Remove at a later date when no bug reports exist related to |
ac0e5b7a6 remove PAGE_GROUP... |
811 |
* grouping pages by mobility |
c361be55b Move free pages b... |
812 813 814 815 816 |
*/ BUG_ON(page_zone(start_page) != page_zone(end_page)); #endif for (page = start_page; page <= end_page;) { |
344c790e3 mm: make setup_zo... |
817 818 |
/* Make sure we are not inadvertently changing nodes */ VM_BUG_ON(page_to_nid(page) != zone_to_nid(zone)); |
c361be55b Move free pages b... |
819 820 821 822 823 824 825 826 827 828 829 |
if (!pfn_valid_within(page_to_pfn(page))) { page++; continue; } if (!PageBuddy(page)) { page++; continue; } order = page_order(page); |
84be48d84 mm/page_alloc.c: ... |
830 831 |
list_move(&page->lru, &zone->free_area[order].free_list[migratetype]); |
c361be55b Move free pages b... |
832 |
page += 1 << order; |
d100313fd Fix calculation i... |
833 |
pages_moved += 1 << order; |
c361be55b Move free pages b... |
834 |
} |
d100313fd Fix calculation i... |
835 |
return pages_moved; |
c361be55b Move free pages b... |
836 |
} |
b69a7288e mm/page_alloc.c: ... |
837 838 |
static int move_freepages_block(struct zone *zone, struct page *page, int migratetype) |
c361be55b Move free pages b... |
839 840 841 842 843 |
{ unsigned long start_pfn, end_pfn; struct page *start_page, *end_page; start_pfn = page_to_pfn(page); |
d9c234005 Do not depend on ... |
844 |
start_pfn = start_pfn & ~(pageblock_nr_pages-1); |
c361be55b Move free pages b... |
845 |
start_page = pfn_to_page(start_pfn); |
d9c234005 Do not depend on ... |
846 847 |
end_page = start_page + pageblock_nr_pages - 1; end_pfn = start_pfn + pageblock_nr_pages - 1; |
c361be55b Move free pages b... |
848 849 850 851 852 853 854 855 856 |
/* Do not cross zone boundaries */ if (start_pfn < zone->zone_start_pfn) start_page = page; if (end_pfn >= zone->zone_start_pfn + zone->spanned_pages) return 0; return move_freepages(zone, start_page, end_page, migratetype); } |
2f66a68f3 page-allocator: c... |
857 858 859 860 861 862 863 864 865 866 |
static void change_pageblock_range(struct page *pageblock_page, int start_order, int migratetype) { int nr_pageblocks = 1 << (start_order - pageblock_order); while (nr_pageblocks--) { set_pageblock_migratetype(pageblock_page, migratetype); pageblock_page += pageblock_nr_pages; } } |
b2a0ac887 Split the free li... |
867 |
/* Remove an element from the buddy allocator from the fallback list */ |
0ac3a4099 page allocator: i... |
868 869 |
static inline struct page * __rmqueue_fallback(struct zone *zone, int order, int start_migratetype) |
b2a0ac887 Split the free li... |
870 871 872 873 874 875 876 877 878 879 880 |
{ struct free_area * area; int current_order; struct page *page; int migratetype, i; /* Find the largest possible block of pages in the other list */ for (current_order = MAX_ORDER-1; current_order >= order; --current_order) { for (i = 0; i < MIGRATE_TYPES - 1; i++) { migratetype = fallbacks[start_migratetype][i]; |
56fd56b86 Bias the location... |
881 882 883 |
/* MIGRATE_RESERVE handled later if necessary */ if (migratetype == MIGRATE_RESERVE) continue; |
e010487db Group high-order ... |
884 |
|
b2a0ac887 Split the free li... |
885 886 887 888 889 890 891 892 893 |
area = &(zone->free_area[current_order]); if (list_empty(&area->free_list[migratetype])) continue; page = list_entry(area->free_list[migratetype].next, struct page, lru); area->nr_free--; /* |
c361be55b Move free pages b... |
894 |
* If breaking a large block of pages, move all free |
46dafbca2 Be more agressive... |
895 896 |
* pages to the preferred allocation list. If falling * back for a reclaimable kernel allocation, be more |
25985edce Fix common misspe... |
897 |
* aggressive about taking ownership of free pages |
b2a0ac887 Split the free li... |
898 |
*/ |
d9c234005 Do not depend on ... |
899 |
if (unlikely(current_order >= (pageblock_order >> 1)) || |
dd5d241ea page-allocator: a... |
900 901 |
start_migratetype == MIGRATE_RECLAIMABLE || page_group_by_mobility_disabled) { |
46dafbca2 Be more agressive... |
902 903 904 905 906 |
unsigned long pages; pages = move_freepages_block(zone, page, start_migratetype); /* Claim the whole block if over half of it is free */ |
dd5d241ea page-allocator: a... |
907 908 |
if (pages >= (1 << (pageblock_order-1)) || page_group_by_mobility_disabled) |
46dafbca2 Be more agressive... |
909 910 |
set_pageblock_migratetype(page, start_migratetype); |
b2a0ac887 Split the free li... |
911 |
migratetype = start_migratetype; |
c361be55b Move free pages b... |
912 |
} |
b2a0ac887 Split the free li... |
913 914 915 916 |
/* Remove the page from the freelists */ list_del(&page->lru); rmv_page_order(page); |
b2a0ac887 Split the free li... |
917 |
|
2f66a68f3 page-allocator: c... |
918 919 920 |
/* Take ownership for orders >= pageblock_order */ if (current_order >= pageblock_order) change_pageblock_range(page, current_order, |
b2a0ac887 Split the free li... |
921 922 923 |
start_migratetype); expand(zone, page, order, current_order, area, migratetype); |
e0fff1bd1 tracing, page-all... |
924 925 926 |
trace_mm_page_alloc_extfrag(page, order, current_order, start_migratetype, migratetype); |
b2a0ac887 Split the free li... |
927 928 929 |
return page; } } |
728ec980f page allocator: i... |
930 |
return NULL; |
b2a0ac887 Split the free li... |
931 |
} |
56fd56b86 Bias the location... |
932 |
/* |
1da177e4c Linux-2.6.12-rc2 |
933 934 935 |
* Do the hard work of removing an element from the buddy allocator. * Call me with the zone->lock already held. */ |
b2a0ac887 Split the free li... |
936 937 |
static struct page *__rmqueue(struct zone *zone, unsigned int order, int migratetype) |
1da177e4c Linux-2.6.12-rc2 |
938 |
{ |
1da177e4c Linux-2.6.12-rc2 |
939 |
struct page *page; |
728ec980f page allocator: i... |
940 |
retry_reserve: |
56fd56b86 Bias the location... |
941 |
page = __rmqueue_smallest(zone, order, migratetype); |
b2a0ac887 Split the free li... |
942 |
|
728ec980f page allocator: i... |
943 |
if (unlikely(!page) && migratetype != MIGRATE_RESERVE) { |
56fd56b86 Bias the location... |
944 |
page = __rmqueue_fallback(zone, order, migratetype); |
b2a0ac887 Split the free li... |
945 |
|
728ec980f page allocator: i... |
946 947 948 949 950 951 952 953 954 955 |
/* * Use MIGRATE_RESERVE rather than fail an allocation. goto * is used because __rmqueue_smallest is an inline function * and we want just one call site */ if (!page) { migratetype = MIGRATE_RESERVE; goto retry_reserve; } } |
0d3d062a6 tracing, page-all... |
956 |
trace_mm_page_alloc_zone_locked(page, order, migratetype); |
b2a0ac887 Split the free li... |
957 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
958 959 960 961 962 963 964 965 |
} /* * Obtain a specified number of elements from the buddy allocator, all under * a single hold of the lock, for efficiency. Add them to the supplied list. * Returns the number of new pages which were placed at *list. */ static int rmqueue_bulk(struct zone *zone, unsigned int order, |
b2a0ac887 Split the free li... |
966 |
unsigned long count, struct list_head *list, |
e084b2d95 page-allocator: p... |
967 |
int migratetype, int cold) |
1da177e4c Linux-2.6.12-rc2 |
968 |
{ |
1da177e4c Linux-2.6.12-rc2 |
969 |
int i; |
1da177e4c Linux-2.6.12-rc2 |
970 |
|
c54ad30c7 [PATCH] mm: pagea... |
971 |
spin_lock(&zone->lock); |
1da177e4c Linux-2.6.12-rc2 |
972 |
for (i = 0; i < count; ++i) { |
b2a0ac887 Split the free li... |
973 |
struct page *page = __rmqueue(zone, order, migratetype); |
085cc7d5d [PATCH] mm: page_... |
974 |
if (unlikely(page == NULL)) |
1da177e4c Linux-2.6.12-rc2 |
975 |
break; |
81eabcbe0 mm: fix page allo... |
976 977 978 979 980 981 982 983 984 985 |
/* * Split buddy pages returned by expand() are received here * in physical page order. The page is added to the callers and * list and the list head then moves forward. From the callers * perspective, the linked list is ordered by page number in * some conditions. This is useful for IO devices that can * merge IO requests if the physical pages are ordered * properly. */ |
e084b2d95 page-allocator: p... |
986 987 988 989 |
if (likely(cold == 0)) list_add(&page->lru, list); else list_add_tail(&page->lru, list); |
535131e69 Choose pages from... |
990 |
set_page_private(page, migratetype); |
81eabcbe0 mm: fix page allo... |
991 |
list = &page->lru; |
1da177e4c Linux-2.6.12-rc2 |
992 |
} |
f2260e6b1 page allocator: u... |
993 |
__mod_zone_page_state(zone, NR_FREE_PAGES, -(i << order)); |
c54ad30c7 [PATCH] mm: pagea... |
994 |
spin_unlock(&zone->lock); |
085cc7d5d [PATCH] mm: page_... |
995 |
return i; |
1da177e4c Linux-2.6.12-rc2 |
996 |
} |
4ae7c0394 [PATCH] Periodica... |
997 |
#ifdef CONFIG_NUMA |
8fce4d8e3 [PATCH] slab: Nod... |
998 |
/* |
4037d4522 Move remote node ... |
999 1000 1001 1002 |
* Called from the vmstat counter updater to drain pagesets of this * currently executing processor on remote nodes after they have * expired. * |
879336c39 [PATCH] drain_nod... |
1003 1004 |
* Note that this function must be called with the thread pinned to * a single processor. |
8fce4d8e3 [PATCH] slab: Nod... |
1005 |
*/ |
4037d4522 Move remote node ... |
1006 |
void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp) |
4ae7c0394 [PATCH] Periodica... |
1007 |
{ |
4ae7c0394 [PATCH] Periodica... |
1008 |
unsigned long flags; |
4037d4522 Move remote node ... |
1009 |
int to_drain; |
4ae7c0394 [PATCH] Periodica... |
1010 |
|
4037d4522 Move remote node ... |
1011 1012 1013 1014 1015 |
local_irq_save(flags); if (pcp->count >= pcp->batch) to_drain = pcp->batch; else to_drain = pcp->count; |
5f8dcc212 page-allocator: s... |
1016 |
free_pcppages_bulk(zone, to_drain, pcp); |
4037d4522 Move remote node ... |
1017 1018 |
pcp->count -= to_drain; local_irq_restore(flags); |
4ae7c0394 [PATCH] Periodica... |
1019 1020 |
} #endif |
9f8f21725 Page allocator: c... |
1021 1022 1023 1024 1025 1026 1027 1028 |
/* * Drain pages of the indicated processor. * * The processor must either be the current processor and the * thread pinned to the current processor or a processor that * is not online. */ static void drain_pages(unsigned int cpu) |
1da177e4c Linux-2.6.12-rc2 |
1029 |
{ |
c54ad30c7 [PATCH] mm: pagea... |
1030 |
unsigned long flags; |
1da177e4c Linux-2.6.12-rc2 |
1031 |
struct zone *zone; |
1da177e4c Linux-2.6.12-rc2 |
1032 |
|
ee99c71c5 mm: introduce for... |
1033 |
for_each_populated_zone(zone) { |
1da177e4c Linux-2.6.12-rc2 |
1034 |
struct per_cpu_pageset *pset; |
3dfa5721f Page allocator: g... |
1035 |
struct per_cpu_pages *pcp; |
1da177e4c Linux-2.6.12-rc2 |
1036 |
|
99dcc3e5a this_cpu: Page al... |
1037 1038 |
local_irq_save(flags); pset = per_cpu_ptr(zone->pageset, cpu); |
3dfa5721f Page allocator: g... |
1039 1040 |
pcp = &pset->pcp; |
2ff754fa8 mm: clear pages_s... |
1041 1042 1043 1044 |
if (pcp->count) { free_pcppages_bulk(zone, pcp->count, pcp); pcp->count = 0; } |
3dfa5721f Page allocator: g... |
1045 |
local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
1046 1047 |
} } |
1da177e4c Linux-2.6.12-rc2 |
1048 |
|
9f8f21725 Page allocator: c... |
1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 |
/* * Spill all of this CPU's per-cpu pages back into the buddy allocator. */ void drain_local_pages(void *arg) { drain_pages(smp_processor_id()); } /* * Spill all the per-cpu pages from all CPUs back into the buddy allocator */ void drain_all_pages(void) { |
15c8b6c1a on_each_cpu(): ki... |
1062 |
on_each_cpu(drain_local_pages, NULL, 1); |
9f8f21725 Page allocator: c... |
1063 |
} |
296699de6 Introduce CONFIG_... |
1064 |
#ifdef CONFIG_HIBERNATION |
1da177e4c Linux-2.6.12-rc2 |
1065 1066 1067 |
void mark_free_pages(struct zone *zone) { |
f623f0db8 [PATCH] swsusp: F... |
1068 1069 |
unsigned long pfn, max_zone_pfn; unsigned long flags; |
b2a0ac887 Split the free li... |
1070 |
int order, t; |
1da177e4c Linux-2.6.12-rc2 |
1071 1072 1073 1074 1075 1076 |
struct list_head *curr; if (!zone->spanned_pages) return; spin_lock_irqsave(&zone->lock, flags); |
f623f0db8 [PATCH] swsusp: F... |
1077 1078 1079 1080 1081 |
max_zone_pfn = zone->zone_start_pfn + zone->spanned_pages; for (pfn = zone->zone_start_pfn; pfn < max_zone_pfn; pfn++) if (pfn_valid(pfn)) { struct page *page = pfn_to_page(pfn); |
7be982349 swsusp: use inlin... |
1082 1083 |
if (!swsusp_page_is_forbidden(page)) swsusp_unset_page_free(page); |
f623f0db8 [PATCH] swsusp: F... |
1084 |
} |
1da177e4c Linux-2.6.12-rc2 |
1085 |
|
b2a0ac887 Split the free li... |
1086 1087 |
for_each_migratetype_order(order, t) { list_for_each(curr, &zone->free_area[order].free_list[t]) { |
f623f0db8 [PATCH] swsusp: F... |
1088 |
unsigned long i; |
1da177e4c Linux-2.6.12-rc2 |
1089 |
|
f623f0db8 [PATCH] swsusp: F... |
1090 1091 |
pfn = page_to_pfn(list_entry(curr, struct page, lru)); for (i = 0; i < (1UL << order); i++) |
7be982349 swsusp: use inlin... |
1092 |
swsusp_set_page_free(pfn_to_page(pfn + i)); |
f623f0db8 [PATCH] swsusp: F... |
1093 |
} |
b2a0ac887 Split the free li... |
1094 |
} |
1da177e4c Linux-2.6.12-rc2 |
1095 1096 |
spin_unlock_irqrestore(&zone->lock, flags); } |
e2c55dc87 Drain per-cpu lis... |
1097 |
#endif /* CONFIG_PM */ |
1da177e4c Linux-2.6.12-rc2 |
1098 1099 |
/* |
1da177e4c Linux-2.6.12-rc2 |
1100 |
* Free a 0-order page |
fc91668ea mm: remove free_h... |
1101 |
* cold == 1 ? free a cold page : free a hot page |
1da177e4c Linux-2.6.12-rc2 |
1102 |
*/ |
fc91668ea mm: remove free_h... |
1103 |
void free_hot_cold_page(struct page *page, int cold) |
1da177e4c Linux-2.6.12-rc2 |
1104 1105 1106 1107 |
{ struct zone *zone = page_zone(page); struct per_cpu_pages *pcp; unsigned long flags; |
5f8dcc212 page-allocator: s... |
1108 |
int migratetype; |
451ea25da mm: perform non-a... |
1109 |
int wasMlocked = __TestClearPageMlocked(page); |
1da177e4c Linux-2.6.12-rc2 |
1110 |
|
ec95f53aa mm: introduce fre... |
1111 |
if (!free_pages_prepare(page, 0)) |
689bcebfd [PATCH] unpaged: ... |
1112 |
return; |
5f8dcc212 page-allocator: s... |
1113 1114 |
migratetype = get_pageblock_migratetype(page); set_page_private(page, migratetype); |
1da177e4c Linux-2.6.12-rc2 |
1115 |
local_irq_save(flags); |
c277331d5 mm: page_alloc: c... |
1116 |
if (unlikely(wasMlocked)) |
da456f14d page allocator: d... |
1117 |
free_page_mlock(page); |
f8891e5e1 [PATCH] Light wei... |
1118 |
__count_vm_event(PGFREE); |
da456f14d page allocator: d... |
1119 |
|
5f8dcc212 page-allocator: s... |
1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 |
/* * We only track unmovable, reclaimable and movable on pcp lists. * Free ISOLATE pages back to the allocator because they are being * offlined but treat RESERVE as movable pages so we can get those * areas back if necessary. Otherwise, we may have to free * excessively into the page allocator */ if (migratetype >= MIGRATE_PCPTYPES) { if (unlikely(migratetype == MIGRATE_ISOLATE)) { free_one_page(zone, page, 0, migratetype); goto out; } migratetype = MIGRATE_MOVABLE; } |
99dcc3e5a this_cpu: Page al... |
1134 |
pcp = &this_cpu_ptr(zone->pageset)->pcp; |
3dfa5721f Page allocator: g... |
1135 |
if (cold) |
5f8dcc212 page-allocator: s... |
1136 |
list_add_tail(&page->lru, &pcp->lists[migratetype]); |
3dfa5721f Page allocator: g... |
1137 |
else |
5f8dcc212 page-allocator: s... |
1138 |
list_add(&page->lru, &pcp->lists[migratetype]); |
1da177e4c Linux-2.6.12-rc2 |
1139 |
pcp->count++; |
48db57f8f [PATCH] mm: free_... |
1140 |
if (pcp->count >= pcp->high) { |
5f8dcc212 page-allocator: s... |
1141 |
free_pcppages_bulk(zone, pcp->batch, pcp); |
48db57f8f [PATCH] mm: free_... |
1142 1143 |
pcp->count -= pcp->batch; } |
5f8dcc212 page-allocator: s... |
1144 1145 |
out: |
1da177e4c Linux-2.6.12-rc2 |
1146 |
local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
1147 |
} |
8dfcc9ba2 [PATCH] mm: split... |
1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 |
/* * split_page takes a non-compound higher-order page, and splits it into * n (1<<order) sub-pages: page[0..n] * Each sub-page must be freed individually. * * Note: this is probably too low level an operation for use in drivers. * Please consult with lkml before using this in your driver. */ void split_page(struct page *page, unsigned int order) { int i; |
725d704ec [PATCH] mm: VM_BU... |
1159 1160 |
VM_BUG_ON(PageCompound(page)); VM_BUG_ON(!page_count(page)); |
b1eeab676 kmemcheck: add ho... |
1161 1162 1163 1164 1165 1166 1167 1168 1169 |
#ifdef CONFIG_KMEMCHECK /* * Split shadow pages too, because free(page[0]) would * otherwise free the whole shadow. */ if (kmemcheck_page_is_tracked(page)) split_page(virt_to_page(page[0].shadow), order); #endif |
7835e98b2 [PATCH] remove se... |
1170 1171 |
for (i = 1; i < (1 << order); i++) set_page_refcounted(page + i); |
8dfcc9ba2 [PATCH] mm: split... |
1172 |
} |
8dfcc9ba2 [PATCH] mm: split... |
1173 |
|
1da177e4c Linux-2.6.12-rc2 |
1174 |
/* |
748446bb6 mm: compaction: m... |
1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 |
* Similar to split_page except the page is already free. As this is only * being used for migration, the migratetype of the block also changes. * As this is called with interrupts disabled, the caller is responsible * for calling arch_alloc_page() and kernel_map_page() after interrupts * are enabled. * * Note: this is probably too low level an operation for use in drivers. * Please consult with lkml before using this in your driver. */ int split_free_page(struct page *page) { unsigned int order; unsigned long watermark; struct zone *zone; BUG_ON(!PageBuddy(page)); zone = page_zone(page); order = page_order(page); /* Obey watermarks as if the page was being allocated */ watermark = low_wmark_pages(zone) + (1 << order); if (!zone_watermark_ok(zone, 0, watermark, 0, 0)) return 0; /* Remove page from free list */ list_del(&page->lru); zone->free_area[order].nr_free--; rmv_page_order(page); __mod_zone_page_state(zone, NR_FREE_PAGES, -(1UL << order)); /* Split into individual pages */ set_page_refcounted(page); split_page(page, order); if (order >= pageblock_order - 1) { struct page *endpage = page + (1 << order) - 1; for (; page < endpage; page += pageblock_nr_pages) set_pageblock_migratetype(page, MIGRATE_MOVABLE); } return 1 << order; } /* |
1da177e4c Linux-2.6.12-rc2 |
1220 1221 1222 1223 |
* Really, prep_compound_page() should be called from __rmqueue_bulk(). But * we cheat by calling it from here, in the order > 0 path. Saves a branch * or two. */ |
0a15c3e9f page allocator: i... |
1224 1225 |
static inline struct page *buffered_rmqueue(struct zone *preferred_zone, |
3dd282669 page allocator: c... |
1226 1227 |
struct zone *zone, int order, gfp_t gfp_flags, int migratetype) |
1da177e4c Linux-2.6.12-rc2 |
1228 1229 |
{ unsigned long flags; |
689bcebfd [PATCH] unpaged: ... |
1230 |
struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
1231 |
int cold = !!(gfp_flags & __GFP_COLD); |
689bcebfd [PATCH] unpaged: ... |
1232 |
again: |
48db57f8f [PATCH] mm: free_... |
1233 |
if (likely(order == 0)) { |
1da177e4c Linux-2.6.12-rc2 |
1234 |
struct per_cpu_pages *pcp; |
5f8dcc212 page-allocator: s... |
1235 |
struct list_head *list; |
1da177e4c Linux-2.6.12-rc2 |
1236 |
|
1da177e4c Linux-2.6.12-rc2 |
1237 |
local_irq_save(flags); |
99dcc3e5a this_cpu: Page al... |
1238 1239 |
pcp = &this_cpu_ptr(zone->pageset)->pcp; list = &pcp->lists[migratetype]; |
5f8dcc212 page-allocator: s... |
1240 |
if (list_empty(list)) { |
535131e69 Choose pages from... |
1241 |
pcp->count += rmqueue_bulk(zone, 0, |
5f8dcc212 page-allocator: s... |
1242 |
pcp->batch, list, |
e084b2d95 page-allocator: p... |
1243 |
migratetype, cold); |
5f8dcc212 page-allocator: s... |
1244 |
if (unlikely(list_empty(list))) |
6fb332fab memory hotplug: e... |
1245 |
goto failed; |
535131e69 Choose pages from... |
1246 |
} |
b92a6edd4 Add a configure o... |
1247 |
|
5f8dcc212 page-allocator: s... |
1248 1249 1250 1251 |
if (cold) page = list_entry(list->prev, struct page, lru); else page = list_entry(list->next, struct page, lru); |
b92a6edd4 Add a configure o... |
1252 1253 |
list_del(&page->lru); pcp->count--; |
7fb1d9fca [PATCH] mm: __all... |
1254 |
} else { |
dab48dab3 page-allocator: w... |
1255 1256 1257 1258 1259 1260 1261 1262 |
if (unlikely(gfp_flags & __GFP_NOFAIL)) { /* * __GFP_NOFAIL is not to be used in new code. * * All __GFP_NOFAIL callers should be fixed so that they * properly detect and handle allocation failures. * * We most definitely don't want callers attempting to |
4923abf9f Don't warn about ... |
1263 |
* allocate greater than order-1 page units with |
dab48dab3 page-allocator: w... |
1264 1265 |
* __GFP_NOFAIL. */ |
4923abf9f Don't warn about ... |
1266 |
WARN_ON_ONCE(order > 1); |
dab48dab3 page-allocator: w... |
1267 |
} |
1da177e4c Linux-2.6.12-rc2 |
1268 |
spin_lock_irqsave(&zone->lock, flags); |
b2a0ac887 Split the free li... |
1269 |
page = __rmqueue(zone, order, migratetype); |
a74609faf [PATCH] mm: page_... |
1270 1271 1272 |
spin_unlock(&zone->lock); if (!page) goto failed; |
6ccf80eb1 page allocator: u... |
1273 |
__mod_zone_page_state(zone, NR_FREE_PAGES, -(1 << order)); |
1da177e4c Linux-2.6.12-rc2 |
1274 |
} |
f8891e5e1 [PATCH] Light wei... |
1275 |
__count_zone_vm_events(PGALLOC, zone, 1 << order); |
78afd5612 mm: add __GFP_OTH... |
1276 |
zone_statistics(preferred_zone, zone, gfp_flags); |
a74609faf [PATCH] mm: page_... |
1277 |
local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
1278 |
|
725d704ec [PATCH] mm: VM_BU... |
1279 |
VM_BUG_ON(bad_range(zone, page)); |
17cf44064 [PATCH] mm: clean... |
1280 |
if (prep_new_page(page, order, gfp_flags)) |
a74609faf [PATCH] mm: page_... |
1281 |
goto again; |
1da177e4c Linux-2.6.12-rc2 |
1282 |
return page; |
a74609faf [PATCH] mm: page_... |
1283 1284 1285 |
failed: local_irq_restore(flags); |
a74609faf [PATCH] mm: page_... |
1286 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1287 |
} |
418589663 page allocator: u... |
1288 1289 1290 1291 1292 1293 1294 1295 |
/* The ALLOC_WMARK bits are used as an index to zone->watermark */ #define ALLOC_WMARK_MIN WMARK_MIN #define ALLOC_WMARK_LOW WMARK_LOW #define ALLOC_WMARK_HIGH WMARK_HIGH #define ALLOC_NO_WATERMARKS 0x04 /* don't check watermarks at all */ /* Mask to get the watermark bits */ #define ALLOC_WMARK_MASK (ALLOC_NO_WATERMARKS-1) |
3148890bf [PATCH] mm: __all... |
1296 1297 1298 |
#define ALLOC_HARDER 0x10 /* try to alloc harder */ #define ALLOC_HIGH 0x20 /* __GFP_HIGH set */ #define ALLOC_CPUSET 0x40 /* check for correct cpuset */ |
7fb1d9fca [PATCH] mm: __all... |
1299 |
|
933e312e7 [PATCH] fault-inj... |
1300 |
#ifdef CONFIG_FAIL_PAGE_ALLOC |
b2588c4b4 fail_page_alloc: ... |
1301 |
static struct { |
933e312e7 [PATCH] fault-inj... |
1302 1303 1304 1305 |
struct fault_attr attr; u32 ignore_gfp_highmem; u32 ignore_gfp_wait; |
54114994f fault-injection: ... |
1306 |
u32 min_order; |
933e312e7 [PATCH] fault-inj... |
1307 1308 |
} fail_page_alloc = { .attr = FAULT_ATTR_INITIALIZER, |
6b1b60f41 [PATCH] fault-inj... |
1309 1310 |
.ignore_gfp_wait = 1, .ignore_gfp_highmem = 1, |
54114994f fault-injection: ... |
1311 |
.min_order = 1, |
933e312e7 [PATCH] fault-inj... |
1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 |
}; static int __init setup_fail_page_alloc(char *str) { return setup_fault_attr(&fail_page_alloc.attr, str); } __setup("fail_page_alloc=", setup_fail_page_alloc); static int should_fail_alloc_page(gfp_t gfp_mask, unsigned int order) { |
54114994f fault-injection: ... |
1322 1323 |
if (order < fail_page_alloc.min_order) return 0; |
933e312e7 [PATCH] fault-inj... |
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 |
if (gfp_mask & __GFP_NOFAIL) return 0; if (fail_page_alloc.ignore_gfp_highmem && (gfp_mask & __GFP_HIGHMEM)) return 0; if (fail_page_alloc.ignore_gfp_wait && (gfp_mask & __GFP_WAIT)) return 0; return should_fail(&fail_page_alloc.attr, 1 << order); } #ifdef CONFIG_FAULT_INJECTION_DEBUG_FS static int __init fail_page_alloc_debugfs(void) { mode_t mode = S_IFREG | S_IRUSR | S_IWUSR; struct dentry *dir; |
933e312e7 [PATCH] fault-inj... |
1340 |
|
dd48c085c fault-injection: ... |
1341 1342 1343 1344 |
dir = fault_create_debugfs_attr("fail_page_alloc", NULL, &fail_page_alloc.attr); if (IS_ERR(dir)) return PTR_ERR(dir); |
933e312e7 [PATCH] fault-inj... |
1345 |
|
b2588c4b4 fail_page_alloc: ... |
1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 |
if (!debugfs_create_bool("ignore-gfp-wait", mode, dir, &fail_page_alloc.ignore_gfp_wait)) goto fail; if (!debugfs_create_bool("ignore-gfp-highmem", mode, dir, &fail_page_alloc.ignore_gfp_highmem)) goto fail; if (!debugfs_create_u32("min-order", mode, dir, &fail_page_alloc.min_order)) goto fail; return 0; fail: |
dd48c085c fault-injection: ... |
1358 |
debugfs_remove_recursive(dir); |
933e312e7 [PATCH] fault-inj... |
1359 |
|
b2588c4b4 fail_page_alloc: ... |
1360 |
return -ENOMEM; |
933e312e7 [PATCH] fault-inj... |
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 |
} late_initcall(fail_page_alloc_debugfs); #endif /* CONFIG_FAULT_INJECTION_DEBUG_FS */ #else /* CONFIG_FAIL_PAGE_ALLOC */ static inline int should_fail_alloc_page(gfp_t gfp_mask, unsigned int order) { return 0; } #endif /* CONFIG_FAIL_PAGE_ALLOC */ |
1da177e4c Linux-2.6.12-rc2 |
1375 |
/* |
88f5acf88 mm: page allocato... |
1376 |
* Return true if free pages are above 'mark'. This takes into account the order |
1da177e4c Linux-2.6.12-rc2 |
1377 1378 |
* of the allocation. */ |
88f5acf88 mm: page allocato... |
1379 1380 |
static bool __zone_watermark_ok(struct zone *z, int order, unsigned long mark, int classzone_idx, int alloc_flags, long free_pages) |
1da177e4c Linux-2.6.12-rc2 |
1381 1382 |
{ /* free_pages my go negative - that's OK */ |
d23ad4232 [PATCH] Use ZVC f... |
1383 |
long min = mark; |
1da177e4c Linux-2.6.12-rc2 |
1384 |
int o; |
88f5acf88 mm: page allocato... |
1385 |
free_pages -= (1 << order) + 1; |
7fb1d9fca [PATCH] mm: __all... |
1386 |
if (alloc_flags & ALLOC_HIGH) |
1da177e4c Linux-2.6.12-rc2 |
1387 |
min -= min / 2; |
7fb1d9fca [PATCH] mm: __all... |
1388 |
if (alloc_flags & ALLOC_HARDER) |
1da177e4c Linux-2.6.12-rc2 |
1389 1390 1391 |
min -= min / 4; if (free_pages <= min + z->lowmem_reserve[classzone_idx]) |
88f5acf88 mm: page allocato... |
1392 |
return false; |
1da177e4c Linux-2.6.12-rc2 |
1393 1394 1395 1396 1397 1398 1399 1400 |
for (o = 0; o < order; o++) { /* At the next order, this order's pages become unavailable */ free_pages -= z->free_area[o].nr_free << o; /* Require fewer higher order pages to be free */ min >>= 1; if (free_pages <= min) |
88f5acf88 mm: page allocato... |
1401 |
return false; |
1da177e4c Linux-2.6.12-rc2 |
1402 |
} |
88f5acf88 mm: page allocato... |
1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 |
return true; } bool zone_watermark_ok(struct zone *z, int order, unsigned long mark, int classzone_idx, int alloc_flags) { return __zone_watermark_ok(z, order, mark, classzone_idx, alloc_flags, zone_page_state(z, NR_FREE_PAGES)); } bool zone_watermark_ok_safe(struct zone *z, int order, unsigned long mark, int classzone_idx, int alloc_flags) { long free_pages = zone_page_state(z, NR_FREE_PAGES); if (z->percpu_drift_mark && free_pages < z->percpu_drift_mark) free_pages = zone_page_state_snapshot(z, NR_FREE_PAGES); return __zone_watermark_ok(z, order, mark, classzone_idx, alloc_flags, free_pages); |
1da177e4c Linux-2.6.12-rc2 |
1423 |
} |
9276b1bc9 [PATCH] memory pa... |
1424 1425 1426 1427 1428 1429 |
#ifdef CONFIG_NUMA /* * zlc_setup - Setup for "zonelist cache". Uses cached zone data to * skip over zones that are not allowed by the cpuset, or that have * been recently (in last second) found to be nearly full. See further * comments in mmzone.h. Reduces cache footprint of zonelist scans |
183ff22bb spelling fixes: mm/ |
1430 |
* that have to skip over a lot of full or unallowed zones. |
9276b1bc9 [PATCH] memory pa... |
1431 1432 1433 |
* * If the zonelist cache is present in the passed in zonelist, then * returns a pointer to the allowed node mask (either the current |
37b07e416 memoryless nodes:... |
1434 |
* tasks mems_allowed, or node_states[N_HIGH_MEMORY].) |
9276b1bc9 [PATCH] memory pa... |
1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 |
* * If the zonelist cache is not available for this zonelist, does * nothing and returns NULL. * * If the fullzones BITMAP in the zonelist cache is stale (more than * a second since last zap'd) then we zap it out (clear its bits.) * * We hold off even calling zlc_setup, until after we've checked the * first zone in the zonelist, on the theory that most allocations will * be satisfied from that first zone, so best to examine that zone as * quickly as we can. */ static nodemask_t *zlc_setup(struct zonelist *zonelist, int alloc_flags) { struct zonelist_cache *zlc; /* cached zonelist speedup info */ nodemask_t *allowednodes; /* zonelist_cache approximation */ zlc = zonelist->zlcache_ptr; if (!zlc) return NULL; |
f05111f50 mm/page_alloc.c: ... |
1455 |
if (time_after(jiffies, zlc->last_full_zap + HZ)) { |
9276b1bc9 [PATCH] memory pa... |
1456 1457 1458 1459 1460 1461 |
bitmap_zero(zlc->fullzones, MAX_ZONES_PER_ZONELIST); zlc->last_full_zap = jiffies; } allowednodes = !in_interrupt() && (alloc_flags & ALLOC_CPUSET) ? &cpuset_current_mems_allowed : |
37b07e416 memoryless nodes:... |
1462 |
&node_states[N_HIGH_MEMORY]; |
9276b1bc9 [PATCH] memory pa... |
1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 |
return allowednodes; } /* * Given 'z' scanning a zonelist, run a couple of quick checks to see * if it is worth looking at further for free memory: * 1) Check that the zone isn't thought to be full (doesn't have its * bit set in the zonelist_cache fullzones BITMAP). * 2) Check that the zones node (obtained from the zonelist_cache * z_to_n[] mapping) is allowed in the passed in allowednodes mask. * Return true (non-zero) if zone is worth looking at further, or * else return false (zero) if it is not. * * This check -ignores- the distinction between various watermarks, * such as GFP_HIGH, GFP_ATOMIC, PF_MEMALLOC, ... If a zone is * found to be full for any variation of these watermarks, it will * be considered full for up to one second by all requests, unless * we are so low on memory on all allowed nodes that we are forced * into the second scan of the zonelist. * * In the second scan we ignore this zonelist cache and exactly * apply the watermarks to all zones, even it is slower to do so. * We are low on memory in the second scan, and should leave no stone * unturned looking for a free page. */ |
dd1a239f6 mm: have zonelist... |
1488 |
static int zlc_zone_worth_trying(struct zonelist *zonelist, struct zoneref *z, |
9276b1bc9 [PATCH] memory pa... |
1489 1490 1491 1492 1493 1494 1495 1496 1497 |
nodemask_t *allowednodes) { struct zonelist_cache *zlc; /* cached zonelist speedup info */ int i; /* index of *z in zonelist zones */ int n; /* node that zone *z is on */ zlc = zonelist->zlcache_ptr; if (!zlc) return 1; |
dd1a239f6 mm: have zonelist... |
1498 |
i = z - zonelist->_zonerefs; |
9276b1bc9 [PATCH] memory pa... |
1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 |
n = zlc->z_to_n[i]; /* This zone is worth trying if it is allowed but not full */ return node_isset(n, *allowednodes) && !test_bit(i, zlc->fullzones); } /* * Given 'z' scanning a zonelist, set the corresponding bit in * zlc->fullzones, so that subsequent attempts to allocate a page * from that zone don't waste time re-examining it. */ |
dd1a239f6 mm: have zonelist... |
1510 |
static void zlc_mark_zone_full(struct zonelist *zonelist, struct zoneref *z) |
9276b1bc9 [PATCH] memory pa... |
1511 1512 1513 1514 1515 1516 1517 |
{ struct zonelist_cache *zlc; /* cached zonelist speedup info */ int i; /* index of *z in zonelist zones */ zlc = zonelist->zlcache_ptr; if (!zlc) return; |
dd1a239f6 mm: have zonelist... |
1518 |
i = z - zonelist->_zonerefs; |
9276b1bc9 [PATCH] memory pa... |
1519 1520 1521 |
set_bit(i, zlc->fullzones); } |
76d3fbf8f mm: page allocato... |
1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 |
/* * clear all zones full, called after direct reclaim makes progress so that * a zone that was recently full is not skipped over for up to a second */ static void zlc_clear_zones_full(struct zonelist *zonelist) { struct zonelist_cache *zlc; /* cached zonelist speedup info */ zlc = zonelist->zlcache_ptr; if (!zlc) return; bitmap_zero(zlc->fullzones, MAX_ZONES_PER_ZONELIST); } |
9276b1bc9 [PATCH] memory pa... |
1536 1537 1538 1539 1540 1541 |
#else /* CONFIG_NUMA */ static nodemask_t *zlc_setup(struct zonelist *zonelist, int alloc_flags) { return NULL; } |
dd1a239f6 mm: have zonelist... |
1542 |
static int zlc_zone_worth_trying(struct zonelist *zonelist, struct zoneref *z, |
9276b1bc9 [PATCH] memory pa... |
1543 1544 1545 1546 |
nodemask_t *allowednodes) { return 1; } |
dd1a239f6 mm: have zonelist... |
1547 |
static void zlc_mark_zone_full(struct zonelist *zonelist, struct zoneref *z) |
9276b1bc9 [PATCH] memory pa... |
1548 1549 |
{ } |
76d3fbf8f mm: page allocato... |
1550 1551 1552 1553 |
static void zlc_clear_zones_full(struct zonelist *zonelist) { } |
9276b1bc9 [PATCH] memory pa... |
1554 |
#endif /* CONFIG_NUMA */ |
7fb1d9fca [PATCH] mm: __all... |
1555 |
/* |
0798e5193 [PATCH] memory pa... |
1556 |
* get_page_from_freelist goes through the zonelist trying to allocate |
7fb1d9fca [PATCH] mm: __all... |
1557 1558 1559 |
* a page. */ static struct page * |
19770b326 mm: filter based ... |
1560 |
get_page_from_freelist(gfp_t gfp_mask, nodemask_t *nodemask, unsigned int order, |
5117f45d1 page allocator: c... |
1561 |
struct zonelist *zonelist, int high_zoneidx, int alloc_flags, |
3dd282669 page allocator: c... |
1562 |
struct zone *preferred_zone, int migratetype) |
753ee7289 [PATCH] VM: early... |
1563 |
{ |
dd1a239f6 mm: have zonelist... |
1564 |
struct zoneref *z; |
7fb1d9fca [PATCH] mm: __all... |
1565 |
struct page *page = NULL; |
54a6eb5c4 mm: use two zonel... |
1566 |
int classzone_idx; |
5117f45d1 page allocator: c... |
1567 |
struct zone *zone; |
9276b1bc9 [PATCH] memory pa... |
1568 1569 1570 |
nodemask_t *allowednodes = NULL;/* zonelist_cache approximation */ int zlc_active = 0; /* set if using zonelist_cache */ int did_zlc_setup = 0; /* just call zlc_setup() one time */ |
54a6eb5c4 mm: use two zonel... |
1571 |
|
19770b326 mm: filter based ... |
1572 |
classzone_idx = zone_idx(preferred_zone); |
9276b1bc9 [PATCH] memory pa... |
1573 |
zonelist_scan: |
7fb1d9fca [PATCH] mm: __all... |
1574 |
/* |
9276b1bc9 [PATCH] memory pa... |
1575 |
* Scan zonelist, looking for a zone with enough free. |
7fb1d9fca [PATCH] mm: __all... |
1576 1577 |
* See also cpuset_zone_allowed() comment in kernel/cpuset.c. */ |
19770b326 mm: filter based ... |
1578 1579 |
for_each_zone_zonelist_nodemask(zone, z, zonelist, high_zoneidx, nodemask) { |
9276b1bc9 [PATCH] memory pa... |
1580 1581 1582 |
if (NUMA_BUILD && zlc_active && !zlc_zone_worth_trying(zonelist, z, allowednodes)) continue; |
7fb1d9fca [PATCH] mm: __all... |
1583 |
if ((alloc_flags & ALLOC_CPUSET) && |
02a0e53d8 [PATCH] cpuset: r... |
1584 |
!cpuset_zone_allowed_softwall(zone, gfp_mask)) |
cd38b115d mm: page allocato... |
1585 |
continue; |
7fb1d9fca [PATCH] mm: __all... |
1586 |
|
418589663 page allocator: u... |
1587 |
BUILD_BUG_ON(ALLOC_NO_WATERMARKS < NR_WMARK); |
7fb1d9fca [PATCH] mm: __all... |
1588 |
if (!(alloc_flags & ALLOC_NO_WATERMARKS)) { |
3148890bf [PATCH] mm: __all... |
1589 |
unsigned long mark; |
fa5e084e4 vmscan: do not un... |
1590 |
int ret; |
418589663 page allocator: u... |
1591 |
mark = zone->watermark[alloc_flags & ALLOC_WMARK_MASK]; |
fa5e084e4 vmscan: do not un... |
1592 1593 1594 |
if (zone_watermark_ok(zone, order, mark, classzone_idx, alloc_flags)) goto try_this_zone; |
cd38b115d mm: page allocato... |
1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 |
if (NUMA_BUILD && !did_zlc_setup && nr_online_nodes > 1) { /* * we do zlc_setup if there are multiple nodes * and before considering the first zone allowed * by the cpuset. */ allowednodes = zlc_setup(zonelist, alloc_flags); zlc_active = 1; did_zlc_setup = 1; } |
fa5e084e4 vmscan: do not un... |
1605 1606 |
if (zone_reclaim_mode == 0) goto this_zone_full; |
cd38b115d mm: page allocato... |
1607 1608 1609 1610 1611 1612 1613 |
/* * As we may have just activated ZLC, check if the first * eligible zone has failed zone_reclaim recently. */ if (NUMA_BUILD && zlc_active && !zlc_zone_worth_trying(zonelist, z, allowednodes)) continue; |
fa5e084e4 vmscan: do not un... |
1614 1615 1616 1617 |
ret = zone_reclaim(zone, gfp_mask, order); switch (ret) { case ZONE_RECLAIM_NOSCAN: /* did not scan */ |
cd38b115d mm: page allocato... |
1618 |
continue; |
fa5e084e4 vmscan: do not un... |
1619 1620 |
case ZONE_RECLAIM_FULL: /* scanned but unreclaimable */ |
cd38b115d mm: page allocato... |
1621 |
continue; |
fa5e084e4 vmscan: do not un... |
1622 1623 1624 1625 |
default: /* did we reclaim enough */ if (!zone_watermark_ok(zone, order, mark, classzone_idx, alloc_flags)) |
9276b1bc9 [PATCH] memory pa... |
1626 |
goto this_zone_full; |
0798e5193 [PATCH] memory pa... |
1627 |
} |
7fb1d9fca [PATCH] mm: __all... |
1628 |
} |
fa5e084e4 vmscan: do not un... |
1629 |
try_this_zone: |
3dd282669 page allocator: c... |
1630 1631 |
page = buffered_rmqueue(preferred_zone, zone, order, gfp_mask, migratetype); |
0798e5193 [PATCH] memory pa... |
1632 |
if (page) |
7fb1d9fca [PATCH] mm: __all... |
1633 |
break; |
9276b1bc9 [PATCH] memory pa... |
1634 1635 1636 |
this_zone_full: if (NUMA_BUILD) zlc_mark_zone_full(zonelist, z); |
54a6eb5c4 mm: use two zonel... |
1637 |
} |
9276b1bc9 [PATCH] memory pa... |
1638 1639 1640 1641 1642 1643 |
if (unlikely(NUMA_BUILD && page == NULL && zlc_active)) { /* Disable zlc cache for second zonelist scan */ zlc_active = 0; goto zonelist_scan; } |
7fb1d9fca [PATCH] mm: __all... |
1644 |
return page; |
753ee7289 [PATCH] VM: early... |
1645 |
} |
29423e77c oom: suppress sho... |
1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 |
/* * Large machines with many possible nodes should not always dump per-node * meminfo in irq context. */ static inline bool should_suppress_show_mem(void) { bool ret = false; #if NODES_SHIFT > 8 ret = in_interrupt(); #endif return ret; } |
a238ab5b0 mm: break out pag... |
1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 |
static DEFINE_RATELIMIT_STATE(nopage_rs, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); void warn_alloc_failed(gfp_t gfp_mask, int order, const char *fmt, ...) { va_list args; unsigned int filter = SHOW_MEM_FILTER_NODES; if ((gfp_mask & __GFP_NOWARN) || !__ratelimit(&nopage_rs)) return; /* * This documents exceptions given to allocations in certain * contexts that are allowed to allocate outside current's set * of allowed nodes. */ if (!(gfp_mask & __GFP_NOMEMALLOC)) if (test_thread_flag(TIF_MEMDIE) || (current->flags & (PF_MEMALLOC | PF_EXITING))) filter &= ~SHOW_MEM_FILTER_NODES; if (in_interrupt() || !(gfp_mask & __GFP_WAIT)) filter &= ~SHOW_MEM_FILTER_NODES; if (fmt) { printk(KERN_WARNING); va_start(args, fmt); vprintk(fmt, args); va_end(args); } pr_warning("%s: page allocation failure: order:%d, mode:0x%x ", current->comm, order, gfp_mask); dump_stack(); if (!should_suppress_show_mem()) show_mem(filter); } |
11e33f6a5 page allocator: b... |
1698 1699 1700 |
static inline int should_alloc_retry(gfp_t gfp_mask, unsigned int order, unsigned long pages_reclaimed) |
1da177e4c Linux-2.6.12-rc2 |
1701 |
{ |
11e33f6a5 page allocator: b... |
1702 1703 1704 |
/* Do not loop if specifically requested */ if (gfp_mask & __GFP_NORETRY) return 0; |
1da177e4c Linux-2.6.12-rc2 |
1705 |
|
11e33f6a5 page allocator: b... |
1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 |
/* * In this implementation, order <= PAGE_ALLOC_COSTLY_ORDER * means __GFP_NOFAIL, but that may not be true in other * implementations. */ if (order <= PAGE_ALLOC_COSTLY_ORDER) return 1; /* * For order > PAGE_ALLOC_COSTLY_ORDER, if __GFP_REPEAT is * specified, then we retry until we no longer reclaim any pages * (above), or we've reclaimed an order of pages at least as * large as the allocation's order. In both cases, if the * allocation still fails, we stop retrying. */ if (gfp_mask & __GFP_REPEAT && pages_reclaimed < (1 << order)) return 1; |
cf40bd16f lockdep: annotate... |
1723 |
|
11e33f6a5 page allocator: b... |
1724 1725 1726 1727 1728 1729 |
/* * Don't let big-order allocations loop unless the caller * explicitly requests that. */ if (gfp_mask & __GFP_NOFAIL) return 1; |
1da177e4c Linux-2.6.12-rc2 |
1730 |
|
11e33f6a5 page allocator: b... |
1731 1732 |
return 0; } |
933e312e7 [PATCH] fault-inj... |
1733 |
|
11e33f6a5 page allocator: b... |
1734 1735 1736 |
static inline struct page * __alloc_pages_may_oom(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, enum zone_type high_zoneidx, |
3dd282669 page allocator: c... |
1737 1738 |
nodemask_t *nodemask, struct zone *preferred_zone, int migratetype) |
11e33f6a5 page allocator: b... |
1739 1740 1741 1742 |
{ struct page *page; /* Acquire the OOM killer lock for the zones in zonelist */ |
ff321feac mm: rename try_se... |
1743 |
if (!try_set_zonelist_oom(zonelist, gfp_mask)) { |
11e33f6a5 page allocator: b... |
1744 |
schedule_timeout_uninterruptible(1); |
1da177e4c Linux-2.6.12-rc2 |
1745 1746 |
return NULL; } |
6b1de9161 [PATCH] VM: fix z... |
1747 |
|
11e33f6a5 page allocator: b... |
1748 1749 1750 1751 1752 1753 1754 |
/* * Go through the zonelist yet one more time, keep very high watermark * here, this is only to catch a parallel oom killing, we must fail if * we're still under heavy pressure. */ page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, nodemask, order, zonelist, high_zoneidx, |
5117f45d1 page allocator: c... |
1755 |
ALLOC_WMARK_HIGH|ALLOC_CPUSET, |
3dd282669 page allocator: c... |
1756 |
preferred_zone, migratetype); |
7fb1d9fca [PATCH] mm: __all... |
1757 |
if (page) |
11e33f6a5 page allocator: b... |
1758 |
goto out; |
4365a5676 oom-kill: fix NUM... |
1759 1760 1761 1762 |
if (!(gfp_mask & __GFP_NOFAIL)) { /* The OOM killer will not help higher order allocs */ if (order > PAGE_ALLOC_COSTLY_ORDER) goto out; |
03668b3ce oom: avoid oom ki... |
1763 1764 1765 |
/* The OOM killer does not needlessly kill tasks for lowmem */ if (high_zoneidx < ZONE_NORMAL) goto out; |
4365a5676 oom-kill: fix NUM... |
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 |
/* * GFP_THISNODE contains __GFP_NORETRY and we never hit this. * Sanity check for bare calls of __GFP_THISNODE, not real OOM. * The caller should handle page allocation failure by itself if * it specifies __GFP_THISNODE. * Note: Hugepage uses it but will hit PAGE_ALLOC_COSTLY_ORDER. */ if (gfp_mask & __GFP_THISNODE) goto out; } |
11e33f6a5 page allocator: b... |
1776 |
/* Exhausted what can be done so it's blamo time */ |
4365a5676 oom-kill: fix NUM... |
1777 |
out_of_memory(zonelist, gfp_mask, order, nodemask); |
11e33f6a5 page allocator: b... |
1778 1779 1780 1781 1782 |
out: clear_zonelist_oom(zonelist, gfp_mask); return page; } |
56de7263f mm: compaction: d... |
1783 1784 1785 1786 1787 1788 |
#ifdef CONFIG_COMPACTION /* Try memory compaction for high-order allocations before reclaim */ static struct page * __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, enum zone_type high_zoneidx, nodemask_t *nodemask, int alloc_flags, struct zone *preferred_zone, |
77f1fe6b0 mm: migration: al... |
1789 1790 |
int migratetype, unsigned long *did_some_progress, bool sync_migration) |
56de7263f mm: compaction: d... |
1791 1792 |
{ struct page *page; |
4f92e2586 mm: compaction: d... |
1793 |
if (!order || compaction_deferred(preferred_zone)) |
56de7263f mm: compaction: d... |
1794 |
return NULL; |
c06b1fca1 mm/page_alloc.c: ... |
1795 |
current->flags |= PF_MEMALLOC; |
56de7263f mm: compaction: d... |
1796 |
*did_some_progress = try_to_compact_pages(zonelist, order, gfp_mask, |
77f1fe6b0 mm: migration: al... |
1797 |
nodemask, sync_migration); |
c06b1fca1 mm/page_alloc.c: ... |
1798 |
current->flags &= ~PF_MEMALLOC; |
56de7263f mm: compaction: d... |
1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 |
if (*did_some_progress != COMPACT_SKIPPED) { /* Page migration frees to the PCP lists but we want merging */ drain_pages(get_cpu()); put_cpu(); page = get_page_from_freelist(gfp_mask, nodemask, order, zonelist, high_zoneidx, alloc_flags, preferred_zone, migratetype); if (page) { |
4f92e2586 mm: compaction: d... |
1810 1811 |
preferred_zone->compact_considered = 0; preferred_zone->compact_defer_shift = 0; |
56de7263f mm: compaction: d... |
1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 |
count_vm_event(COMPACTSUCCESS); return page; } /* * It's bad if compaction run occurs and fails. * The most likely reason is that pages exist, * but not enough to satisfy watermarks. */ count_vm_event(COMPACTFAIL); |
4f92e2586 mm: compaction: d... |
1822 |
defer_compaction(preferred_zone); |
56de7263f mm: compaction: d... |
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 |
cond_resched(); } return NULL; } #else static inline struct page * __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, enum zone_type high_zoneidx, nodemask_t *nodemask, int alloc_flags, struct zone *preferred_zone, |
77f1fe6b0 mm: migration: al... |
1834 1835 |
int migratetype, unsigned long *did_some_progress, bool sync_migration) |
56de7263f mm: compaction: d... |
1836 1837 1838 1839 |
{ return NULL; } #endif /* CONFIG_COMPACTION */ |
11e33f6a5 page allocator: b... |
1840 1841 1842 1843 |
/* The really slow allocator path where we enter direct reclaim */ static inline struct page * __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, enum zone_type high_zoneidx, |
5117f45d1 page allocator: c... |
1844 |
nodemask_t *nodemask, int alloc_flags, struct zone *preferred_zone, |
3dd282669 page allocator: c... |
1845 |
int migratetype, unsigned long *did_some_progress) |
11e33f6a5 page allocator: b... |
1846 1847 1848 |
{ struct page *page = NULL; struct reclaim_state reclaim_state; |
9ee493ce0 mm: page allocato... |
1849 |
bool drained = false; |
11e33f6a5 page allocator: b... |
1850 1851 1852 1853 1854 |
cond_resched(); /* We now go into synchronous reclaim */ cpuset_memory_pressure_bump(); |
c06b1fca1 mm/page_alloc.c: ... |
1855 |
current->flags |= PF_MEMALLOC; |
11e33f6a5 page allocator: b... |
1856 1857 |
lockdep_set_current_reclaim_state(gfp_mask); reclaim_state.reclaimed_slab = 0; |
c06b1fca1 mm/page_alloc.c: ... |
1858 |
current->reclaim_state = &reclaim_state; |
11e33f6a5 page allocator: b... |
1859 1860 |
*did_some_progress = try_to_free_pages(zonelist, order, gfp_mask, nodemask); |
c06b1fca1 mm/page_alloc.c: ... |
1861 |
current->reclaim_state = NULL; |
11e33f6a5 page allocator: b... |
1862 |
lockdep_clear_current_reclaim_state(); |
c06b1fca1 mm/page_alloc.c: ... |
1863 |
current->flags &= ~PF_MEMALLOC; |
11e33f6a5 page allocator: b... |
1864 1865 |
cond_resched(); |
9ee493ce0 mm: page allocato... |
1866 1867 |
if (unlikely(!(*did_some_progress))) return NULL; |
11e33f6a5 page allocator: b... |
1868 |
|
76d3fbf8f mm: page allocato... |
1869 1870 1871 |
/* After successful reclaim, reconsider all zones for allocation */ if (NUMA_BUILD) zlc_clear_zones_full(zonelist); |
9ee493ce0 mm: page allocato... |
1872 1873 |
retry: page = get_page_from_freelist(gfp_mask, nodemask, order, |
5117f45d1 page allocator: c... |
1874 |
zonelist, high_zoneidx, |
3dd282669 page allocator: c... |
1875 1876 |
alloc_flags, preferred_zone, migratetype); |
9ee493ce0 mm: page allocato... |
1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 |
/* * If an allocation failed after direct reclaim, it could be because * pages are pinned on the per-cpu lists. Drain them and try again */ if (!page && !drained) { drain_all_pages(); drained = true; goto retry; } |
11e33f6a5 page allocator: b... |
1887 1888 |
return page; } |
1da177e4c Linux-2.6.12-rc2 |
1889 |
/* |
11e33f6a5 page allocator: b... |
1890 1891 |
* This is called in the allocator slow-path if the allocation request is of * sufficient urgency to ignore watermarks and take other desperate measures |
1da177e4c Linux-2.6.12-rc2 |
1892 |
*/ |
11e33f6a5 page allocator: b... |
1893 1894 1895 |
static inline struct page * __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, enum zone_type high_zoneidx, |
3dd282669 page allocator: c... |
1896 1897 |
nodemask_t *nodemask, struct zone *preferred_zone, int migratetype) |
11e33f6a5 page allocator: b... |
1898 1899 1900 1901 1902 |
{ struct page *page; do { page = get_page_from_freelist(gfp_mask, nodemask, order, |
5117f45d1 page allocator: c... |
1903 |
zonelist, high_zoneidx, ALLOC_NO_WATERMARKS, |
3dd282669 page allocator: c... |
1904 |
preferred_zone, migratetype); |
11e33f6a5 page allocator: b... |
1905 1906 |
if (!page && gfp_mask & __GFP_NOFAIL) |
0e093d997 writeback: do not... |
1907 |
wait_iff_congested(preferred_zone, BLK_RW_ASYNC, HZ/50); |
11e33f6a5 page allocator: b... |
1908 1909 1910 1911 1912 1913 1914 |
} while (!page && (gfp_mask & __GFP_NOFAIL)); return page; } static inline void wake_all_kswapd(unsigned int order, struct zonelist *zonelist, |
995047488 mm: kswapd: stop ... |
1915 1916 |
enum zone_type high_zoneidx, enum zone_type classzone_idx) |
1da177e4c Linux-2.6.12-rc2 |
1917 |
{ |
dd1a239f6 mm: have zonelist... |
1918 1919 |
struct zoneref *z; struct zone *zone; |
1da177e4c Linux-2.6.12-rc2 |
1920 |
|
11e33f6a5 page allocator: b... |
1921 |
for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) |
995047488 mm: kswapd: stop ... |
1922 |
wakeup_kswapd(zone, order, classzone_idx); |
11e33f6a5 page allocator: b... |
1923 |
} |
cf40bd16f lockdep: annotate... |
1924 |
|
341ce06f6 page allocator: c... |
1925 1926 1927 |
static inline int gfp_to_alloc_flags(gfp_t gfp_mask) { |
341ce06f6 page allocator: c... |
1928 1929 |
int alloc_flags = ALLOC_WMARK_MIN | ALLOC_CPUSET; const gfp_t wait = gfp_mask & __GFP_WAIT; |
1da177e4c Linux-2.6.12-rc2 |
1930 |
|
a56f57ff9 page allocator: r... |
1931 |
/* __GFP_HIGH is assumed to be the same as ALLOC_HIGH to save a branch. */ |
e6223a3b1 mm: add casts to/... |
1932 |
BUILD_BUG_ON(__GFP_HIGH != (__force gfp_t) ALLOC_HIGH); |
933e312e7 [PATCH] fault-inj... |
1933 |
|
341ce06f6 page allocator: c... |
1934 1935 1936 1937 1938 1939 |
/* * The caller may dip into page reserves a bit more if the caller * cannot run direct reclaim, or if the caller has realtime scheduling * policy or is asking for __GFP_HIGH memory. GFP_ATOMIC requests will * set both ALLOC_HARDER (!wait) and ALLOC_HIGH (__GFP_HIGH). */ |
e6223a3b1 mm: add casts to/... |
1940 |
alloc_flags |= (__force int) (gfp_mask & __GFP_HIGH); |
1da177e4c Linux-2.6.12-rc2 |
1941 |
|
341ce06f6 page allocator: c... |
1942 |
if (!wait) { |
5c3240d92 thp: don't alloc ... |
1943 1944 1945 1946 1947 1948 |
/* * Not worth trying to allocate harder for * __GFP_NOMEMALLOC even if it can't schedule. */ if (!(gfp_mask & __GFP_NOMEMALLOC)) alloc_flags |= ALLOC_HARDER; |
523b94585 Memoryless nodes:... |
1949 |
/* |
341ce06f6 page allocator: c... |
1950 1951 |
* Ignore cpuset if GFP_ATOMIC (!wait) rather than fail alloc. * See also cpuset_zone_allowed() comment in kernel/cpuset.c. |
523b94585 Memoryless nodes:... |
1952 |
*/ |
341ce06f6 page allocator: c... |
1953 |
alloc_flags &= ~ALLOC_CPUSET; |
c06b1fca1 mm/page_alloc.c: ... |
1954 |
} else if (unlikely(rt_task(current)) && !in_interrupt()) |
341ce06f6 page allocator: c... |
1955 1956 1957 1958 |
alloc_flags |= ALLOC_HARDER; if (likely(!(gfp_mask & __GFP_NOMEMALLOC))) { if (!in_interrupt() && |
c06b1fca1 mm/page_alloc.c: ... |
1959 |
((current->flags & PF_MEMALLOC) || |
341ce06f6 page allocator: c... |
1960 1961 |
unlikely(test_thread_flag(TIF_MEMDIE)))) alloc_flags |= ALLOC_NO_WATERMARKS; |
1da177e4c Linux-2.6.12-rc2 |
1962 |
} |
6b1de9161 [PATCH] VM: fix z... |
1963 |
|
341ce06f6 page allocator: c... |
1964 1965 |
return alloc_flags; } |
11e33f6a5 page allocator: b... |
1966 1967 1968 |
static inline struct page * __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, enum zone_type high_zoneidx, |
3dd282669 page allocator: c... |
1969 1970 |
nodemask_t *nodemask, struct zone *preferred_zone, int migratetype) |
11e33f6a5 page allocator: b... |
1971 1972 1973 1974 1975 1976 |
{ const gfp_t wait = gfp_mask & __GFP_WAIT; struct page *page = NULL; int alloc_flags; unsigned long pages_reclaimed = 0; unsigned long did_some_progress; |
77f1fe6b0 mm: migration: al... |
1977 |
bool sync_migration = false; |
1da177e4c Linux-2.6.12-rc2 |
1978 |
|
952f3b51b [PATCH] GFP_THISN... |
1979 |
/* |
72807a74c page allocator: s... |
1980 1981 1982 1983 1984 |
* In the slowpath, we sanity check order to avoid ever trying to * reclaim >= MAX_ORDER areas which will never succeed. Callers may * be using allocators in order of preference for an area that is * too large. */ |
1fc28b70f page-allocator: a... |
1985 1986 |
if (order >= MAX_ORDER) { WARN_ON_ONCE(!(gfp_mask & __GFP_NOWARN)); |
72807a74c page allocator: s... |
1987 |
return NULL; |
1fc28b70f page-allocator: a... |
1988 |
} |
1da177e4c Linux-2.6.12-rc2 |
1989 |
|
952f3b51b [PATCH] GFP_THISN... |
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 |
/* * GFP_THISNODE (meaning __GFP_THISNODE, __GFP_NORETRY and * __GFP_NOWARN set) should not cause reclaim since the subsystem * (f.e. slab) using GFP_THISNODE may choose to trigger reclaim * using a larger set of nodes after it has established that the * allowed per node queues are empty and that nodes are * over allocated. */ if (NUMA_BUILD && (gfp_mask & GFP_THISNODE) == GFP_THISNODE) goto nopage; |
cc4a68514 page allocator: a... |
2000 |
restart: |
32dba98e0 thp: _GFP_NO_KSWAPD |
2001 2002 |
if (!(gfp_mask & __GFP_NO_KSWAPD)) wake_all_kswapd(order, zonelist, high_zoneidx, |
995047488 mm: kswapd: stop ... |
2003 |
zone_idx(preferred_zone)); |
1da177e4c Linux-2.6.12-rc2 |
2004 |
|
9bf2229f8 [PATCH] cpusets: ... |
2005 |
/* |
7fb1d9fca [PATCH] mm: __all... |
2006 2007 2008 |
* OK, we're below the kswapd watermark and have kicked background * reclaim. Now things get more complex, so set up alloc_flags according * to how we want to proceed. |
9bf2229f8 [PATCH] cpusets: ... |
2009 |
*/ |
341ce06f6 page allocator: c... |
2010 |
alloc_flags = gfp_to_alloc_flags(gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
2011 |
|
f33261d75 mm: fix deferred ... |
2012 2013 2014 2015 2016 2017 2018 |
/* * Find the true preferred zone if the allocation is unconstrained by * cpusets. */ if (!(alloc_flags & ALLOC_CPUSET) && !nodemask) first_zones_zonelist(zonelist, high_zoneidx, NULL, &preferred_zone); |
cfa54a0fc mm/page_alloc.c: ... |
2019 |
rebalance: |
341ce06f6 page allocator: c... |
2020 |
/* This is the last chance, in general, before the goto nopage. */ |
19770b326 mm: filter based ... |
2021 |
page = get_page_from_freelist(gfp_mask, nodemask, order, zonelist, |
341ce06f6 page allocator: c... |
2022 2023 |
high_zoneidx, alloc_flags & ~ALLOC_NO_WATERMARKS, preferred_zone, migratetype); |
7fb1d9fca [PATCH] mm: __all... |
2024 2025 |
if (page) goto got_pg; |
1da177e4c Linux-2.6.12-rc2 |
2026 |
|
11e33f6a5 page allocator: b... |
2027 |
/* Allocate without watermarks if the context allows */ |
341ce06f6 page allocator: c... |
2028 2029 2030 2031 2032 2033 |
if (alloc_flags & ALLOC_NO_WATERMARKS) { page = __alloc_pages_high_priority(gfp_mask, order, zonelist, high_zoneidx, nodemask, preferred_zone, migratetype); if (page) goto got_pg; |
1da177e4c Linux-2.6.12-rc2 |
2034 2035 2036 2037 2038 |
} /* Atomic allocations - we can't balance anything */ if (!wait) goto nopage; |
341ce06f6 page allocator: c... |
2039 |
/* Avoid recursion of direct reclaim */ |
c06b1fca1 mm/page_alloc.c: ... |
2040 |
if (current->flags & PF_MEMALLOC) |
341ce06f6 page allocator: c... |
2041 |
goto nopage; |
6583bb64f mm: avoid endless... |
2042 2043 2044 |
/* Avoid allocations with no watermarks from looping endlessly */ if (test_thread_flag(TIF_MEMDIE) && !(gfp_mask & __GFP_NOFAIL)) goto nopage; |
77f1fe6b0 mm: migration: al... |
2045 2046 2047 2048 |
/* * Try direct compaction. The first pass is asynchronous. Subsequent * attempts after direct reclaim are synchronous */ |
56de7263f mm: compaction: d... |
2049 2050 2051 2052 |
page = __alloc_pages_direct_compact(gfp_mask, order, zonelist, high_zoneidx, nodemask, alloc_flags, preferred_zone, |
77f1fe6b0 mm: migration: al... |
2053 2054 |
migratetype, &did_some_progress, sync_migration); |
56de7263f mm: compaction: d... |
2055 2056 |
if (page) goto got_pg; |
c6a140bf1 mm/compaction: re... |
2057 |
sync_migration = true; |
56de7263f mm: compaction: d... |
2058 |
|
11e33f6a5 page allocator: b... |
2059 2060 2061 2062 |
/* Try direct reclaim and then allocating */ page = __alloc_pages_direct_reclaim(gfp_mask, order, zonelist, high_zoneidx, nodemask, |
5117f45d1 page allocator: c... |
2063 |
alloc_flags, preferred_zone, |
3dd282669 page allocator: c... |
2064 |
migratetype, &did_some_progress); |
11e33f6a5 page allocator: b... |
2065 2066 |
if (page) goto got_pg; |
1da177e4c Linux-2.6.12-rc2 |
2067 |
|
e33c3b5e1 cpusets: update m... |
2068 |
/* |
11e33f6a5 page allocator: b... |
2069 2070 |
* If we failed to make any progress reclaiming, then we are * running out of options and have to consider going OOM |
e33c3b5e1 cpusets: update m... |
2071 |
*/ |
11e33f6a5 page allocator: b... |
2072 2073 |
if (!did_some_progress) { if ((gfp_mask & __GFP_FS) && !(gfp_mask & __GFP_NORETRY)) { |
7f33d49a2 mm, PM/Freezer: D... |
2074 2075 |
if (oom_killer_disabled) goto nopage; |
11e33f6a5 page allocator: b... |
2076 2077 |
page = __alloc_pages_may_oom(gfp_mask, order, zonelist, high_zoneidx, |
3dd282669 page allocator: c... |
2078 2079 |
nodemask, preferred_zone, migratetype); |
11e33f6a5 page allocator: b... |
2080 2081 |
if (page) goto got_pg; |
1da177e4c Linux-2.6.12-rc2 |
2082 |
|
03668b3ce oom: avoid oom ki... |
2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 |
if (!(gfp_mask & __GFP_NOFAIL)) { /* * The oom killer is not called for high-order * allocations that may fail, so if no progress * is being made, there are no other options and * retrying is unlikely to help. */ if (order > PAGE_ALLOC_COSTLY_ORDER) goto nopage; /* * The oom killer is not called for lowmem * allocations to prevent needlessly killing * innocent tasks. */ if (high_zoneidx < ZONE_NORMAL) goto nopage; } |
e2c55dc87 Drain per-cpu lis... |
2100 |
|
ff0ceb9de oom: serialize ou... |
2101 2102 |
goto restart; } |
1da177e4c Linux-2.6.12-rc2 |
2103 |
} |
11e33f6a5 page allocator: b... |
2104 |
/* Check if we should retry the allocation */ |
a41f24ea9 page allocator: s... |
2105 |
pages_reclaimed += did_some_progress; |
11e33f6a5 page allocator: b... |
2106 2107 |
if (should_alloc_retry(gfp_mask, order, pages_reclaimed)) { /* Wait for some write requests to complete then retry */ |
0e093d997 writeback: do not... |
2108 |
wait_iff_congested(preferred_zone, BLK_RW_ASYNC, HZ/50); |
1da177e4c Linux-2.6.12-rc2 |
2109 |
goto rebalance; |
3e7d34497 mm: vmscan: recla... |
2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 |
} else { /* * High-order allocations do not necessarily loop after * direct reclaim and reclaim/compaction depends on compaction * being called after reclaim so call directly if necessary */ page = __alloc_pages_direct_compact(gfp_mask, order, zonelist, high_zoneidx, nodemask, alloc_flags, preferred_zone, |
77f1fe6b0 mm: migration: al... |
2120 2121 |
migratetype, &did_some_progress, sync_migration); |
3e7d34497 mm: vmscan: recla... |
2122 2123 |
if (page) goto got_pg; |
1da177e4c Linux-2.6.12-rc2 |
2124 2125 2126 |
} nopage: |
a238ab5b0 mm: break out pag... |
2127 |
warn_alloc_failed(gfp_mask, order, NULL); |
b1eeab676 kmemcheck: add ho... |
2128 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
2129 |
got_pg: |
b1eeab676 kmemcheck: add ho... |
2130 2131 |
if (kmemcheck_enabled) kmemcheck_pagealloc_alloc(page, order, gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
2132 |
return page; |
11e33f6a5 page allocator: b... |
2133 |
|
1da177e4c Linux-2.6.12-rc2 |
2134 |
} |
11e33f6a5 page allocator: b... |
2135 2136 2137 2138 2139 2140 2141 2142 2143 |
/* * This is the 'heart' of the zoned buddy allocator. */ struct page * __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, nodemask_t *nodemask) { enum zone_type high_zoneidx = gfp_zone(gfp_mask); |
5117f45d1 page allocator: c... |
2144 |
struct zone *preferred_zone; |
11e33f6a5 page allocator: b... |
2145 |
struct page *page; |
3dd282669 page allocator: c... |
2146 |
int migratetype = allocflags_to_migratetype(gfp_mask); |
11e33f6a5 page allocator: b... |
2147 |
|
dcce284a2 mm: Extend gfp ma... |
2148 |
gfp_mask &= gfp_allowed_mask; |
11e33f6a5 page allocator: b... |
2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 |
lockdep_trace_alloc(gfp_mask); might_sleep_if(gfp_mask & __GFP_WAIT); if (should_fail_alloc_page(gfp_mask, order)) return NULL; /* * Check the zones suitable for the gfp_mask contain at least one * valid zone. It's possible to have an empty zonelist as a result * of GFP_THISNODE and a memoryless node */ if (unlikely(!zonelist->_zonerefs->zone)) return NULL; |
c0ff7453b cpuset,mm: fix no... |
2163 |
get_mems_allowed(); |
5117f45d1 page allocator: c... |
2164 |
/* The preferred zone is used for statistics later */ |
f33261d75 mm: fix deferred ... |
2165 2166 2167 |
first_zones_zonelist(zonelist, high_zoneidx, nodemask ? : &cpuset_current_mems_allowed, &preferred_zone); |
c0ff7453b cpuset,mm: fix no... |
2168 2169 |
if (!preferred_zone) { put_mems_allowed(); |
5117f45d1 page allocator: c... |
2170 |
return NULL; |
c0ff7453b cpuset,mm: fix no... |
2171 |
} |
5117f45d1 page allocator: c... |
2172 2173 |
/* First allocation attempt */ |
11e33f6a5 page allocator: b... |
2174 |
page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, nodemask, order, |
5117f45d1 page allocator: c... |
2175 |
zonelist, high_zoneidx, ALLOC_WMARK_LOW|ALLOC_CPUSET, |
3dd282669 page allocator: c... |
2176 |
preferred_zone, migratetype); |
11e33f6a5 page allocator: b... |
2177 2178 |
if (unlikely(!page)) page = __alloc_pages_slowpath(gfp_mask, order, |
5117f45d1 page allocator: c... |
2179 |
zonelist, high_zoneidx, nodemask, |
3dd282669 page allocator: c... |
2180 |
preferred_zone, migratetype); |
c0ff7453b cpuset,mm: fix no... |
2181 |
put_mems_allowed(); |
11e33f6a5 page allocator: b... |
2182 |
|
4b4f278c0 tracing, page-all... |
2183 |
trace_mm_page_alloc(page, order, gfp_mask, migratetype); |
11e33f6a5 page allocator: b... |
2184 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
2185 |
} |
d239171e4 page allocator: r... |
2186 |
EXPORT_SYMBOL(__alloc_pages_nodemask); |
1da177e4c Linux-2.6.12-rc2 |
2187 2188 2189 2190 |
/* * Common helper functions. */ |
920c7a5d0 mm: remove fastca... |
2191 |
unsigned long __get_free_pages(gfp_t gfp_mask, unsigned int order) |
1da177e4c Linux-2.6.12-rc2 |
2192 |
{ |
945a11136 mm: add gfp mask ... |
2193 2194 2195 2196 2197 2198 2199 |
struct page *page; /* * __get_free_pages() returns a 32-bit address, which cannot represent * a highmem page */ VM_BUG_ON((gfp_mask & __GFP_HIGHMEM) != 0); |
1da177e4c Linux-2.6.12-rc2 |
2200 2201 2202 2203 2204 |
page = alloc_pages(gfp_mask, order); if (!page) return 0; return (unsigned long) page_address(page); } |
1da177e4c Linux-2.6.12-rc2 |
2205 |
EXPORT_SYMBOL(__get_free_pages); |
920c7a5d0 mm: remove fastca... |
2206 |
unsigned long get_zeroed_page(gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
2207 |
{ |
945a11136 mm: add gfp mask ... |
2208 |
return __get_free_pages(gfp_mask | __GFP_ZERO, 0); |
1da177e4c Linux-2.6.12-rc2 |
2209 |
} |
1da177e4c Linux-2.6.12-rc2 |
2210 2211 2212 2213 2214 |
EXPORT_SYMBOL(get_zeroed_page); void __pagevec_free(struct pagevec *pvec) { int i = pagevec_count(pvec); |
4b4f278c0 tracing, page-all... |
2215 2216 |
while (--i >= 0) { trace_mm_pagevec_free(pvec->pages[i], pvec->cold); |
1da177e4c Linux-2.6.12-rc2 |
2217 |
free_hot_cold_page(pvec->pages[i], pvec->cold); |
4b4f278c0 tracing, page-all... |
2218 |
} |
1da177e4c Linux-2.6.12-rc2 |
2219 |
} |
920c7a5d0 mm: remove fastca... |
2220 |
void __free_pages(struct page *page, unsigned int order) |
1da177e4c Linux-2.6.12-rc2 |
2221 |
{ |
b5810039a [PATCH] core remo... |
2222 |
if (put_page_testzero(page)) { |
1da177e4c Linux-2.6.12-rc2 |
2223 |
if (order == 0) |
fc91668ea mm: remove free_h... |
2224 |
free_hot_cold_page(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
2225 2226 2227 2228 2229 2230 |
else __free_pages_ok(page, order); } } EXPORT_SYMBOL(__free_pages); |
920c7a5d0 mm: remove fastca... |
2231 |
void free_pages(unsigned long addr, unsigned int order) |
1da177e4c Linux-2.6.12-rc2 |
2232 2233 |
{ if (addr != 0) { |
725d704ec [PATCH] mm: VM_BU... |
2234 |
VM_BUG_ON(!virt_addr_valid((void *)addr)); |
1da177e4c Linux-2.6.12-rc2 |
2235 2236 2237 2238 2239 |
__free_pages(virt_to_page((void *)addr), order); } } EXPORT_SYMBOL(free_pages); |
ee85c2e14 mm: add alloc_pag... |
2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 |
static void *make_alloc_exact(unsigned long addr, unsigned order, size_t size) { if (addr) { unsigned long alloc_end = addr + (PAGE_SIZE << order); unsigned long used = addr + PAGE_ALIGN(size); split_page(virt_to_page((void *)addr), order); while (used < alloc_end) { free_page(used); used += PAGE_SIZE; } } return (void *)addr; } |
2be0ffe2b mm: add alloc_pag... |
2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 |
/** * alloc_pages_exact - allocate an exact number physically-contiguous pages. * @size: the number of bytes to allocate * @gfp_mask: GFP flags for the allocation * * This function is similar to alloc_pages(), except that it allocates the * minimum number of pages to satisfy the request. alloc_pages() can only * allocate memory in power-of-two pages. * * This function is also limited by MAX_ORDER. * * Memory allocated by this function must be released by free_pages_exact(). */ void *alloc_pages_exact(size_t size, gfp_t gfp_mask) { unsigned int order = get_order(size); unsigned long addr; addr = __get_free_pages(gfp_mask, order); |
ee85c2e14 mm: add alloc_pag... |
2273 |
return make_alloc_exact(addr, order, size); |
2be0ffe2b mm: add alloc_pag... |
2274 2275 2276 2277 |
} EXPORT_SYMBOL(alloc_pages_exact); /** |
ee85c2e14 mm: add alloc_pag... |
2278 2279 |
* alloc_pages_exact_nid - allocate an exact number of physically-contiguous * pages on a node. |
b5e6ab589 mm: fix kernel-do... |
2280 |
* @nid: the preferred node ID where memory should be allocated |
ee85c2e14 mm: add alloc_pag... |
2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 |
* @size: the number of bytes to allocate * @gfp_mask: GFP flags for the allocation * * Like alloc_pages_exact(), but try to allocate on node nid first before falling * back. * Note this is not alloc_pages_exact_node() which allocates on a specific node, * but is not exact. */ void *alloc_pages_exact_nid(int nid, size_t size, gfp_t gfp_mask) { unsigned order = get_order(size); struct page *p = alloc_pages_node(nid, gfp_mask, order); if (!p) return NULL; return make_alloc_exact((unsigned long)page_address(p), order, size); } EXPORT_SYMBOL(alloc_pages_exact_nid); /** |
2be0ffe2b mm: add alloc_pag... |
2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 |
* free_pages_exact - release memory allocated via alloc_pages_exact() * @virt: the value returned by alloc_pages_exact. * @size: size of allocation, same value as passed to alloc_pages_exact(). * * Release the memory allocated by a previous call to alloc_pages_exact. */ void free_pages_exact(void *virt, size_t size) { unsigned long addr = (unsigned long)virt; unsigned long end = addr + PAGE_ALIGN(size); while (addr < end) { free_page(addr); addr += PAGE_SIZE; } } EXPORT_SYMBOL(free_pages_exact); |
1da177e4c Linux-2.6.12-rc2 |
2317 2318 |
static unsigned int nr_free_zone_pages(int offset) { |
dd1a239f6 mm: have zonelist... |
2319 |
struct zoneref *z; |
54a6eb5c4 mm: use two zonel... |
2320 |
struct zone *zone; |
e310fd432 [PATCH] Fix NUMA ... |
2321 |
/* Just pick one node, since fallback list is circular */ |
1da177e4c Linux-2.6.12-rc2 |
2322 |
unsigned int sum = 0; |
0e88460da mm: introduce nod... |
2323 |
struct zonelist *zonelist = node_zonelist(numa_node_id(), GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
2324 |
|
54a6eb5c4 mm: use two zonel... |
2325 |
for_each_zone_zonelist(zone, z, zonelist, offset) { |
e310fd432 [PATCH] Fix NUMA ... |
2326 |
unsigned long size = zone->present_pages; |
418589663 page allocator: u... |
2327 |
unsigned long high = high_wmark_pages(zone); |
e310fd432 [PATCH] Fix NUMA ... |
2328 2329 |
if (size > high) sum += size - high; |
1da177e4c Linux-2.6.12-rc2 |
2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 |
} return sum; } /* * Amount of free RAM allocatable within ZONE_DMA and ZONE_NORMAL */ unsigned int nr_free_buffer_pages(void) { |
af4ca457e [PATCH] gfp_t: in... |
2340 |
return nr_free_zone_pages(gfp_zone(GFP_USER)); |
1da177e4c Linux-2.6.12-rc2 |
2341 |
} |
c2f1a551d knfsd: nfsd4: var... |
2342 |
EXPORT_SYMBOL_GPL(nr_free_buffer_pages); |
1da177e4c Linux-2.6.12-rc2 |
2343 2344 2345 2346 2347 2348 |
/* * Amount of free RAM allocatable within all zones */ unsigned int nr_free_pagecache_pages(void) { |
2a1e274ac Create the ZONE_M... |
2349 |
return nr_free_zone_pages(gfp_zone(GFP_HIGHUSER_MOVABLE)); |
1da177e4c Linux-2.6.12-rc2 |
2350 |
} |
08e0f6a97 [PATCH] Add NUMA_... |
2351 2352 |
static inline void show_node(struct zone *zone) |
1da177e4c Linux-2.6.12-rc2 |
2353 |
{ |
08e0f6a97 [PATCH] Add NUMA_... |
2354 |
if (NUMA_BUILD) |
25ba77c14 [PATCH] numa node... |
2355 |
printk("Node %d ", zone_to_nid(zone)); |
1da177e4c Linux-2.6.12-rc2 |
2356 |
} |
1da177e4c Linux-2.6.12-rc2 |
2357 |
|
1da177e4c Linux-2.6.12-rc2 |
2358 2359 2360 2361 |
void si_meminfo(struct sysinfo *val) { val->totalram = totalram_pages; val->sharedram = 0; |
d23ad4232 [PATCH] Use ZVC f... |
2362 |
val->freeram = global_page_state(NR_FREE_PAGES); |
1da177e4c Linux-2.6.12-rc2 |
2363 |
val->bufferram = nr_blockdev_pages(); |
1da177e4c Linux-2.6.12-rc2 |
2364 2365 |
val->totalhigh = totalhigh_pages; val->freehigh = nr_free_highpages(); |
1da177e4c Linux-2.6.12-rc2 |
2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 |
val->mem_unit = PAGE_SIZE; } EXPORT_SYMBOL(si_meminfo); #ifdef CONFIG_NUMA void si_meminfo_node(struct sysinfo *val, int nid) { pg_data_t *pgdat = NODE_DATA(nid); val->totalram = pgdat->node_present_pages; |
d23ad4232 [PATCH] Use ZVC f... |
2377 |
val->freeram = node_page_state(nid, NR_FREE_PAGES); |
98d2b0ebd [PATCH] reduce MA... |
2378 |
#ifdef CONFIG_HIGHMEM |
1da177e4c Linux-2.6.12-rc2 |
2379 |
val->totalhigh = pgdat->node_zones[ZONE_HIGHMEM].present_pages; |
d23ad4232 [PATCH] Use ZVC f... |
2380 2381 |
val->freehigh = zone_page_state(&pgdat->node_zones[ZONE_HIGHMEM], NR_FREE_PAGES); |
98d2b0ebd [PATCH] reduce MA... |
2382 2383 2384 2385 |
#else val->totalhigh = 0; val->freehigh = 0; #endif |
1da177e4c Linux-2.6.12-rc2 |
2386 2387 2388 |
val->mem_unit = PAGE_SIZE; } #endif |
ddd588b5d oom: suppress nod... |
2389 |
/* |
7bf02ea22 arch, mm: filter ... |
2390 2391 |
* Determine whether the node should be displayed or not, depending on whether * SHOW_MEM_FILTER_NODES was passed to show_free_areas(). |
ddd588b5d oom: suppress nod... |
2392 |
*/ |
7bf02ea22 arch, mm: filter ... |
2393 |
bool skip_free_areas_node(unsigned int flags, int nid) |
ddd588b5d oom: suppress nod... |
2394 2395 2396 2397 2398 2399 2400 |
{ bool ret = false; if (!(flags & SHOW_MEM_FILTER_NODES)) goto out; get_mems_allowed(); |
7bf02ea22 arch, mm: filter ... |
2401 |
ret = !node_isset(nid, cpuset_current_mems_allowed); |
ddd588b5d oom: suppress nod... |
2402 2403 2404 2405 |
put_mems_allowed(); out: return ret; } |
1da177e4c Linux-2.6.12-rc2 |
2406 2407 2408 2409 2410 2411 |
#define K(x) ((x) << (PAGE_SHIFT-10)) /* * Show free area list (used inside shift_scroll-lock stuff) * We also calculate the percentage fragmentation. We do this by counting the * memory on each free list with the exception of the first item on the list. |
ddd588b5d oom: suppress nod... |
2412 2413 |
* Suppresses nodes that are not allowed by current's cpuset if * SHOW_MEM_FILTER_NODES is passed. |
1da177e4c Linux-2.6.12-rc2 |
2414 |
*/ |
7bf02ea22 arch, mm: filter ... |
2415 |
void show_free_areas(unsigned int filter) |
1da177e4c Linux-2.6.12-rc2 |
2416 |
{ |
c72419138 [PATCH] Condense ... |
2417 |
int cpu; |
1da177e4c Linux-2.6.12-rc2 |
2418 |
struct zone *zone; |
ee99c71c5 mm: introduce for... |
2419 |
for_each_populated_zone(zone) { |
7bf02ea22 arch, mm: filter ... |
2420 |
if (skip_free_areas_node(filter, zone_to_nid(zone))) |
ddd588b5d oom: suppress nod... |
2421 |
continue; |
c72419138 [PATCH] Condense ... |
2422 2423 2424 |
show_node(zone); printk("%s per-cpu: ", zone->name); |
1da177e4c Linux-2.6.12-rc2 |
2425 |
|
6b482c677 [PATCH] Don't pri... |
2426 |
for_each_online_cpu(cpu) { |
1da177e4c Linux-2.6.12-rc2 |
2427 |
struct per_cpu_pageset *pageset; |
99dcc3e5a this_cpu: Page al... |
2428 |
pageset = per_cpu_ptr(zone->pageset, cpu); |
1da177e4c Linux-2.6.12-rc2 |
2429 |
|
3dfa5721f Page allocator: g... |
2430 2431 2432 2433 |
printk("CPU %4d: hi:%5d, btch:%4d usd:%4d ", cpu, pageset->pcp.high, pageset->pcp.batch, pageset->pcp.count); |
1da177e4c Linux-2.6.12-rc2 |
2434 2435 |
} } |
a731286de mm: vmstat: add i... |
2436 2437 2438 2439 |
printk("active_anon:%lu inactive_anon:%lu isolated_anon:%lu " " active_file:%lu inactive_file:%lu isolated_file:%lu " |
7b854121e Unevictable LRU P... |
2440 |
" unevictable:%lu" |
b76146ed1 revert "mm: oom a... |
2441 2442 |
" dirty:%lu writeback:%lu unstable:%lu " |
3701b0332 mm: show_free_are... |
2443 2444 |
" free:%lu slab_reclaimable:%lu slab_unreclaimable:%lu " |
4b02108ac mm: oom analysis:... |
2445 2446 |
" mapped:%lu shmem:%lu pagetables:%lu bounce:%lu ", |
4f98a2fee vmscan: split LRU... |
2447 |
global_page_state(NR_ACTIVE_ANON), |
4f98a2fee vmscan: split LRU... |
2448 |
global_page_state(NR_INACTIVE_ANON), |
a731286de mm: vmstat: add i... |
2449 2450 |
global_page_state(NR_ISOLATED_ANON), global_page_state(NR_ACTIVE_FILE), |
4f98a2fee vmscan: split LRU... |
2451 |
global_page_state(NR_INACTIVE_FILE), |
a731286de mm: vmstat: add i... |
2452 |
global_page_state(NR_ISOLATED_FILE), |
7b854121e Unevictable LRU P... |
2453 |
global_page_state(NR_UNEVICTABLE), |
b1e7a8fd8 [PATCH] zoned vm ... |
2454 |
global_page_state(NR_FILE_DIRTY), |
ce866b34a [PATCH] zoned vm ... |
2455 |
global_page_state(NR_WRITEBACK), |
fd39fc856 [PATCH] zoned vm ... |
2456 |
global_page_state(NR_UNSTABLE_NFS), |
d23ad4232 [PATCH] Use ZVC f... |
2457 |
global_page_state(NR_FREE_PAGES), |
3701b0332 mm: show_free_are... |
2458 2459 |
global_page_state(NR_SLAB_RECLAIMABLE), global_page_state(NR_SLAB_UNRECLAIMABLE), |
65ba55f50 [PATCH] zoned vm ... |
2460 |
global_page_state(NR_FILE_MAPPED), |
4b02108ac mm: oom analysis:... |
2461 |
global_page_state(NR_SHMEM), |
a25700a53 [PATCH] mm: show ... |
2462 2463 |
global_page_state(NR_PAGETABLE), global_page_state(NR_BOUNCE)); |
1da177e4c Linux-2.6.12-rc2 |
2464 |
|
ee99c71c5 mm: introduce for... |
2465 |
for_each_populated_zone(zone) { |
1da177e4c Linux-2.6.12-rc2 |
2466 |
int i; |
7bf02ea22 arch, mm: filter ... |
2467 |
if (skip_free_areas_node(filter, zone_to_nid(zone))) |
ddd588b5d oom: suppress nod... |
2468 |
continue; |
1da177e4c Linux-2.6.12-rc2 |
2469 2470 2471 2472 2473 2474 |
show_node(zone); printk("%s" " free:%lukB" " min:%lukB" " low:%lukB" " high:%lukB" |
4f98a2fee vmscan: split LRU... |
2475 2476 2477 2478 |
" active_anon:%lukB" " inactive_anon:%lukB" " active_file:%lukB" " inactive_file:%lukB" |
7b854121e Unevictable LRU P... |
2479 |
" unevictable:%lukB" |
a731286de mm: vmstat: add i... |
2480 2481 |
" isolated(anon):%lukB" " isolated(file):%lukB" |
1da177e4c Linux-2.6.12-rc2 |
2482 |
" present:%lukB" |
4a0aa73f1 mm: oom analysis:... |
2483 2484 2485 2486 |
" mlocked:%lukB" " dirty:%lukB" " writeback:%lukB" " mapped:%lukB" |
4b02108ac mm: oom analysis:... |
2487 |
" shmem:%lukB" |
4a0aa73f1 mm: oom analysis:... |
2488 2489 |
" slab_reclaimable:%lukB" " slab_unreclaimable:%lukB" |
c6a7f5728 mm: oom analysis:... |
2490 |
" kernel_stack:%lukB" |
4a0aa73f1 mm: oom analysis:... |
2491 2492 2493 2494 |
" pagetables:%lukB" " unstable:%lukB" " bounce:%lukB" " writeback_tmp:%lukB" |
1da177e4c Linux-2.6.12-rc2 |
2495 2496 2497 2498 2499 |
" pages_scanned:%lu" " all_unreclaimable? %s" " ", zone->name, |
88f5acf88 mm: page allocato... |
2500 |
K(zone_page_state(zone, NR_FREE_PAGES)), |
418589663 page allocator: u... |
2501 2502 2503 |
K(min_wmark_pages(zone)), K(low_wmark_pages(zone)), K(high_wmark_pages(zone)), |
4f98a2fee vmscan: split LRU... |
2504 2505 2506 2507 |
K(zone_page_state(zone, NR_ACTIVE_ANON)), K(zone_page_state(zone, NR_INACTIVE_ANON)), K(zone_page_state(zone, NR_ACTIVE_FILE)), K(zone_page_state(zone, NR_INACTIVE_FILE)), |
7b854121e Unevictable LRU P... |
2508 |
K(zone_page_state(zone, NR_UNEVICTABLE)), |
a731286de mm: vmstat: add i... |
2509 2510 |
K(zone_page_state(zone, NR_ISOLATED_ANON)), K(zone_page_state(zone, NR_ISOLATED_FILE)), |
1da177e4c Linux-2.6.12-rc2 |
2511 |
K(zone->present_pages), |
4a0aa73f1 mm: oom analysis:... |
2512 2513 2514 2515 |
K(zone_page_state(zone, NR_MLOCK)), K(zone_page_state(zone, NR_FILE_DIRTY)), K(zone_page_state(zone, NR_WRITEBACK)), K(zone_page_state(zone, NR_FILE_MAPPED)), |
4b02108ac mm: oom analysis:... |
2516 |
K(zone_page_state(zone, NR_SHMEM)), |
4a0aa73f1 mm: oom analysis:... |
2517 2518 |
K(zone_page_state(zone, NR_SLAB_RECLAIMABLE)), K(zone_page_state(zone, NR_SLAB_UNRECLAIMABLE)), |
c6a7f5728 mm: oom analysis:... |
2519 2520 |
zone_page_state(zone, NR_KERNEL_STACK) * THREAD_SIZE / 1024, |
4a0aa73f1 mm: oom analysis:... |
2521 2522 2523 2524 |
K(zone_page_state(zone, NR_PAGETABLE)), K(zone_page_state(zone, NR_UNSTABLE_NFS)), K(zone_page_state(zone, NR_BOUNCE)), K(zone_page_state(zone, NR_WRITEBACK_TEMP)), |
1da177e4c Linux-2.6.12-rc2 |
2525 |
zone->pages_scanned, |
93e4a89a8 mm: restore zone-... |
2526 |
(zone->all_unreclaimable ? "yes" : "no") |
1da177e4c Linux-2.6.12-rc2 |
2527 2528 2529 2530 2531 2532 2533 |
); printk("lowmem_reserve[]:"); for (i = 0; i < MAX_NR_ZONES; i++) printk(" %lu", zone->lowmem_reserve[i]); printk(" "); } |
ee99c71c5 mm: introduce for... |
2534 |
for_each_populated_zone(zone) { |
8f9de51a4 [PATCH] printk() ... |
2535 |
unsigned long nr[MAX_ORDER], flags, order, total = 0; |
1da177e4c Linux-2.6.12-rc2 |
2536 |
|
7bf02ea22 arch, mm: filter ... |
2537 |
if (skip_free_areas_node(filter, zone_to_nid(zone))) |
ddd588b5d oom: suppress nod... |
2538 |
continue; |
1da177e4c Linux-2.6.12-rc2 |
2539 2540 |
show_node(zone); printk("%s: ", zone->name); |
1da177e4c Linux-2.6.12-rc2 |
2541 2542 2543 |
spin_lock_irqsave(&zone->lock, flags); for (order = 0; order < MAX_ORDER; order++) { |
8f9de51a4 [PATCH] printk() ... |
2544 2545 |
nr[order] = zone->free_area[order].nr_free; total += nr[order] << order; |
1da177e4c Linux-2.6.12-rc2 |
2546 2547 |
} spin_unlock_irqrestore(&zone->lock, flags); |
8f9de51a4 [PATCH] printk() ... |
2548 2549 |
for (order = 0; order < MAX_ORDER; order++) printk("%lu*%lukB ", nr[order], K(1UL) << order); |
1da177e4c Linux-2.6.12-rc2 |
2550 2551 2552 |
printk("= %lukB ", K(total)); } |
e6f3602d2 Include count of ... |
2553 2554 |
printk("%ld total pagecache pages ", global_page_state(NR_FILE_PAGES)); |
1da177e4c Linux-2.6.12-rc2 |
2555 2556 |
show_swap_cache_info(); } |
19770b326 mm: filter based ... |
2557 2558 2559 2560 2561 |
static void zoneref_set_zone(struct zone *zone, struct zoneref *zoneref) { zoneref->zone = zone; zoneref->zone_idx = zone_idx(zone); } |
1da177e4c Linux-2.6.12-rc2 |
2562 2563 |
/* * Builds allocation fallback zone lists. |
1a93205bd [PATCH] mm: simpl... |
2564 2565 |
* * Add all populated zones of a node to the zonelist. |
1da177e4c Linux-2.6.12-rc2 |
2566 |
*/ |
f0c0b2b80 change zonelist o... |
2567 2568 |
static int build_zonelists_node(pg_data_t *pgdat, struct zonelist *zonelist, int nr_zones, enum zone_type zone_type) |
1da177e4c Linux-2.6.12-rc2 |
2569 |
{ |
1a93205bd [PATCH] mm: simpl... |
2570 |
struct zone *zone; |
98d2b0ebd [PATCH] reduce MA... |
2571 |
BUG_ON(zone_type >= MAX_NR_ZONES); |
2f6726e54 [PATCH] Apply typ... |
2572 |
zone_type++; |
02a68a5eb [PATCH] Fix zone ... |
2573 2574 |
do { |
2f6726e54 [PATCH] Apply typ... |
2575 |
zone_type--; |
070f80326 [PATCH] build_zon... |
2576 |
zone = pgdat->node_zones + zone_type; |
1a93205bd [PATCH] mm: simpl... |
2577 |
if (populated_zone(zone)) { |
dd1a239f6 mm: have zonelist... |
2578 2579 |
zoneref_set_zone(zone, &zonelist->_zonerefs[nr_zones++]); |
070f80326 [PATCH] build_zon... |
2580 |
check_highest_zone(zone_type); |
1da177e4c Linux-2.6.12-rc2 |
2581 |
} |
02a68a5eb [PATCH] Fix zone ... |
2582 |
|
2f6726e54 [PATCH] Apply typ... |
2583 |
} while (zone_type); |
070f80326 [PATCH] build_zon... |
2584 |
return nr_zones; |
1da177e4c Linux-2.6.12-rc2 |
2585 |
} |
f0c0b2b80 change zonelist o... |
2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 |
/* * zonelist_order: * 0 = automatic detection of better ordering. * 1 = order by ([node] distance, -zonetype) * 2 = order by (-zonetype, [node] distance) * * If not NUMA, ZONELIST_ORDER_ZONE and ZONELIST_ORDER_NODE will create * the same zonelist. So only NUMA can configure this param. */ #define ZONELIST_ORDER_DEFAULT 0 #define ZONELIST_ORDER_NODE 1 #define ZONELIST_ORDER_ZONE 2 /* zonelist order in the kernel. * set_zonelist_order() will set this to NODE or ZONE. */ static int current_zonelist_order = ZONELIST_ORDER_DEFAULT; static char zonelist_order_name[3][8] = {"Default", "Node", "Zone"}; |
1da177e4c Linux-2.6.12-rc2 |
2605 |
#ifdef CONFIG_NUMA |
f0c0b2b80 change zonelist o... |
2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 |
/* The value user specified ....changed by config */ static int user_zonelist_order = ZONELIST_ORDER_DEFAULT; /* string for sysctl */ #define NUMA_ZONELIST_ORDER_LEN 16 char numa_zonelist_order[16] = "default"; /* * interface for configure zonelist ordering. * command line option "numa_zonelist_order" * = "[dD]efault - default, automatic configuration. * = "[nN]ode - order by node locality, then by zone within node * = "[zZ]one - order by zone, then by locality within zone */ static int __parse_numa_zonelist_order(char *s) { if (*s == 'd' || *s == 'D') { user_zonelist_order = ZONELIST_ORDER_DEFAULT; } else if (*s == 'n' || *s == 'N') { user_zonelist_order = ZONELIST_ORDER_NODE; } else if (*s == 'z' || *s == 'Z') { user_zonelist_order = ZONELIST_ORDER_ZONE; } else { printk(KERN_WARNING "Ignoring invalid numa_zonelist_order value: " "%s ", s); return -EINVAL; } return 0; } static __init int setup_numa_zonelist_order(char *s) { |
ecb256f81 mm: set correct n... |
2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 |
int ret; if (!s) return 0; ret = __parse_numa_zonelist_order(s); if (ret == 0) strlcpy(numa_zonelist_order, s, NUMA_ZONELIST_ORDER_LEN); return ret; |
f0c0b2b80 change zonelist o... |
2650 2651 2652 2653 2654 2655 2656 |
} early_param("numa_zonelist_order", setup_numa_zonelist_order); /* * sysctl handler for numa_zonelist_order */ int numa_zonelist_order_handler(ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
2657 |
void __user *buffer, size_t *length, |
f0c0b2b80 change zonelist o... |
2658 2659 2660 2661 |
loff_t *ppos) { char saved_string[NUMA_ZONELIST_ORDER_LEN]; int ret; |
443c6f145 SYSCTL: Add a mut... |
2662 |
static DEFINE_MUTEX(zl_order_mutex); |
f0c0b2b80 change zonelist o... |
2663 |
|
443c6f145 SYSCTL: Add a mut... |
2664 |
mutex_lock(&zl_order_mutex); |
f0c0b2b80 change zonelist o... |
2665 |
if (write) |
443c6f145 SYSCTL: Add a mut... |
2666 |
strcpy(saved_string, (char*)table->data); |
8d65af789 sysctl: remove "s... |
2667 |
ret = proc_dostring(table, write, buffer, length, ppos); |
f0c0b2b80 change zonelist o... |
2668 |
if (ret) |
443c6f145 SYSCTL: Add a mut... |
2669 |
goto out; |
f0c0b2b80 change zonelist o... |
2670 2671 2672 2673 2674 2675 2676 2677 2678 |
if (write) { int oldval = user_zonelist_order; if (__parse_numa_zonelist_order((char*)table->data)) { /* * bogus value. restore saved string */ strncpy((char*)table->data, saved_string, NUMA_ZONELIST_ORDER_LEN); user_zonelist_order = oldval; |
4eaf3f643 mem-hotplug: fix ... |
2679 2680 |
} else if (oldval != user_zonelist_order) { mutex_lock(&zonelists_mutex); |
1f522509c mem-hotplug: avoi... |
2681 |
build_all_zonelists(NULL); |
4eaf3f643 mem-hotplug: fix ... |
2682 2683 |
mutex_unlock(&zonelists_mutex); } |
f0c0b2b80 change zonelist o... |
2684 |
} |
443c6f145 SYSCTL: Add a mut... |
2685 2686 2687 |
out: mutex_unlock(&zl_order_mutex); return ret; |
f0c0b2b80 change zonelist o... |
2688 |
} |
62bc62a87 page allocator: u... |
2689 |
#define MAX_NODE_LOAD (nr_online_nodes) |
f0c0b2b80 change zonelist o... |
2690 |
static int node_load[MAX_NUMNODES]; |
1da177e4c Linux-2.6.12-rc2 |
2691 |
/** |
4dc3b16ba [PATCH] DocBook: ... |
2692 |
* find_next_best_node - find the next node that should appear in a given node's fallback list |
1da177e4c Linux-2.6.12-rc2 |
2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 |
* @node: node whose fallback list we're appending * @used_node_mask: nodemask_t of already used nodes * * We use a number of factors to determine which is the next node that should * appear on a given node's fallback list. The node should not have appeared * already in @node's fallback list, and it should be the next closest node * according to the distance array (which contains arbitrary distance values * from each node to each node in the system), and should also prefer nodes * with no CPUs, since presumably they'll have very little allocation pressure * on them otherwise. * It returns -1 if no node is found. */ |
f0c0b2b80 change zonelist o... |
2705 |
static int find_next_best_node(int node, nodemask_t *used_node_mask) |
1da177e4c Linux-2.6.12-rc2 |
2706 |
{ |
4cf808eb4 [PATCH] Handle ho... |
2707 |
int n, val; |
1da177e4c Linux-2.6.12-rc2 |
2708 2709 |
int min_val = INT_MAX; int best_node = -1; |
a70f73028 cpumask: replace ... |
2710 |
const struct cpumask *tmp = cpumask_of_node(0); |
1da177e4c Linux-2.6.12-rc2 |
2711 |
|
4cf808eb4 [PATCH] Handle ho... |
2712 2713 2714 2715 2716 |
/* Use the local node if we haven't already */ if (!node_isset(node, *used_node_mask)) { node_set(node, *used_node_mask); return node; } |
1da177e4c Linux-2.6.12-rc2 |
2717 |
|
37b07e416 memoryless nodes:... |
2718 |
for_each_node_state(n, N_HIGH_MEMORY) { |
1da177e4c Linux-2.6.12-rc2 |
2719 2720 2721 2722 |
/* Don't want a node to appear more than once */ if (node_isset(n, *used_node_mask)) continue; |
1da177e4c Linux-2.6.12-rc2 |
2723 2724 |
/* Use the distance array to find the distance */ val = node_distance(node, n); |
4cf808eb4 [PATCH] Handle ho... |
2725 2726 |
/* Penalize nodes under us ("prefer the next node") */ val += (n < node); |
1da177e4c Linux-2.6.12-rc2 |
2727 |
/* Give preference to headless and unused nodes */ |
a70f73028 cpumask: replace ... |
2728 2729 |
tmp = cpumask_of_node(n); if (!cpumask_empty(tmp)) |
1da177e4c Linux-2.6.12-rc2 |
2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 |
val += PENALTY_FOR_NODE_WITH_CPUS; /* Slight preference for less loaded node */ val *= (MAX_NODE_LOAD*MAX_NUMNODES); val += node_load[n]; if (val < min_val) { min_val = val; best_node = n; } } if (best_node >= 0) node_set(best_node, *used_node_mask); return best_node; } |
f0c0b2b80 change zonelist o... |
2747 2748 2749 2750 2751 2752 2753 |
/* * Build zonelists ordered by node and zones within node. * This results in maximum locality--normal zone overflows into local * DMA zone, if any--but risks exhausting DMA zone. */ static void build_zonelists_in_node_order(pg_data_t *pgdat, int node) |
1da177e4c Linux-2.6.12-rc2 |
2754 |
{ |
f0c0b2b80 change zonelist o... |
2755 |
int j; |
1da177e4c Linux-2.6.12-rc2 |
2756 |
struct zonelist *zonelist; |
f0c0b2b80 change zonelist o... |
2757 |
|
54a6eb5c4 mm: use two zonel... |
2758 |
zonelist = &pgdat->node_zonelists[0]; |
dd1a239f6 mm: have zonelist... |
2759 |
for (j = 0; zonelist->_zonerefs[j].zone != NULL; j++) |
54a6eb5c4 mm: use two zonel... |
2760 2761 2762 |
; j = build_zonelists_node(NODE_DATA(node), zonelist, j, MAX_NR_ZONES - 1); |
dd1a239f6 mm: have zonelist... |
2763 2764 |
zonelist->_zonerefs[j].zone = NULL; zonelist->_zonerefs[j].zone_idx = 0; |
f0c0b2b80 change zonelist o... |
2765 2766 2767 |
} /* |
523b94585 Memoryless nodes:... |
2768 2769 2770 2771 |
* Build gfp_thisnode zonelists */ static void build_thisnode_zonelists(pg_data_t *pgdat) { |
523b94585 Memoryless nodes:... |
2772 2773 |
int j; struct zonelist *zonelist; |
54a6eb5c4 mm: use two zonel... |
2774 2775 |
zonelist = &pgdat->node_zonelists[1]; j = build_zonelists_node(pgdat, zonelist, 0, MAX_NR_ZONES - 1); |
dd1a239f6 mm: have zonelist... |
2776 2777 |
zonelist->_zonerefs[j].zone = NULL; zonelist->_zonerefs[j].zone_idx = 0; |
523b94585 Memoryless nodes:... |
2778 2779 2780 |
} /* |
f0c0b2b80 change zonelist o... |
2781 2782 2783 2784 2785 2786 2787 2788 2789 |
* Build zonelists ordered by zone and nodes within zones. * This results in conserving DMA zone[s] until all Normal memory is * exhausted, but results in overflowing to remote node while memory * may still exist in local DMA zone. */ static int node_order[MAX_NUMNODES]; static void build_zonelists_in_zone_order(pg_data_t *pgdat, int nr_nodes) { |
f0c0b2b80 change zonelist o... |
2790 2791 2792 2793 |
int pos, j, node; int zone_type; /* needs to be signed */ struct zone *z; struct zonelist *zonelist; |
54a6eb5c4 mm: use two zonel... |
2794 2795 2796 2797 2798 2799 2800 |
zonelist = &pgdat->node_zonelists[0]; pos = 0; for (zone_type = MAX_NR_ZONES - 1; zone_type >= 0; zone_type--) { for (j = 0; j < nr_nodes; j++) { node = node_order[j]; z = &NODE_DATA(node)->node_zones[zone_type]; if (populated_zone(z)) { |
dd1a239f6 mm: have zonelist... |
2801 2802 |
zoneref_set_zone(z, &zonelist->_zonerefs[pos++]); |
54a6eb5c4 mm: use two zonel... |
2803 |
check_highest_zone(zone_type); |
f0c0b2b80 change zonelist o... |
2804 2805 |
} } |
f0c0b2b80 change zonelist o... |
2806 |
} |
dd1a239f6 mm: have zonelist... |
2807 2808 |
zonelist->_zonerefs[pos].zone = NULL; zonelist->_zonerefs[pos].zone_idx = 0; |
f0c0b2b80 change zonelist o... |
2809 2810 2811 2812 2813 2814 2815 2816 2817 |
} static int default_zonelist_order(void) { int nid, zone_type; unsigned long low_kmem_size,total_size; struct zone *z; int average_size; /* |
883931612 Fix typos in comm... |
2818 |
* ZONE_DMA and ZONE_DMA32 can be very small area in the system. |
f0c0b2b80 change zonelist o... |
2819 2820 |
* If they are really small and used heavily, the system can fall * into OOM very easily. |
e325c90ff mm: default to no... |
2821 |
* This function detect ZONE_DMA/DMA32 size and configures zone order. |
f0c0b2b80 change zonelist o... |
2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 |
*/ /* Is there ZONE_NORMAL ? (ex. ppc has only DMA zone..) */ low_kmem_size = 0; total_size = 0; for_each_online_node(nid) { for (zone_type = 0; zone_type < MAX_NR_ZONES; zone_type++) { z = &NODE_DATA(nid)->node_zones[zone_type]; if (populated_zone(z)) { if (zone_type < ZONE_NORMAL) low_kmem_size += z->present_pages; total_size += z->present_pages; |
e325c90ff mm: default to no... |
2833 2834 2835 2836 2837 2838 2839 2840 2841 |
} else if (zone_type == ZONE_NORMAL) { /* * If any node has only lowmem, then node order * is preferred to allow kernel allocations * locally; otherwise, they can easily infringe * on other nodes when there is an abundance of * lowmem available to allocate from. */ return ZONELIST_ORDER_NODE; |
f0c0b2b80 change zonelist o... |
2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 |
} } } if (!low_kmem_size || /* there are no DMA area. */ low_kmem_size > total_size/2) /* DMA/DMA32 is big. */ return ZONELIST_ORDER_NODE; /* * look into each node's config. * If there is a node whose DMA/DMA32 memory is very big area on * local memory, NODE_ORDER may be suitable. */ |
37b07e416 memoryless nodes:... |
2853 2854 |
average_size = total_size / (nodes_weight(node_states[N_HIGH_MEMORY]) + 1); |
f0c0b2b80 change zonelist o... |
2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 |
for_each_online_node(nid) { low_kmem_size = 0; total_size = 0; for (zone_type = 0; zone_type < MAX_NR_ZONES; zone_type++) { z = &NODE_DATA(nid)->node_zones[zone_type]; if (populated_zone(z)) { if (zone_type < ZONE_NORMAL) low_kmem_size += z->present_pages; total_size += z->present_pages; } } if (low_kmem_size && total_size > average_size && /* ignore small node */ low_kmem_size > total_size * 70/100) return ZONELIST_ORDER_NODE; } return ZONELIST_ORDER_ZONE; } static void set_zonelist_order(void) { if (user_zonelist_order == ZONELIST_ORDER_DEFAULT) current_zonelist_order = default_zonelist_order(); else current_zonelist_order = user_zonelist_order; } static void build_zonelists(pg_data_t *pgdat) { int j, node, load; enum zone_type i; |
1da177e4c Linux-2.6.12-rc2 |
2886 |
nodemask_t used_mask; |
f0c0b2b80 change zonelist o... |
2887 2888 2889 |
int local_node, prev_node; struct zonelist *zonelist; int order = current_zonelist_order; |
1da177e4c Linux-2.6.12-rc2 |
2890 2891 |
/* initialize zonelists */ |
523b94585 Memoryless nodes:... |
2892 |
for (i = 0; i < MAX_ZONELISTS; i++) { |
1da177e4c Linux-2.6.12-rc2 |
2893 |
zonelist = pgdat->node_zonelists + i; |
dd1a239f6 mm: have zonelist... |
2894 2895 |
zonelist->_zonerefs[0].zone = NULL; zonelist->_zonerefs[0].zone_idx = 0; |
1da177e4c Linux-2.6.12-rc2 |
2896 2897 2898 2899 |
} /* NUMA-aware ordering of nodes */ local_node = pgdat->node_id; |
62bc62a87 page allocator: u... |
2900 |
load = nr_online_nodes; |
1da177e4c Linux-2.6.12-rc2 |
2901 2902 |
prev_node = local_node; nodes_clear(used_mask); |
f0c0b2b80 change zonelist o... |
2903 |
|
f0c0b2b80 change zonelist o... |
2904 2905 |
memset(node_order, 0, sizeof(node_order)); j = 0; |
1da177e4c Linux-2.6.12-rc2 |
2906 |
while ((node = find_next_best_node(local_node, &used_mask)) >= 0) { |
9eeff2395 [PATCH] Zone recl... |
2907 2908 2909 2910 2911 2912 2913 2914 |
int distance = node_distance(local_node, node); /* * If another node is sufficiently far away then it is better * to reclaim pages in a zone before going off node. */ if (distance > RECLAIM_DISTANCE) zone_reclaim_mode = 1; |
1da177e4c Linux-2.6.12-rc2 |
2915 2916 2917 2918 2919 |
/* * We don't want to pressure a particular node. * So adding penalty to the first node in same * distance group to make it round-robin. */ |
9eeff2395 [PATCH] Zone recl... |
2920 |
if (distance != node_distance(local_node, prev_node)) |
f0c0b2b80 change zonelist o... |
2921 |
node_load[node] = load; |
1da177e4c Linux-2.6.12-rc2 |
2922 2923 |
prev_node = node; load--; |
f0c0b2b80 change zonelist o... |
2924 2925 2926 2927 2928 |
if (order == ZONELIST_ORDER_NODE) build_zonelists_in_node_order(pgdat, node); else node_order[j++] = node; /* remember order */ } |
1da177e4c Linux-2.6.12-rc2 |
2929 |
|
f0c0b2b80 change zonelist o... |
2930 2931 2932 |
if (order == ZONELIST_ORDER_ZONE) { /* calculate node order -- i.e., DMA last! */ build_zonelists_in_zone_order(pgdat, j); |
1da177e4c Linux-2.6.12-rc2 |
2933 |
} |
523b94585 Memoryless nodes:... |
2934 2935 |
build_thisnode_zonelists(pgdat); |
1da177e4c Linux-2.6.12-rc2 |
2936 |
} |
9276b1bc9 [PATCH] memory pa... |
2937 |
/* Construct the zonelist performance cache - see further mmzone.h */ |
f0c0b2b80 change zonelist o... |
2938 |
static void build_zonelist_cache(pg_data_t *pgdat) |
9276b1bc9 [PATCH] memory pa... |
2939 |
{ |
54a6eb5c4 mm: use two zonel... |
2940 2941 |
struct zonelist *zonelist; struct zonelist_cache *zlc; |
dd1a239f6 mm: have zonelist... |
2942 |
struct zoneref *z; |
9276b1bc9 [PATCH] memory pa... |
2943 |
|
54a6eb5c4 mm: use two zonel... |
2944 2945 2946 |
zonelist = &pgdat->node_zonelists[0]; zonelist->zlcache_ptr = zlc = &zonelist->zlcache; bitmap_zero(zlc->fullzones, MAX_ZONES_PER_ZONELIST); |
dd1a239f6 mm: have zonelist... |
2947 2948 |
for (z = zonelist->_zonerefs; z->zone; z++) zlc->z_to_n[z - zonelist->_zonerefs] = zonelist_node_idx(z); |
9276b1bc9 [PATCH] memory pa... |
2949 |
} |
7aac78988 numa: introduce n... |
2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 |
#ifdef CONFIG_HAVE_MEMORYLESS_NODES /* * Return node id of node used for "local" allocations. * I.e., first node id of first zone in arg node's generic zonelist. * Used for initializing percpu 'numa_mem', which is used primarily * for kernel allocations, so use GFP_KERNEL flags to locate zonelist. */ int local_memory_node(int node) { struct zone *zone; (void)first_zones_zonelist(node_zonelist(node, GFP_KERNEL), gfp_zone(GFP_KERNEL), NULL, &zone); return zone->node; } #endif |
f0c0b2b80 change zonelist o... |
2968 |
|
1da177e4c Linux-2.6.12-rc2 |
2969 |
#else /* CONFIG_NUMA */ |
f0c0b2b80 change zonelist o... |
2970 2971 2972 2973 2974 2975 |
static void set_zonelist_order(void) { current_zonelist_order = ZONELIST_ORDER_ZONE; } static void build_zonelists(pg_data_t *pgdat) |
1da177e4c Linux-2.6.12-rc2 |
2976 |
{ |
19655d348 [PATCH] linearly ... |
2977 |
int node, local_node; |
54a6eb5c4 mm: use two zonel... |
2978 2979 |
enum zone_type j; struct zonelist *zonelist; |
1da177e4c Linux-2.6.12-rc2 |
2980 2981 |
local_node = pgdat->node_id; |
1da177e4c Linux-2.6.12-rc2 |
2982 |
|
54a6eb5c4 mm: use two zonel... |
2983 2984 |
zonelist = &pgdat->node_zonelists[0]; j = build_zonelists_node(pgdat, zonelist, 0, MAX_NR_ZONES - 1); |
1da177e4c Linux-2.6.12-rc2 |
2985 |
|
54a6eb5c4 mm: use two zonel... |
2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 |
/* * Now we build the zonelist so that it contains the zones * of all the other nodes. * We don't want to pressure a particular node, so when * building the zones for node N, we make sure that the * zones coming right after the local ones are those from * node N+1 (modulo N) */ for (node = local_node + 1; node < MAX_NUMNODES; node++) { if (!node_online(node)) continue; j = build_zonelists_node(NODE_DATA(node), zonelist, j, MAX_NR_ZONES - 1); |
1da177e4c Linux-2.6.12-rc2 |
2999 |
} |
54a6eb5c4 mm: use two zonel... |
3000 3001 3002 3003 3004 3005 |
for (node = 0; node < local_node; node++) { if (!node_online(node)) continue; j = build_zonelists_node(NODE_DATA(node), zonelist, j, MAX_NR_ZONES - 1); } |
dd1a239f6 mm: have zonelist... |
3006 3007 |
zonelist->_zonerefs[j].zone = NULL; zonelist->_zonerefs[j].zone_idx = 0; |
1da177e4c Linux-2.6.12-rc2 |
3008 |
} |
9276b1bc9 [PATCH] memory pa... |
3009 |
/* non-NUMA variant of zonelist performance cache - just NULL zlcache_ptr */ |
f0c0b2b80 change zonelist o... |
3010 |
static void build_zonelist_cache(pg_data_t *pgdat) |
9276b1bc9 [PATCH] memory pa... |
3011 |
{ |
54a6eb5c4 mm: use two zonel... |
3012 |
pgdat->node_zonelists[0].zlcache_ptr = NULL; |
9276b1bc9 [PATCH] memory pa... |
3013 |
} |
1da177e4c Linux-2.6.12-rc2 |
3014 |
#endif /* CONFIG_NUMA */ |
99dcc3e5a this_cpu: Page al... |
3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 |
/* * Boot pageset table. One per cpu which is going to be used for all * zones and all nodes. The parameters will be set in such a way * that an item put on a list will immediately be handed over to * the buddy list. This is safe since pageset manipulation is done * with interrupts disabled. * * The boot_pagesets must be kept even after bootup is complete for * unused processors and/or zones. They do play a role for bootstrapping * hotplugged processors. * * zoneinfo_show() and maybe other functions do * not check if the processor is online before following the pageset pointer. * Other parts of the kernel may not check if the zone is available. */ static void setup_pageset(struct per_cpu_pageset *p, unsigned long batch); static DEFINE_PER_CPU(struct per_cpu_pageset, boot_pageset); |
1f522509c mem-hotplug: avoi... |
3032 |
static void setup_zone_pageset(struct zone *zone); |
99dcc3e5a this_cpu: Page al... |
3033 |
|
4eaf3f643 mem-hotplug: fix ... |
3034 3035 3036 3037 3038 |
/* * Global mutex to protect against size modification of zonelists * as well as to serialize pageset setup for the new populated zone. */ DEFINE_MUTEX(zonelists_mutex); |
9b1a4d383 stop_machine: Wea... |
3039 |
/* return values int ....just for stop_machine() */ |
1f522509c mem-hotplug: avoi... |
3040 |
static __init_refok int __build_all_zonelists(void *data) |
1da177e4c Linux-2.6.12-rc2 |
3041 |
{ |
6811378e7 [PATCH] wait_tabl... |
3042 |
int nid; |
99dcc3e5a this_cpu: Page al... |
3043 |
int cpu; |
9276b1bc9 [PATCH] memory pa... |
3044 |
|
7f9cfb310 mm: build_zonelis... |
3045 3046 3047 |
#ifdef CONFIG_NUMA memset(node_load, 0, sizeof(node_load)); #endif |
9276b1bc9 [PATCH] memory pa... |
3048 |
for_each_online_node(nid) { |
7ea1530ab Memoryless nodes:... |
3049 3050 3051 3052 |
pg_data_t *pgdat = NODE_DATA(nid); build_zonelists(pgdat); build_zonelist_cache(pgdat); |
9276b1bc9 [PATCH] memory pa... |
3053 |
} |
99dcc3e5a this_cpu: Page al... |
3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 |
/* * Initialize the boot_pagesets that are going to be used * for bootstrapping processors. The real pagesets for * each zone will be allocated later when the per cpu * allocator is available. * * boot_pagesets are used also for bootstrapping offline * cpus if the system is already booted because the pagesets * are needed to initialize allocators on a specific cpu too. * F.e. the percpu allocator needs the page allocator which * needs the percpu allocator in order to allocate its pagesets * (a chicken-egg dilemma). */ |
7aac78988 numa: introduce n... |
3068 |
for_each_possible_cpu(cpu) { |
99dcc3e5a this_cpu: Page al... |
3069 |
setup_pageset(&per_cpu(boot_pageset, cpu), 0); |
7aac78988 numa: introduce n... |
3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 |
#ifdef CONFIG_HAVE_MEMORYLESS_NODES /* * We now know the "local memory node" for each node-- * i.e., the node of the first zone in the generic zonelist. * Set up numa_mem percpu variable for on-line cpus. During * boot, only the boot cpu should be on-line; we'll init the * secondary cpus' numa_mem as they come on-line. During * node/memory hotplug, we'll fixup all on-line cpus. */ if (cpu_online(cpu)) set_cpu_numa_mem(cpu, local_memory_node(cpu_to_node(cpu))); #endif } |
6811378e7 [PATCH] wait_tabl... |
3083 3084 |
return 0; } |
4eaf3f643 mem-hotplug: fix ... |
3085 3086 3087 3088 |
/* * Called with zonelists_mutex held always * unless system_state == SYSTEM_BOOTING. */ |
9f6ae448b mm/page_alloc.c: ... |
3089 |
void __ref build_all_zonelists(void *data) |
6811378e7 [PATCH] wait_tabl... |
3090 |
{ |
f0c0b2b80 change zonelist o... |
3091 |
set_zonelist_order(); |
6811378e7 [PATCH] wait_tabl... |
3092 |
if (system_state == SYSTEM_BOOTING) { |
423b41d77 [PATCH] mm/page_a... |
3093 |
__build_all_zonelists(NULL); |
68ad8df42 mm: print out the... |
3094 |
mminit_verify_zonelist(); |
6811378e7 [PATCH] wait_tabl... |
3095 3096 |
cpuset_init_current_mems_allowed(); } else { |
183ff22bb spelling fixes: mm/ |
3097 |
/* we have to stop all cpus to guarantee there is no user |
6811378e7 [PATCH] wait_tabl... |
3098 |
of zonelist */ |
e9959f0f3 mm/page_alloc.c: ... |
3099 3100 3101 3102 3103 |
#ifdef CONFIG_MEMORY_HOTPLUG if (data) setup_zone_pageset((struct zone *)data); #endif stop_machine(__build_all_zonelists, NULL, NULL); |
6811378e7 [PATCH] wait_tabl... |
3104 3105 |
/* cpuset refresh routine should be here */ } |
bd1e22b8e [PATCH] initialis... |
3106 |
vm_total_pages = nr_free_pagecache_pages(); |
9ef9acb05 Do not group page... |
3107 3108 3109 3110 3111 3112 3113 |
/* * Disable grouping by mobility if the number of pages in the * system is too low to allow the mechanism to work. It would be * more accurate, but expensive to check per-zone. This check is * made on memory-hotadd so a system can start with mobility * disabled and enable it later */ |
d9c234005 Do not depend on ... |
3114 |
if (vm_total_pages < (pageblock_nr_pages * MIGRATE_TYPES)) |
9ef9acb05 Do not group page... |
3115 3116 3117 3118 3119 3120 3121 |
page_group_by_mobility_disabled = 1; else page_group_by_mobility_disabled = 0; printk("Built %i zonelists in %s order, mobility grouping %s. " "Total pages: %ld ", |
62bc62a87 page allocator: u... |
3122 |
nr_online_nodes, |
f0c0b2b80 change zonelist o... |
3123 |
zonelist_order_name[current_zonelist_order], |
9ef9acb05 Do not group page... |
3124 |
page_group_by_mobility_disabled ? "off" : "on", |
f0c0b2b80 change zonelist o... |
3125 3126 3127 3128 3129 |
vm_total_pages); #ifdef CONFIG_NUMA printk("Policy zone: %s ", zone_names[policy_zone]); #endif |
1da177e4c Linux-2.6.12-rc2 |
3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 |
} /* * Helper functions to size the waitqueue hash table. * Essentially these want to choose hash table sizes sufficiently * large so that collisions trying to wait on pages are rare. * But in fact, the number of active page waitqueues on typical * systems is ridiculously low, less than 200. So this is even * conservative, even though it seems large. * * The constant PAGES_PER_WAITQUEUE specifies the ratio of pages to * waitqueues, i.e. the size of the waitq table given the number of pages. */ #define PAGES_PER_WAITQUEUE 256 |
cca448fe9 [PATCH] wait_tabl... |
3144 |
#ifndef CONFIG_MEMORY_HOTPLUG |
02b694dea [PATCH] wait_tabl... |
3145 |
static inline unsigned long wait_table_hash_nr_entries(unsigned long pages) |
1da177e4c Linux-2.6.12-rc2 |
3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 |
{ unsigned long size = 1; pages /= PAGES_PER_WAITQUEUE; while (size < pages) size <<= 1; /* * Once we have dozens or even hundreds of threads sleeping * on IO we've got bigger problems than wait queue collision. * Limit the size of the wait table to a reasonable size. */ size = min(size, 4096UL); return max(size, 4UL); } |
cca448fe9 [PATCH] wait_tabl... |
3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 |
#else /* * A zone's size might be changed by hot-add, so it is not possible to determine * a suitable size for its wait_table. So we use the maximum size now. * * The max wait table size = 4096 x sizeof(wait_queue_head_t). ie: * * i386 (preemption config) : 4096 x 16 = 64Kbyte. * ia64, x86-64 (no preemption): 4096 x 20 = 80Kbyte. * ia64, x86-64 (preemption) : 4096 x 24 = 96Kbyte. * * The maximum entries are prepared when a zone's memory is (512K + 256) pages * or more by the traditional way. (See above). It equals: * * i386, x86-64, powerpc(4K page size) : = ( 2G + 1M)byte. * ia64(16K page size) : = ( 8G + 4M)byte. * powerpc (64K page size) : = (32G +16M)byte. */ static inline unsigned long wait_table_hash_nr_entries(unsigned long pages) { return 4096UL; } #endif |
1da177e4c Linux-2.6.12-rc2 |
3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 |
/* * This is an integer logarithm so that shifts can be used later * to extract the more random high bits from the multiplicative * hash function before the remainder is taken. */ static inline unsigned long wait_table_bits(unsigned long size) { return ffz(~size); } #define LONG_ALIGN(x) (((x)+(sizeof(long))-1)&~((sizeof(long))-1)) |
56fd56b86 Bias the location... |
3198 |
/* |
6d3163ce8 mm: check if any ... |
3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 |
* Check if a pageblock contains reserved pages */ static int pageblock_is_reserved(unsigned long start_pfn, unsigned long end_pfn) { unsigned long pfn; for (pfn = start_pfn; pfn < end_pfn; pfn++) { if (!pfn_valid_within(pfn) || PageReserved(pfn_to_page(pfn))) return 1; } return 0; } /* |
d9c234005 Do not depend on ... |
3213 |
* Mark a number of pageblocks as MIGRATE_RESERVE. The number |
418589663 page allocator: u... |
3214 3215 |
* of blocks reserved is based on min_wmark_pages(zone). The memory within * the reserve will tend to store contiguous free pages. Setting min_free_kbytes |
56fd56b86 Bias the location... |
3216 3217 3218 3219 3220 |
* higher will lead to a bigger reserve which will get freed as contiguous * blocks as reclaim kicks in */ static void setup_zone_migrate_reserve(struct zone *zone) { |
6d3163ce8 mm: check if any ... |
3221 |
unsigned long start_pfn, pfn, end_pfn, block_end_pfn; |
56fd56b86 Bias the location... |
3222 |
struct page *page; |
78986a678 page-allocator: l... |
3223 3224 |
unsigned long block_migratetype; int reserve; |
56fd56b86 Bias the location... |
3225 3226 3227 3228 |
/* Get the start pfn, end pfn and the number of blocks to reserve */ start_pfn = zone->zone_start_pfn; end_pfn = start_pfn + zone->spanned_pages; |
418589663 page allocator: u... |
3229 |
reserve = roundup(min_wmark_pages(zone), pageblock_nr_pages) >> |
d9c234005 Do not depend on ... |
3230 |
pageblock_order; |
56fd56b86 Bias the location... |
3231 |
|
78986a678 page-allocator: l... |
3232 3233 3234 3235 3236 3237 3238 3239 |
/* * Reserve blocks are generally in place to help high-order atomic * allocations that are short-lived. A min_free_kbytes value that * would result in more than 2 reserve blocks for atomic allocations * is assumed to be in place to help anti-fragmentation for the * future allocation of hugepages at runtime. */ reserve = min(2, reserve); |
d9c234005 Do not depend on ... |
3240 |
for (pfn = start_pfn; pfn < end_pfn; pfn += pageblock_nr_pages) { |
56fd56b86 Bias the location... |
3241 3242 3243 |
if (!pfn_valid(pfn)) continue; page = pfn_to_page(pfn); |
344c790e3 mm: make setup_zo... |
3244 3245 3246 |
/* Watch out for overlapping nodes */ if (page_to_nid(page) != zone_to_nid(zone)) continue; |
56fd56b86 Bias the location... |
3247 |
/* Blocks with reserved pages will never free, skip them. */ |
6d3163ce8 mm: check if any ... |
3248 3249 |
block_end_pfn = min(pfn + pageblock_nr_pages, end_pfn); if (pageblock_is_reserved(pfn, block_end_pfn)) |
56fd56b86 Bias the location... |
3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 |
continue; block_migratetype = get_pageblock_migratetype(page); /* If this block is reserved, account for it */ if (reserve > 0 && block_migratetype == MIGRATE_RESERVE) { reserve--; continue; } /* Suitable for reserving if this block is movable */ if (reserve > 0 && block_migratetype == MIGRATE_MOVABLE) { set_pageblock_migratetype(page, MIGRATE_RESERVE); move_freepages_block(zone, page, MIGRATE_RESERVE); reserve--; continue; } /* * If the reserve is met and this is a previous reserved block, * take it back */ if (block_migratetype == MIGRATE_RESERVE) { set_pageblock_migratetype(page, MIGRATE_MOVABLE); move_freepages_block(zone, page, MIGRATE_MOVABLE); } } } |
ac0e5b7a6 remove PAGE_GROUP... |
3278 |
|
1da177e4c Linux-2.6.12-rc2 |
3279 3280 3281 3282 3283 |
/* * Initially all pages are reserved - free ones are freed * up by free_all_bootmem() once the early boot process is * done. Non-atomic initialization, single-pass. */ |
c09b42404 [PATCH] x86_64: a... |
3284 |
void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone, |
a2f3aa025 [PATCH] Fix spars... |
3285 |
unsigned long start_pfn, enum memmap_context context) |
1da177e4c Linux-2.6.12-rc2 |
3286 |
{ |
1da177e4c Linux-2.6.12-rc2 |
3287 |
struct page *page; |
29751f699 [PATCH] sparsemem... |
3288 3289 |
unsigned long end_pfn = start_pfn + size; unsigned long pfn; |
86051ca5e mm: fix usemap in... |
3290 |
struct zone *z; |
1da177e4c Linux-2.6.12-rc2 |
3291 |
|
22b31eec6 badpage: vm_norma... |
3292 3293 |
if (highest_memmap_pfn < end_pfn - 1) highest_memmap_pfn = end_pfn - 1; |
86051ca5e mm: fix usemap in... |
3294 |
z = &NODE_DATA(nid)->node_zones[zone]; |
cbe8dd4af [PATCH] memmap_in... |
3295 |
for (pfn = start_pfn; pfn < end_pfn; pfn++) { |
a2f3aa025 [PATCH] Fix spars... |
3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 |
/* * There can be holes in boot-time mem_map[]s * handed to this function. They do not * exist on hotplugged memory. */ if (context == MEMMAP_EARLY) { if (!early_pfn_valid(pfn)) continue; if (!early_pfn_in_nid(pfn, nid)) continue; } |
d41dee369 [PATCH] sparsemem... |
3307 3308 |
page = pfn_to_page(pfn); set_page_links(page, zone, nid, pfn); |
708614e61 mm: verify the pa... |
3309 |
mminit_verify_page_links(page, zone, nid, pfn); |
7835e98b2 [PATCH] remove se... |
3310 |
init_page_count(page); |
1da177e4c Linux-2.6.12-rc2 |
3311 3312 |
reset_page_mapcount(page); SetPageReserved(page); |
b2a0ac887 Split the free li... |
3313 3314 3315 3316 3317 |
/* * Mark the block movable so that blocks are reserved for * movable at startup. This will force kernel allocations * to reserve their blocks rather than leaking throughout * the address space during boot when many long-lived |
56fd56b86 Bias the location... |
3318 3319 3320 |
* kernel allocations are made. Later some blocks near * the start are marked MIGRATE_RESERVE by * setup_zone_migrate_reserve() |
86051ca5e mm: fix usemap in... |
3321 3322 3323 3324 3325 |
* * bitmap is created for zone's valid pfn range. but memmap * can be created for invalid pages (for alignment) * check here not to call set_pageblock_migratetype() against * pfn out of zone. |
b2a0ac887 Split the free li... |
3326 |
*/ |
86051ca5e mm: fix usemap in... |
3327 3328 3329 |
if ((z->zone_start_pfn <= pfn) && (pfn < z->zone_start_pfn + z->spanned_pages) && !(pfn & (pageblock_nr_pages - 1))) |
56fd56b86 Bias the location... |
3330 |
set_pageblock_migratetype(page, MIGRATE_MOVABLE); |
b2a0ac887 Split the free li... |
3331 |
|
1da177e4c Linux-2.6.12-rc2 |
3332 3333 3334 3335 |
INIT_LIST_HEAD(&page->lru); #ifdef WANT_PAGE_VIRTUAL /* The shift won't overflow because ZONE_NORMAL is below 4G. */ if (!is_highmem_idx(zone)) |
3212c6be2 [PATCH] fix WANT_... |
3336 |
set_page_address(page, __va(pfn << PAGE_SHIFT)); |
1da177e4c Linux-2.6.12-rc2 |
3337 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
3338 3339 |
} } |
1e548deb5 page allocator: r... |
3340 |
static void __meminit zone_init_free_lists(struct zone *zone) |
1da177e4c Linux-2.6.12-rc2 |
3341 |
{ |
b2a0ac887 Split the free li... |
3342 3343 3344 |
int order, t; for_each_migratetype_order(order, t) { INIT_LIST_HEAD(&zone->free_area[order].free_list[t]); |
1da177e4c Linux-2.6.12-rc2 |
3345 3346 3347 3348 3349 3350 |
zone->free_area[order].nr_free = 0; } } #ifndef __HAVE_ARCH_MEMMAP_INIT #define memmap_init(size, nid, zone, start_pfn) \ |
a2f3aa025 [PATCH] Fix spars... |
3351 |
memmap_init_zone((size), (nid), (zone), (start_pfn), MEMMAP_EARLY) |
1da177e4c Linux-2.6.12-rc2 |
3352 |
#endif |
1d6f4e60e mm: fix section m... |
3353 |
static int zone_batchsize(struct zone *zone) |
e7c8d5c99 [PATCH] node loca... |
3354 |
{ |
3a6be87fd nommu: clamp zone... |
3355 |
#ifdef CONFIG_MMU |
e7c8d5c99 [PATCH] node loca... |
3356 3357 3358 3359 |
int batch; /* * The per-cpu-pages pools are set to around 1000th of the |
ba56e91c9 [PATCH] mm: page_... |
3360 |
* size of the zone. But no more than 1/2 of a meg. |
e7c8d5c99 [PATCH] node loca... |
3361 3362 3363 3364 |
* * OK, so we don't know how big the cache is. So guess. */ batch = zone->present_pages / 1024; |
ba56e91c9 [PATCH] mm: page_... |
3365 3366 |
if (batch * PAGE_SIZE > 512 * 1024) batch = (512 * 1024) / PAGE_SIZE; |
e7c8d5c99 [PATCH] node loca... |
3367 3368 3369 3370 3371 |
batch /= 4; /* We effectively *= 4 below */ if (batch < 1) batch = 1; /* |
0ceaacc97 [PATCH] Fix up pe... |
3372 3373 3374 |
* Clamp the batch to a 2^n - 1 value. Having a power * of 2 value was found to be more likely to have * suboptimal cache aliasing properties in some cases. |
e7c8d5c99 [PATCH] node loca... |
3375 |
* |
0ceaacc97 [PATCH] Fix up pe... |
3376 3377 3378 3379 |
* For example if 2 tasks are alternately allocating * batches of pages, one task can end up with a lot * of pages of one half of the possible page colors * and the other with pages of the other colors. |
e7c8d5c99 [PATCH] node loca... |
3380 |
*/ |
9155203a5 mm: use roundown_... |
3381 |
batch = rounddown_pow_of_two(batch + batch/2) - 1; |
ba56e91c9 [PATCH] mm: page_... |
3382 |
|
e7c8d5c99 [PATCH] node loca... |
3383 |
return batch; |
3a6be87fd nommu: clamp zone... |
3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 |
#else /* The deferral and batching of frees should be suppressed under NOMMU * conditions. * * The problem is that NOMMU needs to be able to allocate large chunks * of contiguous memory as there's no hardware page translation to * assemble apparent contiguous memory from discontiguous pages. * * Queueing large contiguous runs of pages for batching, however, * causes the pages to actually be freed in smaller chunks. As there * can be a significant delay between the individual batches being * recycled, this leads to the once large chunks of space being * fragmented and becoming unavailable for high-order allocations. */ return 0; #endif |
e7c8d5c99 [PATCH] node loca... |
3401 |
} |
b69a7288e mm/page_alloc.c: ... |
3402 |
static void setup_pageset(struct per_cpu_pageset *p, unsigned long batch) |
2caaad41e [PATCH] Reduce si... |
3403 3404 |
{ struct per_cpu_pages *pcp; |
5f8dcc212 page-allocator: s... |
3405 |
int migratetype; |
2caaad41e [PATCH] Reduce si... |
3406 |
|
1c6fe9465 [PATCH] NUMA: bro... |
3407 |
memset(p, 0, sizeof(*p)); |
3dfa5721f Page allocator: g... |
3408 |
pcp = &p->pcp; |
2caaad41e [PATCH] Reduce si... |
3409 |
pcp->count = 0; |
2caaad41e [PATCH] Reduce si... |
3410 3411 |
pcp->high = 6 * batch; pcp->batch = max(1UL, 1 * batch); |
5f8dcc212 page-allocator: s... |
3412 3413 |
for (migratetype = 0; migratetype < MIGRATE_PCPTYPES; migratetype++) INIT_LIST_HEAD(&pcp->lists[migratetype]); |
2caaad41e [PATCH] Reduce si... |
3414 |
} |
8ad4b1fb8 [PATCH] Make high... |
3415 3416 3417 3418 3419 3420 3421 3422 3423 |
/* * setup_pagelist_highmark() sets the high water mark for hot per_cpu_pagelist * to the value high for the pageset p. */ static void setup_pagelist_highmark(struct per_cpu_pageset *p, unsigned long high) { struct per_cpu_pages *pcp; |
3dfa5721f Page allocator: g... |
3424 |
pcp = &p->pcp; |
8ad4b1fb8 [PATCH] Make high... |
3425 3426 3427 3428 3429 |
pcp->high = high; pcp->batch = max(1UL, high/4); if ((high/4) > (PAGE_SHIFT * 8)) pcp->batch = PAGE_SHIFT * 8; } |
58c2ee400 mm: Fix section m... |
3430 |
static void setup_zone_pageset(struct zone *zone) |
319774e25 mem-hotplug: sepa... |
3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 |
{ int cpu; zone->pageset = alloc_percpu(struct per_cpu_pageset); for_each_possible_cpu(cpu) { struct per_cpu_pageset *pcp = per_cpu_ptr(zone->pageset, cpu); setup_pageset(pcp, zone_batchsize(zone)); if (percpu_pagelist_fraction) setup_pagelist_highmark(pcp, (zone->present_pages / percpu_pagelist_fraction)); } } |
2caaad41e [PATCH] Reduce si... |
3447 |
/* |
99dcc3e5a this_cpu: Page al... |
3448 3449 |
* Allocate per cpu pagesets and initialize them. * Before this call only boot pagesets were available. |
e7c8d5c99 [PATCH] node loca... |
3450 |
*/ |
99dcc3e5a this_cpu: Page al... |
3451 |
void __init setup_per_cpu_pageset(void) |
e7c8d5c99 [PATCH] node loca... |
3452 |
{ |
99dcc3e5a this_cpu: Page al... |
3453 |
struct zone *zone; |
e7c8d5c99 [PATCH] node loca... |
3454 |
|
319774e25 mem-hotplug: sepa... |
3455 3456 |
for_each_populated_zone(zone) setup_zone_pageset(zone); |
e7c8d5c99 [PATCH] node loca... |
3457 |
} |
577a32f62 mm: fix section m... |
3458 |
static noinline __init_refok |
cca448fe9 [PATCH] wait_tabl... |
3459 |
int zone_wait_table_init(struct zone *zone, unsigned long zone_size_pages) |
ed8ece2ec [PATCH] memory ho... |
3460 3461 3462 |
{ int i; struct pglist_data *pgdat = zone->zone_pgdat; |
cca448fe9 [PATCH] wait_tabl... |
3463 |
size_t alloc_size; |
ed8ece2ec [PATCH] memory ho... |
3464 3465 3466 3467 3468 |
/* * The per-page waitqueue mechanism uses hashed waitqueues * per zone. */ |
02b694dea [PATCH] wait_tabl... |
3469 3470 3471 3472 |
zone->wait_table_hash_nr_entries = wait_table_hash_nr_entries(zone_size_pages); zone->wait_table_bits = wait_table_bits(zone->wait_table_hash_nr_entries); |
cca448fe9 [PATCH] wait_tabl... |
3473 3474 |
alloc_size = zone->wait_table_hash_nr_entries * sizeof(wait_queue_head_t); |
cd94b9dbf memory hotplug: f... |
3475 |
if (!slab_is_available()) { |
cca448fe9 [PATCH] wait_tabl... |
3476 |
zone->wait_table = (wait_queue_head_t *) |
8f389a99b mm: use alloc_boo... |
3477 |
alloc_bootmem_node_nopanic(pgdat, alloc_size); |
cca448fe9 [PATCH] wait_tabl... |
3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 |
} else { /* * This case means that a zone whose size was 0 gets new memory * via memory hot-add. * But it may be the case that a new node was hot-added. In * this case vmalloc() will not be able to use this new node's * memory - this wait_table must be initialized to use this new * node itself as well. * To use this new node's memory, further consideration will be * necessary. */ |
8691f3a72 mm: no need to ca... |
3489 |
zone->wait_table = vmalloc(alloc_size); |
cca448fe9 [PATCH] wait_tabl... |
3490 3491 3492 |
} if (!zone->wait_table) return -ENOMEM; |
ed8ece2ec [PATCH] memory ho... |
3493 |
|
02b694dea [PATCH] wait_tabl... |
3494 |
for(i = 0; i < zone->wait_table_hash_nr_entries; ++i) |
ed8ece2ec [PATCH] memory ho... |
3495 |
init_waitqueue_head(zone->wait_table + i); |
cca448fe9 [PATCH] wait_tabl... |
3496 3497 |
return 0; |
ed8ece2ec [PATCH] memory ho... |
3498 |
} |
112067f09 memory hotplug: u... |
3499 3500 3501 3502 3503 |
static int __zone_pcp_update(void *data) { struct zone *zone = data; int cpu; unsigned long batch = zone_batchsize(zone), flags; |
2d30a1f63 mm: do not iterat... |
3504 |
for_each_possible_cpu(cpu) { |
112067f09 memory hotplug: u... |
3505 3506 |
struct per_cpu_pageset *pset; struct per_cpu_pages *pcp; |
99dcc3e5a this_cpu: Page al... |
3507 |
pset = per_cpu_ptr(zone->pageset, cpu); |
112067f09 memory hotplug: u... |
3508 3509 3510 |
pcp = &pset->pcp; local_irq_save(flags); |
5f8dcc212 page-allocator: s... |
3511 |
free_pcppages_bulk(zone, pcp->count, pcp); |
112067f09 memory hotplug: u... |
3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 |
setup_pageset(pset, batch); local_irq_restore(flags); } return 0; } void zone_pcp_update(struct zone *zone) { stop_machine(__zone_pcp_update, zone, NULL); } |
c09b42404 [PATCH] x86_64: a... |
3522 |
static __meminit void zone_pcp_init(struct zone *zone) |
ed8ece2ec [PATCH] memory ho... |
3523 |
{ |
99dcc3e5a this_cpu: Page al... |
3524 3525 3526 3527 3528 3529 |
/* * per cpu subsystem is not up at this point. The following code * relies on the ability of the linker to provide the * offset of a (static) per cpu variable into the per cpu area. */ zone->pageset = &boot_pageset; |
ed8ece2ec [PATCH] memory ho... |
3530 |
|
f5335c0f1 [PATCH] quieten z... |
3531 |
if (zone->present_pages) |
99dcc3e5a this_cpu: Page al... |
3532 3533 3534 3535 |
printk(KERN_DEBUG " %s zone: %lu pages, LIFO batch:%u ", zone->name, zone->present_pages, zone_batchsize(zone)); |
ed8ece2ec [PATCH] memory ho... |
3536 |
} |
718127cc3 [PATCH] wait_tabl... |
3537 3538 |
__meminit int init_currently_empty_zone(struct zone *zone, unsigned long zone_start_pfn, |
a2f3aa025 [PATCH] Fix spars... |
3539 3540 |
unsigned long size, enum memmap_context context) |
ed8ece2ec [PATCH] memory ho... |
3541 3542 |
{ struct pglist_data *pgdat = zone->zone_pgdat; |
cca448fe9 [PATCH] wait_tabl... |
3543 3544 3545 3546 |
int ret; ret = zone_wait_table_init(zone, size); if (ret) return ret; |
ed8ece2ec [PATCH] memory ho... |
3547 |
pgdat->nr_zones = zone_idx(zone) + 1; |
ed8ece2ec [PATCH] memory ho... |
3548 |
zone->zone_start_pfn = zone_start_pfn; |
708614e61 mm: verify the pa... |
3549 3550 3551 3552 3553 3554 |
mminit_dprintk(MMINIT_TRACE, "memmap_init", "Initialising map node %d zone %lu pfns %lu -> %lu ", pgdat->node_id, (unsigned long)zone_idx(zone), zone_start_pfn, (zone_start_pfn + size)); |
1e548deb5 page allocator: r... |
3555 |
zone_init_free_lists(zone); |
718127cc3 [PATCH] wait_tabl... |
3556 3557 |
return 0; |
ed8ece2ec [PATCH] memory ho... |
3558 |
} |
c713216de [PATCH] Introduce... |
3559 3560 3561 3562 3563 |
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP /* * Basic iterator support. Return the first range of PFNs for a node * Note: nid == MAX_NUMNODES returns first region regardless of node */ |
a3142c8e1 Fix section misma... |
3564 |
static int __meminit first_active_region_index_in_nid(int nid) |
c713216de [PATCH] Introduce... |
3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 |
{ int i; for (i = 0; i < nr_nodemap_entries; i++) if (nid == MAX_NUMNODES || early_node_map[i].nid == nid) return i; return -1; } /* * Basic iterator support. Return the next active range of PFNs for a node |
183ff22bb spelling fixes: mm/ |
3577 |
* Note: nid == MAX_NUMNODES returns next region regardless of node |
c713216de [PATCH] Introduce... |
3578 |
*/ |
a3142c8e1 Fix section misma... |
3579 |
static int __meminit next_active_region_index_in_nid(int index, int nid) |
c713216de [PATCH] Introduce... |
3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 |
{ for (index = index + 1; index < nr_nodemap_entries; index++) if (nid == MAX_NUMNODES || early_node_map[index].nid == nid) return index; return -1; } #ifndef CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID /* * Required by SPARSEMEM. Given a PFN, return what node the PFN is on. * Architectures may implement their own version but if add_active_range() * was used and there are no special requirements, this is a convenient * alternative */ |
f2dbcfa73 mm: clean up for ... |
3595 |
int __meminit __early_pfn_to_nid(unsigned long pfn) |
c713216de [PATCH] Introduce... |
3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 |
{ int i; for (i = 0; i < nr_nodemap_entries; i++) { unsigned long start_pfn = early_node_map[i].start_pfn; unsigned long end_pfn = early_node_map[i].end_pfn; if (start_pfn <= pfn && pfn < end_pfn) return early_node_map[i].nid; } |
cc2559bcc mm: fix memmap in... |
3606 3607 |
/* This is a memory hole */ return -1; |
c713216de [PATCH] Introduce... |
3608 3609 |
} #endif /* CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID */ |
f2dbcfa73 mm: clean up for ... |
3610 3611 |
int __meminit early_pfn_to_nid(unsigned long pfn) { |
cc2559bcc mm: fix memmap in... |
3612 3613 3614 3615 3616 3617 3618 |
int nid; nid = __early_pfn_to_nid(pfn); if (nid >= 0) return nid; /* just returns 0 */ return 0; |
f2dbcfa73 mm: clean up for ... |
3619 |
} |
cc2559bcc mm: fix memmap in... |
3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 |
#ifdef CONFIG_NODES_SPAN_OTHER_NODES bool __meminit early_pfn_in_nid(unsigned long pfn, int node) { int nid; nid = __early_pfn_to_nid(pfn); if (nid >= 0 && nid != node) return false; return true; } #endif |
f2dbcfa73 mm: clean up for ... |
3631 |
|
c713216de [PATCH] Introduce... |
3632 3633 3634 3635 3636 3637 3638 |
/* Basic iterator support to walk early_node_map[] */ #define for_each_active_range_index_in_nid(i, nid) \ for (i = first_active_region_index_in_nid(nid); i != -1; \ i = next_active_region_index_in_nid(i, nid)) /** * free_bootmem_with_active_regions - Call free_bootmem_node for each active range |
88ca3b94e [PATCH] page_allo... |
3639 3640 |
* @nid: The node to free memory on. If MAX_NUMNODES, all nodes are freed. * @max_low_pfn: The highest PFN that will be passed to free_bootmem_node |
c713216de [PATCH] Introduce... |
3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 |
* * If an architecture guarantees that all ranges registered with * add_active_ranges() contain no holes and may be freed, this * this function may be used instead of calling free_bootmem() manually. */ void __init free_bootmem_with_active_regions(int nid, unsigned long max_low_pfn) { int i; for_each_active_range_index_in_nid(i, nid) { unsigned long size_pages = 0; unsigned long end_pfn = early_node_map[i].end_pfn; if (early_node_map[i].start_pfn >= max_low_pfn) continue; if (end_pfn > max_low_pfn) end_pfn = max_low_pfn; size_pages = end_pfn - early_node_map[i].start_pfn; free_bootmem_node(NODE_DATA(early_node_map[i].nid), PFN_PHYS(early_node_map[i].start_pfn), size_pages << PAGE_SHIFT); } } |
edbe7d23b memblock: Add fin... |
3667 |
#ifdef CONFIG_HAVE_MEMBLOCK |
cc2898943 mm: Move early_no... |
3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 |
/* * Basic iterator support. Return the last range of PFNs for a node * Note: nid == MAX_NUMNODES returns last region regardless of node */ static int __meminit last_active_region_index_in_nid(int nid) { int i; for (i = nr_nodemap_entries - 1; i >= 0; i--) if (nid == MAX_NUMNODES || early_node_map[i].nid == nid) return i; return -1; } /* * Basic iterator support. Return the previous active range of PFNs for a node * Note: nid == MAX_NUMNODES returns next region regardless of node */ static int __meminit previous_active_region_index_in_nid(int index, int nid) { for (index = index - 1; index >= 0; index--) if (nid == MAX_NUMNODES || early_node_map[index].nid == nid) return index; return -1; } #define for_each_active_range_index_in_nid_reverse(i, nid) \ for (i = last_active_region_index_in_nid(nid); i != -1; \ i = previous_active_region_index_in_nid(i, nid)) |
edbe7d23b memblock: Add fin... |
3699 3700 3701 3702 3703 3704 |
u64 __init find_memory_core_early(int nid, u64 size, u64 align, u64 goal, u64 limit) { int i; /* Need to go over early_node_map to find out good range for node */ |
1a4a678b1 memblock: Make fi... |
3705 |
for_each_active_range_index_in_nid_reverse(i, nid) { |
edbe7d23b memblock: Add fin... |
3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 |
u64 addr; u64 ei_start, ei_last; u64 final_start, final_end; ei_last = early_node_map[i].end_pfn; ei_last <<= PAGE_SHIFT; ei_start = early_node_map[i].start_pfn; ei_start <<= PAGE_SHIFT; final_start = max(ei_start, goal); final_end = min(ei_last, limit); if (final_start >= final_end) continue; addr = memblock_find_in_range(final_start, final_end, size, align); if (addr == MEMBLOCK_ERROR) continue; return addr; } return MEMBLOCK_ERROR; } #endif |
08677214e x86: Make 64 bit ... |
3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 |
int __init add_from_early_node_map(struct range *range, int az, int nr_range, int nid) { int i; u64 start, end; /* need to go over early_node_map to find out good range for node */ for_each_active_range_index_in_nid(i, nid) { start = early_node_map[i].start_pfn; end = early_node_map[i].end_pfn; nr_range = add_range(range, az, nr_range, start, end); } return nr_range; } |
b5bc6c0e5 x86, mm: use add_... |
3746 3747 3748 |
void __init work_with_active_regions(int nid, work_fn_t work_fn, void *data) { int i; |
d52d53b8a RFC x86: try to r... |
3749 |
int ret; |
b5bc6c0e5 x86, mm: use add_... |
3750 |
|
d52d53b8a RFC x86: try to r... |
3751 3752 3753 3754 3755 3756 |
for_each_active_range_index_in_nid(i, nid) { ret = work_fn(early_node_map[i].start_pfn, early_node_map[i].end_pfn, data); if (ret) break; } |
b5bc6c0e5 x86, mm: use add_... |
3757 |
} |
c713216de [PATCH] Introduce... |
3758 3759 |
/** * sparse_memory_present_with_active_regions - Call memory_present for each active range |
88ca3b94e [PATCH] page_allo... |
3760 |
* @nid: The node to call memory_present for. If MAX_NUMNODES, all nodes will be used. |
c713216de [PATCH] Introduce... |
3761 3762 3763 |
* * If an architecture guarantees that all ranges registered with * add_active_ranges() contain no holes and may be freed, this |
88ca3b94e [PATCH] page_allo... |
3764 |
* function may be used instead of calling memory_present() manually. |
c713216de [PATCH] Introduce... |
3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 |
*/ void __init sparse_memory_present_with_active_regions(int nid) { int i; for_each_active_range_index_in_nid(i, nid) memory_present(early_node_map[i].nid, early_node_map[i].start_pfn, early_node_map[i].end_pfn); } /** * get_pfn_range_for_nid - Return the start and end page frames for a node |
88ca3b94e [PATCH] page_allo... |
3778 3779 3780 |
* @nid: The nid to return the range for. If MAX_NUMNODES, the min and max PFN are returned. * @start_pfn: Passed by reference. On return, it will have the node start_pfn. * @end_pfn: Passed by reference. On return, it will have the node end_pfn. |
c713216de [PATCH] Introduce... |
3781 3782 3783 3784 |
* * It returns the start and end page frame of a node based on information * provided by an arch calling add_active_range(). If called for a node * with no available memory, a warning is printed and the start and end |
88ca3b94e [PATCH] page_allo... |
3785 |
* PFNs will be 0. |
c713216de [PATCH] Introduce... |
3786 |
*/ |
a3142c8e1 Fix section misma... |
3787 |
void __meminit get_pfn_range_for_nid(unsigned int nid, |
c713216de [PATCH] Introduce... |
3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 |
unsigned long *start_pfn, unsigned long *end_pfn) { int i; *start_pfn = -1UL; *end_pfn = 0; for_each_active_range_index_in_nid(i, nid) { *start_pfn = min(*start_pfn, early_node_map[i].start_pfn); *end_pfn = max(*end_pfn, early_node_map[i].end_pfn); } |
633c0666b Memoryless nodes:... |
3798 |
if (*start_pfn == -1UL) |
c713216de [PATCH] Introduce... |
3799 |
*start_pfn = 0; |
c713216de [PATCH] Introduce... |
3800 3801 3802 |
} /* |
2a1e274ac Create the ZONE_M... |
3803 3804 3805 3806 |
* This finds a zone that can be used for ZONE_MOVABLE pages. The * assumption is made that zones within a node are ordered in monotonic * increasing memory addresses so that the "highest" populated zone is used */ |
b69a7288e mm/page_alloc.c: ... |
3807 |
static void __init find_usable_zone_for_movable(void) |
2a1e274ac Create the ZONE_M... |
3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 |
{ int zone_index; for (zone_index = MAX_NR_ZONES - 1; zone_index >= 0; zone_index--) { if (zone_index == ZONE_MOVABLE) continue; if (arch_zone_highest_possible_pfn[zone_index] > arch_zone_lowest_possible_pfn[zone_index]) break; } VM_BUG_ON(zone_index == -1); movable_zone = zone_index; } /* * The zone ranges provided by the architecture do not include ZONE_MOVABLE |
25985edce Fix common misspe... |
3825 |
* because it is sized independent of architecture. Unlike the other zones, |
2a1e274ac Create the ZONE_M... |
3826 3827 3828 3829 3830 3831 3832 |
* the starting point for ZONE_MOVABLE is not fixed. It may be different * in each node depending on the size of each node and how evenly kernelcore * is distributed. This helper function adjusts the zone ranges * provided by the architecture for a given node by using the end of the * highest usable zone for ZONE_MOVABLE. This preserves the assumption that * zones within a node are in order of monotonic increases memory addresses */ |
b69a7288e mm/page_alloc.c: ... |
3833 |
static void __meminit adjust_zone_range_for_zone_movable(int nid, |
2a1e274ac Create the ZONE_M... |
3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 |
unsigned long zone_type, unsigned long node_start_pfn, unsigned long node_end_pfn, unsigned long *zone_start_pfn, unsigned long *zone_end_pfn) { /* Only adjust if ZONE_MOVABLE is on this node */ if (zone_movable_pfn[nid]) { /* Size ZONE_MOVABLE */ if (zone_type == ZONE_MOVABLE) { *zone_start_pfn = zone_movable_pfn[nid]; *zone_end_pfn = min(node_end_pfn, arch_zone_highest_possible_pfn[movable_zone]); /* Adjust for ZONE_MOVABLE starting within this range */ } else if (*zone_start_pfn < zone_movable_pfn[nid] && *zone_end_pfn > zone_movable_pfn[nid]) { *zone_end_pfn = zone_movable_pfn[nid]; /* Check if this whole range is within ZONE_MOVABLE */ } else if (*zone_start_pfn >= zone_movable_pfn[nid]) *zone_start_pfn = *zone_end_pfn; } } /* |
c713216de [PATCH] Introduce... |
3860 3861 3862 |
* Return the number of pages a zone spans in a node, including holes * present_pages = zone_spanned_pages_in_node() - zone_absent_pages_in_node() */ |
6ea6e6887 mm: more __memini... |
3863 |
static unsigned long __meminit zone_spanned_pages_in_node(int nid, |
c713216de [PATCH] Introduce... |
3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 |
unsigned long zone_type, unsigned long *ignored) { unsigned long node_start_pfn, node_end_pfn; unsigned long zone_start_pfn, zone_end_pfn; /* Get the start and end of the node and zone */ get_pfn_range_for_nid(nid, &node_start_pfn, &node_end_pfn); zone_start_pfn = arch_zone_lowest_possible_pfn[zone_type]; zone_end_pfn = arch_zone_highest_possible_pfn[zone_type]; |
2a1e274ac Create the ZONE_M... |
3874 3875 3876 |
adjust_zone_range_for_zone_movable(nid, zone_type, node_start_pfn, node_end_pfn, &zone_start_pfn, &zone_end_pfn); |
c713216de [PATCH] Introduce... |
3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 |
/* Check that this node has pages within the zone's required range */ if (zone_end_pfn < node_start_pfn || zone_start_pfn > node_end_pfn) return 0; /* Move the zone boundaries inside the node if necessary */ zone_end_pfn = min(zone_end_pfn, node_end_pfn); zone_start_pfn = max(zone_start_pfn, node_start_pfn); /* Return the spanned pages */ return zone_end_pfn - zone_start_pfn; } /* * Return the number of holes in a range on a node. If nid is MAX_NUMNODES, |
88ca3b94e [PATCH] page_allo... |
3892 |
* then all holes in the requested range will be accounted for. |
c713216de [PATCH] Introduce... |
3893 |
*/ |
329962503 x86: Fix checking... |
3894 |
unsigned long __meminit __absent_pages_in_range(int nid, |
c713216de [PATCH] Introduce... |
3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 |
unsigned long range_start_pfn, unsigned long range_end_pfn) { int i = 0; unsigned long prev_end_pfn = 0, hole_pages = 0; unsigned long start_pfn; /* Find the end_pfn of the first active range of pfns in the node */ i = first_active_region_index_in_nid(nid); if (i == -1) return 0; |
b5445f956 Allow nodes to ex... |
3906 |
prev_end_pfn = min(early_node_map[i].start_pfn, range_end_pfn); |
9c7cd6877 [PATCH] Account f... |
3907 3908 |
/* Account for ranges before physical memory on this node */ if (early_node_map[i].start_pfn > range_start_pfn) |
b5445f956 Allow nodes to ex... |
3909 |
hole_pages = prev_end_pfn - range_start_pfn; |
c713216de [PATCH] Introduce... |
3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 |
/* Find all holes for the zone within the node */ for (; i != -1; i = next_active_region_index_in_nid(i, nid)) { /* No need to continue if prev_end_pfn is outside the zone */ if (prev_end_pfn >= range_end_pfn) break; /* Make sure the end of the zone is not within the hole */ start_pfn = min(early_node_map[i].start_pfn, range_end_pfn); prev_end_pfn = max(prev_end_pfn, range_start_pfn); /* Update the hole size cound and move on */ if (start_pfn > range_start_pfn) { BUG_ON(prev_end_pfn > start_pfn); hole_pages += start_pfn - prev_end_pfn; } prev_end_pfn = early_node_map[i].end_pfn; } |
9c7cd6877 [PATCH] Account f... |
3929 3930 |
/* Account for ranges past physical memory on this node */ if (range_end_pfn > prev_end_pfn) |
0c6cb9746 [PATCH] Calculati... |
3931 |
hole_pages += range_end_pfn - |
9c7cd6877 [PATCH] Account f... |
3932 |
max(range_start_pfn, prev_end_pfn); |
c713216de [PATCH] Introduce... |
3933 3934 3935 3936 3937 3938 3939 3940 |
return hole_pages; } /** * absent_pages_in_range - Return number of page frames in holes within a range * @start_pfn: The start PFN to start searching for holes * @end_pfn: The end PFN to stop searching for holes * |
88ca3b94e [PATCH] page_allo... |
3941 |
* It returns the number of pages frames in memory holes within a range. |
c713216de [PATCH] Introduce... |
3942 3943 3944 3945 3946 3947 3948 3949 |
*/ unsigned long __init absent_pages_in_range(unsigned long start_pfn, unsigned long end_pfn) { return __absent_pages_in_range(MAX_NUMNODES, start_pfn, end_pfn); } /* Return the number of page frames in holes in a zone on a node */ |
6ea6e6887 mm: more __memini... |
3950 |
static unsigned long __meminit zone_absent_pages_in_node(int nid, |
c713216de [PATCH] Introduce... |
3951 3952 3953 |
unsigned long zone_type, unsigned long *ignored) { |
9c7cd6877 [PATCH] Account f... |
3954 3955 3956 3957 3958 3959 3960 3961 |
unsigned long node_start_pfn, node_end_pfn; unsigned long zone_start_pfn, zone_end_pfn; get_pfn_range_for_nid(nid, &node_start_pfn, &node_end_pfn); zone_start_pfn = max(arch_zone_lowest_possible_pfn[zone_type], node_start_pfn); zone_end_pfn = min(arch_zone_highest_possible_pfn[zone_type], node_end_pfn); |
2a1e274ac Create the ZONE_M... |
3962 3963 3964 |
adjust_zone_range_for_zone_movable(nid, zone_type, node_start_pfn, node_end_pfn, &zone_start_pfn, &zone_end_pfn); |
9c7cd6877 [PATCH] Account f... |
3965 |
return __absent_pages_in_range(nid, zone_start_pfn, zone_end_pfn); |
c713216de [PATCH] Introduce... |
3966 |
} |
0e0b864e0 [PATCH] Account f... |
3967 |
|
c713216de [PATCH] Introduce... |
3968 |
#else |
6ea6e6887 mm: more __memini... |
3969 |
static inline unsigned long __meminit zone_spanned_pages_in_node(int nid, |
c713216de [PATCH] Introduce... |
3970 3971 3972 3973 3974 |
unsigned long zone_type, unsigned long *zones_size) { return zones_size[zone_type]; } |
6ea6e6887 mm: more __memini... |
3975 |
static inline unsigned long __meminit zone_absent_pages_in_node(int nid, |
c713216de [PATCH] Introduce... |
3976 3977 3978 3979 3980 3981 3982 3983 |
unsigned long zone_type, unsigned long *zholes_size) { if (!zholes_size) return 0; return zholes_size[zone_type]; } |
0e0b864e0 [PATCH] Account f... |
3984 |
|
c713216de [PATCH] Introduce... |
3985 |
#endif |
a3142c8e1 Fix section misma... |
3986 |
static void __meminit calculate_node_totalpages(struct pglist_data *pgdat, |
c713216de [PATCH] Introduce... |
3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 |
unsigned long *zones_size, unsigned long *zholes_size) { unsigned long realtotalpages, totalpages = 0; enum zone_type i; for (i = 0; i < MAX_NR_ZONES; i++) totalpages += zone_spanned_pages_in_node(pgdat->node_id, i, zones_size); pgdat->node_spanned_pages = totalpages; realtotalpages = totalpages; for (i = 0; i < MAX_NR_ZONES; i++) realtotalpages -= zone_absent_pages_in_node(pgdat->node_id, i, zholes_size); pgdat->node_present_pages = realtotalpages; printk(KERN_DEBUG "On node %d totalpages: %lu ", pgdat->node_id, realtotalpages); } |
835c134ec Add a bitmap that... |
4007 4008 4009 |
#ifndef CONFIG_SPARSEMEM /* * Calculate the size of the zone->blockflags rounded to an unsigned long |
d9c234005 Do not depend on ... |
4010 4011 |
* Start by making sure zonesize is a multiple of pageblock_order by rounding * up. Then use 1 NR_PAGEBLOCK_BITS worth of bits per pageblock, finally |
835c134ec Add a bitmap that... |
4012 4013 4014 4015 4016 4017 |
* round what is now in bits to nearest long in bits, then return it in * bytes. */ static unsigned long __init usemap_size(unsigned long zonesize) { unsigned long usemapsize; |
d9c234005 Do not depend on ... |
4018 4019 |
usemapsize = roundup(zonesize, pageblock_nr_pages); usemapsize = usemapsize >> pageblock_order; |
835c134ec Add a bitmap that... |
4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 |
usemapsize *= NR_PAGEBLOCK_BITS; usemapsize = roundup(usemapsize, 8 * sizeof(unsigned long)); return usemapsize / 8; } static void __init setup_usemap(struct pglist_data *pgdat, struct zone *zone, unsigned long zonesize) { unsigned long usemapsize = usemap_size(zonesize); zone->pageblock_flags = NULL; |
58a01a457 mm/page_alloc.c: ... |
4031 |
if (usemapsize) |
8f389a99b mm: use alloc_boo... |
4032 4033 |
zone->pageblock_flags = alloc_bootmem_node_nopanic(pgdat, usemapsize); |
835c134ec Add a bitmap that... |
4034 4035 |
} #else |
fa9f90be7 Kill off a bunch ... |
4036 |
static inline void setup_usemap(struct pglist_data *pgdat, |
835c134ec Add a bitmap that... |
4037 4038 |
struct zone *zone, unsigned long zonesize) {} #endif /* CONFIG_SPARSEMEM */ |
d9c234005 Do not depend on ... |
4039 |
#ifdef CONFIG_HUGETLB_PAGE_SIZE_VARIABLE |
ba72cb8cb Fix boot problem ... |
4040 4041 4042 4043 4044 4045 4046 4047 4048 |
/* Return a sensible default order for the pageblock size. */ static inline int pageblock_default_order(void) { if (HPAGE_SHIFT > PAGE_SHIFT) return HUGETLB_PAGE_ORDER; return MAX_ORDER-1; } |
d9c234005 Do not depend on ... |
4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 |
/* Initialise the number of pages represented by NR_PAGEBLOCK_BITS */ static inline void __init set_pageblock_order(unsigned int order) { /* Check that pageblock_nr_pages has not already been setup */ if (pageblock_order) return; /* * Assume the largest contiguous order of interest is a huge page. * This value may be variable depending on boot parameters on IA64 */ pageblock_order = order; } #else /* CONFIG_HUGETLB_PAGE_SIZE_VARIABLE */ |
ba72cb8cb Fix boot problem ... |
4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 |
/* * When CONFIG_HUGETLB_PAGE_SIZE_VARIABLE is not set, set_pageblock_order() * and pageblock_default_order() are unused as pageblock_order is set * at compile-time. See include/linux/pageblock-flags.h for the values of * pageblock_order based on the kernel config */ static inline int pageblock_default_order(unsigned int order) { return MAX_ORDER-1; } |
d9c234005 Do not depend on ... |
4073 4074 4075 |
#define set_pageblock_order(x) do {} while (0) #endif /* CONFIG_HUGETLB_PAGE_SIZE_VARIABLE */ |
1da177e4c Linux-2.6.12-rc2 |
4076 4077 4078 4079 4080 4081 |
/* * Set up the zone data structures: * - mark all pages reserved * - mark all memory queues empty * - clear the memory bitmaps */ |
b5a0e0113 Solve section mis... |
4082 |
static void __paginginit free_area_init_core(struct pglist_data *pgdat, |
1da177e4c Linux-2.6.12-rc2 |
4083 4084 |
unsigned long *zones_size, unsigned long *zholes_size) { |
2f1b62486 [PATCH] reduce MA... |
4085 |
enum zone_type j; |
ed8ece2ec [PATCH] memory ho... |
4086 |
int nid = pgdat->node_id; |
1da177e4c Linux-2.6.12-rc2 |
4087 |
unsigned long zone_start_pfn = pgdat->node_start_pfn; |
718127cc3 [PATCH] wait_tabl... |
4088 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
4089 |
|
208d54e55 [PATCH] memory ho... |
4090 |
pgdat_resize_init(pgdat); |
1da177e4c Linux-2.6.12-rc2 |
4091 4092 4093 |
pgdat->nr_zones = 0; init_waitqueue_head(&pgdat->kswapd_wait); pgdat->kswapd_max_order = 0; |
52d4b9ac0 memcg: allocate a... |
4094 |
pgdat_page_cgroup_init(pgdat); |
1da177e4c Linux-2.6.12-rc2 |
4095 4096 4097 |
for (j = 0; j < MAX_NR_ZONES; j++) { struct zone *zone = pgdat->node_zones + j; |
0e0b864e0 [PATCH] Account f... |
4098 |
unsigned long size, realsize, memmap_pages; |
b69408e88 vmscan: Use an in... |
4099 |
enum lru_list l; |
1da177e4c Linux-2.6.12-rc2 |
4100 |
|
c713216de [PATCH] Introduce... |
4101 4102 4103 |
size = zone_spanned_pages_in_node(nid, j, zones_size); realsize = size - zone_absent_pages_in_node(nid, j, zholes_size); |
1da177e4c Linux-2.6.12-rc2 |
4104 |
|
0e0b864e0 [PATCH] Account f... |
4105 4106 4107 4108 4109 |
/* * Adjust realsize so that it accounts for how much memory * is used by this zone for memmap. This affects the watermark * and per-cpu initialisations */ |
f72321541 mm: don't drop a ... |
4110 4111 |
memmap_pages = PAGE_ALIGN(size * sizeof(struct page)) >> PAGE_SHIFT; |
0e0b864e0 [PATCH] Account f... |
4112 4113 |
if (realsize >= memmap_pages) { realsize -= memmap_pages; |
5594c8c81 mm: print out mem... |
4114 4115 4116 4117 4118 |
if (memmap_pages) printk(KERN_DEBUG " %s zone: %lu pages used for memmap ", zone_names[j], memmap_pages); |
0e0b864e0 [PATCH] Account f... |
4119 4120 4121 4122 4123 |
} else printk(KERN_WARNING " %s zone: %lu pages exceeds realsize %lu ", zone_names[j], memmap_pages, realsize); |
6267276f3 [PATCH] optional ... |
4124 4125 |
/* Account for reserved pages */ if (j == 0 && realsize > dma_reserve) { |
0e0b864e0 [PATCH] Account f... |
4126 |
realsize -= dma_reserve; |
d903ef9f3 mm: print out mem... |
4127 4128 |
printk(KERN_DEBUG " %s zone: %lu pages reserved ", |
6267276f3 [PATCH] optional ... |
4129 |
zone_names[0], dma_reserve); |
0e0b864e0 [PATCH] Account f... |
4130 |
} |
98d2b0ebd [PATCH] reduce MA... |
4131 |
if (!is_highmem_idx(j)) |
1da177e4c Linux-2.6.12-rc2 |
4132 4133 4134 4135 4136 |
nr_kernel_pages += realsize; nr_all_pages += realsize; zone->spanned_pages = size; zone->present_pages = realsize; |
9614634fe [PATCH] ZVC/zone_... |
4137 |
#ifdef CONFIG_NUMA |
d5f541ed6 [PATCH] Add node ... |
4138 |
zone->node = nid; |
8417bba4b [PATCH] Replace m... |
4139 |
zone->min_unmapped_pages = (realsize*sysctl_min_unmapped_ratio) |
9614634fe [PATCH] ZVC/zone_... |
4140 |
/ 100; |
0ff38490c [PATCH] zone_recl... |
4141 |
zone->min_slab_pages = (realsize * sysctl_min_slab_ratio) / 100; |
9614634fe [PATCH] ZVC/zone_... |
4142 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
4143 4144 4145 |
zone->name = zone_names[j]; spin_lock_init(&zone->lock); spin_lock_init(&zone->lru_lock); |
bdc8cb984 [PATCH] memory ho... |
4146 |
zone_seqlock_init(zone); |
1da177e4c Linux-2.6.12-rc2 |
4147 |
zone->zone_pgdat = pgdat; |
1da177e4c Linux-2.6.12-rc2 |
4148 |
|
ed8ece2ec [PATCH] memory ho... |
4149 |
zone_pcp_init(zone); |
246e87a93 memcg: fix get_sc... |
4150 |
for_each_lru(l) |
b69408e88 vmscan: Use an in... |
4151 |
INIT_LIST_HEAD(&zone->lru[l].list); |
6e9015716 mm: introduce zon... |
4152 4153 4154 4155 |
zone->reclaim_stat.recent_rotated[0] = 0; zone->reclaim_stat.recent_rotated[1] = 0; zone->reclaim_stat.recent_scanned[0] = 0; zone->reclaim_stat.recent_scanned[1] = 0; |
2244b95a7 [PATCH] zoned vm ... |
4156 |
zap_zone_vm_stats(zone); |
e815af95f oom: change all_u... |
4157 |
zone->flags = 0; |
1da177e4c Linux-2.6.12-rc2 |
4158 4159 |
if (!size) continue; |
ba72cb8cb Fix boot problem ... |
4160 |
set_pageblock_order(pageblock_default_order()); |
835c134ec Add a bitmap that... |
4161 |
setup_usemap(pgdat, zone, size); |
a2f3aa025 [PATCH] Fix spars... |
4162 4163 |
ret = init_currently_empty_zone(zone, zone_start_pfn, size, MEMMAP_EARLY); |
718127cc3 [PATCH] wait_tabl... |
4164 |
BUG_ON(ret); |
76cdd58e5 memory_hotplug: a... |
4165 |
memmap_init(size, nid, j, zone_start_pfn); |
1da177e4c Linux-2.6.12-rc2 |
4166 |
zone_start_pfn += size; |
1da177e4c Linux-2.6.12-rc2 |
4167 4168 |
} } |
577a32f62 mm: fix section m... |
4169 |
static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat) |
1da177e4c Linux-2.6.12-rc2 |
4170 |
{ |
1da177e4c Linux-2.6.12-rc2 |
4171 4172 4173 |
/* Skip empty nodes */ if (!pgdat->node_spanned_pages) return; |
d41dee369 [PATCH] sparsemem... |
4174 |
#ifdef CONFIG_FLAT_NODE_MEM_MAP |
1da177e4c Linux-2.6.12-rc2 |
4175 4176 |
/* ia64 gets its own node_mem_map, before this, without bootmem */ if (!pgdat->node_mem_map) { |
e984bb43f [PATCH] Align the... |
4177 |
unsigned long size, start, end; |
d41dee369 [PATCH] sparsemem... |
4178 |
struct page *map; |
e984bb43f [PATCH] Align the... |
4179 4180 4181 4182 4183 4184 4185 4186 4187 |
/* * The zone's endpoints aren't required to be MAX_ORDER * aligned but the node_mem_map endpoints must be in order * for the buddy allocator to function correctly. */ start = pgdat->node_start_pfn & ~(MAX_ORDER_NR_PAGES - 1); end = pgdat->node_start_pfn + pgdat->node_spanned_pages; end = ALIGN(end, MAX_ORDER_NR_PAGES); size = (end - start) * sizeof(struct page); |
6f167ec72 [PATCH] sparsemem... |
4188 4189 |
map = alloc_remap(pgdat->node_id, size); if (!map) |
8f389a99b mm: use alloc_boo... |
4190 |
map = alloc_bootmem_node_nopanic(pgdat, size); |
e984bb43f [PATCH] Align the... |
4191 |
pgdat->node_mem_map = map + (pgdat->node_start_pfn - start); |
1da177e4c Linux-2.6.12-rc2 |
4192 |
} |
12d810c1b m68k: discontinuo... |
4193 |
#ifndef CONFIG_NEED_MULTIPLE_NODES |
1da177e4c Linux-2.6.12-rc2 |
4194 4195 4196 |
/* * With no DISCONTIG, the global mem_map is just set as node 0's */ |
c713216de [PATCH] Introduce... |
4197 |
if (pgdat == NODE_DATA(0)) { |
1da177e4c Linux-2.6.12-rc2 |
4198 |
mem_map = NODE_DATA(0)->node_mem_map; |
c713216de [PATCH] Introduce... |
4199 4200 |
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP if (page_to_pfn(mem_map) != pgdat->node_start_pfn) |
467bc461d Fix crash with FL... |
4201 |
mem_map -= (pgdat->node_start_pfn - ARCH_PFN_OFFSET); |
c713216de [PATCH] Introduce... |
4202 4203 |
#endif /* CONFIG_ARCH_POPULATES_NODE_MAP */ } |
1da177e4c Linux-2.6.12-rc2 |
4204 |
#endif |
d41dee369 [PATCH] sparsemem... |
4205 |
#endif /* CONFIG_FLAT_NODE_MEM_MAP */ |
1da177e4c Linux-2.6.12-rc2 |
4206 |
} |
9109fb7b3 mm: drop unneeded... |
4207 4208 |
void __paginginit free_area_init_node(int nid, unsigned long *zones_size, unsigned long node_start_pfn, unsigned long *zholes_size) |
1da177e4c Linux-2.6.12-rc2 |
4209 |
{ |
9109fb7b3 mm: drop unneeded... |
4210 |
pg_data_t *pgdat = NODE_DATA(nid); |
1da177e4c Linux-2.6.12-rc2 |
4211 4212 |
pgdat->node_id = nid; pgdat->node_start_pfn = node_start_pfn; |
c713216de [PATCH] Introduce... |
4213 |
calculate_node_totalpages(pgdat, zones_size, zholes_size); |
1da177e4c Linux-2.6.12-rc2 |
4214 4215 |
alloc_node_mem_map(pgdat); |
e8c27ac91 x86, numa, 32-bit... |
4216 4217 4218 4219 4220 4221 |
#ifdef CONFIG_FLAT_NODE_MEM_MAP printk(KERN_DEBUG "free_area_init_node: node %d, pgdat %08lx, node_mem_map %08lx ", nid, (unsigned long)pgdat, (unsigned long)pgdat->node_mem_map); #endif |
1da177e4c Linux-2.6.12-rc2 |
4222 4223 4224 |
free_area_init_core(pgdat, zones_size, zholes_size); } |
c713216de [PATCH] Introduce... |
4225 |
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP |
418508c13 fix unused setup_... |
4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 |
#if MAX_NUMNODES > 1 /* * Figure out the number of possible node ids. */ static void __init setup_nr_node_ids(void) { unsigned int node; unsigned int highest = 0; for_each_node_mask(node, node_possible_map) highest = node; nr_node_ids = highest + 1; } #else static inline void setup_nr_node_ids(void) { } #endif |
c713216de [PATCH] Introduce... |
4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 |
/** * add_active_range - Register a range of PFNs backed by physical memory * @nid: The node ID the range resides on * @start_pfn: The start PFN of the available physical memory * @end_pfn: The end PFN of the available physical memory * * These ranges are stored in an early_node_map[] and later used by * free_area_init_nodes() to calculate zone sizes and holes. If the * range spans a memory hole, it is up to the architecture to ensure * the memory is not freed by the bootmem allocator. If possible * the range being registered will be merged with existing ranges. */ void __init add_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn) { int i; |
6b74ab97b mm: add a basic d... |
4261 4262 4263 4264 4265 4266 |
mminit_dprintk(MMINIT_TRACE, "memory_register", "Entering add_active_range(%d, %#lx, %#lx) " "%d entries of %d used ", nid, start_pfn, end_pfn, nr_nodemap_entries, MAX_ACTIVE_REGIONS); |
c713216de [PATCH] Introduce... |
4267 |
|
2dbb51c49 mm: make defensiv... |
4268 |
mminit_validate_memmodel_limits(&start_pfn, &end_pfn); |
c713216de [PATCH] Introduce... |
4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 |
/* Merge with existing active regions if possible */ for (i = 0; i < nr_nodemap_entries; i++) { if (early_node_map[i].nid != nid) continue; /* Skip if an existing region covers this new one */ if (start_pfn >= early_node_map[i].start_pfn && end_pfn <= early_node_map[i].end_pfn) return; /* Merge forward if suitable */ if (start_pfn <= early_node_map[i].end_pfn && end_pfn > early_node_map[i].end_pfn) { early_node_map[i].end_pfn = end_pfn; return; } /* Merge backward if suitable */ |
d2dbe08dd mm/page_alloc: fi... |
4287 |
if (start_pfn < early_node_map[i].start_pfn && |
c713216de [PATCH] Introduce... |
4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 |
end_pfn >= early_node_map[i].start_pfn) { early_node_map[i].start_pfn = start_pfn; return; } } /* Check that early_node_map is large enough */ if (i >= MAX_ACTIVE_REGIONS) { printk(KERN_CRIT "More than %d memory regions, truncating ", MAX_ACTIVE_REGIONS); return; } early_node_map[i].nid = nid; early_node_map[i].start_pfn = start_pfn; early_node_map[i].end_pfn = end_pfn; nr_nodemap_entries = i + 1; } /** |
cc1050baf x86: replace shri... |
4309 |
* remove_active_range - Shrink an existing registered range of PFNs |
c713216de [PATCH] Introduce... |
4310 |
* @nid: The node id the range is on that should be shrunk |
cc1050baf x86: replace shri... |
4311 4312 |
* @start_pfn: The new PFN of the range * @end_pfn: The new PFN of the range |
c713216de [PATCH] Introduce... |
4313 4314 |
* * i386 with NUMA use alloc_remap() to store a node_mem_map on a local node. |
cc1a9d86c mm, x86: shrink_a... |
4315 4316 4317 |
* The map is kept near the end physical page range that has already been * registered. This function allows an arch to shrink an existing registered * range. |
c713216de [PATCH] Introduce... |
4318 |
*/ |
cc1050baf x86: replace shri... |
4319 4320 |
void __init remove_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn) |
c713216de [PATCH] Introduce... |
4321 |
{ |
cc1a9d86c mm, x86: shrink_a... |
4322 4323 |
int i, j; int removed = 0; |
c713216de [PATCH] Introduce... |
4324 |
|
cc1050baf x86: replace shri... |
4325 4326 4327 |
printk(KERN_DEBUG "remove_active_range (%d, %lu, %lu) ", nid, start_pfn, end_pfn); |
c713216de [PATCH] Introduce... |
4328 |
/* Find the old active region end and shrink */ |
cc1a9d86c mm, x86: shrink_a... |
4329 |
for_each_active_range_index_in_nid(i, nid) { |
cc1050baf x86: replace shri... |
4330 4331 |
if (early_node_map[i].start_pfn >= start_pfn && early_node_map[i].end_pfn <= end_pfn) { |
cc1a9d86c mm, x86: shrink_a... |
4332 |
/* clear it */ |
cc1050baf x86: replace shri... |
4333 |
early_node_map[i].start_pfn = 0; |
cc1a9d86c mm, x86: shrink_a... |
4334 4335 4336 4337 |
early_node_map[i].end_pfn = 0; removed = 1; continue; } |
cc1050baf x86: replace shri... |
4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 |
if (early_node_map[i].start_pfn < start_pfn && early_node_map[i].end_pfn > start_pfn) { unsigned long temp_end_pfn = early_node_map[i].end_pfn; early_node_map[i].end_pfn = start_pfn; if (temp_end_pfn > end_pfn) add_active_range(nid, end_pfn, temp_end_pfn); continue; } if (early_node_map[i].start_pfn >= start_pfn && early_node_map[i].end_pfn > end_pfn && early_node_map[i].start_pfn < end_pfn) { early_node_map[i].start_pfn = end_pfn; |
cc1a9d86c mm, x86: shrink_a... |
4350 |
continue; |
c713216de [PATCH] Introduce... |
4351 |
} |
cc1a9d86c mm, x86: shrink_a... |
4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 |
} if (!removed) return; /* remove the blank ones */ for (i = nr_nodemap_entries - 1; i > 0; i--) { if (early_node_map[i].nid != nid) continue; if (early_node_map[i].end_pfn) continue; /* we found it, get rid of it */ for (j = i; j < nr_nodemap_entries - 1; j++) memcpy(&early_node_map[j], &early_node_map[j+1], sizeof(early_node_map[j])); j = nr_nodemap_entries - 1; memset(&early_node_map[j], 0, sizeof(early_node_map[j])); nr_nodemap_entries--; } |
c713216de [PATCH] Introduce... |
4371 4372 4373 4374 |
} /** * remove_all_active_ranges - Remove all currently registered regions |
88ca3b94e [PATCH] page_allo... |
4375 |
* |
c713216de [PATCH] Introduce... |
4376 4377 4378 4379 |
* During discovery, it may be found that a table like SRAT is invalid * and an alternative discovery method must be used. This function removes * all currently registered regions. */ |
88ca3b94e [PATCH] page_allo... |
4380 |
void __init remove_all_active_ranges(void) |
c713216de [PATCH] Introduce... |
4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 |
{ memset(early_node_map, 0, sizeof(early_node_map)); nr_nodemap_entries = 0; } /* Compare two active node_active_regions */ static int __init cmp_node_active_region(const void *a, const void *b) { struct node_active_region *arange = (struct node_active_region *)a; struct node_active_region *brange = (struct node_active_region *)b; /* Done this way to avoid overflows */ if (arange->start_pfn > brange->start_pfn) return 1; if (arange->start_pfn < brange->start_pfn) return -1; return 0; } /* sort the node_map by start_pfn */ |
329962503 x86: Fix checking... |
4402 |
void __init sort_node_map(void) |
c713216de [PATCH] Introduce... |
4403 4404 4405 4406 4407 |
{ sort(early_node_map, (size_t)nr_nodemap_entries, sizeof(struct node_active_region), cmp_node_active_region, NULL); } |
1e01979c8 x86, numa: Implem... |
4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 |
/** * node_map_pfn_alignment - determine the maximum internode alignment * * This function should be called after node map is populated and sorted. * It calculates the maximum power of two alignment which can distinguish * all the nodes. * * For example, if all nodes are 1GiB and aligned to 1GiB, the return value * would indicate 1GiB alignment with (1 << (30 - PAGE_SHIFT)). If the * nodes are shifted by 256MiB, 256MiB. Note that if only the last node is * shifted, 1GiB is enough and this function will indicate so. * * This is used to test whether pfn -> nid mapping of the chosen memory * model has fine enough granularity to avoid incorrect mapping for the * populated node map. * * Returns the determined alignment in pfn's. 0 if there is no alignment * requirement (single node). */ unsigned long __init node_map_pfn_alignment(void) { unsigned long accl_mask = 0, last_end = 0; int last_nid = -1; int i; for_each_active_range_index_in_nid(i, MAX_NUMNODES) { int nid = early_node_map[i].nid; unsigned long start = early_node_map[i].start_pfn; unsigned long end = early_node_map[i].end_pfn; unsigned long mask; if (!start || last_nid < 0 || last_nid == nid) { last_nid = nid; last_end = end; continue; } /* * Start with a mask granular enough to pin-point to the * start pfn and tick off bits one-by-one until it becomes * too coarse to separate the current node from the last. */ mask = ~((1 << __ffs(start)) - 1); while (mask && last_end <= (start & (mask << 1))) mask <<= 1; /* accumulate all internode masks */ accl_mask |= mask; } /* convert mask to number of pages */ return ~accl_mask + 1; } |
a6af2bc3d [PATCH] Avoid exc... |
4461 |
/* Find the lowest pfn for a node */ |
b69a7288e mm/page_alloc.c: ... |
4462 |
static unsigned long __init find_min_pfn_for_node(int nid) |
c713216de [PATCH] Introduce... |
4463 4464 |
{ int i; |
a6af2bc3d [PATCH] Avoid exc... |
4465 |
unsigned long min_pfn = ULONG_MAX; |
1abbfb412 [PATCH] x86_64: f... |
4466 |
|
c713216de [PATCH] Introduce... |
4467 4468 |
/* Assuming a sorted map, the first range found has the starting pfn */ for_each_active_range_index_in_nid(i, nid) |
a6af2bc3d [PATCH] Avoid exc... |
4469 |
min_pfn = min(min_pfn, early_node_map[i].start_pfn); |
c713216de [PATCH] Introduce... |
4470 |
|
a6af2bc3d [PATCH] Avoid exc... |
4471 4472 |
if (min_pfn == ULONG_MAX) { printk(KERN_WARNING |
2bc0d2615 x86 boot: more co... |
4473 4474 |
"Could not find start_pfn for node %d ", nid); |
a6af2bc3d [PATCH] Avoid exc... |
4475 4476 4477 4478 |
return 0; } return min_pfn; |
c713216de [PATCH] Introduce... |
4479 4480 4481 4482 4483 4484 |
} /** * find_min_pfn_with_active_regions - Find the minimum PFN registered * * It returns the minimum PFN based on information provided via |
88ca3b94e [PATCH] page_allo... |
4485 |
* add_active_range(). |
c713216de [PATCH] Introduce... |
4486 4487 4488 4489 4490 |
*/ unsigned long __init find_min_pfn_with_active_regions(void) { return find_min_pfn_for_node(MAX_NUMNODES); } |
37b07e416 memoryless nodes:... |
4491 4492 4493 4494 4495 |
/* * early_calculate_totalpages() * Sum pages in active regions for movable zone. * Populate N_HIGH_MEMORY for calculating usable_nodes. */ |
484f51f82 mm/page_alloc.c: ... |
4496 |
static unsigned long __init early_calculate_totalpages(void) |
7e63efef8 Add a movablecore... |
4497 4498 4499 |
{ int i; unsigned long totalpages = 0; |
37b07e416 memoryless nodes:... |
4500 4501 |
for (i = 0; i < nr_nodemap_entries; i++) { unsigned long pages = early_node_map[i].end_pfn - |
7e63efef8 Add a movablecore... |
4502 |
early_node_map[i].start_pfn; |
37b07e416 memoryless nodes:... |
4503 4504 4505 4506 4507 |
totalpages += pages; if (pages) node_set_state(early_node_map[i].nid, N_HIGH_MEMORY); } return totalpages; |
7e63efef8 Add a movablecore... |
4508 |
} |
2a1e274ac Create the ZONE_M... |
4509 4510 4511 4512 4513 4514 |
/* * Find the PFN the Movable zone begins in each node. Kernel memory * is spread evenly between nodes as long as the nodes have enough * memory. When they don't, some nodes will have more kernelcore than * others */ |
b69a7288e mm/page_alloc.c: ... |
4515 |
static void __init find_zone_movable_pfns_for_nodes(unsigned long *movable_pfn) |
2a1e274ac Create the ZONE_M... |
4516 4517 4518 4519 |
{ int i, nid; unsigned long usable_startpfn; unsigned long kernelcore_node, kernelcore_remaining; |
66918dcdf x86: only clear n... |
4520 4521 |
/* save the state before borrow the nodemask */ nodemask_t saved_node_state = node_states[N_HIGH_MEMORY]; |
37b07e416 memoryless nodes:... |
4522 4523 |
unsigned long totalpages = early_calculate_totalpages(); int usable_nodes = nodes_weight(node_states[N_HIGH_MEMORY]); |
2a1e274ac Create the ZONE_M... |
4524 |
|
7e63efef8 Add a movablecore... |
4525 4526 4527 4528 4529 4530 4531 4532 4533 |
/* * If movablecore was specified, calculate what size of * kernelcore that corresponds so that memory usable for * any allocation type is evenly spread. If both kernelcore * and movablecore are specified, then the value of kernelcore * will be used for required_kernelcore if it's greater than * what movablecore would have allowed. */ if (required_movablecore) { |
7e63efef8 Add a movablecore... |
4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 |
unsigned long corepages; /* * Round-up so that ZONE_MOVABLE is at least as large as what * was requested by the user */ required_movablecore = roundup(required_movablecore, MAX_ORDER_NR_PAGES); corepages = totalpages - required_movablecore; required_kernelcore = max(required_kernelcore, corepages); } |
2a1e274ac Create the ZONE_M... |
4546 4547 |
/* If kernelcore was not specified, there is no ZONE_MOVABLE */ if (!required_kernelcore) |
66918dcdf x86: only clear n... |
4548 |
goto out; |
2a1e274ac Create the ZONE_M... |
4549 4550 4551 4552 4553 4554 4555 4556 |
/* usable_startpfn is the lowest possible pfn ZONE_MOVABLE can be at */ find_usable_zone_for_movable(); usable_startpfn = arch_zone_lowest_possible_pfn[movable_zone]; restart: /* Spread kernelcore memory as evenly as possible throughout nodes */ kernelcore_node = required_kernelcore / usable_nodes; |
37b07e416 memoryless nodes:... |
4557 |
for_each_node_state(nid, N_HIGH_MEMORY) { |
2a1e274ac Create the ZONE_M... |
4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 |
/* * Recalculate kernelcore_node if the division per node * now exceeds what is necessary to satisfy the requested * amount of memory for the kernel */ if (required_kernelcore < kernelcore_node) kernelcore_node = required_kernelcore / usable_nodes; /* * As the map is walked, we track how much memory is usable * by the kernel using kernelcore_remaining. When it is * 0, the rest of the node is usable by ZONE_MOVABLE */ kernelcore_remaining = kernelcore_node; /* Go through each range of PFNs within this node */ for_each_active_range_index_in_nid(i, nid) { unsigned long start_pfn, end_pfn; unsigned long size_pages; start_pfn = max(early_node_map[i].start_pfn, zone_movable_pfn[nid]); end_pfn = early_node_map[i].end_pfn; if (start_pfn >= end_pfn) continue; /* Account for what is only usable for kernelcore */ if (start_pfn < usable_startpfn) { unsigned long kernel_pages; kernel_pages = min(end_pfn, usable_startpfn) - start_pfn; kernelcore_remaining -= min(kernel_pages, kernelcore_remaining); required_kernelcore -= min(kernel_pages, required_kernelcore); /* Continue if range is now fully accounted */ if (end_pfn <= usable_startpfn) { /* * Push zone_movable_pfn to the end so * that if we have to rebalance * kernelcore across nodes, we will * not double account here */ zone_movable_pfn[nid] = end_pfn; continue; } start_pfn = usable_startpfn; } /* * The usable PFN range for ZONE_MOVABLE is from * start_pfn->end_pfn. Calculate size_pages as the * number of pages used as kernelcore */ size_pages = end_pfn - start_pfn; if (size_pages > kernelcore_remaining) size_pages = kernelcore_remaining; zone_movable_pfn[nid] = start_pfn + size_pages; /* * Some kernelcore has been met, update counts and * break if the kernelcore for this node has been * satisified */ required_kernelcore -= min(required_kernelcore, size_pages); kernelcore_remaining -= size_pages; if (!kernelcore_remaining) break; } } /* * If there is still required_kernelcore, we do another pass with one * less node in the count. This will push zone_movable_pfn[nid] further * along on the nodes that still have memory until kernelcore is * satisified */ usable_nodes--; if (usable_nodes && required_kernelcore > usable_nodes) goto restart; /* Align start of ZONE_MOVABLE on all nids to MAX_ORDER_NR_PAGES */ for (nid = 0; nid < MAX_NUMNODES; nid++) zone_movable_pfn[nid] = roundup(zone_movable_pfn[nid], MAX_ORDER_NR_PAGES); |
66918dcdf x86: only clear n... |
4647 4648 4649 4650 |
out: /* restore the node_state */ node_states[N_HIGH_MEMORY] = saved_node_state; |
2a1e274ac Create the ZONE_M... |
4651 |
} |
37b07e416 memoryless nodes:... |
4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 |
/* Any regular memory on that node ? */ static void check_for_regular_memory(pg_data_t *pgdat) { #ifdef CONFIG_HIGHMEM enum zone_type zone_type; for (zone_type = 0; zone_type <= ZONE_NORMAL; zone_type++) { struct zone *zone = &pgdat->node_zones[zone_type]; if (zone->present_pages) node_set_state(zone_to_nid(zone), N_NORMAL_MEMORY); } #endif } |
c713216de [PATCH] Introduce... |
4665 4666 |
/** * free_area_init_nodes - Initialise all pg_data_t and zone data |
88ca3b94e [PATCH] page_allo... |
4667 |
* @max_zone_pfn: an array of max PFNs for each zone |
c713216de [PATCH] Introduce... |
4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 |
* * This will call free_area_init_node() for each active node in the system. * Using the page ranges provided by add_active_range(), the size of each * zone in each node and their holes is calculated. If the maximum PFN * between two adjacent zones match, it is assumed that the zone is empty. * For example, if arch_max_dma_pfn == arch_max_dma32_pfn, it is assumed * that arch_max_dma32_pfn has no pages. It is also assumed that a zone * starts where the previous one ended. For example, ZONE_DMA32 starts * at arch_max_dma_pfn. */ void __init free_area_init_nodes(unsigned long *max_zone_pfn) { unsigned long nid; |
db99100d2 mm/page_alloc.c:f... |
4681 |
int i; |
c713216de [PATCH] Introduce... |
4682 |
|
a6af2bc3d [PATCH] Avoid exc... |
4683 4684 |
/* Sort early_node_map as initialisation assumes it is sorted */ sort_node_map(); |
c713216de [PATCH] Introduce... |
4685 4686 4687 4688 4689 4690 4691 4692 |
/* Record where the zone boundaries are */ memset(arch_zone_lowest_possible_pfn, 0, sizeof(arch_zone_lowest_possible_pfn)); memset(arch_zone_highest_possible_pfn, 0, sizeof(arch_zone_highest_possible_pfn)); arch_zone_lowest_possible_pfn[0] = find_min_pfn_with_active_regions(); arch_zone_highest_possible_pfn[0] = max_zone_pfn[0]; for (i = 1; i < MAX_NR_ZONES; i++) { |
2a1e274ac Create the ZONE_M... |
4693 4694 |
if (i == ZONE_MOVABLE) continue; |
c713216de [PATCH] Introduce... |
4695 4696 4697 4698 4699 |
arch_zone_lowest_possible_pfn[i] = arch_zone_highest_possible_pfn[i-1]; arch_zone_highest_possible_pfn[i] = max(max_zone_pfn[i], arch_zone_lowest_possible_pfn[i]); } |
2a1e274ac Create the ZONE_M... |
4700 4701 4702 4703 4704 4705 |
arch_zone_lowest_possible_pfn[ZONE_MOVABLE] = 0; arch_zone_highest_possible_pfn[ZONE_MOVABLE] = 0; /* Find the PFNs that ZONE_MOVABLE begins at in each node */ memset(zone_movable_pfn, 0, sizeof(zone_movable_pfn)); find_zone_movable_pfns_for_nodes(zone_movable_pfn); |
c713216de [PATCH] Introduce... |
4706 |
|
c713216de [PATCH] Introduce... |
4707 4708 4709 |
/* Print out the zone ranges */ printk("Zone PFN ranges: "); |
2a1e274ac Create the ZONE_M... |
4710 4711 4712 |
for (i = 0; i < MAX_NR_ZONES; i++) { if (i == ZONE_MOVABLE) continue; |
72f0ba025 mm: suppress pfn ... |
4713 4714 4715 4716 4717 4718 4719 4720 |
printk(" %-8s ", zone_names[i]); if (arch_zone_lowest_possible_pfn[i] == arch_zone_highest_possible_pfn[i]) printk("empty "); else printk("%0#10lx -> %0#10lx ", |
c713216de [PATCH] Introduce... |
4721 4722 |
arch_zone_lowest_possible_pfn[i], arch_zone_highest_possible_pfn[i]); |
2a1e274ac Create the ZONE_M... |
4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 |
} /* Print out the PFNs ZONE_MOVABLE begins at in each node */ printk("Movable zone start PFN for each node "); for (i = 0; i < MAX_NUMNODES; i++) { if (zone_movable_pfn[i]) printk(" Node %d: %lu ", i, zone_movable_pfn[i]); } |
c713216de [PATCH] Introduce... |
4733 4734 4735 4736 4737 |
/* Print out the early_node_map[] */ printk("early_node_map[%d] active PFN ranges ", nr_nodemap_entries); for (i = 0; i < nr_nodemap_entries; i++) |
5dab8ec13 mm, generic, x86 ... |
4738 4739 |
printk(" %3d: %0#10lx -> %0#10lx ", early_node_map[i].nid, |
c713216de [PATCH] Introduce... |
4740 4741 4742 4743 |
early_node_map[i].start_pfn, early_node_map[i].end_pfn); /* Initialise every node */ |
708614e61 mm: verify the pa... |
4744 |
mminit_verify_pageflags_layout(); |
8ef828668 [PATCH] slab: red... |
4745 |
setup_nr_node_ids(); |
c713216de [PATCH] Introduce... |
4746 4747 |
for_each_online_node(nid) { pg_data_t *pgdat = NODE_DATA(nid); |
9109fb7b3 mm: drop unneeded... |
4748 |
free_area_init_node(nid, NULL, |
c713216de [PATCH] Introduce... |
4749 |
find_min_pfn_for_node(nid), NULL); |
37b07e416 memoryless nodes:... |
4750 4751 4752 4753 4754 |
/* Any memory on that node */ if (pgdat->node_present_pages) node_set_state(nid, N_HIGH_MEMORY); check_for_regular_memory(pgdat); |
c713216de [PATCH] Introduce... |
4755 4756 |
} } |
2a1e274ac Create the ZONE_M... |
4757 |
|
7e63efef8 Add a movablecore... |
4758 |
static int __init cmdline_parse_core(char *p, unsigned long *core) |
2a1e274ac Create the ZONE_M... |
4759 4760 4761 4762 4763 4764 |
{ unsigned long long coremem; if (!p) return -EINVAL; coremem = memparse(p, &p); |
7e63efef8 Add a movablecore... |
4765 |
*core = coremem >> PAGE_SHIFT; |
2a1e274ac Create the ZONE_M... |
4766 |
|
7e63efef8 Add a movablecore... |
4767 |
/* Paranoid check that UL is enough for the coremem value */ |
2a1e274ac Create the ZONE_M... |
4768 4769 4770 4771 |
WARN_ON((coremem >> PAGE_SHIFT) > ULONG_MAX); return 0; } |
ed7ed3651 handle kernelcore... |
4772 |
|
7e63efef8 Add a movablecore... |
4773 4774 4775 4776 4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 |
/* * kernelcore=size sets the amount of memory for use for allocations that * cannot be reclaimed or migrated. */ static int __init cmdline_parse_kernelcore(char *p) { return cmdline_parse_core(p, &required_kernelcore); } /* * movablecore=size sets the amount of memory for use for allocations that * can be reclaimed or migrated. */ static int __init cmdline_parse_movablecore(char *p) { return cmdline_parse_core(p, &required_movablecore); } |
ed7ed3651 handle kernelcore... |
4790 |
early_param("kernelcore", cmdline_parse_kernelcore); |
7e63efef8 Add a movablecore... |
4791 |
early_param("movablecore", cmdline_parse_movablecore); |
ed7ed3651 handle kernelcore... |
4792 |
|
c713216de [PATCH] Introduce... |
4793 |
#endif /* CONFIG_ARCH_POPULATES_NODE_MAP */ |
0e0b864e0 [PATCH] Account f... |
4794 |
/** |
88ca3b94e [PATCH] page_allo... |
4795 4796 |
* set_dma_reserve - set the specified number of pages reserved in the first zone * @new_dma_reserve: The number of pages to mark reserved |
0e0b864e0 [PATCH] Account f... |
4797 4798 4799 4800 |
* * The per-cpu batchsize and zone watermarks are determined by present_pages. * In the DMA zone, a significant percentage may be consumed by kernel image * and other unfreeable allocations which can skew the watermarks badly. This |
88ca3b94e [PATCH] page_allo... |
4801 4802 4803 |
* function may optionally be used to account for unfreeable pages in the * first zone (e.g., ZONE_DMA). The effect will be lower watermarks and * smaller per-cpu batchsize. |
0e0b864e0 [PATCH] Account f... |
4804 4805 4806 4807 4808 |
*/ void __init set_dma_reserve(unsigned long new_dma_reserve) { dma_reserve = new_dma_reserve; } |
1da177e4c Linux-2.6.12-rc2 |
4809 4810 |
void __init free_area_init(unsigned long *zones_size) { |
9109fb7b3 mm: drop unneeded... |
4811 |
free_area_init_node(0, zones_size, |
1da177e4c Linux-2.6.12-rc2 |
4812 4813 |
__pa(PAGE_OFFSET) >> PAGE_SHIFT, NULL); } |
1da177e4c Linux-2.6.12-rc2 |
4814 |
|
1da177e4c Linux-2.6.12-rc2 |
4815 4816 4817 4818 |
static int page_alloc_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) { int cpu = (unsigned long)hcpu; |
1da177e4c Linux-2.6.12-rc2 |
4819 |
|
8bb784428 Add suspend-relat... |
4820 |
if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) { |
9f8f21725 Page allocator: c... |
4821 4822 4823 4824 4825 4826 4827 4828 |
drain_pages(cpu); /* * Spill the event counters of the dead processor * into the current processors event counters. * This artificially elevates the count of the current * processor. */ |
f8891e5e1 [PATCH] Light wei... |
4829 |
vm_events_fold_cpu(cpu); |
9f8f21725 Page allocator: c... |
4830 4831 4832 4833 4834 4835 4836 4837 |
/* * Zero the differential counters of the dead processor * so that the vm statistics are consistent. * * This is only okay since the processor is dead and cannot * race with what we are doing. */ |
2244b95a7 [PATCH] zoned vm ... |
4838 |
refresh_cpu_vm_stats(cpu); |
1da177e4c Linux-2.6.12-rc2 |
4839 4840 4841 |
} return NOTIFY_OK; } |
1da177e4c Linux-2.6.12-rc2 |
4842 4843 4844 4845 4846 4847 4848 |
void __init page_alloc_init(void) { hotcpu_notifier(page_alloc_cpu_notify, 0); } /* |
cb45b0e96 [PATCH] overcommi... |
4849 4850 4851 4852 4853 4854 4855 |
* calculate_totalreserve_pages - called when sysctl_lower_zone_reserve_ratio * or min_free_kbytes changes. */ static void calculate_totalreserve_pages(void) { struct pglist_data *pgdat; unsigned long reserve_pages = 0; |
2f6726e54 [PATCH] Apply typ... |
4856 |
enum zone_type i, j; |
cb45b0e96 [PATCH] overcommi... |
4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 |
for_each_online_pgdat(pgdat) { for (i = 0; i < MAX_NR_ZONES; i++) { struct zone *zone = pgdat->node_zones + i; unsigned long max = 0; /* Find valid and maximum lowmem_reserve in the zone */ for (j = i; j < MAX_NR_ZONES; j++) { if (zone->lowmem_reserve[j] > max) max = zone->lowmem_reserve[j]; } |
418589663 page allocator: u... |
4868 4869 |
/* we treat the high watermark as reserved pages. */ max += high_wmark_pages(zone); |
cb45b0e96 [PATCH] overcommi... |
4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 |
if (max > zone->present_pages) max = zone->present_pages; reserve_pages += max; } } totalreserve_pages = reserve_pages; } /* |
1da177e4c Linux-2.6.12-rc2 |
4880 4881 4882 4883 4884 4885 4886 4887 |
* setup_per_zone_lowmem_reserve - called whenever * sysctl_lower_zone_reserve_ratio changes. Ensures that each zone * has a correct pages reserved value, so an adequate number of * pages are left in the zone after a successful __alloc_pages(). */ static void setup_per_zone_lowmem_reserve(void) { struct pglist_data *pgdat; |
2f6726e54 [PATCH] Apply typ... |
4888 |
enum zone_type j, idx; |
1da177e4c Linux-2.6.12-rc2 |
4889 |
|
ec936fc56 [PATCH] for_each_... |
4890 |
for_each_online_pgdat(pgdat) { |
1da177e4c Linux-2.6.12-rc2 |
4891 4892 4893 4894 4895 |
for (j = 0; j < MAX_NR_ZONES; j++) { struct zone *zone = pgdat->node_zones + j; unsigned long present_pages = zone->present_pages; zone->lowmem_reserve[j] = 0; |
2f6726e54 [PATCH] Apply typ... |
4896 4897 |
idx = j; while (idx) { |
1da177e4c Linux-2.6.12-rc2 |
4898 |
struct zone *lower_zone; |
2f6726e54 [PATCH] Apply typ... |
4899 |
idx--; |
1da177e4c Linux-2.6.12-rc2 |
4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 |
if (sysctl_lowmem_reserve_ratio[idx] < 1) sysctl_lowmem_reserve_ratio[idx] = 1; lower_zone = pgdat->node_zones + idx; lower_zone->lowmem_reserve[j] = present_pages / sysctl_lowmem_reserve_ratio[idx]; present_pages += lower_zone->present_pages; } } } |
cb45b0e96 [PATCH] overcommi... |
4910 4911 4912 |
/* update totalreserve_pages */ calculate_totalreserve_pages(); |
1da177e4c Linux-2.6.12-rc2 |
4913 |
} |
88ca3b94e [PATCH] page_allo... |
4914 |
/** |
bc75d33f0 page-allocator: c... |
4915 |
* setup_per_zone_wmarks - called when min_free_kbytes changes |
bce7394a3 page-allocator: r... |
4916 |
* or when memory is hot-{added|removed} |
88ca3b94e [PATCH] page_allo... |
4917 |
* |
bc75d33f0 page-allocator: c... |
4918 4919 |
* Ensures that the watermark[min,low,high] values for each zone are set * correctly with respect to min_free_kbytes. |
1da177e4c Linux-2.6.12-rc2 |
4920 |
*/ |
bc75d33f0 page-allocator: c... |
4921 |
void setup_per_zone_wmarks(void) |
1da177e4c Linux-2.6.12-rc2 |
4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 |
{ unsigned long pages_min = min_free_kbytes >> (PAGE_SHIFT - 10); unsigned long lowmem_pages = 0; struct zone *zone; unsigned long flags; /* Calculate total number of !ZONE_HIGHMEM pages */ for_each_zone(zone) { if (!is_highmem(zone)) lowmem_pages += zone->present_pages; } for_each_zone(zone) { |
ac924c603 [PATCH] setup_per... |
4935 |
u64 tmp; |
1125b4e39 setup_per_zone_pa... |
4936 |
spin_lock_irqsave(&zone->lock, flags); |
ac924c603 [PATCH] setup_per... |
4937 4938 |
tmp = (u64)pages_min * zone->present_pages; do_div(tmp, lowmem_pages); |
1da177e4c Linux-2.6.12-rc2 |
4939 4940 |
if (is_highmem(zone)) { /* |
669ed1752 [PATCH] mm: highm... |
4941 4942 4943 4944 |
* __GFP_HIGH and PF_MEMALLOC allocations usually don't * need highmem pages, so cap pages_min to a small * value here. * |
418589663 page allocator: u... |
4945 |
* The WMARK_HIGH-WMARK_LOW and (WMARK_LOW-WMARK_MIN) |
669ed1752 [PATCH] mm: highm... |
4946 4947 |
* deltas controls asynch page reclaim, and so should * not be capped for highmem. |
1da177e4c Linux-2.6.12-rc2 |
4948 4949 4950 4951 4952 4953 4954 4955 |
*/ int min_pages; min_pages = zone->present_pages / 1024; if (min_pages < SWAP_CLUSTER_MAX) min_pages = SWAP_CLUSTER_MAX; if (min_pages > 128) min_pages = 128; |
418589663 page allocator: u... |
4956 |
zone->watermark[WMARK_MIN] = min_pages; |
1da177e4c Linux-2.6.12-rc2 |
4957 |
} else { |
669ed1752 [PATCH] mm: highm... |
4958 4959 |
/* * If it's a lowmem zone, reserve a number of pages |
1da177e4c Linux-2.6.12-rc2 |
4960 4961 |
* proportionate to the zone's size. */ |
418589663 page allocator: u... |
4962 |
zone->watermark[WMARK_MIN] = tmp; |
1da177e4c Linux-2.6.12-rc2 |
4963 |
} |
418589663 page allocator: u... |
4964 4965 |
zone->watermark[WMARK_LOW] = min_wmark_pages(zone) + (tmp >> 2); zone->watermark[WMARK_HIGH] = min_wmark_pages(zone) + (tmp >> 1); |
56fd56b86 Bias the location... |
4966 |
setup_zone_migrate_reserve(zone); |
1125b4e39 setup_per_zone_pa... |
4967 |
spin_unlock_irqrestore(&zone->lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
4968 |
} |
cb45b0e96 [PATCH] overcommi... |
4969 4970 4971 |
/* update totalreserve_pages */ calculate_totalreserve_pages(); |
1da177e4c Linux-2.6.12-rc2 |
4972 |
} |
55a4462af page_alloc: fix k... |
4973 |
/* |
556adecba vmscan: second ch... |
4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 |
* The inactive anon list should be small enough that the VM never has to * do too much work, but large enough that each inactive page has a chance * to be referenced again before it is swapped out. * * The inactive_anon ratio is the target ratio of ACTIVE_ANON to * INACTIVE_ANON pages on this zone's LRU, maintained by the * pageout code. A zone->inactive_ratio of 3 means 3:1 or 25% of * the anonymous pages are kept on the inactive list. * * total target max * memory ratio inactive anon * ------------------------------------- * 10MB 1 5MB * 100MB 1 50MB * 1GB 3 250MB * 10GB 10 0.9GB * 100GB 31 3GB * 1TB 101 10GB * 10TB 320 32GB */ |
1b79acc91 mm, mem-hotplug: ... |
4994 |
static void __meminit calculate_zone_inactive_ratio(struct zone *zone) |
556adecba vmscan: second ch... |
4995 |
{ |
96cb4df5d page-allocator: a... |
4996 |
unsigned int gb, ratio; |
556adecba vmscan: second ch... |
4997 |
|
96cb4df5d page-allocator: a... |
4998 4999 5000 |
/* Zone size in gigabytes */ gb = zone->present_pages >> (30 - PAGE_SHIFT); if (gb) |
556adecba vmscan: second ch... |
5001 |
ratio = int_sqrt(10 * gb); |
96cb4df5d page-allocator: a... |
5002 5003 |
else ratio = 1; |
556adecba vmscan: second ch... |
5004 |
|
96cb4df5d page-allocator: a... |
5005 5006 |
zone->inactive_ratio = ratio; } |
556adecba vmscan: second ch... |
5007 |
|
839a4fcc8 mm, mem-hotplug: ... |
5008 |
static void __meminit setup_per_zone_inactive_ratio(void) |
96cb4df5d page-allocator: a... |
5009 5010 5011 5012 5013 |
{ struct zone *zone; for_each_zone(zone) calculate_zone_inactive_ratio(zone); |
556adecba vmscan: second ch... |
5014 |
} |
1da177e4c Linux-2.6.12-rc2 |
5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 |
/* * Initialise min_free_kbytes. * * For small machines we want it small (128k min). For large machines * we want it large (64MB max). But it is not linear, because network * bandwidth does not increase linearly with machine size. We use * * min_free_kbytes = 4 * sqrt(lowmem_kbytes), for better accuracy: * min_free_kbytes = sqrt(lowmem_kbytes * 16) * * which yields * * 16MB: 512k * 32MB: 724k * 64MB: 1024k * 128MB: 1448k * 256MB: 2048k * 512MB: 2896k * 1024MB: 4096k * 2048MB: 5792k * 4096MB: 8192k * 8192MB: 11584k * 16384MB: 16384k */ |
1b79acc91 mm, mem-hotplug: ... |
5039 |
int __meminit init_per_zone_wmark_min(void) |
1da177e4c Linux-2.6.12-rc2 |
5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 |
{ unsigned long lowmem_kbytes; lowmem_kbytes = nr_free_buffer_pages() * (PAGE_SIZE >> 10); min_free_kbytes = int_sqrt(lowmem_kbytes * 16); if (min_free_kbytes < 128) min_free_kbytes = 128; if (min_free_kbytes > 65536) min_free_kbytes = 65536; |
bc75d33f0 page-allocator: c... |
5050 |
setup_per_zone_wmarks(); |
a6cccdc36 mm, mem-hotplug: ... |
5051 |
refresh_zone_stat_thresholds(); |
1da177e4c Linux-2.6.12-rc2 |
5052 |
setup_per_zone_lowmem_reserve(); |
556adecba vmscan: second ch... |
5053 |
setup_per_zone_inactive_ratio(); |
1da177e4c Linux-2.6.12-rc2 |
5054 5055 |
return 0; } |
bc75d33f0 page-allocator: c... |
5056 |
module_init(init_per_zone_wmark_min) |
1da177e4c Linux-2.6.12-rc2 |
5057 5058 5059 5060 5061 5062 5063 |
/* * min_free_kbytes_sysctl_handler - just a wrapper around proc_dointvec() so * that we can call two helper functions whenever min_free_kbytes * changes. */ int min_free_kbytes_sysctl_handler(ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
5064 |
void __user *buffer, size_t *length, loff_t *ppos) |
1da177e4c Linux-2.6.12-rc2 |
5065 |
{ |
8d65af789 sysctl: remove "s... |
5066 |
proc_dointvec(table, write, buffer, length, ppos); |
3b1d92c56 Do not disable in... |
5067 |
if (write) |
bc75d33f0 page-allocator: c... |
5068 |
setup_per_zone_wmarks(); |
1da177e4c Linux-2.6.12-rc2 |
5069 5070 |
return 0; } |
9614634fe [PATCH] ZVC/zone_... |
5071 5072 |
#ifdef CONFIG_NUMA int sysctl_min_unmapped_ratio_sysctl_handler(ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
5073 |
void __user *buffer, size_t *length, loff_t *ppos) |
9614634fe [PATCH] ZVC/zone_... |
5074 5075 5076 |
{ struct zone *zone; int rc; |
8d65af789 sysctl: remove "s... |
5077 |
rc = proc_dointvec_minmax(table, write, buffer, length, ppos); |
9614634fe [PATCH] ZVC/zone_... |
5078 5079 5080 5081 |
if (rc) return rc; for_each_zone(zone) |
8417bba4b [PATCH] Replace m... |
5082 |
zone->min_unmapped_pages = (zone->present_pages * |
9614634fe [PATCH] ZVC/zone_... |
5083 5084 5085 |
sysctl_min_unmapped_ratio) / 100; return 0; } |
0ff38490c [PATCH] zone_recl... |
5086 5087 |
int sysctl_min_slab_ratio_sysctl_handler(ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
5088 |
void __user *buffer, size_t *length, loff_t *ppos) |
0ff38490c [PATCH] zone_recl... |
5089 5090 5091 |
{ struct zone *zone; int rc; |
8d65af789 sysctl: remove "s... |
5092 |
rc = proc_dointvec_minmax(table, write, buffer, length, ppos); |
0ff38490c [PATCH] zone_recl... |
5093 5094 5095 5096 5097 5098 5099 5100 |
if (rc) return rc; for_each_zone(zone) zone->min_slab_pages = (zone->present_pages * sysctl_min_slab_ratio) / 100; return 0; } |
9614634fe [PATCH] ZVC/zone_... |
5101 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
5102 5103 5104 5105 5106 5107 |
/* * lowmem_reserve_ratio_sysctl_handler - just a wrapper around * proc_dointvec() so that we can call setup_per_zone_lowmem_reserve() * whenever sysctl_lowmem_reserve_ratio changes. * * The reserve ratio obviously has absolutely no relation with the |
418589663 page allocator: u... |
5108 |
* minimum watermarks. The lowmem reserve ratio can only make sense |
1da177e4c Linux-2.6.12-rc2 |
5109 5110 5111 |
* if in function of the boot time zone sizes. */ int lowmem_reserve_ratio_sysctl_handler(ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
5112 |
void __user *buffer, size_t *length, loff_t *ppos) |
1da177e4c Linux-2.6.12-rc2 |
5113 |
{ |
8d65af789 sysctl: remove "s... |
5114 |
proc_dointvec_minmax(table, write, buffer, length, ppos); |
1da177e4c Linux-2.6.12-rc2 |
5115 5116 5117 |
setup_per_zone_lowmem_reserve(); return 0; } |
8ad4b1fb8 [PATCH] Make high... |
5118 5119 5120 5121 5122 5123 5124 |
/* * percpu_pagelist_fraction - changes the pcp->high for each zone on each * cpu. It is the fraction of total pages in each zone that a hot per cpu pagelist * can have before it gets flushed back to buddy allocator. */ int percpu_pagelist_fraction_sysctl_handler(ctl_table *table, int write, |
8d65af789 sysctl: remove "s... |
5125 |
void __user *buffer, size_t *length, loff_t *ppos) |
8ad4b1fb8 [PATCH] Make high... |
5126 5127 5128 5129 |
{ struct zone *zone; unsigned int cpu; int ret; |
8d65af789 sysctl: remove "s... |
5130 |
ret = proc_dointvec_minmax(table, write, buffer, length, ppos); |
8ad4b1fb8 [PATCH] Make high... |
5131 5132 |
if (!write || (ret == -EINVAL)) return ret; |
364df0ebf mm: fix handling ... |
5133 |
for_each_populated_zone(zone) { |
99dcc3e5a this_cpu: Page al... |
5134 |
for_each_possible_cpu(cpu) { |
8ad4b1fb8 [PATCH] Make high... |
5135 5136 |
unsigned long high; high = zone->present_pages / percpu_pagelist_fraction; |
99dcc3e5a this_cpu: Page al... |
5137 5138 |
setup_pagelist_highmark( per_cpu_ptr(zone->pageset, cpu), high); |
8ad4b1fb8 [PATCH] Make high... |
5139 5140 5141 5142 |
} } return 0; } |
f034b5d4e [XFRM]: Dynamic x... |
5143 |
int hashdist = HASHDIST_DEFAULT; |
1da177e4c Linux-2.6.12-rc2 |
5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 |
#ifdef CONFIG_NUMA static int __init set_hashdist(char *str) { if (!str) return 0; hashdist = simple_strtoul(str, &str, 0); return 1; } __setup("hashdist=", set_hashdist); #endif /* * allocate a large system hash table from bootmem * - it is assumed that the hash table must contain an exact power-of-2 * quantity of entries * - limit is the number of hash buckets, not the total allocation size */ void *__init alloc_large_system_hash(const char *tablename, unsigned long bucketsize, unsigned long numentries, int scale, int flags, unsigned int *_hash_shift, unsigned int *_hash_mask, unsigned long limit) { unsigned long long max = limit; unsigned long log2qty, size; void *table = NULL; /* allow the kernel cmdline to have a say */ if (!numentries) { /* round applicable memory size up to nearest megabyte */ |
049036643 [PATCH] remove HA... |
5178 |
numentries = nr_kernel_pages; |
1da177e4c Linux-2.6.12-rc2 |
5179 5180 5181 5182 5183 5184 5185 5186 5187 |
numentries += (1UL << (20 - PAGE_SHIFT)) - 1; numentries >>= 20 - PAGE_SHIFT; numentries <<= 20 - PAGE_SHIFT; /* limit to 1 bucket per 2^scale bytes of low memory */ if (scale > PAGE_SHIFT) numentries >>= (scale - PAGE_SHIFT); else numentries <<= (PAGE_SHIFT - scale); |
9ab37b8f2 [PATCH] Sanely si... |
5188 5189 |
/* Make sure we've got at least a 0-order allocation.. */ |
2c85f51d2 mm: also use allo... |
5190 5191 5192 5193 5194 5195 5196 5197 |
if (unlikely(flags & HASH_SMALL)) { /* Makes no sense without HASH_EARLY */ WARN_ON(!(flags & HASH_EARLY)); if (!(numentries >> *_hash_shift)) { numentries = 1UL << *_hash_shift; BUG_ON(!numentries); } } else if (unlikely((numentries * bucketsize) < PAGE_SIZE)) |
9ab37b8f2 [PATCH] Sanely si... |
5198 |
numentries = PAGE_SIZE / bucketsize; |
1da177e4c Linux-2.6.12-rc2 |
5199 |
} |
6e692ed37 [PATCH] fix alloc... |
5200 |
numentries = roundup_pow_of_two(numentries); |
1da177e4c Linux-2.6.12-rc2 |
5201 5202 5203 5204 5205 5206 5207 5208 5209 |
/* limit allocation size to 1/16 total memory by default */ if (max == 0) { max = ((unsigned long long)nr_all_pages << PAGE_SHIFT) >> 4; do_div(max, bucketsize); } if (numentries > max) numentries = max; |
f0d1b0b30 [PATCH] LOG2: Imp... |
5210 |
log2qty = ilog2(numentries); |
1da177e4c Linux-2.6.12-rc2 |
5211 5212 5213 5214 |
do { size = bucketsize << log2qty; if (flags & HASH_EARLY) |
74768ed83 page allocator: u... |
5215 |
table = alloc_bootmem_nopanic(size); |
1da177e4c Linux-2.6.12-rc2 |
5216 5217 5218 |
else if (hashdist) table = __vmalloc(size, GFP_ATOMIC, PAGE_KERNEL); else { |
1037b83bd MM: alloc_large_s... |
5219 5220 |
/* * If bucketsize is not a power-of-two, we may free |
a1dd268cf mm: use alloc_pag... |
5221 5222 |
* some pages at the end of hash table which * alloc_pages_exact() automatically does |
1037b83bd MM: alloc_large_s... |
5223 |
*/ |
264ef8a90 kmemleak: Remove ... |
5224 |
if (get_order(size) < MAX_ORDER) { |
a1dd268cf mm: use alloc_pag... |
5225 |
table = alloc_pages_exact(size, GFP_ATOMIC); |
264ef8a90 kmemleak: Remove ... |
5226 5227 |
kmemleak_alloc(table, size, 1, GFP_ATOMIC); } |
1da177e4c Linux-2.6.12-rc2 |
5228 5229 5230 5231 5232 5233 |
} } while (!table && size > PAGE_SIZE && --log2qty); if (!table) panic("Failed to allocate %s hash table ", tablename); |
f241e6607 mm: alloc_large_s... |
5234 5235 |
printk(KERN_INFO "%s hash table entries: %ld (order: %d, %lu bytes) ", |
1da177e4c Linux-2.6.12-rc2 |
5236 |
tablename, |
f241e6607 mm: alloc_large_s... |
5237 |
(1UL << log2qty), |
f0d1b0b30 [PATCH] LOG2: Imp... |
5238 |
ilog2(size) - PAGE_SHIFT, |
1da177e4c Linux-2.6.12-rc2 |
5239 5240 5241 5242 5243 5244 5245 5246 5247 |
size); if (_hash_shift) *_hash_shift = log2qty; if (_hash_mask) *_hash_mask = (1 << log2qty) - 1; return table; } |
a117e66ed [PATCH] unify pfn... |
5248 |
|
835c134ec Add a bitmap that... |
5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 |
/* Return a pointer to the bitmap storing bits affecting a block of pages */ static inline unsigned long *get_pageblock_bitmap(struct zone *zone, unsigned long pfn) { #ifdef CONFIG_SPARSEMEM return __pfn_to_section(pfn)->pageblock_flags; #else return zone->pageblock_flags; #endif /* CONFIG_SPARSEMEM */ } static inline int pfn_to_bitidx(struct zone *zone, unsigned long pfn) { #ifdef CONFIG_SPARSEMEM pfn &= (PAGES_PER_SECTION-1); |
d9c234005 Do not depend on ... |
5264 |
return (pfn >> pageblock_order) * NR_PAGEBLOCK_BITS; |
835c134ec Add a bitmap that... |
5265 5266 |
#else pfn = pfn - zone->zone_start_pfn; |
d9c234005 Do not depend on ... |
5267 |
return (pfn >> pageblock_order) * NR_PAGEBLOCK_BITS; |
835c134ec Add a bitmap that... |
5268 5269 5270 5271 |
#endif /* CONFIG_SPARSEMEM */ } /** |
d9c234005 Do not depend on ... |
5272 |
* get_pageblock_flags_group - Return the requested group of flags for the pageblock_nr_pages block of pages |
835c134ec Add a bitmap that... |
5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 |
* @page: The page within the block of interest * @start_bitidx: The first bit of interest to retrieve * @end_bitidx: The last bit of interest * returns pageblock_bits flags */ unsigned long get_pageblock_flags_group(struct page *page, int start_bitidx, int end_bitidx) { struct zone *zone; unsigned long *bitmap; unsigned long pfn, bitidx; unsigned long flags = 0; unsigned long value = 1; zone = page_zone(page); pfn = page_to_pfn(page); bitmap = get_pageblock_bitmap(zone, pfn); bitidx = pfn_to_bitidx(zone, pfn); for (; start_bitidx <= end_bitidx; start_bitidx++, value <<= 1) if (test_bit(bitidx + start_bitidx, bitmap)) flags |= value; |
6220ec784 [PATCH] highest_p... |
5295 |
|
835c134ec Add a bitmap that... |
5296 5297 5298 5299 |
return flags; } /** |
d9c234005 Do not depend on ... |
5300 |
* set_pageblock_flags_group - Set the requested group of flags for a pageblock_nr_pages block of pages |
835c134ec Add a bitmap that... |
5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 |
* @page: The page within the block of interest * @start_bitidx: The first bit of interest * @end_bitidx: The last bit of interest * @flags: The flags to set */ void set_pageblock_flags_group(struct page *page, unsigned long flags, int start_bitidx, int end_bitidx) { struct zone *zone; unsigned long *bitmap; unsigned long pfn, bitidx; unsigned long value = 1; zone = page_zone(page); pfn = page_to_pfn(page); bitmap = get_pageblock_bitmap(zone, pfn); bitidx = pfn_to_bitidx(zone, pfn); |
86051ca5e mm: fix usemap in... |
5318 5319 |
VM_BUG_ON(pfn < zone->zone_start_pfn); VM_BUG_ON(pfn >= zone->zone_start_pfn + zone->spanned_pages); |
835c134ec Add a bitmap that... |
5320 5321 5322 5323 5324 5325 5326 |
for (; start_bitidx <= end_bitidx; start_bitidx++, value <<= 1) if (flags & value) __set_bit(bitidx + start_bitidx, bitmap); else __clear_bit(bitidx + start_bitidx, bitmap); } |
a5d76b54a memory unplug: pa... |
5327 5328 5329 5330 5331 5332 |
/* * This is designed as sub function...plz see page_isolation.c also. * set/clear page block's type to be ISOLATE. * page allocater never alloc memory from ISOLATE block. */ |
49ac82558 memory hotplug: u... |
5333 5334 5335 5336 5337 5338 5339 5340 5341 5342 5343 5344 5345 5346 5347 5348 5349 |
static int __count_immobile_pages(struct zone *zone, struct page *page, int count) { unsigned long pfn, iter, found; /* * For avoiding noise data, lru_add_drain_all() should be called * If ZONE_MOVABLE, the zone never contains immobile pages */ if (zone_idx(zone) == ZONE_MOVABLE) return true; if (get_pageblock_migratetype(page) == MIGRATE_MOVABLE) return true; pfn = page_to_pfn(page); for (found = 0, iter = 0; iter < pageblock_nr_pages; iter++) { unsigned long check = pfn + iter; |
29723fccc mm: fix dubious c... |
5350 |
if (!pfn_valid_within(check)) |
49ac82558 memory hotplug: u... |
5351 |
continue; |
29723fccc mm: fix dubious c... |
5352 |
|
49ac82558 memory hotplug: u... |
5353 5354 5355 5356 5357 5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 |
page = pfn_to_page(check); if (!page_count(page)) { if (PageBuddy(page)) iter += (1 << page_order(page)) - 1; continue; } if (!PageLRU(page)) found++; /* * If there are RECLAIMABLE pages, we need to check it. * But now, memory offline itself doesn't call shrink_slab() * and it still to be fixed. */ /* * If the page is not RAM, page_count()should be 0. * we don't need more check. This is an _used_ not-movable page. * * The problematic thing here is PG_reserved pages. PG_reserved * is set to both of a memory hole page and a _used_ kernel * page at boot. */ if (found > count) return false; } return true; } bool is_pageblock_removable_nolock(struct page *page) { struct zone *zone = page_zone(page); return __count_immobile_pages(zone, page, 0); } |
a5d76b54a memory unplug: pa... |
5385 5386 5387 |
int set_migratetype_isolate(struct page *page) { struct zone *zone; |
49ac82558 memory hotplug: u... |
5388 |
unsigned long flags, pfn; |
925cc71e5 mm: Add notifier ... |
5389 5390 |
struct memory_isolate_notify arg; int notifier_ret; |
a5d76b54a memory unplug: pa... |
5391 5392 5393 |
int ret = -EBUSY; zone = page_zone(page); |
925cc71e5 mm: Add notifier ... |
5394 |
|
a5d76b54a memory unplug: pa... |
5395 |
spin_lock_irqsave(&zone->lock, flags); |
925cc71e5 mm: Add notifier ... |
5396 5397 5398 5399 5400 |
pfn = page_to_pfn(page); arg.start_pfn = pfn; arg.nr_pages = pageblock_nr_pages; arg.pages_found = 0; |
a5d76b54a memory unplug: pa... |
5401 |
/* |
925cc71e5 mm: Add notifier ... |
5402 5403 5404 5405 5406 5407 5408 5409 5410 |
* It may be possible to isolate a pageblock even if the * migratetype is not MIGRATE_MOVABLE. The memory isolation * notifier chain is used by balloon drivers to return the * number of pages in a range that are held by the balloon * driver to shrink memory. If all the pages are accounted for * by balloons, are free, or on the LRU, isolation can continue. * Later, for example, when memory hotplug notifier runs, these * pages reported as "can be isolated" should be isolated(freed) * by the balloon driver through the memory notifier chain. |
a5d76b54a memory unplug: pa... |
5411 |
*/ |
925cc71e5 mm: Add notifier ... |
5412 5413 |
notifier_ret = memory_isolate_notify(MEM_ISOLATE_COUNT, &arg); notifier_ret = notifier_to_errno(notifier_ret); |
4b20477f5 memory hotplug: f... |
5414 |
if (notifier_ret) |
a5d76b54a memory unplug: pa... |
5415 |
goto out; |
49ac82558 memory hotplug: u... |
5416 5417 5418 5419 5420 |
/* * FIXME: Now, memory hotplug doesn't call shrink_slab() by itself. * We just check MOVABLE pages. */ if (__count_immobile_pages(zone, page, arg.pages_found)) |
925cc71e5 mm: Add notifier ... |
5421 |
ret = 0; |
49ac82558 memory hotplug: u... |
5422 5423 5424 5425 |
/* * immobile means "not-on-lru" paes. If immobile is larger than * removable-by-driver pages reported by notifier, we'll fail. */ |
a5d76b54a memory unplug: pa... |
5426 |
out: |
925cc71e5 mm: Add notifier ... |
5427 5428 5429 5430 |
if (!ret) { set_pageblock_migratetype(page, MIGRATE_ISOLATE); move_freepages_block(zone, page, MIGRATE_ISOLATE); } |
a5d76b54a memory unplug: pa... |
5431 5432 |
spin_unlock_irqrestore(&zone->lock, flags); if (!ret) |
9f8f21725 Page allocator: c... |
5433 |
drain_all_pages(); |
a5d76b54a memory unplug: pa... |
5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449 |
return ret; } void unset_migratetype_isolate(struct page *page) { struct zone *zone; unsigned long flags; zone = page_zone(page); spin_lock_irqsave(&zone->lock, flags); if (get_pageblock_migratetype(page) != MIGRATE_ISOLATE) goto out; set_pageblock_migratetype(page, MIGRATE_MOVABLE); move_freepages_block(zone, page, MIGRATE_MOVABLE); out: spin_unlock_irqrestore(&zone->lock, flags); } |
0c0e61958 memory unplug: pa... |
5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474 5475 5476 5477 5478 5479 5480 5481 5482 5483 5484 5485 5486 5487 5488 5489 5490 5491 5492 5493 5494 5495 5496 5497 |
#ifdef CONFIG_MEMORY_HOTREMOVE /* * All pages in the range must be isolated before calling this. */ void __offline_isolated_pages(unsigned long start_pfn, unsigned long end_pfn) { struct page *page; struct zone *zone; int order, i; unsigned long pfn; unsigned long flags; /* find the first valid pfn */ for (pfn = start_pfn; pfn < end_pfn; pfn++) if (pfn_valid(pfn)) break; if (pfn == end_pfn) return; zone = page_zone(pfn_to_page(pfn)); spin_lock_irqsave(&zone->lock, flags); pfn = start_pfn; while (pfn < end_pfn) { if (!pfn_valid(pfn)) { pfn++; continue; } page = pfn_to_page(pfn); BUG_ON(page_count(page)); BUG_ON(!PageBuddy(page)); order = page_order(page); #ifdef CONFIG_DEBUG_VM printk(KERN_INFO "remove from free list %lx %d %lx ", pfn, 1 << order, end_pfn); #endif list_del(&page->lru); rmv_page_order(page); zone->free_area[order].nr_free--; __mod_zone_page_state(zone, NR_FREE_PAGES, - (1UL << order)); for (i = 0; i < (1 << order); i++) SetPageReserved((page+i)); pfn += (1 << order); } spin_unlock_irqrestore(&zone->lock, flags); } #endif |
8d22ba1b7 HWPOISON: detect ... |
5498 5499 5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 |
#ifdef CONFIG_MEMORY_FAILURE bool is_free_buddy_page(struct page *page) { struct zone *zone = page_zone(page); unsigned long pfn = page_to_pfn(page); unsigned long flags; int order; spin_lock_irqsave(&zone->lock, flags); for (order = 0; order < MAX_ORDER; order++) { struct page *page_head = page - (pfn & ((1 << order) - 1)); if (PageBuddy(page_head) && page_order(page_head) >= order) break; } spin_unlock_irqrestore(&zone->lock, flags); return order < MAX_ORDER; } #endif |
718a38211 mm: introduce dum... |
5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541 5542 5543 |
static struct trace_print_flags pageflag_names[] = { {1UL << PG_locked, "locked" }, {1UL << PG_error, "error" }, {1UL << PG_referenced, "referenced" }, {1UL << PG_uptodate, "uptodate" }, {1UL << PG_dirty, "dirty" }, {1UL << PG_lru, "lru" }, {1UL << PG_active, "active" }, {1UL << PG_slab, "slab" }, {1UL << PG_owner_priv_1, "owner_priv_1" }, {1UL << PG_arch_1, "arch_1" }, {1UL << PG_reserved, "reserved" }, {1UL << PG_private, "private" }, {1UL << PG_private_2, "private_2" }, {1UL << PG_writeback, "writeback" }, #ifdef CONFIG_PAGEFLAGS_EXTENDED {1UL << PG_head, "head" }, {1UL << PG_tail, "tail" }, #else {1UL << PG_compound, "compound" }, #endif {1UL << PG_swapcache, "swapcache" }, {1UL << PG_mappedtodisk, "mappedtodisk" }, {1UL << PG_reclaim, "reclaim" }, |
718a38211 mm: introduce dum... |
5544 5545 5546 5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557 5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 5568 5569 5570 5571 5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 |
{1UL << PG_swapbacked, "swapbacked" }, {1UL << PG_unevictable, "unevictable" }, #ifdef CONFIG_MMU {1UL << PG_mlocked, "mlocked" }, #endif #ifdef CONFIG_ARCH_USES_PG_UNCACHED {1UL << PG_uncached, "uncached" }, #endif #ifdef CONFIG_MEMORY_FAILURE {1UL << PG_hwpoison, "hwpoison" }, #endif {-1UL, NULL }, }; static void dump_page_flags(unsigned long flags) { const char *delim = ""; unsigned long mask; int i; printk(KERN_ALERT "page flags: %#lx(", flags); /* remove zone id */ flags &= (1UL << NR_PAGEFLAGS) - 1; for (i = 0; pageflag_names[i].name && flags; i++) { mask = pageflag_names[i].mask; if ((flags & mask) != mask) continue; flags &= ~mask; printk("%s%s", delim, pageflag_names[i].name); delim = "|"; } /* check for left over flags */ if (flags) printk("%s%#lx", delim, flags); printk(") "); } void dump_page(struct page *page) { printk(KERN_ALERT "page:%p count:%d mapcount:%d mapping:%p index:%#lx ", |
4e9f64c42 thp: fix bad_page... |
5593 |
page, atomic_read(&page->_count), page_mapcount(page), |
718a38211 mm: introduce dum... |
5594 5595 |
page->mapping, page->index); dump_page_flags(page->flags); |
f212ad7cf memcg: add memcg ... |
5596 |
mem_cgroup_print_bad_page(page); |
718a38211 mm: introduce dum... |
5597 |
} |