Blame view
mm/mempolicy.c
59.3 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 |
/* * Simple NUMA memory policy for the Linux kernel. * * Copyright 2003,2004 Andi Kleen, SuSE Labs. |
8bccd85ff [PATCH] Implement... |
5 |
* (C) Copyright 2005 Christoph Lameter, Silicon Graphics, Inc. |
1da177e4c Linux-2.6.12-rc2 |
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
* Subject to the GNU Public License, version 2. * * NUMA policy allows the user to give hints in which node(s) memory should * be allocated. * * Support four policies per VMA and per process: * * The VMA policy has priority over the process policy for a page fault. * * interleave Allocate memory interleaved over a set of nodes, * with normal fallback if it fails. * For VMA based allocations this interleaves based on the * offset into the backing object or offset into the mapping * for anonymous memory. For process policy an process counter * is used. |
8bccd85ff [PATCH] Implement... |
21 |
* |
1da177e4c Linux-2.6.12-rc2 |
22 23 |
* bind Only allocate memory on a specific set of nodes, * no fallback. |
8bccd85ff [PATCH] Implement... |
24 25 26 27 |
* FIXME: memory is allocated starting with the first node * to the last. It would be better if bind would truly restrict * the allocation to memory nodes instead * |
1da177e4c Linux-2.6.12-rc2 |
28 29 30 31 32 |
* preferred Try a specific node first before normal fallback. * As a special case node -1 here means do the allocation * on the local CPU. This is normally identical to default, * but useful to set in a VMA when you have a non default * process policy. |
8bccd85ff [PATCH] Implement... |
33 |
* |
1da177e4c Linux-2.6.12-rc2 |
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
* default Allocate on the local node first, or when on a VMA * use the process policy. This is what Linux always did * in a NUMA aware kernel and still does by, ahem, default. * * The process policy is applied for most non interrupt memory allocations * in that process' context. Interrupts ignore the policies and always * try to allocate on the local CPU. The VMA policy is only applied for memory * allocations for a VMA in the VM. * * Currently there are a few corner cases in swapping where the policy * is not applied, but the majority should be handled. When process policy * is used it is not remembered over swap outs/swap ins. * * Only the highest zone in the zone hierarchy gets policied. Allocations * requesting a lower zone just use default policy. This implies that * on systems with highmem kernel lowmem allocation don't get policied. * Same with GFP_DMA allocations. * * For shmfs/tmpfs/hugetlbfs shared memory the policy is shared between * all users and remembered even when nobody has memory mapped. */ /* Notebook: fix mmap readahead to honour policy and enable policy for any page cache object statistics for bigpages global policy for page cache? currently it uses process policy. Requires first item above. handle mremap for shared memory (currently ignored for the policy) grows down? make bind policy root only? It can trigger oom much faster and the kernel is not always grateful with that. |
1da177e4c Linux-2.6.12-rc2 |
66 67 68 69 70 71 72 73 |
*/ #include <linux/mempolicy.h> #include <linux/mm.h> #include <linux/highmem.h> #include <linux/hugetlb.h> #include <linux/kernel.h> #include <linux/sched.h> |
1da177e4c Linux-2.6.12-rc2 |
74 75 76 77 78 79 |
#include <linux/nodemask.h> #include <linux/cpuset.h> #include <linux/gfp.h> #include <linux/slab.h> #include <linux/string.h> #include <linux/module.h> |
b488893a3 pid namespaces: c... |
80 |
#include <linux/nsproxy.h> |
1da177e4c Linux-2.6.12-rc2 |
81 82 83 |
#include <linux/interrupt.h> #include <linux/init.h> #include <linux/compat.h> |
dc9aa5b9d [PATCH] Swap Migr... |
84 |
#include <linux/swap.h> |
1a75a6c82 [PATCH] Fold numa... |
85 86 |
#include <linux/seq_file.h> #include <linux/proc_fs.h> |
b20a35035 [PATCH] page migr... |
87 |
#include <linux/migrate.h> |
95a402c38 [PATCH] page migr... |
88 |
#include <linux/rmap.h> |
86c3a7645 [PATCH] SELinux: ... |
89 |
#include <linux/security.h> |
dbcb0f19c mm/mempolicy.c: c... |
90 |
#include <linux/syscalls.h> |
095f1fc4e mempolicy: rework... |
91 |
#include <linux/ctype.h> |
dc9aa5b9d [PATCH] Swap Migr... |
92 |
|
1da177e4c Linux-2.6.12-rc2 |
93 94 |
#include <asm/tlbflush.h> #include <asm/uaccess.h> |
62695a84e vmscan: move isol... |
95 |
#include "internal.h" |
38e35860d [PATCH] mempolici... |
96 |
/* Internal flags */ |
dc9aa5b9d [PATCH] Swap Migr... |
97 |
#define MPOL_MF_DISCONTIG_OK (MPOL_MF_INTERNAL << 0) /* Skip checks for continuous vmas */ |
38e35860d [PATCH] mempolici... |
98 |
#define MPOL_MF_INVERT (MPOL_MF_INTERNAL << 1) /* Invert check for nodemask */ |
1a75a6c82 [PATCH] Fold numa... |
99 |
#define MPOL_MF_STATS (MPOL_MF_INTERNAL << 2) /* Gather statistics */ |
dc9aa5b9d [PATCH] Swap Migr... |
100 |
|
fcc234f88 [PATCH] mm: kill ... |
101 102 |
static struct kmem_cache *policy_cache; static struct kmem_cache *sn_cache; |
1da177e4c Linux-2.6.12-rc2 |
103 |
|
1da177e4c Linux-2.6.12-rc2 |
104 105 |
/* Highest zone. An specific allocation for a zone below that is not policied. */ |
6267276f3 [PATCH] optional ... |
106 |
enum zone_type policy_zone = 0; |
1da177e4c Linux-2.6.12-rc2 |
107 |
|
bea904d54 mempolicy: use MP... |
108 109 110 |
/* * run-time system-wide default policy => local allocation */ |
d42c69972 [PATCH] PCI: Run ... |
111 |
struct mempolicy default_policy = { |
1da177e4c Linux-2.6.12-rc2 |
112 |
.refcnt = ATOMIC_INIT(1), /* never free it */ |
bea904d54 mempolicy: use MP... |
113 |
.mode = MPOL_PREFERRED, |
fc36b8d3d mempolicy: use MP... |
114 |
.flags = MPOL_F_LOCAL, |
1da177e4c Linux-2.6.12-rc2 |
115 |
}; |
37012946d mempolicy: create... |
116 117 118 119 |
static const struct mempolicy_operations { int (*create)(struct mempolicy *pol, const nodemask_t *nodes); void (*rebind)(struct mempolicy *pol, const nodemask_t *nodes); } mpol_ops[MPOL_MAX]; |
19770b326 mm: filter based ... |
120 |
/* Check that the nodemask contains at least one populated zone */ |
37012946d mempolicy: create... |
121 |
static int is_valid_nodemask(const nodemask_t *nodemask) |
1da177e4c Linux-2.6.12-rc2 |
122 |
{ |
19770b326 mm: filter based ... |
123 |
int nd, k; |
1da177e4c Linux-2.6.12-rc2 |
124 |
|
19770b326 mm: filter based ... |
125 126 127 128 129 130 131 132 133 134 |
/* Check that there is something useful in this mask */ k = policy_zone; for_each_node_mask(nd, *nodemask) { struct zone *z; for (k = 0; k <= policy_zone; k++) { z = &NODE_DATA(nd)->node_zones[k]; if (z->present_pages > 0) return 1; |
dd942ae33 [PATCH] Handle al... |
135 |
} |
8af5e2eb3 [PATCH] fix mempo... |
136 |
} |
19770b326 mm: filter based ... |
137 138 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
139 |
} |
f5b087b52 mempolicy: add MP... |
140 141 |
static inline int mpol_store_user_nodemask(const struct mempolicy *pol) { |
4c50bc011 mempolicy: add MP... |
142 143 144 145 146 147 148 149 150 |
return pol->flags & (MPOL_F_STATIC_NODES | MPOL_F_RELATIVE_NODES); } static void mpol_relative_nodemask(nodemask_t *ret, const nodemask_t *orig, const nodemask_t *rel) { nodemask_t tmp; nodes_fold(tmp, *orig, nodes_weight(*rel)); nodes_onto(*ret, tmp, *rel); |
f5b087b52 mempolicy: add MP... |
151 |
} |
37012946d mempolicy: create... |
152 153 154 155 156 157 158 159 160 161 162 |
static int mpol_new_interleave(struct mempolicy *pol, const nodemask_t *nodes) { if (nodes_empty(*nodes)) return -EINVAL; pol->v.nodes = *nodes; return 0; } static int mpol_new_preferred(struct mempolicy *pol, const nodemask_t *nodes) { if (!nodes) |
fc36b8d3d mempolicy: use MP... |
163 |
pol->flags |= MPOL_F_LOCAL; /* local allocation */ |
37012946d mempolicy: create... |
164 165 166 167 168 169 170 171 172 173 174 175 176 177 |
else if (nodes_empty(*nodes)) return -EINVAL; /* no allowed nodes */ else pol->v.preferred_node = first_node(*nodes); return 0; } static int mpol_new_bind(struct mempolicy *pol, const nodemask_t *nodes) { if (!is_valid_nodemask(nodes)) return -EINVAL; pol->v.nodes = *nodes; return 0; } |
1da177e4c Linux-2.6.12-rc2 |
178 |
/* Create a new policy */ |
028fec414 mempolicy: suppor... |
179 180 |
static struct mempolicy *mpol_new(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
181 182 |
{ struct mempolicy *policy; |
f5b087b52 mempolicy: add MP... |
183 |
nodemask_t cpuset_context_nmask; |
37012946d mempolicy: create... |
184 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
185 |
|
028fec414 mempolicy: suppor... |
186 187 188 |
pr_debug("setting mode %d flags %d nodes[0] %lx ", mode, flags, nodes ? nodes_addr(*nodes)[0] : -1); |
140d5a490 numa: mempolicy: ... |
189 |
|
3e1f06456 mempolicy: disall... |
190 191 |
if (mode == MPOL_DEFAULT) { if (nodes && !nodes_empty(*nodes)) |
37012946d mempolicy: create... |
192 |
return ERR_PTR(-EINVAL); |
bea904d54 mempolicy: use MP... |
193 |
return NULL; /* simply delete any existing policy */ |
37012946d mempolicy: create... |
194 |
} |
3e1f06456 mempolicy: disall... |
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 |
VM_BUG_ON(!nodes); /* * MPOL_PREFERRED cannot be used with MPOL_F_STATIC_NODES or * MPOL_F_RELATIVE_NODES if the nodemask is empty (local allocation). * All other modes require a valid pointer to a non-empty nodemask. */ if (mode == MPOL_PREFERRED) { if (nodes_empty(*nodes)) { if (((flags & MPOL_F_STATIC_NODES) || (flags & MPOL_F_RELATIVE_NODES))) return ERR_PTR(-EINVAL); nodes = NULL; /* flag local alloc */ } } else if (nodes_empty(*nodes)) return ERR_PTR(-EINVAL); |
1da177e4c Linux-2.6.12-rc2 |
211 212 213 214 |
policy = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!policy) return ERR_PTR(-ENOMEM); atomic_set(&policy->refcnt, 1); |
45c4745af mempolicy: rename... |
215 |
policy->mode = mode; |
3e1f06456 mempolicy: disall... |
216 |
policy->flags = flags; |
37012946d mempolicy: create... |
217 |
|
3e1f06456 mempolicy: disall... |
218 219 220 221 |
if (nodes) { /* * cpuset related setup doesn't apply to local allocation */ |
37012946d mempolicy: create... |
222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 |
cpuset_update_task_memory_state(); if (flags & MPOL_F_RELATIVE_NODES) mpol_relative_nodemask(&cpuset_context_nmask, nodes, &cpuset_current_mems_allowed); else nodes_and(cpuset_context_nmask, *nodes, cpuset_current_mems_allowed); if (mpol_store_user_nodemask(policy)) policy->w.user_nodemask = *nodes; else policy->w.cpuset_mems_allowed = cpuset_mems_allowed(current); } ret = mpol_ops[mode].create(policy, |
3e1f06456 mempolicy: disall... |
237 |
nodes ? &cpuset_context_nmask : NULL); |
37012946d mempolicy: create... |
238 239 240 241 |
if (ret < 0) { kmem_cache_free(policy_cache, policy); return ERR_PTR(ret); } |
1da177e4c Linux-2.6.12-rc2 |
242 |
return policy; |
37012946d mempolicy: create... |
243 |
} |
52cd3b074 mempolicy: rework... |
244 245 246 247 248 |
/* Slow path of a mpol destructor. */ void __mpol_put(struct mempolicy *p) { if (!atomic_dec_and_test(&p->refcnt)) return; |
52cd3b074 mempolicy: rework... |
249 250 |
kmem_cache_free(policy_cache, p); } |
37012946d mempolicy: create... |
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 |
static void mpol_rebind_default(struct mempolicy *pol, const nodemask_t *nodes) { } static void mpol_rebind_nodemask(struct mempolicy *pol, const nodemask_t *nodes) { nodemask_t tmp; if (pol->flags & MPOL_F_STATIC_NODES) nodes_and(tmp, pol->w.user_nodemask, *nodes); else if (pol->flags & MPOL_F_RELATIVE_NODES) mpol_relative_nodemask(&tmp, &pol->w.user_nodemask, nodes); else { nodes_remap(tmp, pol->v.nodes, pol->w.cpuset_mems_allowed, *nodes); pol->w.cpuset_mems_allowed = *nodes; } |
f5b087b52 mempolicy: add MP... |
269 |
|
37012946d mempolicy: create... |
270 271 272 273 274 275 276 277 278 279 280 281 282 283 |
pol->v.nodes = tmp; if (!node_isset(current->il_next, tmp)) { current->il_next = next_node(current->il_next, tmp); if (current->il_next >= MAX_NUMNODES) current->il_next = first_node(tmp); if (current->il_next >= MAX_NUMNODES) current->il_next = numa_node_id(); } } static void mpol_rebind_preferred(struct mempolicy *pol, const nodemask_t *nodes) { nodemask_t tmp; |
37012946d mempolicy: create... |
284 285 |
if (pol->flags & MPOL_F_STATIC_NODES) { int node = first_node(pol->w.user_nodemask); |
fc36b8d3d mempolicy: use MP... |
286 |
if (node_isset(node, *nodes)) { |
37012946d mempolicy: create... |
287 |
pol->v.preferred_node = node; |
fc36b8d3d mempolicy: use MP... |
288 289 290 |
pol->flags &= ~MPOL_F_LOCAL; } else pol->flags |= MPOL_F_LOCAL; |
37012946d mempolicy: create... |
291 292 293 |
} else if (pol->flags & MPOL_F_RELATIVE_NODES) { mpol_relative_nodemask(&tmp, &pol->w.user_nodemask, nodes); pol->v.preferred_node = first_node(tmp); |
fc36b8d3d mempolicy: use MP... |
294 |
} else if (!(pol->flags & MPOL_F_LOCAL)) { |
37012946d mempolicy: create... |
295 296 297 298 299 |
pol->v.preferred_node = node_remap(pol->v.preferred_node, pol->w.cpuset_mems_allowed, *nodes); pol->w.cpuset_mems_allowed = *nodes; } |
1da177e4c Linux-2.6.12-rc2 |
300 |
} |
1d0d2680a mempolicy: move r... |
301 302 303 304 |
/* Migrate a policy to a different set of nodes */ static void mpol_rebind_policy(struct mempolicy *pol, const nodemask_t *newmask) { |
1d0d2680a mempolicy: move r... |
305 306 |
if (!pol) return; |
1d0d2680a mempolicy: move r... |
307 308 309 |
if (!mpol_store_user_nodemask(pol) && nodes_equal(pol->w.cpuset_mems_allowed, *newmask)) return; |
45c4745af mempolicy: rename... |
310 |
mpol_ops[pol->mode].rebind(pol, newmask); |
1d0d2680a mempolicy: move r... |
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 |
} /* * Wrapper for mpol_rebind_policy() that just requires task * pointer, and updates task mempolicy. */ void mpol_rebind_task(struct task_struct *tsk, const nodemask_t *new) { mpol_rebind_policy(tsk->mempolicy, new); } /* * Rebind each vma in mm to new nodemask. * * Call holding a reference to mm. Takes mm->mmap_sem during call. */ void mpol_rebind_mm(struct mm_struct *mm, nodemask_t *new) { struct vm_area_struct *vma; down_write(&mm->mmap_sem); for (vma = mm->mmap; vma; vma = vma->vm_next) mpol_rebind_policy(vma->vm_policy, new); up_write(&mm->mmap_sem); } |
37012946d mempolicy: create... |
338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 |
static const struct mempolicy_operations mpol_ops[MPOL_MAX] = { [MPOL_DEFAULT] = { .rebind = mpol_rebind_default, }, [MPOL_INTERLEAVE] = { .create = mpol_new_interleave, .rebind = mpol_rebind_nodemask, }, [MPOL_PREFERRED] = { .create = mpol_new_preferred, .rebind = mpol_rebind_preferred, }, [MPOL_BIND] = { .create = mpol_new_bind, .rebind = mpol_rebind_nodemask, }, }; |
397874dfe [PATCH] numa_maps... |
355 |
static void gather_stats(struct page *, void *, int pte_dirty); |
fc3012896 [PATCH] Simplify ... |
356 357 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags); |
1a75a6c82 [PATCH] Fold numa... |
358 |
|
38e35860d [PATCH] mempolici... |
359 |
/* Scan through pages checking if pages follow certain conditions. */ |
b5810039a [PATCH] core remo... |
360 |
static int check_pte_range(struct vm_area_struct *vma, pmd_t *pmd, |
dc9aa5b9d [PATCH] Swap Migr... |
361 362 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
363 |
void *private) |
1da177e4c Linux-2.6.12-rc2 |
364 |
{ |
91612e0df [PATCH] mbind: ch... |
365 366 |
pte_t *orig_pte; pte_t *pte; |
705e87c0c [PATCH] mm: pte_o... |
367 |
spinlock_t *ptl; |
941150a32 [PATCH] mbind: fi... |
368 |
|
705e87c0c [PATCH] mm: pte_o... |
369 |
orig_pte = pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); |
91612e0df [PATCH] mbind: ch... |
370 |
do { |
6aab341e0 mm: re-architect ... |
371 |
struct page *page; |
25ba77c14 [PATCH] numa node... |
372 |
int nid; |
91612e0df [PATCH] mbind: ch... |
373 374 |
if (!pte_present(*pte)) |
1da177e4c Linux-2.6.12-rc2 |
375 |
continue; |
6aab341e0 mm: re-architect ... |
376 377 |
page = vm_normal_page(vma, addr, *pte); if (!page) |
1da177e4c Linux-2.6.12-rc2 |
378 |
continue; |
053837fce [PATCH] mm: migra... |
379 380 381 382 383 384 385 386 387 388 389 |
/* * The check for PageReserved here is important to avoid * handling zero pages and other pages that may have been * marked special by the system. * * If the PageReserved would not be checked here then f.e. * the location of the zero page could have an influence * on MPOL_MF_STRICT, zero pages would be counted for * the per node stats, and there would be useless attempts * to put zero pages on the migration list. */ |
f4598c8b3 [PATCH] migration... |
390 391 |
if (PageReserved(page)) continue; |
6aab341e0 mm: re-architect ... |
392 |
nid = page_to_nid(page); |
38e35860d [PATCH] mempolici... |
393 394 |
if (node_isset(nid, *nodes) == !!(flags & MPOL_MF_INVERT)) continue; |
1a75a6c82 [PATCH] Fold numa... |
395 |
if (flags & MPOL_MF_STATS) |
397874dfe [PATCH] numa_maps... |
396 |
gather_stats(page, private, pte_dirty(*pte)); |
053837fce [PATCH] mm: migra... |
397 |
else if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) |
fc3012896 [PATCH] Simplify ... |
398 |
migrate_page_add(page, private, flags); |
38e35860d [PATCH] mempolici... |
399 400 |
else break; |
91612e0df [PATCH] mbind: ch... |
401 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
705e87c0c [PATCH] mm: pte_o... |
402 |
pte_unmap_unlock(orig_pte, ptl); |
91612e0df [PATCH] mbind: ch... |
403 404 |
return addr != end; } |
b5810039a [PATCH] core remo... |
405 |
static inline int check_pmd_range(struct vm_area_struct *vma, pud_t *pud, |
dc9aa5b9d [PATCH] Swap Migr... |
406 407 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
408 |
void *private) |
91612e0df [PATCH] mbind: ch... |
409 410 411 412 413 414 415 416 417 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; |
dc9aa5b9d [PATCH] Swap Migr... |
418 |
if (check_pte_range(vma, pmd, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
419 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
420 421 422 423 |
return -EIO; } while (pmd++, addr = next, addr != end); return 0; } |
b5810039a [PATCH] core remo... |
424 |
static inline int check_pud_range(struct vm_area_struct *vma, pgd_t *pgd, |
dc9aa5b9d [PATCH] Swap Migr... |
425 426 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
427 |
void *private) |
91612e0df [PATCH] mbind: ch... |
428 429 430 431 432 433 434 435 436 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
dc9aa5b9d [PATCH] Swap Migr... |
437 |
if (check_pmd_range(vma, pud, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
438 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
439 440 441 442 |
return -EIO; } while (pud++, addr = next, addr != end); return 0; } |
b5810039a [PATCH] core remo... |
443 |
static inline int check_pgd_range(struct vm_area_struct *vma, |
dc9aa5b9d [PATCH] Swap Migr... |
444 445 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
446 |
void *private) |
91612e0df [PATCH] mbind: ch... |
447 448 449 |
{ pgd_t *pgd; unsigned long next; |
b5810039a [PATCH] core remo... |
450 |
pgd = pgd_offset(vma->vm_mm, addr); |
91612e0df [PATCH] mbind: ch... |
451 452 453 454 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
dc9aa5b9d [PATCH] Swap Migr... |
455 |
if (check_pud_range(vma, pgd, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
456 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
457 458 459 |
return -EIO; } while (pgd++, addr = next, addr != end); return 0; |
1da177e4c Linux-2.6.12-rc2 |
460 |
} |
dc9aa5b9d [PATCH] Swap Migr... |
461 462 463 464 465 |
/* * Check if all pages in a range are on a set of nodes. * If pagelist != NULL then isolate pages from the LRU and * put them on the pagelist. */ |
1da177e4c Linux-2.6.12-rc2 |
466 467 |
static struct vm_area_struct * check_range(struct mm_struct *mm, unsigned long start, unsigned long end, |
38e35860d [PATCH] mempolici... |
468 |
const nodemask_t *nodes, unsigned long flags, void *private) |
1da177e4c Linux-2.6.12-rc2 |
469 470 471 |
{ int err; struct vm_area_struct *first, *vma, *prev; |
053837fce [PATCH] mm: migra... |
472 |
|
1da177e4c Linux-2.6.12-rc2 |
473 474 475 476 477 |
first = find_vma(mm, start); if (!first) return ERR_PTR(-EFAULT); prev = NULL; for (vma = first; vma && vma->vm_start < end; vma = vma->vm_next) { |
dc9aa5b9d [PATCH] Swap Migr... |
478 479 480 481 482 483 484 485 486 487 |
if (!(flags & MPOL_MF_DISCONTIG_OK)) { if (!vma->vm_next && vma->vm_end < end) return ERR_PTR(-EFAULT); if (prev && prev->vm_end < vma->vm_start) return ERR_PTR(-EFAULT); } if (!is_vm_hugetlb_page(vma) && ((flags & MPOL_MF_STRICT) || ((flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) && vma_migratable(vma)))) { |
5b952b3c1 [PATCH] Fix MPOL_... |
488 |
unsigned long endvma = vma->vm_end; |
dc9aa5b9d [PATCH] Swap Migr... |
489 |
|
5b952b3c1 [PATCH] Fix MPOL_... |
490 491 492 493 |
if (endvma > end) endvma = end; if (vma->vm_start > start) start = vma->vm_start; |
dc9aa5b9d [PATCH] Swap Migr... |
494 |
err = check_pgd_range(vma, start, endvma, nodes, |
38e35860d [PATCH] mempolici... |
495 |
flags, private); |
1da177e4c Linux-2.6.12-rc2 |
496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 |
if (err) { first = ERR_PTR(err); break; } } prev = vma; } return first; } /* Apply policy to a single VMA */ static int policy_vma(struct vm_area_struct *vma, struct mempolicy *new) { int err = 0; struct mempolicy *old = vma->vm_policy; |
140d5a490 numa: mempolicy: ... |
511 512 |
pr_debug("vma %lx-%lx/%lx vm_ops %p vm_file %p set_policy %p ", |
1da177e4c Linux-2.6.12-rc2 |
513 514 515 516 517 518 519 520 521 |
vma->vm_start, vma->vm_end, vma->vm_pgoff, vma->vm_ops, vma->vm_file, vma->vm_ops ? vma->vm_ops->set_policy : NULL); if (vma->vm_ops && vma->vm_ops->set_policy) err = vma->vm_ops->set_policy(vma, new); if (!err) { mpol_get(new); vma->vm_policy = new; |
f0be3d32b mempolicy: rename... |
522 |
mpol_put(old); |
1da177e4c Linux-2.6.12-rc2 |
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 |
} return err; } /* Step 2: apply policy to a range and do splits. */ static int mbind_range(struct vm_area_struct *vma, unsigned long start, unsigned long end, struct mempolicy *new) { struct vm_area_struct *next; int err; err = 0; for (; vma && vma->vm_start < end; vma = next) { next = vma->vm_next; if (vma->vm_start < start) err = split_vma(vma->vm_mm, vma, start, 1); if (!err && vma->vm_end > end) err = split_vma(vma->vm_mm, vma, end, 0); if (!err) err = policy_vma(vma, new); if (err) break; } return err; } |
c61afb181 [PATCH] cpuset me... |
548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 |
/* * Update task->flags PF_MEMPOLICY bit: set iff non-default * mempolicy. Allows more rapid checking of this (combined perhaps * with other PF_* flag bits) on memory allocation hot code paths. * * If called from outside this file, the task 'p' should -only- be * a newly forked child not yet visible on the task list, because * manipulating the task flags of a visible task is not safe. * * The above limitation is why this routine has the funny name * mpol_fix_fork_child_flag(). * * It is also safe to call this with a task pointer of current, * which the static wrapper mpol_set_task_struct_flag() does, * for use within this file. */ void mpol_fix_fork_child_flag(struct task_struct *p) { if (p->mempolicy) p->flags |= PF_MEMPOLICY; else p->flags &= ~PF_MEMPOLICY; } static void mpol_set_task_struct_flag(void) { mpol_fix_fork_child_flag(current); } |
1da177e4c Linux-2.6.12-rc2 |
577 |
/* Set the process memory policy */ |
028fec414 mempolicy: suppor... |
578 579 |
static long do_set_mempolicy(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
580 |
{ |
1da177e4c Linux-2.6.12-rc2 |
581 |
struct mempolicy *new; |
f4e53d910 mempolicy: write ... |
582 |
struct mm_struct *mm = current->mm; |
1da177e4c Linux-2.6.12-rc2 |
583 |
|
028fec414 mempolicy: suppor... |
584 |
new = mpol_new(mode, flags, nodes); |
1da177e4c Linux-2.6.12-rc2 |
585 586 |
if (IS_ERR(new)) return PTR_ERR(new); |
f4e53d910 mempolicy: write ... |
587 588 589 590 591 592 593 594 595 |
/* * prevent changing our mempolicy while show_numa_maps() * is using it. * Note: do_set_mempolicy() can be called at init time * with no 'mm'. */ if (mm) down_write(&mm->mmap_sem); |
f0be3d32b mempolicy: rename... |
596 |
mpol_put(current->mempolicy); |
1da177e4c Linux-2.6.12-rc2 |
597 |
current->mempolicy = new; |
c61afb181 [PATCH] cpuset me... |
598 |
mpol_set_task_struct_flag(); |
45c4745af mempolicy: rename... |
599 |
if (new && new->mode == MPOL_INTERLEAVE && |
f5b087b52 mempolicy: add MP... |
600 |
nodes_weight(new->v.nodes)) |
dfcd3c0dc [PATCH] Convert m... |
601 |
current->il_next = first_node(new->v.nodes); |
f4e53d910 mempolicy: write ... |
602 603 |
if (mm) up_write(&mm->mmap_sem); |
1da177e4c Linux-2.6.12-rc2 |
604 605 |
return 0; } |
bea904d54 mempolicy: use MP... |
606 607 608 609 |
/* * Return nodemask for policy for get_mempolicy() query */ static void get_policy_nodemask(struct mempolicy *p, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
610 |
{ |
dfcd3c0dc [PATCH] Convert m... |
611 |
nodes_clear(*nodes); |
bea904d54 mempolicy: use MP... |
612 613 |
if (p == &default_policy) return; |
45c4745af mempolicy: rename... |
614 |
switch (p->mode) { |
19770b326 mm: filter based ... |
615 616 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
617 |
case MPOL_INTERLEAVE: |
dfcd3c0dc [PATCH] Convert m... |
618 |
*nodes = p->v.nodes; |
1da177e4c Linux-2.6.12-rc2 |
619 620 |
break; case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
621 |
if (!(p->flags & MPOL_F_LOCAL)) |
dfcd3c0dc [PATCH] Convert m... |
622 |
node_set(p->v.preferred_node, *nodes); |
53f2556b6 mempolicy: mPOL_P... |
623 |
/* else return empty node mask for local allocation */ |
1da177e4c Linux-2.6.12-rc2 |
624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 |
break; default: BUG(); } } static int lookup_node(struct mm_struct *mm, unsigned long addr) { struct page *p; int err; err = get_user_pages(current, mm, addr & PAGE_MASK, 1, 0, 0, &p, NULL); if (err >= 0) { err = page_to_nid(p); put_page(p); } return err; } |
1da177e4c Linux-2.6.12-rc2 |
642 |
/* Retrieve NUMA policy */ |
dbcb0f19c mm/mempolicy.c: c... |
643 644 |
static long do_get_mempolicy(int *policy, nodemask_t *nmask, unsigned long addr, unsigned long flags) |
1da177e4c Linux-2.6.12-rc2 |
645 |
{ |
8bccd85ff [PATCH] Implement... |
646 |
int err; |
1da177e4c Linux-2.6.12-rc2 |
647 648 649 |
struct mm_struct *mm = current->mm; struct vm_area_struct *vma = NULL; struct mempolicy *pol = current->mempolicy; |
cf2a473c4 [PATCH] cpuset: c... |
650 |
cpuset_update_task_memory_state(); |
754af6f5a Mem Policy: add M... |
651 652 |
if (flags & ~(unsigned long)(MPOL_F_NODE|MPOL_F_ADDR|MPOL_F_MEMS_ALLOWED)) |
1da177e4c Linux-2.6.12-rc2 |
653 |
return -EINVAL; |
754af6f5a Mem Policy: add M... |
654 655 656 657 658 659 660 661 |
if (flags & MPOL_F_MEMS_ALLOWED) { if (flags & (MPOL_F_NODE|MPOL_F_ADDR)) return -EINVAL; *policy = 0; /* just so it's initialized */ *nmask = cpuset_current_mems_allowed; return 0; } |
1da177e4c Linux-2.6.12-rc2 |
662 |
if (flags & MPOL_F_ADDR) { |
bea904d54 mempolicy: use MP... |
663 664 665 666 667 |
/* * Do NOT fall back to task policy if the * vma/shared policy at addr is NULL. We * want to return MPOL_DEFAULT in this case. */ |
1da177e4c Linux-2.6.12-rc2 |
668 669 670 671 672 673 674 675 676 677 678 679 680 681 |
down_read(&mm->mmap_sem); vma = find_vma_intersection(mm, addr, addr+1); if (!vma) { up_read(&mm->mmap_sem); return -EFAULT; } if (vma->vm_ops && vma->vm_ops->get_policy) pol = vma->vm_ops->get_policy(vma, addr); else pol = vma->vm_policy; } else if (addr) return -EINVAL; if (!pol) |
bea904d54 mempolicy: use MP... |
682 |
pol = &default_policy; /* indicates default behavior */ |
1da177e4c Linux-2.6.12-rc2 |
683 684 685 686 687 688 |
if (flags & MPOL_F_NODE) { if (flags & MPOL_F_ADDR) { err = lookup_node(mm, addr); if (err < 0) goto out; |
8bccd85ff [PATCH] Implement... |
689 |
*policy = err; |
1da177e4c Linux-2.6.12-rc2 |
690 |
} else if (pol == current->mempolicy && |
45c4745af mempolicy: rename... |
691 |
pol->mode == MPOL_INTERLEAVE) { |
8bccd85ff [PATCH] Implement... |
692 |
*policy = current->il_next; |
1da177e4c Linux-2.6.12-rc2 |
693 694 695 696 |
} else { err = -EINVAL; goto out; } |
bea904d54 mempolicy: use MP... |
697 698 699 |
} else { *policy = pol == &default_policy ? MPOL_DEFAULT : pol->mode; |
d79df630f mempolicy: mask o... |
700 701 702 703 704 |
/* * Internal mempolicy flags must be masked off before exposing * the policy to userspace. */ *policy |= (pol->flags & MPOL_MODE_FLAGS); |
bea904d54 mempolicy: use MP... |
705 |
} |
1da177e4c Linux-2.6.12-rc2 |
706 707 708 709 710 |
if (vma) { up_read(¤t->mm->mmap_sem); vma = NULL; } |
1da177e4c Linux-2.6.12-rc2 |
711 |
err = 0; |
8bccd85ff [PATCH] Implement... |
712 |
if (nmask) |
bea904d54 mempolicy: use MP... |
713 |
get_policy_nodemask(pol, nmask); |
1da177e4c Linux-2.6.12-rc2 |
714 715 |
out: |
52cd3b074 mempolicy: rework... |
716 |
mpol_cond_put(pol); |
1da177e4c Linux-2.6.12-rc2 |
717 718 719 720 |
if (vma) up_read(¤t->mm->mmap_sem); return err; } |
b20a35035 [PATCH] page migr... |
721 |
#ifdef CONFIG_MIGRATION |
8bccd85ff [PATCH] Implement... |
722 |
/* |
6ce3c4c0f [PATCH] Move page... |
723 724 |
* page migration */ |
fc3012896 [PATCH] Simplify ... |
725 726 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
727 728 |
{ /* |
fc3012896 [PATCH] Simplify ... |
729 |
* Avoid migrating a page that is shared with others. |
6ce3c4c0f [PATCH] Move page... |
730 |
*/ |
62695a84e vmscan: move isol... |
731 732 733 734 735 |
if ((flags & MPOL_MF_MOVE_ALL) || page_mapcount(page) == 1) { if (!isolate_lru_page(page)) { list_add_tail(&page->lru, pagelist); } } |
7e2ab150d [PATCH] Direct Mi... |
736 |
} |
6ce3c4c0f [PATCH] Move page... |
737 |
|
742755a1d [PATCH] page migr... |
738 |
static struct page *new_node_page(struct page *page, unsigned long node, int **x) |
95a402c38 [PATCH] page migr... |
739 |
{ |
769848c03 Add __GFP_MOVABLE... |
740 |
return alloc_pages_node(node, GFP_HIGHUSER_MOVABLE, 0); |
95a402c38 [PATCH] page migr... |
741 |
} |
7e2ab150d [PATCH] Direct Mi... |
742 743 744 745 |
/* * Migrate pages from one node to a target node. * Returns error or the number of pages not migrated. */ |
dbcb0f19c mm/mempolicy.c: c... |
746 747 |
static int migrate_to_node(struct mm_struct *mm, int source, int dest, int flags) |
7e2ab150d [PATCH] Direct Mi... |
748 749 750 751 752 753 754 |
{ nodemask_t nmask; LIST_HEAD(pagelist); int err = 0; nodes_clear(nmask); node_set(source, nmask); |
6ce3c4c0f [PATCH] Move page... |
755 |
|
7e2ab150d [PATCH] Direct Mi... |
756 757 |
check_range(mm, mm->mmap->vm_start, TASK_SIZE, &nmask, flags | MPOL_MF_DISCONTIG_OK, &pagelist); |
aaa994b30 [PATCH] page migr... |
758 |
if (!list_empty(&pagelist)) |
95a402c38 [PATCH] page migr... |
759 |
err = migrate_pages(&pagelist, new_node_page, dest); |
7e2ab150d [PATCH] Direct Mi... |
760 |
return err; |
6ce3c4c0f [PATCH] Move page... |
761 762 763 |
} /* |
7e2ab150d [PATCH] Direct Mi... |
764 765 |
* Move pages between the two nodesets so as to preserve the physical * layout as much as possible. |
39743889a [PATCH] Swap Migr... |
766 767 768 769 770 771 |
* * Returns the number of page that could not be moved. */ int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from_nodes, const nodemask_t *to_nodes, int flags) { |
7e2ab150d [PATCH] Direct Mi... |
772 |
int busy = 0; |
0aedadf91 mm: move migrate_... |
773 |
int err; |
7e2ab150d [PATCH] Direct Mi... |
774 |
nodemask_t tmp; |
39743889a [PATCH] Swap Migr... |
775 |
|
0aedadf91 mm: move migrate_... |
776 777 778 |
err = migrate_prep(); if (err) return err; |
53f2556b6 mempolicy: mPOL_P... |
779 |
down_read(&mm->mmap_sem); |
39743889a [PATCH] Swap Migr... |
780 |
|
7b2259b3e [PATCH] page migr... |
781 782 783 |
err = migrate_vmas(mm, from_nodes, to_nodes, flags); if (err) goto out; |
7e2ab150d [PATCH] Direct Mi... |
784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 |
/* * Find a 'source' bit set in 'tmp' whose corresponding 'dest' * bit in 'to' is not also set in 'tmp'. Clear the found 'source' * bit in 'tmp', and return that <source, dest> pair for migration. * The pair of nodemasks 'to' and 'from' define the map. * * If no pair of bits is found that way, fallback to picking some * pair of 'source' and 'dest' bits that are not the same. If the * 'source' and 'dest' bits are the same, this represents a node * that will be migrating to itself, so no pages need move. * * If no bits are left in 'tmp', or if all remaining bits left * in 'tmp' correspond to the same bit in 'to', return false * (nothing left to migrate). * * This lets us pick a pair of nodes to migrate between, such that * if possible the dest node is not already occupied by some other * source node, minimizing the risk of overloading the memory on a * node that would happen if we migrated incoming memory to a node * before migrating outgoing memory source that same node. * * A single scan of tmp is sufficient. As we go, we remember the * most recent <s, d> pair that moved (s != d). If we find a pair * that not only moved, but what's better, moved to an empty slot * (d is not set in tmp), then we break out then, with that pair. * Otherwise when we finish scannng from_tmp, we at least have the * most recent <s, d> pair that moved. If we get all the way through * the scan of tmp without finding any node that moved, much less * moved to an empty node, then there is nothing left worth migrating. */ |
d49847113 [PATCH] SwapMig: ... |
814 |
|
7e2ab150d [PATCH] Direct Mi... |
815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 |
tmp = *from_nodes; while (!nodes_empty(tmp)) { int s,d; int source = -1; int dest = 0; for_each_node_mask(s, tmp) { d = node_remap(s, *from_nodes, *to_nodes); if (s == d) continue; source = s; /* Node moved. Memorize */ dest = d; /* dest not in remaining from nodes? */ if (!node_isset(dest, tmp)) break; } if (source == -1) break; node_clear(source, tmp); err = migrate_to_node(mm, source, dest, flags); if (err > 0) busy += err; if (err < 0) break; |
39743889a [PATCH] Swap Migr... |
842 |
} |
7b2259b3e [PATCH] page migr... |
843 |
out: |
39743889a [PATCH] Swap Migr... |
844 |
up_read(&mm->mmap_sem); |
7e2ab150d [PATCH] Direct Mi... |
845 846 847 |
if (err < 0) return err; return busy; |
b20a35035 [PATCH] page migr... |
848 849 |
} |
3ad33b243 Migration: find c... |
850 851 852 853 854 855 856 |
/* * Allocate a new page for page migration based on vma policy. * Start assuming that page is mapped by vma pointed to by @private. * Search forward from there, if not. N.B., this assumes that the * list of pages handed to migrate_pages()--which is how we get here-- * is in virtual address order. */ |
742755a1d [PATCH] page migr... |
857 |
static struct page *new_vma_page(struct page *page, unsigned long private, int **x) |
95a402c38 [PATCH] page migr... |
858 859 |
{ struct vm_area_struct *vma = (struct vm_area_struct *)private; |
3ad33b243 Migration: find c... |
860 |
unsigned long uninitialized_var(address); |
95a402c38 [PATCH] page migr... |
861 |
|
3ad33b243 Migration: find c... |
862 863 864 865 866 867 868 869 870 871 872 |
while (vma) { address = page_address_in_vma(page, vma); if (address != -EFAULT) break; vma = vma->vm_next; } /* * if !vma, alloc_page_vma() will use task or system default policy */ return alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); |
95a402c38 [PATCH] page migr... |
873 |
} |
b20a35035 [PATCH] page migr... |
874 875 876 877 878 |
#else static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) { |
39743889a [PATCH] Swap Migr... |
879 |
} |
b20a35035 [PATCH] page migr... |
880 881 882 883 884 |
int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from_nodes, const nodemask_t *to_nodes, int flags) { return -ENOSYS; } |
95a402c38 [PATCH] page migr... |
885 |
|
699397499 [PATCH] Fix do_mb... |
886 |
static struct page *new_vma_page(struct page *page, unsigned long private, int **x) |
95a402c38 [PATCH] page migr... |
887 888 889 |
{ return NULL; } |
b20a35035 [PATCH] page migr... |
890 |
#endif |
dbcb0f19c mm/mempolicy.c: c... |
891 |
static long do_mbind(unsigned long start, unsigned long len, |
028fec414 mempolicy: suppor... |
892 893 |
unsigned short mode, unsigned short mode_flags, nodemask_t *nmask, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
894 895 896 897 898 899 900 |
{ struct vm_area_struct *vma; struct mm_struct *mm = current->mm; struct mempolicy *new; unsigned long end; int err; LIST_HEAD(pagelist); |
a3b51e014 mempolicy: conver... |
901 902 |
if (flags & ~(unsigned long)(MPOL_MF_STRICT | MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) |
6ce3c4c0f [PATCH] Move page... |
903 |
return -EINVAL; |
74c002410 [PATCH] Consisten... |
904 |
if ((flags & MPOL_MF_MOVE_ALL) && !capable(CAP_SYS_NICE)) |
6ce3c4c0f [PATCH] Move page... |
905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 |
return -EPERM; if (start & ~PAGE_MASK) return -EINVAL; if (mode == MPOL_DEFAULT) flags &= ~MPOL_MF_STRICT; len = (len + PAGE_SIZE - 1) & PAGE_MASK; end = start + len; if (end < start) return -EINVAL; if (end == start) return 0; |
028fec414 mempolicy: suppor... |
920 |
new = mpol_new(mode, mode_flags, nmask); |
6ce3c4c0f [PATCH] Move page... |
921 922 923 924 925 926 927 928 929 |
if (IS_ERR(new)) return PTR_ERR(new); /* * If we are using the default policy then operation * on discontinuous address spaces is okay after all */ if (!new) flags |= MPOL_MF_DISCONTIG_OK; |
028fec414 mempolicy: suppor... |
930 931 932 933 |
pr_debug("mbind %lx-%lx mode:%d flags:%d nodes:%lx ", start, start + len, mode, mode_flags, nmask ? nodes_addr(*nmask)[0] : -1); |
6ce3c4c0f [PATCH] Move page... |
934 |
|
0aedadf91 mm: move migrate_... |
935 936 937 938 939 940 |
if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) { err = migrate_prep(); if (err) return err; } |
6ce3c4c0f [PATCH] Move page... |
941 942 943 944 945 946 947 948 949 |
down_write(&mm->mmap_sem); vma = check_range(mm, start, end, nmask, flags | MPOL_MF_INVERT, &pagelist); err = PTR_ERR(vma); if (!IS_ERR(vma)) { int nr_failed = 0; err = mbind_range(vma, start, end, new); |
7e2ab150d [PATCH] Direct Mi... |
950 |
|
6ce3c4c0f [PATCH] Move page... |
951 |
if (!list_empty(&pagelist)) |
95a402c38 [PATCH] page migr... |
952 953 |
nr_failed = migrate_pages(&pagelist, new_vma_page, (unsigned long)vma); |
6ce3c4c0f [PATCH] Move page... |
954 955 956 957 |
if (!err && nr_failed && (flags & MPOL_MF_STRICT)) err = -EIO; } |
b20a35035 [PATCH] page migr... |
958 |
|
6ce3c4c0f [PATCH] Move page... |
959 |
up_write(&mm->mmap_sem); |
f0be3d32b mempolicy: rename... |
960 |
mpol_put(new); |
6ce3c4c0f [PATCH] Move page... |
961 962 |
return err; } |
39743889a [PATCH] Swap Migr... |
963 |
/* |
8bccd85ff [PATCH] Implement... |
964 965 966 967 |
* User space interface with variable sized bitmaps for nodelists. */ /* Copy a node mask from user space. */ |
39743889a [PATCH] Swap Migr... |
968 |
static int get_nodes(nodemask_t *nodes, const unsigned long __user *nmask, |
8bccd85ff [PATCH] Implement... |
969 970 971 972 973 974 975 976 977 978 |
unsigned long maxnode) { unsigned long k; unsigned long nlongs; unsigned long endmask; --maxnode; nodes_clear(*nodes); if (maxnode == 0 || !nmask) return 0; |
a9c930bac [PATCH] Fix units... |
979 |
if (maxnode > PAGE_SIZE*BITS_PER_BYTE) |
636f13c17 [PATCH] sys_mbind... |
980 |
return -EINVAL; |
8bccd85ff [PATCH] Implement... |
981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 |
nlongs = BITS_TO_LONGS(maxnode); if ((maxnode % BITS_PER_LONG) == 0) endmask = ~0UL; else endmask = (1UL << (maxnode % BITS_PER_LONG)) - 1; /* When the user specified more nodes than supported just check if the non supported part is all zero. */ if (nlongs > BITS_TO_LONGS(MAX_NUMNODES)) { if (nlongs > PAGE_SIZE/sizeof(long)) return -EINVAL; for (k = BITS_TO_LONGS(MAX_NUMNODES); k < nlongs; k++) { unsigned long t; if (get_user(t, nmask + k)) return -EFAULT; if (k == nlongs - 1) { if (t & endmask) return -EINVAL; } else if (t) return -EINVAL; } nlongs = BITS_TO_LONGS(MAX_NUMNODES); endmask = ~0UL; } if (copy_from_user(nodes_addr(*nodes), nmask, nlongs*sizeof(unsigned long))) return -EFAULT; nodes_addr(*nodes)[nlongs-1] &= endmask; return 0; } /* Copy a kernel node mask to user space */ static int copy_nodes_to_user(unsigned long __user *mask, unsigned long maxnode, nodemask_t *nodes) { unsigned long copy = ALIGN(maxnode-1, 64) / 8; const int nbytes = BITS_TO_LONGS(MAX_NUMNODES) * sizeof(long); if (copy > nbytes) { if (copy > PAGE_SIZE) return -EINVAL; if (clear_user((char __user *)mask + nbytes, copy - nbytes)) return -EFAULT; copy = nbytes; } return copy_to_user(mask, nodes_addr(*nodes), copy) ? -EFAULT : 0; } asmlinkage long sys_mbind(unsigned long start, unsigned long len, unsigned long mode, unsigned long __user *nmask, unsigned long maxnode, unsigned flags) { nodemask_t nodes; int err; |
028fec414 mempolicy: suppor... |
1037 |
unsigned short mode_flags; |
8bccd85ff [PATCH] Implement... |
1038 |
|
028fec414 mempolicy: suppor... |
1039 1040 |
mode_flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; |
a3b51e014 mempolicy: conver... |
1041 1042 |
if (mode >= MPOL_MAX) return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1043 1044 1045 |
if ((mode_flags & MPOL_F_STATIC_NODES) && (mode_flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1046 1047 1048 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1049 |
return do_mbind(start, len, mode, mode_flags, &nodes, flags); |
8bccd85ff [PATCH] Implement... |
1050 1051 1052 1053 1054 1055 1056 1057 |
} /* Set the process memory policy */ asmlinkage long sys_set_mempolicy(int mode, unsigned long __user *nmask, unsigned long maxnode) { int err; nodemask_t nodes; |
028fec414 mempolicy: suppor... |
1058 |
unsigned short flags; |
8bccd85ff [PATCH] Implement... |
1059 |
|
028fec414 mempolicy: suppor... |
1060 1061 1062 |
flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; if ((unsigned int)mode >= MPOL_MAX) |
8bccd85ff [PATCH] Implement... |
1063 |
return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1064 1065 |
if ((flags & MPOL_F_STATIC_NODES) && (flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1066 1067 1068 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1069 |
return do_set_mempolicy(mode, flags, &nodes); |
8bccd85ff [PATCH] Implement... |
1070 |
} |
39743889a [PATCH] Swap Migr... |
1071 1072 1073 1074 |
asmlinkage long sys_migrate_pages(pid_t pid, unsigned long maxnode, const unsigned long __user *old_nodes, const unsigned long __user *new_nodes) { |
c69e8d9c0 CRED: Use RCU to ... |
1075 |
const struct cred *cred = current_cred(), *tcred; |
39743889a [PATCH] Swap Migr... |
1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 |
struct mm_struct *mm; struct task_struct *task; nodemask_t old; nodemask_t new; nodemask_t task_nodes; int err; err = get_nodes(&old, old_nodes, maxnode); if (err) return err; err = get_nodes(&new, new_nodes, maxnode); if (err) return err; /* Find the mm_struct */ read_lock(&tasklist_lock); |
228ebcbe6 Uninline find_tas... |
1093 |
task = pid ? find_task_by_vpid(pid) : current; |
39743889a [PATCH] Swap Migr... |
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 |
if (!task) { read_unlock(&tasklist_lock); return -ESRCH; } mm = get_task_mm(task); read_unlock(&tasklist_lock); if (!mm) return -EINVAL; /* * Check if this process has the right to modify the specified * process. The right exists if the process has administrative |
7f927fcc2 [PATCH] Typo fixes |
1107 |
* capabilities, superuser privileges or the same |
39743889a [PATCH] Swap Migr... |
1108 1109 |
* userid as the target process. */ |
c69e8d9c0 CRED: Use RCU to ... |
1110 1111 |
rcu_read_lock(); tcred = __task_cred(task); |
b6dff3ec5 CRED: Separate ta... |
1112 1113 |
if (cred->euid != tcred->suid && cred->euid != tcred->uid && cred->uid != tcred->suid && cred->uid != tcred->uid && |
74c002410 [PATCH] Consisten... |
1114 |
!capable(CAP_SYS_NICE)) { |
c69e8d9c0 CRED: Use RCU to ... |
1115 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1116 1117 1118 |
err = -EPERM; goto out; } |
c69e8d9c0 CRED: Use RCU to ... |
1119 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1120 1121 1122 |
task_nodes = cpuset_mems_allowed(task); /* Is the user allowed to access the target nodes? */ |
74c002410 [PATCH] Consisten... |
1123 |
if (!nodes_subset(new, task_nodes) && !capable(CAP_SYS_NICE)) { |
39743889a [PATCH] Swap Migr... |
1124 1125 1126 |
err = -EPERM; goto out; } |
37b07e416 memoryless nodes:... |
1127 |
if (!nodes_subset(new, node_states[N_HIGH_MEMORY])) { |
3b42d28b2 Page migration: D... |
1128 1129 1130 |
err = -EINVAL; goto out; } |
86c3a7645 [PATCH] SELinux: ... |
1131 1132 1133 |
err = security_task_movememory(task); if (err) goto out; |
511030bcd [PATCH] Fix sys_m... |
1134 |
err = do_migrate_pages(mm, &old, &new, |
74c002410 [PATCH] Consisten... |
1135 |
capable(CAP_SYS_NICE) ? MPOL_MF_MOVE_ALL : MPOL_MF_MOVE); |
39743889a [PATCH] Swap Migr... |
1136 1137 1138 1139 |
out: mmput(mm); return err; } |
8bccd85ff [PATCH] Implement... |
1140 1141 1142 1143 1144 1145 |
/* Retrieve NUMA policy */ asmlinkage long sys_get_mempolicy(int __user *policy, unsigned long __user *nmask, unsigned long maxnode, unsigned long addr, unsigned long flags) { |
dbcb0f19c mm/mempolicy.c: c... |
1146 1147 |
int err; int uninitialized_var(pval); |
8bccd85ff [PATCH] Implement... |
1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 |
nodemask_t nodes; if (nmask != NULL && maxnode < MAX_NUMNODES) return -EINVAL; err = do_get_mempolicy(&pval, &nodes, addr, flags); if (err) return err; if (policy && put_user(pval, policy)) return -EFAULT; if (nmask) err = copy_nodes_to_user(nmask, maxnode, &nodes); return err; } |
1da177e4c Linux-2.6.12-rc2 |
1166 1167 1168 1169 1170 1171 1172 1173 1174 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 1220 1221 1222 1223 1224 1225 |
#ifdef CONFIG_COMPAT asmlinkage long compat_sys_get_mempolicy(int __user *policy, compat_ulong_t __user *nmask, compat_ulong_t maxnode, compat_ulong_t addr, compat_ulong_t flags) { long err; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; DECLARE_BITMAP(bm, MAX_NUMNODES); nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) nm = compat_alloc_user_space(alloc_size); err = sys_get_mempolicy(policy, nm, nr_bits+1, addr, flags); if (!err && nmask) { err = copy_from_user(bm, nm, alloc_size); /* ensure entire bitmap is zeroed */ err |= clear_user(nmask, ALIGN(maxnode-1, 8) / 8); err |= compat_put_bitmap(nmask, bm, nr_bits); } return err; } asmlinkage long compat_sys_set_mempolicy(int mode, compat_ulong_t __user *nmask, compat_ulong_t maxnode) { long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; DECLARE_BITMAP(bm, MAX_NUMNODES); nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) { err = compat_get_bitmap(bm, nmask, nr_bits); nm = compat_alloc_user_space(alloc_size); err |= copy_to_user(nm, bm, alloc_size); } if (err) return -EFAULT; return sys_set_mempolicy(mode, nm, nr_bits+1); } asmlinkage long compat_sys_mbind(compat_ulong_t start, compat_ulong_t len, compat_ulong_t mode, compat_ulong_t __user *nmask, compat_ulong_t maxnode, compat_ulong_t flags) { long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; |
dfcd3c0dc [PATCH] Convert m... |
1226 |
nodemask_t bm; |
1da177e4c Linux-2.6.12-rc2 |
1227 1228 1229 1230 1231 |
nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) { |
dfcd3c0dc [PATCH] Convert m... |
1232 |
err = compat_get_bitmap(nodes_addr(bm), nmask, nr_bits); |
1da177e4c Linux-2.6.12-rc2 |
1233 |
nm = compat_alloc_user_space(alloc_size); |
dfcd3c0dc [PATCH] Convert m... |
1234 |
err |= copy_to_user(nm, nodes_addr(bm), alloc_size); |
1da177e4c Linux-2.6.12-rc2 |
1235 1236 1237 1238 1239 1240 1241 1242 1243 |
} if (err) return -EFAULT; return sys_mbind(start, len, mode, nm, nr_bits+1, flags); } #endif |
480eccf9a Fix NUMA Memory P... |
1244 1245 1246 1247 1248 1249 1250 1251 |
/* * get_vma_policy(@task, @vma, @addr) * @task - task for fallback if vma policy == default * @vma - virtual memory area whose policy is sought * @addr - address in @vma for shared policy lookup * * Returns effective policy for a VMA at specified address. * Falls back to @task or system default policy, as necessary. |
52cd3b074 mempolicy: rework... |
1252 1253 1254 1255 1256 1257 1258 |
* Current or other task's task mempolicy and non-shared vma policies * are protected by the task's mmap_sem, which must be held for read by * the caller. * Shared policies [those marked as MPOL_F_SHARED] require an extra reference * count--added by the get_policy() vm_op, as appropriate--to protect against * freeing by another task. It is the caller's responsibility to free the * extra reference for shared policies. |
480eccf9a Fix NUMA Memory P... |
1259 |
*/ |
ae4d8c16a mempolicy: fixup ... |
1260 |
static struct mempolicy *get_vma_policy(struct task_struct *task, |
48fce3429 [PATCH] mempolici... |
1261 |
struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
1262 |
{ |
6e21c8f14 [PATCH] /proc/<pi... |
1263 |
struct mempolicy *pol = task->mempolicy; |
1da177e4c Linux-2.6.12-rc2 |
1264 1265 |
if (vma) { |
480eccf9a Fix NUMA Memory P... |
1266 |
if (vma->vm_ops && vma->vm_ops->get_policy) { |
ae4d8c16a mempolicy: fixup ... |
1267 1268 1269 1270 |
struct mempolicy *vpol = vma->vm_ops->get_policy(vma, addr); if (vpol) pol = vpol; |
bea904d54 mempolicy: use MP... |
1271 |
} else if (vma->vm_policy) |
1da177e4c Linux-2.6.12-rc2 |
1272 1273 1274 1275 1276 1277 |
pol = vma->vm_policy; } if (!pol) pol = &default_policy; return pol; } |
52cd3b074 mempolicy: rework... |
1278 1279 1280 1281 1282 |
/* * Return a nodemask representing a mempolicy for filtering nodes for * page allocation */ static nodemask_t *policy_nodemask(gfp_t gfp, struct mempolicy *policy) |
19770b326 mm: filter based ... |
1283 1284 |
{ /* Lower zones don't get a nodemask applied for MPOL_BIND */ |
45c4745af mempolicy: rename... |
1285 |
if (unlikely(policy->mode == MPOL_BIND) && |
19770b326 mm: filter based ... |
1286 1287 1288 1289 1290 1291 |
gfp_zone(gfp) >= policy_zone && cpuset_nodemask_valid_mems_allowed(&policy->v.nodes)) return &policy->v.nodes; return NULL; } |
52cd3b074 mempolicy: rework... |
1292 1293 |
/* Return a zonelist indicated by gfp for node representing a mempolicy */ static struct zonelist *policy_zonelist(gfp_t gfp, struct mempolicy *policy) |
1da177e4c Linux-2.6.12-rc2 |
1294 |
{ |
fc36b8d3d mempolicy: use MP... |
1295 |
int nd = numa_node_id(); |
1da177e4c Linux-2.6.12-rc2 |
1296 |
|
45c4745af mempolicy: rename... |
1297 |
switch (policy->mode) { |
1da177e4c Linux-2.6.12-rc2 |
1298 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1299 1300 |
if (!(policy->flags & MPOL_F_LOCAL)) nd = policy->v.preferred_node; |
1da177e4c Linux-2.6.12-rc2 |
1301 1302 |
break; case MPOL_BIND: |
19770b326 mm: filter based ... |
1303 |
/* |
52cd3b074 mempolicy: rework... |
1304 1305 1306 1307 |
* Normally, MPOL_BIND allocations are node-local within the * allowed nodemask. However, if __GFP_THISNODE is set and the * current node is part of the mask, we use the zonelist for * the first node in the mask instead. |
19770b326 mm: filter based ... |
1308 |
*/ |
19770b326 mm: filter based ... |
1309 1310 1311 1312 |
if (unlikely(gfp & __GFP_THISNODE) && unlikely(!node_isset(nd, policy->v.nodes))) nd = first_node(policy->v.nodes); break; |
1da177e4c Linux-2.6.12-rc2 |
1313 |
case MPOL_INTERLEAVE: /* should not happen */ |
1da177e4c Linux-2.6.12-rc2 |
1314 1315 |
break; default: |
1da177e4c Linux-2.6.12-rc2 |
1316 1317 |
BUG(); } |
0e88460da mm: introduce nod... |
1318 |
return node_zonelist(nd, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1319 1320 1321 1322 1323 1324 1325 1326 1327 |
} /* Do dynamic interleaving for a process */ static unsigned interleave_nodes(struct mempolicy *policy) { unsigned nid, next; struct task_struct *me = current; nid = me->il_next; |
dfcd3c0dc [PATCH] Convert m... |
1328 |
next = next_node(nid, policy->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1329 |
if (next >= MAX_NUMNODES) |
dfcd3c0dc [PATCH] Convert m... |
1330 |
next = first_node(policy->v.nodes); |
f5b087b52 mempolicy: add MP... |
1331 1332 |
if (next < MAX_NUMNODES) me->il_next = next; |
1da177e4c Linux-2.6.12-rc2 |
1333 1334 |
return nid; } |
dc85da15d [PATCH] NUMA poli... |
1335 1336 1337 |
/* * Depending on the memory policy provide a node from which to allocate the * next slab entry. |
52cd3b074 mempolicy: rework... |
1338 1339 1340 1341 |
* @policy must be protected by freeing by the caller. If @policy is * the current task's mempolicy, this protection is implicit, as only the * task can change it's policy. The system default policy requires no * such protection. |
dc85da15d [PATCH] NUMA poli... |
1342 1343 1344 |
*/ unsigned slab_node(struct mempolicy *policy) { |
fc36b8d3d mempolicy: use MP... |
1345 |
if (!policy || policy->flags & MPOL_F_LOCAL) |
bea904d54 mempolicy: use MP... |
1346 1347 1348 1349 |
return numa_node_id(); switch (policy->mode) { case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1350 1351 1352 1353 |
/* * handled MPOL_F_LOCAL above */ return policy->v.preferred_node; |
765c4507a [PATCH] GFP_THISN... |
1354 |
|
dc85da15d [PATCH] NUMA poli... |
1355 1356 |
case MPOL_INTERLEAVE: return interleave_nodes(policy); |
dd1a239f6 mm: have zonelist... |
1357 |
case MPOL_BIND: { |
dc85da15d [PATCH] NUMA poli... |
1358 1359 1360 1361 |
/* * Follow bind policy behavior and start allocation at the * first node. */ |
19770b326 mm: filter based ... |
1362 1363 1364 1365 1366 1367 1368 1369 |
struct zonelist *zonelist; struct zone *zone; enum zone_type highest_zoneidx = gfp_zone(GFP_KERNEL); zonelist = &NODE_DATA(numa_node_id())->node_zonelists[0]; (void)first_zones_zonelist(zonelist, highest_zoneidx, &policy->v.nodes, &zone); return zone->node; |
dd1a239f6 mm: have zonelist... |
1370 |
} |
dc85da15d [PATCH] NUMA poli... |
1371 |
|
dc85da15d [PATCH] NUMA poli... |
1372 |
default: |
bea904d54 mempolicy: use MP... |
1373 |
BUG(); |
dc85da15d [PATCH] NUMA poli... |
1374 1375 |
} } |
1da177e4c Linux-2.6.12-rc2 |
1376 1377 1378 1379 |
/* Do static interleaving for a VMA with known offset. */ static unsigned offset_il_node(struct mempolicy *pol, struct vm_area_struct *vma, unsigned long off) { |
dfcd3c0dc [PATCH] Convert m... |
1380 |
unsigned nnodes = nodes_weight(pol->v.nodes); |
f5b087b52 mempolicy: add MP... |
1381 |
unsigned target; |
1da177e4c Linux-2.6.12-rc2 |
1382 1383 |
int c; int nid = -1; |
f5b087b52 mempolicy: add MP... |
1384 1385 1386 |
if (!nnodes) return numa_node_id(); target = (unsigned int)off % nnodes; |
1da177e4c Linux-2.6.12-rc2 |
1387 1388 |
c = 0; do { |
dfcd3c0dc [PATCH] Convert m... |
1389 |
nid = next_node(nid, pol->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1390 1391 |
c++; } while (c <= target); |
1da177e4c Linux-2.6.12-rc2 |
1392 1393 |
return nid; } |
5da7ca860 [PATCH] Add NUMA ... |
1394 1395 1396 1397 1398 1399 |
/* Determine a node number for interleave */ static inline unsigned interleave_nid(struct mempolicy *pol, struct vm_area_struct *vma, unsigned long addr, int shift) { if (vma) { unsigned long off; |
3b98b087f [PATCH] fix NUMA ... |
1400 1401 1402 1403 1404 1405 1406 1407 1408 |
/* * for small pages, there is no difference between * shift and PAGE_SHIFT, so the bit-shift is safe. * for huge pages, since vm_pgoff is in units of small * pages, we need to shift off the always 0 bits to get * a useful offset. */ BUG_ON(shift < PAGE_SHIFT); off = vma->vm_pgoff >> (shift - PAGE_SHIFT); |
5da7ca860 [PATCH] Add NUMA ... |
1409 1410 1411 1412 1413 |
off += (addr - vma->vm_start) >> shift; return offset_il_node(pol, vma, off); } else return interleave_nodes(pol); } |
00ac59adf [PATCH] x86_64: F... |
1414 |
#ifdef CONFIG_HUGETLBFS |
480eccf9a Fix NUMA Memory P... |
1415 1416 1417 1418 1419 |
/* * huge_zonelist(@vma, @addr, @gfp_flags, @mpol) * @vma = virtual memory area whose policy is sought * @addr = address in @vma for shared policy lookup and interleave policy * @gfp_flags = for requested zone |
19770b326 mm: filter based ... |
1420 1421 |
* @mpol = pointer to mempolicy pointer for reference counted mempolicy * @nodemask = pointer to nodemask pointer for MPOL_BIND nodemask |
480eccf9a Fix NUMA Memory P... |
1422 |
* |
52cd3b074 mempolicy: rework... |
1423 1424 1425 1426 |
* Returns a zonelist suitable for a huge page allocation and a pointer * to the struct mempolicy for conditional unref after allocation. * If the effective policy is 'BIND, returns a pointer to the mempolicy's * @nodemask for filtering the zonelist. |
480eccf9a Fix NUMA Memory P... |
1427 |
*/ |
396faf030 Allow huge page a... |
1428 |
struct zonelist *huge_zonelist(struct vm_area_struct *vma, unsigned long addr, |
19770b326 mm: filter based ... |
1429 1430 |
gfp_t gfp_flags, struct mempolicy **mpol, nodemask_t **nodemask) |
5da7ca860 [PATCH] Add NUMA ... |
1431 |
{ |
480eccf9a Fix NUMA Memory P... |
1432 |
struct zonelist *zl; |
5da7ca860 [PATCH] Add NUMA ... |
1433 |
|
52cd3b074 mempolicy: rework... |
1434 |
*mpol = get_vma_policy(current, vma, addr); |
19770b326 mm: filter based ... |
1435 |
*nodemask = NULL; /* assume !MPOL_BIND */ |
5da7ca860 [PATCH] Add NUMA ... |
1436 |
|
52cd3b074 mempolicy: rework... |
1437 1438 |
if (unlikely((*mpol)->mode == MPOL_INTERLEAVE)) { zl = node_zonelist(interleave_nid(*mpol, vma, addr, |
a55164389 hugetlb: modular ... |
1439 |
huge_page_shift(hstate_vma(vma))), gfp_flags); |
52cd3b074 mempolicy: rework... |
1440 1441 1442 1443 |
} else { zl = policy_zonelist(gfp_flags, *mpol); if ((*mpol)->mode == MPOL_BIND) *nodemask = &(*mpol)->v.nodes; |
480eccf9a Fix NUMA Memory P... |
1444 1445 |
} return zl; |
5da7ca860 [PATCH] Add NUMA ... |
1446 |
} |
00ac59adf [PATCH] x86_64: F... |
1447 |
#endif |
5da7ca860 [PATCH] Add NUMA ... |
1448 |
|
1da177e4c Linux-2.6.12-rc2 |
1449 1450 |
/* Allocate a page in interleaved policy. Own path because it needs to do special accounting. */ |
662f3a0b9 [PATCH] Remove ne... |
1451 1452 |
static struct page *alloc_page_interleave(gfp_t gfp, unsigned order, unsigned nid) |
1da177e4c Linux-2.6.12-rc2 |
1453 1454 1455 |
{ struct zonelist *zl; struct page *page; |
0e88460da mm: introduce nod... |
1456 |
zl = node_zonelist(nid, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1457 |
page = __alloc_pages(gfp, order, zl); |
dd1a239f6 mm: have zonelist... |
1458 |
if (page && page_zone(page) == zonelist_zone(&zl->_zonerefs[0])) |
ca889e6c4 [PATCH] Use Zoned... |
1459 |
inc_zone_page_state(page, NUMA_INTERLEAVE_HIT); |
1da177e4c Linux-2.6.12-rc2 |
1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 |
return page; } /** * alloc_page_vma - Allocate a page for a VMA. * * @gfp: * %GFP_USER user allocation. * %GFP_KERNEL kernel allocations, * %GFP_HIGHMEM highmem/user allocations, * %GFP_FS allocation should not call back into a file system. * %GFP_ATOMIC don't sleep. * * @vma: Pointer to VMA or NULL if not available. * @addr: Virtual Address of the allocation. Must be inside the VMA. * * This function allocates a page from the kernel page pool and applies * a NUMA policy associated with the VMA or the current process. * When VMA is not NULL caller must hold down_read on the mmap_sem of the * mm_struct of the VMA to prevent it from going away. Should be used for * all allocations for pages that will be mapped into * user space. Returns NULL when no page can be allocated. * * Should be called with the mm_sem of the vma hold. */ struct page * |
dd0fc66fb [PATCH] gfp flags... |
1486 |
alloc_page_vma(gfp_t gfp, struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
1487 |
{ |
6e21c8f14 [PATCH] /proc/<pi... |
1488 |
struct mempolicy *pol = get_vma_policy(current, vma, addr); |
480eccf9a Fix NUMA Memory P... |
1489 |
struct zonelist *zl; |
1da177e4c Linux-2.6.12-rc2 |
1490 |
|
cf2a473c4 [PATCH] cpuset: c... |
1491 |
cpuset_update_task_memory_state(); |
1da177e4c Linux-2.6.12-rc2 |
1492 |
|
45c4745af mempolicy: rename... |
1493 |
if (unlikely(pol->mode == MPOL_INTERLEAVE)) { |
1da177e4c Linux-2.6.12-rc2 |
1494 |
unsigned nid; |
5da7ca860 [PATCH] Add NUMA ... |
1495 1496 |
nid = interleave_nid(pol, vma, addr, PAGE_SHIFT); |
52cd3b074 mempolicy: rework... |
1497 |
mpol_cond_put(pol); |
1da177e4c Linux-2.6.12-rc2 |
1498 1499 |
return alloc_page_interleave(gfp, 0, nid); } |
52cd3b074 mempolicy: rework... |
1500 1501 |
zl = policy_zonelist(gfp, pol); if (unlikely(mpol_needs_cond_ref(pol))) { |
480eccf9a Fix NUMA Memory P... |
1502 |
/* |
52cd3b074 mempolicy: rework... |
1503 |
* slow path: ref counted shared policy |
480eccf9a Fix NUMA Memory P... |
1504 |
*/ |
19770b326 mm: filter based ... |
1505 |
struct page *page = __alloc_pages_nodemask(gfp, 0, |
52cd3b074 mempolicy: rework... |
1506 |
zl, policy_nodemask(gfp, pol)); |
f0be3d32b mempolicy: rename... |
1507 |
__mpol_put(pol); |
480eccf9a Fix NUMA Memory P... |
1508 1509 1510 1511 1512 |
return page; } /* * fast path: default or task policy */ |
52cd3b074 mempolicy: rework... |
1513 |
return __alloc_pages_nodemask(gfp, 0, zl, policy_nodemask(gfp, pol)); |
1da177e4c Linux-2.6.12-rc2 |
1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 |
} /** * alloc_pages_current - Allocate pages. * * @gfp: * %GFP_USER user allocation, * %GFP_KERNEL kernel allocation, * %GFP_HIGHMEM highmem allocation, * %GFP_FS don't call back into a file system. * %GFP_ATOMIC don't sleep. * @order: Power of two of allocation size in pages. 0 is a single page. * * Allocate a page from the kernel page pool. When not in * interrupt context and apply the current process NUMA policy. * Returns NULL when no page can be allocated. * |
cf2a473c4 [PATCH] cpuset: c... |
1531 |
* Don't call cpuset_update_task_memory_state() unless |
1da177e4c Linux-2.6.12-rc2 |
1532 1533 1534 |
* 1) it's ok to take cpuset_sem (can WAIT), and * 2) allocating for current task (not interrupt). */ |
dd0fc66fb [PATCH] gfp flags... |
1535 |
struct page *alloc_pages_current(gfp_t gfp, unsigned order) |
1da177e4c Linux-2.6.12-rc2 |
1536 1537 1538 1539 |
{ struct mempolicy *pol = current->mempolicy; if ((gfp & __GFP_WAIT) && !in_interrupt()) |
cf2a473c4 [PATCH] cpuset: c... |
1540 |
cpuset_update_task_memory_state(); |
9b819d204 [PATCH] Add __GFP... |
1541 |
if (!pol || in_interrupt() || (gfp & __GFP_THISNODE)) |
1da177e4c Linux-2.6.12-rc2 |
1542 |
pol = &default_policy; |
52cd3b074 mempolicy: rework... |
1543 1544 1545 1546 1547 |
/* * No reference counting needed for current->mempolicy * nor system default_policy */ |
45c4745af mempolicy: rename... |
1548 |
if (pol->mode == MPOL_INTERLEAVE) |
1da177e4c Linux-2.6.12-rc2 |
1549 |
return alloc_page_interleave(gfp, order, interleave_nodes(pol)); |
19770b326 mm: filter based ... |
1550 |
return __alloc_pages_nodemask(gfp, order, |
52cd3b074 mempolicy: rework... |
1551 |
policy_zonelist(gfp, pol), policy_nodemask(gfp, pol)); |
1da177e4c Linux-2.6.12-rc2 |
1552 1553 |
} EXPORT_SYMBOL(alloc_pages_current); |
4225399a6 [PATCH] cpuset: r... |
1554 |
/* |
846a16bf0 mempolicy: rename... |
1555 |
* If mpol_dup() sees current->cpuset == cpuset_being_rebound, then it |
4225399a6 [PATCH] cpuset: r... |
1556 1557 1558 1559 1560 |
* rebinds the mempolicy its copying by calling mpol_rebind_policy() * with the mems_allowed returned by cpuset_mems_allowed(). This * keeps mempolicies cpuset relative after its cpuset moves. See * further kernel/cpuset.c update_nodemask(). */ |
4225399a6 [PATCH] cpuset: r... |
1561 |
|
846a16bf0 mempolicy: rename... |
1562 1563 |
/* Slow path of a mempolicy duplicate */ struct mempolicy *__mpol_dup(struct mempolicy *old) |
1da177e4c Linux-2.6.12-rc2 |
1564 1565 1566 1567 1568 |
{ struct mempolicy *new = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!new) return ERR_PTR(-ENOMEM); |
4225399a6 [PATCH] cpuset: r... |
1569 1570 1571 1572 |
if (current_cpuset_is_being_rebound()) { nodemask_t mems = cpuset_mems_allowed(current); mpol_rebind_policy(old, &mems); } |
1da177e4c Linux-2.6.12-rc2 |
1573 1574 |
*new = *old; atomic_set(&new->refcnt, 1); |
1da177e4c Linux-2.6.12-rc2 |
1575 1576 |
return new; } |
52cd3b074 mempolicy: rework... |
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 |
/* * If *frompol needs [has] an extra ref, copy *frompol to *tompol , * eliminate the * MPOL_F_* flags that require conditional ref and * [NOTE!!!] drop the extra ref. Not safe to reference *frompol directly * after return. Use the returned value. * * Allows use of a mempolicy for, e.g., multiple allocations with a single * policy lookup, even if the policy needs/has extra ref on lookup. * shmem_readahead needs this. */ struct mempolicy *__mpol_cond_copy(struct mempolicy *tompol, struct mempolicy *frompol) { if (!mpol_needs_cond_ref(frompol)) return frompol; *tompol = *frompol; tompol->flags &= ~MPOL_F_SHARED; /* copy doesn't need unref */ __mpol_put(frompol); return tompol; } |
f5b087b52 mempolicy: add MP... |
1598 1599 1600 1601 1602 1603 1604 1605 1606 |
static int mpol_match_intent(const struct mempolicy *a, const struct mempolicy *b) { if (a->flags != b->flags) return 0; if (!mpol_store_user_nodemask(a)) return 1; return nodes_equal(a->w.user_nodemask, b->w.user_nodemask); } |
1da177e4c Linux-2.6.12-rc2 |
1607 1608 1609 1610 1611 |
/* Slow path of a mempolicy comparison */ int __mpol_equal(struct mempolicy *a, struct mempolicy *b) { if (!a || !b) return 0; |
45c4745af mempolicy: rename... |
1612 |
if (a->mode != b->mode) |
1da177e4c Linux-2.6.12-rc2 |
1613 |
return 0; |
45c4745af mempolicy: rename... |
1614 |
if (a->mode != MPOL_DEFAULT && !mpol_match_intent(a, b)) |
f5b087b52 mempolicy: add MP... |
1615 |
return 0; |
45c4745af mempolicy: rename... |
1616 |
switch (a->mode) { |
19770b326 mm: filter based ... |
1617 1618 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
1619 |
case MPOL_INTERLEAVE: |
dfcd3c0dc [PATCH] Convert m... |
1620 |
return nodes_equal(a->v.nodes, b->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1621 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1622 1623 |
return a->v.preferred_node == b->v.preferred_node && a->flags == b->flags; |
1da177e4c Linux-2.6.12-rc2 |
1624 1625 1626 1627 1628 |
default: BUG(); return 0; } } |
1da177e4c Linux-2.6.12-rc2 |
1629 |
/* |
1da177e4c Linux-2.6.12-rc2 |
1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 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 |
* Shared memory backing store policy support. * * Remember policies even when nobody has shared memory mapped. * The policies are kept in Red-Black tree linked from the inode. * They are protected by the sp->lock spinlock, which should be held * for any accesses to the tree. */ /* lookup first element intersecting start-end */ /* Caller holds sp->lock */ static struct sp_node * sp_lookup(struct shared_policy *sp, unsigned long start, unsigned long end) { struct rb_node *n = sp->root.rb_node; while (n) { struct sp_node *p = rb_entry(n, struct sp_node, nd); if (start >= p->end) n = n->rb_right; else if (end <= p->start) n = n->rb_left; else break; } if (!n) return NULL; for (;;) { struct sp_node *w = NULL; struct rb_node *prev = rb_prev(n); if (!prev) break; w = rb_entry(prev, struct sp_node, nd); if (w->end <= start) break; n = prev; } return rb_entry(n, struct sp_node, nd); } /* Insert a new shared policy into the list. */ /* Caller holds sp->lock */ static void sp_insert(struct shared_policy *sp, struct sp_node *new) { struct rb_node **p = &sp->root.rb_node; struct rb_node *parent = NULL; struct sp_node *nd; while (*p) { parent = *p; nd = rb_entry(parent, struct sp_node, nd); if (new->start < nd->start) p = &(*p)->rb_left; else if (new->end > nd->end) p = &(*p)->rb_right; else BUG(); } rb_link_node(&new->nd, parent, p); rb_insert_color(&new->nd, &sp->root); |
140d5a490 numa: mempolicy: ... |
1690 1691 |
pr_debug("inserting %lx-%lx: %d ", new->start, new->end, |
45c4745af mempolicy: rename... |
1692 |
new->policy ? new->policy->mode : 0); |
1da177e4c Linux-2.6.12-rc2 |
1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 |
} /* Find shared policy intersecting idx */ struct mempolicy * mpol_shared_policy_lookup(struct shared_policy *sp, unsigned long idx) { struct mempolicy *pol = NULL; struct sp_node *sn; if (!sp->root.rb_node) return NULL; spin_lock(&sp->lock); sn = sp_lookup(sp, idx, idx+1); if (sn) { mpol_get(sn->policy); pol = sn->policy; } spin_unlock(&sp->lock); return pol; } static void sp_delete(struct shared_policy *sp, struct sp_node *n) { |
140d5a490 numa: mempolicy: ... |
1716 1717 |
pr_debug("deleting %lx-l%lx ", n->start, n->end); |
1da177e4c Linux-2.6.12-rc2 |
1718 |
rb_erase(&n->nd, &sp->root); |
f0be3d32b mempolicy: rename... |
1719 |
mpol_put(n->policy); |
1da177e4c Linux-2.6.12-rc2 |
1720 1721 |
kmem_cache_free(sn_cache, n); } |
dbcb0f19c mm/mempolicy.c: c... |
1722 1723 |
static struct sp_node *sp_alloc(unsigned long start, unsigned long end, struct mempolicy *pol) |
1da177e4c Linux-2.6.12-rc2 |
1724 1725 1726 1727 1728 1729 1730 1731 |
{ struct sp_node *n = kmem_cache_alloc(sn_cache, GFP_KERNEL); if (!n) return NULL; n->start = start; n->end = end; mpol_get(pol); |
aab0b1029 mempolicy: mark s... |
1732 |
pol->flags |= MPOL_F_SHARED; /* for unref */ |
1da177e4c Linux-2.6.12-rc2 |
1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 |
n->policy = pol; return n; } /* Replace a policy range. */ static int shared_policy_replace(struct shared_policy *sp, unsigned long start, unsigned long end, struct sp_node *new) { struct sp_node *n, *new2 = NULL; restart: spin_lock(&sp->lock); n = sp_lookup(sp, start, end); /* Take care of old policies in the same range. */ while (n && n->start < end) { struct rb_node *next = rb_next(&n->nd); if (n->start >= start) { if (n->end <= end) sp_delete(sp, n); else n->start = end; } else { /* Old policy spanning whole new range. */ if (n->end > end) { if (!new2) { spin_unlock(&sp->lock); new2 = sp_alloc(end, n->end, n->policy); if (!new2) return -ENOMEM; goto restart; } n->end = start; sp_insert(sp, new2); new2 = NULL; break; } else n->end = start; } if (!next) break; n = rb_entry(next, struct sp_node, nd); } if (new) sp_insert(sp, new); spin_unlock(&sp->lock); if (new2) { |
f0be3d32b mempolicy: rename... |
1779 |
mpol_put(new2->policy); |
1da177e4c Linux-2.6.12-rc2 |
1780 1781 1782 1783 |
kmem_cache_free(sn_cache, new2); } return 0; } |
71fe804b6 mempolicy: use st... |
1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 |
/** * mpol_shared_policy_init - initialize shared policy for inode * @sp: pointer to inode shared policy * @mpol: struct mempolicy to install * * Install non-NULL @mpol in inode's shared policy rb-tree. * On entry, the current task has a reference on a non-NULL @mpol. * This must be released on exit. */ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) { sp->root = RB_ROOT; /* empty tree == default mempolicy */ spin_lock_init(&sp->lock); if (mpol) { struct vm_area_struct pvma; struct mempolicy *new; /* contextualize the tmpfs mount point mempolicy */ new = mpol_new(mpol->mode, mpol->flags, &mpol->w.user_nodemask); mpol_put(mpol); /* drop our ref on sb mpol */ if (IS_ERR(new)) return; /* no valid nodemask intersection */ /* Create pseudo-vma that contains just the policy */ memset(&pvma, 0, sizeof(struct vm_area_struct)); pvma.vm_end = TASK_SIZE; /* policy covers entire file */ mpol_set_shared_policy(sp, &pvma, new); /* adds ref */ mpol_put(new); /* drop initial ref */ |
7339ff830 [PATCH] Add tmpfs... |
1813 1814 |
} } |
1da177e4c Linux-2.6.12-rc2 |
1815 1816 1817 1818 1819 1820 |
int mpol_set_shared_policy(struct shared_policy *info, struct vm_area_struct *vma, struct mempolicy *npol) { int err; struct sp_node *new = NULL; unsigned long sz = vma_pages(vma); |
028fec414 mempolicy: suppor... |
1821 1822 |
pr_debug("set_shared_policy %lx sz %lu %d %d %lx ", |
1da177e4c Linux-2.6.12-rc2 |
1823 |
vma->vm_pgoff, |
45c4745af mempolicy: rename... |
1824 |
sz, npol ? npol->mode : -1, |
028fec414 mempolicy: suppor... |
1825 |
npol ? npol->flags : -1, |
140d5a490 numa: mempolicy: ... |
1826 |
npol ? nodes_addr(npol->v.nodes)[0] : -1); |
1da177e4c Linux-2.6.12-rc2 |
1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 |
if (npol) { new = sp_alloc(vma->vm_pgoff, vma->vm_pgoff + sz, npol); if (!new) return -ENOMEM; } err = shared_policy_replace(info, vma->vm_pgoff, vma->vm_pgoff+sz, new); if (err && new) kmem_cache_free(sn_cache, new); return err; } /* Free a backing policy store on inode delete. */ void mpol_free_shared_policy(struct shared_policy *p) { struct sp_node *n; struct rb_node *next; if (!p->root.rb_node) return; spin_lock(&p->lock); next = rb_first(&p->root); while (next) { n = rb_entry(next, struct sp_node, nd); next = rb_next(&n->nd); |
90c5029e4 [PATCH] Undo memp... |
1852 |
rb_erase(&n->nd, &p->root); |
f0be3d32b mempolicy: rename... |
1853 |
mpol_put(n->policy); |
1da177e4c Linux-2.6.12-rc2 |
1854 1855 1856 |
kmem_cache_free(sn_cache, n); } spin_unlock(&p->lock); |
1da177e4c Linux-2.6.12-rc2 |
1857 1858 1859 1860 1861 |
} /* assumes fs == KERNEL_DS */ void __init numa_policy_init(void) { |
b71636e29 numa: mempolicy: ... |
1862 1863 1864 |
nodemask_t interleave_nodes; unsigned long largest = 0; int nid, prefer = 0; |
1da177e4c Linux-2.6.12-rc2 |
1865 1866 |
policy_cache = kmem_cache_create("numa_policy", sizeof(struct mempolicy), |
20c2df83d mm: Remove slab d... |
1867 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
1868 1869 1870 |
sn_cache = kmem_cache_create("shared_policy_node", sizeof(struct sp_node), |
20c2df83d mm: Remove slab d... |
1871 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
1872 |
|
b71636e29 numa: mempolicy: ... |
1873 1874 1875 1876 1877 1878 |
/* * Set interleaving policy for system init. Interleaving is only * enabled across suitably sized nodes (default is >= 16MB), or * fall back to the largest node if they're all smaller. */ nodes_clear(interleave_nodes); |
56bbd65df Memoryless nodes:... |
1879 |
for_each_node_state(nid, N_HIGH_MEMORY) { |
b71636e29 numa: mempolicy: ... |
1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 |
unsigned long total_pages = node_present_pages(nid); /* Preserve the largest node */ if (largest < total_pages) { largest = total_pages; prefer = nid; } /* Interleave this node? */ if ((total_pages << PAGE_SHIFT) >= (16 << 20)) node_set(nid, interleave_nodes); } /* All too small, use the largest */ if (unlikely(nodes_empty(interleave_nodes))) node_set(prefer, interleave_nodes); |
1da177e4c Linux-2.6.12-rc2 |
1896 |
|
028fec414 mempolicy: suppor... |
1897 |
if (do_set_mempolicy(MPOL_INTERLEAVE, 0, &interleave_nodes)) |
1da177e4c Linux-2.6.12-rc2 |
1898 1899 1900 |
printk("numa_policy_init: interleaving failed "); } |
8bccd85ff [PATCH] Implement... |
1901 |
/* Reset policy of current process to default */ |
1da177e4c Linux-2.6.12-rc2 |
1902 1903 |
void numa_default_policy(void) { |
028fec414 mempolicy: suppor... |
1904 |
do_set_mempolicy(MPOL_DEFAULT, 0, NULL); |
1da177e4c Linux-2.6.12-rc2 |
1905 |
} |
68860ec10 [PATCH] cpusets: ... |
1906 |
|
4225399a6 [PATCH] cpuset: r... |
1907 |
/* |
095f1fc4e mempolicy: rework... |
1908 1909 1910 1911 |
* Parse and format mempolicy from/to strings */ /* |
fc36b8d3d mempolicy: use MP... |
1912 |
* "local" is pseudo-policy: MPOL_PREFERRED with MPOL_F_LOCAL flag |
3f226aa1c mempolicy: suppor... |
1913 |
* Used only for mpol_parse_str() and mpol_to_str() |
1a75a6c82 [PATCH] Fold numa... |
1914 |
*/ |
53f2556b6 mempolicy: mPOL_P... |
1915 |
#define MPOL_LOCAL (MPOL_INTERLEAVE + 1) |
15ad7cdcf [PATCH] struct se... |
1916 |
static const char * const policy_types[] = |
53f2556b6 mempolicy: mPOL_P... |
1917 |
{ "default", "prefer", "bind", "interleave", "local" }; |
1a75a6c82 [PATCH] Fold numa... |
1918 |
|
095f1fc4e mempolicy: rework... |
1919 1920 1921 1922 1923 |
#ifdef CONFIG_TMPFS /** * mpol_parse_str - parse string to mempolicy * @str: string containing mempolicy to parse |
71fe804b6 mempolicy: use st... |
1924 1925 |
* @mpol: pointer to struct mempolicy pointer, returned on success. * @no_context: flag whether to "contextualize" the mempolicy |
095f1fc4e mempolicy: rework... |
1926 1927 1928 1929 |
* * Format of input: * <mode>[=<flags>][:<nodelist>] * |
71fe804b6 mempolicy: use st... |
1930 1931 1932 1933 1934 1935 1936 1937 |
* if @no_context is true, save the input nodemask in w.user_nodemask in * the returned mempolicy. This will be used to "clone" the mempolicy in * a specific context [cpuset] at a later time. Used to parse tmpfs mpol * mount option. Note that if 'static' or 'relative' mode flags were * specified, the input nodemask will already have been saved. Saving * it again is redundant, but safe. * * On success, returns 0, else 1 |
095f1fc4e mempolicy: rework... |
1938 |
*/ |
71fe804b6 mempolicy: use st... |
1939 |
int mpol_parse_str(char *str, struct mempolicy **mpol, int no_context) |
095f1fc4e mempolicy: rework... |
1940 |
{ |
71fe804b6 mempolicy: use st... |
1941 1942 1943 1944 |
struct mempolicy *new = NULL; unsigned short uninitialized_var(mode); unsigned short uninitialized_var(mode_flags); nodemask_t nodes; |
095f1fc4e mempolicy: rework... |
1945 1946 1947 1948 1949 1950 1951 1952 |
char *nodelist = strchr(str, ':'); char *flags = strchr(str, '='); int i; int err = 1; if (nodelist) { /* NUL-terminate mode or flags string */ *nodelist++ = '\0'; |
71fe804b6 mempolicy: use st... |
1953 |
if (nodelist_parse(nodelist, nodes)) |
095f1fc4e mempolicy: rework... |
1954 |
goto out; |
71fe804b6 mempolicy: use st... |
1955 |
if (!nodes_subset(nodes, node_states[N_HIGH_MEMORY])) |
095f1fc4e mempolicy: rework... |
1956 |
goto out; |
71fe804b6 mempolicy: use st... |
1957 1958 |
} else nodes_clear(nodes); |
095f1fc4e mempolicy: rework... |
1959 1960 |
if (flags) *flags++ = '\0'; /* terminate mode string */ |
3f226aa1c mempolicy: suppor... |
1961 |
for (i = 0; i <= MPOL_LOCAL; i++) { |
095f1fc4e mempolicy: rework... |
1962 |
if (!strcmp(str, policy_types[i])) { |
71fe804b6 mempolicy: use st... |
1963 |
mode = i; |
095f1fc4e mempolicy: rework... |
1964 1965 1966 |
break; } } |
3f226aa1c mempolicy: suppor... |
1967 |
if (i > MPOL_LOCAL) |
095f1fc4e mempolicy: rework... |
1968 |
goto out; |
71fe804b6 mempolicy: use st... |
1969 |
switch (mode) { |
095f1fc4e mempolicy: rework... |
1970 |
case MPOL_PREFERRED: |
71fe804b6 mempolicy: use st... |
1971 1972 1973 |
/* * Insist on a nodelist of one node only */ |
095f1fc4e mempolicy: rework... |
1974 1975 1976 1977 1978 1979 1980 1981 |
if (nodelist) { char *rest = nodelist; while (isdigit(*rest)) rest++; if (!*rest) err = 0; } break; |
095f1fc4e mempolicy: rework... |
1982 1983 1984 1985 1986 |
case MPOL_INTERLEAVE: /* * Default to online nodes with memory if no nodelist */ if (!nodelist) |
71fe804b6 mempolicy: use st... |
1987 |
nodes = node_states[N_HIGH_MEMORY]; |
095f1fc4e mempolicy: rework... |
1988 |
err = 0; |
3f226aa1c mempolicy: suppor... |
1989 |
break; |
71fe804b6 mempolicy: use st... |
1990 |
case MPOL_LOCAL: |
3f226aa1c mempolicy: suppor... |
1991 |
/* |
71fe804b6 mempolicy: use st... |
1992 |
* Don't allow a nodelist; mpol_new() checks flags |
3f226aa1c mempolicy: suppor... |
1993 |
*/ |
71fe804b6 mempolicy: use st... |
1994 |
if (nodelist) |
3f226aa1c mempolicy: suppor... |
1995 |
goto out; |
71fe804b6 mempolicy: use st... |
1996 |
mode = MPOL_PREFERRED; |
3f226aa1c mempolicy: suppor... |
1997 |
break; |
71fe804b6 mempolicy: use st... |
1998 1999 2000 2001 2002 |
/* * case MPOL_BIND: mpol_new() enforces non-empty nodemask. * case MPOL_DEFAULT: mpol_new() enforces empty nodemask, ignores flags. */ |
095f1fc4e mempolicy: rework... |
2003 |
} |
71fe804b6 mempolicy: use st... |
2004 |
mode_flags = 0; |
095f1fc4e mempolicy: rework... |
2005 2006 2007 2008 2009 2010 |
if (flags) { /* * Currently, we only support two mutually exclusive * mode flags. */ if (!strcmp(flags, "static")) |
71fe804b6 mempolicy: use st... |
2011 |
mode_flags |= MPOL_F_STATIC_NODES; |
095f1fc4e mempolicy: rework... |
2012 |
else if (!strcmp(flags, "relative")) |
71fe804b6 mempolicy: use st... |
2013 |
mode_flags |= MPOL_F_RELATIVE_NODES; |
095f1fc4e mempolicy: rework... |
2014 2015 2016 |
else err = 1; } |
71fe804b6 mempolicy: use st... |
2017 2018 2019 2020 2021 2022 |
new = mpol_new(mode, mode_flags, &nodes); if (IS_ERR(new)) err = 1; else if (no_context) new->w.user_nodemask = nodes; /* save for contextualization */ |
095f1fc4e mempolicy: rework... |
2023 2024 2025 2026 2027 2028 |
out: /* Restore string for error message */ if (nodelist) *--nodelist = ':'; if (flags) *--flags = '='; |
71fe804b6 mempolicy: use st... |
2029 2030 |
if (!err) *mpol = new; |
095f1fc4e mempolicy: rework... |
2031 2032 2033 |
return err; } #endif /* CONFIG_TMPFS */ |
71fe804b6 mempolicy: use st... |
2034 2035 2036 2037 2038 2039 2040 |
/** * mpol_to_str - format a mempolicy structure for printing * @buffer: to contain formatted mempolicy string * @maxlen: length of @buffer * @pol: pointer to mempolicy to be formatted * @no_context: "context free" mempolicy - use nodemask in w.user_nodemask * |
1a75a6c82 [PATCH] Fold numa... |
2041 2042 2043 2044 |
* Convert a mempolicy into a string. * Returns the number of characters in buffer (if positive) * or an error (negative) */ |
71fe804b6 mempolicy: use st... |
2045 |
int mpol_to_str(char *buffer, int maxlen, struct mempolicy *pol, int no_context) |
1a75a6c82 [PATCH] Fold numa... |
2046 2047 2048 2049 |
{ char *p = buffer; int l; nodemask_t nodes; |
bea904d54 mempolicy: use MP... |
2050 |
unsigned short mode; |
f5b087b52 mempolicy: add MP... |
2051 |
unsigned short flags = pol ? pol->flags : 0; |
1a75a6c82 [PATCH] Fold numa... |
2052 |
|
2291990ab mempolicy: clean-... |
2053 2054 2055 2056 |
/* * Sanity check: room for longest mode, flag and some nodes */ VM_BUG_ON(maxlen < strlen("interleave") + strlen("relative") + 16); |
bea904d54 mempolicy: use MP... |
2057 2058 2059 2060 |
if (!pol || pol == &default_policy) mode = MPOL_DEFAULT; else mode = pol->mode; |
1a75a6c82 [PATCH] Fold numa... |
2061 2062 2063 2064 2065 2066 2067 |
switch (mode) { case MPOL_DEFAULT: nodes_clear(nodes); break; case MPOL_PREFERRED: nodes_clear(nodes); |
fc36b8d3d mempolicy: use MP... |
2068 |
if (flags & MPOL_F_LOCAL) |
53f2556b6 mempolicy: mPOL_P... |
2069 2070 |
mode = MPOL_LOCAL; /* pseudo-policy */ else |
fc36b8d3d mempolicy: use MP... |
2071 |
node_set(pol->v.preferred_node, nodes); |
1a75a6c82 [PATCH] Fold numa... |
2072 2073 2074 |
break; case MPOL_BIND: |
19770b326 mm: filter based ... |
2075 |
/* Fall through */ |
1a75a6c82 [PATCH] Fold numa... |
2076 |
case MPOL_INTERLEAVE: |
71fe804b6 mempolicy: use st... |
2077 2078 2079 2080 |
if (no_context) nodes = pol->w.user_nodemask; else nodes = pol->v.nodes; |
1a75a6c82 [PATCH] Fold numa... |
2081 2082 2083 2084 |
break; default: BUG(); |
1a75a6c82 [PATCH] Fold numa... |
2085 2086 2087 |
} l = strlen(policy_types[mode]); |
53f2556b6 mempolicy: mPOL_P... |
2088 2089 |
if (buffer + maxlen < p + l + 1) return -ENOSPC; |
1a75a6c82 [PATCH] Fold numa... |
2090 2091 2092 |
strcpy(p, policy_types[mode]); p += l; |
fc36b8d3d mempolicy: use MP... |
2093 |
if (flags & MPOL_MODE_FLAGS) { |
f5b087b52 mempolicy: add MP... |
2094 2095 2096 |
if (buffer + maxlen < p + 2) return -ENOSPC; *p++ = '='; |
2291990ab mempolicy: clean-... |
2097 2098 2099 |
/* * Currently, the only defined flags are mutually exclusive */ |
f5b087b52 mempolicy: add MP... |
2100 |
if (flags & MPOL_F_STATIC_NODES) |
2291990ab mempolicy: clean-... |
2101 2102 2103 |
p += snprintf(p, buffer + maxlen - p, "static"); else if (flags & MPOL_F_RELATIVE_NODES) p += snprintf(p, buffer + maxlen - p, "relative"); |
f5b087b52 mempolicy: add MP... |
2104 |
} |
1a75a6c82 [PATCH] Fold numa... |
2105 2106 2107 |
if (!nodes_empty(nodes)) { if (buffer + maxlen < p + 2) return -ENOSPC; |
095f1fc4e mempolicy: rework... |
2108 |
*p++ = ':'; |
1a75a6c82 [PATCH] Fold numa... |
2109 2110 2111 2112 2113 2114 2115 2116 |
p += nodelist_scnprintf(p, buffer + maxlen - p, nodes); } return p - buffer; } struct numa_maps { unsigned long pages; unsigned long anon; |
397874dfe [PATCH] numa_maps... |
2117 2118 |
unsigned long active; unsigned long writeback; |
1a75a6c82 [PATCH] Fold numa... |
2119 |
unsigned long mapcount_max; |
397874dfe [PATCH] numa_maps... |
2120 2121 |
unsigned long dirty; unsigned long swapcache; |
1a75a6c82 [PATCH] Fold numa... |
2122 2123 |
unsigned long node[MAX_NUMNODES]; }; |
397874dfe [PATCH] numa_maps... |
2124 |
static void gather_stats(struct page *page, void *private, int pte_dirty) |
1a75a6c82 [PATCH] Fold numa... |
2125 2126 2127 |
{ struct numa_maps *md = private; int count = page_mapcount(page); |
397874dfe [PATCH] numa_maps... |
2128 2129 2130 |
md->pages++; if (pte_dirty || PageDirty(page)) md->dirty++; |
1a75a6c82 [PATCH] Fold numa... |
2131 |
|
397874dfe [PATCH] numa_maps... |
2132 2133 |
if (PageSwapCache(page)) md->swapcache++; |
1a75a6c82 [PATCH] Fold numa... |
2134 |
|
894bc3104 Unevictable LRU I... |
2135 |
if (PageActive(page) || PageUnevictable(page)) |
397874dfe [PATCH] numa_maps... |
2136 2137 2138 2139 |
md->active++; if (PageWriteback(page)) md->writeback++; |
1a75a6c82 [PATCH] Fold numa... |
2140 2141 2142 |
if (PageAnon(page)) md->anon++; |
397874dfe [PATCH] numa_maps... |
2143 2144 |
if (count > md->mapcount_max) md->mapcount_max = count; |
1a75a6c82 [PATCH] Fold numa... |
2145 |
md->node[page_to_nid(page)]++; |
1a75a6c82 [PATCH] Fold numa... |
2146 |
} |
7f709ed0e [PATCH] numa_maps... |
2147 |
#ifdef CONFIG_HUGETLB_PAGE |
397874dfe [PATCH] numa_maps... |
2148 2149 2150 2151 2152 2153 |
static void check_huge_range(struct vm_area_struct *vma, unsigned long start, unsigned long end, struct numa_maps *md) { unsigned long addr; struct page *page; |
a55164389 hugetlb: modular ... |
2154 2155 |
struct hstate *h = hstate_vma(vma); unsigned long sz = huge_page_size(h); |
397874dfe [PATCH] numa_maps... |
2156 |
|
a55164389 hugetlb: modular ... |
2157 2158 2159 |
for (addr = start; addr < end; addr += sz) { pte_t *ptep = huge_pte_offset(vma->vm_mm, addr & huge_page_mask(h)); |
397874dfe [PATCH] numa_maps... |
2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 |
pte_t pte; if (!ptep) continue; pte = *ptep; if (pte_none(pte)) continue; page = pte_page(pte); if (!page) continue; gather_stats(page, md, pte_dirty(*ptep)); } } |
7f709ed0e [PATCH] numa_maps... |
2176 2177 2178 2179 2180 2181 2182 |
#else static inline void check_huge_range(struct vm_area_struct *vma, unsigned long start, unsigned long end, struct numa_maps *md) { } #endif |
397874dfe [PATCH] numa_maps... |
2183 |
|
53f2556b6 mempolicy: mPOL_P... |
2184 2185 2186 |
/* * Display pages allocated per node and memory policy via /proc. */ |
1a75a6c82 [PATCH] Fold numa... |
2187 2188 |
int show_numa_map(struct seq_file *m, void *v) { |
99f895518 [PATCH] proc: don... |
2189 |
struct proc_maps_private *priv = m->private; |
1a75a6c82 [PATCH] Fold numa... |
2190 2191 |
struct vm_area_struct *vma = v; struct numa_maps *md; |
397874dfe [PATCH] numa_maps... |
2192 2193 |
struct file *file = vma->vm_file; struct mm_struct *mm = vma->vm_mm; |
480eccf9a Fix NUMA Memory P... |
2194 |
struct mempolicy *pol; |
1a75a6c82 [PATCH] Fold numa... |
2195 2196 |
int n; char buffer[50]; |
397874dfe [PATCH] numa_maps... |
2197 |
if (!mm) |
1a75a6c82 [PATCH] Fold numa... |
2198 2199 2200 2201 2202 |
return 0; md = kzalloc(sizeof(struct numa_maps), GFP_KERNEL); if (!md) return 0; |
480eccf9a Fix NUMA Memory P... |
2203 |
pol = get_vma_policy(priv->task, vma, vma->vm_start); |
71fe804b6 mempolicy: use st... |
2204 |
mpol_to_str(buffer, sizeof(buffer), pol, 0); |
52cd3b074 mempolicy: rework... |
2205 |
mpol_cond_put(pol); |
397874dfe [PATCH] numa_maps... |
2206 2207 2208 2209 2210 |
seq_printf(m, "%08lx %s", vma->vm_start, buffer); if (file) { seq_printf(m, " file="); |
c32c2f63a d_path: Make seq_... |
2211 2212 |
seq_path(m, &file->f_path, " \t= "); |
397874dfe [PATCH] numa_maps... |
2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 |
} else if (vma->vm_start <= mm->brk && vma->vm_end >= mm->start_brk) { seq_printf(m, " heap"); } else if (vma->vm_start <= mm->start_stack && vma->vm_end >= mm->start_stack) { seq_printf(m, " stack"); } if (is_vm_hugetlb_page(vma)) { check_huge_range(vma, vma->vm_start, vma->vm_end, md); seq_printf(m, " huge"); } else { |
a57ebfdb2 [PATCH] numa_maps... |
2224 |
check_pgd_range(vma, vma->vm_start, vma->vm_end, |
56bbd65df Memoryless nodes:... |
2225 |
&node_states[N_HIGH_MEMORY], MPOL_MF_STATS, md); |
397874dfe [PATCH] numa_maps... |
2226 2227 2228 2229 |
} if (!md->pages) goto out; |
1a75a6c82 [PATCH] Fold numa... |
2230 |
|
397874dfe [PATCH] numa_maps... |
2231 2232 |
if (md->anon) seq_printf(m," anon=%lu",md->anon); |
1a75a6c82 [PATCH] Fold numa... |
2233 |
|
397874dfe [PATCH] numa_maps... |
2234 2235 |
if (md->dirty) seq_printf(m," dirty=%lu",md->dirty); |
1a75a6c82 [PATCH] Fold numa... |
2236 |
|
397874dfe [PATCH] numa_maps... |
2237 2238 |
if (md->pages != md->anon && md->pages != md->dirty) seq_printf(m, " mapped=%lu", md->pages); |
1a75a6c82 [PATCH] Fold numa... |
2239 |
|
397874dfe [PATCH] numa_maps... |
2240 2241 |
if (md->mapcount_max > 1) seq_printf(m, " mapmax=%lu", md->mapcount_max); |
1a75a6c82 [PATCH] Fold numa... |
2242 |
|
397874dfe [PATCH] numa_maps... |
2243 2244 2245 2246 2247 2248 2249 2250 |
if (md->swapcache) seq_printf(m," swapcache=%lu", md->swapcache); if (md->active < md->pages && !is_vm_hugetlb_page(vma)) seq_printf(m," active=%lu", md->active); if (md->writeback) seq_printf(m," writeback=%lu", md->writeback); |
56bbd65df Memoryless nodes:... |
2251 |
for_each_node_state(n, N_HIGH_MEMORY) |
397874dfe [PATCH] numa_maps... |
2252 2253 2254 2255 2256 |
if (md->node[n]) seq_printf(m, " N%d=%lu", n, md->node[n]); out: seq_putc(m, ' '); |
1a75a6c82 [PATCH] Fold numa... |
2257 2258 2259 |
kfree(md); if (m->count < m->size) |
99f895518 [PATCH] proc: don... |
2260 |
m->version = (vma != priv->tail_vma) ? vma->vm_start : 0; |
1a75a6c82 [PATCH] Fold numa... |
2261 2262 |
return 0; } |