Blame view
mm/mempolicy.c
71.6 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 |
* preferred Try a specific node first before normal fallback. |
00ef2d2f8 mm: use NUMA_NO_NODE |
29 |
* As a special case NUMA_NO_NODE here means do the allocation |
1da177e4c Linux-2.6.12-rc2 |
30 31 32 |
* 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 |
#include <linux/nodemask.h> #include <linux/cpuset.h> |
1da177e4c Linux-2.6.12-rc2 |
76 77 |
#include <linux/slab.h> #include <linux/string.h> |
b95f1b31b mm: Map most file... |
78 |
#include <linux/export.h> |
b488893a3 pid namespaces: c... |
79 |
#include <linux/nsproxy.h> |
1da177e4c Linux-2.6.12-rc2 |
80 81 82 |
#include <linux/interrupt.h> #include <linux/init.h> #include <linux/compat.h> |
dc9aa5b9d [PATCH] Swap Migr... |
83 |
#include <linux/swap.h> |
1a75a6c82 [PATCH] Fold numa... |
84 85 |
#include <linux/seq_file.h> #include <linux/proc_fs.h> |
b20a35035 [PATCH] page migr... |
86 |
#include <linux/migrate.h> |
62b61f611 ksm: memory hotre... |
87 |
#include <linux/ksm.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> |
6d9c285a6 mm: move inc_zone... |
92 |
#include <linux/mm_inline.h> |
b24f53a0b mm: mempolicy: Ad... |
93 |
#include <linux/mmu_notifier.h> |
dc9aa5b9d [PATCH] Swap Migr... |
94 |
|
1da177e4c Linux-2.6.12-rc2 |
95 96 |
#include <asm/tlbflush.h> #include <asm/uaccess.h> |
778d3b0ff cpusets: randomiz... |
97 |
#include <linux/random.h> |
1da177e4c Linux-2.6.12-rc2 |
98 |
|
62695a84e vmscan: move isol... |
99 |
#include "internal.h" |
38e35860d [PATCH] mempolici... |
100 |
/* Internal flags */ |
dc9aa5b9d [PATCH] Swap Migr... |
101 |
#define MPOL_MF_DISCONTIG_OK (MPOL_MF_INTERNAL << 0) /* Skip checks for continuous vmas */ |
38e35860d [PATCH] mempolici... |
102 |
#define MPOL_MF_INVERT (MPOL_MF_INTERNAL << 1) /* Invert check for nodemask */ |
dc9aa5b9d [PATCH] Swap Migr... |
103 |
|
fcc234f88 [PATCH] mm: kill ... |
104 105 |
static struct kmem_cache *policy_cache; static struct kmem_cache *sn_cache; |
1da177e4c Linux-2.6.12-rc2 |
106 |
|
1da177e4c Linux-2.6.12-rc2 |
107 108 |
/* Highest zone. An specific allocation for a zone below that is not policied. */ |
6267276f3 [PATCH] optional ... |
109 |
enum zone_type policy_zone = 0; |
1da177e4c Linux-2.6.12-rc2 |
110 |
|
bea904d54 mempolicy: use MP... |
111 112 113 |
/* * run-time system-wide default policy => local allocation */ |
e754d79d3 mm/mempolicy.c: q... |
114 |
static struct mempolicy default_policy = { |
1da177e4c Linux-2.6.12-rc2 |
115 |
.refcnt = ATOMIC_INIT(1), /* never free it */ |
bea904d54 mempolicy: use MP... |
116 |
.mode = MPOL_PREFERRED, |
fc36b8d3d mempolicy: use MP... |
117 |
.flags = MPOL_F_LOCAL, |
1da177e4c Linux-2.6.12-rc2 |
118 |
}; |
5606e3877 mm: numa: Migrate... |
119 120 121 122 123 |
static struct mempolicy preferred_node_policy[MAX_NUMNODES]; static struct mempolicy *get_task_policy(struct task_struct *p) { struct mempolicy *pol = p->mempolicy; |
5606e3877 mm: numa: Migrate... |
124 125 |
if (!pol) { |
1da6f0e1b mm/mempolicy: ret... |
126 |
int node = numa_node_id(); |
5606e3877 mm: numa: Migrate... |
127 |
|
1da6f0e1b mm/mempolicy: ret... |
128 129 130 131 132 133 134 135 136 |
if (node != NUMA_NO_NODE) { pol = &preferred_node_policy[node]; /* * preferred_node_policy is not initialised early in * boot */ if (!pol->mode) pol = NULL; } |
5606e3877 mm: numa: Migrate... |
137 138 139 140 |
} return pol; } |
37012946d mempolicy: create... |
141 142 |
static const struct mempolicy_operations { int (*create)(struct mempolicy *pol, const nodemask_t *nodes); |
708c1bbc9 mempolicy: restru... |
143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 |
/* * If read-side task has no lock to protect task->mempolicy, write-side * task will rebind the task->mempolicy by two step. The first step is * setting all the newly nodes, and the second step is cleaning all the * disallowed nodes. In this way, we can avoid finding no node to alloc * page. * If we have a lock to protect task->mempolicy in read-side, we do * rebind directly. * * step: * MPOL_REBIND_ONCE - do rebind work at once * MPOL_REBIND_STEP1 - set all the newly nodes * MPOL_REBIND_STEP2 - clean all the disallowed nodes */ void (*rebind)(struct mempolicy *pol, const nodemask_t *nodes, enum mpol_rebind_step step); |
37012946d mempolicy: create... |
159 |
} mpol_ops[MPOL_MAX]; |
19770b326 mm: filter based ... |
160 |
/* Check that the nodemask contains at least one populated zone */ |
37012946d mempolicy: create... |
161 |
static int is_valid_nodemask(const nodemask_t *nodemask) |
1da177e4c Linux-2.6.12-rc2 |
162 |
{ |
d3eb1570a mempolicy: fix is... |
163 |
return nodes_intersects(*nodemask, node_states[N_MEMORY]); |
1da177e4c Linux-2.6.12-rc2 |
164 |
} |
f5b087b52 mempolicy: add MP... |
165 166 |
static inline int mpol_store_user_nodemask(const struct mempolicy *pol) { |
6d556294d mempolicy: remove... |
167 |
return pol->flags & MPOL_MODE_FLAGS; |
4c50bc011 mempolicy: add MP... |
168 169 170 171 172 173 174 175 |
} 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... |
176 |
} |
37012946d mempolicy: create... |
177 178 179 180 181 182 183 184 185 186 187 |
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... |
188 |
pol->flags |= MPOL_F_LOCAL; /* local allocation */ |
37012946d mempolicy: create... |
189 190 191 192 193 194 195 196 197 198 199 200 201 202 |
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; } |
58568d2a8 cpuset,mm: update... |
203 204 205 206 207 208 209 210 211 |
/* * mpol_set_nodemask is called after mpol_new() to set up the nodemask, if * any, for the new policy. mpol_new() has already validated the nodes * parameter with respect to the policy mode and flags. But, we need to * handle an empty nodemask with MPOL_PREFERRED here. * * Must be called holding task's alloc_lock to protect task's mems_allowed * and mempolicy. May also be called holding the mmap_semaphore for write. */ |
4bfc44958 mm: make set_memp... |
212 213 |
static int mpol_set_nodemask(struct mempolicy *pol, const nodemask_t *nodes, struct nodemask_scratch *nsc) |
58568d2a8 cpuset,mm: update... |
214 |
{ |
58568d2a8 cpuset,mm: update... |
215 216 217 218 219 |
int ret; /* if mode is MPOL_DEFAULT, pol is NULL. This is right. */ if (pol == NULL) return 0; |
01f13bd60 mempolicy: use N_... |
220 |
/* Check N_MEMORY */ |
4bfc44958 mm: make set_memp... |
221 |
nodes_and(nsc->mask1, |
01f13bd60 mempolicy: use N_... |
222 |
cpuset_current_mems_allowed, node_states[N_MEMORY]); |
58568d2a8 cpuset,mm: update... |
223 224 225 226 227 228 |
VM_BUG_ON(!nodes); if (pol->mode == MPOL_PREFERRED && nodes_empty(*nodes)) nodes = NULL; /* explicit local allocation */ else { if (pol->flags & MPOL_F_RELATIVE_NODES) |
4bfc44958 mm: make set_memp... |
229 |
mpol_relative_nodemask(&nsc->mask2, nodes,&nsc->mask1); |
58568d2a8 cpuset,mm: update... |
230 |
else |
4bfc44958 mm: make set_memp... |
231 |
nodes_and(nsc->mask2, *nodes, nsc->mask1); |
58568d2a8 cpuset,mm: update... |
232 233 234 235 236 237 |
if (mpol_store_user_nodemask(pol)) pol->w.user_nodemask = *nodes; else pol->w.cpuset_mems_allowed = cpuset_current_mems_allowed; } |
4bfc44958 mm: make set_memp... |
238 239 240 241 |
if (nodes) ret = mpol_ops[pol->mode].create(pol, &nsc->mask2); else ret = mpol_ops[pol->mode].create(pol, NULL); |
58568d2a8 cpuset,mm: update... |
242 243 244 245 246 247 248 |
return ret; } /* * This function just creates a new policy, does some check and simple * initialization. You must invoke mpol_set_nodemask() to set nodes. */ |
028fec414 mempolicy: suppor... |
249 250 |
static struct mempolicy *mpol_new(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
251 252 |
{ struct mempolicy *policy; |
028fec414 mempolicy: suppor... |
253 254 |
pr_debug("setting mode %d flags %d nodes[0] %lx ", |
00ef2d2f8 mm: use NUMA_NO_NODE |
255 |
mode, flags, nodes ? nodes_addr(*nodes)[0] : NUMA_NO_NODE); |
140d5a490 numa: mempolicy: ... |
256 |
|
3e1f06456 mempolicy: disall... |
257 258 |
if (mode == MPOL_DEFAULT) { if (nodes && !nodes_empty(*nodes)) |
37012946d mempolicy: create... |
259 |
return ERR_PTR(-EINVAL); |
d3a710337 mm: mempolicy: Ad... |
260 |
return NULL; |
37012946d mempolicy: create... |
261 |
} |
3e1f06456 mempolicy: disall... |
262 263 264 265 266 267 268 269 270 271 272 273 |
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); |
3e1f06456 mempolicy: disall... |
274 |
} |
479e2802d mm: mempolicy: Ma... |
275 276 277 278 |
} else if (mode == MPOL_LOCAL) { if (!nodes_empty(*nodes)) return ERR_PTR(-EINVAL); mode = MPOL_PREFERRED; |
3e1f06456 mempolicy: disall... |
279 280 |
} else if (nodes_empty(*nodes)) return ERR_PTR(-EINVAL); |
1da177e4c Linux-2.6.12-rc2 |
281 282 283 284 |
policy = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!policy) return ERR_PTR(-ENOMEM); atomic_set(&policy->refcnt, 1); |
45c4745af mempolicy: rename... |
285 |
policy->mode = mode; |
3e1f06456 mempolicy: disall... |
286 |
policy->flags = flags; |
37012946d mempolicy: create... |
287 |
|
1da177e4c Linux-2.6.12-rc2 |
288 |
return policy; |
37012946d mempolicy: create... |
289 |
} |
52cd3b074 mempolicy: rework... |
290 291 292 293 294 |
/* Slow path of a mpol destructor. */ void __mpol_put(struct mempolicy *p) { if (!atomic_dec_and_test(&p->refcnt)) return; |
52cd3b074 mempolicy: rework... |
295 296 |
kmem_cache_free(policy_cache, p); } |
708c1bbc9 mempolicy: restru... |
297 298 |
static void mpol_rebind_default(struct mempolicy *pol, const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
299 300 |
{ } |
708c1bbc9 mempolicy: restru... |
301 302 303 304 305 306 307 308 |
/* * step: * MPOL_REBIND_ONCE - do rebind work at once * MPOL_REBIND_STEP1 - set all the newly nodes * MPOL_REBIND_STEP2 - clean all the disallowed nodes */ static void mpol_rebind_nodemask(struct mempolicy *pol, const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
309 310 311 312 313 314 315 316 |
{ 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 { |
708c1bbc9 mempolicy: restru... |
317 318 319 320 321 322 323 324 325 326 327 328 329 |
/* * if step == 1, we use ->w.cpuset_mems_allowed to cache the * result */ if (step == MPOL_REBIND_ONCE || step == MPOL_REBIND_STEP1) { nodes_remap(tmp, pol->v.nodes, pol->w.cpuset_mems_allowed, *nodes); pol->w.cpuset_mems_allowed = step ? tmp : *nodes; } else if (step == MPOL_REBIND_STEP2) { tmp = pol->w.cpuset_mems_allowed; pol->w.cpuset_mems_allowed = *nodes; } else BUG(); |
37012946d mempolicy: create... |
330 |
} |
f5b087b52 mempolicy: add MP... |
331 |
|
708c1bbc9 mempolicy: restru... |
332 333 334 335 336 337 338 339 340 |
if (nodes_empty(tmp)) tmp = *nodes; if (step == MPOL_REBIND_STEP1) nodes_or(pol->v.nodes, pol->v.nodes, tmp); else if (step == MPOL_REBIND_ONCE || step == MPOL_REBIND_STEP2) pol->v.nodes = tmp; else BUG(); |
37012946d mempolicy: create... |
341 342 343 344 345 346 347 348 349 350 |
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, |
708c1bbc9 mempolicy: restru... |
351 352 |
const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
353 354 |
{ nodemask_t tmp; |
37012946d mempolicy: create... |
355 356 |
if (pol->flags & MPOL_F_STATIC_NODES) { int node = first_node(pol->w.user_nodemask); |
fc36b8d3d mempolicy: use MP... |
357 |
if (node_isset(node, *nodes)) { |
37012946d mempolicy: create... |
358 |
pol->v.preferred_node = node; |
fc36b8d3d mempolicy: use MP... |
359 360 361 |
pol->flags &= ~MPOL_F_LOCAL; } else pol->flags |= MPOL_F_LOCAL; |
37012946d mempolicy: create... |
362 363 364 |
} 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... |
365 |
} else if (!(pol->flags & MPOL_F_LOCAL)) { |
37012946d mempolicy: create... |
366 367 368 369 370 |
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 |
371 |
} |
708c1bbc9 mempolicy: restru... |
372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 |
/* * mpol_rebind_policy - Migrate a policy to a different set of nodes * * If read-side task has no lock to protect task->mempolicy, write-side * task will rebind the task->mempolicy by two step. The first step is * setting all the newly nodes, and the second step is cleaning all the * disallowed nodes. In this way, we can avoid finding no node to alloc * page. * If we have a lock to protect task->mempolicy in read-side, we do * rebind directly. * * step: * MPOL_REBIND_ONCE - do rebind work at once * MPOL_REBIND_STEP1 - set all the newly nodes * MPOL_REBIND_STEP2 - clean all the disallowed nodes */ static void mpol_rebind_policy(struct mempolicy *pol, const nodemask_t *newmask, enum mpol_rebind_step step) |
1d0d2680a mempolicy: move r... |
390 |
{ |
1d0d2680a mempolicy: move r... |
391 392 |
if (!pol) return; |
89c522c78 mm/mempolicy.c: u... |
393 |
if (!mpol_store_user_nodemask(pol) && step == MPOL_REBIND_ONCE && |
1d0d2680a mempolicy: move r... |
394 395 |
nodes_equal(pol->w.cpuset_mems_allowed, *newmask)) return; |
708c1bbc9 mempolicy: restru... |
396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 |
if (step == MPOL_REBIND_STEP1 && (pol->flags & MPOL_F_REBINDING)) return; if (step == MPOL_REBIND_STEP2 && !(pol->flags & MPOL_F_REBINDING)) BUG(); if (step == MPOL_REBIND_STEP1) pol->flags |= MPOL_F_REBINDING; else if (step == MPOL_REBIND_STEP2) pol->flags &= ~MPOL_F_REBINDING; else if (step >= MPOL_REBIND_NSTEP) BUG(); mpol_ops[pol->mode].rebind(pol, newmask, step); |
1d0d2680a mempolicy: move r... |
411 412 413 414 415 |
} /* * Wrapper for mpol_rebind_policy() that just requires task * pointer, and updates task mempolicy. |
58568d2a8 cpuset,mm: update... |
416 417 |
* * Called with task's alloc_lock held. |
1d0d2680a mempolicy: move r... |
418 |
*/ |
708c1bbc9 mempolicy: restru... |
419 420 |
void mpol_rebind_task(struct task_struct *tsk, const nodemask_t *new, enum mpol_rebind_step step) |
1d0d2680a mempolicy: move r... |
421 |
{ |
708c1bbc9 mempolicy: restru... |
422 |
mpol_rebind_policy(tsk->mempolicy, new, step); |
1d0d2680a mempolicy: move r... |
423 424 425 426 427 428 429 430 431 432 433 434 435 436 |
} /* * 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) |
708c1bbc9 mempolicy: restru... |
437 |
mpol_rebind_policy(vma->vm_policy, new, MPOL_REBIND_ONCE); |
1d0d2680a mempolicy: move r... |
438 439 |
up_write(&mm->mmap_sem); } |
37012946d mempolicy: create... |
440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 |
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, }, }; |
fc3012896 [PATCH] Simplify ... |
457 458 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags); |
1a75a6c82 [PATCH] Fold numa... |
459 |
|
980949457 mm/mempolicy: ren... |
460 461 462 463 464 |
/* * Scan through pages checking if pages follow certain conditions, * and move them to the pagelist if they do. */ static int queue_pages_pte_range(struct vm_area_struct *vma, pmd_t *pmd, |
dc9aa5b9d [PATCH] Swap Migr... |
465 466 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
467 |
void *private) |
1da177e4c Linux-2.6.12-rc2 |
468 |
{ |
91612e0df [PATCH] mbind: ch... |
469 470 |
pte_t *orig_pte; pte_t *pte; |
705e87c0c [PATCH] mm: pte_o... |
471 |
spinlock_t *ptl; |
941150a32 [PATCH] mbind: fi... |
472 |
|
705e87c0c [PATCH] mm: pte_o... |
473 |
orig_pte = pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); |
91612e0df [PATCH] mbind: ch... |
474 |
do { |
6aab341e0 mm: re-architect ... |
475 |
struct page *page; |
25ba77c14 [PATCH] numa node... |
476 |
int nid; |
91612e0df [PATCH] mbind: ch... |
477 478 |
if (!pte_present(*pte)) |
1da177e4c Linux-2.6.12-rc2 |
479 |
continue; |
6aab341e0 mm: re-architect ... |
480 481 |
page = vm_normal_page(vma, addr, *pte); if (!page) |
1da177e4c Linux-2.6.12-rc2 |
482 |
continue; |
053837fce [PATCH] mm: migra... |
483 |
/* |
62b61f611 ksm: memory hotre... |
484 485 |
* vm_normal_page() filters out zero pages, but there might * still be PageReserved pages to skip, perhaps in a VDSO. |
053837fce [PATCH] mm: migra... |
486 |
*/ |
b79bc0a0c ksm: enable KSM p... |
487 |
if (PageReserved(page)) |
f4598c8b3 [PATCH] migration... |
488 |
continue; |
6aab341e0 mm: re-architect ... |
489 |
nid = page_to_nid(page); |
38e35860d [PATCH] mempolici... |
490 491 |
if (node_isset(nid, *nodes) == !!(flags & MPOL_MF_INVERT)) continue; |
b1f72d185 mm: remove MPOL_M... |
492 |
if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) |
fc3012896 [PATCH] Simplify ... |
493 |
migrate_page_add(page, private, flags); |
38e35860d [PATCH] mempolici... |
494 495 |
else break; |
91612e0df [PATCH] mbind: ch... |
496 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
705e87c0c [PATCH] mm: pte_o... |
497 |
pte_unmap_unlock(orig_pte, ptl); |
91612e0df [PATCH] mbind: ch... |
498 499 |
return addr != end; } |
980949457 mm/mempolicy: ren... |
500 501 |
static void queue_pages_hugetlb_pmd_range(struct vm_area_struct *vma, pmd_t *pmd, const nodemask_t *nodes, unsigned long flags, |
e2d8cf405 migrate: add huge... |
502 503 504 505 506 |
void *private) { #ifdef CONFIG_HUGETLB_PAGE int nid; struct page *page; |
cb900f412 mm, hugetlb: conv... |
507 |
spinlock_t *ptl; |
e2d8cf405 migrate: add huge... |
508 |
|
cb900f412 mm, hugetlb: conv... |
509 |
ptl = huge_pte_lock(hstate_vma(vma), vma->vm_mm, (pte_t *)pmd); |
e2d8cf405 migrate: add huge... |
510 511 512 513 514 515 516 517 518 |
page = pte_page(huge_ptep_get((pte_t *)pmd)); nid = page_to_nid(page); if (node_isset(nid, *nodes) == !!(flags & MPOL_MF_INVERT)) goto unlock; /* With MPOL_MF_MOVE, we migrate only unshared hugepage. */ if (flags & (MPOL_MF_MOVE_ALL) || (flags & MPOL_MF_MOVE && page_mapcount(page) == 1)) isolate_huge_page(page, private); unlock: |
cb900f412 mm, hugetlb: conv... |
519 |
spin_unlock(ptl); |
e2d8cf405 migrate: add huge... |
520 521 522 523 |
#else BUG(); #endif } |
980949457 mm/mempolicy: ren... |
524 |
static inline int queue_pages_pmd_range(struct vm_area_struct *vma, pud_t *pud, |
dc9aa5b9d [PATCH] Swap Migr... |
525 526 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
527 |
void *private) |
91612e0df [PATCH] mbind: ch... |
528 529 530 531 532 533 534 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
e2d8cf405 migrate: add huge... |
535 536 537 |
if (!pmd_present(*pmd)) continue; if (pmd_huge(*pmd) && is_vm_hugetlb_page(vma)) { |
980949457 mm/mempolicy: ren... |
538 |
queue_pages_hugetlb_pmd_range(vma, pmd, nodes, |
e2d8cf405 migrate: add huge... |
539 540 541 |
flags, private); continue; } |
e180377f1 thp: change split... |
542 |
split_huge_page_pmd(vma, addr, pmd); |
1a5a9906d mm: thp: fix pmd_... |
543 |
if (pmd_none_or_trans_huge_or_clear_bad(pmd)) |
91612e0df [PATCH] mbind: ch... |
544 |
continue; |
980949457 mm/mempolicy: ren... |
545 |
if (queue_pages_pte_range(vma, pmd, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
546 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
547 548 549 550 |
return -EIO; } while (pmd++, addr = next, addr != end); return 0; } |
980949457 mm/mempolicy: ren... |
551 |
static inline int queue_pages_pud_range(struct vm_area_struct *vma, pgd_t *pgd, |
dc9aa5b9d [PATCH] Swap Migr... |
552 553 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
554 |
void *private) |
91612e0df [PATCH] mbind: ch... |
555 556 557 558 559 560 561 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); |
e2d8cf405 migrate: add huge... |
562 563 |
if (pud_huge(*pud) && is_vm_hugetlb_page(vma)) continue; |
91612e0df [PATCH] mbind: ch... |
564 565 |
if (pud_none_or_clear_bad(pud)) continue; |
980949457 mm/mempolicy: ren... |
566 |
if (queue_pages_pmd_range(vma, pud, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
567 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
568 569 570 571 |
return -EIO; } while (pud++, addr = next, addr != end); return 0; } |
980949457 mm/mempolicy: ren... |
572 |
static inline int queue_pages_pgd_range(struct vm_area_struct *vma, |
dc9aa5b9d [PATCH] Swap Migr... |
573 574 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
575 |
void *private) |
91612e0df [PATCH] mbind: ch... |
576 577 578 |
{ pgd_t *pgd; unsigned long next; |
b5810039a [PATCH] core remo... |
579 |
pgd = pgd_offset(vma->vm_mm, addr); |
91612e0df [PATCH] mbind: ch... |
580 581 582 583 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
980949457 mm/mempolicy: ren... |
584 |
if (queue_pages_pud_range(vma, pgd, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
585 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
586 587 588 |
return -EIO; } while (pgd++, addr = next, addr != end); return 0; |
1da177e4c Linux-2.6.12-rc2 |
589 |
} |
5877231f6 mm: Move change_p... |
590 |
#ifdef CONFIG_NUMA_BALANCING |
b24f53a0b mm: mempolicy: Ad... |
591 |
/* |
4b10e7d56 mm: mempolicy: Im... |
592 593 594 595 596 597 598 |
* This is used to mark a range of virtual addresses to be inaccessible. * These are later cleared by a NUMA hinting fault. Depending on these * faults, pages may be migrated for better NUMA placement. * * This is assuming that NUMA faults are handled using PROT_NONE. If * an architecture makes a different choice, it will need further * changes to the core. |
b24f53a0b mm: mempolicy: Ad... |
599 |
*/ |
4b10e7d56 mm: mempolicy: Im... |
600 601 |
unsigned long change_prot_numa(struct vm_area_struct *vma, unsigned long addr, unsigned long end) |
b24f53a0b mm: mempolicy: Ad... |
602 |
{ |
4b10e7d56 mm: mempolicy: Im... |
603 |
int nr_updated; |
b24f53a0b mm: mempolicy: Ad... |
604 |
|
4b10e7d56 mm: mempolicy: Im... |
605 |
nr_updated = change_protection(vma, addr, end, vma->vm_page_prot, 0, 1); |
03c5a6e16 mm: numa: Add pte... |
606 607 |
if (nr_updated) count_vm_numa_events(NUMA_PTE_UPDATES, nr_updated); |
b24f53a0b mm: mempolicy: Ad... |
608 |
|
4b10e7d56 mm: mempolicy: Im... |
609 |
return nr_updated; |
b24f53a0b mm: mempolicy: Ad... |
610 611 612 613 614 615 616 |
} #else static unsigned long change_prot_numa(struct vm_area_struct *vma, unsigned long addr, unsigned long end) { return 0; } |
5877231f6 mm: Move change_p... |
617 |
#endif /* CONFIG_NUMA_BALANCING */ |
b24f53a0b mm: mempolicy: Ad... |
618 |
|
dc9aa5b9d [PATCH] Swap Migr... |
619 |
/* |
980949457 mm/mempolicy: ren... |
620 621 622 623 624 |
* Walk through page tables and collect pages to be migrated. * * If pages found in a given range are on a set of nodes (determined by * @nodes and @flags,) it's isolated and queued to the pagelist which is * passed via @private.) |
dc9aa5b9d [PATCH] Swap Migr... |
625 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
626 |
static struct vm_area_struct * |
980949457 mm/mempolicy: ren... |
627 |
queue_pages_range(struct mm_struct *mm, unsigned long start, unsigned long end, |
38e35860d [PATCH] mempolici... |
628 |
const nodemask_t *nodes, unsigned long flags, void *private) |
1da177e4c Linux-2.6.12-rc2 |
629 630 631 |
{ int err; struct vm_area_struct *first, *vma, *prev; |
053837fce [PATCH] mm: migra... |
632 |
|
1da177e4c Linux-2.6.12-rc2 |
633 634 635 636 637 |
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) { |
b24f53a0b mm: mempolicy: Ad... |
638 639 640 641 642 643 |
unsigned long endvma = vma->vm_end; if (endvma > end) endvma = end; if (vma->vm_start > start) start = vma->vm_start; |
dc9aa5b9d [PATCH] Swap Migr... |
644 645 646 647 648 649 |
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); } |
b24f53a0b mm: mempolicy: Ad... |
650 |
|
b24f53a0b mm: mempolicy: Ad... |
651 652 653 654 655 656 |
if (flags & MPOL_MF_LAZY) { change_prot_numa(vma, start, endvma); goto next; } if ((flags & MPOL_MF_STRICT) || |
dc9aa5b9d [PATCH] Swap Migr... |
657 |
((flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) && |
b24f53a0b mm: mempolicy: Ad... |
658 |
vma_migratable(vma))) { |
dc9aa5b9d [PATCH] Swap Migr... |
659 |
|
980949457 mm/mempolicy: ren... |
660 |
err = queue_pages_pgd_range(vma, start, endvma, nodes, |
38e35860d [PATCH] mempolici... |
661 |
flags, private); |
1da177e4c Linux-2.6.12-rc2 |
662 663 664 665 666 |
if (err) { first = ERR_PTR(err); break; } } |
b24f53a0b mm: mempolicy: Ad... |
667 |
next: |
1da177e4c Linux-2.6.12-rc2 |
668 669 670 671 |
prev = vma; } return first; } |
869833f2c mempolicy: remove... |
672 673 674 675 676 677 |
/* * Apply policy to a single VMA * This must be called with the mmap_sem held for writing. */ static int vma_replace_policy(struct vm_area_struct *vma, struct mempolicy *pol) |
8d34694c1 revert "mm: mempo... |
678 |
{ |
869833f2c mempolicy: remove... |
679 680 681 |
int err; struct mempolicy *old; struct mempolicy *new; |
8d34694c1 revert "mm: mempo... |
682 683 684 685 686 687 |
pr_debug("vma %lx-%lx/%lx vm_ops %p vm_file %p set_policy %p ", vma->vm_start, vma->vm_end, vma->vm_pgoff, vma->vm_ops, vma->vm_file, vma->vm_ops ? vma->vm_ops->set_policy : NULL); |
869833f2c mempolicy: remove... |
688 689 690 691 692 |
new = mpol_dup(pol); if (IS_ERR(new)) return PTR_ERR(new); if (vma->vm_ops && vma->vm_ops->set_policy) { |
8d34694c1 revert "mm: mempo... |
693 |
err = vma->vm_ops->set_policy(vma, new); |
869833f2c mempolicy: remove... |
694 695 |
if (err) goto err_out; |
8d34694c1 revert "mm: mempo... |
696 |
} |
869833f2c mempolicy: remove... |
697 698 699 700 701 702 703 704 |
old = vma->vm_policy; vma->vm_policy = new; /* protected by mmap_sem */ mpol_put(old); return 0; err_out: mpol_put(new); |
8d34694c1 revert "mm: mempo... |
705 706 |
return err; } |
1da177e4c Linux-2.6.12-rc2 |
707 |
/* Step 2: apply policy to a range and do splits. */ |
9d8cebd4b mm: fix mbind vma... |
708 709 |
static int mbind_range(struct mm_struct *mm, unsigned long start, unsigned long end, struct mempolicy *new_pol) |
1da177e4c Linux-2.6.12-rc2 |
710 711 |
{ struct vm_area_struct *next; |
9d8cebd4b mm: fix mbind vma... |
712 713 714 |
struct vm_area_struct *prev; struct vm_area_struct *vma; int err = 0; |
e26a51148 mm/mempolicy.c: r... |
715 |
pgoff_t pgoff; |
9d8cebd4b mm: fix mbind vma... |
716 717 |
unsigned long vmstart; unsigned long vmend; |
1da177e4c Linux-2.6.12-rc2 |
718 |
|
097d59106 vm: avoid using f... |
719 |
vma = find_vma(mm, start); |
9d8cebd4b mm: fix mbind vma... |
720 721 |
if (!vma || vma->vm_start > start) return -EFAULT; |
097d59106 vm: avoid using f... |
722 |
prev = vma->vm_prev; |
e26a51148 mm/mempolicy.c: r... |
723 724 |
if (start > vma->vm_start) prev = vma; |
9d8cebd4b mm: fix mbind vma... |
725 |
for (; vma && vma->vm_start < end; prev = vma, vma = next) { |
1da177e4c Linux-2.6.12-rc2 |
726 |
next = vma->vm_next; |
9d8cebd4b mm: fix mbind vma... |
727 728 |
vmstart = max(start, vma->vm_start); vmend = min(end, vma->vm_end); |
e26a51148 mm/mempolicy.c: r... |
729 730 731 732 733 |
if (mpol_equal(vma_policy(vma), new_pol)) continue; pgoff = vma->vm_pgoff + ((vmstart - vma->vm_start) >> PAGE_SHIFT); |
9d8cebd4b mm: fix mbind vma... |
734 |
prev = vma_merge(mm, prev, vmstart, vmend, vma->vm_flags, |
e26a51148 mm/mempolicy.c: r... |
735 |
vma->anon_vma, vma->vm_file, pgoff, |
8aacc9f55 mm/mempolicy.c: f... |
736 |
new_pol); |
9d8cebd4b mm: fix mbind vma... |
737 738 739 |
if (prev) { vma = prev; next = vma->vm_next; |
3964acd0d mm: mempolicy: fi... |
740 741 742 743 |
if (mpol_equal(vma_policy(vma), new_pol)) continue; /* vma_merge() joined vma && vma->next, case 8 */ goto replace; |
9d8cebd4b mm: fix mbind vma... |
744 745 746 747 748 749 750 751 752 753 754 |
} if (vma->vm_start != vmstart) { err = split_vma(vma->vm_mm, vma, vmstart, 1); if (err) goto out; } if (vma->vm_end != vmend) { err = split_vma(vma->vm_mm, vma, vmend, 0); if (err) goto out; } |
3964acd0d mm: mempolicy: fi... |
755 |
replace: |
869833f2c mempolicy: remove... |
756 |
err = vma_replace_policy(vma, new_pol); |
8d34694c1 revert "mm: mempo... |
757 758 |
if (err) goto out; |
1da177e4c Linux-2.6.12-rc2 |
759 |
} |
9d8cebd4b mm: fix mbind vma... |
760 761 |
out: |
1da177e4c Linux-2.6.12-rc2 |
762 763 |
return err; } |
1da177e4c Linux-2.6.12-rc2 |
764 |
/* Set the process memory policy */ |
028fec414 mempolicy: suppor... |
765 766 |
static long do_set_mempolicy(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
767 |
{ |
58568d2a8 cpuset,mm: update... |
768 |
struct mempolicy *new, *old; |
f4e53d910 mempolicy: write ... |
769 |
struct mm_struct *mm = current->mm; |
4bfc44958 mm: make set_memp... |
770 |
NODEMASK_SCRATCH(scratch); |
58568d2a8 cpuset,mm: update... |
771 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
772 |
|
4bfc44958 mm: make set_memp... |
773 774 |
if (!scratch) return -ENOMEM; |
f4e53d910 mempolicy: write ... |
775 |
|
4bfc44958 mm: make set_memp... |
776 777 778 779 780 |
new = mpol_new(mode, flags, nodes); if (IS_ERR(new)) { ret = PTR_ERR(new); goto out; } |
f4e53d910 mempolicy: write ... |
781 782 783 784 785 786 787 788 |
/* * 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); |
58568d2a8 cpuset,mm: update... |
789 |
task_lock(current); |
4bfc44958 mm: make set_memp... |
790 |
ret = mpol_set_nodemask(new, nodes, scratch); |
58568d2a8 cpuset,mm: update... |
791 792 793 794 795 |
if (ret) { task_unlock(current); if (mm) up_write(&mm->mmap_sem); mpol_put(new); |
4bfc44958 mm: make set_memp... |
796 |
goto out; |
58568d2a8 cpuset,mm: update... |
797 798 |
} old = current->mempolicy; |
1da177e4c Linux-2.6.12-rc2 |
799 |
current->mempolicy = new; |
45c4745af mempolicy: rename... |
800 |
if (new && new->mode == MPOL_INTERLEAVE && |
f5b087b52 mempolicy: add MP... |
801 |
nodes_weight(new->v.nodes)) |
dfcd3c0dc [PATCH] Convert m... |
802 |
current->il_next = first_node(new->v.nodes); |
58568d2a8 cpuset,mm: update... |
803 |
task_unlock(current); |
f4e53d910 mempolicy: write ... |
804 805 |
if (mm) up_write(&mm->mmap_sem); |
58568d2a8 cpuset,mm: update... |
806 |
mpol_put(old); |
4bfc44958 mm: make set_memp... |
807 808 809 810 |
ret = 0; out: NODEMASK_SCRATCH_FREE(scratch); return ret; |
1da177e4c Linux-2.6.12-rc2 |
811 |
} |
bea904d54 mempolicy: use MP... |
812 813 |
/* * Return nodemask for policy for get_mempolicy() query |
58568d2a8 cpuset,mm: update... |
814 815 |
* * Called with task's alloc_lock held |
bea904d54 mempolicy: use MP... |
816 817 |
*/ static void get_policy_nodemask(struct mempolicy *p, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
818 |
{ |
dfcd3c0dc [PATCH] Convert m... |
819 |
nodes_clear(*nodes); |
bea904d54 mempolicy: use MP... |
820 821 |
if (p == &default_policy) return; |
45c4745af mempolicy: rename... |
822 |
switch (p->mode) { |
19770b326 mm: filter based ... |
823 824 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
825 |
case MPOL_INTERLEAVE: |
dfcd3c0dc [PATCH] Convert m... |
826 |
*nodes = p->v.nodes; |
1da177e4c Linux-2.6.12-rc2 |
827 828 |
break; case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
829 |
if (!(p->flags & MPOL_F_LOCAL)) |
dfcd3c0dc [PATCH] Convert m... |
830 |
node_set(p->v.preferred_node, *nodes); |
53f2556b6 mempolicy: mPOL_P... |
831 |
/* else return empty node mask for local allocation */ |
1da177e4c Linux-2.6.12-rc2 |
832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 |
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 |
850 |
/* Retrieve NUMA policy */ |
dbcb0f19c mm/mempolicy.c: c... |
851 852 |
static long do_get_mempolicy(int *policy, nodemask_t *nmask, unsigned long addr, unsigned long flags) |
1da177e4c Linux-2.6.12-rc2 |
853 |
{ |
8bccd85ff [PATCH] Implement... |
854 |
int err; |
1da177e4c Linux-2.6.12-rc2 |
855 856 857 |
struct mm_struct *mm = current->mm; struct vm_area_struct *vma = NULL; struct mempolicy *pol = current->mempolicy; |
754af6f5a Mem Policy: add M... |
858 859 |
if (flags & ~(unsigned long)(MPOL_F_NODE|MPOL_F_ADDR|MPOL_F_MEMS_ALLOWED)) |
1da177e4c Linux-2.6.12-rc2 |
860 |
return -EINVAL; |
754af6f5a Mem Policy: add M... |
861 862 863 864 865 |
if (flags & MPOL_F_MEMS_ALLOWED) { if (flags & (MPOL_F_NODE|MPOL_F_ADDR)) return -EINVAL; *policy = 0; /* just so it's initialized */ |
58568d2a8 cpuset,mm: update... |
866 |
task_lock(current); |
754af6f5a Mem Policy: add M... |
867 |
*nmask = cpuset_current_mems_allowed; |
58568d2a8 cpuset,mm: update... |
868 |
task_unlock(current); |
754af6f5a Mem Policy: add M... |
869 870 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
871 |
if (flags & MPOL_F_ADDR) { |
bea904d54 mempolicy: use MP... |
872 873 874 875 876 |
/* * 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 |
877 878 879 880 881 882 883 884 885 886 887 888 889 890 |
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... |
891 |
pol = &default_policy; /* indicates default behavior */ |
1da177e4c Linux-2.6.12-rc2 |
892 893 894 895 896 897 |
if (flags & MPOL_F_NODE) { if (flags & MPOL_F_ADDR) { err = lookup_node(mm, addr); if (err < 0) goto out; |
8bccd85ff [PATCH] Implement... |
898 |
*policy = err; |
1da177e4c Linux-2.6.12-rc2 |
899 |
} else if (pol == current->mempolicy && |
45c4745af mempolicy: rename... |
900 |
pol->mode == MPOL_INTERLEAVE) { |
8bccd85ff [PATCH] Implement... |
901 |
*policy = current->il_next; |
1da177e4c Linux-2.6.12-rc2 |
902 903 904 905 |
} else { err = -EINVAL; goto out; } |
bea904d54 mempolicy: use MP... |
906 907 908 |
} else { *policy = pol == &default_policy ? MPOL_DEFAULT : pol->mode; |
d79df630f mempolicy: mask o... |
909 910 911 912 913 |
/* * Internal mempolicy flags must be masked off before exposing * the policy to userspace. */ *policy |= (pol->flags & MPOL_MODE_FLAGS); |
bea904d54 mempolicy: use MP... |
914 |
} |
1da177e4c Linux-2.6.12-rc2 |
915 916 917 918 919 |
if (vma) { up_read(¤t->mm->mmap_sem); vma = NULL; } |
1da177e4c Linux-2.6.12-rc2 |
920 |
err = 0; |
58568d2a8 cpuset,mm: update... |
921 |
if (nmask) { |
c6b6ef8bb mempolicy: fix ge... |
922 923 924 925 926 927 928 |
if (mpol_store_user_nodemask(pol)) { *nmask = pol->w.user_nodemask; } else { task_lock(current); get_policy_nodemask(pol, nmask); task_unlock(current); } |
58568d2a8 cpuset,mm: update... |
929 |
} |
1da177e4c Linux-2.6.12-rc2 |
930 931 |
out: |
52cd3b074 mempolicy: rework... |
932 |
mpol_cond_put(pol); |
1da177e4c Linux-2.6.12-rc2 |
933 934 935 936 |
if (vma) up_read(¤t->mm->mmap_sem); return err; } |
b20a35035 [PATCH] page migr... |
937 |
#ifdef CONFIG_MIGRATION |
8bccd85ff [PATCH] Implement... |
938 |
/* |
6ce3c4c0f [PATCH] Move page... |
939 940 |
* page migration */ |
fc3012896 [PATCH] Simplify ... |
941 942 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
943 944 |
{ /* |
fc3012896 [PATCH] Simplify ... |
945 |
* Avoid migrating a page that is shared with others. |
6ce3c4c0f [PATCH] Move page... |
946 |
*/ |
62695a84e vmscan: move isol... |
947 948 949 |
if ((flags & MPOL_MF_MOVE_ALL) || page_mapcount(page) == 1) { if (!isolate_lru_page(page)) { list_add_tail(&page->lru, pagelist); |
6d9c285a6 mm: move inc_zone... |
950 951 |
inc_zone_page_state(page, NR_ISOLATED_ANON + page_is_file_cache(page)); |
62695a84e vmscan: move isol... |
952 953 |
} } |
7e2ab150d [PATCH] Direct Mi... |
954 |
} |
6ce3c4c0f [PATCH] Move page... |
955 |
|
742755a1d [PATCH] page migr... |
956 |
static struct page *new_node_page(struct page *page, unsigned long node, int **x) |
95a402c38 [PATCH] page migr... |
957 |
{ |
e2d8cf405 migrate: add huge... |
958 959 960 961 962 |
if (PageHuge(page)) return alloc_huge_page_node(page_hstate(compound_head(page)), node); else return alloc_pages_exact_node(node, GFP_HIGHUSER_MOVABLE, 0); |
95a402c38 [PATCH] page migr... |
963 |
} |
7e2ab150d [PATCH] Direct Mi... |
964 965 966 967 |
/* * Migrate pages from one node to a target node. * Returns error or the number of pages not migrated. */ |
dbcb0f19c mm/mempolicy.c: c... |
968 969 |
static int migrate_to_node(struct mm_struct *mm, int source, int dest, int flags) |
7e2ab150d [PATCH] Direct Mi... |
970 971 972 973 974 975 976 |
{ nodemask_t nmask; LIST_HEAD(pagelist); int err = 0; nodes_clear(nmask); node_set(source, nmask); |
6ce3c4c0f [PATCH] Move page... |
977 |
|
082708072 mm: revert 0def08... |
978 979 980 981 982 983 |
/* * This does not "check" the range but isolates all pages that * need migration. Between passing in the full user address * space range and MPOL_MF_DISCONTIG_OK, this call can not fail. */ VM_BUG_ON(!(flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL))); |
980949457 mm/mempolicy: ren... |
984 |
queue_pages_range(mm, mm->mmap->vm_start, mm->task_size, &nmask, |
7e2ab150d [PATCH] Direct Mi... |
985 |
flags | MPOL_MF_DISCONTIG_OK, &pagelist); |
cf608ac19 mm: compaction: f... |
986 |
if (!list_empty(&pagelist)) { |
7f0f24967 mm: migration: cl... |
987 |
err = migrate_pages(&pagelist, new_node_page, dest, |
9c620e2bc mm: remove offlin... |
988 |
MIGRATE_SYNC, MR_SYSCALL); |
cf608ac19 mm: compaction: f... |
989 |
if (err) |
e2d8cf405 migrate: add huge... |
990 |
putback_movable_pages(&pagelist); |
cf608ac19 mm: compaction: f... |
991 |
} |
95a402c38 [PATCH] page migr... |
992 |
|
7e2ab150d [PATCH] Direct Mi... |
993 |
return err; |
6ce3c4c0f [PATCH] Move page... |
994 995 996 |
} /* |
7e2ab150d [PATCH] Direct Mi... |
997 998 |
* Move pages between the two nodesets so as to preserve the physical * layout as much as possible. |
39743889a [PATCH] Swap Migr... |
999 1000 1001 |
* * Returns the number of page that could not be moved. */ |
0ce72d4f7 mm: do_migrate_pa... |
1002 1003 |
int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from, const nodemask_t *to, int flags) |
39743889a [PATCH] Swap Migr... |
1004 |
{ |
7e2ab150d [PATCH] Direct Mi... |
1005 |
int busy = 0; |
0aedadf91 mm: move migrate_... |
1006 |
int err; |
7e2ab150d [PATCH] Direct Mi... |
1007 |
nodemask_t tmp; |
39743889a [PATCH] Swap Migr... |
1008 |
|
0aedadf91 mm: move migrate_... |
1009 1010 1011 |
err = migrate_prep(); if (err) return err; |
53f2556b6 mempolicy: mPOL_P... |
1012 |
down_read(&mm->mmap_sem); |
39743889a [PATCH] Swap Migr... |
1013 |
|
0ce72d4f7 mm: do_migrate_pa... |
1014 |
err = migrate_vmas(mm, from, to, flags); |
7b2259b3e [PATCH] page migr... |
1015 1016 |
if (err) goto out; |
da0aa1389 mm/mempolicy.c: f... |
1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 |
/* * 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. |
ae0e47f02 Remove one to man... |
1042 |
* Otherwise when we finish scanning from_tmp, we at least have the |
da0aa1389 mm/mempolicy.c: f... |
1043 1044 1045 1046 |
* 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: ... |
1047 |
|
0ce72d4f7 mm: do_migrate_pa... |
1048 |
tmp = *from; |
7e2ab150d [PATCH] Direct Mi... |
1049 1050 |
while (!nodes_empty(tmp)) { int s,d; |
b76ac7e73 mm/mempolicy: use... |
1051 |
int source = NUMA_NO_NODE; |
7e2ab150d [PATCH] Direct Mi... |
1052 1053 1054 |
int dest = 0; for_each_node_mask(s, tmp) { |
4a5b18cc1 mm: do_migrate_pa... |
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 |
/* * do_migrate_pages() tries to maintain the relative * node relationship of the pages established between * threads and memory areas. * * However if the number of source nodes is not equal to * the number of destination nodes we can not preserve * this node relative relationship. In that case, skip * copying memory from a node that is in the destination * mask. * * Example: [2,3,4] -> [3,4,5] moves everything. * [0-7] - > [3,4,5] moves only 0,1,2,6,7. */ |
0ce72d4f7 mm: do_migrate_pa... |
1070 1071 |
if ((nodes_weight(*from) != nodes_weight(*to)) && (node_isset(s, *to))) |
4a5b18cc1 mm: do_migrate_pa... |
1072 |
continue; |
0ce72d4f7 mm: do_migrate_pa... |
1073 |
d = node_remap(s, *from, *to); |
7e2ab150d [PATCH] Direct Mi... |
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 |
if (s == d) continue; source = s; /* Node moved. Memorize */ dest = d; /* dest not in remaining from nodes? */ if (!node_isset(dest, tmp)) break; } |
b76ac7e73 mm/mempolicy: use... |
1084 |
if (source == NUMA_NO_NODE) |
7e2ab150d [PATCH] Direct Mi... |
1085 1086 1087 1088 1089 1090 1091 1092 |
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... |
1093 |
} |
7b2259b3e [PATCH] page migr... |
1094 |
out: |
39743889a [PATCH] Swap Migr... |
1095 |
up_read(&mm->mmap_sem); |
7e2ab150d [PATCH] Direct Mi... |
1096 1097 1098 |
if (err < 0) return err; return busy; |
b20a35035 [PATCH] page migr... |
1099 1100 |
} |
3ad33b243 Migration: find c... |
1101 1102 1103 1104 1105 1106 1107 |
/* * 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... |
1108 |
static struct page *new_vma_page(struct page *page, unsigned long private, int **x) |
95a402c38 [PATCH] page migr... |
1109 1110 |
{ struct vm_area_struct *vma = (struct vm_area_struct *)private; |
3ad33b243 Migration: find c... |
1111 |
unsigned long uninitialized_var(address); |
95a402c38 [PATCH] page migr... |
1112 |
|
3ad33b243 Migration: find c... |
1113 1114 1115 1116 1117 1118 |
while (vma) { address = page_address_in_vma(page, vma); if (address != -EFAULT) break; vma = vma->vm_next; } |
11c731e81 mm/mempolicy: fix... |
1119 1120 |
if (PageHuge(page)) { |
cc81717ed mm: new_vma_page(... |
1121 1122 |
BUG_ON(!vma); return alloc_huge_page_noerr(vma, address, 1); |
11c731e81 mm/mempolicy: fix... |
1123 |
} |
0bf598d86 mbind: add BUG_ON... |
1124 |
/* |
11c731e81 mm/mempolicy: fix... |
1125 |
* if !vma, alloc_page_vma() will use task or system default policy |
0bf598d86 mbind: add BUG_ON... |
1126 |
*/ |
3ad33b243 Migration: find c... |
1127 |
return alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); |
95a402c38 [PATCH] page migr... |
1128 |
} |
b20a35035 [PATCH] page migr... |
1129 1130 1131 1132 1133 |
#else static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) { |
39743889a [PATCH] Swap Migr... |
1134 |
} |
0ce72d4f7 mm: do_migrate_pa... |
1135 1136 |
int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from, const nodemask_t *to, int flags) |
b20a35035 [PATCH] page migr... |
1137 1138 1139 |
{ return -ENOSYS; } |
95a402c38 [PATCH] page migr... |
1140 |
|
699397499 [PATCH] Fix do_mb... |
1141 |
static struct page *new_vma_page(struct page *page, unsigned long private, int **x) |
95a402c38 [PATCH] page migr... |
1142 1143 1144 |
{ return NULL; } |
b20a35035 [PATCH] page migr... |
1145 |
#endif |
dbcb0f19c mm/mempolicy.c: c... |
1146 |
static long do_mbind(unsigned long start, unsigned long len, |
028fec414 mempolicy: suppor... |
1147 1148 |
unsigned short mode, unsigned short mode_flags, nodemask_t *nmask, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
1149 1150 1151 1152 1153 1154 1155 |
{ struct vm_area_struct *vma; struct mm_struct *mm = current->mm; struct mempolicy *new; unsigned long end; int err; LIST_HEAD(pagelist); |
b24f53a0b mm: mempolicy: Ad... |
1156 |
if (flags & ~(unsigned long)MPOL_MF_VALID) |
6ce3c4c0f [PATCH] Move page... |
1157 |
return -EINVAL; |
74c002410 [PATCH] Consisten... |
1158 |
if ((flags & MPOL_MF_MOVE_ALL) && !capable(CAP_SYS_NICE)) |
6ce3c4c0f [PATCH] Move page... |
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
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... |
1174 |
new = mpol_new(mode, mode_flags, nmask); |
6ce3c4c0f [PATCH] Move page... |
1175 1176 |
if (IS_ERR(new)) return PTR_ERR(new); |
b24f53a0b mm: mempolicy: Ad... |
1177 1178 |
if (flags & MPOL_MF_LAZY) new->flags |= MPOL_F_MOF; |
6ce3c4c0f [PATCH] Move page... |
1179 1180 1181 1182 1183 1184 |
/* * 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... |
1185 1186 1187 |
pr_debug("mbind %lx-%lx mode:%d flags:%d nodes:%lx ", start, start + len, mode, mode_flags, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1188 |
nmask ? nodes_addr(*nmask)[0] : NUMA_NO_NODE); |
6ce3c4c0f [PATCH] Move page... |
1189 |
|
0aedadf91 mm: move migrate_... |
1190 1191 1192 1193 |
if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) { err = migrate_prep(); if (err) |
b05ca7385 do_mbind(): fix m... |
1194 |
goto mpol_out; |
0aedadf91 mm: move migrate_... |
1195 |
} |
4bfc44958 mm: make set_memp... |
1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 |
{ NODEMASK_SCRATCH(scratch); if (scratch) { down_write(&mm->mmap_sem); task_lock(current); err = mpol_set_nodemask(new, nmask, scratch); task_unlock(current); if (err) up_write(&mm->mmap_sem); } else err = -ENOMEM; NODEMASK_SCRATCH_FREE(scratch); } |
b05ca7385 do_mbind(): fix m... |
1209 1210 |
if (err) goto mpol_out; |
980949457 mm/mempolicy: ren... |
1211 |
vma = queue_pages_range(mm, start, end, nmask, |
6ce3c4c0f [PATCH] Move page... |
1212 |
flags | MPOL_MF_INVERT, &pagelist); |
b24f53a0b mm: mempolicy: Ad... |
1213 |
err = PTR_ERR(vma); /* maybe ... */ |
a720094de mm: mempolicy: Hi... |
1214 |
if (!IS_ERR(vma)) |
9d8cebd4b mm: fix mbind vma... |
1215 |
err = mbind_range(mm, start, end, new); |
7e2ab150d [PATCH] Direct Mi... |
1216 |
|
b24f53a0b mm: mempolicy: Ad... |
1217 1218 |
if (!err) { int nr_failed = 0; |
cf608ac19 mm: compaction: f... |
1219 |
if (!list_empty(&pagelist)) { |
b24f53a0b mm: mempolicy: Ad... |
1220 |
WARN_ON_ONCE(flags & MPOL_MF_LAZY); |
95a402c38 [PATCH] page migr... |
1221 |
nr_failed = migrate_pages(&pagelist, new_vma_page, |
9c620e2bc mm: remove offlin... |
1222 1223 |
(unsigned long)vma, MIGRATE_SYNC, MR_MEMPOLICY_MBIND); |
cf608ac19 mm: compaction: f... |
1224 |
if (nr_failed) |
74060e4d7 mm: mbind: add hu... |
1225 |
putback_movable_pages(&pagelist); |
cf608ac19 mm: compaction: f... |
1226 |
} |
6ce3c4c0f [PATCH] Move page... |
1227 |
|
b24f53a0b mm: mempolicy: Ad... |
1228 |
if (nr_failed && (flags & MPOL_MF_STRICT)) |
6ce3c4c0f [PATCH] Move page... |
1229 |
err = -EIO; |
ab8a3e14e mbind(): fix leak... |
1230 |
} else |
b0e5fd735 mm/mempolicy: cor... |
1231 |
putback_movable_pages(&pagelist); |
b20a35035 [PATCH] page migr... |
1232 |
|
6ce3c4c0f [PATCH] Move page... |
1233 |
up_write(&mm->mmap_sem); |
b05ca7385 do_mbind(): fix m... |
1234 |
mpol_out: |
f0be3d32b mempolicy: rename... |
1235 |
mpol_put(new); |
6ce3c4c0f [PATCH] Move page... |
1236 1237 |
return err; } |
39743889a [PATCH] Swap Migr... |
1238 |
/* |
8bccd85ff [PATCH] Implement... |
1239 1240 1241 1242 |
* User space interface with variable sized bitmaps for nodelists. */ /* Copy a node mask from user space. */ |
39743889a [PATCH] Swap Migr... |
1243 |
static int get_nodes(nodemask_t *nodes, const unsigned long __user *nmask, |
8bccd85ff [PATCH] Implement... |
1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 |
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... |
1254 |
if (maxnode > PAGE_SIZE*BITS_PER_BYTE) |
636f13c17 [PATCH] sys_mbind... |
1255 |
return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 |
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; } |
938bb9f5e [CVE-2009-0029] S... |
1304 1305 1306 |
SYSCALL_DEFINE6(mbind, unsigned long, start, unsigned long, len, unsigned long, mode, unsigned long __user *, nmask, unsigned long, maxnode, unsigned, flags) |
8bccd85ff [PATCH] Implement... |
1307 1308 1309 |
{ nodemask_t nodes; int err; |
028fec414 mempolicy: suppor... |
1310 |
unsigned short mode_flags; |
8bccd85ff [PATCH] Implement... |
1311 |
|
028fec414 mempolicy: suppor... |
1312 1313 |
mode_flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; |
a3b51e014 mempolicy: conver... |
1314 1315 |
if (mode >= MPOL_MAX) return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1316 1317 1318 |
if ((mode_flags & MPOL_F_STATIC_NODES) && (mode_flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1319 1320 1321 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1322 |
return do_mbind(start, len, mode, mode_flags, &nodes, flags); |
8bccd85ff [PATCH] Implement... |
1323 1324 1325 |
} /* Set the process memory policy */ |
938bb9f5e [CVE-2009-0029] S... |
1326 1327 |
SYSCALL_DEFINE3(set_mempolicy, int, mode, unsigned long __user *, nmask, unsigned long, maxnode) |
8bccd85ff [PATCH] Implement... |
1328 1329 1330 |
{ int err; nodemask_t nodes; |
028fec414 mempolicy: suppor... |
1331 |
unsigned short flags; |
8bccd85ff [PATCH] Implement... |
1332 |
|
028fec414 mempolicy: suppor... |
1333 1334 1335 |
flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; if ((unsigned int)mode >= MPOL_MAX) |
8bccd85ff [PATCH] Implement... |
1336 |
return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1337 1338 |
if ((flags & MPOL_F_STATIC_NODES) && (flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1339 1340 1341 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1342 |
return do_set_mempolicy(mode, flags, &nodes); |
8bccd85ff [PATCH] Implement... |
1343 |
} |
938bb9f5e [CVE-2009-0029] S... |
1344 1345 1346 |
SYSCALL_DEFINE4(migrate_pages, pid_t, pid, unsigned long, maxnode, const unsigned long __user *, old_nodes, const unsigned long __user *, new_nodes) |
39743889a [PATCH] Swap Migr... |
1347 |
{ |
c69e8d9c0 CRED: Use RCU to ... |
1348 |
const struct cred *cred = current_cred(), *tcred; |
596d7cfa2 mempolicy: reduce... |
1349 |
struct mm_struct *mm = NULL; |
39743889a [PATCH] Swap Migr... |
1350 |
struct task_struct *task; |
39743889a [PATCH] Swap Migr... |
1351 1352 |
nodemask_t task_nodes; int err; |
596d7cfa2 mempolicy: reduce... |
1353 1354 1355 1356 1357 1358 |
nodemask_t *old; nodemask_t *new; NODEMASK_SCRATCH(scratch); if (!scratch) return -ENOMEM; |
39743889a [PATCH] Swap Migr... |
1359 |
|
596d7cfa2 mempolicy: reduce... |
1360 1361 1362 1363 |
old = &scratch->mask1; new = &scratch->mask2; err = get_nodes(old, old_nodes, maxnode); |
39743889a [PATCH] Swap Migr... |
1364 |
if (err) |
596d7cfa2 mempolicy: reduce... |
1365 |
goto out; |
39743889a [PATCH] Swap Migr... |
1366 |
|
596d7cfa2 mempolicy: reduce... |
1367 |
err = get_nodes(new, new_nodes, maxnode); |
39743889a [PATCH] Swap Migr... |
1368 |
if (err) |
596d7cfa2 mempolicy: reduce... |
1369 |
goto out; |
39743889a [PATCH] Swap Migr... |
1370 1371 |
/* Find the mm_struct */ |
55cfaa3cb mm/mempolicy.c: a... |
1372 |
rcu_read_lock(); |
228ebcbe6 Uninline find_tas... |
1373 |
task = pid ? find_task_by_vpid(pid) : current; |
39743889a [PATCH] Swap Migr... |
1374 |
if (!task) { |
55cfaa3cb mm/mempolicy.c: a... |
1375 |
rcu_read_unlock(); |
596d7cfa2 mempolicy: reduce... |
1376 1377 |
err = -ESRCH; goto out; |
39743889a [PATCH] Swap Migr... |
1378 |
} |
3268c63ed mm: fix move/migr... |
1379 |
get_task_struct(task); |
39743889a [PATCH] Swap Migr... |
1380 |
|
596d7cfa2 mempolicy: reduce... |
1381 |
err = -EINVAL; |
39743889a [PATCH] Swap Migr... |
1382 1383 1384 1385 |
/* * Check if this process has the right to modify the specified * process. The right exists if the process has administrative |
7f927fcc2 [PATCH] Typo fixes |
1386 |
* capabilities, superuser privileges or the same |
39743889a [PATCH] Swap Migr... |
1387 1388 |
* userid as the target process. */ |
c69e8d9c0 CRED: Use RCU to ... |
1389 |
tcred = __task_cred(task); |
b38a86eb1 userns: Convert t... |
1390 1391 |
if (!uid_eq(cred->euid, tcred->suid) && !uid_eq(cred->euid, tcred->uid) && !uid_eq(cred->uid, tcred->suid) && !uid_eq(cred->uid, tcred->uid) && |
74c002410 [PATCH] Consisten... |
1392 |
!capable(CAP_SYS_NICE)) { |
c69e8d9c0 CRED: Use RCU to ... |
1393 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1394 |
err = -EPERM; |
3268c63ed mm: fix move/migr... |
1395 |
goto out_put; |
39743889a [PATCH] Swap Migr... |
1396 |
} |
c69e8d9c0 CRED: Use RCU to ... |
1397 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1398 1399 1400 |
task_nodes = cpuset_mems_allowed(task); /* Is the user allowed to access the target nodes? */ |
596d7cfa2 mempolicy: reduce... |
1401 |
if (!nodes_subset(*new, task_nodes) && !capable(CAP_SYS_NICE)) { |
39743889a [PATCH] Swap Migr... |
1402 |
err = -EPERM; |
3268c63ed mm: fix move/migr... |
1403 |
goto out_put; |
39743889a [PATCH] Swap Migr... |
1404 |
} |
01f13bd60 mempolicy: use N_... |
1405 |
if (!nodes_subset(*new, node_states[N_MEMORY])) { |
3b42d28b2 Page migration: D... |
1406 |
err = -EINVAL; |
3268c63ed mm: fix move/migr... |
1407 |
goto out_put; |
3b42d28b2 Page migration: D... |
1408 |
} |
86c3a7645 [PATCH] SELinux: ... |
1409 1410 |
err = security_task_movememory(task); if (err) |
3268c63ed mm: fix move/migr... |
1411 |
goto out_put; |
86c3a7645 [PATCH] SELinux: ... |
1412 |
|
3268c63ed mm: fix move/migr... |
1413 1414 |
mm = get_task_mm(task); put_task_struct(task); |
f2a9ef880 mm: fix NULL ptr ... |
1415 1416 |
if (!mm) { |
3268c63ed mm: fix move/migr... |
1417 |
err = -EINVAL; |
f2a9ef880 mm: fix NULL ptr ... |
1418 1419 1420 1421 1422 |
goto out; } err = do_migrate_pages(mm, old, new, capable(CAP_SYS_NICE) ? MPOL_MF_MOVE_ALL : MPOL_MF_MOVE); |
3268c63ed mm: fix move/migr... |
1423 1424 1425 |
mmput(mm); out: |
596d7cfa2 mempolicy: reduce... |
1426 |
NODEMASK_SCRATCH_FREE(scratch); |
39743889a [PATCH] Swap Migr... |
1427 |
return err; |
3268c63ed mm: fix move/migr... |
1428 1429 1430 1431 |
out_put: put_task_struct(task); goto out; |
39743889a [PATCH] Swap Migr... |
1432 |
} |
8bccd85ff [PATCH] Implement... |
1433 |
/* Retrieve NUMA policy */ |
938bb9f5e [CVE-2009-0029] S... |
1434 1435 1436 |
SYSCALL_DEFINE5(get_mempolicy, int __user *, policy, unsigned long __user *, nmask, unsigned long, maxnode, unsigned long, addr, unsigned long, flags) |
8bccd85ff [PATCH] Implement... |
1437 |
{ |
dbcb0f19c mm/mempolicy.c: c... |
1438 1439 |
int err; int uninitialized_var(pval); |
8bccd85ff [PATCH] Implement... |
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 |
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 |
1458 |
#ifdef CONFIG_COMPAT |
c93e0f6c8 mm/compat: conver... |
1459 1460 1461 1462 |
COMPAT_SYSCALL_DEFINE5(get_mempolicy, int __user *, policy, compat_ulong_t __user *, nmask, compat_ulong_t, maxnode, compat_ulong_t, addr, compat_ulong_t, flags) |
1da177e4c Linux-2.6.12-rc2 |
1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 |
{ 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) { |
2bbff6c76 mm/mempolicy.c: m... |
1478 1479 1480 |
unsigned long copy_size; copy_size = min_t(unsigned long, sizeof(bm), alloc_size); err = copy_from_user(bm, nm, copy_size); |
1da177e4c Linux-2.6.12-rc2 |
1481 1482 1483 1484 1485 1486 1487 |
/* ensure entire bitmap is zeroed */ err |= clear_user(nmask, ALIGN(maxnode-1, 8) / 8); err |= compat_put_bitmap(nmask, bm, nr_bits); } return err; } |
c93e0f6c8 mm/compat: conver... |
1488 1489 |
COMPAT_SYSCALL_DEFINE3(set_mempolicy, int, mode, compat_ulong_t __user *, nmask, compat_ulong_t, maxnode) |
1da177e4c Linux-2.6.12-rc2 |
1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 |
{ 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); } |
c93e0f6c8 mm/compat: conver... |
1510 1511 1512 |
COMPAT_SYSCALL_DEFINE6(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) |
1da177e4c Linux-2.6.12-rc2 |
1513 1514 1515 1516 |
{ long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; |
dfcd3c0dc [PATCH] Convert m... |
1517 |
nodemask_t bm; |
1da177e4c Linux-2.6.12-rc2 |
1518 1519 1520 1521 1522 |
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... |
1523 |
err = compat_get_bitmap(nodes_addr(bm), nmask, nr_bits); |
1da177e4c Linux-2.6.12-rc2 |
1524 |
nm = compat_alloc_user_space(alloc_size); |
dfcd3c0dc [PATCH] Convert m... |
1525 |
err |= copy_to_user(nm, nodes_addr(bm), alloc_size); |
1da177e4c Linux-2.6.12-rc2 |
1526 1527 1528 1529 1530 1531 1532 1533 1534 |
} if (err) return -EFAULT; return sys_mbind(start, len, mode, nm, nr_bits+1, flags); } #endif |
480eccf9a Fix NUMA Memory P... |
1535 1536 1537 1538 1539 1540 1541 1542 |
/* * 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. |
32f8516a8 mm, mempolicy: fi... |
1543 1544 |
* Current or other task's task mempolicy and non-shared vma policies must be * protected by task_lock(task) by the caller. |
52cd3b074 mempolicy: rework... |
1545 1546 1547 1548 |
* 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... |
1549 |
*/ |
d98f6cb67 mm: export get_vm... |
1550 |
struct mempolicy *get_vma_policy(struct task_struct *task, |
48fce3429 [PATCH] mempolici... |
1551 |
struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
1552 |
{ |
5606e3877 mm: numa: Migrate... |
1553 |
struct mempolicy *pol = get_task_policy(task); |
1da177e4c Linux-2.6.12-rc2 |
1554 1555 |
if (vma) { |
480eccf9a Fix NUMA Memory P... |
1556 |
if (vma->vm_ops && vma->vm_ops->get_policy) { |
ae4d8c16a mempolicy: fixup ... |
1557 1558 1559 1560 |
struct mempolicy *vpol = vma->vm_ops->get_policy(vma, addr); if (vpol) pol = vpol; |
00442ad04 mempolicy: fix a ... |
1561 |
} else if (vma->vm_policy) { |
1da177e4c Linux-2.6.12-rc2 |
1562 |
pol = vma->vm_policy; |
00442ad04 mempolicy: fix a ... |
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 |
/* * shmem_alloc_page() passes MPOL_F_SHARED policy with * a pseudo vma whose vma->vm_ops=NULL. Take a reference * count on these policies which will be dropped by * mpol_cond_put() later */ if (mpol_needs_cond_ref(pol)) mpol_get(pol); } |
1da177e4c Linux-2.6.12-rc2 |
1573 1574 1575 1576 1577 |
} if (!pol) pol = &default_policy; return pol; } |
fc3147245 mm: numa: Limit N... |
1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 |
bool vma_policy_mof(struct task_struct *task, struct vm_area_struct *vma) { struct mempolicy *pol = get_task_policy(task); if (vma) { if (vma->vm_ops && vma->vm_ops->get_policy) { bool ret = false; pol = vma->vm_ops->get_policy(vma, vma->vm_start); if (pol && (pol->flags & MPOL_F_MOF)) ret = true; mpol_cond_put(pol); return ret; } else if (vma->vm_policy) { pol = vma->vm_policy; } } if (!pol) return default_policy.flags & MPOL_F_MOF; return pol->flags & MPOL_F_MOF; } |
d3eb1570a mempolicy: fix is... |
1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 |
static int apply_policy_zone(struct mempolicy *policy, enum zone_type zone) { enum zone_type dynamic_policy_zone = policy_zone; BUG_ON(dynamic_policy_zone == ZONE_MOVABLE); /* * if policy->v.nodes has movable memory only, * we apply policy when gfp_zone(gfp) = ZONE_MOVABLE only. * * policy->v.nodes is intersect with node_states[N_MEMORY]. * so if the following test faile, it implies * policy->v.nodes has movable memory only. */ if (!nodes_intersects(policy->v.nodes, node_states[N_HIGH_MEMORY])) dynamic_policy_zone = ZONE_MOVABLE; return zone >= dynamic_policy_zone; } |
52cd3b074 mempolicy: rework... |
1620 1621 1622 1623 1624 |
/* * 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 ... |
1625 1626 |
{ /* Lower zones don't get a nodemask applied for MPOL_BIND */ |
45c4745af mempolicy: rename... |
1627 |
if (unlikely(policy->mode == MPOL_BIND) && |
d3eb1570a mempolicy: fix is... |
1628 |
apply_policy_zone(policy, gfp_zone(gfp)) && |
19770b326 mm: filter based ... |
1629 1630 1631 1632 1633 |
cpuset_nodemask_valid_mems_allowed(&policy->v.nodes)) return &policy->v.nodes; return NULL; } |
52cd3b074 mempolicy: rework... |
1634 |
/* Return a zonelist indicated by gfp for node representing a mempolicy */ |
2f5f9486f mm: change alloc_... |
1635 1636 |
static struct zonelist *policy_zonelist(gfp_t gfp, struct mempolicy *policy, int nd) |
1da177e4c Linux-2.6.12-rc2 |
1637 |
{ |
45c4745af mempolicy: rename... |
1638 |
switch (policy->mode) { |
1da177e4c Linux-2.6.12-rc2 |
1639 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1640 1641 |
if (!(policy->flags & MPOL_F_LOCAL)) nd = policy->v.preferred_node; |
1da177e4c Linux-2.6.12-rc2 |
1642 1643 |
break; case MPOL_BIND: |
19770b326 mm: filter based ... |
1644 |
/* |
52cd3b074 mempolicy: rework... |
1645 1646 |
* Normally, MPOL_BIND allocations are node-local within the * allowed nodemask. However, if __GFP_THISNODE is set and the |
6eb27e1fd mempolicy: remove... |
1647 |
* current node isn't part of the mask, we use the zonelist for |
52cd3b074 mempolicy: rework... |
1648 |
* the first node in the mask instead. |
19770b326 mm: filter based ... |
1649 |
*/ |
19770b326 mm: filter based ... |
1650 1651 1652 1653 |
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 |
1654 |
default: |
1da177e4c Linux-2.6.12-rc2 |
1655 1656 |
BUG(); } |
0e88460da mm: introduce nod... |
1657 |
return node_zonelist(nd, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1658 1659 1660 1661 1662 1663 1664 1665 1666 |
} /* 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... |
1667 |
next = next_node(nid, policy->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1668 |
if (next >= MAX_NUMNODES) |
dfcd3c0dc [PATCH] Convert m... |
1669 |
next = first_node(policy->v.nodes); |
f5b087b52 mempolicy: add MP... |
1670 1671 |
if (next < MAX_NUMNODES) me->il_next = next; |
1da177e4c Linux-2.6.12-rc2 |
1672 1673 |
return nid; } |
dc85da15d [PATCH] NUMA poli... |
1674 1675 1676 1677 |
/* * Depending on the memory policy provide a node from which to allocate the * next slab entry. */ |
2a389610a mm, mempolicy: re... |
1678 |
unsigned int mempolicy_slab_node(void) |
dc85da15d [PATCH] NUMA poli... |
1679 |
{ |
e7b691b08 slab/mempolicy: a... |
1680 |
struct mempolicy *policy; |
2a389610a mm, mempolicy: re... |
1681 |
int node = numa_mem_id(); |
e7b691b08 slab/mempolicy: a... |
1682 1683 |
if (in_interrupt()) |
2a389610a mm, mempolicy: re... |
1684 |
return node; |
e7b691b08 slab/mempolicy: a... |
1685 1686 |
policy = current->mempolicy; |
fc36b8d3d mempolicy: use MP... |
1687 |
if (!policy || policy->flags & MPOL_F_LOCAL) |
2a389610a mm, mempolicy: re... |
1688 |
return node; |
bea904d54 mempolicy: use MP... |
1689 1690 1691 |
switch (policy->mode) { case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1692 1693 1694 1695 |
/* * handled MPOL_F_LOCAL above */ return policy->v.preferred_node; |
765c4507a [PATCH] GFP_THISN... |
1696 |
|
dc85da15d [PATCH] NUMA poli... |
1697 1698 |
case MPOL_INTERLEAVE: return interleave_nodes(policy); |
dd1a239f6 mm: have zonelist... |
1699 |
case MPOL_BIND: { |
dc85da15d [PATCH] NUMA poli... |
1700 1701 1702 1703 |
/* * Follow bind policy behavior and start allocation at the * first node. */ |
19770b326 mm: filter based ... |
1704 1705 1706 |
struct zonelist *zonelist; struct zone *zone; enum zone_type highest_zoneidx = gfp_zone(GFP_KERNEL); |
2a389610a mm, mempolicy: re... |
1707 |
zonelist = &NODE_DATA(node)->node_zonelists[0]; |
19770b326 mm: filter based ... |
1708 1709 1710 |
(void)first_zones_zonelist(zonelist, highest_zoneidx, &policy->v.nodes, &zone); |
2a389610a mm, mempolicy: re... |
1711 |
return zone ? zone->node : node; |
dd1a239f6 mm: have zonelist... |
1712 |
} |
dc85da15d [PATCH] NUMA poli... |
1713 |
|
dc85da15d [PATCH] NUMA poli... |
1714 |
default: |
bea904d54 mempolicy: use MP... |
1715 |
BUG(); |
dc85da15d [PATCH] NUMA poli... |
1716 1717 |
} } |
1da177e4c Linux-2.6.12-rc2 |
1718 1719 1720 1721 |
/* 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... |
1722 |
unsigned nnodes = nodes_weight(pol->v.nodes); |
f5b087b52 mempolicy: add MP... |
1723 |
unsigned target; |
1da177e4c Linux-2.6.12-rc2 |
1724 |
int c; |
b76ac7e73 mm/mempolicy: use... |
1725 |
int nid = NUMA_NO_NODE; |
1da177e4c Linux-2.6.12-rc2 |
1726 |
|
f5b087b52 mempolicy: add MP... |
1727 1728 1729 |
if (!nnodes) return numa_node_id(); target = (unsigned int)off % nnodes; |
1da177e4c Linux-2.6.12-rc2 |
1730 1731 |
c = 0; do { |
dfcd3c0dc [PATCH] Convert m... |
1732 |
nid = next_node(nid, pol->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1733 1734 |
c++; } while (c <= target); |
1da177e4c Linux-2.6.12-rc2 |
1735 1736 |
return nid; } |
5da7ca860 [PATCH] Add NUMA ... |
1737 1738 1739 1740 1741 1742 |
/* 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 ... |
1743 1744 1745 1746 1747 1748 1749 1750 1751 |
/* * 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 ... |
1752 1753 1754 1755 1756 |
off += (addr - vma->vm_start) >> shift; return offset_il_node(pol, vma, off); } else return interleave_nodes(pol); } |
778d3b0ff cpusets: randomiz... |
1757 1758 |
/* * Return the bit number of a random bit set in the nodemask. |
b76ac7e73 mm/mempolicy: use... |
1759 |
* (returns NUMA_NO_NODE if nodemask is empty) |
778d3b0ff cpusets: randomiz... |
1760 1761 1762 |
*/ int node_random(const nodemask_t *maskp) { |
b76ac7e73 mm/mempolicy: use... |
1763 |
int w, bit = NUMA_NO_NODE; |
778d3b0ff cpusets: randomiz... |
1764 1765 1766 1767 1768 1769 1770 |
w = nodes_weight(*maskp); if (w) bit = bitmap_ord_to_pos(maskp->bits, get_random_int() % w, MAX_NUMNODES); return bit; } |
00ac59adf [PATCH] x86_64: F... |
1771 |
#ifdef CONFIG_HUGETLBFS |
480eccf9a Fix NUMA Memory P... |
1772 1773 1774 1775 1776 |
/* * 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 ... |
1777 1778 |
* @mpol = pointer to mempolicy pointer for reference counted mempolicy * @nodemask = pointer to nodemask pointer for MPOL_BIND nodemask |
480eccf9a Fix NUMA Memory P... |
1779 |
* |
52cd3b074 mempolicy: rework... |
1780 1781 1782 1783 |
* 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. |
c0ff7453b cpuset,mm: fix no... |
1784 |
* |
d26914d11 mm: optimize put_... |
1785 |
* Must be protected by read_mems_allowed_begin() |
480eccf9a Fix NUMA Memory P... |
1786 |
*/ |
396faf030 Allow huge page a... |
1787 |
struct zonelist *huge_zonelist(struct vm_area_struct *vma, unsigned long addr, |
19770b326 mm: filter based ... |
1788 1789 |
gfp_t gfp_flags, struct mempolicy **mpol, nodemask_t **nodemask) |
5da7ca860 [PATCH] Add NUMA ... |
1790 |
{ |
480eccf9a Fix NUMA Memory P... |
1791 |
struct zonelist *zl; |
5da7ca860 [PATCH] Add NUMA ... |
1792 |
|
52cd3b074 mempolicy: rework... |
1793 |
*mpol = get_vma_policy(current, vma, addr); |
19770b326 mm: filter based ... |
1794 |
*nodemask = NULL; /* assume !MPOL_BIND */ |
5da7ca860 [PATCH] Add NUMA ... |
1795 |
|
52cd3b074 mempolicy: rework... |
1796 1797 |
if (unlikely((*mpol)->mode == MPOL_INTERLEAVE)) { zl = node_zonelist(interleave_nid(*mpol, vma, addr, |
a55164389 hugetlb: modular ... |
1798 |
huge_page_shift(hstate_vma(vma))), gfp_flags); |
52cd3b074 mempolicy: rework... |
1799 |
} else { |
2f5f9486f mm: change alloc_... |
1800 |
zl = policy_zonelist(gfp_flags, *mpol, numa_node_id()); |
52cd3b074 mempolicy: rework... |
1801 1802 |
if ((*mpol)->mode == MPOL_BIND) *nodemask = &(*mpol)->v.nodes; |
480eccf9a Fix NUMA Memory P... |
1803 1804 |
} return zl; |
5da7ca860 [PATCH] Add NUMA ... |
1805 |
} |
06808b082 hugetlb: derive h... |
1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 |
/* * init_nodemask_of_mempolicy * * If the current task's mempolicy is "default" [NULL], return 'false' * to indicate default policy. Otherwise, extract the policy nodemask * for 'bind' or 'interleave' policy into the argument nodemask, or * initialize the argument nodemask to contain the single node for * 'preferred' or 'local' policy and return 'true' to indicate presence * of non-default mempolicy. * * We don't bother with reference counting the mempolicy [mpol_get/put] * because the current task is examining it's own mempolicy and a task's * mempolicy is only ever changed by the task itself. * * N.B., it is the caller's responsibility to free a returned nodemask. */ bool init_nodemask_of_mempolicy(nodemask_t *mask) { struct mempolicy *mempolicy; int nid; if (!(mask && current->mempolicy)) return false; |
c0ff7453b cpuset,mm: fix no... |
1830 |
task_lock(current); |
06808b082 hugetlb: derive h... |
1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 |
mempolicy = current->mempolicy; switch (mempolicy->mode) { case MPOL_PREFERRED: if (mempolicy->flags & MPOL_F_LOCAL) nid = numa_node_id(); else nid = mempolicy->v.preferred_node; init_nodemask_of_node(mask, nid); break; case MPOL_BIND: /* Fall through */ case MPOL_INTERLEAVE: *mask = mempolicy->v.nodes; break; default: BUG(); } |
c0ff7453b cpuset,mm: fix no... |
1850 |
task_unlock(current); |
06808b082 hugetlb: derive h... |
1851 1852 1853 |
return true; } |
00ac59adf [PATCH] x86_64: F... |
1854 |
#endif |
5da7ca860 [PATCH] Add NUMA ... |
1855 |
|
6f48d0ebd oom: select task ... |
1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 |
/* * mempolicy_nodemask_intersects * * If tsk's mempolicy is "default" [NULL], return 'true' to indicate default * policy. Otherwise, check for intersection between mask and the policy * nodemask for 'bind' or 'interleave' policy. For 'perferred' or 'local' * policy, always return true since it may allocate elsewhere on fallback. * * Takes task_lock(tsk) to prevent freeing of its mempolicy. */ bool mempolicy_nodemask_intersects(struct task_struct *tsk, const nodemask_t *mask) { struct mempolicy *mempolicy; bool ret = true; if (!mask) return ret; task_lock(tsk); mempolicy = tsk->mempolicy; if (!mempolicy) goto out; switch (mempolicy->mode) { case MPOL_PREFERRED: /* * MPOL_PREFERRED and MPOL_F_LOCAL are only preferred nodes to * allocate from, they may fallback to other nodes when oom. * Thus, it's possible for tsk to have allocated memory from * nodes in mask. */ break; case MPOL_BIND: case MPOL_INTERLEAVE: ret = nodes_intersects(mempolicy->v.nodes, *mask); break; default: BUG(); } out: task_unlock(tsk); return ret; } |
1da177e4c Linux-2.6.12-rc2 |
1899 1900 |
/* Allocate a page in interleaved policy. Own path because it needs to do special accounting. */ |
662f3a0b9 [PATCH] Remove ne... |
1901 1902 |
static struct page *alloc_page_interleave(gfp_t gfp, unsigned order, unsigned nid) |
1da177e4c Linux-2.6.12-rc2 |
1903 1904 1905 |
{ struct zonelist *zl; struct page *page; |
0e88460da mm: introduce nod... |
1906 |
zl = node_zonelist(nid, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1907 |
page = __alloc_pages(gfp, order, zl); |
dd1a239f6 mm: have zonelist... |
1908 |
if (page && page_zone(page) == zonelist_zone(&zl->_zonerefs[0])) |
ca889e6c4 [PATCH] Use Zoned... |
1909 |
inc_zone_page_state(page, NUMA_INTERLEAVE_HIT); |
1da177e4c Linux-2.6.12-rc2 |
1910 1911 1912 1913 |
return page; } /** |
0bbbc0b33 thp: add numa awa... |
1914 |
* alloc_pages_vma - Allocate a page for a VMA. |
1da177e4c Linux-2.6.12-rc2 |
1915 1916 1917 1918 1919 1920 1921 1922 |
* * @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. * |
0bbbc0b33 thp: add numa awa... |
1923 |
* @order:Order of the GFP allocation. |
1da177e4c Linux-2.6.12-rc2 |
1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 |
* @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 * |
0bbbc0b33 thp: add numa awa... |
1937 |
alloc_pages_vma(gfp_t gfp, int order, struct vm_area_struct *vma, |
2f5f9486f mm: change alloc_... |
1938 |
unsigned long addr, int node) |
1da177e4c Linux-2.6.12-rc2 |
1939 |
{ |
cc9a6c877 cpuset: mm: reduc... |
1940 |
struct mempolicy *pol; |
c0ff7453b cpuset,mm: fix no... |
1941 |
struct page *page; |
cc9a6c877 cpuset: mm: reduc... |
1942 1943 1944 1945 |
unsigned int cpuset_mems_cookie; retry_cpuset: pol = get_vma_policy(current, vma, addr); |
d26914d11 mm: optimize put_... |
1946 |
cpuset_mems_cookie = read_mems_allowed_begin(); |
1da177e4c Linux-2.6.12-rc2 |
1947 |
|
45c4745af mempolicy: rename... |
1948 |
if (unlikely(pol->mode == MPOL_INTERLEAVE)) { |
1da177e4c Linux-2.6.12-rc2 |
1949 |
unsigned nid; |
5da7ca860 [PATCH] Add NUMA ... |
1950 |
|
8eac563c1 thp: fix interlea... |
1951 |
nid = interleave_nid(pol, vma, addr, PAGE_SHIFT + order); |
52cd3b074 mempolicy: rework... |
1952 |
mpol_cond_put(pol); |
0bbbc0b33 thp: add numa awa... |
1953 |
page = alloc_page_interleave(gfp, order, nid); |
d26914d11 mm: optimize put_... |
1954 |
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) |
cc9a6c877 cpuset: mm: reduc... |
1955 |
goto retry_cpuset; |
c0ff7453b cpuset,mm: fix no... |
1956 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1957 |
} |
212a0a6f2 mm, mempolicy: re... |
1958 1959 |
page = __alloc_pages_nodemask(gfp, order, policy_zonelist(gfp, pol, node), |
0bbbc0b33 thp: add numa awa... |
1960 |
policy_nodemask(gfp, pol)); |
212a0a6f2 mm, mempolicy: re... |
1961 1962 |
if (unlikely(mpol_needs_cond_ref(pol))) __mpol_put(pol); |
d26914d11 mm: optimize put_... |
1963 |
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) |
cc9a6c877 cpuset: mm: reduc... |
1964 |
goto retry_cpuset; |
c0ff7453b cpuset,mm: fix no... |
1965 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 |
} /** * 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... |
1983 |
* Don't call cpuset_update_task_memory_state() unless |
1da177e4c Linux-2.6.12-rc2 |
1984 1985 1986 |
* 1) it's ok to take cpuset_sem (can WAIT), and * 2) allocating for current task (not interrupt). */ |
dd0fc66fb [PATCH] gfp flags... |
1987 |
struct page *alloc_pages_current(gfp_t gfp, unsigned order) |
1da177e4c Linux-2.6.12-rc2 |
1988 |
{ |
5606e3877 mm: numa: Migrate... |
1989 |
struct mempolicy *pol = get_task_policy(current); |
c0ff7453b cpuset,mm: fix no... |
1990 |
struct page *page; |
cc9a6c877 cpuset: mm: reduc... |
1991 |
unsigned int cpuset_mems_cookie; |
1da177e4c Linux-2.6.12-rc2 |
1992 |
|
9b819d204 [PATCH] Add __GFP... |
1993 |
if (!pol || in_interrupt() || (gfp & __GFP_THISNODE)) |
1da177e4c Linux-2.6.12-rc2 |
1994 |
pol = &default_policy; |
52cd3b074 mempolicy: rework... |
1995 |
|
cc9a6c877 cpuset: mm: reduc... |
1996 |
retry_cpuset: |
d26914d11 mm: optimize put_... |
1997 |
cpuset_mems_cookie = read_mems_allowed_begin(); |
cc9a6c877 cpuset: mm: reduc... |
1998 |
|
52cd3b074 mempolicy: rework... |
1999 2000 2001 2002 |
/* * No reference counting needed for current->mempolicy * nor system default_policy */ |
45c4745af mempolicy: rename... |
2003 |
if (pol->mode == MPOL_INTERLEAVE) |
c0ff7453b cpuset,mm: fix no... |
2004 2005 2006 |
page = alloc_page_interleave(gfp, order, interleave_nodes(pol)); else page = __alloc_pages_nodemask(gfp, order, |
5c4b4be3b mm: use correct n... |
2007 2008 |
policy_zonelist(gfp, pol, numa_node_id()), policy_nodemask(gfp, pol)); |
cc9a6c877 cpuset: mm: reduc... |
2009 |
|
d26914d11 mm: optimize put_... |
2010 |
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) |
cc9a6c877 cpuset: mm: reduc... |
2011 |
goto retry_cpuset; |
c0ff7453b cpuset,mm: fix no... |
2012 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
2013 2014 |
} EXPORT_SYMBOL(alloc_pages_current); |
ef0855d33 mm: mempolicy: tu... |
2015 2016 2017 2018 2019 2020 2021 2022 2023 |
int vma_dup_policy(struct vm_area_struct *src, struct vm_area_struct *dst) { struct mempolicy *pol = mpol_dup(vma_policy(src)); if (IS_ERR(pol)) return PTR_ERR(pol); dst->vm_policy = pol; return 0; } |
4225399a6 [PATCH] cpuset: r... |
2024 |
/* |
846a16bf0 mempolicy: rename... |
2025 |
* If mpol_dup() sees current->cpuset == cpuset_being_rebound, then it |
4225399a6 [PATCH] cpuset: r... |
2026 2027 2028 2029 |
* 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(). |
708c1bbc9 mempolicy: restru... |
2030 2031 2032 |
* * current's mempolicy may be rebinded by the other task(the task that changes * cpuset's mems), so we needn't do rebind work for current task. |
4225399a6 [PATCH] cpuset: r... |
2033 |
*/ |
4225399a6 [PATCH] cpuset: r... |
2034 |
|
846a16bf0 mempolicy: rename... |
2035 2036 |
/* Slow path of a mempolicy duplicate */ struct mempolicy *__mpol_dup(struct mempolicy *old) |
1da177e4c Linux-2.6.12-rc2 |
2037 2038 2039 2040 2041 |
{ struct mempolicy *new = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!new) return ERR_PTR(-ENOMEM); |
708c1bbc9 mempolicy: restru... |
2042 2043 2044 2045 2046 2047 2048 2049 |
/* task's mempolicy is protected by alloc_lock */ if (old == current->mempolicy) { task_lock(current); *new = *old; task_unlock(current); } else *new = *old; |
99ee4ca74 rcu: Suppress __m... |
2050 |
rcu_read_lock(); |
4225399a6 [PATCH] cpuset: r... |
2051 2052 |
if (current_cpuset_is_being_rebound()) { nodemask_t mems = cpuset_mems_allowed(current); |
708c1bbc9 mempolicy: restru... |
2053 2054 2055 2056 |
if (new->flags & MPOL_F_REBINDING) mpol_rebind_policy(new, &mems, MPOL_REBIND_STEP2); else mpol_rebind_policy(new, &mems, MPOL_REBIND_ONCE); |
4225399a6 [PATCH] cpuset: r... |
2057 |
} |
99ee4ca74 rcu: Suppress __m... |
2058 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
2059 |
atomic_set(&new->refcnt, 1); |
1da177e4c Linux-2.6.12-rc2 |
2060 2061 2062 2063 |
return new; } /* Slow path of a mempolicy comparison */ |
fcfb4dcc9 mm/mempolicy.c: m... |
2064 |
bool __mpol_equal(struct mempolicy *a, struct mempolicy *b) |
1da177e4c Linux-2.6.12-rc2 |
2065 2066 |
{ if (!a || !b) |
fcfb4dcc9 mm/mempolicy.c: m... |
2067 |
return false; |
45c4745af mempolicy: rename... |
2068 |
if (a->mode != b->mode) |
fcfb4dcc9 mm/mempolicy.c: m... |
2069 |
return false; |
198005025 mempolicy: remove... |
2070 |
if (a->flags != b->flags) |
fcfb4dcc9 mm/mempolicy.c: m... |
2071 |
return false; |
198005025 mempolicy: remove... |
2072 2073 |
if (mpol_store_user_nodemask(a)) if (!nodes_equal(a->w.user_nodemask, b->w.user_nodemask)) |
fcfb4dcc9 mm/mempolicy.c: m... |
2074 |
return false; |
198005025 mempolicy: remove... |
2075 |
|
45c4745af mempolicy: rename... |
2076 |
switch (a->mode) { |
19770b326 mm: filter based ... |
2077 2078 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
2079 |
case MPOL_INTERLEAVE: |
fcfb4dcc9 mm/mempolicy.c: m... |
2080 |
return !!nodes_equal(a->v.nodes, b->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
2081 |
case MPOL_PREFERRED: |
757196618 mempolicy: remove... |
2082 |
return a->v.preferred_node == b->v.preferred_node; |
1da177e4c Linux-2.6.12-rc2 |
2083 2084 |
default: BUG(); |
fcfb4dcc9 mm/mempolicy.c: m... |
2085 |
return false; |
1da177e4c Linux-2.6.12-rc2 |
2086 2087 |
} } |
1da177e4c Linux-2.6.12-rc2 |
2088 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2089 2090 2091 2092 2093 2094 2095 2096 2097 |
* 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 */ |
42288fe36 mm: mempolicy: Co... |
2098 |
/* Caller holds sp->lock */ |
1da177e4c Linux-2.6.12-rc2 |
2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 |
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: ... |
2149 2150 |
pr_debug("inserting %lx-%lx: %d ", new->start, new->end, |
45c4745af mempolicy: rename... |
2151 |
new->policy ? new->policy->mode : 0); |
1da177e4c Linux-2.6.12-rc2 |
2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 |
} /* 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; |
42288fe36 mm: mempolicy: Co... |
2163 |
spin_lock(&sp->lock); |
1da177e4c Linux-2.6.12-rc2 |
2164 2165 2166 2167 2168 |
sn = sp_lookup(sp, idx, idx+1); if (sn) { mpol_get(sn->policy); pol = sn->policy; } |
42288fe36 mm: mempolicy: Co... |
2169 |
spin_unlock(&sp->lock); |
1da177e4c Linux-2.6.12-rc2 |
2170 2171 |
return pol; } |
63f74ca21 mempolicy: fix re... |
2172 2173 2174 2175 2176 |
static void sp_free(struct sp_node *n) { mpol_put(n->policy); kmem_cache_free(sn_cache, n); } |
771fb4d80 mm: mempolicy: Ch... |
2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 |
/** * mpol_misplaced - check whether current page node is valid in policy * * @page - page to be checked * @vma - vm area where page mapped * @addr - virtual address where page mapped * * Lookup current policy node id for vma,addr and "compare to" page's * node id. * * Returns: * -1 - not misplaced, page is in the right node * node - node id where the page should be * * Policy determination "mimics" alloc_page_vma(). * Called from fault path where we know the vma and faulting address. */ int mpol_misplaced(struct page *page, struct vm_area_struct *vma, unsigned long addr) { struct mempolicy *pol; struct zone *zone; int curnid = page_to_nid(page); unsigned long pgoff; |
90572890d mm: numa: Change ... |
2200 2201 |
int thiscpu = raw_smp_processor_id(); int thisnid = cpu_to_node(thiscpu); |
771fb4d80 mm: mempolicy: Ch... |
2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 |
int polnid = -1; int ret = -1; BUG_ON(!vma); pol = get_vma_policy(current, vma, addr); if (!(pol->flags & MPOL_F_MOF)) goto out; switch (pol->mode) { case MPOL_INTERLEAVE: BUG_ON(addr >= vma->vm_end); BUG_ON(addr < vma->vm_start); pgoff = vma->vm_pgoff; pgoff += (addr - vma->vm_start) >> PAGE_SHIFT; polnid = offset_il_node(pol, vma, pgoff); break; case MPOL_PREFERRED: if (pol->flags & MPOL_F_LOCAL) polnid = numa_node_id(); else polnid = pol->v.preferred_node; break; case MPOL_BIND: /* * allows binding to multiple nodes. * use current page if in policy nodemask, * else select nearest allowed node, if any. * If no allowed nodes, use current [!misplaced]. */ if (node_isset(curnid, pol->v.nodes)) goto out; (void)first_zones_zonelist( node_zonelist(numa_node_id(), GFP_HIGHUSER), gfp_zone(GFP_HIGHUSER), &pol->v.nodes, &zone); polnid = zone->node; break; default: BUG(); } |
5606e3877 mm: numa: Migrate... |
2247 2248 |
/* Migrate the page towards the node whose CPU is referencing it */ |
e42c8ff29 mm: numa: Use a t... |
2249 |
if (pol->flags & MPOL_F_MORON) { |
90572890d mm: numa: Change ... |
2250 |
polnid = thisnid; |
5606e3877 mm: numa: Migrate... |
2251 |
|
10f390427 sched/numa, mm: U... |
2252 |
if (!should_numa_migrate_memory(current, page, curnid, thiscpu)) |
de1c9ce6f sched/numa: Skip ... |
2253 |
goto out; |
e42c8ff29 mm: numa: Use a t... |
2254 |
} |
771fb4d80 mm: mempolicy: Ch... |
2255 2256 2257 2258 2259 2260 2261 |
if (curnid != polnid) ret = polnid; out: mpol_cond_put(pol); return ret; } |
1da177e4c Linux-2.6.12-rc2 |
2262 2263 |
static void sp_delete(struct shared_policy *sp, struct sp_node *n) { |
140d5a490 numa: mempolicy: ... |
2264 2265 |
pr_debug("deleting %lx-l%lx ", n->start, n->end); |
1da177e4c Linux-2.6.12-rc2 |
2266 |
rb_erase(&n->nd, &sp->root); |
63f74ca21 mempolicy: fix re... |
2267 |
sp_free(n); |
1da177e4c Linux-2.6.12-rc2 |
2268 |
} |
42288fe36 mm: mempolicy: Co... |
2269 2270 2271 2272 2273 2274 2275 |
static void sp_node_init(struct sp_node *node, unsigned long start, unsigned long end, struct mempolicy *pol) { node->start = start; node->end = end; node->policy = pol; } |
dbcb0f19c mm/mempolicy.c: c... |
2276 2277 |
static struct sp_node *sp_alloc(unsigned long start, unsigned long end, struct mempolicy *pol) |
1da177e4c Linux-2.6.12-rc2 |
2278 |
{ |
869833f2c mempolicy: remove... |
2279 2280 |
struct sp_node *n; struct mempolicy *newpol; |
1da177e4c Linux-2.6.12-rc2 |
2281 |
|
869833f2c mempolicy: remove... |
2282 |
n = kmem_cache_alloc(sn_cache, GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
2283 2284 |
if (!n) return NULL; |
869833f2c mempolicy: remove... |
2285 2286 2287 2288 2289 2290 2291 |
newpol = mpol_dup(pol); if (IS_ERR(newpol)) { kmem_cache_free(sn_cache, n); return NULL; } newpol->flags |= MPOL_F_SHARED; |
42288fe36 mm: mempolicy: Co... |
2292 |
sp_node_init(n, start, end, newpol); |
869833f2c mempolicy: remove... |
2293 |
|
1da177e4c Linux-2.6.12-rc2 |
2294 2295 2296 2297 2298 2299 2300 |
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) { |
b22d127a3 mempolicy: fix a ... |
2301 |
struct sp_node *n; |
42288fe36 mm: mempolicy: Co... |
2302 2303 |
struct sp_node *n_new = NULL; struct mempolicy *mpol_new = NULL; |
b22d127a3 mempolicy: fix a ... |
2304 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2305 |
|
42288fe36 mm: mempolicy: Co... |
2306 2307 |
restart: spin_lock(&sp->lock); |
1da177e4c Linux-2.6.12-rc2 |
2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 |
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) { |
42288fe36 mm: mempolicy: Co... |
2320 2321 2322 2323 2324 |
if (!n_new) goto alloc_new; *mpol_new = *n->policy; atomic_set(&mpol_new->refcnt, 1); |
7880639c3 mm/mempolicy.c: f... |
2325 |
sp_node_init(n_new, end, n->end, mpol_new); |
1da177e4c Linux-2.6.12-rc2 |
2326 |
n->end = start; |
5ca395751 mm/mempolicy.c: f... |
2327 |
sp_insert(sp, n_new); |
42288fe36 mm: mempolicy: Co... |
2328 2329 |
n_new = NULL; mpol_new = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 |
break; } else n->end = start; } if (!next) break; n = rb_entry(next, struct sp_node, nd); } if (new) sp_insert(sp, new); |
42288fe36 mm: mempolicy: Co... |
2340 2341 2342 2343 2344 2345 2346 2347 |
spin_unlock(&sp->lock); ret = 0; err_out: if (mpol_new) mpol_put(mpol_new); if (n_new) kmem_cache_free(sn_cache, n_new); |
b22d127a3 mempolicy: fix a ... |
2348 |
return ret; |
42288fe36 mm: mempolicy: Co... |
2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 |
alloc_new: spin_unlock(&sp->lock); ret = -ENOMEM; n_new = kmem_cache_alloc(sn_cache, GFP_KERNEL); if (!n_new) goto err_out; mpol_new = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!mpol_new) goto err_out; goto restart; |
1da177e4c Linux-2.6.12-rc2 |
2360 |
} |
71fe804b6 mempolicy: use st... |
2361 2362 2363 2364 2365 2366 2367 2368 |
/** * 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. |
4bfc44958 mm: make set_memp... |
2369 |
* This is called at get_inode() calls and we can use GFP_KERNEL. |
71fe804b6 mempolicy: use st... |
2370 2371 2372 |
*/ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) { |
58568d2a8 cpuset,mm: update... |
2373 |
int ret; |
71fe804b6 mempolicy: use st... |
2374 |
sp->root = RB_ROOT; /* empty tree == default mempolicy */ |
42288fe36 mm: mempolicy: Co... |
2375 |
spin_lock_init(&sp->lock); |
71fe804b6 mempolicy: use st... |
2376 2377 2378 2379 |
if (mpol) { struct vm_area_struct pvma; struct mempolicy *new; |
4bfc44958 mm: make set_memp... |
2380 |
NODEMASK_SCRATCH(scratch); |
71fe804b6 mempolicy: use st... |
2381 |
|
4bfc44958 mm: make set_memp... |
2382 |
if (!scratch) |
5c0c16549 mempolicy: fix da... |
2383 |
goto put_mpol; |
71fe804b6 mempolicy: use st... |
2384 2385 |
/* contextualize the tmpfs mount point mempolicy */ new = mpol_new(mpol->mode, mpol->flags, &mpol->w.user_nodemask); |
15d77835a mempolicy: factor... |
2386 |
if (IS_ERR(new)) |
0cae3457b mempolicy: ERR_PT... |
2387 |
goto free_scratch; /* no valid nodemask intersection */ |
58568d2a8 cpuset,mm: update... |
2388 2389 |
task_lock(current); |
4bfc44958 mm: make set_memp... |
2390 |
ret = mpol_set_nodemask(new, &mpol->w.user_nodemask, scratch); |
58568d2a8 cpuset,mm: update... |
2391 |
task_unlock(current); |
15d77835a mempolicy: factor... |
2392 |
if (ret) |
5c0c16549 mempolicy: fix da... |
2393 |
goto put_new; |
71fe804b6 mempolicy: use st... |
2394 2395 2396 2397 2398 |
/* 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 */ |
15d77835a mempolicy: factor... |
2399 |
|
5c0c16549 mempolicy: fix da... |
2400 |
put_new: |
71fe804b6 mempolicy: use st... |
2401 |
mpol_put(new); /* drop initial ref */ |
0cae3457b mempolicy: ERR_PT... |
2402 |
free_scratch: |
4bfc44958 mm: make set_memp... |
2403 |
NODEMASK_SCRATCH_FREE(scratch); |
5c0c16549 mempolicy: fix da... |
2404 2405 |
put_mpol: mpol_put(mpol); /* drop our incoming ref on sb mpol */ |
7339ff830 [PATCH] Add tmpfs... |
2406 2407 |
} } |
1da177e4c Linux-2.6.12-rc2 |
2408 2409 2410 2411 2412 2413 |
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... |
2414 2415 |
pr_debug("set_shared_policy %lx sz %lu %d %d %lx ", |
1da177e4c Linux-2.6.12-rc2 |
2416 |
vma->vm_pgoff, |
45c4745af mempolicy: rename... |
2417 |
sz, npol ? npol->mode : -1, |
028fec414 mempolicy: suppor... |
2418 |
npol ? npol->flags : -1, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2419 |
npol ? nodes_addr(npol->v.nodes)[0] : NUMA_NO_NODE); |
1da177e4c Linux-2.6.12-rc2 |
2420 2421 2422 2423 2424 2425 2426 2427 |
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) |
63f74ca21 mempolicy: fix re... |
2428 |
sp_free(new); |
1da177e4c Linux-2.6.12-rc2 |
2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 |
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; |
42288fe36 mm: mempolicy: Co... |
2440 |
spin_lock(&p->lock); |
1da177e4c Linux-2.6.12-rc2 |
2441 2442 2443 2444 |
next = rb_first(&p->root); while (next) { n = rb_entry(next, struct sp_node, nd); next = rb_next(&n->nd); |
63f74ca21 mempolicy: fix re... |
2445 |
sp_delete(p, n); |
1da177e4c Linux-2.6.12-rc2 |
2446 |
} |
42288fe36 mm: mempolicy: Co... |
2447 |
spin_unlock(&p->lock); |
1da177e4c Linux-2.6.12-rc2 |
2448 |
} |
1a687c2e9 mm: sched: numa: ... |
2449 |
#ifdef CONFIG_NUMA_BALANCING |
c297663c0 mm: numa: initial... |
2450 |
static int __initdata numabalancing_override; |
1a687c2e9 mm: sched: numa: ... |
2451 2452 2453 2454 2455 2456 2457 |
static void __init check_numabalancing_enable(void) { bool numabalancing_default = false; if (IS_ENABLED(CONFIG_NUMA_BALANCING_DEFAULT_ENABLED)) numabalancing_default = true; |
c297663c0 mm: numa: initial... |
2458 2459 2460 |
/* Parsed by setup_numabalancing. override == 1 enables, -1 disables */ if (numabalancing_override) set_numabalancing_state(numabalancing_override == 1); |
1a687c2e9 mm: sched: numa: ... |
2461 |
if (nr_node_ids > 1 && !numabalancing_override) { |
4a404bea9 mm/mempolicy.c: c... |
2462 |
pr_info("%s automatic NUMA balancing. " |
c297663c0 mm: numa: initial... |
2463 2464 2465 |
"Configure with numa_balancing= or the " "kernel.numa_balancing sysctl", numabalancing_default ? "Enabling" : "Disabling"); |
1a687c2e9 mm: sched: numa: ... |
2466 2467 2468 2469 2470 2471 2472 2473 2474 |
set_numabalancing_state(numabalancing_default); } } static int __init setup_numabalancing(char *str) { int ret = 0; if (!str) goto out; |
1a687c2e9 mm: sched: numa: ... |
2475 2476 |
if (!strcmp(str, "enable")) { |
c297663c0 mm: numa: initial... |
2477 |
numabalancing_override = 1; |
1a687c2e9 mm: sched: numa: ... |
2478 2479 |
ret = 1; } else if (!strcmp(str, "disable")) { |
c297663c0 mm: numa: initial... |
2480 |
numabalancing_override = -1; |
1a687c2e9 mm: sched: numa: ... |
2481 2482 2483 2484 |
ret = 1; } out: if (!ret) |
4a404bea9 mm/mempolicy.c: c... |
2485 2486 |
pr_warn("Unable to parse numa_balancing= "); |
1a687c2e9 mm: sched: numa: ... |
2487 2488 2489 2490 2491 2492 2493 2494 2495 |
return ret; } __setup("numa_balancing=", setup_numabalancing); #else static inline void __init check_numabalancing_enable(void) { } #endif /* CONFIG_NUMA_BALANCING */ |
1da177e4c Linux-2.6.12-rc2 |
2496 2497 2498 |
/* assumes fs == KERNEL_DS */ void __init numa_policy_init(void) { |
b71636e29 numa: mempolicy: ... |
2499 2500 2501 |
nodemask_t interleave_nodes; unsigned long largest = 0; int nid, prefer = 0; |
1da177e4c Linux-2.6.12-rc2 |
2502 2503 |
policy_cache = kmem_cache_create("numa_policy", sizeof(struct mempolicy), |
20c2df83d mm: Remove slab d... |
2504 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2505 2506 2507 |
sn_cache = kmem_cache_create("shared_policy_node", sizeof(struct sp_node), |
20c2df83d mm: Remove slab d... |
2508 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2509 |
|
5606e3877 mm: numa: Migrate... |
2510 2511 2512 2513 2514 2515 2516 2517 |
for_each_node(nid) { preferred_node_policy[nid] = (struct mempolicy) { .refcnt = ATOMIC_INIT(1), .mode = MPOL_PREFERRED, .flags = MPOL_F_MOF | MPOL_F_MORON, .v = { .preferred_node = nid, }, }; } |
b71636e29 numa: mempolicy: ... |
2518 2519 2520 2521 2522 2523 |
/* * 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); |
01f13bd60 mempolicy: use N_... |
2524 |
for_each_node_state(nid, N_MEMORY) { |
b71636e29 numa: mempolicy: ... |
2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 |
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 |
2541 |
|
028fec414 mempolicy: suppor... |
2542 |
if (do_set_mempolicy(MPOL_INTERLEAVE, 0, &interleave_nodes)) |
1da177e4c Linux-2.6.12-rc2 |
2543 2544 |
printk("numa_policy_init: interleaving failed "); |
1a687c2e9 mm: sched: numa: ... |
2545 2546 |
check_numabalancing_enable(); |
1da177e4c Linux-2.6.12-rc2 |
2547 |
} |
8bccd85ff [PATCH] Implement... |
2548 |
/* Reset policy of current process to default */ |
1da177e4c Linux-2.6.12-rc2 |
2549 2550 |
void numa_default_policy(void) { |
028fec414 mempolicy: suppor... |
2551 |
do_set_mempolicy(MPOL_DEFAULT, 0, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2552 |
} |
68860ec10 [PATCH] cpusets: ... |
2553 |
|
4225399a6 [PATCH] cpuset: r... |
2554 |
/* |
095f1fc4e mempolicy: rework... |
2555 2556 2557 2558 |
* Parse and format mempolicy from/to strings */ /* |
f2a07f40d tmpfs mempolicy: ... |
2559 |
* "local" is implemented internally by MPOL_PREFERRED with MPOL_F_LOCAL flag. |
1a75a6c82 [PATCH] Fold numa... |
2560 |
*/ |
345ace9c7 mempolicy: rename... |
2561 2562 2563 2564 2565 2566 |
static const char * const policy_modes[] = { [MPOL_DEFAULT] = "default", [MPOL_PREFERRED] = "prefer", [MPOL_BIND] = "bind", [MPOL_INTERLEAVE] = "interleave", |
d3a710337 mm: mempolicy: Ad... |
2567 |
[MPOL_LOCAL] = "local", |
345ace9c7 mempolicy: rename... |
2568 |
}; |
1a75a6c82 [PATCH] Fold numa... |
2569 |
|
095f1fc4e mempolicy: rework... |
2570 2571 2572 |
#ifdef CONFIG_TMPFS /** |
f2a07f40d tmpfs mempolicy: ... |
2573 |
* mpol_parse_str - parse string to mempolicy, for tmpfs mpol mount option. |
095f1fc4e mempolicy: rework... |
2574 |
* @str: string containing mempolicy to parse |
71fe804b6 mempolicy: use st... |
2575 |
* @mpol: pointer to struct mempolicy pointer, returned on success. |
095f1fc4e mempolicy: rework... |
2576 2577 2578 2579 |
* * Format of input: * <mode>[=<flags>][:<nodelist>] * |
71fe804b6 mempolicy: use st... |
2580 |
* On success, returns 0, else 1 |
095f1fc4e mempolicy: rework... |
2581 |
*/ |
a7a88b237 mempolicy: remove... |
2582 |
int mpol_parse_str(char *str, struct mempolicy **mpol) |
095f1fc4e mempolicy: rework... |
2583 |
{ |
71fe804b6 mempolicy: use st... |
2584 |
struct mempolicy *new = NULL; |
b4652e842 mempolicy: lose u... |
2585 |
unsigned short mode; |
f2a07f40d tmpfs mempolicy: ... |
2586 |
unsigned short mode_flags; |
71fe804b6 mempolicy: use st... |
2587 |
nodemask_t nodes; |
095f1fc4e mempolicy: rework... |
2588 2589 |
char *nodelist = strchr(str, ':'); char *flags = strchr(str, '='); |
095f1fc4e mempolicy: rework... |
2590 2591 2592 2593 2594 |
int err = 1; if (nodelist) { /* NUL-terminate mode or flags string */ *nodelist++ = '\0'; |
71fe804b6 mempolicy: use st... |
2595 |
if (nodelist_parse(nodelist, nodes)) |
095f1fc4e mempolicy: rework... |
2596 |
goto out; |
01f13bd60 mempolicy: use N_... |
2597 |
if (!nodes_subset(nodes, node_states[N_MEMORY])) |
095f1fc4e mempolicy: rework... |
2598 |
goto out; |
71fe804b6 mempolicy: use st... |
2599 2600 |
} else nodes_clear(nodes); |
095f1fc4e mempolicy: rework... |
2601 2602 |
if (flags) *flags++ = '\0'; /* terminate mode string */ |
479e2802d mm: mempolicy: Ma... |
2603 |
for (mode = 0; mode < MPOL_MAX; mode++) { |
345ace9c7 mempolicy: rename... |
2604 |
if (!strcmp(str, policy_modes[mode])) { |
095f1fc4e mempolicy: rework... |
2605 2606 2607 |
break; } } |
a720094de mm: mempolicy: Hi... |
2608 |
if (mode >= MPOL_MAX) |
095f1fc4e mempolicy: rework... |
2609 |
goto out; |
71fe804b6 mempolicy: use st... |
2610 |
switch (mode) { |
095f1fc4e mempolicy: rework... |
2611 |
case MPOL_PREFERRED: |
71fe804b6 mempolicy: use st... |
2612 2613 2614 |
/* * Insist on a nodelist of one node only */ |
095f1fc4e mempolicy: rework... |
2615 2616 2617 2618 |
if (nodelist) { char *rest = nodelist; while (isdigit(*rest)) rest++; |
926f2ae04 tmpfs: cleanup mp... |
2619 2620 |
if (*rest) goto out; |
095f1fc4e mempolicy: rework... |
2621 2622 |
} break; |
095f1fc4e mempolicy: rework... |
2623 2624 2625 2626 2627 |
case MPOL_INTERLEAVE: /* * Default to online nodes with memory if no nodelist */ if (!nodelist) |
01f13bd60 mempolicy: use N_... |
2628 |
nodes = node_states[N_MEMORY]; |
3f226aa1c mempolicy: suppor... |
2629 |
break; |
71fe804b6 mempolicy: use st... |
2630 |
case MPOL_LOCAL: |
3f226aa1c mempolicy: suppor... |
2631 |
/* |
71fe804b6 mempolicy: use st... |
2632 |
* Don't allow a nodelist; mpol_new() checks flags |
3f226aa1c mempolicy: suppor... |
2633 |
*/ |
71fe804b6 mempolicy: use st... |
2634 |
if (nodelist) |
3f226aa1c mempolicy: suppor... |
2635 |
goto out; |
71fe804b6 mempolicy: use st... |
2636 |
mode = MPOL_PREFERRED; |
3f226aa1c mempolicy: suppor... |
2637 |
break; |
413b43dea tmpfs: fix oops o... |
2638 2639 2640 2641 2642 2643 2644 |
case MPOL_DEFAULT: /* * Insist on a empty nodelist */ if (!nodelist) err = 0; goto out; |
d69b2e63e tmpfs: mpol=bind:... |
2645 2646 2647 2648 2649 2650 |
case MPOL_BIND: /* * Insist on a nodelist */ if (!nodelist) goto out; |
095f1fc4e mempolicy: rework... |
2651 |
} |
71fe804b6 mempolicy: use st... |
2652 |
mode_flags = 0; |
095f1fc4e mempolicy: rework... |
2653 2654 2655 2656 2657 2658 |
if (flags) { /* * Currently, we only support two mutually exclusive * mode flags. */ if (!strcmp(flags, "static")) |
71fe804b6 mempolicy: use st... |
2659 |
mode_flags |= MPOL_F_STATIC_NODES; |
095f1fc4e mempolicy: rework... |
2660 |
else if (!strcmp(flags, "relative")) |
71fe804b6 mempolicy: use st... |
2661 |
mode_flags |= MPOL_F_RELATIVE_NODES; |
095f1fc4e mempolicy: rework... |
2662 |
else |
926f2ae04 tmpfs: cleanup mp... |
2663 |
goto out; |
095f1fc4e mempolicy: rework... |
2664 |
} |
71fe804b6 mempolicy: use st... |
2665 2666 2667 |
new = mpol_new(mode, mode_flags, &nodes); if (IS_ERR(new)) |
926f2ae04 tmpfs: cleanup mp... |
2668 |
goto out; |
f2a07f40d tmpfs mempolicy: ... |
2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 |
/* * Save nodes for mpol_to_str() to show the tmpfs mount options * for /proc/mounts, /proc/pid/mounts and /proc/pid/mountinfo. */ if (mode != MPOL_PREFERRED) new->v.nodes = nodes; else if (nodelist) new->v.preferred_node = first_node(nodes); else new->flags |= MPOL_F_LOCAL; /* * Save nodes for contextualization: this will be used to "clone" * the mempolicy in a specific context [cpuset] at a later time. */ new->w.user_nodemask = nodes; |
926f2ae04 tmpfs: cleanup mp... |
2685 |
err = 0; |
71fe804b6 mempolicy: use st... |
2686 |
|
095f1fc4e mempolicy: rework... |
2687 2688 2689 2690 2691 2692 |
out: /* Restore string for error message */ if (nodelist) *--nodelist = ':'; if (flags) *--flags = '='; |
71fe804b6 mempolicy: use st... |
2693 2694 |
if (!err) *mpol = new; |
095f1fc4e mempolicy: rework... |
2695 2696 2697 |
return err; } #endif /* CONFIG_TMPFS */ |
71fe804b6 mempolicy: use st... |
2698 2699 2700 2701 2702 |
/** * 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 |
71fe804b6 mempolicy: use st... |
2703 |
* |
948927ee9 mm, mempolicy: ma... |
2704 2705 2706 |
* Convert @pol into a string. If @buffer is too short, truncate the string. * Recommend a @maxlen of at least 32 for the longest mode, "interleave", the * longest flag, "relative", and to display at least a few node ids. |
1a75a6c82 [PATCH] Fold numa... |
2707 |
*/ |
948927ee9 mm, mempolicy: ma... |
2708 |
void mpol_to_str(char *buffer, int maxlen, struct mempolicy *pol) |
1a75a6c82 [PATCH] Fold numa... |
2709 2710 |
{ char *p = buffer; |
948927ee9 mm, mempolicy: ma... |
2711 2712 2713 |
nodemask_t nodes = NODE_MASK_NONE; unsigned short mode = MPOL_DEFAULT; unsigned short flags = 0; |
2291990ab mempolicy: clean-... |
2714 |
|
8790c71a1 mm/mempolicy.c: f... |
2715 |
if (pol && pol != &default_policy && !(pol->flags & MPOL_F_MORON)) { |
bea904d54 mempolicy: use MP... |
2716 |
mode = pol->mode; |
948927ee9 mm, mempolicy: ma... |
2717 2718 |
flags = pol->flags; } |
bea904d54 mempolicy: use MP... |
2719 |
|
1a75a6c82 [PATCH] Fold numa... |
2720 2721 |
switch (mode) { case MPOL_DEFAULT: |
1a75a6c82 [PATCH] Fold numa... |
2722 |
break; |
1a75a6c82 [PATCH] Fold numa... |
2723 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
2724 |
if (flags & MPOL_F_LOCAL) |
f2a07f40d tmpfs mempolicy: ... |
2725 |
mode = MPOL_LOCAL; |
53f2556b6 mempolicy: mPOL_P... |
2726 |
else |
fc36b8d3d mempolicy: use MP... |
2727 |
node_set(pol->v.preferred_node, nodes); |
1a75a6c82 [PATCH] Fold numa... |
2728 |
break; |
1a75a6c82 [PATCH] Fold numa... |
2729 |
case MPOL_BIND: |
1a75a6c82 [PATCH] Fold numa... |
2730 |
case MPOL_INTERLEAVE: |
f2a07f40d tmpfs mempolicy: ... |
2731 |
nodes = pol->v.nodes; |
1a75a6c82 [PATCH] Fold numa... |
2732 |
break; |
1a75a6c82 [PATCH] Fold numa... |
2733 |
default: |
948927ee9 mm, mempolicy: ma... |
2734 2735 2736 |
WARN_ON_ONCE(1); snprintf(p, maxlen, "unknown"); return; |
1a75a6c82 [PATCH] Fold numa... |
2737 |
} |
b7a9f420e mm, mempolicy: si... |
2738 |
p += snprintf(p, maxlen, "%s", policy_modes[mode]); |
1a75a6c82 [PATCH] Fold numa... |
2739 |
|
fc36b8d3d mempolicy: use MP... |
2740 |
if (flags & MPOL_MODE_FLAGS) { |
948927ee9 mm, mempolicy: ma... |
2741 |
p += snprintf(p, buffer + maxlen - p, "="); |
f5b087b52 mempolicy: add MP... |
2742 |
|
2291990ab mempolicy: clean-... |
2743 2744 2745 |
/* * Currently, the only defined flags are mutually exclusive */ |
f5b087b52 mempolicy: add MP... |
2746 |
if (flags & MPOL_F_STATIC_NODES) |
2291990ab mempolicy: clean-... |
2747 2748 2749 |
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... |
2750 |
} |
1a75a6c82 [PATCH] Fold numa... |
2751 |
if (!nodes_empty(nodes)) { |
948927ee9 mm, mempolicy: ma... |
2752 |
p += snprintf(p, buffer + maxlen - p, ":"); |
1a75a6c82 [PATCH] Fold numa... |
2753 2754 |
p += nodelist_scnprintf(p, buffer + maxlen - p, nodes); } |
1a75a6c82 [PATCH] Fold numa... |
2755 |
} |