Blame view
mm/mempolicy.c
67.2 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 |
/* * Simple NUMA memory policy for the Linux kernel. * * Copyright 2003,2004 Andi Kleen, SuSE Labs. |
8bccd85ff [PATCH] Implement... |
5 |
* (C) Copyright 2005 Christoph Lameter, Silicon Graphics, Inc. |
1da177e4c Linux-2.6.12-rc2 |
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
* Subject to the GNU Public License, version 2. * * NUMA policy allows the user to give hints in which node(s) memory should * be allocated. * * Support four policies per VMA and per process: * * The VMA policy has priority over the process policy for a page fault. * * interleave Allocate memory interleaved over a set of nodes, * with normal fallback if it fails. * For VMA based allocations this interleaves based on the * offset into the backing object or offset into the mapping * for anonymous memory. For process policy an process counter * is used. |
8bccd85ff [PATCH] Implement... |
21 |
* |
1da177e4c Linux-2.6.12-rc2 |
22 23 |
* bind Only allocate memory on a specific set of nodes, * no fallback. |
8bccd85ff [PATCH] Implement... |
24 25 26 27 |
* FIXME: memory is allocated starting with the first node * to the last. It would be better if bind would truly restrict * the allocation to memory nodes instead * |
1da177e4c Linux-2.6.12-rc2 |
28 29 30 31 32 |
* preferred Try a specific node first before normal fallback. * As a special case node -1 here means do the allocation * on the local CPU. This is normally identical to default, * but useful to set in a VMA when you have a non default * process policy. |
8bccd85ff [PATCH] Implement... |
33 |
* |
1da177e4c Linux-2.6.12-rc2 |
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
* default Allocate on the local node first, or when on a VMA * use the process policy. This is what Linux always did * in a NUMA aware kernel and still does by, ahem, default. * * The process policy is applied for most non interrupt memory allocations * in that process' context. Interrupts ignore the policies and always * try to allocate on the local CPU. The VMA policy is only applied for memory * allocations for a VMA in the VM. * * Currently there are a few corner cases in swapping where the policy * is not applied, but the majority should be handled. When process policy * is used it is not remembered over swap outs/swap ins. * * Only the highest zone in the zone hierarchy gets policied. Allocations * requesting a lower zone just use default policy. This implies that * on systems with highmem kernel lowmem allocation don't get policied. * Same with GFP_DMA allocations. * * For shmfs/tmpfs/hugetlbfs shared memory the policy is shared between * all users and remembered even when nobody has memory mapped. */ /* Notebook: fix mmap readahead to honour policy and enable policy for any page cache object statistics for bigpages global policy for page cache? currently it uses process policy. Requires first item above. handle mremap for shared memory (currently ignored for the policy) grows down? make bind policy root only? It can trigger oom much faster and the kernel is not always grateful with that. |
1da177e4c Linux-2.6.12-rc2 |
66 67 68 69 70 71 72 73 |
*/ #include <linux/mempolicy.h> #include <linux/mm.h> #include <linux/highmem.h> #include <linux/hugetlb.h> #include <linux/kernel.h> #include <linux/sched.h> |
1da177e4c Linux-2.6.12-rc2 |
74 75 |
#include <linux/nodemask.h> #include <linux/cpuset.h> |
1da177e4c Linux-2.6.12-rc2 |
76 77 78 |
#include <linux/slab.h> #include <linux/string.h> #include <linux/module.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> |
dc9aa5b9d [PATCH] Swap Migr... |
93 |
|
1da177e4c Linux-2.6.12-rc2 |
94 95 |
#include <asm/tlbflush.h> #include <asm/uaccess.h> |
62695a84e vmscan: move isol... |
96 |
#include "internal.h" |
38e35860d [PATCH] mempolici... |
97 |
/* Internal flags */ |
dc9aa5b9d [PATCH] Swap Migr... |
98 |
#define MPOL_MF_DISCONTIG_OK (MPOL_MF_INTERNAL << 0) /* Skip checks for continuous vmas */ |
38e35860d [PATCH] mempolici... |
99 |
#define MPOL_MF_INVERT (MPOL_MF_INTERNAL << 1) /* Invert check for nodemask */ |
1a75a6c82 [PATCH] Fold numa... |
100 |
#define MPOL_MF_STATS (MPOL_MF_INTERNAL << 2) /* Gather statistics */ |
dc9aa5b9d [PATCH] Swap Migr... |
101 |
|
fcc234f88 [PATCH] mm: kill ... |
102 103 |
static struct kmem_cache *policy_cache; static struct kmem_cache *sn_cache; |
1da177e4c Linux-2.6.12-rc2 |
104 |
|
1da177e4c Linux-2.6.12-rc2 |
105 106 |
/* Highest zone. An specific allocation for a zone below that is not policied. */ |
6267276f3 [PATCH] optional ... |
107 |
enum zone_type policy_zone = 0; |
1da177e4c Linux-2.6.12-rc2 |
108 |
|
bea904d54 mempolicy: use MP... |
109 110 111 |
/* * run-time system-wide default policy => local allocation */ |
d42c69972 [PATCH] PCI: Run ... |
112 |
struct mempolicy default_policy = { |
1da177e4c Linux-2.6.12-rc2 |
113 |
.refcnt = ATOMIC_INIT(1), /* never free it */ |
bea904d54 mempolicy: use MP... |
114 |
.mode = MPOL_PREFERRED, |
fc36b8d3d mempolicy: use MP... |
115 |
.flags = MPOL_F_LOCAL, |
1da177e4c Linux-2.6.12-rc2 |
116 |
}; |
37012946d mempolicy: create... |
117 118 |
static const struct mempolicy_operations { int (*create)(struct mempolicy *pol, const nodemask_t *nodes); |
708c1bbc9 mempolicy: restru... |
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 |
/* * 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... |
135 |
} mpol_ops[MPOL_MAX]; |
19770b326 mm: filter based ... |
136 |
/* Check that the nodemask contains at least one populated zone */ |
37012946d mempolicy: create... |
137 |
static int is_valid_nodemask(const nodemask_t *nodemask) |
1da177e4c Linux-2.6.12-rc2 |
138 |
{ |
19770b326 mm: filter based ... |
139 |
int nd, k; |
1da177e4c Linux-2.6.12-rc2 |
140 |
|
19770b326 mm: filter based ... |
141 142 143 144 145 146 147 |
for_each_node_mask(nd, *nodemask) { struct zone *z; for (k = 0; k <= policy_zone; k++) { z = &NODE_DATA(nd)->node_zones[k]; if (z->present_pages > 0) return 1; |
dd942ae33 [PATCH] Handle al... |
148 |
} |
8af5e2eb3 [PATCH] fix mempo... |
149 |
} |
19770b326 mm: filter based ... |
150 151 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
152 |
} |
f5b087b52 mempolicy: add MP... |
153 154 |
static inline int mpol_store_user_nodemask(const struct mempolicy *pol) { |
6d556294d mempolicy: remove... |
155 |
return pol->flags & MPOL_MODE_FLAGS; |
4c50bc011 mempolicy: add MP... |
156 157 158 159 160 161 162 163 |
} 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... |
164 |
} |
37012946d mempolicy: create... |
165 166 167 168 169 170 171 172 173 174 175 |
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... |
176 |
pol->flags |= MPOL_F_LOCAL; /* local allocation */ |
37012946d mempolicy: create... |
177 178 179 180 181 182 183 184 185 186 187 188 189 190 |
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... |
191 192 193 194 195 196 197 198 199 |
/* * 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... |
200 201 |
static int mpol_set_nodemask(struct mempolicy *pol, const nodemask_t *nodes, struct nodemask_scratch *nsc) |
58568d2a8 cpuset,mm: update... |
202 |
{ |
58568d2a8 cpuset,mm: update... |
203 204 205 206 207 |
int ret; /* if mode is MPOL_DEFAULT, pol is NULL. This is right. */ if (pol == NULL) return 0; |
4bfc44958 mm: make set_memp... |
208 209 210 |
/* Check N_HIGH_MEMORY */ nodes_and(nsc->mask1, cpuset_current_mems_allowed, node_states[N_HIGH_MEMORY]); |
58568d2a8 cpuset,mm: update... |
211 212 213 214 215 216 |
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... |
217 |
mpol_relative_nodemask(&nsc->mask2, nodes,&nsc->mask1); |
58568d2a8 cpuset,mm: update... |
218 |
else |
4bfc44958 mm: make set_memp... |
219 |
nodes_and(nsc->mask2, *nodes, nsc->mask1); |
58568d2a8 cpuset,mm: update... |
220 221 222 223 224 225 |
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... |
226 227 228 229 |
if (nodes) ret = mpol_ops[pol->mode].create(pol, &nsc->mask2); else ret = mpol_ops[pol->mode].create(pol, NULL); |
58568d2a8 cpuset,mm: update... |
230 231 232 233 234 235 236 |
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... |
237 238 |
static struct mempolicy *mpol_new(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
239 240 |
{ struct mempolicy *policy; |
028fec414 mempolicy: suppor... |
241 242 243 |
pr_debug("setting mode %d flags %d nodes[0] %lx ", mode, flags, nodes ? nodes_addr(*nodes)[0] : -1); |
140d5a490 numa: mempolicy: ... |
244 |
|
3e1f06456 mempolicy: disall... |
245 246 |
if (mode == MPOL_DEFAULT) { if (nodes && !nodes_empty(*nodes)) |
37012946d mempolicy: create... |
247 |
return ERR_PTR(-EINVAL); |
bea904d54 mempolicy: use MP... |
248 |
return NULL; /* simply delete any existing policy */ |
37012946d mempolicy: create... |
249 |
} |
3e1f06456 mempolicy: disall... |
250 251 252 253 254 255 256 257 258 259 260 261 |
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... |
262 263 264 |
} } else if (nodes_empty(*nodes)) return ERR_PTR(-EINVAL); |
1da177e4c Linux-2.6.12-rc2 |
265 266 267 268 |
policy = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!policy) return ERR_PTR(-ENOMEM); atomic_set(&policy->refcnt, 1); |
45c4745af mempolicy: rename... |
269 |
policy->mode = mode; |
3e1f06456 mempolicy: disall... |
270 |
policy->flags = flags; |
37012946d mempolicy: create... |
271 |
|
1da177e4c Linux-2.6.12-rc2 |
272 |
return policy; |
37012946d mempolicy: create... |
273 |
} |
52cd3b074 mempolicy: rework... |
274 275 276 277 278 |
/* Slow path of a mpol destructor. */ void __mpol_put(struct mempolicy *p) { if (!atomic_dec_and_test(&p->refcnt)) return; |
52cd3b074 mempolicy: rework... |
279 280 |
kmem_cache_free(policy_cache, p); } |
708c1bbc9 mempolicy: restru... |
281 282 |
static void mpol_rebind_default(struct mempolicy *pol, const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
283 284 |
{ } |
708c1bbc9 mempolicy: restru... |
285 286 287 288 289 290 291 292 |
/* * 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... |
293 294 295 296 297 298 299 300 |
{ 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... |
301 302 303 304 305 306 307 308 309 310 311 312 313 |
/* * 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... |
314 |
} |
f5b087b52 mempolicy: add MP... |
315 |
|
708c1bbc9 mempolicy: restru... |
316 317 318 319 320 321 322 323 324 |
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... |
325 326 327 328 329 330 331 332 333 334 |
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... |
335 336 |
const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
337 338 |
{ nodemask_t tmp; |
37012946d mempolicy: create... |
339 340 |
if (pol->flags & MPOL_F_STATIC_NODES) { int node = first_node(pol->w.user_nodemask); |
fc36b8d3d mempolicy: use MP... |
341 |
if (node_isset(node, *nodes)) { |
37012946d mempolicy: create... |
342 |
pol->v.preferred_node = node; |
fc36b8d3d mempolicy: use MP... |
343 344 345 |
pol->flags &= ~MPOL_F_LOCAL; } else pol->flags |= MPOL_F_LOCAL; |
37012946d mempolicy: create... |
346 347 348 |
} 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... |
349 |
} else if (!(pol->flags & MPOL_F_LOCAL)) { |
37012946d mempolicy: create... |
350 351 352 353 354 |
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 |
355 |
} |
708c1bbc9 mempolicy: restru... |
356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 |
/* * 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... |
374 |
{ |
1d0d2680a mempolicy: move r... |
375 376 |
if (!pol) return; |
708c1bbc9 mempolicy: restru... |
377 |
if (!mpol_store_user_nodemask(pol) && step == 0 && |
1d0d2680a mempolicy: move r... |
378 379 |
nodes_equal(pol->w.cpuset_mems_allowed, *newmask)) return; |
708c1bbc9 mempolicy: restru... |
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 |
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... |
395 396 397 398 399 |
} /* * Wrapper for mpol_rebind_policy() that just requires task * pointer, and updates task mempolicy. |
58568d2a8 cpuset,mm: update... |
400 401 |
* * Called with task's alloc_lock held. |
1d0d2680a mempolicy: move r... |
402 |
*/ |
708c1bbc9 mempolicy: restru... |
403 404 |
void mpol_rebind_task(struct task_struct *tsk, const nodemask_t *new, enum mpol_rebind_step step) |
1d0d2680a mempolicy: move r... |
405 |
{ |
708c1bbc9 mempolicy: restru... |
406 |
mpol_rebind_policy(tsk->mempolicy, new, step); |
1d0d2680a mempolicy: move r... |
407 408 409 410 411 412 413 414 415 416 417 418 419 420 |
} /* * 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... |
421 |
mpol_rebind_policy(vma->vm_policy, new, MPOL_REBIND_ONCE); |
1d0d2680a mempolicy: move r... |
422 423 |
up_write(&mm->mmap_sem); } |
37012946d mempolicy: create... |
424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 |
static const struct mempolicy_operations mpol_ops[MPOL_MAX] = { [MPOL_DEFAULT] = { .rebind = mpol_rebind_default, }, [MPOL_INTERLEAVE] = { .create = mpol_new_interleave, .rebind = mpol_rebind_nodemask, }, [MPOL_PREFERRED] = { .create = mpol_new_preferred, .rebind = mpol_rebind_preferred, }, [MPOL_BIND] = { .create = mpol_new_bind, .rebind = mpol_rebind_nodemask, }, }; |
397874dfe [PATCH] numa_maps... |
441 |
static void gather_stats(struct page *, void *, int pte_dirty); |
fc3012896 [PATCH] Simplify ... |
442 443 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags); |
1a75a6c82 [PATCH] Fold numa... |
444 |
|
38e35860d [PATCH] mempolici... |
445 |
/* Scan through pages checking if pages follow certain conditions. */ |
b5810039a [PATCH] core remo... |
446 |
static int check_pte_range(struct vm_area_struct *vma, pmd_t *pmd, |
dc9aa5b9d [PATCH] Swap Migr... |
447 448 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
449 |
void *private) |
1da177e4c Linux-2.6.12-rc2 |
450 |
{ |
91612e0df [PATCH] mbind: ch... |
451 452 |
pte_t *orig_pte; pte_t *pte; |
705e87c0c [PATCH] mm: pte_o... |
453 |
spinlock_t *ptl; |
941150a32 [PATCH] mbind: fi... |
454 |
|
705e87c0c [PATCH] mm: pte_o... |
455 |
orig_pte = pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); |
91612e0df [PATCH] mbind: ch... |
456 |
do { |
6aab341e0 mm: re-architect ... |
457 |
struct page *page; |
25ba77c14 [PATCH] numa node... |
458 |
int nid; |
91612e0df [PATCH] mbind: ch... |
459 460 |
if (!pte_present(*pte)) |
1da177e4c Linux-2.6.12-rc2 |
461 |
continue; |
6aab341e0 mm: re-architect ... |
462 463 |
page = vm_normal_page(vma, addr, *pte); if (!page) |
1da177e4c Linux-2.6.12-rc2 |
464 |
continue; |
053837fce [PATCH] mm: migra... |
465 |
/* |
62b61f611 ksm: memory hotre... |
466 467 468 |
* vm_normal_page() filters out zero pages, but there might * still be PageReserved pages to skip, perhaps in a VDSO. * And we cannot move PageKsm pages sensibly or safely yet. |
053837fce [PATCH] mm: migra... |
469 |
*/ |
62b61f611 ksm: memory hotre... |
470 |
if (PageReserved(page) || PageKsm(page)) |
f4598c8b3 [PATCH] migration... |
471 |
continue; |
6aab341e0 mm: re-architect ... |
472 |
nid = page_to_nid(page); |
38e35860d [PATCH] mempolici... |
473 474 |
if (node_isset(nid, *nodes) == !!(flags & MPOL_MF_INVERT)) continue; |
1a75a6c82 [PATCH] Fold numa... |
475 |
if (flags & MPOL_MF_STATS) |
397874dfe [PATCH] numa_maps... |
476 |
gather_stats(page, private, pte_dirty(*pte)); |
053837fce [PATCH] mm: migra... |
477 |
else if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) |
fc3012896 [PATCH] Simplify ... |
478 |
migrate_page_add(page, private, flags); |
38e35860d [PATCH] mempolici... |
479 480 |
else break; |
91612e0df [PATCH] mbind: ch... |
481 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
705e87c0c [PATCH] mm: pte_o... |
482 |
pte_unmap_unlock(orig_pte, ptl); |
91612e0df [PATCH] mbind: ch... |
483 484 |
return addr != end; } |
b5810039a [PATCH] core remo... |
485 |
static inline int check_pmd_range(struct vm_area_struct *vma, pud_t *pud, |
dc9aa5b9d [PATCH] Swap Migr... |
486 487 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
488 |
void *private) |
91612e0df [PATCH] mbind: ch... |
489 490 491 492 493 494 495 496 497 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; |
dc9aa5b9d [PATCH] Swap Migr... |
498 |
if (check_pte_range(vma, pmd, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
499 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
500 501 502 503 |
return -EIO; } while (pmd++, addr = next, addr != end); return 0; } |
b5810039a [PATCH] core remo... |
504 |
static inline int check_pud_range(struct vm_area_struct *vma, pgd_t *pgd, |
dc9aa5b9d [PATCH] Swap Migr... |
505 506 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
507 |
void *private) |
91612e0df [PATCH] mbind: ch... |
508 509 510 511 512 513 514 515 516 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
dc9aa5b9d [PATCH] Swap Migr... |
517 |
if (check_pmd_range(vma, pud, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
518 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
519 520 521 522 |
return -EIO; } while (pud++, addr = next, addr != end); return 0; } |
b5810039a [PATCH] core remo... |
523 |
static inline int check_pgd_range(struct vm_area_struct *vma, |
dc9aa5b9d [PATCH] Swap Migr... |
524 525 |
unsigned long addr, unsigned long end, const nodemask_t *nodes, unsigned long flags, |
38e35860d [PATCH] mempolici... |
526 |
void *private) |
91612e0df [PATCH] mbind: ch... |
527 528 529 |
{ pgd_t *pgd; unsigned long next; |
b5810039a [PATCH] core remo... |
530 |
pgd = pgd_offset(vma->vm_mm, addr); |
91612e0df [PATCH] mbind: ch... |
531 532 533 534 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
dc9aa5b9d [PATCH] Swap Migr... |
535 |
if (check_pud_range(vma, pgd, addr, next, nodes, |
38e35860d [PATCH] mempolici... |
536 |
flags, private)) |
91612e0df [PATCH] mbind: ch... |
537 538 539 |
return -EIO; } while (pgd++, addr = next, addr != end); return 0; |
1da177e4c Linux-2.6.12-rc2 |
540 |
} |
dc9aa5b9d [PATCH] Swap Migr... |
541 542 543 544 545 |
/* * Check if all pages in a range are on a set of nodes. * If pagelist != NULL then isolate pages from the LRU and * put them on the pagelist. */ |
1da177e4c Linux-2.6.12-rc2 |
546 547 |
static struct vm_area_struct * check_range(struct mm_struct *mm, unsigned long start, unsigned long end, |
38e35860d [PATCH] mempolici... |
548 |
const nodemask_t *nodes, unsigned long flags, void *private) |
1da177e4c Linux-2.6.12-rc2 |
549 550 551 |
{ int err; struct vm_area_struct *first, *vma, *prev; |
053837fce [PATCH] mm: migra... |
552 |
|
1da177e4c Linux-2.6.12-rc2 |
553 554 555 556 557 |
first = find_vma(mm, start); if (!first) return ERR_PTR(-EFAULT); prev = NULL; for (vma = first; vma && vma->vm_start < end; vma = vma->vm_next) { |
dc9aa5b9d [PATCH] Swap Migr... |
558 559 560 561 562 563 564 565 566 567 |
if (!(flags & MPOL_MF_DISCONTIG_OK)) { if (!vma->vm_next && vma->vm_end < end) return ERR_PTR(-EFAULT); if (prev && prev->vm_end < vma->vm_start) return ERR_PTR(-EFAULT); } if (!is_vm_hugetlb_page(vma) && ((flags & MPOL_MF_STRICT) || ((flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) && vma_migratable(vma)))) { |
5b952b3c1 [PATCH] Fix MPOL_... |
568 |
unsigned long endvma = vma->vm_end; |
dc9aa5b9d [PATCH] Swap Migr... |
569 |
|
5b952b3c1 [PATCH] Fix MPOL_... |
570 571 572 573 |
if (endvma > end) endvma = end; if (vma->vm_start > start) start = vma->vm_start; |
dc9aa5b9d [PATCH] Swap Migr... |
574 |
err = check_pgd_range(vma, start, endvma, nodes, |
38e35860d [PATCH] mempolici... |
575 |
flags, private); |
1da177e4c Linux-2.6.12-rc2 |
576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 |
if (err) { first = ERR_PTR(err); break; } } prev = vma; } return first; } /* Apply policy to a single VMA */ static int policy_vma(struct vm_area_struct *vma, struct mempolicy *new) { int err = 0; struct mempolicy *old = vma->vm_policy; |
140d5a490 numa: mempolicy: ... |
591 592 |
pr_debug("vma %lx-%lx/%lx vm_ops %p vm_file %p set_policy %p ", |
1da177e4c Linux-2.6.12-rc2 |
593 594 595 596 597 598 599 600 601 |
vma->vm_start, vma->vm_end, vma->vm_pgoff, vma->vm_ops, vma->vm_file, vma->vm_ops ? vma->vm_ops->set_policy : NULL); if (vma->vm_ops && vma->vm_ops->set_policy) err = vma->vm_ops->set_policy(vma, new); if (!err) { mpol_get(new); vma->vm_policy = new; |
f0be3d32b mempolicy: rename... |
602 |
mpol_put(old); |
1da177e4c Linux-2.6.12-rc2 |
603 604 605 606 607 |
} return err; } /* Step 2: apply policy to a range and do splits. */ |
9d8cebd4b mm: fix mbind vma... |
608 609 |
static int mbind_range(struct mm_struct *mm, unsigned long start, unsigned long end, struct mempolicy *new_pol) |
1da177e4c Linux-2.6.12-rc2 |
610 611 |
{ struct vm_area_struct *next; |
9d8cebd4b mm: fix mbind vma... |
612 613 614 615 616 617 |
struct vm_area_struct *prev; struct vm_area_struct *vma; int err = 0; pgoff_t pgoff; unsigned long vmstart; unsigned long vmend; |
1da177e4c Linux-2.6.12-rc2 |
618 |
|
9d8cebd4b mm: fix mbind vma... |
619 620 621 622 623 |
vma = find_vma_prev(mm, start, &prev); if (!vma || vma->vm_start > start) return -EFAULT; for (; vma && vma->vm_start < end; prev = vma, vma = next) { |
1da177e4c Linux-2.6.12-rc2 |
624 |
next = vma->vm_next; |
9d8cebd4b mm: fix mbind vma... |
625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 |
vmstart = max(start, vma->vm_start); vmend = min(end, vma->vm_end); pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT); prev = vma_merge(mm, prev, vmstart, vmend, vma->vm_flags, vma->anon_vma, vma->vm_file, pgoff, new_pol); if (prev) { vma = prev; next = vma->vm_next; continue; } 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; } err = policy_vma(vma, new_pol); |
1da177e4c Linux-2.6.12-rc2 |
647 |
if (err) |
9d8cebd4b mm: fix mbind vma... |
648 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
649 |
} |
9d8cebd4b mm: fix mbind vma... |
650 651 |
out: |
1da177e4c Linux-2.6.12-rc2 |
652 653 |
return err; } |
c61afb181 [PATCH] cpuset me... |
654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 |
/* * Update task->flags PF_MEMPOLICY bit: set iff non-default * mempolicy. Allows more rapid checking of this (combined perhaps * with other PF_* flag bits) on memory allocation hot code paths. * * If called from outside this file, the task 'p' should -only- be * a newly forked child not yet visible on the task list, because * manipulating the task flags of a visible task is not safe. * * The above limitation is why this routine has the funny name * mpol_fix_fork_child_flag(). * * It is also safe to call this with a task pointer of current, * which the static wrapper mpol_set_task_struct_flag() does, * for use within this file. */ void mpol_fix_fork_child_flag(struct task_struct *p) { if (p->mempolicy) p->flags |= PF_MEMPOLICY; else p->flags &= ~PF_MEMPOLICY; } static void mpol_set_task_struct_flag(void) { mpol_fix_fork_child_flag(current); } |
1da177e4c Linux-2.6.12-rc2 |
683 |
/* Set the process memory policy */ |
028fec414 mempolicy: suppor... |
684 685 |
static long do_set_mempolicy(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
686 |
{ |
58568d2a8 cpuset,mm: update... |
687 |
struct mempolicy *new, *old; |
f4e53d910 mempolicy: write ... |
688 |
struct mm_struct *mm = current->mm; |
4bfc44958 mm: make set_memp... |
689 |
NODEMASK_SCRATCH(scratch); |
58568d2a8 cpuset,mm: update... |
690 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
691 |
|
4bfc44958 mm: make set_memp... |
692 693 |
if (!scratch) return -ENOMEM; |
f4e53d910 mempolicy: write ... |
694 |
|
4bfc44958 mm: make set_memp... |
695 696 697 698 699 |
new = mpol_new(mode, flags, nodes); if (IS_ERR(new)) { ret = PTR_ERR(new); goto out; } |
f4e53d910 mempolicy: write ... |
700 701 702 703 704 705 706 707 |
/* * 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... |
708 |
task_lock(current); |
4bfc44958 mm: make set_memp... |
709 |
ret = mpol_set_nodemask(new, nodes, scratch); |
58568d2a8 cpuset,mm: update... |
710 711 712 713 714 |
if (ret) { task_unlock(current); if (mm) up_write(&mm->mmap_sem); mpol_put(new); |
4bfc44958 mm: make set_memp... |
715 |
goto out; |
58568d2a8 cpuset,mm: update... |
716 717 |
} old = current->mempolicy; |
1da177e4c Linux-2.6.12-rc2 |
718 |
current->mempolicy = new; |
c61afb181 [PATCH] cpuset me... |
719 |
mpol_set_task_struct_flag(); |
45c4745af mempolicy: rename... |
720 |
if (new && new->mode == MPOL_INTERLEAVE && |
f5b087b52 mempolicy: add MP... |
721 |
nodes_weight(new->v.nodes)) |
dfcd3c0dc [PATCH] Convert m... |
722 |
current->il_next = first_node(new->v.nodes); |
58568d2a8 cpuset,mm: update... |
723 |
task_unlock(current); |
f4e53d910 mempolicy: write ... |
724 725 |
if (mm) up_write(&mm->mmap_sem); |
58568d2a8 cpuset,mm: update... |
726 |
mpol_put(old); |
4bfc44958 mm: make set_memp... |
727 728 729 730 |
ret = 0; out: NODEMASK_SCRATCH_FREE(scratch); return ret; |
1da177e4c Linux-2.6.12-rc2 |
731 |
} |
bea904d54 mempolicy: use MP... |
732 733 |
/* * Return nodemask for policy for get_mempolicy() query |
58568d2a8 cpuset,mm: update... |
734 735 |
* * Called with task's alloc_lock held |
bea904d54 mempolicy: use MP... |
736 737 |
*/ static void get_policy_nodemask(struct mempolicy *p, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
738 |
{ |
dfcd3c0dc [PATCH] Convert m... |
739 |
nodes_clear(*nodes); |
bea904d54 mempolicy: use MP... |
740 741 |
if (p == &default_policy) return; |
45c4745af mempolicy: rename... |
742 |
switch (p->mode) { |
19770b326 mm: filter based ... |
743 744 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
745 |
case MPOL_INTERLEAVE: |
dfcd3c0dc [PATCH] Convert m... |
746 |
*nodes = p->v.nodes; |
1da177e4c Linux-2.6.12-rc2 |
747 748 |
break; case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
749 |
if (!(p->flags & MPOL_F_LOCAL)) |
dfcd3c0dc [PATCH] Convert m... |
750 |
node_set(p->v.preferred_node, *nodes); |
53f2556b6 mempolicy: mPOL_P... |
751 |
/* else return empty node mask for local allocation */ |
1da177e4c Linux-2.6.12-rc2 |
752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 |
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 |
770 |
/* Retrieve NUMA policy */ |
dbcb0f19c mm/mempolicy.c: c... |
771 772 |
static long do_get_mempolicy(int *policy, nodemask_t *nmask, unsigned long addr, unsigned long flags) |
1da177e4c Linux-2.6.12-rc2 |
773 |
{ |
8bccd85ff [PATCH] Implement... |
774 |
int err; |
1da177e4c Linux-2.6.12-rc2 |
775 776 777 |
struct mm_struct *mm = current->mm; struct vm_area_struct *vma = NULL; struct mempolicy *pol = current->mempolicy; |
754af6f5a Mem Policy: add M... |
778 779 |
if (flags & ~(unsigned long)(MPOL_F_NODE|MPOL_F_ADDR|MPOL_F_MEMS_ALLOWED)) |
1da177e4c Linux-2.6.12-rc2 |
780 |
return -EINVAL; |
754af6f5a Mem Policy: add M... |
781 782 783 784 785 |
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... |
786 |
task_lock(current); |
754af6f5a Mem Policy: add M... |
787 |
*nmask = cpuset_current_mems_allowed; |
58568d2a8 cpuset,mm: update... |
788 |
task_unlock(current); |
754af6f5a Mem Policy: add M... |
789 790 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
791 |
if (flags & MPOL_F_ADDR) { |
bea904d54 mempolicy: use MP... |
792 793 794 795 796 |
/* * 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 |
797 798 799 800 801 802 803 804 805 806 807 808 809 810 |
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... |
811 |
pol = &default_policy; /* indicates default behavior */ |
1da177e4c Linux-2.6.12-rc2 |
812 813 814 815 816 817 |
if (flags & MPOL_F_NODE) { if (flags & MPOL_F_ADDR) { err = lookup_node(mm, addr); if (err < 0) goto out; |
8bccd85ff [PATCH] Implement... |
818 |
*policy = err; |
1da177e4c Linux-2.6.12-rc2 |
819 |
} else if (pol == current->mempolicy && |
45c4745af mempolicy: rename... |
820 |
pol->mode == MPOL_INTERLEAVE) { |
8bccd85ff [PATCH] Implement... |
821 |
*policy = current->il_next; |
1da177e4c Linux-2.6.12-rc2 |
822 823 824 825 |
} else { err = -EINVAL; goto out; } |
bea904d54 mempolicy: use MP... |
826 827 828 |
} else { *policy = pol == &default_policy ? MPOL_DEFAULT : pol->mode; |
d79df630f mempolicy: mask o... |
829 830 831 832 833 |
/* * Internal mempolicy flags must be masked off before exposing * the policy to userspace. */ *policy |= (pol->flags & MPOL_MODE_FLAGS); |
bea904d54 mempolicy: use MP... |
834 |
} |
1da177e4c Linux-2.6.12-rc2 |
835 836 837 838 839 |
if (vma) { up_read(¤t->mm->mmap_sem); vma = NULL; } |
1da177e4c Linux-2.6.12-rc2 |
840 |
err = 0; |
58568d2a8 cpuset,mm: update... |
841 |
if (nmask) { |
c6b6ef8bb mempolicy: fix ge... |
842 843 844 845 846 847 848 |
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... |
849 |
} |
1da177e4c Linux-2.6.12-rc2 |
850 851 |
out: |
52cd3b074 mempolicy: rework... |
852 |
mpol_cond_put(pol); |
1da177e4c Linux-2.6.12-rc2 |
853 854 855 856 |
if (vma) up_read(¤t->mm->mmap_sem); return err; } |
b20a35035 [PATCH] page migr... |
857 |
#ifdef CONFIG_MIGRATION |
8bccd85ff [PATCH] Implement... |
858 |
/* |
6ce3c4c0f [PATCH] Move page... |
859 860 |
* page migration */ |
fc3012896 [PATCH] Simplify ... |
861 862 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
863 864 |
{ /* |
fc3012896 [PATCH] Simplify ... |
865 |
* Avoid migrating a page that is shared with others. |
6ce3c4c0f [PATCH] Move page... |
866 |
*/ |
62695a84e vmscan: move isol... |
867 868 869 |
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... |
870 871 |
inc_zone_page_state(page, NR_ISOLATED_ANON + page_is_file_cache(page)); |
62695a84e vmscan: move isol... |
872 873 |
} } |
7e2ab150d [PATCH] Direct Mi... |
874 |
} |
6ce3c4c0f [PATCH] Move page... |
875 |
|
742755a1d [PATCH] page migr... |
876 |
static struct page *new_node_page(struct page *page, unsigned long node, int **x) |
95a402c38 [PATCH] page migr... |
877 |
{ |
6484eb3e2 page allocator: d... |
878 |
return alloc_pages_exact_node(node, GFP_HIGHUSER_MOVABLE, 0); |
95a402c38 [PATCH] page migr... |
879 |
} |
7e2ab150d [PATCH] Direct Mi... |
880 881 882 883 |
/* * Migrate pages from one node to a target node. * Returns error or the number of pages not migrated. */ |
dbcb0f19c mm/mempolicy.c: c... |
884 885 |
static int migrate_to_node(struct mm_struct *mm, int source, int dest, int flags) |
7e2ab150d [PATCH] Direct Mi... |
886 887 888 889 890 891 892 |
{ nodemask_t nmask; LIST_HEAD(pagelist); int err = 0; nodes_clear(nmask); node_set(source, nmask); |
6ce3c4c0f [PATCH] Move page... |
893 |
|
6ec3a1271 mm: consider the ... |
894 |
check_range(mm, mm->mmap->vm_start, mm->task_size, &nmask, |
7e2ab150d [PATCH] Direct Mi... |
895 |
flags | MPOL_MF_DISCONTIG_OK, &pagelist); |
aaa994b30 [PATCH] page migr... |
896 |
if (!list_empty(&pagelist)) |
62b61f611 ksm: memory hotre... |
897 |
err = migrate_pages(&pagelist, new_node_page, dest, 0); |
95a402c38 [PATCH] page migr... |
898 |
|
7e2ab150d [PATCH] Direct Mi... |
899 |
return err; |
6ce3c4c0f [PATCH] Move page... |
900 901 902 |
} /* |
7e2ab150d [PATCH] Direct Mi... |
903 904 |
* Move pages between the two nodesets so as to preserve the physical * layout as much as possible. |
39743889a [PATCH] Swap Migr... |
905 906 907 908 909 910 |
* * Returns the number of page that could not be moved. */ int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from_nodes, const nodemask_t *to_nodes, int flags) { |
7e2ab150d [PATCH] Direct Mi... |
911 |
int busy = 0; |
0aedadf91 mm: move migrate_... |
912 |
int err; |
7e2ab150d [PATCH] Direct Mi... |
913 |
nodemask_t tmp; |
39743889a [PATCH] Swap Migr... |
914 |
|
0aedadf91 mm: move migrate_... |
915 916 917 |
err = migrate_prep(); if (err) return err; |
53f2556b6 mempolicy: mPOL_P... |
918 |
down_read(&mm->mmap_sem); |
39743889a [PATCH] Swap Migr... |
919 |
|
7b2259b3e [PATCH] page migr... |
920 921 922 |
err = migrate_vmas(mm, from_nodes, to_nodes, flags); if (err) goto out; |
da0aa1389 mm/mempolicy.c: f... |
923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 |
/* * Find a 'source' bit set in 'tmp' whose corresponding 'dest' * bit in 'to' is not also set in 'tmp'. Clear the found 'source' * bit in 'tmp', and return that <source, dest> pair for migration. * The pair of nodemasks 'to' and 'from' define the map. * * If no pair of bits is found that way, fallback to picking some * pair of 'source' and 'dest' bits that are not the same. If the * 'source' and 'dest' bits are the same, this represents a node * that will be migrating to itself, so no pages need move. * * If no bits are left in 'tmp', or if all remaining bits left * in 'tmp' correspond to the same bit in 'to', return false * (nothing left to migrate). * * This lets us pick a pair of nodes to migrate between, such that * if possible the dest node is not already occupied by some other * source node, minimizing the risk of overloading the memory on a * node that would happen if we migrated incoming memory to a node * before migrating outgoing memory source that same node. * * A single scan of tmp is sufficient. As we go, we remember the * most recent <s, d> pair that moved (s != d). If we find a pair * that not only moved, but what's better, moved to an empty slot * (d is not set in tmp), then we break out then, with that pair. * Otherwise when we finish scannng from_tmp, we at least have the * most recent <s, d> pair that moved. If we get all the way through * the scan of tmp without finding any node that moved, much less * moved to an empty node, then there is nothing left worth migrating. */ |
d49847113 [PATCH] SwapMig: ... |
953 |
|
7e2ab150d [PATCH] Direct Mi... |
954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 |
tmp = *from_nodes; while (!nodes_empty(tmp)) { int s,d; int source = -1; int dest = 0; for_each_node_mask(s, tmp) { d = node_remap(s, *from_nodes, *to_nodes); if (s == d) continue; source = s; /* Node moved. Memorize */ dest = d; /* dest not in remaining from nodes? */ if (!node_isset(dest, tmp)) break; } if (source == -1) break; node_clear(source, tmp); err = migrate_to_node(mm, source, dest, flags); if (err > 0) busy += err; if (err < 0) break; |
39743889a [PATCH] Swap Migr... |
981 |
} |
7b2259b3e [PATCH] page migr... |
982 |
out: |
39743889a [PATCH] Swap Migr... |
983 |
up_read(&mm->mmap_sem); |
7e2ab150d [PATCH] Direct Mi... |
984 985 986 |
if (err < 0) return err; return busy; |
b20a35035 [PATCH] page migr... |
987 988 |
} |
3ad33b243 Migration: find c... |
989 990 991 992 993 994 995 |
/* * 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... |
996 |
static struct page *new_vma_page(struct page *page, unsigned long private, int **x) |
95a402c38 [PATCH] page migr... |
997 998 |
{ struct vm_area_struct *vma = (struct vm_area_struct *)private; |
3ad33b243 Migration: find c... |
999 |
unsigned long uninitialized_var(address); |
95a402c38 [PATCH] page migr... |
1000 |
|
3ad33b243 Migration: find c... |
1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 |
while (vma) { address = page_address_in_vma(page, vma); if (address != -EFAULT) break; vma = vma->vm_next; } /* * if !vma, alloc_page_vma() will use task or system default policy */ return alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); |
95a402c38 [PATCH] page migr... |
1012 |
} |
b20a35035 [PATCH] page migr... |
1013 1014 1015 1016 1017 |
#else static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) { |
39743889a [PATCH] Swap Migr... |
1018 |
} |
b20a35035 [PATCH] page migr... |
1019 1020 1021 1022 1023 |
int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from_nodes, const nodemask_t *to_nodes, int flags) { return -ENOSYS; } |
95a402c38 [PATCH] page migr... |
1024 |
|
699397499 [PATCH] Fix do_mb... |
1025 |
static struct page *new_vma_page(struct page *page, unsigned long private, int **x) |
95a402c38 [PATCH] page migr... |
1026 1027 1028 |
{ return NULL; } |
b20a35035 [PATCH] page migr... |
1029 |
#endif |
dbcb0f19c mm/mempolicy.c: c... |
1030 |
static long do_mbind(unsigned long start, unsigned long len, |
028fec414 mempolicy: suppor... |
1031 1032 |
unsigned short mode, unsigned short mode_flags, nodemask_t *nmask, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
1033 1034 1035 1036 1037 1038 1039 |
{ struct vm_area_struct *vma; struct mm_struct *mm = current->mm; struct mempolicy *new; unsigned long end; int err; LIST_HEAD(pagelist); |
a3b51e014 mempolicy: conver... |
1040 1041 |
if (flags & ~(unsigned long)(MPOL_MF_STRICT | MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) |
6ce3c4c0f [PATCH] Move page... |
1042 |
return -EINVAL; |
74c002410 [PATCH] Consisten... |
1043 |
if ((flags & MPOL_MF_MOVE_ALL) && !capable(CAP_SYS_NICE)) |
6ce3c4c0f [PATCH] Move page... |
1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 |
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... |
1059 |
new = mpol_new(mode, mode_flags, nmask); |
6ce3c4c0f [PATCH] Move page... |
1060 1061 1062 1063 1064 1065 1066 1067 1068 |
if (IS_ERR(new)) return PTR_ERR(new); /* * If we are using the default policy then operation * on discontinuous address spaces is okay after all */ if (!new) flags |= MPOL_MF_DISCONTIG_OK; |
028fec414 mempolicy: suppor... |
1069 1070 1071 1072 |
pr_debug("mbind %lx-%lx mode:%d flags:%d nodes:%lx ", start, start + len, mode, mode_flags, nmask ? nodes_addr(*nmask)[0] : -1); |
6ce3c4c0f [PATCH] Move page... |
1073 |
|
0aedadf91 mm: move migrate_... |
1074 1075 1076 1077 |
if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) { err = migrate_prep(); if (err) |
b05ca7385 do_mbind(): fix m... |
1078 |
goto mpol_out; |
0aedadf91 mm: move migrate_... |
1079 |
} |
4bfc44958 mm: make set_memp... |
1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 |
{ 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... |
1093 1094 |
if (err) goto mpol_out; |
6ce3c4c0f [PATCH] Move page... |
1095 1096 1097 1098 1099 1100 |
vma = check_range(mm, start, end, nmask, flags | MPOL_MF_INVERT, &pagelist); err = PTR_ERR(vma); if (!IS_ERR(vma)) { int nr_failed = 0; |
9d8cebd4b mm: fix mbind vma... |
1101 |
err = mbind_range(mm, start, end, new); |
7e2ab150d [PATCH] Direct Mi... |
1102 |
|
6ce3c4c0f [PATCH] Move page... |
1103 |
if (!list_empty(&pagelist)) |
95a402c38 [PATCH] page migr... |
1104 |
nr_failed = migrate_pages(&pagelist, new_vma_page, |
62b61f611 ksm: memory hotre... |
1105 |
(unsigned long)vma, 0); |
6ce3c4c0f [PATCH] Move page... |
1106 1107 1108 |
if (!err && nr_failed && (flags & MPOL_MF_STRICT)) err = -EIO; |
ab8a3e14e mbind(): fix leak... |
1109 1110 |
} else putback_lru_pages(&pagelist); |
b20a35035 [PATCH] page migr... |
1111 |
|
6ce3c4c0f [PATCH] Move page... |
1112 |
up_write(&mm->mmap_sem); |
b05ca7385 do_mbind(): fix m... |
1113 |
mpol_out: |
f0be3d32b mempolicy: rename... |
1114 |
mpol_put(new); |
6ce3c4c0f [PATCH] Move page... |
1115 1116 |
return err; } |
39743889a [PATCH] Swap Migr... |
1117 |
/* |
8bccd85ff [PATCH] Implement... |
1118 1119 1120 1121 |
* User space interface with variable sized bitmaps for nodelists. */ /* Copy a node mask from user space. */ |
39743889a [PATCH] Swap Migr... |
1122 |
static int get_nodes(nodemask_t *nodes, const unsigned long __user *nmask, |
8bccd85ff [PATCH] Implement... |
1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 |
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... |
1133 |
if (maxnode > PAGE_SIZE*BITS_PER_BYTE) |
636f13c17 [PATCH] sys_mbind... |
1134 |
return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 |
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... |
1183 1184 1185 |
SYSCALL_DEFINE6(mbind, unsigned long, start, unsigned long, len, unsigned long, mode, unsigned long __user *, nmask, unsigned long, maxnode, unsigned, flags) |
8bccd85ff [PATCH] Implement... |
1186 1187 1188 |
{ nodemask_t nodes; int err; |
028fec414 mempolicy: suppor... |
1189 |
unsigned short mode_flags; |
8bccd85ff [PATCH] Implement... |
1190 |
|
028fec414 mempolicy: suppor... |
1191 1192 |
mode_flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; |
a3b51e014 mempolicy: conver... |
1193 1194 |
if (mode >= MPOL_MAX) return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1195 1196 1197 |
if ((mode_flags & MPOL_F_STATIC_NODES) && (mode_flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1198 1199 1200 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1201 |
return do_mbind(start, len, mode, mode_flags, &nodes, flags); |
8bccd85ff [PATCH] Implement... |
1202 1203 1204 |
} /* Set the process memory policy */ |
938bb9f5e [CVE-2009-0029] S... |
1205 1206 |
SYSCALL_DEFINE3(set_mempolicy, int, mode, unsigned long __user *, nmask, unsigned long, maxnode) |
8bccd85ff [PATCH] Implement... |
1207 1208 1209 |
{ int err; nodemask_t nodes; |
028fec414 mempolicy: suppor... |
1210 |
unsigned short flags; |
8bccd85ff [PATCH] Implement... |
1211 |
|
028fec414 mempolicy: suppor... |
1212 1213 1214 |
flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; if ((unsigned int)mode >= MPOL_MAX) |
8bccd85ff [PATCH] Implement... |
1215 |
return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1216 1217 |
if ((flags & MPOL_F_STATIC_NODES) && (flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1218 1219 1220 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1221 |
return do_set_mempolicy(mode, flags, &nodes); |
8bccd85ff [PATCH] Implement... |
1222 |
} |
938bb9f5e [CVE-2009-0029] S... |
1223 1224 1225 |
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... |
1226 |
{ |
c69e8d9c0 CRED: Use RCU to ... |
1227 |
const struct cred *cred = current_cred(), *tcred; |
596d7cfa2 mempolicy: reduce... |
1228 |
struct mm_struct *mm = NULL; |
39743889a [PATCH] Swap Migr... |
1229 |
struct task_struct *task; |
39743889a [PATCH] Swap Migr... |
1230 1231 |
nodemask_t task_nodes; int err; |
596d7cfa2 mempolicy: reduce... |
1232 1233 1234 1235 1236 1237 |
nodemask_t *old; nodemask_t *new; NODEMASK_SCRATCH(scratch); if (!scratch) return -ENOMEM; |
39743889a [PATCH] Swap Migr... |
1238 |
|
596d7cfa2 mempolicy: reduce... |
1239 1240 1241 1242 |
old = &scratch->mask1; new = &scratch->mask2; err = get_nodes(old, old_nodes, maxnode); |
39743889a [PATCH] Swap Migr... |
1243 |
if (err) |
596d7cfa2 mempolicy: reduce... |
1244 |
goto out; |
39743889a [PATCH] Swap Migr... |
1245 |
|
596d7cfa2 mempolicy: reduce... |
1246 |
err = get_nodes(new, new_nodes, maxnode); |
39743889a [PATCH] Swap Migr... |
1247 |
if (err) |
596d7cfa2 mempolicy: reduce... |
1248 |
goto out; |
39743889a [PATCH] Swap Migr... |
1249 1250 1251 |
/* Find the mm_struct */ read_lock(&tasklist_lock); |
228ebcbe6 Uninline find_tas... |
1252 |
task = pid ? find_task_by_vpid(pid) : current; |
39743889a [PATCH] Swap Migr... |
1253 1254 |
if (!task) { read_unlock(&tasklist_lock); |
596d7cfa2 mempolicy: reduce... |
1255 1256 |
err = -ESRCH; goto out; |
39743889a [PATCH] Swap Migr... |
1257 1258 1259 |
} mm = get_task_mm(task); read_unlock(&tasklist_lock); |
596d7cfa2 mempolicy: reduce... |
1260 |
err = -EINVAL; |
39743889a [PATCH] Swap Migr... |
1261 |
if (!mm) |
596d7cfa2 mempolicy: reduce... |
1262 |
goto out; |
39743889a [PATCH] Swap Migr... |
1263 1264 1265 1266 |
/* * Check if this process has the right to modify the specified * process. The right exists if the process has administrative |
7f927fcc2 [PATCH] Typo fixes |
1267 |
* capabilities, superuser privileges or the same |
39743889a [PATCH] Swap Migr... |
1268 1269 |
* userid as the target process. */ |
c69e8d9c0 CRED: Use RCU to ... |
1270 1271 |
rcu_read_lock(); tcred = __task_cred(task); |
b6dff3ec5 CRED: Separate ta... |
1272 1273 |
if (cred->euid != tcred->suid && cred->euid != tcred->uid && cred->uid != tcred->suid && cred->uid != tcred->uid && |
74c002410 [PATCH] Consisten... |
1274 |
!capable(CAP_SYS_NICE)) { |
c69e8d9c0 CRED: Use RCU to ... |
1275 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1276 1277 1278 |
err = -EPERM; goto out; } |
c69e8d9c0 CRED: Use RCU to ... |
1279 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1280 1281 1282 |
task_nodes = cpuset_mems_allowed(task); /* Is the user allowed to access the target nodes? */ |
596d7cfa2 mempolicy: reduce... |
1283 |
if (!nodes_subset(*new, task_nodes) && !capable(CAP_SYS_NICE)) { |
39743889a [PATCH] Swap Migr... |
1284 1285 1286 |
err = -EPERM; goto out; } |
596d7cfa2 mempolicy: reduce... |
1287 |
if (!nodes_subset(*new, node_states[N_HIGH_MEMORY])) { |
3b42d28b2 Page migration: D... |
1288 1289 1290 |
err = -EINVAL; goto out; } |
86c3a7645 [PATCH] SELinux: ... |
1291 1292 1293 |
err = security_task_movememory(task); if (err) goto out; |
596d7cfa2 mempolicy: reduce... |
1294 |
err = do_migrate_pages(mm, old, new, |
74c002410 [PATCH] Consisten... |
1295 |
capable(CAP_SYS_NICE) ? MPOL_MF_MOVE_ALL : MPOL_MF_MOVE); |
39743889a [PATCH] Swap Migr... |
1296 |
out: |
596d7cfa2 mempolicy: reduce... |
1297 1298 1299 |
if (mm) mmput(mm); NODEMASK_SCRATCH_FREE(scratch); |
39743889a [PATCH] Swap Migr... |
1300 1301 |
return err; } |
8bccd85ff [PATCH] Implement... |
1302 |
/* Retrieve NUMA policy */ |
938bb9f5e [CVE-2009-0029] S... |
1303 1304 1305 |
SYSCALL_DEFINE5(get_mempolicy, int __user *, policy, unsigned long __user *, nmask, unsigned long, maxnode, unsigned long, addr, unsigned long, flags) |
8bccd85ff [PATCH] Implement... |
1306 |
{ |
dbcb0f19c mm/mempolicy.c: c... |
1307 1308 |
int err; int uninitialized_var(pval); |
8bccd85ff [PATCH] Implement... |
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 |
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 |
1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
#ifdef CONFIG_COMPAT asmlinkage long compat_sys_get_mempolicy(int __user *policy, compat_ulong_t __user *nmask, compat_ulong_t maxnode, compat_ulong_t addr, compat_ulong_t flags) { long err; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; DECLARE_BITMAP(bm, MAX_NUMNODES); nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) nm = compat_alloc_user_space(alloc_size); err = sys_get_mempolicy(policy, nm, nr_bits+1, addr, flags); if (!err && nmask) { err = copy_from_user(bm, nm, alloc_size); /* ensure entire bitmap is zeroed */ err |= clear_user(nmask, ALIGN(maxnode-1, 8) / 8); err |= compat_put_bitmap(nmask, bm, nr_bits); } return err; } asmlinkage long compat_sys_set_mempolicy(int mode, compat_ulong_t __user *nmask, compat_ulong_t maxnode) { long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; DECLARE_BITMAP(bm, MAX_NUMNODES); nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) { err = compat_get_bitmap(bm, nmask, nr_bits); nm = compat_alloc_user_space(alloc_size); err |= copy_to_user(nm, bm, alloc_size); } if (err) return -EFAULT; return sys_set_mempolicy(mode, nm, nr_bits+1); } asmlinkage long compat_sys_mbind(compat_ulong_t start, compat_ulong_t len, compat_ulong_t mode, compat_ulong_t __user *nmask, compat_ulong_t maxnode, compat_ulong_t flags) { long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; |
dfcd3c0dc [PATCH] Convert m... |
1387 |
nodemask_t bm; |
1da177e4c Linux-2.6.12-rc2 |
1388 1389 1390 1391 1392 |
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... |
1393 |
err = compat_get_bitmap(nodes_addr(bm), nmask, nr_bits); |
1da177e4c Linux-2.6.12-rc2 |
1394 |
nm = compat_alloc_user_space(alloc_size); |
dfcd3c0dc [PATCH] Convert m... |
1395 |
err |= copy_to_user(nm, nodes_addr(bm), alloc_size); |
1da177e4c Linux-2.6.12-rc2 |
1396 1397 1398 1399 1400 1401 1402 1403 1404 |
} if (err) return -EFAULT; return sys_mbind(start, len, mode, nm, nr_bits+1, flags); } #endif |
480eccf9a Fix NUMA Memory P... |
1405 1406 1407 1408 1409 1410 1411 1412 |
/* * get_vma_policy(@task, @vma, @addr) * @task - task for fallback if vma policy == default * @vma - virtual memory area whose policy is sought * @addr - address in @vma for shared policy lookup * * Returns effective policy for a VMA at specified address. * Falls back to @task or system default policy, as necessary. |
52cd3b074 mempolicy: rework... |
1413 1414 1415 1416 1417 1418 1419 |
* Current or other task's task mempolicy and non-shared vma policies * are protected by the task's mmap_sem, which must be held for read by * the caller. * Shared policies [those marked as MPOL_F_SHARED] require an extra reference * count--added by the get_policy() vm_op, as appropriate--to protect against * freeing by another task. It is the caller's responsibility to free the * extra reference for shared policies. |
480eccf9a Fix NUMA Memory P... |
1420 |
*/ |
ae4d8c16a mempolicy: fixup ... |
1421 |
static struct mempolicy *get_vma_policy(struct task_struct *task, |
48fce3429 [PATCH] mempolici... |
1422 |
struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
1423 |
{ |
6e21c8f14 [PATCH] /proc/<pi... |
1424 |
struct mempolicy *pol = task->mempolicy; |
1da177e4c Linux-2.6.12-rc2 |
1425 1426 |
if (vma) { |
480eccf9a Fix NUMA Memory P... |
1427 |
if (vma->vm_ops && vma->vm_ops->get_policy) { |
ae4d8c16a mempolicy: fixup ... |
1428 1429 1430 1431 |
struct mempolicy *vpol = vma->vm_ops->get_policy(vma, addr); if (vpol) pol = vpol; |
bea904d54 mempolicy: use MP... |
1432 |
} else if (vma->vm_policy) |
1da177e4c Linux-2.6.12-rc2 |
1433 1434 1435 1436 1437 1438 |
pol = vma->vm_policy; } if (!pol) pol = &default_policy; return pol; } |
52cd3b074 mempolicy: rework... |
1439 1440 1441 1442 1443 |
/* * 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 ... |
1444 1445 |
{ /* Lower zones don't get a nodemask applied for MPOL_BIND */ |
45c4745af mempolicy: rename... |
1446 |
if (unlikely(policy->mode == MPOL_BIND) && |
19770b326 mm: filter based ... |
1447 1448 1449 1450 1451 1452 |
gfp_zone(gfp) >= policy_zone && cpuset_nodemask_valid_mems_allowed(&policy->v.nodes)) return &policy->v.nodes; return NULL; } |
52cd3b074 mempolicy: rework... |
1453 1454 |
/* Return a zonelist indicated by gfp for node representing a mempolicy */ static struct zonelist *policy_zonelist(gfp_t gfp, struct mempolicy *policy) |
1da177e4c Linux-2.6.12-rc2 |
1455 |
{ |
fc36b8d3d mempolicy: use MP... |
1456 |
int nd = numa_node_id(); |
1da177e4c Linux-2.6.12-rc2 |
1457 |
|
45c4745af mempolicy: rename... |
1458 |
switch (policy->mode) { |
1da177e4c Linux-2.6.12-rc2 |
1459 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1460 1461 |
if (!(policy->flags & MPOL_F_LOCAL)) nd = policy->v.preferred_node; |
1da177e4c Linux-2.6.12-rc2 |
1462 1463 |
break; case MPOL_BIND: |
19770b326 mm: filter based ... |
1464 |
/* |
52cd3b074 mempolicy: rework... |
1465 1466 |
* Normally, MPOL_BIND allocations are node-local within the * allowed nodemask. However, if __GFP_THISNODE is set and the |
6eb27e1fd mempolicy: remove... |
1467 |
* current node isn't part of the mask, we use the zonelist for |
52cd3b074 mempolicy: rework... |
1468 |
* the first node in the mask instead. |
19770b326 mm: filter based ... |
1469 |
*/ |
19770b326 mm: filter based ... |
1470 1471 1472 1473 |
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 |
1474 |
default: |
1da177e4c Linux-2.6.12-rc2 |
1475 1476 |
BUG(); } |
0e88460da mm: introduce nod... |
1477 |
return node_zonelist(nd, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1478 1479 1480 1481 1482 1483 1484 1485 1486 |
} /* 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... |
1487 |
next = next_node(nid, policy->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1488 |
if (next >= MAX_NUMNODES) |
dfcd3c0dc [PATCH] Convert m... |
1489 |
next = first_node(policy->v.nodes); |
f5b087b52 mempolicy: add MP... |
1490 1491 |
if (next < MAX_NUMNODES) me->il_next = next; |
1da177e4c Linux-2.6.12-rc2 |
1492 1493 |
return nid; } |
dc85da15d [PATCH] NUMA poli... |
1494 1495 1496 |
/* * Depending on the memory policy provide a node from which to allocate the * next slab entry. |
52cd3b074 mempolicy: rework... |
1497 1498 1499 1500 |
* @policy must be protected by freeing by the caller. If @policy is * the current task's mempolicy, this protection is implicit, as only the * task can change it's policy. The system default policy requires no * such protection. |
dc85da15d [PATCH] NUMA poli... |
1501 1502 1503 |
*/ unsigned slab_node(struct mempolicy *policy) { |
fc36b8d3d mempolicy: use MP... |
1504 |
if (!policy || policy->flags & MPOL_F_LOCAL) |
bea904d54 mempolicy: use MP... |
1505 1506 1507 1508 |
return numa_node_id(); switch (policy->mode) { case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1509 1510 1511 1512 |
/* * handled MPOL_F_LOCAL above */ return policy->v.preferred_node; |
765c4507a [PATCH] GFP_THISN... |
1513 |
|
dc85da15d [PATCH] NUMA poli... |
1514 1515 |
case MPOL_INTERLEAVE: return interleave_nodes(policy); |
dd1a239f6 mm: have zonelist... |
1516 |
case MPOL_BIND: { |
dc85da15d [PATCH] NUMA poli... |
1517 1518 1519 1520 |
/* * Follow bind policy behavior and start allocation at the * first node. */ |
19770b326 mm: filter based ... |
1521 1522 1523 1524 1525 1526 1527 1528 |
struct zonelist *zonelist; struct zone *zone; enum zone_type highest_zoneidx = gfp_zone(GFP_KERNEL); zonelist = &NODE_DATA(numa_node_id())->node_zonelists[0]; (void)first_zones_zonelist(zonelist, highest_zoneidx, &policy->v.nodes, &zone); return zone->node; |
dd1a239f6 mm: have zonelist... |
1529 |
} |
dc85da15d [PATCH] NUMA poli... |
1530 |
|
dc85da15d [PATCH] NUMA poli... |
1531 |
default: |
bea904d54 mempolicy: use MP... |
1532 |
BUG(); |
dc85da15d [PATCH] NUMA poli... |
1533 1534 |
} } |
1da177e4c Linux-2.6.12-rc2 |
1535 1536 1537 1538 |
/* 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... |
1539 |
unsigned nnodes = nodes_weight(pol->v.nodes); |
f5b087b52 mempolicy: add MP... |
1540 |
unsigned target; |
1da177e4c Linux-2.6.12-rc2 |
1541 1542 |
int c; int nid = -1; |
f5b087b52 mempolicy: add MP... |
1543 1544 1545 |
if (!nnodes) return numa_node_id(); target = (unsigned int)off % nnodes; |
1da177e4c Linux-2.6.12-rc2 |
1546 1547 |
c = 0; do { |
dfcd3c0dc [PATCH] Convert m... |
1548 |
nid = next_node(nid, pol->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1549 1550 |
c++; } while (c <= target); |
1da177e4c Linux-2.6.12-rc2 |
1551 1552 |
return nid; } |
5da7ca860 [PATCH] Add NUMA ... |
1553 1554 1555 1556 1557 1558 |
/* 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 ... |
1559 1560 1561 1562 1563 1564 1565 1566 1567 |
/* * 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 ... |
1568 1569 1570 1571 1572 |
off += (addr - vma->vm_start) >> shift; return offset_il_node(pol, vma, off); } else return interleave_nodes(pol); } |
00ac59adf [PATCH] x86_64: F... |
1573 |
#ifdef CONFIG_HUGETLBFS |
480eccf9a Fix NUMA Memory P... |
1574 1575 1576 1577 1578 |
/* * 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 ... |
1579 1580 |
* @mpol = pointer to mempolicy pointer for reference counted mempolicy * @nodemask = pointer to nodemask pointer for MPOL_BIND nodemask |
480eccf9a Fix NUMA Memory P... |
1581 |
* |
52cd3b074 mempolicy: rework... |
1582 1583 1584 1585 |
* 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... |
1586 1587 |
* * Must be protected by get_mems_allowed() |
480eccf9a Fix NUMA Memory P... |
1588 |
*/ |
396faf030 Allow huge page a... |
1589 |
struct zonelist *huge_zonelist(struct vm_area_struct *vma, unsigned long addr, |
19770b326 mm: filter based ... |
1590 1591 |
gfp_t gfp_flags, struct mempolicy **mpol, nodemask_t **nodemask) |
5da7ca860 [PATCH] Add NUMA ... |
1592 |
{ |
480eccf9a Fix NUMA Memory P... |
1593 |
struct zonelist *zl; |
5da7ca860 [PATCH] Add NUMA ... |
1594 |
|
52cd3b074 mempolicy: rework... |
1595 |
*mpol = get_vma_policy(current, vma, addr); |
19770b326 mm: filter based ... |
1596 |
*nodemask = NULL; /* assume !MPOL_BIND */ |
5da7ca860 [PATCH] Add NUMA ... |
1597 |
|
52cd3b074 mempolicy: rework... |
1598 1599 |
if (unlikely((*mpol)->mode == MPOL_INTERLEAVE)) { zl = node_zonelist(interleave_nid(*mpol, vma, addr, |
a55164389 hugetlb: modular ... |
1600 |
huge_page_shift(hstate_vma(vma))), gfp_flags); |
52cd3b074 mempolicy: rework... |
1601 1602 1603 1604 |
} else { zl = policy_zonelist(gfp_flags, *mpol); if ((*mpol)->mode == MPOL_BIND) *nodemask = &(*mpol)->v.nodes; |
480eccf9a Fix NUMA Memory P... |
1605 1606 |
} return zl; |
5da7ca860 [PATCH] Add NUMA ... |
1607 |
} |
06808b082 hugetlb: derive h... |
1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 |
/* * 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... |
1632 |
task_lock(current); |
06808b082 hugetlb: derive h... |
1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 |
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... |
1652 |
task_unlock(current); |
06808b082 hugetlb: derive h... |
1653 1654 1655 |
return true; } |
00ac59adf [PATCH] x86_64: F... |
1656 |
#endif |
5da7ca860 [PATCH] Add NUMA ... |
1657 |
|
6f48d0ebd oom: select task ... |
1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 |
/* * 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 |
1701 1702 |
/* Allocate a page in interleaved policy. Own path because it needs to do special accounting. */ |
662f3a0b9 [PATCH] Remove ne... |
1703 1704 |
static struct page *alloc_page_interleave(gfp_t gfp, unsigned order, unsigned nid) |
1da177e4c Linux-2.6.12-rc2 |
1705 1706 1707 |
{ struct zonelist *zl; struct page *page; |
0e88460da mm: introduce nod... |
1708 |
zl = node_zonelist(nid, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1709 |
page = __alloc_pages(gfp, order, zl); |
dd1a239f6 mm: have zonelist... |
1710 |
if (page && page_zone(page) == zonelist_zone(&zl->_zonerefs[0])) |
ca889e6c4 [PATCH] Use Zoned... |
1711 |
inc_zone_page_state(page, NUMA_INTERLEAVE_HIT); |
1da177e4c Linux-2.6.12-rc2 |
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 |
return page; } /** * alloc_page_vma - Allocate a page for a VMA. * * @gfp: * %GFP_USER user allocation. * %GFP_KERNEL kernel allocations, * %GFP_HIGHMEM highmem/user allocations, * %GFP_FS allocation should not call back into a file system. * %GFP_ATOMIC don't sleep. * * @vma: Pointer to VMA or NULL if not available. * @addr: Virtual Address of the allocation. Must be inside the VMA. * * This function allocates a page from the kernel page pool and applies * a NUMA policy associated with the VMA or the current process. * When VMA is not NULL caller must hold down_read on the mmap_sem of the * mm_struct of the VMA to prevent it from going away. Should be used for * all allocations for pages that will be mapped into * user space. Returns NULL when no page can be allocated. * * Should be called with the mm_sem of the vma hold. */ struct page * |
dd0fc66fb [PATCH] gfp flags... |
1738 |
alloc_page_vma(gfp_t gfp, struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
1739 |
{ |
6e21c8f14 [PATCH] /proc/<pi... |
1740 |
struct mempolicy *pol = get_vma_policy(current, vma, addr); |
480eccf9a Fix NUMA Memory P... |
1741 |
struct zonelist *zl; |
c0ff7453b cpuset,mm: fix no... |
1742 |
struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
1743 |
|
c0ff7453b cpuset,mm: fix no... |
1744 |
get_mems_allowed(); |
45c4745af mempolicy: rename... |
1745 |
if (unlikely(pol->mode == MPOL_INTERLEAVE)) { |
1da177e4c Linux-2.6.12-rc2 |
1746 |
unsigned nid; |
5da7ca860 [PATCH] Add NUMA ... |
1747 1748 |
nid = interleave_nid(pol, vma, addr, PAGE_SHIFT); |
52cd3b074 mempolicy: rework... |
1749 |
mpol_cond_put(pol); |
c0ff7453b cpuset,mm: fix no... |
1750 1751 1752 |
page = alloc_page_interleave(gfp, 0, nid); put_mems_allowed(); return page; |
1da177e4c Linux-2.6.12-rc2 |
1753 |
} |
52cd3b074 mempolicy: rework... |
1754 1755 |
zl = policy_zonelist(gfp, pol); if (unlikely(mpol_needs_cond_ref(pol))) { |
480eccf9a Fix NUMA Memory P... |
1756 |
/* |
52cd3b074 mempolicy: rework... |
1757 |
* slow path: ref counted shared policy |
480eccf9a Fix NUMA Memory P... |
1758 |
*/ |
19770b326 mm: filter based ... |
1759 |
struct page *page = __alloc_pages_nodemask(gfp, 0, |
52cd3b074 mempolicy: rework... |
1760 |
zl, policy_nodemask(gfp, pol)); |
f0be3d32b mempolicy: rename... |
1761 |
__mpol_put(pol); |
c0ff7453b cpuset,mm: fix no... |
1762 |
put_mems_allowed(); |
480eccf9a Fix NUMA Memory P... |
1763 1764 1765 1766 1767 |
return page; } /* * fast path: default or task policy */ |
c0ff7453b cpuset,mm: fix no... |
1768 1769 1770 |
page = __alloc_pages_nodemask(gfp, 0, zl, policy_nodemask(gfp, pol)); put_mems_allowed(); return page; |
1da177e4c Linux-2.6.12-rc2 |
1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 |
} /** * 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... |
1788 |
* Don't call cpuset_update_task_memory_state() unless |
1da177e4c Linux-2.6.12-rc2 |
1789 1790 1791 |
* 1) it's ok to take cpuset_sem (can WAIT), and * 2) allocating for current task (not interrupt). */ |
dd0fc66fb [PATCH] gfp flags... |
1792 |
struct page *alloc_pages_current(gfp_t gfp, unsigned order) |
1da177e4c Linux-2.6.12-rc2 |
1793 1794 |
{ struct mempolicy *pol = current->mempolicy; |
c0ff7453b cpuset,mm: fix no... |
1795 |
struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
1796 |
|
9b819d204 [PATCH] Add __GFP... |
1797 |
if (!pol || in_interrupt() || (gfp & __GFP_THISNODE)) |
1da177e4c Linux-2.6.12-rc2 |
1798 |
pol = &default_policy; |
52cd3b074 mempolicy: rework... |
1799 |
|
c0ff7453b cpuset,mm: fix no... |
1800 |
get_mems_allowed(); |
52cd3b074 mempolicy: rework... |
1801 1802 1803 1804 |
/* * No reference counting needed for current->mempolicy * nor system default_policy */ |
45c4745af mempolicy: rename... |
1805 |
if (pol->mode == MPOL_INTERLEAVE) |
c0ff7453b cpuset,mm: fix no... |
1806 1807 1808 |
page = alloc_page_interleave(gfp, order, interleave_nodes(pol)); else page = __alloc_pages_nodemask(gfp, order, |
52cd3b074 mempolicy: rework... |
1809 |
policy_zonelist(gfp, pol), policy_nodemask(gfp, pol)); |
c0ff7453b cpuset,mm: fix no... |
1810 1811 |
put_mems_allowed(); return page; |
1da177e4c Linux-2.6.12-rc2 |
1812 1813 |
} EXPORT_SYMBOL(alloc_pages_current); |
4225399a6 [PATCH] cpuset: r... |
1814 |
/* |
846a16bf0 mempolicy: rename... |
1815 |
* If mpol_dup() sees current->cpuset == cpuset_being_rebound, then it |
4225399a6 [PATCH] cpuset: r... |
1816 1817 1818 1819 |
* 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... |
1820 1821 1822 |
* * 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... |
1823 |
*/ |
4225399a6 [PATCH] cpuset: r... |
1824 |
|
846a16bf0 mempolicy: rename... |
1825 1826 |
/* Slow path of a mempolicy duplicate */ struct mempolicy *__mpol_dup(struct mempolicy *old) |
1da177e4c Linux-2.6.12-rc2 |
1827 1828 1829 1830 1831 |
{ struct mempolicy *new = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!new) return ERR_PTR(-ENOMEM); |
708c1bbc9 mempolicy: restru... |
1832 1833 1834 1835 1836 1837 1838 1839 |
/* 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... |
1840 |
rcu_read_lock(); |
4225399a6 [PATCH] cpuset: r... |
1841 1842 |
if (current_cpuset_is_being_rebound()) { nodemask_t mems = cpuset_mems_allowed(current); |
708c1bbc9 mempolicy: restru... |
1843 1844 1845 1846 |
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... |
1847 |
} |
99ee4ca74 rcu: Suppress __m... |
1848 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1849 |
atomic_set(&new->refcnt, 1); |
1da177e4c Linux-2.6.12-rc2 |
1850 1851 |
return new; } |
52cd3b074 mempolicy: rework... |
1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 |
/* * If *frompol needs [has] an extra ref, copy *frompol to *tompol , * eliminate the * MPOL_F_* flags that require conditional ref and * [NOTE!!!] drop the extra ref. Not safe to reference *frompol directly * after return. Use the returned value. * * Allows use of a mempolicy for, e.g., multiple allocations with a single * policy lookup, even if the policy needs/has extra ref on lookup. * shmem_readahead needs this. */ struct mempolicy *__mpol_cond_copy(struct mempolicy *tompol, struct mempolicy *frompol) { if (!mpol_needs_cond_ref(frompol)) return frompol; *tompol = *frompol; tompol->flags &= ~MPOL_F_SHARED; /* copy doesn't need unref */ __mpol_put(frompol); return tompol; } |
1da177e4c Linux-2.6.12-rc2 |
1873 1874 1875 1876 1877 |
/* Slow path of a mempolicy comparison */ int __mpol_equal(struct mempolicy *a, struct mempolicy *b) { if (!a || !b) return 0; |
45c4745af mempolicy: rename... |
1878 |
if (a->mode != b->mode) |
1da177e4c Linux-2.6.12-rc2 |
1879 |
return 0; |
198005025 mempolicy: remove... |
1880 |
if (a->flags != b->flags) |
f5b087b52 mempolicy: add MP... |
1881 |
return 0; |
198005025 mempolicy: remove... |
1882 1883 1884 |
if (mpol_store_user_nodemask(a)) if (!nodes_equal(a->w.user_nodemask, b->w.user_nodemask)) return 0; |
45c4745af mempolicy: rename... |
1885 |
switch (a->mode) { |
19770b326 mm: filter based ... |
1886 1887 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
1888 |
case MPOL_INTERLEAVE: |
dfcd3c0dc [PATCH] Convert m... |
1889 |
return nodes_equal(a->v.nodes, b->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1890 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1891 1892 |
return a->v.preferred_node == b->v.preferred_node && a->flags == b->flags; |
1da177e4c Linux-2.6.12-rc2 |
1893 1894 1895 1896 1897 |
default: BUG(); return 0; } } |
1da177e4c Linux-2.6.12-rc2 |
1898 |
/* |
1da177e4c Linux-2.6.12-rc2 |
1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 |
* Shared memory backing store policy support. * * Remember policies even when nobody has shared memory mapped. * The policies are kept in Red-Black tree linked from the inode. * They are protected by the sp->lock spinlock, which should be held * for any accesses to the tree. */ /* lookup first element intersecting start-end */ /* Caller holds sp->lock */ static struct sp_node * sp_lookup(struct shared_policy *sp, unsigned long start, unsigned long end) { struct rb_node *n = sp->root.rb_node; while (n) { struct sp_node *p = rb_entry(n, struct sp_node, nd); if (start >= p->end) n = n->rb_right; else if (end <= p->start) n = n->rb_left; else break; } if (!n) return NULL; for (;;) { struct sp_node *w = NULL; struct rb_node *prev = rb_prev(n); if (!prev) break; w = rb_entry(prev, struct sp_node, nd); if (w->end <= start) break; n = prev; } return rb_entry(n, struct sp_node, nd); } /* Insert a new shared policy into the list. */ /* Caller holds sp->lock */ static void sp_insert(struct shared_policy *sp, struct sp_node *new) { struct rb_node **p = &sp->root.rb_node; struct rb_node *parent = NULL; struct sp_node *nd; while (*p) { parent = *p; nd = rb_entry(parent, struct sp_node, nd); if (new->start < nd->start) p = &(*p)->rb_left; else if (new->end > nd->end) p = &(*p)->rb_right; else BUG(); } rb_link_node(&new->nd, parent, p); rb_insert_color(&new->nd, &sp->root); |
140d5a490 numa: mempolicy: ... |
1959 1960 |
pr_debug("inserting %lx-%lx: %d ", new->start, new->end, |
45c4745af mempolicy: rename... |
1961 |
new->policy ? new->policy->mode : 0); |
1da177e4c Linux-2.6.12-rc2 |
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 |
} /* Find shared policy intersecting idx */ struct mempolicy * mpol_shared_policy_lookup(struct shared_policy *sp, unsigned long idx) { struct mempolicy *pol = NULL; struct sp_node *sn; if (!sp->root.rb_node) return NULL; spin_lock(&sp->lock); sn = sp_lookup(sp, idx, idx+1); if (sn) { mpol_get(sn->policy); pol = sn->policy; } spin_unlock(&sp->lock); return pol; } static void sp_delete(struct shared_policy *sp, struct sp_node *n) { |
140d5a490 numa: mempolicy: ... |
1985 1986 |
pr_debug("deleting %lx-l%lx ", n->start, n->end); |
1da177e4c Linux-2.6.12-rc2 |
1987 |
rb_erase(&n->nd, &sp->root); |
f0be3d32b mempolicy: rename... |
1988 |
mpol_put(n->policy); |
1da177e4c Linux-2.6.12-rc2 |
1989 1990 |
kmem_cache_free(sn_cache, n); } |
dbcb0f19c mm/mempolicy.c: c... |
1991 1992 |
static struct sp_node *sp_alloc(unsigned long start, unsigned long end, struct mempolicy *pol) |
1da177e4c Linux-2.6.12-rc2 |
1993 1994 1995 1996 1997 1998 1999 2000 |
{ struct sp_node *n = kmem_cache_alloc(sn_cache, GFP_KERNEL); if (!n) return NULL; n->start = start; n->end = end; mpol_get(pol); |
aab0b1029 mempolicy: mark s... |
2001 |
pol->flags |= MPOL_F_SHARED; /* for unref */ |
1da177e4c Linux-2.6.12-rc2 |
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 |
n->policy = pol; return n; } /* Replace a policy range. */ static int shared_policy_replace(struct shared_policy *sp, unsigned long start, unsigned long end, struct sp_node *new) { struct sp_node *n, *new2 = NULL; restart: spin_lock(&sp->lock); n = sp_lookup(sp, start, end); /* Take care of old policies in the same range. */ while (n && n->start < end) { struct rb_node *next = rb_next(&n->nd); if (n->start >= start) { if (n->end <= end) sp_delete(sp, n); else n->start = end; } else { /* Old policy spanning whole new range. */ if (n->end > end) { if (!new2) { spin_unlock(&sp->lock); new2 = sp_alloc(end, n->end, n->policy); if (!new2) return -ENOMEM; goto restart; } n->end = start; sp_insert(sp, new2); new2 = NULL; break; } else n->end = start; } if (!next) break; n = rb_entry(next, struct sp_node, nd); } if (new) sp_insert(sp, new); spin_unlock(&sp->lock); if (new2) { |
f0be3d32b mempolicy: rename... |
2048 |
mpol_put(new2->policy); |
1da177e4c Linux-2.6.12-rc2 |
2049 2050 2051 2052 |
kmem_cache_free(sn_cache, new2); } return 0; } |
71fe804b6 mempolicy: use st... |
2053 2054 2055 2056 2057 2058 2059 2060 |
/** * 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... |
2061 |
* This is called at get_inode() calls and we can use GFP_KERNEL. |
71fe804b6 mempolicy: use st... |
2062 2063 2064 |
*/ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) { |
58568d2a8 cpuset,mm: update... |
2065 |
int ret; |
71fe804b6 mempolicy: use st... |
2066 2067 2068 2069 2070 2071 |
sp->root = RB_ROOT; /* empty tree == default mempolicy */ spin_lock_init(&sp->lock); if (mpol) { struct vm_area_struct pvma; struct mempolicy *new; |
4bfc44958 mm: make set_memp... |
2072 |
NODEMASK_SCRATCH(scratch); |
71fe804b6 mempolicy: use st... |
2073 |
|
4bfc44958 mm: make set_memp... |
2074 |
if (!scratch) |
5c0c16549 mempolicy: fix da... |
2075 |
goto put_mpol; |
71fe804b6 mempolicy: use st... |
2076 2077 |
/* contextualize the tmpfs mount point mempolicy */ new = mpol_new(mpol->mode, mpol->flags, &mpol->w.user_nodemask); |
15d77835a mempolicy: factor... |
2078 |
if (IS_ERR(new)) |
0cae3457b mempolicy: ERR_PT... |
2079 |
goto free_scratch; /* no valid nodemask intersection */ |
58568d2a8 cpuset,mm: update... |
2080 2081 |
task_lock(current); |
4bfc44958 mm: make set_memp... |
2082 |
ret = mpol_set_nodemask(new, &mpol->w.user_nodemask, scratch); |
58568d2a8 cpuset,mm: update... |
2083 |
task_unlock(current); |
15d77835a mempolicy: factor... |
2084 |
if (ret) |
5c0c16549 mempolicy: fix da... |
2085 |
goto put_new; |
71fe804b6 mempolicy: use st... |
2086 2087 2088 2089 2090 |
/* 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... |
2091 |
|
5c0c16549 mempolicy: fix da... |
2092 |
put_new: |
71fe804b6 mempolicy: use st... |
2093 |
mpol_put(new); /* drop initial ref */ |
0cae3457b mempolicy: ERR_PT... |
2094 |
free_scratch: |
4bfc44958 mm: make set_memp... |
2095 |
NODEMASK_SCRATCH_FREE(scratch); |
5c0c16549 mempolicy: fix da... |
2096 2097 |
put_mpol: mpol_put(mpol); /* drop our incoming ref on sb mpol */ |
7339ff830 [PATCH] Add tmpfs... |
2098 2099 |
} } |
1da177e4c Linux-2.6.12-rc2 |
2100 2101 2102 2103 2104 2105 |
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... |
2106 2107 |
pr_debug("set_shared_policy %lx sz %lu %d %d %lx ", |
1da177e4c Linux-2.6.12-rc2 |
2108 |
vma->vm_pgoff, |
45c4745af mempolicy: rename... |
2109 |
sz, npol ? npol->mode : -1, |
028fec414 mempolicy: suppor... |
2110 |
npol ? npol->flags : -1, |
140d5a490 numa: mempolicy: ... |
2111 |
npol ? nodes_addr(npol->v.nodes)[0] : -1); |
1da177e4c Linux-2.6.12-rc2 |
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 |
if (npol) { new = sp_alloc(vma->vm_pgoff, vma->vm_pgoff + sz, npol); if (!new) return -ENOMEM; } err = shared_policy_replace(info, vma->vm_pgoff, vma->vm_pgoff+sz, new); if (err && new) kmem_cache_free(sn_cache, new); return err; } /* Free a backing policy store on inode delete. */ void mpol_free_shared_policy(struct shared_policy *p) { struct sp_node *n; struct rb_node *next; if (!p->root.rb_node) return; spin_lock(&p->lock); next = rb_first(&p->root); while (next) { n = rb_entry(next, struct sp_node, nd); next = rb_next(&n->nd); |
90c5029e4 [PATCH] Undo memp... |
2137 |
rb_erase(&n->nd, &p->root); |
f0be3d32b mempolicy: rename... |
2138 |
mpol_put(n->policy); |
1da177e4c Linux-2.6.12-rc2 |
2139 2140 2141 |
kmem_cache_free(sn_cache, n); } spin_unlock(&p->lock); |
1da177e4c Linux-2.6.12-rc2 |
2142 2143 2144 2145 2146 |
} /* assumes fs == KERNEL_DS */ void __init numa_policy_init(void) { |
b71636e29 numa: mempolicy: ... |
2147 2148 2149 |
nodemask_t interleave_nodes; unsigned long largest = 0; int nid, prefer = 0; |
1da177e4c Linux-2.6.12-rc2 |
2150 2151 |
policy_cache = kmem_cache_create("numa_policy", sizeof(struct mempolicy), |
20c2df83d mm: Remove slab d... |
2152 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2153 2154 2155 |
sn_cache = kmem_cache_create("shared_policy_node", sizeof(struct sp_node), |
20c2df83d mm: Remove slab d... |
2156 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2157 |
|
b71636e29 numa: mempolicy: ... |
2158 2159 2160 2161 2162 2163 |
/* * Set interleaving policy for system init. Interleaving is only * enabled across suitably sized nodes (default is >= 16MB), or * fall back to the largest node if they're all smaller. */ nodes_clear(interleave_nodes); |
56bbd65df Memoryless nodes:... |
2164 |
for_each_node_state(nid, N_HIGH_MEMORY) { |
b71636e29 numa: mempolicy: ... |
2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 |
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 |
2181 |
|
028fec414 mempolicy: suppor... |
2182 |
if (do_set_mempolicy(MPOL_INTERLEAVE, 0, &interleave_nodes)) |
1da177e4c Linux-2.6.12-rc2 |
2183 2184 2185 |
printk("numa_policy_init: interleaving failed "); } |
8bccd85ff [PATCH] Implement... |
2186 |
/* Reset policy of current process to default */ |
1da177e4c Linux-2.6.12-rc2 |
2187 2188 |
void numa_default_policy(void) { |
028fec414 mempolicy: suppor... |
2189 |
do_set_mempolicy(MPOL_DEFAULT, 0, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2190 |
} |
68860ec10 [PATCH] cpusets: ... |
2191 |
|
4225399a6 [PATCH] cpuset: r... |
2192 |
/* |
095f1fc4e mempolicy: rework... |
2193 2194 2195 2196 |
* Parse and format mempolicy from/to strings */ /* |
fc36b8d3d mempolicy: use MP... |
2197 |
* "local" is pseudo-policy: MPOL_PREFERRED with MPOL_F_LOCAL flag |
3f226aa1c mempolicy: suppor... |
2198 |
* Used only for mpol_parse_str() and mpol_to_str() |
1a75a6c82 [PATCH] Fold numa... |
2199 |
*/ |
345ace9c7 mempolicy: rename... |
2200 2201 2202 2203 2204 2205 2206 2207 2208 |
#define MPOL_LOCAL MPOL_MAX static const char * const policy_modes[] = { [MPOL_DEFAULT] = "default", [MPOL_PREFERRED] = "prefer", [MPOL_BIND] = "bind", [MPOL_INTERLEAVE] = "interleave", [MPOL_LOCAL] = "local" }; |
1a75a6c82 [PATCH] Fold numa... |
2209 |
|
095f1fc4e mempolicy: rework... |
2210 2211 2212 2213 2214 |
#ifdef CONFIG_TMPFS /** * mpol_parse_str - parse string to mempolicy * @str: string containing mempolicy to parse |
71fe804b6 mempolicy: use st... |
2215 2216 |
* @mpol: pointer to struct mempolicy pointer, returned on success. * @no_context: flag whether to "contextualize" the mempolicy |
095f1fc4e mempolicy: rework... |
2217 2218 2219 2220 |
* * Format of input: * <mode>[=<flags>][:<nodelist>] * |
71fe804b6 mempolicy: use st... |
2221 2222 2223 2224 2225 2226 2227 2228 |
* if @no_context is true, save the input nodemask in w.user_nodemask in * the returned mempolicy. This will be used to "clone" the mempolicy in * a specific context [cpuset] at a later time. Used to parse tmpfs mpol * mount option. Note that if 'static' or 'relative' mode flags were * specified, the input nodemask will already have been saved. Saving * it again is redundant, but safe. * * On success, returns 0, else 1 |
095f1fc4e mempolicy: rework... |
2229 |
*/ |
71fe804b6 mempolicy: use st... |
2230 |
int mpol_parse_str(char *str, struct mempolicy **mpol, int no_context) |
095f1fc4e mempolicy: rework... |
2231 |
{ |
71fe804b6 mempolicy: use st... |
2232 |
struct mempolicy *new = NULL; |
b4652e842 mempolicy: lose u... |
2233 |
unsigned short mode; |
71fe804b6 mempolicy: use st... |
2234 2235 |
unsigned short uninitialized_var(mode_flags); nodemask_t nodes; |
095f1fc4e mempolicy: rework... |
2236 2237 |
char *nodelist = strchr(str, ':'); char *flags = strchr(str, '='); |
095f1fc4e mempolicy: rework... |
2238 2239 2240 2241 2242 |
int err = 1; if (nodelist) { /* NUL-terminate mode or flags string */ *nodelist++ = '\0'; |
71fe804b6 mempolicy: use st... |
2243 |
if (nodelist_parse(nodelist, nodes)) |
095f1fc4e mempolicy: rework... |
2244 |
goto out; |
71fe804b6 mempolicy: use st... |
2245 |
if (!nodes_subset(nodes, node_states[N_HIGH_MEMORY])) |
095f1fc4e mempolicy: rework... |
2246 |
goto out; |
71fe804b6 mempolicy: use st... |
2247 2248 |
} else nodes_clear(nodes); |
095f1fc4e mempolicy: rework... |
2249 2250 |
if (flags) *flags++ = '\0'; /* terminate mode string */ |
b4652e842 mempolicy: lose u... |
2251 |
for (mode = 0; mode <= MPOL_LOCAL; mode++) { |
345ace9c7 mempolicy: rename... |
2252 |
if (!strcmp(str, policy_modes[mode])) { |
095f1fc4e mempolicy: rework... |
2253 2254 2255 |
break; } } |
b4652e842 mempolicy: lose u... |
2256 |
if (mode > MPOL_LOCAL) |
095f1fc4e mempolicy: rework... |
2257 |
goto out; |
71fe804b6 mempolicy: use st... |
2258 |
switch (mode) { |
095f1fc4e mempolicy: rework... |
2259 |
case MPOL_PREFERRED: |
71fe804b6 mempolicy: use st... |
2260 2261 2262 |
/* * Insist on a nodelist of one node only */ |
095f1fc4e mempolicy: rework... |
2263 2264 2265 2266 |
if (nodelist) { char *rest = nodelist; while (isdigit(*rest)) rest++; |
926f2ae04 tmpfs: cleanup mp... |
2267 2268 |
if (*rest) goto out; |
095f1fc4e mempolicy: rework... |
2269 2270 |
} break; |
095f1fc4e mempolicy: rework... |
2271 2272 2273 2274 2275 |
case MPOL_INTERLEAVE: /* * Default to online nodes with memory if no nodelist */ if (!nodelist) |
71fe804b6 mempolicy: use st... |
2276 |
nodes = node_states[N_HIGH_MEMORY]; |
3f226aa1c mempolicy: suppor... |
2277 |
break; |
71fe804b6 mempolicy: use st... |
2278 |
case MPOL_LOCAL: |
3f226aa1c mempolicy: suppor... |
2279 |
/* |
71fe804b6 mempolicy: use st... |
2280 |
* Don't allow a nodelist; mpol_new() checks flags |
3f226aa1c mempolicy: suppor... |
2281 |
*/ |
71fe804b6 mempolicy: use st... |
2282 |
if (nodelist) |
3f226aa1c mempolicy: suppor... |
2283 |
goto out; |
71fe804b6 mempolicy: use st... |
2284 |
mode = MPOL_PREFERRED; |
3f226aa1c mempolicy: suppor... |
2285 |
break; |
413b43dea tmpfs: fix oops o... |
2286 2287 2288 2289 2290 2291 2292 |
case MPOL_DEFAULT: /* * Insist on a empty nodelist */ if (!nodelist) err = 0; goto out; |
d69b2e63e tmpfs: mpol=bind:... |
2293 2294 2295 2296 2297 2298 |
case MPOL_BIND: /* * Insist on a nodelist */ if (!nodelist) goto out; |
095f1fc4e mempolicy: rework... |
2299 |
} |
71fe804b6 mempolicy: use st... |
2300 |
mode_flags = 0; |
095f1fc4e mempolicy: rework... |
2301 2302 2303 2304 2305 2306 |
if (flags) { /* * Currently, we only support two mutually exclusive * mode flags. */ if (!strcmp(flags, "static")) |
71fe804b6 mempolicy: use st... |
2307 |
mode_flags |= MPOL_F_STATIC_NODES; |
095f1fc4e mempolicy: rework... |
2308 |
else if (!strcmp(flags, "relative")) |
71fe804b6 mempolicy: use st... |
2309 |
mode_flags |= MPOL_F_RELATIVE_NODES; |
095f1fc4e mempolicy: rework... |
2310 |
else |
926f2ae04 tmpfs: cleanup mp... |
2311 |
goto out; |
095f1fc4e mempolicy: rework... |
2312 |
} |
71fe804b6 mempolicy: use st... |
2313 2314 2315 |
new = mpol_new(mode, mode_flags, &nodes); if (IS_ERR(new)) |
926f2ae04 tmpfs: cleanup mp... |
2316 |
goto out; |
e17f74af3 mempolicy: don't ... |
2317 2318 2319 2320 |
if (no_context) { /* save for contextualization */ new->w.user_nodemask = nodes; } else { |
58568d2a8 cpuset,mm: update... |
2321 |
int ret; |
4bfc44958 mm: make set_memp... |
2322 2323 2324 2325 2326 2327 2328 2329 2330 |
NODEMASK_SCRATCH(scratch); if (scratch) { task_lock(current); ret = mpol_set_nodemask(new, &nodes, scratch); task_unlock(current); } else ret = -ENOMEM; NODEMASK_SCRATCH_FREE(scratch); if (ret) { |
4bfc44958 mm: make set_memp... |
2331 |
mpol_put(new); |
926f2ae04 tmpfs: cleanup mp... |
2332 |
goto out; |
58568d2a8 cpuset,mm: update... |
2333 2334 |
} } |
926f2ae04 tmpfs: cleanup mp... |
2335 |
err = 0; |
71fe804b6 mempolicy: use st... |
2336 |
|
095f1fc4e mempolicy: rework... |
2337 2338 2339 2340 2341 2342 |
out: /* Restore string for error message */ if (nodelist) *--nodelist = ':'; if (flags) *--flags = '='; |
71fe804b6 mempolicy: use st... |
2343 2344 |
if (!err) *mpol = new; |
095f1fc4e mempolicy: rework... |
2345 2346 2347 |
return err; } #endif /* CONFIG_TMPFS */ |
71fe804b6 mempolicy: use st... |
2348 2349 2350 2351 2352 2353 2354 |
/** * mpol_to_str - format a mempolicy structure for printing * @buffer: to contain formatted mempolicy string * @maxlen: length of @buffer * @pol: pointer to mempolicy to be formatted * @no_context: "context free" mempolicy - use nodemask in w.user_nodemask * |
1a75a6c82 [PATCH] Fold numa... |
2355 2356 2357 2358 |
* Convert a mempolicy into a string. * Returns the number of characters in buffer (if positive) * or an error (negative) */ |
71fe804b6 mempolicy: use st... |
2359 |
int mpol_to_str(char *buffer, int maxlen, struct mempolicy *pol, int no_context) |
1a75a6c82 [PATCH] Fold numa... |
2360 2361 2362 2363 |
{ char *p = buffer; int l; nodemask_t nodes; |
bea904d54 mempolicy: use MP... |
2364 |
unsigned short mode; |
f5b087b52 mempolicy: add MP... |
2365 |
unsigned short flags = pol ? pol->flags : 0; |
1a75a6c82 [PATCH] Fold numa... |
2366 |
|
2291990ab mempolicy: clean-... |
2367 2368 2369 2370 |
/* * Sanity check: room for longest mode, flag and some nodes */ VM_BUG_ON(maxlen < strlen("interleave") + strlen("relative") + 16); |
bea904d54 mempolicy: use MP... |
2371 2372 2373 2374 |
if (!pol || pol == &default_policy) mode = MPOL_DEFAULT; else mode = pol->mode; |
1a75a6c82 [PATCH] Fold numa... |
2375 2376 2377 2378 2379 2380 2381 |
switch (mode) { case MPOL_DEFAULT: nodes_clear(nodes); break; case MPOL_PREFERRED: nodes_clear(nodes); |
fc36b8d3d mempolicy: use MP... |
2382 |
if (flags & MPOL_F_LOCAL) |
53f2556b6 mempolicy: mPOL_P... |
2383 2384 |
mode = MPOL_LOCAL; /* pseudo-policy */ else |
fc36b8d3d mempolicy: use MP... |
2385 |
node_set(pol->v.preferred_node, nodes); |
1a75a6c82 [PATCH] Fold numa... |
2386 2387 2388 |
break; case MPOL_BIND: |
19770b326 mm: filter based ... |
2389 |
/* Fall through */ |
1a75a6c82 [PATCH] Fold numa... |
2390 |
case MPOL_INTERLEAVE: |
71fe804b6 mempolicy: use st... |
2391 2392 2393 2394 |
if (no_context) nodes = pol->w.user_nodemask; else nodes = pol->v.nodes; |
1a75a6c82 [PATCH] Fold numa... |
2395 2396 2397 2398 |
break; default: BUG(); |
1a75a6c82 [PATCH] Fold numa... |
2399 |
} |
345ace9c7 mempolicy: rename... |
2400 |
l = strlen(policy_modes[mode]); |
53f2556b6 mempolicy: mPOL_P... |
2401 2402 |
if (buffer + maxlen < p + l + 1) return -ENOSPC; |
1a75a6c82 [PATCH] Fold numa... |
2403 |
|
345ace9c7 mempolicy: rename... |
2404 |
strcpy(p, policy_modes[mode]); |
1a75a6c82 [PATCH] Fold numa... |
2405 |
p += l; |
fc36b8d3d mempolicy: use MP... |
2406 |
if (flags & MPOL_MODE_FLAGS) { |
f5b087b52 mempolicy: add MP... |
2407 2408 2409 |
if (buffer + maxlen < p + 2) return -ENOSPC; *p++ = '='; |
2291990ab mempolicy: clean-... |
2410 2411 2412 |
/* * Currently, the only defined flags are mutually exclusive */ |
f5b087b52 mempolicy: add MP... |
2413 |
if (flags & MPOL_F_STATIC_NODES) |
2291990ab mempolicy: clean-... |
2414 2415 2416 |
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... |
2417 |
} |
1a75a6c82 [PATCH] Fold numa... |
2418 2419 2420 |
if (!nodes_empty(nodes)) { if (buffer + maxlen < p + 2) return -ENOSPC; |
095f1fc4e mempolicy: rework... |
2421 |
*p++ = ':'; |
1a75a6c82 [PATCH] Fold numa... |
2422 2423 2424 2425 2426 2427 2428 2429 |
p += nodelist_scnprintf(p, buffer + maxlen - p, nodes); } return p - buffer; } struct numa_maps { unsigned long pages; unsigned long anon; |
397874dfe [PATCH] numa_maps... |
2430 2431 |
unsigned long active; unsigned long writeback; |
1a75a6c82 [PATCH] Fold numa... |
2432 |
unsigned long mapcount_max; |
397874dfe [PATCH] numa_maps... |
2433 2434 |
unsigned long dirty; unsigned long swapcache; |
1a75a6c82 [PATCH] Fold numa... |
2435 2436 |
unsigned long node[MAX_NUMNODES]; }; |
397874dfe [PATCH] numa_maps... |
2437 |
static void gather_stats(struct page *page, void *private, int pte_dirty) |
1a75a6c82 [PATCH] Fold numa... |
2438 2439 2440 |
{ struct numa_maps *md = private; int count = page_mapcount(page); |
397874dfe [PATCH] numa_maps... |
2441 2442 2443 |
md->pages++; if (pte_dirty || PageDirty(page)) md->dirty++; |
1a75a6c82 [PATCH] Fold numa... |
2444 |
|
397874dfe [PATCH] numa_maps... |
2445 2446 |
if (PageSwapCache(page)) md->swapcache++; |
1a75a6c82 [PATCH] Fold numa... |
2447 |
|
894bc3104 Unevictable LRU I... |
2448 |
if (PageActive(page) || PageUnevictable(page)) |
397874dfe [PATCH] numa_maps... |
2449 2450 2451 2452 |
md->active++; if (PageWriteback(page)) md->writeback++; |
1a75a6c82 [PATCH] Fold numa... |
2453 2454 2455 |
if (PageAnon(page)) md->anon++; |
397874dfe [PATCH] numa_maps... |
2456 2457 |
if (count > md->mapcount_max) md->mapcount_max = count; |
1a75a6c82 [PATCH] Fold numa... |
2458 |
md->node[page_to_nid(page)]++; |
1a75a6c82 [PATCH] Fold numa... |
2459 |
} |
7f709ed0e [PATCH] numa_maps... |
2460 |
#ifdef CONFIG_HUGETLB_PAGE |
397874dfe [PATCH] numa_maps... |
2461 2462 2463 2464 2465 2466 |
static void check_huge_range(struct vm_area_struct *vma, unsigned long start, unsigned long end, struct numa_maps *md) { unsigned long addr; struct page *page; |
a55164389 hugetlb: modular ... |
2467 2468 |
struct hstate *h = hstate_vma(vma); unsigned long sz = huge_page_size(h); |
397874dfe [PATCH] numa_maps... |
2469 |
|
a55164389 hugetlb: modular ... |
2470 2471 2472 |
for (addr = start; addr < end; addr += sz) { pte_t *ptep = huge_pte_offset(vma->vm_mm, addr & huge_page_mask(h)); |
397874dfe [PATCH] numa_maps... |
2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 |
pte_t pte; if (!ptep) continue; pte = *ptep; if (pte_none(pte)) continue; page = pte_page(pte); if (!page) continue; gather_stats(page, md, pte_dirty(*ptep)); } } |
7f709ed0e [PATCH] numa_maps... |
2489 2490 2491 2492 2493 2494 2495 |
#else static inline void check_huge_range(struct vm_area_struct *vma, unsigned long start, unsigned long end, struct numa_maps *md) { } #endif |
397874dfe [PATCH] numa_maps... |
2496 |
|
53f2556b6 mempolicy: mPOL_P... |
2497 2498 2499 |
/* * Display pages allocated per node and memory policy via /proc. */ |
1a75a6c82 [PATCH] Fold numa... |
2500 2501 |
int show_numa_map(struct seq_file *m, void *v) { |
99f895518 [PATCH] proc: don... |
2502 |
struct proc_maps_private *priv = m->private; |
1a75a6c82 [PATCH] Fold numa... |
2503 2504 |
struct vm_area_struct *vma = v; struct numa_maps *md; |
397874dfe [PATCH] numa_maps... |
2505 2506 |
struct file *file = vma->vm_file; struct mm_struct *mm = vma->vm_mm; |
480eccf9a Fix NUMA Memory P... |
2507 |
struct mempolicy *pol; |
1a75a6c82 [PATCH] Fold numa... |
2508 2509 |
int n; char buffer[50]; |
397874dfe [PATCH] numa_maps... |
2510 |
if (!mm) |
1a75a6c82 [PATCH] Fold numa... |
2511 2512 2513 2514 2515 |
return 0; md = kzalloc(sizeof(struct numa_maps), GFP_KERNEL); if (!md) return 0; |
480eccf9a Fix NUMA Memory P... |
2516 |
pol = get_vma_policy(priv->task, vma, vma->vm_start); |
71fe804b6 mempolicy: use st... |
2517 |
mpol_to_str(buffer, sizeof(buffer), pol, 0); |
52cd3b074 mempolicy: rework... |
2518 |
mpol_cond_put(pol); |
397874dfe [PATCH] numa_maps... |
2519 2520 2521 2522 2523 |
seq_printf(m, "%08lx %s", vma->vm_start, buffer); if (file) { seq_printf(m, " file="); |
c32c2f63a d_path: Make seq_... |
2524 2525 |
seq_path(m, &file->f_path, " \t= "); |
397874dfe [PATCH] numa_maps... |
2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 |
} else if (vma->vm_start <= mm->brk && vma->vm_end >= mm->start_brk) { seq_printf(m, " heap"); } else if (vma->vm_start <= mm->start_stack && vma->vm_end >= mm->start_stack) { seq_printf(m, " stack"); } if (is_vm_hugetlb_page(vma)) { check_huge_range(vma, vma->vm_start, vma->vm_end, md); seq_printf(m, " huge"); } else { |
a57ebfdb2 [PATCH] numa_maps... |
2537 |
check_pgd_range(vma, vma->vm_start, vma->vm_end, |
56bbd65df Memoryless nodes:... |
2538 |
&node_states[N_HIGH_MEMORY], MPOL_MF_STATS, md); |
397874dfe [PATCH] numa_maps... |
2539 2540 2541 2542 |
} if (!md->pages) goto out; |
1a75a6c82 [PATCH] Fold numa... |
2543 |
|
397874dfe [PATCH] numa_maps... |
2544 2545 |
if (md->anon) seq_printf(m," anon=%lu",md->anon); |
1a75a6c82 [PATCH] Fold numa... |
2546 |
|
397874dfe [PATCH] numa_maps... |
2547 2548 |
if (md->dirty) seq_printf(m," dirty=%lu",md->dirty); |
1a75a6c82 [PATCH] Fold numa... |
2549 |
|
397874dfe [PATCH] numa_maps... |
2550 2551 |
if (md->pages != md->anon && md->pages != md->dirty) seq_printf(m, " mapped=%lu", md->pages); |
1a75a6c82 [PATCH] Fold numa... |
2552 |
|
397874dfe [PATCH] numa_maps... |
2553 2554 |
if (md->mapcount_max > 1) seq_printf(m, " mapmax=%lu", md->mapcount_max); |
1a75a6c82 [PATCH] Fold numa... |
2555 |
|
397874dfe [PATCH] numa_maps... |
2556 2557 2558 2559 2560 2561 2562 2563 |
if (md->swapcache) seq_printf(m," swapcache=%lu", md->swapcache); if (md->active < md->pages && !is_vm_hugetlb_page(vma)) seq_printf(m," active=%lu", md->active); if (md->writeback) seq_printf(m," writeback=%lu", md->writeback); |
56bbd65df Memoryless nodes:... |
2564 |
for_each_node_state(n, N_HIGH_MEMORY) |
397874dfe [PATCH] numa_maps... |
2565 2566 2567 2568 2569 |
if (md->node[n]) seq_printf(m, " N%d=%lu", n, md->node[n]); out: seq_putc(m, ' '); |
1a75a6c82 [PATCH] Fold numa... |
2570 2571 2572 |
kfree(md); if (m->count < m->size) |
99f895518 [PATCH] proc: don... |
2573 |
m->version = (vma != priv->tail_vma) ? vma->vm_start : 0; |
1a75a6c82 [PATCH] Fold numa... |
2574 2575 |
return 0; } |