Blame view
mm/mempolicy.c
70.6 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 |
/* * Simple NUMA memory policy for the Linux kernel. * * Copyright 2003,2004 Andi Kleen, SuSE Labs. |
8bccd85ff [PATCH] Implement... |
5 |
* (C) Copyright 2005 Christoph Lameter, Silicon Graphics, Inc. |
1da177e4c Linux-2.6.12-rc2 |
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
* Subject to the GNU Public License, version 2. * * NUMA policy allows the user to give hints in which node(s) memory should * be allocated. * * Support four policies per VMA and per process: * * The VMA policy has priority over the process policy for a page fault. * * interleave Allocate memory interleaved over a set of nodes, * with normal fallback if it fails. * For VMA based allocations this interleaves based on the * offset into the backing object or offset into the mapping * for anonymous memory. For process policy an process counter * is used. |
8bccd85ff [PATCH] Implement... |
21 |
* |
1da177e4c Linux-2.6.12-rc2 |
22 23 |
* bind Only allocate memory on a specific set of nodes, * no fallback. |
8bccd85ff [PATCH] Implement... |
24 25 26 27 |
* FIXME: memory is allocated starting with the first node * to the last. It would be better if bind would truly restrict * the allocation to memory nodes instead * |
1da177e4c Linux-2.6.12-rc2 |
28 |
* preferred Try a specific node first before normal fallback. |
00ef2d2f8 mm: use NUMA_NO_NODE |
29 |
* As a special case NUMA_NO_NODE here means do the allocation |
1da177e4c Linux-2.6.12-rc2 |
30 31 32 |
* on the local CPU. This is normally identical to default, * but useful to set in a VMA when you have a non default * process policy. |
8bccd85ff [PATCH] Implement... |
33 |
* |
1da177e4c Linux-2.6.12-rc2 |
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
* default Allocate on the local node first, or when on a VMA * use the process policy. This is what Linux always did * in a NUMA aware kernel and still does by, ahem, default. * * The process policy is applied for most non interrupt memory allocations * in that process' context. Interrupts ignore the policies and always * try to allocate on the local CPU. The VMA policy is only applied for memory * allocations for a VMA in the VM. * * Currently there are a few corner cases in swapping where the policy * is not applied, but the majority should be handled. When process policy * is used it is not remembered over swap outs/swap ins. * * Only the highest zone in the zone hierarchy gets policied. Allocations * requesting a lower zone just use default policy. This implies that * on systems with highmem kernel lowmem allocation don't get policied. * Same with GFP_DMA allocations. * * For shmfs/tmpfs/hugetlbfs shared memory the policy is shared between * all users and remembered even when nobody has memory mapped. */ /* Notebook: fix mmap readahead to honour policy and enable policy for any page cache object statistics for bigpages global policy for page cache? currently it uses process policy. Requires first item above. handle mremap for shared memory (currently ignored for the policy) grows down? make bind policy root only? It can trigger oom much faster and the kernel is not always grateful with that. |
1da177e4c Linux-2.6.12-rc2 |
66 |
*/ |
b1de0d139 mm: convert some ... |
67 |
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt |
1da177e4c Linux-2.6.12-rc2 |
68 69 70 71 72 73 |
#include <linux/mempolicy.h> #include <linux/mm.h> #include <linux/highmem.h> #include <linux/hugetlb.h> #include <linux/kernel.h> #include <linux/sched.h> |
1da177e4c Linux-2.6.12-rc2 |
74 75 |
#include <linux/nodemask.h> #include <linux/cpuset.h> |
1da177e4c Linux-2.6.12-rc2 |
76 77 |
#include <linux/slab.h> #include <linux/string.h> |
b95f1b31b mm: Map most file... |
78 |
#include <linux/export.h> |
b488893a3 pid namespaces: c... |
79 |
#include <linux/nsproxy.h> |
1da177e4c Linux-2.6.12-rc2 |
80 81 82 |
#include <linux/interrupt.h> #include <linux/init.h> #include <linux/compat.h> |
dc9aa5b9d [PATCH] Swap Migr... |
83 |
#include <linux/swap.h> |
1a75a6c82 [PATCH] Fold numa... |
84 85 |
#include <linux/seq_file.h> #include <linux/proc_fs.h> |
b20a35035 [PATCH] page migr... |
86 |
#include <linux/migrate.h> |
62b61f611 ksm: memory hotre... |
87 |
#include <linux/ksm.h> |
95a402c38 [PATCH] page migr... |
88 |
#include <linux/rmap.h> |
86c3a7645 [PATCH] SELinux: ... |
89 |
#include <linux/security.h> |
dbcb0f19c mm/mempolicy.c: c... |
90 |
#include <linux/syscalls.h> |
095f1fc4e mempolicy: rework... |
91 |
#include <linux/ctype.h> |
6d9c285a6 mm: move inc_zone... |
92 |
#include <linux/mm_inline.h> |
b24f53a0b mm: mempolicy: Ad... |
93 |
#include <linux/mmu_notifier.h> |
b1de0d139 mm: convert some ... |
94 |
#include <linux/printk.h> |
dc9aa5b9d [PATCH] Swap Migr... |
95 |
|
1da177e4c Linux-2.6.12-rc2 |
96 97 |
#include <asm/tlbflush.h> #include <asm/uaccess.h> |
778d3b0ff cpusets: randomiz... |
98 |
#include <linux/random.h> |
1da177e4c Linux-2.6.12-rc2 |
99 |
|
62695a84e vmscan: move isol... |
100 |
#include "internal.h" |
38e35860d [PATCH] mempolici... |
101 |
/* Internal flags */ |
dc9aa5b9d [PATCH] Swap Migr... |
102 |
#define MPOL_MF_DISCONTIG_OK (MPOL_MF_INTERNAL << 0) /* Skip checks for continuous vmas */ |
38e35860d [PATCH] mempolici... |
103 |
#define MPOL_MF_INVERT (MPOL_MF_INTERNAL << 1) /* Invert check for nodemask */ |
dc9aa5b9d [PATCH] Swap Migr... |
104 |
|
fcc234f88 [PATCH] mm: kill ... |
105 106 |
static struct kmem_cache *policy_cache; static struct kmem_cache *sn_cache; |
1da177e4c Linux-2.6.12-rc2 |
107 |
|
1da177e4c Linux-2.6.12-rc2 |
108 109 |
/* Highest zone. An specific allocation for a zone below that is not policied. */ |
6267276f3 [PATCH] optional ... |
110 |
enum zone_type policy_zone = 0; |
1da177e4c Linux-2.6.12-rc2 |
111 |
|
bea904d54 mempolicy: use MP... |
112 113 114 |
/* * run-time system-wide default policy => local allocation */ |
e754d79d3 mm/mempolicy.c: q... |
115 |
static struct mempolicy default_policy = { |
1da177e4c Linux-2.6.12-rc2 |
116 |
.refcnt = ATOMIC_INIT(1), /* never free it */ |
bea904d54 mempolicy: use MP... |
117 |
.mode = MPOL_PREFERRED, |
fc36b8d3d mempolicy: use MP... |
118 |
.flags = MPOL_F_LOCAL, |
1da177e4c Linux-2.6.12-rc2 |
119 |
}; |
5606e3877 mm: numa: Migrate... |
120 |
static struct mempolicy preferred_node_policy[MAX_NUMNODES]; |
74d2c3a05 mempolicy: introd... |
121 |
struct mempolicy *get_task_policy(struct task_struct *p) |
5606e3877 mm: numa: Migrate... |
122 123 |
{ struct mempolicy *pol = p->mempolicy; |
f15ca78e3 mempolicy: change... |
124 |
int node; |
5606e3877 mm: numa: Migrate... |
125 |
|
f15ca78e3 mempolicy: change... |
126 127 |
if (pol) return pol; |
5606e3877 mm: numa: Migrate... |
128 |
|
f15ca78e3 mempolicy: change... |
129 130 131 132 133 134 |
node = numa_node_id(); if (node != NUMA_NO_NODE) { pol = &preferred_node_policy[node]; /* preferred_node_policy is not initialised early in boot */ if (pol->mode) return pol; |
5606e3877 mm: numa: Migrate... |
135 |
} |
f15ca78e3 mempolicy: change... |
136 |
return &default_policy; |
5606e3877 mm: numa: Migrate... |
137 |
} |
37012946d mempolicy: create... |
138 139 |
static const struct mempolicy_operations { int (*create)(struct mempolicy *pol, const nodemask_t *nodes); |
708c1bbc9 mempolicy: restru... |
140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 |
/* * 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... |
156 |
} mpol_ops[MPOL_MAX]; |
f5b087b52 mempolicy: add MP... |
157 158 |
static inline int mpol_store_user_nodemask(const struct mempolicy *pol) { |
6d556294d mempolicy: remove... |
159 |
return pol->flags & MPOL_MODE_FLAGS; |
4c50bc011 mempolicy: add MP... |
160 161 162 163 164 165 166 167 |
} 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... |
168 |
} |
37012946d mempolicy: create... |
169 170 171 172 173 174 175 176 177 178 179 |
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... |
180 |
pol->flags |= MPOL_F_LOCAL; /* local allocation */ |
37012946d mempolicy: create... |
181 182 183 184 185 186 187 188 189 |
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) { |
859f7ef14 mm/mempolicy.c: r... |
190 |
if (nodes_empty(*nodes)) |
37012946d mempolicy: create... |
191 192 193 194 |
return -EINVAL; pol->v.nodes = *nodes; return 0; } |
58568d2a8 cpuset,mm: update... |
195 196 197 198 199 200 201 202 203 |
/* * 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... |
204 205 |
static int mpol_set_nodemask(struct mempolicy *pol, const nodemask_t *nodes, struct nodemask_scratch *nsc) |
58568d2a8 cpuset,mm: update... |
206 |
{ |
58568d2a8 cpuset,mm: update... |
207 208 209 210 211 |
int ret; /* if mode is MPOL_DEFAULT, pol is NULL. This is right. */ if (pol == NULL) return 0; |
01f13bd60 mempolicy: use N_... |
212 |
/* Check N_MEMORY */ |
4bfc44958 mm: make set_memp... |
213 |
nodes_and(nsc->mask1, |
01f13bd60 mempolicy: use N_... |
214 |
cpuset_current_mems_allowed, node_states[N_MEMORY]); |
58568d2a8 cpuset,mm: update... |
215 216 217 218 219 220 |
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) |
859f7ef14 mm/mempolicy.c: r... |
221 |
mpol_relative_nodemask(&nsc->mask2, nodes, &nsc->mask1); |
58568d2a8 cpuset,mm: update... |
222 |
else |
4bfc44958 mm: make set_memp... |
223 |
nodes_and(nsc->mask2, *nodes, nsc->mask1); |
58568d2a8 cpuset,mm: update... |
224 225 226 227 228 229 |
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... |
230 231 232 233 |
if (nodes) ret = mpol_ops[pol->mode].create(pol, &nsc->mask2); else ret = mpol_ops[pol->mode].create(pol, NULL); |
58568d2a8 cpuset,mm: update... |
234 235 236 237 238 239 240 |
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... |
241 242 |
static struct mempolicy *mpol_new(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
243 244 |
{ struct mempolicy *policy; |
028fec414 mempolicy: suppor... |
245 246 |
pr_debug("setting mode %d flags %d nodes[0] %lx ", |
00ef2d2f8 mm: use NUMA_NO_NODE |
247 |
mode, flags, nodes ? nodes_addr(*nodes)[0] : NUMA_NO_NODE); |
140d5a490 numa: mempolicy: ... |
248 |
|
3e1f06456 mempolicy: disall... |
249 250 |
if (mode == MPOL_DEFAULT) { if (nodes && !nodes_empty(*nodes)) |
37012946d mempolicy: create... |
251 |
return ERR_PTR(-EINVAL); |
d3a710337 mm: mempolicy: Ad... |
252 |
return NULL; |
37012946d mempolicy: create... |
253 |
} |
3e1f06456 mempolicy: disall... |
254 255 256 257 258 259 260 261 262 263 264 265 |
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... |
266 |
} |
479e2802d mm: mempolicy: Ma... |
267 268 269 270 |
} else if (mode == MPOL_LOCAL) { if (!nodes_empty(*nodes)) return ERR_PTR(-EINVAL); mode = MPOL_PREFERRED; |
3e1f06456 mempolicy: disall... |
271 272 |
} else if (nodes_empty(*nodes)) return ERR_PTR(-EINVAL); |
1da177e4c Linux-2.6.12-rc2 |
273 274 275 276 |
policy = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!policy) return ERR_PTR(-ENOMEM); atomic_set(&policy->refcnt, 1); |
45c4745af mempolicy: rename... |
277 |
policy->mode = mode; |
3e1f06456 mempolicy: disall... |
278 |
policy->flags = flags; |
37012946d mempolicy: create... |
279 |
|
1da177e4c Linux-2.6.12-rc2 |
280 |
return policy; |
37012946d mempolicy: create... |
281 |
} |
52cd3b074 mempolicy: rework... |
282 283 284 285 286 |
/* Slow path of a mpol destructor. */ void __mpol_put(struct mempolicy *p) { if (!atomic_dec_and_test(&p->refcnt)) return; |
52cd3b074 mempolicy: rework... |
287 288 |
kmem_cache_free(policy_cache, p); } |
708c1bbc9 mempolicy: restru... |
289 290 |
static void mpol_rebind_default(struct mempolicy *pol, const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
291 292 |
{ } |
708c1bbc9 mempolicy: restru... |
293 294 295 296 297 298 299 300 |
/* * 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... |
301 302 303 304 305 306 307 308 |
{ 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... |
309 310 311 312 313 314 315 316 317 318 319 320 321 |
/* * 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... |
322 |
} |
f5b087b52 mempolicy: add MP... |
323 |
|
708c1bbc9 mempolicy: restru... |
324 325 326 327 328 329 330 331 332 |
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... |
333 334 335 336 337 338 339 340 341 342 |
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... |
343 344 |
const nodemask_t *nodes, enum mpol_rebind_step step) |
37012946d mempolicy: create... |
345 346 |
{ nodemask_t tmp; |
37012946d mempolicy: create... |
347 348 |
if (pol->flags & MPOL_F_STATIC_NODES) { int node = first_node(pol->w.user_nodemask); |
fc36b8d3d mempolicy: use MP... |
349 |
if (node_isset(node, *nodes)) { |
37012946d mempolicy: create... |
350 |
pol->v.preferred_node = node; |
fc36b8d3d mempolicy: use MP... |
351 352 353 |
pol->flags &= ~MPOL_F_LOCAL; } else pol->flags |= MPOL_F_LOCAL; |
37012946d mempolicy: create... |
354 355 356 |
} 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... |
357 |
} else if (!(pol->flags & MPOL_F_LOCAL)) { |
37012946d mempolicy: create... |
358 359 360 361 362 |
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 |
363 |
} |
708c1bbc9 mempolicy: restru... |
364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 |
/* * 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... |
382 |
{ |
1d0d2680a mempolicy: move r... |
383 384 |
if (!pol) return; |
89c522c78 mm/mempolicy.c: u... |
385 |
if (!mpol_store_user_nodemask(pol) && step == MPOL_REBIND_ONCE && |
1d0d2680a mempolicy: move r... |
386 387 |
nodes_equal(pol->w.cpuset_mems_allowed, *newmask)) return; |
708c1bbc9 mempolicy: restru... |
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 |
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... |
403 404 405 406 407 |
} /* * Wrapper for mpol_rebind_policy() that just requires task * pointer, and updates task mempolicy. |
58568d2a8 cpuset,mm: update... |
408 409 |
* * Called with task's alloc_lock held. |
1d0d2680a mempolicy: move r... |
410 |
*/ |
708c1bbc9 mempolicy: restru... |
411 412 |
void mpol_rebind_task(struct task_struct *tsk, const nodemask_t *new, enum mpol_rebind_step step) |
1d0d2680a mempolicy: move r... |
413 |
{ |
708c1bbc9 mempolicy: restru... |
414 |
mpol_rebind_policy(tsk->mempolicy, new, step); |
1d0d2680a mempolicy: move r... |
415 416 417 418 419 420 421 422 423 424 425 426 427 428 |
} /* * 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... |
429 |
mpol_rebind_policy(vma->vm_policy, new, MPOL_REBIND_ONCE); |
1d0d2680a mempolicy: move r... |
430 431 |
up_write(&mm->mmap_sem); } |
37012946d mempolicy: create... |
432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 |
static const struct mempolicy_operations mpol_ops[MPOL_MAX] = { [MPOL_DEFAULT] = { .rebind = mpol_rebind_default, }, [MPOL_INTERLEAVE] = { .create = mpol_new_interleave, .rebind = mpol_rebind_nodemask, }, [MPOL_PREFERRED] = { .create = mpol_new_preferred, .rebind = mpol_rebind_preferred, }, [MPOL_BIND] = { .create = mpol_new_bind, .rebind = mpol_rebind_nodemask, }, }; |
fc3012896 [PATCH] Simplify ... |
449 450 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags); |
1a75a6c82 [PATCH] Fold numa... |
451 |
|
6f4576e36 mempolicy: apply ... |
452 453 454 455 456 457 |
struct queue_pages { struct list_head *pagelist; unsigned long flags; nodemask_t *nmask; struct vm_area_struct *prev; }; |
980949457 mm/mempolicy: ren... |
458 459 460 461 |
/* * Scan through pages checking if pages follow certain conditions, * and move them to the pagelist if they do. */ |
6f4576e36 mempolicy: apply ... |
462 463 |
static int queue_pages_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, struct mm_walk *walk) |
1da177e4c Linux-2.6.12-rc2 |
464 |
{ |
6f4576e36 mempolicy: apply ... |
465 466 467 468 469 |
struct vm_area_struct *vma = walk->vma; struct page *page; struct queue_pages *qp = walk->private; unsigned long flags = qp->flags; int nid; |
91612e0df [PATCH] mbind: ch... |
470 |
pte_t *pte; |
705e87c0c [PATCH] mm: pte_o... |
471 |
spinlock_t *ptl; |
941150a32 [PATCH] mbind: fi... |
472 |
|
6f4576e36 mempolicy: apply ... |
473 474 475 |
split_huge_page_pmd(vma, addr, pmd); if (pmd_trans_unstable(pmd)) return 0; |
91612e0df [PATCH] mbind: ch... |
476 |
|
6f4576e36 mempolicy: apply ... |
477 478 |
pte = pte_offset_map_lock(walk->mm, pmd, addr, &ptl); for (; addr != end; pte++, addr += PAGE_SIZE) { |
91612e0df [PATCH] mbind: ch... |
479 |
if (!pte_present(*pte)) |
1da177e4c Linux-2.6.12-rc2 |
480 |
continue; |
6aab341e0 mm: re-architect ... |
481 482 |
page = vm_normal_page(vma, addr, *pte); if (!page) |
1da177e4c Linux-2.6.12-rc2 |
483 |
continue; |
053837fce [PATCH] mm: migra... |
484 |
/* |
62b61f611 ksm: memory hotre... |
485 486 |
* vm_normal_page() filters out zero pages, but there might * still be PageReserved pages to skip, perhaps in a VDSO. |
053837fce [PATCH] mm: migra... |
487 |
*/ |
b79bc0a0c ksm: enable KSM p... |
488 |
if (PageReserved(page)) |
f4598c8b3 [PATCH] migration... |
489 |
continue; |
6aab341e0 mm: re-architect ... |
490 |
nid = page_to_nid(page); |
6f4576e36 mempolicy: apply ... |
491 |
if (node_isset(nid, *qp->nmask) == !!(flags & MPOL_MF_INVERT)) |
38e35860d [PATCH] mempolici... |
492 |
continue; |
b1f72d185 mm: remove MPOL_M... |
493 |
if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) |
6f4576e36 mempolicy: apply ... |
494 495 496 497 498 |
migrate_page_add(page, qp->pagelist, flags); } pte_unmap_unlock(pte - 1, ptl); cond_resched(); return 0; |
91612e0df [PATCH] mbind: ch... |
499 |
} |
6f4576e36 mempolicy: apply ... |
500 501 502 |
static int queue_pages_hugetlb(pte_t *pte, unsigned long hmask, unsigned long addr, unsigned long end, struct mm_walk *walk) |
e2d8cf405 migrate: add huge... |
503 504 |
{ #ifdef CONFIG_HUGETLB_PAGE |
6f4576e36 mempolicy: apply ... |
505 506 |
struct queue_pages *qp = walk->private; unsigned long flags = qp->flags; |
e2d8cf405 migrate: add huge... |
507 508 |
int nid; struct page *page; |
cb900f412 mm, hugetlb: conv... |
509 |
spinlock_t *ptl; |
d4c54919e mm: add !pte_pres... |
510 |
pte_t entry; |
e2d8cf405 migrate: add huge... |
511 |
|
6f4576e36 mempolicy: apply ... |
512 513 |
ptl = huge_pte_lock(hstate_vma(walk->vma), walk->mm, pte); entry = huge_ptep_get(pte); |
d4c54919e mm: add !pte_pres... |
514 515 516 |
if (!pte_present(entry)) goto unlock; page = pte_page(entry); |
e2d8cf405 migrate: add huge... |
517 |
nid = page_to_nid(page); |
6f4576e36 mempolicy: apply ... |
518 |
if (node_isset(nid, *qp->nmask) == !!(flags & MPOL_MF_INVERT)) |
e2d8cf405 migrate: add huge... |
519 520 521 522 |
goto unlock; /* With MPOL_MF_MOVE, we migrate only unshared hugepage. */ if (flags & (MPOL_MF_MOVE_ALL) || (flags & MPOL_MF_MOVE && page_mapcount(page) == 1)) |
6f4576e36 mempolicy: apply ... |
523 |
isolate_huge_page(page, qp->pagelist); |
e2d8cf405 migrate: add huge... |
524 |
unlock: |
cb900f412 mm, hugetlb: conv... |
525 |
spin_unlock(ptl); |
e2d8cf405 migrate: add huge... |
526 527 528 |
#else BUG(); #endif |
91612e0df [PATCH] mbind: ch... |
529 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
530 |
} |
5877231f6 mm: Move change_p... |
531 |
#ifdef CONFIG_NUMA_BALANCING |
b24f53a0b mm: mempolicy: Ad... |
532 |
/* |
4b10e7d56 mm: mempolicy: Im... |
533 534 535 536 537 538 539 |
* This is used to mark a range of virtual addresses to be inaccessible. * These are later cleared by a NUMA hinting fault. Depending on these * faults, pages may be migrated for better NUMA placement. * * This is assuming that NUMA faults are handled using PROT_NONE. If * an architecture makes a different choice, it will need further * changes to the core. |
b24f53a0b mm: mempolicy: Ad... |
540 |
*/ |
4b10e7d56 mm: mempolicy: Im... |
541 542 |
unsigned long change_prot_numa(struct vm_area_struct *vma, unsigned long addr, unsigned long end) |
b24f53a0b mm: mempolicy: Ad... |
543 |
{ |
4b10e7d56 mm: mempolicy: Im... |
544 |
int nr_updated; |
b24f53a0b mm: mempolicy: Ad... |
545 |
|
4d9424669 mm: convert p[te|... |
546 |
nr_updated = change_protection(vma, addr, end, PAGE_NONE, 0, 1); |
03c5a6e16 mm: numa: Add pte... |
547 548 |
if (nr_updated) count_vm_numa_events(NUMA_PTE_UPDATES, nr_updated); |
b24f53a0b mm: mempolicy: Ad... |
549 |
|
4b10e7d56 mm: mempolicy: Im... |
550 |
return nr_updated; |
b24f53a0b mm: mempolicy: Ad... |
551 552 553 554 555 556 557 |
} #else static unsigned long change_prot_numa(struct vm_area_struct *vma, unsigned long addr, unsigned long end) { return 0; } |
5877231f6 mm: Move change_p... |
558 |
#endif /* CONFIG_NUMA_BALANCING */ |
b24f53a0b mm: mempolicy: Ad... |
559 |
|
6f4576e36 mempolicy: apply ... |
560 561 562 563 564 565 566 |
static int queue_pages_test_walk(unsigned long start, unsigned long end, struct mm_walk *walk) { struct vm_area_struct *vma = walk->vma; struct queue_pages *qp = walk->private; unsigned long endvma = vma->vm_end; unsigned long flags = qp->flags; |
48684a65b mm: pagewalk: fix... |
567 568 |
if (vma->vm_flags & VM_PFNMAP) return 1; |
6f4576e36 mempolicy: apply ... |
569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 |
if (endvma > end) endvma = end; if (vma->vm_start > start) start = vma->vm_start; if (!(flags & MPOL_MF_DISCONTIG_OK)) { if (!vma->vm_next && vma->vm_end < end) return -EFAULT; if (qp->prev && qp->prev->vm_end < vma->vm_start) return -EFAULT; } qp->prev = vma; if (vma->vm_flags & VM_PFNMAP) return 1; if (flags & MPOL_MF_LAZY) { /* Similar to task_numa_work, skip inaccessible VMAs */ if (vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)) change_prot_numa(vma, start, endvma); return 1; } if ((flags & MPOL_MF_STRICT) || ((flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) && vma_migratable(vma))) /* queue pages from current vma */ return 0; return 1; } |
dc9aa5b9d [PATCH] Swap Migr... |
600 |
/* |
980949457 mm/mempolicy: ren... |
601 602 603 604 605 |
* Walk through page tables and collect pages to be migrated. * * If pages found in a given range are on a set of nodes (determined by * @nodes and @flags,) it's isolated and queued to the pagelist which is * passed via @private.) |
dc9aa5b9d [PATCH] Swap Migr... |
606 |
*/ |
d05f0cdcb mm: fix crashes f... |
607 |
static int |
980949457 mm/mempolicy: ren... |
608 |
queue_pages_range(struct mm_struct *mm, unsigned long start, unsigned long end, |
6f4576e36 mempolicy: apply ... |
609 610 |
nodemask_t *nodes, unsigned long flags, struct list_head *pagelist) |
1da177e4c Linux-2.6.12-rc2 |
611 |
{ |
6f4576e36 mempolicy: apply ... |
612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 |
struct queue_pages qp = { .pagelist = pagelist, .flags = flags, .nmask = nodes, .prev = NULL, }; struct mm_walk queue_pages_walk = { .hugetlb_entry = queue_pages_hugetlb, .pmd_entry = queue_pages_pte_range, .test_walk = queue_pages_test_walk, .mm = mm, .private = &qp, }; return walk_page_range(start, end, &queue_pages_walk); |
1da177e4c Linux-2.6.12-rc2 |
627 |
} |
869833f2c mempolicy: remove... |
628 629 630 631 632 633 |
/* * Apply policy to a single VMA * This must be called with the mmap_sem held for writing. */ static int vma_replace_policy(struct vm_area_struct *vma, struct mempolicy *pol) |
8d34694c1 revert "mm: mempo... |
634 |
{ |
869833f2c mempolicy: remove... |
635 636 637 |
int err; struct mempolicy *old; struct mempolicy *new; |
8d34694c1 revert "mm: mempo... |
638 639 640 641 642 643 |
pr_debug("vma %lx-%lx/%lx vm_ops %p vm_file %p set_policy %p ", vma->vm_start, vma->vm_end, vma->vm_pgoff, vma->vm_ops, vma->vm_file, vma->vm_ops ? vma->vm_ops->set_policy : NULL); |
869833f2c mempolicy: remove... |
644 645 646 647 648 |
new = mpol_dup(pol); if (IS_ERR(new)) return PTR_ERR(new); if (vma->vm_ops && vma->vm_ops->set_policy) { |
8d34694c1 revert "mm: mempo... |
649 |
err = vma->vm_ops->set_policy(vma, new); |
869833f2c mempolicy: remove... |
650 651 |
if (err) goto err_out; |
8d34694c1 revert "mm: mempo... |
652 |
} |
869833f2c mempolicy: remove... |
653 654 655 656 657 658 659 660 |
old = vma->vm_policy; vma->vm_policy = new; /* protected by mmap_sem */ mpol_put(old); return 0; err_out: mpol_put(new); |
8d34694c1 revert "mm: mempo... |
661 662 |
return err; } |
1da177e4c Linux-2.6.12-rc2 |
663 |
/* Step 2: apply policy to a range and do splits. */ |
9d8cebd4b mm: fix mbind vma... |
664 665 |
static int mbind_range(struct mm_struct *mm, unsigned long start, unsigned long end, struct mempolicy *new_pol) |
1da177e4c Linux-2.6.12-rc2 |
666 667 |
{ struct vm_area_struct *next; |
9d8cebd4b mm: fix mbind vma... |
668 669 670 |
struct vm_area_struct *prev; struct vm_area_struct *vma; int err = 0; |
e26a51148 mm/mempolicy.c: r... |
671 |
pgoff_t pgoff; |
9d8cebd4b mm: fix mbind vma... |
672 673 |
unsigned long vmstart; unsigned long vmend; |
1da177e4c Linux-2.6.12-rc2 |
674 |
|
097d59106 vm: avoid using f... |
675 |
vma = find_vma(mm, start); |
9d8cebd4b mm: fix mbind vma... |
676 677 |
if (!vma || vma->vm_start > start) return -EFAULT; |
097d59106 vm: avoid using f... |
678 |
prev = vma->vm_prev; |
e26a51148 mm/mempolicy.c: r... |
679 680 |
if (start > vma->vm_start) prev = vma; |
9d8cebd4b mm: fix mbind vma... |
681 |
for (; vma && vma->vm_start < end; prev = vma, vma = next) { |
1da177e4c Linux-2.6.12-rc2 |
682 |
next = vma->vm_next; |
9d8cebd4b mm: fix mbind vma... |
683 684 |
vmstart = max(start, vma->vm_start); vmend = min(end, vma->vm_end); |
e26a51148 mm/mempolicy.c: r... |
685 686 687 688 689 |
if (mpol_equal(vma_policy(vma), new_pol)) continue; pgoff = vma->vm_pgoff + ((vmstart - vma->vm_start) >> PAGE_SHIFT); |
9d8cebd4b mm: fix mbind vma... |
690 |
prev = vma_merge(mm, prev, vmstart, vmend, vma->vm_flags, |
e26a51148 mm/mempolicy.c: r... |
691 |
vma->anon_vma, vma->vm_file, pgoff, |
247ec28ed mm: add a field t... |
692 |
new_pol, vma_get_anon_name(vma)); |
9d8cebd4b mm: fix mbind vma... |
693 694 695 |
if (prev) { vma = prev; next = vma->vm_next; |
3964acd0d mm: mempolicy: fi... |
696 697 698 699 |
if (mpol_equal(vma_policy(vma), new_pol)) continue; /* vma_merge() joined vma && vma->next, case 8 */ goto replace; |
9d8cebd4b mm: fix mbind vma... |
700 701 702 703 704 705 706 707 708 709 710 |
} if (vma->vm_start != vmstart) { err = split_vma(vma->vm_mm, vma, vmstart, 1); if (err) goto out; } if (vma->vm_end != vmend) { err = split_vma(vma->vm_mm, vma, vmend, 0); if (err) goto out; } |
3964acd0d mm: mempolicy: fi... |
711 |
replace: |
869833f2c mempolicy: remove... |
712 |
err = vma_replace_policy(vma, new_pol); |
8d34694c1 revert "mm: mempo... |
713 714 |
if (err) goto out; |
1da177e4c Linux-2.6.12-rc2 |
715 |
} |
9d8cebd4b mm: fix mbind vma... |
716 717 |
out: |
1da177e4c Linux-2.6.12-rc2 |
718 719 |
return err; } |
1da177e4c Linux-2.6.12-rc2 |
720 |
/* Set the process memory policy */ |
028fec414 mempolicy: suppor... |
721 722 |
static long do_set_mempolicy(unsigned short mode, unsigned short flags, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
723 |
{ |
58568d2a8 cpuset,mm: update... |
724 |
struct mempolicy *new, *old; |
4bfc44958 mm: make set_memp... |
725 |
NODEMASK_SCRATCH(scratch); |
58568d2a8 cpuset,mm: update... |
726 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
727 |
|
4bfc44958 mm: make set_memp... |
728 729 |
if (!scratch) return -ENOMEM; |
f4e53d910 mempolicy: write ... |
730 |
|
4bfc44958 mm: make set_memp... |
731 732 733 734 735 |
new = mpol_new(mode, flags, nodes); if (IS_ERR(new)) { ret = PTR_ERR(new); goto out; } |
2c7c3a7d0 mempolicy: kill d... |
736 |
|
58568d2a8 cpuset,mm: update... |
737 |
task_lock(current); |
4bfc44958 mm: make set_memp... |
738 |
ret = mpol_set_nodemask(new, nodes, scratch); |
58568d2a8 cpuset,mm: update... |
739 740 |
if (ret) { task_unlock(current); |
58568d2a8 cpuset,mm: update... |
741 |
mpol_put(new); |
4bfc44958 mm: make set_memp... |
742 |
goto out; |
58568d2a8 cpuset,mm: update... |
743 744 |
} old = current->mempolicy; |
1da177e4c Linux-2.6.12-rc2 |
745 |
current->mempolicy = new; |
45c4745af mempolicy: rename... |
746 |
if (new && new->mode == MPOL_INTERLEAVE && |
f5b087b52 mempolicy: add MP... |
747 |
nodes_weight(new->v.nodes)) |
dfcd3c0dc [PATCH] Convert m... |
748 |
current->il_next = first_node(new->v.nodes); |
58568d2a8 cpuset,mm: update... |
749 |
task_unlock(current); |
58568d2a8 cpuset,mm: update... |
750 |
mpol_put(old); |
4bfc44958 mm: make set_memp... |
751 752 753 754 |
ret = 0; out: NODEMASK_SCRATCH_FREE(scratch); return ret; |
1da177e4c Linux-2.6.12-rc2 |
755 |
} |
bea904d54 mempolicy: use MP... |
756 757 |
/* * Return nodemask for policy for get_mempolicy() query |
58568d2a8 cpuset,mm: update... |
758 759 |
* * Called with task's alloc_lock held |
bea904d54 mempolicy: use MP... |
760 761 |
*/ static void get_policy_nodemask(struct mempolicy *p, nodemask_t *nodes) |
1da177e4c Linux-2.6.12-rc2 |
762 |
{ |
dfcd3c0dc [PATCH] Convert m... |
763 |
nodes_clear(*nodes); |
bea904d54 mempolicy: use MP... |
764 765 |
if (p == &default_policy) return; |
45c4745af mempolicy: rename... |
766 |
switch (p->mode) { |
19770b326 mm: filter based ... |
767 768 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
769 |
case MPOL_INTERLEAVE: |
dfcd3c0dc [PATCH] Convert m... |
770 |
*nodes = p->v.nodes; |
1da177e4c Linux-2.6.12-rc2 |
771 772 |
break; case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
773 |
if (!(p->flags & MPOL_F_LOCAL)) |
dfcd3c0dc [PATCH] Convert m... |
774 |
node_set(p->v.preferred_node, *nodes); |
53f2556b6 mempolicy: mPOL_P... |
775 |
/* else return empty node mask for local allocation */ |
1da177e4c Linux-2.6.12-rc2 |
776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 |
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 |
794 |
/* Retrieve NUMA policy */ |
dbcb0f19c mm/mempolicy.c: c... |
795 796 |
static long do_get_mempolicy(int *policy, nodemask_t *nmask, unsigned long addr, unsigned long flags) |
1da177e4c Linux-2.6.12-rc2 |
797 |
{ |
8bccd85ff [PATCH] Implement... |
798 |
int err; |
1da177e4c Linux-2.6.12-rc2 |
799 800 801 |
struct mm_struct *mm = current->mm; struct vm_area_struct *vma = NULL; struct mempolicy *pol = current->mempolicy; |
754af6f5a Mem Policy: add M... |
802 803 |
if (flags & ~(unsigned long)(MPOL_F_NODE|MPOL_F_ADDR|MPOL_F_MEMS_ALLOWED)) |
1da177e4c Linux-2.6.12-rc2 |
804 |
return -EINVAL; |
754af6f5a Mem Policy: add M... |
805 806 807 808 809 |
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... |
810 |
task_lock(current); |
754af6f5a Mem Policy: add M... |
811 |
*nmask = cpuset_current_mems_allowed; |
58568d2a8 cpuset,mm: update... |
812 |
task_unlock(current); |
754af6f5a Mem Policy: add M... |
813 814 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
815 |
if (flags & MPOL_F_ADDR) { |
bea904d54 mempolicy: use MP... |
816 817 818 819 820 |
/* * 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 |
821 822 823 824 825 826 827 828 829 830 831 832 833 834 |
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... |
835 |
pol = &default_policy; /* indicates default behavior */ |
1da177e4c Linux-2.6.12-rc2 |
836 837 838 839 840 841 |
if (flags & MPOL_F_NODE) { if (flags & MPOL_F_ADDR) { err = lookup_node(mm, addr); if (err < 0) goto out; |
8bccd85ff [PATCH] Implement... |
842 |
*policy = err; |
1da177e4c Linux-2.6.12-rc2 |
843 |
} else if (pol == current->mempolicy && |
45c4745af mempolicy: rename... |
844 |
pol->mode == MPOL_INTERLEAVE) { |
8bccd85ff [PATCH] Implement... |
845 |
*policy = current->il_next; |
1da177e4c Linux-2.6.12-rc2 |
846 847 848 849 |
} else { err = -EINVAL; goto out; } |
bea904d54 mempolicy: use MP... |
850 851 852 |
} else { *policy = pol == &default_policy ? MPOL_DEFAULT : pol->mode; |
d79df630f mempolicy: mask o... |
853 854 855 856 857 |
/* * Internal mempolicy flags must be masked off before exposing * the policy to userspace. */ *policy |= (pol->flags & MPOL_MODE_FLAGS); |
bea904d54 mempolicy: use MP... |
858 |
} |
1da177e4c Linux-2.6.12-rc2 |
859 860 861 862 863 |
if (vma) { up_read(¤t->mm->mmap_sem); vma = NULL; } |
1da177e4c Linux-2.6.12-rc2 |
864 |
err = 0; |
58568d2a8 cpuset,mm: update... |
865 |
if (nmask) { |
c6b6ef8bb mempolicy: fix ge... |
866 867 868 869 870 871 872 |
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... |
873 |
} |
1da177e4c Linux-2.6.12-rc2 |
874 875 |
out: |
52cd3b074 mempolicy: rework... |
876 |
mpol_cond_put(pol); |
1da177e4c Linux-2.6.12-rc2 |
877 878 879 880 |
if (vma) up_read(¤t->mm->mmap_sem); return err; } |
b20a35035 [PATCH] page migr... |
881 |
#ifdef CONFIG_MIGRATION |
8bccd85ff [PATCH] Implement... |
882 |
/* |
6ce3c4c0f [PATCH] Move page... |
883 884 |
* page migration */ |
fc3012896 [PATCH] Simplify ... |
885 886 |
static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
887 888 |
{ /* |
fc3012896 [PATCH] Simplify ... |
889 |
* Avoid migrating a page that is shared with others. |
6ce3c4c0f [PATCH] Move page... |
890 |
*/ |
62695a84e vmscan: move isol... |
891 892 893 |
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... |
894 895 |
inc_zone_page_state(page, NR_ISOLATED_ANON + page_is_file_cache(page)); |
62695a84e vmscan: move isol... |
896 897 |
} } |
7e2ab150d [PATCH] Direct Mi... |
898 |
} |
6ce3c4c0f [PATCH] Move page... |
899 |
|
742755a1d [PATCH] page migr... |
900 |
static struct page *new_node_page(struct page *page, unsigned long node, int **x) |
95a402c38 [PATCH] page migr... |
901 |
{ |
e2d8cf405 migrate: add huge... |
902 903 904 905 |
if (PageHuge(page)) return alloc_huge_page_node(page_hstate(compound_head(page)), node); else |
b360edb43 mm, mempolicy: mi... |
906 907 |
return alloc_pages_exact_node(node, GFP_HIGHUSER_MOVABLE | __GFP_THISNODE, 0); |
95a402c38 [PATCH] page migr... |
908 |
} |
7e2ab150d [PATCH] Direct Mi... |
909 910 911 912 |
/* * Migrate pages from one node to a target node. * Returns error or the number of pages not migrated. */ |
dbcb0f19c mm/mempolicy.c: c... |
913 914 |
static int migrate_to_node(struct mm_struct *mm, int source, int dest, int flags) |
7e2ab150d [PATCH] Direct Mi... |
915 916 917 918 919 920 921 |
{ nodemask_t nmask; LIST_HEAD(pagelist); int err = 0; nodes_clear(nmask); node_set(source, nmask); |
6ce3c4c0f [PATCH] Move page... |
922 |
|
082708072 mm: revert 0def08... |
923 924 925 926 927 928 |
/* * This does not "check" the range but isolates all pages that * need migration. Between passing in the full user address * space range and MPOL_MF_DISCONTIG_OK, this call can not fail. */ VM_BUG_ON(!(flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL))); |
980949457 mm/mempolicy: ren... |
929 |
queue_pages_range(mm, mm->mmap->vm_start, mm->task_size, &nmask, |
7e2ab150d [PATCH] Direct Mi... |
930 |
flags | MPOL_MF_DISCONTIG_OK, &pagelist); |
cf608ac19 mm: compaction: f... |
931 |
if (!list_empty(&pagelist)) { |
68711a746 mm, migration: ad... |
932 |
err = migrate_pages(&pagelist, new_node_page, NULL, dest, |
9c620e2bc mm: remove offlin... |
933 |
MIGRATE_SYNC, MR_SYSCALL); |
cf608ac19 mm: compaction: f... |
934 |
if (err) |
e2d8cf405 migrate: add huge... |
935 |
putback_movable_pages(&pagelist); |
cf608ac19 mm: compaction: f... |
936 |
} |
95a402c38 [PATCH] page migr... |
937 |
|
7e2ab150d [PATCH] Direct Mi... |
938 |
return err; |
6ce3c4c0f [PATCH] Move page... |
939 940 941 |
} /* |
7e2ab150d [PATCH] Direct Mi... |
942 943 |
* Move pages between the two nodesets so as to preserve the physical * layout as much as possible. |
39743889a [PATCH] Swap Migr... |
944 945 946 |
* * Returns the number of page that could not be moved. */ |
0ce72d4f7 mm: do_migrate_pa... |
947 948 |
int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from, const nodemask_t *to, int flags) |
39743889a [PATCH] Swap Migr... |
949 |
{ |
7e2ab150d [PATCH] Direct Mi... |
950 |
int busy = 0; |
0aedadf91 mm: move migrate_... |
951 |
int err; |
7e2ab150d [PATCH] Direct Mi... |
952 |
nodemask_t tmp; |
39743889a [PATCH] Swap Migr... |
953 |
|
0aedadf91 mm: move migrate_... |
954 955 956 |
err = migrate_prep(); if (err) return err; |
53f2556b6 mempolicy: mPOL_P... |
957 |
down_read(&mm->mmap_sem); |
39743889a [PATCH] Swap Migr... |
958 |
|
da0aa1389 mm/mempolicy.c: f... |
959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 |
/* * Find a 'source' bit set in 'tmp' whose corresponding 'dest' * bit in 'to' is not also set in 'tmp'. Clear the found 'source' * bit in 'tmp', and return that <source, dest> pair for migration. * The pair of nodemasks 'to' and 'from' define the map. * * If no pair of bits is found that way, fallback to picking some * pair of 'source' and 'dest' bits that are not the same. If the * 'source' and 'dest' bits are the same, this represents a node * that will be migrating to itself, so no pages need move. * * If no bits are left in 'tmp', or if all remaining bits left * in 'tmp' correspond to the same bit in 'to', return false * (nothing left to migrate). * * This lets us pick a pair of nodes to migrate between, such that * if possible the dest node is not already occupied by some other * source node, minimizing the risk of overloading the memory on a * node that would happen if we migrated incoming memory to a node * before migrating outgoing memory source that same node. * * A single scan of tmp is sufficient. As we go, we remember the * most recent <s, d> pair that moved (s != d). If we find a pair * that not only moved, but what's better, moved to an empty slot * (d is not set in tmp), then we break out then, with that pair. |
ae0e47f02 Remove one to man... |
984 |
* Otherwise when we finish scanning from_tmp, we at least have the |
da0aa1389 mm/mempolicy.c: f... |
985 986 987 988 |
* 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: ... |
989 |
|
0ce72d4f7 mm: do_migrate_pa... |
990 |
tmp = *from; |
7e2ab150d [PATCH] Direct Mi... |
991 992 |
while (!nodes_empty(tmp)) { int s,d; |
b76ac7e73 mm/mempolicy: use... |
993 |
int source = NUMA_NO_NODE; |
7e2ab150d [PATCH] Direct Mi... |
994 995 996 |
int dest = 0; for_each_node_mask(s, tmp) { |
4a5b18cc1 mm: do_migrate_pa... |
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 |
/* * do_migrate_pages() tries to maintain the relative * node relationship of the pages established between * threads and memory areas. * * However if the number of source nodes is not equal to * the number of destination nodes we can not preserve * this node relative relationship. In that case, skip * copying memory from a node that is in the destination * mask. * * Example: [2,3,4] -> [3,4,5] moves everything. * [0-7] - > [3,4,5] moves only 0,1,2,6,7. */ |
0ce72d4f7 mm: do_migrate_pa... |
1012 1013 |
if ((nodes_weight(*from) != nodes_weight(*to)) && (node_isset(s, *to))) |
4a5b18cc1 mm: do_migrate_pa... |
1014 |
continue; |
0ce72d4f7 mm: do_migrate_pa... |
1015 |
d = node_remap(s, *from, *to); |
7e2ab150d [PATCH] Direct Mi... |
1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 |
if (s == d) continue; source = s; /* Node moved. Memorize */ dest = d; /* dest not in remaining from nodes? */ if (!node_isset(dest, tmp)) break; } |
b76ac7e73 mm/mempolicy: use... |
1026 |
if (source == NUMA_NO_NODE) |
7e2ab150d [PATCH] Direct Mi... |
1027 1028 1029 1030 1031 1032 1033 1034 |
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... |
1035 1036 |
} up_read(&mm->mmap_sem); |
7e2ab150d [PATCH] Direct Mi... |
1037 1038 1039 |
if (err < 0) return err; return busy; |
b20a35035 [PATCH] page migr... |
1040 1041 |
} |
3ad33b243 Migration: find c... |
1042 1043 |
/* * Allocate a new page for page migration based on vma policy. |
d05f0cdcb mm: fix crashes f... |
1044 |
* Start by assuming the page is mapped by the same vma as contains @start. |
3ad33b243 Migration: find c... |
1045 1046 1047 1048 |
* 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. */ |
d05f0cdcb mm: fix crashes f... |
1049 |
static struct page *new_page(struct page *page, unsigned long start, int **x) |
95a402c38 [PATCH] page migr... |
1050 |
{ |
d05f0cdcb mm: fix crashes f... |
1051 |
struct vm_area_struct *vma; |
3ad33b243 Migration: find c... |
1052 |
unsigned long uninitialized_var(address); |
95a402c38 [PATCH] page migr... |
1053 |
|
d05f0cdcb mm: fix crashes f... |
1054 |
vma = find_vma(current->mm, start); |
3ad33b243 Migration: find c... |
1055 1056 1057 1058 1059 1060 |
while (vma) { address = page_address_in_vma(page, vma); if (address != -EFAULT) break; vma = vma->vm_next; } |
11c731e81 mm/mempolicy: fix... |
1061 1062 |
if (PageHuge(page)) { |
cc81717ed mm: new_vma_page(... |
1063 1064 |
BUG_ON(!vma); return alloc_huge_page_noerr(vma, address, 1); |
11c731e81 mm/mempolicy: fix... |
1065 |
} |
0bf598d86 mbind: add BUG_ON... |
1066 |
/* |
11c731e81 mm/mempolicy: fix... |
1067 |
* if !vma, alloc_page_vma() will use task or system default policy |
0bf598d86 mbind: add BUG_ON... |
1068 |
*/ |
3ad33b243 Migration: find c... |
1069 |
return alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); |
95a402c38 [PATCH] page migr... |
1070 |
} |
b20a35035 [PATCH] page migr... |
1071 1072 1073 1074 1075 |
#else static void migrate_page_add(struct page *page, struct list_head *pagelist, unsigned long flags) { |
39743889a [PATCH] Swap Migr... |
1076 |
} |
0ce72d4f7 mm: do_migrate_pa... |
1077 1078 |
int do_migrate_pages(struct mm_struct *mm, const nodemask_t *from, const nodemask_t *to, int flags) |
b20a35035 [PATCH] page migr... |
1079 1080 1081 |
{ return -ENOSYS; } |
95a402c38 [PATCH] page migr... |
1082 |
|
d05f0cdcb mm: fix crashes f... |
1083 |
static struct page *new_page(struct page *page, unsigned long start, int **x) |
95a402c38 [PATCH] page migr... |
1084 1085 1086 |
{ return NULL; } |
b20a35035 [PATCH] page migr... |
1087 |
#endif |
dbcb0f19c mm/mempolicy.c: c... |
1088 |
static long do_mbind(unsigned long start, unsigned long len, |
028fec414 mempolicy: suppor... |
1089 1090 |
unsigned short mode, unsigned short mode_flags, nodemask_t *nmask, unsigned long flags) |
6ce3c4c0f [PATCH] Move page... |
1091 |
{ |
6ce3c4c0f [PATCH] Move page... |
1092 1093 1094 1095 1096 |
struct mm_struct *mm = current->mm; struct mempolicy *new; unsigned long end; int err; LIST_HEAD(pagelist); |
b24f53a0b mm: mempolicy: Ad... |
1097 |
if (flags & ~(unsigned long)MPOL_MF_VALID) |
6ce3c4c0f [PATCH] Move page... |
1098 |
return -EINVAL; |
74c002410 [PATCH] Consisten... |
1099 |
if ((flags & MPOL_MF_MOVE_ALL) && !capable(CAP_SYS_NICE)) |
6ce3c4c0f [PATCH] Move page... |
1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 |
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... |
1115 |
new = mpol_new(mode, mode_flags, nmask); |
6ce3c4c0f [PATCH] Move page... |
1116 1117 |
if (IS_ERR(new)) return PTR_ERR(new); |
b24f53a0b mm: mempolicy: Ad... |
1118 1119 |
if (flags & MPOL_MF_LAZY) new->flags |= MPOL_F_MOF; |
6ce3c4c0f [PATCH] Move page... |
1120 1121 1122 1123 1124 1125 |
/* * 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... |
1126 1127 1128 |
pr_debug("mbind %lx-%lx mode:%d flags:%d nodes:%lx ", start, start + len, mode, mode_flags, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1129 |
nmask ? nodes_addr(*nmask)[0] : NUMA_NO_NODE); |
6ce3c4c0f [PATCH] Move page... |
1130 |
|
0aedadf91 mm: move migrate_... |
1131 1132 1133 1134 |
if (flags & (MPOL_MF_MOVE | MPOL_MF_MOVE_ALL)) { err = migrate_prep(); if (err) |
b05ca7385 do_mbind(): fix m... |
1135 |
goto mpol_out; |
0aedadf91 mm: move migrate_... |
1136 |
} |
4bfc44958 mm: make set_memp... |
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 |
{ 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... |
1150 1151 |
if (err) goto mpol_out; |
d05f0cdcb mm: fix crashes f... |
1152 |
err = queue_pages_range(mm, start, end, nmask, |
6ce3c4c0f [PATCH] Move page... |
1153 |
flags | MPOL_MF_INVERT, &pagelist); |
d05f0cdcb mm: fix crashes f... |
1154 |
if (!err) |
9d8cebd4b mm: fix mbind vma... |
1155 |
err = mbind_range(mm, start, end, new); |
7e2ab150d [PATCH] Direct Mi... |
1156 |
|
b24f53a0b mm: mempolicy: Ad... |
1157 1158 |
if (!err) { int nr_failed = 0; |
cf608ac19 mm: compaction: f... |
1159 |
if (!list_empty(&pagelist)) { |
b24f53a0b mm: mempolicy: Ad... |
1160 |
WARN_ON_ONCE(flags & MPOL_MF_LAZY); |
d05f0cdcb mm: fix crashes f... |
1161 1162 |
nr_failed = migrate_pages(&pagelist, new_page, NULL, start, MIGRATE_SYNC, MR_MEMPOLICY_MBIND); |
cf608ac19 mm: compaction: f... |
1163 |
if (nr_failed) |
74060e4d7 mm: mbind: add hu... |
1164 |
putback_movable_pages(&pagelist); |
cf608ac19 mm: compaction: f... |
1165 |
} |
6ce3c4c0f [PATCH] Move page... |
1166 |
|
b24f53a0b mm: mempolicy: Ad... |
1167 |
if (nr_failed && (flags & MPOL_MF_STRICT)) |
6ce3c4c0f [PATCH] Move page... |
1168 |
err = -EIO; |
ab8a3e14e mbind(): fix leak... |
1169 |
} else |
b0e5fd735 mm/mempolicy: cor... |
1170 |
putback_movable_pages(&pagelist); |
b20a35035 [PATCH] page migr... |
1171 |
|
6ce3c4c0f [PATCH] Move page... |
1172 |
up_write(&mm->mmap_sem); |
b05ca7385 do_mbind(): fix m... |
1173 |
mpol_out: |
f0be3d32b mempolicy: rename... |
1174 |
mpol_put(new); |
6ce3c4c0f [PATCH] Move page... |
1175 1176 |
return err; } |
39743889a [PATCH] Swap Migr... |
1177 |
/* |
8bccd85ff [PATCH] Implement... |
1178 1179 1180 1181 |
* User space interface with variable sized bitmaps for nodelists. */ /* Copy a node mask from user space. */ |
39743889a [PATCH] Swap Migr... |
1182 |
static int get_nodes(nodemask_t *nodes, const unsigned long __user *nmask, |
8bccd85ff [PATCH] Implement... |
1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 |
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... |
1193 |
if (maxnode > PAGE_SIZE*BITS_PER_BYTE) |
636f13c17 [PATCH] sys_mbind... |
1194 |
return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 |
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... |
1243 |
SYSCALL_DEFINE6(mbind, unsigned long, start, unsigned long, len, |
f7f28ca98 mm: constify nmas... |
1244 |
unsigned long, mode, const unsigned long __user *, nmask, |
938bb9f5e [CVE-2009-0029] S... |
1245 |
unsigned long, maxnode, unsigned, flags) |
8bccd85ff [PATCH] Implement... |
1246 1247 1248 |
{ nodemask_t nodes; int err; |
028fec414 mempolicy: suppor... |
1249 |
unsigned short mode_flags; |
8bccd85ff [PATCH] Implement... |
1250 |
|
028fec414 mempolicy: suppor... |
1251 1252 |
mode_flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; |
a3b51e014 mempolicy: conver... |
1253 1254 |
if (mode >= MPOL_MAX) return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1255 1256 1257 |
if ((mode_flags & MPOL_F_STATIC_NODES) && (mode_flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1258 1259 1260 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1261 |
return do_mbind(start, len, mode, mode_flags, &nodes, flags); |
8bccd85ff [PATCH] Implement... |
1262 1263 1264 |
} /* Set the process memory policy */ |
23c8902d4 mm: constify nmas... |
1265 |
SYSCALL_DEFINE3(set_mempolicy, int, mode, const unsigned long __user *, nmask, |
938bb9f5e [CVE-2009-0029] S... |
1266 |
unsigned long, maxnode) |
8bccd85ff [PATCH] Implement... |
1267 1268 1269 |
{ int err; nodemask_t nodes; |
028fec414 mempolicy: suppor... |
1270 |
unsigned short flags; |
8bccd85ff [PATCH] Implement... |
1271 |
|
028fec414 mempolicy: suppor... |
1272 1273 1274 |
flags = mode & MPOL_MODE_FLAGS; mode &= ~MPOL_MODE_FLAGS; if ((unsigned int)mode >= MPOL_MAX) |
8bccd85ff [PATCH] Implement... |
1275 |
return -EINVAL; |
4c50bc011 mempolicy: add MP... |
1276 1277 |
if ((flags & MPOL_F_STATIC_NODES) && (flags & MPOL_F_RELATIVE_NODES)) return -EINVAL; |
8bccd85ff [PATCH] Implement... |
1278 1279 1280 |
err = get_nodes(&nodes, nmask, maxnode); if (err) return err; |
028fec414 mempolicy: suppor... |
1281 |
return do_set_mempolicy(mode, flags, &nodes); |
8bccd85ff [PATCH] Implement... |
1282 |
} |
938bb9f5e [CVE-2009-0029] S... |
1283 1284 1285 |
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... |
1286 |
{ |
c69e8d9c0 CRED: Use RCU to ... |
1287 |
const struct cred *cred = current_cred(), *tcred; |
596d7cfa2 mempolicy: reduce... |
1288 |
struct mm_struct *mm = NULL; |
39743889a [PATCH] Swap Migr... |
1289 |
struct task_struct *task; |
39743889a [PATCH] Swap Migr... |
1290 1291 |
nodemask_t task_nodes; int err; |
596d7cfa2 mempolicy: reduce... |
1292 1293 1294 1295 1296 1297 |
nodemask_t *old; nodemask_t *new; NODEMASK_SCRATCH(scratch); if (!scratch) return -ENOMEM; |
39743889a [PATCH] Swap Migr... |
1298 |
|
596d7cfa2 mempolicy: reduce... |
1299 1300 1301 1302 |
old = &scratch->mask1; new = &scratch->mask2; err = get_nodes(old, old_nodes, maxnode); |
39743889a [PATCH] Swap Migr... |
1303 |
if (err) |
596d7cfa2 mempolicy: reduce... |
1304 |
goto out; |
39743889a [PATCH] Swap Migr... |
1305 |
|
596d7cfa2 mempolicy: reduce... |
1306 |
err = get_nodes(new, new_nodes, maxnode); |
39743889a [PATCH] Swap Migr... |
1307 |
if (err) |
596d7cfa2 mempolicy: reduce... |
1308 |
goto out; |
39743889a [PATCH] Swap Migr... |
1309 1310 |
/* Find the mm_struct */ |
55cfaa3cb mm/mempolicy.c: a... |
1311 |
rcu_read_lock(); |
228ebcbe6 Uninline find_tas... |
1312 |
task = pid ? find_task_by_vpid(pid) : current; |
39743889a [PATCH] Swap Migr... |
1313 |
if (!task) { |
55cfaa3cb mm/mempolicy.c: a... |
1314 |
rcu_read_unlock(); |
596d7cfa2 mempolicy: reduce... |
1315 1316 |
err = -ESRCH; goto out; |
39743889a [PATCH] Swap Migr... |
1317 |
} |
3268c63ed mm: fix move/migr... |
1318 |
get_task_struct(task); |
39743889a [PATCH] Swap Migr... |
1319 |
|
596d7cfa2 mempolicy: reduce... |
1320 |
err = -EINVAL; |
39743889a [PATCH] Swap Migr... |
1321 1322 1323 1324 |
/* * Check if this process has the right to modify the specified * process. The right exists if the process has administrative |
7f927fcc2 [PATCH] Typo fixes |
1325 |
* capabilities, superuser privileges or the same |
39743889a [PATCH] Swap Migr... |
1326 1327 |
* userid as the target process. */ |
c69e8d9c0 CRED: Use RCU to ... |
1328 |
tcred = __task_cred(task); |
b38a86eb1 userns: Convert t... |
1329 1330 |
if (!uid_eq(cred->euid, tcred->suid) && !uid_eq(cred->euid, tcred->uid) && !uid_eq(cred->uid, tcred->suid) && !uid_eq(cred->uid, tcred->uid) && |
74c002410 [PATCH] Consisten... |
1331 |
!capable(CAP_SYS_NICE)) { |
c69e8d9c0 CRED: Use RCU to ... |
1332 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1333 |
err = -EPERM; |
3268c63ed mm: fix move/migr... |
1334 |
goto out_put; |
39743889a [PATCH] Swap Migr... |
1335 |
} |
c69e8d9c0 CRED: Use RCU to ... |
1336 |
rcu_read_unlock(); |
39743889a [PATCH] Swap Migr... |
1337 1338 1339 |
task_nodes = cpuset_mems_allowed(task); /* Is the user allowed to access the target nodes? */ |
596d7cfa2 mempolicy: reduce... |
1340 |
if (!nodes_subset(*new, task_nodes) && !capable(CAP_SYS_NICE)) { |
39743889a [PATCH] Swap Migr... |
1341 |
err = -EPERM; |
3268c63ed mm: fix move/migr... |
1342 |
goto out_put; |
39743889a [PATCH] Swap Migr... |
1343 |
} |
01f13bd60 mempolicy: use N_... |
1344 |
if (!nodes_subset(*new, node_states[N_MEMORY])) { |
3b42d28b2 Page migration: D... |
1345 |
err = -EINVAL; |
3268c63ed mm: fix move/migr... |
1346 |
goto out_put; |
3b42d28b2 Page migration: D... |
1347 |
} |
86c3a7645 [PATCH] SELinux: ... |
1348 1349 |
err = security_task_movememory(task); if (err) |
3268c63ed mm: fix move/migr... |
1350 |
goto out_put; |
86c3a7645 [PATCH] SELinux: ... |
1351 |
|
3268c63ed mm: fix move/migr... |
1352 1353 |
mm = get_task_mm(task); put_task_struct(task); |
f2a9ef880 mm: fix NULL ptr ... |
1354 1355 |
if (!mm) { |
3268c63ed mm: fix move/migr... |
1356 |
err = -EINVAL; |
f2a9ef880 mm: fix NULL ptr ... |
1357 1358 1359 1360 1361 |
goto out; } err = do_migrate_pages(mm, old, new, capable(CAP_SYS_NICE) ? MPOL_MF_MOVE_ALL : MPOL_MF_MOVE); |
3268c63ed mm: fix move/migr... |
1362 1363 1364 |
mmput(mm); out: |
596d7cfa2 mempolicy: reduce... |
1365 |
NODEMASK_SCRATCH_FREE(scratch); |
39743889a [PATCH] Swap Migr... |
1366 |
return err; |
3268c63ed mm: fix move/migr... |
1367 1368 1369 1370 |
out_put: put_task_struct(task); goto out; |
39743889a [PATCH] Swap Migr... |
1371 |
} |
8bccd85ff [PATCH] Implement... |
1372 |
/* Retrieve NUMA policy */ |
938bb9f5e [CVE-2009-0029] S... |
1373 1374 1375 |
SYSCALL_DEFINE5(get_mempolicy, int __user *, policy, unsigned long __user *, nmask, unsigned long, maxnode, unsigned long, addr, unsigned long, flags) |
8bccd85ff [PATCH] Implement... |
1376 |
{ |
dbcb0f19c mm/mempolicy.c: c... |
1377 1378 |
int err; int uninitialized_var(pval); |
8bccd85ff [PATCH] Implement... |
1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 |
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 |
1397 |
#ifdef CONFIG_COMPAT |
c93e0f6c8 mm/compat: conver... |
1398 1399 1400 1401 |
COMPAT_SYSCALL_DEFINE5(get_mempolicy, int __user *, policy, compat_ulong_t __user *, nmask, compat_ulong_t, maxnode, compat_ulong_t, addr, compat_ulong_t, flags) |
1da177e4c Linux-2.6.12-rc2 |
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 |
{ long err; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; DECLARE_BITMAP(bm, MAX_NUMNODES); nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) nm = compat_alloc_user_space(alloc_size); err = sys_get_mempolicy(policy, nm, nr_bits+1, addr, flags); if (!err && nmask) { |
2bbff6c76 mm/mempolicy.c: m... |
1417 1418 1419 |
unsigned long copy_size; copy_size = min_t(unsigned long, sizeof(bm), alloc_size); err = copy_from_user(bm, nm, copy_size); |
1da177e4c Linux-2.6.12-rc2 |
1420 1421 1422 1423 1424 1425 1426 |
/* ensure entire bitmap is zeroed */ err |= clear_user(nmask, ALIGN(maxnode-1, 8) / 8); err |= compat_put_bitmap(nmask, bm, nr_bits); } return err; } |
c93e0f6c8 mm/compat: conver... |
1427 1428 |
COMPAT_SYSCALL_DEFINE3(set_mempolicy, int, mode, compat_ulong_t __user *, nmask, compat_ulong_t, maxnode) |
1da177e4c Linux-2.6.12-rc2 |
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 |
{ long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; DECLARE_BITMAP(bm, MAX_NUMNODES); nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) { err = compat_get_bitmap(bm, nmask, nr_bits); nm = compat_alloc_user_space(alloc_size); err |= copy_to_user(nm, bm, alloc_size); } if (err) return -EFAULT; return sys_set_mempolicy(mode, nm, nr_bits+1); } |
c93e0f6c8 mm/compat: conver... |
1449 1450 1451 |
COMPAT_SYSCALL_DEFINE6(mbind, compat_ulong_t, start, compat_ulong_t, len, compat_ulong_t, mode, compat_ulong_t __user *, nmask, compat_ulong_t, maxnode, compat_ulong_t, flags) |
1da177e4c Linux-2.6.12-rc2 |
1452 1453 1454 1455 |
{ long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; |
dfcd3c0dc [PATCH] Convert m... |
1456 |
nodemask_t bm; |
1da177e4c Linux-2.6.12-rc2 |
1457 1458 1459 1460 1461 |
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... |
1462 |
err = compat_get_bitmap(nodes_addr(bm), nmask, nr_bits); |
1da177e4c Linux-2.6.12-rc2 |
1463 |
nm = compat_alloc_user_space(alloc_size); |
dfcd3c0dc [PATCH] Convert m... |
1464 |
err |= copy_to_user(nm, nodes_addr(bm), alloc_size); |
1da177e4c Linux-2.6.12-rc2 |
1465 1466 1467 1468 1469 1470 1471 1472 1473 |
} if (err) return -EFAULT; return sys_mbind(start, len, mode, nm, nr_bits+1, flags); } #endif |
74d2c3a05 mempolicy: introd... |
1474 1475 |
struct mempolicy *__get_vma_policy(struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
1476 |
{ |
8d90274b3 mempolicy: saniti... |
1477 |
struct mempolicy *pol = NULL; |
1da177e4c Linux-2.6.12-rc2 |
1478 1479 |
if (vma) { |
480eccf9a Fix NUMA Memory P... |
1480 |
if (vma->vm_ops && vma->vm_ops->get_policy) { |
8d90274b3 mempolicy: saniti... |
1481 |
pol = vma->vm_ops->get_policy(vma, addr); |
00442ad04 mempolicy: fix a ... |
1482 |
} else if (vma->vm_policy) { |
1da177e4c Linux-2.6.12-rc2 |
1483 |
pol = vma->vm_policy; |
00442ad04 mempolicy: fix a ... |
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 |
/* * shmem_alloc_page() passes MPOL_F_SHARED policy with * a pseudo vma whose vma->vm_ops=NULL. Take a reference * count on these policies which will be dropped by * mpol_cond_put() later */ if (mpol_needs_cond_ref(pol)) mpol_get(pol); } |
1da177e4c Linux-2.6.12-rc2 |
1494 |
} |
f15ca78e3 mempolicy: change... |
1495 |
|
74d2c3a05 mempolicy: introd... |
1496 1497 1498 1499 |
return pol; } /* |
dd6eecb91 mempolicy: unexpo... |
1500 |
* get_vma_policy(@vma, @addr) |
74d2c3a05 mempolicy: introd... |
1501 1502 1503 1504 |
* @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. |
dd6eecb91 mempolicy: unexpo... |
1505 |
* Falls back to current->mempolicy or system default policy, as necessary. |
74d2c3a05 mempolicy: introd... |
1506 1507 1508 1509 1510 |
* 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. */ |
dd6eecb91 mempolicy: unexpo... |
1511 1512 |
static struct mempolicy *get_vma_policy(struct vm_area_struct *vma, unsigned long addr) |
74d2c3a05 mempolicy: introd... |
1513 1514 |
{ struct mempolicy *pol = __get_vma_policy(vma, addr); |
8d90274b3 mempolicy: saniti... |
1515 |
if (!pol) |
dd6eecb91 mempolicy: unexpo... |
1516 |
pol = get_task_policy(current); |
8d90274b3 mempolicy: saniti... |
1517 |
|
1da177e4c Linux-2.6.12-rc2 |
1518 1519 |
return pol; } |
6b6482bbf mempolicy: remove... |
1520 |
bool vma_policy_mof(struct vm_area_struct *vma) |
fc3147245 mm: numa: Limit N... |
1521 |
{ |
6b6482bbf mempolicy: remove... |
1522 |
struct mempolicy *pol; |
fc3147245 mm: numa: Limit N... |
1523 |
|
6b6482bbf mempolicy: remove... |
1524 1525 |
if (vma->vm_ops && vma->vm_ops->get_policy) { bool ret = false; |
fc3147245 mm: numa: Limit N... |
1526 |
|
6b6482bbf mempolicy: remove... |
1527 1528 1529 1530 |
pol = vma->vm_ops->get_policy(vma, vma->vm_start); if (pol && (pol->flags & MPOL_F_MOF)) ret = true; mpol_cond_put(pol); |
8d90274b3 mempolicy: saniti... |
1531 |
|
6b6482bbf mempolicy: remove... |
1532 |
return ret; |
fc3147245 mm: numa: Limit N... |
1533 |
} |
6b6482bbf mempolicy: remove... |
1534 |
pol = vma->vm_policy; |
8d90274b3 mempolicy: saniti... |
1535 |
if (!pol) |
6b6482bbf mempolicy: remove... |
1536 |
pol = get_task_policy(current); |
8d90274b3 mempolicy: saniti... |
1537 |
|
fc3147245 mm: numa: Limit N... |
1538 1539 |
return pol->flags & MPOL_F_MOF; } |
d3eb1570a mempolicy: fix is... |
1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 |
static int apply_policy_zone(struct mempolicy *policy, enum zone_type zone) { enum zone_type dynamic_policy_zone = policy_zone; BUG_ON(dynamic_policy_zone == ZONE_MOVABLE); /* * if policy->v.nodes has movable memory only, * we apply policy when gfp_zone(gfp) = ZONE_MOVABLE only. * * policy->v.nodes is intersect with node_states[N_MEMORY]. * so if the following test faile, it implies * policy->v.nodes has movable memory only. */ if (!nodes_intersects(policy->v.nodes, node_states[N_HIGH_MEMORY])) dynamic_policy_zone = ZONE_MOVABLE; return zone >= dynamic_policy_zone; } |
52cd3b074 mempolicy: rework... |
1559 1560 1561 1562 1563 |
/* * 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 ... |
1564 1565 |
{ /* Lower zones don't get a nodemask applied for MPOL_BIND */ |
45c4745af mempolicy: rename... |
1566 |
if (unlikely(policy->mode == MPOL_BIND) && |
d3eb1570a mempolicy: fix is... |
1567 |
apply_policy_zone(policy, gfp_zone(gfp)) && |
19770b326 mm: filter based ... |
1568 1569 1570 1571 1572 |
cpuset_nodemask_valid_mems_allowed(&policy->v.nodes)) return &policy->v.nodes; return NULL; } |
52cd3b074 mempolicy: rework... |
1573 |
/* Return a zonelist indicated by gfp for node representing a mempolicy */ |
2f5f9486f mm: change alloc_... |
1574 1575 |
static struct zonelist *policy_zonelist(gfp_t gfp, struct mempolicy *policy, int nd) |
1da177e4c Linux-2.6.12-rc2 |
1576 |
{ |
45c4745af mempolicy: rename... |
1577 |
switch (policy->mode) { |
1da177e4c Linux-2.6.12-rc2 |
1578 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1579 1580 |
if (!(policy->flags & MPOL_F_LOCAL)) nd = policy->v.preferred_node; |
1da177e4c Linux-2.6.12-rc2 |
1581 1582 |
break; case MPOL_BIND: |
19770b326 mm: filter based ... |
1583 |
/* |
52cd3b074 mempolicy: rework... |
1584 1585 |
* Normally, MPOL_BIND allocations are node-local within the * allowed nodemask. However, if __GFP_THISNODE is set and the |
6eb27e1fd mempolicy: remove... |
1586 |
* current node isn't part of the mask, we use the zonelist for |
52cd3b074 mempolicy: rework... |
1587 |
* the first node in the mask instead. |
19770b326 mm: filter based ... |
1588 |
*/ |
19770b326 mm: filter based ... |
1589 1590 1591 1592 |
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 |
1593 |
default: |
1da177e4c Linux-2.6.12-rc2 |
1594 1595 |
BUG(); } |
0e88460da mm: introduce nod... |
1596 |
return node_zonelist(nd, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1597 1598 1599 1600 1601 1602 1603 1604 1605 |
} /* 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... |
1606 |
next = next_node(nid, policy->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1607 |
if (next >= MAX_NUMNODES) |
dfcd3c0dc [PATCH] Convert m... |
1608 |
next = first_node(policy->v.nodes); |
f5b087b52 mempolicy: add MP... |
1609 1610 |
if (next < MAX_NUMNODES) me->il_next = next; |
1da177e4c Linux-2.6.12-rc2 |
1611 1612 |
return nid; } |
dc85da15d [PATCH] NUMA poli... |
1613 1614 1615 1616 |
/* * Depending on the memory policy provide a node from which to allocate the * next slab entry. */ |
2a389610a mm, mempolicy: re... |
1617 |
unsigned int mempolicy_slab_node(void) |
dc85da15d [PATCH] NUMA poli... |
1618 |
{ |
e7b691b08 slab/mempolicy: a... |
1619 |
struct mempolicy *policy; |
2a389610a mm, mempolicy: re... |
1620 |
int node = numa_mem_id(); |
e7b691b08 slab/mempolicy: a... |
1621 1622 |
if (in_interrupt()) |
2a389610a mm, mempolicy: re... |
1623 |
return node; |
e7b691b08 slab/mempolicy: a... |
1624 1625 |
policy = current->mempolicy; |
fc36b8d3d mempolicy: use MP... |
1626 |
if (!policy || policy->flags & MPOL_F_LOCAL) |
2a389610a mm, mempolicy: re... |
1627 |
return node; |
bea904d54 mempolicy: use MP... |
1628 1629 1630 |
switch (policy->mode) { case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
1631 1632 1633 1634 |
/* * handled MPOL_F_LOCAL above */ return policy->v.preferred_node; |
765c4507a [PATCH] GFP_THISN... |
1635 |
|
dc85da15d [PATCH] NUMA poli... |
1636 1637 |
case MPOL_INTERLEAVE: return interleave_nodes(policy); |
dd1a239f6 mm: have zonelist... |
1638 |
case MPOL_BIND: { |
dc85da15d [PATCH] NUMA poli... |
1639 1640 1641 1642 |
/* * Follow bind policy behavior and start allocation at the * first node. */ |
19770b326 mm: filter based ... |
1643 1644 1645 |
struct zonelist *zonelist; struct zone *zone; enum zone_type highest_zoneidx = gfp_zone(GFP_KERNEL); |
2a389610a mm, mempolicy: re... |
1646 |
zonelist = &NODE_DATA(node)->node_zonelists[0]; |
19770b326 mm: filter based ... |
1647 1648 1649 |
(void)first_zones_zonelist(zonelist, highest_zoneidx, &policy->v.nodes, &zone); |
2a389610a mm, mempolicy: re... |
1650 |
return zone ? zone->node : node; |
dd1a239f6 mm: have zonelist... |
1651 |
} |
dc85da15d [PATCH] NUMA poli... |
1652 |
|
dc85da15d [PATCH] NUMA poli... |
1653 |
default: |
bea904d54 mempolicy: use MP... |
1654 |
BUG(); |
dc85da15d [PATCH] NUMA poli... |
1655 1656 |
} } |
1da177e4c Linux-2.6.12-rc2 |
1657 1658 1659 1660 |
/* 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... |
1661 |
unsigned nnodes = nodes_weight(pol->v.nodes); |
f5b087b52 mempolicy: add MP... |
1662 |
unsigned target; |
1da177e4c Linux-2.6.12-rc2 |
1663 |
int c; |
b76ac7e73 mm/mempolicy: use... |
1664 |
int nid = NUMA_NO_NODE; |
1da177e4c Linux-2.6.12-rc2 |
1665 |
|
f5b087b52 mempolicy: add MP... |
1666 1667 1668 |
if (!nnodes) return numa_node_id(); target = (unsigned int)off % nnodes; |
1da177e4c Linux-2.6.12-rc2 |
1669 1670 |
c = 0; do { |
dfcd3c0dc [PATCH] Convert m... |
1671 |
nid = next_node(nid, pol->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
1672 1673 |
c++; } while (c <= target); |
1da177e4c Linux-2.6.12-rc2 |
1674 1675 |
return nid; } |
5da7ca860 [PATCH] Add NUMA ... |
1676 1677 1678 1679 1680 1681 |
/* 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 ... |
1682 1683 1684 1685 1686 1687 1688 1689 1690 |
/* * 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 ... |
1691 1692 1693 1694 1695 |
off += (addr - vma->vm_start) >> shift; return offset_il_node(pol, vma, off); } else return interleave_nodes(pol); } |
778d3b0ff cpusets: randomiz... |
1696 1697 |
/* * Return the bit number of a random bit set in the nodemask. |
b76ac7e73 mm/mempolicy: use... |
1698 |
* (returns NUMA_NO_NODE if nodemask is empty) |
778d3b0ff cpusets: randomiz... |
1699 1700 1701 |
*/ int node_random(const nodemask_t *maskp) { |
b76ac7e73 mm/mempolicy: use... |
1702 |
int w, bit = NUMA_NO_NODE; |
778d3b0ff cpusets: randomiz... |
1703 1704 1705 1706 1707 1708 1709 |
w = nodes_weight(*maskp); if (w) bit = bitmap_ord_to_pos(maskp->bits, get_random_int() % w, MAX_NUMNODES); return bit; } |
00ac59adf [PATCH] x86_64: F... |
1710 |
#ifdef CONFIG_HUGETLBFS |
480eccf9a Fix NUMA Memory P... |
1711 1712 |
/* * huge_zonelist(@vma, @addr, @gfp_flags, @mpol) |
b46e14acb mm/mempolicy.c: p... |
1713 1714 1715 1716 1717 |
* @vma: virtual memory area whose policy is sought * @addr: address in @vma for shared policy lookup and interleave policy * @gfp_flags: for requested zone * @mpol: pointer to mempolicy pointer for reference counted mempolicy * @nodemask: pointer to nodemask pointer for MPOL_BIND nodemask |
480eccf9a Fix NUMA Memory P... |
1718 |
* |
52cd3b074 mempolicy: rework... |
1719 1720 1721 1722 |
* 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... |
1723 |
* |
d26914d11 mm: optimize put_... |
1724 |
* Must be protected by read_mems_allowed_begin() |
480eccf9a Fix NUMA Memory P... |
1725 |
*/ |
396faf030 Allow huge page a... |
1726 |
struct zonelist *huge_zonelist(struct vm_area_struct *vma, unsigned long addr, |
19770b326 mm: filter based ... |
1727 1728 |
gfp_t gfp_flags, struct mempolicy **mpol, nodemask_t **nodemask) |
5da7ca860 [PATCH] Add NUMA ... |
1729 |
{ |
480eccf9a Fix NUMA Memory P... |
1730 |
struct zonelist *zl; |
5da7ca860 [PATCH] Add NUMA ... |
1731 |
|
dd6eecb91 mempolicy: unexpo... |
1732 |
*mpol = get_vma_policy(vma, addr); |
19770b326 mm: filter based ... |
1733 |
*nodemask = NULL; /* assume !MPOL_BIND */ |
5da7ca860 [PATCH] Add NUMA ... |
1734 |
|
52cd3b074 mempolicy: rework... |
1735 1736 |
if (unlikely((*mpol)->mode == MPOL_INTERLEAVE)) { zl = node_zonelist(interleave_nid(*mpol, vma, addr, |
a55164389 hugetlb: modular ... |
1737 |
huge_page_shift(hstate_vma(vma))), gfp_flags); |
52cd3b074 mempolicy: rework... |
1738 |
} else { |
2f5f9486f mm: change alloc_... |
1739 |
zl = policy_zonelist(gfp_flags, *mpol, numa_node_id()); |
52cd3b074 mempolicy: rework... |
1740 1741 |
if ((*mpol)->mode == MPOL_BIND) *nodemask = &(*mpol)->v.nodes; |
480eccf9a Fix NUMA Memory P... |
1742 1743 |
} return zl; |
5da7ca860 [PATCH] Add NUMA ... |
1744 |
} |
06808b082 hugetlb: derive h... |
1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 |
/* * 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... |
1769 |
task_lock(current); |
06808b082 hugetlb: derive h... |
1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 |
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... |
1789 |
task_unlock(current); |
06808b082 hugetlb: derive h... |
1790 1791 1792 |
return true; } |
00ac59adf [PATCH] x86_64: F... |
1793 |
#endif |
5da7ca860 [PATCH] Add NUMA ... |
1794 |
|
6f48d0ebd oom: select task ... |
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 |
/* * 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 |
1838 1839 |
/* Allocate a page in interleaved policy. Own path because it needs to do special accounting. */ |
662f3a0b9 [PATCH] Remove ne... |
1840 1841 |
static struct page *alloc_page_interleave(gfp_t gfp, unsigned order, unsigned nid) |
1da177e4c Linux-2.6.12-rc2 |
1842 1843 1844 |
{ struct zonelist *zl; struct page *page; |
0e88460da mm: introduce nod... |
1845 |
zl = node_zonelist(nid, gfp); |
1da177e4c Linux-2.6.12-rc2 |
1846 |
page = __alloc_pages(gfp, order, zl); |
dd1a239f6 mm: have zonelist... |
1847 |
if (page && page_zone(page) == zonelist_zone(&zl->_zonerefs[0])) |
ca889e6c4 [PATCH] Use Zoned... |
1848 |
inc_zone_page_state(page, NUMA_INTERLEAVE_HIT); |
1da177e4c Linux-2.6.12-rc2 |
1849 1850 1851 1852 |
return page; } /** |
0bbbc0b33 thp: add numa awa... |
1853 |
* alloc_pages_vma - Allocate a page for a VMA. |
1da177e4c Linux-2.6.12-rc2 |
1854 1855 1856 1857 1858 1859 1860 1861 |
* * @gfp: * %GFP_USER user allocation. * %GFP_KERNEL kernel allocations, * %GFP_HIGHMEM highmem/user allocations, * %GFP_FS allocation should not call back into a file system. * %GFP_ATOMIC don't sleep. * |
0bbbc0b33 thp: add numa awa... |
1862 |
* @order:Order of the GFP allocation. |
1da177e4c Linux-2.6.12-rc2 |
1863 1864 |
* @vma: Pointer to VMA or NULL if not available. * @addr: Virtual Address of the allocation. Must be inside the VMA. |
be97a41b2 mm/mempolicy.c: m... |
1865 1866 |
* @node: Which node to prefer for allocation (modulo policy). * @hugepage: for hugepages try only the preferred node if possible |
1da177e4c Linux-2.6.12-rc2 |
1867 1868 1869 1870 1871 |
* * 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 |
be97a41b2 mm/mempolicy.c: m... |
1872 1873 |
* all allocations for pages that will be mapped into user space. Returns * NULL when no page can be allocated. |
1da177e4c Linux-2.6.12-rc2 |
1874 1875 |
*/ struct page * |
0bbbc0b33 thp: add numa awa... |
1876 |
alloc_pages_vma(gfp_t gfp, int order, struct vm_area_struct *vma, |
be97a41b2 mm/mempolicy.c: m... |
1877 |
unsigned long addr, int node, bool hugepage) |
1da177e4c Linux-2.6.12-rc2 |
1878 |
{ |
cc9a6c877 cpuset: mm: reduc... |
1879 |
struct mempolicy *pol; |
c0ff7453b cpuset,mm: fix no... |
1880 |
struct page *page; |
cc9a6c877 cpuset: mm: reduc... |
1881 |
unsigned int cpuset_mems_cookie; |
be97a41b2 mm/mempolicy.c: m... |
1882 1883 |
struct zonelist *zl; nodemask_t *nmask; |
cc9a6c877 cpuset: mm: reduc... |
1884 1885 |
retry_cpuset: |
dd6eecb91 mempolicy: unexpo... |
1886 |
pol = get_vma_policy(vma, addr); |
d26914d11 mm: optimize put_... |
1887 |
cpuset_mems_cookie = read_mems_allowed_begin(); |
1da177e4c Linux-2.6.12-rc2 |
1888 |
|
1021c9720 mm, thp: respect ... |
1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 |
if (pol->mode == MPOL_INTERLEAVE) { unsigned nid; nid = interleave_nid(pol, vma, addr, PAGE_SHIFT + order); mpol_cond_put(pol); page = alloc_page_interleave(gfp, order, nid); goto out; } if (unlikely(IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE) && hugepage)) { int hpage_node = node; |
be97a41b2 mm/mempolicy.c: m... |
1900 1901 |
/* * For hugepage allocation and non-interleave policy which |
1021c9720 mm, thp: respect ... |
1902 1903 1904 1905 |
* allows the current node (or other explicitly preferred * node) we only try to allocate from the current/preferred * node and don't fall back to other nodes, as the cost of * remote accesses would likely offset THP benefits. |
be97a41b2 mm/mempolicy.c: m... |
1906 1907 1908 1909 |
* * If the policy is interleave, or does not allow the current * node in its nodemask, we allocate the standard way. */ |
1021c9720 mm, thp: respect ... |
1910 1911 1912 |
if (pol->mode == MPOL_PREFERRED && !(pol->flags & MPOL_F_LOCAL)) hpage_node = pol->v.preferred_node; |
be97a41b2 mm/mempolicy.c: m... |
1913 |
nmask = policy_nodemask(gfp, pol); |
1021c9720 mm, thp: respect ... |
1914 |
if (!nmask || node_isset(hpage_node, *nmask)) { |
be97a41b2 mm/mempolicy.c: m... |
1915 |
mpol_cond_put(pol); |
1021c9720 mm, thp: respect ... |
1916 |
page = alloc_pages_exact_node(hpage_node, |
5265047ac mm, thp: really l... |
1917 |
gfp | __GFP_THISNODE, order); |
be97a41b2 mm/mempolicy.c: m... |
1918 1919 1920 |
goto out; } } |
be97a41b2 mm/mempolicy.c: m... |
1921 1922 |
nmask = policy_nodemask(gfp, pol); zl = policy_zonelist(gfp, pol, node); |
2386740d1 mempolicy: change... |
1923 |
mpol_cond_put(pol); |
be97a41b2 mm/mempolicy.c: m... |
1924 1925 |
page = __alloc_pages_nodemask(gfp, order, zl, nmask); out: |
d26914d11 mm: optimize put_... |
1926 |
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) |
cc9a6c877 cpuset: mm: reduc... |
1927 |
goto retry_cpuset; |
c0ff7453b cpuset,mm: fix no... |
1928 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 |
} /** * 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... |
1946 |
* Don't call cpuset_update_task_memory_state() unless |
1da177e4c Linux-2.6.12-rc2 |
1947 1948 1949 |
* 1) it's ok to take cpuset_sem (can WAIT), and * 2) allocating for current task (not interrupt). */ |
dd0fc66fb [PATCH] gfp flags... |
1950 |
struct page *alloc_pages_current(gfp_t gfp, unsigned order) |
1da177e4c Linux-2.6.12-rc2 |
1951 |
{ |
8d90274b3 mempolicy: saniti... |
1952 |
struct mempolicy *pol = &default_policy; |
c0ff7453b cpuset,mm: fix no... |
1953 |
struct page *page; |
cc9a6c877 cpuset: mm: reduc... |
1954 |
unsigned int cpuset_mems_cookie; |
1da177e4c Linux-2.6.12-rc2 |
1955 |
|
8d90274b3 mempolicy: saniti... |
1956 1957 |
if (!in_interrupt() && !(gfp & __GFP_THISNODE)) pol = get_task_policy(current); |
52cd3b074 mempolicy: rework... |
1958 |
|
cc9a6c877 cpuset: mm: reduc... |
1959 |
retry_cpuset: |
d26914d11 mm: optimize put_... |
1960 |
cpuset_mems_cookie = read_mems_allowed_begin(); |
cc9a6c877 cpuset: mm: reduc... |
1961 |
|
52cd3b074 mempolicy: rework... |
1962 1963 1964 1965 |
/* * No reference counting needed for current->mempolicy * nor system default_policy */ |
45c4745af mempolicy: rename... |
1966 |
if (pol->mode == MPOL_INTERLEAVE) |
c0ff7453b cpuset,mm: fix no... |
1967 1968 1969 |
page = alloc_page_interleave(gfp, order, interleave_nodes(pol)); else page = __alloc_pages_nodemask(gfp, order, |
5c4b4be3b mm: use correct n... |
1970 1971 |
policy_zonelist(gfp, pol, numa_node_id()), policy_nodemask(gfp, pol)); |
cc9a6c877 cpuset: mm: reduc... |
1972 |
|
d26914d11 mm: optimize put_... |
1973 |
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) |
cc9a6c877 cpuset: mm: reduc... |
1974 |
goto retry_cpuset; |
c0ff7453b cpuset,mm: fix no... |
1975 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1976 1977 |
} EXPORT_SYMBOL(alloc_pages_current); |
ef0855d33 mm: mempolicy: tu... |
1978 1979 1980 1981 1982 1983 1984 1985 1986 |
int vma_dup_policy(struct vm_area_struct *src, struct vm_area_struct *dst) { struct mempolicy *pol = mpol_dup(vma_policy(src)); if (IS_ERR(pol)) return PTR_ERR(pol); dst->vm_policy = pol; return 0; } |
4225399a6 [PATCH] cpuset: r... |
1987 |
/* |
846a16bf0 mempolicy: rename... |
1988 |
* If mpol_dup() sees current->cpuset == cpuset_being_rebound, then it |
4225399a6 [PATCH] cpuset: r... |
1989 1990 1991 1992 |
* 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... |
1993 1994 1995 |
* * 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... |
1996 |
*/ |
4225399a6 [PATCH] cpuset: r... |
1997 |
|
846a16bf0 mempolicy: rename... |
1998 1999 |
/* Slow path of a mempolicy duplicate */ struct mempolicy *__mpol_dup(struct mempolicy *old) |
1da177e4c Linux-2.6.12-rc2 |
2000 2001 2002 2003 2004 |
{ struct mempolicy *new = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!new) return ERR_PTR(-ENOMEM); |
708c1bbc9 mempolicy: restru... |
2005 2006 2007 2008 2009 2010 2011 2012 |
/* task's mempolicy is protected by alloc_lock */ if (old == current->mempolicy) { task_lock(current); *new = *old; task_unlock(current); } else *new = *old; |
4225399a6 [PATCH] cpuset: r... |
2013 2014 |
if (current_cpuset_is_being_rebound()) { nodemask_t mems = cpuset_mems_allowed(current); |
708c1bbc9 mempolicy: restru... |
2015 2016 2017 2018 |
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... |
2019 |
} |
1da177e4c Linux-2.6.12-rc2 |
2020 |
atomic_set(&new->refcnt, 1); |
1da177e4c Linux-2.6.12-rc2 |
2021 2022 2023 2024 |
return new; } /* Slow path of a mempolicy comparison */ |
fcfb4dcc9 mm/mempolicy.c: m... |
2025 |
bool __mpol_equal(struct mempolicy *a, struct mempolicy *b) |
1da177e4c Linux-2.6.12-rc2 |
2026 2027 |
{ if (!a || !b) |
fcfb4dcc9 mm/mempolicy.c: m... |
2028 |
return false; |
45c4745af mempolicy: rename... |
2029 |
if (a->mode != b->mode) |
fcfb4dcc9 mm/mempolicy.c: m... |
2030 |
return false; |
198005025 mempolicy: remove... |
2031 |
if (a->flags != b->flags) |
fcfb4dcc9 mm/mempolicy.c: m... |
2032 |
return false; |
198005025 mempolicy: remove... |
2033 2034 |
if (mpol_store_user_nodemask(a)) if (!nodes_equal(a->w.user_nodemask, b->w.user_nodemask)) |
fcfb4dcc9 mm/mempolicy.c: m... |
2035 |
return false; |
198005025 mempolicy: remove... |
2036 |
|
45c4745af mempolicy: rename... |
2037 |
switch (a->mode) { |
19770b326 mm: filter based ... |
2038 2039 |
case MPOL_BIND: /* Fall through */ |
1da177e4c Linux-2.6.12-rc2 |
2040 |
case MPOL_INTERLEAVE: |
fcfb4dcc9 mm/mempolicy.c: m... |
2041 |
return !!nodes_equal(a->v.nodes, b->v.nodes); |
1da177e4c Linux-2.6.12-rc2 |
2042 |
case MPOL_PREFERRED: |
757196618 mempolicy: remove... |
2043 |
return a->v.preferred_node == b->v.preferred_node; |
1da177e4c Linux-2.6.12-rc2 |
2044 2045 |
default: BUG(); |
fcfb4dcc9 mm/mempolicy.c: m... |
2046 |
return false; |
1da177e4c Linux-2.6.12-rc2 |
2047 2048 |
} } |
1da177e4c Linux-2.6.12-rc2 |
2049 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2050 2051 2052 2053 2054 2055 2056 2057 2058 |
* Shared memory backing store policy support. * * Remember policies even when nobody has shared memory mapped. * The policies are kept in Red-Black tree linked from the inode. * They are protected by the sp->lock spinlock, which should be held * for any accesses to the tree. */ /* lookup first element intersecting start-end */ |
42288fe36 mm: mempolicy: Co... |
2059 |
/* Caller holds sp->lock */ |
1da177e4c Linux-2.6.12-rc2 |
2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 |
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: ... |
2110 2111 |
pr_debug("inserting %lx-%lx: %d ", new->start, new->end, |
45c4745af mempolicy: rename... |
2112 |
new->policy ? new->policy->mode : 0); |
1da177e4c Linux-2.6.12-rc2 |
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 |
} /* Find shared policy intersecting idx */ struct mempolicy * mpol_shared_policy_lookup(struct shared_policy *sp, unsigned long idx) { struct mempolicy *pol = NULL; struct sp_node *sn; if (!sp->root.rb_node) return NULL; |
42288fe36 mm: mempolicy: Co... |
2124 |
spin_lock(&sp->lock); |
1da177e4c Linux-2.6.12-rc2 |
2125 2126 2127 2128 2129 |
sn = sp_lookup(sp, idx, idx+1); if (sn) { mpol_get(sn->policy); pol = sn->policy; } |
42288fe36 mm: mempolicy: Co... |
2130 |
spin_unlock(&sp->lock); |
1da177e4c Linux-2.6.12-rc2 |
2131 2132 |
return pol; } |
63f74ca21 mempolicy: fix re... |
2133 2134 2135 2136 2137 |
static void sp_free(struct sp_node *n) { mpol_put(n->policy); kmem_cache_free(sn_cache, n); } |
771fb4d80 mm: mempolicy: Ch... |
2138 2139 2140 |
/** * mpol_misplaced - check whether current page node is valid in policy * |
b46e14acb mm/mempolicy.c: p... |
2141 2142 2143 |
* @page: page to be checked * @vma: vm area where page mapped * @addr: virtual address where page mapped |
771fb4d80 mm: mempolicy: Ch... |
2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 |
* * Lookup current policy node id for vma,addr and "compare to" page's * node id. * * Returns: * -1 - not misplaced, page is in the right node * node - node id where the page should be * * Policy determination "mimics" alloc_page_vma(). * Called from fault path where we know the vma and faulting address. */ int mpol_misplaced(struct page *page, struct vm_area_struct *vma, unsigned long addr) { struct mempolicy *pol; struct zone *zone; int curnid = page_to_nid(page); unsigned long pgoff; |
90572890d mm: numa: Change ... |
2161 2162 |
int thiscpu = raw_smp_processor_id(); int thisnid = cpu_to_node(thiscpu); |
771fb4d80 mm: mempolicy: Ch... |
2163 2164 2165 2166 |
int polnid = -1; int ret = -1; BUG_ON(!vma); |
dd6eecb91 mempolicy: unexpo... |
2167 |
pol = get_vma_policy(vma, addr); |
771fb4d80 mm: mempolicy: Ch... |
2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 |
if (!(pol->flags & MPOL_F_MOF)) goto out; switch (pol->mode) { case MPOL_INTERLEAVE: BUG_ON(addr >= vma->vm_end); BUG_ON(addr < vma->vm_start); pgoff = vma->vm_pgoff; pgoff += (addr - vma->vm_start) >> PAGE_SHIFT; polnid = offset_il_node(pol, vma, pgoff); break; case MPOL_PREFERRED: if (pol->flags & MPOL_F_LOCAL) polnid = numa_node_id(); else polnid = pol->v.preferred_node; break; case MPOL_BIND: /* * allows binding to multiple nodes. * use current page if in policy nodemask, * else select nearest allowed node, if any. * If no allowed nodes, use current [!misplaced]. */ if (node_isset(curnid, pol->v.nodes)) goto out; (void)first_zones_zonelist( node_zonelist(numa_node_id(), GFP_HIGHUSER), gfp_zone(GFP_HIGHUSER), &pol->v.nodes, &zone); polnid = zone->node; break; default: BUG(); } |
5606e3877 mm: numa: Migrate... |
2207 2208 |
/* Migrate the page towards the node whose CPU is referencing it */ |
e42c8ff29 mm: numa: Use a t... |
2209 |
if (pol->flags & MPOL_F_MORON) { |
90572890d mm: numa: Change ... |
2210 |
polnid = thisnid; |
5606e3877 mm: numa: Migrate... |
2211 |
|
10f390427 sched/numa, mm: U... |
2212 |
if (!should_numa_migrate_memory(current, page, curnid, thiscpu)) |
de1c9ce6f sched/numa: Skip ... |
2213 |
goto out; |
e42c8ff29 mm: numa: Use a t... |
2214 |
} |
771fb4d80 mm: mempolicy: Ch... |
2215 2216 2217 2218 2219 2220 2221 |
if (curnid != polnid) ret = polnid; out: mpol_cond_put(pol); return ret; } |
1da177e4c Linux-2.6.12-rc2 |
2222 2223 |
static void sp_delete(struct shared_policy *sp, struct sp_node *n) { |
140d5a490 numa: mempolicy: ... |
2224 2225 |
pr_debug("deleting %lx-l%lx ", n->start, n->end); |
1da177e4c Linux-2.6.12-rc2 |
2226 |
rb_erase(&n->nd, &sp->root); |
63f74ca21 mempolicy: fix re... |
2227 |
sp_free(n); |
1da177e4c Linux-2.6.12-rc2 |
2228 |
} |
42288fe36 mm: mempolicy: Co... |
2229 2230 2231 2232 2233 2234 2235 |
static void sp_node_init(struct sp_node *node, unsigned long start, unsigned long end, struct mempolicy *pol) { node->start = start; node->end = end; node->policy = pol; } |
dbcb0f19c mm/mempolicy.c: c... |
2236 2237 |
static struct sp_node *sp_alloc(unsigned long start, unsigned long end, struct mempolicy *pol) |
1da177e4c Linux-2.6.12-rc2 |
2238 |
{ |
869833f2c mempolicy: remove... |
2239 2240 |
struct sp_node *n; struct mempolicy *newpol; |
1da177e4c Linux-2.6.12-rc2 |
2241 |
|
869833f2c mempolicy: remove... |
2242 |
n = kmem_cache_alloc(sn_cache, GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
2243 2244 |
if (!n) return NULL; |
869833f2c mempolicy: remove... |
2245 2246 2247 2248 2249 2250 2251 |
newpol = mpol_dup(pol); if (IS_ERR(newpol)) { kmem_cache_free(sn_cache, n); return NULL; } newpol->flags |= MPOL_F_SHARED; |
42288fe36 mm: mempolicy: Co... |
2252 |
sp_node_init(n, start, end, newpol); |
869833f2c mempolicy: remove... |
2253 |
|
1da177e4c Linux-2.6.12-rc2 |
2254 2255 2256 2257 2258 2259 2260 |
return n; } /* Replace a policy range. */ static int shared_policy_replace(struct shared_policy *sp, unsigned long start, unsigned long end, struct sp_node *new) { |
b22d127a3 mempolicy: fix a ... |
2261 |
struct sp_node *n; |
42288fe36 mm: mempolicy: Co... |
2262 2263 |
struct sp_node *n_new = NULL; struct mempolicy *mpol_new = NULL; |
b22d127a3 mempolicy: fix a ... |
2264 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2265 |
|
42288fe36 mm: mempolicy: Co... |
2266 2267 |
restart: spin_lock(&sp->lock); |
1da177e4c Linux-2.6.12-rc2 |
2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 |
n = sp_lookup(sp, start, end); /* Take care of old policies in the same range. */ while (n && n->start < end) { struct rb_node *next = rb_next(&n->nd); if (n->start >= start) { if (n->end <= end) sp_delete(sp, n); else n->start = end; } else { /* Old policy spanning whole new range. */ if (n->end > end) { |
42288fe36 mm: mempolicy: Co... |
2280 2281 2282 2283 2284 |
if (!n_new) goto alloc_new; *mpol_new = *n->policy; atomic_set(&mpol_new->refcnt, 1); |
7880639c3 mm/mempolicy.c: f... |
2285 |
sp_node_init(n_new, end, n->end, mpol_new); |
1da177e4c Linux-2.6.12-rc2 |
2286 |
n->end = start; |
5ca395751 mm/mempolicy.c: f... |
2287 |
sp_insert(sp, n_new); |
42288fe36 mm: mempolicy: Co... |
2288 2289 |
n_new = NULL; mpol_new = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 |
break; } else n->end = start; } if (!next) break; n = rb_entry(next, struct sp_node, nd); } if (new) sp_insert(sp, new); |
42288fe36 mm: mempolicy: Co... |
2300 2301 2302 2303 2304 2305 2306 2307 |
spin_unlock(&sp->lock); ret = 0; err_out: if (mpol_new) mpol_put(mpol_new); if (n_new) kmem_cache_free(sn_cache, n_new); |
b22d127a3 mempolicy: fix a ... |
2308 |
return ret; |
42288fe36 mm: mempolicy: Co... |
2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 |
alloc_new: spin_unlock(&sp->lock); ret = -ENOMEM; n_new = kmem_cache_alloc(sn_cache, GFP_KERNEL); if (!n_new) goto err_out; mpol_new = kmem_cache_alloc(policy_cache, GFP_KERNEL); if (!mpol_new) goto err_out; goto restart; |
1da177e4c Linux-2.6.12-rc2 |
2320 |
} |
71fe804b6 mempolicy: use st... |
2321 2322 2323 2324 2325 2326 2327 2328 |
/** * 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... |
2329 |
* This is called at get_inode() calls and we can use GFP_KERNEL. |
71fe804b6 mempolicy: use st... |
2330 2331 2332 |
*/ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol) { |
58568d2a8 cpuset,mm: update... |
2333 |
int ret; |
71fe804b6 mempolicy: use st... |
2334 |
sp->root = RB_ROOT; /* empty tree == default mempolicy */ |
42288fe36 mm: mempolicy: Co... |
2335 |
spin_lock_init(&sp->lock); |
71fe804b6 mempolicy: use st... |
2336 2337 2338 2339 |
if (mpol) { struct vm_area_struct pvma; struct mempolicy *new; |
4bfc44958 mm: make set_memp... |
2340 |
NODEMASK_SCRATCH(scratch); |
71fe804b6 mempolicy: use st... |
2341 |
|
4bfc44958 mm: make set_memp... |
2342 |
if (!scratch) |
5c0c16549 mempolicy: fix da... |
2343 |
goto put_mpol; |
71fe804b6 mempolicy: use st... |
2344 2345 |
/* contextualize the tmpfs mount point mempolicy */ new = mpol_new(mpol->mode, mpol->flags, &mpol->w.user_nodemask); |
15d77835a mempolicy: factor... |
2346 |
if (IS_ERR(new)) |
0cae3457b mempolicy: ERR_PT... |
2347 |
goto free_scratch; /* no valid nodemask intersection */ |
58568d2a8 cpuset,mm: update... |
2348 2349 |
task_lock(current); |
4bfc44958 mm: make set_memp... |
2350 |
ret = mpol_set_nodemask(new, &mpol->w.user_nodemask, scratch); |
58568d2a8 cpuset,mm: update... |
2351 |
task_unlock(current); |
15d77835a mempolicy: factor... |
2352 |
if (ret) |
5c0c16549 mempolicy: fix da... |
2353 |
goto put_new; |
71fe804b6 mempolicy: use st... |
2354 2355 2356 2357 2358 |
/* 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... |
2359 |
|
5c0c16549 mempolicy: fix da... |
2360 |
put_new: |
71fe804b6 mempolicy: use st... |
2361 |
mpol_put(new); /* drop initial ref */ |
0cae3457b mempolicy: ERR_PT... |
2362 |
free_scratch: |
4bfc44958 mm: make set_memp... |
2363 |
NODEMASK_SCRATCH_FREE(scratch); |
5c0c16549 mempolicy: fix da... |
2364 2365 |
put_mpol: mpol_put(mpol); /* drop our incoming ref on sb mpol */ |
7339ff830 [PATCH] Add tmpfs... |
2366 2367 |
} } |
1da177e4c Linux-2.6.12-rc2 |
2368 2369 2370 2371 2372 2373 |
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... |
2374 2375 |
pr_debug("set_shared_policy %lx sz %lu %d %d %lx ", |
1da177e4c Linux-2.6.12-rc2 |
2376 |
vma->vm_pgoff, |
45c4745af mempolicy: rename... |
2377 |
sz, npol ? npol->mode : -1, |
028fec414 mempolicy: suppor... |
2378 |
npol ? npol->flags : -1, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2379 |
npol ? nodes_addr(npol->v.nodes)[0] : NUMA_NO_NODE); |
1da177e4c Linux-2.6.12-rc2 |
2380 2381 2382 2383 2384 2385 2386 2387 |
if (npol) { new = sp_alloc(vma->vm_pgoff, vma->vm_pgoff + sz, npol); if (!new) return -ENOMEM; } err = shared_policy_replace(info, vma->vm_pgoff, vma->vm_pgoff+sz, new); if (err && new) |
63f74ca21 mempolicy: fix re... |
2388 |
sp_free(new); |
1da177e4c Linux-2.6.12-rc2 |
2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 |
return err; } /* Free a backing policy store on inode delete. */ void mpol_free_shared_policy(struct shared_policy *p) { struct sp_node *n; struct rb_node *next; if (!p->root.rb_node) return; |
42288fe36 mm: mempolicy: Co... |
2400 |
spin_lock(&p->lock); |
1da177e4c Linux-2.6.12-rc2 |
2401 2402 2403 2404 |
next = rb_first(&p->root); while (next) { n = rb_entry(next, struct sp_node, nd); next = rb_next(&n->nd); |
63f74ca21 mempolicy: fix re... |
2405 |
sp_delete(p, n); |
1da177e4c Linux-2.6.12-rc2 |
2406 |
} |
42288fe36 mm: mempolicy: Co... |
2407 |
spin_unlock(&p->lock); |
1da177e4c Linux-2.6.12-rc2 |
2408 |
} |
1a687c2e9 mm: sched: numa: ... |
2409 |
#ifdef CONFIG_NUMA_BALANCING |
c297663c0 mm: numa: initial... |
2410 |
static int __initdata numabalancing_override; |
1a687c2e9 mm: sched: numa: ... |
2411 2412 2413 2414 2415 2416 2417 |
static void __init check_numabalancing_enable(void) { bool numabalancing_default = false; if (IS_ENABLED(CONFIG_NUMA_BALANCING_DEFAULT_ENABLED)) numabalancing_default = true; |
c297663c0 mm: numa: initial... |
2418 2419 2420 |
/* Parsed by setup_numabalancing. override == 1 enables, -1 disables */ if (numabalancing_override) set_numabalancing_state(numabalancing_override == 1); |
b0dc2b9bb mm, numa: really ... |
2421 |
if (num_online_nodes() > 1 && !numabalancing_override) { |
4a404bea9 mm/mempolicy.c: c... |
2422 |
pr_info("%s automatic NUMA balancing. " |
c297663c0 mm: numa: initial... |
2423 2424 2425 |
"Configure with numa_balancing= or the " "kernel.numa_balancing sysctl", numabalancing_default ? "Enabling" : "Disabling"); |
1a687c2e9 mm: sched: numa: ... |
2426 2427 2428 2429 2430 2431 2432 2433 2434 |
set_numabalancing_state(numabalancing_default); } } static int __init setup_numabalancing(char *str) { int ret = 0; if (!str) goto out; |
1a687c2e9 mm: sched: numa: ... |
2435 2436 |
if (!strcmp(str, "enable")) { |
c297663c0 mm: numa: initial... |
2437 |
numabalancing_override = 1; |
1a687c2e9 mm: sched: numa: ... |
2438 2439 |
ret = 1; } else if (!strcmp(str, "disable")) { |
c297663c0 mm: numa: initial... |
2440 |
numabalancing_override = -1; |
1a687c2e9 mm: sched: numa: ... |
2441 2442 2443 2444 |
ret = 1; } out: if (!ret) |
4a404bea9 mm/mempolicy.c: c... |
2445 2446 |
pr_warn("Unable to parse numa_balancing= "); |
1a687c2e9 mm: sched: numa: ... |
2447 2448 2449 2450 2451 2452 2453 2454 2455 |
return ret; } __setup("numa_balancing=", setup_numabalancing); #else static inline void __init check_numabalancing_enable(void) { } #endif /* CONFIG_NUMA_BALANCING */ |
1da177e4c Linux-2.6.12-rc2 |
2456 2457 2458 |
/* assumes fs == KERNEL_DS */ void __init numa_policy_init(void) { |
b71636e29 numa: mempolicy: ... |
2459 2460 2461 |
nodemask_t interleave_nodes; unsigned long largest = 0; int nid, prefer = 0; |
1da177e4c Linux-2.6.12-rc2 |
2462 2463 |
policy_cache = kmem_cache_create("numa_policy", sizeof(struct mempolicy), |
20c2df83d mm: Remove slab d... |
2464 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2465 2466 2467 |
sn_cache = kmem_cache_create("shared_policy_node", sizeof(struct sp_node), |
20c2df83d mm: Remove slab d... |
2468 |
0, SLAB_PANIC, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2469 |
|
5606e3877 mm: numa: Migrate... |
2470 2471 2472 2473 2474 2475 2476 2477 |
for_each_node(nid) { preferred_node_policy[nid] = (struct mempolicy) { .refcnt = ATOMIC_INIT(1), .mode = MPOL_PREFERRED, .flags = MPOL_F_MOF | MPOL_F_MORON, .v = { .preferred_node = nid, }, }; } |
b71636e29 numa: mempolicy: ... |
2478 2479 2480 2481 2482 2483 |
/* * Set interleaving policy for system init. Interleaving is only * enabled across suitably sized nodes (default is >= 16MB), or * fall back to the largest node if they're all smaller. */ nodes_clear(interleave_nodes); |
01f13bd60 mempolicy: use N_... |
2484 |
for_each_node_state(nid, N_MEMORY) { |
b71636e29 numa: mempolicy: ... |
2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 |
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 |
2501 |
|
028fec414 mempolicy: suppor... |
2502 |
if (do_set_mempolicy(MPOL_INTERLEAVE, 0, &interleave_nodes)) |
b1de0d139 mm: convert some ... |
2503 2504 |
pr_err("%s: interleaving failed ", __func__); |
1a687c2e9 mm: sched: numa: ... |
2505 2506 |
check_numabalancing_enable(); |
1da177e4c Linux-2.6.12-rc2 |
2507 |
} |
8bccd85ff [PATCH] Implement... |
2508 |
/* Reset policy of current process to default */ |
1da177e4c Linux-2.6.12-rc2 |
2509 2510 |
void numa_default_policy(void) { |
028fec414 mempolicy: suppor... |
2511 |
do_set_mempolicy(MPOL_DEFAULT, 0, NULL); |
1da177e4c Linux-2.6.12-rc2 |
2512 |
} |
68860ec10 [PATCH] cpusets: ... |
2513 |
|
4225399a6 [PATCH] cpuset: r... |
2514 |
/* |
095f1fc4e mempolicy: rework... |
2515 2516 2517 2518 |
* Parse and format mempolicy from/to strings */ /* |
f2a07f40d tmpfs mempolicy: ... |
2519 |
* "local" is implemented internally by MPOL_PREFERRED with MPOL_F_LOCAL flag. |
1a75a6c82 [PATCH] Fold numa... |
2520 |
*/ |
345ace9c7 mempolicy: rename... |
2521 2522 2523 2524 2525 2526 |
static const char * const policy_modes[] = { [MPOL_DEFAULT] = "default", [MPOL_PREFERRED] = "prefer", [MPOL_BIND] = "bind", [MPOL_INTERLEAVE] = "interleave", |
d3a710337 mm: mempolicy: Ad... |
2527 |
[MPOL_LOCAL] = "local", |
345ace9c7 mempolicy: rename... |
2528 |
}; |
1a75a6c82 [PATCH] Fold numa... |
2529 |
|
095f1fc4e mempolicy: rework... |
2530 2531 2532 |
#ifdef CONFIG_TMPFS /** |
f2a07f40d tmpfs mempolicy: ... |
2533 |
* mpol_parse_str - parse string to mempolicy, for tmpfs mpol mount option. |
095f1fc4e mempolicy: rework... |
2534 |
* @str: string containing mempolicy to parse |
71fe804b6 mempolicy: use st... |
2535 |
* @mpol: pointer to struct mempolicy pointer, returned on success. |
095f1fc4e mempolicy: rework... |
2536 2537 2538 2539 |
* * Format of input: * <mode>[=<flags>][:<nodelist>] * |
71fe804b6 mempolicy: use st... |
2540 |
* On success, returns 0, else 1 |
095f1fc4e mempolicy: rework... |
2541 |
*/ |
a7a88b237 mempolicy: remove... |
2542 |
int mpol_parse_str(char *str, struct mempolicy **mpol) |
095f1fc4e mempolicy: rework... |
2543 |
{ |
71fe804b6 mempolicy: use st... |
2544 |
struct mempolicy *new = NULL; |
b4652e842 mempolicy: lose u... |
2545 |
unsigned short mode; |
f2a07f40d tmpfs mempolicy: ... |
2546 |
unsigned short mode_flags; |
71fe804b6 mempolicy: use st... |
2547 |
nodemask_t nodes; |
095f1fc4e mempolicy: rework... |
2548 2549 |
char *nodelist = strchr(str, ':'); char *flags = strchr(str, '='); |
095f1fc4e mempolicy: rework... |
2550 2551 2552 2553 2554 |
int err = 1; if (nodelist) { /* NUL-terminate mode or flags string */ *nodelist++ = '\0'; |
71fe804b6 mempolicy: use st... |
2555 |
if (nodelist_parse(nodelist, nodes)) |
095f1fc4e mempolicy: rework... |
2556 |
goto out; |
01f13bd60 mempolicy: use N_... |
2557 |
if (!nodes_subset(nodes, node_states[N_MEMORY])) |
095f1fc4e mempolicy: rework... |
2558 |
goto out; |
71fe804b6 mempolicy: use st... |
2559 2560 |
} else nodes_clear(nodes); |
095f1fc4e mempolicy: rework... |
2561 2562 |
if (flags) *flags++ = '\0'; /* terminate mode string */ |
479e2802d mm: mempolicy: Ma... |
2563 |
for (mode = 0; mode < MPOL_MAX; mode++) { |
345ace9c7 mempolicy: rename... |
2564 |
if (!strcmp(str, policy_modes[mode])) { |
095f1fc4e mempolicy: rework... |
2565 2566 2567 |
break; } } |
a720094de mm: mempolicy: Hi... |
2568 |
if (mode >= MPOL_MAX) |
095f1fc4e mempolicy: rework... |
2569 |
goto out; |
71fe804b6 mempolicy: use st... |
2570 |
switch (mode) { |
095f1fc4e mempolicy: rework... |
2571 |
case MPOL_PREFERRED: |
71fe804b6 mempolicy: use st... |
2572 2573 2574 |
/* * Insist on a nodelist of one node only */ |
095f1fc4e mempolicy: rework... |
2575 2576 2577 2578 |
if (nodelist) { char *rest = nodelist; while (isdigit(*rest)) rest++; |
926f2ae04 tmpfs: cleanup mp... |
2579 2580 |
if (*rest) goto out; |
095f1fc4e mempolicy: rework... |
2581 2582 |
} break; |
095f1fc4e mempolicy: rework... |
2583 2584 2585 2586 2587 |
case MPOL_INTERLEAVE: /* * Default to online nodes with memory if no nodelist */ if (!nodelist) |
01f13bd60 mempolicy: use N_... |
2588 |
nodes = node_states[N_MEMORY]; |
3f226aa1c mempolicy: suppor... |
2589 |
break; |
71fe804b6 mempolicy: use st... |
2590 |
case MPOL_LOCAL: |
3f226aa1c mempolicy: suppor... |
2591 |
/* |
71fe804b6 mempolicy: use st... |
2592 |
* Don't allow a nodelist; mpol_new() checks flags |
3f226aa1c mempolicy: suppor... |
2593 |
*/ |
71fe804b6 mempolicy: use st... |
2594 |
if (nodelist) |
3f226aa1c mempolicy: suppor... |
2595 |
goto out; |
71fe804b6 mempolicy: use st... |
2596 |
mode = MPOL_PREFERRED; |
3f226aa1c mempolicy: suppor... |
2597 |
break; |
413b43dea tmpfs: fix oops o... |
2598 2599 2600 2601 2602 2603 2604 |
case MPOL_DEFAULT: /* * Insist on a empty nodelist */ if (!nodelist) err = 0; goto out; |
d69b2e63e tmpfs: mpol=bind:... |
2605 2606 2607 2608 2609 2610 |
case MPOL_BIND: /* * Insist on a nodelist */ if (!nodelist) goto out; |
095f1fc4e mempolicy: rework... |
2611 |
} |
71fe804b6 mempolicy: use st... |
2612 |
mode_flags = 0; |
095f1fc4e mempolicy: rework... |
2613 2614 2615 2616 2617 2618 |
if (flags) { /* * Currently, we only support two mutually exclusive * mode flags. */ if (!strcmp(flags, "static")) |
71fe804b6 mempolicy: use st... |
2619 |
mode_flags |= MPOL_F_STATIC_NODES; |
095f1fc4e mempolicy: rework... |
2620 |
else if (!strcmp(flags, "relative")) |
71fe804b6 mempolicy: use st... |
2621 |
mode_flags |= MPOL_F_RELATIVE_NODES; |
095f1fc4e mempolicy: rework... |
2622 |
else |
926f2ae04 tmpfs: cleanup mp... |
2623 |
goto out; |
095f1fc4e mempolicy: rework... |
2624 |
} |
71fe804b6 mempolicy: use st... |
2625 2626 2627 |
new = mpol_new(mode, mode_flags, &nodes); if (IS_ERR(new)) |
926f2ae04 tmpfs: cleanup mp... |
2628 |
goto out; |
f2a07f40d tmpfs mempolicy: ... |
2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 |
/* * Save nodes for mpol_to_str() to show the tmpfs mount options * for /proc/mounts, /proc/pid/mounts and /proc/pid/mountinfo. */ if (mode != MPOL_PREFERRED) new->v.nodes = nodes; else if (nodelist) new->v.preferred_node = first_node(nodes); else new->flags |= MPOL_F_LOCAL; /* * Save nodes for contextualization: this will be used to "clone" * the mempolicy in a specific context [cpuset] at a later time. */ new->w.user_nodemask = nodes; |
926f2ae04 tmpfs: cleanup mp... |
2645 |
err = 0; |
71fe804b6 mempolicy: use st... |
2646 |
|
095f1fc4e mempolicy: rework... |
2647 2648 2649 2650 2651 2652 |
out: /* Restore string for error message */ if (nodelist) *--nodelist = ':'; if (flags) *--flags = '='; |
71fe804b6 mempolicy: use st... |
2653 2654 |
if (!err) *mpol = new; |
095f1fc4e mempolicy: rework... |
2655 2656 2657 |
return err; } #endif /* CONFIG_TMPFS */ |
71fe804b6 mempolicy: use st... |
2658 2659 2660 2661 2662 |
/** * mpol_to_str - format a mempolicy structure for printing * @buffer: to contain formatted mempolicy string * @maxlen: length of @buffer * @pol: pointer to mempolicy to be formatted |
71fe804b6 mempolicy: use st... |
2663 |
* |
948927ee9 mm, mempolicy: ma... |
2664 2665 2666 |
* Convert @pol into a string. If @buffer is too short, truncate the string. * Recommend a @maxlen of at least 32 for the longest mode, "interleave", the * longest flag, "relative", and to display at least a few node ids. |
1a75a6c82 [PATCH] Fold numa... |
2667 |
*/ |
948927ee9 mm, mempolicy: ma... |
2668 |
void mpol_to_str(char *buffer, int maxlen, struct mempolicy *pol) |
1a75a6c82 [PATCH] Fold numa... |
2669 2670 |
{ char *p = buffer; |
948927ee9 mm, mempolicy: ma... |
2671 2672 2673 |
nodemask_t nodes = NODE_MASK_NONE; unsigned short mode = MPOL_DEFAULT; unsigned short flags = 0; |
2291990ab mempolicy: clean-... |
2674 |
|
8790c71a1 mm/mempolicy.c: f... |
2675 |
if (pol && pol != &default_policy && !(pol->flags & MPOL_F_MORON)) { |
bea904d54 mempolicy: use MP... |
2676 |
mode = pol->mode; |
948927ee9 mm, mempolicy: ma... |
2677 2678 |
flags = pol->flags; } |
bea904d54 mempolicy: use MP... |
2679 |
|
1a75a6c82 [PATCH] Fold numa... |
2680 2681 |
switch (mode) { case MPOL_DEFAULT: |
1a75a6c82 [PATCH] Fold numa... |
2682 |
break; |
1a75a6c82 [PATCH] Fold numa... |
2683 |
case MPOL_PREFERRED: |
fc36b8d3d mempolicy: use MP... |
2684 |
if (flags & MPOL_F_LOCAL) |
f2a07f40d tmpfs mempolicy: ... |
2685 |
mode = MPOL_LOCAL; |
53f2556b6 mempolicy: mPOL_P... |
2686 |
else |
fc36b8d3d mempolicy: use MP... |
2687 |
node_set(pol->v.preferred_node, nodes); |
1a75a6c82 [PATCH] Fold numa... |
2688 |
break; |
1a75a6c82 [PATCH] Fold numa... |
2689 |
case MPOL_BIND: |
1a75a6c82 [PATCH] Fold numa... |
2690 |
case MPOL_INTERLEAVE: |
f2a07f40d tmpfs mempolicy: ... |
2691 |
nodes = pol->v.nodes; |
1a75a6c82 [PATCH] Fold numa... |
2692 |
break; |
1a75a6c82 [PATCH] Fold numa... |
2693 |
default: |
948927ee9 mm, mempolicy: ma... |
2694 2695 2696 |
WARN_ON_ONCE(1); snprintf(p, maxlen, "unknown"); return; |
1a75a6c82 [PATCH] Fold numa... |
2697 |
} |
b7a9f420e mm, mempolicy: si... |
2698 |
p += snprintf(p, maxlen, "%s", policy_modes[mode]); |
1a75a6c82 [PATCH] Fold numa... |
2699 |
|
fc36b8d3d mempolicy: use MP... |
2700 |
if (flags & MPOL_MODE_FLAGS) { |
948927ee9 mm, mempolicy: ma... |
2701 |
p += snprintf(p, buffer + maxlen - p, "="); |
f5b087b52 mempolicy: add MP... |
2702 |
|
2291990ab mempolicy: clean-... |
2703 2704 2705 |
/* * Currently, the only defined flags are mutually exclusive */ |
f5b087b52 mempolicy: add MP... |
2706 |
if (flags & MPOL_F_STATIC_NODES) |
2291990ab mempolicy: clean-... |
2707 2708 2709 |
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... |
2710 |
} |
9e763e0f4 mm: use %*pb[l] t... |
2711 2712 2713 |
if (!nodes_empty(nodes)) p += scnprintf(p, buffer + maxlen - p, ":%*pbl", nodemask_pr_args(&nodes)); |
1a75a6c82 [PATCH] Fold numa... |
2714 |
} |