Blame view
mm/memory-failure.c
49.1 KB
6a46079cf
|
1 2 3 4 5 6 7 8 9 |
/* * Copyright (C) 2008, 2009 Intel Corporation * Authors: Andi Kleen, Fengguang Wu * * This software may be redistributed and/or modified under the terms of * the GNU General Public License ("GPL") version 2 only as published by the * Free Software Foundation. * * High level machine check handler. Handles pages reported by the |
1c80b990a
|
10 |
* hardware as being corrupted usually due to a multi-bit ECC memory or cache |
6a46079cf
|
11 |
* failure. |
1c80b990a
|
12 13 14 |
* * In addition there is a "soft offline" entry point that allows stop using * not-yet-corrupted-by-suspicious pages without killing anything. |
6a46079cf
|
15 16 |
* * Handles page cache pages in various states. The tricky part |
1c80b990a
|
17 18 19 20 21 22 |
* here is that we can access any page asynchronously in respect to * other VM users, because memory failures could happen anytime and * anywhere. This could violate some of their assumptions. This is why * this code has to be extremely careful. Generally it tries to use * normal locking rules, as in get the standard locks, even if that means * the error handling takes potentially a long time. |
e0de78dfb
|
23 24 25 26 27 28 29 30 |
* * It can be very tempting to add handling for obscure cases here. * In general any code for handling new cases should only be added iff: * - You know how to test it. * - You have a test that can be added to mce-test * https://git.kernel.org/cgit/utils/cpu/mce/mce-test.git/ * - The case actually shows up as a frequent (top 10) page state in * tools/vm/page-types when running a real workload. |
1c80b990a
|
31 32 33 34 35 36 37 |
* * There are several operations here with exponential complexity because * of unsuitable VM data structures. For example the operation to map back * from RMAP chains to processes has to walk the complete process list and * has non linear complexity with the number. But since memory corruptions * are rare we hope to get away with this. This avoids impacting the core * VM. |
6a46079cf
|
38 |
*/ |
6a46079cf
|
39 40 41 |
#include <linux/kernel.h> #include <linux/mm.h> #include <linux/page-flags.h> |
478c5ffc0
|
42 |
#include <linux/kernel-page-flags.h> |
6a46079cf
|
43 |
#include <linux/sched.h> |
01e00f880
|
44 |
#include <linux/ksm.h> |
6a46079cf
|
45 |
#include <linux/rmap.h> |
b9e15bafd
|
46 |
#include <linux/export.h> |
6a46079cf
|
47 48 49 |
#include <linux/pagemap.h> #include <linux/swap.h> #include <linux/backing-dev.h> |
facb6011f
|
50 51 52 |
#include <linux/migrate.h> #include <linux/page-isolation.h> #include <linux/suspend.h> |
5a0e3ad6a
|
53 |
#include <linux/slab.h> |
bf998156d
|
54 |
#include <linux/swapops.h> |
7af446a84
|
55 |
#include <linux/hugetlb.h> |
20d6c96b5
|
56 |
#include <linux/memory_hotplug.h> |
5db8a73a8
|
57 |
#include <linux/mm_inline.h> |
ea8f5fb8a
|
58 |
#include <linux/kfifo.h> |
a5f651090
|
59 |
#include <linux/ratelimit.h> |
6a46079cf
|
60 |
#include "internal.h" |
97f0b1345
|
61 |
#include "ras/ras_event.h" |
6a46079cf
|
62 63 64 65 |
int sysctl_memory_failure_early_kill __read_mostly = 0; int sysctl_memory_failure_recovery __read_mostly = 1; |
293c07e31
|
66 |
atomic_long_t num_poisoned_pages __read_mostly = ATOMIC_LONG_INIT(0); |
6a46079cf
|
67 |
|
27df5068e
|
68 |
#if defined(CONFIG_HWPOISON_INJECT) || defined(CONFIG_HWPOISON_INJECT_MODULE) |
1bfe5febe
|
69 |
u32 hwpoison_filter_enable = 0; |
7c116f2b0
|
70 71 |
u32 hwpoison_filter_dev_major = ~0U; u32 hwpoison_filter_dev_minor = ~0U; |
478c5ffc0
|
72 73 |
u64 hwpoison_filter_flags_mask; u64 hwpoison_filter_flags_value; |
1bfe5febe
|
74 |
EXPORT_SYMBOL_GPL(hwpoison_filter_enable); |
7c116f2b0
|
75 76 |
EXPORT_SYMBOL_GPL(hwpoison_filter_dev_major); EXPORT_SYMBOL_GPL(hwpoison_filter_dev_minor); |
478c5ffc0
|
77 78 |
EXPORT_SYMBOL_GPL(hwpoison_filter_flags_mask); EXPORT_SYMBOL_GPL(hwpoison_filter_flags_value); |
7c116f2b0
|
79 80 81 82 83 84 85 86 87 88 89 |
static int hwpoison_filter_dev(struct page *p) { struct address_space *mapping; dev_t dev; if (hwpoison_filter_dev_major == ~0U && hwpoison_filter_dev_minor == ~0U) return 0; /* |
1c80b990a
|
90 |
* page_mapping() does not accept slab pages. |
7c116f2b0
|
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 |
*/ if (PageSlab(p)) return -EINVAL; mapping = page_mapping(p); if (mapping == NULL || mapping->host == NULL) return -EINVAL; dev = mapping->host->i_sb->s_dev; if (hwpoison_filter_dev_major != ~0U && hwpoison_filter_dev_major != MAJOR(dev)) return -EINVAL; if (hwpoison_filter_dev_minor != ~0U && hwpoison_filter_dev_minor != MINOR(dev)) return -EINVAL; return 0; } |
478c5ffc0
|
109 110 111 112 113 114 115 116 117 118 119 |
static int hwpoison_filter_flags(struct page *p) { if (!hwpoison_filter_flags_mask) return 0; if ((stable_page_flags(p) & hwpoison_filter_flags_mask) == hwpoison_filter_flags_value) return 0; else return -EINVAL; } |
4fd466eb4
|
120 121 122 123 124 125 126 127 128 129 |
/* * This allows stress tests to limit test scope to a collection of tasks * by putting them under some memcg. This prevents killing unrelated/important * processes such as /sbin/init. Note that the target task may share clean * pages with init (eg. libc text), which is harmless. If the target task * share _dirty_ pages with another task B, the test scheme must make sure B * is also included in the memcg. At last, due to race conditions this filter * can only guarantee that the page either belongs to the memcg tasks, or is * a freed page. */ |
94a59fb36
|
130 |
#ifdef CONFIG_MEMCG |
4fd466eb4
|
131 132 133 134 |
u64 hwpoison_filter_memcg; EXPORT_SYMBOL_GPL(hwpoison_filter_memcg); static int hwpoison_filter_task(struct page *p) { |
4fd466eb4
|
135 136 |
if (!hwpoison_filter_memcg) return 0; |
94a59fb36
|
137 |
if (page_cgroup_ino(p) != hwpoison_filter_memcg) |
4fd466eb4
|
138 139 140 141 142 143 144 |
return -EINVAL; return 0; } #else static int hwpoison_filter_task(struct page *p) { return 0; } #endif |
7c116f2b0
|
145 146 |
int hwpoison_filter(struct page *p) { |
1bfe5febe
|
147 148 |
if (!hwpoison_filter_enable) return 0; |
7c116f2b0
|
149 150 |
if (hwpoison_filter_dev(p)) return -EINVAL; |
478c5ffc0
|
151 152 |
if (hwpoison_filter_flags(p)) return -EINVAL; |
4fd466eb4
|
153 154 |
if (hwpoison_filter_task(p)) return -EINVAL; |
7c116f2b0
|
155 156 |
return 0; } |
27df5068e
|
157 158 159 160 161 162 |
#else int hwpoison_filter(struct page *p) { return 0; } #endif |
7c116f2b0
|
163 |
EXPORT_SYMBOL_GPL(hwpoison_filter); |
6a46079cf
|
164 |
/* |
7329bbeb9
|
165 166 167 |
* Send all the processes who have the page mapped a signal. * ``action optional'' if they are not immediately affected by the error * ``action required'' if error happened in current execution context |
6a46079cf
|
168 |
*/ |
7329bbeb9
|
169 170 |
static int kill_proc(struct task_struct *t, unsigned long addr, int trapno, unsigned long pfn, struct page *page, int flags) |
6a46079cf
|
171 172 173 |
{ struct siginfo si; int ret; |
495367c05
|
174 175 176 |
pr_err("Memory failure: %#lx: Killing %s:%d due to hardware memory corruption ", pfn, t->comm, t->pid); |
6a46079cf
|
177 178 |
si.si_signo = SIGBUS; si.si_errno = 0; |
6a46079cf
|
179 180 181 182 |
si.si_addr = (void *)addr; #ifdef __ARCH_SI_TRAPNO si.si_trapno = trapno; #endif |
f9121153f
|
183 |
si.si_addr_lsb = compound_order(compound_head(page)) + PAGE_SHIFT; |
7329bbeb9
|
184 |
|
a70ffcac7
|
185 |
if ((flags & MF_ACTION_REQUIRED) && t->mm == current->mm) { |
7329bbeb9
|
186 |
si.si_code = BUS_MCEERR_AR; |
a70ffcac7
|
187 |
ret = force_sig_info(SIGBUS, &si, current); |
7329bbeb9
|
188 189 190 191 192 193 194 195 196 197 |
} else { /* * Don't use force here, it's convenient if the signal * can be temporarily blocked. * This could cause a loop when the user sets SIGBUS * to SIG_IGN, but hopefully no one will do that? */ si.si_code = BUS_MCEERR_AO; ret = send_sig_info(SIGBUS, &si, t); /* synchronous? */ } |
6a46079cf
|
198 |
if (ret < 0) |
495367c05
|
199 200 |
pr_info("Memory failure: Error sending signal to %s:%d: %d ", |
1170532bb
|
201 |
t->comm, t->pid, ret); |
6a46079cf
|
202 203 204 205 |
return ret; } /* |
588f9ce6c
|
206 207 208 |
* When a unknown page type is encountered drain as many buffers as possible * in the hope to turn the page into a LRU or free page, which we can handle. */ |
facb6011f
|
209 |
void shake_page(struct page *p, int access) |
588f9ce6c
|
210 211 212 213 214 |
{ if (!PageSlab(p)) { lru_add_drain_all(); if (PageLRU(p)) return; |
c05543293
|
215 |
drain_all_pages(page_zone(p)); |
588f9ce6c
|
216 217 218 |
if (PageLRU(p) || is_free_buddy_page(p)) return; } |
facb6011f
|
219 |
|
588f9ce6c
|
220 |
/* |
6b4f7799c
|
221 222 |
* Only call shrink_node_slabs here (which would also shrink * other caches) if access is not potentially fatal. |
588f9ce6c
|
223 |
*/ |
cb731d6c6
|
224 225 |
if (access) drop_slab_node(page_to_nid(p)); |
588f9ce6c
|
226 227 228 229 |
} EXPORT_SYMBOL_GPL(shake_page); /* |
6a46079cf
|
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 |
* Kill all processes that have a poisoned page mapped and then isolate * the page. * * General strategy: * Find all processes having the page mapped and kill them. * But we keep a page reference around so that the page is not * actually freed yet. * Then stash the page away * * There's no convenient way to get back to mapped processes * from the VMAs. So do a brute-force search over all * running processes. * * Remember that machine checks are not common (or rather * if they are common you have other problems), so this shouldn't * be a performance issue. * * Also there are some races possible while we get from the * error detection to actually handle it. */ struct to_kill { struct list_head nd; struct task_struct *tsk; unsigned long addr; |
9033ae164
|
255 |
char addr_valid; |
6a46079cf
|
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 |
}; /* * Failure handling: if we can't find or can't kill a process there's * not much we can do. We just print a message and ignore otherwise. */ /* * Schedule a process for later kill. * Uses GFP_ATOMIC allocations to avoid potential recursions in the VM. * TBD would GFP_NOIO be enough? */ static void add_to_kill(struct task_struct *tsk, struct page *p, struct vm_area_struct *vma, struct list_head *to_kill, struct to_kill **tkc) { struct to_kill *tk; if (*tkc) { tk = *tkc; *tkc = NULL; } else { tk = kmalloc(sizeof(struct to_kill), GFP_ATOMIC); if (!tk) { |
495367c05
|
281 282 |
pr_err("Memory failure: Out of memory while machine check handling "); |
6a46079cf
|
283 284 285 286 287 288 289 290 291 292 293 294 295 |
return; } } tk->addr = page_address_in_vma(p, vma); tk->addr_valid = 1; /* * In theory we don't have to kill when the page was * munmaped. But it could be also a mremap. Since that's * likely very rare kill anyways just out of paranoia, but use * a SIGKILL because the error is not contained anymore. */ if (tk->addr == -EFAULT) { |
495367c05
|
296 297 |
pr_info("Memory failure: Unable to find user space address %lx in %s ", |
6a46079cf
|
298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 |
page_to_pfn(p), tsk->comm); tk->addr_valid = 0; } get_task_struct(tsk); tk->tsk = tsk; list_add_tail(&tk->nd, to_kill); } /* * Kill the processes that have been collected earlier. * * Only do anything when DOIT is set, otherwise just free the list * (this is used for clean pages which do not need killing) * Also when FAIL is set do a force kill because something went * wrong earlier. */ |
6751ed65d
|
314 |
static void kill_procs(struct list_head *to_kill, int forcekill, int trapno, |
7329bbeb9
|
315 316 |
int fail, struct page *page, unsigned long pfn, int flags) |
6a46079cf
|
317 318 319 320 |
{ struct to_kill *tk, *next; list_for_each_entry_safe (tk, next, to_kill, nd) { |
6751ed65d
|
321 |
if (forcekill) { |
6a46079cf
|
322 |
/* |
af901ca18
|
323 |
* In case something went wrong with munmapping |
6a46079cf
|
324 325 |
* make sure the process doesn't catch the * signal and then access the memory. Just kill it. |
6a46079cf
|
326 327 |
*/ if (fail || tk->addr_valid == 0) { |
495367c05
|
328 329 |
pr_err("Memory failure: %#lx: forcibly killing %s:%d because of failure to unmap corrupted page ", |
1170532bb
|
330 |
pfn, tk->tsk->comm, tk->tsk->pid); |
6a46079cf
|
331 332 333 334 335 336 337 338 339 |
force_sig(SIGKILL, tk->tsk); } /* * In theory the process could have mapped * something else on the address in-between. We could * check for that, but we need to tell the * process anyways. */ |
7329bbeb9
|
340 341 |
else if (kill_proc(tk->tsk, tk->addr, trapno, pfn, page, flags) < 0) |
495367c05
|
342 343 |
pr_err("Memory failure: %#lx: Cannot send advisory machine check signal to %s:%d ", |
1170532bb
|
344 |
pfn, tk->tsk->comm, tk->tsk->pid); |
6a46079cf
|
345 346 347 348 349 |
} put_task_struct(tk->tsk); kfree(tk); } } |
3ba08129e
|
350 351 352 353 354 355 356 357 358 |
/* * Find a dedicated thread which is supposed to handle SIGBUS(BUS_MCEERR_AO) * on behalf of the thread group. Return task_struct of the (first found) * dedicated thread if found, and return NULL otherwise. * * We already hold read_lock(&tasklist_lock) in the caller, so we don't * have to call rcu_read_lock/unlock() in this function. */ static struct task_struct *find_early_kill_thread(struct task_struct *tsk) |
6a46079cf
|
359 |
{ |
3ba08129e
|
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 |
struct task_struct *t; for_each_thread(tsk, t) if ((t->flags & PF_MCE_PROCESS) && (t->flags & PF_MCE_EARLY)) return t; return NULL; } /* * Determine whether a given process is "early kill" process which expects * to be signaled when some page under the process is hwpoisoned. * Return task_struct of the dedicated thread (main thread unless explicitly * specified) if the process is "early kill," and otherwise returns NULL. */ static struct task_struct *task_early_kill(struct task_struct *tsk, int force_early) { struct task_struct *t; |
6a46079cf
|
378 |
if (!tsk->mm) |
3ba08129e
|
379 |
return NULL; |
74614de17
|
380 |
if (force_early) |
3ba08129e
|
381 382 383 384 385 386 387 |
return tsk; t = find_early_kill_thread(tsk); if (t) return t; if (sysctl_memory_failure_early_kill) return tsk; return NULL; |
6a46079cf
|
388 389 390 391 392 393 |
} /* * Collect processes when the error hit an anonymous page. */ static void collect_procs_anon(struct page *page, struct list_head *to_kill, |
74614de17
|
394 |
struct to_kill **tkc, int force_early) |
6a46079cf
|
395 396 397 398 |
{ struct vm_area_struct *vma; struct task_struct *tsk; struct anon_vma *av; |
bf181b9f9
|
399 |
pgoff_t pgoff; |
6a46079cf
|
400 |
|
4fc3f1d66
|
401 |
av = page_lock_anon_vma_read(page); |
6a46079cf
|
402 |
if (av == NULL) /* Not actually mapped anymore */ |
9b679320a
|
403 |
return; |
a0f7a756c
|
404 |
pgoff = page_to_pgoff(page); |
9b679320a
|
405 |
read_lock(&tasklist_lock); |
6a46079cf
|
406 |
for_each_process (tsk) { |
5beb49305
|
407 |
struct anon_vma_chain *vmac; |
3ba08129e
|
408 |
struct task_struct *t = task_early_kill(tsk, force_early); |
5beb49305
|
409 |
|
3ba08129e
|
410 |
if (!t) |
6a46079cf
|
411 |
continue; |
bf181b9f9
|
412 413 |
anon_vma_interval_tree_foreach(vmac, &av->rb_root, pgoff, pgoff) { |
5beb49305
|
414 |
vma = vmac->vma; |
6a46079cf
|
415 416 |
if (!page_mapped_in_vma(page, vma)) continue; |
3ba08129e
|
417 418 |
if (vma->vm_mm == t->mm) add_to_kill(t, page, vma, to_kill, tkc); |
6a46079cf
|
419 420 |
} } |
6a46079cf
|
421 |
read_unlock(&tasklist_lock); |
4fc3f1d66
|
422 |
page_unlock_anon_vma_read(av); |
6a46079cf
|
423 424 425 426 427 428 |
} /* * Collect processes when the error hit a file mapped page. */ static void collect_procs_file(struct page *page, struct list_head *to_kill, |
74614de17
|
429 |
struct to_kill **tkc, int force_early) |
6a46079cf
|
430 431 432 |
{ struct vm_area_struct *vma; struct task_struct *tsk; |
6a46079cf
|
433 |
struct address_space *mapping = page->mapping; |
d28eb9c86
|
434 |
i_mmap_lock_read(mapping); |
9b679320a
|
435 |
read_lock(&tasklist_lock); |
6a46079cf
|
436 |
for_each_process(tsk) { |
a0f7a756c
|
437 |
pgoff_t pgoff = page_to_pgoff(page); |
3ba08129e
|
438 |
struct task_struct *t = task_early_kill(tsk, force_early); |
6a46079cf
|
439 |
|
3ba08129e
|
440 |
if (!t) |
6a46079cf
|
441 |
continue; |
6b2dbba8b
|
442 |
vma_interval_tree_foreach(vma, &mapping->i_mmap, pgoff, |
6a46079cf
|
443 444 445 446 447 448 449 450 |
pgoff) { /* * Send early kill signal to tasks where a vma covers * the page but the corrupted page is not necessarily * mapped it in its pte. * Assume applications who requested early kill want * to be informed of all such data corruptions. */ |
3ba08129e
|
451 452 |
if (vma->vm_mm == t->mm) add_to_kill(t, page, vma, to_kill, tkc); |
6a46079cf
|
453 454 |
} } |
6a46079cf
|
455 |
read_unlock(&tasklist_lock); |
d28eb9c86
|
456 |
i_mmap_unlock_read(mapping); |
6a46079cf
|
457 458 459 460 461 462 463 464 |
} /* * Collect the processes who have the corrupted page mapped to kill. * This is done in two steps for locking reasons. * First preallocate one tokill structure outside the spin locks, * so that we can kill at least one process reasonably reliable. */ |
74614de17
|
465 466 |
static void collect_procs(struct page *page, struct list_head *tokill, int force_early) |
6a46079cf
|
467 468 469 470 471 472 473 474 475 476 |
{ struct to_kill *tk; if (!page->mapping) return; tk = kmalloc(sizeof(struct to_kill), GFP_NOIO); if (!tk) return; if (PageAnon(page)) |
74614de17
|
477 |
collect_procs_anon(page, tokill, &tk, force_early); |
6a46079cf
|
478 |
else |
74614de17
|
479 |
collect_procs_file(page, tokill, &tk, force_early); |
6a46079cf
|
480 481 |
kfree(tk); } |
6a46079cf
|
482 |
static const char *action_name[] = { |
cc637b170
|
483 484 485 486 |
[MF_IGNORED] = "Ignored", [MF_FAILED] = "Failed", [MF_DELAYED] = "Delayed", [MF_RECOVERED] = "Recovered", |
64d37a2ba
|
487 488 489 |
}; static const char * const action_page_types[] = { |
cc637b170
|
490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 |
[MF_MSG_KERNEL] = "reserved kernel page", [MF_MSG_KERNEL_HIGH_ORDER] = "high-order kernel page", [MF_MSG_SLAB] = "kernel slab page", [MF_MSG_DIFFERENT_COMPOUND] = "different compound page after locking", [MF_MSG_POISONED_HUGE] = "huge page already hardware poisoned", [MF_MSG_HUGE] = "huge page", [MF_MSG_FREE_HUGE] = "free huge page", [MF_MSG_UNMAP_FAILED] = "unmapping failed page", [MF_MSG_DIRTY_SWAPCACHE] = "dirty swapcache page", [MF_MSG_CLEAN_SWAPCACHE] = "clean swapcache page", [MF_MSG_DIRTY_MLOCKED_LRU] = "dirty mlocked LRU page", [MF_MSG_CLEAN_MLOCKED_LRU] = "clean mlocked LRU page", [MF_MSG_DIRTY_UNEVICTABLE_LRU] = "dirty unevictable LRU page", [MF_MSG_CLEAN_UNEVICTABLE_LRU] = "clean unevictable LRU page", [MF_MSG_DIRTY_LRU] = "dirty LRU page", [MF_MSG_CLEAN_LRU] = "clean LRU page", [MF_MSG_TRUNCATED_LRU] = "already truncated LRU page", [MF_MSG_BUDDY] = "free buddy page", [MF_MSG_BUDDY_2ND] = "free buddy page (2nd try)", [MF_MSG_UNKNOWN] = "unknown page", |
64d37a2ba
|
510 |
}; |
6a46079cf
|
511 |
/* |
dc2a1cbf7
|
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 |
* XXX: It is possible that a page is isolated from LRU cache, * and then kept in swap cache or failed to remove from page cache. * The page count will stop it from being freed by unpoison. * Stress tests should be aware of this memory leak problem. */ static int delete_from_lru_cache(struct page *p) { if (!isolate_lru_page(p)) { /* * Clear sensible page flags, so that the buddy system won't * complain when the page is unpoison-and-freed. */ ClearPageActive(p); ClearPageUnevictable(p); /* * drop the page count elevated by isolate_lru_page() */ |
09cbfeaf1
|
529 |
put_page(p); |
dc2a1cbf7
|
530 531 532 533 534 535 |
return 0; } return -EIO; } /* |
6a46079cf
|
536 537 538 539 540 541 |
* Error hit kernel page. * Do nothing, try to be lucky and not touch this instead. For a few cases we * could be more sophisticated. */ static int me_kernel(struct page *p, unsigned long pfn) { |
cc637b170
|
542 |
return MF_IGNORED; |
6a46079cf
|
543 544 545 546 547 548 549 |
} /* * Page in unknown state. Do nothing. */ static int me_unknown(struct page *p, unsigned long pfn) { |
495367c05
|
550 551 |
pr_err("Memory failure: %#lx: Unknown page state ", pfn); |
cc637b170
|
552 |
return MF_FAILED; |
6a46079cf
|
553 554 555 |
} /* |
6a46079cf
|
556 557 558 559 560 |
* Clean (or cleaned) page cache page. */ static int me_pagecache_clean(struct page *p, unsigned long pfn) { int err; |
cc637b170
|
561 |
int ret = MF_FAILED; |
6a46079cf
|
562 |
struct address_space *mapping; |
dc2a1cbf7
|
563 |
delete_from_lru_cache(p); |
6a46079cf
|
564 565 566 567 568 |
/* * For anonymous pages we're done the only reference left * should be the one m_f() holds. */ if (PageAnon(p)) |
cc637b170
|
569 |
return MF_RECOVERED; |
6a46079cf
|
570 571 572 573 574 575 576 577 578 579 580 581 582 |
/* * Now truncate the page in the page cache. This is really * more like a "temporary hole punch" * Don't do this for block devices when someone else * has a reference, because it could be file system metadata * and that's not safe to truncate. */ mapping = page_mapping(p); if (!mapping) { /* * Page has been teared down in the meanwhile */ |
cc637b170
|
583 |
return MF_FAILED; |
6a46079cf
|
584 585 586 587 588 589 590 591 592 593 |
} /* * Truncation is a bit tricky. Enable it per file system for now. * * Open: to take i_mutex or not for this? Right now we don't. */ if (mapping->a_ops->error_remove_page) { err = mapping->a_ops->error_remove_page(mapping, p); if (err != 0) { |
495367c05
|
594 595 |
pr_info("Memory failure: %#lx: Failed to punch page: %d ", |
1170532bb
|
596 |
pfn, err); |
6a46079cf
|
597 598 |
} else if (page_has_private(p) && !try_to_release_page(p, GFP_NOIO)) { |
495367c05
|
599 600 601 |
pr_info("Memory failure: %#lx: failed to release buffers ", pfn); |
6a46079cf
|
602 |
} else { |
cc637b170
|
603 |
ret = MF_RECOVERED; |
6a46079cf
|
604 605 606 607 608 609 610 |
} } else { /* * If the file system doesn't support it just invalidate * This fails on dirty or anything with private pages */ if (invalidate_inode_page(p)) |
cc637b170
|
611 |
ret = MF_RECOVERED; |
6a46079cf
|
612 |
else |
495367c05
|
613 614 615 |
pr_info("Memory failure: %#lx: Failed to invalidate ", pfn); |
6a46079cf
|
616 617 618 619 620 |
} return ret; } /* |
549543dff
|
621 |
* Dirty pagecache page |
6a46079cf
|
622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 |
* Issues: when the error hit a hole page the error is not properly * propagated. */ static int me_pagecache_dirty(struct page *p, unsigned long pfn) { struct address_space *mapping = page_mapping(p); SetPageError(p); /* TBD: print more information about the file. */ if (mapping) { /* * IO error will be reported by write(), fsync(), etc. * who check the mapping. * This way the application knows that something went * wrong with its dirty file data. * * There's one open issue: * * The EIO will be only reported on the next IO * operation and then cleared through the IO map. * Normally Linux has two mechanisms to pass IO error * first through the AS_EIO flag in the address space * and then through the PageError flag in the page. * Since we drop pages on memory failure handling the * only mechanism open to use is through AS_AIO. * * This has the disadvantage that it gets cleared on * the first operation that returns an error, while * the PageError bit is more sticky and only cleared * when the page is reread or dropped. If an * application assumes it will always get error on * fsync, but does other operations on the fd before |
25985edce
|
654 |
* and the page is dropped between then the error |
6a46079cf
|
655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 |
* will not be properly reported. * * This can already happen even without hwpoisoned * pages: first on metadata IO errors (which only * report through AS_EIO) or when the page is dropped * at the wrong time. * * So right now we assume that the application DTRT on * the first EIO, but we're not worse than other parts * of the kernel. */ mapping_set_error(mapping, EIO); } return me_pagecache_clean(p, pfn); } /* * Clean and dirty swap cache. * * Dirty swap cache page is tricky to handle. The page could live both in page * cache and swap cache(ie. page is freshly swapped in). So it could be * referenced concurrently by 2 types of PTEs: * normal PTEs and swap PTEs. We try to handle them consistently by calling * try_to_unmap(TTU_IGNORE_HWPOISON) to convert the normal PTEs to swap PTEs, * and then * - clear dirty bit to prevent IO * - remove from LRU * - but keep in the swap cache, so that when we return to it on * a later page fault, we know the application is accessing * corrupted data and shall be killed (we installed simple * interception code in do_swap_page to catch it). * * Clean swap cache pages can be directly isolated. A later page fault will * bring in the known good data from disk. */ static int me_swapcache_dirty(struct page *p, unsigned long pfn) { |
6a46079cf
|
693 694 695 |
ClearPageDirty(p); /* Trigger EIO in shmem: */ ClearPageUptodate(p); |
dc2a1cbf7
|
696 |
if (!delete_from_lru_cache(p)) |
cc637b170
|
697 |
return MF_DELAYED; |
dc2a1cbf7
|
698 |
else |
cc637b170
|
699 |
return MF_FAILED; |
6a46079cf
|
700 701 702 703 |
} static int me_swapcache_clean(struct page *p, unsigned long pfn) { |
6a46079cf
|
704 |
delete_from_swap_cache(p); |
e43c3afb3
|
705 |
|
dc2a1cbf7
|
706 |
if (!delete_from_lru_cache(p)) |
cc637b170
|
707 |
return MF_RECOVERED; |
dc2a1cbf7
|
708 |
else |
cc637b170
|
709 |
return MF_FAILED; |
6a46079cf
|
710 711 712 713 714 |
} /* * Huge pages. Needs work. * Issues: |
93f70f900
|
715 716 |
* - Error on hugepage is contained in hugepage unit (not in raw page unit.) * To narrow down kill region to one page, we need to break up pmd. |
6a46079cf
|
717 718 719 |
*/ static int me_huge_page(struct page *p, unsigned long pfn) { |
6de2b1aab
|
720 |
int res = 0; |
93f70f900
|
721 |
struct page *hpage = compound_head(p); |
2491ffee9
|
722 723 724 |
if (!PageHuge(hpage)) return MF_DELAYED; |
93f70f900
|
725 726 727 728 729 730 731 |
/* * We can safely recover from error on free or reserved (i.e. * not in-use) hugepage by dequeuing it from freelist. * To check whether a hugepage is in-use or not, we can't use * page->lru because it can be used in other hugepage operations, * such as __unmap_hugepage_range() and gather_surplus_pages(). * So instead we use page_mapping() and PageAnon(). |
93f70f900
|
732 733 |
*/ if (!(page_mapping(hpage) || PageAnon(hpage))) { |
6de2b1aab
|
734 735 |
res = dequeue_hwpoisoned_huge_page(hpage); if (!res) |
cc637b170
|
736 |
return MF_RECOVERED; |
93f70f900
|
737 |
} |
cc637b170
|
738 |
return MF_DELAYED; |
6a46079cf
|
739 740 741 742 743 744 745 746 747 |
} /* * Various page states we can handle. * * A page state is defined by its current page->flags bits. * The table matches them in order and calls the right handler. * * This is quite tricky because we can access page at any time |
25985edce
|
748 |
* in its live cycle, so all accesses have to be extremely careful. |
6a46079cf
|
749 750 751 752 753 754 755 756 757 758 759 760 761 |
* * This is not complete. More states could be added. * For any missing state don't attempt recovery. */ #define dirty (1UL << PG_dirty) #define sc (1UL << PG_swapcache) #define unevict (1UL << PG_unevictable) #define mlock (1UL << PG_mlocked) #define writeback (1UL << PG_writeback) #define lru (1UL << PG_lru) #define swapbacked (1UL << PG_swapbacked) #define head (1UL << PG_head) |
6a46079cf
|
762 |
#define slab (1UL << PG_slab) |
6a46079cf
|
763 764 765 766 767 |
#define reserved (1UL << PG_reserved) static struct page_state { unsigned long mask; unsigned long res; |
cc637b170
|
768 |
enum mf_action_page_type type; |
6a46079cf
|
769 770 |
int (*action)(struct page *p, unsigned long pfn); } error_states[] = { |
cc637b170
|
771 |
{ reserved, reserved, MF_MSG_KERNEL, me_kernel }, |
95d01fc66
|
772 773 774 775 |
/* * free pages are specially detected outside this table: * PG_buddy pages only make a small fraction of all free pages. */ |
6a46079cf
|
776 777 778 779 780 781 |
/* * Could in theory check if slab page is free or if we can drop * currently unused objects without touching them. But just * treat it as standard kernel for now. */ |
cc637b170
|
782 |
{ slab, slab, MF_MSG_SLAB, me_kernel }, |
6a46079cf
|
783 |
|
cc637b170
|
784 |
{ head, head, MF_MSG_HUGE, me_huge_page }, |
6a46079cf
|
785 |
|
cc637b170
|
786 787 |
{ sc|dirty, sc|dirty, MF_MSG_DIRTY_SWAPCACHE, me_swapcache_dirty }, { sc|dirty, sc, MF_MSG_CLEAN_SWAPCACHE, me_swapcache_clean }, |
6a46079cf
|
788 |
|
cc637b170
|
789 790 |
{ mlock|dirty, mlock|dirty, MF_MSG_DIRTY_MLOCKED_LRU, me_pagecache_dirty }, { mlock|dirty, mlock, MF_MSG_CLEAN_MLOCKED_LRU, me_pagecache_clean }, |
6a46079cf
|
791 |
|
cc637b170
|
792 793 |
{ unevict|dirty, unevict|dirty, MF_MSG_DIRTY_UNEVICTABLE_LRU, me_pagecache_dirty }, { unevict|dirty, unevict, MF_MSG_CLEAN_UNEVICTABLE_LRU, me_pagecache_clean }, |
5f4b9fc5c
|
794 |
|
cc637b170
|
795 796 |
{ lru|dirty, lru|dirty, MF_MSG_DIRTY_LRU, me_pagecache_dirty }, { lru|dirty, lru, MF_MSG_CLEAN_LRU, me_pagecache_clean }, |
6a46079cf
|
797 798 799 800 |
/* * Catchall entry: must be at end. */ |
cc637b170
|
801 |
{ 0, 0, MF_MSG_UNKNOWN, me_unknown }, |
6a46079cf
|
802 |
}; |
2326c467d
|
803 804 805 806 807 808 809 810 |
#undef dirty #undef sc #undef unevict #undef mlock #undef writeback #undef lru #undef swapbacked #undef head |
2326c467d
|
811 812 |
#undef slab #undef reserved |
ff604cf6d
|
813 814 815 816 |
/* * "Dirty/Clean" indication is not 100% accurate due to the possibility of * setting PG_dirty outside page lock. See also comment above set_page_dirty(). */ |
cc3e2af42
|
817 818 |
static void action_result(unsigned long pfn, enum mf_action_page_type type, enum mf_result result) |
6a46079cf
|
819 |
{ |
97f0b1345
|
820 |
trace_memory_failure_event(pfn, type, result); |
495367c05
|
821 822 |
pr_err("Memory failure: %#lx: recovery action for %s: %s ", |
64d37a2ba
|
823 |
pfn, action_page_types[type], action_name[result]); |
6a46079cf
|
824 825 826 |
} static int page_action(struct page_state *ps, struct page *p, |
bd1ce5f91
|
827 |
unsigned long pfn) |
6a46079cf
|
828 829 |
{ int result; |
7456b0405
|
830 |
int count; |
6a46079cf
|
831 832 |
result = ps->action(p, pfn); |
7456b0405
|
833 |
|
bd1ce5f91
|
834 |
count = page_count(p) - 1; |
cc637b170
|
835 |
if (ps->action == me_swapcache_dirty && result == MF_DELAYED) |
138ce286e
|
836 837 |
count--; if (count != 0) { |
495367c05
|
838 839 |
pr_err("Memory failure: %#lx: %s still referenced by %d users ", |
64d37a2ba
|
840 |
pfn, action_page_types[ps->type], count); |
cc637b170
|
841 |
result = MF_FAILED; |
138ce286e
|
842 |
} |
64d37a2ba
|
843 |
action_result(pfn, ps->type, result); |
6a46079cf
|
844 845 846 847 848 |
/* Could do more checks here if page looks ok */ /* * Could adjust zone counters here to correct for the missing page. */ |
cc637b170
|
849 |
return (result == MF_RECOVERED || result == MF_DELAYED) ? 0 : -EBUSY; |
6a46079cf
|
850 |
} |
ead07f6a8
|
851 852 853 854 855 856 857 858 859 860 |
/** * get_hwpoison_page() - Get refcount for memory error handling: * @page: raw error page (hit by memory error) * * Return: return 0 if failed to grab the refcount, otherwise true (some * non-zero value.) */ int get_hwpoison_page(struct page *page) { struct page *head = compound_head(page); |
4e41a30c6
|
861 |
if (!PageHuge(head) && PageTransHuge(head)) { |
98ed2b005
|
862 863 864 865 866 867 868 |
/* * Non anonymous thp exists only in allocation/free time. We * can't handle such a case correctly, so let's give it up. * This should be better than triggering BUG_ON when kernel * tries to touch the "partially handled" page. */ if (!PageAnon(head)) { |
495367c05
|
869 870 |
pr_err("Memory failure: %#lx: non anonymous thp ", |
98ed2b005
|
871 872 873 |
page_to_pfn(page)); return 0; } |
ead07f6a8
|
874 |
} |
c2e7e00b7
|
875 876 877 |
if (get_page_unless_zero(head)) { if (head == compound_head(page)) return 1; |
495367c05
|
878 879 880 |
pr_info("Memory failure: %#lx cannot catch tail ", page_to_pfn(page)); |
c2e7e00b7
|
881 882 883 884 |
put_page(head); } return 0; |
ead07f6a8
|
885 886 |
} EXPORT_SYMBOL_GPL(get_hwpoison_page); |
6a46079cf
|
887 888 889 890 |
/* * Do all that is necessary to remove user space mappings. Unmap * the pages and send SIGBUS to the processes if the data was dirty. */ |
1668bfd5b
|
891 |
static int hwpoison_user_mappings(struct page *p, unsigned long pfn, |
54b9dd14d
|
892 |
int trapno, int flags, struct page **hpagep) |
6a46079cf
|
893 894 895 896 897 |
{ enum ttu_flags ttu = TTU_UNMAP | TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS; struct address_space *mapping; LIST_HEAD(tokill); int ret; |
6751ed65d
|
898 |
int kill = 1, forcekill; |
54b9dd14d
|
899 |
struct page *hpage = *hpagep; |
6a46079cf
|
900 |
|
93a9eb39f
|
901 902 903 904 905 906 907 |
/* * Here we are interested only in user-mapped pages, so skip any * other types of pages. */ if (PageReserved(p) || PageSlab(p)) return SWAP_SUCCESS; if (!(PageLRU(hpage) || PageHuge(p))) |
1668bfd5b
|
908 |
return SWAP_SUCCESS; |
6a46079cf
|
909 |
|
6a46079cf
|
910 911 912 913 |
/* * This check implies we don't kill processes if their pages * are in the swap cache early. Those are always late kills. */ |
7af446a84
|
914 |
if (!page_mapped(hpage)) |
1668bfd5b
|
915 |
return SWAP_SUCCESS; |
52089b14c
|
916 |
if (PageKsm(p)) { |
495367c05
|
917 918 |
pr_err("Memory failure: %#lx: can't handle KSM pages. ", pfn); |
1668bfd5b
|
919 |
return SWAP_FAIL; |
52089b14c
|
920 |
} |
6a46079cf
|
921 922 |
if (PageSwapCache(p)) { |
495367c05
|
923 924 925 |
pr_err("Memory failure: %#lx: keeping poisoned page in swap cache ", pfn); |
6a46079cf
|
926 927 928 929 930 931 |
ttu |= TTU_IGNORE_HWPOISON; } /* * Propagate the dirty bit from PTEs to struct page first, because we * need this to decide if we should kill or just drop the page. |
db0480b3a
|
932 933 |
* XXX: the dirty test could be racy: set_page_dirty() may not always * be called inside page lock (it's recommended but not enforced). |
6a46079cf
|
934 |
*/ |
7af446a84
|
935 |
mapping = page_mapping(hpage); |
6751ed65d
|
936 |
if (!(flags & MF_MUST_KILL) && !PageDirty(hpage) && mapping && |
7af446a84
|
937 938 939 |
mapping_cap_writeback_dirty(mapping)) { if (page_mkclean(hpage)) { SetPageDirty(hpage); |
6a46079cf
|
940 941 942 |
} else { kill = 0; ttu |= TTU_IGNORE_HWPOISON; |
495367c05
|
943 944 |
pr_info("Memory failure: %#lx: corrupted page was clean: dropped without side effects ", |
6a46079cf
|
945 946 947 |
pfn); } } |
a6d30ddda
|
948 |
/* |
6a46079cf
|
949 950 951 952 953 954 955 956 |
* First collect all the processes that have the page * mapped in dirty form. This has to be done before try_to_unmap, * because ttu takes the rmap data structures down. * * Error handling: We ignore errors here because * there's nothing that can be done. */ if (kill) |
415c64c14
|
957 |
collect_procs(hpage, &tokill, flags & MF_ACTION_REQUIRED); |
6a46079cf
|
958 |
|
415c64c14
|
959 |
ret = try_to_unmap(hpage, ttu); |
6a46079cf
|
960 |
if (ret != SWAP_SUCCESS) |
495367c05
|
961 962 |
pr_err("Memory failure: %#lx: failed to unmap page (mapcount=%d) ", |
1170532bb
|
963 |
pfn, page_mapcount(hpage)); |
a6d30ddda
|
964 |
|
6a46079cf
|
965 966 967 968 |
/* * Now that the dirty bit has been propagated to the * struct page and all unmaps done we can decide if * killing is needed or not. Only kill when the page |
6751ed65d
|
969 970 |
* was dirty or the process is not restartable, * otherwise the tokill list is merely |
6a46079cf
|
971 972 973 974 |
* freed. When there was a problem unmapping earlier * use a more force-full uncatchable kill to prevent * any accesses to the poisoned memory. */ |
415c64c14
|
975 |
forcekill = PageDirty(hpage) || (flags & MF_MUST_KILL); |
6751ed65d
|
976 |
kill_procs(&tokill, forcekill, trapno, |
7329bbeb9
|
977 |
ret != SWAP_SUCCESS, p, pfn, flags); |
1668bfd5b
|
978 979 |
return ret; |
6a46079cf
|
980 |
} |
7013febc8
|
981 982 983 |
static void set_page_hwpoison_huge_page(struct page *hpage) { int i; |
f9121153f
|
984 |
int nr_pages = 1 << compound_order(hpage); |
7013febc8
|
985 986 987 988 989 990 991 |
for (i = 0; i < nr_pages; i++) SetPageHWPoison(hpage + i); } static void clear_page_hwpoison_huge_page(struct page *hpage) { int i; |
f9121153f
|
992 |
int nr_pages = 1 << compound_order(hpage); |
7013febc8
|
993 994 995 |
for (i = 0; i < nr_pages; i++) ClearPageHWPoison(hpage + i); } |
cd42f4a3b
|
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 |
/** * memory_failure - Handle memory failure of a page. * @pfn: Page Number of the corrupted page * @trapno: Trap number reported in the signal to user space. * @flags: fine tune action taken * * This function is called by the low level machine check code * of an architecture when it detects hardware memory corruption * of a page. It tries its best to recover, which includes * dropping pages, killing processes etc. * * The function is primarily of use for corruptions that * happen outside the current execution context (e.g. when * detected by a background scrubber) * * Must run in process context (e.g. a work queue) with interrupts * enabled and no spinlocks hold. */ int memory_failure(unsigned long pfn, int trapno, int flags) |
6a46079cf
|
1015 1016 1017 |
{ struct page_state *ps; struct page *p; |
7af446a84
|
1018 |
struct page *hpage; |
415c64c14
|
1019 |
struct page *orig_head; |
6a46079cf
|
1020 |
int res; |
c9fbdd5f1
|
1021 |
unsigned int nr_pages; |
524fca1e7
|
1022 |
unsigned long page_flags; |
6a46079cf
|
1023 1024 1025 1026 1027 |
if (!sysctl_memory_failure_recovery) panic("Memory failure from trap %d on page %lx", trapno, pfn); if (!pfn_valid(pfn)) { |
495367c05
|
1028 1029 1030 |
pr_err("Memory failure: %#lx: memory outside kernel control ", pfn); |
a7560fc80
|
1031 |
return -ENXIO; |
6a46079cf
|
1032 1033 1034 |
} p = pfn_to_page(pfn); |
415c64c14
|
1035 |
orig_head = hpage = compound_head(p); |
6a46079cf
|
1036 |
if (TestSetPageHWPoison(p)) { |
495367c05
|
1037 1038 1039 |
pr_err("Memory failure: %#lx: already hardware poisoned ", pfn); |
6a46079cf
|
1040 1041 |
return 0; } |
4db0e950c
|
1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 |
/* * Currently errors on hugetlbfs pages are measured in hugepage units, * so nr_pages should be 1 << compound_order. OTOH when errors are on * transparent hugepages, they are supposed to be split and error * measurement is done in normal page units. So nr_pages should be one * in this case. */ if (PageHuge(p)) nr_pages = 1 << compound_order(hpage); else /* normal page or thp */ nr_pages = 1; |
8e30456b6
|
1053 |
num_poisoned_pages_add(nr_pages); |
6a46079cf
|
1054 1055 1056 1057 1058 |
/* * We need/can do nothing about count=0 pages. * 1) it's a free page, and therefore in safe hand: * prep_new_page() will be the gate keeper. |
8c6c2ecb4
|
1059 1060 1061 1062 |
* 2) it's a free hugepage, which is also safe: * an affected hugepage will be dequeued from hugepage freelist, * so there's no concern about reusing it ever after. * 3) it's part of a non-compound high order page. |
6a46079cf
|
1063 1064 1065 1066 1067 1068 |
* Implies some kernel user: cannot stop them from * R/W the page; let's pray that the page has been * used and will be freed some time later. * In fact it's dangerous to directly bump up page count from 0, * that may make page_freeze_refs()/page_unfreeze_refs() mismatch. */ |
ead07f6a8
|
1069 |
if (!(flags & MF_COUNT_INCREASED) && !get_hwpoison_page(p)) { |
8d22ba1b7
|
1070 |
if (is_free_buddy_page(p)) { |
cc637b170
|
1071 |
action_result(pfn, MF_MSG_BUDDY, MF_DELAYED); |
8d22ba1b7
|
1072 |
return 0; |
8c6c2ecb4
|
1073 1074 |
} else if (PageHuge(hpage)) { /* |
b985194c8
|
1075 |
* Check "filter hit" and "race with other subpage." |
8c6c2ecb4
|
1076 |
*/ |
7eaceacca
|
1077 |
lock_page(hpage); |
b985194c8
|
1078 1079 1080 |
if (PageHWPoison(hpage)) { if ((hwpoison_filter(p) && TestClearPageHWPoison(p)) || (p != hpage && TestSetPageHWPoison(hpage))) { |
8e30456b6
|
1081 |
num_poisoned_pages_sub(nr_pages); |
b985194c8
|
1082 1083 1084 |
unlock_page(hpage); return 0; } |
8c6c2ecb4
|
1085 1086 1087 |
} set_page_hwpoison_huge_page(hpage); res = dequeue_hwpoisoned_huge_page(hpage); |
cc637b170
|
1088 1089 |
action_result(pfn, MF_MSG_FREE_HUGE, res ? MF_IGNORED : MF_DELAYED); |
8c6c2ecb4
|
1090 1091 |
unlock_page(hpage); return res; |
8d22ba1b7
|
1092 |
} else { |
cc637b170
|
1093 |
action_result(pfn, MF_MSG_KERNEL_HIGH_ORDER, MF_IGNORED); |
8d22ba1b7
|
1094 1095 |
return -EBUSY; } |
6a46079cf
|
1096 |
} |
415c64c14
|
1097 |
if (!PageHuge(p) && PageTransHuge(hpage)) { |
4d2fa9654
|
1098 |
lock_page(hpage); |
7f6bf39bb
|
1099 |
if (!PageAnon(hpage) || unlikely(split_huge_page(hpage))) { |
4d2fa9654
|
1100 |
unlock_page(hpage); |
7f6bf39bb
|
1101 |
if (!PageAnon(hpage)) |
495367c05
|
1102 1103 1104 |
pr_err("Memory failure: %#lx: non anonymous thp ", pfn); |
7f6bf39bb
|
1105 |
else |
495367c05
|
1106 1107 1108 |
pr_err("Memory failure: %#lx: thp split failed ", pfn); |
ead07f6a8
|
1109 |
if (TestClearPageHWPoison(p)) |
8e30456b6
|
1110 |
num_poisoned_pages_sub(nr_pages); |
665d9da7f
|
1111 |
put_hwpoison_page(p); |
415c64c14
|
1112 1113 |
return -EBUSY; } |
4d2fa9654
|
1114 |
unlock_page(hpage); |
4e41a30c6
|
1115 1116 |
get_hwpoison_page(p); put_hwpoison_page(hpage); |
415c64c14
|
1117 1118 1119 |
VM_BUG_ON_PAGE(!page_count(p), p); hpage = compound_head(p); } |
6a46079cf
|
1120 |
/* |
e43c3afb3
|
1121 1122 |
* We ignore non-LRU pages for good reasons. * - PG_locked is only well defined for LRU pages and a few others |
48c935ad8
|
1123 |
* - to avoid races with __SetPageLocked() |
e43c3afb3
|
1124 1125 1126 1127 |
* - to avoid races with __SetPageSlab*() (and more non-atomic ops) * The check (unnecessarily) ignores LRU pages being isolated and * walked by the page reclaim code, however that's not a big loss. */ |
09789e5de
|
1128 |
if (!PageHuge(p)) { |
415c64c14
|
1129 1130 1131 |
if (!PageLRU(p)) shake_page(p, 0); if (!PageLRU(p)) { |
af241a083
|
1132 1133 1134 1135 |
/* * shake_page could have turned it free. */ if (is_free_buddy_page(p)) { |
2d421acd1
|
1136 |
if (flags & MF_COUNT_INCREASED) |
cc637b170
|
1137 |
action_result(pfn, MF_MSG_BUDDY, MF_DELAYED); |
2d421acd1
|
1138 |
else |
cc637b170
|
1139 1140 |
action_result(pfn, MF_MSG_BUDDY_2ND, MF_DELAYED); |
af241a083
|
1141 1142 |
return 0; } |
0474a60ec
|
1143 |
} |
e43c3afb3
|
1144 |
} |
e43c3afb3
|
1145 |
|
7eaceacca
|
1146 |
lock_page(hpage); |
847ce401d
|
1147 1148 |
/* |
f37d4298a
|
1149 1150 1151 |
* The page could have changed compound pages during the locking. * If this happens just bail out. */ |
415c64c14
|
1152 |
if (PageCompound(p) && compound_head(p) != orig_head) { |
cc637b170
|
1153 |
action_result(pfn, MF_MSG_DIFFERENT_COMPOUND, MF_IGNORED); |
f37d4298a
|
1154 1155 1156 1157 1158 |
res = -EBUSY; goto out; } /* |
524fca1e7
|
1159 1160 1161 1162 1163 1164 1165 1166 1167 |
* We use page flags to determine what action should be taken, but * the flags can be modified by the error containment action. One * example is an mlocked page, where PG_mlocked is cleared by * page_remove_rmap() in try_to_unmap_one(). So to determine page status * correctly, we save a copy of the page flags at this time. */ page_flags = p->flags; /* |
847ce401d
|
1168 1169 1170 |
* unpoison always clear PG_hwpoison inside page lock */ if (!PageHWPoison(p)) { |
495367c05
|
1171 1172 |
pr_err("Memory failure: %#lx: just unpoisoned ", pfn); |
8e30456b6
|
1173 |
num_poisoned_pages_sub(nr_pages); |
a09233f3e
|
1174 |
unlock_page(hpage); |
665d9da7f
|
1175 |
put_hwpoison_page(hpage); |
a09233f3e
|
1176 |
return 0; |
847ce401d
|
1177 |
} |
7c116f2b0
|
1178 1179 |
if (hwpoison_filter(p)) { if (TestClearPageHWPoison(p)) |
8e30456b6
|
1180 |
num_poisoned_pages_sub(nr_pages); |
7af446a84
|
1181 |
unlock_page(hpage); |
665d9da7f
|
1182 |
put_hwpoison_page(hpage); |
7c116f2b0
|
1183 1184 |
return 0; } |
847ce401d
|
1185 |
|
0bc1f8b06
|
1186 1187 |
if (!PageHuge(p) && !PageTransTail(p) && !PageLRU(p)) goto identify_page_state; |
7013febc8
|
1188 1189 1190 1191 |
/* * For error on the tail page, we should set PG_hwpoison * on the head page to show that the hugepage is hwpoisoned */ |
a6d30ddda
|
1192 |
if (PageHuge(p) && PageTail(p) && TestSetPageHWPoison(hpage)) { |
cc637b170
|
1193 |
action_result(pfn, MF_MSG_POISONED_HUGE, MF_IGNORED); |
7013febc8
|
1194 |
unlock_page(hpage); |
665d9da7f
|
1195 |
put_hwpoison_page(hpage); |
7013febc8
|
1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 |
return 0; } /* * Set PG_hwpoison on all pages in an error hugepage, * because containment is done in hugepage unit for now. * Since we have done TestSetPageHWPoison() for the head page with * page lock held, we can safely set PG_hwpoison bits on tail pages. */ if (PageHuge(p)) set_page_hwpoison_huge_page(hpage); |
6edd6cc66
|
1206 1207 1208 1209 |
/* * It's very difficult to mess with pages currently under IO * and in many cases impossible, so we just avoid it here. */ |
6a46079cf
|
1210 1211 1212 1213 |
wait_on_page_writeback(p); /* * Now take care of user space mappings. |
e64a782fe
|
1214 |
* Abort on fail: __delete_from_page_cache() assumes unmapped page. |
54b9dd14d
|
1215 1216 1217 |
* * When the raw error page is thp tail page, hpage points to the raw * page after thp split. |
6a46079cf
|
1218 |
*/ |
54b9dd14d
|
1219 1220 |
if (hwpoison_user_mappings(p, pfn, trapno, flags, &hpage) != SWAP_SUCCESS) { |
cc637b170
|
1221 |
action_result(pfn, MF_MSG_UNMAP_FAILED, MF_IGNORED); |
1668bfd5b
|
1222 1223 1224 |
res = -EBUSY; goto out; } |
6a46079cf
|
1225 1226 1227 1228 |
/* * Torn down by someone else? */ |
dc2a1cbf7
|
1229 |
if (PageLRU(p) && !PageSwapCache(p) && p->mapping == NULL) { |
cc637b170
|
1230 |
action_result(pfn, MF_MSG_TRUNCATED_LRU, MF_IGNORED); |
d95ea51e3
|
1231 |
res = -EBUSY; |
6a46079cf
|
1232 1233 |
goto out; } |
0bc1f8b06
|
1234 |
identify_page_state: |
6a46079cf
|
1235 |
res = -EBUSY; |
524fca1e7
|
1236 1237 1238 1239 1240 1241 1242 |
/* * The first check uses the current page flags which may not have any * relevant information. The second check with the saved page flagss is * carried out only if the first check can't determine the page status. */ for (ps = error_states;; ps++) if ((p->flags & ps->mask) == ps->res) |
6a46079cf
|
1243 |
break; |
841fcc583
|
1244 1245 |
page_flags |= (p->flags & (1UL << PG_dirty)); |
524fca1e7
|
1246 1247 1248 1249 1250 |
if (!ps->mask) for (ps = error_states;; ps++) if ((page_flags & ps->mask) == ps->res) break; res = page_action(ps, p, pfn); |
6a46079cf
|
1251 |
out: |
7af446a84
|
1252 |
unlock_page(hpage); |
6a46079cf
|
1253 1254 |
return res; } |
cd42f4a3b
|
1255 |
EXPORT_SYMBOL_GPL(memory_failure); |
847ce401d
|
1256 |
|
ea8f5fb8a
|
1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 |
#define MEMORY_FAILURE_FIFO_ORDER 4 #define MEMORY_FAILURE_FIFO_SIZE (1 << MEMORY_FAILURE_FIFO_ORDER) struct memory_failure_entry { unsigned long pfn; int trapno; int flags; }; struct memory_failure_cpu { DECLARE_KFIFO(fifo, struct memory_failure_entry, MEMORY_FAILURE_FIFO_SIZE); spinlock_t lock; struct work_struct work; }; static DEFINE_PER_CPU(struct memory_failure_cpu, memory_failure_cpu); /** * memory_failure_queue - Schedule handling memory failure of a page. * @pfn: Page Number of the corrupted page * @trapno: Trap number reported in the signal to user space. * @flags: Flags for memory failure handling * * This function is called by the low level hardware error handler * when it detects hardware memory corruption of a page. It schedules * the recovering of error page, including dropping pages, killing * processes etc. * * The function is primarily of use for corruptions that * happen outside the current execution context (e.g. when * detected by a background scrubber) * * Can run in IRQ context. */ void memory_failure_queue(unsigned long pfn, int trapno, int flags) { struct memory_failure_cpu *mf_cpu; unsigned long proc_flags; struct memory_failure_entry entry = { .pfn = pfn, .trapno = trapno, .flags = flags, }; mf_cpu = &get_cpu_var(memory_failure_cpu); spin_lock_irqsave(&mf_cpu->lock, proc_flags); |
498d319bb
|
1304 |
if (kfifo_put(&mf_cpu->fifo, entry)) |
ea8f5fb8a
|
1305 1306 |
schedule_work_on(smp_processor_id(), &mf_cpu->work); else |
8e33a52fa
|
1307 1308 |
pr_err("Memory failure: buffer overflow when queuing memory failure at %#lx ", |
ea8f5fb8a
|
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 |
pfn); spin_unlock_irqrestore(&mf_cpu->lock, proc_flags); put_cpu_var(memory_failure_cpu); } EXPORT_SYMBOL_GPL(memory_failure_queue); static void memory_failure_work_func(struct work_struct *work) { struct memory_failure_cpu *mf_cpu; struct memory_failure_entry entry = { 0, }; unsigned long proc_flags; int gotten; |
7c8e0181e
|
1321 |
mf_cpu = this_cpu_ptr(&memory_failure_cpu); |
ea8f5fb8a
|
1322 1323 1324 1325 1326 1327 |
for (;;) { spin_lock_irqsave(&mf_cpu->lock, proc_flags); gotten = kfifo_get(&mf_cpu->fifo, &entry); spin_unlock_irqrestore(&mf_cpu->lock, proc_flags); if (!gotten) break; |
cf870c70a
|
1328 1329 1330 1331 |
if (entry.flags & MF_SOFT_OFFLINE) soft_offline_page(pfn_to_page(entry.pfn), entry.flags); else memory_failure(entry.pfn, entry.trapno, entry.flags); |
ea8f5fb8a
|
1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 |
} } static int __init memory_failure_init(void) { struct memory_failure_cpu *mf_cpu; int cpu; for_each_possible_cpu(cpu) { mf_cpu = &per_cpu(memory_failure_cpu, cpu); spin_lock_init(&mf_cpu->lock); INIT_KFIFO(mf_cpu->fifo); INIT_WORK(&mf_cpu->work, memory_failure_work_func); } return 0; } core_initcall(memory_failure_init); |
a5f651090
|
1350 1351 1352 1353 1354 |
#define unpoison_pr_info(fmt, pfn, rs) \ ({ \ if (__ratelimit(rs)) \ pr_info(fmt, pfn); \ }) |
847ce401d
|
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 |
/** * unpoison_memory - Unpoison a previously poisoned page * @pfn: Page number of the to be unpoisoned page * * Software-unpoison a page that has been poisoned by * memory_failure() earlier. * * This is only done on the software-level, so it only works * for linux injected failures, not real hardware failures * * Returns 0 for success, otherwise -errno. */ int unpoison_memory(unsigned long pfn) { struct page *page; struct page *p; int freeit = 0; |
c9fbdd5f1
|
1372 |
unsigned int nr_pages; |
a5f651090
|
1373 1374 |
static DEFINE_RATELIMIT_STATE(unpoison_rs, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); |
847ce401d
|
1375 1376 1377 1378 1379 1380 1381 1382 |
if (!pfn_valid(pfn)) return -ENXIO; p = pfn_to_page(pfn); page = compound_head(p); if (!PageHWPoison(p)) { |
495367c05
|
1383 1384 |
unpoison_pr_info("Unpoison: Page was already unpoisoned %#lx ", |
a5f651090
|
1385 |
pfn, &unpoison_rs); |
847ce401d
|
1386 1387 |
return 0; } |
230ac719c
|
1388 |
if (page_count(page) > 1) { |
495367c05
|
1389 1390 |
unpoison_pr_info("Unpoison: Someone grabs the hwpoison page %#lx ", |
a5f651090
|
1391 |
pfn, &unpoison_rs); |
230ac719c
|
1392 1393 1394 1395 |
return 0; } if (page_mapped(page)) { |
495367c05
|
1396 1397 |
unpoison_pr_info("Unpoison: Someone maps the hwpoison page %#lx ", |
a5f651090
|
1398 |
pfn, &unpoison_rs); |
230ac719c
|
1399 1400 1401 1402 |
return 0; } if (page_mapping(page)) { |
495367c05
|
1403 1404 |
unpoison_pr_info("Unpoison: the hwpoison page has non-NULL mapping %#lx ", |
a5f651090
|
1405 |
pfn, &unpoison_rs); |
230ac719c
|
1406 1407 |
return 0; } |
0cea3fdc4
|
1408 1409 1410 1411 1412 |
/* * unpoison_memory() can encounter thp only when the thp is being * worked by memory_failure() and the page lock is not held yet. * In such case, we yield to memory_failure() and make unpoison fail. */ |
e76d30e20
|
1413 |
if (!PageHuge(page) && PageTransHuge(page)) { |
495367c05
|
1414 1415 |
unpoison_pr_info("Unpoison: Memory failure is now running on %#lx ", |
a5f651090
|
1416 |
pfn, &unpoison_rs); |
ead07f6a8
|
1417 |
return 0; |
0cea3fdc4
|
1418 |
} |
f9121153f
|
1419 |
nr_pages = 1 << compound_order(page); |
c9fbdd5f1
|
1420 |
|
ead07f6a8
|
1421 |
if (!get_hwpoison_page(p)) { |
8c6c2ecb4
|
1422 1423 1424 1425 1426 1427 1428 |
/* * Since HWPoisoned hugepage should have non-zero refcount, * race between memory failure and unpoison seems to happen. * In such case unpoison fails and memory failure runs * to the end. */ if (PageHuge(page)) { |
495367c05
|
1429 1430 |
unpoison_pr_info("Unpoison: Memory failure is now running on free hugepage %#lx ", |
a5f651090
|
1431 |
pfn, &unpoison_rs); |
8c6c2ecb4
|
1432 1433 |
return 0; } |
847ce401d
|
1434 |
if (TestClearPageHWPoison(p)) |
8e30456b6
|
1435 |
num_poisoned_pages_dec(); |
495367c05
|
1436 1437 |
unpoison_pr_info("Unpoison: Software-unpoisoned free page %#lx ", |
a5f651090
|
1438 |
pfn, &unpoison_rs); |
847ce401d
|
1439 1440 |
return 0; } |
7eaceacca
|
1441 |
lock_page(page); |
847ce401d
|
1442 1443 1444 1445 1446 1447 |
/* * This test is racy because PG_hwpoison is set outside of page lock. * That's acceptable because that won't trigger kernel panic. Instead, * the PG_hwpoison page will be caught and isolated on the entrance to * the free buddy page pool. */ |
c9fbdd5f1
|
1448 |
if (TestClearPageHWPoison(page)) { |
495367c05
|
1449 1450 |
unpoison_pr_info("Unpoison: Software-unpoisoned page %#lx ", |
a5f651090
|
1451 |
pfn, &unpoison_rs); |
8e30456b6
|
1452 |
num_poisoned_pages_sub(nr_pages); |
847ce401d
|
1453 |
freeit = 1; |
6a90181c7
|
1454 1455 |
if (PageHuge(page)) clear_page_hwpoison_huge_page(page); |
847ce401d
|
1456 1457 |
} unlock_page(page); |
665d9da7f
|
1458 |
put_hwpoison_page(page); |
3ba5eebc4
|
1459 |
if (freeit && !(pfn == my_zero_pfn(0) && page_count(p) == 1)) |
665d9da7f
|
1460 |
put_hwpoison_page(page); |
847ce401d
|
1461 1462 1463 1464 |
return 0; } EXPORT_SYMBOL(unpoison_memory); |
facb6011f
|
1465 1466 1467 |
static struct page *new_page(struct page *p, unsigned long private, int **x) { |
12686d153
|
1468 |
int nid = page_to_nid(p); |
d950b9588
|
1469 1470 1471 1472 |
if (PageHuge(p)) return alloc_huge_page_node(page_hstate(compound_head(p)), nid); else |
96db800f5
|
1473 |
return __alloc_pages_node(nid, GFP_HIGHUSER_MOVABLE, 0); |
facb6011f
|
1474 1475 1476 1477 1478 1479 1480 1481 |
} /* * Safely get reference count of an arbitrary page. * Returns 0 for a free page, -EIO for a zero refcount page * that is not free, and 1 for any other page type. * For 1 the page is returned with increased page count, otherwise not. */ |
af8fae7c0
|
1482 |
static int __get_any_page(struct page *p, unsigned long pfn, int flags) |
facb6011f
|
1483 1484 1485 1486 1487 1488 1489 |
{ int ret; if (flags & MF_COUNT_INCREASED) return 1; /* |
d950b9588
|
1490 1491 1492 |
* When the target page is a free hugepage, just remove it * from free hugepage list. */ |
ead07f6a8
|
1493 |
if (!get_hwpoison_page(p)) { |
d950b9588
|
1494 |
if (PageHuge(p)) { |
71dd0b8ae
|
1495 1496 |
pr_info("%s: %#lx free huge page ", __func__, pfn); |
af8fae7c0
|
1497 |
ret = 0; |
d950b9588
|
1498 |
} else if (is_free_buddy_page(p)) { |
71dd0b8ae
|
1499 1500 |
pr_info("%s: %#lx free buddy page ", __func__, pfn); |
facb6011f
|
1501 1502 |
ret = 0; } else { |
71dd0b8ae
|
1503 1504 1505 |
pr_info("%s: %#lx: unknown zero refcount page type %lx ", __func__, pfn, p->flags); |
facb6011f
|
1506 1507 1508 1509 1510 1511 |
ret = -EIO; } } else { /* Not a free page */ ret = 1; } |
facb6011f
|
1512 1513 |
return ret; } |
af8fae7c0
|
1514 1515 1516 1517 1518 1519 1520 1521 |
static int get_any_page(struct page *page, unsigned long pfn, int flags) { int ret = __get_any_page(page, pfn, flags); if (ret == 1 && !PageHuge(page) && !PageLRU(page)) { /* * Try to free it. */ |
665d9da7f
|
1522 |
put_hwpoison_page(page); |
af8fae7c0
|
1523 1524 1525 1526 1527 1528 |
shake_page(page, 1); /* * Did it turn free? */ ret = __get_any_page(page, pfn, 0); |
d96b339f4
|
1529 |
if (ret == 1 && !PageLRU(page)) { |
4f32be677
|
1530 |
/* Drop page reference which is from __get_any_page() */ |
665d9da7f
|
1531 |
put_hwpoison_page(page); |
af8fae7c0
|
1532 1533 1534 1535 1536 1537 1538 1539 |
pr_info("soft_offline: %#lx: unknown non LRU page type %lx ", pfn, page->flags); return -EIO; } } return ret; } |
d950b9588
|
1540 1541 1542 1543 1544 |
static int soft_offline_huge_page(struct page *page, int flags) { int ret; unsigned long pfn = page_to_pfn(page); struct page *hpage = compound_head(page); |
b8ec1cee5
|
1545 |
LIST_HEAD(pagelist); |
d950b9588
|
1546 |
|
af8fae7c0
|
1547 1548 1549 1550 1551 |
/* * This double-check of PageHWPoison is to avoid the race with * memory_failure(). See also comment in __soft_offline_page(). */ lock_page(hpage); |
0ebff32c3
|
1552 |
if (PageHWPoison(hpage)) { |
af8fae7c0
|
1553 |
unlock_page(hpage); |
665d9da7f
|
1554 |
put_hwpoison_page(hpage); |
0ebff32c3
|
1555 1556 |
pr_info("soft offline: %#lx hugepage already poisoned ", pfn); |
af8fae7c0
|
1557 |
return -EBUSY; |
0ebff32c3
|
1558 |
} |
af8fae7c0
|
1559 |
unlock_page(hpage); |
d950b9588
|
1560 |
|
bcc542223
|
1561 |
ret = isolate_huge_page(hpage, &pagelist); |
036138080
|
1562 1563 1564 1565 |
/* * get_any_page() and isolate_huge_page() takes a refcount each, * so need to drop one here. */ |
665d9da7f
|
1566 |
put_hwpoison_page(hpage); |
036138080
|
1567 |
if (!ret) { |
bcc542223
|
1568 1569 1570 1571 |
pr_info("soft offline: %#lx hugepage failed to isolate ", pfn); return -EBUSY; } |
68711a746
|
1572 |
ret = migrate_pages(&pagelist, new_page, NULL, MPOL_MF_MOVE_ALL, |
b8ec1cee5
|
1573 |
MIGRATE_SYNC, MR_MEMORY_FAILURE); |
d950b9588
|
1574 |
if (ret) { |
dd73e85f6
|
1575 1576 1577 |
pr_info("soft offline: %#lx: migration failed %d, type %lx ", pfn, ret, page->flags); |
b8ec1cee5
|
1578 1579 1580 1581 1582 1583 1584 1585 |
/* * We know that soft_offline_huge_page() tries to migrate * only one hugepage pointed to by hpage, so we need not * run through the pagelist here. */ putback_active_hugepage(hpage); if (ret > 0) ret = -EIO; |
af8fae7c0
|
1586 |
} else { |
a49ecbcd7
|
1587 1588 1589 1590 |
/* overcommit hugetlb page will be freed to buddy */ if (PageHuge(page)) { set_page_hwpoison_huge_page(hpage); dequeue_hwpoisoned_huge_page(hpage); |
8e30456b6
|
1591 |
num_poisoned_pages_add(1 << compound_order(hpage)); |
a49ecbcd7
|
1592 1593 |
} else { SetPageHWPoison(page); |
8e30456b6
|
1594 |
num_poisoned_pages_inc(); |
a49ecbcd7
|
1595 |
} |
d950b9588
|
1596 |
} |
d950b9588
|
1597 1598 |
return ret; } |
af8fae7c0
|
1599 1600 1601 1602 |
static int __soft_offline_page(struct page *page, int flags) { int ret; unsigned long pfn = page_to_pfn(page); |
facb6011f
|
1603 |
|
facb6011f
|
1604 |
/* |
af8fae7c0
|
1605 1606 1607 1608 |
* Check PageHWPoison again inside page lock because PageHWPoison * is set by memory_failure() outside page lock. Note that * memory_failure() also double-checks PageHWPoison inside page lock, * so there's no race between soft_offline_page() and memory_failure(). |
facb6011f
|
1609 |
*/ |
0ebff32c3
|
1610 1611 |
lock_page(page); wait_on_page_writeback(page); |
af8fae7c0
|
1612 1613 |
if (PageHWPoison(page)) { unlock_page(page); |
665d9da7f
|
1614 |
put_hwpoison_page(page); |
af8fae7c0
|
1615 1616 1617 1618 |
pr_info("soft offline: %#lx page already poisoned ", pfn); return -EBUSY; } |
facb6011f
|
1619 1620 1621 1622 1623 1624 |
/* * Try to invalidate first. This should work for * non dirty unmapped page cache pages. */ ret = invalidate_inode_page(page); unlock_page(page); |
facb6011f
|
1625 |
/* |
facb6011f
|
1626 1627 1628 |
* RED-PEN would be better to keep it isolated here, but we * would need to fix isolation locking first. */ |
facb6011f
|
1629 |
if (ret == 1) { |
665d9da7f
|
1630 |
put_hwpoison_page(page); |
fb46e7352
|
1631 1632 |
pr_info("soft_offline: %#lx: invalidated ", pfn); |
af8fae7c0
|
1633 |
SetPageHWPoison(page); |
8e30456b6
|
1634 |
num_poisoned_pages_inc(); |
af8fae7c0
|
1635 |
return 0; |
facb6011f
|
1636 1637 1638 1639 1640 1641 1642 1643 |
} /* * Simple invalidation didn't work. * Try to migrate to a new page instead. migrate.c * handles a large number of cases for us. */ ret = isolate_lru_page(page); |
bd486285f
|
1644 1645 1646 1647 |
/* * Drop page reference which is came from get_any_page() * successful isolate_lru_page() already took another one. */ |
665d9da7f
|
1648 |
put_hwpoison_page(page); |
facb6011f
|
1649 1650 |
if (!ret) { LIST_HEAD(pagelist); |
599d0c954
|
1651 |
inc_node_page_state(page, NR_ISOLATED_ANON + |
9c620e2bc
|
1652 |
page_is_file_cache(page)); |
facb6011f
|
1653 |
list_add(&page->lru, &pagelist); |
68711a746
|
1654 |
ret = migrate_pages(&pagelist, new_page, NULL, MPOL_MF_MOVE_ALL, |
9c620e2bc
|
1655 |
MIGRATE_SYNC, MR_MEMORY_FAILURE); |
facb6011f
|
1656 |
if (ret) { |
59c82b70d
|
1657 1658 |
if (!list_empty(&pagelist)) { list_del(&page->lru); |
599d0c954
|
1659 |
dec_node_page_state(page, NR_ISOLATED_ANON + |
59c82b70d
|
1660 1661 1662 |
page_is_file_cache(page)); putback_lru_page(page); } |
fb46e7352
|
1663 1664 |
pr_info("soft offline: %#lx: migration failed %d, type %lx ", |
facb6011f
|
1665 1666 1667 1668 1669 |
pfn, ret, page->flags); if (ret > 0) ret = -EIO; } } else { |
fb46e7352
|
1670 1671 |
pr_info("soft offline: %#lx: isolation failed: %d, page count %d, type %lx ", |
dd73e85f6
|
1672 |
pfn, ret, page_count(page), page->flags); |
facb6011f
|
1673 |
} |
facb6011f
|
1674 1675 |
return ret; } |
86e057734
|
1676 |
|
acc14dc4b
|
1677 1678 1679 1680 1681 1682 1683 |
static int soft_offline_in_use_page(struct page *page, int flags) { int ret; struct page *hpage = compound_head(page); if (!PageHuge(page) && PageTransHuge(hpage)) { lock_page(hpage); |
98fd1ef42
|
1684 1685 1686 1687 1688 1689 1690 1691 1692 |
if (!PageAnon(hpage) || unlikely(split_huge_page(hpage))) { unlock_page(hpage); if (!PageAnon(hpage)) pr_info("soft offline: %#lx: non anonymous thp ", page_to_pfn(page)); else pr_info("soft offline: %#lx: thp split failed ", page_to_pfn(page)); put_hwpoison_page(hpage); |
acc14dc4b
|
1693 1694 |
return -EBUSY; } |
98fd1ef42
|
1695 |
unlock_page(hpage); |
acc14dc4b
|
1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 |
get_hwpoison_page(page); put_hwpoison_page(hpage); } if (PageHuge(page)) ret = soft_offline_huge_page(page, flags); else ret = __soft_offline_page(page, flags); return ret; } static void soft_offline_free_page(struct page *page) { if (PageHuge(page)) { struct page *hpage = compound_head(page); set_page_hwpoison_huge_page(hpage); if (!dequeue_hwpoisoned_huge_page(hpage)) num_poisoned_pages_add(1 << compound_order(hpage)); } else { if (!TestSetPageHWPoison(page)) num_poisoned_pages_inc(); } } |
86e057734
|
1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 |
/** * soft_offline_page - Soft offline a page. * @page: page to offline * @flags: flags. Same as memory_failure(). * * Returns 0 on success, otherwise negated errno. * * Soft offline a page, by migration or invalidation, * without killing anything. This is for the case when * a page is not corrupted yet (so it's still valid to access), * but has had a number of corrected errors and is better taken * out. * * The actual policy on when to do that is maintained by * user space. * * This should never impact any application or cause data loss, * however it might take some time. * * This is not a 100% solution for all memory, but tries to be * ``good enough'' for the majority of memory. */ int soft_offline_page(struct page *page, int flags) { int ret; unsigned long pfn = page_to_pfn(page); |
86e057734
|
1747 1748 1749 1750 |
if (PageHWPoison(page)) { pr_info("soft offline: %#lx page already poisoned ", pfn); |
1e0e635be
|
1751 |
if (flags & MF_COUNT_INCREASED) |
665d9da7f
|
1752 |
put_hwpoison_page(page); |
86e057734
|
1753 1754 |
return -EBUSY; } |
86e057734
|
1755 |
|
bfc8c9013
|
1756 |
get_online_mems(); |
86e057734
|
1757 |
ret = get_any_page(page, pfn, flags); |
bfc8c9013
|
1758 |
put_online_mems(); |
4e41a30c6
|
1759 |
|
acc14dc4b
|
1760 1761 1762 1763 |
if (ret > 0) ret = soft_offline_in_use_page(page, flags); else if (ret == 0) soft_offline_free_page(page); |
4e41a30c6
|
1764 |
|
86e057734
|
1765 1766 |
return ret; } |