Blame view
mm/memory-failure.c
52.1 KB
1439f94c5 treewide: Replace... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
6a46079cf HWPOISON: The hig... |
2 3 4 5 |
/* * Copyright (C) 2008, 2009 Intel Corporation * Authors: Andi Kleen, Fengguang Wu * |
6a46079cf HWPOISON: The hig... |
6 |
* High level machine check handler. Handles pages reported by the |
1c80b990a HWPOISON: Improve... |
7 |
* hardware as being corrupted usually due to a multi-bit ECC memory or cache |
6a46079cf HWPOISON: The hig... |
8 |
* failure. |
1c80b990a HWPOISON: Improve... |
9 10 11 |
* * In addition there is a "soft offline" entry point that allows stop using * not-yet-corrupted-by-suspicious pages without killing anything. |
6a46079cf HWPOISON: The hig... |
12 13 |
* * Handles page cache pages in various states. The tricky part |
1c80b990a HWPOISON: Improve... |
14 15 16 17 18 19 |
* 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 mm, hwpoison: add... |
20 21 22 23 24 25 26 27 |
* * 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 HWPOISON: Improve... |
28 29 30 31 32 33 34 |
* * 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 HWPOISON: The hig... |
35 |
*/ |
6a46079cf HWPOISON: The hig... |
36 37 38 |
#include <linux/kernel.h> #include <linux/mm.h> #include <linux/page-flags.h> |
478c5ffc0 HWPOISON: add pag... |
39 |
#include <linux/kernel-page-flags.h> |
3f07c0144 sched/headers: Pr... |
40 |
#include <linux/sched/signal.h> |
299300258 sched/headers: Pr... |
41 |
#include <linux/sched/task.h> |
01e00f880 HWPOISON: fix oop... |
42 |
#include <linux/ksm.h> |
6a46079cf HWPOISON: The hig... |
43 |
#include <linux/rmap.h> |
b9e15bafd mm: Add export.h ... |
44 |
#include <linux/export.h> |
6a46079cf HWPOISON: The hig... |
45 46 47 |
#include <linux/pagemap.h> #include <linux/swap.h> #include <linux/backing-dev.h> |
facb6011f HWPOISON: Add sof... |
48 |
#include <linux/migrate.h> |
facb6011f HWPOISON: Add sof... |
49 |
#include <linux/suspend.h> |
5a0e3ad6a include cleanup: ... |
50 |
#include <linux/slab.h> |
bf998156d KVM: Avoid killin... |
51 |
#include <linux/swapops.h> |
7af446a84 HWPOISON, hugetlb... |
52 |
#include <linux/hugetlb.h> |
20d6c96b5 mem-hotplug: intr... |
53 |
#include <linux/memory_hotplug.h> |
5db8a73a8 mm/memory-failure... |
54 |
#include <linux/mm_inline.h> |
6100e34b2 mm, memory_failur... |
55 |
#include <linux/memremap.h> |
ea8f5fb8a HWPoison: add mem... |
56 |
#include <linux/kfifo.h> |
a5f651090 mm: hwpoison: rat... |
57 |
#include <linux/ratelimit.h> |
d4ae9916e mm: soft-offline:... |
58 |
#include <linux/page-isolation.h> |
6a46079cf HWPOISON: The hig... |
59 |
#include "internal.h" |
97f0b1345 tracing: add trac... |
60 |
#include "ras/ras_event.h" |
6a46079cf HWPOISON: The hig... |
61 62 63 64 |
int sysctl_memory_failure_early_kill __read_mostly = 0; int sysctl_memory_failure_recovery __read_mostly = 1; |
293c07e31 memory-failure: u... |
65 |
atomic_long_t num_poisoned_pages __read_mostly = ATOMIC_LONG_INIT(0); |
6a46079cf HWPOISON: The hig... |
66 |
|
27df5068e HWPOISON: Add PRO... |
67 |
#if defined(CONFIG_HWPOISON_INJECT) || defined(CONFIG_HWPOISON_INJECT_MODULE) |
1bfe5febe HWPOISON: add an ... |
68 |
u32 hwpoison_filter_enable = 0; |
7c116f2b0 HWPOISON: add fs/... |
69 70 |
u32 hwpoison_filter_dev_major = ~0U; u32 hwpoison_filter_dev_minor = ~0U; |
478c5ffc0 HWPOISON: add pag... |
71 72 |
u64 hwpoison_filter_flags_mask; u64 hwpoison_filter_flags_value; |
1bfe5febe HWPOISON: add an ... |
73 |
EXPORT_SYMBOL_GPL(hwpoison_filter_enable); |
7c116f2b0 HWPOISON: add fs/... |
74 75 |
EXPORT_SYMBOL_GPL(hwpoison_filter_dev_major); EXPORT_SYMBOL_GPL(hwpoison_filter_dev_minor); |
478c5ffc0 HWPOISON: add pag... |
76 77 |
EXPORT_SYMBOL_GPL(hwpoison_filter_flags_mask); EXPORT_SYMBOL_GPL(hwpoison_filter_flags_value); |
7c116f2b0 HWPOISON: add fs/... |
78 79 80 81 82 83 84 85 86 87 88 |
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 HWPOISON: Improve... |
89 |
* page_mapping() does not accept slab pages. |
7c116f2b0 HWPOISON: add fs/... |
90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
*/ 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 HWPOISON: add pag... |
108 109 110 111 112 113 114 115 116 117 118 |
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 HWPOISON: add mem... |
119 120 121 122 123 124 125 126 127 128 |
/* * 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 hwpoison: use pag... |
129 |
#ifdef CONFIG_MEMCG |
4fd466eb4 HWPOISON: add mem... |
130 131 132 133 |
u64 hwpoison_filter_memcg; EXPORT_SYMBOL_GPL(hwpoison_filter_memcg); static int hwpoison_filter_task(struct page *p) { |
4fd466eb4 HWPOISON: add mem... |
134 135 |
if (!hwpoison_filter_memcg) return 0; |
94a59fb36 hwpoison: use pag... |
136 |
if (page_cgroup_ino(p) != hwpoison_filter_memcg) |
4fd466eb4 HWPOISON: add mem... |
137 138 139 140 141 142 143 |
return -EINVAL; return 0; } #else static int hwpoison_filter_task(struct page *p) { return 0; } #endif |
7c116f2b0 HWPOISON: add fs/... |
144 145 |
int hwpoison_filter(struct page *p) { |
1bfe5febe HWPOISON: add an ... |
146 147 |
if (!hwpoison_filter_enable) return 0; |
7c116f2b0 HWPOISON: add fs/... |
148 149 |
if (hwpoison_filter_dev(p)) return -EINVAL; |
478c5ffc0 HWPOISON: add pag... |
150 151 |
if (hwpoison_filter_flags(p)) return -EINVAL; |
4fd466eb4 HWPOISON: add mem... |
152 153 |
if (hwpoison_filter_task(p)) return -EINVAL; |
7c116f2b0 HWPOISON: add fs/... |
154 155 |
return 0; } |
27df5068e HWPOISON: Add PRO... |
156 157 158 159 160 161 |
#else int hwpoison_filter(struct page *p) { return 0; } #endif |
7c116f2b0 HWPOISON: add fs/... |
162 |
EXPORT_SYMBOL_GPL(hwpoison_filter); |
6a46079cf HWPOISON: The hig... |
163 |
/* |
ae1139ece mm, memory_failur... |
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 |
* 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; short size_shift; |
ae1139ece mm, memory_failur... |
190 191 192 |
}; /* |
7329bbeb9 HWPOISON: Add cod... |
193 194 195 |
* 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 HWPOISON: The hig... |
196 |
*/ |
ae1139ece mm, memory_failur... |
197 |
static int kill_proc(struct to_kill *tk, unsigned long pfn, int flags) |
6a46079cf HWPOISON: The hig... |
198 |
{ |
ae1139ece mm, memory_failur... |
199 200 |
struct task_struct *t = tk->tsk; short addr_lsb = tk->size_shift; |
6a46079cf HWPOISON: The hig... |
201 |
int ret; |
135e53514 mm/memory-failure... |
202 203 |
pr_err("Memory failure: %#lx: Sending SIGBUS to %s:%d due to hardware memory corruption ", |
495367c05 mm/memory-failure... |
204 |
pfn, t->comm, t->pid); |
7329bbeb9 HWPOISON: Add cod... |
205 |
|
a70ffcac7 mm/memory-failure... |
206 |
if ((flags & MF_ACTION_REQUIRED) && t->mm == current->mm) { |
ae1139ece mm, memory_failur... |
207 |
ret = force_sig_mceerr(BUS_MCEERR_AR, (void __user *)tk->addr, |
f8eac9011 signal: Remove ta... |
208 |
addr_lsb); |
7329bbeb9 HWPOISON: Add cod... |
209 210 211 212 213 214 215 |
} 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? */ |
ae1139ece mm, memory_failur... |
216 |
ret = send_sig_mceerr(BUS_MCEERR_AO, (void __user *)tk->addr, |
c0f45555b signal/memory-fai... |
217 |
addr_lsb, t); /* synchronous? */ |
7329bbeb9 HWPOISON: Add cod... |
218 |
} |
6a46079cf HWPOISON: The hig... |
219 |
if (ret < 0) |
495367c05 mm/memory-failure... |
220 221 |
pr_info("Memory failure: Error sending signal to %s:%d: %d ", |
1170532bb mm: convert print... |
222 |
t->comm, t->pid, ret); |
6a46079cf HWPOISON: The hig... |
223 224 225 226 |
return ret; } /* |
588f9ce6c HWPOISON: Be more... |
227 228 229 |
* 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 HWPOISON: Add sof... |
230 |
void shake_page(struct page *p, int access) |
588f9ce6c HWPOISON: Be more... |
231 |
{ |
8bcb74de7 mm: hwpoison: cal... |
232 233 |
if (PageHuge(p)) return; |
588f9ce6c HWPOISON: Be more... |
234 235 236 237 |
if (!PageSlab(p)) { lru_add_drain_all(); if (PageLRU(p)) return; |
c05543293 mm, memory_hotplu... |
238 |
drain_all_pages(page_zone(p)); |
588f9ce6c HWPOISON: Be more... |
239 240 241 |
if (PageLRU(p) || is_free_buddy_page(p)) return; } |
facb6011f HWPOISON: Add sof... |
242 |
|
588f9ce6c HWPOISON: Be more... |
243 |
/* |
6b4f7799c mm: vmscan: invok... |
244 245 |
* Only call shrink_node_slabs here (which would also shrink * other caches) if access is not potentially fatal. |
588f9ce6c HWPOISON: Be more... |
246 |
*/ |
cb731d6c6 vmscan: per memor... |
247 248 |
if (access) drop_slab_node(page_to_nid(p)); |
588f9ce6c HWPOISON: Be more... |
249 250 |
} EXPORT_SYMBOL_GPL(shake_page); |
6100e34b2 mm, memory_failur... |
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 |
static unsigned long dev_pagemap_mapping_shift(struct page *page, struct vm_area_struct *vma) { unsigned long address = vma_address(page, vma); pgd_t *pgd; p4d_t *p4d; pud_t *pud; pmd_t *pmd; pte_t *pte; pgd = pgd_offset(vma->vm_mm, address); if (!pgd_present(*pgd)) return 0; p4d = p4d_offset(pgd, address); if (!p4d_present(*p4d)) return 0; pud = pud_offset(p4d, address); if (!pud_present(*pud)) return 0; if (pud_devmap(*pud)) return PUD_SHIFT; pmd = pmd_offset(pud, address); if (!pmd_present(*pmd)) return 0; if (pmd_devmap(*pmd)) return PMD_SHIFT; pte = pte_offset_map(pmd, address); if (!pte_present(*pte)) return 0; if (pte_devmap(*pte)) return PAGE_SHIFT; return 0; } |
6a46079cf HWPOISON: The hig... |
284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 |
/* * 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 mm/memory-failure... |
308 309 |
pr_err("Memory failure: Out of memory while machine check handling "); |
6a46079cf HWPOISON: The hig... |
310 311 312 313 |
return; } } tk->addr = page_address_in_vma(p, vma); |
6100e34b2 mm, memory_failur... |
314 315 316 317 |
if (is_zone_device_page(p)) tk->size_shift = dev_pagemap_mapping_shift(p, vma); else tk->size_shift = compound_order(compound_head(p)) + PAGE_SHIFT; |
6a46079cf HWPOISON: The hig... |
318 319 |
/* |
3d7fed4ad mm/memory-failure... |
320 321 322 323 324 325 326 327 |
* Send SIGKILL if "tk->addr == -EFAULT". Also, as * "tk->size_shift" is always non-zero for !is_zone_device_page(), * so "tk->size_shift == 0" effectively checks no mapping on * ZONE_DEVICE. Indeed, when a devdax page is mmapped N times * to a process' address space, it's possible not all N VMAs * contain mappings for the page, but at least one VMA does. * Only deliver SIGBUS with payload derived from the VMA that * has a mapping for the page. |
6a46079cf HWPOISON: The hig... |
328 |
*/ |
3d7fed4ad mm/memory-failure... |
329 |
if (tk->addr == -EFAULT) { |
495367c05 mm/memory-failure... |
330 331 |
pr_info("Memory failure: Unable to find user space address %lx in %s ", |
6a46079cf HWPOISON: The hig... |
332 |
page_to_pfn(p), tsk->comm); |
3d7fed4ad mm/memory-failure... |
333 334 335 |
} else if (tk->size_shift == 0) { kfree(tk); return; |
6a46079cf HWPOISON: The hig... |
336 337 338 339 340 341 342 343 344 345 346 347 348 349 |
} 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. */ |
ae1139ece mm, memory_failur... |
350 351 |
static void kill_procs(struct list_head *to_kill, int forcekill, bool fail, unsigned long pfn, int flags) |
6a46079cf HWPOISON: The hig... |
352 353 354 355 |
{ struct to_kill *tk, *next; list_for_each_entry_safe (tk, next, to_kill, nd) { |
6751ed65d x86/mce: Fix sigi... |
356 |
if (forcekill) { |
6a46079cf HWPOISON: The hig... |
357 |
/* |
af901ca18 tree-wide: fix as... |
358 |
* In case something went wrong with munmapping |
6a46079cf HWPOISON: The hig... |
359 360 |
* make sure the process doesn't catch the * signal and then access the memory. Just kill it. |
6a46079cf HWPOISON: The hig... |
361 |
*/ |
3d7fed4ad mm/memory-failure... |
362 |
if (fail || tk->addr == -EFAULT) { |
495367c05 mm/memory-failure... |
363 364 |
pr_err("Memory failure: %#lx: forcibly killing %s:%d because of failure to unmap corrupted page ", |
1170532bb mm: convert print... |
365 |
pfn, tk->tsk->comm, tk->tsk->pid); |
6376360ec mm: hwpoison: use... |
366 367 |
do_send_sig_info(SIGKILL, SEND_SIG_PRIV, tk->tsk, PIDTYPE_PID); |
6a46079cf HWPOISON: The hig... |
368 369 370 371 372 373 374 375 |
} /* * 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. */ |
ae1139ece mm, memory_failur... |
376 |
else if (kill_proc(tk, pfn, flags) < 0) |
495367c05 mm/memory-failure... |
377 378 |
pr_err("Memory failure: %#lx: Cannot send advisory machine check signal to %s:%d ", |
1170532bb mm: convert print... |
379 |
pfn, tk->tsk->comm, tk->tsk->pid); |
6a46079cf HWPOISON: The hig... |
380 381 382 383 384 |
} put_task_struct(tk->tsk); kfree(tk); } } |
3ba08129e mm/memory-failure... |
385 386 387 388 389 390 391 392 393 |
/* * 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 HWPOISON: The hig... |
394 |
{ |
3ba08129e mm/memory-failure... |
395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 |
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 HWPOISON: The hig... |
413 |
if (!tsk->mm) |
3ba08129e mm/memory-failure... |
414 |
return NULL; |
74614de17 mm/memory-failure... |
415 |
if (force_early) |
3ba08129e mm/memory-failure... |
416 417 418 419 420 421 422 |
return tsk; t = find_early_kill_thread(tsk); if (t) return t; if (sysctl_memory_failure_early_kill) return tsk; return NULL; |
6a46079cf HWPOISON: The hig... |
423 424 425 426 427 428 |
} /* * Collect processes when the error hit an anonymous page. */ static void collect_procs_anon(struct page *page, struct list_head *to_kill, |
74614de17 mm/memory-failure... |
429 |
struct to_kill **tkc, int force_early) |
6a46079cf HWPOISON: The hig... |
430 431 432 433 |
{ struct vm_area_struct *vma; struct task_struct *tsk; struct anon_vma *av; |
bf181b9f9 mm anon rmap: rep... |
434 |
pgoff_t pgoff; |
6a46079cf HWPOISON: The hig... |
435 |
|
4fc3f1d66 mm/rmap, migratio... |
436 |
av = page_lock_anon_vma_read(page); |
6a46079cf HWPOISON: The hig... |
437 |
if (av == NULL) /* Not actually mapped anymore */ |
9b679320a mm/memory-failure... |
438 |
return; |
a0f7a756c mm/rmap.c: fix pg... |
439 |
pgoff = page_to_pgoff(page); |
9b679320a mm/memory-failure... |
440 |
read_lock(&tasklist_lock); |
6a46079cf HWPOISON: The hig... |
441 |
for_each_process (tsk) { |
5beb49305 mm: change anon_v... |
442 |
struct anon_vma_chain *vmac; |
3ba08129e mm/memory-failure... |
443 |
struct task_struct *t = task_early_kill(tsk, force_early); |
5beb49305 mm: change anon_v... |
444 |
|
3ba08129e mm/memory-failure... |
445 |
if (!t) |
6a46079cf HWPOISON: The hig... |
446 |
continue; |
bf181b9f9 mm anon rmap: rep... |
447 448 |
anon_vma_interval_tree_foreach(vmac, &av->rb_root, pgoff, pgoff) { |
5beb49305 mm: change anon_v... |
449 |
vma = vmac->vma; |
6a46079cf HWPOISON: The hig... |
450 451 |
if (!page_mapped_in_vma(page, vma)) continue; |
3ba08129e mm/memory-failure... |
452 453 |
if (vma->vm_mm == t->mm) add_to_kill(t, page, vma, to_kill, tkc); |
6a46079cf HWPOISON: The hig... |
454 455 |
} } |
6a46079cf HWPOISON: The hig... |
456 |
read_unlock(&tasklist_lock); |
4fc3f1d66 mm/rmap, migratio... |
457 |
page_unlock_anon_vma_read(av); |
6a46079cf HWPOISON: The hig... |
458 459 460 461 462 463 |
} /* * Collect processes when the error hit a file mapped page. */ static void collect_procs_file(struct page *page, struct list_head *to_kill, |
74614de17 mm/memory-failure... |
464 |
struct to_kill **tkc, int force_early) |
6a46079cf HWPOISON: The hig... |
465 466 467 |
{ struct vm_area_struct *vma; struct task_struct *tsk; |
6a46079cf HWPOISON: The hig... |
468 |
struct address_space *mapping = page->mapping; |
d28eb9c86 mm/memory-failure... |
469 |
i_mmap_lock_read(mapping); |
9b679320a mm/memory-failure... |
470 |
read_lock(&tasklist_lock); |
6a46079cf HWPOISON: The hig... |
471 |
for_each_process(tsk) { |
a0f7a756c mm/rmap.c: fix pg... |
472 |
pgoff_t pgoff = page_to_pgoff(page); |
3ba08129e mm/memory-failure... |
473 |
struct task_struct *t = task_early_kill(tsk, force_early); |
6a46079cf HWPOISON: The hig... |
474 |
|
3ba08129e mm/memory-failure... |
475 |
if (!t) |
6a46079cf HWPOISON: The hig... |
476 |
continue; |
6b2dbba8b mm: replace vma p... |
477 |
vma_interval_tree_foreach(vma, &mapping->i_mmap, pgoff, |
6a46079cf HWPOISON: The hig... |
478 479 480 481 482 483 484 485 |
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 mm/memory-failure... |
486 487 |
if (vma->vm_mm == t->mm) add_to_kill(t, page, vma, to_kill, tkc); |
6a46079cf HWPOISON: The hig... |
488 489 |
} } |
6a46079cf HWPOISON: The hig... |
490 |
read_unlock(&tasklist_lock); |
d28eb9c86 mm/memory-failure... |
491 |
i_mmap_unlock_read(mapping); |
6a46079cf HWPOISON: The hig... |
492 493 494 495 496 497 498 499 |
} /* * 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 mm/memory-failure... |
500 501 |
static void collect_procs(struct page *page, struct list_head *tokill, int force_early) |
6a46079cf HWPOISON: The hig... |
502 503 504 505 506 507 508 509 510 511 |
{ struct to_kill *tk; if (!page->mapping) return; tk = kmalloc(sizeof(struct to_kill), GFP_NOIO); if (!tk) return; if (PageAnon(page)) |
74614de17 mm/memory-failure... |
512 |
collect_procs_anon(page, tokill, &tk, force_early); |
6a46079cf HWPOISON: The hig... |
513 |
else |
74614de17 mm/memory-failure... |
514 |
collect_procs_file(page, tokill, &tk, force_early); |
6a46079cf HWPOISON: The hig... |
515 516 |
kfree(tk); } |
6a46079cf HWPOISON: The hig... |
517 |
static const char *action_name[] = { |
cc637b170 memory-failure: e... |
518 519 520 521 |
[MF_IGNORED] = "Ignored", [MF_FAILED] = "Failed", [MF_DELAYED] = "Delayed", [MF_RECOVERED] = "Recovered", |
64d37a2ba mm/memory-failure... |
522 523 524 |
}; static const char * const action_page_types[] = { |
cc637b170 memory-failure: e... |
525 526 527 528 529 530 531 |
[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", |
31286a848 mm: hwpoison: dis... |
532 |
[MF_MSG_NON_PMD_HUGE] = "non-pmd-sized huge page", |
cc637b170 memory-failure: e... |
533 534 535 536 537 538 539 540 541 542 543 544 |
[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)", |
6100e34b2 mm, memory_failur... |
545 |
[MF_MSG_DAX] = "dax page", |
cc637b170 memory-failure: e... |
546 |
[MF_MSG_UNKNOWN] = "unknown page", |
64d37a2ba mm/memory-failure... |
547 |
}; |
6a46079cf HWPOISON: The hig... |
548 |
/* |
dc2a1cbf7 HWPOISON: introdu... |
549 550 551 552 553 554 555 556 557 558 559 560 561 562 |
* 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); |
18365225f hwpoison, memcg: ... |
563 564 565 566 567 568 |
/* * Poisoned page might never drop its ref count to 0 so we have * to uncharge it manually from its memcg. */ mem_cgroup_uncharge(p); |
dc2a1cbf7 HWPOISON: introdu... |
569 570 571 |
/* * drop the page count elevated by isolate_lru_page() */ |
09cbfeaf1 mm, fs: get rid o... |
572 |
put_page(p); |
dc2a1cbf7 HWPOISON: introdu... |
573 574 575 576 |
return 0; } return -EIO; } |
78bb92034 mm: hwpoison: dis... |
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 |
static int truncate_error_page(struct page *p, unsigned long pfn, struct address_space *mapping) { int ret = MF_FAILED; if (mapping->a_ops->error_remove_page) { int err = mapping->a_ops->error_remove_page(mapping, p); if (err != 0) { pr_info("Memory failure: %#lx: Failed to punch page: %d ", pfn, err); } else if (page_has_private(p) && !try_to_release_page(p, GFP_NOIO)) { pr_info("Memory failure: %#lx: failed to release buffers ", pfn); } else { ret = MF_RECOVERED; } } 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)) ret = MF_RECOVERED; else pr_info("Memory failure: %#lx: Failed to invalidate ", pfn); } return ret; } |
dc2a1cbf7 HWPOISON: introdu... |
612 |
/* |
6a46079cf HWPOISON: The hig... |
613 614 615 616 617 618 |
* 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 memory-failure: e... |
619 |
return MF_IGNORED; |
6a46079cf HWPOISON: The hig... |
620 621 622 623 624 625 626 |
} /* * Page in unknown state. Do nothing. */ static int me_unknown(struct page *p, unsigned long pfn) { |
495367c05 mm/memory-failure... |
627 628 |
pr_err("Memory failure: %#lx: Unknown page state ", pfn); |
cc637b170 memory-failure: e... |
629 |
return MF_FAILED; |
6a46079cf HWPOISON: The hig... |
630 631 632 |
} /* |
6a46079cf HWPOISON: The hig... |
633 634 635 636 |
* Clean (or cleaned) page cache page. */ static int me_pagecache_clean(struct page *p, unsigned long pfn) { |
6a46079cf HWPOISON: The hig... |
637 |
struct address_space *mapping; |
dc2a1cbf7 HWPOISON: introdu... |
638 |
delete_from_lru_cache(p); |
6a46079cf HWPOISON: The hig... |
639 640 641 642 643 |
/* * For anonymous pages we're done the only reference left * should be the one m_f() holds. */ if (PageAnon(p)) |
cc637b170 memory-failure: e... |
644 |
return MF_RECOVERED; |
6a46079cf HWPOISON: The hig... |
645 646 647 648 649 650 651 652 653 654 655 656 657 |
/* * 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 memory-failure: e... |
658 |
return MF_FAILED; |
6a46079cf HWPOISON: The hig... |
659 660 661 662 663 664 665 |
} /* * 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. */ |
78bb92034 mm: hwpoison: dis... |
666 |
return truncate_error_page(p, pfn, mapping); |
6a46079cf HWPOISON: The hig... |
667 668 669 |
} /* |
549543dff mm, memory-failur... |
670 |
* Dirty pagecache page |
6a46079cf HWPOISON: The hig... |
671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 |
* 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 Fix common misspe... |
703 |
* and the page is dropped between then the error |
6a46079cf HWPOISON: The hig... |
704 705 706 707 708 709 710 711 712 713 714 |
* 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. */ |
af21bfaf7 mm: fix mapping_s... |
715 |
mapping_set_error(mapping, -EIO); |
6a46079cf HWPOISON: The hig... |
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 |
} 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 HWPOISON: The hig... |
742 743 744 |
ClearPageDirty(p); /* Trigger EIO in shmem: */ ClearPageUptodate(p); |
dc2a1cbf7 HWPOISON: introdu... |
745 |
if (!delete_from_lru_cache(p)) |
cc637b170 memory-failure: e... |
746 |
return MF_DELAYED; |
dc2a1cbf7 HWPOISON: introdu... |
747 |
else |
cc637b170 memory-failure: e... |
748 |
return MF_FAILED; |
6a46079cf HWPOISON: The hig... |
749 750 751 752 |
} static int me_swapcache_clean(struct page *p, unsigned long pfn) { |
6a46079cf HWPOISON: The hig... |
753 |
delete_from_swap_cache(p); |
e43c3afb3 HWPOISON: return ... |
754 |
|
dc2a1cbf7 HWPOISON: introdu... |
755 |
if (!delete_from_lru_cache(p)) |
cc637b170 memory-failure: e... |
756 |
return MF_RECOVERED; |
dc2a1cbf7 HWPOISON: introdu... |
757 |
else |
cc637b170 memory-failure: e... |
758 |
return MF_FAILED; |
6a46079cf HWPOISON: The hig... |
759 760 761 762 763 |
} /* * Huge pages. Needs work. * Issues: |
93f70f900 HWPOISON, hugetlb... |
764 765 |
* - 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 HWPOISON: The hig... |
766 767 768 |
*/ static int me_huge_page(struct page *p, unsigned long pfn) { |
6de2b1aab HWPOISON, hugetlb... |
769 |
int res = 0; |
93f70f900 HWPOISON, hugetlb... |
770 |
struct page *hpage = compound_head(p); |
78bb92034 mm: hwpoison: dis... |
771 |
struct address_space *mapping; |
2491ffee9 mm/memory-failure... |
772 773 774 |
if (!PageHuge(hpage)) return MF_DELAYED; |
78bb92034 mm: hwpoison: dis... |
775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 |
mapping = page_mapping(hpage); if (mapping) { res = truncate_error_page(hpage, pfn, mapping); } else { unlock_page(hpage); /* * migration entry prevents later access on error anonymous * hugepage, so we can free and dissolve it into buddy to * save healthy subpages. */ if (PageAnon(hpage)) put_page(hpage); dissolve_free_huge_page(p); res = MF_RECOVERED; lock_page(hpage); |
93f70f900 HWPOISON, hugetlb... |
790 |
} |
78bb92034 mm: hwpoison: dis... |
791 792 |
return res; |
6a46079cf HWPOISON: The hig... |
793 794 795 796 797 798 799 800 801 |
} /* * 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 Fix common misspe... |
802 |
* in its live cycle, so all accesses have to be extremely careful. |
6a46079cf HWPOISON: The hig... |
803 804 805 806 807 808 |
* * This is not complete. More states could be added. * For any missing state don't attempt recovery. */ #define dirty (1UL << PG_dirty) |
6326fec11 mm: Use owner_pri... |
809 |
#define sc ((1UL << PG_swapcache) | (1UL << PG_swapbacked)) |
6a46079cf HWPOISON: The hig... |
810 811 812 813 |
#define unevict (1UL << PG_unevictable) #define mlock (1UL << PG_mlocked) #define writeback (1UL << PG_writeback) #define lru (1UL << PG_lru) |
6a46079cf HWPOISON: The hig... |
814 |
#define head (1UL << PG_head) |
6a46079cf HWPOISON: The hig... |
815 |
#define slab (1UL << PG_slab) |
6a46079cf HWPOISON: The hig... |
816 817 818 819 820 |
#define reserved (1UL << PG_reserved) static struct page_state { unsigned long mask; unsigned long res; |
cc637b170 memory-failure: e... |
821 |
enum mf_action_page_type type; |
6a46079cf HWPOISON: The hig... |
822 823 |
int (*action)(struct page *p, unsigned long pfn); } error_states[] = { |
cc637b170 memory-failure: e... |
824 |
{ reserved, reserved, MF_MSG_KERNEL, me_kernel }, |
95d01fc66 HWPOISON: remove ... |
825 826 827 828 |
/* * free pages are specially detected outside this table: * PG_buddy pages only make a small fraction of all free pages. */ |
6a46079cf HWPOISON: The hig... |
829 830 831 832 833 834 |
/* * 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 memory-failure: e... |
835 |
{ slab, slab, MF_MSG_SLAB, me_kernel }, |
6a46079cf HWPOISON: The hig... |
836 |
|
cc637b170 memory-failure: e... |
837 |
{ head, head, MF_MSG_HUGE, me_huge_page }, |
6a46079cf HWPOISON: The hig... |
838 |
|
cc637b170 memory-failure: e... |
839 840 |
{ sc|dirty, sc|dirty, MF_MSG_DIRTY_SWAPCACHE, me_swapcache_dirty }, { sc|dirty, sc, MF_MSG_CLEAN_SWAPCACHE, me_swapcache_clean }, |
6a46079cf HWPOISON: The hig... |
841 |
|
cc637b170 memory-failure: e... |
842 843 |
{ mlock|dirty, mlock|dirty, MF_MSG_DIRTY_MLOCKED_LRU, me_pagecache_dirty }, { mlock|dirty, mlock, MF_MSG_CLEAN_MLOCKED_LRU, me_pagecache_clean }, |
6a46079cf HWPOISON: The hig... |
844 |
|
cc637b170 memory-failure: e... |
845 846 |
{ unevict|dirty, unevict|dirty, MF_MSG_DIRTY_UNEVICTABLE_LRU, me_pagecache_dirty }, { unevict|dirty, unevict, MF_MSG_CLEAN_UNEVICTABLE_LRU, me_pagecache_clean }, |
5f4b9fc5c HWPOISON: change ... |
847 |
|
cc637b170 memory-failure: e... |
848 849 |
{ lru|dirty, lru|dirty, MF_MSG_DIRTY_LRU, me_pagecache_dirty }, { lru|dirty, lru, MF_MSG_CLEAN_LRU, me_pagecache_clean }, |
6a46079cf HWPOISON: The hig... |
850 851 852 853 |
/* * Catchall entry: must be at end. */ |
cc637b170 memory-failure: e... |
854 |
{ 0, 0, MF_MSG_UNKNOWN, me_unknown }, |
6a46079cf HWPOISON: The hig... |
855 |
}; |
2326c467d HWPOISON: Undefin... |
856 857 858 859 860 861 |
#undef dirty #undef sc #undef unevict #undef mlock #undef writeback #undef lru |
2326c467d HWPOISON: Undefin... |
862 |
#undef head |
2326c467d HWPOISON: Undefin... |
863 864 |
#undef slab #undef reserved |
ff604cf6d mm: hwpoison: fix... |
865 866 867 868 |
/* * "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 memory-failure: c... |
869 870 |
static void action_result(unsigned long pfn, enum mf_action_page_type type, enum mf_result result) |
6a46079cf HWPOISON: The hig... |
871 |
{ |
97f0b1345 tracing: add trac... |
872 |
trace_memory_failure_event(pfn, type, result); |
495367c05 mm/memory-failure... |
873 874 |
pr_err("Memory failure: %#lx: recovery action for %s: %s ", |
64d37a2ba mm/memory-failure... |
875 |
pfn, action_page_types[type], action_name[result]); |
6a46079cf HWPOISON: The hig... |
876 877 878 |
} static int page_action(struct page_state *ps, struct page *p, |
bd1ce5f91 HWPOISON: avoid g... |
879 |
unsigned long pfn) |
6a46079cf HWPOISON: The hig... |
880 881 |
{ int result; |
7456b0405 HWPOISON: fix inv... |
882 |
int count; |
6a46079cf HWPOISON: The hig... |
883 884 |
result = ps->action(p, pfn); |
7456b0405 HWPOISON: fix inv... |
885 |
|
bd1ce5f91 HWPOISON: avoid g... |
886 |
count = page_count(p) - 1; |
cc637b170 memory-failure: e... |
887 |
if (ps->action == me_swapcache_dirty && result == MF_DELAYED) |
138ce286e HWPOISON: return ... |
888 |
count--; |
78bb92034 mm: hwpoison: dis... |
889 |
if (count > 0) { |
495367c05 mm/memory-failure... |
890 891 |
pr_err("Memory failure: %#lx: %s still referenced by %d users ", |
64d37a2ba mm/memory-failure... |
892 |
pfn, action_page_types[ps->type], count); |
cc637b170 memory-failure: e... |
893 |
result = MF_FAILED; |
138ce286e HWPOISON: return ... |
894 |
} |
64d37a2ba mm/memory-failure... |
895 |
action_result(pfn, ps->type, result); |
6a46079cf HWPOISON: The hig... |
896 897 898 899 900 |
/* Could do more checks here if page looks ok */ /* * Could adjust zone counters here to correct for the missing page. */ |
cc637b170 memory-failure: e... |
901 |
return (result == MF_RECOVERED || result == MF_DELAYED) ? 0 : -EBUSY; |
6a46079cf HWPOISON: The hig... |
902 |
} |
ead07f6a8 mm/memory-failure... |
903 904 905 906 907 908 909 910 911 912 |
/** * 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 mm: hwpoison: adj... |
913 |
if (!PageHuge(head) && PageTransHuge(head)) { |
98ed2b005 mm/memory-failure... |
914 915 916 917 918 919 920 |
/* * 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 mm/memory-failure... |
921 922 |
pr_err("Memory failure: %#lx: non anonymous thp ", |
98ed2b005 mm/memory-failure... |
923 924 925 |
page_to_pfn(page)); return 0; } |
ead07f6a8 mm/memory-failure... |
926 |
} |
c2e7e00b7 mm/memory-failure... |
927 928 929 |
if (get_page_unless_zero(head)) { if (head == compound_head(page)) return 1; |
495367c05 mm/memory-failure... |
930 931 932 |
pr_info("Memory failure: %#lx cannot catch tail ", page_to_pfn(page)); |
c2e7e00b7 mm/memory-failure... |
933 934 935 936 |
put_page(head); } return 0; |
ead07f6a8 mm/memory-failure... |
937 938 |
} EXPORT_SYMBOL_GPL(get_hwpoison_page); |
6a46079cf HWPOISON: The hig... |
939 940 941 942 |
/* * Do all that is necessary to remove user space mappings. Unmap * the pages and send SIGBUS to the processes if the data was dirty. */ |
666e5a406 mm: make ttu's re... |
943 |
static bool hwpoison_user_mappings(struct page *p, unsigned long pfn, |
83b57531c mm/memory_failure... |
944 |
int flags, struct page **hpagep) |
6a46079cf HWPOISON: The hig... |
945 |
{ |
a128ca71f mm: delete unnece... |
946 |
enum ttu_flags ttu = TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS; |
6a46079cf HWPOISON: The hig... |
947 948 |
struct address_space *mapping; LIST_HEAD(tokill); |
ddeaab32a hugetlbfs: revert... |
949 |
bool unmap_success; |
6751ed65d x86/mce: Fix sigi... |
950 |
int kill = 1, forcekill; |
54b9dd14d mm/memory-failure... |
951 |
struct page *hpage = *hpagep; |
286c469a9 mm: hwpoison: cal... |
952 |
bool mlocked = PageMlocked(hpage); |
6a46079cf HWPOISON: The hig... |
953 |
|
93a9eb39f hwpoison: fix hug... |
954 955 956 957 958 |
/* * Here we are interested only in user-mapped pages, so skip any * other types of pages. */ if (PageReserved(p) || PageSlab(p)) |
666e5a406 mm: make ttu's re... |
959 |
return true; |
93a9eb39f hwpoison: fix hug... |
960 |
if (!(PageLRU(hpage) || PageHuge(p))) |
666e5a406 mm: make ttu's re... |
961 |
return true; |
6a46079cf HWPOISON: The hig... |
962 |
|
6a46079cf HWPOISON: The hig... |
963 964 965 966 |
/* * This check implies we don't kill processes if their pages * are in the swap cache early. Those are always late kills. */ |
7af446a84 HWPOISON, hugetlb... |
967 |
if (!page_mapped(hpage)) |
666e5a406 mm: make ttu's re... |
968 |
return true; |
1668bfd5b HWPOISON: abort o... |
969 |
|
52089b14c hwpoison: call ac... |
970 |
if (PageKsm(p)) { |
495367c05 mm/memory-failure... |
971 972 |
pr_err("Memory failure: %#lx: can't handle KSM pages. ", pfn); |
666e5a406 mm: make ttu's re... |
973 |
return false; |
52089b14c hwpoison: call ac... |
974 |
} |
6a46079cf HWPOISON: The hig... |
975 976 |
if (PageSwapCache(p)) { |
495367c05 mm/memory-failure... |
977 978 979 |
pr_err("Memory failure: %#lx: keeping poisoned page in swap cache ", pfn); |
6a46079cf HWPOISON: The hig... |
980 981 982 983 984 985 |
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 HWPOISON: comment... |
986 987 |
* 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 HWPOISON: The hig... |
988 |
*/ |
7af446a84 HWPOISON, hugetlb... |
989 |
mapping = page_mapping(hpage); |
6751ed65d x86/mce: Fix sigi... |
990 |
if (!(flags & MF_MUST_KILL) && !PageDirty(hpage) && mapping && |
7af446a84 HWPOISON, hugetlb... |
991 992 993 |
mapping_cap_writeback_dirty(mapping)) { if (page_mkclean(hpage)) { SetPageDirty(hpage); |
6a46079cf HWPOISON: The hig... |
994 995 996 |
} else { kill = 0; ttu |= TTU_IGNORE_HWPOISON; |
495367c05 mm/memory-failure... |
997 998 |
pr_info("Memory failure: %#lx: corrupted page was clean: dropped without side effects ", |
6a46079cf HWPOISON: The hig... |
999 1000 1001 |
pfn); } } |
a6d30ddda thp: fix the wron... |
1002 |
/* |
6a46079cf HWPOISON: The hig... |
1003 1004 1005 1006 1007 1008 1009 1010 |
* 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 mm/memory-failure... |
1011 |
collect_procs(hpage, &tokill, flags & MF_ACTION_REQUIRED); |
6a46079cf HWPOISON: The hig... |
1012 |
|
ddeaab32a hugetlbfs: revert... |
1013 |
unmap_success = try_to_unmap(hpage, ttu); |
666e5a406 mm: make ttu's re... |
1014 |
if (!unmap_success) |
495367c05 mm/memory-failure... |
1015 1016 |
pr_err("Memory failure: %#lx: failed to unmap page (mapcount=%d) ", |
1170532bb mm: convert print... |
1017 |
pfn, page_mapcount(hpage)); |
a6d30ddda thp: fix the wron... |
1018 |
|
6a46079cf HWPOISON: The hig... |
1019 |
/* |
286c469a9 mm: hwpoison: cal... |
1020 1021 1022 1023 1024 1025 1026 |
* try_to_unmap() might put mlocked page in lru cache, so call * shake_page() again to ensure that it's flushed. */ if (mlocked) shake_page(hpage, 0); /* |
6a46079cf HWPOISON: The hig... |
1027 1028 1029 |
* 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 x86/mce: Fix sigi... |
1030 1031 |
* was dirty or the process is not restartable, * otherwise the tokill list is merely |
6a46079cf HWPOISON: The hig... |
1032 1033 1034 1035 |
* freed. When there was a problem unmapping earlier * use a more force-full uncatchable kill to prevent * any accesses to the poisoned memory. */ |
415c64c14 mm/memory-failure... |
1036 |
forcekill = PageDirty(hpage) || (flags & MF_MUST_KILL); |
ae1139ece mm, memory_failur... |
1037 |
kill_procs(&tokill, forcekill, !unmap_success, pfn, flags); |
1668bfd5b HWPOISON: abort o... |
1038 |
|
666e5a406 mm: make ttu's re... |
1039 |
return unmap_success; |
6a46079cf HWPOISON: The hig... |
1040 |
} |
0348d2ebe mm: hwpoison: int... |
1041 1042 |
static int identify_page_state(unsigned long pfn, struct page *p, unsigned long page_flags) |
761ad8d7c mm: hwpoison: int... |
1043 1044 |
{ struct page_state *ps; |
0348d2ebe mm: hwpoison: int... |
1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 |
/* * The first check uses the current page flags which may not have any * relevant information. The second check with the saved page flags 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) break; page_flags |= (p->flags & (1UL << PG_dirty)); if (!ps->mask) for (ps = error_states;; ps++) if ((page_flags & ps->mask) == ps->res) break; return page_action(ps, p, pfn); } |
83b57531c mm/memory_failure... |
1063 |
static int memory_failure_hugetlb(unsigned long pfn, int flags) |
0348d2ebe mm: hwpoison: int... |
1064 |
{ |
761ad8d7c mm: hwpoison: int... |
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 |
struct page *p = pfn_to_page(pfn); struct page *head = compound_head(p); int res; unsigned long page_flags; if (TestSetPageHWPoison(head)) { pr_err("Memory failure: %#lx: already hardware poisoned ", pfn); return 0; } num_poisoned_pages_inc(); if (!(flags & MF_COUNT_INCREASED) && !get_hwpoison_page(p)) { /* * Check "filter hit" and "race with other subpage." */ lock_page(head); if (PageHWPoison(head)) { if ((hwpoison_filter(p) && TestClearPageHWPoison(p)) || (p != head && TestSetPageHWPoison(head))) { num_poisoned_pages_dec(); unlock_page(head); return 0; } } unlock_page(head); dissolve_free_huge_page(p); action_result(pfn, MF_MSG_FREE_HUGE, MF_DELAYED); return 0; } lock_page(head); page_flags = head->flags; if (!PageHWPoison(head)) { pr_err("Memory failure: %#lx: just unpoisoned ", pfn); num_poisoned_pages_dec(); unlock_page(head); put_hwpoison_page(head); return 0; } |
31286a848 mm: hwpoison: dis... |
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 |
/* * TODO: hwpoison for pud-sized hugetlb doesn't work right now, so * simply disable it. In order to make it work properly, we need * make sure that: * - conversion of a pud that maps an error hugetlb into hwpoison * entry properly works, and * - other mm code walking over page table is aware of pud-aligned * hwpoison entries. */ if (huge_page_size(page_hstate(head)) > PMD_SIZE) { action_result(pfn, MF_MSG_NON_PMD_HUGE, MF_IGNORED); res = -EBUSY; goto out; } |
83b57531c mm/memory_failure... |
1123 |
if (!hwpoison_user_mappings(p, pfn, flags, &head)) { |
761ad8d7c mm: hwpoison: int... |
1124 1125 1126 1127 |
action_result(pfn, MF_MSG_UNMAP_FAILED, MF_IGNORED); res = -EBUSY; goto out; } |
0348d2ebe mm: hwpoison: int... |
1128 |
res = identify_page_state(pfn, p, page_flags); |
761ad8d7c mm: hwpoison: int... |
1129 1130 1131 1132 |
out: unlock_page(head); return res; } |
6100e34b2 mm, memory_failur... |
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 |
static int memory_failure_dev_pagemap(unsigned long pfn, int flags, struct dev_pagemap *pgmap) { struct page *page = pfn_to_page(pfn); const bool unmap_success = true; unsigned long size = 0; struct to_kill *tk; LIST_HEAD(tokill); int rc = -EBUSY; loff_t start; |
27359fd6e dax: Fix unlock m... |
1143 |
dax_entry_t cookie; |
6100e34b2 mm, memory_failur... |
1144 1145 1146 1147 1148 1149 1150 1151 |
/* * Prevent the inode from being freed while we are interrogating * the address_space, typically this would be handled by * lock_page(), but dax pages do not use the page lock. This * also prevents changes to the mapping of this pfn until * poison signaling is complete. */ |
27359fd6e dax: Fix unlock m... |
1152 1153 |
cookie = dax_lock_page(page); if (!cookie) |
6100e34b2 mm, memory_failur... |
1154 1155 1156 1157 1158 1159 |
goto out; if (hwpoison_filter(page)) { rc = 0; goto unlock; } |
25b2995a3 mm: remove MEMORY... |
1160 |
if (pgmap->type == MEMORY_DEVICE_PRIVATE) { |
6100e34b2 mm, memory_failur... |
1161 1162 1163 1164 1165 |
/* * TODO: Handle HMM pages which may need coordination * with device-side memory. */ goto unlock; |
6100e34b2 mm, memory_failur... |
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 |
} /* * Use this flag as an indication that the dax page has been * remapped UC to prevent speculative consumption of poison. */ SetPageHWPoison(page); /* * Unlike System-RAM there is no possibility to swap in a * different physical page at a given virtual address, so all * userspace consumption of ZONE_DEVICE memory necessitates * SIGBUS (i.e. MF_MUST_KILL) */ flags |= MF_ACTION_REQUIRED | MF_MUST_KILL; collect_procs(page, &tokill, flags & MF_ACTION_REQUIRED); list_for_each_entry(tk, &tokill, nd) if (tk->size_shift) size = max(size, 1UL << tk->size_shift); if (size) { /* * Unmap the largest mapping to avoid breaking up * device-dax mappings which are constant size. The * actual size of the mapping being torn down is * communicated in siginfo, see kill_proc() */ start = (page->index << PAGE_SHIFT) & ~(size - 1); unmap_mapping_range(page->mapping, start, start + size, 0); } kill_procs(&tokill, flags & MF_MUST_KILL, !unmap_success, pfn, flags); rc = 0; unlock: |
27359fd6e dax: Fix unlock m... |
1199 |
dax_unlock_page(page, cookie); |
6100e34b2 mm, memory_failur... |
1200 1201 1202 1203 1204 1205 |
out: /* drop pgmap ref acquired in caller */ put_dev_pagemap(pgmap); action_result(pfn, MF_MSG_DAX, rc ? MF_FAILED : MF_RECOVERED); return rc; } |
cd42f4a3b HWPOISON: Clean u... |
1206 1207 1208 |
/** * memory_failure - Handle memory failure of a page. * @pfn: Page Number of the corrupted page |
cd42f4a3b HWPOISON: Clean u... |
1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 |
* @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. */ |
83b57531c mm/memory_failure... |
1223 |
int memory_failure(unsigned long pfn, int flags) |
6a46079cf HWPOISON: The hig... |
1224 |
{ |
6a46079cf HWPOISON: The hig... |
1225 |
struct page *p; |
7af446a84 HWPOISON, hugetlb... |
1226 |
struct page *hpage; |
415c64c14 mm/memory-failure... |
1227 |
struct page *orig_head; |
6100e34b2 mm, memory_failur... |
1228 |
struct dev_pagemap *pgmap; |
6a46079cf HWPOISON: The hig... |
1229 |
int res; |
524fca1e7 HWPOISON: fix mis... |
1230 |
unsigned long page_flags; |
6a46079cf HWPOISON: The hig... |
1231 1232 |
if (!sysctl_memory_failure_recovery) |
83b57531c mm/memory_failure... |
1233 |
panic("Memory failure on page %lx", pfn); |
6a46079cf HWPOISON: The hig... |
1234 |
|
96c804a6a mm/memory-failure... |
1235 1236 1237 1238 1239 1240 1241 1242 |
p = pfn_to_online_page(pfn); if (!p) { if (pfn_valid(pfn)) { pgmap = get_dev_pagemap(pfn, NULL); if (pgmap) return memory_failure_dev_pagemap(pfn, flags, pgmap); } |
495367c05 mm/memory-failure... |
1243 1244 1245 |
pr_err("Memory failure: %#lx: memory outside kernel control ", pfn); |
a7560fc80 HWPOISON: return ... |
1246 |
return -ENXIO; |
6a46079cf HWPOISON: The hig... |
1247 |
} |
761ad8d7c mm: hwpoison: int... |
1248 |
if (PageHuge(p)) |
83b57531c mm/memory_failure... |
1249 |
return memory_failure_hugetlb(pfn, flags); |
6a46079cf HWPOISON: The hig... |
1250 |
if (TestSetPageHWPoison(p)) { |
495367c05 mm/memory-failure... |
1251 1252 1253 |
pr_err("Memory failure: %#lx: already hardware poisoned ", pfn); |
6a46079cf HWPOISON: The hig... |
1254 1255 |
return 0; } |
761ad8d7c mm: hwpoison: int... |
1256 |
orig_head = hpage = compound_head(p); |
b37ff71cc mm: hwpoison: cha... |
1257 |
num_poisoned_pages_inc(); |
6a46079cf HWPOISON: The hig... |
1258 1259 1260 1261 1262 |
/* * 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. |
761ad8d7c mm: hwpoison: int... |
1263 |
* 2) it's part of a non-compound high order page. |
6a46079cf HWPOISON: The hig... |
1264 1265 1266 1267 |
* 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, |
1c4c3b99c mm: fix page_free... |
1268 |
* that may make page_ref_freeze()/page_ref_unfreeze() mismatch. |
6a46079cf HWPOISON: The hig... |
1269 |
*/ |
ead07f6a8 mm/memory-failure... |
1270 |
if (!(flags & MF_COUNT_INCREASED) && !get_hwpoison_page(p)) { |
8d22ba1b7 HWPOISON: detect ... |
1271 |
if (is_free_buddy_page(p)) { |
cc637b170 memory-failure: e... |
1272 |
action_result(pfn, MF_MSG_BUDDY, MF_DELAYED); |
8d22ba1b7 HWPOISON: detect ... |
1273 1274 |
return 0; } else { |
cc637b170 memory-failure: e... |
1275 |
action_result(pfn, MF_MSG_KERNEL_HIGH_ORDER, MF_IGNORED); |
8d22ba1b7 HWPOISON: detect ... |
1276 1277 |
return -EBUSY; } |
6a46079cf HWPOISON: The hig... |
1278 |
} |
761ad8d7c mm: hwpoison: int... |
1279 |
if (PageTransHuge(hpage)) { |
c3901e722 mm: hwpoison: fix... |
1280 1281 1282 1283 |
lock_page(p); if (!PageAnon(p) || unlikely(split_huge_page(p))) { unlock_page(p); if (!PageAnon(p)) |
495367c05 mm/memory-failure... |
1284 1285 1286 |
pr_err("Memory failure: %#lx: non anonymous thp ", pfn); |
7f6bf39bb mm/hwpoison: fix ... |
1287 |
else |
495367c05 mm/memory-failure... |
1288 1289 1290 |
pr_err("Memory failure: %#lx: thp split failed ", pfn); |
ead07f6a8 mm/memory-failure... |
1291 |
if (TestClearPageHWPoison(p)) |
b37ff71cc mm: hwpoison: cha... |
1292 |
num_poisoned_pages_dec(); |
665d9da7f mm/hwpoison: repl... |
1293 |
put_hwpoison_page(p); |
415c64c14 mm/memory-failure... |
1294 1295 |
return -EBUSY; } |
c3901e722 mm: hwpoison: fix... |
1296 |
unlock_page(p); |
415c64c14 mm/memory-failure... |
1297 1298 1299 |
VM_BUG_ON_PAGE(!page_count(p), p); hpage = compound_head(p); } |
6a46079cf HWPOISON: The hig... |
1300 |
/* |
e43c3afb3 HWPOISON: return ... |
1301 1302 |
* We ignore non-LRU pages for good reasons. * - PG_locked is only well defined for LRU pages and a few others |
48c935ad8 page-flags: defin... |
1303 |
* - to avoid races with __SetPageLocked() |
e43c3afb3 HWPOISON: return ... |
1304 1305 1306 1307 |
* - 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. */ |
8bcb74de7 mm: hwpoison: cal... |
1308 1309 1310 1311 1312 1313 1314 1315 |
shake_page(p, 0); /* shake_page could have turned it free. */ if (!PageLRU(p) && is_free_buddy_page(p)) { if (flags & MF_COUNT_INCREASED) action_result(pfn, MF_MSG_BUDDY, MF_DELAYED); else action_result(pfn, MF_MSG_BUDDY_2ND, MF_DELAYED); return 0; |
e43c3afb3 HWPOISON: return ... |
1316 |
} |
e43c3afb3 HWPOISON: return ... |
1317 |
|
761ad8d7c mm: hwpoison: int... |
1318 |
lock_page(p); |
847ce401d HWPOISON: Add unp... |
1319 1320 |
/* |
f37d4298a hwpoison: fix rac... |
1321 1322 1323 |
* The page could have changed compound pages during the locking. * If this happens just bail out. */ |
415c64c14 mm/memory-failure... |
1324 |
if (PageCompound(p) && compound_head(p) != orig_head) { |
cc637b170 memory-failure: e... |
1325 |
action_result(pfn, MF_MSG_DIFFERENT_COMPOUND, MF_IGNORED); |
f37d4298a hwpoison: fix rac... |
1326 1327 1328 1329 1330 |
res = -EBUSY; goto out; } /* |
524fca1e7 HWPOISON: fix mis... |
1331 1332 1333 1334 1335 1336 |
* 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. */ |
7258ae5c5 mm/memory-failure... |
1337 1338 1339 1340 |
if (PageHuge(p)) page_flags = hpage->flags; else page_flags = p->flags; |
524fca1e7 HWPOISON: fix mis... |
1341 1342 |
/* |
847ce401d HWPOISON: Add unp... |
1343 1344 1345 |
* unpoison always clear PG_hwpoison inside page lock */ if (!PageHWPoison(p)) { |
495367c05 mm/memory-failure... |
1346 1347 |
pr_err("Memory failure: %#lx: just unpoisoned ", pfn); |
b37ff71cc mm: hwpoison: cha... |
1348 |
num_poisoned_pages_dec(); |
761ad8d7c mm: hwpoison: int... |
1349 1350 |
unlock_page(p); put_hwpoison_page(p); |
a09233f3e mm/memory-failure... |
1351 |
return 0; |
847ce401d HWPOISON: Add unp... |
1352 |
} |
7c116f2b0 HWPOISON: add fs/... |
1353 1354 |
if (hwpoison_filter(p)) { if (TestClearPageHWPoison(p)) |
b37ff71cc mm: hwpoison: cha... |
1355 |
num_poisoned_pages_dec(); |
761ad8d7c mm: hwpoison: int... |
1356 1357 |
unlock_page(p); put_hwpoison_page(p); |
7c116f2b0 HWPOISON: add fs/... |
1358 1359 |
return 0; } |
847ce401d HWPOISON: Add unp... |
1360 |
|
761ad8d7c mm: hwpoison: int... |
1361 |
if (!PageTransTail(p) && !PageLRU(p)) |
0bc1f8b06 hwpoison: fix the... |
1362 |
goto identify_page_state; |
7013febc8 HWPOISON, hugetlb... |
1363 |
/* |
6edd6cc66 mm/memory-failure... |
1364 1365 1366 |
* It's very difficult to mess with pages currently under IO * and in many cases impossible, so we just avoid it here. */ |
6a46079cf HWPOISON: The hig... |
1367 1368 1369 1370 |
wait_on_page_writeback(p); /* * Now take care of user space mappings. |
e64a782fe mm: change __remo... |
1371 |
* Abort on fail: __delete_from_page_cache() assumes unmapped page. |
54b9dd14d mm/memory-failure... |
1372 1373 1374 |
* * When the raw error page is thp tail page, hpage points to the raw * page after thp split. |
6a46079cf HWPOISON: The hig... |
1375 |
*/ |
83b57531c mm/memory_failure... |
1376 |
if (!hwpoison_user_mappings(p, pfn, flags, &hpage)) { |
cc637b170 memory-failure: e... |
1377 |
action_result(pfn, MF_MSG_UNMAP_FAILED, MF_IGNORED); |
1668bfd5b HWPOISON: abort o... |
1378 1379 1380 |
res = -EBUSY; goto out; } |
6a46079cf HWPOISON: The hig... |
1381 1382 1383 1384 |
/* * Torn down by someone else? */ |
dc2a1cbf7 HWPOISON: introdu... |
1385 |
if (PageLRU(p) && !PageSwapCache(p) && p->mapping == NULL) { |
cc637b170 memory-failure: e... |
1386 |
action_result(pfn, MF_MSG_TRUNCATED_LRU, MF_IGNORED); |
d95ea51e3 HWPOISON: make se... |
1387 |
res = -EBUSY; |
6a46079cf HWPOISON: The hig... |
1388 1389 |
goto out; } |
0bc1f8b06 hwpoison: fix the... |
1390 |
identify_page_state: |
0348d2ebe mm: hwpoison: int... |
1391 |
res = identify_page_state(pfn, p, page_flags); |
6a46079cf HWPOISON: The hig... |
1392 |
out: |
761ad8d7c mm: hwpoison: int... |
1393 |
unlock_page(p); |
6a46079cf HWPOISON: The hig... |
1394 1395 |
return res; } |
cd42f4a3b HWPOISON: Clean u... |
1396 |
EXPORT_SYMBOL_GPL(memory_failure); |
847ce401d HWPOISON: Add unp... |
1397 |
|
ea8f5fb8a HWPoison: add mem... |
1398 1399 1400 1401 1402 |
#define MEMORY_FAILURE_FIFO_ORDER 4 #define MEMORY_FAILURE_FIFO_SIZE (1 << MEMORY_FAILURE_FIFO_ORDER) struct memory_failure_entry { unsigned long pfn; |
ea8f5fb8a HWPoison: add mem... |
1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 |
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 |
ea8f5fb8a HWPoison: add mem... |
1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 |
* @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. */ |
83b57531c mm/memory_failure... |
1431 |
void memory_failure_queue(unsigned long pfn, int flags) |
ea8f5fb8a HWPoison: add mem... |
1432 1433 1434 1435 1436 |
{ struct memory_failure_cpu *mf_cpu; unsigned long proc_flags; struct memory_failure_entry entry = { .pfn = pfn, |
ea8f5fb8a HWPoison: add mem... |
1437 1438 1439 1440 1441 |
.flags = flags, }; mf_cpu = &get_cpu_var(memory_failure_cpu); spin_lock_irqsave(&mf_cpu->lock, proc_flags); |
498d319bb kfifo API type sa... |
1442 |
if (kfifo_put(&mf_cpu->fifo, entry)) |
ea8f5fb8a HWPoison: add mem... |
1443 1444 |
schedule_work_on(smp_processor_id(), &mf_cpu->work); else |
8e33a52fa treewide: Fix pri... |
1445 1446 |
pr_err("Memory failure: buffer overflow when queuing memory failure at %#lx ", |
ea8f5fb8a HWPoison: add mem... |
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 |
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 mm: replace __get... |
1459 |
mf_cpu = this_cpu_ptr(&memory_failure_cpu); |
ea8f5fb8a HWPoison: add mem... |
1460 1461 1462 1463 1464 1465 |
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 mce: acpi/apei: S... |
1466 1467 1468 |
if (entry.flags & MF_SOFT_OFFLINE) soft_offline_page(pfn_to_page(entry.pfn), entry.flags); else |
83b57531c mm/memory_failure... |
1469 |
memory_failure(entry.pfn, entry.flags); |
ea8f5fb8a HWPoison: add mem... |
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 |
} } 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 mm: hwpoison: rat... |
1488 1489 1490 1491 1492 |
#define unpoison_pr_info(fmt, pfn, rs) \ ({ \ if (__ratelimit(rs)) \ pr_info(fmt, pfn); \ }) |
847ce401d HWPOISON: Add unp... |
1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 |
/** * 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; |
a5f651090 mm: hwpoison: rat... |
1510 1511 |
static DEFINE_RATELIMIT_STATE(unpoison_rs, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); |
847ce401d HWPOISON: Add unp... |
1512 1513 1514 1515 1516 1517 1518 1519 |
if (!pfn_valid(pfn)) return -ENXIO; p = pfn_to_page(pfn); page = compound_head(p); if (!PageHWPoison(p)) { |
495367c05 mm/memory-failure... |
1520 1521 |
unpoison_pr_info("Unpoison: Page was already unpoisoned %#lx ", |
a5f651090 mm: hwpoison: rat... |
1522 |
pfn, &unpoison_rs); |
847ce401d HWPOISON: Add unp... |
1523 1524 |
return 0; } |
230ac719c mm/hwpoison: don'... |
1525 |
if (page_count(page) > 1) { |
495367c05 mm/memory-failure... |
1526 1527 |
unpoison_pr_info("Unpoison: Someone grabs the hwpoison page %#lx ", |
a5f651090 mm: hwpoison: rat... |
1528 |
pfn, &unpoison_rs); |
230ac719c mm/hwpoison: don'... |
1529 1530 1531 1532 |
return 0; } if (page_mapped(page)) { |
495367c05 mm/memory-failure... |
1533 1534 |
unpoison_pr_info("Unpoison: Someone maps the hwpoison page %#lx ", |
a5f651090 mm: hwpoison: rat... |
1535 |
pfn, &unpoison_rs); |
230ac719c mm/hwpoison: don'... |
1536 1537 1538 1539 |
return 0; } if (page_mapping(page)) { |
495367c05 mm/memory-failure... |
1540 1541 |
unpoison_pr_info("Unpoison: the hwpoison page has non-NULL mapping %#lx ", |
a5f651090 mm: hwpoison: rat... |
1542 |
pfn, &unpoison_rs); |
230ac719c mm/hwpoison: don'... |
1543 1544 |
return 0; } |
0cea3fdc4 mm/hwpoison: fix ... |
1545 1546 1547 1548 1549 |
/* * 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 mm/hwpoison: fix ... |
1550 |
if (!PageHuge(page) && PageTransHuge(page)) { |
495367c05 mm/memory-failure... |
1551 1552 |
unpoison_pr_info("Unpoison: Memory failure is now running on %#lx ", |
a5f651090 mm: hwpoison: rat... |
1553 |
pfn, &unpoison_rs); |
ead07f6a8 mm/memory-failure... |
1554 |
return 0; |
0cea3fdc4 mm/hwpoison: fix ... |
1555 |
} |
ead07f6a8 mm/memory-failure... |
1556 |
if (!get_hwpoison_page(p)) { |
847ce401d HWPOISON: Add unp... |
1557 |
if (TestClearPageHWPoison(p)) |
8e30456b6 mm/hwpoison: intr... |
1558 |
num_poisoned_pages_dec(); |
495367c05 mm/memory-failure... |
1559 1560 |
unpoison_pr_info("Unpoison: Software-unpoisoned free page %#lx ", |
a5f651090 mm: hwpoison: rat... |
1561 |
pfn, &unpoison_rs); |
847ce401d HWPOISON: Add unp... |
1562 1563 |
return 0; } |
7eaceacca block: remove per... |
1564 |
lock_page(page); |
847ce401d HWPOISON: Add unp... |
1565 1566 1567 1568 1569 1570 |
/* * 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 HWPOISON, hugetlb... |
1571 |
if (TestClearPageHWPoison(page)) { |
495367c05 mm/memory-failure... |
1572 1573 |
unpoison_pr_info("Unpoison: Software-unpoisoned page %#lx ", |
a5f651090 mm: hwpoison: rat... |
1574 |
pfn, &unpoison_rs); |
b37ff71cc mm: hwpoison: cha... |
1575 |
num_poisoned_pages_dec(); |
847ce401d HWPOISON: Add unp... |
1576 1577 1578 |
freeit = 1; } unlock_page(page); |
665d9da7f mm/hwpoison: repl... |
1579 |
put_hwpoison_page(page); |
3ba5eebc4 mm/memory-failure... |
1580 |
if (freeit && !(pfn == my_zero_pfn(0) && page_count(p) == 1)) |
665d9da7f mm/hwpoison: repl... |
1581 |
put_hwpoison_page(page); |
847ce401d HWPOISON: Add unp... |
1582 1583 1584 1585 |
return 0; } EXPORT_SYMBOL(unpoison_memory); |
facb6011f HWPOISON: Add sof... |
1586 |
|
666feb21a mm, migrate: remo... |
1587 |
static struct page *new_page(struct page *p, unsigned long private) |
facb6011f HWPOISON: Add sof... |
1588 |
{ |
12686d153 HWPOISON: Try to ... |
1589 |
int nid = page_to_nid(p); |
94310cbca mm/madvise: enabl... |
1590 |
|
ef77ba5ce mm, hugetlb, soft... |
1591 |
return new_page_nodemask(p, nid, &node_states[N_MEMORY]); |
facb6011f HWPOISON: Add sof... |
1592 1593 1594 1595 1596 1597 1598 1599 |
} /* * 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 mm/memory-failure... |
1600 |
static int __get_any_page(struct page *p, unsigned long pfn, int flags) |
facb6011f HWPOISON: Add sof... |
1601 1602 1603 1604 1605 1606 1607 |
{ int ret; if (flags & MF_COUNT_INCREASED) return 1; /* |
d950b9588 HWPOISON, hugetlb... |
1608 1609 1610 |
* When the target page is a free hugepage, just remove it * from free hugepage list. */ |
ead07f6a8 mm/memory-failure... |
1611 |
if (!get_hwpoison_page(p)) { |
d950b9588 HWPOISON, hugetlb... |
1612 |
if (PageHuge(p)) { |
71dd0b8ae mm/memory_failure... |
1613 1614 |
pr_info("%s: %#lx free huge page ", __func__, pfn); |
af8fae7c0 mm/memory-failure... |
1615 |
ret = 0; |
d950b9588 HWPOISON, hugetlb... |
1616 |
} else if (is_free_buddy_page(p)) { |
71dd0b8ae mm/memory_failure... |
1617 1618 |
pr_info("%s: %#lx free buddy page ", __func__, pfn); |
facb6011f HWPOISON: Add sof... |
1619 1620 |
ret = 0; } else { |
71dd0b8ae mm/memory_failure... |
1621 1622 1623 |
pr_info("%s: %#lx: unknown zero refcount page type %lx ", __func__, pfn, p->flags); |
facb6011f HWPOISON: Add sof... |
1624 1625 1626 1627 1628 1629 |
ret = -EIO; } } else { /* Not a free page */ ret = 1; } |
facb6011f HWPOISON: Add sof... |
1630 1631 |
return ret; } |
af8fae7c0 mm/memory-failure... |
1632 1633 1634 |
static int get_any_page(struct page *page, unsigned long pfn, int flags) { int ret = __get_any_page(page, pfn, flags); |
85fbe5d1b HWPOISON: soft of... |
1635 1636 |
if (ret == 1 && !PageHuge(page) && !PageLRU(page) && !__PageMovable(page)) { |
af8fae7c0 mm/memory-failure... |
1637 1638 1639 |
/* * Try to free it. */ |
665d9da7f mm/hwpoison: repl... |
1640 |
put_hwpoison_page(page); |
af8fae7c0 mm/memory-failure... |
1641 1642 1643 1644 1645 1646 |
shake_page(page, 1); /* * Did it turn free? */ ret = __get_any_page(page, pfn, 0); |
d96b339f4 mm: soft-offline:... |
1647 |
if (ret == 1 && !PageLRU(page)) { |
4f32be677 mm/hwpoison: fix ... |
1648 |
/* Drop page reference which is from __get_any_page() */ |
665d9da7f mm/hwpoison: repl... |
1649 |
put_hwpoison_page(page); |
82a2481e8 mm/memory-failure... |
1650 1651 1652 |
pr_info("soft_offline: %#lx: unknown non LRU page type %lx (%pGp) ", pfn, page->flags, &page->flags); |
af8fae7c0 mm/memory-failure... |
1653 1654 1655 1656 1657 |
return -EIO; } } return ret; } |
d950b9588 HWPOISON, hugetlb... |
1658 1659 1660 1661 1662 |
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 mm: soft-offline:... |
1663 |
LIST_HEAD(pagelist); |
d950b9588 HWPOISON, hugetlb... |
1664 |
|
af8fae7c0 mm/memory-failure... |
1665 1666 1667 1668 1669 |
/* * This double-check of PageHWPoison is to avoid the race with * memory_failure(). See also comment in __soft_offline_page(). */ lock_page(hpage); |
0ebff32c3 memory-failure: f... |
1670 |
if (PageHWPoison(hpage)) { |
af8fae7c0 mm/memory-failure... |
1671 |
unlock_page(hpage); |
665d9da7f mm/hwpoison: repl... |
1672 |
put_hwpoison_page(hpage); |
0ebff32c3 memory-failure: f... |
1673 1674 |
pr_info("soft offline: %#lx hugepage already poisoned ", pfn); |
af8fae7c0 mm/memory-failure... |
1675 |
return -EBUSY; |
0ebff32c3 memory-failure: f... |
1676 |
} |
af8fae7c0 mm/memory-failure... |
1677 |
unlock_page(hpage); |
d950b9588 HWPOISON, hugetlb... |
1678 |
|
bcc542223 mm: hugetlb: intr... |
1679 |
ret = isolate_huge_page(hpage, &pagelist); |
036138080 mm/hwpoison: fix ... |
1680 1681 1682 1683 |
/* * get_any_page() and isolate_huge_page() takes a refcount each, * so need to drop one here. */ |
665d9da7f mm/hwpoison: repl... |
1684 |
put_hwpoison_page(hpage); |
036138080 mm/hwpoison: fix ... |
1685 |
if (!ret) { |
bcc542223 mm: hugetlb: intr... |
1686 1687 1688 1689 |
pr_info("soft offline: %#lx hugepage failed to isolate ", pfn); return -EBUSY; } |
68711a746 mm, migration: ad... |
1690 |
ret = migrate_pages(&pagelist, new_page, NULL, MPOL_MF_MOVE_ALL, |
b8ec1cee5 mm: soft-offline:... |
1691 |
MIGRATE_SYNC, MR_MEMORY_FAILURE); |
d950b9588 HWPOISON, hugetlb... |
1692 |
if (ret) { |
b6b18aa87 mm, soft_offline:... |
1693 1694 |
pr_info("soft offline: %#lx: hugepage migration failed %d, type %lx (%pGp) ", |
82a2481e8 mm/memory-failure... |
1695 |
pfn, ret, page->flags, &page->flags); |
30809f559 mm/migrate: fix r... |
1696 1697 |
if (!list_empty(&pagelist)) putback_movable_pages(&pagelist); |
b8ec1cee5 mm: soft-offline:... |
1698 1699 |
if (ret > 0) ret = -EIO; |
af8fae7c0 mm/memory-failure... |
1700 |
} else { |
6bc9b5643 mm: fix race on s... |
1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 |
/* * We set PG_hwpoison only when the migration source hugepage * was successfully dissolved, because otherwise hwpoisoned * hugepage remains on free hugepage list, then userspace will * find it as SIGBUS by allocation failure. That's not expected * in soft-offlining. */ ret = dissolve_free_huge_page(page); if (!ret) { if (set_hwpoison_free_buddy_page(page)) num_poisoned_pages_inc(); |
b38e5962f mm: soft-offline:... |
1712 1713 |
else ret = -EBUSY; |
6bc9b5643 mm: fix race on s... |
1714 |
} |
d950b9588 HWPOISON, hugetlb... |
1715 |
} |
d950b9588 HWPOISON, hugetlb... |
1716 1717 |
return ret; } |
af8fae7c0 mm/memory-failure... |
1718 1719 1720 1721 |
static int __soft_offline_page(struct page *page, int flags) { int ret; unsigned long pfn = page_to_pfn(page); |
facb6011f HWPOISON: Add sof... |
1722 |
|
facb6011f HWPOISON: Add sof... |
1723 |
/* |
af8fae7c0 mm/memory-failure... |
1724 1725 1726 1727 |
* 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 HWPOISON: Add sof... |
1728 |
*/ |
0ebff32c3 memory-failure: f... |
1729 1730 |
lock_page(page); wait_on_page_writeback(page); |
af8fae7c0 mm/memory-failure... |
1731 1732 |
if (PageHWPoison(page)) { unlock_page(page); |
665d9da7f mm/hwpoison: repl... |
1733 |
put_hwpoison_page(page); |
af8fae7c0 mm/memory-failure... |
1734 1735 1736 1737 |
pr_info("soft offline: %#lx page already poisoned ", pfn); return -EBUSY; } |
facb6011f HWPOISON: Add sof... |
1738 1739 1740 1741 1742 1743 |
/* * Try to invalidate first. This should work for * non dirty unmapped page cache pages. */ ret = invalidate_inode_page(page); unlock_page(page); |
facb6011f HWPOISON: Add sof... |
1744 |
/* |
facb6011f HWPOISON: Add sof... |
1745 1746 1747 |
* RED-PEN would be better to keep it isolated here, but we * would need to fix isolation locking first. */ |
facb6011f HWPOISON: Add sof... |
1748 |
if (ret == 1) { |
665d9da7f mm/hwpoison: repl... |
1749 |
put_hwpoison_page(page); |
fb46e7352 HWPOISON: Convert... |
1750 1751 |
pr_info("soft_offline: %#lx: invalidated ", pfn); |
af8fae7c0 mm/memory-failure... |
1752 |
SetPageHWPoison(page); |
8e30456b6 mm/hwpoison: intr... |
1753 |
num_poisoned_pages_inc(); |
af8fae7c0 mm/memory-failure... |
1754 |
return 0; |
facb6011f HWPOISON: Add sof... |
1755 1756 1757 1758 1759 1760 1761 |
} /* * Simple invalidation didn't work. * Try to migrate to a new page instead. migrate.c * handles a large number of cases for us. */ |
85fbe5d1b HWPOISON: soft of... |
1762 1763 1764 1765 |
if (PageLRU(page)) ret = isolate_lru_page(page); else ret = isolate_movable_page(page, ISOLATE_UNEVICTABLE); |
bd486285f mem-hwpoison: fix... |
1766 1767 1768 1769 |
/* * Drop page reference which is came from get_any_page() * successful isolate_lru_page() already took another one. */ |
665d9da7f mm/hwpoison: repl... |
1770 |
put_hwpoison_page(page); |
facb6011f HWPOISON: Add sof... |
1771 1772 |
if (!ret) { LIST_HEAD(pagelist); |
85fbe5d1b HWPOISON: soft of... |
1773 1774 1775 1776 1777 1778 1779 1780 |
/* * After isolated lru page, the PageLRU will be cleared, * so use !__PageMovable instead for LRU page's mapping * cannot have PAGE_MAPPING_MOVABLE. */ if (!__PageMovable(page)) inc_node_page_state(page, NR_ISOLATED_ANON + page_is_file_cache(page)); |
facb6011f HWPOISON: Add sof... |
1781 |
list_add(&page->lru, &pagelist); |
68711a746 mm, migration: ad... |
1782 |
ret = migrate_pages(&pagelist, new_page, NULL, MPOL_MF_MOVE_ALL, |
9c620e2bc mm: remove offlin... |
1783 |
MIGRATE_SYNC, MR_MEMORY_FAILURE); |
facb6011f HWPOISON: Add sof... |
1784 |
if (ret) { |
85fbe5d1b HWPOISON: soft of... |
1785 1786 |
if (!list_empty(&pagelist)) putback_movable_pages(&pagelist); |
59c82b70d mm/migrate: remov... |
1787 |
|
82a2481e8 mm/memory-failure... |
1788 1789 1790 |
pr_info("soft offline: %#lx: migration failed %d, type %lx (%pGp) ", pfn, ret, page->flags, &page->flags); |
facb6011f HWPOISON: Add sof... |
1791 1792 1793 1794 |
if (ret > 0) ret = -EIO; } } else { |
82a2481e8 mm/memory-failure... |
1795 1796 1797 |
pr_info("soft offline: %#lx: isolation failed: %d, page count %d, type %lx (%pGp) ", pfn, ret, page_count(page), page->flags, &page->flags); |
facb6011f HWPOISON: Add sof... |
1798 |
} |
facb6011f HWPOISON: Add sof... |
1799 1800 |
return ret; } |
86e057734 mm/hwpoison: drop... |
1801 |
|
acc14dc4b mm: soft-offline:... |
1802 1803 1804 |
static int soft_offline_in_use_page(struct page *page, int flags) { int ret; |
d4ae9916e mm: soft-offline:... |
1805 |
int mt; |
acc14dc4b mm: soft-offline:... |
1806 1807 1808 |
struct page *hpage = compound_head(page); if (!PageHuge(page) && PageTransHuge(hpage)) { |
46612b751 mm: hwpoison: fix... |
1809 1810 1811 1812 |
lock_page(page); if (!PageAnon(page) || unlikely(split_huge_page(page))) { unlock_page(page); if (!PageAnon(page)) |
98fd1ef42 mm: soft-offline:... |
1813 1814 1815 1816 1817 |
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)); |
46612b751 mm: hwpoison: fix... |
1818 |
put_hwpoison_page(page); |
acc14dc4b mm: soft-offline:... |
1819 1820 |
return -EBUSY; } |
46612b751 mm: hwpoison: fix... |
1821 |
unlock_page(page); |
acc14dc4b mm: soft-offline:... |
1822 |
} |
d4ae9916e mm: soft-offline:... |
1823 1824 1825 1826 1827 1828 1829 1830 1831 |
/* * Setting MIGRATE_ISOLATE here ensures that the page will be linked * to free list immediately (not via pcplist) when released after * successful page migration. Otherwise we can't guarantee that the * page is really free after put_page() returns, so * set_hwpoison_free_buddy_page() highly likely fails. */ mt = get_pageblock_migratetype(page); set_pageblock_migratetype(page, MIGRATE_ISOLATE); |
acc14dc4b mm: soft-offline:... |
1832 1833 1834 1835 |
if (PageHuge(page)) ret = soft_offline_huge_page(page, flags); else ret = __soft_offline_page(page, flags); |
d4ae9916e mm: soft-offline:... |
1836 |
set_pageblock_migratetype(page, mt); |
acc14dc4b mm: soft-offline:... |
1837 1838 |
return ret; } |
d4ae9916e mm: soft-offline:... |
1839 |
static int soft_offline_free_page(struct page *page) |
acc14dc4b mm: soft-offline:... |
1840 |
{ |
faf53def3 mm: hugetlb: soft... |
1841 |
int rc = dissolve_free_huge_page(page); |
acc14dc4b mm: soft-offline:... |
1842 |
|
d4ae9916e mm: soft-offline:... |
1843 1844 1845 1846 1847 1848 1849 |
if (!rc) { if (set_hwpoison_free_buddy_page(page)) num_poisoned_pages_inc(); else rc = -EBUSY; } return rc; |
acc14dc4b mm: soft-offline:... |
1850 |
} |
86e057734 mm/hwpoison: drop... |
1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 |
/** * 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 mm/hwpoison: drop... |
1877 |
|
86a66810b mm, madvise_injec... |
1878 1879 1880 1881 1882 1883 1884 1885 |
if (is_zone_device_page(page)) { pr_debug_ratelimited("soft_offline: %#lx page is device page ", pfn); if (flags & MF_COUNT_INCREASED) put_page(page); return -EIO; } |
86e057734 mm/hwpoison: drop... |
1886 1887 1888 |
if (PageHWPoison(page)) { pr_info("soft offline: %#lx page already poisoned ", pfn); |
1e0e635be mm/hwpoison: fix ... |
1889 |
if (flags & MF_COUNT_INCREASED) |
665d9da7f mm/hwpoison: repl... |
1890 |
put_hwpoison_page(page); |
86e057734 mm/hwpoison: drop... |
1891 1892 |
return -EBUSY; } |
86e057734 mm/hwpoison: drop... |
1893 |
|
bfc8c9013 mem-hotplug: impl... |
1894 |
get_online_mems(); |
86e057734 mm/hwpoison: drop... |
1895 |
ret = get_any_page(page, pfn, flags); |
bfc8c9013 mem-hotplug: impl... |
1896 |
put_online_mems(); |
4e41a30c6 mm: hwpoison: adj... |
1897 |
|
acc14dc4b mm: soft-offline:... |
1898 1899 1900 |
if (ret > 0) ret = soft_offline_in_use_page(page, flags); else if (ret == 0) |
d4ae9916e mm: soft-offline:... |
1901 |
ret = soft_offline_free_page(page); |
4e41a30c6 mm: hwpoison: adj... |
1902 |
|
86e057734 mm/hwpoison: drop... |
1903 1904 |
return ret; } |