Blame view
mm/mmu_notifier.c
9.49 KB
cddb8a5c1
|
1 2 3 4 5 6 7 8 9 10 11 12 13 |
/* * linux/mm/mmu_notifier.c * * Copyright (C) 2008 Qumranet, Inc. * Copyright (C) 2008 SGI * Christoph Lameter <clameter@sgi.com> * * This work is licensed under the terms of the GNU GPL, version 2. See * the COPYING file in the top-level directory. */ #include <linux/rculist.h> #include <linux/mmu_notifier.h> |
b95f1b31b
|
14 |
#include <linux/export.h> |
cddb8a5c1
|
15 16 |
#include <linux/mm.h> #include <linux/err.h> |
21a92735f
|
17 |
#include <linux/srcu.h> |
cddb8a5c1
|
18 19 |
#include <linux/rcupdate.h> #include <linux/sched.h> |
5a0e3ad6a
|
20 |
#include <linux/slab.h> |
cddb8a5c1
|
21 |
|
21a92735f
|
22 |
/* global SRCU for all MMs */ |
70400303c
|
23 |
static struct srcu_struct srcu; |
21a92735f
|
24 |
|
cddb8a5c1
|
25 26 27 28 29 30 31 |
/* * This function can't run concurrently against mmu_notifier_register * because mm->mm_users > 0 during mmu_notifier_register and exit_mmap * runs with mm_users == 0. Other tasks may still invoke mmu notifiers * in parallel despite there being no task using this mm any more, * through the vmas outside of the exit_mmap context, such as with * vmtruncate. This serializes against mmu_notifier_unregister with |
21a92735f
|
32 33 |
* the mmu_notifier_mm->lock in addition to SRCU and it serializes * against the other mmu notifiers with SRCU. struct mmu_notifier_mm |
cddb8a5c1
|
34 35 36 37 38 39 |
* can't go away from under us as exit_mmap holds an mm_count pin * itself. */ void __mmu_notifier_release(struct mm_struct *mm) { struct mmu_notifier *mn; |
21a92735f
|
40 |
int id; |
3ad3d901b
|
41 42 |
/* |
d34883d4e
|
43 44 |
* SRCU here will block mmu_notifier_unregister until * ->release returns. |
3ad3d901b
|
45 |
*/ |
21a92735f
|
46 |
id = srcu_read_lock(&srcu); |
d34883d4e
|
47 48 49 50 51 52 53 54 55 56 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) /* * If ->release runs before mmu_notifier_unregister it must be * handled, as it's the only way for the driver to flush all * existing sptes and stop the driver from establishing any more * sptes before all the pages in the mm are freed. */ if (mn->ops->release) mn->ops->release(mn, mm); srcu_read_unlock(&srcu, id); |
cddb8a5c1
|
57 58 59 60 61 62 |
spin_lock(&mm->mmu_notifier_mm->lock); while (unlikely(!hlist_empty(&mm->mmu_notifier_mm->list))) { mn = hlist_entry(mm->mmu_notifier_mm->list.first, struct mmu_notifier, hlist); /* |
d34883d4e
|
63 64 65 66 |
* We arrived before mmu_notifier_unregister so * mmu_notifier_unregister will do nothing other than to wait * for ->release to finish and for mmu_notifier_unregister to * return. |
cddb8a5c1
|
67 68 |
*/ hlist_del_init_rcu(&mn->hlist); |
cddb8a5c1
|
69 70 71 72 |
} spin_unlock(&mm->mmu_notifier_mm->lock); /* |
d34883d4e
|
73 74 75 76 77 78 79 |
* synchronize_srcu here prevents mmu_notifier_release from returning to * exit_mmap (which would proceed with freeing all pages in the mm) * until the ->release method returns, if it was invoked by * mmu_notifier_unregister. * * The mmu_notifier_mm can't go away from under us because one mm_count * is held by exit_mmap. |
cddb8a5c1
|
80 |
*/ |
21a92735f
|
81 |
synchronize_srcu(&srcu); |
cddb8a5c1
|
82 83 84 85 86 87 88 89 90 91 92 |
} /* * If no young bitflag is supported by the hardware, ->clear_flush_young can * unmap the address and return 1 or 0 depending if the mapping previously * existed or not. */ int __mmu_notifier_clear_flush_young(struct mm_struct *mm, unsigned long address) { struct mmu_notifier *mn; |
21a92735f
|
93 |
int young = 0, id; |
cddb8a5c1
|
94 |
|
21a92735f
|
95 |
id = srcu_read_lock(&srcu); |
b67bfe0d4
|
96 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) { |
cddb8a5c1
|
97 98 99 |
if (mn->ops->clear_flush_young) young |= mn->ops->clear_flush_young(mn, mm, address); } |
21a92735f
|
100 |
srcu_read_unlock(&srcu, id); |
cddb8a5c1
|
101 102 103 |
return young; } |
8ee53820e
|
104 105 106 107 |
int __mmu_notifier_test_young(struct mm_struct *mm, unsigned long address) { struct mmu_notifier *mn; |
21a92735f
|
108 |
int young = 0, id; |
8ee53820e
|
109 |
|
21a92735f
|
110 |
id = srcu_read_lock(&srcu); |
b67bfe0d4
|
111 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) { |
8ee53820e
|
112 113 114 115 116 117 |
if (mn->ops->test_young) { young = mn->ops->test_young(mn, mm, address); if (young) break; } } |
21a92735f
|
118 |
srcu_read_unlock(&srcu, id); |
8ee53820e
|
119 120 121 |
return young; } |
828502d30
|
122 123 124 125 |
void __mmu_notifier_change_pte(struct mm_struct *mm, unsigned long address, pte_t pte) { struct mmu_notifier *mn; |
21a92735f
|
126 |
int id; |
828502d30
|
127 |
|
21a92735f
|
128 |
id = srcu_read_lock(&srcu); |
b67bfe0d4
|
129 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) { |
828502d30
|
130 131 |
if (mn->ops->change_pte) mn->ops->change_pte(mn, mm, address, pte); |
828502d30
|
132 |
} |
21a92735f
|
133 |
srcu_read_unlock(&srcu, id); |
828502d30
|
134 |
} |
cddb8a5c1
|
135 136 137 138 |
void __mmu_notifier_invalidate_page(struct mm_struct *mm, unsigned long address) { struct mmu_notifier *mn; |
21a92735f
|
139 |
int id; |
cddb8a5c1
|
140 |
|
21a92735f
|
141 |
id = srcu_read_lock(&srcu); |
b67bfe0d4
|
142 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) { |
cddb8a5c1
|
143 144 145 |
if (mn->ops->invalidate_page) mn->ops->invalidate_page(mn, mm, address); } |
21a92735f
|
146 |
srcu_read_unlock(&srcu, id); |
cddb8a5c1
|
147 148 149 150 151 152 |
} void __mmu_notifier_invalidate_range_start(struct mm_struct *mm, unsigned long start, unsigned long end) { struct mmu_notifier *mn; |
21a92735f
|
153 |
int id; |
cddb8a5c1
|
154 |
|
21a92735f
|
155 |
id = srcu_read_lock(&srcu); |
b67bfe0d4
|
156 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) { |
cddb8a5c1
|
157 158 159 |
if (mn->ops->invalidate_range_start) mn->ops->invalidate_range_start(mn, mm, start, end); } |
21a92735f
|
160 |
srcu_read_unlock(&srcu, id); |
cddb8a5c1
|
161 |
} |
fa794199e
|
162 |
EXPORT_SYMBOL_GPL(__mmu_notifier_invalidate_range_start); |
cddb8a5c1
|
163 164 165 166 167 |
void __mmu_notifier_invalidate_range_end(struct mm_struct *mm, unsigned long start, unsigned long end) { struct mmu_notifier *mn; |
21a92735f
|
168 |
int id; |
cddb8a5c1
|
169 |
|
21a92735f
|
170 |
id = srcu_read_lock(&srcu); |
b67bfe0d4
|
171 |
hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list, hlist) { |
cddb8a5c1
|
172 173 174 |
if (mn->ops->invalidate_range_end) mn->ops->invalidate_range_end(mn, mm, start, end); } |
21a92735f
|
175 |
srcu_read_unlock(&srcu, id); |
cddb8a5c1
|
176 |
} |
fa794199e
|
177 |
EXPORT_SYMBOL_GPL(__mmu_notifier_invalidate_range_end); |
cddb8a5c1
|
178 179 180 181 182 183 184 185 186 |
static int do_mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm, int take_mmap_sem) { struct mmu_notifier_mm *mmu_notifier_mm; int ret; BUG_ON(atomic_read(&mm->mm_users) <= 0); |
21a92735f
|
187 |
/* |
35cfa2b0b
|
188 189 190 |
* Verify that mmu_notifier_init() already run and the global srcu is * initialized. */ |
21a92735f
|
191 |
BUG_ON(!srcu.per_cpu_ref); |
35cfa2b0b
|
192 193 194 195 |
ret = -ENOMEM; mmu_notifier_mm = kmalloc(sizeof(struct mmu_notifier_mm), GFP_KERNEL); if (unlikely(!mmu_notifier_mm)) goto out; |
cddb8a5c1
|
196 197 198 199 |
if (take_mmap_sem) down_write(&mm->mmap_sem); ret = mm_take_all_locks(mm); if (unlikely(ret)) |
35cfa2b0b
|
200 |
goto out_clean; |
cddb8a5c1
|
201 202 203 204 |
if (!mm_has_notifiers(mm)) { INIT_HLIST_HEAD(&mmu_notifier_mm->list); spin_lock_init(&mmu_notifier_mm->lock); |
e0f3c3f78
|
205 |
|
cddb8a5c1
|
206 |
mm->mmu_notifier_mm = mmu_notifier_mm; |
35cfa2b0b
|
207 |
mmu_notifier_mm = NULL; |
cddb8a5c1
|
208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 |
} atomic_inc(&mm->mm_count); /* * Serialize the update against mmu_notifier_unregister. A * side note: mmu_notifier_release can't run concurrently with * us because we hold the mm_users pin (either implicitly as * current->mm or explicitly with get_task_mm() or similar). * We can't race against any other mmu notifier method either * thanks to mm_take_all_locks(). */ spin_lock(&mm->mmu_notifier_mm->lock); hlist_add_head(&mn->hlist, &mm->mmu_notifier_mm->list); spin_unlock(&mm->mmu_notifier_mm->lock); mm_drop_all_locks(mm); |
35cfa2b0b
|
224 |
out_clean: |
cddb8a5c1
|
225 226 |
if (take_mmap_sem) up_write(&mm->mmap_sem); |
35cfa2b0b
|
227 228 |
kfree(mmu_notifier_mm); out: |
cddb8a5c1
|
229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 |
BUG_ON(atomic_read(&mm->mm_users) <= 0); return ret; } /* * Must not hold mmap_sem nor any other VM related lock when calling * this registration function. Must also ensure mm_users can't go down * to zero while this runs to avoid races with mmu_notifier_release, * so mm has to be current->mm or the mm should be pinned safely such * as with get_task_mm(). If the mm is not current->mm, the mm_users * pin should be released by calling mmput after mmu_notifier_register * returns. mmu_notifier_unregister must be always called to * unregister the notifier. mm_count is automatically pinned to allow * mmu_notifier_unregister to safely run at any time later, before or * after exit_mmap. ->release will always be called before exit_mmap * frees the pages. */ int mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm) { return do_mmu_notifier_register(mn, mm, 1); } EXPORT_SYMBOL_GPL(mmu_notifier_register); /* * Same as mmu_notifier_register but here the caller must hold the * mmap_sem in write mode. */ int __mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm) { return do_mmu_notifier_register(mn, mm, 0); } EXPORT_SYMBOL_GPL(__mmu_notifier_register); /* this is called after the last mmu_notifier_unregister() returned */ void __mmu_notifier_mm_destroy(struct mm_struct *mm) { BUG_ON(!hlist_empty(&mm->mmu_notifier_mm->list)); kfree(mm->mmu_notifier_mm); mm->mmu_notifier_mm = LIST_POISON1; /* debug */ } /* * This releases the mm_count pin automatically and frees the mm * structure if it was the last user of it. It serializes against |
21a92735f
|
273 274 |
* running mmu notifiers with SRCU and against mmu_notifier_unregister * with the unregister lock + SRCU. All sptes must be dropped before |
cddb8a5c1
|
275 276 277 278 279 280 281 282 |
* calling mmu_notifier_unregister. ->release or any other notifier * method may be invoked concurrently with mmu_notifier_unregister, * and only after mmu_notifier_unregister returned we're guaranteed * that ->release or any other method can't run anymore. */ void mmu_notifier_unregister(struct mmu_notifier *mn, struct mm_struct *mm) { BUG_ON(atomic_read(&mm->mm_count) <= 0); |
cddb8a5c1
|
283 |
if (!hlist_unhashed(&mn->hlist)) { |
d34883d4e
|
284 285 286 287 |
/* * SRCU here will force exit_mmap to wait for ->release to * finish before freeing the pages. */ |
21a92735f
|
288 |
int id; |
3ad3d901b
|
289 |
|
d34883d4e
|
290 |
id = srcu_read_lock(&srcu); |
cddb8a5c1
|
291 |
/* |
d34883d4e
|
292 293 |
* exit_mmap will block in mmu_notifier_release to guarantee * that ->release is called before freeing the pages. |
cddb8a5c1
|
294 295 296 |
*/ if (mn->ops->release) mn->ops->release(mn, mm); |
d34883d4e
|
297 |
srcu_read_unlock(&srcu, id); |
3ad3d901b
|
298 |
|
d34883d4e
|
299 |
spin_lock(&mm->mmu_notifier_mm->lock); |
751efd861
|
300 |
/* |
d34883d4e
|
301 302 |
* Can not use list_del_rcu() since __mmu_notifier_release * can delete it before we hold the lock. |
751efd861
|
303 |
*/ |
d34883d4e
|
304 |
hlist_del_init_rcu(&mn->hlist); |
cddb8a5c1
|
305 |
spin_unlock(&mm->mmu_notifier_mm->lock); |
d34883d4e
|
306 |
} |
cddb8a5c1
|
307 308 |
/* |
d34883d4e
|
309 |
* Wait for any running method to finish, of course including |
83a35e360
|
310 |
* ->release if it was run by mmu_notifier_release instead of us. |
cddb8a5c1
|
311 |
*/ |
21a92735f
|
312 |
synchronize_srcu(&srcu); |
cddb8a5c1
|
313 314 315 316 317 318 |
BUG_ON(atomic_read(&mm->mm_count) <= 0); mmdrop(mm); } EXPORT_SYMBOL_GPL(mmu_notifier_unregister); |
21a92735f
|
319 320 321 322 323 324 325 |
static int __init mmu_notifier_init(void) { return init_srcu_struct(&srcu); } module_init(mmu_notifier_init); |