Blame view
mm/backing-dev.c
27.7 KB
3fcfab16c [PATCH] separate ... |
1 2 3 |
#include <linux/wait.h> #include <linux/backing-dev.h> |
03ba3782e writeback: switch... |
4 5 |
#include <linux/kthread.h> #include <linux/freezer.h> |
3fcfab16c [PATCH] separate ... |
6 |
#include <linux/fs.h> |
26160158d Move the default_... |
7 |
#include <linux/pagemap.h> |
03ba3782e writeback: switch... |
8 |
#include <linux/mm.h> |
3fcfab16c [PATCH] separate ... |
9 10 |
#include <linux/sched.h> #include <linux/module.h> |
cf0ca9fe5 mm: bdi: export B... |
11 12 |
#include <linux/writeback.h> #include <linux/device.h> |
455b28646 writeback: Initia... |
13 |
#include <trace/events/writeback.h> |
cf0ca9fe5 mm: bdi: export B... |
14 |
|
5129a469a Catch filesystems... |
15 16 |
struct backing_dev_info noop_backing_dev_info = { .name = "noop", |
976e48f8a bdi: Initialize n... |
17 |
.capabilities = BDI_CAP_NO_ACCT_AND_WRITEBACK, |
5129a469a Catch filesystems... |
18 |
}; |
a212b105b bdi: make inode_t... |
19 |
EXPORT_SYMBOL_GPL(noop_backing_dev_info); |
5129a469a Catch filesystems... |
20 |
|
cf0ca9fe5 mm: bdi: export B... |
21 |
static struct class *bdi_class; |
cfc4ba536 writeback: use RC... |
22 23 |
/* |
181387da2 writeback: remove... |
24 |
* bdi_lock protects updates to bdi_list. bdi_list has RCU reader side |
cfc4ba536 writeback: use RC... |
25 26 |
* locking. */ |
03ba3782e writeback: switch... |
27 |
DEFINE_SPINLOCK(bdi_lock); |
66f3b8e2e writeback: move d... |
28 |
LIST_HEAD(bdi_list); |
03ba3782e writeback: switch... |
29 |
|
839a8e866 writeback: replac... |
30 31 |
/* bdi_wq serves all asynchronous writeback tasks */ struct workqueue_struct *bdi_wq; |
76f1418b4 mm: bdi: move sta... |
32 33 34 35 36 37 38 39 40 41 42 43 44 45 |
#ifdef CONFIG_DEBUG_FS #include <linux/debugfs.h> #include <linux/seq_file.h> static struct dentry *bdi_debug_root; static void bdi_debug_init(void) { bdi_debug_root = debugfs_create_dir("bdi", NULL); } static int bdi_debug_stats_show(struct seq_file *m, void *v) { struct backing_dev_info *bdi = m->private; |
c1955ce32 writeback: remove... |
46 |
struct bdi_writeback *wb = &bdi->wb; |
364aeb284 mm: change dirty ... |
47 48 |
unsigned long background_thresh; unsigned long dirty_thresh; |
0d960a383 writeback: clean ... |
49 |
unsigned long wb_thresh; |
0ae45f63d vfs: add support ... |
50 |
unsigned long nr_dirty, nr_io, nr_more_io, nr_dirty_time; |
f09b00d3e writeback: add so... |
51 |
struct inode *inode; |
0ae45f63d vfs: add support ... |
52 |
nr_dirty = nr_io = nr_more_io = nr_dirty_time = 0; |
f758eeabe writeback: split ... |
53 |
spin_lock(&wb->list_lock); |
c7f540849 inode: rename i_w... |
54 |
list_for_each_entry(inode, &wb->b_dirty, i_io_list) |
c1955ce32 writeback: remove... |
55 |
nr_dirty++; |
c7f540849 inode: rename i_w... |
56 |
list_for_each_entry(inode, &wb->b_io, i_io_list) |
c1955ce32 writeback: remove... |
57 |
nr_io++; |
c7f540849 inode: rename i_w... |
58 |
list_for_each_entry(inode, &wb->b_more_io, i_io_list) |
c1955ce32 writeback: remove... |
59 |
nr_more_io++; |
c7f540849 inode: rename i_w... |
60 |
list_for_each_entry(inode, &wb->b_dirty_time, i_io_list) |
0ae45f63d vfs: add support ... |
61 62 |
if (inode->i_state & I_DIRTY_TIME) nr_dirty_time++; |
f758eeabe writeback: split ... |
63 |
spin_unlock(&wb->list_lock); |
76f1418b4 mm: bdi: move sta... |
64 |
|
16c4042f0 writeback: avoid ... |
65 |
global_dirty_limits(&background_thresh, &dirty_thresh); |
0d960a383 writeback: clean ... |
66 |
wb_thresh = wb_calc_thresh(wb, dirty_thresh); |
76f1418b4 mm: bdi: move sta... |
67 68 69 |
#define K(x) ((x) << (PAGE_SHIFT - 10)) seq_printf(m, |
00821b002 writeback: show b... |
70 71 72 73 74 75 76 77 78 79 |
"BdiWriteback: %10lu kB " "BdiReclaimable: %10lu kB " "BdiDirtyThresh: %10lu kB " "DirtyThresh: %10lu kB " "BackgroundThresh: %10lu kB " |
c8e28ce04 writeback: accoun... |
80 81 |
"BdiDirtied: %10lu kB " |
00821b002 writeback: show b... |
82 83 84 85 86 87 88 89 90 91 |
"BdiWritten: %10lu kB " "BdiWriteBandwidth: %10lu kBps " "b_dirty: %10lu " "b_io: %10lu " "b_more_io: %10lu " |
0ae45f63d vfs: add support ... |
92 93 |
"b_dirty_time: %10lu " |
00821b002 writeback: show b... |
94 95 96 97 |
"bdi_list: %10u " "state: %10lx ", |
93f78d882 writeback: move b... |
98 99 |
(unsigned long) K(wb_stat(wb, WB_WRITEBACK)), (unsigned long) K(wb_stat(wb, WB_RECLAIMABLE)), |
0d960a383 writeback: clean ... |
100 |
K(wb_thresh), |
f7d2b1ecd writeback: accoun... |
101 102 |
K(dirty_thresh), K(background_thresh), |
93f78d882 writeback: move b... |
103 104 |
(unsigned long) K(wb_stat(wb, WB_DIRTIED)), (unsigned long) K(wb_stat(wb, WB_WRITTEN)), |
a88a341a7 writeback: move b... |
105 |
(unsigned long) K(wb->write_bandwidth), |
f7d2b1ecd writeback: accoun... |
106 107 108 |
nr_dirty, nr_io, nr_more_io, |
0ae45f63d vfs: add support ... |
109 |
nr_dirty_time, |
4452226ea writeback: move b... |
110 |
!list_empty(&bdi->bdi_list), bdi->wb.state); |
76f1418b4 mm: bdi: move sta... |
111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 |
#undef K return 0; } static int bdi_debug_stats_open(struct inode *inode, struct file *file) { return single_open(file, bdi_debug_stats_show, inode->i_private); } static const struct file_operations bdi_debug_stats_fops = { .open = bdi_debug_stats_open, .read = seq_read, .llseek = seq_lseek, .release = single_release, }; static void bdi_debug_register(struct backing_dev_info *bdi, const char *name) { bdi->debug_dir = debugfs_create_dir(name, bdi_debug_root); bdi->debug_stats = debugfs_create_file("stats", 0444, bdi->debug_dir, bdi, &bdi_debug_stats_fops); } static void bdi_debug_unregister(struct backing_dev_info *bdi) { debugfs_remove(bdi->debug_stats); debugfs_remove(bdi->debug_dir); } #else static inline void bdi_debug_init(void) { } static inline void bdi_debug_register(struct backing_dev_info *bdi, const char *name) { } static inline void bdi_debug_unregister(struct backing_dev_info *bdi) { } #endif |
cf0ca9fe5 mm: bdi: export B... |
152 153 154 155 156 |
static ssize_t read_ahead_kb_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { struct backing_dev_info *bdi = dev_get_drvdata(dev); |
cf0ca9fe5 mm: bdi: export B... |
157 |
unsigned long read_ahead_kb; |
7034ed132 backing-dev: use ... |
158 |
ssize_t ret; |
cf0ca9fe5 mm: bdi: export B... |
159 |
|
7034ed132 backing-dev: use ... |
160 161 162 163 164 165 166 |
ret = kstrtoul(buf, 10, &read_ahead_kb); if (ret < 0) return ret; bdi->ra_pages = read_ahead_kb >> (PAGE_SHIFT - 10); return count; |
cf0ca9fe5 mm: bdi: export B... |
167 168 169 170 171 172 173 174 175 176 177 178 |
} #define K(pages) ((pages) << (PAGE_SHIFT - 10)) #define BDI_SHOW(name, expr) \ static ssize_t name##_show(struct device *dev, \ struct device_attribute *attr, char *page) \ { \ struct backing_dev_info *bdi = dev_get_drvdata(dev); \ \ return snprintf(page, PAGE_SIZE-1, "%lld ", (long long)expr); \ |
d9e1241e4 backing-dev: conv... |
179 180 |
} \ static DEVICE_ATTR_RW(name); |
cf0ca9fe5 mm: bdi: export B... |
181 182 |
BDI_SHOW(read_ahead_kb, K(bdi->ra_pages)) |
189d3c4a9 mm: bdi: allow se... |
183 184 185 186 |
static ssize_t min_ratio_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { struct backing_dev_info *bdi = dev_get_drvdata(dev); |
189d3c4a9 mm: bdi: allow se... |
187 |
unsigned int ratio; |
7034ed132 backing-dev: use ... |
188 189 190 191 192 193 194 195 196 |
ssize_t ret; ret = kstrtouint(buf, 10, &ratio); if (ret < 0) return ret; ret = bdi_set_min_ratio(bdi, ratio); if (!ret) ret = count; |
189d3c4a9 mm: bdi: allow se... |
197 |
|
189d3c4a9 mm: bdi: allow se... |
198 199 200 |
return ret; } BDI_SHOW(min_ratio, bdi->min_ratio) |
a42dde041 mm: bdi: allow se... |
201 202 203 204 |
static ssize_t max_ratio_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { struct backing_dev_info *bdi = dev_get_drvdata(dev); |
a42dde041 mm: bdi: allow se... |
205 |
unsigned int ratio; |
7034ed132 backing-dev: use ... |
206 207 208 209 210 211 212 213 214 |
ssize_t ret; ret = kstrtouint(buf, 10, &ratio); if (ret < 0) return ret; ret = bdi_set_max_ratio(bdi, ratio); if (!ret) ret = count; |
a42dde041 mm: bdi: allow se... |
215 |
|
a42dde041 mm: bdi: allow se... |
216 217 218 |
return ret; } BDI_SHOW(max_ratio, bdi->max_ratio) |
7d311cdab bdi: allow block ... |
219 220 221 222 223 224 225 226 227 228 |
static ssize_t stable_pages_required_show(struct device *dev, struct device_attribute *attr, char *page) { struct backing_dev_info *bdi = dev_get_drvdata(dev); return snprintf(page, PAGE_SIZE-1, "%d ", bdi_cap_stable_pages_required(bdi) ? 1 : 0); } |
d9e1241e4 backing-dev: conv... |
229 230 231 232 233 234 235 236 |
static DEVICE_ATTR_RO(stable_pages_required); static struct attribute *bdi_dev_attrs[] = { &dev_attr_read_ahead_kb.attr, &dev_attr_min_ratio.attr, &dev_attr_max_ratio.attr, &dev_attr_stable_pages_required.attr, NULL, |
cf0ca9fe5 mm: bdi: export B... |
237 |
}; |
d9e1241e4 backing-dev: conv... |
238 |
ATTRIBUTE_GROUPS(bdi_dev); |
cf0ca9fe5 mm: bdi: export B... |
239 240 241 242 |
static __init int bdi_class_init(void) { bdi_class = class_create(THIS_MODULE, "bdi"); |
144214537 backing-dev: Hand... |
243 244 |
if (IS_ERR(bdi_class)) return PTR_ERR(bdi_class); |
d9e1241e4 backing-dev: conv... |
245 |
bdi_class->dev_groups = bdi_dev_groups; |
76f1418b4 mm: bdi: move sta... |
246 |
bdi_debug_init(); |
d03f6cdc1 block: Dynamicall... |
247 |
|
cf0ca9fe5 mm: bdi: export B... |
248 249 |
return 0; } |
76f1418b4 mm: bdi: move sta... |
250 |
postcore_initcall(bdi_class_init); |
cf0ca9fe5 mm: bdi: export B... |
251 |
|
2e82b84c0 block: Remove unu... |
252 |
static int bdi_init(struct backing_dev_info *bdi); |
26160158d Move the default_... |
253 254 255 |
static int __init default_bdi_init(void) { int err; |
839a8e866 writeback: replac... |
256 |
bdi_wq = alloc_workqueue("writeback", WQ_MEM_RECLAIM | WQ_FREEZABLE | |
b5c872ddb writeback: expose... |
257 |
WQ_UNBOUND | WQ_SYSFS, 0); |
839a8e866 writeback: replac... |
258 259 |
if (!bdi_wq) return -ENOMEM; |
976e48f8a bdi: Initialize n... |
260 |
err = bdi_init(&noop_backing_dev_info); |
26160158d Move the default_... |
261 262 263 264 |
return err; } subsys_initcall(default_bdi_init); |
6467716a3 writeback: optimi... |
265 |
/* |
f0054bb1e writeback: move b... |
266 |
* This function is used when the first inode for this wb is marked dirty. It |
6467716a3 writeback: optimi... |
267 268 269 270 271 272 273 274 |
* wakes-up the corresponding bdi thread which should then take care of the * periodic background write-out of dirty inodes. Since the write-out would * starts only 'dirty_writeback_interval' centisecs from now anyway, we just * set up a timer which wakes the bdi thread up later. * * Note, we wouldn't bother setting up the timer, but this function is on the * fast-path (used by '__mark_inode_dirty()'), so we save few context switches * by delaying the wake-up. |
6ca738d60 backing_dev: fix ... |
275 276 277 |
* * We have to be careful not to postpone flush work if it is scheduled for * earlier. Thus we use queue_delayed_work(). |
6467716a3 writeback: optimi... |
278 |
*/ |
f0054bb1e writeback: move b... |
279 |
void wb_wakeup_delayed(struct bdi_writeback *wb) |
6467716a3 writeback: optimi... |
280 281 282 283 |
{ unsigned long timeout; timeout = msecs_to_jiffies(dirty_writeback_interval * 10); |
f0054bb1e writeback: move b... |
284 285 286 287 |
spin_lock_bh(&wb->work_lock); if (test_bit(WB_registered, &wb->state)) queue_delayed_work(bdi_wq, &wb->dwork, timeout); spin_unlock_bh(&wb->work_lock); |
03ba3782e writeback: switch... |
288 |
} |
cfc4ba536 writeback: use RC... |
289 |
/* |
a88a341a7 writeback: move b... |
290 |
* Initial write bandwidth: 100 MB/s |
cfc4ba536 writeback: use RC... |
291 |
*/ |
a88a341a7 writeback: move b... |
292 |
#define INIT_BW (100 << (20 - PAGE_SHIFT)) |
cfc4ba536 writeback: use RC... |
293 |
|
8395cd9f8 writeback: add @g... |
294 |
static int wb_init(struct bdi_writeback *wb, struct backing_dev_info *bdi, |
a13f35e87 writeback: don't ... |
295 |
int blkcg_id, gfp_t gfp) |
cf0ca9fe5 mm: bdi: export B... |
296 |
{ |
93f78d882 writeback: move b... |
297 |
int i, err; |
cf0ca9fe5 mm: bdi: export B... |
298 |
|
6467716a3 writeback: optimi... |
299 |
memset(wb, 0, sizeof(*wb)); |
f1d0b063d bdi: register sys... |
300 |
|
810df54a6 bdi: Make wb->bdi... |
301 302 |
if (wb != &bdi->wb) bdi_get(bdi); |
6467716a3 writeback: optimi... |
303 304 305 306 307 |
wb->bdi = bdi; wb->last_old_flush = jiffies; INIT_LIST_HEAD(&wb->b_dirty); INIT_LIST_HEAD(&wb->b_io); INIT_LIST_HEAD(&wb->b_more_io); |
0ae45f63d vfs: add support ... |
308 |
INIT_LIST_HEAD(&wb->b_dirty_time); |
f758eeabe writeback: split ... |
309 |
spin_lock_init(&wb->list_lock); |
66f3b8e2e writeback: move d... |
310 |
|
a88a341a7 writeback: move b... |
311 312 313 314 315 |
wb->bw_time_stamp = jiffies; wb->balanced_dirty_ratelimit = INIT_BW; wb->dirty_ratelimit = INIT_BW; wb->write_bandwidth = INIT_BW; wb->avg_write_bandwidth = INIT_BW; |
cf0ca9fe5 mm: bdi: export B... |
316 |
|
f0054bb1e writeback: move b... |
317 318 319 |
spin_lock_init(&wb->work_lock); INIT_LIST_HEAD(&wb->work_list); INIT_DELAYED_WORK(&wb->dwork, wb_workfn); |
b57d74aff writeback: track ... |
320 |
wb->dirty_sleep = jiffies; |
c284de61d writeback: cleanu... |
321 |
|
a13f35e87 writeback: don't ... |
322 |
wb->congested = wb_congested_get_create(bdi, blkcg_id, gfp); |
810df54a6 bdi: Make wb->bdi... |
323 324 325 326 |
if (!wb->congested) { err = -ENOMEM; goto out_put_bdi; } |
a13f35e87 writeback: don't ... |
327 |
|
8395cd9f8 writeback: add @g... |
328 |
err = fprop_local_init_percpu(&wb->completions, gfp); |
a88a341a7 writeback: move b... |
329 |
if (err) |
a13f35e87 writeback: don't ... |
330 |
goto out_put_cong; |
c284de61d writeback: cleanu... |
331 |
|
93f78d882 writeback: move b... |
332 |
for (i = 0; i < NR_WB_STAT_ITEMS; i++) { |
8395cd9f8 writeback: add @g... |
333 |
err = percpu_counter_init(&wb->stat[i], 0, gfp); |
a13f35e87 writeback: don't ... |
334 335 |
if (err) goto out_destroy_stat; |
93f78d882 writeback: move b... |
336 |
} |
cf0ca9fe5 mm: bdi: export B... |
337 |
|
93f78d882 writeback: move b... |
338 |
return 0; |
a13f35e87 writeback: don't ... |
339 340 |
out_destroy_stat: |
078c6c3a5 mm/backing-dev.c:... |
341 |
while (i--) |
a13f35e87 writeback: don't ... |
342 343 344 345 |
percpu_counter_destroy(&wb->stat[i]); fprop_local_destroy_percpu(&wb->completions); out_put_cong: wb_congested_put(wb->congested); |
810df54a6 bdi: Make wb->bdi... |
346 347 348 |
out_put_bdi: if (wb != &bdi->wb) bdi_put(bdi); |
a13f35e87 writeback: don't ... |
349 |
return err; |
cf0ca9fe5 mm: bdi: export B... |
350 |
} |
cf0ca9fe5 mm: bdi: export B... |
351 |
|
e8cb72b32 bdi: Unify bdi->w... |
352 |
static void cgwb_remove_from_bdi_list(struct bdi_writeback *wb); |
03ba3782e writeback: switch... |
353 354 355 |
/* * Remove bdi from the global list and shutdown any threads we have running */ |
461000714 writeback: reorga... |
356 |
static void wb_shutdown(struct bdi_writeback *wb) |
66f3b8e2e writeback: move d... |
357 |
{ |
c4db59d31 fs: don't reassig... |
358 |
/* Make sure nobody queues further work */ |
461000714 writeback: reorga... |
359 360 361 |
spin_lock_bh(&wb->work_lock); if (!test_and_clear_bit(WB_registered, &wb->state)) { spin_unlock_bh(&wb->work_lock); |
03ba3782e writeback: switch... |
362 |
return; |
c4db59d31 fs: don't reassig... |
363 |
} |
461000714 writeback: reorga... |
364 |
spin_unlock_bh(&wb->work_lock); |
03ba3782e writeback: switch... |
365 |
|
e8cb72b32 bdi: Unify bdi->w... |
366 |
cgwb_remove_from_bdi_list(wb); |
03ba3782e writeback: switch... |
367 |
/* |
461000714 writeback: reorga... |
368 369 370 |
* Drain work list and shutdown the delayed_work. !WB_registered * tells wb_workfn() that @wb is dying and its work_list needs to * be drained no matter what. |
03ba3782e writeback: switch... |
371 |
*/ |
461000714 writeback: reorga... |
372 373 374 375 |
mod_delayed_work(bdi_wq, &wb->dwork, 0); flush_delayed_work(&wb->dwork); WARN_ON(!list_empty(&wb->work_list)); } |
f0054bb1e writeback: move b... |
376 |
static void wb_exit(struct bdi_writeback *wb) |
93f78d882 writeback: move b... |
377 378 379 380 381 382 383 |
{ int i; WARN_ON(delayed_work_pending(&wb->dwork)); for (i = 0; i < NR_WB_STAT_ITEMS; i++) percpu_counter_destroy(&wb->stat[i]); |
6467716a3 writeback: optimi... |
384 |
|
a88a341a7 writeback: move b... |
385 |
fprop_local_destroy_percpu(&wb->completions); |
a13f35e87 writeback: don't ... |
386 |
wb_congested_put(wb->congested); |
810df54a6 bdi: Make wb->bdi... |
387 388 |
if (wb != &wb->bdi->wb) bdi_put(wb->bdi); |
a88a341a7 writeback: move b... |
389 |
} |
e98be2d59 writeback: bdi wr... |
390 |
|
52ebea749 writeback: make b... |
391 392 393 394 395 396 397 |
#ifdef CONFIG_CGROUP_WRITEBACK #include <linux/memcontrol.h> /* * cgwb_lock protects bdi->cgwb_tree, bdi->cgwb_congested_tree, * blkcg->cgwb_list, and memcg->cgwb_list. bdi->cgwb_tree is also RCU |
4514451e7 bdi: Do not wait ... |
398 |
* protected. |
52ebea749 writeback: make b... |
399 400 |
*/ static DEFINE_SPINLOCK(cgwb_lock); |
67b46304b bdi: Move cgroup ... |
401 |
static struct workqueue_struct *cgwb_release_wq; |
52ebea749 writeback: make b... |
402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 |
/** * wb_congested_get_create - get or create a wb_congested * @bdi: associated bdi * @blkcg_id: ID of the associated blkcg * @gfp: allocation mask * * Look up the wb_congested for @blkcg_id on @bdi. If missing, create one. * The returned wb_congested has its reference count incremented. Returns * NULL on failure. */ struct bdi_writeback_congested * wb_congested_get_create(struct backing_dev_info *bdi, int blkcg_id, gfp_t gfp) { struct bdi_writeback_congested *new_congested = NULL, *congested; struct rb_node **node, *parent; unsigned long flags; |
52ebea749 writeback: make b... |
419 420 421 422 423 424 425 426 |
retry: spin_lock_irqsave(&cgwb_lock, flags); node = &bdi->cgwb_congested_tree.rb_node; parent = NULL; while (*node != NULL) { parent = *node; |
bc71226b0 mm/backing-dev.c:... |
427 428 |
congested = rb_entry(parent, struct bdi_writeback_congested, rb_node); |
52ebea749 writeback: make b... |
429 430 431 432 433 434 435 436 437 438 439 440 441 442 |
if (congested->blkcg_id < blkcg_id) node = &parent->rb_left; else if (congested->blkcg_id > blkcg_id) node = &parent->rb_right; else goto found; } if (new_congested) { /* !found and storage for new one already allocated, insert */ congested = new_congested; new_congested = NULL; rb_link_node(&congested->rb_node, parent, node); rb_insert_color(&congested->rb_node, &bdi->cgwb_congested_tree); |
52ebea749 writeback: make b... |
443 444 445 446 447 448 449 450 451 452 453 |
goto found; } spin_unlock_irqrestore(&cgwb_lock, flags); /* allocate storage for new one and retry */ new_congested = kzalloc(sizeof(*new_congested), gfp); if (!new_congested) return NULL; atomic_set(&new_congested->refcnt, 0); |
b7d680d7b bdi: Mark congest... |
454 |
new_congested->__bdi = bdi; |
52ebea749 writeback: make b... |
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 |
new_congested->blkcg_id = blkcg_id; goto retry; found: atomic_inc(&congested->refcnt); spin_unlock_irqrestore(&cgwb_lock, flags); kfree(new_congested); return congested; } /** * wb_congested_put - put a wb_congested * @congested: wb_congested to put * * Put @congested and destroy it if the refcnt reaches zero. */ void wb_congested_put(struct bdi_writeback_congested *congested) { |
52ebea749 writeback: make b... |
473 |
unsigned long flags; |
52ebea749 writeback: make b... |
474 475 476 477 478 |
local_irq_save(flags); if (!atomic_dec_and_lock(&congested->refcnt, &cgwb_lock)) { local_irq_restore(flags); return; } |
a20135ffb writeback: don't ... |
479 |
/* bdi might already have been destroyed leaving @congested unlinked */ |
b7d680d7b bdi: Mark congest... |
480 |
if (congested->__bdi) { |
a20135ffb writeback: don't ... |
481 |
rb_erase(&congested->rb_node, |
b7d680d7b bdi: Mark congest... |
482 483 |
&congested->__bdi->cgwb_congested_tree); congested->__bdi = NULL; |
a20135ffb writeback: don't ... |
484 |
} |
52ebea749 writeback: make b... |
485 486 |
spin_unlock_irqrestore(&cgwb_lock, flags); kfree(congested); |
52ebea749 writeback: make b... |
487 488 489 490 491 492 |
} static void cgwb_release_workfn(struct work_struct *work) { struct bdi_writeback *wb = container_of(work, struct bdi_writeback, release_work); |
52ebea749 writeback: make b... |
493 |
|
1bbe05e27 bdi: Fix another ... |
494 |
mutex_lock(&wb->bdi->cgwb_release_mutex); |
52ebea749 writeback: make b... |
495 496 497 498 |
wb_shutdown(wb); css_put(wb->memcg_css); css_put(wb->blkcg_css); |
1bbe05e27 bdi: Fix another ... |
499 |
mutex_unlock(&wb->bdi->cgwb_release_mutex); |
52ebea749 writeback: make b... |
500 |
|
841710aa6 writeback: implem... |
501 |
fprop_local_destroy_percpu(&wb->memcg_completions); |
52ebea749 writeback: make b... |
502 503 504 |
percpu_ref_exit(&wb->refcnt); wb_exit(wb); kfree_rcu(wb, rcu); |
52ebea749 writeback: make b... |
505 506 507 508 509 510 |
} static void cgwb_release(struct percpu_ref *refcnt) { struct bdi_writeback *wb = container_of(refcnt, struct bdi_writeback, refcnt); |
67b46304b bdi: Move cgroup ... |
511 |
queue_work(cgwb_release_wq, &wb->release_work); |
52ebea749 writeback: make b... |
512 513 514 515 516 517 518 519 520 521 522 |
} static void cgwb_kill(struct bdi_writeback *wb) { lockdep_assert_held(&cgwb_lock); WARN_ON(!radix_tree_delete(&wb->bdi->cgwb_tree, wb->memcg_css->id)); list_del(&wb->memcg_node); list_del(&wb->blkcg_node); percpu_ref_kill(&wb->refcnt); } |
e8cb72b32 bdi: Unify bdi->w... |
523 524 525 526 527 528 |
static void cgwb_remove_from_bdi_list(struct bdi_writeback *wb) { spin_lock_irq(&cgwb_lock); list_del_rcu(&wb->bdi_node); spin_unlock_irq(&cgwb_lock); } |
52ebea749 writeback: make b... |
529 530 531 532 533 534 535 536 537 538 539 540 |
static int cgwb_create(struct backing_dev_info *bdi, struct cgroup_subsys_state *memcg_css, gfp_t gfp) { struct mem_cgroup *memcg; struct cgroup_subsys_state *blkcg_css; struct blkcg *blkcg; struct list_head *memcg_cgwb_list, *blkcg_cgwb_list; struct bdi_writeback *wb; unsigned long flags; int ret = 0; memcg = mem_cgroup_from_css(memcg_css); |
c165b3e3c blkcg: rename sub... |
541 |
blkcg_css = cgroup_get_e_css(memcg_css->cgroup, &io_cgrp_subsys); |
52ebea749 writeback: make b... |
542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 |
blkcg = css_to_blkcg(blkcg_css); memcg_cgwb_list = mem_cgroup_cgwb_list(memcg); blkcg_cgwb_list = &blkcg->cgwb_list; /* look up again under lock and discard on blkcg mismatch */ spin_lock_irqsave(&cgwb_lock, flags); wb = radix_tree_lookup(&bdi->cgwb_tree, memcg_css->id); if (wb && wb->blkcg_css != blkcg_css) { cgwb_kill(wb); wb = NULL; } spin_unlock_irqrestore(&cgwb_lock, flags); if (wb) goto out_put; /* need to create a new one */ wb = kmalloc(sizeof(*wb), gfp); |
0b045bd1c mm/backing-dev.c:... |
559 560 561 562 |
if (!wb) { ret = -ENOMEM; goto out_put; } |
52ebea749 writeback: make b... |
563 |
|
a13f35e87 writeback: don't ... |
564 |
ret = wb_init(wb, bdi, blkcg_css->id, gfp); |
52ebea749 writeback: make b... |
565 566 567 568 569 570 |
if (ret) goto err_free; ret = percpu_ref_init(&wb->refcnt, cgwb_release, 0, gfp); if (ret) goto err_wb_exit; |
841710aa6 writeback: implem... |
571 572 573 |
ret = fprop_local_init_percpu(&wb->memcg_completions, gfp); if (ret) goto err_ref_exit; |
52ebea749 writeback: make b... |
574 575 576 577 |
wb->memcg_css = memcg_css; wb->blkcg_css = blkcg_css; INIT_WORK(&wb->release_work, cgwb_release_workfn); set_bit(WB_registered, &wb->state); |
03ba3782e writeback: switch... |
578 579 |
/* |
52ebea749 writeback: make b... |
580 581 582 583 |
* The root wb determines the registered state of the whole bdi and * memcg_cgwb_list and blkcg_cgwb_list's next pointers indicate * whether they're still online. Don't link @wb if any is dead. * See wb_memcg_offline() and wb_blkcg_offline(). |
03ba3782e writeback: switch... |
584 |
*/ |
52ebea749 writeback: make b... |
585 586 587 588 589 590 591 |
ret = -ENODEV; spin_lock_irqsave(&cgwb_lock, flags); if (test_bit(WB_registered, &bdi->wb.state) && blkcg_cgwb_list->next && memcg_cgwb_list->next) { /* we might have raced another instance of this function */ ret = radix_tree_insert(&bdi->cgwb_tree, memcg_css->id, wb); if (!ret) { |
b817525a4 writeback: bdi_wr... |
592 |
list_add_tail_rcu(&wb->bdi_node, &bdi->wb_list); |
52ebea749 writeback: make b... |
593 594 595 596 597 598 599 600 601 602 |
list_add(&wb->memcg_node, memcg_cgwb_list); list_add(&wb->blkcg_node, blkcg_cgwb_list); css_get(memcg_css); css_get(blkcg_css); } } spin_unlock_irqrestore(&cgwb_lock, flags); if (ret) { if (ret == -EEXIST) ret = 0; |
a13f35e87 writeback: don't ... |
603 |
goto err_fprop_exit; |
52ebea749 writeback: make b... |
604 605 |
} goto out_put; |
841710aa6 writeback: implem... |
606 607 |
err_fprop_exit: fprop_local_destroy_percpu(&wb->memcg_completions); |
52ebea749 writeback: make b... |
608 609 610 611 612 613 614 615 616 |
err_ref_exit: percpu_ref_exit(&wb->refcnt); err_wb_exit: wb_exit(wb); err_free: kfree(wb); out_put: css_put(blkcg_css); return ret; |
66f3b8e2e writeback: move d... |
617 |
} |
52ebea749 writeback: make b... |
618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 |
/** * wb_get_create - get wb for a given memcg, create if necessary * @bdi: target bdi * @memcg_css: cgroup_subsys_state of the target memcg (must have positive ref) * @gfp: allocation mask to use * * Try to get the wb for @memcg_css on @bdi. If it doesn't exist, try to * create one. The returned wb has its refcount incremented. * * This function uses css_get() on @memcg_css and thus expects its refcnt * to be positive on invocation. IOW, rcu_read_lock() protection on * @memcg_css isn't enough. try_get it before calling this function. * * A wb is keyed by its associated memcg. As blkcg implicitly enables * memcg on the default hierarchy, memcg association is guaranteed to be * more specific (equal or descendant to the associated blkcg) and thus can * identify both the memcg and blkcg associations. * * Because the blkcg associated with a memcg may change as blkcg is enabled * and disabled closer to root in the hierarchy, each wb keeps track of * both the memcg and blkcg associated with it and verifies the blkcg on * each lookup. On mismatch, the existing wb is discarded and a new one is * created. */ struct bdi_writeback *wb_get_create(struct backing_dev_info *bdi, struct cgroup_subsys_state *memcg_css, gfp_t gfp) |
6467716a3 writeback: optimi... |
645 |
{ |
52ebea749 writeback: make b... |
646 |
struct bdi_writeback *wb; |
d0164adc8 mm, page_alloc: d... |
647 |
might_sleep_if(gfpflags_allow_blocking(gfp)); |
52ebea749 writeback: make b... |
648 649 650 651 652 653 654 655 656 657 658 659 |
if (!memcg_css->parent) return &bdi->wb; do { rcu_read_lock(); wb = radix_tree_lookup(&bdi->cgwb_tree, memcg_css->id); if (wb) { struct cgroup_subsys_state *blkcg_css; /* see whether the blkcg association has changed */ blkcg_css = cgroup_get_e_css(memcg_css->cgroup, |
c165b3e3c blkcg: rename sub... |
660 |
&io_cgrp_subsys); |
52ebea749 writeback: make b... |
661 662 663 664 665 666 667 668 669 670 |
if (unlikely(wb->blkcg_css != blkcg_css || !wb_tryget(wb))) wb = NULL; css_put(blkcg_css); } rcu_read_unlock(); } while (!wb && !cgwb_create(bdi, memcg_css, gfp)); return wb; } |
6467716a3 writeback: optimi... |
671 |
|
a13f35e87 writeback: don't ... |
672 |
static int cgwb_bdi_init(struct backing_dev_info *bdi) |
52ebea749 writeback: make b... |
673 |
{ |
a13f35e87 writeback: don't ... |
674 |
int ret; |
52ebea749 writeback: make b... |
675 676 |
INIT_RADIX_TREE(&bdi->cgwb_tree, GFP_ATOMIC); bdi->cgwb_congested_tree = RB_ROOT; |
1bbe05e27 bdi: Fix another ... |
677 |
mutex_init(&bdi->cgwb_release_mutex); |
a13f35e87 writeback: don't ... |
678 679 680 |
ret = wb_init(&bdi->wb, bdi, 1, GFP_KERNEL); if (!ret) { |
7d828602e mm: memcontrol: e... |
681 |
bdi->wb.memcg_css = &root_mem_cgroup->css; |
a13f35e87 writeback: don't ... |
682 683 684 |
bdi->wb.blkcg_css = blkcg_root_css; } return ret; |
6467716a3 writeback: optimi... |
685 |
} |
b1c51afc0 bdi: Rename cgwb_... |
686 |
static void cgwb_bdi_unregister(struct backing_dev_info *bdi) |
52ebea749 writeback: make b... |
687 688 689 |
{ struct radix_tree_iter iter; void **slot; |
5318ce7d4 bdi: Shutdown wri... |
690 |
struct bdi_writeback *wb; |
52ebea749 writeback: make b... |
691 692 693 694 695 696 |
WARN_ON(test_bit(WB_registered, &bdi->wb.state)); spin_lock_irq(&cgwb_lock); radix_tree_for_each_slot(slot, &bdi->cgwb_tree, &iter, 0) cgwb_kill(*slot); |
1bbe05e27 bdi: Fix another ... |
697 |
spin_unlock_irq(&cgwb_lock); |
5318ce7d4 bdi: Shutdown wri... |
698 |
|
1bbe05e27 bdi: Fix another ... |
699 700 |
mutex_lock(&bdi->cgwb_release_mutex); spin_lock_irq(&cgwb_lock); |
5318ce7d4 bdi: Shutdown wri... |
701 702 703 704 705 706 707 |
while (!list_empty(&bdi->wb_list)) { wb = list_first_entry(&bdi->wb_list, struct bdi_writeback, bdi_node); spin_unlock_irq(&cgwb_lock); wb_shutdown(wb); spin_lock_irq(&cgwb_lock); } |
52ebea749 writeback: make b... |
708 |
spin_unlock_irq(&cgwb_lock); |
1bbe05e27 bdi: Fix another ... |
709 |
mutex_unlock(&bdi->cgwb_release_mutex); |
52ebea749 writeback: make b... |
710 711 712 713 714 715 716 |
} /** * wb_memcg_offline - kill all wb's associated with a memcg being offlined * @memcg: memcg being offlined * * Also prevents creation of any new wb's associated with @memcg. |
e98be2d59 writeback: bdi wr... |
717 |
*/ |
52ebea749 writeback: make b... |
718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 |
void wb_memcg_offline(struct mem_cgroup *memcg) { LIST_HEAD(to_destroy); struct list_head *memcg_cgwb_list = mem_cgroup_cgwb_list(memcg); struct bdi_writeback *wb, *next; spin_lock_irq(&cgwb_lock); list_for_each_entry_safe(wb, next, memcg_cgwb_list, memcg_node) cgwb_kill(wb); memcg_cgwb_list->next = NULL; /* prevent new wb's */ spin_unlock_irq(&cgwb_lock); } /** * wb_blkcg_offline - kill all wb's associated with a blkcg being offlined * @blkcg: blkcg being offlined * * Also prevents creation of any new wb's associated with @blkcg. */ void wb_blkcg_offline(struct blkcg *blkcg) { LIST_HEAD(to_destroy); struct bdi_writeback *wb, *next; spin_lock_irq(&cgwb_lock); list_for_each_entry_safe(wb, next, &blkcg->cgwb_list, blkcg_node) cgwb_kill(wb); blkcg->cgwb_list.next = NULL; /* prevent new wb's */ spin_unlock_irq(&cgwb_lock); } |
df23de556 bdi: Fix use-afte... |
748 749 750 751 752 753 754 755 756 757 |
static void cgwb_bdi_exit(struct backing_dev_info *bdi) { struct rb_node *rbn; spin_lock_irq(&cgwb_lock); while ((rbn = rb_first(&bdi->cgwb_congested_tree))) { struct bdi_writeback_congested *congested = rb_entry(rbn, struct bdi_writeback_congested, rb_node); rb_erase(rbn, &bdi->cgwb_congested_tree); |
b7d680d7b bdi: Mark congest... |
758 |
congested->__bdi = NULL; /* mark @congested unlinked */ |
df23de556 bdi: Fix use-afte... |
759 760 761 |
} spin_unlock_irq(&cgwb_lock); } |
e8cb72b32 bdi: Unify bdi->w... |
762 763 764 765 766 767 |
static void cgwb_bdi_register(struct backing_dev_info *bdi) { spin_lock_irq(&cgwb_lock); list_add_tail_rcu(&bdi->wb.bdi_node, &bdi->wb_list); spin_unlock_irq(&cgwb_lock); } |
67b46304b bdi: Move cgroup ... |
768 769 770 771 772 773 774 775 776 777 778 779 780 781 |
static int __init cgwb_init(void) { /* * There can be many concurrent release work items overwhelming * system_wq. Put them in a separate wq and limit concurrency. * There's no point in executing many of these in parallel. */ cgwb_release_wq = alloc_workqueue("cgwb_release", 0, 1); if (!cgwb_release_wq) return -ENOMEM; return 0; } subsys_initcall(cgwb_init); |
52ebea749 writeback: make b... |
782 |
#else /* CONFIG_CGROUP_WRITEBACK */ |
a13f35e87 writeback: don't ... |
783 784 785 786 787 788 789 |
static int cgwb_bdi_init(struct backing_dev_info *bdi) { int err; bdi->wb_congested = kzalloc(sizeof(*bdi->wb_congested), GFP_KERNEL); if (!bdi->wb_congested) return -ENOMEM; |
5f478e4ea block: fix double... |
790 |
atomic_set(&bdi->wb_congested->refcnt, 1); |
a13f35e87 writeback: don't ... |
791 792 |
err = wb_init(&bdi->wb, bdi, 1, GFP_KERNEL); if (err) { |
5f478e4ea block: fix double... |
793 |
wb_congested_put(bdi->wb_congested); |
a13f35e87 writeback: don't ... |
794 795 796 797 |
return err; } return 0; } |
b1c51afc0 bdi: Rename cgwb_... |
798 |
static void cgwb_bdi_unregister(struct backing_dev_info *bdi) { } |
df23de556 bdi: Fix use-afte... |
799 800 |
static void cgwb_bdi_exit(struct backing_dev_info *bdi) |
5f478e4ea block: fix double... |
801 802 803 |
{ wb_congested_put(bdi->wb_congested); } |
52ebea749 writeback: make b... |
804 |
|
e8cb72b32 bdi: Unify bdi->w... |
805 806 807 808 809 810 811 812 813 |
static void cgwb_bdi_register(struct backing_dev_info *bdi) { list_add_tail_rcu(&bdi->wb.bdi_node, &bdi->wb_list); } static void cgwb_remove_from_bdi_list(struct bdi_writeback *wb) { list_del_rcu(&wb->bdi_node); } |
52ebea749 writeback: make b... |
814 |
#endif /* CONFIG_CGROUP_WRITEBACK */ |
e98be2d59 writeback: bdi wr... |
815 |
|
2e82b84c0 block: Remove unu... |
816 |
static int bdi_init(struct backing_dev_info *bdi) |
b2e8fb6ef mm: scalable bdi ... |
817 |
{ |
b817525a4 writeback: bdi_wr... |
818 |
int ret; |
cf0ca9fe5 mm: bdi: export B... |
819 |
bdi->dev = NULL; |
d03f6cdc1 block: Dynamicall... |
820 |
kref_init(&bdi->refcnt); |
189d3c4a9 mm: bdi: allow se... |
821 |
bdi->min_ratio = 0; |
a42dde041 mm: bdi: allow se... |
822 |
bdi->max_ratio = 100; |
eb608e3a3 block: Convert BD... |
823 |
bdi->max_prop_frac = FPROP_FRAC_BASE; |
66f3b8e2e writeback: move d... |
824 |
INIT_LIST_HEAD(&bdi->bdi_list); |
b817525a4 writeback: bdi_wr... |
825 |
INIT_LIST_HEAD(&bdi->wb_list); |
cc395d7f1 writeback: implem... |
826 |
init_waitqueue_head(&bdi->wb_waitq); |
03ba3782e writeback: switch... |
827 |
|
b817525a4 writeback: bdi_wr... |
828 |
ret = cgwb_bdi_init(bdi); |
b817525a4 writeback: bdi_wr... |
829 |
return ret; |
b2e8fb6ef mm: scalable bdi ... |
830 |
} |
e98be2d59 writeback: bdi wr... |
831 |
|
d03f6cdc1 block: Dynamicall... |
832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 |
struct backing_dev_info *bdi_alloc_node(gfp_t gfp_mask, int node_id) { struct backing_dev_info *bdi; bdi = kmalloc_node(sizeof(struct backing_dev_info), gfp_mask | __GFP_ZERO, node_id); if (!bdi) return NULL; if (bdi_init(bdi)) { kfree(bdi); return NULL; } return bdi; } |
62bf42adc bdi: Export bdi_a... |
847 |
EXPORT_SYMBOL(bdi_alloc_node); |
d03f6cdc1 block: Dynamicall... |
848 |
|
7c4cc3002 bdi: Drop 'parent... |
849 |
int bdi_register_va(struct backing_dev_info *bdi, const char *fmt, va_list args) |
461000714 writeback: reorga... |
850 |
{ |
461000714 writeback: reorga... |
851 |
struct device *dev; |
e98be2d59 writeback: bdi wr... |
852 |
|
461000714 writeback: reorga... |
853 854 |
if (bdi->dev) /* The driver needs to use separate queues per device */ return 0; |
e98be2d59 writeback: bdi wr... |
855 |
|
7c4cc3002 bdi: Drop 'parent... |
856 |
dev = device_create_vargs(bdi_class, NULL, MKDEV(0, 0), bdi, fmt, args); |
461000714 writeback: reorga... |
857 858 |
if (IS_ERR(dev)) return PTR_ERR(dev); |
04fbfdc14 mm: per device di... |
859 |
|
e8cb72b32 bdi: Unify bdi->w... |
860 |
cgwb_bdi_register(bdi); |
461000714 writeback: reorga... |
861 |
bdi->dev = dev; |
b2e8fb6ef mm: scalable bdi ... |
862 |
|
461000714 writeback: reorga... |
863 864 865 866 867 868 869 870 871 |
bdi_debug_register(bdi, dev_name(dev)); set_bit(WB_registered, &bdi->wb.state); spin_lock_bh(&bdi_lock); list_add_tail_rcu(&bdi->bdi_list, &bdi_list); spin_unlock_bh(&bdi_lock); trace_writeback_bdi_register(bdi); return 0; |
b2e8fb6ef mm: scalable bdi ... |
872 |
} |
baf7a616d bdi: Provide bdi_... |
873 |
EXPORT_SYMBOL(bdi_register_va); |
7c4cc3002 bdi: Drop 'parent... |
874 |
int bdi_register(struct backing_dev_info *bdi, const char *fmt, ...) |
baf7a616d bdi: Provide bdi_... |
875 876 877 878 879 |
{ va_list args; int ret; va_start(args, fmt); |
7c4cc3002 bdi: Drop 'parent... |
880 |
ret = bdi_register_va(bdi, fmt, args); |
baf7a616d bdi: Provide bdi_... |
881 882 883 |
va_end(args); return ret; } |
461000714 writeback: reorga... |
884 |
EXPORT_SYMBOL(bdi_register); |
b2e8fb6ef mm: scalable bdi ... |
885 |
|
df08c32ce block: fix bdi vs... |
886 887 888 |
int bdi_register_owner(struct backing_dev_info *bdi, struct device *owner) { int rc; |
7c4cc3002 bdi: Drop 'parent... |
889 |
rc = bdi_register(bdi, "%u:%u", MAJOR(owner->devt), MINOR(owner->devt)); |
df08c32ce block: fix bdi vs... |
890 891 |
if (rc) return rc; |
b6f8fec44 block: Allow bdi ... |
892 893 |
/* Leaking owner reference... */ WARN_ON(bdi->owner); |
df08c32ce block: fix bdi vs... |
894 895 896 897 898 |
bdi->owner = owner; get_device(owner); return 0; } EXPORT_SYMBOL(bdi_register_owner); |
461000714 writeback: reorga... |
899 900 901 902 903 904 905 906 |
/* * Remove bdi from bdi_list, and ensure that it is no longer visible */ static void bdi_remove_from_list(struct backing_dev_info *bdi) { spin_lock_bh(&bdi_lock); list_del_rcu(&bdi->bdi_list); spin_unlock_bh(&bdi_lock); |
b2e8fb6ef mm: scalable bdi ... |
907 |
|
461000714 writeback: reorga... |
908 909 |
synchronize_rcu_expedited(); } |
cf0ca9fe5 mm: bdi: export B... |
910 |
|
b02176f30 block: don't rele... |
911 |
void bdi_unregister(struct backing_dev_info *bdi) |
b2e8fb6ef mm: scalable bdi ... |
912 |
{ |
f0054bb1e writeback: move b... |
913 914 915 |
/* make sure nobody finds us on the bdi_list anymore */ bdi_remove_from_list(bdi); wb_shutdown(&bdi->wb); |
b1c51afc0 bdi: Rename cgwb_... |
916 |
cgwb_bdi_unregister(bdi); |
7a401a972 backing-dev: ensu... |
917 |
|
c4db59d31 fs: don't reassig... |
918 919 920 921 922 |
if (bdi->dev) { bdi_debug_unregister(bdi); device_unregister(bdi->dev); bdi->dev = NULL; } |
df08c32ce block: fix bdi vs... |
923 924 925 926 927 |
if (bdi->owner) { put_device(bdi->owner); bdi->owner = NULL; } |
b02176f30 block: don't rele... |
928 |
} |
c4db59d31 fs: don't reassig... |
929 |
|
d03f6cdc1 block: Dynamicall... |
930 931 932 933 |
static void release_bdi(struct kref *ref) { struct backing_dev_info *bdi = container_of(ref, struct backing_dev_info, refcnt); |
5af110b2f block: Unregister... |
934 935 |
if (test_bit(WB_registered, &bdi->wb.state)) bdi_unregister(bdi); |
2e82b84c0 block: Remove unu... |
936 937 938 |
WARN_ON_ONCE(bdi->dev); wb_exit(&bdi->wb); cgwb_bdi_exit(bdi); |
d03f6cdc1 block: Dynamicall... |
939 940 941 942 943 944 945 |
kfree(bdi); } void bdi_put(struct backing_dev_info *bdi) { kref_put(&bdi->refcnt, release_bdi); } |
62bf42adc bdi: Export bdi_a... |
946 |
EXPORT_SYMBOL(bdi_put); |
d03f6cdc1 block: Dynamicall... |
947 |
|
3fcfab16c [PATCH] separate ... |
948 949 950 951 |
static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[0]), __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; |
ec8a6f264 writeback: make c... |
952 |
static atomic_t nr_wb_congested[2]; |
3fcfab16c [PATCH] separate ... |
953 |
|
ec8a6f264 writeback: make c... |
954 |
void clear_wb_congested(struct bdi_writeback_congested *congested, int sync) |
3fcfab16c [PATCH] separate ... |
955 |
{ |
1faa16d22 block: change the... |
956 |
wait_queue_head_t *wqh = &congestion_wqh[sync]; |
c877ef8ae writeback: fix th... |
957 |
enum wb_congested_state bit; |
3fcfab16c [PATCH] separate ... |
958 |
|
4452226ea writeback: move b... |
959 |
bit = sync ? WB_sync_congested : WB_async_congested; |
ec8a6f264 writeback: make c... |
960 961 |
if (test_and_clear_bit(bit, &congested->state)) atomic_dec(&nr_wb_congested[sync]); |
4e857c58e arch: Mass conver... |
962 |
smp_mb__after_atomic(); |
3fcfab16c [PATCH] separate ... |
963 964 965 |
if (waitqueue_active(wqh)) wake_up(wqh); } |
ec8a6f264 writeback: make c... |
966 |
EXPORT_SYMBOL(clear_wb_congested); |
3fcfab16c [PATCH] separate ... |
967 |
|
ec8a6f264 writeback: make c... |
968 |
void set_wb_congested(struct bdi_writeback_congested *congested, int sync) |
3fcfab16c [PATCH] separate ... |
969 |
{ |
c877ef8ae writeback: fix th... |
970 |
enum wb_congested_state bit; |
3fcfab16c [PATCH] separate ... |
971 |
|
4452226ea writeback: move b... |
972 |
bit = sync ? WB_sync_congested : WB_async_congested; |
ec8a6f264 writeback: make c... |
973 974 |
if (!test_and_set_bit(bit, &congested->state)) atomic_inc(&nr_wb_congested[sync]); |
3fcfab16c [PATCH] separate ... |
975 |
} |
ec8a6f264 writeback: make c... |
976 |
EXPORT_SYMBOL(set_wb_congested); |
3fcfab16c [PATCH] separate ... |
977 978 979 |
/** * congestion_wait - wait for a backing_dev to become uncongested |
8aa7e847d Fix congestion_wa... |
980 |
* @sync: SYNC or ASYNC IO |
3fcfab16c [PATCH] separate ... |
981 982 983 984 985 986 |
* @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ |
8aa7e847d Fix congestion_wa... |
987 |
long congestion_wait(int sync, long timeout) |
3fcfab16c [PATCH] separate ... |
988 989 |
{ long ret; |
52bb91986 writeback: accoun... |
990 |
unsigned long start = jiffies; |
3fcfab16c [PATCH] separate ... |
991 |
DEFINE_WAIT(wait); |
8aa7e847d Fix congestion_wa... |
992 |
wait_queue_head_t *wqh = &congestion_wqh[sync]; |
3fcfab16c [PATCH] separate ... |
993 994 995 996 |
prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE); ret = io_schedule_timeout(timeout); finish_wait(wqh, &wait); |
52bb91986 writeback: accoun... |
997 998 999 |
trace_writeback_congestion_wait(jiffies_to_usecs(timeout), jiffies_to_usecs(jiffies - start)); |
3fcfab16c [PATCH] separate ... |
1000 1001 1002 |
return ret; } EXPORT_SYMBOL(congestion_wait); |
04fbfdc14 mm: per device di... |
1003 |
|
0e093d997 writeback: do not... |
1004 |
/** |
599d0c954 mm, vmscan: move ... |
1005 1006 |
* wait_iff_congested - Conditionally wait for a backing_dev to become uncongested or a pgdat to complete writes * @pgdat: A pgdat to check if it is heavily congested |
0e093d997 writeback: do not... |
1007 1008 1009 1010 |
* @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * In the event of a congested backing_dev (any backing_dev) and the given |
599d0c954 mm, vmscan: move ... |
1011 |
* @pgdat has experienced recent congestion, this waits for up to @timeout |
0e093d997 writeback: do not... |
1012 1013 1014 |
* jiffies for either a BDI to exit congestion of the given @sync queue * or a write to complete. * |
599d0c954 mm, vmscan: move ... |
1015 |
* In the absence of pgdat congestion, cond_resched() is called to yield |
ede377137 mm: throttle on I... |
1016 |
* the processor if necessary but otherwise does not sleep. |
0e093d997 writeback: do not... |
1017 1018 1019 1020 1021 |
* * The return value is 0 if the sleep is for the full timeout. Otherwise, * it is the number of jiffies that were still remaining when the function * returned. return_value == timeout implies the function did not sleep. */ |
599d0c954 mm, vmscan: move ... |
1022 |
long wait_iff_congested(struct pglist_data *pgdat, int sync, long timeout) |
0e093d997 writeback: do not... |
1023 1024 1025 1026 1027 1028 1029 1030 |
{ long ret; unsigned long start = jiffies; DEFINE_WAIT(wait); wait_queue_head_t *wqh = &congestion_wqh[sync]; /* * If there is no congestion, or heavy congestion is not being |
599d0c954 mm, vmscan: move ... |
1031 |
* encountered in the current pgdat, yield if necessary instead |
0e093d997 writeback: do not... |
1032 1033 |
* of sleeping on the congestion queue */ |
ec8a6f264 writeback: make c... |
1034 |
if (atomic_read(&nr_wb_congested[sync]) == 0 || |
599d0c954 mm, vmscan: move ... |
1035 |
!test_bit(PGDAT_CONGESTED, &pgdat->flags)) { |
ede377137 mm: throttle on I... |
1036 |
cond_resched(); |
599d0c954 mm, vmscan: move ... |
1037 |
|
0e093d997 writeback: do not... |
1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 |
/* In case we scheduled, work out time remaining */ ret = timeout - (jiffies - start); if (ret < 0) ret = 0; goto out; } /* Sleep until uncongested or a write happens */ prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE); ret = io_schedule_timeout(timeout); finish_wait(wqh, &wait); out: trace_writeback_wait_iff_congested(jiffies_to_usecs(timeout), jiffies_to_usecs(jiffies - start)); return ret; } EXPORT_SYMBOL(wait_iff_congested); |
3965c9ae4 mm: prepare for r... |
1058 1059 1060 1061 1062 1063 |
int pdflush_proc_obsolete(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { char kbuf[] = "0 "; |
4c3bffc27 mm/backing-dev.c:... |
1064 |
if (*ppos || *lenp < sizeof(kbuf)) { |
3965c9ae4 mm: prepare for r... |
1065 1066 1067 1068 1069 1070 |
*lenp = 0; return 0; } if (copy_to_user(buffer, kbuf, sizeof(kbuf))) return -EFAULT; |
1170532bb mm: convert print... |
1071 1072 1073 |
pr_warn_once("%s exported in /proc is scheduled for removal ", table->procname); |
3965c9ae4 mm: prepare for r... |
1074 1075 1076 1077 1078 |
*lenp = 2; *ppos += *lenp; return 2; } |