Root/mm/readahead.c

1/*
2 * mm/readahead.c - address_space-level file readahead.
3 *
4 * Copyright (C) 2002, Linus Torvalds
5 *
6 * 09Apr2002 Andrew Morton
7 * Initial version.
8 */
9
10#include <linux/kernel.h>
11#include <linux/fs.h>
12#include <linux/gfp.h>
13#include <linux/mm.h>
14#include <linux/module.h>
15#include <linux/blkdev.h>
16#include <linux/backing-dev.h>
17#include <linux/task_io_accounting_ops.h>
18#include <linux/pagevec.h>
19#include <linux/pagemap.h>
20
21/*
22 * Initialise a struct file's readahead state. Assumes that the caller has
23 * memset *ra to zero.
24 */
25void
26file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping)
27{
28    ra->ra_pages = mapping->backing_dev_info->ra_pages;
29    ra->prev_pos = -1;
30}
31EXPORT_SYMBOL_GPL(file_ra_state_init);
32
33#define list_to_page(head) (list_entry((head)->prev, struct page, lru))
34
35/*
36 * see if a page needs releasing upon read_cache_pages() failure
37 * - the caller of read_cache_pages() may have set PG_private or PG_fscache
38 * before calling, such as the NFS fs marking pages that are cached locally
39 * on disk, thus we need to give the fs a chance to clean up in the event of
40 * an error
41 */
42static void read_cache_pages_invalidate_page(struct address_space *mapping,
43                         struct page *page)
44{
45    if (page_has_private(page)) {
46        if (!trylock_page(page))
47            BUG();
48        page->mapping = mapping;
49        do_invalidatepage(page, 0);
50        page->mapping = NULL;
51        unlock_page(page);
52    }
53    page_cache_release(page);
54}
55
56/*
57 * release a list of pages, invalidating them first if need be
58 */
59static void read_cache_pages_invalidate_pages(struct address_space *mapping,
60                          struct list_head *pages)
61{
62    struct page *victim;
63
64    while (!list_empty(pages)) {
65        victim = list_to_page(pages);
66        list_del(&victim->lru);
67        read_cache_pages_invalidate_page(mapping, victim);
68    }
69}
70
71/**
72 * read_cache_pages - populate an address space with some pages & start reads against them
73 * @mapping: the address_space
74 * @pages: The address of a list_head which contains the target pages. These
75 * pages have their ->index populated and are otherwise uninitialised.
76 * @filler: callback routine for filling a single page.
77 * @data: private data for the callback routine.
78 *
79 * Hides the details of the LRU cache etc from the filesystems.
80 */
81int read_cache_pages(struct address_space *mapping, struct list_head *pages,
82            int (*filler)(void *, struct page *), void *data)
83{
84    struct page *page;
85    int ret = 0;
86
87    while (!list_empty(pages)) {
88        page = list_to_page(pages);
89        list_del(&page->lru);
90        if (add_to_page_cache_lru(page, mapping,
91                    page->index, GFP_KERNEL)) {
92            read_cache_pages_invalidate_page(mapping, page);
93            continue;
94        }
95        page_cache_release(page);
96
97        ret = filler(data, page);
98        if (unlikely(ret)) {
99            read_cache_pages_invalidate_pages(mapping, pages);
100            break;
101        }
102        task_io_account_read(PAGE_CACHE_SIZE);
103    }
104    return ret;
105}
106
107EXPORT_SYMBOL(read_cache_pages);
108
109static int read_pages(struct address_space *mapping, struct file *filp,
110        struct list_head *pages, unsigned nr_pages)
111{
112    unsigned page_idx;
113    int ret;
114
115    if (mapping->a_ops->readpages) {
116        ret = mapping->a_ops->readpages(filp, mapping, pages, nr_pages);
117        /* Clean up the remaining pages */
118        put_pages_list(pages);
119        goto out;
120    }
121
122    for (page_idx = 0; page_idx < nr_pages; page_idx++) {
123        struct page *page = list_to_page(pages);
124        list_del(&page->lru);
125        if (!add_to_page_cache_lru(page, mapping,
126                    page->index, GFP_KERNEL)) {
127            mapping->a_ops->readpage(filp, page);
128        }
129        page_cache_release(page);
130    }
131    ret = 0;
132out:
133    return ret;
134}
135
136/*
137 * __do_page_cache_readahead() actually reads a chunk of disk. It allocates all
138 * the pages first, then submits them all for I/O. This avoids the very bad
139 * behaviour which would occur if page allocations are causing VM writeback.
140 * We really don't want to intermingle reads and writes like that.
141 *
142 * Returns the number of pages requested, or the maximum amount of I/O allowed.
143 */
144static int
145__do_page_cache_readahead(struct address_space *mapping, struct file *filp,
146            pgoff_t offset, unsigned long nr_to_read,
147            unsigned long lookahead_size)
148{
149    struct inode *inode = mapping->host;
150    struct page *page;
151    unsigned long end_index; /* The last page we want to read */
152    LIST_HEAD(page_pool);
153    int page_idx;
154    int ret = 0;
155    loff_t isize = i_size_read(inode);
156
157    if (isize == 0)
158        goto out;
159
160    end_index = ((isize - 1) >> PAGE_CACHE_SHIFT);
161
162    /*
163     * Preallocate as many pages as we will need.
164     */
165    for (page_idx = 0; page_idx < nr_to_read; page_idx++) {
166        pgoff_t page_offset = offset + page_idx;
167
168        if (page_offset > end_index)
169            break;
170
171        rcu_read_lock();
172        page = radix_tree_lookup(&mapping->page_tree, page_offset);
173        rcu_read_unlock();
174        if (page)
175            continue;
176
177        page = page_cache_alloc_cold(mapping);
178        if (!page)
179            break;
180        page->index = page_offset;
181        list_add(&page->lru, &page_pool);
182        if (page_idx == nr_to_read - lookahead_size)
183            SetPageReadahead(page);
184        ret++;
185    }
186
187    /*
188     * Now start the IO. We ignore I/O errors - if the page is not
189     * uptodate then the caller will launch readpage again, and
190     * will then handle the error.
191     */
192    if (ret)
193        read_pages(mapping, filp, &page_pool, ret);
194    BUG_ON(!list_empty(&page_pool));
195out:
196    return ret;
197}
198
199/*
200 * Chunk the readahead into 2 megabyte units, so that we don't pin too much
201 * memory at once.
202 */
203int force_page_cache_readahead(struct address_space *mapping, struct file *filp,
204        pgoff_t offset, unsigned long nr_to_read)
205{
206    int ret = 0;
207
208    if (unlikely(!mapping->a_ops->readpage && !mapping->a_ops->readpages))
209        return -EINVAL;
210
211    nr_to_read = max_sane_readahead(nr_to_read);
212    while (nr_to_read) {
213        int err;
214
215        unsigned long this_chunk = (2 * 1024 * 1024) / PAGE_CACHE_SIZE;
216
217        if (this_chunk > nr_to_read)
218            this_chunk = nr_to_read;
219        err = __do_page_cache_readahead(mapping, filp,
220                        offset, this_chunk, 0);
221        if (err < 0) {
222            ret = err;
223            break;
224        }
225        ret += err;
226        offset += this_chunk;
227        nr_to_read -= this_chunk;
228    }
229    return ret;
230}
231
232/*
233 * Given a desired number of PAGE_CACHE_SIZE readahead pages, return a
234 * sensible upper limit.
235 */
236unsigned long max_sane_readahead(unsigned long nr)
237{
238    return min(nr, (node_page_state(numa_node_id(), NR_INACTIVE_FILE)
239        + node_page_state(numa_node_id(), NR_FREE_PAGES)) / 2);
240}
241
242/*
243 * Submit IO for the read-ahead request in file_ra_state.
244 */
245unsigned long ra_submit(struct file_ra_state *ra,
246               struct address_space *mapping, struct file *filp)
247{
248    int actual;
249
250    actual = __do_page_cache_readahead(mapping, filp,
251                    ra->start, ra->size, ra->async_size);
252
253    return actual;
254}
255
256/*
257 * Set the initial window size, round to next power of 2 and square
258 * for small size, x 4 for medium, and x 2 for large
259 * for 128k (32 page) max ra
260 * 1-8 page = 32k initial, > 8 page = 128k initial
261 */
262static unsigned long get_init_ra_size(unsigned long size, unsigned long max)
263{
264    unsigned long newsize = roundup_pow_of_two(size);
265
266    if (newsize <= max / 32)
267        newsize = newsize * 4;
268    else if (newsize <= max / 4)
269        newsize = newsize * 2;
270    else
271        newsize = max;
272
273    return newsize;
274}
275
276/*
277 * Get the previous window size, ramp it up, and
278 * return it as the new window size.
279 */
280static unsigned long get_next_ra_size(struct file_ra_state *ra,
281                        unsigned long max)
282{
283    unsigned long cur = ra->size;
284    unsigned long newsize;
285
286    if (cur < max / 16)
287        newsize = 4 * cur;
288    else
289        newsize = 2 * cur;
290
291    return min(newsize, max);
292}
293
294/*
295 * On-demand readahead design.
296 *
297 * The fields in struct file_ra_state represent the most-recently-executed
298 * readahead attempt:
299 *
300 * |<----- async_size ---------|
301 * |------------------- size -------------------->|
302 * |==================#===========================|
303 * ^start ^page marked with PG_readahead
304 *
305 * To overlap application thinking time and disk I/O time, we do
306 * `readahead pipelining': Do not wait until the application consumed all
307 * readahead pages and stalled on the missing page at readahead_index;
308 * Instead, submit an asynchronous readahead I/O as soon as there are
309 * only async_size pages left in the readahead window. Normally async_size
310 * will be equal to size, for maximum pipelining.
311 *
312 * In interleaved sequential reads, concurrent streams on the same fd can
313 * be invalidating each other's readahead state. So we flag the new readahead
314 * page at (start+size-async_size) with PG_readahead, and use it as readahead
315 * indicator. The flag won't be set on already cached pages, to avoid the
316 * readahead-for-nothing fuss, saving pointless page cache lookups.
317 *
318 * prev_pos tracks the last visited byte in the _previous_ read request.
319 * It should be maintained by the caller, and will be used for detecting
320 * small random reads. Note that the readahead algorithm checks loosely
321 * for sequential patterns. Hence interleaved reads might be served as
322 * sequential ones.
323 *
324 * There is a special-case: if the first page which the application tries to
325 * read happens to be the first page of the file, it is assumed that a linear
326 * read is about to happen and the window is immediately set to the initial size
327 * based on I/O request size and the max_readahead.
328 *
329 * The code ramps up the readahead size aggressively at first, but slow down as
330 * it approaches max_readhead.
331 */
332
333/*
334 * Count contiguously cached pages from @offset-1 to @offset-@max,
335 * this count is a conservative estimation of
336 * - length of the sequential read sequence, or
337 * - thrashing threshold in memory tight systems
338 */
339static pgoff_t count_history_pages(struct address_space *mapping,
340                   struct file_ra_state *ra,
341                   pgoff_t offset, unsigned long max)
342{
343    pgoff_t head;
344
345    rcu_read_lock();
346    head = radix_tree_prev_hole(&mapping->page_tree, offset - 1, max);
347    rcu_read_unlock();
348
349    return offset - 1 - head;
350}
351
352/*
353 * page cache context based read-ahead
354 */
355static int try_context_readahead(struct address_space *mapping,
356                 struct file_ra_state *ra,
357                 pgoff_t offset,
358                 unsigned long req_size,
359                 unsigned long max)
360{
361    pgoff_t size;
362
363    size = count_history_pages(mapping, ra, offset, max);
364
365    /*
366     * no history pages:
367     * it could be a random read
368     */
369    if (!size)
370        return 0;
371
372    /*
373     * starts from beginning of file:
374     * it is a strong indication of long-run stream (or whole-file-read)
375     */
376    if (size >= offset)
377        size *= 2;
378
379    ra->start = offset;
380    ra->size = get_init_ra_size(size + req_size, max);
381    ra->async_size = ra->size;
382
383    return 1;
384}
385
386/*
387 * A minimal readahead algorithm for trivial sequential/random reads.
388 */
389static unsigned long
390ondemand_readahead(struct address_space *mapping,
391           struct file_ra_state *ra, struct file *filp,
392           bool hit_readahead_marker, pgoff_t offset,
393           unsigned long req_size)
394{
395    unsigned long max = max_sane_readahead(ra->ra_pages);
396
397    /*
398     * start of file
399     */
400    if (!offset)
401        goto initial_readahead;
402
403    /*
404     * It's the expected callback offset, assume sequential access.
405     * Ramp up sizes, and push forward the readahead window.
406     */
407    if ((offset == (ra->start + ra->size - ra->async_size) ||
408         offset == (ra->start + ra->size))) {
409        ra->start += ra->size;
410        ra->size = get_next_ra_size(ra, max);
411        ra->async_size = ra->size;
412        goto readit;
413    }
414
415    /*
416     * Hit a marked page without valid readahead state.
417     * E.g. interleaved reads.
418     * Query the pagecache for async_size, which normally equals to
419     * readahead size. Ramp it up and use it as the new readahead size.
420     */
421    if (hit_readahead_marker) {
422        pgoff_t start;
423
424        rcu_read_lock();
425        start = radix_tree_next_hole(&mapping->page_tree, offset+1,max);
426        rcu_read_unlock();
427
428        if (!start || start - offset > max)
429            return 0;
430
431        ra->start = start;
432        ra->size = start - offset; /* old async_size */
433        ra->size += req_size;
434        ra->size = get_next_ra_size(ra, max);
435        ra->async_size = ra->size;
436        goto readit;
437    }
438
439    /*
440     * oversize read
441     */
442    if (req_size > max)
443        goto initial_readahead;
444
445    /*
446     * sequential cache miss
447     */
448    if (offset - (ra->prev_pos >> PAGE_CACHE_SHIFT) <= 1UL)
449        goto initial_readahead;
450
451    /*
452     * Query the page cache and look for the traces(cached history pages)
453     * that a sequential stream would leave behind.
454     */
455    if (try_context_readahead(mapping, ra, offset, req_size, max))
456        goto readit;
457
458    /*
459     * standalone, small random read
460     * Read as is, and do not pollute the readahead state.
461     */
462    return __do_page_cache_readahead(mapping, filp, offset, req_size, 0);
463
464initial_readahead:
465    ra->start = offset;
466    ra->size = get_init_ra_size(req_size, max);
467    ra->async_size = ra->size > req_size ? ra->size - req_size : ra->size;
468
469readit:
470    /*
471     * Will this read hit the readahead marker made by itself?
472     * If so, trigger the readahead marker hit now, and merge
473     * the resulted next readahead window into the current one.
474     */
475    if (offset == ra->start && ra->size == ra->async_size) {
476        ra->async_size = get_next_ra_size(ra, max);
477        ra->size += ra->async_size;
478    }
479
480    return ra_submit(ra, mapping, filp);
481}
482
483/**
484 * page_cache_sync_readahead - generic file readahead
485 * @mapping: address_space which holds the pagecache and I/O vectors
486 * @ra: file_ra_state which holds the readahead state
487 * @filp: passed on to ->readpage() and ->readpages()
488 * @offset: start offset into @mapping, in pagecache page-sized units
489 * @req_size: hint: total size of the read which the caller is performing in
490 * pagecache pages
491 *
492 * page_cache_sync_readahead() should be called when a cache miss happened:
493 * it will submit the read. The readahead logic may decide to piggyback more
494 * pages onto the read request if access patterns suggest it will improve
495 * performance.
496 */
497void page_cache_sync_readahead(struct address_space *mapping,
498                   struct file_ra_state *ra, struct file *filp,
499                   pgoff_t offset, unsigned long req_size)
500{
501    /* no read-ahead */
502    if (!ra->ra_pages)
503        return;
504
505    /* be dumb */
506    if (filp && (filp->f_mode & FMODE_RANDOM)) {
507        force_page_cache_readahead(mapping, filp, offset, req_size);
508        return;
509    }
510
511    /* do read-ahead */
512    ondemand_readahead(mapping, ra, filp, false, offset, req_size);
513}
514EXPORT_SYMBOL_GPL(page_cache_sync_readahead);
515
516/**
517 * page_cache_async_readahead - file readahead for marked pages
518 * @mapping: address_space which holds the pagecache and I/O vectors
519 * @ra: file_ra_state which holds the readahead state
520 * @filp: passed on to ->readpage() and ->readpages()
521 * @page: the page at @offset which has the PG_readahead flag set
522 * @offset: start offset into @mapping, in pagecache page-sized units
523 * @req_size: hint: total size of the read which the caller is performing in
524 * pagecache pages
525 *
526 * page_cache_async_readahead() should be called when a page is used which
527 * has the PG_readahead flag; this is a marker to suggest that the application
528 * has used up enough of the readahead window that we should start pulling in
529 * more pages.
530 */
531void
532page_cache_async_readahead(struct address_space *mapping,
533               struct file_ra_state *ra, struct file *filp,
534               struct page *page, pgoff_t offset,
535               unsigned long req_size)
536{
537    /* no read-ahead */
538    if (!ra->ra_pages)
539        return;
540
541    /*
542     * Same bit is used for PG_readahead and PG_reclaim.
543     */
544    if (PageWriteback(page))
545        return;
546
547    ClearPageReadahead(page);
548
549    /*
550     * Defer asynchronous read-ahead on IO congestion.
551     */
552    if (bdi_read_congested(mapping->backing_dev_info))
553        return;
554
555    /* do read-ahead */
556    ondemand_readahead(mapping, ra, filp, true, offset, req_size);
557
558#ifdef CONFIG_BLOCK
559    /*
560     * Normally the current page is !uptodate and lock_page() will be
561     * immediately called to implicitly unplug the device. However this
562     * is not always true for RAID conifgurations, where data arrives
563     * not strictly in their submission order. In this case we need to
564     * explicitly kick off the IO.
565     */
566    if (PageUptodate(page))
567        blk_run_backing_dev(mapping->backing_dev_info, NULL);
568#endif
569}
570EXPORT_SYMBOL_GPL(page_cache_async_readahead);
571

Archive Download this file



interactive