Commit cf914a7d authored by Rusty Russell's avatar Rusty Russell Committed by Linus Torvalds

readahead: split ondemand readahead interface into two functions

Split ondemand readahead interface into two functions.  I think this makes it
a little clearer for non-readahead experts (like Rusty).

Internally they both call ondemand_readahead(), but the page argument is
changed to an obvious boolean flag.
Signed-off-by: default avatarRusty Russell <rusty@rustcorp.com.au>
Signed-off-by: default avatarFengguang Wu <wfg@mail.ustc.edu.cn>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent fe3cba17
...@@ -139,10 +139,10 @@ static int ext3_readdir(struct file * filp, ...@@ -139,10 +139,10 @@ static int ext3_readdir(struct file * filp,
pgoff_t index = map_bh.b_blocknr >> pgoff_t index = map_bh.b_blocknr >>
(PAGE_CACHE_SHIFT - inode->i_blkbits); (PAGE_CACHE_SHIFT - inode->i_blkbits);
if (!ra_has_index(&filp->f_ra, index)) if (!ra_has_index(&filp->f_ra, index))
page_cache_readahead_ondemand( page_cache_sync_readahead(
sb->s_bdev->bd_inode->i_mapping, sb->s_bdev->bd_inode->i_mapping,
&filp->f_ra, filp, &filp->f_ra, filp,
NULL, index, 1); index, 1);
filp->f_ra.prev_index = index; filp->f_ra.prev_index = index;
bh = ext3_bread(NULL, inode, blk, 0, &err); bh = ext3_bread(NULL, inode, blk, 0, &err);
} }
......
...@@ -138,10 +138,10 @@ static int ext4_readdir(struct file * filp, ...@@ -138,10 +138,10 @@ static int ext4_readdir(struct file * filp,
pgoff_t index = map_bh.b_blocknr >> pgoff_t index = map_bh.b_blocknr >>
(PAGE_CACHE_SHIFT - inode->i_blkbits); (PAGE_CACHE_SHIFT - inode->i_blkbits);
if (!ra_has_index(&filp->f_ra, index)) if (!ra_has_index(&filp->f_ra, index))
page_cache_readahead_ondemand( page_cache_sync_readahead(
sb->s_bdev->bd_inode->i_mapping, sb->s_bdev->bd_inode->i_mapping,
&filp->f_ra, filp, &filp->f_ra, filp,
NULL, index, 1); index, 1);
filp->f_ra.prev_index = index; filp->f_ra.prev_index = index;
bh = ext4_bread(NULL, inode, blk, 0, &err); bh = ext4_bread(NULL, inode, blk, 0, &err);
} }
......
...@@ -295,8 +295,8 @@ __generic_file_splice_read(struct file *in, loff_t *ppos, ...@@ -295,8 +295,8 @@ __generic_file_splice_read(struct file *in, loff_t *ppos,
* readahead/allocate the rest and fill in the holes. * readahead/allocate the rest and fill in the holes.
*/ */
if (spd.nr_pages < nr_pages) if (spd.nr_pages < nr_pages)
page_cache_readahead_ondemand(mapping, &in->f_ra, in, page_cache_sync_readahead(mapping, &in->f_ra, in,
NULL, index, req_pages - spd.nr_pages); index, req_pages - spd.nr_pages);
error = 0; error = 0;
while (spd.nr_pages < nr_pages) { while (spd.nr_pages < nr_pages) {
...@@ -352,7 +352,7 @@ __generic_file_splice_read(struct file *in, loff_t *ppos, ...@@ -352,7 +352,7 @@ __generic_file_splice_read(struct file *in, loff_t *ppos,
page = pages[page_nr]; page = pages[page_nr];
if (PageReadahead(page)) if (PageReadahead(page))
page_cache_readahead_ondemand(mapping, &in->f_ra, in, page_cache_async_readahead(mapping, &in->f_ra, in,
page, index, req_pages - page_nr); page, index, req_pages - page_nr);
/* /*
......
...@@ -1138,12 +1138,20 @@ int do_page_cache_readahead(struct address_space *mapping, struct file *filp, ...@@ -1138,12 +1138,20 @@ int do_page_cache_readahead(struct address_space *mapping, struct file *filp,
pgoff_t offset, unsigned long nr_to_read); pgoff_t offset, unsigned long nr_to_read);
int force_page_cache_readahead(struct address_space *mapping, struct file *filp, int force_page_cache_readahead(struct address_space *mapping, struct file *filp,
pgoff_t offset, unsigned long nr_to_read); pgoff_t offset, unsigned long nr_to_read);
unsigned long page_cache_readahead_ondemand(struct address_space *mapping,
struct file_ra_state *ra, void page_cache_sync_readahead(struct address_space *mapping,
struct file *filp, struct file_ra_state *ra,
struct page *page, struct file *filp,
pgoff_t offset, pgoff_t offset,
unsigned long size); unsigned long size);
void page_cache_async_readahead(struct address_space *mapping,
struct file_ra_state *ra,
struct file *filp,
struct page *pg,
pgoff_t offset,
unsigned long size);
unsigned long max_sane_readahead(unsigned long nr); unsigned long max_sane_readahead(unsigned long nr);
/* Do stack extension */ /* Do stack extension */
......
...@@ -894,15 +894,15 @@ void do_generic_mapping_read(struct address_space *mapping, ...@@ -894,15 +894,15 @@ void do_generic_mapping_read(struct address_space *mapping,
find_page: find_page:
page = find_get_page(mapping, index); page = find_get_page(mapping, index);
if (!page) { if (!page) {
page_cache_readahead_ondemand(mapping, page_cache_sync_readahead(mapping,
&ra, filp, page, &ra, filp,
index, last_index - index); index, last_index - index);
page = find_get_page(mapping, index); page = find_get_page(mapping, index);
if (unlikely(page == NULL)) if (unlikely(page == NULL))
goto no_cached_page; goto no_cached_page;
} }
if (PageReadahead(page)) { if (PageReadahead(page)) {
page_cache_readahead_ondemand(mapping, page_cache_async_readahead(mapping,
&ra, filp, page, &ra, filp, page,
index, last_index - index); index, last_index - index);
} }
...@@ -1348,14 +1348,14 @@ int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) ...@@ -1348,14 +1348,14 @@ int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
*/ */
if (VM_SequentialReadHint(vma)) { if (VM_SequentialReadHint(vma)) {
if (!page) { if (!page) {
page_cache_readahead_ondemand(mapping, ra, file, page, page_cache_sync_readahead(mapping, ra, file,
vmf->pgoff, 1); vmf->pgoff, 1);
page = find_lock_page(mapping, vmf->pgoff); page = find_lock_page(mapping, vmf->pgoff);
if (!page) if (!page)
goto no_cached_page; goto no_cached_page;
} }
if (PageReadahead(page)) { if (PageReadahead(page)) {
page_cache_readahead_ondemand(mapping, ra, file, page, page_cache_async_readahead(mapping, ra, file, page,
vmf->pgoff, 1); vmf->pgoff, 1);
} }
} }
......
...@@ -359,7 +359,7 @@ static unsigned long get_next_ra_size(struct file_ra_state *ra, ...@@ -359,7 +359,7 @@ static unsigned long get_next_ra_size(struct file_ra_state *ra,
static unsigned long static unsigned long
ondemand_readahead(struct address_space *mapping, ondemand_readahead(struct address_space *mapping,
struct file_ra_state *ra, struct file *filp, struct file_ra_state *ra, struct file *filp,
struct page *page, pgoff_t offset, bool hit_readahead_marker, pgoff_t offset,
unsigned long req_size) unsigned long req_size)
{ {
unsigned long max; /* max readahead pages */ unsigned long max; /* max readahead pages */
...@@ -387,7 +387,7 @@ ondemand_readahead(struct address_space *mapping, ...@@ -387,7 +387,7 @@ ondemand_readahead(struct address_space *mapping,
* Standalone, small read. * Standalone, small read.
* Read as is, and do not pollute the readahead state. * Read as is, and do not pollute the readahead state.
*/ */
if (!page && !sequential) { if (!hit_readahead_marker && !sequential) {
return __do_page_cache_readahead(mapping, filp, return __do_page_cache_readahead(mapping, filp,
offset, req_size, 0); offset, req_size, 0);
} }
...@@ -408,7 +408,7 @@ ondemand_readahead(struct address_space *mapping, ...@@ -408,7 +408,7 @@ ondemand_readahead(struct address_space *mapping,
* E.g. interleaved reads. * E.g. interleaved reads.
* Not knowing its readahead pos/size, bet on the minimal possible one. * Not knowing its readahead pos/size, bet on the minimal possible one.
*/ */
if (page) { if (hit_readahead_marker) {
ra_index++; ra_index++;
ra_size = min(4 * ra_size, max); ra_size = min(4 * ra_size, max);
} }
...@@ -421,50 +421,71 @@ ondemand_readahead(struct address_space *mapping, ...@@ -421,50 +421,71 @@ ondemand_readahead(struct address_space *mapping,
} }
/** /**
* page_cache_readahead_ondemand - generic file readahead * page_cache_sync_readahead - generic file readahead
* @mapping: address_space which holds the pagecache and I/O vectors * @mapping: address_space which holds the pagecache and I/O vectors
* @ra: file_ra_state which holds the readahead state * @ra: file_ra_state which holds the readahead state
* @filp: passed on to ->readpage() and ->readpages() * @filp: passed on to ->readpage() and ->readpages()
* @page: the page at @offset, or NULL if non-present * @offset: start offset into @mapping, in pagecache page-sized units
* @offset: start offset into @mapping, in PAGE_CACHE_SIZE units
* @req_size: hint: total size of the read which the caller is performing in * @req_size: hint: total size of the read which the caller is performing in
* PAGE_CACHE_SIZE units * pagecache pages
* *
* page_cache_readahead_ondemand() is the entry point of readahead logic. * page_cache_sync_readahead() should be called when a cache miss happened:
* This function should be called when it is time to perform readahead: * it will submit the read. The readahead logic may decide to piggyback more
* 1) @page == NULL * pages onto the read request if access patterns suggest it will improve
* A cache miss happened, time for synchronous readahead. * performance.
* 2) @page != NULL && PageReadahead(@page)
* A look-ahead hit occured, time for asynchronous readahead.
*/ */
unsigned long void page_cache_sync_readahead(struct address_space *mapping,
page_cache_readahead_ondemand(struct address_space *mapping, struct file_ra_state *ra, struct file *filp,
struct file_ra_state *ra, struct file *filp, pgoff_t offset, unsigned long req_size)
struct page *page, pgoff_t offset,
unsigned long req_size)
{ {
/* no read-ahead */ /* no read-ahead */
if (!ra->ra_pages) if (!ra->ra_pages)
return 0; return;
if (page) { /* do read-ahead */
/* ondemand_readahead(mapping, ra, filp, false, offset, req_size);
* It can be PG_reclaim. }
*/ EXPORT_SYMBOL_GPL(page_cache_sync_readahead);
if (PageWriteback(page))
return 0; /**
* page_cache_async_readahead - file readahead for marked pages
ClearPageReadahead(page); * @mapping: address_space which holds the pagecache and I/O vectors
* @ra: file_ra_state which holds the readahead state
/* * @filp: passed on to ->readpage() and ->readpages()
* Defer asynchronous read-ahead on IO congestion. * @page: the page at @offset which has the PG_readahead flag set
*/ * @offset: start offset into @mapping, in pagecache page-sized units
if (bdi_read_congested(mapping->backing_dev_info)) * @req_size: hint: total size of the read which the caller is performing in
return 0; * pagecache pages
} *
* page_cache_async_ondemand() should be called when a page is used which
* has the PG_readahead flag: this is a marker to suggest that the application
* has used up enough of the readahead window that we should start pulling in
* more pages. */
void
page_cache_async_readahead(struct address_space *mapping,
struct file_ra_state *ra, struct file *filp,
struct page *page, pgoff_t offset,
unsigned long req_size)
{
/* no read-ahead */
if (!ra->ra_pages)
return;
/*
* Same bit is used for PG_readahead and PG_reclaim.
*/
if (PageWriteback(page))
return;
ClearPageReadahead(page);
/*
* Defer asynchronous read-ahead on IO congestion.
*/
if (bdi_read_congested(mapping->backing_dev_info))
return;
/* do read-ahead */ /* do read-ahead */
return ondemand_readahead(mapping, ra, filp, page, ondemand_readahead(mapping, ra, filp, true, offset, req_size);
offset, req_size);
} }
EXPORT_SYMBOL_GPL(page_cache_readahead_ondemand); EXPORT_SYMBOL_GPL(page_cache_async_readahead);
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment