From: clameter@sgi.com
To: linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
Cc: Christoph Hellwig <hch@lst.de>, Mel Gorman <mel@skynet.ie>,
Fengguang Wu <fengguang.wu@gmail.com>
Cc: William Lee Irwin III <wli@holomorphy.com>, David Chinner <dgc@sgi.com>
Cc: Jens Axboe <jens.axboe@oracle.com>, Badari Pulavarty <pbadari@gmail.com>
Cc: Maxim Levitsky <maximlevitsky@gmail.com>
Subject: [32/37] Readahead changes to support large blocksize.
Date: Wed, 20 Jun 2007 11:29:39 -0700 [thread overview]
Message-ID: <20070620183013.496868776@sgi.com> (raw)
In-Reply-To: 20070620182907.506775016@sgi.com
[-- Attachment #1: vps-readahead.patch --]
[-- Type: text/plain, Size: 5537 bytes --]
Fix up readhead for large I/O operations.
Only calculate the readahead until the 2M boundary then fall back to
one page.
Signed-off-by: Fengguang Wu <fengguang.wu@gmail.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
===================================================================
---
include/linux/mm.h | 2 +-
mm/fadvise.c | 4 ++--
mm/filemap.c | 5 ++---
mm/madvise.c | 2 +-
mm/readahead.c | 22 ++++++++++++++--------
5 files changed, 20 insertions(+), 15 deletions(-)
Index: linux-2.6.22-rc4-mm2/mm/fadvise.c
===================================================================
--- linux-2.6.22-rc4-mm2.orig/mm/fadvise.c 2007-06-18 23:09:37.000000000 -0700
+++ linux-2.6.22-rc4-mm2/mm/fadvise.c 2007-06-19 20:01:53.000000000 -0700
@@ -86,10 +86,10 @@ asmlinkage long sys_fadvise64_64(int fd,
nrpages = end_index - start_index + 1;
if (!nrpages)
nrpages = ~0UL;
-
+
ret = force_page_cache_readahead(mapping, file,
start_index,
- max_sane_readahead(nrpages));
+ nrpages);
if (ret > 0)
ret = 0;
break;
Index: linux-2.6.22-rc4-mm2/mm/filemap.c
===================================================================
--- linux-2.6.22-rc4-mm2.orig/mm/filemap.c 2007-06-19 19:28:15.000000000 -0700
+++ linux-2.6.22-rc4-mm2/mm/filemap.c 2007-06-19 20:01:53.000000000 -0700
@@ -1288,8 +1288,7 @@ do_readahead(struct address_space *mappi
if (!mapping || !mapping->a_ops || !mapping->a_ops->readpage)
return -EINVAL;
- force_page_cache_readahead(mapping, filp, index,
- max_sane_readahead(nr));
+ force_page_cache_readahead(mapping, filp, index, nr);
return 0;
}
@@ -1427,7 +1426,7 @@ retry_find:
count_vm_event(PGMAJFAULT);
}
did_readaround = 1;
- ra_pages = max_sane_readahead(file->f_ra.ra_pages);
+ ra_pages = file->f_ra.ra_pages;
if (ra_pages) {
pgoff_t start = 0;
Index: linux-2.6.22-rc4-mm2/mm/madvise.c
===================================================================
--- linux-2.6.22-rc4-mm2.orig/mm/madvise.c 2007-06-04 17:57:25.000000000 -0700
+++ linux-2.6.22-rc4-mm2/mm/madvise.c 2007-06-19 20:01:53.000000000 -0700
@@ -124,7 +124,7 @@ static long madvise_willneed(struct vm_a
end = ((end - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
force_page_cache_readahead(file->f_mapping,
- file, start, max_sane_readahead(end - start));
+ file, start, end - start);
return 0;
}
Index: linux-2.6.22-rc4-mm2/mm/readahead.c
===================================================================
--- linux-2.6.22-rc4-mm2.orig/mm/readahead.c 2007-06-15 17:35:33.000000000 -0700
+++ linux-2.6.22-rc4-mm2/mm/readahead.c 2007-06-19 20:01:53.000000000 -0700
@@ -44,7 +44,8 @@ EXPORT_SYMBOL_GPL(default_backing_dev_in
void
file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping)
{
- ra->ra_pages = mapping->backing_dev_info->ra_pages;
+ ra->ra_pages = DIV_ROUND_UP(mapping->backing_dev_info->ra_pages,
+ page_cache_size(mapping));
ra->prev_index = -1;
}
EXPORT_SYMBOL_GPL(file_ra_state_init);
@@ -82,7 +83,7 @@ int read_cache_pages(struct address_spac
put_pages_list(pages);
break;
}
- task_io_account_read(PAGE_CACHE_SIZE);
+ task_io_account_read(page_cache_size(mapping));
}
return ret;
}
@@ -143,7 +144,7 @@ __do_page_cache_readahead(struct address
if (isize == 0)
goto out;
- end_index = ((isize - 1) >> PAGE_CACHE_SHIFT);
+ end_index = page_cache_index(mapping, isize - 1);
/*
* Preallocate as many pages as we will need.
@@ -196,10 +197,12 @@ int force_page_cache_readahead(struct ad
if (unlikely(!mapping->a_ops->readpage && !mapping->a_ops->readpages))
return -EINVAL;
+ nr_to_read = max_sane_readahead(nr_to_read, mapping_order(mapping));
while (nr_to_read) {
int err;
- unsigned long this_chunk = (2 * 1024 * 1024) / PAGE_CACHE_SIZE;
+ unsigned long this_chunk = DIV_ROUND_UP(2 * 1024 * 1024,
+ page_cache_size(mapping));
if (this_chunk > nr_to_read)
this_chunk = nr_to_read;
@@ -229,17 +232,20 @@ int do_page_cache_readahead(struct addre
if (bdi_read_congested(mapping->backing_dev_info))
return -1;
+ nr_to_read = max_sane_readahead(nr_to_read, mapping_order(mapping));
return __do_page_cache_readahead(mapping, filp, offset, nr_to_read, 0);
}
/*
- * Given a desired number of PAGE_CACHE_SIZE readahead pages, return a
+ * Given a desired number of page order readahead pages, return a
* sensible upper limit.
*/
-unsigned long max_sane_readahead(unsigned long nr)
+unsigned long max_sane_readahead(unsigned long nr, int order)
{
- return min(nr, (node_page_state(numa_node_id(), NR_INACTIVE)
- + node_page_state(numa_node_id(), NR_FREE_PAGES)) / 2);
+ unsigned long base_pages = node_page_state(numa_node_id(), NR_INACTIVE)
+ + node_page_state(numa_node_id(), NR_FREE_PAGES);
+
+ return min(nr, (base_pages / 2) >> order);
}
/*
Index: linux-2.6.22-rc4-mm2/include/linux/mm.h
===================================================================
--- linux-2.6.22-rc4-mm2.orig/include/linux/mm.h 2007-06-18 23:09:37.000000000 -0700
+++ linux-2.6.22-rc4-mm2/include/linux/mm.h 2007-06-19 20:01:53.000000000 -0700
@@ -1167,7 +1167,7 @@ unsigned long page_cache_readahead_ondem
struct page *page,
pgoff_t offset,
unsigned long size);
-unsigned long max_sane_readahead(unsigned long nr);
+unsigned long max_sane_readahead(unsigned long nr, int order);
/* Do stack extension */
extern int expand_stack(struct vm_area_struct *vma, unsigned long address);
--
next prev parent reply other threads:[~2007-06-20 18:30 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-06-20 18:29 [00/37] Large Blocksize Support V4 clameter
2007-06-20 18:29 ` [01/37] Define functions for page cache handling clameter
2007-06-20 18:29 ` [02/37] Pagecache zeroing: zero_user_segment, zero_user_segments and zero_user clameter
2007-06-20 18:29 ` [03/37] Use page_cache_xxx function in mm/filemap.c clameter
2007-06-20 18:29 ` [04/37] Use page_cache_xxx in mm/page-writeback.c clameter
2007-06-20 18:29 ` [05/37] Use page_cache_xxx in mm/truncate.c clameter
2007-06-20 18:29 ` [06/37] Use page_cache_xxx in mm/rmap.c clameter
2007-06-20 18:29 ` [07/37] Use page_cache_xxx in mm/filemap_xip.c clameter
2007-06-20 18:29 ` [08/37] Use page_cache_xxx in mm/migrate.c clameter
2007-06-20 18:29 ` [09/37] Use page_cache_xxx in fs/libfs.c clameter
2007-06-20 18:29 ` [10/37] Use page_cache_xxx in fs/sync clameter
2007-06-20 18:29 ` [11/37] Use page_cache_xxx in fs/buffer.c clameter
2007-06-20 18:29 ` [12/37] Use page_cache_xxx in mm/mpage.c clameter
2007-06-20 18:29 ` [13/37] Use page_cache_xxx in mm/fadvise.c clameter
2007-06-20 18:29 ` [14/37] Use page_cache_xxx in fs/splice.c clameter
2007-06-20 18:29 ` [15/37] Use page_cache_xxx functions in fs/ext2 clameter
2007-06-20 18:29 ` [16/37] Use page_cache_xxx in fs/ext3 clameter
2007-06-20 18:29 ` [17/37] Use page_cache_xxx in fs/ext4 clameter
2007-06-20 18:29 ` [18/37] Use page_cache_xxx in fs/reiserfs clameter
2007-06-20 18:29 ` [19/37] Use page_cache_xxx for fs/xfs clameter
2007-06-20 18:29 ` [20/37] Fix PAGE SIZE assumption in miscellaneous places clameter
2007-06-20 18:29 ` [21/37] Use page_cache_xxx in drivers/block/loop.c clameter
2007-06-20 18:29 ` [22/37] Use page_cache_xxx in drivers/block/rd.c clameter
2007-06-20 18:29 ` [23/37] compound pages: PageHead/PageTail instead of PageCompound clameter
2007-06-20 18:29 ` [24/37] compound pages: Add new support functions clameter
2007-06-20 18:29 ` [25/37] compound pages: vmstat support clameter
2007-06-20 18:29 ` [26/37] compound pages: Use new compound vmstat functions in SLUB clameter
2007-06-20 18:29 ` [27/37] compound pages: Allow use of get_page_unless_zero with compound pages clameter
2007-06-20 18:29 ` [28/37] compound pages: Allow freeing of compound pages via pagevec clameter
2007-06-20 18:29 ` [29/37] Large blocksize support: Fix up reclaim counters clameter
2007-06-20 18:29 ` [30/37] Add VM_BUG_ONs to check for correct page order clameter
2007-06-20 18:29 ` [31/37] Large blocksize support: Core piece clameter
2007-06-21 0:20 ` Bob Picco
2007-06-21 5:26 ` Christoph Lameter
2007-06-20 18:29 ` clameter [this message]
2007-06-20 18:29 ` [33/37] Large blocksize: Compound page zeroing and flushing clameter
2007-06-20 18:29 ` [34/37] Large blocksize support in ramfs clameter
2007-06-20 20:50 ` Andreas Dilger
2007-06-20 21:29 ` Christoph Lameter
2007-06-20 18:29 ` [35/37] Large blocksize support in XFS clameter
2007-06-20 18:29 ` [36/37] Large blocksize support for ext2 clameter
2007-06-20 20:56 ` Andreas Dilger
2007-06-20 21:27 ` Christoph Lameter
2007-06-20 22:19 ` Andreas Dilger
2007-06-20 18:29 ` [37/37] Reiserfs: Fix up for mapping_set_gfp_mask clameter
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20070620183013.496868776@sgi.com \
--to=clameter@sgi.com \
--cc=fengguang.wu@gmail.com \
--cc=hch@lst.de \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mel@skynet.ie \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).