Add readahead support in CoW for reflinked files.

PATCH Superseded Tao Ma 2 participants 15 messages Jun 29, 2010 – Jul 1, 2010 OCFS2

Share & tools

Tao Ma

author

2010-06-29 08:27 UTC #0 lore raw

cover Add readahead support in CoW for reflinked files.

Hi all,
When we try to CoW some clusters for a reflinked file, we have to read
the contents first, allocate some clusters and then map these pages to the new clusters. Currently, we use block_read_full_page to read it, but it is a little bit slower.
So this patch set try to add readahead support for CoW. Before we
start, we call readahead first so that the pages can be read at the very first. And during CoW, when we find a readahead page, we know that we need to move the readahead window, so a new asyncreadahead is called.

I have a small test to show how readahead speed up CoW. readahead_test()
{
MNT_DIR=/mnt/ocfs2
DEVICE=/dev/sda8
echo 'y'|mkfs.ocfs2 --fs-features=local,refcount $DEVICE mount -t ocfs2 $DEVICE $MNT_DIR FILE=$MNT_DIR/$RANDOM
REFLINK=$MNT_DIR/$RANDOM
dd if=/dev/zero of=$FILE bs=1M count=1000 reflink $FILE $REFLINK
dd if=/dev/zero of=$REFLINK bs=1M count=1000 conv=notrunc umount $MNT_DIR
}

Without these patch set, the 2nd dd has a i/o speed of 22MB/s. with the patch set, the i/o speed is increased to about 40MB/s.

Any comments are welcomed.

Regards,
Tao

Tao Ma

author

2010-06-29 08:35 UTC #1 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Add a new parameter 'struct file *' to ocfs2_refcount_cow so that we can add readahead support later.

Signed-off-by Tao Ma <[email protected]>

---
 fs/ocfs2/aops.c         |    2 +-
 fs/ocfs2/file.c         |   13 ++++++++-----
 fs/ocfs2/refcounttree.c |    4 +++-
 fs/ocfs2/refcounttree.h |    3 ++-
 4 files changed, 14 insertions(+), 8 deletions(-)

fs/ocfs2/aops.c +1 -1

index 4dfaa6e..5f94923 100644
@@ -1691,7 +1691,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 		mlog_errno(ret);
 		goto out;
 	} else if (ret == 1) {
-		ret = ocfs2_refcount_cow(inode, di_bh,
+		ret = ocfs2_refcount_cow(inode, NULL, di_bh,
 					 wc->w_cpos, wc->w_clen, UINT_MAX);
 		if (ret) {
 			mlog_errno(ret);

fs/ocfs2/file.c +8 -5

index ca527ca..06780f0 100644
@@ -361,7 +361,7 @@ static int ocfs2_cow_file_pos(struct inode *inode,
 	if (!(ext_flags & OCFS2_EXT_REFCOUNTED))
 		goto out;
 
-	return ocfs2_refcount_cow(inode, fe_bh, cpos, 1, cpos+1);
+	return ocfs2_refcount_cow(inode, NULL, fe_bh, cpos, 1, cpos+1);
 
 out:
 	return status;

@@ -1864,6 +1864,7 @@ out:
 }
 
 static int ocfs2_prepare_inode_for_refcount(struct inode *inode,
+					    struct file *file,
 					    loff_t pos, size_t count,
 					    int *meta_level)
 {

@@ -1881,7 +1882,7 @@ static int ocfs2_prepare_inode_for_refcount(struct inode *inode,
 
 	*meta_level = 1;
 
-	ret = ocfs2_refcount_cow(inode, di_bh, cpos, clusters, UINT_MAX);
+	ret = ocfs2_refcount_cow(inode, file, di_bh, cpos, clusters, UINT_MAX);
 	if (ret)
 		mlog_errno(ret);
 out:

@@ -1889,7 +1890,7 @@ out:
 	return ret;
 }
 
-static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
+static int ocfs2_prepare_inode_for_write(struct file *file,
 					 loff_t *ppos,
 					 size_t count,
 					 int appending,

@@ -1897,6 +1898,7 @@ static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
 					 int *has_refcount)
 {
 	int ret = 0, meta_level = 0;
+	struct dentry *dentry = file->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
 	loff_t saved_pos, end;

@@ -1952,6 +1954,7 @@ static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
 			meta_level = -1;
 
 			ret = ocfs2_prepare_inode_for_refcount(inode,
+							       file,
 							       saved_pos,
 							       count,
 							       &meta_level);

@@ -2066,7 +2069,7 @@ relock:
 	}
 
 	can_do_direct = direct_io;
-	ret = ocfs2_prepare_inode_for_write(file->f_path.dentry, ppos,
+	ret = ocfs2_prepare_inode_for_write(file, ppos,
 					    iocb->ki_left, appending,
 					    &can_do_direct, &has_refcount);
 	if (ret < 0) {

@@ -2193,7 +2196,7 @@ static int ocfs2_splice_to_file(struct pipe_inode_info *pipe,
 {
 	int ret;
 
-	ret = ocfs2_prepare_inode_for_write(out->f_path.dentry,	&sd->pos,
+	ret = ocfs2_prepare_inode_for_write(out, &sd->pos,
 					    sd->total_len, 0, NULL, NULL);
 	if (ret < 0) {
 		mlog_errno(ret);

fs/ocfs2/refcounttree.c +3 -1

index 4793f36..7636174 100644
@@ -3398,6 +3398,7 @@ static int ocfs2_replace_cow(struct ocfs2_cow_context *context)
  * unrefcounted extent.
  */
 static int ocfs2_refcount_cow_hunk(struct inode *inode,
+				   struct file *file,
 				   struct buffer_head *di_bh,
 				   u32 cpos, u32 write_len, u32 max_cpos)
 {

@@ -3475,6 +3476,7 @@ out:
  * clusters between cpos and cpos+write_len are safe to modify.
  */
 int ocfs2_refcount_cow(struct inode *inode,
+		       struct file *file,
 		       struct buffer_head *di_bh,
 		       u32 cpos, u32 write_len, u32 max_cpos)
 {

@@ -3494,7 +3496,7 @@ int ocfs2_refcount_cow(struct inode *inode,
 			num_clusters = write_len;
 
 		if (ext_flags & OCFS2_EXT_REFCOUNTED) {
-			ret = ocfs2_refcount_cow_hunk(inode, di_bh, cpos,
+			ret = ocfs2_refcount_cow_hunk(inode, file, di_bh, cpos,
 						      num_clusters, max_cpos);
 			if (ret) {
 				mlog_errno(ret);

fs/ocfs2/refcounttree.h +2 -1

index 9983ba1..29cba0e 100644
@@ -52,7 +52,8 @@ int ocfs2_prepare_refcount_change_for_del(struct inode *inode,
 					  u32 clusters,
 					  int *credits,
 					  int *ref_blocks);
-int ocfs2_refcount_cow(struct inode *inode, struct buffer_head *di_bh,
+int ocfs2_refcount_cow(struct inode *inode,
+		       struct file *filep, struct buffer_head *di_bh,
 		       u32 cpos, u32 write_len, u32 max_cpos);
 
 typedef int (ocfs2_post_refcount_func)(struct inode *inode,

Signature · 1 line

1.7.1.GIT

Joel Becker

2010-06-29 23:14 UTC #2 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Quoted 16 lines · Tao Ma

Add a new parameter 'struct file *' to ocfs2_refcount_cow so that we can add readahead support later.
Signed-off-by Tao Ma <[email protected]>
---
 fs/ocfs2/aops.c         |    2 +-
 fs/ocfs2/file.c         |   13 ++++++++-----
 fs/ocfs2/refcounttree.c |    4 +++-
 fs/ocfs2/refcounttree.h |    3 ++-
 4 files changed, 14 insertions(+), 8 deletions(-)
fs/ocfs2/aops.c +0 -0
index 4dfaa6e..5f94923 100644

fs/ocfs2/aops.c int ocfs2_write_begin_nolock(struct address_space *mapping, · line 1695

Quoted 6 lines

fs/ocfs2/aops.c +1 -1

@@ -1691,7 +1691,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 		mlog_errno(ret);
 		goto out;
 	} else if (ret == 1) {
-		ret = ocfs2_refcount_cow(inode, di_bh,
+		ret = ocfs2_refcount_cow(inode, NULL, di_bh,
 					 wc->w_cpos, wc->w_clen, UINT_MAX);

You should be replacing the inode parameter with the file
parameter. You can always get back to the inode from the filp. When I first saw this, I had to read through your entire series to figure out if a NULL filp was valid. It's not. In the end, you change write_begin_nolock() to take (inode, filp), which is pointless, because write_begin() always gets a valid filp. So does page_mkwrite().
So in your first patch of the series, change
ocfs2_write_begin_nolock() and __ocfs2_page_mkwrite() to take the filp instead of the inode. Then in your later patches you can always expect the filp to be valid.

Joel

Signature · 10 lines

Life's Little Instruction Book #464

	"Don't miss the magic of the moment by focusing on what's
	 to come."

Joel Becker
Consulting Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127

Tao Ma

author

2010-06-30 01:34 UTC #3 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Quoted 26 lines · Joel Becker

Quoted 16 lines · Tao Ma
Add a new parameter 'struct file *' to ocfs2_refcount_cow so that we can add readahead support later.
Signed-off-by Tao Ma<[email protected]>
---
  fs/ocfs2/aops.c         |    2 +-
  fs/ocfs2/file.c         |   13 ++++++++-----
  fs/ocfs2/refcounttree.c |    4 +++-
  fs/ocfs2/refcounttree.h |    3 ++-
  4 files changed, 14 insertions(+), 8 deletions(-)
fs/ocfs2/aops.c +0 -0
index 4dfaa6e..5f94923 100644
fs/ocfs2/aops.c int ocfs2_write_begin_nolock(struct address_space *mapping, · line 1695
Quoted 6 lines
fs/ocfs2/aops.c +1 -1
@@ -1691,7 +1691,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
  		mlog_errno(ret);
  		goto out;
  	} else if (ret == 1) {
-		ret = ocfs2_refcount_cow(inode, di_bh,
+		ret = ocfs2_refcount_cow(inode, NULL, di_bh,
  					 wc->w_cpos, wc->w_clen, UINT_MAX);
You should be replacing the inode parameter with the file
parameter. You can always get back to the inode from the filp. When I

Quoted 8 lines

first saw this, I had to read through your entire series to figure out if a NULL filp was valid. It's not. In the end, you change write_begin_nolock() to take (inode, filp), which is pointless, because write_begin() always gets a valid filp. So does page_mkwrite().
So in your first patch of the series, change
ocfs2_write_begin_nolock() and __ocfs2_page_mkwrite() to take the filp instead of the inode. Then in your later patches you can always expect the filp to be valid.

make sense. I will change it. thanks.

btw, I forget to say this series is targeted to the next merge window. ;)

Regards,
Tao

Quoted 3 lines

Joel

Tao Ma

author

2010-06-30 03:11 UTC #4 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

In reply to Joel Becker

Hi Joel,

Quoted 26 lines · Joel Becker

Quoted 16 lines · Tao Ma
Add a new parameter 'struct file *' to ocfs2_refcount_cow so that we can add readahead support later.
Signed-off-by Tao Ma<[email protected]>
---
  fs/ocfs2/aops.c         |    2 +-
  fs/ocfs2/file.c         |   13 ++++++++-----
  fs/ocfs2/refcounttree.c |    4 +++-
  fs/ocfs2/refcounttree.h |    3 ++-
  4 files changed, 14 insertions(+), 8 deletions(-)
fs/ocfs2/aops.c +0 -0
index 4dfaa6e..5f94923 100644
fs/ocfs2/aops.c int ocfs2_write_begin_nolock(struct address_space *mapping, · line 1695
Quoted 6 lines
fs/ocfs2/aops.c +1 -1
@@ -1691,7 +1691,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
  		mlog_errno(ret);
  		goto out;
  	} else if (ret == 1) {
-		ret = ocfs2_refcount_cow(inode, di_bh,
+		ret = ocfs2_refcount_cow(inode, NULL, di_bh,
  					 wc->w_cpos, wc->w_clen, UINT_MAX);
You should be replacing the inode parameter with the file
parameter. You can always get back to the inode from the filp. When I

Quoted 8 lines

first saw this, I had to read through your entire series to figure out if a NULL filp was valid. It's not. In the end, you change write_begin_nolock() to take (inode, filp), which is pointless, because write_begin() always gets a valid filp. So does page_mkwrite().
So in your first patch of the series, change
ocfs2_write_begin_nolock() and __ocfs2_page_mkwrite() to take the filp instead of the inode. Then in your later patches you can always expect the filp to be valid.

oh, sorry joel. There does exist one place that we don't have filp. It is in ocfs2_orphan_for_truncate, we have to CoW the cluster containing i_size and pass NULL as filp.

Regards,
Tao

Joel Becker

2010-06-30 12:00 UTC #5 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Quoted 3 lines · Tao Ma

oh, sorry joel. There does exist one place that we don't have filp. It is in ocfs2_orphan_for_truncate, we have to CoW the cluster containing i_size and pass NULL as filp.

Maybe ocfs2_cow_file_pos() needs both, but write_begin_nolock
and page_mkwrite do not.

Joel

Signature · 9 lines

"I don't want to achieve immortality through my work; I want to
 achieve immortality through not dying."
        - Woody Allen

Joel Becker
Consulting Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127

Tao Ma

author

2010-07-01 00:17 UTC #6 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Quoted 10 lines · Joel Becker

Quoted 4 lines · Tao Ma
oh, sorry joel. There does exist one place that we don't have filp. It is in ocfs2_orphan_for_truncate, we have to CoW the cluster containing i_size and pass NULL as filp.
Maybe ocfs2_cow_file_pos() needs both, but write_begin_nolock
and page_mkwrite do not.

yes, so what do you mean? I am puzzled, sorry.
You mean we export 2 different functions from refcountree.c. One is ocfs2_refcount_cow_file(struct file *,...) and another is ocfs2_refcount_cow_inode(struct inode *,...)?

Regards,
Tao

Joel Becker

2010-07-01 09:12 UTC #7 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Quoted 6 lines · Tao Ma

Quoted 5 lines · Joel Becker
Quoted 3 lines · Tao Ma
oh, sorry joel. There does exist one place that we don't have filp. It is in ocfs2_orphan_for_truncate, we have to CoW the cluster containing i_size and pass NULL as filp.

Quoted 6 lines

Quoted 2 lines
Maybe ocfs2_cow_file_pos() needs both, but write_begin_nolock
and page_mkwrite do not.
yes, so what do you mean? I am puzzled, sorry.
You mean we export 2 different functions from refcountree.c. One is ocfs2_refcount_cow_file(struct file *,...) and another is ocfs2_refcount_cow_inode(struct inode *,...)?

No. Just add have refcount_cow() take both an inode and a file,
and pass the NULL file from orphan_for_truncate(). That's fine.
But ocfs2_write_begin_nolock() and and ocfs2_page_mkwrite() can
take a file without having to take an inode.

Joel

Signature · 9 lines

Life's Little Instruction Book #252

	"Take good care of those you love."

Joel Becker
Consulting Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127

Tao Ma

author

2010-07-01 09:20 UTC #8 lore raw

PATCH 1/4 ocfs2: Add struct file to ocfs2_refcount_cow.

Quoted 14 lines · Joel Becker

Quoted 6 lines · Tao Ma
Quoted 5 lines · Joel Becker
Quoted 3 lines · Tao Ma
oh, sorry joel. There does exist one place that we don't have filp. It is in ocfs2_orphan_for_truncate, we have to CoW the cluster containing i_size and pass NULL as filp.
Quoted 6 lines
Quoted 2 lines
Maybe ocfs2_cow_file_pos() needs both, but write_begin_nolock
and page_mkwrite do not.
yes, so what do you mean? I am puzzled, sorry.
You mean we export 2 different functions from refcountree.c. One is ocfs2_refcount_cow_file(struct file *,...) and another is ocfs2_refcount_cow_inode(struct inode *,...)?

Quoted 4 lines

No. Just add have refcount_cow() take both an inode and a file,
and pass the NULL file from orphan_for_truncate(). That's fine.
But ocfs2_write_begin_nolock() and and ocfs2_page_mkwrite() can
take a file without having to take an inode.

oh, I see. thanks.

Regards,
Tao

Tao Ma

author

2010-06-29 08:35 UTC #9 lore raw

PATCH 2/4 ocfs2: Add readahead support for CoW.

In reply to Tao Ma

Add a new function ocfs2_readahead_for_cow so that we start readahead before we start our CoW.

Signed-off-by Tao Ma <[email protected]>

---
 fs/ocfs2/refcounttree.c |   24 ++++++++++++++++++++++++
 1 files changed, 24 insertions(+), 0 deletions(-)

fs/ocfs2/refcounttree.c +24 -0

index 7636174..03ec6ac 100644
@@ -3392,6 +3392,28 @@ static int ocfs2_replace_cow(struct ocfs2_cow_context *context)
 	return ret;
 }
 
+static void ocfs2_readahead_for_cow(struct inode *inode,
+				    struct file *file,
+				    u32 start, u32 len)
+{
+	struct address_space *mapping;
+	pgoff_t index;
+	unsigned long num_pages;
+	int cs_bits = OCFS2_SB(inode->i_sb)->s_clustersize_bits;
+
+	if (!file)
+		return;
+
+	mapping = file->f_mapping;
+	num_pages = (len << cs_bits) >> PAGE_CACHE_SHIFT;
+	if (!num_pages)
+		num_pages = 1;
+
+	index = ((loff_t)start << cs_bits) >> PAGE_CACHE_SHIFT;
+	page_cache_sync_readahead(mapping, &file->f_ra, file,
+				  index, num_pages);
+}
+
 /*
  * Starting at cpos, try to CoW write_len clusters.  Don't CoW
  * past max_cpos.  This will stop when it runs into a hole or an

@@ -3427,6 +3449,8 @@ static int ocfs2_refcount_cow_hunk(struct inode *inode,
 
 	BUG_ON(cow_len == 0);
 
+	ocfs2_readahead_for_cow(inode, file, cow_start, cow_len);
+
 	context = kzalloc(sizeof(struct ocfs2_cow_context), GFP_NOFS);
 	if (!context) {
 		ret = -ENOMEM;

Signature · 1 line

1.7.1.GIT

Tao Ma

author

2010-06-29 08:35 UTC #10 lore raw

PATCH 3/4 ocfs2: Add readhead during CoW.

In reply to Tao Ma

In CoW, when we meet with a readahead page, we know it is time to move the readahead window. So carry out a new readahead.

Signed-off-by Tao Ma <[email protected]>

---
 fs/ocfs2/refcounttree.c |   15 ++++++++++++++-
 1 files changed, 14 insertions(+), 1 deletions(-)

fs/ocfs2/refcounttree.c +14 -1

index 03ec6ac..df47182 100644
@@ -49,6 +49,7 @@
 
 struct ocfs2_cow_context {
 	struct inode *inode;
+	struct file *file;
 	u32 cow_start;
 	u32 cow_len;
 	struct ocfs2_extent_tree data_et;

@@ -2922,13 +2923,16 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle,
 	u64 new_block = ocfs2_clusters_to_blocks(sb, new_cluster);
 	struct page *page;
 	pgoff_t page_index;
-	unsigned int from, to;
+	unsigned int from, to, readahead_pages;
 	loff_t offset, end, map_end;
 	struct address_space *mapping = context->inode->i_mapping;
 
 	mlog(0, "old_cluster %u, new %u, len %u at offset %u\n", old_cluster,
 	     new_cluster, new_len, cpos);
 
+	readahead_pages =
+		(ocfs2_cow_contig_clusters(sb) <<
+		 OCFS2_SB(sb)->s_clustersize_bits) >> PAGE_CACHE_SHIFT;
 	offset = ((loff_t)cpos) << OCFS2_SB(sb)->s_clustersize_bits;
 	end = offset + (new_len << OCFS2_SB(sb)->s_clustersize_bits);

@@ -2953,6 +2957,14 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle,
 		if (PAGE_CACHE_SIZE <= OCFS2_SB(sb)->s_clustersize)
 			BUG_ON(PageDirty(page));
 
+		if (PageReadahead(page) && context->file) {
+			page_cache_async_readahead(mapping,
+						   &context->file->f_ra,
+						   context->file,
+						   page, page_index,
+						   readahead_pages);
+		}
+
 		if (!PageUptodate(page)) {
 			ret = block_read_full_page(page, ocfs2_get_block);
 			if (ret) {

@@ -3472,6 +3484,7 @@ static int ocfs2_refcount_cow_hunk(struct inode *inode,
 	context->ref_root_bh = ref_root_bh;
 	context->cow_duplicate_clusters = ocfs2_duplicate_clusters_by_page;
 	context->get_clusters = ocfs2_di_get_clusters;
+	context->file = file;
 
 	ocfs2_init_dinode_extent_tree(&context->data_et,
 				      INODE_CACHE(inode), di_bh);

Signature · 1 line

1.7.1.GIT

Joel Becker

2010-06-29 23:23 UTC #11 lore raw

PATCH 3/4 ocfs2: Add readhead during CoW.

fs/ocfs2/refcounttree.c static int ocfs2_duplicate_clusters_by_page(handle_t *handle, · line 2966

Quoted 11 lines · Tao Ma

fs/ocfs2/refcounttree.c +7 -0

@@ -2953,6 +2957,14 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle,
 		if (PAGE_CACHE_SIZE <= OCFS2_SB(sb)->s_clustersize)
 			BUG_ON(PageDirty(page));
 
+		if (PageReadahead(page) && context->file) {
+			page_cache_async_readahead(mapping,
+						   &context->file->f_ra,
+						   context->file,
+						   page, page_index,
+						   readahead_pages);
+		}

This is merely re-sending the same pages that were already sent,
right? In the previous patch, you asked the readahead code to try all pages in the hunk. Now you've discovered a page that isn't yet up to date, and you send it (and the 1M next to it) back to readahead.
This is, I assume, because the readahead code doesn't actually
read your entire request from page_cache_sync_readahead(). It just reads some, and this is you hinting that you need the next bit. Am I right?

Joel

Signature · 10 lines

Life's Little Instruction Book #464

	"Don't miss the magic of the moment by focusing on what's
	 to come."

Joel Becker
Consulting Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127

Tao Ma

author

2010-06-30 01:46 UTC #12 lore raw

PATCH 3/4 ocfs2: Add readhead during CoW.

Hi Joel,

Quoted 5 lines · Joel Becker

Quoted 4 lines · Tao Ma

fs/ocfs2/refcounttree.c +0 -0

@@ -2953,6 +2957,14 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle,
  		if (PAGE_CACHE_SIZE<= OCFS2_SB(sb)->s_clustersize)
  			BUG_ON(PageDirty(page));

Quoted 16 lines

fs/ocfs2/refcounttree.c static int ocfs2_duplicate_clusters_by_page(handle_t *handle, · line 2966
Quoted 7 lines
fs/ocfs2/refcounttree.c +7 -0
+		if (PageReadahead(page)&&  context->file) {
+			page_cache_async_readahead(mapping,
+						&context->file->f_ra,
+						   context->file,
+						   page, page_index,
+						   readahead_pages);
+		}
This is merely re-sending the same pages that were already sent,
right? In the previous patch, you asked the readahead code to try all pages in the hunk. Now you've discovered a page that isn't yet up to date, and you send it (and the 1M next to it) back to readahead.
This is, I assume, because the readahead code doesn't actually
read your entire request from page_cache_sync_readahead(). It just reads some, and this is you hinting that you need the next bit. Am I right?

The first previous patch just let the caller to do the readahead for the whole hunk and set PG_readahead to a page as the start of readahead window. So when we meet with a page with PG_readahead flag set, we know it's time to move our readahead window so a new readahead is issued here.

Fengguang Wu has a document named "On the Design of a New Linux Readahead Framework", you can refer to it and hope I read it clearly.

Regards,
Tao

Tao Ma

author

2010-06-29 08:35 UTC #13 lore raw

PATCH 4/4 ocfs2: make mmap CoW work with readahead.

In reply to Tao Ma

refcount_cow need file * to do the readahead.
So we just add 'struct file *' to ocfs2_write_begin_nolock and __ocfs2_page_mkwrite and let them pass this parameter to ocfs2_refcount_cow.

Signed-off-by Tao Ma <[email protected]>

---
 fs/ocfs2/aops.c |    7 ++++---
 fs/ocfs2/aops.h |    3 ++-
 fs/ocfs2/mmap.c |    8 ++++----
 3 files changed, 10 insertions(+), 8 deletions(-)

fs/ocfs2/aops.c +4 -3

index 5f94923..337e874 100644
@@ -1645,7 +1645,8 @@ static int ocfs2_expand_nonsparse_inode(struct inode *inode, loff_t pos,
 	return ret;
 }
 
-int ocfs2_write_begin_nolock(struct address_space *mapping,
+int ocfs2_write_begin_nolock(struct file *file,
+			     struct address_space *mapping,
 			     loff_t pos, unsigned len, unsigned flags,
 			     struct page **pagep, void **fsdata,
 			     struct buffer_head *di_bh, struct page *mmap_page)

@@ -1691,7 +1692,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 		mlog_errno(ret);
 		goto out;
 	} else if (ret == 1) {
-		ret = ocfs2_refcount_cow(inode, NULL, di_bh,
+		ret = ocfs2_refcount_cow(inode, file, di_bh,
 					 wc->w_cpos, wc->w_clen, UINT_MAX);
 		if (ret) {
 			mlog_errno(ret);

@@ -1853,7 +1854,7 @@ static int ocfs2_write_begin(struct file *file, struct address_space *mapping,
 	 */
 	down_write(&OCFS2_I(inode)->ip_alloc_sem);
 
-	ret = ocfs2_write_begin_nolock(mapping, pos, len, flags, pagep,
+	ret = ocfs2_write_begin_nolock(file, mapping, pos, len, flags, pagep,
 				       fsdata, di_bh, NULL);
 	if (ret) {
 		mlog_errno(ret);

fs/ocfs2/aops.h +2 -1

index c48e93f..e99ac0f 100644
@@ -48,7 +48,8 @@ int ocfs2_write_end_nolock(struct address_space *mapping,
 			   loff_t pos, unsigned len, unsigned copied,
 			   struct page *page, void *fsdata);
 
-int ocfs2_write_begin_nolock(struct address_space *mapping,
+int ocfs2_write_begin_nolock(struct file *file,
+			     struct address_space *mapping,
 			     loff_t pos, unsigned len, unsigned flags,
 			     struct page **pagep, void **fsdata,
 			     struct buffer_head *di_bh, struct page *mmap_page);

fs/ocfs2/mmap.c +4 -4

index af2b8fe..7fcb107 100644
@@ -59,8 +59,8 @@ static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
 	return ret;
 }
 
-static int __ocfs2_page_mkwrite(struct inode *inode, struct buffer_head *di_bh,
-				struct page *page)
+static int __ocfs2_page_mkwrite(struct file *file, struct inode *inode,
+				struct buffer_head *di_bh, struct page *page)
 {
 	int ret;
 	struct address_space *mapping = inode->i_mapping;

@@ -109,7 +109,7 @@ static int __ocfs2_page_mkwrite(struct inode *inode, struct buffer_head *di_bh,
 	if (page->index == last_index)
 		len = size & ~PAGE_CACHE_MASK;
 
-	ret = ocfs2_write_begin_nolock(mapping, pos, len, 0, &locked_page,
+	ret = ocfs2_write_begin_nolock(file, mapping, pos, len, 0, &locked_page,
 				       &fsdata, di_bh, page);
 	if (ret) {
 		if (ret != -ENOSPC)

@@ -157,7 +157,7 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	 */
 	down_write(&OCFS2_I(inode)->ip_alloc_sem);
 
-	ret = __ocfs2_page_mkwrite(inode, di_bh, page);
+	ret = __ocfs2_page_mkwrite(vma->vm_file, inode, di_bh, page);
 
 	up_write(&OCFS2_I(inode)->ip_alloc_sem);

Signature · 1 line

1.7.1.GIT

Joel Becker

2010-06-29 23:24 UTC #14 lore raw

PATCH 4/4 ocfs2: make mmap CoW work with readahead.

Quoted 4 lines · Tao Ma

refcount_cow need file * to do the readahead.
So we just add 'struct file *' to ocfs2_write_begin_nolock and __ocfs2_page_mkwrite and let them pass this parameter to ocfs2_refcount_cow.

As I stated in my comment on the first patch, these functions
only need the file*. The inode can be gotten off of that.

Joel

Signature · 8 lines

"I have never let my schooling interfere with my education."
        - Mark Twain

Joel Becker
Consulting Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127