[PATCH v2 3/7] f2fs: add tracepoint for tracing the page i/o operations

[PATCH v2 3/7] f2fs: add tracepoint for tracing the page i/o operations

[Namjae Jeon]

From: Namjae Jeon <namjae.jeon@samsung.com>

Add tracepoints for page i/o operations and block allocation
tracing during page read operation.

Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
Signed-off-by: Pankaj Kumar <pankaj.km@samsung.com>
---
 fs/f2fs/data.c              |   11 ++++++--
 include/trace/events/f2fs.h |   58 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 02ad450..bb9117b 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -22,6 +22,7 @@
 #include "f2fs.h"
 #include "node.h"
 #include "segment.h"
+#include <trace/events/f2fs.h>
 
 /*
  * Lock ordering for the change of data block address:
@@ -335,6 +336,9 @@ int f2fs_readpage(struct f2fs_sb_info *sbi, struct page *page,
 	bool sync = (type == READ_SYNC);
 	struct bio *bio;
 
+	if (page->mapping)
+		trace_f2fs_readpage(page);
+
 	/* This page can be already read by other threads */
 	if (PageUptodate(page)) {
 		if (!sync)
@@ -385,6 +389,7 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
 	pgoff_t pgofs;
 	int err;
 
+	trace_f2fs_get_data_block_enter(inode, iblock, 0);
 	/* Get the page offset from the block offset(iblock) */
 	pgofs =	(pgoff_t)(iblock >> (PAGE_CACHE_SHIFT - blkbits));
 
@@ -394,9 +399,10 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
 	/* When reading holes, we need its node page */
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = get_dnode_of_data(&dn, pgofs, LOOKUP_NODE_RA);
-	if (err)
+	if (err) {
+		trace_f2fs_get_data_block_exit(inode, iblock, err);
 		return (err == -ENOENT) ? 0 : err;
-
+	}
 	/* It does not support data allocation */
 	BUG_ON(create);
 
@@ -420,6 +426,7 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
 		bh_result->b_size = (i << blkbits);
 	}
 	f2fs_put_dnode(&dn);
+	trace_f2fs_get_data_block_exit(inode, iblock, 0);
 	return 0;
 }
 
diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h
index 0d39f58..d60f3ed 100644
--- a/include/trace/events/f2fs.h
+++ b/include/trace/events/f2fs.h
@@ -266,6 +266,64 @@ TRACE_EVENT(f2fs_truncate,
 		  (unsigned long) __entry->ino)
 );
 
+TRACE_EVENT(f2fs_readpage,
+	TP_PROTO(struct page *page),
+
+	TP_ARGS(page),
+
+	TP_STRUCT__entry(
+		__field(pgoff_t, index)
+		__field(ino_t,	ino)
+		__field(dev_t,	dev)
+
+	),
+
+	TP_fast_assign(
+		__entry->index	= page->index;
+		__entry->ino	= page->mapping->host->i_ino;
+		__entry->dev	= page->mapping->host->i_sb->s_dev;
+	),
+
+	TP_printk("dev %d,%d ino %lu page_index %lu",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  (unsigned long) __entry->ino,
+		  (unsigned long) __entry->index)
+);
+
+DECLARE_EVENT_CLASS(f2fs_data_block,
+	TP_PROTO(struct inode *inode, sector_t block, int ret),
+
+	TP_ARGS(inode, block, ret),
+
+	TP_STRUCT__entry(
+		__field(dev_t,	dev)
+		__field(ino_t,	ino)
+		__field(sector_t,	block)
+		__field(int,	ret)
+	),
+
+	TP_fast_assign(
+		__entry->dev	= inode->i_sb->s_dev;
+		__entry->ino	= inode->i_ino;
+		__entry->block	= block;
+		__entry->ret	= ret;
+	),
+
+	TP_printk("dev %d,%d ino %lu block number %llu error %d",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  (unsigned long) __entry->ino, (unsigned long long) __entry->block, __entry->ret)
+);
+
+DEFINE_EVENT(f2fs_data_block, f2fs_get_data_block_enter,
+	TP_PROTO(struct inode *inode, sector_t block, int ret),
+	TP_ARGS(inode, block, ret)
+);
+
+DEFINE_EVENT(f2fs_data_block, f2fs_get_data_block_exit,
+	TP_PROTO(struct inode *inode, sector_t block, int ret),
+	TP_ARGS(inode, block, ret)
+);
+
 #endif /* _TRACE_F2FS_H */
 
  /* This part must be outside protection */
-- 
1.7.9.5

Re: [PATCH v2 3/7] f2fs: add tracepoint for tracing the page i/o operations

[Steven Rostedt]

On Sun, 2013-03-17 at 17:40 +0900, Namjae Jeon wrote:
> From: Namjae Jeon <namjae.jeon@samsung.com>
> 
> Add tracepoints for page i/o operations and block allocation
> tracing during page read operation.
> 
> Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
> Signed-off-by: Pankaj Kumar <pankaj.km@samsung.com>
> ---
>  fs/f2fs/data.c              |   11 ++++++--
>  include/trace/events/f2fs.h |   58 +++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 67 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
> index 02ad450..bb9117b 100644
> --- a/fs/f2fs/data.c
> +++ b/fs/f2fs/data.c
> @@ -22,6 +22,7 @@
>  #include "f2fs.h"
>  #include "node.h"
>  #include "segment.h"
> +#include <trace/events/f2fs.h>
>  
>  /*
>   * Lock ordering for the change of data block address:
> @@ -335,6 +336,9 @@ int f2fs_readpage(struct f2fs_sb_info *sbi, struct page *page,
>  	bool sync = (type == READ_SYNC);
>  	struct bio *bio;
>  
> +	if (page->mapping)
> +		trace_f2fs_readpage(page);

Since the only thing that happens whet page->mapping is set is to trace
the page, you can move that check into the tracepoint itself. That is,
move the branch into the tracepoint, making the branch a nop when the
tracepoint isn't active. That's what TRACE_EVENT_CONDITION() does. Make
this into:

	trace_f2fs_readpage(page);

and then see below.

> +
>  	/* This page can be already read by other threads */
>  	if (PageUptodate(page)) {
>  		if (!sync)
> @@ -385,6 +389,7 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
>  	pgoff_t pgofs;
>  	int err;
>  
> +	trace_f2fs_get_data_block_enter(inode, iblock, 0);
>  	/* Get the page offset from the block offset(iblock) */
>  	pgofs =	(pgoff_t)(iblock >> (PAGE_CACHE_SHIFT - blkbits));
>  
> @@ -394,9 +399,10 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
>  	/* When reading holes, we need its node page */
>  	set_new_dnode(&dn, inode, NULL, NULL, 0);
>  	err = get_dnode_of_data(&dn, pgofs, LOOKUP_NODE_RA);
> -	if (err)
> +	if (err) {
> +		trace_f2fs_get_data_block_exit(inode, iblock, err);
>  		return (err == -ENOENT) ? 0 : err;
> -
> +	}
>  	/* It does not support data allocation */
>  	BUG_ON(create);
>  
> @@ -420,6 +426,7 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
>  		bh_result->b_size = (i << blkbits);
>  	}
>  	f2fs_put_dnode(&dn);
> +	trace_f2fs_get_data_block_exit(inode, iblock, 0);
>  	return 0;
>  }
>  
> diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h
> index 0d39f58..d60f3ed 100644
> --- a/include/trace/events/f2fs.h
> +++ b/include/trace/events/f2fs.h
> @@ -266,6 +266,64 @@ TRACE_EVENT(f2fs_truncate,
>  		  (unsigned long) __entry->ino)
>  );
>  
> +TRACE_EVENT(f2fs_readpage,

TRACE_EVENT_CONDITIO(f2fs_readpage,

> +	TP_PROTO(struct page *page),
> +
> +	TP_ARGS(page),

	TP_CONDITION(page->mapping),

That's it!

Now the tracepoint will only be called if page->mapping is set, but
checking page->mapping will be done only if the tracepoint is enabled.
Keeping the original code untouched.

-- Steve

> +
> +	TP_STRUCT__entry(
> +		__field(pgoff_t, index)
> +		__field(ino_t,	ino)
> +		__field(dev_t,	dev)
> +
> +	),
> +
> +	TP_fast_assign(
> +		__entry->index	= page->index;
> +		__entry->ino	= page->mapping->host->i_ino;
> +		__entry->dev	= page->mapping->host->i_sb->s_dev;
> +	),
> +
> +	TP_printk("dev %d,%d ino %lu page_index %lu",
> +		  MAJOR(__entry->dev), MINOR(__entry->dev),
> +		  (unsigned long) __entry->ino,
> +		  (unsigned long) __entry->index)
> +);
> +
> +DECLARE_EVENT_CLASS(f2fs_data_block,
> +	TP_PROTO(struct inode *inode, sector_t block, int ret),
> +
> +	TP_ARGS(inode, block, ret),
> +
> +	TP_STRUCT__entry(
> +		__field(dev_t,	dev)
> +		__field(ino_t,	ino)
> +		__field(sector_t,	block)
> +		__field(int,	ret)
> +	),
> +
> +	TP_fast_assign(
> +		__entry->dev	= inode->i_sb->s_dev;
> +		__entry->ino	= inode->i_ino;
> +		__entry->block	= block;
> +		__entry->ret	= ret;
> +	),
> +
> +	TP_printk("dev %d,%d ino %lu block number %llu error %d",
> +		  MAJOR(__entry->dev), MINOR(__entry->dev),
> +		  (unsigned long) __entry->ino, (unsigned long long) __entry->block, __entry->ret)
> +);
> +
> +DEFINE_EVENT(f2fs_data_block, f2fs_get_data_block_enter,
> +	TP_PROTO(struct inode *inode, sector_t block, int ret),
> +	TP_ARGS(inode, block, ret)
> +);
> +
> +DEFINE_EVENT(f2fs_data_block, f2fs_get_data_block_exit,
> +	TP_PROTO(struct inode *inode, sector_t block, int ret),
> +	TP_ARGS(inode, block, ret)
> +);
> +
>  #endif /* _TRACE_F2FS_H */
>  
>   /* This part must be outside protection */

Re: [PATCH v2 3/7] f2fs: add tracepoint for tracing the page i/o operations

[Namjae Jeon]

2013/3/17, Steven Rostedt <rostedt@goodmis.org>:
> On Sun, 2013-03-17 at 17:40 +0900, Namjae Jeon wrote:
>> From: Namjae Jeon <namjae.jeon@samsung.com>
>>
>> Add tracepoints for page i/o operations and block allocation
>> tracing during page read operation.
>>
>> Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com>
>> Signed-off-by: Pankaj Kumar <pankaj.km@samsung.com>
>> ---
>>  fs/f2fs/data.c              |   11 ++++++--
>>  include/trace/events/f2fs.h |   58
>> +++++++++++++++++++++++++++++++++++++++++++
>>  2 files changed, 67 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
>> index 02ad450..bb9117b 100644
>> --- a/fs/f2fs/data.c
>> +++ b/fs/f2fs/data.c
>> @@ -22,6 +22,7 @@
>>  #include "f2fs.h"
>>  #include "node.h"
>>  #include "segment.h"
>> +#include <trace/events/f2fs.h>
>>
>>  /*
>>   * Lock ordering for the change of data block address:
>> @@ -335,6 +336,9 @@ int f2fs_readpage(struct f2fs_sb_info *sbi, struct
>> page *page,
>>  	bool sync = (type == READ_SYNC);
>>  	struct bio *bio;
>>
>> +	if (page->mapping)
>> +		trace_f2fs_readpage(page);
>
> Since the only thing that happens whet page->mapping is set is to trace
> the page, you can move that check into the tracepoint itself. That is,
> move the branch into the tracepoint, making the branch a nop when the
> tracepoint isn't active. That's what TRACE_EVENT_CONDITION() does. Make
> this into:
>
> 	trace_f2fs_readpage(page);
>
> and then see below.
>
>> +
>>  	/* This page can be already read by other threads */
>>  	if (PageUptodate(page)) {
>>  		if (!sync)
>> @@ -385,6 +389,7 @@ static int get_data_block_ro(struct inode *inode,
>> sector_t iblock,
>>  	pgoff_t pgofs;
>>  	int err;
>>
>> +	trace_f2fs_get_data_block_enter(inode, iblock, 0);
>>  	/* Get the page offset from the block offset(iblock) */
>>  	pgofs =	(pgoff_t)(iblock >> (PAGE_CACHE_SHIFT - blkbits));
>>
>> @@ -394,9 +399,10 @@ static int get_data_block_ro(struct inode *inode,
>> sector_t iblock,
>>  	/* When reading holes, we need its node page */
>>  	set_new_dnode(&dn, inode, NULL, NULL, 0);
>>  	err = get_dnode_of_data(&dn, pgofs, LOOKUP_NODE_RA);
>> -	if (err)
>> +	if (err) {
>> +		trace_f2fs_get_data_block_exit(inode, iblock, err);
>>  		return (err == -ENOENT) ? 0 : err;
>> -
>> +	}
>>  	/* It does not support data allocation */
>>  	BUG_ON(create);
>>
>> @@ -420,6 +426,7 @@ static int get_data_block_ro(struct inode *inode,
>> sector_t iblock,
>>  		bh_result->b_size = (i << blkbits);
>>  	}
>>  	f2fs_put_dnode(&dn);
>> +	trace_f2fs_get_data_block_exit(inode, iblock, 0);
>>  	return 0;
>>  }
>>
>> diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h
>> index 0d39f58..d60f3ed 100644
>> --- a/include/trace/events/f2fs.h
>> +++ b/include/trace/events/f2fs.h
>> @@ -266,6 +266,64 @@ TRACE_EVENT(f2fs_truncate,
>>  		  (unsigned long) __entry->ino)
>>  );
>>
>> +TRACE_EVENT(f2fs_readpage,
>
> TRACE_EVENT_CONDITIO(f2fs_readpage,
>
>> +	TP_PROTO(struct page *page),
>> +
>> +	TP_ARGS(page),
>
> 	TP_CONDITION(page->mapping),
>
> That's it!
>
> Now the tracepoint will only be called if page->mapping is set, but
> checking page->mapping will be done only if the tracepoint is enabled.
> Keeping the original code untouched.
Hi Steve.
Okay :) I will fix it. Thanks for your review !
>
> -- Steve
>
>> +
>> +	TP_STRUCT__entry(
>> +		__field(pgoff_t, index)
>> +		__field(ino_t,	ino)
>> +		__field(dev_t,	dev)
>> +
>> +	),
>> +
>> +	TP_fast_assign(
>> +		__entry->index	= page->index;
>> +		__entry->ino	= page->mapping->host->i_ino;
>> +		__entry->dev	= page->mapping->host->i_sb->s_dev;
>> +	),
>> +
>> +	TP_printk("dev %d,%d ino %lu page_index %lu",
>> +		  MAJOR(__entry->dev), MINOR(__entry->dev),
>> +		  (unsigned long) __entry->ino,
>> +		  (unsigned long) __entry->index)
>> +);
>> +
>> +DECLARE_EVENT_CLASS(f2fs_data_block,
>> +	TP_PROTO(struct inode *inode, sector_t block, int ret),
>> +
>> +	TP_ARGS(inode, block, ret),
>> +
>> +	TP_STRUCT__entry(
>> +		__field(dev_t,	dev)
>> +		__field(ino_t,	ino)
>> +		__field(sector_t,	block)
>> +		__field(int,	ret)
>> +	),
>> +
>> +	TP_fast_assign(
>> +		__entry->dev	= inode->i_sb->s_dev;
>> +		__entry->ino	= inode->i_ino;
>> +		__entry->block	= block;
>> +		__entry->ret	= ret;
>> +	),
>> +
>> +	TP_printk("dev %d,%d ino %lu block number %llu error %d",
>> +		  MAJOR(__entry->dev), MINOR(__entry->dev),
>> +		  (unsigned long) __entry->ino, (unsigned long long) __entry->block,
>> __entry->ret)
>> +);
>> +
>> +DEFINE_EVENT(f2fs_data_block, f2fs_get_data_block_enter,
>> +	TP_PROTO(struct inode *inode, sector_t block, int ret),
>> +	TP_ARGS(inode, block, ret)
>> +);
>> +
>> +DEFINE_EVENT(f2fs_data_block, f2fs_get_data_block_exit,
>> +	TP_PROTO(struct inode *inode, sector_t block, int ret),
>> +	TP_ARGS(inode, block, ret)
>> +);
>> +
>>  #endif /* _TRACE_F2FS_H */
>>
>>   /* This part must be outside protection */
>
>
>