[v20] Implement copy offload support

[PATCH v20 05/12] fs/read_write: Enable copy_file_range for block device.

Posted by Nitesh Shetty 1 year, 3 months ago

From: Anuj Gupta <anuj20.g@samsung.com>

This is a prep patch. Allow copy_file_range to work for block devices.
Relaxing generic_copy_file_checks allows us to reuse the existing infra,
instead of adding a new user interface for block copy offload.
Change generic_copy_file_checks to use ->f_mapping->host for both inode_in
and inode_out. Allow block device in generic_file_rw_checks.

Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
---
 fs/read_write.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/fs/read_write.c b/fs/read_write.c
index ef6339391351..31645ca5ed58 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1413,8 +1413,8 @@ static int generic_copy_file_checks(struct file *file_in, loff_t pos_in,
 				    struct file *file_out, loff_t pos_out,
 				    size_t *req_count, unsigned int flags)
 {
-	struct inode *inode_in = file_inode(file_in);
-	struct inode *inode_out = file_inode(file_out);
+	struct inode *inode_in = file_in->f_mapping->host;
+	struct inode *inode_out = file_out->f_mapping->host;
 	uint64_t count = *req_count;
 	loff_t size_in;
 	int ret;
@@ -1726,7 +1726,9 @@ int generic_file_rw_checks(struct file *file_in, struct file *file_out)
 	/* Don't copy dirs, pipes, sockets... */
 	if (S_ISDIR(inode_in->i_mode) || S_ISDIR(inode_out->i_mode))
 		return -EISDIR;
-	if (!S_ISREG(inode_in->i_mode) || !S_ISREG(inode_out->i_mode))
+	if (!S_ISREG(inode_in->i_mode) && !S_ISBLK(inode_in->i_mode))
+		return -EINVAL;
+	if ((inode_in->i_mode & S_IFMT) != (inode_out->i_mode & S_IFMT))
 		return -EINVAL;
 
 	if (!(file_in->f_mode & FMODE_READ) ||
-- 
2.17.1

Re: [PATCH v20 05/12] fs/read_write: Enable copy_file_range for block device.

Posted by Dave Chinner 1 year, 3 months ago

On Mon, May 20, 2024 at 03:50:18PM +0530, Nitesh Shetty wrote:
> From: Anuj Gupta <anuj20.g@samsung.com>
> 
> This is a prep patch. Allow copy_file_range to work for block devices.
> Relaxing generic_copy_file_checks allows us to reuse the existing infra,
> instead of adding a new user interface for block copy offload.
> Change generic_copy_file_checks to use ->f_mapping->host for both inode_in
> and inode_out. Allow block device in generic_file_rw_checks.
> 
> Reviewed-by: Hannes Reinecke <hare@suse.de>
> Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
> Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
> ---
>  fs/read_write.c | 8 +++++---
>  1 file changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/read_write.c b/fs/read_write.c
> index ef6339391351..31645ca5ed58 100644
> --- a/fs/read_write.c
> +++ b/fs/read_write.c
> @@ -1413,8 +1413,8 @@ static int generic_copy_file_checks(struct file *file_in, loff_t pos_in,
>  				    struct file *file_out, loff_t pos_out,
>  				    size_t *req_count, unsigned int flags)
>  {
> -	struct inode *inode_in = file_inode(file_in);
> -	struct inode *inode_out = file_inode(file_out);
> +	struct inode *inode_in = file_in->f_mapping->host;
> +	struct inode *inode_out = file_out->f_mapping->host;
>  	uint64_t count = *req_count;
>  	loff_t size_in;
>  	int ret;

Ok, so this changes from file->f_inode to file->mapping->host. No
doubt this is because of how bdev inode mappings are munged.
However, the first code that is run here is:

	ret = generic_file_rw_checks(file_in, file_out);

and that function still uses file_inode().

Hence there checks:

> @@ -1726,7 +1726,9 @@ int generic_file_rw_checks(struct file *file_in, struct file *file_out)
>  	/* Don't copy dirs, pipes, sockets... */
>  	if (S_ISDIR(inode_in->i_mode) || S_ISDIR(inode_out->i_mode))
>  		return -EISDIR;
> -	if (!S_ISREG(inode_in->i_mode) || !S_ISREG(inode_out->i_mode))
> +	if (!S_ISREG(inode_in->i_mode) && !S_ISBLK(inode_in->i_mode))
> +		return -EINVAL;
> +	if ((inode_in->i_mode & S_IFMT) != (inode_out->i_mode & S_IFMT))
>  		return -EINVAL;

.... are being done on different inodes to the rest of
generic_copy_file_checks() when block devices are used.

Is this correct? If so, this needs a pair of comments (one for each
function) to explain why the specific inode used for these functions
is correct for block devices....

-Dave.
-- 
Dave Chinner
david@fromorbit.com

Re: [PATCH v20 05/12] fs/read_write: Enable copy_file_range for block device.

Posted by Nitesh Shetty 1 year, 3 months ago

On 26/05/24 09:02AM, Dave Chinner wrote:
>On Mon, May 20, 2024 at 03:50:18PM +0530, Nitesh Shetty wrote:
>> From: Anuj Gupta <anuj20.g@samsung.com>
>>
>> This is a prep patch. Allow copy_file_range to work for block devices.
>> Relaxing generic_copy_file_checks allows us to reuse the existing infra,
>> instead of adding a new user interface for block copy offload.
>> Change generic_copy_file_checks to use ->f_mapping->host for both inode_in
>> and inode_out. Allow block device in generic_file_rw_checks.
>>
>> Reviewed-by: Hannes Reinecke <hare@suse.de>
>> Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
>> Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
>> ---
>>  fs/read_write.c | 8 +++++---
>>  1 file changed, 5 insertions(+), 3 deletions(-)
>>
>> diff --git a/fs/read_write.c b/fs/read_write.c
>> index ef6339391351..31645ca5ed58 100644
>> --- a/fs/read_write.c
>> +++ b/fs/read_write.c
>> @@ -1413,8 +1413,8 @@ static int generic_copy_file_checks(struct file *file_in, loff_t pos_in,
>>  				    struct file *file_out, loff_t pos_out,
>>  				    size_t *req_count, unsigned int flags)
>>  {
>> -	struct inode *inode_in = file_inode(file_in);
>> -	struct inode *inode_out = file_inode(file_out);
>> +	struct inode *inode_in = file_in->f_mapping->host;
>> +	struct inode *inode_out = file_out->f_mapping->host;
>>  	uint64_t count = *req_count;
>>  	loff_t size_in;
>>  	int ret;
>
>Ok, so this changes from file->f_inode to file->mapping->host. No
>doubt this is because of how bdev inode mappings are munged.
>However, the first code that is run here is:
>
>	ret = generic_file_rw_checks(file_in, file_out);
>
>and that function still uses file_inode().
>
>Hence there checks:
>
>> @@ -1726,7 +1726,9 @@ int generic_file_rw_checks(struct file *file_in, struct file *file_out)
>>  	/* Don't copy dirs, pipes, sockets... */
>>  	if (S_ISDIR(inode_in->i_mode) || S_ISDIR(inode_out->i_mode))
>>  		return -EISDIR;
>> -	if (!S_ISREG(inode_in->i_mode) || !S_ISREG(inode_out->i_mode))
>> +	if (!S_ISREG(inode_in->i_mode) && !S_ISBLK(inode_in->i_mode))
>> +		return -EINVAL;
>> +	if ((inode_in->i_mode & S_IFMT) != (inode_out->i_mode & S_IFMT))
>>  		return -EINVAL;
>
>.... are being done on different inodes to the rest of
>generic_copy_file_checks() when block devices are used.
>
>Is this correct? If so, this needs a pair of comments (one for each
>function) to explain why the specific inode used for these functions
>is correct for block devices....
>
We were getting wrong size with file_inode() for block device, but we
missed to do it here in generic_file_rw_checks.
We will change the generic_file_rw_checks to use file->mapping->host
to make it consistent.

Thank You,
Nitesh Shetty

Re: [PATCH v20 05/12] fs/read_write: Enable copy_file_range for block device.

Posted by Hannes Reinecke 1 year, 3 months ago

On 5/20/24 12:20, Nitesh Shetty wrote:
> From: Anuj Gupta <anuj20.g@samsung.com>
> 
> This is a prep patch. Allow copy_file_range to work for block devices.
> Relaxing generic_copy_file_checks allows us to reuse the existing infra,
> instead of adding a new user interface for block copy offload.
> Change generic_copy_file_checks to use ->f_mapping->host for both inode_in
> and inode_out. Allow block device in generic_file_rw_checks.
> 
> Reviewed-by: Hannes Reinecke <hare@suse.de>
> Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
> Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
> ---
>   fs/read_write.c | 8 +++++---
>   1 file changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/read_write.c b/fs/read_write.c
> index ef6339391351..31645ca5ed58 100644
> --- a/fs/read_write.c
> +++ b/fs/read_write.c
> @@ -1413,8 +1413,8 @@ static int generic_copy_file_checks(struct file *file_in, loff_t pos_in,
>   				    struct file *file_out, loff_t pos_out,
>   				    size_t *req_count, unsigned int flags)
>   {
> -	struct inode *inode_in = file_inode(file_in);
> -	struct inode *inode_out = file_inode(file_out);
> +	struct inode *inode_in = file_in->f_mapping->host;
> +	struct inode *inode_out = file_out->f_mapping->host;
>   	uint64_t count = *req_count;
>   	loff_t size_in;
>   	int ret;
> @@ -1726,7 +1726,9 @@ int generic_file_rw_checks(struct file *file_in, struct file *file_out)
>   	/* Don't copy dirs, pipes, sockets... */
>   	if (S_ISDIR(inode_in->i_mode) || S_ISDIR(inode_out->i_mode))
>   		return -EISDIR;
> -	if (!S_ISREG(inode_in->i_mode) || !S_ISREG(inode_out->i_mode))
> +	if (!S_ISREG(inode_in->i_mode) && !S_ISBLK(inode_in->i_mode))
> +		return -EINVAL;
> +	if ((inode_in->i_mode & S_IFMT) != (inode_out->i_mode & S_IFMT))
>   		return -EINVAL;
>   
>   	if (!(file_in->f_mode & FMODE_READ) ||

Reviewed-by: Hannes Reinecke <hare@suse.de>

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                  Kernel Storage Architect
hare@suse.de                                +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich