[PATCH] scsi: scsi_devinfo: return an error when the proc write fails

[PATCH] scsi: scsi_devinfo: return an error when the proc write fails

[shao.mingyin]

From: Shao Mingyin <shao.mingyin@zte.com.cn>

proc_scsi_devinfo_write() calls scsi_dev_info_list_add_str() but ignores
its return value, so a write to /proc/scsi/device_info always reports
success, even when the string is malformed and no entry was added.  Users
and scripts then believe that the quirk is in effect.

scsi_init_devinfo() already checks the return value of the same function
and fails the initialization.  Return the parser error, but keep returning
the number of bytes written on success: a proc write handler that returns
zero for a non-empty write would look like a short write to the caller.

Signed-off-by: Shao Mingyin <shao.mingyin@zte.com.cn>
---
diff --git a/drivers/scsi/scsi_devinfo.c b/drivers/scsi/scsi_devinfo.c
index cfc259fd0141..b1f3bcd739b4 100644
--- a/drivers/scsi/scsi_devinfo.c
+++ b/drivers/scsi/scsi_devinfo.c
@@ -688,6 +688,7 @@ static ssize_t proc_scsi_devinfo_write(struct file *file,
 {
 	char *buffer;
 	ssize_t err = length;
+	int ret;

 	if (!buf || length>PAGE_SIZE)
 		return -EINVAL;
@@ -706,7 +707,9 @@ static ssize_t proc_scsi_devinfo_write(struct file *file,
 		goto out;
 	}

-	scsi_dev_info_list_add_str(buffer);
+	ret = scsi_dev_info_list_add_str(buffer);
+	if (ret)
+		err = ret;

 out:
 	kfree(buffer);

Re: [PATCH] scsi: scsi_devinfo: return an error when the proc write fails

[Laurence Oberman]

On Sat, 2026-10-31 at 22:21 +0800, shao.mingyin@zte.com.cn wrote:
> From: Shao Mingyin <shao.mingyin@zte.com.cn>
> 
> proc_scsi_devinfo_write() calls scsi_dev_info_list_add_str() but
> ignores
> its return value, so a write to /proc/scsi/device_info always reports
> success, even when the string is malformed and no entry was added. 
> Users
> and scripts then believe that the quirk is in effect.
> 
> scsi_init_devinfo() already checks the return value of the same
> function
> and fails the initialization.  Return the parser error, but keep
> returning
> the number of bytes written on success: a proc write handler that
> returns
> zero for a non-empty write would look like a short write to the
> caller.
> 
> Signed-off-by: Shao Mingyin <shao.mingyin@zte.com.cn>
> ---
> diff --git a/drivers/scsi/scsi_devinfo.c
> b/drivers/scsi/scsi_devinfo.c
> index cfc259fd0141..b1f3bcd739b4 100644
> --- a/drivers/scsi/scsi_devinfo.c
> +++ b/drivers/scsi/scsi_devinfo.c
> @@ -688,6 +688,7 @@ static ssize_t proc_scsi_devinfo_write(struct
> file *file,
>  {
>  	char *buffer;
>  	ssize_t err = length;
> +	int ret;
> 
>  	if (!buf || length>PAGE_SIZE)
>  		return -EINVAL;
> @@ -706,7 +707,9 @@ static ssize_t proc_scsi_devinfo_write(struct
> file *file,
>  		goto out;
>  	}
> 
> -	scsi_dev_info_list_add_str(buffer);
> +	ret = scsi_dev_info_list_add_str(buffer);
> +	if (ret)
> +		err = ret;
> 
>  out:
>  	kfree(buffer);

Looks good to me.

Note that a multi-entry write can still partially apply (entries before
the bad one are added, then -EINVAL is returned), but reporting the
error is a clear improvement over silent success.

Reviewed-by: Laurence Oberman <loberman@redhat.com>