[PATCH] 2.6.15-mm3 - make useless quota error message informative

[PATCH] 2.6.15-mm3 - make useless quota error message informative

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 972 bytes --]

fs/quota_v2.c can, under some conditions, issue a kernel message that
says, in totality, 'failed read'.  This patch does the following:

1) Gives a hint who issued the error message, so people reading the logs
don't have to go grepping the entire kernel tree (with 11 false positives).

2) Say what amount of data we expected, and actually got.

Signed-off-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
---
--- linux-2.6.15-mm3/fs/quota_v2.c.quot-bug	2006-01-02 22:21:10.000000000 -0500
+++ linux-2.6.15-mm3/fs/quota_v2.c	2006-01-12 15:26:43.000000000 -0500
@@ -35,7 +35,8 @@ static int v2_check_quota_file(struct su
  
 	size = sb->s_op->quota_read(sb, type, (char *)&dqhead, sizeof(struct v2_disk_dqheader), 0);
 	if (size != sizeof(struct v2_disk_dqheader)) {
-		printk("failed read\n");
+		printk("quota_v2: failed read expected=%d got=%d\n",
+			sizeof(struct v2_disk_dqheader), size);
 		return 0;
 	}
 	if (le32_to_cpu(dqhead.dqh_magic) != quota_magics[type] ||
 

[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: [PATCH] 2.6.15-mm3 - make useless quota error message informative

[Jan Kara]

> fs/quota_v2.c can, under some conditions, issue a kernel message that
> says, in totality, 'failed read'.  This patch does the following:
> 
> 1) Gives a hint who issued the error message, so people reading the logs
> don't have to go grepping the entire kernel tree (with 11 false positives).
> 
> 2) Say what amount of data we expected, and actually got.
> 
> Signed-off-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
> ---
> --- linux-2.6.15-mm3/fs/quota_v2.c.quot-bug	2006-01-02 22:21:10.000000000 -0500
> +++ linux-2.6.15-mm3/fs/quota_v2.c	2006-01-12 15:26:43.000000000 -0500
> @@ -35,7 +35,8 @@ static int v2_check_quota_file(struct su
>   
>  	size = sb->s_op->quota_read(sb, type, (char *)&dqhead, sizeof(struct v2_disk_dqheader), 0);
>  	if (size != sizeof(struct v2_disk_dqheader)) {
> -		printk("failed read\n");
> +		printk("quota_v2: failed read expected=%d got=%d\n",
> +			sizeof(struct v2_disk_dqheader), size);
>  		return 0;
>  	}
>  	if (le32_to_cpu(dqhead.dqh_magic) != quota_magics[type] ||
>  

  The patch is fine. Thanks for fixing. I wonder how I managed to write such
error message ;).

									Honza

-- 
Jan Kara <jack@suse.cz>
SuSE CR Labs