From mboxrd@z Thu Jan  1 00:00:00 1970
From: Steven Whitehouse <swhiteho@redhat.com>
Date: Fri, 11 Jul 2014 17:07:51 +0100
Subject: [Cluster-devel] [gfs2-utils PATCH] fsck.gfs2: Issue read-ahead
 for dinodes in each bitmap
In-Reply-To: <2099118726.6310652.1405011363612.JavaMail.zimbra@redhat.com>
References: <2099118726.6310652.1405011363612.JavaMail.zimbra@redhat.com>
Message-ID: <53C00BD7.9030502@redhat.com>
List-Id: <cluster-devel.redhat.com>
To: cluster-devel.redhat.com
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

Hi,

On 10/07/14 17:56, Bob Peterson wrote:
> Hi,
>
> This patch speeds up fsck.gfs2 by issuing read-ahead for each of the
> dinodes found in each of the bitmaps, prior to processing the bitmap.
Looks good to me,

Steve.

> Regards,
>
> Bob Peterson
> Red Hat File Systems
>
> Signed-off-by: Bob Peterson <rpeterso@redhat.com>
> ---
>   gfs2/fsck/pass1.c | 4 ++++
>   1 file changed, 4 insertions(+)
>
> diff --git a/gfs2/fsck/pass1.c b/gfs2/fsck/pass1.c
> index fec2f64..90ca357 100644
> --- a/gfs2/fsck/pass1.c
> +++ b/gfs2/fsck/pass1.c
> @@ -1442,6 +1442,10 @@ static int pass1_process_bitmap(struct gfs2_sbd *sdp, struct rgrp_tree *rgd, uin
>   	struct gfs2_inode *ip;
>   	uint8_t q;
>   
> +	/* Issue read-ahead for all dinodes in this bitmap */
> +	for (i = 0; i < n; i++)
> +		posix_fadvise(sdp->device_fd, ibuf[i] * sdp->bsize, sdp->bsize,
> +			      POSIX_FADV_WILLNEED);
>   	for (i = 0; i < n; i++) {
>   		int is_inode;
>   		uint32_t check_magic;
>