From mboxrd@z Thu Jan  1 00:00:00 1970
From: Bob Peterson <rpeterso@redhat.com>
Date: Fri, 7 Sep 2018 08:46:19 -0400 (EDT)
Subject: [Cluster-devel] [GFS2 PATCH 4/4] gfs2: read journal in large
 chunks	to locate the head
In-Reply-To: <1536253357-8967-5-git-send-email-adas@redhat.com>
References: <1536253357-8967-1-git-send-email-adas@redhat.com>
	<1536253357-8967-5-git-send-email-adas@redhat.com>
Message-ID: <1446531282.11122692.1536324379848.JavaMail.zimbra@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 Abhi,

Overall, I like the patch set. What Andreas said still applies, plus:
----- Original Message -----
> Use bio(s) to read in the journal sequentially in large chunks and
> locate the head of the journal.
> This is faster in most cases when compared to the existing bisect
> method which operates one block at a time.
> 
> Signed-off-by: Abhi Das <adas@redhat.com>
> ---
(snip)
> +	for (i=0; i<blocks; i++) {

Just a nit: Can you put some spaces in here like this?
for (i = 0; i < blocks; i++) {

(snip)

> +	if (jd->jd_jhead.lh_sequence == 0)
> +		error = 1;

Since gfs2_find_jhead is called by remount, which is called from vfs,
I think we should probably use a return code more meaningful than 1.
I'm not sure what we ought to use in this case. Perhaps -EPERM or -EACCES ?

Regards,

Bob Peterson
Red Hat File Systems