linux-ext4 - Re: [PATCH] ext4: don't give the "disabling delalloc" if not explicitly specified

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4E497BCA.4030300@redhat.com>
Date:	Mon, 15 Aug 2011 15:04:26 -0500
From:	Eric Sandeen <sandeen@...hat.com>
To:	"Ted Ts'o" <tytso@....edu>
CC:	Lukas Czerner <lczerner@...hat.com>,
	Ext4 Developers List <linux-ext4@...r.kernel.org>
Subject: Re: [PATCH] ext4: don't give the "disabling delalloc" if not explicitly
 specified

On 8/15/11 12:54 PM, Ted Ts'o wrote:
> On Mon, Aug 15, 2011 at 10:59:02AM -0500, Eric Sandeen wrote:
>>
>> The giant behavior-options switch in ext4 is confusing enough; if enabling
>> one option disables another default option, I think that explicitly stating
>> it in the logs is useful.  Doing so silently just covers up the behavior.
>>
>> If users are unhappy with the message, it's probably more because of 
>> the fact of the matter, and not because of the presentation of the fact.  :)
> 
> Most users probably have no idea what "delalloc" actually means.  So
> when they get a message that saying that data=journalled has disabled
> delalloc, it could easily be seen as noise.  I was moved to do it
> because I got tired of seeing the message over, and over, and over
> again when running xfstests.
> 
> Maybe an improvement would be (1) to document what data=journal
> implies in the Documentation/filesystems/ext4.txt, (2) change the
> message to explicitly say "delayed allocation" instead of "delalloc"
> (although many people won't have any idea what "delayed allocation"
> means either), and (3) make it a printk_once thing.
> 
> I guess I don't agree with the fundamental presumption which is that
> users should be looking at the dmesg output to understand what various
> things mean, and if they didn't explicitly specify delalloc, why
> should we complain about the fact that both delalloc and data=journal
> were specified (when in fact it wasn't specified).

Well, just my $0.02, I won't fight it.

One thing I do want, though, is to be able to look at logs and know
what mode we're running in.  I guess we do print all specified options,
so those in the know, will know that delalloc is off if data=journal is on.

But I think we could use some consistency here at least:

[root@...de ~]# mount -o data=journal,dioread_nolock /dev/sdb5 /mnt/test
[root@...de ~]#  /* YAY it worked! */

[root@...de ~]# dmesg | tail
...
[269530.183245] EXT4-fs (sdb5): Ignoring delalloc option - requested data journaling mode
[269530.191170] EXT4-fs (sdb5): Ignoring dioread_nolock option - requested data journaling mode

/* boo it didn't work */

Better hit that one too, I guess, and any other option which disables
another default ...  (honestly, I think it'd be better to fail the mount
in cases like this if conflicting options are specified).

-Eric


-Eric
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html