linux-ext4 - Re: [PATCH] e2fsck: skip extent optimization by default

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: <3B21D37E-AA2D-4D54-B4C7-8D094FFB766D@dilger.ca>
Date:   Thu, 1 Oct 2020 20:11:06 -0600
From:   Andreas Dilger <adilger@...ger.ca>
To:     "Theodore Y. Ts'o" <tytso@....EDU>
Cc:     Ext4 Developers List <linux-ext4@...r.kernel.org>
Subject: Re: [PATCH] e2fsck: skip extent optimization by default

On Oct 1, 2020, at 12:03 PM, Theodore Y. Ts'o <tytso@....EDU> wrote:
> On Mon, Sep 21, 2020 at 04:16:02PM -0600, adilger@...mcloud.com wrote:
>> From: Andreas Dilger <adilger@...mcloud.com>
>> 
>> The e2fsck error message:
>> 
>>    inode nnn extent tree (at level 1) could be narrower. Optimize<y>?
>> 
>> can be fairly verbose at times, and leads users to think that there
>> may be something wrong with the filesystem.  Basically, almost any
>> message printed by e2fsck makes users nervous when they are facing
>> other corruption, and a few thousand of these printed may hide other
>> errors.  It also isn't clear that saving a few blocks optimizing the
>> extent tree noticeably improves performance.
>> 
>> This message has previously been annoying enough for Ted to add the
>> "-E no_optimize_extents" option to disable it.  Just enable this
>> option by default, similar to the "-D" directory optimization option.
>> 
>> Signed-off-by: Andreas Dilger <adilger@...ger.ca>
> 
> Applying this patch causes a whole bunch of tests fail:
> 
> 348 tests succeeded 9 tests failed
> Tests failed: d_punch_bigalloc d_punch f_collapse_extent_tree
>      f_compress_extent_tree_level f_extent_bad_node f_extent_int_bad_magic
>      f_extent_leaf_bad_magic f_extent_oobounds f_quota_extent_opt

Sorry about that, I usually *do* run the tests after every patch, I'm not
sure why I didn't for this patch.

>> @@ -1051,6 +1053,11 @@ static errcode_t PRS(int argc, char *argv[], e2fsck_t *ret_ctx)
>> 	if (c)
>> 		ctx->options |= E2F_OPT_NOOPT_EXTENTS;
>> 
>> +	profile_get_boolean(ctx->profile, "options", "optimize_extents",
>> +			    0, 0, &c);
>> +	if (c)
>> +		ctx->options &= ~E2F_OPT_NOOPT_EXTENTS;
>> +
> 
> We already have a no_optimize_extents option supported in e2fsck.conf.
> So if we want to change the default, a simpler way to do this might be
> to edit e2fsck.conf.5.in to simply add "no_optimize_extents=true" to
> the default version of e2fsck.conf defined by default.

Does that mean you *don't* want a refresh of this patch that fixes the
test cases?  Lukas had also been discussing how to change e2fsck so it
still fixed the inodes, but didn't print a message for each one, though
it wasn't clear to me that there is much benefit to this at all.

> As a reminder, for future changes, when we add a new tunable to
> e2fsck.conf or mke2fs.conf, the man page should be edited.

Yes, I did edit the e2fsck.8.in man page to describe the change in
default behavior.

Cheers, Andreas






Download attachment "signature.asc" of type "application/pgp-signature" (874 bytes)