linux-ext4 - Re: [PATCH] ext4: unconditionally enable the i

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] [day] [month] [year] [list]
Date:   Tue, 26 Jul 2022 07:15:32 -0400
From:   Jeff Layton <jlayton@...nel.org>
To:     Lukas Czerner <lczerner@...hat.com>
Cc:     tytso@....edu, adilger.kernel@...ger.ca,
        linux-ext4@...r.kernel.org, Dave Chinner <david@...morbit.com>,
        Benjamin Coddington <bcodding@...hat.com>,
        Christoph Hellwig <hch@...radead.org>, kzak@...hat.com
Subject: Re: [PATCH] ext4: unconditionally enable the i_version counter

On Tue, 2022-07-26 at 09:10 +0200, Lukas Czerner wrote:
> On Mon, Jul 25, 2022 at 03:29:46PM -0400, Jeff Layton wrote:
> > The original i_version implementation was pretty expensive, requiring a
> > log flush on every change. Because of this, it was gated behind a mount
> > option (implemented via the MS_I_VERSION mountoption flag).
> > 
> > Commit ae5e165d855d (fs: new API for handling inode->i_version) made the
> > i_version flag much less expensive, so there is no longer a performance
> > penalty from enabling it.
> > 
> > Have ext4 ignore the SB_I_VERSION flag, and just enable it
> > unconditionally.
> 
> Hi Jeff,
> 
> the ext4 specific i_version option is deprecated and it's perfect time
> to remove it as well.
> 
> As for enabling iversion by default, as I said before I am fine with
> that. With this patch we're left with no means of disabling it, which
> might be ok, I don't have a strong opinion.
> 
> However I don't like the fact that the noiversion mount option is going
> to be just silently ignored and only on some file systems. The
> inconsistency bugs me and will create confusion when things go wrong.
> 
> Shouldn't we introduce something like SB_NOIVERSION to give fs a clean
> way to inform user we're ignoring it, or even change the default? Or
> perhaps 'noinversion' should be removed?
> 

Ordinarily, I'd agree here. I like leaving some mechanism to revert an
unconditional change like this.

Unfortunately, the problem here is that the original mount option was
implemented via a mountflag and was parsed by the userland helper.
Trying to rework how "-o {no}iversion" options work is going to be
messy. We can't rely on /bin/mount and the kernel being in sync.

If we did want to add something like this, then we might need to add a
new mount option that doesn't match the old one (e.g. -o no_i_version).
Alternately, could that be made settable via mkfs/tune2fs?

> 
> > 
> > Cc: Dave Chinner <david@...morbit.com>
> > Cc: Lukas Czerner <lczerner@...hat.com>
> > Cc: Benjamin Coddington <bcodding@...hat.com>
> > Cc: Christoph Hellwig <hch@...radead.org>
> > Signed-off-by: Jeff Layton <jlayton@...nel.org>
> > ---
> >  fs/ext4/inode.c | 5 ++---
> >  fs/ext4/super.c | 8 ++++----
> >  2 files changed, 6 insertions(+), 7 deletions(-)
> > 
> > diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> > index 84c0eb55071d..c785c0b72116 100644
> > --- a/fs/ext4/inode.c
> > +++ b/fs/ext4/inode.c
> > @@ -5411,7 +5411,7 @@ int ext4_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
> >  			return -EINVAL;
> >  		}
> >  
> > -		if (IS_I_VERSION(inode) && attr->ia_size != inode->i_size)
> > +		if (attr->ia_size != inode->i_size)
> >  			inode_inc_iversion(inode);
> >  
> >  		if (shrink) {
> > @@ -5717,8 +5717,7 @@ int ext4_mark_iloc_dirty(handle_t *handle,
> >  	}
> >  	ext4_fc_track_inode(handle, inode);
> >  
> > -	if (IS_I_VERSION(inode))
> > -		inode_inc_iversion(inode);
> > +	inode_inc_iversion(inode);
> >  
> >  	/* the do_update_inode consumes one bh->b_count */
> >  	get_bh(iloc->bh);
> > diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> > index 845f2f8aee5f..30645d4343b6 100644
> > --- a/fs/ext4/super.c
> > +++ b/fs/ext4/super.c
> > @@ -2142,8 +2142,7 @@ static int ext4_parse_param(struct fs_context *fc, struct fs_parameter *param)
> >  		return 0;
> >  	case Opt_i_version:
> >  		ext4_msg(NULL, KERN_WARNING, deprecated_msg, param->key, "5.20");
> > -		ext4_msg(NULL, KERN_WARNING, "Use iversion instead\n");
> > -		ctx_set_flags(ctx, SB_I_VERSION);
> > +		ext4_msg(NULL, KERN_WARNING, "i_version counter is always enabled.\n");
> >  		return 0;
> >  	case Opt_inlinecrypt:
> >  #ifdef CONFIG_FS_ENCRYPTION_INLINE_CRYPT
> > @@ -2970,8 +2969,6 @@ static int _ext4_show_options(struct seq_file *seq, struct super_block *sb,
> >  		SEQ_OPTS_PRINT("min_batch_time=%u", sbi->s_min_batch_time);
> >  	if (nodefs || sbi->s_max_batch_time != EXT4_DEF_MAX_BATCH_TIME)
> >  		SEQ_OPTS_PRINT("max_batch_time=%u", sbi->s_max_batch_time);
> > -	if (sb->s_flags & SB_I_VERSION)
> > -		SEQ_OPTS_PUTS("i_version");
> >  	if (nodefs || sbi->s_stripe)
> >  		SEQ_OPTS_PRINT("stripe=%lu", sbi->s_stripe);
> >  	if (nodefs || EXT4_MOUNT_DATA_FLAGS &
> > @@ -4630,6 +4627,9 @@ static int __ext4_fill_super(struct fs_context *fc, struct super_block *sb)
> >  	sb->s_flags = (sb->s_flags & ~SB_POSIXACL) |
> >  		(test_opt(sb, POSIX_ACL) ? SB_POSIXACL : 0);
> >  
> > +	/* i_version is always enabled now */
> > +	sb->s_flags |= SB_I_VERSION;
> > +
> >  	if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV &&
> >  	    (ext4_has_compat_features(sb) ||
> >  	     ext4_has_ro_compat_features(sb) ||
> > -- 
> > 2.37.1
> > 
> 

-- 
Jeff Layton <jlayton@...nel.org>