[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180711174153.GB9955@thunk.org>
Date: Wed, 11 Jul 2018 13:41:53 -0400
From: "Theodore Y. Ts'o" <tytso@....edu>
To: Gabriel Krisman Bertazi <krisman@...labora.co.uk>
Cc: linux-ext4@...r.kernel.org, darrick.wong@...cle.com,
kernel@...labora.com
Subject: Re: [PATCH 20/20] ext4: Implement encoding-aware dcache hooks
On Tue, Jul 03, 2018 at 01:07:00PM -0400, Gabriel Krisman Bertazi wrote:
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 53db9b6c7e33..f292cc5bacda 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -4358,6 +4358,10 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
> iput(root);
> goto failed_mount4;
> }
> +
> + if (sbi->encoding)
> + sb->s_d_op = &ext4_dentry_ops;
> +
This is going to be potentially problematic for fscrypt (e.g., when
ext4's encryption is enabled, as will be the case in Android's File
Based Encryption). See the call to d_set_d_op() in
__fscrypt_prepare_lookup in fs/crypto/hooks.c, since d_set_d_op is
going to overwrite sb->s_d_op.
This probably means that the fscrypt code is going to have to be
case-sensitivity aware. Or we would have to make case folding and fs
encryption to be mutually exclusive, which would be unfortunate, since
Android is going to be a potential user of case folding.
- Ted
Powered by blists - more mailing lists