linux-ext4.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Theodore Ts'o <tytso@mit.edu>
To: Daniel Axtens <dja@axtens.net>
Cc: linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org,
	viro@zeniv.linux.org.uk, miklos@szeredi.hu,
	linux-unionfs@vger.kernel.org
Subject: Re: ext4_file_open: Inconsistent encryption contexts (commit ff978b09f973) breaking Docker
Date: Thu, 10 Mar 2016 21:15:06 -0500	[thread overview]
Message-ID: <20160311021506.GA32214@thunk.org> (raw)
In-Reply-To: <87io0t3ks9.fsf@gamma.ozlabs.ibm.com>

On Fri, Mar 11, 2016 at 11:44:54AM +1100, Daniel Axtens wrote:
> Hi,
> 
> Trying to run a Docker container on a mainline kernel is failing
> intermittently, in interesting and exciting ways, such as:
> 
> $ docker run -it --rm --env PACKAGE=sinatra npmtest
> operation not permitted
> docker: Error response from daemon: Cannot start container 4fc0120a6389f25241f84527a0d31854806f6fe4fd98d019f790cea0ae7e230b: [10] System error: operation not permitted.
> 
> EXT4-fs warning (device sda2): ext4_file_open:402: Inconsistent encryption contexts: 27842/3691208

This could only happen if the EXT4_ENCRYPT_FL flag is set.  (I assume
you weren't actually trying to use ext4 encryption.)  The flag can't
be set using the FS_IOC_SETFLAGS ioctl.  It can only be set using
EXT4_IOC_SET_ENCRYPTION_POLICY.

The only thing I can think of is that overlayfs is somehow setting or
otherwise corrupting the i_flags.

So if you could try changing the relevant code in ext4_file_open() to
look like this:

	if (ext4_encrypted_inode(dir) &&
	    !ext4_is_child_context_consistent_with_parent(dir, inode)) {
		char tmpbuf[128];

		ext4_warning(inode->i_sb,
			     "Inconsistent encryption contexts: %lu/%lu\n",
			     (unsigned long) dir->i_ino,
			     (unsigned long) inode->i_ino);
		cp = d_path(&filp->f_path, tmpbuf, sizeof(tmpbuf));
		if (!IS_ERR(cp))
			pr_err("pathname: %s\n", cp);
		pr_err("inode flags: %lu\n", EXT4_I(inode)->i_flags);
		WARN_ON(1);
		return -EPERM;
	}

This should give you the stack trace as well as the pathname and inode
flag information.  This will hopefully give us a bit more context
about how this is happening.

						- Ted

  reply	other threads:[~2016-03-11  2:15 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-11  0:44 ext4_file_open: Inconsistent encryption contexts (commit ff978b09f973) breaking Docker Daniel Axtens
2016-03-11  2:15 ` Theodore Ts'o [this message]
2016-03-11 15:34   ` Miklos Szeredi
2016-03-11 23:32     ` Daniel Axtens
2016-03-14  6:47     ` Daniel Axtens
2016-03-14 10:27       ` Miklos Szeredi
2016-03-14 22:49         ` Daniel Axtens
2016-03-31 20:39         ` Marc Haber

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160311021506.GA32214@thunk.org \
    --to=tytso@mit.edu \
    --cc=dja@axtens.net \
    --cc=linux-ext4@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-unionfs@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=viro@zeniv.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).