All of lore.kernel.org
 help / color / mirror / Atom feed
From: "William H. Taber" <wtaber@us.ibm.com>
To: Ian Kent <raven@themaw.net>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>,
	Jeff Moyer <jmoyer@redhat.com>, Ram Pai <linuxram@us.ibm.com>,
	autofs mailing list <autofs@linux.kernel.org>,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>
Subject: Re: [autofs] [RFC PATCH]autofs4: hang and proposed fix
Date: Thu, 01 Dec 2005 11:30:19 -0500	[thread overview]
Message-ID: <438F251B.7060602@us.ibm.com> (raw)
In-Reply-To: <Pine.LNX.4.63.0512011917010.3189@donald.themaw.net>

Ian Kent wrote:
> On Wed, 30 Nov 2005, William H. Taber wrote:
> 
> 
>>Trond Myklebust wrote:
>>
>>>On Wed, 2005-11-30 at 15:32 -0500, William H. Taber wrote:
>>>
>>>
>>>
>>>>Not only is there this case, but the original premise is wrong as well.
>>>>There is a second case in which a d_revalidate function is called with the
>>>>parent i_sem and that is when it is called from inside of lookup_one_len.
>>>>What makes this tricky is that lookup_one_len is called from
>>>>nfs_sillyrename from inside of nfs_rename which is called, naturally
>>>>enough by sys_rename.  The rename code is very careful about the order in
>>>>which it obtains the parent semaphores because it needs to get two of
>>>>them.  It must always obtain the locks in the same order so that does not
>>>>get into a deadly embrace.  If we start arbitrarily releasing a parent
>>>>semaphore in cached_lookup and taking it again after the revalidate, we
>>>>risk breaking the lock ordering and creating a deadly embrace.
>>>>
>>>>When I started writing this I thought that it would be safe for the autofs
>>>>revalidate code to release the parent semaphore because they do not have a
>>>>rename callback.  But I looked again at the rename code and it calls
>>>>lookup_hash on the final source and destination files after locking the
>>>>parents so the potential for a deadly embrace still exists unless there is
>>>>some other assurance that these final lookups will never pend waiting on
>>>>the automounter in either their revalidate or lookup routines.  (Actually
>>>>the requirement is that they never give up the parent i_sem lock, but the
>>>>lookup code has to give up the lock so that the autofs demon can run and
>>>>perform the mount so it amounts to the same thing.)
>>>>
>>>>The same issue exists for devfs which also releases the parent i_sem lock
>>>>so that it can wait inside its revalidation routine.
>>>
>>>
>>>So exactly why does autofs4 want to hold the dir->i_sem in d_revalidate
>>>in the first place? Can't we move any code that requires dir->i_sem to
>>>be held into a ->lookup() method?
>>
>>It's not that d_revalidate wants or doesn't want to hold the lock.  The caller
>>of lookup_one_len is required to get the lock and this function calls
>>lookup_hash which calls cached_lookup which calls d_revalidate.
>>
>>
>>>Trivially, if you have a d_revalidate that does something like
>>>
>>>int autofs_revalidate(struct dentry *dentry, struct nameidata *nd)
>>>{
>>>  d_drop(dentry);
>>>  return 0;
>>>}
>>>
>>>then the VFS will currently allocate a new dentry with the same name,
>>>and call ->lookup() on it without dropping dir->i_sem. If you still need
>>>to reference the old dentry, then put it on a private list somewhere.
>>>That would also allow you to return the old dentry as the result of the
>>>->lookup() operation if that is desirable.
>>
>>Problem with that, as I understand it and Ian Kent knows better than I, is
>>that the autofs lookup code creates the dentry and fills it in partially and
>>marks it as waiting for mounting and wakes up the automount demon.  The demon
>>completes the mount and finishes filling in the dentry.  So we cannot have
>>some other lookup coming in and removing the dentry on us.  At least that is
>>what I understand from Ian's answer when I proposed the same sort of thing to
>>him.   Even if  they end up doing something like that in a future version of
>>the automounter, I would still like a simple patch that can be applied to
>>existing systems as an interim fix.
> 
> 
> Lets see if I can keep this explaination simple.
> 
> The user space process using the autofs filesystem (autodir or automount) 
> needs to be able to call mkdir at mount time as a result of a callback 
> from revalidate. Sometimes this comes indirectly from lookup (if the 
> directory does not already exist).
> 
> lookup_one_len requires the i_sem to be held so two instances of a 
> filesystem calling it lead to a deadlock when mkdir is called from 
> userspace (the third process). In the case we are discussing this happens 
> because the first process calls lookup which releases the i_sem and 
> calls revalidate itself. The second calls revalidate which doesn't release 
> the i_sem and is places on a wait queue for mount completion. Consequently 
> the mkdir blocks.
> 
> So the requirement is that autofs release the i_sem during the callback, 
> not obtain it.
> 
> Will believes that it is not safe for autofs to release i_sem for 
> the callback to user space because it is possible that path that aquired 
> it may not be the path that has called revalidate and I can see his point.
> 
> Never the less I'm still not convinced that this is possible given the 
> restrictions of autofs.
> 
> Let me try and describe this, hopefully more clearly than I've done so 
> far.
> 
> The only operations defined for autofs are:
> 
> mkdir, rmdir, symlink and unlink 
> 
> and the only processes that can do these operations must be in the same 
> process group that mounted the filesystem. EACCESS is returned for all 
> other processes attempting these operations.
> 
> The other functionality is read-only (and perhaps triggers a mount) 
> being lookup, revalidate and readdir.
> 
> So the question is, can anyone provide an example of a path that, upon 
> calling autofs revalidate or lookup with the i_sem held, not be the path 
> that aquired it?

Any other process calling lookup_one_len on a file in /net.

Will

  reply	other threads:[~2005-12-01 16:30 UTC|newest]

Thread overview: 95+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-11-16 10:17 [RFC PATCH]autofs4: hang and proposed fix Ram Pai
2005-11-16 12:41 ` [autofs] " Ian Kent
2005-11-16 16:50   ` Ram Pai
2005-11-16 22:57     ` Ian Kent
2005-11-17  1:52       ` [autofs] " Ram Pai
2005-11-17 18:50         ` Ian Kent
2005-11-17 19:19           ` William H. Taber
2005-11-17 20:39             ` Ram Pai
2005-11-17 22:31               ` William H. Taber
2005-11-18 14:57                 ` Ian Kent
2005-11-18 14:54               ` Ian Kent
2005-11-18 14:44             ` Ian Kent
2005-11-18 15:20               ` William H. Taber
2005-11-18 16:30                 ` Ian Kent
2005-11-18 17:12                   ` William H. Taber
2005-11-18 18:57                     ` Ram Pai
2005-11-18 20:08                       ` William H. Taber
2005-11-19  2:52                         ` Ian Kent
2005-11-21 16:40                           ` William H. Taber
2005-11-22 13:13                             ` Ian Kent
2005-11-22 17:48                               ` [autofs] " William H. Taber
2005-11-23 14:11                                 ` Ian Kent
2005-11-23 16:42                                   ` William H. Taber
2005-11-23 17:52                                     ` Ian Kent
2005-11-23 18:47                                       ` William H. Taber
2005-11-23 17:52                                     ` Ian Kent
2005-11-19  1:40                     ` [autofs] " Ian Kent
2005-11-16 15:22 ` Jeff Moyer
2005-11-16 15:22   ` Jeff Moyer
2005-11-16 17:00   ` [autofs] " Ram Pai
2005-11-16 18:25     ` Jeff Moyer
2005-11-16 19:24       ` William H. Taber
2005-11-16 19:51         ` Ram Pai
2005-11-27 10:47 ` Ian Kent
2005-11-28 17:19   ` William H. Taber
2005-11-28 23:12     ` Badari Pulavarty
2005-11-29 14:19       ` Ian Kent
2005-11-29 16:34         ` William H. Taber
2005-11-30 14:02           ` Ian Kent
2005-11-30 16:49             ` Badari Pulavarty
2005-11-30 17:04               ` Trond Myklebust
2005-11-30 21:10                 ` William H. Taber
2005-11-29 14:20     ` Ian Kent
2005-11-30  1:16 ` [autofs] " Jeff Moyer
2005-11-30  1:16   ` Jeff Moyer
2005-11-30  1:56   ` Trond Myklebust
2005-11-30  4:15     ` Jeff Moyer
2005-11-30  6:14       ` Trond Myklebust
2005-11-30 15:44         ` Ian Kent
2005-11-30 15:53           ` [autofs] " Trond Myklebust
2005-11-30 16:12             ` Ian Kent
2005-11-30 16:27               ` Ian Kent
2005-11-30 16:45               ` [autofs] " Trond Myklebust
2005-11-30 20:32     ` William H. Taber
2005-11-30 20:53       ` Trond Myklebust
2005-11-30 21:30         ` William H. Taber
2005-11-30 22:32           ` Trond Myklebust
2005-12-01 16:27             ` William H. Taber
2005-12-01 12:09           ` Ian Kent
2005-12-01 16:30             ` William H. Taber [this message]
2005-12-02 13:49               ` Ian Kent
2005-12-02 14:07                 ` Jeff Moyer
2005-12-02 15:21                   ` Ian Kent
2005-12-02 16:35                     ` [autofs] " Will Taber
2005-12-02 17:11                       ` Ian Kent
2005-12-02 15:34                 ` Will Taber
2005-12-02 17:29                   ` Ian Kent
2005-12-02 18:12                     ` Trond Myklebust
2005-12-04 12:56                       ` Christoph Hellwig
2005-12-04 12:57                         ` Christoph Hellwig
2005-12-04 14:58                           ` Ian Kent
2005-12-04 17:17                             ` [autofs] " Christoph Hellwig
2005-12-05 14:02                               ` Ian Kent
2005-12-06 21:20                               ` Jeff Moyer
2005-12-06 21:40                                 ` Christoph Hellwig
2005-12-06 22:37                                   ` Jeff Moyer
2005-12-07 14:52                                   ` Will Taber
2005-12-07 15:18                                     ` Christoph Hellwig
2005-12-07 15:22                                   ` Brian Long
2005-12-07 15:25                                     ` Christoph Hellwig
2005-12-07 17:46                                     ` Will Taber
2005-12-08 14:16                                       ` Ian Kent
2005-12-09 12:12                                       ` Christoph Hellwig
2005-12-09 13:33                                         ` John T. Kohl
2005-12-13 18:39                                           ` Christoph Hellwig
2005-12-04 14:56                         ` Ian Kent
2005-12-02 19:04                     ` [autofs] " Will Taber
2005-12-04  9:39                       ` Ian Kent
2005-12-02 16:04                 ` [autofs] " Jeff Moyer
2005-12-02 17:36                   ` Ian Kent
2005-12-02 18:33                     ` [autofs] " Will Taber
2005-12-04  9:52                       ` Ian Kent
2005-12-04 14:54                         ` Ian Kent
2005-12-05 15:40                           ` Ian Kent
2005-11-30 14:48   ` [autofs] " Ian Kent

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=438F251B.7060602@us.ibm.com \
    --to=wtaber@us.ibm.com \
    --cc=autofs@linux.kernel.org \
    --cc=jmoyer@redhat.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linuxram@us.ibm.com \
    --cc=raven@themaw.net \
    --cc=trond.myklebust@fys.uio.no \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.