When the autofs protocol version 5 packet type was added in commit
5c0a32fc2c ("autofs4: add new packet type for v5 communications"), it
obvously tried quite hard to be word-size agnostic, and uses explicitly
sized fields that are all correctly aligned.
However, with the final "char name[NAME_MAX+1]" array at the end, the
actual size of the structure ends up being not very well defined:
because the struct isn't marked 'packed', doing a "sizeof()" on it will
align the size of the struct up to the biggest alignment of the members
it has.
And despite all the members being the same, the alignment of them is
different: a "__u64" has 4-byte alignment on x86-32, but native 8-byte
alignment on x86-64. And while 'NAME_MAX+1' ends up being a nice round
number (256), the name[] array starts out a 4-byte aligned.
End result: the "packed" size of the structure is 300 bytes: 4-byte, but
not 8-byte aligned.
As a result, despite all the fields being in the same place on all
architectures, sizeof() will round up that size to 304 bytes on
architectures that have 8-byte alignment for u64.
Note that this is *not* a problem for 32-bit compat mode on POWER, since
there __u64 is 8-byte aligned even in 32-bit mode. But on x86, 32-bit
and 64-bit alignment is different for 64-bit entities, and as a result
the structure that has exactly the same layout has different sizes.
So on x86-64, but no other architecture, we will just subtract 4 from
the size of the structure when running in a compat task. That way we
will write the properly sized packet that user mode expects.
Not pretty. Sadly, this very subtle, and unnecessary, size difference
has been encoded in user space that wants to read packets of *exactly*
the right size, and will refuse to touch anything else.
Reported-and-tested-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When recursing down the locks when traversing a tree/list in
get_next_positive_dentry() or get_next_positive_subdir() a lock can
change from being nested to being a parent which breaks lockdep. This
patch tells lockdep about what we did.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
I don't know how I missed this obvious mistake when I
reviewed Als' patches, sorry.
[ Quoting Al:
Grr... Note to self: do git status *and* git stash show -p
before git push. Nothing like "WTF? I'd fixed that braino"
feeling ;-/
Al sent the same patch - it got broken in commit d668dc5663:
"autofs4: deal with autofs4_write/autofs4_write races". ]
Reported-and-tested-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Just serialize the actual writing of packets into pipe on
a new mutex, independent from everything else in the locking
hierarchy. As soon as something has started feeding a piece
of packet into the pipe to daemon, we *want* everything else
about to try the same to wait until we are done.
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
we need to hold ->wq_mutex while we are forming the packet to send,
lest we have autofs4_catatonic_mode() setting wq->name.name to NULL
just as autofs4_notify_daemon() decides to memcpy() from it...
We do have check for catatonic mode immediately after that (under
->wq_mutex, as it ought to be) and packet won't be actually sent,
but it'll be too late for us if we oops on that memcpy() from NULL...
Fix is obvious - just extend the area covered by ->wq_mutex over
that switch and check whether it's catatonic *before* doing anything
else.
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
We need to recheck ->catatonic after autofs4_wait() got ->wq_mutex
for good, or we might end up with wq inserted into queue after
autofs4_catatonic_mode() had done its thing. It will stick there
forever, since there won't be anything to clear its ->name.name.
A bit of a complication: validate_request() drops and regains ->wq_mutex.
It actually ends up the most convenient place to stick the check into...
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
vfs_mkdir() gets int, but immediately drops everything that might not
fit into umode_t and that's the only caller of ->mkdir()...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Replace remaining direct i_nlink updates with a new set_nlink()
updater function.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Toshiyuki Okajima <toshi.okajima@jp.fujitsu.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The previous comit made the autofs4 debug printouts check types against
the printout format, and uncovered this bug:
fs/autofs4/waitq.c:106:2: warning: format ‘%08lx’ expects type ‘long unsigned int’, but argument 4 has type ‘autofs_wqt_t’
which is due to the insane type for wait_queue_token. That thing should
be some fixed well-defined size (preferably just 'unsigned int' or
'u32') but for unexplained reasons it is randomly either 'unsigned long'
or 'unsigned int' depending on the architecture.
For now, cast it to 'unsigned long' for printing, the way we do
elsewhere. Somebody else can try to explain the typedef mess.
(There's a reason we don't support excessive use of typedefs in the
kernel: it's usually just a good way of confusing yourself).
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use 'pr_debug()' for DPRINTK, which will do the proper type checking on
the arguments (without generating code) even when DEBUG isn't #defined.
Also, use the standard __VA_ARGS__ for the macros, and stop the
pointless abuse of 'do { xyz } while (0)' when the macro is already a
perfectly well-formed single statement.
Reported-by: David Howells <dhowells@redhat.com>
Suggested-by: Joe Perches <joe@perches.com>
Cc: Ian Kent <raven@themaw.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Only a few file systems need this. Start by pushing it down into each
fs rmdir method (except gfs2 and xfs) so it can be dealt with on a per-fs
basis.
This does not change behavior for any in-tree file systems.
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In fs/autofs4/dev-ioctl.c::autofs_dev_ioctl_setpipefd() we call fget(),
which may return NULL, but we do not explicitly test for that NULL return
so we may end up dereferencing a NULL pointer - bad.
When I originally submitted this patch I had chosen EBUSY as the return
value to use if this happens. Ian Kent was kind enough to explain why that
would most likely be wrong and why EBADF should most likely be used
instead. This version of the patch uses EBADF.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The autofs4_lock introduced by the rcu-walk changes has unnecessarily
broad scope. The locking is better handled by the per-autofs super
block lookup_lock.
Signed-off-by: Ian Kent <raven@themaw.net>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The daemon never needs to block and, in the rcu-walk case an error
return isn't used, so always return zero.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The vfs-scale changes changed the traversal used in
autofs4_expire_indirect() from a list to a depth first tree traversal
which isn't right.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
There is a missing dput() when returning from autofs4_expire_direct()
when we see that the dentry is already a pending mount.
Signed-off-by: Ian Kent <raven@themaw.net>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
When direct (and offset) mounts were introduced the the last used
timeout could no longer be updated in ->d_revalidate(). This is
because covered direct mounts would be followed over without calling
the autofs file system. As a result the definition of the busyness
check for all entries was changed to be "actually busy" being an open
file or working directory within the automount. But now we have a call
back in the follow so the last used update on any access can be
re-instated. This requires DCACHE_MANAGE_TRANSIT to always be set.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The latter is called only when both ino and dentry are about to
be freed, so cleaning ->d_fsdata and ->dentry is pointless.
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
split init_ino into new_ino and clean_ino; the former is
what used to be init_ino(NULL, sbi), the latter is for cases
where we passed non-NULL ino. Lose unused arguments.
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
It's used only to pass the length of symlink body to
autofs4_get_inode() in autofs4_dir_symlink(). We can
bloody well set inode->i_size in autofs4_dir_symlink()
directly and be done with that.
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In all cases we'd set inf->mode to know value just before
passing it to autofs4_get_inode(). That kills the need
to store it in autofs_info and pass it to autofs_init_ino()
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Kill it. Mind you, it's been an obfuscated call of autofs4_init_ino()
ever since 2.3.99pre6-4...
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
gets rid of all ->free()/->u.symlink machinery in autofs; we simply
keep symlink bodies in inode->i_private and free them in ->evict_inode().
Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
oz_mode isn't defined any more, use autofs4_oz_mode(sbi) instead.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The initialization condition in fs/autofs4/expire.c:get_next_positive_dentry()
appears to be incorrect. If prev == NULL I believe that root should be
returned.
Further down, at the current dentry check for it being simple_positive()
it looks like the d_lock for dentry p should be dropped instead of dentry
ret, otherwise when p is assinged to ret we end up with no lock on p and
a lost lock on ret, which leads to a deadlock.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Merge the remaining autofs4 dentry ops tables. It doesn't matter if
d_automount and d_manage are present on something that's not mountable or
holdable as these ops are only used if the appropriate flags are set in
dentry->d_flags.
[AV] switch to ->s_d_op, since now _everything_ on autofs4 is using the
same dentry_operations.
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Allow d_manage() to be called from pathwalk when it is in RCU-walk mode as well
as when it is in Ref-walk mode. This permits __follow_mount_rcu() to call
d_manage() directly. d_manage() needs a parameter to indicate that it is in
RCU-walk mode as it isn't allowed to sleep if in that mode (but should return
-ECHILD instead).
autofs4_d_manage() can then be set to retain RCU-walk mode if the daemon
accesses it and otherwise request dropping back to ref-walk mode.
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Version 4 of autofs provides a pseudo direct mount implementation
that relies on directories at the leaves of a directory tree under
an indirect mount to trigger mounts.
This patch adds support for that functionality.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
It is possible for the check in wait.c:validate_request() to return
an incorrect result if the dentry that was mounted upon has changed
during the callback.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
When this function is called the local reference count does't need to
be updated since the dentry is going away and dput definitely must
not be called here.
Also the autofs info struct field inode isn't used so remove it.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
There are now two distinct dentry operations uses. One for dentrys
that trigger mounts and one for dentrys that do not.
Rationalize the use of these dentry operations and rename them to
reflect their function.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Since the use of ->follow_link() has been eliminated there is no
need to separate the indirect and direct inode operations.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Remove code that is not used due to the use of ->d_automount()
and ->d_manage().
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This patch required a previous patch to add the ->d_automount()
dentry operation.
Add a function to use the newly defined ->d_manage() dentry operation
for blocking during mount and expire.
Whether the VFS calls the dentry operations d_automount() and d_manage()
is controled by the DMANAGED_AUTOMOUNT and DMANAGED_TRANSIT flags. autofs
uses the d_automount() operation to callback to user space to request
mount operations and the d_manage() operation to block walks into mounts
that are under construction or destruction.
In order to prevent these functions from being called unnecessarily the
DMANAGED_* flags are cleared for cases which would cause this. In the
common case the DMANAGED_AUTOMOUNT and DMANAGED_TRANSIT flags are both
set for dentrys waiting to be mounted. The DMANAGED_TRANSIT flag is
cleared upon successful mount request completion and set during expire
runs, both during the dentry expire check, and if selected for expire,
is left set until a subsequent successful mount request completes.
The exception to this is the so-called rootless multi-mount which has
no actual mount at its base. In this case the DMANAGED_AUTOMOUNT flag
is cleared upon successful mount request completion as well and set
again after a successful expire.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Add a function to use the newly defined ->d_automount() dentry operation
for triggering mounts instead of doing the user space callback in ->lookup()
and ->d_revalidate().
Note, to be useful the subsequent patch to add the ->d_manage() dentry
operation is also needed so the discussion of functionality is deferred to
that patch.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Add a dentry op (d_manage) to permit a filesystem to hold a process and make it
sleep when it tries to transit away from one of that filesystem's directories
during a pathwalk. The operation is keyed off a new dentry flag
(DCACHE_MANAGE_TRANSIT).
The filesystem is allowed to be selective about which processes it holds and
which it permits to continue on or prohibits from transiting from each flagged
directory. This will allow autofs to hold up client processes whilst letting
its userspace daemon through to maintain the directory or the stuff behind it
or mounted upon it.
The ->d_manage() dentry operation:
int (*d_manage)(struct path *path, bool mounting_here);
takes a pointer to the directory about to be transited away from and a flag
indicating whether the transit is undertaken by do_add_mount() or
do_move_mount() skipping through a pile of filesystems mounted on a mountpoint.
It should return 0 if successful and to let the process continue on its way;
-EISDIR to prohibit the caller from skipping to overmounted filesystems or
automounting, and to use this directory; or some other error code to return to
the user.
->d_manage() is called with namespace_sem writelocked if mounting_here is true
and no other locks held, so it may sleep. However, if mounting_here is true,
it may not initiate or wait for a mount or unmount upon the parameter
directory, even if the act is actually performed by userspace.
Within fs/namei.c, follow_managed() is extended to check with d_manage() first
on each managed directory, before transiting away from it or attempting to
automount upon it.
follow_down() is renamed follow_down_one() and should only be used where the
filesystem deliberately intends to avoid management steps (e.g. autofs).
A new follow_down() is added that incorporates the loop done by all other
callers of follow_down() (do_add/move_mount(), autofs and NFSD; whilst AFS, NFS
and CIFS do use it, their use is removed by converting them to use
d_automount()). The new follow_down() calls d_manage() as appropriate. It
also takes an extra parameter to indicate if it is being called from mount code
(with namespace_sem writelocked) which it passes to d_manage(). follow_down()
ignores automount points so that it can be used to mount on them.
__follow_mount_rcu() is made to abort rcu-walk mode if it hits a directory with
DCACHE_MANAGE_TRANSIT set on the basis that we're probably going to have to
sleep. It would be possible to enter d_manage() in rcu-walk mode too, and have
that determine whether to abort or not itself. That would allow the autofs
daemon to continue on in rcu-walk mode.
Note that DCACHE_MANAGE_TRANSIT on a directory should be cleared when it isn't
required as every tranist from that directory will cause d_manage() to be
invoked. It can always be set again when necessary.
==========================
WHAT THIS MEANS FOR AUTOFS
==========================
Autofs currently uses the lookup() inode op and the d_revalidate() dentry op to
trigger the automounting of indirect mounts, and both of these can be called
with i_mutex held.
autofs knows that the i_mutex will be held by the caller in lookup(), and so
can drop it before invoking the daemon - but this isn't so for d_revalidate(),
since the lock is only held on _some_ of the code paths that call it. This
means that autofs can't risk dropping i_mutex from its d_revalidate() function
before it calls the daemon.
The bug could manifest itself as, for example, a process that's trying to
validate an automount dentry that gets made to wait because that dentry is
expired and needs cleaning up:
mkdir S ffffffff8014e05a 0 32580 24956
Call Trace:
[<ffffffff885371fd>] :autofs4:autofs4_wait+0x674/0x897
[<ffffffff80127f7d>] avc_has_perm+0x46/0x58
[<ffffffff8009fdcf>] autoremove_wake_function+0x0/0x2e
[<ffffffff88537be6>] :autofs4:autofs4_expire_wait+0x41/0x6b
[<ffffffff88535cfc>] :autofs4:autofs4_revalidate+0x91/0x149
[<ffffffff80036d96>] __lookup_hash+0xa0/0x12f
[<ffffffff80057a2f>] lookup_create+0x46/0x80
[<ffffffff800e6e31>] sys_mkdirat+0x56/0xe4
versus the automount daemon which wants to remove that dentry, but can't
because the normal process is holding the i_mutex lock:
automount D ffffffff8014e05a 0 32581 1 32561
Call Trace:
[<ffffffff80063c3f>] __mutex_lock_slowpath+0x60/0x9b
[<ffffffff8000ccf1>] do_path_lookup+0x2ca/0x2f1
[<ffffffff80063c89>] .text.lock.mutex+0xf/0x14
[<ffffffff800e6d55>] do_rmdir+0x77/0xde
[<ffffffff8005d229>] tracesys+0x71/0xe0
[<ffffffff8005d28d>] tracesys+0xd5/0xe0
which means that the system is deadlocked.
This patch allows autofs to hold up normal processes whilst the daemon goes
ahead and does things to the dentry tree behind the automouter point without
risking a deadlock as almost no locks are held in d_manage() and none in
d_automount().
Signed-off-by: David Howells <dhowells@redhat.com>
Was-Acked-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Require filesystems be aware of .d_revalidate being called in rcu-walk
mode (nd->flags & LOOKUP_RCU). For now do a simple push down, returning
-ECHILD from all implementations.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
Reduce some branches and memory accesses in dcache lookup by adding dentry
flags to indicate common d_ops are set, rather than having to check them.
This saves a pointer memory access (dentry->d_op) in common path lookup
situations, and saves another pointer load and branch in cases where we
have d_op but not the particular operation.
Patched with:
git grep -E '[.>]([[:space:]])*d_op([[:space:]])*=' | xargs sed -e 's/\([^\t ]*\)->d_op = \(.*\);/d_set_d_op(\1, \2);/' -e 's/\([^\t ]*\)\.d_op = \(.*\);/d_set_d_op(\&\1, \2);/' -i
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
Rather than keep a d_mounted count in the dentry, set a dentry flag instead.
The flag can be cleared by checking the hash table to see if there are any
mounts left, which is not time critical because it is performed at detach time.
The mounted state of a dentry is only used to speculatively take a look in the
mount hash table if it is set -- before following the mount, vfsmount lock is
taken and mount re-checked without races.
This saves 4 bytes on 32-bit, nothing on 64-bit but it does provide a hole I
might use later (and some configs have larger than 32-bit spinlocks which might
make use of the hole).
Autofs4 conversion and changelog by Ian Kent <raven@themaw.net>:
In autofs4, when expring direct (or offset) mounts we need to ensure that we
block user path walks into the autofs mount, which is covered by another mount.
To do this we clear the mounted status so that follows stop before walking into
the mount and are essentially blocked until the expire is completed. The
automount daemon still finds the correct dentry for the umount due to the
follow mount logic in fs/autofs4/root.c:autofs4_follow_link(), which is set as
an inode operation for direct and offset mounts only and is called following
the lookup that stopped at the covered mount.
At the end of the expire the covering mount probably has gone away so the
mounted status need not be restored. But we need to check this and only restore
the mounted status if the expire failed.
XXX: autofs may not work right if we have other mounts go over the top of it?
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
The remaining usages for dcache_lock is to allow atomic, multi-step read-side
operations over the directory tree by excluding modifications to the tree.
Also, to walk in the leaf->root direction in the tree where we don't have
a natural d_lock ordering.
This could be accomplished by taking every d_lock, but this would mean a
huge number of locks and actually gets very tricky.
Solve this instead by using the rename seqlock for multi-step read-side
operations, retry in case of a rename so we don't walk up the wrong parent.
Concurrent dentry insertions are not serialised against. Concurrent deletes
are tricky when walking up the directory: our parent might have been deleted
when dropping locks so also need to check and retry for that.
We can also use the rename lock in cases where livelock is a worry (and it
is introduced in subsequent patch).
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
Protect d_subdirs and d_child with d_lock, except in filesystems that aren't
using dcache_lock for these anyway (eg. using i_mutex).
Note: if we change the locking rule in future so that ->d_child protection is
provided only with ->d_parent->d_lock, it may allow us to reduce some locking.
But it would be an exception to an otherwise regular locking scheme, so we'd
have to see some good results. Probably not worthwhile.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
