The file_lock_list is only used for /proc/locks. The vastly common case
is for locks to be put onto the list and come off again, without ever
being traversed.
Help optimize for this use-case by moving to percpu hlist_head-s. At the
same time, we can make the locking less contentious by moving to an
lglock. When iterating over the lists for /proc/locks, we must take the
global lock and then iterate over each CPU's list in turn.
This change necessitates a new fl_link_cpu field to keep track of which
CPU the entry is on. On x86_64 at least, this field is placed within an
existing hole in the struct to avoid growing the size.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
There's no reason we have to protect the blocked_hash and file_lock_list
with the same spinlock. With the tests I have, breaking it in two gives
a barely measurable performance benefit, but it seems reasonable to make
this locking as granular as possible.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Currently, the hashing that the locking code uses to add these values
to the blocked_hash is simply calculated using fl_owner field. That's
valid in most cases except for server-side lockd, which validates the
owner of a lock based on fl_owner and fl_pid.
In the case where you have a small number of NFS clients doing a lot
of locking between different processes, you could end up with all
the blocked requests sitting in a very small number of hash buckets.
Add a new lm_owner_key operation to the lock_manager_operations that
will generate an unsigned long to use as the key in the hashtable.
That function is only implemented for server-side lockd, and simply
XORs the fl_owner and fl_pid.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Break up the blocked_list into a hashtable, using the fl_owner as a key.
This speeds up searching the hash chains, which is especially significant
for deadlock detection.
Note that the initial implementation assumes that hashing on fl_owner is
sufficient. In most cases it should be, with the notable exception being
server-side lockd, which compares ownership using a tuple of the
nlm_host and the pid sent in the lock request. So, this may degrade to a
single hash bucket when you only have a single NFS client. That will be
addressed in a later patch.
The careful observer may note that this patch leaves the file_lock_list
alone. There's much less of a case for turning the file_lock_list into a
hashtable. The only user of that list is the code that generates
/proc/locks, and it always walks the entire list.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Testing has shown that iterating over the blocked_list for deadlock
detection turns out to be a bottleneck. In order to alleviate that,
begin the process of turning it into a hashtable. We start by turning
the fl_link into a hlist_node and the global lists into hlists. A later
patch will do the conversion of the blocked_list to a hashtable.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Since we always hold the i_lock when inserting a new waiter onto the
fl_block list, we can avoid taking the global lock at all if we find
that it's empty when we go to wake up blocked waiters.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Having a global lock that protects all of this code is a clear
scalability problem. Instead of doing that, move most of the code to be
protected by the i_lock instead. The exceptions are the global lists
that the ->fl_link sits on, and the ->fl_block list.
->fl_link is what connects these structures to the
global lists, so we must ensure that we hold those locks when iterating
over or updating these lists.
Furthermore, sound deadlock detection requires that we hold the
blocked_list state steady while checking for loops. We also must ensure
that the search and update to the list are atomic.
For the checking and insertion side of the blocked_list, push the
acquisition of the global lock into __posix_lock_file and ensure that
checking and update of the blocked_list is done without dropping the
lock in between.
On the removal side, when waking up blocked lock waiters, take the
global lock before walking the blocked list and dequeue the waiters from
the global list prior to removal from the fl_block list.
With this, deadlock detection should be race free while we minimize
excessive file_lock_lock thrashing.
Finally, in order to avoid a lock inversion problem when handling
/proc/locks output we must ensure that manipulations of the fl_block
list are also protected by the file_lock_lock.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Move the fl_link list handling routines into a separate set of helpers.
Also ensure that locks and requests are always put on global lists
last (after fully initializing them) and are taken off before unintializing
them.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
commit 66189be74 (CIFS: Fix VFS lock usage for oplocked files) exported
the locks_delete_block symbol. There's already an exported helper
function that provides this capability however, so make cifs use that
instead and turn locks_delete_block back into a static function.
Note that if fl->fl_next == NULL then this lock has already been through
locks_delete_block(), so we should be OK to ignore an ENOENT error here
and simply not retry the lock.
Cc: Pavel Shilovsky <piastryyy@gmail.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: J. Bruce Fields <bfields@fieldses.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This is to complete part of the Userspace API (UAPI) disintegration for which
the preparatory patches were pulled recently. After these patches, userspace
headers will be segregated into:
include/uapi/linux/.../foo.h
for the userspace interface stuff, and:
include/linux/.../foo.h
for the strictly kernel internal stuff.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
The rules for fl_type are rather convoluted. Typically it's treated as
holding specific values, except in the case of LOCK_MAND, in which case
it can be or'ed with LOCK_READ|LOCK_WRITE.
On some arches F_WRLCK == 2 and F_UNLCK == 3, so and'ing with F_WRLCK will also
catch the F_UNLCK case. It's unlikely in either case here that we'd ever see
F_UNLCK since those shouldn't end up on any lists, but it's still best to be
consistent.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
In commit 3b6e2723f3 ("locks: prevent side-effects of
locks_release_private before file_lock is initialized") we removed the
last user of lm_release_private without removing the field itself.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull nfsd changes from J. Bruce Fields:
"This has been an unusually quiet cycle--mostly bugfixes and cleanup.
The one large piece is Stanislav's work to containerize the server's
grace period--but that in itself is just one more step in a
not-yet-complete project to allow fully containerized nfs service.
There are a number of outstanding delegation, container, v4 state, and
gss patches that aren't quite ready yet; 3.7 may be wilder."
* 'nfsd-next' of git://linux-nfs.org/~bfields/linux: (35 commits)
NFSd: make boot_time variable per network namespace
NFSd: make grace end flag per network namespace
Lockd: move grace period management from lockd() to per-net functions
LockD: pass actual network namespace to grace period management functions
LockD: manage grace list per network namespace
SUNRPC: service request network namespace helper introduced
NFSd: make nfsd4_manager allocated per network namespace context.
LockD: make lockd manager allocated per network namespace
LockD: manage grace period per network namespace
Lockd: add more debug to host shutdown functions
Lockd: host complaining function introduced
LockD: manage used host count per networks namespace
LockD: manage garbage collection timeout per networks namespace
LockD: make garbage collector network namespace aware.
LockD: mark host per network namespace on garbage collect
nfsd4: fix missing fault_inject.h include
locks: move lease-specific code out of locks_delete_lock
locks: prevent side-effects of locks_release_private before file_lock is initialized
NFSd: set nfsd_serv to NULL after service destruction
NFSd: introduce nfsd_destroy() helper
...
When calling fcntl(fd, F_SETLEASE, lck) [with lck=F_WRLCK or F_RDLCK],
the custom signal or owner (if any were previously set using F_SETSIG
or F_SETOWN fcntls) would be reset when F_SETLEASE was called for the
second time on the same file descriptor.
This bug is a regression of 2.6.37 and is described here:
https://bugzilla.kernel.org/show_bug.cgi?id=43336
This patch reverts a commit from Oct 2004 (with subject "nfs4 lease:
move the f_delown processing") which originally introduced the
lm_release_private callback.
Signed-off-by: Filipe Brandenburger <filbranden@gmail.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
The only checks of the long argument passed to fcntl(fd,F_SETLEASE,.)
are done after converting the long to an int. Thus some illegal values
may be let through and cause problems in later code.
[ They actually *don't* cause problems in mainline, as of Dave Jones's
commit 8d657eb3b4 "Remove easily user-triggerable BUG from
generic_setlease", but we should fix this anyway. And this patch will
be necessary to fix real bugs on earlier kernels. ]
Cc: stable@vger.kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This can be trivially triggered from userspace by passing in something unexpected.
kernel BUG at fs/locks.c:1468!
invalid opcode: 0000 [#1] SMP
RIP: 0010:generic_setlease+0xc2/0x100
Call Trace:
__vfs_setlease+0x35/0x40
fcntl_setlease+0x76/0x150
sys_fcntl+0x1c6/0x810
system_call_fastpath+0x1a/0x1f
Signed-off-by: Dave Jones <davej@redhat.com>
Cc: stable@kernel.org # 3.2+
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
