When adding or removing a sysfs_dirent, the user used to be required
to call link/unlink separately. It was for two reasons - code looked
like that before sysfs_addrm_cxt conversion and to avoid looping
through parent_sd->children list twice during removal.
Performance optimization during removal just isn't worth it. Make
sysfs_add/remove_one() call sysfs_link/unlink_sibing() implicitly.
This makes code simpler albeit slightly less efficient. This change
doesn't introduce any noticeable behavior change.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
With the shadow directories gone, sysfs_rename_dir() can be simplified.
* parent doesn't need to be grabbed separately. Just access
old_dentry->d_parent.
* parent sd can never change. Remove code to move under the new
parent.
* Massage comments a bit.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
* remove space between * and symbol name in variable declaration.
* kill unnecessary new line.
* kill 'found' and test 'sd' instead.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
While shadow directories appear to be a good idea, the current scheme
of controlling their creation and destruction outside of sysfs appears
to be a locking and maintenance nightmare in the face of sysfs
directories dynamically coming and going. Which can now occur for
directories containing network devices when CONFIG_SYSFS_DEPRECATED is
not set.
This patch removes everything from the initial shadow directory support
that allowed the shadow directory creation to be controlled at a higher
level. So except for a few bits of sysfs_rename_dir everything from
commit b592fcfe7f is now gone.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Use mutex instead of semaphore in sysfs/file.c : sys_buffer.
Signed-off-by: Dave Young <hidave.darkstar@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Allows debugfs helper functions to have a hex output, rather than just decimal
Signed-off-by: Robin Getz <rgetz@blackfin.uclinux.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
A kset should not have its name set directly, so dynamically set the
name at runtime.
This is needed to remove the static array in the kobject structure which
will be changed in a future patch.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
A number of different drivers incorrect access the kobject name field
directly. This is not correct as the name might not be in the array.
Use the proper accessor function instead.
Modify ocfs2_dir_foreach_blk() to optionally return any error from the
filldir callback. This way ocfs2_dirforeach() can terminate early, as
opposed to always passing through the entire directory. This fixes a bug
introduced during a previous code refactor where ocfs2_empty_dir() would
loop infinitely.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Create all new directories with OCFS2_INLINE_DATA_FL and the inline data
bytes formatted as an empty directory. Inode size field reflects the actual
amount of inline data available, which makes searching for dirent space
very similar to the regular directory search.
Inline-data directories are automatically pushed out to extents on any
insert request which is too large for the available space.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
This splits out extent based directory read support and implements
inline-data versions of those functions. All knowledge of inline-data versus
extent based directories is internalized. For lookups the code uses
ocfs2_find_entry_id(), full dir iterations make use of
ocfs2_dir_foreach_blk_id().
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
This fixes up write, truncate, mmap, and RESVSP/UNRESVP to understand inline
inode data.
For the most part, the changes to the core write code can be relied on to do
the heavy lifting. Any code calling ocfs2_write_begin (including shared
writeable mmap) can count on it doing the right thing with respect to
growing inline data to an extent tree.
Size reducing truncates, including UNRESVP can simply zero that portion of
the inode block being removed. Size increasing truncatesm, including RESVP
have to be a little bit smarter and grow the inode to an extent tree if
necessary.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
This hooks up ocfs2_readpage() to populate a page with data from an inode
block. Direct IO reads from inline data are modified to fall back to
buffered I/O. Appropriate checks are also placed in the extent map code to
avoid reading an extent list when inline data might be stored.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
Add the disk, network and memory structures needed to support data in inode.
Struct ocfs2_inline_data is defined and embedded in ocfs2_dinode for storing
inline data.
A new inode field, i_dyn_features, is added to facilitate tracking of
dynamic inode state. Since it will be used often, we want to mirror it on
ocfs2_inode_info, and transfer it via the meta data lvb.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
The check to see if a new dirent would fit in an old one is pretty ugly, and
it's done at least twice. Clean things up by putting this in it's own
easier-to-read function.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
ocfs2_rename() does direct manipulation of the dirent it's gotten back from
a directory search. Wrap this manipulation inside of a function so that we
can transparently change directory update behavior in the future. As an
added bonus, this gets rid of an ugly macro.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
A couple paths which needed to just match a parent dir + name pair to an
inode number were a bit messy because they had to deal with
ocfs2_find_files_on_disk() which returns a larger number of values. Provide
a convenience function, ocfs2_lookup_ino_from_name() which internalizes all
the extra accounting.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
We can preserve the behavior of ocfs2_empty_dir(), while getting rid of the
open coded directory walk by just providing a smart filldir callback. This
also automatically gets to use the dir readahead code, though in this case
any advantage is minor at best.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
ocfs2_queue_orphans() has an open coded readdir loop which can easily just
use a directory accessor function.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
filldir_t can take this, so don't turn de->inode into a 32 bit value. Right
now this doesn't make a difference since no ocfs2 inodes overflow that, but
it could be a nasty surprise later on if some kernel code is calling
ocfs2_dir_foreach_blk() and expecting real inode numbers back...
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
Put this in it's own function so that the functionality can be overridden.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
The code for adding, removing, deleting directory entries was splattered all
over namei.c. I'd rather have this all centralized, so that it's easier to
make changes for inline dir data, and eventually indexed directories.
None of the code in any of the functions was changed. I only removed the
static keyword from some prototypes so that they could be exported.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
We'll want to reuse most of this when pushing inline data back out to an
extent. Keeping this part as a seperate patch helps to keep the upcoming
changes for write support uncluttered.
The core portion of ocfs2_zero_cluster_pages() responsible for making sure a
page is mapped and properly dirtied is abstracted out into it's own
function, ocfs2_map_and_dirty_page(). Actual functionality doesn't change,
though zeroing becomes optional.
We also turn part of ocfs2_free_write_ctxt() into a common function for
unlocking and freeing a page array. This operation is very common (and
uniform) for Ocfs2 cluster sizes greater than page size, so it makes sense
to keep the code in one place.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
By doing this, we can remove any higher level logic which has to have
knowledge of btree functionality - any callers of ocfs2_write_begin() can
now expect it to do anything necessary to prepare the inode for new data.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Reviewed-by: Joel Becker <joel.becker@oracle.com>
Implement sops->show_options() so as to allow /proc/mounts to show the mount
options.
Signed-off-by: Sunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
This is technically harmless (recovery will clean it out later), but leaves
a bogus entry in the slot_map which really shouldn't be there.
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
delete_tail_recs in ocfs2_try_to_merge_extent() was only ever set, remove
it.
Signed-off-by: Tao Mao <tao.ma@oracle.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
ocfs2_insert_type->ins_free_records was only used in one place, and was set
incorrectly in most places. We can free up some memory and lose some code by
removing this.
* Small warning fixup contributed by Andrew Mortom <akpm@linux-foundation.org>
Signed-off-by: Tao Mao <tao.ma@oracle.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
Big thanks go to Mathias Kolehmainen for reporting the bug, providing
debug output and testing the patches I sent him to get it working.
The fix was to stop calling ntfs_attr_set() at mount time as that causes
balance_dirty_pages_ratelimited() to be called which on systems with
little memory actually tries to go and balance the dirty pages which tries
to take the s_umount semaphore but because we are still in fill_super()
across which the VFS holds s_umount for writing this results in a
deadlock.
We now do the dirty work by hand by submitting individual buffers. This
has the annoying "feature" that mounting can take a few seconds if the
journal is large as we have clear it all. One day someone should improve
on this by deferring the journal clearing to a helper kernel thread so it
can be done in the background but I don't have time for this at the moment
and the current solution works fine so I am leaving it like this for now.
Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw: (51 commits)
[DLM] block dlm_recv in recovery transition
[DLM] don't overwrite castparam if it's NULL
[GFS2] Get superblock a different way
[GFS2] Don't try to remove buffers that don't exist
[GFS2] Alternate gfs2_iget to avoid looking up inodes being freed
[GFS2] Data corruption fix
[GFS2] Clean up journaled data writing
[GFS2] GFS2: chmod hung - fix race in thread creation
[DLM] Make dlm_sendd cond_resched more
[GFS2] Move inode deletion out of blocking_cb
[GFS2] flocks from same process trip kernel BUG at fs/gfs2/glock.c:1118!
[GFS2] Clean up gfs2_trans_add_revoke()
[GFS2] Use slab operations for all gfs2_bufdata allocations
[GFS2] Replace revoke structure with bufdata structure
[GFS2] Fix ordering of dirty/journal for ordered buffer unstuffing
[GFS2] Clean up ordered write code
[GFS2] Move pin/unpin into lops.c, clean up locking
[GFS2] Don't mark jdata dirty in gfs2_unstuffer_page()
[GFS2] Introduce gfs2_remove_from_ail
[GFS2] Correct lock ordering in unlink
...
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc: (408 commits)
[POWERPC] Add memchr() to the bootwrapper
[POWERPC] Implement logging of unhandled signals
[POWERPC] Add legacy serial support for OPB with flattened device tree
[POWERPC] Use 1TB segments
[POWERPC] XilinxFB: Allow fixed framebuffer base address
[POWERPC] XilinxFB: Add support for custom screen resolution
[POWERPC] XilinxFB: Use pdata to pass around framebuffer parameters
[POWERPC] PCI: Add 64-bit physical address support to setup_indirect_pci
[POWERPC] 4xx: Kilauea defconfig file
[POWERPC] 4xx: Kilauea DTS
[POWERPC] 4xx: Add AMCC Kilauea eval board support to platforms/40x
[POWERPC] 4xx: Add AMCC 405EX support to cputable.c
[POWERPC] Adjust TASK_SIZE on ppc32 systems to 3GB that are capable
[POWERPC] Use PAGE_OFFSET to tell if an address is user/kernel in SW TLB handlers
[POWERPC] 85xx: Enable FP emulation in MPC8560 ADS defconfig
[POWERPC] 85xx: Killed <asm/mpc85xx.h>
[POWERPC] 85xx: Add cpm nodes for 8541/8555 CDS
[POWERPC] 85xx: Convert mpc8560ads to the new CPM binding.
[POWERPC] mpc8272ads: Remove muram from the CPM reg property.
[POWERPC] Make clockevents work on PPC601 processors
...
Fixed up conflict in Documentation/powerpc/booting-without-of.txt manually.
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (867 commits)
[SKY2]: status polling loop (post merge)
[NET]: Fix NAPI completion handling in some drivers.
[TCP]: Limit processing lost_retrans loop to work-to-do cases
[TCP]: Fix lost_retrans loop vs fastpath problems
[TCP]: No need to re-count fackets_out/sacked_out at RTO
[TCP]: Extract tcp_match_queue_to_sack from sacktag code
[TCP]: Kill almost unused variable pcount from sacktag
[TCP]: Fix mark_head_lost to ignore R-bit when trying to mark L
[TCP]: Add bytes_acked (ABC) clearing to FRTO too
[IPv6]: Update setsockopt(IPV6_MULTICAST_IF) to support RFC 3493, try2
[NETFILTER]: x_tables: add missing ip6t_modulename aliases
[NETFILTER]: nf_conntrack_tcp: fix connection reopening
[QETH]: fix qeth_main.c
[NETLINK]: fib_frontend build fixes
[IPv6]: Export userland ND options through netlink (RDNSS support)
[9P]: build fix with !CONFIG_SYSCTL
[NET]: Fix dev_put() and dev_hold() comments
[NET]: make netlink user -> kernel interface synchronious
[NET]: unify netlink kernel socket recognition
[NET]: cleanup 3rd argument in netlink_sendskb
...
Fix up conflicts manually in Documentation/feature-removal-schedule.txt
and my new least favourite crap, the "mod_devicetable" support in the
files include/linux/mod_devicetable.h and scripts/mod/file2alias.c.
(The latter files seem to be explicitly _designed_ to get conflicts when
different subsystems work with them - that have an absolutely horrid
lack of subsystem separation!)
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
On Fri, 2007-07-13 at 02:05 -0700, Andrew Morton wrote:
> Except lockdep doesn't know about journal_start(), which has ranking
> requirements similar to a semaphore.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
We were intending to put the previous instance of delegation->cred
before setting a new one.
Thanks to David Howells for spotting this.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
This patch make processing netlink user -> kernel messages synchronious.
This change was inspired by the talk with Alexey Kuznetsov about current
netlink messages processing. He says that he was badly wrong when introduced
asynchronious user -> kernel communication.
The call netlink_unicast is the only path to send message to the kernel
netlink socket. But, unfortunately, it is also used to send data to the
user.
Before this change the user message has been attached to the socket queue
and sk->sk_data_ready was called. The process has been blocked until all
pending messages were processed. The bad thing is that this processing
may occur in the arbitrary process context.
This patch changes nlk->data_ready callback to get 1 skb and force packet
processing right in the netlink_unicast.
Kernel -> user path in netlink_unicast remains untouched.
EINTR processing for in netlink_run_queue was changed. It forces rtnl_lock
drop, but the process remains in the cycle until the message will be fully
processed. So, there is no need to use this kludges now.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
This function allocates the zeroed chunk of memory and
call seq_open(). The __seq_open_private() helper returns
the allocated memory to make it possible for the caller
to initialize it.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the net namespaces many code leaved the __init section,
thus making the kernel occupy more memory than it did before.
Since we have a config option that prohibits the namespace
creation, the functions that initialize/finalize some netns
stuff are simply not needed and can be freed after the boot.
Currently, this is almost not noticeable, since few calls
are no longer in __init, but when the namespaces will be
merged it will be possible to free more code. I propose to
use the __net_init, __net_exit and __net_initdata "attributes"
for functions/variables that are not used if the CONFIG_NET_NS
is not set to save more space in memory.
The exiting functions cannot just reside in the __exit section,
as noticed by David, since the init section will have
references on it and the compilation will fail due to modpost
checks. These references can exist, since the init namespace
never dies and the exit callbacks are never called. So I
introduce the __exit_refok attribute just like it is already
done with the __init_refok.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The problem: proc_net files remember which network namespace the are
against but do not remember hold a reference count (as that would pin
the network namespace). So we currently have a small window where
the reference count on a network namespace may be incremented when opening
a /proc file when it has already gone to zero.
To fix this introduce maybe_get_net and get_proc_net.
maybe_get_net increments the network namespace reference count only if it is
greater then zero, ensuring we don't increment a reference count after it
has gone to zero.
get_proc_net handles all of the magic to go from a proc inode to the network
namespace instance and call maybe_get_net on it.
PROC_NET the old accessor is removed so that we don't get confused and use
the wrong helper function.
Then I fix up the callers to use get_proc_net and handle the case case
where get_proc_net returns NULL. In that case I return -ENXIO because
effectively the network namespace has already gone away so the files
we are trying to access don't exist anymore.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the appropriate EXPORT_SYMBOLS for proc_net_create,
proc_net_fops_create and proc_net_remove to fix errors when
compiling allmodconfig
Signed-off-by: Mark Nelson <markn@au1.ibm.com>
Acked-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch makes most of the generic device layer network
namespace safe. This patch makes dev_base_head a
network namespace variable, and then it picks up
a few associated variables. The functions:
dev_getbyhwaddr
dev_getfirsthwbytype
dev_get_by_flags
dev_get_by_name
__dev_get_by_name
dev_get_by_index
__dev_get_by_index
dev_ioctl
dev_ethtool
dev_load
wireless_process_ioctl
were modified to take a network namespace argument, and
deal with it.
vlan_ioctl_set and brioctl_set were modified so their
hooks will receive a network namespace argument.
So basically anthing in the core of the network stack that was
affected to by the change of dev_base was modified to handle
multiple network namespaces. The rest of the network stack was
simply modified to explicitly use &init_net the initial network
namespace. This can be fixed when those components of the network
stack are modified to handle multiple network namespaces.
For now the ifindex generator is left global.
Fundametally ifindex numbers are per namespace, or else
we will have corner case problems with migration when
we get that far.
At the same time there are assumptions in the network stack
that the ifindex of a network device won't change. Making
the ifindex number global seems a good compromise until
the network stack can cope with ifindex changes when
you change namespaces, and the like.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Each netlink socket will live in exactly one network namespace,
this includes the controlling kernel sockets.
This patch updates all of the existing netlink protocols
to only support the initial network namespace. Request
by clients in other namespaces will get -ECONREFUSED.
As they would if the kernel did not have the support for
that netlink protocol compiled in.
As each netlink protocol is updated to be multiple network
namespace safe it can register multiple kernel sockets
to acquire a presence in the rest of the network namespaces.
The implementation in af_netlink is a simple filter implementation
at hash table insertion and hash table look up time.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch makes /proc/net per network namespace. It modifies the global
variables proc_net and proc_net_stat to be per network namespace.
The proc_net file helpers are modified to take a network namespace argument,
and all of their callers are fixed to pass &init_net for that argument.
This ensures that all of the /proc/net files are only visible and
usable in the initial network namespace until the code behind them
has been updated to be handle multiple network namespaces.
Making /proc/net per namespace is necessary as at least some files
in /proc/net depend upon the set of network devices which is per
network namespace, and even more files in /proc/net have contents
that are relevant to a single network namespace.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current implementation of dev_ifname makes maintenance difficult
because updates to the implementation of the ioctl have to made in two
places. So this patch updates dev_ifname32 to do a classic 32/64
structure conversion and call sys_ioctl like the rest of the
compat calls do.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce a per-lockspace rwsem that's held in read mode by dlm_recv
threads while working in the dlm. This allows dlm_recv activity to be
suspended when the lockspace transitions to, from and between recovery
cycles.
The specific bug prompting this change is one where an in-progress
recovery cycle is aborted by a new recovery cycle. While dlm_recv was
processing a recovery message, the recovery cycle was aborted and
dlm_recoverd began cleaning up. dlm_recv decremented recover_locks_count
on an rsb after dlm_recoverd had reset it to zero. This is fixed by
suspending dlm_recv (taking write lock on the rwsem) before aborting the
current recovery.
The transitions to/from normal and recovery modes are simplified by using
this new ability to block dlm_recv. The switch from normal to recovery
mode means dlm_recv goes from processing locking messages, to saving them
for later, and vice versa. Races are avoided by blocking dlm_recv when
setting the flag that switches between modes.
Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
If the castaddr passed to the userland API is NULL then don't overwrite the
existing castparam. This allows a different thread to cancel a lock request and
the CANCEL AST gets delivered to the original thread.
bz#306391 (for RHEL4) refers.
Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The mapping may be NULL by the time the I/O has completed, so
we now get the superblock by a different route (via the bd and glock)
to avoid this problem.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Wendy Cheng <wcheng@redhat.com>
There is a possible deadlock between two processes on the same node, where one
process is deleting an inode, and another process is looking for allocated but
unused inodes to delete in order to create more space.
process A does an iput() on inode X, and it's i_count drops to 0. This causes
iput_final() to be called, which puts an inode into state I_FREEING at
generic_delete_inode(). There no point between when iput_final() is called, and
when I_FREEING is set where GFS2 could acquire any glocks. Once I_FREEING is
set, no other process on that node can successfully look up that inode until
the delete finishes.
process B locks the the resource group for the same inode in get_local_rgrp(),
which is called by gfs2_inplace_reserve_i()
process A tries to lock the resource group for the inode in
gfs2_dinode_dealloc(), but it's already locked by process B
process B waits in find_inode for the inode to have the I_FREEING state cleared.
Deadlock.
This patch solves the problem by adding an alternative to gfs2_iget(),
gfs2_iget_skip(), that simply skips any inodes that are in the I_FREEING
state.o The alternate test function is just like the original one, except that
it fails if the inode is being freed, and sets a skipped flag. The alternate
set function is just like the original, except that it fails if the skipped
flag is set. Only try_rgrp_unlink() calls gfs2_iget_skip() instead of
gfs2_iget().
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
* GFS2 has been using i_cache array to store its indirect meta blocks.
Its flush routine doesn't correctly clean up all the entries. The
problem would show while multiple nodes do simultaneous writes to the
same file. Upon glock exclusive lock transfer, if the file is a sparse
file with large file size where the indirect meta blocks span multiple
array entries with "zero" entries in between. The flush routine
prematurely stops the flushing that leaves old (stale) entries around.
This leads to several nasty issues, including data corruption.
* Fix gfs2_get_block_noalloc checking to correctly return EIO upon
unmapped buffer.
Signed-off-by: Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch cleans up the code for writing journaled data into the log.
It also removes the need to allocate a small "tag" structure for each
block written into the log. Instead we just keep count of the outstanding
I/O so that we can be sure that its all been written at the correct time.
Another result of this patch is that a number of ll_rw_block() calls
have become submit_bh() calls, closing some races at the same time.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The problem boiled down to a race between the gdlm_init_threads()
function initializing thread1 and its setting of blist = 1.
Essentially, "if (current == ls->thread1)" was checked by the thread
before the thread creator set ls->thread1.
Since thread1 is the only thread who is allowed to work on the
blocking queue, and since neither thread thought it was thread1, no one
was working on the queue. So everything just sat.
This patch reuses the ls->async_lock spin_lock to fix the race,
and it fixes the problem. I've done more than 2000 iterations of the
loop that was recreating the failure and it seems to work.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
--
Under high recovery loads dlm_sendd can monopolise the CPU and cause soft lockups.
This one extra and one moved cond_resched() make it yield a little more during
such times keeping work moving.
Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Move inode deletion code out of blocking_cb handle_callback route to
avoid racy conditions that end up blocking lock_dlm1 thread. Fix
bugzilla 286821.
Signed-off-by: Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch adds a new flag to the gfs2_holder structure GL_FLOCK.
It is set on holders of glocks representing flocks. This flag is
checked in add_to_queue() and a process is permitted to queue more
than one holder onto a glock if it is set. This solves the issue
of a process not being able to do multiple flocks on the same file.
Through a single descriptor, a process can now promote and demote
flocks. Through multiple descriptors a process can now queue
multiple flocks on the same file. There's still the problem of
a process deadlocking itself (because gfs2 blocking locks are not
interruptible) by queueing incompatible deadlock.
Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The following alters gfs2_trans_add_revoke() to take a struct
gfs2_bufdata as an argument. This eliminates the memory allocation which
was previously required by making use of the already existing struct
gfs2_bufdata. It makes some sanity checks to ensure that the
gfs2_bufdata has been removed from all the lists before its recycled as
a revoke structure. This saves one memory allocation and one free per
revoke structure.
Also as a result, and to simplify the locking, since there is no longer
any blocking code in gfs2_trans_add_revoke() we must hold the log lock
whenever this function is called. This reduces the amount of times we
take and unlock the log lock.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The old revoke structure was allocated using kalloc/kfree but
there is a slab cache for gfs2_bufdata, so we should use that
now that the structures have been converted.
This is part two of the patch series to merge the revoke
and gfs2_bufdata structures.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Both the revoke structure and the bufdata structure are quite similar.
They are basically small tags which are put on lists. In addition to
which the revoke structure is always allocated when there is a bufdata
structure which is (or can be) freed. As such it should be possible to
reduce the number of frees and allocations by using the same structure
for both purposes.
This patch is the first step along that path. It replaces existing uses
of the revoke structure with the bufdata structure.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The following patch removes the ordered write processing from
databuf_lo_before_commit() and moves it to log.c. This has the effect of
greatly simplyfying databuf_lo_before_commit() and well as potentially
making the ordered write code more efficient.
As a side effect of this, its now possible to remove ordered buffers
from the ordered buffer list at any time, so we now make use of this in
invalidatepage and releasepage to ensure timely release of these
buffers.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
gfs2_pin and gfs2_unpin are only used in lops.c, despite being
defined in meta_io.c, so this patch moves them into lops.c and
makes them static. At the same time, its possible to clean up
the locking in the buf and databuf _lo_add() functions so that
we only need to grab the spinlock once. Also we have to move
lock_buffer() around the _lo_add() functions since we can't
do that in gfs2_pin() any more since we hold the spinlock
for the duration of that function.
As a result, the code shrinks by 12 lines and we do far fewer
operations when adding buffers to the log. It also makes the
code somewhat easier to read & understand.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This collects together the operations required to remove a gfs2_bufdata
from the ail lists. Its only called from two places to start with, but
expect to see more of this function in future.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch corrects the lock ordering in unlink to be the same as
that in the rest of GFS2, i.e. parent -> child -> rgrp.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Fix a nasty inode meta data corruption issue by keeping the buffer head in
icache array. This buffer needs to stay in memory until journal flush occurs
Otherwise, gfs2_meta_inode_buffer could do a disk read before the inode hits
disk. It ends up with meta data corruptions. The buffer will be released as
part of the existing journal flush logic.
Signed-off-by: S. Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When a lot of IO, with some distributed mmap IO, is run on a GFS2 filesystem in
a cluster, it will deadlock. The reason is that do_no_page() will repeatedly
call gfs2_sharewrite_nopage(), because each node keeps giving up the glock
too early, and is forced to call unmap_mapping_range(). This bumps the
mapping->truncate_count sequence count, forcing do_no_page() to retry. This
patch institutes a minimum glock hold time a tenth a second. This insures
that even in heavy contention cases, the node has enough time to get some
useful work done before it gives up the glock.
A second issue is that when gfs2_glock_dq() is called from within a page fault
to demote a lock, and the associated page needs to be written out, it will
try to acqire a lock on it, but it has already been locked at a higher level.
This patch puts makes gfs2_glock_dq() use the work queue as well, to avoid this
issue. This is the same patch as Steve Whitehouse originally proposed to fix
this issue, execpt that gfs2_glock_dq() now grabs a reference to the glock
before it queues up the work on it.
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When you try to mount gfs2 with -o garbage, the mount fails and the gfs2
superblock is deallocated and becomes NULL. The vfs comes around later
on and calls gfs2_kill_sb. At this point the hidden gfs2 superblock
pointer (sb->s_fs_info) is NULL and dereferencing it through
gfs2_meta_syncfs causes the panic. (the other function call to
gfs2_delete_debugfs_file() succeeds because this function already checks
for a NULL pointer)
Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is a patch to GFS2 to protect sd_log_num_jdata with the
gfs2_log_lock. Without this patch, there is a timing window
where you can get hit the following assert from function
gfs2_log_flush():
gfs2_assert_withdraw(sdp,
sdp->sd_log_num_buf + sdp->sd_log_num_jdata ==
sdp->sd_log_commited_buf +
sdp->sd_log_commited_databuf);
I've tested it on my roth cluster and it fixes the problem.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
With this patch, gfs2 glockdump through the debugfs filesystem will only
dump glocks for the specified filesystem instead of all glocks. Also, to
aid debugging, the glock number is dumped in hex instead of decimal.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: S. Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Abhijith Das <adas@redhat.com>
This patch fixes the slight mess made in lowcomms closing by previous patches
and fixes all sorts of DLM hangs.
Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Current GFS2 setattr call unconditionally invokes do_shrink even the
requested size and actual file size are equal. This has generated large
amount of extra IOs found during NFS benchmark runs. This patch moves
the relevant logic out of shrink code path. Since setattr is a system
call, the time stamps update is still required.
Signed-off-by: S. Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
match_token() was returning garbage data instead of a fail value. This data
happened to match a valid option id for an option that required an argument (in
this case, lockproto=%s) For match_token() to correctly fail if the option
doesn't match any of the tokens, the token table must end with a NULL entry.
This patch adds the NULL entry.
Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This was missing from the dir_split_leaf() function although in
most cases its not a problem due to other functions having
already previously called gfs2_trans_add_bh. This makes certain
that it is correct.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Wendy Cheng <wcheng@redhat.com>
This patch fixes some bugs relating to journaled data files by cleaning
up the gfs2_invalidatepage() and gfs2_releasepage() functions. We now
never block during gfs2_releasepage(), instead we always either release
or refuse to release depending on the status of the buffers.
This fixes Red Hat bugzillas #248969 and #252392.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
This is the filesystem part of the patches to fix this bz. There are
additional userland patches (gfs2_quota, libgfs2) for the complete
solution. This patch adds a new field qu_ll_next to the gfs2_quota
structure. This field allows us to create linked lists of quotas in the
ondisk quota inode. Instead of scanning through the entire sparse quota
file for valid quotas, we can now simply walk through the user and group
quota linked lists to perform the do_list operation.
Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch forcibly unstuffs (if stuffed) the hidden quota inode at the
first availble opportunity. In any practical scenario the quota inode
won't be stuffed, so this is ok to do. Unstuffing the quota inode allows
us to ignore the case of a stuffed quota inode in gfs2_adjust_quota().
Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
the original code could work, but I think this code could work better.
Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
sb->s_fs_info is a void pointer, thus the type cast is not needed.
Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is for bugzilla bug #248176: GFS2: invalid metadata block
Patches 1 thru 3 were accepted upstream, but there were problems
with 4 and 5. Those issues have been resolved and now the recovery
tests are passing without errors. This code has gone through
41 * 3 successful gfs2 recovery tests before it hit an
unrelated (openais) problem. I'm continuing to test it.
This is a complete rewrite of patch 5 for bug #248176, written by
Steve Whitehouse. This is referred to in the bugzilla record as
"new 6" and "a different solution".
The problem was that the journal inodes, although protected by
a glock, were not synched with the other nodes because they don't
use the inode glock synch operations (i.e. no "glops" were defined).
Therefore, journal recovery on a journal-recovering node were causing
the blocks to get out of sync with the node that was actually trying
to use that journal as it comes back up from a reboot.
There are two possible solutions: (1) To make the journals use the
normal inode glock sync operations, or (2) To make the journal
operations take effect immediately (i.e. no caching). Although
option 1 works, it turns out to be a lot more code. Steve opted
for option 2, which is much simpler and therefore less prone to
regression errors.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
--
This is for bugzilla bug #248176: GFS2: invalid metadata block
Patches 1 thru 3 were accepted upstream, but there were problems
with 4 and 5. Those issues have been resolved and now the recovery
tests are passing without errors. This code has gone through
41 * 3 successful gfs2 recovery tests before it hit an
unrelated (openais) problem.
This is a complete rewrite of patch 4 for bug #248176.
Part of the problem was that inodes were being recycled
before their buffers were flushed to the journal logs.
Another problem was that the clone bitmaps were being
searched for deleted inodes to recycle, but only the
"real" bitmaps should be searched for that purpose.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
We only need a single gfs2_scand process rather than the one
per filesystem which we had previously. As a result the parameter
determining the frequency of gfs2_scand runs becomes a module
parameter rather than a mount parameter as it was before.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
these struct *_operations are all method tables, thus should be const.
Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is patch 5 of 5 for bug #248176
Metadata corruption was occurring because page references weren't
being removed in all cases. I previously added a function called
detach_bufdata, but I discovered there already WAS a function out
there to do the job. It's called gfs2_meta_cache_flush. So I added
a call to that to remove the page references.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
this is more clear.
Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is patch three of five for bug #248176.
The try_rgrp_unlink code in rgrp.c had an infinite loop. This was
caused because the bitmap function rgblk_search can return a block
less than the "goal" block, in which case it was looping. The fix is
to make it always march forward as needed.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is patch 2 of 5 for bug #248176.
The list_move code previously concocted in log.c for bug #238162
(see https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=238162#c23)
never runs as bh can now never be NULL at this point.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is the first of five patches for bug #248176:
There were still some critical variables being manipulated outside
the log_lock spinlock. That usually resulted in a hang.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This fixes an oops which was occurring during glock dumping due to the
seq file code not taking a reference to the glock. Also this fixes a
memory leak which occurred in certain cases, in turn preventing the
filesystem from unmounting.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When looking at an unrelated problem, I noticed that nfsd does not
set nameidata pointer on create (ie nd is NULL). This should
cause an oops in some cases in which when NFSd is mounted over GFS2.
Signed-off-by: Steve French <sfrench@us.ibm.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch cleans up duplicate includes in
fs/gfs2/
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
If a glock is in the exclusive state and a request for demote to
deferred has been received, then further requests for demote to
shared are being ignored. This patch fixes that by ensuring that
we demote to unlocked in that case.
Signed-off-by: Josef Whiter <jwhiter@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
One of the races relates to referencing a variable while not holding
its protecting spinlock. The patch simply moves the test inside the
spin lock. The other races occurs when a demote to unlocked request
occurs during the time a demote to shared request is already running.
This of course only happens in the case that the lock was in the
exclusive mode to start with. The patch adds a check to see if another
demote request has occurred in the mean time and if it has, then it
performs a second demote.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The floppy ioctls are used by multiple drivers, so they should be
handled in a shared location. Also, add minor cleanups.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
These are shared by all cd-rom drivers and should have common
handlers. Do slight cosmetic cleanups in the process.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
BLKPG is common to all block devices, so it should be handled
by common code.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
These are common to multiple block drivers, so they should
be handled by the block layer.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
blk_trace_setup is broken on x86_64 compat systems,
this makes the code work correctly on all 64 bit architectures
in compat mode.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Handle those blockdev ioctl calls that are compatible
directly from the compat_blkdev_ioctl() function, instead
of having to go through the compat_ioctl hash lookup.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Make compat_blkdev_ioctl and blkdev_ioctl reflect the respective
native versions. This is somewhat more efficient and makes it easier
to keep the two in sync.
Also get rid of the bogus handling for broken_blkgetsize and the
duplicate entry for BLKRASET.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
As bi_end_io is only called once when the reqeust is complete,
the 'size' argument is now redundant. Remove it.
Now there is no need for bio_endio to subtract the size completed
from bi_size. So don't do that either.
While we are at it, change bi_end_io to return void.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
The only caller of bio_endio that does not pass the full bi_size
is end_that_request_first. Also, no ->bi_end_io method is really
interested in bi_size being decremented.
So move the decrement and related code into ll_rw_blk and merge it
with order_bio_endio to form req_bio_endio which does endio functionality
specific to request completion.
As some ->bi_end_io methods do check bi_size of 0, we set it thus for
now, but that will go in the next patch.
Signed-off-by: Neil Brown <neilb@suse.de>
### Diffstat output
./block/ll_rw_blk.c | 42 +++++++++++++++++++++++++++---------------
./fs/bio.c | 23 +++++++++++------------
2 files changed, 38 insertions(+), 27 deletions(-)
diff .prev/block/ll_rw_blk.c ./block/ll_rw_blk.c
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Currently bi_end_io can be called multiple times as sub-requests
complete. However no ->bi_end_io function wants to know about that.
So only call when the bio is complete.
Signed-off-by: Neil Brown <neilb@suse.de>
### Diffstat output
./fs/bio.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff .prev/fs/bio.c ./fs/bio.c
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Hide everything in blkdev.h with CONFIG_BLOCK isn't set, and fixup
the (few) files that fail to build because they were relying on blkdev.h
pulling in extra includes for them.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This macro is only used in one place; in this place it seems simpler to
put open-code it and move the comment to where it's used.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Currently /proc/locks is shown with a proc_read function, but its behavior
is rather complex as it has to manually handle current offset and buffer
length. On the other hand, files that show objects from lists can be
easily reimplemented using the sequential files and the seq_list_XXX()
helpers.
This saves (as usually) 16 lines of code and more than 200 from
the .text section.
[akpm@linux-foundation.org: no externs in C]
[akpm@linux-foundation.org: warning fixes]
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
fs/locks.c: use list_for_each_entry() instead of list_for_each() in
posix_locks_deadlock() and get_locks_status()
Signed-off-by: Matthias Kaehlcke <matthias.kaehlcke@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The __mandatory_lock(inode) macro makes the same check, but makes the code
more readable.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The __mandatory_lock(inode) macro makes the same check, but makes the code
more readable.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The __mandatory_lock(inode) macro makes the same check, but makes the code
more readable.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Eric Van Hensbergen <ericvh@gmail.com>
Cc: Ron Minnich <rminnich@sandia.gov>
Cc: Latchesar Ionkov <lucho@ionkov.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The __mandatory_lock(inode) function makes the same check, but makes the code
more readable.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The combination of S_ISGID bit set and S_IXGRP bit unset is used to mark the
inode as "mandatory lockable" and there's a macro for this check called
MANDATORY_LOCK(inode). However, fs/locks.c and some filesystems still perform
the explicit i_mode checking. Besides, Andrew pointed out, that this macro is
buggy itself, as it dereferences the inode arg twice.
Convert this macro into static inline function and switch its users to it,
making the code shorter and more readable.
The __mandatory_lock() helper is to be used in places where the IS_MANDLOCK()
for superblock is already known to be true.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Eric Van Hensbergen <ericvh@gmail.com>
Cc: Ron Minnich <rminnich@sandia.gov>
Cc: Latchesar Ionkov <lucho@ionkov.net>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
This code is run under lock_kernel(), which is dropped during
sleeping operations, so the following race is possible:
CPU1: CPU2:
vfs_setlease(); vfs_setlease();
lock_kernel();
lock_kernel(); /* spin */
generic_setlease():
...
for (before = ...)
/* here we found some lease after
* which we will insert the new one
*/
fl = locks_alloc_lock();
/* go to sleep in this allocation and
* drop the BKL
*/
generic_setlease():
...
for (before = ...)
/* here we find the "before" pointing
* at the one we found on CPU1
*/
->fl_change(my_before, arg);
lease_modify();
locks_free_lock();
/* and we freed it */
...
unlock_kernel();
locks_insert_lock(before, fl);
/* OOPS! We have just tried to add the lease
* at the tail of already removed one
*/
The similar races are already handled in other code - all the
allocations are performed before any checks/updates.
Thanks to Kamalesh Babulal for testing and for a bug report on an
earlier version.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Cc: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
This routine deletes all the elements from the list
with the "while (!list_empty())" loop, and we already
have a list_first_entry() macro to help it look nicer :)
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
This comment wasn't updated when lease support was added, and it makes
essentially the same mistake that the code made before a recent bugfix.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
When the flock_lock_file() is called to change the flock
from F_RDLCK to F_WRLCK or vice versa the existing flock
can be removed without appropriate warning.
Look:
for_each_lock(inode, before) {
struct file_lock *fl = *before;
if (IS_POSIX(fl))
break;
if (IS_LEASE(fl))
continue;
if (filp != fl->fl_file)
continue;
if (request->fl_type == fl->fl_type)
goto out;
found = 1;
locks_delete_lock(before); <<<<<< !
break;
}
if after this point the subsequent locks_alloc_lock() will
fail the return code will be -ENOMEM, but the existing lock
is already removed.
This is a known feature that such "re-locking" is not atomic,
but in the racy case the file should stay locked (although by
some other process), but in this case the file will be unlocked.
The proposal is to prepare the lock in advance keeping no chance
to fail in the future code.
Found during making the flocks pid-namespaces aware.
(Note: Thanks to Reuben Farrelly for finding a bug in an earlier version
of this patch.)
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Cc: Reuben Farrelly <reuben-linuxkernel@reub.net>
There's no need for another variable local to this loop; we can use the
variable (of the same name!) already declared at the top of the function,
and not used till later (at which point it's initialized, so this is safe).
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
The first argument to posix_locks_conflict() is meant to be a lock request,
and the second a lock from an inode's lock request. It doesn't really
make a difference which order you call them in, since the only
asymmetric test in posix_lock_conflict() is the check whether the second
argument is a posix lock--and every caller already does that check for
some reason.
But may as well fix posix_test_lock() to call posix_locks_conflict()
with the arguments in the same order as everywhere else.
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Without this we always return 2^32-1 as the the maximum namelength.
Thanks to Andreas Gruenbacher for bug report and testing.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Cc: Andreas Gruenbacher <agruen@suse.de>
It's not enough to take a reference on the delegation object itself; we
need to ensure that the rpc_client won't go away just as we're about to
make an rpc call.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
If a callback still holds a reference on the client, then it may be
about to perform an rpc call, so it isn't safe to call rpc_shutdown().
(Though rpc_shutdown() does wait for any outstanding rpc's, it can't
know if a new rpc is about to be issued with that client.)
So, wait to shutdown the rpc_client until the reference count on the
client has gone to zero.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Currently there's a race that can cause an oops in generic_setlease.
(In detail: nfsd, when it removes a lease, does so by calling
vfs_setlease() with F_UNLCK and a pointer to the fl_flock field, which
in turn points to nfsd's existing lease; but the first thing the
setlease code does is call time_out_leases(). If the lease happens to
already be beyond the lease break time, that will free the lease and (in
nfsd's release_private callback) set fl_flock to NULL, leading to a NULL
deference soon after in vfs_setlease().)
There are probably other things to fix here too, but it seems inherently
racy to allow either locks.c or nfsd to time out this lease. Instead
just set the fl_break_time to 0 (preventing locks.c from ever timing out
this lock) and leave it up to nfsd's laundromat thread to deal with it.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Modify the NFS server code to support 64 bit ino's, as
appropriate for the system and the NFS protocol version.
The gist of the changes is to query the underlying file system
for attributes and not just to use the cached attributes in the
inode. For this specific purpose, the inode only contains an
ino field which unsigned long, which is large enough on 64 bit
platforms, but is not large enough on 32 bit platforms.
I haven't been able to find any reason why ->getattr can't be called
while i_mutex. The specification indicates that i_mutex is not
required to be held in order to invoke ->getattr, but it doesn't say
that i_mutex can't be held while invoking ->getattr.
I also haven't come to any conclusions regarding the value of
lease_get_mtime() and whether it should or should not be invoked
by fill_post_wcc() too. I chose not to change this because I
thought that it was safer to leave well enough alone. If we
decide to make a change, it can be done separately.
Signed-off-by: Peter Staubach <staubach@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
Each branch of this if-then-else has a bunch of duplicated code that we
could just put at the end.
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
fs/nfsd/nfsctl.c: In function 'write_filehandle':
fs/nfsd/nfsctl.c:301: warning: 'maxsize' may be used uninitialized in this function
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
It doesn't make sense to make the callback with credentials that the
client made the setclientid with. Instead the spec requires that the
callback occur with the credentials the client authenticated *to*.
It probably doesn't matter what we use for auth_unix, and some more
infrastructure will be needed for auth_gss, so let's just remove the
cred lookup for now.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
We have some slabs that the nfs4 server uses to store state objects.
We're currently creating and destroying those slabs whenever the server
is brought up or down. That seems excessive; may as well just do that
in module initialization and exit.
Also add some minor header cleanup. (Thanks to Andrew Morton for that
and a compile fix.)
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
We want to allow gss on the callback channel, so people using krb5 can
still get the benefits of delegations.
But looking up the rpc credential can take some time in that case. And
we shouldn't delay the response to setclientid_confirm while we wait.
It may be inefficient, but for now the simplest solution is just to
spawn a new thread as necessary for the purpose.
(Thanks to Adrian Bunk for catching a missing static here.)
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Cc: Adrian Bunk <bunk@kernel.org>
Note that qword_get() returns length or -1, not an -ERROR.
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
To quote a recent mail from Andrew Morton:
Look: if there's a way in which an unprivileged user can trigger
a printk we fix it, end of story.
OK. I assume that goes double for printk()s that might be triggered by
random hosts on the internet. So, disable some printk()s that look like
they could be triggered by malfunctioning or malicious clients. For
now, just downgrade them to dprintk()s.
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
Benny Halevy suggested renaming cmp_* to same_* to make the meaning of
the return value clearer.
Fix some nearby style deviations while we're at it, including a small
swath of creative indentation in nfs4_preprocess_seqid_op().
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
I moved this check into map_new_errors, but forgot to delete the
original. Oops.
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
The nfserr_dropit happens routinely on upcalls (so a kmalloc failure is
almost never the actual cause), but I occasionally get a complant from
some tester that's worried because they ran across this message after
turning on debugging to research some unrelated problem.
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Acked-by: Neil Brown <neilb@suse.de>
This boot parameter will allow legacy 32-bit applications which call stat()
to continue to function even if the NFSv3/v4 server uses 64-bit inode
numbers.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
If the cached attributes match the ones supplied in the fattr, then assume
we've revalidated the inode.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
We're failing basic test6 against Linux servers because they lack a correct
change attribute. The fix is to assume that we always want to invalidate
the readdir caches when we call update_changeattr and/or
nfs_post_op_update_inode on a directory.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
nfs_post_op_update_inode() is really only meant to be used if we expect the
inode and its attributes to have changed in some way.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
- NFS_READTIME, NFS_CHANGE_ATTR are completely unused.
- Inline the few remaining uses of NFS_ATTRTIMEO, and remove.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Tejun Heo