Remove unused function ext4_ext_check_cache() and merge the code back to
the ext4_ext_in_cache().
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Similar situation to that of __alloc_fd(); do not use unless you
really have to. You should not touch any descriptor table other
than your own; it's a sure sign of a really bad API design.
As with __alloc_fd(), you *must* use a first-class reference to
struct files_struct; something obtained by get_files_struct(some task)
(let alone direct task->files) will not do. It must be either
current->files, or obtained by get_files_struct(current) by the
owner of that sucker and given to you.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
At that point nobody can see us anyway; everything that
looks at files_fdtable(files) is separated from the
guts of put_files_struct(files) - either since files is
current->files or because we fetched it under task_lock()
and hadn't dropped that yet, or because we'd bumped
files->count while holding task_lock()...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Essentially, alloc_fd() in a files_struct we own a reference to.
Most of the time wanting to use it is a sign of lousy API
design (such as android/binder). It's *not* a general-purpose
interface; better that than open-coding its guts, but again,
playing with other process' descriptor table is a sign of bad
design.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
... except for one in android, where the check is different
and already done in caller. No need to recalculate rlimit
many times in alloc_fd() either.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
* do copy_to_user() before prepare_for_access_response(); that kills
the need in remove_access_response().
* don't do fd_install() until we are past the last possible failure
exit. Don't use sys_close() on cleanup side - just put_unused_fd()
and fput(). Less racy that way...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
don't mess with sys_close() if copy_to_user() fails; just postpone
fd_install() until we know it hasn't.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The only difference between autofs_dev_ioctl_fd_install() and
fd_install() is __set_close_on_exec() done by the latter. Just
use get_unused_fd_flags(O_CLOEXEC) to allocate the descriptor
and be done with that...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Using kmem_cache_zalloc() instead of kmem_cache_alloc() and memset().
spatch with a semantic match is used to found this problem.
(http://coccinelle.lip6.fr/)
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
As inode64 is the default option now, and was also made remountable
previously, inode32 can also be remounted on-the-fly when it is needed.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
To make inode32 a remountable option, xfs_set_inode32() should be able
to make a transition from inode64 option, disabling inode allocation on
higher AGs.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
With the changes made on xfs_set_inode64(), to make it behave as
xfs_set_inode32() (now leaving to the caller the responsibility to update
mp->m_maxagi), we use the return value of xfs_set_inode64() to update
mp->m_maxagi during remount.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Add xfs_set_inode32() to be used to enable inode32 allocation mode. this
will reduce the amount of duplicated code needed to mount/remount a
filesystem with inode32 option. This patch also changes
xfs_set_inode64() to return the maximum AG number that inodes can be
allocated instead of set mp->m_maxagi by itself, so that the behaviour
is the same as xfs_set_inode32(). This simplifies code that calls these
functions and needs to know the maximum AG that inodes can be allocated
in.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
since 64-bit inodes can be accessed while using inode32, and these can
also be used on 32-bit kernels, there is no reason to still keep inode32
as the default mount option. If the filesystem cannot handle 64bit
inode numbers (i.e CONFIG_LBDAF is not enabled and BITS_PER_LONG == 32),
XFS_MOUNT_SMALL_INUMS will still be set by default, so inode64 is not an
unconditional default value.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
xfs_ialloc_next_ag() currently resets m_agirotor when it is equal to
m_maxagi:
if (++mp->m_agirotor == mp->m_maxagi)
mp->m_agirotor = 0;
But, if for some reason mp->m_maxagi changes to a lower value than
current m_agirotor, this condition will never be true, causing
m_agirotor to exceed the maximum allowed value (m_maxagi).
This implies mainly during lookups for xfs_perag structs in its radix
tree, since the agno value used for the lookup is based on m_agirotor.
An out-of-range m_agirotor may cause a lookup failure which in case will
return NULL.
As an example, the value of m_maxagi is decreased during
inode64->inode32 remount process, case where I've found this problem.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Actually, there is no reason about why a user must umount and mount a
XFS filesystem to enable 'inode64' option. So, this patch makes this a
remountable option.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
ERRnoresource is an ERRSRV level (aka server-side) error and means "No
resources currently available for request". Currently that maps to POSIX
-ENOBUFS. No NT errors map to it currently.
NT_STATUS_INSUFFICIENT_RESOURCES and NT_STATUS_INSUFF_SERVER_RESOURCES
are also similar in meaning. Currently the client maps those to
ERRnomem, which maps to -ENOMEM in POSIX.
All of these mappings seem to be quite wrong to me and are confusing for
users. All of the above errors indicate problems on the server, not the
client. Reporting -ENOMEM or -ENOBUFS implies that the client is running
out of resources.
This patch changes those mappings. The NT_* errors are changed to map to
the SRV level ERRnoresource. That error is in turn changed to return
-EREMOTEIO which is the only POSIX error I could find that conveys that
something went wrong on the server. While we're at it, change the SMB2
equivalent error to return the same.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: Suresh Jayaraman <sjayaraman@suse.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Uninitialized extent may became initialized(parallel writeback task)
at any moment after we drop i_data_sem, so we have to recheck extent's
state after we hold page's lock and i_data_sem.
If we about to change page's mapping we must hold page's lock in order to
serialize other users.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Non-full list of bugs:
1) uninitialized extent optimization does not hold page's lock,
and simply replace brunches after that writeback code goes
crazy because block mapping changed under it's feets
kernel BUG at fs/ext4/inode.c:1434! ( 288'th xfstress)
2) uninitialized extent may became initialized right after we
drop i_data_sem, so extent state must be rechecked
3) Locked pages goes uptodate via following sequence:
->readpage(page); lock_page(page); use_that_page(page)
But after readpage() one may invalidate it because it is
uptodate and unlocked (reclaimer does that)
As result kernel bug at include/linux/buffer_head.c:133!
4) We call write_begin() with already opened stansaction which
result in following deadlock:
->move_extent_per_page()
->ext4_journal_start()-> hold journal transaction
->write_begin()
->ext4_da_write_begin()
->ext4_nonda_switch()
->writeback_inodes_sb_if_idle() --> will wait for journal_stop()
5) try_to_release_page() may fail and it does fail if one of page's bh was
pinned by journal
6) If we about to change page's mapping we MUST hold it's lock during entire
remapping procedure, this is true for both pages(original and donor one)
Fixes:
- Avoid (1) and (2) simply by temproraly drop uninitialized extent handling
optimization, this will be reimplemented later.
- Fix (3) by manually forcing page to uptodate state w/o dropping it's lock
- Fix (4) by rearranging existing locking:
from: journal_start(); ->write_begin
to: write_begin(); journal_extend()
- Fix (5) simply by checking retvalue
- Fix (6) by locking both (original and donor one) pages during extent swap
with help of mext_page_double_lock()
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Proper block swap for inodes with full journaling enabled is
truly non obvious task. In order to be on a safe side let's
explicitly disable it for now.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
- Remove usless checks, because it is too late to check that inode != NULL
at the moment it was referenced several times.
- Double lock routines looks very ugly and locking ordering relays on
order of i_ino, but other kernel code rely on order of pointers.
Let's make them simple and clean.
- check that inodes belongs to the same SB as soon as possible.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
When performing an online resize, we add a bunch of groups at one time
in ext4_flex_group_add, so in most cases a lot of group descriptors
will be in the same group block. But in the end of this function,
update_backups will be called for every group descriptor and the same
block will be copied and journalled again and again. It is really a
waste.
Fix things so we only update a particular bg descriptor block once and
skip subsequent updates of the same block.
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
bh_submit_read() is responsible for unlock bh on endio. In addition,
we need to use bh_uptodate_or_lock() to avoid races.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Wrap the use of TIOCSRS485 and TIOCGRS485 in #ifdef so that we avoid
adding undefined IOCTLs to the ioctl pointer list as compatible
ioctls.
This change was motivated by a build error on a MIPS build.
tree: git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty.git
tty-next
head: ac57e7f38e
commit: 84c3b84860 [10/16] compat_ioctl:
Add RS-485 IOCTLs to the list
config: mips-fuloong2e_defconfig
All related error/warning messages:
fs/compat_ioctl.c:869:1: error: 'TIOCSRS485' undeclared here (not in a
function)
fs/compat_ioctl.c:870:1: error: 'TIOCGRS485' undeclared here (not in a
function)
vim +869 fs/compat_ioctl.c
863 COMPATIBLE_IOCTL(TIOCSPGRP)
864 COMPATIBLE_IOCTL(TIOCGPGRP)
865 COMPATIBLE_IOCTL(TIOCGPTN)
866 COMPATIBLE_IOCTL(TIOCSPTLCK)
867 COMPATIBLE_IOCTL(TIOCSERGETLSR)
868 COMPATIBLE_IOCTL(TIOCSIG)
> 869 COMPATIBLE_IOCTL(TIOCSRS485)
870 COMPATIBLE_IOCTL(TIOCGRS485)
871 #ifdef TCGETS2
872 COMPATIBLE_IOCTL(TCGETS2)
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Jaeden Amero <jaeden.amero@ni.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
...and make the default cache=strict as promised for 3.7.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
and add missed increments of failed async read and write requests.
Signed-off-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <smfrench@gmail.com>
by making it __le64 rather than __u64 in FILE_AL_INFO structure.
Signed-off-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <smfrench@gmail.com>
Some trivial endian fixes for the SMB2 code. One
warning remains which I asked Pavel to look at.
Reviewed-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <smfrench@gmail.com>
Now that the merge of the remaining pieces needed for
SMB2 (SMB2.1 dialect) are in, and most test cases pass,
we can consider SMB2.1 EXPERIMENTAL rather than "BROKEN."
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Reviewed-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <smfrench@gmail.com>
With SMB2 support, update from version 1.79 to 2.0 to make
it easier for users to recognize which version has SMB2 support.
Signed-off-by: Steve French <sfrench@us.ibm.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
FL_CLOSE is quite common when you close a file on which you hold a
lock. The spurious "Unknown lock flags" message in cFYI is
confusing in this case.
Reported-by: Alexander Bokovoy <abokovoy@redhat.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Reviewed-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <sfrench@us.ibm.com>
The string for "unc=" in /proc/mounts needs to be escaped. The current
behaviour can create problems in cases when mounting a share starting
with a number.
example:
>mount -t cifs -o username=test,password=x vm140-31:/17000-test /mnt
>mount -o remount,password=x /mnt
mount error: could not resolve address for vm140-31x00-test: Unknown
error
The sub-string "\170" which is part of the unc for the mount above in
/proc/mounts is interpreted as character'x' in the case above. Escaping
the string fixes the problem.
Signed-off-by: Sachin Prabhu <sprabhu@redhat.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Rename inode pointers for better clarity. Move the d_instantiate call to
the end of the function to prevent other tasks from seeing it before
we've finished constructing it. Since we should have exclusive access to
the inode at this point, remove the spinlock around i_nlink update.
Reviewed-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Now we walk though cifsFileInfo's list for every incoming lease
break and look for an equivalent there. That approach misses lease
breaks that come just after an open response - we don't have time
to populate new cifsFileInfo structure to the list. Fix this by
adding new list of pending opens and look for a lease there if we
didn't find it in the list of cifsFileInfo structures.
Signed-off-by: Pavel Shilovsky <pshilovsky@etersoft.ru>
Signed-off-by: Steve French <sfrench@us.ibm.com>
When we have a file opened with read oplock and we are writing a data
to this file, we need to store the data in the cache and then send to
the server to ensure that the next read operation will get a coherent
data.
Also mark it as CONFIG_CIFS_SMB2 because it's more suitable for SMB2
code but can fix some CIFS problems too (when server delays sending
an oplock break after a write request). We can drop this ifdefs
dependence in future.
Signed-off-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Currently CIFS code accept read/write ops on mandatory locked area
when two processes use the same file descriptor - it's wrong.
Fix this by serializing io and brlock operations on the inode.
Signed-off-by: Pavel Shilovsky <pshilovsky@etersoft.ru>
Signed-off-by: Steve French <sfrench@us.ibm.com>
and allow several processes to walk through the lock list and read
can_cache_brlcks value if they are not going to modify them.
Signed-off-by: Pavel Shilovsky <pshilovsky@etersoft.ru>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Now we need to lock/unlock a spinlock while processing brlock ops
on the inode. Move brlocks of a fid to a separate list and attach
all such lists to the inode. This let us not hold a spinlock.
Signed-off-by: Pavel Shilovsky <pshilovsky@samba.org>
Now that we aren't abusing the kmap address space, there's no need for
this lock or to impose a limit on the rsize.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Replace the "marshal_iov" function with a "read_into_pages" function.
That function will copy the read data off the socket and into the
pages array, kmapping and reading pages one at a time.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
We'll need an array to put into a smb_rqst, so convert this into an array
instead of (ab)using the lru list_head.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Eventually, we're going to want to append a list of pages to
cifs_readdata instead of a list of kvecs. To prepare for that, turn
the kvec array allocation into a separate one and just keep a
pointer to it in the readdata.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Now that we're using TCP_CORK on the socket, there's no value in
continuting to support this option. Schedule it for removal in 3.9.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Now that we're not kmapping so much at once, there's no need to cap
the wsize at the amount that can be simultaneously kmapped.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
For now, none of the callers populate rq_pages. That will be done for
writes in a later patch. While we're at it, change the prototype of
setup_async_request not to need a return pointer argument. Just
return the pointer to the mid_q_entry or an ERR_PTR.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Use the smb_send_rqst helper function to kmap each page in the array
and update the hash for that chunk.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Add code that allows smb_send_rqst to send an array of pages after the
initial kvec array has been sent. For now, we simply kmap the page
array and send it using the standard smb_send_kvec function. Eventually,
we may want to convert this code to use kernel_sendpage under the hood
and avoid the kmap altogether for the page data.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
We want to send SMBs as "atomically" as possible. Prior to sending any
data on the socket, cork it to make sure that no non-full frames go
out. Afterward, uncork it to make sure all of the data gets pushed out
to the wire.
Note that this more or less renders the socket=TCP_NODELAY mount option
obsolete. When TCP_CORK and TCP_NODELAY are used on the same socket,
TCP_NODELAY is essentially ignored.
Acked-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Again, just a change in the arguments and some function renaming here.
In later patches, we'll change this code to deal with page arrays.
In this patch, we add a new smb_send_rqst wrapper and have smb_sendv
call that. Then we move most of the existing smb_sendv code into a new
function -- smb_send_kvec. This seems a little redundant, but later
we'll flesh this out to deal with arrays of pages.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
We need a way to represent a call to be sent on the wire that does not
require having all of the page data kmapped. Behold the smb_rqst struct.
This new struct represents an array of kvecs immediately followed by an
array of pages.
Convert the signing routines to use these structs under the hood and
turn the existing functions for this into wrappers around that. For now,
we're just changing these functions to take different args. Later, we'll
teach them how to deal with arrays of pages.
Reviewed-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Signed-off-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Use hmac-sha256 and rather than hmac-md5 that is used for CIFS/SMB.
Signature field in SMB2 header is 16 bytes instead of 8 bytes.
Automatically enable signing by client when requested by the server
when signing ability is available to the client.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Sachin Prabhu <sprabhu@redhat.com>
Signed-off-by: Pavel Shilovsky <piastryyy@gmail.com>
Signed-off-by: Steve French <smfrench@gmail.com>
Signed-off-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Steve French <smfrench@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
and make cifs_get_file_info(_unix) calls static.
Signed-off-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Steve French <smfrench@gmail.com>
This is help us to extend the code for future protocols that can use
another fid mechanism (as SMB2 that has it divided into two parts:
persistent and violatile).
Also rename variables and refactor the code around the changes.
Reviewed-by: Jeff Layton <jlayton@samba.org>
Signed-off-by: Pavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: Steve French <smfrench@gmail.com>
If a dirty GFS2 inode was being deleted but was in use by another node, its
metadata was not getting written out before GFS2 checked for dirty buffers in
gfs2_ail_flush(). GFS2 was relying on inode_go_sync() to write out the
metadata when the other node tried to free the file, but it failed the error
check before it got that far. This patch writes out the metadata before calling
gfs2_ail_flush()
Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
gfs2_ail_empty_gl() contains an "inline version" of gfs2_trans_begin(),
so it needs an explicit sb_start_intwrite() as well, to balance the
sb_end_intwrite() which will be called by gfs2_trans_end().
With this, xfstest 068 passes on lock_nolock local gfs2.
Without it, we reach a writer count of -1 and get stuck.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch fixes an infinite loop in gfs2_rbm_find that was introduced
by the previous patch. The problem occurred when the length was less
than 3 but the rbm block was byte-aligned, causing it to improperly
return a extent length of zero, which caused it to spin.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Tested-by: Barry Marson <bmarson@redhat.com>
With the recently added block reservation code, an additional function
was added to search for free blocks. This had a restriction of only being
able to search for aligned extents of free blocks. As a result the
allocation patterns when reserving blocks were suboptimal when the
existing allocation of blocks for an inode was not aligned to the same
boundary.
This patch resolves that problem by adding the ability for gfs2_rbm_find
to search for extents of a particular minimum size. We can then use
gfs2_rbm_find for both looking for reservations, and also looking for
free blocks on an individual basis when we actually come to do the
allocation later on. As a result we only need a single set of code
to deal with both situations.
The function gfs2_rbm_from_block() is moved up rgrp.c so that it
occurs before all of its callers.
Many thanks are due to Bob for helping track down the final issue in
this patch. That fix to the rb_tree traversal and to not share
block reservations from a dirctory to its children is included here.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
GFS2 uses i_mutex on its system quota inode to synchronize writes to
quota file. Since this is an internal inode to GFS2 (not part of directory
hiearchy or visible by user) we are safe to define locking rules for it. So
let's just get it its own locking class to make it clear.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch stops multiple block allocations if a nonzero
return code is received from gfs2_rbm_from_block. Without
this patch, if enough pressure is put on the file system,
you get a kernel warning quickly followed by:
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<ffffffffa04f47e8>] gfs2_alloc_blocks+0x2c8/0x880 [gfs2]
With this patch, things run normally.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When rgd->rd_free_clone is less than rgd->rd_reserved, the
unclaimed_blocks() calculation would wrap and produce
incorrect results. This patch checks for this condition
when this function is called from gfs2_mblk_search()
In addition, the use of this particular function in other
places in the code has been dropped by means of a general
clean up of gfs2_inplace_reserve(). This function is now
much easier to follow.
Also the setting of the rgd->rd_last_alloc field is corrected.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch improves the tracing of block reservations by
removing some corner cases and also providing more useful
detail in the traces.
A new field is added to the reservation structure to contain
the inode number. This is used since in certain contexts it is
not possible to access the inode itself to obtain this information.
As a result we can then display the inode number for all tracepoints
and also in case we dump the resource group.
The "del" tracepoint operation has been removed. This could be called
with the reservation rgrp set to NULL. That resulted in not printing
the device number, and thus making the information largely useless
anyway. Also, the conditional on the rgrp being NULL can then be
removed from the tracepoint. After this change, all the block
reservation tracepoint calls will be called with the rgrp information.
The existing ins,clm and tdel calls to the block reservation tracepoint
are sufficient to track the entire life of the block reservation.
In gfs2_block_alloc() the error detection is updated to print out
the inode number of the problematic inode. This can then be compared
against the information in the glock dump,tracepoints, etc.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When we get to the stage of allocating blocks, we know that the
resource group in question must contain enough free blocks, otherwise
gfs2_inplace_reserve() would have failed. So if we are left with only
free blocks which are reserved, then we must use those. This can happen
if another node has sneeked in and use some blocks reserved on this
node, for example. Generally this will happen very rarely and only
when the resouce group is nearly full.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The ->show_options() function for GFS2 was not correctly displaying
the value when statfs slow in in use.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Reported-by: Milos Jakubicek <xjakub@fi.muni.cz>
Use the rbm structure for gfs2_setbit() in order to simplify the
arguments to the function. We have to add a bool to control whether
the clone bitmap should be updated (if it exists) but otherwise it
is a more or less direct substitution.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Change the arguments to gfs2_testbit() so that it now just takes an
rbm specifying the position of the two bit entry to return.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Function gfs2_bitfit was checking for state > 3, but that's
impossible since it is only called from rgblk_search, which receives
only GFS2_BLKST_ constants.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Function add_to_queue was checking may_grant for the passed-in
holder for every iteration of its gh2 loop. Now it only checks it
once at the beginning to see if a try lock is futile.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Function gfs2_glock_dq_wait called two-line function wait_on_demote,
so they were combined.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Function gfs2_glock_wait only called function wait_on_holder and
returned its return code, so they were combined for readability.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Since function gfs2_glock_schedule_for_reclaim is only two
significant lines, we can eliminate it, simplifying the code
and making it more readable.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch changes function gfs2_direct_IO so that it uses a normal
call to gfs2_glock_dq rather than a call to a multiple-dq of one item.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch fixes a few small rbm related things. First, it fixes
a corner case where the rbm needs to switch bitmaps and wasn't
adjusting its buffer pointer. Second, there's a white space issue
fixed. Third, the logic in function gfs2_rbm_from_block was optimized
a bit. Lastly, a check for goal block overflows was added to function
gfs2_alloc_blocks.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
One corner case which the original patch failed to take into
account was when there is a reservation which ended such that
the following block was one beyond the end of the rgrp in
question. This extra test fixes that case.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Reported-by: Bob Peterson <rpeterso@redhat.com>
Tested-by: Bob Peterson <rpeterso@redhat.com>
gfs2 calls RB_EMPTY_NODE() to check if nodes are not on an rbtree.
The corresponding initialization function is RB_CLEAR_NODE().
rb_init_node() was never clearly defined and is going away.
Signed-off-by: Michel Lespinasse <walken@google.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is part of a series of patches which are introducing the
gfs2_rbm structure throughout the block allocation code. The
main aim of this part is to create a search function which can
deal directly with struct gfs2_rbm. In this case it specifies
the initial position at which to start the search and also the
point at which the search terminates.
The net result of this is to clean up the search code and make
it rather more readable, and the various possible exceptions which
may occur during the search are partitioned into their own functions.
There are some bug fixes too. We should not be checking the reservations
while allocating extents - the time for that is when we are searching
for where to put the extent, not when we've already made that decision.
Also, rgblk_search had two uses, and in only one of those cases did
it make sense to check for reservations. This is fixed in the new
gfs2_rbm_find function, which has a cleaner interface.
The reservation checking has been improved by always checking for
contiguous reservations, and returning the first free block after
all contiguous reservations. This is done under the spin lock to
ensure consistancy of the tree.
The allocation of extents is now in all cases done by the existing
allocation code, and if there is an active reservation, that is updated
after the fact. Again this is done under the spin lock, since it entails
changing the lookup key for the reservation in question.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch introduces a new structure, gfs2_rbm, which is a
tuple of a resource group, a bitmap within the resource group
and an offset within that bitmap. This is designed to make
manipulating these sets of variables easier. There is also a
new helper function which converts this representation back
to a disk block address.
In addition, the rbtree nodes which are used for the reservations
were not being correctly initialised, which is now fixed. Also,
the tracing was not passing through the inode where it should
have been. That is mostly fixed aside from one corner case. This
needs to be revisited since there can also be a NULL rgrp in
some cases which results in the device being incorrect in the
trace.
This is intended to be the first step towards cleaning up some
of the allocation code, and some further bug fixes.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The rs_requested field is left over from the original allocation
code, however this should have been a parameter passed to the
various functions from gfs2_inplace_reserve() and not a member of the
reservation structure as the value is not required after the
initial allocation.
This also helps simplify the code since we no longer need to set
the rs_requested to zero. Also the gfs2_inplace_release()
function can also be simplified since the reservation structure
will always be defined when it is called, and the only remaining
task is to unlock the rgrp if required. It can also now be
called unconditionally too, resulting in a further simplification.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
There were two functions in the xattr code which were nearly
identical, the only difference being that one was copy data into
the unstuffed xattrs and the other was copying data out from it.
This patch merges the two functions such that the code which deal
with iteration over the unstuffed xattrs is no longer duplicated.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Recently, I ecountered some corrupted filesystems in which some
groups' free inode counts were 65535, it seemed that free inode
count was overflow. This patch teaches ext4 to check free inode
count before allocaing an inode.
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Free block counters should be checked before doing allocation.
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
The crash was caused by a variable being erronously declared static in
token2str().
In addition to /proc/mounts, the problem can also be easily replicated
by accessing /proc/fs/ext4/<partition>/options in parallel:
$ cat /proc/fs/ext4/<partition>/options > options.txt
... and then running the following command in two different terminals:
$ while diff /proc/fs/ext4/<partition>/options options.txt; do true; done
This is also the cause of the following a crash while running xfstests
#234, as reported in the following bug reports:
https://bugs.launchpad.net/bugs/1053019https://bugzilla.kernel.org/show_bug.cgi?id=47731
Signed-off-by: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: Brad Figg <brad.figg@canonical.com>
Cc: stable@vger.kernel.org
normally we deal with lock_mount()/umount races by checking that
mountpoint to be is still in our namespace after lock_mount() has
been done. However, do_add_mount() skips that check when called
with MNT_SHRINKABLE in flags (i.e. from finish_automount()). The
reason is that ->mnt_ns may be a temporary namespace created exactly
to contain automounts a-la NFS4 referral handling. It's not the
namespace of the caller, though, so check_mnt() would fail here.
We still need to check that ->mnt_ns is non-NULL in that case,
though.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
- fix a regression related to xfs_sync_worker racing with unmount.
- fix a race while discarding xfs buffers.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)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=ZcGP
-----END PGP SIGNATURE-----
Merge tag 'for-linus-v3.6-rc7' of git://oss.sgi.com/xfs/xfs
Pull xfs bugfixes from Ben Myers:
- fix a regression related to xfs_sync_worker racing with unmount.
- fix a race while discarding xfs buffers.
* tag 'for-linus-v3.6-rc7' of git://oss.sgi.com/xfs/xfs:
xfs: stop the sync worker before xfs_unmountfs
xfs: fix race while discarding buffers [V4]
The format_array_alloc() function is fundamentally racy, in that it
prints the array twice: once to figure out how much space to allocate
for the buffer, and the second time to actually print out the data.
If any of the array contents changes in between, the allocation size may
be wrong, and the end result may be truncated in odd ways.
Just don't do it. Allocate a maximum-sized array up-front, and just
format the array contents once. The only user of the u32_array
interfaces is the Xen spinlock statistics code, and it has 31 entries in
the arrays, so the maximum size really isn't that big, and the end
result is much simpler code without the bug.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
u32_array_open() is racy when multiple threads read from a file with a
seek position of zero, i.e. when two or more simultaneous reads are
occurring after the non-seekable files are created. It is possible that
file->private_data is double-freed because the threads races between
kfree(file->private-data);
and
file->private_data = NULL;
The fix is to only do format_array_alloc() when the file is opened and
free it when it is closed.
Note that because the file has always been non-seekable, you can't open
it and read it multiple times anyway, so the data has always been
generated just once. The difference is that now it is generated at open
time rather than at the time of the first read, and that avoids the
race.
Reported-by: Dave Jones <davej@redhat.com>
Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Tested-by: Raghavendra <raghavendra.kt@linux.vnet.ibm.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The RS-485 TIOCSRS485 and TIOCGRS485 ioctls are 32-bit compatible, so
in order to call them on 64-bit systems from 32-bit user mode, we add
them to the ioctl pointer list as compatible ioctls.
Signed-off-by: Jaeden Amero <jaeden.amero@ni.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
- General routine uid/gid conversion work
- When storing posix acls treat ACL_USER and ACL_GROUP separately
so I can call from_kuid or from_kgid as appropriate.
- When reading posix acls treat ACL_USER and ACL_GROUP separately
so I can call make_kuid or make_kgid as appropriate.
Cc: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: Chris Mason <chris.mason@fusionio.com>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: "Tigran A. Aivazian" <tigran@aivazian.fsnet.co.uk>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Don't make the security modules deal with raw user space uid and
gids instead pass in a kuid_t and a kgid_t so that security modules
only have to deal with internal kernel uids and gids.
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: James Morris <james.l.morris@oracle.com>
Cc: John Johansen <john.johansen@canonical.com>
Cc: Kentaro Takeda <takedakn@nttdata.co.jp>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Acked-by: Bob Copeland <me@bobcopeland.com>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Note sysctl_hugetlb_shm_group can only be written in the root user
in the initial user namespace, so we can assume sysctl_hugetlb_shm_group
is in the initial user namespace.
Cc: William Irwin <wli@holomorphy.com>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
The action field has been merged into struct btrfs_delayed_ref_node,
and no struct btrfs_delayed_ref is available now.
Signed-off-by: Wang Sheng-Hui <shhuiw@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Like 8250 driver, when pstore is registered as a console,
to avoid recursive spinlocks when panic happening, change the
spin_lock_irqsave to spin_trylock_irqsave when oops_in_progress
is true.
Signed-off-by: liu chuansheng <chuansheng.liu@intel.com>
Signed-off-by: Anton Vorontsov <anton.vorontsov@linaro.org>
The update_backups() function is used to backup all the metadata
blocks, so we should not take it for granted that 'data' is pointed to
a super block and use ext4_superblock_csum_set to calculate the
checksum there. In case where the data is a group descriptor block,
it will corrupt the last group descriptor, and then e2fsck will
complain about it it.
As all the metadata checksums should already be OK when we do the
backup, remove the wrong ext4_superblock_csum_set and it should be
just fine.
Reported-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
In ext4_nonda_switch(), if the file system is getting full we used to
call writeback_inodes_sb_if_idle(). The problem is that we can be
holding i_mutex already, and this causes a potential deadlock when
writeback_inodes_sb_if_idle() when it tries to take s_umount. (See
lockdep output below).
As it turns out we don't need need to hold s_umount; the fact that we
are in the middle of the write(2) system call will keep the superblock
pinned. Unfortunately writeback_inodes_sb() checks to make sure
s_umount is taken, and the VFS uses a different mechanism for making
sure the file system doesn't get unmounted out from under us. The
simplest way of dealing with this is to just simply grab s_umount
using a trylock, and skip kicking the writeback flusher thread in the
very unlikely case that we can't take a read lock on s_umount without
blocking.
Also, we now check the cirteria for kicking the writeback thread
before we decide to whether to fall back to non-delayed writeback, so
if there are any outstanding delayed allocation writes, we try to get
them resolved as soon as possible.
[ INFO: possible circular locking dependency detected ]
3.6.0-rc1-00042-gce894ca #367 Not tainted
-------------------------------------------------------
dd/8298 is trying to acquire lock:
(&type->s_umount_key#18){++++..}, at: [<c02277d4>] writeback_inodes_sb_if_idle+0x28/0x46
but task is already holding lock:
(&sb->s_type->i_mutex_key#8){+.+...}, at: [<c01ddcce>] generic_file_aio_write+0x5f/0xd3
which lock already depends on the new lock.
2 locks held by dd/8298:
#0: (sb_writers#2){.+.+.+}, at: [<c01ddcc5>] generic_file_aio_write+0x56/0xd3
#1: (&sb->s_type->i_mutex_key#8){+.+...}, at: [<c01ddcce>] generic_file_aio_write+0x5f/0xd3
stack backtrace:
Pid: 8298, comm: dd Not tainted 3.6.0-rc1-00042-gce894ca #367
Call Trace:
[<c015b79c>] ? console_unlock+0x345/0x372
[<c06d62a1>] print_circular_bug+0x190/0x19d
[<c019906c>] __lock_acquire+0x86d/0xb6c
[<c01999db>] ? mark_held_locks+0x5c/0x7b
[<c0199724>] lock_acquire+0x66/0xb9
[<c02277d4>] ? writeback_inodes_sb_if_idle+0x28/0x46
[<c06db935>] down_read+0x28/0x58
[<c02277d4>] ? writeback_inodes_sb_if_idle+0x28/0x46
[<c02277d4>] writeback_inodes_sb_if_idle+0x28/0x46
[<c026f3b2>] ext4_nonda_switch+0xe1/0xf4
[<c0271ece>] ext4_da_write_begin+0x27/0x193
[<c01dcdb0>] generic_file_buffered_write+0xc8/0x1bb
[<c01ddc47>] __generic_file_aio_write+0x1dd/0x205
[<c01ddce7>] generic_file_aio_write+0x78/0xd3
[<c026d336>] ext4_file_write+0x480/0x4a6
[<c0198c1d>] ? __lock_acquire+0x41e/0xb6c
[<c0180944>] ? sched_clock_cpu+0x11a/0x13e
[<c01967e9>] ? trace_hardirqs_off+0xb/0xd
[<c018099f>] ? local_clock+0x37/0x4e
[<c0209f2c>] do_sync_write+0x67/0x9d
[<c0209ec5>] ? wait_on_retry_sync_kiocb+0x44/0x44
[<c020a7b9>] vfs_write+0x7b/0xe6
[<c020a9a6>] sys_write+0x3b/0x64
[<c06dd4bd>] syscall_call+0x7/0xb
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
Do not iterate over data blocks scanning for bh's to forget as they're
never exist. This improves time taken by unlink / truncate syscall.
Tested by continuously truncating file that is being written by dd.
Another test is rm -rf of linux tree while tar unpacks it. With
ordered data mode condition unlikely(!tbh) was always met in
ext4_free_blocks. With journal data mode tbh was found only few times,
so optimisation is also possible.
Unlinking fallocated 60G file after doing sync && echo 3 >
/proc/sys/vm/drop_caches && time rm --help
X86 before (linux 3.6-rc4):
# time rm -f test1
real 0m2.710s
user 0m0.000s
sys 0m1.530s
X86 after:
# time rm -f test1
real 0m0.644s
user 0m0.003s
sys 0m0.060s
MIPS before (linux 2.6.37):
# time rm -f test1
real 0m 4.93s
user 0m 0.00s
sys 0m 4.61s
MIPS after:
# time rm -f test1
real 0m 0.16s
user 0m 0.00s
sys 0m 0.06s
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Andrey Sidorov <qrxd43@motorola.com>
Commit 1c6bd7173d introduced a regression where an online resize
operation which did not change the number of block groups would fail,
i.e:
mke2fs -t /dev/vdc 60000
mount /dev/vdc
resize2fs /dev/vdc 60001
This was due to a bug in the logic regarding when to try converting
the filesystem to use meta_bg.
Also fix up a number of other minor issues with the online resizing
code: (a) Fix a sparse warning; (b) only check to make sure the device
is large enough once, instead of multiple times through the resize
loop.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cancel work of the xfs_sync_worker before teardown of the log in
xfs_unmountfs. This prevents occasional crashes on unmount like so:
PID: 21602 TASK: ee9df060 CPU: 0 COMMAND: "kworker/0:3"
#0 [c5377d28] crash_kexec at c0292c94
#1 [c5377d80] oops_end at c07090c2
#2 [c5377d98] no_context at c06f614e
#3 [c5377dbc] __bad_area_nosemaphore at c06f6281
#4 [c5377df4] bad_area_nosemaphore at c06f629b
#5 [c5377e00] do_page_fault at c070b0cb
#6 [c5377e7c] error_code (via page_fault) at c070892c
EAX: f300c6a8 EBX: f300c6a8 ECX: 000000c0 EDX: 000000c0 EBP: c5377ed0
DS: 007b ESI: 00000000 ES: 007b EDI: 00000001 GS: ffffad20
CS: 0060 EIP: c0481ad0 ERR: ffffffff EFLAGS: 00010246
#7 [c5377eb0] atomic64_read_cx8 at c0481ad0
#8 [c5377ebc] xlog_assign_tail_lsn_locked at f7cc7c6e [xfs]
#9 [c5377ed4] xfs_trans_ail_delete_bulk at f7ccd520 [xfs]
#10 [c5377f0c] xfs_buf_iodone at f7ccb602 [xfs]
#11 [c5377f24] xfs_buf_do_callbacks at f7cca524 [xfs]
#12 [c5377f30] xfs_buf_iodone_callbacks at f7cca5da [xfs]
#13 [c5377f4c] xfs_buf_iodone_work at f7c718d0 [xfs]
#14 [c5377f58] process_one_work at c024ee4c
#15 [c5377f98] worker_thread at c024f43d
#16 [c5377fbc] kthread at c025326b
#17 [c5377fe8] kernel_thread_helper at c070e834
PID: 26653 TASK: e79143b0 CPU: 3 COMMAND: "umount"
#0 [cde0fda0] __schedule at c0706595
#1 [cde0fe28] schedule at c0706b89
#2 [cde0fe30] schedule_timeout at c0705600
#3 [cde0fe94] __down_common at c0706098
#4 [cde0fec8] __down at c0706122
#5 [cde0fed0] down at c025936f
#6 [cde0fee0] xfs_buf_lock at f7c7131d [xfs]
#7 [cde0ff00] xfs_freesb at f7cc2236 [xfs]
#8 [cde0ff10] xfs_fs_put_super at f7c80f21 [xfs]
#9 [cde0ff1c] generic_shutdown_super at c0333d7a
#10 [cde0ff38] kill_block_super at c0333e0f
#11 [cde0ff48] deactivate_locked_super at c0334218
#12 [cde0ff58] deactivate_super at c033495d
#13 [cde0ff68] mntput_no_expire at c034bc13
#14 [cde0ff7c] sys_umount at c034cc69
#15 [cde0ffa0] sys_oldumount at c034ccd4
#16 [cde0ffb0] system_call at c0707e66
commit 11159a05 added this to xfs_log_unmount and needs to be cleaned up
at a later date.
Signed-off-by: Ben Myers <bpm@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
This function returns the wrong value, which causes the callers to get
the length of the resulting pathname wrong when it contains non-ASCII
characters.
This seems to fix https://bugzilla.samba.org/show_bug.cgi?id=6767
Cc: <stable@vger.kernel.org>
Reported-by: Baldvin Kovacs <baldvin.kovacs@gmail.com>
Reported-and-Tested-by: Nicolas Lefebvre <nico.lefebvre@gmail.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
IBM reported a soft lockup after applying the fix for the rename_lock
deadlock. Commit c83ce989cb ("VFS: Fix the nfs sillyrename regression
in kernel 2.6.38") was found to be the culprit.
The nfs sillyrename fix used DCACHE_DISCONNECTED to indicate that the
dentry was killed. This flag can be set on non-killed dentries too,
which results in infinite retries when trying to traverse the dentry
tree.
This patch introduces a separate flag: DCACHE_DENTRY_KILLED, which is
only set in d_kill() and makes try_to_ascend() test only this flag.
IBM reported successful test results with this patch.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Instead of checking whether the handle is valid, we check if journal
is enabled. This avoids taking the s_orphan_lock mutex in all cases
when there is no journal in use, including the error paths where
ext4_orphan_del() is called with a handle set to NULL.
Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This is a revert of commit b56ff9d397, which removed the call to
ext4_issue_discard() to fix a BUG reported because
ext4_issue_discard() was being called from inside a block group
spinlock. As it turns out this bug had already been fixed by Lukas
Czerner in commit 53fdcf992d by the simple expedient of moving when
we call ext4_issue_discard() outside the spinlock.
So it should be safe to re-enable this functionality, which I tested
by putting an BUG_ON(in_atomic) just after the restored callsite to
ext4_issue_discard().
Addresses-Google-Bug: #6750518
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: Anatol Pomozov <anatol.pomozov@gmail.com>
In a hasty fix to replace a 64-bit division with do_div, I
unintentionally assigned the divisor to a 32-bit variable.
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
Cc: Tino Reichardt <milky-kernel@mcmilk.de>
Now that the type changes are done, here is the final set of
changes to make the quota code work when user namespaces are enabled.
Small cleanups and fixes to make the code build when user namespaces
are enabled.
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Convert w_dq_id to be a struct kquid and remove the now unncessary
w_dq_type.
This is a simple conversion and enough other places have already
been converted that this actually reduces the code complexity
by a little bit, when removing now unnecessary type conversions.
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Change struct dquot dq_id to a struct kqid and remove the now
unecessary dq_type.
Make minimal changes to dquot, quota_tree, quota_v1, quota_v2, ext3,
ext4, and ocfs2 to deal with the change in quota structures and
signatures. The ocfs2 changes are larger than most because of the
extensive tracing throughout the ocfs2 quota code that prints out
dq_id.
quota_tree.c:get_index is modified to take a struct kqid instead of a
qid_t because all of it's callers pass in dquot->dq_id and it allows
me to introduce only a single conversion.
The rest of the changes are either just replacing dq_type with dq_id.type,
adding conversions to deal with the change in type and occassionally
adding qid_eq to allow quota id comparisons in a user namespace safe way.
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Theodore Tso <tytso@mit.edu>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Modify dqget to take struct kqid instead of a type and an identifier
pair.
Modify the callers of dqget in ocfs2 and dquot to take generate
a struct kqid so they can continue to call dqget. The conversion
to create struct kqid should all be the final conversions that
are needed in those code paths.
Cc: Jan Kara <jack@suse.cz>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Modify quota_send_warning to take struct kqid instead a type and
identifier pair.
When sending netlink broadcasts always convert uids and quota
identifiers into the intial user namespace. There is as yet no way to
send a netlink broadcast message with different contents to receivers
in different namespaces, so for the time being just map all of the
identifiers into the initial user namespace which preserves the
current behavior.
Change the callers of quota_send_warning in gfs2, xfs and dquot
to generate a struct kqid to pass to quota send warning. When
all of the user namespaces convesions are complete a struct kqid
values will be availbe without need for conversion, but a conversion
is needed now to avoid needing to convert everything at once.
Cc: Ben Myers <bpm@sgi.com>
Cc: Alex Elder <elder@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Update the quotactl user space interface to successfull compile with
user namespaces support enabled and to hand off quota identifiers to
lower layers of the kernel in struct kqid instead of type and qid
pairs.
The quota on function is not converted because while it takes a quota
type and an id. The id is the on disk quota format to use, which
is something completely different.
The signature of two struct quotactl_ops methods were changed to take
struct kqid argumetns get_dqblk and set_dqblk.
The dquot, xfs, and ocfs2 implementations of get_dqblk and set_dqblk
are minimally changed so that the code continues to work with
the change in parameter type.
This is the first in a series of changes to always store quota
identifiers in the kernel in struct kqid and only use raw type and qid
values when interacting with on disk structures or userspace. Always
using struct kqid internally makes it hard to miss places that need
conversion to or from the kernel internal values.
Cc: Jan Kara <jack@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Ben Myers <bpm@sgi.com>
Cc: Alex Elder <elder@kernel.org>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Add the data type struct kqid which holds the kernel internal form of
the owning identifier of a quota. struct kqid is a replacement for
the implicit union of uid, gid and project id stored in an unsigned
int and the quota type field that is was used in the quota data
structures. Making the data type explicit allows the kuid_t and
kgid_t type safety to propogate more thoroughly through the code,
revealing more places where uid/gid conversions need be made.
Along with the data type struct kqid comes the helper functions
qid_eq, qid_lt, from_kqid, from_kqid_munged, qid_valid, make_kqid,
make_kqid_invalid, make_kqid_uid, make_kqid_gid.
Cc: Jan Kara <jack@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Implement kprojid_t a cousin of the kuid_t and kgid_t.
The per user namespace mapping of project id values can be set with
/proc/<pid>/projid_map.
A full compliment of helpers is provided: make_kprojid, from_kprojid,
from_kprojid_munged, kporjid_has_mapping, projid_valid, projid_eq,
projid_eq, projid_lt.
Project identifiers are part of the generic disk quota interface,
although it appears only xfs implements project identifiers currently.
The xfs code allows anyone who has permission to set the project
identifier on a file to use any project identifier so when
setting up the user namespace project identifier mappings I do
not require a capability.
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Convert ext2, ext3, and ext4 to fully support the posix acl changes,
using e_uid e_gid instead e_id.
Enabled building with posix acls enabled, all filesystems supporting
user namespaces, now also support posix acls when user namespaces are enabled.
Cc: Theodore Tso <tytso@mit.edu>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Jan Kara <jack@suse.cz>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
