Provide an internal interface for the removal of arbitrary file regions.

ocfs2_remove_inode_range() takes a byte range within a file and will remove
existing extents within that range. Partial clusters will be zeroed so that
any read from within the region will return zeros.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

The partial cluster zeroing code used during truncate usually assumes that
the rightmost byte in the range to be zeroed lies on a cluster boundary.
This makes sense for truncate, but punching holes might require zeroing on
non-aligned rightmost boundaries.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Add code to the btree paths to support the removal of arbitrary regions
within an existing extent. With proper higher level support this can be used
to "punch holes" in a file. Truncate (a special case of hole punching) could
also be converted to use these methods.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

This can now be trivially supported with re-use of our existing extend code.

ocfs2_allocate_unwritten_extents() takes a start offset and a byte length
and iterates over the inode, adding extents (marked as unwritten) until len
is reached. Existing extents are skipped over.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Update the write code to detect when the user is asking to write to an
unwritten extent. Like writing to a hole, we must zero the region between
the write and the cluster boundaries. Most of the existing cluster zeroing
logic can be re-used with some additional checks for the unwritten flag on
extent records.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

We can easily seperate out the write descriptor setup and manipulation
into helper functions.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Writes to a region marked as unwritten might result in a record split or
merge. We can support splits by making minor changes to the existing insert
code. Merges require left rotations which mostly re-use right rotation
support functions.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

The top level calls and logic for growing a tree can easily be abstracted
out of ocfs2_insert_extent() into a seperate function - ocfs2_grow_tree().

This allows future code to easily grow btrees when needed.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Now that we have a method to deallocate blocks from them, each node should
allocate extent blocks from their local suballocator file.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Deallocation of suballocator blocks, most notably extent blocks, might
involve multiple suballocator inodes.

The locking for this can get extremely complicated, especially when the
suballocator inodes to delete from aren't known until deep within an
unrelated codepath.

Implement a simple scheme for recording the blocks to be unlinked so that
the actual deallocation can be done in a context which won't deadlock.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

We don't want to submit buffer_new blocks for read i/o. This actually won't
happen right now because those requests during an allocating write are all nicely
aligned. It's probably a good idea to provide an explicit check though.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Implement cluster consistent shared writeable mappings using the
->page_mkwrite() callback.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ocfs2_mkwrite() will want this so that it can add some mmap specific checks
before asking for a write.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Use some ideas from the new-aops patch series and turn
ocfs2_buffered_write_cluster() into a 2 stage operation with the caller
copying data in between. The code now understands multiple cluster writes as
a result of having to deal with a full page write for greater than 4k pages.

This sets us up to easily call into the write path during ->page_mkwrite().
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Use of the alloc sem during truncate was too narrow - we want to protect
the i_size change and page truncation against mmap now.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ocfs2 will attempt to assign the node the slot# provided in the mount
option. Failure to assign the preferred slot is not an error. This small
feature can be useful for automated testing.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Replacing memset(<addr>,0,PAGE_SIZE) with clear_page() in
fs/ocfs2/dlm/dlmrecovery.c
Signed-off-by: NShani Moideen <shani.moideen@wipro.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Tell o2cb_region_dev_write() to wake up if rmdir(2) happens on the
heartbeat region while it is starting up.  Then o2hb_region_dev_write()
can check to see if it is alive and act accordingly.  This prevents a hang
(not being woken) and a crash (if it's woken by a signal).
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Removing the local node configuration out from underneath a running
heartbeat is "bad".  Provide an API in the ocfs2 nodemanager to request
a configfs dependancy on the local node, then use it in heartbeat.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ocfs2 mounts require a heartbeat region.  Use the new configfs_depend_item()
facility to actually depend on them so they can't go away from under us.

First, teach cluster/nodemanager.c to depend an item on the o2cb subsystem.
Then teach o2hb_register_callbacks to take a UUID and depend on the
appropriate region.  Finally, teach all users of o2hb to pass a UUID or
NULL if they don't require a pin.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Sometimes other drivers depend on particular configfs items.  For
example, ocfs2 mounts depend on a heartbeat region item.  If that
region item is removed with rmdir(2), the ocfs2 mount must BUG or go
readonly.  Not happy.

This provides two additional API calls: configfs_depend_item() and
configfs_undepend_item().  A client driver can call
configfs_depend_item() on an existing item to tell configfs that it is
depended on.  configfs will then return -EBUSY from rmdir(2) for that
item.  When the item is no longer depended on, the client driver calls
configfs_undepend_item() on it.

These API cannot be called underneath any configfs callbacks, as
they will conflict.  They can block and allocate.  A client driver
probably shouldn't calling them of its own gumption.  Rather it should
be providing an API that external subsystems call.

How does this work?  Imagine the ocfs2 mount process.  When it mounts,
it asks for a heart region item.  This is done via a call into the
heartbeat code.  Inside the heartbeat code, the region item is looked
up.  Here, the heartbeat code calls configfs_depend_item().  If it
succeeds, then heartbeat knows the region is safe to give to ocfs2.
If it fails, it was being torn down anyway, and heartbeat can gracefully
pass up an error.

[ Fixed some bad whitespace in configfs.txt. --Mark ]
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Add a notification callback, ops->disconnect_notify(). It has the same
prototype as ->drop_item(), but it will be called just before the item
linkage is broken. This way, configfs users who want to do work while
the object is still in the heirarchy have a chance.

Client drivers will still need to config_item_put() in their
->drop_item(), if they implement it.  They need do nothing in
->disconnect_notify().  They don't have to provide it if they don't
care.  But someone who wants to be notified before ci_parent is set to
NULL can now be notified.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Seems copied from sysfs, but I don't see a reason here nor there to use
a semaphore instead of a mutex. Convert.
Signed-off-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Convert the su_sem member of struct configfs_subsystem to a struct
mutex, as that's what it is. Also convert all the users and update
Documentation/configfs.txt and Documentation/configfs_example.c
accordingly.

[ Conflict in fs/dlm/config.c with commit
  3168b078 manually resolved. --Mark ]
Inspired-by: NSatyam Sharma <ssatyam@cse.iitk.ac.in>
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Configfs being based upon sysfs code, config_group_find_obj() is probably
so named because of the similar kset_find_obj() in sysfs. However,
"kobject"s in sysfs become "config_item"s in configfs, so let's call it
config_group_find_item() instead, for sake of uniformity, and make
corresponding change in the users of this function.

BTW a crucial difference between kset_find_obj and config_group_find_item
is in locking expectations. kset_find_obj does its locking by itself, but
config_group_find_item expects the *caller* to do the locking. The reason
for this: kset's have their own locks, config_group's don't but instead
rely on the subsystem mutex. And, subsystem needn't necessarily be around
when config_group_find_item() is called.

So let's state these locking semantics explicitly, and rectify the comment,
otherwise bugs could continue to occur in future, as they did in the past
(refer commit d82b8191e238 in gfs2-2.6-fixes.git).

[ I also took the opportunity to fix some bad whitespace and
double-empty lines. --Joel ]

[ Conflict in fs/dlm/config.c with commit
  3168b078 manually resolved. --Mark ]
Signed-off-by: NSatyam Sharma <ssatyam@cse.iitk.ac.in>
Cc: David Teigland <teigland@redhat.com>
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

fs/dlm/config.c contains a useful generic macro called __CONFIGFS_ATTR
that is similar to sysfs' __ATTR macro that makes defining attributes
easy for any user of configfs. Separate it out into configfs.h so that
other users (forthcoming in dynamic netconsole patchset) can use it too.
Signed-off-by: NSatyam Sharma <ssatyam@cse.iitk.ac.in>
Cc: David Teigland <teigland@redhat.com>
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

1. item.c:config_item_cleanup() is a private function (only called by
config_item_release() in same file). However, it is spuriously
exported in include/linux/configfs.h, so remove that export and make
it static in item.c. Also, it is no longer exported / interface
function, so no need to give comment for this function (the comment
was stating obvious thing, anyway).

2. Kernel-doc comment format does not allow empty line between end of
comment and start of function (declaration line). There were several
such spurious empty lines in item.c, so fix them.

  fs/configfs/item.c       |   15 +++------------
  include/linux/configfs.h |    1 -
  2 files changed, 3 insertions(+), 13 deletions(-)
Signed-off-by: NSatyam Sharma <ssatyam@cse.iitk.ac.in>
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

The attribute store/show code currently limits attributes at PAGE_SIZE.
This code comes from sysfs, where it still works that way.

However, PAGE_SIZE is not constant.  A 16k attribute string works on
ia64 but not on x86.  Really a subsystem shouldn't allow different
attribute sizes based on platform.

As such, limit all simple attributes to 4k.  This works on all
platforms, and is consistent with all current code.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc:
  mmc: at91_mci: fix hanging and rework to match flowcharts
  mmc: at91_mci typo
  sdhci: Fix "Unexpected interrupt" handling
  mmc: fix silly copy-and-paste error
  mmc: move layer init and workqueue to core file
  mmc: refactor host class handling
  mmc: refactor bus operations
  sdhci: add ene controller id
  mmc: bounce requests for simple hosts

* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: (40 commits)
  bonding/bond_main.c: make 2 functions static
  ps3: gigabit ethernet driver for PS3, take3
  [netdrvr] Fix dependencies for ax88796 ne2k clone driver
  eHEA: Capability flag for DLPAR support
  Remove sk98lin ethernet driver.
  sunhme.c:quattro_pci_find() must be __devinit
  bonding / ipv6: no addrconf for slaves separately from master
  atl1: remove write-only var in tx handler
  macmace: use "unsigned long flags;"
  Cleanup usbnet_probe() return value handling
  netxen: deinline and sparse fix
  eeprom_93cx6: shorten pulse timing to match spec (bis)
  phylib: Add Marvell 88E1112 phy id
  phylib: cleanup marvell.c a bit
  AX88796 network driver
  IOC3: Switch to pci refcounting safe APIs
  e100: Fix Tyan motherboard e100 not receiving IPMI commands
  QE Ethernet driver writes to wrong register to mask interrupts
  rrunner.c:rr_init() must be __devinit
  tokenring/3c359.c:xl_init() must be __devinit
  ...

* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev: (32 commits)
  [libata] sata_mv: print out additional chip info during probe
  [libata] Use ATA_UDMAx standard masks when filling driver's udma_mask info
  [libata] AHCI: Add support for Marvell AHCI-like chips (initially 6145)
  [libata] Clean up driver udma_mask initializers
  libata: Support chips with 64K PRD quirk
  Add a PCI ID for santa rosa's PATA controller.
  sata_sil24: sil24_interrupt() micro-optimisation
  Add irq_flags to struct pata_platform_info
  sata_promise: cleanups
  [libata] pata_ixp4xx: kill unused var
  ata_piix: fix pio/mwdma programming
  [libata] ahci: minor internal cleanups
  [ATA] Add named constant for ATAPI command DEVICE RESET
  [libata] sata_sx4, sata_via: minor documentation updates
  [libata] ahci: minor internal cleanups
  [libata] ahci: Factor out SATA port init into a separate function
  [libata] pata_sil680: minor cleanups from benh
  [libata] sata_sx4: named constant cleanup
  [libata] pata_ixp4xx: convert to new EH
  [libata] pdc_adma: Reorder initializers with a couple structs
  ...

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus: (62 commits)
  [MIPS] PNX8550: Cleanup proc code.
  [MIPS] WRPPMC: Fix build.
  [MIPS] Yosemite: Fix modpost warnings.
  [MIPS] Change names of local variables to silence sparse
  [MIPS] SB1: Fix modpost warning.
  [MIPS] PNX: Fix modpost warnings.
  [MIPS] Alchemy: Fix modpost warnings.
  [MIPS] Non-FPAFF: Fix warning.
  [MIPS] DEC: Fix modpost warning.
  [MIPS] MIPSsim: Enable MIPSsim virtual network driver.
  [MIPS] Delete Ocelot 3 support.
  [MIPS] remove LASAT Networks platforms support
  [MIPS] Early check for SMTC kernel on non-MT processor
  [MIPS] Add debugfs files to show fpuemu statistics
  [MIPS] Add some debugfs files to debug unaligned accesses
  [MIPS] rbtx4938: Fix secondary PCIC and glue internal NICs
  [MIPS] tc35815: Load MAC address via platform_device
  [MIPS] Move FPU affinity code into separate file.
  [MIPS] Make ioremap() work on TX39/49 special unmapped segment
  [MIPS] rbtx4938: Update and minimize defconfig
  ...

* git://git.infradead.org/~dwmw2/battery-2.6:
  [BATTERY] ds2760 W1 slave
  [BATTERY] One Laptop Per Child power/battery driver
  [BATTERY] Apple PMU driver
  [BATTERY] 1-Wire ds2760 chip battery driver
  [BATTERY] APM emulation driver for class batteries
  [BATTERY] pda_power platform driver
  [BATTERY] Universal power supply class (was: battery class)

* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6:
  [S390] vmlogrdr function annotation.
  [S390] s390: rename CPU_IDLE to S390_CPU_IDLE
  [S390] cio: Remove prototype for non-existing function cmf_reset().
  [S390] zcrypt: fix request timeout handling
  [S390] system call optimization.
  [S390] dasd: Avoid compile warnings on !CONFIG_DASD_PROFILE
  [S390] Remove volatile from atomic_t
  [S390] Program check in diag 210 under 31 bit
  [S390] Bogomips calculation for 64 bit.
  [S390] smp: Merge smp_count_cpus() and smp_get_save_areas().
  [S390] zcore: Fix __user annotation.
  [S390] fixed cdl-format detection.
  [S390] sclp: Test facility list before executing a service call.
  [S390] sclp: introduce some new interfaces.
  [S390] Fixed comment typo.
  [S390] vmcp cleanup

* git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw: (57 commits)
  [GFS2] Accept old format NFS filehandles
  [GFS2] Small fixes to logging code
  [DLM] dump more lock values
  [GFS2] Remove i_mode passing from NFS File Handle
  [GFS2] Obtaining no_formal_ino from directory entry
  [GFS2] git-gfs2-nmw-build-fix
  [GFS2] System won't suspend with GFS2 file system mounted
  [GFS2] remounting w/o acl option leaves acls enabled
  [GFS2] inode size inconsistency
  [DLM] Telnet to port 21064 can stop all lockspaces
  [GFS2] Fix gfs2_block_truncate_page err return
  [GFS2] Addendum to the journaled file/unmount patch
  [GFS2] Simplify multiple glock aquisition
  [GFS2] assertion failure after writing to journaled file, umount
  [GFS2] Use zero_user_page() in stuffed_readpage()
  [GFS2] Remove bogus '\0' in rgrp.c
  [GFS2] Journaled file write/unstuff bug
  [DLM] don't require FS flag on all nodes
  [GFS2] Fix deallocation issues
  [GFS2] return conflicts for GETLK
  ...

* 'splice-2.6.23' of git://git.kernel.dk/data/git/linux-2.6-block:
  pipe: add documentation and comments
  pipe: change the ->pin() operation to ->confirm()
  Remove remnants of sendfile()
  xip sendfile removal
  splice: completely document external interface with kerneldoc
  sendfile: remove bad_sendfile() from bad_file_ops
  shmem: convert to using splice instead of sendfile()
  relay: use splice_to_pipe() instead of open-coding the pipe loop
  pipe: allow passing around of ops private pointer
  splice: divorce the splice structure/function definitions from the pipe header
  splice: relay support
  sendfile: convert nfsd to splice_direct_to_actor()
  sendfile: convert nfs to using splice_read()
  loop: convert to using splice_direct_to_actor() instead of sendfile()
  splice: add void cookie to the actor data
  sendfile: kill generic_file_sendfile()
  sendfile: remove .sendfile from filesystems that use generic_file_sendfile()
  sys_sendfile: switch to using ->splice_read, if available
  vmsplice: add vmsplice-to-user support
  splice: abstract out actor data

* 'trivial-2.6.23' of git://git.kernel.dk/data/git/linux-2.6-block:
  Documentation/block/barrier.txt is not in sync with the actual code: - blk_queue_ordered() no longer has a gfp_mask parameter - blk_queue_ordered_locked() no longer exists - sd_prepare_flush() looks slightly different
  Use list_for_each_entry() instead of list_for_each() in the block device
  Make a "menuconfig" out of the Kconfig objects "menu, ..., endmenu",
  block/Kconfig already has its own "menuconfig" so remove these
  Use menuconfigs instead of menus, so the whole menu can be disabled at once
  cfq-iosched: fix async queue behaviour
  unexport bio_{,un}map_user
  Remove legacy CDROM drivers
  [PATCH] fix request->cmd == INT cases
  cciss: add new controller support for P700m
  [PATCH] Remove acsi.c
  [BLOCK] drop unnecessary bvec rewinding from flush_dry_bio_endio
  [PATCH] cdrom_sysctl_info fix
  blk_hw_contig_segment(): bad segment size checks
  [TRIVIAL PATCH] Kill blk_congestion_wait() stub for !CONFIG_BLOCK

Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Cc: Chad Tindel <ctindel@users.sourceforge.net>
Cc: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>