[Summary]

 Split LRU-list of unused dentries to one per superblock to avoid soft
 lock up during NFS mounts and remounting of any filesystem.

 Previously I posted here:
 http://lkml.org/lkml/2008/3/5/590

[Descriptions]

- background

  dentry_unused is a list of dentries which are not referenced.
  dentry_unused grows up when references on directories or files are
  released.  This list can be very long if there is huge free memory.

- the problem

  When shrink_dcache_sb() is called, it scans all dentry_unused linearly
  under spin_lock(), and if dentry->d_sb is differnt from given
  superblock, scan next dentry.  This scan costs very much if there are
  many entries, and very ineffective if there are many superblocks.

  IOW, When we need to shrink unused dentries on one dentry, but scans
  unused dentries on all superblocks in the system.  For example, we scan
  500 dentries to unmount a filesystem, but scans 1,000,000 or more unused
  dentries on other superblocks.

  In our case , At mounting NFS*, shrink_dcache_sb() is called to shrink
  unused dentries on NFS, but scans 100,000,000 unused dentries on
  superblocks in the system such as local ext3 filesystems.  I hear NFS
  mounting took 1 min on some system in use.

* : NFS uses virtual filesystem in rpc layer, so NFS is affected by
  this problem.

  100,000,000 is possible number on large systems.

  Per-superblock LRU of unused dentried can reduce the cost in
  reasonable manner.

- How to fix

  I found this problem is solved by David Chinner's "Per-superblock
  unused dentry LRU lists V3"(1), so I rebase it and add some fix to
  reclaim with fairness, which is in Andrew Morton's comments(2).

  1) http://lkml.org/lkml/2006/5/25/318
  2) http://lkml.org/lkml/2006/5/25/320

  Split LRU-list of unused dentries to each superblocks.  Then, NFS
  mounting will check dentries under a superblock instead of all.  But
  this spliting will break LRU of dentry-unused.  So, I've attempted to
  make reclaim unused dentrins with fairness by calculate number of
  dentries to scan on this sb based on following way

  number of dentries to scan on this sb =
  count * (number of dentries on this sb / number of dentries in the machine)

- ToDo
 - I have to measuring performance number and do stress tests.

 - When unmount occurs during prune_dcache(), scanning on same
  superblock, It is unable to reach next superblock because it is gone
  away.  We restart scannig superblock from first one, it causes
  unfairness of reclaim unused dentries on first superblock.  But I think
  this happens very rarely.

- Test Results

  Result on 6GB boxes with excessive unused dentries.

Without patch:

$ cat /proc/sys/fs/dentry-state
10181835        10180203        45      0       0       0
# mount -t nfs 10.124.60.70:/work/kernel-src nfs
real    0m1.830s
user    0m0.001s
sys     0m1.653s

 With this patch:
$ cat /proc/sys/fs/dentry-state
10236610        10234751        45      0       0       0
# mount -t nfs 10.124.60.70:/work/kernel-src nfs
real    0m0.106s
user    0m0.002s
sys     0m0.032s

[akpm@linux-foundation.org: fix comments]
Signed-off-by: NKentaro Makita <k-makita@np.css.fujitsu.com>
Cc: Neil Brown <neilb@suse.de>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: David Chinner <dgc@sgi.com>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The double indirection here is not needed anywhere and hence (at least)
confusing.
Signed-off-by: NJan Beulich <jbeulich@novell.com>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: "David S. Miller" <davem@davemloft.net>
Acked-by: NJeremy Fitzhardinge <jeremy@goop.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

In order to be able to debug things like the X server and programs using
the PPC Cell SPUs, the debugger needs to be able to access device memory
through ptrace and /proc/pid/mem.

This patch:

Add the generic_access_phys access function and put the hooks in place
to allow access_process_vm to access device or PPC Cell SPU memory.

[riel@redhat.com: Add documentation for the vm_ops->access function]
Signed-off-by: NRik van Riel <riel@redhat.com>
Signed-off-by: NBenjamin Herrensmidt <benh@kernel.crashing.org>
Cc: Dave Airlie <airlied@linux.ie>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Acked-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

There are no users of nopfn in the tree. Remove it.

[hugh@veritas.com: fix build error]
Signed-off-by: NNick Piggin <npiggin@suse.de>
Signed-off-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch adds proper extern declarations for five variables in
include/linux/vmstat.h
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Two zonelist patch series rewrote __page_alloc() largely.  Now, it is just
a wrapper function.  Inlining them will save a function call.

[akpm@linux-foundation.org: export __alloc_pages_internal]
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NKOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This function has no external callers, so unexport it.  Also fix its naming
inconsistency.
Signed-off-by: NJohannes Weiner <hannes@saeurebad.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Andy Whitcroft <apw@shadowen.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

There are a lot of places that define either a single bootmem descriptor or an
array of them.  Use only one central array with MAX_NUMNODES items instead.
Signed-off-by: NJohannes Weiner <hannes@saeurebad.de>
Acked-by: NRalf Baechle <ralf@linux-mips.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Hirokazu Takata <takata@linux-m32r.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Kyle McMartin <kyle@parisc-linux.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

lib/debugobjects.c has a function to test if an object is on the stack.
The block layer and ide needs it (they need to avoid DMA from/to stack
buffers).  This patch moves the function to include/linux/sched.h so that
everyone can use it.

lib/debugobjects.c uses current->stack but this patch uses a
task_stack_page() accessor, which is a preferable way to access the stack.
Signed-off-by: NFUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Andy Whitcroft <apw@shadowen.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

James Bottomley warns that inclusion of linux/fs.h in a low level
driver was always a danger signal.  This patch moves
memory_read_from_buffer() from fs.h to string.h and fixes includes in
existing memory_read_from_buffer() users.
Signed-off-by: NAkinobu Mita <akinobu.mita@gmail.com>
Cc: James Bottomley <James.Bottomley@hansenpartnership.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Bob Moore <robert.moore@intel.com>
Cc: Thomas Renninger <trenn@suse.de>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

There should be no functionality change resulting from this patch.
Signed-off-by: NBorislav Petkov <petkovbb@gmail.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

while at it, remove PC_FLAG_ZIP_DRIVE from the packed command flags altogether
and query the drive type through drive->atapi_flags.

v2:
ide-floppy fix.

There should be no functionality change resulting from this patch.

[bart: IDE_FLAG_* -> IDE_AFLAG_*, dev_flags -> atapi_flags]
Signed-off-by: NBorislav Petkov <petkovbb@gmail.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Push device flags up into ide_drive_t.

There should be no functionality change resulting from this patch.

[bart: IDE_FLAG_* -> IDE_AFLAG_*, dev_flags -> atapi_flags]
Signed-off-by: NBorislav Petkov <petkovbb@gmail.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

There should be no functionality change resulting from this patch.
Signed-off-by: NBorislav Petkov <petkovbb@gmail.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Refrain from carrying the callback ptr with every packet command since the
callback function is only one anyways. ide_drive_t is probably not the most
suitable place for it right now but is the more sane solution. Besides, these
structs are going to be reorganized anyways during the generic ide rewrite.
Signed-off-by: NBorislav Petkov <petkovbb@gmail.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Add ide_host_free() helper and convert ide_host_remove() to use it.

* Fix handling of ide_host_register() failure in ide_host_add(),
  icside.c, ide-generic.c, falconide.c and sgiioc4.c.

While at it:

* Fix handling of ide_host_alloc_all() failure in ide-generic.c.

* Fix handling of ide_host_alloc() failure in falconide.c
  (also return the correct error value if no device is found).

v2:
* falconide build fix. (From Stephen Rothwell)
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Add ide_host_add() helper which does ide_host_alloc()+ide_host_register(),
then convert ide_setup_pci_device[s](), ide_legacy_device_add() and some
host drivers to use it.

While at it:

* Fix ide_setup_pci_device[s](), ide_arm.c, gayle.c, ide-4drives.c,
  macide.c, q40ide.c, cmd640.c and cs5520.c to return correct error value.

* -ENOENT -> -ENOMEM in rapide.c, ide-h8300.c, ide-generic.c, au1xxx-ide.c
  and pmac.c

* -ENODEV -> -ENOMEM in palm_bk3710.c, ide_platform.c and delkin_cb.c

* -1 -> -ENOMEM in ide-pnp.c
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Add struct ide_host which keeps pointers to host's ports.

* Add ide_host_alloc[_all]() and ide_host_remove() helpers.

* Pass 'struct ide_host *host' instead of 'u8 *idx' to
  ide_device_add[_all]() and rename it to ide_host_register[_all]().

* Convert host drivers and core code to use struct ide_host.

* Remove no longer needed ide_find_port().

* Make ide_find_port_slot() static.

* Unexport ide_unregister().

v2:
* Add missing 'struct ide_host *host' to macide.c.

v3:
* Fix build problem in pmac.c (s/ide_alloc_host/ide_host_alloc/)
  (Noticed by Stephen Rothwell).
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Add struct ide_tp_ops for transport methods.

* Add 'const struct ide_tp_ops *tp_ops' to struct ide_port_info
  and ide_hwif_t.

* Set the default hwif->tp_ops in ide_init_port_data().

* Set host driver specific hwif->tp_ops in ide_init_port().

* Export ide_exec_command(), ide_read_status(), ide_read_altstatus(),
  ide_read_sff_dma_status(), ide_set_irq(), ide_tf_{load,read}()
  and ata_{in,out}put_data().

* Convert host drivers and core code to use struct ide_tp_ops.

* Remove no longer needed default_hwif_transport().

* Cleanup ide_hwif_t from methods that are now in struct ide_tp_ops.

While at it:

* Use struct ide_port_info in falconide.c and q40ide.c.

* Rename ata_{in,out}put_data() to ide_{in,out}put_data().

v2:

* Fix missing convertion in ns87415.c.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Add 'config' field to hw_regs_t and use it to set hwif->config_data in
ide_init_port_hw(), then convert ide_legacy_init_one() to use hw->config.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Filter out "default" transfer mode values (0x00 - default PIO mode,
  0x01 - default PIO mode w/ IORDY disabled) in write handler for obsoleted
  /proc/ide/hd?/settings:current_speed setting.

  Allowing "default" transfer mode values is a dangerous thing to do as
  we don't support programming controller to the "default" transfer mode
  and devices often use different values for the default and maximum PIO
  mode (i.e. PIO2 default and PIO4 maximum) so the controller will stay
  programmed for higher PIO mode while device will use the lower PIO mode.

  There is no functionality loss as by using special IOCTLs device can
  still be programmed to "default" transfer modes (it is only useful for
  debugging/testing purposes anyway).

* Remove no longer needed IDE_HFLAG_ABUSE_SET_DMA_MODE host flag, it was
  previously used by few host drivers to program the controller to PIO0
  timings for "default" transfer mode == 0x01 (although some host drivers
  would program invalid PIO timings instead).

* Cleanup ide_set_xfer_rate() and add BUG_ON().
Suggested-by: NSergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Lets remove dead Virtual DMA support for now so it doesn't clutter
core IDE code (it can be bring back when there is a need for it):

* Remove IDE_HFLAG_VDMA host flag.

* Remove ide_drive_t.vdma flag.

* cs5520.c: remove stale FIXMEs, cs5520_dma_host_set() and cs5520_dma_ops
  (also there is no longer a need to set IDE_HFLAG_NO_ATAPI_DMA).

There should be no functional changes caused by this patch.

Cc: TAKADA Yoshihito <takada@mbf.nifty.com>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Remove no longer needed ->INB, ->OUTB and ->OUTBSYNC methods.

Then:

* Remove no longer used default_hwif_[mm]iops() and ide_[mm_]outbsync().

* Cleanup SuperIO handling in ns87415.c.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Add ide_read_bcount_and_ireason() helper and use it instead of ->INB
in {cdrom_newpc,ide_pc}_intr().
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Add IDE_TFLAG_IN_FEATURE taskfile flag for reading Feature
  register and handle it in ->tf_read.

* Convert ide_read_error() to use ->tf_read instead of ->INB,
  then uninline and export it.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Add ->set_irq method for setting nIEN bit of ATA Device Control
register and use it instead of ide_set_irq().

While at it:

* Use ->set_irq in init_irq() and do_reset1().

* Don't use HWIF() macro in ide_check_pm_state().

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Remove ide_read_altstatus() inline helper.

* Add ->read_altstatus method for reading ATA Alternate Status
  register and use it instead of ->INB.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Remove ide_read_status() inline helper.

* Add ->read_status method for reading ATA Status register
  and use it instead of ->INB.

While at it:

* Don't use HWGROUP() macro.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Add ->exec_command method for writing ATA Command register
and use it instead of ->OUTBSYNC.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Factor out simplex handling from ide_pci_dma_base() to
  ide_pci_check_simplex().

* Set hwif->dma_base early in ->init_dma method / ide_hwif_setup_dma()
  and reset it in ide_init_port() if DMA initialization fails.

* Use ->read_sff_dma_status instead of ->INB in ide_pci_dma_base().

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Export sff_dma_ops and then remove ide_setup_dma().

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Use ->dma_base + offset instead of ->dma_{status,command}
  and remove no longer needed ->dma_{status,command}.

While at it:

* Use ATA_DMA_* defines.

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Add ->read_sff_dma_status method for reading DMA Status register
and use it instead of ->INB.

While at it:

* Use inb() directly in ns87415.c::ns87415_dma_end().

There should be no functional changes caused by this patch.
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

* Add 'hw_regs_t **hws' argument to ide_device_add[_all]() and convert
  host drivers + ide_legacy_init_one() + ide_setup_pci_device[s]() to use
  it instead of calling ide_init_port_hw() directly.

  [ However if host has > 1 port we must still set hwif->chipset to hint
    consecutive ide_find_port() call that the previous slot is occupied. ]

* Unexport ide_init_port_hw().

v2:
* Use defines instead of hard-coded values in buddha.c, gayle.c and q40ide.c.
  (Suggested by Geert Uytterhoeven)

* Better patch description.

v3:
* Fix build problem in ide-cs.c. (Noticed by Stephen Rothwell)

There should be no functional changes caused by this patch.

Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>

This adds a hid usage that is reported by the N-Trig digitizer in the Dell
Latitude XT screen.
Signed-off-by: NRafi Rubin <rafi@seas.upenn.edu>
Signed-off-by: NVojtech Pavlik <vojtech@suse.cz>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

This is alternative implementation of sg content iterator introduced
by commit 83e7d317... from Pierre Ossman in next-20080716.  As there's
already an sg iterator which iterates over sg entries themselves, name
this sg_mapping_iterator.

Slightly edited description from the original implementation follows.

Iteration over a sg list is not that trivial when you take into
account that memory pages might have to be mapped before being used.
Unfortunately, that means that some parts of the kernel restrict
themselves to directly accesible memory just to not have to deal with
the mess.

This patch adds a simple iterator system that allows any code to
easily traverse an sg list and not have to deal with all the details.
The user can decide to consume part of the iteration.  Also, iteration
can be stopped and resumed later if releasing the kmap between
iteration steps is necessary.  These features are useful to implement
piecemeal sg copying for interrupt drive PIO for example.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

This patch adds to ioatdma and dca modules
support for Intel I/OAT DMA engine ver.3 (aka CB3 device).
The main features of I/OAT ver.3 are:
 * 8 single channel DMA devices (8 channels total)
 * 8 DCA providers, each can accept 2 requesters
 * 8-bit TAG values and 32-bit extended APIC IDs
Signed-off-by: NMaciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

get_proc_net() can now become static.
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Acked-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The function header comments have to go with the functions
they are documenting, or things go horribly wrong when we
try to process them with the docbook tools.

Warning(include/linux/netdevice.h:1006): No description found for parameter 'dev_queue'
Warning(include/linux/netdevice.h:1033): No description found for parameter 'dev_queue'
Warning(include/linux/netdevice.h:1067): No description found for parameter 'dev_queue'
Warning(include/linux/netdevice.h:1093): No description found for parameter 'dev_queue'
Warning(include/linux/netdevice.h:1474): No description found for parameter 'txq'
Error(net/core/dev.c:1674): cannot understand prototype: 'u32 simple_tx_hashrnd; '
Signed-off-by: NDave Jones <davej@redhat.com>
Acked-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The mdio_port, mdio_bit, mdc_port and mdc_bit fields in the
fs_mii_bb_platform_info structure are left-overs from the move to the Phy
Abstraction Layer subsystem. They are not used anymore and can be safely
removed.
Signed-off-by: NLaurent Pinchart <laurentp@cse-semaphore.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>