<linux/prio_tree.h> is unsafe for inclusion by userland apps, but it is in
the userland-exposed portion of <linux/fs.h>.  It's only needed in the
__KERNEL__ protected portion of the file, so move the #include down to
there.

lmbench-2.0.4 runs into this issue in 'flushdisk'.
Signed-off-by: Tom Rini <trini@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

From: Vladimir Saveliev <vs@namesys.com>, Jan Kara <jack@suse.cz>

Add missing SMP locking and one lock_buffer() to the reiserfs quota code.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Fix some bugs in device-mapper handling of 64-bit values, replacing
dm_div_up() and dm_round_up() inlines with macros and removing some
avoidable divisions of 64-bit numbers.

The mirror region size is the granularity used to manage and monitor the
data copying, typically 512KB, so 32 bits (of sectors) should be plenty to
hold this.

Taken together with the two earlier patches:
  "fix TB stripe data corruption" (a missing cast) and
  "stripe_width should be sector_t",
I've now had 3 independent reports that this fixes device-mapper
for devices with large stripes.

Still awaiting test results for dm-crypt and dm-raid1.
Signed-Off-By: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

stripe_width should be sector_t to support large devices.
Signed-Off-By: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Fix unbalanced QP reference count decrement (introduced with QP lock
optimization patch)
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <roland@topspin.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Add a missing break statement between RC and UD cases in mthca_post_send().
 This fixes a possible oops for protocols that use the RC transport.
Signed-off-by: Libor Michalek <libor@topspin.com>
Signed-off-by: Roland Dreier <roland@topspin.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

- Fix it for older C compilers.

- Extern decls always, always, always go in header files, please.

Cc: "Brown, Len" <len.brown@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

into ppc970.osdl.org:/home/torvalds/v2.6/linux

into ppc970.osdl.org:/home/torvalds/v2.6/linux

Not that low-level code really tends to care, but since loff_t is
a signed 64-bit entity and size_t is unsigned (and potentially
64-bit), mixing the two isn't very well-defined.. 

The actual user copy will do them too, but only for the
range that ends up being actually copied. That hides
bugs when the range has been clamped by file size or other
issues.

into ppc970.osdl.org:/home/torvalds/v2.6/linux

With Olaf Hering <olh@suse.de>.

Parameters were interchanged.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

From: Prarit Bhargava <prarit@sgi.com>

I discovered an issue where a hwif_init() failure lead to /proc/ide files
being created for devices that failed probes.  This resulted in
oops/WARN_ON/BUG_ON executions through the kernel depending on what
actions were on going.

slightly changed by me (bart)
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

ide_hwgroup_t.polling field added.  0 in poll_timeout field
used to indicate inactive polling but because 0 is a valid
jiffy value, though slim, there's a chance that something
weird can happen.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

ide_drive_t.sleeping field added.  0 in sleep field used to
indicate inactive sleeping but because 0 is a valid jiffy
value, though slim, there's a chance that something can go
weird.  And while at it, explicit jiffy comparisons are
converted to use time_before() macros.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

In ide_error(), drive cannot be NULL.  ide_dump_status() can't
handle NULL drive.

From: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Explicit jiffy comparision converted to time_after() macro.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

In __ide_do_rw_disk(), ide_started used to be returned blindly
after issusing PIO write.  This can cause hang if
pre_task_out_intr() returns ide_stopped due to failed
ide_wait_stat() test.  Fixed to pass the return value of
pre_task_out_intr().
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

In __ide_do_rw_disk(), the shifted block, instead of the
original rq->sector, should be used when checking range for
lba48 dma.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

In drivers/ide/pci/piix.[hc], init_setup_piix() is defined and
used but only one init_setup function is defined and no
demultiplexing is done using init_setup callback.  As other
drivers call ide_setup_pci_device() directly in such cases,
this patch removes init_setup_piix() and makes piix_init_one()
call ide_setup_pci_device() directly.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

In drivers/ide/pci/opti612.[hc], init_setup_opti621() is
declared, defined and referenced but never actually used.
This patch removes the function.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

In drivers/ide/pci/it8172.h, it8172_ratefilter() and
init_setup_it8172() are declared and the latter is referenced
in it8172_chipsets.  Both functions are not defined or used
anywhere.  This patch removes the prototypes and reference.
it8172 should be compilable now.
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Removes drivers/ide/pci/adma100.[hc].  The driver isn't
compilable (missing functions) and no Kconfig actually enables
CONFIG_BLK_DEV_ADMA100.

On Wed, 02 Feb 2005 03:31:59 -0500, Jeff Garzik <jgarzik@pobox.com> wrote:
> Also, the libata-dev-2.6 tree has an "ata_adma" driver which is
> complete, but needs some testing (and I have h/w).
Signed-off-by: Tejun Heo <tj@home-tj.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

This patch contains the following possible cleanups:
- make some needlessly global code static
- ide-dma.c: remove the unneeded EXPORT_SYMBOL(__ide_dma_test_irq)

slightly changed by me (bart)
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

From: Arjan van de Ven <arjan@infradead.org>

* make atapi_{input,output}_bytes() static, fix users to use drive->hwif
* remove ide_read_24() export; it's unused since the ide-io.c reorganization
* add a FIXME comment to the ide_fix_driveid() export

slightly changed by me (bart)
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

PPC_MULTIPLATFORM exists both for ppc32 and ppc64; chrp and pegasos,
OTOH, are ppc32-only.  The ifdef in via82cxxx should be for
PPC_MULTIPLATFORM and PPC32 - it makes no sense for ppc64 multiplatform
and, as the matter of fact, simply won't compile there.
Signed-off-by: Al Viro <viro@parcelfarce.linux.theplanet.co.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

compile fix on sparc64 - missed pgd_val() bites now that we have
STRICT_MM_TYPECHECKS set.  Incidentally, shifting _by_ unsigned long is
useless - type of result is determined by the type of first argument... 
Signed-off-by: Al Viro <viro@parcelfarce.linux.theplanet.co.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

I accidentally introduced a leak in there, because the end_io function
doesn't do a bio_put() like most others. Fortunately it's not critical,
as it only happens on the md super block read. So just drop the initial
bio_get(), since it's sync we can just bio_put() safely at the end after
it has completed.
Signed-off-by: Jens Axboe <axboe@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Markus Trippelsdorf reported a non-working bt848 card with the recent
DVB update. 

It's an initialization order bug which changes the default state of the
tda9887 output ports.

Verified to fix the problem by Markus Trippelsdorf <markus@trippelsdorf.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

isdnhdlc_decode is called multiple times for bigger frames, so
decrementing dsize is a bad idea and can cause a overflow of
the dst buffer.
Signed-off-by: Karsten Keil <kkeil@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

On the one hand APM isn't enabled on all laptops.
On the other hand, this also affects regular PCs with APM support (or
using a distribution kernel with APM support).

The time for the !APM case was already increased from 30msec in 2.4 .
Isn't there a timeout that is suitable for all cases?

Alan Cox answered:
> The five seconds should be just fine for all cases. The smaller value
> with no
> power manglement should help speed up recovery however. It probably
> doesn't belong CONFIG_APM now ACPI and friends are involved either.

Until someone has a real good solution (consider e.g. that most PC users 
might have ACPI support enabled), this patch unconditionally sets 
WAIT_READY to 5 seconds.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

attached is a new patch for the garmin_gps driver. It only updates my email
address in the header, as i am in the process of changing my isp and the
listed address will be offline sometime soon.
The second change makes the configuration entry look more like the
other entries in the menu.
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

From: Nico Huber <nico.h@gmx.de>

The receiver of my Logitech Cordeless Desktop fails to report the
keyboard's class descriptor most times I insert the usb-hid module since I
changed to linux 2.6.  The modell of the receiver is C-BD9-DUAL REV C.  The
request seems not to fail but the count of received characters is zero.

As I said it only fails most times, I worked around making the following
changes in drivers/usb/input/hid-core.c from linux-2.6.11-rc2:

Following the good example of drivers/usb/core/message.c line 575, I
initialized the buffer in hid_get_class_descriptor() to zero.  In the loop
of hid_get_class_descriptor() not waiting for any result but waiting for a
result wich is lower the requested size of the class descriptor (line
1290).  usb_hid_configure() should not try to parse the expected length but
the received (line 1653).

Attached is a patch with these changes
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

Peter Bartosch wrote:
> Hello all,
>
>
> i think i found a possible buffer overflow in hid-core (2.6.10-rc2):

Agreed, and sorry it took so long to reply/patch this.
(from 2004-DEC-01)

> when i interpret this code right, the buffer (buf) will overflow when
> the iManufacturer stringlength is 0 and the iProduct stringlength
> is > 64!

Thanks, here's a patch for it.

Fix potential buffer overflow (by a device descriptor):
buf size is only 64 bytes.
Signed-off-by: Randy Dunlap <rddunlap@osdl.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>