sfrench/cifs-2.6.git
18 years ago[PATCH] USB: Fix regression in core/devio.c
Alan Stern [Wed, 10 Aug 2005 19:15:57 +0000 (15:15 -0400)]
[PATCH] USB: Fix regression in core/devio.c

This patch (as551) fixes another little problem recently added to the
USB core.  Someone didn't fix the type of the first argument to
unregister_chrdev_region.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: real nodes instead of usbfs
Kay Sievers [Sat, 30 Jul 2005 23:05:53 +0000 (01:05 +0200)]
[PATCH] USB: real nodes instead of usbfs

This patch introduces a /sys/class/usb_device/ class
where every connected usb-device will show up:

  tree /sys/class/usb_device/
  /sys/class/usb_device/
  |-- usb1.1
  |   |-- dev
  |   `-- device -> ../../../devices/pci0000:00/0000:00:1d.0/usb1
  |-- usb2.1
  |   |-- dev
  |   `-- device -> ../../../devices/pci0000:00/0000:00:1d.1/usb2
  ...

The presence of the "dev" file lets udev create real device nodes.
  kay@pim:~/src/linux-2.6> tree /dev/bus/usb/
  /dev/bus/usb/
  |-- 1
  |   `-- 1
  |-- 2
  |   `-- 1
  ...

udev rule:
  SUBSYSTEM="usb_device", PROGRAM="/sbin/usb_device %k", NAME="%c"
  (echo $1 | /bin/sed 's/usb\([0-9]*\)\.\([0-9]*\)/bus\/usb\/\1\/\2/')

This makes libusb pick up the real nodes instead of the mounted usbfs:
  export USB_DEVFS_PATH=/dev/bus/usb

Background:
  All this makes it possible to manage usb devices with udev instead of
  the devfs solution. We are currently working on a pam_console/resmgr
  replacement driven by udev and a pam-helper. It applies ACL's to device
  nodes, which is required for modern desktop functionalty like
  "Fast User Switching" or multiple local login support.

New patch with its own major. I've succesfully disabled usbfs and use real
nodes only on my box. With: "export USB_DEVFS_PATH=/dev/bus/usb" libusb picks
up the udev managed nodes instead of reading usbfs files.

This makes udev to provide symlinks for libusb to pick up:
  SUBSYSTEM="usb_device", PROGRAM="/sbin/usbdevice %k", SYMLINK="%c"

/sbin/usbdevice:
  #!/bin/sh
  echo $1 | /bin/sed 's/usbdev\([0-9]*\)\.\([0-9]*\)/bus\/usb\/\1\/\2/'

Signed-off-by: Kay Sievers <kay.sievers@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: Prevent hid-core claiming Apple Bluetooth device on new G4 powerbooks
Andrew de Quincey [Thu, 4 Aug 2005 22:16:12 +0000 (23:16 +0100)]
[PATCH] USB: Prevent hid-core claiming Apple Bluetooth device on new G4 powerbooks

To recap: My new G4 powerbook has a bluetooth device that boots up in
what apppears to be a compatability mode - it looks exactly like an HID
keyboard/mouse device.

A special command sequence is sent to switch it into full bluetooth
mode. When this occurs the original HID device vanishes, and a new
(bluetooth HID) USB device appears on the bus with a different product
ID.

The original thread is here:
http://sourceforge.net/mailarchive/message.php?msg_id=12532263

The attached patch adds the device to the hid-core quirks so that
hid-core ignores it.

Signed-off-by: Andrew de Quincey <adq_dvb@lidskialf.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB ftdi_sio: New IDs for ELV, Xsens and Falcom products
Ian Abbott [Tue, 2 Aug 2005 13:01:27 +0000 (14:01 +0100)]
[PATCH] USB ftdi_sio: New IDs for ELV, Xsens and Falcom products

This patch for the ftdi_sio driver adds a bunch of new devices and fixes
an incorrect PID:

o Fix PID for ELV UO100 (the PID was in fact for ELV UR100).
o Add PID ELV UR100 (see above) and ELV ALC 8500 Expert.
o Add a whole bunch of other PIDs for ELV USB devices, commented out for
   now as they may be used by other drivers eventually.  (Christian Abt
   of ELV.de submitted a full list of devices including an indication of
   which set of drivers are used by default in the MS Windows world.  We
   decided to comment out the devices that use FTDI's D2XX Windows
   drivers by default.)
o Add PIDs for eight devices from Xsens Technologies BV (submitted in a
   patch against 2.6.12.2 by Patrick Riphagen).
o Add PID for Falcom Samba GPRS modem (submitted by Sebastian Schubert).

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB ftdi_sio: user specified VID/PID
Ian Abbott [Thu, 28 Jul 2005 17:40:32 +0000 (18:40 +0100)]
[PATCH] USB ftdi_sio: user specified VID/PID

ftdi_sio: Support one user specified vendor and product ID via a couple
of new module parameters.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: pl2303 driver, makes pl2303HX chip work correctly
Dariusz M [Thu, 28 Jul 2005 16:06:13 +0000 (18:06 +0200)]
[PATCH] USB: pl2303 driver, makes pl2303HX chip work correctly

This trivial patch makes pl2303 driver work correctly with pl2303HX chip.
Apparently some bug in HX version of pl2303 makes the chip loose some
transmitted bytes or stop working at all after reception of
USB_REQ_CLEAR_FEATURE mesage. Logs generated by UsbSnoop application reveal
that windows driver does not send this type of messages to the converter.

From: "Dariusz M." <D.Marcinkiewicz@elka.pw.edu.pl>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: fix keyspan_remote endian bug on probe
Greg Kroah-Hartman [Wed, 27 Jul 2005 08:06:19 +0000 (01:06 -0700)]
[PATCH] USB: fix keyspan_remote endian bug on probe

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: Gadget library: centralize gadget controller numbers
David Brownell [Wed, 13 Jul 2005 22:18:30 +0000 (15:18 -0700)]
[PATCH] USB: Gadget library: centralize gadget controller numbers

This patch centralizes the assignment of bcdDevice numbers for different
gadget controllers.  This won't improve the object code at all, but it
does save a lot of repetitive and error-prone source code ... and will
simplify the work of supporting a new controller driver, since most new
gadget drivers will no longer need patches (unless some hardware quirks
limit USB protocol messaging).

Added minor cleanups and identifer hooks for the UDC in the Freescale
iMX series processors.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: drivers/serial/usb-serial: Remove unneeded void * casts
Tobias Klauser [Mon, 4 Jul 2005 17:32:51 +0000 (19:32 +0200)]
[PATCH] USB: drivers/serial/usb-serial: Remove unneeded void * casts

The following patch removes unneeded casts for the following (void *) pointers:
- tty_struct->driver_data
- void *private argument of usb_serial_port_softint()

Signed-off-by: Tobias Klauser <tklauser@nuerscht.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: option card driver coding style tweaks
Andrew Morton [Wed, 27 Jul 2005 08:08:30 +0000 (01:08 -0700)]
[PATCH] USB: option card driver coding style tweaks

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] USB: whitespace fixes for cypress_m8 driver
Lonnie Mendez [Tue, 12 Jul 2005 22:21:31 +0000 (17:21 -0500)]
[PATCH] USB: whitespace fixes for cypress_m8 driver

Reading this driver I noticed some trailing whitespaces and tabs so I
removed them with some 80th column fitting and a few more similar
things.

From: Carlo Perassi <carlo@linux.it>
Signed-off-by: Lonnie Mendez <dignome@gmail.com>
Signed-off-by: Carlo Perassi <carlo@linux.it>
Signed-off-by: Domen Puncer <domen@coderock.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/w1-2.6
Linus Torvalds [Thu, 8 Sep 2005 22:55:53 +0000 (15:55 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/w1-2.6

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6
Linus Torvalds [Thu, 8 Sep 2005 22:55:23 +0000 (15:55 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/pci-2.6

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6
Linus Torvalds [Thu, 8 Sep 2005 22:54:53 +0000 (15:54 -0700)]
Merge /pub/scm/linux/kernel/git/davem/sparc-2.6

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Thu, 8 Sep 2005 22:52:11 +0000 (15:52 -0700)]
Merge /pub/scm/linux/kernel/git/davem/net-2.6

18 years agoMerge master.kernel.org:/home/rmk/linux-2.6-i2c manually
Linus Torvalds [Thu, 8 Sep 2005 22:43:09 +0000 (15:43 -0700)]
Merge kernel.org:/home/rmk/linux-2.6-i2c manually

Old tree, so the automatic merge had some problems.

18 years agoMerge master.kernel.org:/home/rmk/linux-2.6-serial
Linus Torvalds [Thu, 8 Sep 2005 22:30:14 +0000 (15:30 -0700)]
Merge master.kernel.org:/home/rmk/linux-2.6-serial

18 years agoMerge master.kernel.org:/home/rmk/linux-2.6-mmc
Linus Torvalds [Thu, 8 Sep 2005 22:28:16 +0000 (15:28 -0700)]
Merge master.kernel.org:/home/rmk/linux-2.6-mmc

18 years agoMerge master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Thu, 8 Sep 2005 22:26:48 +0000 (15:26 -0700)]
Merge master.kernel.org:/home/rmk/linux-2.6-arm

18 years ago[XFRM]: Always release dst_entry on error in xfrm_lookup
Patrick McHardy [Thu, 8 Sep 2005 22:11:55 +0000 (15:11 -0700)]
[XFRM]: Always release dst_entry on error in xfrm_lookup

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[TCP]: Fix off by one in tcp_fragment() "already sent" test.
Herbert Xu [Thu, 8 Sep 2005 22:10:52 +0000 (15:10 -0700)]
[TCP]: Fix off by one in tcp_fragment() "already sent" test.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[ARM] 2892/1: remove gcc workaround for direct access to absolute memory addresses
Nicolas Pitre [Thu, 8 Sep 2005 22:07:40 +0000 (23:07 +0100)]
[ARM] 2892/1: remove gcc workaround for direct access to absolute memory addresses

Patch from Nicolas Pitre

It used to make a difference in the gcc-2.95 era.  However these days
modern gcc apparently got better at not being influenced by such constructs
(which is good in general) and therefore such workaround is of no real
advantage anymore.
The good news is that gcc (from version 4.1.0) is now fixed with
regards to the defficiency this workaround was trying to address.
For those interested the patch can easily be backported to older gcc
versions and can be found here:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm.c.diff?r1=1.476&r2=1.478
and also here:
http://savannah.gnu.org/cgi-bin/viewcvs/gcc/gcc/gcc/config/arm/arm.c.diff?r1=text&tr1=1.476&r2=text&tr2=1.478&diff_format=u

Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[ARM] 2891/1: S3C2410 - update s3c2410_defconfig for 2.6.13
Ben Dooks [Thu, 8 Sep 2005 22:07:39 +0000 (23:07 +0100)]
[ARM] 2891/1: S3C2410 - update s3c2410_defconfig for 2.6.13

Patch from Ben Dooks

Updated the s3c2410_defconfig for the 2.6.13-git8
kernel release, as well as adding the Anubis
board to the list of boards built.

Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[ARM] 2890/1: OMAP 1/4: Update omap1 specific files, take 2
Tony Lindgren [Thu, 8 Sep 2005 22:07:38 +0000 (23:07 +0100)]
[ARM] 2890/1: OMAP 1/4: Update omap1 specific files, take 2

Patch from Tony Lindgren

This patch syncs the mainline kernel with linux-omap tree.
The highlights of the patch are:
- Convert more drivers to register resources in board-*.c to take
  advantage of the driver model by David Brownell and Ladislav Michl
- Use set_irq_type() for GPIO interrupts instead of
  omap_set_gpio_edge_ctrl() by David Brownell
- Add minimal support for handling optional add-on boards, such as
  OSK Mistral board with LCD and keypad, by David Brownell
- Minimal support for loading functions to SRAM by Tony Lindgren
- Wake up from serial port by muxing RX lines temporarily into GPIO
  interrupts by Tony Lindgren
- 32KHz sched_clock by Tony Lindgren and Juha Yrjola

Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[PATCH] PCI/libata INTx cleanup
Brett M Russ [Mon, 15 Aug 2005 19:23:41 +0000 (15:23 -0400)]
[PATCH] PCI/libata INTx cleanup

Simple cleanup to eliminate X copies of the pci_enable_intx() function
in libata.  Moved ahci.c's pci_intx() to pci.c and use it throughout
libata and msi.c.

Signed-off-by: Brett Russ <russb@emc.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: Support PCM PM CAP version 3
Daniel Ritz [Wed, 17 Aug 2005 22:32:19 +0000 (15:32 -0700)]
[PATCH] PCI: Support PCM PM CAP version 3

- support PCI PM CAP version 3 (as defined in PCI PM Interface Spec v1.2)

- pci/probe.c sets the PM state initially to 4 which is D3cold.  add a
  PCI_UNKNOWN

- minor cleanups

Signed-off-by: Daniel Ritz <daniel.ritz@gmx.ch>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: Add pci_walk_bus function to PCI core (nonrecursive)
Paul Mackerras [Thu, 18 Aug 2005 04:33:01 +0000 (14:33 +1000)]
[PATCH] PCI: Add pci_walk_bus function to PCI core (nonrecursive)

The PCI error recovery infrastructure needs to be able to contact all
the drivers affected by a PCI error event, which may mean traversing
all the devices under a given PCI-PCI bridge.  This patch adds a
function to the PCI core that traverses all the PCI devices on a PCI
bus and under any PCI-PCI bridges on that bus (and so on), calling a
given function for each device.  This provides a way for the error
recovery code to iterate through all devices that are affected by an
error event.

This version is not implemented as a recursive function.  Instead,
when we reach a PCI-PCI bridge, we set the pointers to start doing the
devices on the bus under the bridge, and when we reach the end of a
bus's devices, we use the bus->self pointer to go back up to the next
higher bus and continue doing its devices.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years agoMerge branch 'master' of /usr/src/linux-2.6
Anton Altaparmakov [Thu, 8 Sep 2005 22:03:30 +0000 (23:03 +0100)]
Merge branch 'master' of /usr/src/linux-2.6

18 years agoNTFS: 2.1.24 release and some minor final fixes.
Anton Altaparmakov [Thu, 8 Sep 2005 22:01:16 +0000 (23:01 +0100)]
NTFS: 2.1.24 release and some minor final fixes.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[PATCH] PCI Hotplug: SGI hotplug driver fixes
Prarit Bhargava [Fri, 12 Aug 2005 14:13:34 +0000 (10:13 -0400)]
[PATCH] PCI Hotplug: SGI hotplug driver fixes

These fixes were suggested by pcihpd-discuss, but were dropped in the
initial checkin of the code.  These fixes include cleaning up the
hotplug driver sysfs filename, and some minor code cleanups.  The driver
also requires at least PROM 4.30, not 4.20.

Signed-off-by: Prarit Bhargava <prarit@sgi.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: remove pci_find_device from parport_pc.c
Jiri Slaby [Wed, 10 Aug 2005 00:09:39 +0000 (02:09 +0200)]
[PATCH] PCI: remove pci_find_device from parport_pc.c

This patch changes pci_find_device to pci_get_device (encapsulated in
for_each_pci_dev).

Signed-off-by: Jiri Slaby <xslaby@fi.muni.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: Fix regression in pci_enable_device_bars
Alan Stern [Wed, 10 Aug 2005 19:18:44 +0000 (15:18 -0400)]
[PATCH] PCI: Fix regression in pci_enable_device_bars

This patch (as552) fixes yet another small problem recently added.  If an
attempt to put a PCI device back into D0 fails because the device doesn't
support PCI PM, it shouldn't count as error.  Without this patch the UHCI
controllers on my Intel motherboard don't work.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: start paying attention to a lot of pci function return values
Greg Kroah-Hartman [Thu, 28 Jul 2005 18:37:33 +0000 (11:37 -0700)]
[PATCH] PCI: start paying attention to a lot of pci function return values

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] Make sparc64 use setup-res.c
David S. Miller [Mon, 8 Aug 2005 20:19:08 +0000 (13:19 -0700)]
[PATCH] Make sparc64 use setup-res.c

There were three changes necessary in order to allow
sparc64 to use setup-res.c:

1) Sparc64 roots the PCI I/O and MEM address space using
   parent resources contained in the PCI controller structure.
   I'm actually surprised no other platforms do this, especially
   ones like Alpha and PPC{,64}.  These resources get linked into the
   iomem/ioport tree when PCI controllers are probed.

   So the hierarchy looks like this:

   iomem --|
   PCI controller 1 MEM space --|
        device 1
device 2
etc.
   PCI controller 2 MEM space --|
        ...
   ioport --|
            PCI controller 1 IO space --|
...
            PCI controller 2 IO space --|
...

   You get the idea.  The drivers/pci/setup-res.c code allocates
   using plain iomem_space and ioport_space as the root, so that
   wouldn't work with the above setup.

   So I added a pcibios_select_root() that is used to handle this.
   It uses the PCI controller struct's io_space and mem_space on
   sparc64, and io{port,mem}_resource on every other platform to
   keep current behavior.

2) quirk_io_region() is buggy.  It takes in raw BUS view addresses
   and tries to use them as a PCI resource.

   pci_claim_resource() expects the resource to be fully formed when
   it gets called.  The sparc64 implementation would do the translation
   but that's absolutely wrong, because if the same resource gets
   released then re-claimed we'll adjust things twice.

   So I fixed up quirk_io_region() to do the proper pcibios_bus_to_resource()
   conversion before passing it on to pci_claim_resource().

3) I was mistakedly __init'ing the function methods the PCI controller
   drivers provide on sparc64 to implement some parts of these
   routines.  This was, of course, easy to fix.

So we end up with the following, and that nasty SPARC64 makefile
ifdef in drivers/pci/Makefile is finally zapped.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: restore BAR values after D3hot->D0 for devices that need it
John W. Linville [Wed, 27 Jul 2005 14:19:44 +0000 (10:19 -0400)]
[PATCH] PCI: restore BAR values after D3hot->D0 for devices that need it

Some PCI devices (e.g. 3c905B, 3c556B) lose all configuration
(including BARs) when transitioning from D3hot->D0.  This leaves such
a device in an inaccessible state.  The patch below causes the BARs
to be restored when enabling such a device, so that its driver will
be able to access it.

The patch also adds pci_restore_bars as a new global symbol, and adds a
correpsonding EXPORT_SYMBOL_GPL for that.

Some firmware (e.g. Thinkpad T21) leaves devices in D3hot after a
(re)boot.  Most drivers call pci_enable_device very early, so devices
left in D3hot that lose configuration during the D3hot->D0 transition
will be inaccessible to their drivers.

Drivers could be modified to account for this, but it would
be difficult to know which drivers need modification.  This is
especially true since often many devices are covered by the same
driver.  It likely would be necessary to replicate code across dozens
of drivers.

The patch below should trigger only when transitioning from D3hot->D0
(or at boot), and only for devices that have the "no soft reset" bit
cleared in the PM control register.  I believe it is safe to include
this patch as part of the PCI infrastructure.

The cleanest implementation of pci_restore_bars was to call
pci_update_resource.  Unfortunately, that does not currently exist
for the sparc64 architecture.  The patch below includes a null
implemenation of pci_update_resource for sparc64.

Some have expressed interest in making general use of the the
pci_restore_bars function, so that has been exported to GPL licensed
modules.

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: use bus_slot number for name
Kristen Accardi [Fri, 5 Aug 2005 19:16:06 +0000 (12:16 -0700)]
[PATCH] PCI Hotplug: use bus_slot number for name

For systems with multiple hotplug controllers, you need to use more than
just the slot number to uniquely name the slot.  Without a unique slot
name, the pci_hp_register() will fail.  This patch adds the bus number
to the name.

Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] arch/386/pci: remap_pfn_range -> io_remap_pfn_range
Michael S. Tsirkin [Sun, 31 Jul 2005 08:51:45 +0000 (11:51 +0300)]
[PATCH] arch/386/pci: remap_pfn_range -> io_remap_pfn_range

Convert i386/pci to use io_remap_pfn_range instead of remap_pfn_range.
This is good for Xen which reuses i386/pci/i386.c for domain 0 code.

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: clean up pci.h and split pci register info to separate header file.
Greg Kroah-Hartman [Thu, 28 Jul 2005 18:37:33 +0000 (11:37 -0700)]
[PATCH] PCI: clean up pci.h and split pci register info to separate header file.

This cleans up some of the #ifdef CONFIG_PCI stuff up, and moves the pci register
info out to a separate file, where it belongs.  Eventually we can stop including
this file from within pci.h, but lots of code needs to be audited first.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: fix up pretty-names removal patch
Andrew Morton [Wed, 27 Jul 2005 07:07:11 +0000 (00:07 -0700)]
[PATCH] PCI: fix up pretty-names removal patch

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: remove CONFIG_PCI_NAMES
Adrian Bunk [Sun, 17 Jul 2005 02:22:20 +0000 (04:22 +0200)]
[PATCH] PCI: remove CONFIG_PCI_NAMES

This patch removes CONFIG_PCI_NAMES.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: Move PCI fixup data into r/o section
akpm@osdl.org [Tue, 26 Jul 2005 06:28:14 +0000 (23:28 -0700)]
[PATCH] PCI: Move PCI fixup data into r/o section

Make PCI fixup data const, so it'll end up in a r/o section.

This also fixes the conversion into ECOFF which gets broken by too many
changes between r/w and r/o sections.  Call it a hack but it's a change
that's correct by itself.

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI: Run PCI driver initialization on local node
Andi Kleen [Wed, 6 Jul 2005 17:56:03 +0000 (19:56 +0200)]
[PATCH] PCI: Run PCI driver initialization on local node

Run PCI driver initialization on local node

Instead of adding messy kmalloc_node()s everywhere run the
PCI driver probe on the node local to the device.

This would not have helped for IDE, but should for
other more clean drivers that do more initialization in probe().
It won't help for drivers that do most of the work
on first open (like many network drivers)

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: rpaphp: Purify hotplug
John Rose [Mon, 25 Jul 2005 15:17:03 +0000 (10:17 -0500)]
[PATCH] PCI Hotplug: rpaphp: Purify hotplug

Currently rpaphp registers the following bus types as hotplug slots:
1) Actual PCI Hotplug slots
2) Embedded/Internal PCI slots
3) PCI Host Bridges

The second and third bus types are not actually direct parents of
removable adapters.  As such, the rpaphp has special case code to fake
results for attributes like power, adapter status, etc.  This patch
removes types 2 and 3 from the rpaphp module.

This patch also changes the DLPAR module so that slots can be
DLPAR-added/removed without having been designated as hotplug-capable.

Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: rpaphp: Export slot enable
John Rose [Mon, 25 Jul 2005 15:16:58 +0000 (10:16 -0500)]
[PATCH] PCI Hotplug: rpaphp: Export slot enable

This patch exports rpaphp_config_pci_adapter() for use by the rpadlpar
module.  It also changes this function by removing any dependencies on
struct slot.  The patch also changes the RPA DLPAR-add path to enable
newly-added slots in a separate step from that which registers them as
hotplug slots.

Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: rpaphp: Remove rpaphp_find_pci
John Rose [Mon, 25 Jul 2005 15:16:53 +0000 (10:16 -0500)]
[PATCH] PCI Hotplug: rpaphp: Remove rpaphp_find_pci

The rpaphp module currently uses a fragile method to find a pci device
by its device node.  This function is unnecessary, so this patch scraps
it.

Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: rpaphp: Change slot pci reference
John Rose [Mon, 25 Jul 2005 16:13:38 +0000 (11:13 -0500)]
[PATCH] PCI Hotplug: rpaphp: Change slot pci reference

The slot structure in the rpaphp module currently references the PCI
contents of the slot using the PCI device of the parent bridge.  This
is unnecessary, since the module is actually interested in the
subordinate bus of the bridge.  The dependency on a PCI bridge device
also prohibits the module from registering hotplug slots that have a
root bridge as a parent, since root bridges on PPC64 don't have PCI
devices.

This patch changes struct slot to reference the PCI subsystem using a
pci_bus rather than a pci_dev.

Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: rpaphp: Move VIO registration
John Rose [Mon, 25 Jul 2005 15:16:42 +0000 (10:16 -0500)]
[PATCH] PCI Hotplug: rpaphp: Move VIO registration

Currently, rpaphp registers Virtual I/O slots as hotplug slots.  The
only purpose of this registration is to ensure that the VIO subsystem
is notified of new VIO buses during DLPAR adds.  Similarly, rpaphp
notifies the VIO subsystem when a VIO bus is DLPAR-removed.  The rpaphp
module has special case code to fake results for attributes like power,
adapter status, etc.

The VIO register/unregister functions could just as easily be made from
the DLPAR module.  This patch moves the VIO registration calls to the
DLPAR module, and removes the VIO fluff from rpaphp altogether.

Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] PCI Hotplug: rpaphp: Remove unused stuff
John Rose [Mon, 25 Jul 2005 15:16:37 +0000 (10:16 -0500)]
[PATCH] PCI Hotplug: rpaphp: Remove unused stuff

Subject line says it all :)

Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[MMC] Add mmc_detect_change() delay support for PXAMCI driver
Richard Purdie [Thu, 8 Sep 2005 21:48:16 +0000 (22:48 +0100)]
[MMC] Add mmc_detect_change() delay support for PXAMCI driver

Allow PXA platforms to pass an appropriate delay value to the
PXA MCI driver for delaying detection changes.

Signed-Off-By: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years agoFix up ARM serial driver compile failure
Linus Torvalds [Thu, 8 Sep 2005 21:47:12 +0000 (14:47 -0700)]
Fix up ARM serial driver compile failure

Proud member of Uglyhacks'R'US.

Acked-by: David S. Miller <davem@davemloft.net>
Acked-by: Russell King <rmk+lkml@arm.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[MMC] Ensure correct mmc_priv() behaviour
Russell King [Thu, 8 Sep 2005 21:46:00 +0000 (22:46 +0100)]
[MMC] Ensure correct mmc_priv() behaviour

mmc_priv() has some nasty effects if the wrong pointer type is
passed to it.  Introduce type checking, which also means we get
the right type.  Also add an additional member to mmc_host which
is used to align host-private data appropriately.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[PATCH] w1_ds2433: Added crc16 protection and read caching.
Evgeniy Polyakov [Wed, 17 Aug 2005 11:24:37 +0000 (15:24 +0400)]
[PATCH] w1_ds2433: Added crc16 protection and read caching.

The changes to ds2433 to add CRC16 protection and read caching.

Signed-off-by: Ben Gardner <bgardner@wabtec.com>
Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: added private family data into w1_slave strucutre.
Evgeniy Polyakov [Wed, 17 Aug 2005 11:19:08 +0000 (15:19 +0400)]
[PATCH] w1: added private family data into w1_slave strucutre.

Add family_data to struct w1_slave.

Signed-off-by: Ben Gardner <bgardner@wabtec.com>
Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] lib/crc16: added crc16 algorithm.
Evgeniy Polyakov [Wed, 17 Aug 2005 11:17:26 +0000 (15:17 +0400)]
[PATCH] lib/crc16: added crc16 algorithm.

Add the crc16 routines, as used by w1 devices.

Signed-off-by: Ben Gardner <bgardner@wabtec.com>
Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Added DS2433 driver - family id update.
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:50 +0000 (17:27 +0400)]
[PATCH] w1: Added DS2433 driver - family id update.

Work by Ben Gardner <bgardner@wabtec.com>.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Added DS2433 driver.
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:50 +0000 (17:27 +0400)]
[PATCH] w1: Added DS2433 driver.

Work by Ben Gardner <bgardner@wabtec.com>.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Decreased debug level.
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:50 +0000 (17:27 +0400)]
[PATCH] w1: Decreased debug level.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Detouching bug fixed.
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:50 +0000 (17:27 +0400)]
[PATCH] w1: Detouching bug fixed.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Added add/remove slave callbacks.
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:50 +0000 (17:27 +0400)]
[PATCH] w1: Added add/remove slave callbacks.

Patch is based on work from Ben Gardner <bgardner@wabtec.com>

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Added w1_reset_select_slave() - Resets the bus and then selects the slave by
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:49 +0000 (17:27 +0400)]
[PATCH] w1: Added w1_reset_select_slave() - Resets the bus and then selects the slave by

sending either a skip rom or a rom match.

Patch from Ben Gardner <bgardner@wabtec.com>

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Added inline functions on top of container_of().
Evgeniy Polyakov [Thu, 11 Aug 2005 13:27:49 +0000 (17:27 +0400)]
[PATCH] w1: Added inline functions on top of container_of().

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: Fixed 64bit compilation warning.
Evgeniy Polyakov [Thu, 11 Aug 2005 09:45:54 +0000 (13:45 +0400)]
[PATCH] w1: Fixed 64bit compilation warning.

Fixed 64bit compilation warning.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] w1: hotplug support.
Evgeniy Polyakov [Thu, 11 Aug 2005 09:20:07 +0000 (13:20 +0400)]
[PATCH] w1: hotplug support.

Here is W1 hotplug in addition to netlink notifications.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] W1: Sync with w1/ds9490 tree.
Evgeniy Polyakov [Wed, 3 Aug 2005 11:14:50 +0000 (15:14 +0400)]
[PATCH] W1: Sync with w1/ds9490 tree.

Whitespace, static/nonstatic cleanups.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[PATCH] W1: w1_netlink: New init/fini netlink callbacks.
Evgeniy Polyakov [Wed, 27 Jul 2005 09:10:11 +0000 (13:10 +0400)]
[PATCH] W1: w1_netlink: New init/fini netlink callbacks.

They are guarded with NETLINK_DISABLE compile time options,
so if CONFIG_NET is disabled, no linking errors occur.
Bug noticed by Adrian Bunk <bunk@stusta.de>.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
18 years ago[SPARC64]: Inline membar()'s again.
David S. Miller [Thu, 8 Sep 2005 21:37:53 +0000 (14:37 -0700)]
[SPARC64]: Inline membar()'s again.

Since GCC has to emit a call and a delay slot to the
out-of-line "membar" routines in arch/sparc64/lib/mb.S
it is much better to just do the necessary predicted
branch inline instead as:

ba,pt %xcc, 1f
 membar #whatever
1:

instead of the current:

call membar_foo
 dslot

because this way GCC is not required to allocate a stack
frame if the function can be a leaf function.

This also makes this bug fix easier to backport to 2.4.x

Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[IPV6]: Don't redo xfrm_lookup for cached dst entries
Patrick McHardy [Thu, 8 Sep 2005 21:27:47 +0000 (14:27 -0700)]
[IPV6]: Don't redo xfrm_lookup for cached dst entries

The xfrm lookup is already done when the dst entry is looked up first and
stored in the cache.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[IA64] Manual merge fix for 3 files
Tony Luck [Thu, 8 Sep 2005 21:27:13 +0000 (14:27 -0700)]
[IA64] Manual merge fix for 3 files

arch/ia64/Kconfig
arch/ia64/kernel/acpi.c
include/asm-ia64/irq.h

Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years agoNTFS: Improve scalability by changing the driver global spin lock in
Anton Altaparmakov [Thu, 8 Sep 2005 21:13:02 +0000 (22:13 +0100)]
NTFS: Improve scalability by changing the driver global spin lock in
      fs/ntfs/aops.c::ntfs_end_buffer_async_read() to a bit spin lock
      in the first buffer head of a page.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Fix page_has_buffers()/page_buffers() handling in fs/ntfs/aops.c.
Anton Altaparmakov [Thu, 8 Sep 2005 21:08:11 +0000 (22:08 +0100)]
NTFS: Fix page_has_buffers()/page_buffers() handling in fs/ntfs/aops.c.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Fixup handling of sparse, compressed, and encrypted attributes in
Anton Altaparmakov [Thu, 8 Sep 2005 21:04:20 +0000 (22:04 +0100)]
NTFS: Fixup handling of sparse, compressed, and encrypted attributes in
      fs/ntfs/aops.c::ntfs_readpage().

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Fix fs/ntfs/aops.c::ntfs_{read,write}_block() to handle the case
Anton Altaparmakov [Thu, 8 Sep 2005 21:00:33 +0000 (22:00 +0100)]
NTFS: Fix fs/ntfs/aops.c::ntfs_{read,write}_block() to handle the case
      where a concurrent truncate has truncated the runlist under our feet.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[IA64] Increase max physical address for SN platforms
Jack Steiner [Thu, 8 Sep 2005 20:28:28 +0000 (15:28 -0500)]
[IA64] Increase max physical address for SN platforms

Increase the value for the maximum physical address on SN systems.

Signed-off-by: Jack Steiner <steiner@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years ago[IA64] ensure XPC and XPNET are loaded on sn2 platforms only
Dean Nelson [Thu, 8 Sep 2005 15:46:58 +0000 (10:46 -0500)]
[IA64] ensure XPC and XPNET are loaded on sn2 platforms only

These are SN2 only drivers.  They should have platform checks to prevent
them from doing evil stuff in GENERIC kernels.

Signed-off-by: Martin Hicks <mort@sgi.com>
Acked-by: Dean Nelson <dcn@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years agoMerge branch 'upstream' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6
Linus Torvalds [Thu, 8 Sep 2005 20:52:24 +0000 (13:52 -0700)]
Merge branch 'upstream' of /linux/kernel/git/jgarzik/netdev-2.6

18 years ago[IA64] defconfig: turn off QLOGIC_FC
Martin Hicks [Thu, 8 Sep 2005 15:28:11 +0000 (11:28 -0400)]
[IA64] defconfig: turn off QLOGIC_FC

Turn off the QLOGIC_FC driver.  Supposedly qla2xxx should support
these devices.  Do any ia64 machines have one of these devices as
the boot device?

Signed-off-by: Martin Hicks <mort@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
18 years agoKconfig: IEEE80211 should not depend on NET_RADIO
Jeff Garzik [Thu, 8 Sep 2005 20:44:33 +0000 (16:44 -0400)]
Kconfig: IEEE80211 should not depend on NET_RADIO

We should not restrict use of ieee80211 to only when wireless drivers
are enabled.  In-development and out-of-tree drivers may wish to use it,
and by removing this restriction we eliminate a circular dependency.

18 years agoNTFS: Optimize fs/ntfs/aops.c::ntfs_write_block() by extending the page
Anton Altaparmakov [Thu, 8 Sep 2005 20:43:47 +0000 (21:43 +0100)]
NTFS: Optimize fs/ntfs/aops.c::ntfs_write_block() by extending the page
      lock protection over the buffer submission for i/o which allows the
      removal of the get_bh()/put_bh() pairs for each buffer.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[AX.25]: Make asc2ax() thread-proof
Ralf Baechle [Thu, 8 Sep 2005 20:40:41 +0000 (13:40 -0700)]
[AX.25]: Make asc2ax() thread-proof

Asc2ax was still using a static buffer for all invocations which isn't
exactly SMP-safe.  Change asc2ax to take an additional result buffer as
the argument.  Change all callers to provide such a buffer.

This one only really is a fix for ROSE and as per recent discussions
there's still much more to fix in ROSE ...

Signed-off-by: Ralf Baechle DL5RB <ralf@linux-mips.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years agoNTFS: Fixup handling of sparse, compressed, and encrypted attributes in
Anton Altaparmakov [Thu, 8 Sep 2005 20:38:05 +0000 (21:38 +0100)]
NTFS: Fixup handling of sparse, compressed, and encrypted attributes in
      fs/ntfs/aops.c::ntfs_writepage().

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[NETFILTER]: ip_conntrack_netbios_ns.c gcc-2.95.x build fix
Andrew Morton [Thu, 8 Sep 2005 20:36:34 +0000 (13:36 -0700)]
[NETFILTER]: ip_conntrack_netbios_ns.c gcc-2.95.x build fix

gcc-2.95.x can't do this sort of initialisation

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[IPV4]: Fix refcount damaging in net/ipv4/route.c
Julian Anastasov [Thu, 8 Sep 2005 20:34:47 +0000 (13:34 -0700)]
[IPV4]: Fix refcount damaging in net/ipv4/route.c

One such place that can damage the dst refcnts is route.c with
CONFIG_IP_ROUTE_MULTIPATH_CACHED enabled, i don't see the user's
.config. In this new code i see that rt_intern_hash is called before
dst->refcnt is set to 1, dst is the 2nd arg to rt_intern_hash.

Arg 2 of rt_intern_hash must come with refcnt 1 as it is added to
table or dropped depending on error/add/update. One such example is
ip_mkroute_input where __mkroute_input return rth with refcnt 0 which
is provided to rt_intern_hash. ip_mkroute_output looks like a 2nd such
place. Appending untested patch for comments and review.  The idea is
to put previous reference as we are going to return next result/error.

Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years agoNTFS: Make ntfs_write_block() not instantiate sparse blocks if they are zero.
Anton Altaparmakov [Thu, 8 Sep 2005 20:25:48 +0000 (21:25 +0100)]
NTFS: Make ntfs_write_block() not instantiate sparse blocks if they are zero.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Fixup handling of sparse, compressed, and encrypted attributes in
Anton Altaparmakov [Thu, 8 Sep 2005 20:19:45 +0000 (21:19 +0100)]
NTFS: Fixup handling of sparse, compressed, and encrypted attributes in
      fs/ntfs/inode.c::ntfs_read_locked_{,attr_,index_}inode().

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[BNX2]: Fix bug in irq handler and add prefetch
Michael Chan [Thu, 8 Sep 2005 20:15:32 +0000 (13:15 -0700)]
[BNX2]: Fix bug in irq handler and add prefetch

Fix bug in bnx2_interrupt() that caused an unnecessary register read.
The BNX2_PCICFG_MISC_STATUS should only be read when the status tag
has not changed.

Add prefetch of the status block in bnx2_msi() similar to tg3_msi().
The status block is not touched in bnx2_msi() and prefetching it will
speed up bnx2_poll() that will run on the same CPU that received the
MSI.

Update version.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years agoNTFS: Truncate {a,c,m}time to the ntfs supported time granularity when
Anton Altaparmakov [Thu, 8 Sep 2005 20:15:09 +0000 (21:15 +0100)]
NTFS: Truncate {a,c,m}time to the ntfs supported time granularity when
      updating the times in the inode in ntfs_setattr().

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Fix cluster (de)allocators to work when the runlist is NULL and more
Anton Altaparmakov [Thu, 8 Sep 2005 20:09:06 +0000 (21:09 +0100)]
NTFS: Fix cluster (de)allocators to work when the runlist is NULL and more
      importantly to take a locked runlist rather than them locking it
      which leads to lock reversal.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[I2C] Clean up i2c-pxa debugging/printks
Russell King [Thu, 8 Sep 2005 20:04:58 +0000 (21:04 +0100)]
[I2C] Clean up i2c-pxa debugging/printks

Fix up comments made by review by gregkh.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years agoNTFS: Fix handling of sparse attributes in ntfs_attr_make_non_resident().
Anton Altaparmakov [Thu, 8 Sep 2005 20:01:17 +0000 (21:01 +0100)]
NTFS: Fix handling of sparse attributes in ntfs_attr_make_non_resident().
      Also, add BUG() checks to ntfs_attr_make_non_resident() and
      ntfs_attr_set() to ensure that these functions are never called
      for compressed or encrypted attributes.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoMerge git://git.skbuff.net/gitroot/yoshfuji/linux-2.6-git-rfc3542
David S. Miller [Thu, 8 Sep 2005 19:59:43 +0000 (12:59 -0700)]
Merge git://git.skbuff.net/gitroot/yoshfuji/linux-2.6-git-rfc3542

18 years ago[IPV6]: Need to use pskb_trim_rcsum().
Stephen Hemminger [Thu, 8 Sep 2005 19:57:43 +0000 (12:57 -0700)]
[IPV6]: Need to use pskb_trim_rcsum().

Fix pskb_trim usage in ipv6. Only the udp one is really
a bug, other places are just doing equivalent code.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years agoNTFS: Fix several bugs in fs/ntfs/attrib.c.
Anton Altaparmakov [Thu, 8 Sep 2005 19:56:09 +0000 (20:56 +0100)]
NTFS: Fix several bugs in fs/ntfs/attrib.c.

- Fix a bug in ntfs_map_runlist_nolock() where we forgot to protect
  access to the allocated size in the ntfs inode with the size lock.
- Fix ntfs_attr_vcn_to_lcn_nolock() and ntfs_attr_find_vcn_nolock() to
  return LCN_ENOENT when there is no runlist and the allocated size is
  zero.
- Fix load_attribute_list() to handle the case of a NULL runlist.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Add fs/ntfs/attrib.[hc]::ntfs_resident_attr_value_resize().
Anton Altaparmakov [Thu, 8 Sep 2005 19:40:32 +0000 (20:40 +0100)]
NTFS: Add fs/ntfs/attrib.[hc]::ntfs_resident_attr_value_resize().

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Remove bogus setting of PageError in ntfs_read_compressed_block().
Anton Altaparmakov [Thu, 8 Sep 2005 19:35:33 +0000 (20:35 +0100)]
NTFS: Remove bogus setting of PageError in ntfs_read_compressed_block().

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years ago[NET]: Need struct sock forward decl in net/compat.h
David S. Miller [Thu, 8 Sep 2005 19:32:46 +0000 (12:32 -0700)]
[NET]: Need struct sock forward decl in net/compat.h

Else we get build failures like:

  CC      arch/sparc64/kernel/sparc64_ksyms.o
In file included from arch/sparc64/kernel/sparc64_ksyms.c:28:
include/net/compat.h:37: warning: "struct sock" declared inside parameter list
include/net/compat.h:37: warning: its scope is only this definition or declaration, which is probably not what you want

Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[IPV4] udp: trim forgets about CHECKSUM_HW
Stephen Hemminger [Thu, 8 Sep 2005 19:32:21 +0000 (12:32 -0700)]
[IPV4] udp: trim forgets about CHECKSUM_HW

A UDP packet may contain extra data that needs to be trimmed off.
But when doing so, UDP forgets to fixup the skb checksum if CHECKSUM_HW
is being used.

I think this explains the case of a NFS receive using skge driver
causing 'udp hw checksum failures' when interacting with a crufty
settop box.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[NET]: Optimize pskb_trim_rcsum()
Stephen Hemminger [Thu, 8 Sep 2005 19:32:03 +0000 (12:32 -0700)]
[NET]: Optimize pskb_trim_rcsum()

Since packets almost never contain extra garbage at the end, it is
worthwhile to optimize for that case.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years agoNTFS: Fix a bug in fs/ntfs/index.c::ntfs_index_lookup(). When the returned
Anton Altaparmakov [Thu, 8 Sep 2005 19:29:50 +0000 (20:29 +0100)]
NTFS: Fix a bug in fs/ntfs/index.c::ntfs_index_lookup().  When the returned
      index entry is in the index root, we forgot to set the @ir pointer in
      the index context.  Thanks for Yura Pakhuchiy for finding this bug.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Add ntfs_rl_punch_nolock() which punches a caller specified hole into a runlist.
Anton Altaparmakov [Thu, 8 Sep 2005 19:26:34 +0000 (20:26 +0100)]
NTFS: Add ntfs_rl_punch_nolock() which punches a caller specified hole into a runlist.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
18 years agoNTFS: Change ntfs_rl_truncate_nolock() to throw away the runlist if the new
Anton Altaparmakov [Thu, 8 Sep 2005 19:23:06 +0000 (20:23 +0100)]
NTFS: Change ntfs_rl_truncate_nolock() to throw away the runlist if the new
      length is zero.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>