Merge branch 'upstream-davem' of /linux/kernel/git/jgarzik/netdev-2.6

Merge branch 'for-jeff' of git://git./linux/kernel/git/chris/linux-2.6 into tmp

WAN: remove extra help text from HDLC_PPP config option

Remove extra help text from HDLC_PPP config option.

Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net/core: Allow receive on active slaves.

If a packet_type specifies an active slave to bonding and not just any
interface, allow it to receive frames that came in on that interface.

Signed-off-by: Joe Eykholt <jre@nuovasystems.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net/core: Allow certain receives on inactive slave.

Allow a packet_type that specifies the exact device to receive
even on an inactive bonding slave devices.  This is important for some
L2 protocols such as LLDP and FCoE.  This can eventually be used
for the bonding special cases as well.

Signed-off-by: Joe Eykholt <jre@nuovasystems.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net/core: Uninline skb_bond().

Otherwise subsequent changes need multiple return values.

Signed-off-by: Joe Eykholt <jre@nuovasystems.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

bonding: Don't destroy bonding master when removing slave via sysfs

It is wrong to destroy a bonding master from a context that uses the sysfs
of that bond. When last IPoIB slave is unenslaved from by writing to a
sysfs file (for bond0 this would be /sys/class/net/bond0/bonding/slaves)
the driver tries to destroy the bond. This is wrong and can lead to a
lockup or a crash.  This fix lets the bonding master stay and relies on
the user to destroy the bonding master if necessary (i.e. before module
ib_ipoib is unloaded)

This patch affects only bonds of IPoIB slaves. Ethernet slaves stay
unaffected.

Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

bonding: refactor mii monitor

Refactor mii monitor.  As with the previous ARP monitor refactor,
the motivation for this is to handle locking rationally (in this case,
removing conditional locking) and generally clean up the code.

This patch breaks up the monolithic mii monitor into two phases:
an inspection phase, followed by an optional commit phase.  The commit phase
is the only portion that requires RTNL or makes changes to state, and is
only called when inspection finds something to change.

Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

tcp: Fix kernel panic when calling tcp_v(4/6)_md5_do_lookup

If the following packet flow happen, kernel will panic.
MathineA MathineB
SYN
---------------------->
         SYN+ACK
<----------------------
ACK(bad seq)
---------------------->
When a bad seq ACK is received, tcp_v4_md5_do_lookup(skb->sk, ip_hdr(skb)->daddr))
is finally called by tcp_v4_reqsk_send_ack(), but the first parameter(skb->sk) is
NULL at that moment, so kernel panic happens.
This patch fixes this bug.

OOPS output is as following:
[  302.812793] IP: [<c05cfaa6>] tcp_v4_md5_do_lookup+0x12/0x42
[  302.817075] Oops: 0000 [#1] SMP
[  302.819815] Modules linked in: ipv6 loop dm_multipath rtc_cmos rtc_core rtc_lib pcspkr pcnet32 mii i2c_piix4 parport_pc i2c_core parport ac button ata_piix libata dm_mod mptspi mptscsih mptbase scsi_transport_spi sd_mod scsi_mod crc_t10dif ext3 jbd mbcache uhci_hcd ohci_hcd ehci_hcd [last unloaded: scsi_wait_scan]
[  302.849946]
[  302.851198] Pid: 0, comm: swapper Not tainted (2.6.27-rc1-guijf #5)
[  302.855184] EIP: 0060:[<c05cfaa6>] EFLAGS: 00010296 CPU: 0
[  302.858296] EIP is at tcp_v4_md5_do_lookup+0x12/0x42
[  302.861027] EAX: 0000001e EBX: 00000000 ECX: 00000046 EDX: 00000046
[  302.864867] ESI: ceb69e00 EDI: 1467a8c0 EBP: cf75f180 ESP: c0792e54
[  302.868333]  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
[  302.871287] Process swapper (pid: 0, ti=c0792000 task=c0712340 task.ti=c0746000)
[  302.875592] Stack: c06f413a 00000000 cf75f180 ceb69e00 00000000 c05d0d86 000016d0 ceac5400
[  302.883275]        c05d28f8 000016d0 ceb69e00 ceb69e20 681bf6e3 00001000 00000000 0a67a8c0
[  302.890971]        ceac5400 c04250a3 c06f413a c0792eb0 c0792edc cf59a620 cf59a620 cf59a634
[  302.900140] Call Trace:
[  302.902392]  [<c05d0d86>] tcp_v4_reqsk_send_ack+0x17/0x35
[  302.907060]  [<c05d28f8>] tcp_check_req+0x156/0x372
[  302.910082]  [<c04250a3>] printk+0x14/0x18
[  302.912868]  [<c05d0aa1>] tcp_v4_do_rcv+0x1d3/0x2bf
[  302.917423]  [<c05d26be>] tcp_v4_rcv+0x563/0x5b9
[  302.920453]  [<c05bb20f>] ip_local_deliver_finish+0xe8/0x183
[  302.923865]  [<c05bb10a>] ip_rcv_finish+0x286/0x2a3
[  302.928569]  [<c059e438>] dev_alloc_skb+0x11/0x25
[  302.931563]  [<c05a211f>] netif_receive_skb+0x2d6/0x33a
[  302.934914]  [<d0917941>] pcnet32_poll+0x333/0x680 [pcnet32]
[  302.938735]  [<c05a3b48>] net_rx_action+0x5c/0xfe
[  302.941792]  [<c042856b>] __do_softirq+0x5d/0xc1
[  302.944788]  [<c042850e>] __do_softirq+0x0/0xc1
[  302.948999]  [<c040564b>] do_softirq+0x55/0x88
[  302.951870]  [<c04501b1>] handle_fasteoi_irq+0x0/0xa4
[  302.954986]  [<c04284da>] irq_exit+0x35/0x69
[  302.959081]  [<c0405717>] do_IRQ+0x99/0xae
[  302.961896]  [<c040422b>] common_interrupt+0x23/0x28
[  302.966279]  [<c040819d>] default_idle+0x2a/0x3d
[  302.969212]  [<c0402552>] cpu_idle+0xb2/0xd2
[  302.972169]  =======================
[  302.974274] Code: fc ff 84 d2 0f 84 df fd ff ff e9 34 fe ff ff 83 c4 0c 5b 5e 5f 5d c3 90 90 57 89 d7 56 53 89 c3 50 68 3a 41 6f c0 e8 e9 55 e5 ff <8b> 93 9c 04 00 00 58 85 d2 59 74 1e 8b 72 10 31 db 31 c9 85 f6
[  303.011610] EIP: [<c05cfaa6>] tcp_v4_md5_do_lookup+0x12/0x42 SS:ESP 0068:c0792e54
[  303.018360] Kernel panic - not syncing: Fatal exception in interrupt

Signed-off-by: Gui Jianfeng <guijianfeng@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

pkt_sched: Fix "parent is root" test in qdisc_create().

As noticed by Stephen Hemminger, the root qdisc is denoted by
TC_H_ROOT, not zero.

Signed-off-by: David S. Miller <davem@davemloft.net>

[netdrvr] Drivers should not set IFF_* flag themselves

Some hardware set promisc when they are requested to set IFF_ALLMULTI flag.
It's ok, but if drivers set IFF_PROMISC flag when they set promisc,
it will broken upper layer handle for promisc and allmulti.
In addition, drivers can use their own hardware programming to make it.
So do not allow drivers to set IFF_* flags.

This is a general driver fix, so I didn't split it to pieces and send
to specific driver maintainers.

Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

dm9000: Support MAC address setting through platform data.

The dm9000 driver reads the chip's MAC address from the attached EEPROM. When
no EEPROM is present, or when the MAC address is invalid, it falls back to
reading the address from the chip.

This patch lets platform code set the desired MAC address through platform
data.

Signed-off-by: Laurent Pinchart <laurentp@cse-semaphore.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

3c59x: use netdev_alloc_skb

Fix possible bug where end of receive buffer could be overwritten.
The allocation needs to allow for the reserved space. This would only happen
if device received packet greater than Ethernet standard MTU.

Change this driver to use netdev_alloc_skb rather than setting skb->dev
directly. For the initial allocation it doesn't need to be GFP_ATOMIC.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ethtool: Expand ethtool_cmd.speed to 32 bits

Introduce the speed_hi field to ethtool_cmd, using the reserved space,
to expand the speed field to 2^32 Megabits/second.

Making this field expansion now gives us plenty of time to fix up the
user-space pieces that use SIOCETHTOOL before hardware faster than 64
Gb/s is available.

Signed-off-by: Brandon Philips <bphilips@suse.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

pegasus: add blacklist support to fix Belkin bluetooth dongle.

Reference: https://launchpad.net/bugs/140511

The Belkin bluetooth dongle unfortunately shares the vendor and device id
with the network adapter which causes lockups whenever the bluetooth
dongle is inserted.

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Ben Collins <ben.collins@canonical.com>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Cc: David Brownell <david-b@pacbell.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

drivers/net/netxen/netxen_nic_hw.c: fix printk warnings

drivers/net/netxen/netxen_nic_hw.c: In function 'netxen_nic_pci_mem_read_direct':
drivers/net/netxen/netxen_nic_hw.c:1414: warning: format '%llx' expects type 'long long unsigned int', but argument 3 has type 'u64'
drivers/net/netxen/netxen_nic_hw.c: In function 'netxen_nic_pci_mem_write_direct':
drivers/net/netxen/netxen_nic_hw.c:1487: warning: format '%llx' expects type 'long long unsigned int', but argument 3 has type 'u64'

You don't know what type was used for u64 hence they cannot be printed without
casting.

Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

hamradio: add missing sanity check to tty operation

Add missing sanity check to tty operation.

Acked-by: Alan Cox <alan@redhat.com>
Signed-off-by: Eugene Teo <eteo@redhat.com>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

drivers/net: coding styles fixes to drivers/net/8390.c

Fix all errors and warnings reported by checkpatch.pl

Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

drivers/net: coding styles fixes to drivers/net/8390p.c

Fix all errors and warnings reported by checkpatch.pl

Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

drivers/net: Remove 'return' of void function NS8390p_init()

We don't need this into a void function.

Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

[netdrvr] sh_eth: Add SH7619 support

Add support SH7619 Internal ethernet controler.

Signed-off-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

METH: fix MAC address setup

Setup of the mac filter lost the upper 16bit of the mac address. This
bug got unconvered by a patch, which fixed the promiscous handling.

Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

xen-netfront: use netif_start_queue() on xennet_open()

xen-netfront never called netif_start_queue() and was was waking the
queue on xennet_open(), triggering the BUG_ON() on __netif_schedule().

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qla3xxx: driver version change.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qla3xxx: cleanup: Remove some unused structure definitions and structure elements.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qla3xxx: cleanup: Remove some unused defined constants in the header file.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qla3xxx: fix: Fix IFF_MULTICAST setting.

The driver was erroneously clearing this bit though the hardware supports multicast.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qla3xxx: fix: Remove unused set_multicast function.

This device is one side of a two-function adapter (NIC and iSCSI).
Promiscuous mode setting/clearing is not allowed from the NIC side.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

netxen: fix promisc mode, mtu setting

For NX3031, multicast filtering, promisc mode, and max frame size
setting is handled by firmware, driver needs to send request to
enable/disable it.

For old chip revisions / firmware, driver still sets it directly.

Added function pointer to set mtu according to chip revision.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

netxen: fix cmd ring init

Initialize producer and consumer indices during netdev open(), only
for old firmware/chip.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

netxen: fix legacy interrupts

Fix legacy interrupt mode for NX3031 chips, read pci interrupt state
in hardware to guard against spurious interrupt.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

netxen: add new board types

Add couple of new board configurations based on NX3031 chip.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

netxen: fix link status, link speed

For NX3031, the phy is managed by firmware, so driver should avoid
setting any phy registers.

Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

remove bogus CONFIG_GFAR_NAPI's

The commit that made the CONFIG_GFAR_NAPI code unconditional was
included at the same time as a new CONFIG_GFAR_NAPI user, resulting
in these bugus #ifdef's.

Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net/fs_enet: remove redundant messages for performance

Currently when we do a packet flood to the Ethernet port, the console
reports error every time when a packet is dropped.  This is too
redundant and cost performance.  Remove message for this type of event.

Signed-off-by: Li Yang <leoli@freescale.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

forcedeth: add tx pause limit

This patch adds support for limiting the number of tx pause frames to a
default of 8. Previously, hardware would send out continuous stream of
pause frames.

Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

forcedeth: add jumbo frame support for mcp79

This patch adds jumbo frame support for MCP79 chipsets.

Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

forcedeth: add new tx stat counters

This patch adds support for new tx statistic counters in the hardware -
unicast, multicast, and broadcast

Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

forcedeth: fix rx error policy

This patch enforces a stricter policy on rx errors. The driver needs to
verify whether there are multiple rx errors versus a single error.

Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Merge branch 'via-velocity' of git://git./linux/kernel/git/romieu/netdev-2.6 into tmp

sky2: fix PM related regressions

Fix the problems reported for 2.6.27-rc1 caused by over aggressive
power management. Turning clock off on PCI Express is problematic for WOL,
and when doing multi-booting.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: remove igb_init_managability as it is deprecated

igb_init_managability does not actually perform any function as the two
registers it attempts to write are both read only on the host.  This patch
removes the function and all references to it from the driver.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: remove three redundant functions left in the code

Three functions were left in the code that are no longer used.  I am removing
these functions just to keep the code clean.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: correct issue of set_mta member of mac.ops not being populated

The igb_mta_set function was not being correctly used

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: fixes 82576 serdes init to correctly support manual flow control changes

This patch changes the PCS configuration for serdes so that the flow
control options change be set via the ethtool -A option.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: fix null pointer dereference seen with fiber NICs

With a fiber or serdes NIC installed the driver was causing a null pointer
dereference on driver unload.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: fix comments

The internal name was used in comments, replaced with silicon part number.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

atl1: deal with hardware rx checksum bug

The L1 hardware contains a bug that flags a fragmented IP packet
as having an incorrect TCP/UDP checksum, even though the packet
is perfectly valid and its checksum is correct.  There's no way to
distinguish between one of these good packets and a packet that
actually contains a TCP/UDP checksum error, so all we can do is
allow the packet to be handed up to the higher layers and let it
be sorted out there.

Add a comment describing this condition and remove the code that
currently fails to handle what may or may not be a checksum error.

Signed-off-by: Jay Cliburn <jacliburn@bellsouth.net>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

e1000e: remove inapplicable test for ioport

There are currently no devices supported by the e1000e driver which need
ioport resources, remove the test for it and all unnecessary code
associated with it (struct e1000_adapter elements, local variables, etc.)

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

e1000e: fix drv load issues

a few people seem to have problems maintaining gigabit link
and it was root caused to an interaction between the managability
firmware on the host and the driver, not communicating.

The form of communication they use is the drv_load bit.

Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

e1000e: perform basic 82573 EEPROM checks for known issues

82573 EEPROMs have been shipped out with known issues. While most
people will never see the issues some people do and we know
how to address them. Warn the user if we find one of these
EEPROM issues.

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

e1000e: convert ndev_ printks to something smaller

The ndev_* printk's are too lenghty and we don't need to specify
the adapter/netdev struct at all, making this a lot more readable.

Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

myri10ge: set 64bits consistent DMA mask

Set 64bits consistent DMA mask since it improves performance
in some cases. No need to check the return value since it is
not required for the driver to work.

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

myri10ge: update firmware headers

Update myri10ge firmware headers.

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ipv4: Fix over-ifdeffing of ip_static_sysctl_init.

Noticed by Paulius Zaleckas.

Signed-off-by: David S. Miller <davem@davemloft.net>

ipsec: Interfamily IPSec BEET, ipv4-inner ipv6-outer

Here's a revised version, based on Herbert's comments, of a fix for
the ipv4-inner, ipv6-outer interfamily ipsec beet mode. It fixes the
network header adjustment during interfamily, as well as makes sure
that we reserve enough room for the new ipv6 header if we might have
something else as the inner family. Also, the ipv4 pseudo header
construction was added.

Signed-off-by: Joakim Koskela <jookos@gmail.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipsec: Interfamily IPSec BEET

Here's a revised version, based on Herbert's comments, of a fix for
the ipv6-inner, ipv4-outer interfamily ipsec beet mode. It fixes the
network header adjustment in interfamily, and doesn't reserve space
for the pseudo header anymore when we have ipv6 as the inner family.

Signed-off-by: Joakim Koskela <jookos@gmail.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

netfilter: fix two recent sysctl problems

Starting with 9043476f726802f4b00c96d0c4f418dde48d1304 ("[PATCH]
sanitize proc_sysctl") we have two netfilter releated problems:

 - WARNING: at kernel/sysctl.c:1966 unregister_sysctl_table+0xcc/0x103(),
   caused by wrong order of ini/fini calls

 - net.netfilter is duplicated and has truncated set of records

Thanks to very useful guidelines from Al Viro, this patch fixes both
of them.

Signed-off-by: Krzysztof Piotr Oledzki <ole@ans.pl>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: replace dst_metric() with dst_mtu() in net/ipv6/route.c.

This patch replaces dst_metric() with dst_mtu() in net/ipv6/route.c.

Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv4: replace dst_metric() with dst_mtu() in net/ipv4/route.c.

This patch replaces dst_metric() with dst_mtu() in net/ipv4/route.c.

Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Linux 2.6.27-rc2

Re-introduce "[SCSI] extend the last_sector_bug flag to cover more sectors"

This re-introduces commit 2b142900784c6e38c8d39fa57d5f95ef08e735d8,
which was reverted due to the regression it caused by commit
fca082c9f1e11ec07efa8d2f9f13688521253f36.

That regression was not root-caused by the original commit, it was just
uncovered by it, and the real fix was done by Alan Stern in commit
580da34847488b404218d1d7f53b156f245f5555 ("Fix USB storage hang on
command abort").

We can thus re-introduce the change that was confirmed by Alan Jenkins
to be still required by his odd card reader.

Cc: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (78 commits)
  AX.25: Fix sysctl registration if !CONFIG_AX25_DAMA_SLAVE
  pktgen: mac count
  pktgen: random flow
  bridge: Eliminate unnecessary forward delay
  bridge: fix compile warning in net/bridge/br_netfilter.c
  ipv4: remove unused field in struct flowi (include/net/flow.h).
  tg3: Fix 'scheduling while atomic' errors
  net: Kill plain NET_XMIT_BYPASS.
  net_sched: Add qdisc __NET_XMIT_BYPASS flag
  net_sched: Add qdisc __NET_XMIT_STOLEN flag
  iwl3945: fix merge mistake for packet injection
  iwlwifi: grap nic access before accessing periphery registers
  iwlwifi: decrement rx skb counter in scan abort handler
  iwlwifi: fix unhandled interrupt when HW rfkill is on
  iwlwifi: implement iwl5000_calc_rssi
  iwlwifi: memory allocation optimization
  iwlwifi: HW bug fixes
  p54: Fix potential concurrent access to private data
  rt2x00: Disable link tuning in rt2500usb
  iwlwifi: Don't use buffer allocated on the stack for led names
  ...

AX.25: Fix sysctl registration if !CONFIG_AX25_DAMA_SLAVE

Since 49ffcf8f99e8d33ec8afb450956804af518fd788 ("sysctl: update
sysctl_check_table") setting struct ctl_table.procname = NULL does no
longer work as it used to the way the AX.25 code is expecting it to
resulting in the AX.25 sysctl registration code to break if
CONFIG_AX25_DAMA_SLAVE was not set as in some distribution kernels.
Kernel releases from 2.6.24 are affected.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

pktgen: mac count

dst_mac_count and src_mac_count patch from Eneas Hunguana
We have sent one mac address to much.

Signed-off-by: Robert Olsson <robert.olsson@its.uu.se>
Signed-off-by: David S. Miller <davem@davemloft.net>

pktgen: random flow

Random flow generation has not worked. This fixes it.

Signed-off-by: Robert Olsson <robert.olsson@its.uu.se>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: Eliminate unnecessary forward delay

From: Stephen Hemminger <shemminger@vyatta.com>

Based upon original patch by Herbert Xu, which contained
the following problem description:

--------------------
When the forward delay is set to zero, we still delay the setting
of the forwarding state by one or possibly two timers depending
on whether STP is enabled.  This could either turn out to be
instantaneous, or horribly slow depending on the load of the
machine.

As there is nothing preventing us from enabling forwarding straight
away, this patch eliminates this potential delay by executing the
code directly if the forward delay is zero.

The effect of this problem is that immediately after the carrier
comes on a port, the bridge will drop all packets received from
that port until it enters forwarding mode, thus causing unnecessary
packet loss.

Note that this patch doesn't fully remove the delay due to the
link watcher.  We should also check the carrier state when we
are about to drop an incoming packet because the port is disabled.
But that's for another patch.
--------------------

This version of the fix takes a different approach, in that
it just does the state change directly.

Signed-off-by: David S. Miller <davem@davemloft.net>

docbook: fix s390 fatal error after header files moved

Fix (fatal error) s390 docbook file to handle renamed header files.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

drivers/media/video/vino.c needs v4l2-ioctl.h

mips allmodconfig:

  drivers/media/video/vino.c: In function `vino_free_buffer_with_count':
  drivers/media/video/vino.c:811: warning: passing arg 1 of `virt_to_phys' makes pointer from integer without a cast
  drivers/media/video/vino.c: In function `vino_allocate_buffer':
  drivers/media/video/vino.c:889: warning: passing arg 1 of `virt_to_phys' makes pointer from integer without a cast
  drivers/media/video/vino.c: In function `vino_ioctl':
  drivers/media/video/vino.c:4364: error: implicit declaration of function `video_usercopy'

The patch fixes the error, but not the warnings.

Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: fix uninitialized variables for find_vma_prepare callers

gcc 4.3.0 correctly emits the following warnings.
When a vma covering addr is found, find_vma_prepare indeed returns without
setting pprev, rb_link, and rb_parent.

  mm/mmap.c: In function `insert_vm_struct':
  mm/mmap.c:2085: warning: `rb_parent' may be used uninitialized in this function
  mm/mmap.c:2085: warning: `rb_link' may be used uninitialized in this function
  mm/mmap.c:2084: warning: `prev' may be used uninitialized in this function
  mm/mmap.c: In function `copy_vma':
  mm/mmap.c:2124: warning: `rb_parent' may be used uninitialized in this function
  mm/mmap.c:2124: warning: `rb_link' may be used uninitialized in this function
  mm/mmap.c:2123: warning: `prev' may be used uninitialized in this function
  mm/mmap.c: In function `do_brk':
  mm/mmap.c:1951: warning: `rb_parent' may be used uninitialized in this function
  mm/mmap.c:1951: warning: `rb_link' may be used uninitialized in this function
  mm/mmap.c:1949: warning: `prev' may be used uninitialized in this function
  mm/mmap.c: In function `mmap_region':
  mm/mmap.c:1092: warning: `rb_parent' may be used uninitialized in this function
  mm/mmap.c:1092: warning: `rb_link' may be used uninitialized in this function
  mm/mmap.c:1089: warning: `prev' may be used uninitialized in this function

Hugh adds: in fact, none of find_vma_prepare's callers use those values
when a vma is found to be already covering addr, it's either an error or
an occasion to munmap and repeat.  Okay, let's quieten the compiler (but I
would prefer it if pprev, rb_link and rb_parent were meaningful in that
case, rather than whatever's in them from descending the tree).

Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: "Ryan Hope" <rmh3093@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

pm_qos: spelling fixes

A documentation cleanup patch.  With a minor tweak to clarify units for
kbs.

[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: mark gross <mgross@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

proc: fix warnings

proc: fix warnings

 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 3 has type 'u64'
 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 4 has type 'u64'
 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 5 has type 'u64'
 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 6 has type 'u64'
 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 7 has type 'u64'
 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 8 has type 'u64'
 fs/proc/base.c:2429: warning: format '%llu' expects type 'long long unsigned int', but argument 9 has type 'u64'

Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Acked-by: Andrea Righi <righi.andrea@gmail.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

drivers/misc/sgi-gru/grutlbpurge.c: removed duplicated #include

Removed duplicated include <linux/delay.h> in
drivers/misc/sgi-gru/grutlbpurge.c.

Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com>
Cc: Jack Steiner <steiner@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

omfs: fix warning

fs/omfs/inode.c:495: warning: format '%llx' expects type 'long long
unsigned int', but argument 2 has type 'u64'
fs/omfs/inode.c:495: warning: format '%llx' expects type 'long
long unsigned int', but argument 3 has type '__be64'

Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Acked-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kernel-doc: skip nested struct/union cleanly

Fix handling of nested structs or unions.  The regex to strip (eliminate)
nested structs or unions was limited to only 0 or 1 matches.  This can
cause an uneven number of left/right braces to be stripped, which causes
this:

Warning(linux-2.6.27-rc1-git2//include/net/mac80211.h:336): No description found for parameter '}'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

dma: fix order calculation in dma_mark_declared_memory_occupied()

get_order() takes byte-sized input, not a page-granular one.

Irrespective of this fix I'm inclined to believe that this doesn't work
right anyway - bitmap_allocate_region() has an implicit assumption of
'pos' being suitable for 'order', which this function doesn't seem to
enforce (and since it's being called with a byte-granular value there's no
reason to believe that the callers would make sure device_addr is passed
accordingly - it's also not documented that way).

Signed-off-by: Jan Beulich <jbeulich@novell.com>
Cc: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Dmitry Baryshkov <dbaryshkov@gmail.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

atyfb_base.c: fix warning

drivers/video/aty/atyfb_base.c:2663: warning: 'aty_resume_chip' defined
but not used

Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Cc: Ville Syrjala <syrjala@sci.fi>
Cc: Antonino A. Daplas <adaplas@pol.net>
Cc: Mark Asselstine <asselsm@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

drivers/video/fsl-diu-fb.c: add missing of_node_put

of_node_put is needed before discarding a value received from
of_find_node_by_type, eg in error handling code.

The semantic patch that makes the change is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@@
struct device_node *n;
struct device_node *n1;
struct device_node *n2;
statement S;
identifier f1,f2;
expression E1,E2;
constant C;
@@

n = of_find_node_by_type(...)
...
if (!n) S
... when != of_node_put(n)
    when != n1 = f1(n,...)
    when != E1 = n
    when any
    when strict
(
+ of_node_put(n);
  return -C;
|
  of_node_put(n);
|
  n2 = f2(n,...)
|
  E2 = n
|
  return ...;
)
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Acked-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Cc: Timur Tabi <timur@freescale.com>
Cc: York Sun <yorksun@freescale.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

radeonfb: give i2c buses nicer names

The name of the i2c buses shows in the output of "i2cdetect -l", so
it's important to give informative names.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fbcon: fix scrolling after logo is cleared

If the 'clear' command is used on the frame buffer with a logo the upper
area is filled by few lines but not scrolled anymore.

Fix this by removing the protected area for the logo if any part of the
logo is cleared.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

radeon: misc corrections

I have a new PCI-E radeon RV380 series card (PCI device ID 5b64) that
hangs in my sparc64 boxes when the init scripts set the font.  The problem
goes away if I disable acceleration.

I haven't figured out that bug yet, but along the way I found some
corrections to make based upon some auditing.

1) The RB2D_DC_FLUSH_ALL value used by the kernel fb driver
   and the XORG video driver differ.  I've made the kernel
   match what XORG is using.

2) In radeonfb_engine_reset() we have top-level code structure
   that roughly looks like:

if (family is 300, 350, or V350)
do this;
else
do that;
...
if (family is NOT 300, OR
    family is NOT 350, OR
    family is NOT V350)
do another thing;

   this last conditional makes no sense, is always true,
   and obviously was likely meant to be "family is NOT
   300, 350, or V350".  So I've made the code match the
   intent.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Tested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Remove the deprecated cli() sti() functions

These functions have been deprecated for some time now but remained until
all legacy callers could be removed.  With a few commits in 2.6.26 this
has happened so now we can remove these deprecated functions.

Signed-off-by: Mark Asselstine <mark.asselstine@windriver.com>
Reviewed-by: Matthew Wilcox <willy@linux.intel.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

backlight: add MODULE_ALIAS() to pwm_backlight driver

Add the missing MODULE_ALIAS() to the pwm_backlight driver.

Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

backlight: add more information output to pwm_backlight

Make the error paths in the pwm_backlight driver more informative in the
probe path, especially for the times that it finds an error.

Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Cc: Richard Purdie <rpurdie@rpsys.net>
Acked-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: drop PIE/stopwatch code since the hardware can only do a max of 1HZ and this same functionality is provided by UIE

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: convert PIE handling to irq_set_state() as pointed out by David Brownell

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: wait for the write complete interrupt complete before sleeping

Since we use the write complete interrupt, wait for it
to complete before sleeping so we don't wake right back up due to it

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: disable the write complete irq upon close

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: don't bother passing the rtc struct down to bfin_rtc_int_{set,clear} since it isnt needed (shaves off ~100bytes)

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: add support for power management framework

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: Fix bug Only RTC interrupt can wake up deeper sleep core

Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blackfin RTC driver: if we dont define irq_set_freq, the common rtc-dev layer will give us the same behavior of returning ENOTTY

Signed-off-by: Mike Frysinger <michael.frysinger@analog.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

genirq: better warning on irqchip->set_type() failure

While I'm glad to finally see the hole fixed whereby passing an invalid
IRQ trigger type to request_irq() would be ignored, the current diagnostic
isn't quite useful.  Fixed by also listing the trigger type which was
rejected.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Acked-by: Uwe Kleine-König <Uwe.Kleine-Koenig@digi.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

semaphore: __down_common: use signal_pending_state()

Change __down_common() to use signal_pending_state() instead of open
coding.

The changes in kernel/semaphore.o are just artifacts, the state checks are
optimized away.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

relay: fix "full buffer with exactly full last subbuffer" accounting problem

In relay's current read implementation, if the buffer is completely full
but hasn't triggered the buffer-full condition (i.e. the last write
didn't cross the subbuffer boundary) and the last subbuffer is exactly
full, the subbuffer accounting code erroneously finds nothing available.
This patch fixes the problem.

Signed-off-by: Tom Zanussi <tzanussi@gmail.com>
Cc: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: Jens Axboe <jens.axboe@oracle.com>
Cc: Mathieu Desnoyers <compudj@krystal.dyndns.org>
Cc: Andrea Righi <righi.andrea@gmail.com>
Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

spi: new orion_spi driver

This adds an SPI driver for the SPI controller found in various Marvell
Orion ARM SoCs.  It currently supports only one slave, which must use SPI
mode 0.

[dbrownell@users.sourceforge.net: cleanups, meet specs, pass "sparse"]
Signed-off-by: Shadi Ammouri <shadi@marvell.com>
Signed-off-by: Saeed Bishara <saeed@marvell.com>
Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

block/cciss.c: remove pointless curr_queue calculation

curr_queue is a local variable in a for loop, and it's being initialized
at the start of each loop.  So any assignment at the end of the loop is
pointless.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Cc: Mike Miller <mike.miller@hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

vt8623fb: fix kernel oops

commit 20e061fb750d36ec0ffcb2e44ed7dafa9018223b
  Author: Ondrej Zajicek <santiago@crfreenet.org>
  Date:   Mon Apr 28 02:15:18 2008 -0700

      fbdev: framebuffer_alloc() fixes

      Correct the dev arg of framebuffer_alloc() in arkfb, s3fb and vt8623fb.

causes a null-pointer deref because "info->dev is NULL, info was just
kzallocated".

Signed-off-by: Ondrej Zajicek <santiago@crfreenet.org>
Reported-by: "MadLoisae@gmx.net" <MadLoisae@gmx.net>
Tested-by: "MadLoisae@gmx.net" <MadLoisae@gmx.net>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Cc: Krzysztof Helt <krzysztof.h1@poczta.fm>
Cc: <stable@kernel.org> [2.6.26.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kdump: report actual value of VMCOREINFO_OSRELEASE in VMCOREINFO

The current implementation reports the structure name as
VMCOREINFO_OSRELEASE in VMCOREINFO, e.g.

        VMCOREINFO_OSRELEASE=init_uts_ns.name.release

That doesn't make sense because it's always the same. Instead, use the
value, e.g.

        VMCOREINFO_OSRELEASE=2.6.26-rc3

That's also what the 'makedumpfile -g' does.

Signed-off-by: Bernhard Walle <bwalle@suse.de>
Cc: "Ken'ichi Ohmichi" <oomichi@mxs.nes.nec.co.jp>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

powerpc: 86xx: mpc8610_hpcd: add watchdog node

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: Wim Van Sebroeck <wim@iguana.be>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>