The RDMA CM is intended to support the use of a loopback address
when establishing a connection; however, the behavior of the CM
when loopback addresses are used is confusing and does not always
work, depending on whether loopback was specified by the server,
the client, or both.

The defined behavior of rdma_bind_addr is to associate an RDMA
device with an rdma_cm_id, as long as the user specified a non-
zero address.  (ie they weren't just trying to reserve a port)
Currently, if the loopback address is passed to rdam_bind_addr,
no device is associated with the rdma_cm_id.  Fix this.

If a loopback address is specified by the client as the destination
address for a connection, it will fail to establish a connection.
This is true even if the server is listing across all addresses or
on the loopback address itself.  The issue is that the server tries
to translate the IP address carried in the REQ message to a local
net_device address, which fails.  The translation is not needed in
this case, since the REQ carries the actual HW address that should
be used.

Finally, cleanup loopback support to be more transport neutral.
Replace separate calls to get/set the sgid and dgid from the
device address to a single call that behaves correctly depending
on the format of the device address.  And support both IPv4 and
IPv6 address formats.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>

[ Fixed RDS build by s/ib_addr_get/rdma_addr_get/  - Roland ]
Signed-off-by: Roland Dreier <rolandd@cisco.com>

The struct rdma_dev_addr stores net_device address information:
the source device address, destination hardware address, and
broadcast address.  For consistency, store the net_device type
rather than converting it to the rdma_node_type.

The type indicates the format of the various hardware addresses,
which is what we're concerned with, and not the RDMA node type
that the address may map to.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

If a source address is provided, verify that the address family matches
that of the destination address.  If the source is not specified, use the
same address family as the destination.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Provide the device interface when resolving route information to
ensure that the correct outbound device is used.  This will also
simplify processing of sin6_scope_id for IPv6 support.

Based on work from:
David Wilder <dwilder@us.ibm.com>
Jason Gunthorpe <jgunthrope@obsidianresearch.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

If joining to an AF_INET6 address, we need to map the address to a MGID
in the same way as the IP stack.  The old code would just fall through to
the IPv4 case and generate garbage.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

RDMA CM treats AF_INET6 addresses that are either 0 or prefixed with
FF1x:A01B::/32 as MGIDs, but the detection for the prefix was buggy;
fix it up.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

for_each_netdev() should be used with RTNL or dev_base_lock held,
or else we risk a crash.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Export rdma_set_ib_paths to user space to allow applications to
manually set the IB path used for connections.  This allows
alternative ways for a user space application or library to obtain
path record information, including retrieving path information
from cached data, avoiding direct interaction with the IB SA.
The IB SA is a single, centralized entity that can limit scaling
on large clusters running MPI applications.

Future changes to the rdma cm can expand on this framework to
support the full range of features allowed by the IB CM, such as
separate forward and reverse paths and APM.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Reviewed-By: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/anholt/drm-intel:
  drm/i915: Ironlake suspend/resume support
  drm/i915: kill warning in intel_find_pll_g4x_dp
  drm/i915: update watermarks before enabling PLLs
  drm/i915: add FIFO watermark support for G4x
  drm/i915: quiet DP i2c init
  drm/i915: fix panel fitting filter coefficient select for Ironlake
  drm/i915: fix to setup display reference clock control on Ironlake
  drm/i915: Install a fence register for fbc on g4x
  drm/i915: save/restore BLC histogram control reg across suspend/resume
  drm/i915: Fix FDI M/N setting according with correct color depth
  drm/i915: disable powersave feature for Ironlake currently
  drm/i915: Fix render reclock availability detection.
  drm/i915: Save and restore the GM45 FBC regs on suspend and resume.
  drm/i915: Set the LVDS_BORDER when using LVDS scaling mode
  drm/i915: disable FBC for Pineview, fixing a boot hang.

* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
  cfq-iosched: limit coop preemption
  cfq-iosched: fix bad return value cfq_should_preempt()
  backing-dev: bdi sb prune should be in the unregister path, not destroy
  Fix bio_alloc() and bio_kmalloc() documentation
  bio_put(): add bio_clone() to the list of functions in the comment

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  sata_via: Remove redundant device ID for VIA VT8261
  drivers/ata/libata: Move dereference after NULL test
  ahci: Enable SB600 64bit DMA on MSI K9A2 Platinum v2

Just remove redundant device ID for VIA VT8261.
The device ID 0x9000 and 0x9040 are redundant (for VT8261).
The 0x9040 is reserved for other usage.
Signed-off-by: Joseph Chan <josephchan@via.com.tw>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

In each case, if the NULL test on qc is needed, then the derefernce
should be after the NULL test.

A simplified version of the semantic match that detects this problem is as
follows (http://coccinelle.lip6.fr/):

// <smpl>
@match exists@
expression x, E;
identifier fld;
@@

* x->fld
  ... when != \(x = E\|&x\)
* x == NULL
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Like the Asus M2A-VM, MSI's K9A2 Platinum (MS-7376) can also support 64bit
DMA. It is a new enough board that all the BIOS releases work correctly with
64bit DMA enabled.
Signed-off-by: Mark Nelson <mdnelson8@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

CFQ has an optimization for cooperated applications. if several
io-context have close requests, they will get boost. But the
optimization get abused. Considering thread a, b, which work on one
file. a reads sectors s, s+2, s+4, ...; b reads sectors s+1, s+3, s
+5, ... Both a and b are sequential read, so they can open idle window.
a reads a sector s and goes to idle window and wakeup b. b reads sector
s+1, since in current implementation, cfq_should_preempt() thinks a and
b are cooperators, b will preempt a. b then reads sector s+1 and goes to
idle window and wakeup a. for the same reason, a will preempt b and
reads s+2. a and b will continue the circle. The circle will be very
long, and a and b will occupy whole disk queue. Other applications will
nearly have no chance to run.

Fix this limiting coop preempt until a queue is scheduled normally
again.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

Commit a6151c3a inadvertently reversed
a preempt condition check, potentially causing a performance regression.
Make the meta check correct again.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

Commit 592b09a4 was different from
the tested path, in that it moved the bdi super_block prune from
unregister to destroy context. This doesn't fully fix the sync hang
bug on unexpected device removal, as need to prune the bdi cache
pointer before killing flusher thread.
Tested-by: Artur Skawina <art.08.09@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

* 'for-linus' of git://github.com/at91linux/linux-2.6-at91:
  at91: at91sam9g45 family: identify several chip versions
  avr32: add two new at91 to cpu.h definition

cpu_is_xxx() macros are identifying generic at91sam9g45 chip. This patch adds
the capacity to differentiate Engineering Samples and final lots through the
inclusion of  at91_cpu_fully_identify() and the related chip IDs with chip
version field preserved.
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Acked-by: Hans-Christian Egtvedt <hans-christian.egtvedt@atmel.com>

Somme common drivers will need those at91 cpu_is_xxx() definitions. As
at91sam9g10 and at91sam9g45 are on the way to linus' tree, here is the patch
that adds those chips to cpu.h in AVR32 architecture.
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Haavard Skinnemoen <haavard.skinnemoen@atmel.com>

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus: (38 commits)
  MIPS: O32: Fix ppoll
  MIPS: Oprofile: Rename cpu_type from godson2 to loongson2
  MIPS: Alchemy: Fix hang with high-frequency edge interrupts
  MIPS: TXx9: Fix spi-baseclk value
  MIPS: bcm63xx: Set the correct BCM3302 CPU name
  MIPS: Loongson 2: Set cpu_has_dc_aliases and cpu_icache_snoops_remote_store
  MIPS: Avoid potential hazard on Context register
  MIPS: Octeon: Use lockless interrupt controller operations when possible.
  MIPS: Octeon: Use write_{un,}lock_irq{restore,save} to set irq affinity
  MIPS: Set S-cache linesize to 64-bytes for MTI's S-cache
  MIPS: SMTC: Avoid queing multiple reschedule IPIs
  MIPS: GCMP: Avoid accessing registers when they are not present
  MIPS: GIC: Random fixes and enhancements.
  MIPS: CMP: Fix memory barriers for correct operation of amon_cpu_start
  MIPS: Fix abs.[sd] and neg.[sd] emulation for NaN operands
  MIPS: SPRAM: Clean up support code a little
  MIPS: 1004K: Enable SPRAM support.
  MIPS: Malta: Enable PCI 2.1 compatibility in PIIX4
  MIPS: Kconfig: Fix duplicate default value for MIPS_L1_CACHE_SHIFT.
  MIPS: MTI: Fix accesses to device registers on MIPS boards
  ...

* 'pm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspend-2.6:
  PM: Remove some debug messages producing too much noise
  PM: Fix warning on suspend errors
  PM / Hibernate: Add newline to load_image() fail path
  PM / Hibernate: Fix error handling in save_image()
  PM / Hibernate: Fix blkdev refleaks
  PM / yenta: Split resume into early and late parts (rev. 4)

nr_processes() returns the sum of the per cpu counter process_counts for
all online CPUs. This counter is incremented for the current CPU on
fork() and decremented for the current CPU on exit(). Since a process
does not necessarily fork and exit on the same CPU the process_count for
an individual CPU can be either positive or negative and effectively has
no meaning in isolation.

Therefore calculating the sum of process_counts over only the online
CPUs omits the processes which were started or stopped on any CPU which
has since been unplugged. Only the sum of process_counts across all
possible CPUs has meaning.

The only caller of nr_processes() is proc_root_getattr() which
calculates the number of links to /proc as
        stat->nlink = proc_root.nlink + nr_processes();

You don't have to be all that unlucky for the nr_processes() to return a
negative value leading to a negative number of links (or rather, an
apparently enormous number of links). If this happens then you can get
failures where things like "ls /proc" start to fail because they got an
-EOVERFLOW from some stat() call.

Example with some debugging inserted to show what goes on:
        # ps haux|wc -l
        nr_processes: CPU0:     90
        nr_processes: CPU1:     1030
        nr_processes: CPU2:     -900
        nr_processes: CPU3:     -136
        nr_processes: TOTAL:    84
        proc_root_getattr. nlink 12 + nr_processes() 84 = 96
        84
        # echo 0 >/sys/devices/system/cpu/cpu1/online
        # ps haux|wc -l
        nr_processes: CPU0:     85
        nr_processes: CPU2:     -901
        nr_processes: CPU3:     -137
        nr_processes: TOTAL:    -953
        proc_root_getattr. nlink 12 + nr_processes() -953 = -941
        75
        # stat /proc/
        nr_processes: CPU0:     84
        nr_processes: CPU2:     -901
        nr_processes: CPU3:     -137
        nr_processes: TOTAL:    -954
        proc_root_getattr. nlink 12 + nr_processes() -954 = -942
          File: `/proc/'
          Size: 0               Blocks: 0          IO Block: 1024   directory
        Device: 3h/3d   Inode: 1           Links: 4294966354
        Access: (0555/dr-xr-xr-x)  Uid: (    0/    root)   Gid: (    0/    root)
        Access: 2009-11-03 09:06:55.000000000 +0000
        Modify: 2009-11-03 09:06:55.000000000 +0000
        Change: 2009-11-03 09:06:55.000000000 +0000

I'm not 100% convinced that the per_cpu regions remain valid for offline
CPUs, although my testing suggests that they do. If not then I think the
correct solution would be to aggregate the process_count for a given CPU
into a global base value in cpu_down().

This bug appears to pre-date the transition to git and it looks like it
may even have been present in linux-2.6.0-test7-bk3 since it looks like
the code Rusty patched in http://lwn.net/Articles/64773/ was already
wrong.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: gpio-keys - use IRQF_SHARED
  Input: winbond-cir - select LEDS_TRIGGERS
  Input: i8042 - try to get stable CTR value when initializing
  Input: atkbd - add a quirk for OQO 01+ multimedia keys

* 'fixes-s3c-2632-rc5' of git://git.fluff.org/bjdooks/linux:
  ARM: S3C2410: Fix sparse warnings in arch/arm/mach-s3c2410/gpio.c
  ARM: S3C2440: mini2440: Fix spare warnings
  ARM: S3C24XX: Fix warnings in arch/arm/plat-s3c24xx/gpio.c
  ARM: S3C2440: mini2440: Fix missing CONFIG_S3C_DEV_USB_HOST
  ARM: S3C24XX: arch/arm/plat-s3c24xx: Move dereference after NULL test
  ARM: S3C: Fix adc function exports
  ARM: S3C2410: Fix link if CONFIG_S3C2410_IOTIMING is not set
  ARM: S3C24XX: Introduce S3C2442B CPU
  ARM: S3C24XX: Define a macro to avoid compilation error
  ARM: S3C: Add info for supporting circular DMA buffers
  ARM: S3C64XX: Set rate of crystal mux
  ARM: S3C64XX: Fix S3C64XX_CLKDIV0_ARM_MASK value

* 'i2c-fixes' of git://git.fluff.org/bjdooks/linux:
  i2c-mpc: Do not generate STOP after read.
  i2c: imx: disable clock when it's possible to save power.
  i2c: imx: only imx1 needs disable delay
  i2c: imx: check busy bit when START/STOP

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
  nilfs2: add zero-fill for new btree node buffers
  nilfs2: fix irregular checkpoint creation due to data flush
  nilfs2: fix dirty page accounting leak causing hang at write

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (21 commits)
  mac80211: check interface is down before type change
  cfg80211: fix NULL ptr deref
  libertas if_usb: Fix crash on 64-bit machines
  mac80211: fix reason code output endianness
  mac80211: fix addba timer
  ath9k: fix misplaced semicolon on rate control
  b43: Fix DMA TX bounce buffer copying
  mac80211: fix BSS leak
  rt73usb.c : more ids
  ipw2200: fix oops on missing firmware
  gre: Fix dev_addr clobbering for gretap
  sky2: set carrier off in probe
  net: fix sk_forward_alloc corruption
  pcnet_cs: add cis of PreMax PE-200 ethernet pcmcia card
  r8169: Fix card drop incoming VLAN tagged MTU byte large jumbo frames
  ibmtr: possible Read buffer overflow?
  net: Fix RPF to work with policy routing
  net: fix kmemcheck annotations
  e1000e: rework disable K1 at 1000Mbps for 82577/82578
  e1000e: config PHY via software after resets
  ...

pm_runtime_idle() is somewhat noisy. Remove debug prints.
Signed-off-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

Fixes the point where we need to complete the power transition when
device suspend fails, so that we don't print warnings about devices
added to the device hierarchy after a failing suspend.

[rjw: Modified changelog.]
Signed-off-by: Romit Dasgupta <romit@ti.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

Finish a line by \n when load_image fails in the middle of loading.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

There are too many retval variables in save_image(). Thus error return
value from snapshot_read_next() may be ignored and only part of the
snapshot (successfully) written.

Remove 'error' variable, invert the condition in the do-while loop
and convert the loop to use only 'ret' variable.

Switch the rest of the function to consider only 'ret'.

Also make sure we end printed line by \n if an error occurs.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

While cruising through the swsusp code I found few blkdev reference
leaks of resume_bdev.

swsusp_read: remove blkdev_put altogether. Some fail paths do
             not do that.
swsusp_check: make sure we always put a reference on fail paths
software_resume: all fail paths between swsusp_check and swsusp_read
                 omit swsusp_close. Add it in those cases. And since
                 swsusp_read doesn't drop the reference anymore, do
                 it here unconditionally.

[rjw: Fixed a small coding style issue.]
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

Commit 0c570cde
(PM / yenta: Fix cardbus suspend/resume regression) caused resume to
fail on systems with two CardBus bridges.  While the exact nature
of the failure is not known at the moment, it can be worked around by
splitting the yenta resume into an early part, executed during the
early phase of resume, that will only resume the socket and power it
up if there was a card in it during suspend, and a late part,
executed during "regular" resume, that will carry out all of the
remaining yenta resume operations.

Fixes http://bugzilla.kernel.org/show_bug.cgi?id=14334, which is a
listed regression from 2.6.31.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Dominik Brodowski <linux@dominikbrodowski.net>
Reported-by: Stephen J. Gowdy <gowdy@cern.ch>
Tested-by: Jose Marino <braket@hotmail.com>

There is nothing that disallows gpio-keys to share it's IRQ line
w/ other drivers. Make it use IRQF_SHARED in request_irq().

An example of other driver with which I'd like to share IRQ line
for GPIO buttons is ledtrig-gpio.
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

drivers/input/misc/winbond-cir.c depends on LEDS_TRIGGERS so
add an appropriate select to drivers/input/misc/Kconfig
Signed-off-by: David Härdeman <david@hardeman.nu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

If user presses keys while i8042 is being initialized there is a chance
that keyboard data will be mistaken for results of Read Control Register
command causing futher troubles. Work around this issue by reading CTR
several times and stop when we get matching results.
Reported-and-tested-by: Dave Young <hidave.darkstar@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

Adds missing initialization of newly allocated b-tree node buffers.
This avoids garbage data to be mixed in b-tree node blocks.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>