This seems to have popped up after the recent merges:

drivers/watchdog/w83697ug_wdt.c: In function ‘w83697ug_select_wd_register’:
drivers/watchdog/w83697ug_wdt.c:105: warning: ‘return’ with a value, in function returning void
Signed-off-by: Alan Cox <alan@redhat.com>
Acked-by: Wim Van Sebroeck <wim@iguana.be>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

- atomic operations which both modify the variable and return something imply
  full smp memory barriers before and after the memory operations involved
  (failing atomic_cmpxchg, atomic_add_unless, etc don't imply a barrier because
  they don't modify the target). See Documentation/atomic_ops.txt.
  So remove extra barriers and branches.

- All architectures support atomic_cmpxchg. This has no relation to
  __HAVE_ARCH_CMPXCHG. We can just take the atomic_cmpxchg path unconditionally

This reduces a simple single threaded fastpath lock+unlock test from 590 cycles
to 203 cycles on a ppc970 system.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* git://git.kernel.org/pub/scm/linux/kernel/git/czankel/xtensa-2.6:
  xtensa: Add config files for Diamond 232L - Rev B processor variant
  xtensa: Fix io regions
  xtensa: Add support for the Sonic Ethernet device for the XT2000 board.
  xtensa: replace remaining __FUNCTION__ occurrences
  xtensa: use newer __SPIN_LOCK_UNLOCKED macro
  XTENSA: warn about including <asm/rwsem.h> directly.

page_cgroup_init() is called from mem_cgroup_init(). But at this
point, we cannot call alloc_bootmem().
(and this caused panic at boot.)

This patch moves page_cgroup_init() to init/main.c.

Time table is following:
==
  parse_args(). # we can trust mem_cgroup_subsys.disabled bit after this.
  ....
  cgroup_init_early()  # "early" init of cgroup.
  ....
  setup_arch()         # memmap is allocated.
  ...
  page_cgroup_init();
  mem_init();   # we cannot call alloc_bootmem after this.
  ....
  cgroup_init() # mem_cgroup is initialized.
==

Before page_cgroup_init(), mem_map must be initialized. So,
I added page_cgroup_init() to init/main.c directly.

(*) maybe this is not very clean but
    - cgroup_init_early() is too early
    - in cgroup_init(), we have to use vmalloc instead of alloc_bootmem().
    use of vmalloc area in x86-32 is important and we should avoid very large
    vmalloc() in x86-32. So, we want to use alloc_bootmem() and added page_cgroup_init()
    directly to init/main.c

[akpm@linux-foundation.org: remove unneeded/bad mem_cgroup_subsys declaration]
[akpm@linux-foundation.org: fix build]
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Tested-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

The __log_wait_for_space function sits in a loop checkpointing
transactions until there is sufficient space free in the journal.
However, if there are no transactions to be processed (e.g.  because the
free space calculation is wrong due to a corrupted filesystem) it will
never progress.

Check for space being required when no transactions are outstanding and
abort the journal instead of endlessly looping.

This patch fixes the bug reported by Sami Liedes at:
http://bugzilla.kernel.org/show_bug.cgi?id=10976Signed-off-by: Duane Griffin <duaneg@dghda.com>
Tested-by: Sami Liedes <sliedes@cc.hut.fi>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

__try_to_free_cp_buf(), __process_buffer(), and __wait_cp_io() test
BH_Uptodate flag to detect write I/O errors on metadata buffers.  But by
commit 95450f5a "ext3: don't read inode
block if the buffer has a write error"(*), BH_Uptodate flag can be set to
inode buffers with BH_Write_EIO in order to avoid reading old inode data.
So now, we have to test BH_Write_EIO flag of checkpointing inode buffers
instead of BH_Uptodate.  This patch does it.
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Acked-by: Jan Kara <jack@suse.cz>
Acked-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

If the journal has aborted due to a checkpointing failure, we have to
keep the contents of the journal space.  Otherwise, the filesystem will
lose uncheckpointed metadata completely and become inconsistent.  To
avoid this, we need to keep needs_recovery flag if checkpoint has
failed.

With this patch, ext3_put_super() detects a checkpointing failure from
the return value of journal_destroy(), then it invokes ext3_abort() to
make the filesystem read only and keep needs_recovery flag.  Errors
from journal_flush() are also handled by this patch in some places.
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Cc: Jan Kara <jack@ucw.cz>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

When a checkpointing IO fails, current JBD code doesn't check the error
and continue journaling.  This means latest metadata can be lost from both
the journal and filesystem.

This patch leaves the failed metadata blocks in the journal space and
aborts journaling in the case of log_do_checkpoint().  To achieve this, we
need to do:

1. don't remove the failed buffer from the checkpoint list where in
   the case of __try_to_free_cp_buf() because it may be released or
   overwritten by a later transaction
2. log_do_checkpoint() is the last chance, remove the failed buffer
   from the checkpoint list and abort the journal
3. when checkpointing fails, don't update the journal super block to
   prevent the journaled contents from being cleaned.  For safety,
   don't update j_tail and j_tail_sequence either
4. when checkpointing fails, notify this error to the ext3 layer so
   that ext3 don't clear the needs_recovery flag, otherwise the
   journaled contents are ignored and cleaned in the recovery phase
5. if the recovery fails, keep the needs_recovery flag
6. prevent cleanup_journal_tail() from being called between
   __journal_drop_transaction() and journal_abort() (a race issue
   between journal_flush() and __log_wait_for_space()
Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Acked-by: Jan Kara <jack@suse.cz>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

In the case where procfs is disabled, create_proc_profile() does not
exist. Stub it in with the others.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/page_cgroup.c: In function 'init_section_page_cgroup':
mm/page_cgroup.c:111: error: implicit declaration of function 'vmalloc_node'
mm/page_cgroup.c:111: warning: assignment makes pointer from integer without a cast
mm/page_cgroup.c: In function '__free_page_cgroup':
mm/page_cgroup.c:140: error: implicit declaration of function 'vfree'
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (53 commits)
  powerpc: Support for relocatable kdump kernel
  powerpc: Don't use a 16G page if beyond mem= limits
  powerpc: Add del_node() for early boot code to prune inapplicable devices.
  powerpc: Further compile fixup for STRICT_MM_TYPECHECKS
  powerpc: Remove empty #else from signal_64.c
  powerpc: Move memory size print into common show_cpuinfo for 32-bit
  hvc_console: Remove __devexit annotation of hvc_remove()
  hvc_console: Add support for tty window resizing
  hvc_console: Fix loop if put_char() returns 0
  hvc_console: Add tty driver flag TTY_DRIVER_RESET_TERMIOS
  hvc_console: Add a hangup notifier for backends
  powerpc/83xx: Add DS1339 RTC support for MPC8349E-mITX boards .dts
  powerpc/83xx: Add support for MCU microcontroller in .dts files
  powerpc/85xx: Move mpc8572ds.dts to address-cells/size-cells = <2>
  of/spi: Support specifying chip select as active high via device tree
  powerpc: Remove device_type = "board_control" properties in .dts files
  i2c-cpm: Suppress autoprobing for devices
  powerpc/85xx: Fix mpc8536ds dma interrupt numbers
  powerpc/85xx: Enable enhanced functions for 8536 TSEC
  powerpc: Delete unused prom_strtoul and prom_memparse
  ...

* 'for-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/dvrabel/uwb: (47 commits)
  uwb: wrong sizeof argument in mac address compare
  uwb: don't use printk_ratelimit() so often
  uwb: use kcalloc where appropriate
  uwb: use time_after() when purging stale beacons
  uwb: add credits for the original developers of the UWB/WUSB/WLP subsystems
  uwb: add entries in the MAINTAINERS file
  uwb: depend on EXPERIMENTAL
  wusb: wusb-cbaf (CBA driver) sysfs ABI simplification
  uwb: document UWB and WUSB sysfs files
  uwb: add symlinks in sysfs between radio controllers and PALs
  uwb: dont tranmit identification IEs
  uwb: i1480/GUWA100U: fix firmware download issues
  uwb: i1480: remove MAC/PHY information checking function
  uwb: add Intel i1480 HWA to the UWB RC quirk table
  uwb: disable command/event filtering for D-Link DUB-1210
  uwb: initialize the debug sub-system
  uwb: Fix handling IEs with empty IE data in uwb_est_get_size()
  wusb: fix bmRequestType for Abort RPipe request
  wusb: fix error path for wusb_set_dev_addr()
  wusb: add HWA host controller driver
  ...

* 'for-next' of git://git.o-hand.com/linux-mfd:
  mfd: check for platform_get_irq() return value in sm501
  mfd: use pci_ioremap_bar() in sm501
  mfd: Don't store volatile bits in WM8350 register cache
  mfd: don't export wm3850 static functions
  mfd: twl4030-gpio driver
  mfd: rtc-twl4030 driver
  mfd: twl4030 IRQ handling update

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  IB/ehca: Reject dynamic memory add/remove when ehca adapter is present
  IB/ehca: Fix reported max number of QPs and CQs in systems with >1 adapter
  IPoIB: Set netdev offload features properly for child (VLAN) interfaces
  IPoIB: Clean up ethtool support
  mlx4_core: Add Ethernet PCI device IDs
  mlx4_en: Add driver for Mellanox ConnectX 10GbE NIC
  mlx4_core: Multiple port type support
  mlx4_core: Ethernet MAC/VLAN management
  mlx4_core: Get ethernet MTU and default address from firmware
  mlx4_core: Support multiple pre-reserved QP regions
  Update NetEffect maintainer emails to Intel emails
  RDMA/cxgb3: Remove cmid reference on tid allocation failures
  IB/mad: Use krealloc() to resize snoop table
  IPoIB: Always initialize poll_timer to avoid crash on unload
  IB/ehca: Don't allow creating UC QP with SRQ
  mlx4_core: Add QP range reservation support
  RDMA/ucma: Test ucma_alloc_multicast() return against NULL, not with IS_ERR()

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  sata_via: load DEVICE register when CTL changes
  libata: set device class to NONE if phys_offline
  libata-eh: fix slave link EH action mask handling
  libata: transfer EHI control flags to slave ehc.i
  libata-sff: fix ata_sff_post_internal_cmd()
  libata: initialize port_task when !CONFIG_ATA_SFF

* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] clps711x: add sparsemem definitions
  [ARM] 5315/1: Fix section mismatch warning (sa1111)
  [ARM] Orion: activate workaround for 88f6183 SPI clock erratum
  [ARM] Orion: instantiate the dsa switch driver
  [ARM] mv78xx0: force link speed/duplex on eth2/eth3
  [ARM] remove extra brace in arch/arm/mach-pxa/trizeps4.c
  [ARM] balance parenthesis in header file
  [ARM] pxa: fix trizeps PCMCIA build
  [ARM] pxa: fix trizeps defconfig
  [ARM] dmabounce requires ZONE_DMA
  [ARM] 5303/1: period_cycles should be greater than 1
  [ARM] 5310/1: Fix cache flush functions for ARMv4
  [ARM] pxa: fix 3bca103a
  [ARM] pxa: fix redefinition of NR_IRQS
  [ARM] S3C24XX: Fix redefine of DEFINE_TIMER() in s3c24xx pwm-clock.c
  [ARM] S3C2443: Fix HCLK rate
  [ARM] S3C24XX: Serial driver debug depends on DEBUG_LL
  [ARM] S3C24XX: pwm-clock set_parent mask fix

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6: (41 commits)
  [IA64] Fix annoying IA64_TR_ALLOC_MAX message.
  [IA64] kill sys32_pipe
  [IA64] remove sys32_pause
  [IA64] Add Variable Page Size and IA64 Support in Intel IOMMU
  ia64/pv_ops: paravirtualized instruction checker.
  ia64/xen: a recipe for using xen/ia64 with pv_ops.
  ia64/pv_ops: update Kconfig for paravirtualized guest and xen.
  ia64/xen: preliminary support for save/restore.
  ia64/xen: define xen machine vector for domU.
  ia64/pv_ops/xen: implement xen pv_time_ops.
  ia64/pv_ops/xen: implement xen pv_irq_ops.
  ia64/pv_ops/xen: define the nubmer of irqs which xen needs.
  ia64/pv_ops/xen: implement xen pv_iosapic_ops.
  ia64/pv_ops/xen: paravirtualize entry.S for ia64/xen.
  ia64/pv_ops/xen: paravirtualize ivt.S for xen.
  ia64/pv_ops/xen: paravirtualize DO_SAVE_MIN for xen.
  ia64/pv_ops/xen: define xen paravirtualized instructions for hand written assembly code
  ia64/pv_ops/xen: define xen pv_cpu_ops.
  ia64/pv_ops/xen: define xen pv_init_ops for various xen initialization.
  ia64/pv_ops/xen: elf note based xen startup.
  ...

VIA controllers clear DEVICE register when IEN changes.  Make sure
DEVICE is updated along with CTL.

This change is separated from Joseph Chan's larger patch.

  http://thread.gmane.org/gmane.linux.kernel.commits.mm/40640Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Joseph Chan <JosephChan@via.com.tw>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Reset methods don't have access to phys link status for slave links
and may incorrectly indicate device presence causing unnecessary probe
failures for unoccupied links.  This patch clears device class to NONE
during post-reset processing if phys link is offline.

As on/offlineness semantics is strictly defined and used in multiple
places by the core layer, this won't change behavior for drivers which
don't use slave links.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Slave link action mask is transferred to master link and all the EH
actions are taken by the master link.  ata_eh_about_to_do() and
ata_eh_done() are called with ATA_EH_ALL_ACTIONS to clear the slave
link actions during transfer.  This always sets ATA_PFLAG_RECOVERED
flag causing spurious "EH complete" messages.

Don't set ATA_PFLAG_RECOVERED for slave link actions.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ATA_EHI_NO_AUTOPSY and ATA_EHI_QUIET are used to control the behavior
of EH.  As only the master link is visible outside EH, these flags are
set only for the master link although they should also apply to the
slave link, which causes spurious EH messages during probe and
suspend/resume.

This patch transfers those two flags to slave ehc.i before performing
slave autopsy and reporting.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ata_sff_post_internal_cmd() needs to grab port lock before calling
ata_bmdma_stop() and also need to clear hsm_task_state.  Fix it.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ap->port_task was not initialized if !CONFIG_ATA_SFF later triggering
lockdep warning.  Make sure it's initialized.

Reported by Larry Finger.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Since the ehca device driver does not support dynamic memory add and
remove operations, the driver must explicitly reject such requests in
order to prevent unpredictable behaviors related to existing memory
regions that cover all of memory being used by InfiniBand protocols in
the kernel.

The solution (for now at least) is to add a memory notifier to the
ehca device driver and if a request for dynamic memory add or remove
comes in, ehca will always reject it.  The user can add or remove
memory by hot-removing the ehca adapter, performing the memory
operation, and then hot-adding the ehca adapter back.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Because ehca adapters can differ in the maximum number of QPs and CQs
we have to save the maximum number of these ressources per adapter and
not globally per ehca driver. This fix introduces 2 new members to the
shca structure to store the maximum value for QPs and CQs per adapter.

The module parameters are now used as initial values for those
variables.  If a user selects an invalid number of CQs or QPs we don't
print an error any longer, instead we will inform the user with a
warning and set the values to the respective maximum supported by the
HW.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Child devices were created without any offload features set, fix this by
moving the code that computes the features into generic function which is
now called through non-child and child device creation.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>

-- v1 has a bug where the 'result' flag in ipoib_vlan_add may be used uninitialized
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Add a get_rx_csum method.  Remove the driver's own get_tso method, as
the ethtool kernel code uses the default one if nothing is provided.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

The Mellanox ConnectX can operate as an InfiniBand adapter, as an
Ethernet NIC, or as a Fibre Channel (FC) HBA.  The kernel has a
low-level driver, mlx4_core, which handles multiplexing access to the
device, and there is also already an InfiniBad driver, mlx4_ib.

This patch adds a new driver, mlx4_en, which implements a standard
Ethernet NIC driver.
Signed-off-by: Liran Liss <liranl@mellanox.co.il>
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Multi-protocol adapters support different port types.  Each consumer
of mlx4_core queries for supported port types; in particular mlx4_ib
can no longer assume that all physical ports belong to it.  Port type
is configured through a sysfs interface.  When the type of a port is
changed, all mlx4 interfaces are unregistered, and then registered
again with the new port types.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Add support for managing MAC and VLAN filters for each port.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Oren Duer <oren@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Get maximum ethernet MTU and default MAC address from the firmware
QUERY_DEV_CAP command.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

For ethernet support, we need to reserve QPs for the ethernet and
fibre channel driver.  The QPs are reserved at the end of the QP
table.  (This way we assure that they are aligned to their size)

We need to consider these reserved ranges in bitmap creation, so we
extend the mlx4 bitmap utility functions to allow reserved ranges at
both the bottom and the top of the range.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

This adds relocatable kernel support for kdump. With this one can
use the same regular kernel to capture the kdump. A signature (0xfeed1234)
is passed in r6 from panic code to the next kernel through kexec_sequence
and purgatory code. The signature is used to differentiate between
kdump kernel and non-kdump kernels.

The purgatory code compares the signature and sets the __kdump_flag in
head_64.S.  During the boot up, kernel code checks __kdump_flag and if it
is set, the kernel will behave as relocatable kdump kernel. This kernel
will boot at the address where it was loaded by kexec-tools ie. at the
address reserved through crashkernel boot parameter.

CONFIG_CRASH_DUMP depends on CONFIG_RELOCATABLE option to build kdump
kernel as relocatable. So the same kernel can be used as production and
kdump kernel.

This patch incorporates the changes suggested by Paul Mackerras to avoid
GOT use and to avoid two copies of the code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Mohan Kumar M <mohan@in.ibm.com>
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

If mem= is used on the boot command line to limit memory then the memory block where a 16G page resides may not be available.

Thanks to Michael Ellerman for finding the problem.
Signed-off-by: Jon Tollefson <kniht@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Some platforms have variants that can share most of a flat device tree but need
a few devices selectively pruned at boot time.  This adds del_node() to ops.h
to allow access to the existing fdt_del_node().
Signed-off-by: Mike Ditto <mditto@consentry.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

A patch of mine was recently committed to fix up STRICT_MM_TYPECHECKS
behaviour on powerpc (f5ea64dc).
However, something which breaks it again seems to have slipped in
afterwards.  So, here's another small fix.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Remove empty/bogus #else from signal_64.c
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Most of the platforms were printing the size of the memory
in their show_cpuinfo implementations. This moves that to
the common show_cpuinfo, so that all 32-bit platforms will
now print the size of memory.  I also update the code
to deal with the fact that total_memory is now a phys_addr_t.
Signed-off-by: Becky Bruce <becky.bruce@freescale.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>