linux-2.6-microblaze.git
12 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 20 Aug 2012 17:36:18 +0000 (10:36 -0700)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Ingo Molnar.

A x32 socket ABI fix with a -stable backport tag among other fixes.

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x32: Use compat shims for {g,s}etsockopt
  Revert "x86-64/efi: Use EFI to deal with platform wall clock"
  x86, apic: fix broken legacy interrupts in the logical apic mode
  x86, build: Globally set -fno-pic
  x86, avx: don't use avx instructions with "noxsave" boot param

12 years agoMerge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 20 Aug 2012 17:35:05 +0000 (10:35 -0700)]
Merge branch 'sched-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull scheduler fixes from Ingo Molnar.

* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Fix migration thread runtime bogosity
  sched,rt: fix isolated CPUs leaving root_task_group indefinitely throttled
  sched,cgroup: Fix up task_groups list
  sched: fix divide by zero at {thread_group,task}_times
  sched, cgroup: Reduce rq->lock hold times for large cgroup hierarchies

12 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 20 Aug 2012 17:34:21 +0000 (10:34 -0700)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 perf fixes from Ingo Molnar.

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86: disable PEBS on a guest entry.
  perf/x86: Add Intel Westmere-EX uncore support
  perf/x86: Fixes for Nehalem-EX uncore driver
  perf, x86: Fix uncore_types_exit section mismatch

12 years agoMerge branch 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 20 Aug 2012 17:30:57 +0000 (10:30 -0700)]
Merge branch 'core-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull a mutex fix from Ingo Molnar.

Fix the fastpath_lock failure contention flag for xchg-based mutexes.

* 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  mutex: Place lock in contended state after fastpath_lock failure

12 years agovfs: missed source of ->f_pos races
Al Viro [Mon, 20 Aug 2012 14:28:00 +0000 (15:28 +0100)]
vfs: missed source of ->f_pos races

compat_sys_{read,write}v() need the same "pass a copy of file->f_pos" thing
as sys_{read,write}{,v}().

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge branch 'alpha' (alpha architecture patches)
Linus Torvalds [Sun, 19 Aug 2012 15:41:29 +0000 (08:41 -0700)]
Merge branch 'alpha' (alpha architecture patches)

Merge alpha architecture update from Michael Cree:
 "The Alpha Maintainer, Matt Turner, is currently unavailable, so I have
  collected up patches that have been posted to the linux-alpha mailing
  list over the last couple of months, and are forwarding them to you in
  the hope that you are prepared to accept them via me.

  The patches by Al Viro and myself I have been running against kernels
  for two months now so have had quite a bit of testing.  All except one
  patch were intended for the 3.5 kernel but because of Matt's
  unavailability never got forwarded to you."

* emailed patches from Michael Cree <mcree@orcon.net.nz>: (9 commits)
  alpha: Fix fall-out from disintegrating asm/system.h
  Redefine ATOMIC_INIT and ATOMIC64_INIT to drop the casts
  alpha: fix fpu.h usage in userspace
  alpha/mm/fault.c: Port OOM changes to do_page_fault
  alpha: take kernel_execve() out of entry.S
  alpha: take a bunch of syscalls into osf_sys.c
  alpha: Use new generic strncpy_from_user() and strnlen_user()
  alpha: Wire up cross memory attach syscalls
  alpha: Don't export SOCK_NONBLOCK to user space.

12 years agoalpha: Fix fall-out from disintegrating asm/system.h
Michael Cree [Sun, 19 Aug 2012 02:41:04 +0000 (14:41 +1200)]
alpha: Fix fall-out from disintegrating asm/system.h

Commit ec2212088c42 ("Disintegrate asm/system.h for Alpha") removed
asm/system.h however arch/alpha/oprofile/common.c requires definitions
that were shifted from asm/system.h to asm/special_insns.h.  Include
that.

Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoRedefine ATOMIC_INIT and ATOMIC64_INIT to drop the casts
Mel Gorman [Sun, 19 Aug 2012 02:41:03 +0000 (14:41 +1200)]
Redefine ATOMIC_INIT and ATOMIC64_INIT to drop the casts

The following build error occurred during an alpha build:

  net/core/sock.c:274:36: error: initializer element is not constant

Dave Anglin says:
> Here is the line in sock.i:
>
> struct static_key memalloc_socks = ((struct static_key) { .enabled =
> ((atomic_t) { (0) }) });

The above line contains two compound literals.  It also uses a designated
initializer to initialize the field enabled.  A compound literal is not a
constant expression.

The location of the above statement isn't fully clear, but if a compound
literal occurs outside the body of a function, the initializer list must
consist of constant expressions.

Cc: <stable@vger.kernel.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha: fix fpu.h usage in userspace
Mike Frysinger [Sun, 19 Aug 2012 02:41:02 +0000 (14:41 +1200)]
alpha: fix fpu.h usage in userspace

After commit ec2212088c42 ("Disintegrate asm/system.h for Alpha"), the
fpu.h header which we install for userland started depending on
special_insns.h which is not installed.

However, fpu.h only uses that for __KERNEL__ code, so protect the
inclusion the same way to avoid build breakage in glibc:

  /usr/include/asm/fpu.h:4:31: fatal error: asm/special_insns.h: No such file or directory

Cc: stable@vger.kernel.org
Reported-by: Matt Turner <mattst88@gentoo.org>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha/mm/fault.c: Port OOM changes to do_page_fault
Kautuk Consul [Sun, 19 Aug 2012 02:41:01 +0000 (14:41 +1200)]
alpha/mm/fault.c: Port OOM changes to do_page_fault

Commit d065bd810b6d ("mm: retry page fault when blocking on disk
transfer") and 37b23e0525d3 ("x86,mm: make pagefault killable")
introduced changes into the x86 pagefault handler for making the page
fault handler retryable as well as killable.

These changes reduce the mmap_sem hold time, which is crucial during OOM
killer invocation.

Port these changes to ALPHA.

Signed-off-by: Mohd. Faris <mohdfarisq2010@gmail.com>
Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha: take kernel_execve() out of entry.S
Al Viro [Sun, 19 Aug 2012 02:41:00 +0000 (14:41 +1200)]
alpha: take kernel_execve() out of entry.S

Signed-off-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha: take a bunch of syscalls into osf_sys.c
Al Viro [Sun, 19 Aug 2012 02:40:59 +0000 (14:40 +1200)]
alpha: take a bunch of syscalls into osf_sys.c

New helper: current_thread_info().  Allows to do a bunch of odd syscalls
in C. While we are at it, there had never been a reason to do
osf_getpriority() in assembler.  We also get "namespace"-aware (read:
consistent with getuid(2), etc.) behaviour from getx?id() syscalls now.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha: Use new generic strncpy_from_user() and strnlen_user()
Michael Cree [Sun, 19 Aug 2012 02:40:58 +0000 (14:40 +1200)]
alpha: Use new generic strncpy_from_user() and strnlen_user()

Similar to x86/sparc/powerpc implementations except:
1) we implement an extremely efficient has_zero()/find_zero()
   sequence with both prep_zero_mask() and create_zero_mask()
   no-operations.
2) Our output from prep_zero_mask() differs in that only the
   lowest eight bits are used to represent the zero bytes
   nevertheless it can be safely ORed with other similar masks
   from prep_zero_mask() and forms input to create_zero_mask(),
   the two fundamental properties prep_zero_mask() must satisfy.

Tests on EV67 and EV68 CPUs revealed that the generic code is
essentially as fast (to within 0.5% of CPU cycles) of the old
Alpha specific code for large quadword-aligned strings, despite
the 30% extra CPU instructions executed.  In contrast, the
generic code for unaligned strings is substantially slower (by
more than a factor of 3) than the old Alpha specific code.

Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha: Wire up cross memory attach syscalls
Michael Cree [Sun, 19 Aug 2012 02:40:57 +0000 (14:40 +1200)]
alpha: Wire up cross memory attach syscalls

Add sys_process_vm_readv and sys_process_vm_writev to Alpha.

Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoalpha: Don't export SOCK_NONBLOCK to user space.
Michael Cree [Sun, 19 Aug 2012 02:40:56 +0000 (14:40 +1200)]
alpha: Don't export SOCK_NONBLOCK to user space.

Currently we export SOCK_NONBLOCK to user space but that conflicts with
the definition from glibc leading to compilation errors in user programs
(e.g.  see Debian bug #658460).

The generic socket.h restricts the definition of SOCK_NONBLOCK to the
kernel, as does the MIPS specific socket.h, so let's do the same on
Alpha.

Signed-off-by: Michael Cree <mcree@orcon.net.nz>
Acked-by: Matt Turner <mattst88@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge branch 'fixes' of git://git.linaro.org/people/rmk/linux-arm
Linus Torvalds [Sat, 18 Aug 2012 23:20:05 +0000 (16:20 -0700)]
Merge branch 'fixes' of git://git.linaro.org/people/rmk/linux-arm

Pull ARM fixes from Russell King:
 "The largest thing in this set of changes is bringing back some of the
  ARMv3 code to fix a compile problem noticed on RiscPC, which we still
  support, even though we only support ARMv4 there.

  (The reason is that the system bus doesn't support ARMv4 half-word
  accesses, so we need the ARMv3 library code for this platform.)

  The rest are all quite minor fixes."

* 'fixes' of git://git.linaro.org/people/rmk/linux-arm:
  ARM: 7490/1: Drop duplicate select for GENERIC_IRQ_PROBE
  ARM: Bring back ARMv3 IO and user access code
  ARM: 7489/1: errata: fix workaround for erratum #720789 on UP systems
  ARM: 7488/1: mm: use 5 bits for swapfile type encoding
  ARM: 7487/1: mm: avoid setting nG bit for user mappings that aren't present
  ARM: 7486/1: sched_clock: update epoch_cyc on resume
  ARM: 7484/1: Don't enable GENERIC_LOCKBREAK with ticket spinlocks
  ARM: 7483/1: vfp: only advertise VFPv4 in hwcaps if CONFIG_VFPv3 is enabled
  ARM: 7482/1: topology: fix section mismatch warning for init_cpu_topology

12 years agoMerge tag 'pm-for-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Sat, 18 Aug 2012 21:39:19 +0000 (14:39 -0700)]
Merge tag 'pm-for-3.6-rc3' of git://git./linux/kernel/git/rafael/linux-pm

Pull power management fixes from Rafael J. Wysocki:
  - Fixes for three obscure problems in the runtime PM core code found
   recently.
 - Two fixes for the new "coupled" cpuidle code from Colin Cross and Jon
   Medhurst.
 - intel_idle driver fix from Konrad Rzeszutek Wilk.

* tag 'pm-for-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  intel_idle: Check cpu_idle_get_driver() for NULL before dereferencing it.
  cpuidle: Prevent null pointer dereference in cpuidle_coupled_cpu_notify
  cpuidle: coupled: fix sleeping while atomic in cpu notifier
  PM / Runtime: Check device PM QoS setting before "no callbacks" check
  PM / Runtime: Clear power.deferred_resume on success in rpm_suspend()
  PM / Runtime: Fix rpm_resume() return value for power.no_callbacks set

12 years agox32: Use compat shims for {g,s}etsockopt
Mike Frysinger [Sat, 18 Aug 2012 20:11:37 +0000 (16:11 -0400)]
x32: Use compat shims for {g,s}etsockopt

Some of the arguments to {g,s}etsockopt are passed in userland pointers.
If we try to use the 64bit entry point, we end up sometimes failing.

For example, dhcpcd doesn't run in x32:
# dhcpcd eth0
dhcpcd[1979]: version 5.5.6 starting
dhcpcd[1979]: eth0: broadcasting for a lease
dhcpcd[1979]: eth0: open_socket: Invalid argument
dhcpcd[1979]: eth0: send_raw_packet: Bad file descriptor

The code in particular is getting back EINVAL when doing:
struct sock_fprog pf;
setsockopt(s, SOL_SOCKET, SO_ATTACH_FILTER, &pf, sizeof(pf));

Diving into the kernel code, we can see:
include/linux/filter.h:
struct sock_fprog {
unsigned short len;
struct sock_filter __user *filter;
};

net/core/sock.c:
case SO_ATTACH_FILTER:
ret = -EINVAL;
if (optlen == sizeof(struct sock_fprog)) {
struct sock_fprog fprog;

ret = -EFAULT;
if (copy_from_user(&fprog, optval, sizeof(fprog)))
break;

ret = sk_attach_filter(&fprog, sk);
}
break;

arch/x86/syscalls/syscall_64.tbl:
54 common setsockopt sys_setsockopt
55 common getsockopt sys_getsockopt

So for x64, sizeof(sock_fprog) is 16 bytes.  For x86/x32, it's 8 bytes.
This comes down to the pointer being 32bit for x32, which means we need
to do structure size translation.  But since x32 comes in directly to
sys_setsockopt, it doesn't get translated like x86.

After changing the syscall table and rebuilding glibc with the new kernel
headers, dhcp runs fine in an x32 userland.

Oddly, it seems like Linus noted the same thing during the initial port,
but I guess that was missed/lost along the way:
https://lkml.org/lkml/2011/8/26/452

[ hpa: tagging for -stable since this is an ABI fix. ]

Bugzilla: https://bugs.gentoo.org/423649
Reported-by: Mads <mads@ab3.no>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Link: http://lkml.kernel.org/r/1345320697-15713-1-git-send-email-vapier@gentoo.org
Cc: H. J. Lu <hjl.tools@gmail.com>
Cc: <stable@vger.kernel.org> v3.4..v3.5
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
12 years agoMerge branch 'vfs-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs
Linus Torvalds [Sat, 18 Aug 2012 17:02:17 +0000 (10:02 -0700)]
Merge branch 'vfs-fixes' of git://git./linux/kernel/git/mszeredi/vfs

Pull vfs fixes from Miklos Szeredi.

This mainly fixes some confusion about whether the open 'mode' variable
passed around should contain the full file type (S_IFREG etc)
information or just the permission mode.  In particular, the lack of
proper file type information had confused fuse.

* 'vfs-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs:
  vfs: fix propagation of atomic_open create error on negative dentry
  fuse: check create mode in atomic open
  vfs: pass right create mode to may_o_create()
  vfs: atomic_open(): fix create mode usage
  vfs: canonicalize create mode in build_open_flags()

12 years agoMerge tag 'md-3.6-fixes' of git://neil.brown.name/md
Linus Torvalds [Sat, 18 Aug 2012 00:47:32 +0000 (17:47 -0700)]
Merge tag 'md-3.6-fixes' of git://neil.brown.name/md

Pull md fixes from NeilBrown:
 "2 fixes for md, tagged for -stable"

* tag 'md-3.6-fixes' of git://neil.brown.name/md:
  md/raid10: fix problem with on-stack allocation of r10bio structure.
  md: Don't truncate size at 4TB for RAID0 and Linear

12 years agomd/raid10: fix problem with on-stack allocation of r10bio structure.
NeilBrown [Fri, 17 Aug 2012 23:51:42 +0000 (09:51 +1000)]
md/raid10: fix problem with on-stack allocation of r10bio structure.

A 'struct r10bio' has an array of per-copy information at the end.
This array is declared with size [0] and r10bio_pool_alloc allocates
enough extra space to store the per-copy information depending on the
number of copies needed.

So declaring a 'struct r10bio on the stack isn't going to work.  It
won't allocate enough space, and memory corruption will ensue.

So in the two places where this is done, declare a sufficiently large
structure and use that instead.

The two call-sites of this bug were introduced in 3.4 and 3.5
so this is suitable for both those kernels.  The patch will have to
be modified for 3.4 as it only has one bug.

Cc: stable@vger.kernel.org
Reported-by: Ivan Vasilyev <ivan.vasilyev@gmail.com>
Tested-by: Ivan Vasilyev <ivan.vasilyev@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
12 years agoMerge tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland...
Linus Torvalds [Fri, 17 Aug 2012 18:45:58 +0000 (11:45 -0700)]
Merge tag 'rdma-for-linus' of git://git./linux/kernel/git/roland/infiniband

Pull infiniband/rdma fixes from Roland Dreier:
 "Grab bag of InfiniBand/RDMA fixes:
   - IPoIB fixes for regressions introduced by path database conversion
   - mlx4 fixes for bugs with large memory systems and regressions from
     SR-IOV patches
   - RDMA CM fix for passing bad event up to userspace
   - Other minor fixes"

* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  IB/mlx4: Check iboe netdev pointer before dereferencing it
  mlx4_core: Clean up buddy bitmap allocation
  mlx4_core: Fix integer overflow issues around MTT table
  mlx4_core: Allow large mlx4_buddy bitmaps
  IB/srp: Fix a race condition
  IB/qib: Fix error return code in qib_init_7322_variables()
  IB: Fix typos in infiniband drivers
  IB/ipoib: Fix RCU pointer dereference of wrong object
  IB/ipoib: Add missing locking when CM object is deleted
  RDMA/ucma.c: Fix for events with wrong context on iWARP
  RDMA/ocrdma: Don't call vlan_dev_real_dev() for non-VLAN netdevs
  IB/mlx4: Fix possible deadlock on sm_lock spinlock

12 years agoMerge tag 'tty-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
Linus Torvalds [Fri, 17 Aug 2012 18:12:28 +0000 (11:12 -0700)]
Merge tag 'tty-3.6-rc3' of git://git./linux/kernel/git/gregkh/tty

Pull TTY fixes from Greg Kroah-Hartman:
 "Here are 4 tiny patches, each fixing a serial driver problem that
  people have reported.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>"
* tag 'tty-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  pmac_zilog,kdb: Fix console poll hook to return instead of loop
  serial: mxs-auart: fix the wrong RTS hardware flow control
  serial: ifx6x60: fix paging fault on spi_register_driver
  serial: Change Kconfig entry for CLPS711X-target

12 years agointel_idle: Check cpu_idle_get_driver() for NULL before dereferencing it.
Konrad Rzeszutek Wilk [Thu, 16 Aug 2012 20:06:55 +0000 (22:06 +0200)]
intel_idle: Check cpu_idle_get_driver() for NULL before dereferencing it.

If the machine is booted without any cpu_idle driver set
(b/c disable_cpuidle() has been called) we should follow
other users of cpu_idle API and check the return value
for NULL before using it.

Reported-and-tested-by: Mark van Dijk <mark@internecto.net>
Suggested-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
12 years agocpuidle: Prevent null pointer dereference in cpuidle_coupled_cpu_notify
Jon Medhurst (Tixy) [Wed, 15 Aug 2012 20:11:00 +0000 (22:11 +0200)]
cpuidle: Prevent null pointer dereference in cpuidle_coupled_cpu_notify

When a kernel is built to support multiple hardware types it's possible
that CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED is set but the hardware the
kernel is run on doesn't support cpuidle and therefore doesn't load a
driver for it. In this case, when the system is shut down,
cpuidle_coupled_cpu_notify() gets called with cpuidle_devices set to
NULL. There are quite possibly other circumstances where this
situation can also occur and we should check for it.

Signed-off-by: Jon Medhurst <tixy@linaro.org>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
12 years agocpuidle: coupled: fix sleeping while atomic in cpu notifier
Colin Cross [Wed, 15 Aug 2012 20:10:50 +0000 (22:10 +0200)]
cpuidle: coupled: fix sleeping while atomic in cpu notifier

The cpu hotplug notifier gets called in both atomic and non-atomic
contexts, it is not always safe to lock a mutex.  Filter out all events
except the six necessary ones, which are all sleepable, before taking
the mutex.

Signed-off-by: Colin Cross <ccross@android.com>
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
12 years agoPM / Runtime: Check device PM QoS setting before "no callbacks" check
Rafael J. Wysocki [Wed, 15 Aug 2012 19:32:04 +0000 (21:32 +0200)]
PM / Runtime: Check device PM QoS setting before "no callbacks" check

If __dev_pm_qos_read_value(dev) returns a negative value,
rpm_suspend() should return -EPERM for dev even if its
power.no_callbacks flag is set.  For this to happen, the device's
power.no_callbacks flag has to be checked after the PM QoS check,
so move the PM QoS check to rpm_check_suspend_allowed() (this will
make it cover idle notifications as well as runtime suspend too).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable@vger.kernel.org
12 years agoPM / Runtime: Clear power.deferred_resume on success in rpm_suspend()
Rafael J. Wysocki [Wed, 15 Aug 2012 19:31:55 +0000 (21:31 +0200)]
PM / Runtime: Clear power.deferred_resume on success in rpm_suspend()

The power.deferred_resume can only be set if the runtime PM status
of device is RPM_SUSPENDING and it should be cleared after its
status has been changed, regardless of whether or not the runtime
suspend has been successful.  However, it only is cleared on
suspend failure, while it may remain set on successful suspend and
is happily leaked to rpm_resume() executed in that case.

That shouldn't happen, so if power.deferred_resume is set in
rpm_suspend() after the status has been changed to RPM_SUSPENDED,
clear it before calling rpm_resume().  Then, it doesn't need to be
cleared before changing the status to RPM_SUSPENDING any more,
because it's always cleared after the status has been changed to
either RPM_SUSPENDED (on success) or RPM_ACTIVE (on failure).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable@vger.kernel.org
12 years agoPM / Runtime: Fix rpm_resume() return value for power.no_callbacks set
Rafael J. Wysocki [Wed, 15 Aug 2012 19:31:45 +0000 (21:31 +0200)]
PM / Runtime: Fix rpm_resume() return value for power.no_callbacks set

For devices whose power.no_callbacks flag is set, rpm_resume()
should return 1 if the device's parent is already active, so that
the callers of pm_runtime_get() don't think that they have to wait
for the device to resume (asynchronously) in that case (the core
won't queue up an asynchronous resume in that case, so there's
nothing to wait for anyway).

Modify the code accordingly (and make sure that an idle notification
will be queued up on success, even if 1 is to be returned).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable@vger.kernel.org
12 years agoMerge tag 'staging-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Fri, 17 Aug 2012 17:17:03 +0000 (10:17 -0700)]
Merge tag 'staging-3.6-rc3' of git://git./linux/kernel/git/gregkh/staging

Pull staging fixes from Greg Kroah-Hartman:
 "Here are some staging driver fixes (and iio driver fixes, they get
  lumped in with the staging stuff due to dependancies) for your 3.6-rc3
  tree.

  Nothing major, just a bunch of fixes that people have reported.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>"
* tag 'staging-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging: (26 commits)
  iio: lm3533-als: Fix build warnings
  staging:iio:ad7780: Mark channels as unsigned
  staging:iio:ad7192: Report offset and scale for temperature channel
  staging:iio:ad7192: Report channel offset
  staging:iio:ad7192: Mark channels as unsigned
  staging:iio:ad7192: Fix setting ACX
  staging:iio:ad7192: Add missing break in switch statement
  staging:iio:ad7793: Fix internal reference value
  staging:iio:ad7793: Follow new IIO naming spec
  staging:iio:ad7793: Fix temperature scale and offset
  staging:iio:ad7793: Report channel offset
  staging:iio:ad7793: Mark channels as unsigned
  staging:iio:ad7793: Add missing break in switch statement
  iio/adjd_s311: Fix potential memory leak in adjd_s311_update_scan_mode()
  iio: frequency: ADF4350: Fix potential reference div factor overflow.
  iio: staging: ad7298_ring: Fix maybe-uninitialized warning
  staging: comedi: usbduxfast: Declare MODULE_FIRMWARE usage
  staging: comedi: usbdux: Declare MODULE_FIRMWARE usage
  staging: comedi: usbduxsigma: Declare MODULE_FIRMWARE usage
  staging: csr: add INET dependancy
  ...

12 years agoMerge tag 'driver-core-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 17 Aug 2012 17:16:30 +0000 (10:16 -0700)]
Merge tag 'driver-core-3.6-rc3' of git://git./linux/kernel/git/gregkh/driver-core

Pull driver core fixes from Greg Kroah-Hartman:
 "Here are two tiny patches, one fixing a dynamic debug problem that the
  printk rework turned up, and the other one fixing an extcon problem
  that people reported.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>"
* tag 'driver-core-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
  extcon: extcon_gpio: Replace gpio_request_one by devm_gpio_request_one
  drivers-core: make structured logging play nice with dynamic-debug

12 years agoMerge tag 'char-misc-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Fri, 17 Aug 2012 17:15:57 +0000 (10:15 -0700)]
Merge tag 'char-misc-3.6-rc3' of git://git./linux/kernel/git/gregkh/char-misc

Pull Char / Misc driver fixes from Greg Kroah-Hartman:
 "Here are some small misc and w1 driver fixes for 3.6-rc3.  Nothing
  major, just some some bugfixes and a new device id for a w1 driver.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>"
* tag 'char-misc-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
  1-Wire: Add support for the maxim ds1825 temperature sensor
  ti-st: Fix check for pdata->chip_awake function pointer
  mei: add mei_quirk_probe function
  mei: fix device stall after wd is stopped

12 years agoMerge tag 'usb-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Fri, 17 Aug 2012 17:14:53 +0000 (10:14 -0700)]
Merge tag 'usb-3.6-rc3' of git://git./linux/kernel/git/gregkh/usb

Pull USB patches from Greg Kroah-Hartman:
 "Here are a number of small USB patches for 3.6-rc3.

  The "large" one is just a number of device id updates to the option
  driver, done by the manufacturer, properly fixing up the device ids
  based on shipping devices.

  Other than that, some gadget driver fixes, the obligitary XHCI
  patches, and some other device ids and bugs fixed.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>"
* tag 'usb-3.6-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (26 commits)
  USB: qcserial: fix port handling on Gobi 1K and 2K+
  USB: serial: Fix mos7840 timeout
  USB: option: add ZTE K5006-Z
  usb: gadget: u_ether: fix kworker 100% CPU issue with still used interfaces in eth_stop
  usb: host: tegra: fix warning messages in ehci_remove
  usb: host: mips: sead3: Update for EHCI register structure.
  usb: renesas_usbhs: fixup resume method for autonomy mode
  usb: renesas_usbhs: mod_host: add missing .bus_suspend/resume
  update MAINTAINERS for Oliver Neukum
  usb: usb_wwan: resume/suspend can be called after port is gone
  usb: serial: prevent suspend/resume from racing against probe/remove
  usb: usb_wwan: replace release and disconnect with a port_remove hook
  usb: serial: mos7840: Fixup mos7840_chars_in_buffer()
  USB: isp1362-hcd.c: usb message always saved in case of underrun
  OMAP: USB : Fix the EHCI enumeration and core retention issue
  usb: chipidea: fix and improve dependencies if usb host or gadget support is built as module
  USB: support the new interfaces of Huawei Data Card devices in option driver
  USB: ftdi_sio: Add VID/PID for Kondo Serial USB
  xhci: Switch PPT ports to EHCI on shutdown.
  xhci: Fix bug after deq ptr set to link TRB.
  ...

12 years agoscripts/kernel-doc: fix fatal script error
Randy Dunlap [Thu, 16 Aug 2012 23:23:20 +0000 (16:23 -0700)]
scripts/kernel-doc: fix fatal script error

Fix fatal error in scripts/kernel-doc by ignoring the "__weak" attribute:

  Error(drivers/pci/pci.c:2820): cannot understand prototype: 'char * __weak pcibios_setup(char *str) '

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris...
Linus Torvalds [Fri, 17 Aug 2012 16:17:11 +0000 (09:17 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/jmorris/linux-security

Pull a Yama bugfix from James Morris.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
  Yama: access task_struct->comm directly

12 years agoMerge tag 'for-linus' of git://linux-c6x.org/git/projects/linux-c6x-upstreaming
Linus Torvalds [Fri, 17 Aug 2012 15:10:12 +0000 (08:10 -0700)]
Merge tag 'for-linus' of git://linux-c6x.org/git/projects/linux-c6x-upstreaming

Pull C6X atomic64 support from Mark Salter:
 "Enable atomic64 ops in C6X
   - define L1_CACHE_SHIFT
   - select GENERIC_ATOMIC64"

* tag 'for-linus' of git://linux-c6x.org/git/projects/linux-c6x-upstreaming:
  C6X: select GENERIC_ATOMIC64
  C6X: add Lx_CACHE_SHIFT defines

12 years agoMerge tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 17 Aug 2012 15:04:47 +0000 (08:04 -0700)]
Merge tag 'ext4_for_linus_stable' of git://git./linux/kernel/git/tytso/ext4

Pull ext4 bug fixes from Ted Ts'o:
 "The following are all bug fixes and regressions.  The most notable are
  the ones which cause problems for ext4 on RAID --- a performance
  problem when mounting very large filesystems, and a kernel OOPS when
  doing an rm -rf on large directory hierarchies on fast devices."

* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: fix kernel BUG on large-scale rm -rf commands
  ext4: fix long mount times on very big file systems
  ext4: don't call ext4_error while block group is locked
  ext4: avoid kmemcheck complaint from reading uninitialized memory
  ext4: make sure the journal sb is written in ext4_clear_journal_err()

12 years agoautofs4 - fix expire check
Ian Kent [Fri, 17 Aug 2012 03:09:04 +0000 (11:09 +0800)]
autofs4 - fix expire check

In some cases when an autofs indirect mount is contained in a file
system that is marked as shared (such as when systemd does the
equivalent of "mount --make-rshared /" early in the boot), mounts
stop expiring.

When this happens the first expiry check on a mountpoint dentry in
autofs_expire_indirect() sees a mountpoint dentry with a higher
than minimal reference count. Consequently the dentry is condidered
busy and the actual expiry check is never done.

This particular check was originally meant as an optimisation to
detect a path walk in progress but with the addition of rcu-walk
it can be ineffective anyway.

Removing the test allows automounts to expire again since the
actual expire check doesn't rely on the dentry reference count.

Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoext4: fix kernel BUG on large-scale rm -rf commands
Theodore Ts'o [Fri, 17 Aug 2012 12:54:52 +0000 (08:54 -0400)]
ext4: fix kernel BUG on large-scale rm -rf commands

Commit 968dee7722: "ext4: fix hole punch failure when depth is greater
than 0" introduced a regression in v3.5.1/v3.6-rc1 which caused kernel
crashes when users ran run "rm -rf" on large directory hierarchy on
ext4 filesystems on RAID devices:

    BUG: unable to handle kernel NULL pointer dereference at 0000000000000028

    Process rm (pid: 18229, threadinfo ffff8801276bc000, task ffff880123631710)
    Call Trace:
     [<ffffffff81236483>] ? __ext4_handle_dirty_metadata+0x83/0x110
     [<ffffffff812353d3>] ext4_ext_truncate+0x193/0x1d0
     [<ffffffff8120a8cf>] ? ext4_mark_inode_dirty+0x7f/0x1f0
     [<ffffffff81207e05>] ext4_truncate+0xf5/0x100
     [<ffffffff8120cd51>] ext4_evict_inode+0x461/0x490
     [<ffffffff811a1312>] evict+0xa2/0x1a0
     [<ffffffff811a1513>] iput+0x103/0x1f0
     [<ffffffff81196d84>] do_unlinkat+0x154/0x1c0
     [<ffffffff8118cc3a>] ? sys_newfstatat+0x2a/0x40
     [<ffffffff81197b0b>] sys_unlinkat+0x1b/0x50
     [<ffffffff816135e9>] system_call_fastpath+0x16/0x1b
    Code: 8b 4d 20 0f b7 41 02 48 8d 04 40 48 8d 04 81 49 89 45 18 0f b7 49 02 48 83 c1 01 49 89 4d 00 e9 ae f8 ff ff 0f 1f 00 49 8b 45 28 <48> 8b 40 28 49 89 45 20 e9 85 f8 ff ff 0f 1f 80 00 00 00

    RIP  [<ffffffff81233164>] ext4_ext_remove_space+0xa34/0xdf0

This could be reproduced as follows:

The problem in commit 968dee7722 was that caused the variable 'i' to
be left uninitialized if the truncate required more space than was
available in the journal.  This resulted in the function
ext4_ext_truncate_extend_restart() returning -EAGAIN, which caused
ext4_ext_remove_space() to restart the truncate operation after
starting a new jbd2 handle.

Reported-by: Maciej Żenczykowski <maze@google.com>
Reported-by: Marti Raudsepp <marti@juffo.org>
Tested-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
12 years agoext4: fix long mount times on very big file systems
Theodore Ts'o [Thu, 16 Aug 2012 15:59:04 +0000 (11:59 -0400)]
ext4: fix long mount times on very big file systems

Commit 8aeb00ff85a: "ext4: fix overhead calculation used by
ext4_statfs()" introduced a O(n**2) calculation which makes very large
file systems take forever to mount.  Fix this with an optimization for
non-bigalloc file systems.  (For bigalloc file systems the overhead
needs to be set in the the superblock.)

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
12 years agoext4: don't call ext4_error while block group is locked
Theodore Ts'o [Fri, 10 Aug 2012 17:57:52 +0000 (13:57 -0400)]
ext4: don't call ext4_error while block group is locked

While in ext4_validate_block_bitmap(), if an block allocation bitmap
is found to be invalid, we call ext4_error() while the block group is
still locked.  This causes ext4_commit_super() to call a function
which might sleep while in an atomic context.

There's no need to keep the block group locked at this point, so hoist
the ext4_error() call up to ext4_validate_block_bitmap() and release
the block group spinlock before calling ext4_error().

The reported stack trace can be found at:

http://article.gmane.org/gmane.comp.file-systems.ext4/33731

Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
12 years agoYama: access task_struct->comm directly
Kees Cook [Wed, 15 Aug 2012 18:41:55 +0000 (11:41 -0700)]
Yama: access task_struct->comm directly

The core ptrace access checking routine holds a task lock, and when
reporting a failure, Yama takes a separate task lock. To avoid a
potential deadlock with two ptracers taking the opposite locks, do not
use get_task_comm() and just use ->comm directly since accuracy is not
important for the report.

Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Suggested-by: Oleg Nesterov <oleg@redhat.com>
CC: stable@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: James Morris <james.l.morris@oracle.com>
12 years agoLinux 3.6-rc2
Linus Torvalds [Thu, 16 Aug 2012 21:51:24 +0000 (14:51 -0700)]
Linux 3.6-rc2

12 years agoMerge tag 'v3.6-rc1-iio-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23...
Greg Kroah-Hartman [Thu, 16 Aug 2012 21:17:08 +0000 (14:17 -0700)]
Merge tag 'v3.6-rc1-iio-fixes' of git://git./linux/kernel/git/jic23/iio into staging-linus

IIO fixes for v3.6-rc1

These mostly consist of fixes from Lars-Peter Clausen that were
the first part of a large series reworking the drivers concerned.
Turns out these drivers had quite a wealth of minor bugs.

Also here are some build warning fixes for lm3533-als and
adjd_s111 (both new drives in this cycle).
Final elements are a a div factor overflow and a warning
related fix in a couple of Analog Devices drivers.

All in all nothing major, but a worthwhile bunch of short
fixes.

12 years ago1-Wire: Add support for the maxim ds1825 temperature sensor
Raphael Assenat [Thu, 16 Aug 2012 16:56:40 +0000 (12:56 -0400)]
1-Wire: Add support for the maxim ds1825 temperature sensor

This patch adds support for maxim ds1825 based 1-wire temperature sensors.

Signed-off-by: Raphael Assenat <raph@8d.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoti-st: Fix check for pdata->chip_awake function pointer
Matthias Kaehlcke [Thu, 2 Aug 2012 20:17:48 +0000 (22:17 +0200)]
ti-st: Fix check for pdata->chip_awake function pointer

ll_device_want_to_wakeup(): Fix the NULL pointer check on pdata->chip_awake,
which is performed on the wrong function pointer

Signed-off-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoiio: lm3533-als: Fix build warnings
Axel Lin [Thu, 2 Aug 2012 10:10:00 +0000 (11:10 +0100)]
iio: lm3533-als: Fix build warnings

Fix below build warnings:
  CC [M]  drivers/iio/light/lm3533-als.o
drivers/iio/light/lm3533-als.c:667:8: warning: initialization from incompatible pointer type [enabled by default]
drivers/iio/light/lm3533-als.c:667:8: warning: (near initialization for 'dev_attr_in_illuminance0_thresh_either_en.show') [enabled by default]
drivers/iio/light/lm3533-als.c:667:8: warning: initialization from incompatible pointer type [enabled by default]
drivers/iio/light/lm3533-als.c:667:8: warning: (near initialization for 'dev_attr_in_illuminance0_thresh_either_en.store') [enabled by default]

Signed-off-by: Axel Lin <axel.lin@gmail.com>
12 years agostaging:iio:ad7780: Mark channels as unsigned
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7780: Mark channels as unsigned

The values reported by the AD7780 are unsigned with a binary offset:

0x000000 is negative fullscale
0x800000 is zeroscale
0xffffff is positive fullscale

So mark the channel in the channel spec as unsigned rather than signed.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7192: Report offset and scale for temperature channel
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7192: Report offset and scale for temperature channel

The temperature channel reports values in degree Kelvin with sensitivity of 5630
codes per degree. If the chip is configured in bipolar mode there is an
additional binary offset of 0x800000 and the sensitivity is divided by two.

Currently the driver does the mapping from the raw value to degree Celsius when
doing a manual conversion. This has several disadvantages, the major one being
that it does not work for buffered mode, also by doing the division by the
sensitivity in the driver the precession of the reported value is needlessly
reduced.

Furthermore the current calculation only works in bipolar mode and the current
scale is of by a factor of 1000.

This patch modifies the driver to report correct offset and scale values in
both unipolar and bipolar mode and to report the raw temperature value
for manual conversions.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7192: Report channel offset
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7192: Report channel offset

In bipolar mode there is a a binary offset of 2**(N-1) (with N being the number
of bits) on the reported value. Currently this value is subtracted when doing a
manual read. While this works for manual channel readings it does not work for
buffered mode. So report the offset in the channels offset property, which will
work in both modes.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
12 years agostaging:iio:ad7192: Mark channels as unsigned
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7192: Mark channels as unsigned

The values reported by the AD7793 are unsigned.
In uniploar mode:
0x000000 is zeroscale
0xffffff is fullscale
In bipolar mode:
0x000000 is negative fullscale
0x800000 is zeroscale
0xffffff is positive fullscale

In bipolar mode there is a binary offset, but the values are still unsigned.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7192: Fix setting ACX
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7192: Fix setting ACX

Write to the correct register when setting the ACX bit.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7192: Add missing break in switch statement
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7192: Add missing break in switch statement

Without the break statement we fall right through to the default case and return
an error value.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7793: Fix internal reference value
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7793: Fix internal reference value

The internal reference for the ad7793 and similar is 1.17V

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7793: Follow new IIO naming spec
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7793: Follow new IIO naming spec

Make the "in-in_scale_available" attribute follow the new naming spec and
rename it to "in_voltage-voltage_scale_available".

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7793: Fix temperature scale and offset
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7793: Fix temperature scale and offset

The temperature channel uses the internal 1.17V reference with 0.81 mv/C. The
reported temperature is in Kevlin, so we need to add the Kelvin to Celcius
offset when reporting the offset for the temperature channel.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7793: Report channel offset
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7793: Report channel offset

In bipolar mode there is a a binary offset of 2**(N-1) (with N being the number
of bits) on the reported value. Currently this value is subtracted when doing a
manual read. While this works for manual channel readings it does not work for
buffered mode. So report the offset in the channels offset property, which will
work in both modes.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7793: Mark channels as unsigned
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7793: Mark channels as unsigned

The values reported by the AD7793 are unsigned.
In uniploar mode:
0x000000 is zeroscale
0xffffff is fullscale
In bipolar mode:
0x000000 is negative fullscale
0x800000 is zeroscale
0xffffff is positive fullscale

In bipolar mode there is a binary offset, but the values are still unsigned.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agostaging:iio:ad7793: Add missing break in switch statement
Lars-Peter Clausen [Fri, 10 Aug 2012 16:36:00 +0000 (17:36 +0100)]
staging:iio:ad7793: Add missing break in switch statement

Without the break statement we fall right through to the default case and return
an error value.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agoiio/adjd_s311: Fix potential memory leak in adjd_s311_update_scan_mode()
Alexey Khoroshilov [Wed, 8 Aug 2012 09:58:00 +0000 (10:58 +0100)]
iio/adjd_s311: Fix potential memory leak in adjd_s311_update_scan_mode()

Do not leak memory by updating pointer with potentially NULL realloc return value.
There is no need to preserve data in the buffer,
so replace krealloc() by kfree()-kmalloc() pair.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Acked-by: Peter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agoiio: frequency: ADF4350: Fix potential reference div factor overflow.
Michael Hennerich [Fri, 20 Jul 2012 08:31:00 +0000 (09:31 +0100)]
iio: frequency: ADF4350: Fix potential reference div factor overflow.

With small channel spacing values and high reference frequencies it is
possible to exceed the range of the 10-bit counter.
Workaround by checking the range and widening some constrains.

We don't use the REG1_PHASE value in this case the datasheet recommends to set
it to 1 if not used.

Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agoiio: staging: ad7298_ring: Fix maybe-uninitialized warning
Michael Hennerich [Mon, 16 Jul 2012 08:42:00 +0000 (09:42 +0100)]
iio: staging: ad7298_ring: Fix maybe-uninitialized warning

drivers/staging/iio/adc/ad7298_ring.c:97:37: warning: 'time_ns' may
be used uninitialized in this function [-Wmaybe-uninitialized]

Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
12 years agopmac_zilog,kdb: Fix console poll hook to return instead of loop
Jason Wessel [Sun, 12 Aug 2012 12:16:43 +0000 (07:16 -0500)]
pmac_zilog,kdb: Fix console poll hook to return instead of loop

kdb <-> kgdb transitioning does not work properly with this UART
driver because the get character routine loops indefinitely as opposed
to returning NO_POLL_CHAR per the expectation of the KDB I/O driver
API.

The symptom is a kernel hang when trying to switch debug modes.

Cc: Alan Cox <alan@linux.intel.com>
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoserial: mxs-auart: fix the wrong RTS hardware flow control
Huang Shijie [Wed, 8 Aug 2012 02:37:59 +0000 (10:37 +0800)]
serial: mxs-auart: fix the wrong RTS hardware flow control

Without checking if the auart supports the hardware flow control or not,
the old mxs_auart_set_mctrl() asserted the RTS pin blindly.

This will causes the auart receives wrong data in the following case:
   The far-end has already started the write operation, and wait for
the auart asserts the RTS pin. Then the auart starts the read operation,
but mxs_auart_set_mctrl() may be called before we set the RTSCTS in the
mxs_auart_settermios(). So the RTS pin is asserted in a wrong situation,
and we get the wrong data in the end.

This bug has been catched when I connect the mx23(DTE) to the mx53(DCE).

This patch also replaces the AUART_CTRL2_RTS with AUART_CTRL2_RTSEN.
We should use the real the hardware flow control, not the software-controled
hardware flow control.

Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoautofs4 - fix get_next_positive_subdir()
Ian Kent [Mon, 6 Aug 2012 01:37:47 +0000 (09:37 +0800)]
autofs4 - fix get_next_positive_subdir()

Following a report of a crash during an automount expire I found that
the locking in fs/autofs4/expire.c:get_next_positive_subdir() was wrong.
Not only is the locking wrong but the function is more complex than it
needs to be.

The function is meant to calculate (and dget) the next entry in the list
of directories contained in the root of an autofs mount point (an autofs
indirect mount to be precise). The main problem was that the d_lock of
the owner of the list was not being taken when walking the list, which
lead to list corruption under load. The only other lock that needs to
be taken is against the next dentry candidate so it can be checked for
usability.

Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge tag 'vfio-for-v3.6-rc1' of git://github.com/awilliam/linux-vfio
Linus Torvalds [Thu, 16 Aug 2012 18:47:42 +0000 (11:47 -0700)]
Merge tag 'vfio-for-v3.6-rc1' of git://github.com/awilliam/linux-vfio

Pull VFIO fix from Alex Williamson:
 "Just a trivial patch to include vfio.h in the installed headers so we
  can complete userspace integration into QEMU."

* tag 'vfio-for-v3.6-rc1' of git://github.com/awilliam/linux-vfio:
  vfio: Include vfio.h in installed headers

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi...
Linus Torvalds [Thu, 16 Aug 2012 18:46:31 +0000 (11:46 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mszeredi/fuse

Pull fuse updates from Miklos Szeredi.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
  fuse: verify all ioctl retry iov elements
  fuse: add missing INIT flag descriptions
  fuse: add missing INIT flags
  fuse: update attributes on aio_read
  fuse: invalidate inode mapping if mtime changes
  fuse: add FUSE_AUTO_INVAL_DATA init flag

12 years agoMerge tag 'stable/for-linus-3.6-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 16 Aug 2012 18:31:59 +0000 (11:31 -0700)]
Merge tag 'stable/for-linus-3.6-rc1-tag' of git://git./linux/kernel/git/konrad/xen

Pull Xen fix from Konrad Rzeszutek Wilk:
 "Way back in v3.5 we added a mechanism to populate back pages that were
  released (they overlapped with MMIO regions), but neglected to reserve
  the proper amount of virtual space for extend_brk to work properly.

  Coincidentally some other commit aligned the _brk space to larger area
  so I didn't trigger this until it was run on a machine with more than
  2GB of MMIO space."

 * On machines with large MMIO/PCI E820 spaces we fail to boot b/c
   we failed to pre-allocate large enough virtual space for extend_brk.

* tag 'stable/for-linus-3.6-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen/p2m: Reserve 8MB of _brk space for P2M leafs when populating back.

12 years agoMerge tag 'sh-for-linus' of git://github.com/pmundt/linux-sh
Linus Torvalds [Thu, 16 Aug 2012 18:31:29 +0000 (11:31 -0700)]
Merge tag 'sh-for-linus' of git://github.com/pmundt/linux-sh

Pull SuperH fixes from Paul Mundt.

* tag 'sh-for-linus' of git://github.com/pmundt/linux-sh:
  sh: intc: Handle domain association for sparseirq pre-allocated vectors.
  sh: sh7269: Fix LCD pinmux
  sh: dma: fix request_irq usage

12 years agoMAINTAINERS: update address for Dan Williams
Dan Williams [Thu, 16 Aug 2012 02:20:02 +0000 (19:20 -0700)]
MAINTAINERS: update address for Dan Williams

Moved to djbw@fb.com

Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: Dan Williams <djbw@fb.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agostaging: comedi: usbduxfast: Declare MODULE_FIRMWARE usage
Tim Gardner [Thu, 16 Aug 2012 17:09:28 +0000 (11:09 -0600)]
staging: comedi: usbduxfast: Declare MODULE_FIRMWARE usage

Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Mori Hess <fmhess@users.sourceforge.net>
Cc: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ravishankar Karkala Mallikarjunayya <ravishankar.km@greenturtles.in>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: comedi: usbdux: Declare MODULE_FIRMWARE usage
Tim Gardner [Thu, 16 Aug 2012 17:15:37 +0000 (11:15 -0600)]
staging: comedi: usbdux: Declare MODULE_FIRMWARE usage

Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Mori Hess <fmhess@users.sourceforge.net>
Cc: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: comedi: usbduxsigma: Declare MODULE_FIRMWARE usage
Tim Gardner [Thu, 16 Aug 2012 17:19:11 +0000 (11:19 -0600)]
staging: comedi: usbduxsigma: Declare MODULE_FIRMWARE usage

Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Mori Hess <fmhess@users.sourceforge.net>
Cc: Bernd Porr <berndporr@f2s.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoscripts/decodecode: Fixup trapping instruction marker
Borislav Petkov [Wed, 15 Aug 2012 11:00:51 +0000 (13:00 +0200)]
scripts/decodecode: Fixup trapping instruction marker

When dumping "Code: " sections from an oops, the trapping instruction
%rip points to can be a string copy

  2b:*  f3 a5                   rep movsl %ds:(%rsi),%es:(%rdi)

and the line contain a bunch of ":".  Current "cut" selects only the and
the second field output looks funnily overlaid this:

  2b:*  f3 a5                   rep movsl %ds     <-- trapping instruction:(%rsi),%es:(%rdi

Fix this by selecting the remaining fields too.

Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-kbuild@vger.kernel.org
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge branch 'fixes' of git://git.infradead.org/users/vkoul/slave-dma
Linus Torvalds [Thu, 16 Aug 2012 18:13:16 +0000 (11:13 -0700)]
Merge branch 'fixes' of git://git.infradead.org/users/vkoul/slave-dma

Pull two slave-dmaengine fixes from Vinod Koul:
 "One fixes the correct use of clock API in imx driver and the other
  enables clock for tegra driver, which is used for other tegra driver
  conversion to dmanegine in -next."

* 'fixes' of git://git.infradead.org/users/vkoul/slave-dma:
  dma: tegra: enable/disable dma clock
  dma: imx-dma: Fix kernel crash due to missing clock conversion

12 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Thu, 16 Aug 2012 18:08:32 +0000 (11:08 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull more drm fixes from Dave Airlie:
 "Just some intel and nouveau ones this time, intel has more edp panel
  fixes for macbooks and nouveau has a suspend/resume regression fix in
  there."

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/i915: Apply post-sync write for pipe control invalidates
  drm/i915: reorder edp disabling to fix ivb MacBook Air
  drm/nv86/fifo: suspend fix
  drm/nouveau: disable copy engine on NVAF
  nouveau: fixup scanout enable in nvc0_pm
  drm/nouveau/aux: mask off higher bits of auxch index in i2c table entry
  drm/nvd0/disp: mask off high 16 bit of negative cursor x-coordinate
  drm/i915: ensure i2c adapter is all set before adding it
  drm/i915: ignore eDP bpc settings from vbt
  drm/i915: Fix blank panel at reopening lid
  drm/nve0/fifo: add support for the flip completion swmthd

12 years agomei: add mei_quirk_probe function
Tomas Winkler [Mon, 6 Aug 2012 12:23:55 +0000 (15:23 +0300)]
mei: add mei_quirk_probe function

The main purpose of this function is to exclude ME devices
without support for MEI/HECI interface from binding

Currently affected systems are C600/X79 based servers
that expose PCI device even though it doesn't supported ME Interface.
MEI driver accessing such nonfunctional device can corrupt
the system.

Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agovfs: fix propagation of atomic_open create error on negative dentry
Sage Weil [Wed, 15 Aug 2012 20:30:12 +0000 (13:30 -0700)]
vfs: fix propagation of atomic_open create error on negative dentry

If ->atomic_open() returns -ENOENT, we take care to return the create
error (e.g., EACCES), if any.  Do the same when ->atomic_open() returns 1
and provides a negative dentry.

This fixes a regression where an unprivileged open O_CREAT fails with
ENOENT instead of EACCES, introduced with the new atomic_open code.  It
is tested by the open/08.t test in the pjd posix test suite, and was
observed on top of fuse (backed by ceph-fuse).

Signed-off-by: Sage Weil <sage@inktank.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
12 years agoextcon: extcon_gpio: Replace gpio_request_one by devm_gpio_request_one
Axel Lin [Tue, 24 Jul 2012 01:26:57 +0000 (09:26 +0800)]
extcon: extcon_gpio: Replace gpio_request_one by devm_gpio_request_one

commit 01eaf24 "extcon: Convert extcon_gpio to devm_gpio_request_one"
missed the replacement for devm_gpio_request_one. fix it.

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agodrivers-core: make structured logging play nice with dynamic-debug
Jim Cromie [Thu, 19 Jul 2012 19:46:21 +0000 (13:46 -0600)]
drivers-core: make structured logging play nice with dynamic-debug

commit c4e00daaa96d3a0786f1f4fe6456281c60ef9a16 changed __dev_printk
in a way that broke dynamic-debug's ability to control the dynamic
prefix of dev_dbg(dev,..), but not dev_dbg(NULL,..) or pr_debug(..),
which is why it wasnt noticed sooner.

When dev==NULL, __dev_printk() just calls printk(), which just works.
But otherwise, it assumed that level was always a string like "<L>"
and just plucked out the 'L', ignoring the rest.  However,
dynamic_emit_prefix() adds "[tid] module:func:line:" to the string,
those additions all got lost.

Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Acked-by: Jason Baron <jbaron@redhat.com>
Cc: stable <stable@vger.kernel.org>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoUSB: qcserial: fix port handling on Gobi 1K and 2K+
Dan Williams [Mon, 23 Jul 2012 19:26:07 +0000 (14:26 -0500)]
USB: qcserial: fix port handling on Gobi 1K and 2K+

Bjorn's latest patchset does break Gobi 1K and 2K because on both
devices as it claims usb interface 0.  That's because usbif 0 is not
handled in the switch statement, and thus the if0 gets claimed when it
should not.  So let's just make things even simpler yet, and handle both
the 1K and 2K+ cases separately.  This patch should not affect the new
Sierra device support, because those devices are matched via
interface-specific matching and thus should never hit the composite
code.

Signed-off-by: Dan Williams <dcbw@redhat.com>
Tested-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agomei: fix device stall after wd is stopped
Tomas Winkler [Thu, 16 Aug 2012 13:25:33 +0000 (16:25 +0300)]
mei: fix device stall after wd is stopped

After watchdog was disabled the driver would stall
due to wrong calculation of credits reduction

The cat&paste bug was introduced in the commit
7bdf72d3d8059a50214069ea4b87c2174645f40f
mei: introduce mei_data2slots wrapper

Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoMerge branches 'cma', 'ipoib', 'misc', 'mlx4', 'ocrdma', 'qib' and 'srp' into for...
Roland Dreier [Thu, 16 Aug 2012 16:38:39 +0000 (09:38 -0700)]
Merge branches 'cma', 'ipoib', 'misc', 'mlx4', 'ocrdma', 'qib' and 'srp' into for-next

12 years agoIB/mlx4: Check iboe netdev pointer before dereferencing it
Kleber Sacilotto de Souza [Fri, 10 Aug 2012 18:25:34 +0000 (18:25 +0000)]
IB/mlx4: Check iboe netdev pointer before dereferencing it

Unlike other parts of the mlx4_ib code, the function build_mlx_header()
doesn't check if the iboe netdev of the given port is valid before
dereferencing it, which can cause a crash if the ethernet interface
has already been taken down.

Fix this by checking for a valid netdev pointer before using it to get
the port MAC address.

Signed-off-by: Kleber Sacilotto de Souza <klebers@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agomd: Don't truncate size at 4TB for RAID0 and Linear
NeilBrown [Thu, 16 Aug 2012 06:46:12 +0000 (16:46 +1000)]
md: Don't truncate size at 4TB for RAID0 and Linear

commit 27a7b260f71439c40546b43588448faac01adb93
   md: Fix handling for devices from 2TB to 4TB in 0.90 metadata.

changed 0.90 metadata handling to truncated size to 4TB as that is
all that 0.90 can record.
However for RAID0 and Linear, 0.90 doesn't need to record the size, so
this truncation is not needed and causes working arrays to become too small.

So avoid the truncation for RAID0 and Linear

This bug was introduced in 3.1 and is suitable for any stable kernels
from then onwards.
As the offending commit was tagged for 'stable', any stable kernel
that it was applied to should also get this patch.  That includes
at least 2.6.32, 2.6.33 and 3.0. (Thanks to Ben Hutchings for
providing that list).

Cc: stable@vger.kernel.org
Signed-off-by: Neil Brown <neilb@suse.de>
12 years agomlx4_core: Clean up buddy bitmap allocation
Roland Dreier [Tue, 14 Aug 2012 22:17:10 +0000 (15:17 -0700)]
mlx4_core: Clean up buddy bitmap allocation

 - Use kcalloc() / vzalloc() instead of an extra bitmap_zero().
 - Add __GFP_NOWARN to kcalloc() since we'll try vzalloc() if it fails.

Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agomlx4_core: Fix integer overflow issues around MTT table
Yishai Hadas [Mon, 13 Aug 2012 08:15:07 +0000 (08:15 +0000)]
mlx4_core: Fix integer overflow issues around MTT table

Fix some issues around int variables used in data structures related
to memory registration.

Handle int overflow in mlx4_init_icm_table by using a u64 intermediate
variable and changing struct mlx4_icm_table num_obj field to be u32.

Change some more fields/variables to use u32 instead of int to prevent
a case where the variable becomes negative when bit 31 is set.

Also subtract log_mtts_per_seg from the exponent when computing
num_mtt, since its added later on in that very same code area.

This and the previous commit fixes some issues which actually prevent
commit db5a7a65c058 ("mlx4_core: Scale size of MTT table with system
RAM") from working.  Now, when the number of MTTs is scaled with the
size of the RAM we can map up to 8TB.

Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Jack Morgenstein <jackm@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agomlx4_core: Allow large mlx4_buddy bitmaps
Yishai Hadas [Mon, 13 Aug 2012 08:15:06 +0000 (08:15 +0000)]
mlx4_core: Allow large mlx4_buddy bitmaps

mlx4_buddy_init uses kmalloc() to allocate bitmaps, which fails when
the required size is beyond the max supported value (or when memory is
too fragmented to handle a huge allocation).  Extend this to use use
vmalloc() if kmalloc() fails, and take that into account when freeing
the bitmaps as well.

This fixes a driver load failure when log num mtt is 26 or higher, and
is a step in the direction of allowing to register huge amounts of
memory on large memory systems.

Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc
Linus Torvalds [Thu, 16 Aug 2012 00:07:01 +0000 (17:07 -0700)]
Merge git://git./linux/kernel/git/davem/sparc

Pull two sparc fixes from David S. Miller.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc64: Be less verbose during vmemmap population.
  sparc64: do not clobber personality flags in sys_sparc64_personality()

12 years agoUSB: serial: Fix mos7840 timeout
Mark Ferrell [Tue, 24 Jul 2012 18:38:31 +0000 (13:38 -0500)]
USB: serial: Fix mos7840 timeout

* mos7840 driver was using multiple of HZ for the timeout handed off to
  usb_control_msg().  Changed the timeout to use msecs instead.

* Remove unused WAIT_FOR_EVER definition

Signed-off-by: Mark Ferrell <mferrell@uplogix.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoUSB: option: add ZTE K5006-Z
Bjørn Mork [Wed, 15 Aug 2012 13:43:33 +0000 (15:43 +0200)]
USB: option: add ZTE K5006-Z

The ZTE (Vodafone) K5006-Z use the following
interface layout:

00 DIAG
01 secondary
02 modem
03 networkcard
04 storage

Ignoring interface #3 which is handled by the qmi_wwan
driver.

Cc: Thomas Schäfer <tschaefer@t-online.de>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoIB/srp: Fix a race condition
Bart Van Assche [Tue, 14 Aug 2012 13:18:53 +0000 (13:18 +0000)]
IB/srp: Fix a race condition

Avoid a crash caused by the scmnd->scsi_done(scmnd) call in
srp_process_rsp() being invoked with scsi_done == NULL.  This can
happen if a reply is received during or after a command abort.

Reported-by: Joseph Glanville <joseph.glanville@orionvm.com.au>
Reference: http://marc.info/?l=linux-rdma&m=134314367801595
Cc: <stable@vger.kernel.org>
Acked-by: David Dillow <dillowda@ornl.gov>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agoIB/qib: Fix error return code in qib_init_7322_variables()
Julia Lawall [Tue, 14 Aug 2012 12:58:35 +0000 (12:58 +0000)]
IB/qib: Fix error return code in qib_init_7322_variables()

Convert a 0 error return code to a negative one, as returned elsewhere
in the function.

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

// <smpl>
@@
identifier ret;
expression e,e1,e2,e3,e4,x;
@@

(
if (\(ret != 0\|ret < 0\) || ...) { ... return ...; }
|
ret = 0
)
... when != ret = e1
*x = \(kmalloc\|kzalloc\|kcalloc\|devm_kzalloc\|ioremap\|ioremap_nocache\|devm_ioremap\|devm_ioremap_nocache\)(...);
... when != x = e2
    when != ret = e3
*if (x == NULL || ...)
{
  ... when != ret = e4
*  return ret;
}
// </smpl>

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Acked-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agoIB: Fix typos in infiniband drivers
Masanari Iida [Fri, 10 Aug 2012 00:07:58 +0000 (00:07 +0000)]
IB: Fix typos in infiniband drivers

Correct spelling typos in comments in drivers/infiniband.

Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
12 years agoC6X: select GENERIC_ATOMIC64
Mark Salter [Wed, 15 Aug 2012 16:12:16 +0000 (12:12 -0400)]
C6X: select GENERIC_ATOMIC64

The generic atomic64 support came in 2009 to support the perf subsystem
with the expectation that all architectures would implement atomic64
support. Since then, other optional parts of the generic kernel have
also come to expect atomic64 support. This patch enables generic atomic64
support for C6X architecture.

Signed-off-by: Mark Salter <msalter@redhat.com>
12 years agoC6X: add Lx_CACHE_SHIFT defines
Mark Salter [Wed, 15 Aug 2012 15:52:01 +0000 (11:52 -0400)]
C6X: add Lx_CACHE_SHIFT defines

C6X currently lacks Lx_CACHE_SHIFT defines which are needed in a
few places in the generic kernel. This patch adds _SHIFT defines
for the various caches and bases the Lx_CACHE_BYTES defines on
them.

Signed-off-by: Mark Salter <msalter@redhat.com>
12 years agofuse: check create mode in atomic open
Miklos Szeredi [Wed, 15 Aug 2012 11:01:24 +0000 (13:01 +0200)]
fuse: check create mode in atomic open

Verify that the VFS is passing us a complete create mode with the S_IFREG to
atomic open.

Reported-by: Steve <steveamigauk@yahoo.co.uk>
Reported-by: Richard W.M. Jones <rjones@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
12 years agovfs: pass right create mode to may_o_create()
Miklos Szeredi [Wed, 15 Aug 2012 11:01:24 +0000 (13:01 +0200)]
vfs: pass right create mode to may_o_create()

Pass the umask-ed create mode to may_o_create() instead of the original one.

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
12 years agovfs: atomic_open(): fix create mode usage
Miklos Szeredi [Wed, 15 Aug 2012 11:01:24 +0000 (13:01 +0200)]
vfs: atomic_open(): fix create mode usage

Don't mask S_ISREG off the create mode before passing to ->atomic_open().  Other
methods (->create, ->mknod) also get the complete file mode and filesystems
expect it.

Reported-by: Steve <steveamigauk@yahoo.co.uk>
Reported-by: Richard W.M. Jones <rjones@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
12 years agovfs: canonicalize create mode in build_open_flags()
Miklos Szeredi [Wed, 15 Aug 2012 11:01:24 +0000 (13:01 +0200)]
vfs: canonicalize create mode in build_open_flags()

Userspace can pass weird create mode in open(2) that we canonicalize to
"(mode & S_IALLUGO) | S_IFREG" in vfs_create().

The problem is that we use the uncanonicalized mode before calling vfs_create()
with unforseen consequences.

So do the canonicalization early in build_open_flags().

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
CC: stable@vger.kernel.org