platform/kernel/linux-stable.git
12 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-block
Linus Torvalds [Sat, 11 Feb 2012 18:07:11 +0000 (10:07 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block

Says Jens:

 "Time to push off some of the pending items.  I really wanted to wait
  until we had the regression nailed, but alas it's not quite there yet.
  But I'm very confident that it's "just" a missing expire on exit, so
  fix from Tejun should be fairly trivial.  I'm headed out for a week on
  the slopes.

  - Killing the barrier part of mtip32xx.  It doesn't really support
    barriers, and it doesn't need them (writes are fully ordered).

  - A few fixes from Dan Carpenter, preventing overflows of integer
    multiplication.

  - A fixup for loop, fixing a previous commit that didn't quite solve
    the partial read problem from Dave Young.

  - A bio integer overflow fix from Kent Overstreet.

  - Improvement/fix of the door "keep locked" part of the cdrom shared
    code from Paolo Benzini.

  - A few cfq fixes from Shaohua Li.

  - A fix for bsg sysfs warning when removing a file it did not create
    from Stanislaw Gruszka.

  - Two fixes for floppy from Vivek, preventing a crash.

  - A few block core fixes from Tejun.  One killing the over-optimized
    ioc exit path, cleaning that up nicely.  Two others fixing an oops
    on elevator switch, due to calling into the scheduler merge check
    code without holding the queue lock."

* 'for-linus' of git://git.kernel.dk/linux-block:
  block: fix lockdep warning on io_context release put_io_context()
  relay: prevent integer overflow in relay_open()
  loop: zero fill bio instead of return -EIO for partial read
  bio: don't overflow in bio_get_nr_vecs()
  floppy: Fix a crash during rmmod
  floppy: Cleanup disk->queue before caling put_disk() if add_disk() was never called
  cdrom: move shared static to cdrom_device_info
  bsg: fix sysfs link remove warning
  block: don't call elevator callbacks for plug merges
  block: separate out blk_rq_merge_ok() and blk_try_merge() from elevator functions
  mtip32xx: removed the irrelevant argument of mtip_hw_submit_io() and the unused member of struct driver_data
  block: strip out locking optimization in put_io_context()
  cdrom: use copy_to_user() without the underscores
  block: fix ioc locking warning
  block: fix NULL icq_cache reference
  block,cfq: change code order

12 years agoblock: fix lockdep warning on io_context release put_io_context()
Tejun Heo [Sat, 11 Feb 2012 11:37:25 +0000 (12:37 +0100)]
block: fix lockdep warning on io_context release put_io_context()

11a3122f6c "block: strip out locking optimization in put_io_context()"
removed ioc_lock depth lockdep annoation along with locking
optimization; however, while recursing from put_io_context() is no
longer possible, ioc_release_fn() may still end up putting the last
reference of another ioc through elevator, which wlil grab ioc->lock
triggering spurious (as the ioc is always different one) A-A deadlock
warning.

As this can only happen one time from ioc_release_fn(), using non-zero
subclass from ioc_release_fn() is enough.  Use subclass 1.

Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Fri, 10 Feb 2012 22:18:46 +0000 (14:18 -0800)]
Merge git://git./linux/kernel/git/davem/net

Quoth David:

1) GRO MAC header comparisons were ethernet specific, breaking other
   link types.  This required a multi-faceted fix to cure the originally
   noted case (Infiniband), because IPoIB was lying about it's actual
   hard header length.  Thanks to Eric Dumazet, Roland Dreier, and
   others.

2) Fix build failure when INET_UDP_DIAG is built in and ipv6 is modular.
   From Anisse Astier.

3) Off by ones and other bug fixes in netprio_cgroup from Neil Horman.

4) ipv4 TCP reset generation needs to respect any network interface
   binding from the socket, otherwise route lookups might give a
   different result than all the other segments received.  From Shawn
   Lu.

5) Fix unintended regression in ipv4 proxy ARP responses, from Thomas
   Graf.

6) Fix SKB under-allocation bug in sh_eth, from Yoshihiro Shimoda.

7) Revert skge PCI mapping changes that are causing crashes for some
   folks, from Stephen Hemminger.

8) IPV4 route lookups fill in the wildcarded fields of the given flow
   lookup key passed in, which is fine most of the time as this is
   exactly what the caller's want.  However there are a few cases that
   want to retain the original flow key values afterwards, so handle
   those cases properly.  Fix from Julian Anastasov.

9) IGB/IXGBE VF lookup bug fixes from Greg Rose.

10) Properly null terminate filename passed to ethtool flash device
    method, from Ben Hutchings.

11) S3 resume fix in via-velocity from David Lv.

12) Fix double SKB free during xmit failure in CAIF, from Dmitry
    Tarnyagin.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (72 commits)
  net: Don't proxy arp respond if iif == rt->dst.dev if private VLAN is disabled
  ipv4: Fix wrong order of ip_rt_get_source() and update iph->daddr.
  netprio_cgroup: fix wrong memory access when NETPRIO_CGROUP=m
  netprio_cgroup: don't allocate prio table when a device is registered
  netprio_cgroup: fix an off-by-one bug
  bna: fix error handling of bnad_get_flash_partition_by_offset()
  isdn: type bug in isdn_net_header()
  net: Make qdisc_skb_cb upper size bound explicit.
  ixgbe: ethtool: stats user buffer overrun
  ixgbe: dcb: up2tc mapping lost on disable/enable CEE DCB state
  ixgbe: do not update real num queues when netdev is going away
  ixgbe: Fix broken dependency on MAX_SKB_FRAGS being related to page size
  ixgbe: Fix case of Tx Hang in PF with 32 VFs
  ixgbe: fix vf lookup
  igb: fix vf lookup
  e1000: add dropped DMA receive enable back in for WoL
  gro: more generic L2 header check
  IPoIB: Stop lying about hard_header_len and use skb->cb to stash LL addresses
  zd1211rw: firmware needs duration_id set to zero for non-pspoll frames
  net: enable TC35815 for MIPS again
  ...

12 years agonet: Don't proxy arp respond if iif == rt->dst.dev if private VLAN is disabled
Thomas Graf [Fri, 10 Feb 2012 04:07:11 +0000 (04:07 +0000)]
net: Don't proxy arp respond if iif == rt->dst.dev if private VLAN is disabled

Commit 653241 (net: RFC3069, private VLAN proxy arp support) changed
the behavior of arp proxy to send arp replies back out on the interface
the request came in even if the private VLAN feature is disabled.

Previously we checked rt->dst.dev != skb->dev for in scenarios, when
proxy arp is enabled on for the netdevice and also when individual proxy
neighbour entries have been added.

This patch adds the check back for the pneigh_lookup() scenario.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Jesper Dangaard Brouer <hawk@comx.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv4: Fix wrong order of ip_rt_get_source() and update iph->daddr.
Li Wei [Thu, 9 Feb 2012 21:15:25 +0000 (21:15 +0000)]
ipv4: Fix wrong order of ip_rt_get_source() and update iph->daddr.

This patch fix a bug which introduced by commit ac8a4810 (ipv4: Save
nexthop address of LSRR/SSRR option to IPCB.).In that patch, we saved
the nexthop of SRR in ip_option->nexthop and update iph->daddr until
we get to ip_forward_options(), but we need to update it before
ip_rt_get_source(), otherwise we may get a wrong src.

Signed-off-by: Li Wei <lw@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetprio_cgroup: fix wrong memory access when NETPRIO_CGROUP=m
Neil Horman [Fri, 10 Feb 2012 05:43:38 +0000 (05:43 +0000)]
netprio_cgroup: fix wrong memory access when NETPRIO_CGROUP=m

When the netprio_cgroup module is not loaded, net_prio_subsys_id
is -1, and so sock_update_prioidx() accesses cgroup_subsys array
with negative index subsys[-1].

Make the code resembles cls_cgroup code, which is bug free.

Origionally-authored-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
CC: "David S. Miller" <davem@davemloft.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetprio_cgroup: don't allocate prio table when a device is registered
Neil Horman [Fri, 10 Feb 2012 05:43:37 +0000 (05:43 +0000)]
netprio_cgroup: don't allocate prio table when a device is registered

So we delay the allocation till the priority is set through cgroup,
and this makes skb_update_priority() faster when it's not set.

This also eliminates an off-by-one bug similar with the one fixed
in the previous patch.

Origionally-authored-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
CC: "David S. Miller" <davem@davemloft.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetprio_cgroup: fix an off-by-one bug
Neil Horman [Fri, 10 Feb 2012 05:43:36 +0000 (05:43 +0000)]
netprio_cgroup: fix an off-by-one bug

# mount -t cgroup xxx /mnt
  # mkdir /mnt/tmp
  # cat /mnt/tmp/net_prio.ifpriomap
  lo 0
  eth0 0
  virbr0 0
  # echo 'lo 999' > /mnt/tmp/net_prio.ifpriomap
  # cat /mnt/tmp/net_prio.ifpriomap
  lo 999
  eth0 0
  virbr0 4101267344

We got weired output, because we exceeded the boundary of the array.
We may even crash the kernel..

Origionally-authored-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
CC: "David S. Miller" <davem@davemloft.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Fri, 10 Feb 2012 17:06:25 +0000 (09:06 -0800)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/i915: fixup interlaced bits clearing in PIPECONF on PCH_SPLIT (v2)
  drm/i915: no lvds quirk for AOpen MP45
  drm/i915: Force explicit bpp selection for intel_dp_link_required
  drm/radeon: do not continue after error from r600_ib_test
  drivers/gpu/drm/drm_ioc32.c: initialize all fields
  drm/i915: fixup interlaced bits clearing in PIPECONF on PCH_SPLIT
  drm/i915:: Disable FBC on SandyBridge

12 years agoMerge tag 'writeback-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/wfg...
Linus Torvalds [Fri, 10 Feb 2012 17:05:52 +0000 (09:05 -0800)]
Merge tag 'writeback-fixes' of git://git./linux/kernel/git/wfg/linux

fix 1 mysterious divide error
fix 3 NULL dereference bugs in writeback tracing, on SD card removal w/o umount

* tag 'writeback-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/wfg/linux:
  writeback: fix dereferencing NULL bdi->dev on trace_writeback_queue
  lib: proportion: lower PROP_MAX_SHIFT to 32 on 64-bit kernel
  writeback: fix NULL bdi->dev in trace writeback_single_inode
  backing-dev: fix wakeup timer races with bdi_unregister()

12 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 10 Feb 2012 17:05:07 +0000 (09:05 -0800)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf: Fix double start/stop in x86_pmu_start()
  perf evsel: Fix an issue where perf report fails to show the proper percentage
  perf tools: Fix prefix matching for kernel maps
  perf tools: Fix perf stack to non executable on x86_64
  perf: Remove deprecated WARN_ON_ONCE()

12 years agoMerge tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck...
Linus Torvalds [Fri, 10 Feb 2012 17:04:37 +0000 (09:04 -0800)]
Merge tag 'hwmon-for-linus' of git://git./linux/kernel/git/groeck/linux-staging

Couple of regressions, couple of zero-day bugs, a minor enhancement.
Nothing really major.

* tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (f75375s) Let f75375_update_device treat pwmX as a measured value
  hwmon: (f75375s) Fix bit shifting in f75375_write16
  hwmon: (f75375s) Fix reading of wrong register when initializing the F75387
  hwmon: (f75375s) Fix automatic pwm mode setting for F75373 & F75375
  hwmon: (w83627ehf) Remove duplicate code

12 years agoMerge branch 'drm-intel-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/keith...
Dave Airlie [Fri, 10 Feb 2012 08:35:19 +0000 (08:35 +0000)]
Merge branch 'drm-intel-fixes' of git://git./linux/kernel/git/keithp/linux into drm-fixes

* 'drm-intel-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/keithp/linux:
  drm/i915: fixup interlaced bits clearing in PIPECONF on PCH_SPLIT (v2)
  drm/i915: no lvds quirk for AOpen MP45
  drm/i915: Force explicit bpp selection for intel_dp_link_required
  drm/i915: fixup interlaced bits clearing in PIPECONF on PCH_SPLIT
  drm/i915:: Disable FBC on SandyBridge

12 years agorelay: prevent integer overflow in relay_open()
Dan Carpenter [Fri, 10 Feb 2012 08:03:58 +0000 (09:03 +0100)]
relay: prevent integer overflow in relay_open()

"subbuf_size" and "n_subbufs" come from the user and they need to be
capped to prevent an integer overflow.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: stable@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agoMerge tag 'tty-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
Linus Torvalds [Thu, 9 Feb 2012 21:52:57 +0000 (13:52 -0800)]
Merge tag 'tty-3.3-rc3' of git://git./linux/kernel/git/gregkh/tty

Serial/TTY fixes for the 3.3-rc3 tree

Just a few new device ids, omap serial driver regression fixes, and a
build fix for the 8250 driver movement.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* tag 'tty-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  tty: serial: omap-serial: wakeup latency constraint is in microseconds, not milliseconds
  tty: serial: OMAP: block idle while the UART is transferring data in PIO mode
  tty: serial: OMAP: use a 1-byte RX FIFO threshold in PIO mode
  m32r: relocate drivers back out of 8250 dir
  tty: fix a build failure on sparc
  serial: samsung: Add support for EXYNOS5250
  serial: samsung: Add support for EXYNOS4212 and EXYNOS4412
  drivers/tty/vt/vt_ioctl.c: fix KDFONTOP 32bit compatibility layer

12 years agoMerge tag 'staging-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Thu, 9 Feb 2012 21:52:18 +0000 (13:52 -0800)]
Merge tag 'staging-3.3-rc3' of git://git./linux/kernel/git/gregkh/staging

Staging tree patches for 3.3-rc3

Big things here is the deletion of the Android pmem driver, as it's
obsolete and no one uses it, the gma500 driver as it's already in the
drm portion of the kernel tree, and the pohmelfs filesystem as it's
obsolete and a rewritten version is being proposed for the fs/ section
of the kernel.

Other than that, a smattering of different bugfixes and regressions, and
some omap drm api merge fixups that were needed due to api changes in
the main portion of the drm tree, allowing this code to build properly
again.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* tag 'staging-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging: (28 commits)
  staging: pohmelfs: remove drivers/staging/pohmelfs
  staging: android/ram_console: Don't build on arches w/o ioremap
  staging: r8712u: Use asynchronous firmware loading
  staging: usbip: fix to prevent potentially using uninitialized spinlock
  staging: r8712u: Fix problem when CONFIG_R8712_AP is set
  staging: tidspbridge: fix incorrect free to drv_datap
  staging: tidspbridge: fix bridge_open memory leaks
  staging: android: lowmemorykiller: Don't wait more than one second for a process to die
  MAINTAINERS: staging: iio: add iio information
  staging: zcache: fix serialization bug in zv stats
  staging: fix go7007-usb license
  Staging: android: binder: Fix crashes when sharing a binder file between processes
  Staging: android: Remove pmem driver
  Staging: asus_oled: fix NULL-ptr crash on unloading
  Staging: asus_oled: fix image processing
  Staging: android: binder: Don't call dump_stack in binder_vma_open
  staging: r8712u: Add new Sitecom UsB ID
  zcache: Set SWIZ_BITS to 8 to reduce tmem bucket lock contention.
  zcache: fix deadlock condition
  staging: drm/omap: fix locking issue
  ...

12 years agoMerge tag 'driver-core-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 9 Feb 2012 21:51:36 +0000 (13:51 -0800)]
Merge tag 'driver-core-3.3-rc3' of git://git./linux/kernel/git/gregkh/driver-core

Driver core fixes for the 3.3-rc3 tree.

A few fixes for kobject warnings that have popped up in the cpu hotplug path,
and a regression fix for the speed of the hotplug memory code.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* tag 'driver-core-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
  driver-core: cpu: fix kobject warning when hotplugging a cpu
  ACPI: remove duplicated lines of merging problems with acpi_processor_add
  docbook: fix fatal errors in device-drivers docbook and add DMA Management section
  drivers/base/memory.c: fix memory_dev_init() long delay
  driver core: cpu: remove kernel warning when removing a cpu

12 years agoMerge tag 'char-misc-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Thu, 9 Feb 2012 21:51:13 +0000 (13:51 -0800)]
Merge tag 'char-misc-3.3-rc3' of git://git./linux/kernel/git/gregkh/char-misc

Minor char-misc fixes for 3.3-rc3

Nothing big here, some Kconfig fixes for the MISC_DEVICES config option
that was being used incorrectly, and some other minor bug fixes.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* tag 'char-misc-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
  mmc: cb710 core: Add missing spin_lock_init for irq_lock of struct cb710_chip
  cs5535-mfgpt: don't call __init function from __devinit
  vmw_balloon: fix for a -Wuninitialized warning
  drivers: misc: Remove MISC_DEVICES config option
  c2port: fix build error for duramar2150 due to missing header.

12 years agoMerge tag 'usb-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Thu, 9 Feb 2012 21:50:54 +0000 (13:50 -0800)]
Merge tag 'usb-3.3-rc3' of git://git./linux/kernel/git/gregkh/usb

USB fixes for 3.3-rc3

Here are a few minor USB fixes and a bunch of device id updates for the
USB drivers.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* tag 'usb-3.3-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  USB: usbserial: add new PID number (0xa951) to the ftdi driver
  usb: ch9.h: usb_endpoint_maxp() uses __le16_to_cpu()
  usb: musb: fix a build error on mips
  uwb & wusb & usb wireless controllers: fix kconfig error & build errors
  usb: Skip PCI USB quirk handling for Netlogic XLP
  powerpc/usb: fix issue of CPU halt when missing USB PHY clock
  usb: otg: mv_otg: Add dependence
  usb: host: Distinguish Kconfig text for Freescale controllers
  USB: add new zte 3g-dongle's pid to option.c
  usb: ch9.h: usb_endpoint_maxp() uses __le16_to_cpu()
  USB: qcserial: don't enable autosuspend
  USB: qcserial: add several new serial devices
  usb: otg: mv_otg: Add dependence
  usb: gadget: zero: fix bug in loopback autoresume handling

12 years agobna: fix error handling of bnad_get_flash_partition_by_offset()
Dan Carpenter [Thu, 9 Feb 2012 00:49:34 +0000 (00:49 +0000)]
bna: fix error handling of bnad_get_flash_partition_by_offset()

The current error handling doesn't work because we flash_part is a u32
so the checks for negative error codes don't work.  I considered making
things signed but I don't know the hardware enough to say if that's a
problem.  Really, we don't use the error codes so just returning zero
for all problems is fine.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoisdn: type bug in isdn_net_header()
Dan Carpenter [Thu, 9 Feb 2012 00:46:47 +0000 (00:46 +0000)]
isdn: type bug in isdn_net_header()

We use len to store the return value from eth_header().  eth_header()
can return -ETH_HLEN (-14).  We want to pass this back instead of
truncating it to 65522 and returning that.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agohwmon: (f75375s) Let f75375_update_device treat pwmX as a measured value
Nikolaus Schulz [Wed, 8 Feb 2012 17:56:11 +0000 (18:56 +0100)]
hwmon: (f75375s) Let f75375_update_device treat pwmX as a measured value

Treat pwmX as a measured value, not as a (mostly static) limit value, so
that it is updated more frequently from the device register.

Signed-off-by: Nikolaus Schulz <mail@microschulz.de>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
12 years agonet: Make qdisc_skb_cb upper size bound explicit.
David S. Miller [Mon, 6 Feb 2012 20:14:37 +0000 (15:14 -0500)]
net: Make qdisc_skb_cb upper size bound explicit.

Just like skb->cb[], so that qdisc_skb_cb can be encapsulated inside
of other data structures.

This is intended to be used by IPoIB so that it can remember
addressing information stored at hard_header_ops->create() time that
it can fetch when the packet gets to the transmit routine.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotty: serial: omap-serial: wakeup latency constraint is in microseconds, not milliseconds
Paul Walmsley [Thu, 26 Jan 2012 02:50:56 +0000 (19:50 -0700)]
tty: serial: omap-serial: wakeup latency constraint is in microseconds, not milliseconds

The receive FIFO wakeup latency estimate in the omap-serial driver is
three orders of magnitude too small.  This effectively prevents the
MPU from going to a low-power state when CONFIG_CPU_IDLE=y.  This is a
major power management regression and masks some other FIFO-related
bugs in the driver.

Fix by correcting the most egregious problem in the RX wakeup latency
estimate.  There are several other flaws in the estimator; these will
be fixed by a separate patch series intended for 3.4.

The difference in low-power states with this patch can be observed via
debugfs in pm_debug/count.

This estimate does not have any effect when CONFIG_CPU_IDLE=n.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: Alan Cox <alan@linux.intel.com>
Acked-by: Govindraj.R <govindraj.raja@ti.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>
Tested-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agotty: serial: OMAP: block idle while the UART is transferring data in PIO mode
Paul Walmsley [Thu, 26 Jan 2012 02:50:52 +0000 (19:50 -0700)]
tty: serial: OMAP: block idle while the UART is transferring data in PIO mode

Prevent OMAP UARTs from going idle while they are still transferring
data in PIO mode.  This works around an oversight in the OMAP UART
hardware present in OMAP34xx and earlier: an idle UART won't send a
wakeup when the TX FIFO threshold is reached.  This causes long delays
during data transmission when the MPU powerdomain enters a low-power
mode.  The MPU interrupt controller is not able to respond to
interrupts when it's in a low-power state, so the TX buffer is not
refilled until another wakeup event occurs.

This fix changes the erratum i291 DMA idle workaround.  Rather than
toggling between force-idle and no-idle, it will toggle between
smart-idle and no-idle.  The important part of the workaround is the
no-idle part, so this shouldn't result in any change in behavior.

This fix should work on all OMAP UARTs.  Future patches intended for
the 3.4 merge window will make this workaround conditional on a
"feature" flag, and will use the OMAP36xx+ TX event wakeup support.

Thanks to Kevin Hilman <khilman@ti.com> for mentioning the erratum i291
workaround, which led to the development of this approach.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Alan Cox <alan@linux.intel.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Acked-by: Govindraj.R <govindraj.raja@ti.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>
Tested-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agotty: serial: OMAP: use a 1-byte RX FIFO threshold in PIO mode
Paul Walmsley [Thu, 26 Jan 2012 02:50:36 +0000 (19:50 -0700)]
tty: serial: OMAP: use a 1-byte RX FIFO threshold in PIO mode

In the (default) PIO mode, use a one-byte RX FIFO threshold.  The OMAP
UART IP blocks do not appear to be capable of waking the system under
an RX timeout condition.  Since the previous RX FIFO threshold was 16
bytes, this meant that omap-serial.c did not become aware of any
received data until all those bytes arrived or until another UART
interrupt occurred.  This made the serial console and presumably other
serial applications (GPS, serial Bluetooth) unusable or extremely
slow.  A 1-byte RX FIFO threshold also allows the MPU to enter a
low-power consumption state while waiting for the FIFO to fill.

This can be verified using the serial console by comparing the
behavior when "0123456789abcde" is pasted in from another window, with
the behavior when "0123456789abcdef" is pasted in.  Since the former
string is less than sixteen bytes long, the string is not echoed for
some time, while the latter string is echoed immediately.

DMA operation is unaffected by this patch.

Thanks to Russell King - ARM Linux <linux@arm.linux.org.uk> for some
additional information on the standard behavior of the RX timeout
event, which was used to improve this commit description.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: Govindraj Raja <govindraj.r@ti.com>
Cc: Alan Cox <alan@linux.intel.com>
Cc: Russell King <linux@arm.linux.org.uk>
Reviewed-by: Kevin Hilman <khilman@ti.com>
Tested-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoReduce the number of expensive division instructions done by _parse_integer()
David Howells [Thu, 9 Feb 2012 15:48:20 +0000 (15:48 +0000)]
Reduce the number of expensive division instructions done by _parse_integer()

_parse_integer() does one or two division instructions (which are slow)
per digit parsed to perform the overflow check.

Furthermore, these are particularly expensive examples of division
instruction as the number of clock cycles required to complete them may
go up with the position of the most significant set bit in the dividend:

if (*res > div_u64(ULLONG_MAX - val, base))

which is as maximal as possible.

Worse, on 32-bit arches, more than one of these division instructions
may be required per digit.

So, assuming we don't support a base of more than 16, skip the check if the
top nibble of the result is not set at this point.

Signed-off-by: David Howells <dhowells@redhat.com>
[ Changed it to not dereference the pointer all the time - even if the
  compiler can and does optimize it away, the code just looks cleaner.
  And edited the top nybble test slightly to make the code generated on
  x86-64 better in the loop - test against a hoisted constant instead of
  shifting and testing the result ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoixgbe: ethtool: stats user buffer overrun
John Fastabend [Sat, 28 Jan 2012 03:32:17 +0000 (03:32 +0000)]
ixgbe: ethtool: stats user buffer overrun

If the number of tx/rx queues changes the ethtool ioctl
ETHTOOL_GSTATS may overrun the userspace buffer. This
occurs because the general practice in user space to
query stats is to issue a ETHTOOL_GSSET cmd to learn the
buffer size needed, allocate the buffer, then call
ETHTOOL_GSTIRNGS and ETHTOOL_GSTATS. If the number of
real_num_queues is changed or flow control attributes
are changed after ETHTOOL_GSSET but before the
ETHTOOL_GSTRINGS/ETHTOOL_GSTATS a user space buffer
overrun occurs.

To fix the overrun always return the max buffer size
needed from get_sset_count() then return all strings
and stats from get_strings()/get_ethtool_stats().

This _will_ change the output from the ioctl() call
which could break applications and script parsing in
theory. I believe these changes should not break existing
tools because the only changes will be more {tx|rx}_queues
and the {tx|rx}_pb_* stats will always be returned.
Existing scripts already need to handle changing number
of queues because this occurs today depending on system
and current features. The {tx|rx}_pb_* stats are at the
end of the output and should be handled by scripts today
regardless.

Finally get_ethtool_stats and get_strings are free-form
outputs tools parsing these outputs should be defensive
anyways. In the end these updates are better then
having a tool segfault because of a buffer overrun.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: dcb: up2tc mapping lost on disable/enable CEE DCB state
John Fastabend [Sat, 28 Jan 2012 01:22:35 +0000 (01:22 +0000)]
ixgbe: dcb: up2tc mapping lost on disable/enable CEE DCB state

Users expect the up2tc mapping to be maintained across a DCB
enable/disable/enable transition. And since we maintain all
the other DCB attributes we should do this for up2tc mappings
as well just to be consistent. Also without this we break
user space applications that expect this to occur that
previously worked.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: do not update real num queues when netdev is going away
Yi Zou [Sat, 7 Jan 2012 08:39:50 +0000 (08:39 +0000)]
ixgbe: do not update real num queues when netdev is going away

If the netdev is already in NETREG_UNREGISTERING/_UNREGISTERED state, do not
update the real num tx queues. netdev_queue_update_kobjects() is already
called via remove_queue_kobjects() at NETREG_UNREGISTERING time. So, when
upper layer driver, e.g., FCoE protocol stack is monitoring the netdev
event of NETDEV_UNREGISTER and calls back to LLD ndo_fcoe_disable() to remove
extra queues allocated for FCoE, the associated txq sysfs kobjects are already
removed, and trying to update the real num queues would cause something like
below:

...
PID: 25138  TASK: ffff88021e64c440  CPU: 3   COMMAND: "kworker/3:3"
 #0 [ffff88021f007760] machine_kexec at ffffffff810226d9
 #1 [ffff88021f0077d0] crash_kexec at ffffffff81089d2d
 #2 [ffff88021f0078a0] oops_end at ffffffff813bca78
 #3 [ffff88021f0078d0] no_context at ffffffff81029e72
 #4 [ffff88021f007920] __bad_area_nosemaphore at ffffffff8102a155
 #5 [ffff88021f0079f0] bad_area_nosemaphore at ffffffff8102a23e
 #6 [ffff88021f007a00] do_page_fault at ffffffff813bf32e
 #7 [ffff88021f007b10] page_fault at ffffffff813bc045
    [exception RIP: sysfs_find_dirent+17]
    RIP: ffffffff81178611  RSP: ffff88021f007bc0  RFLAGS: 00010246
    RAX: ffff88021e64c440  RBX: ffffffff8156cc63  RCX: 0000000000000004
    RDX: ffffffff8156cc63  RSI: 0000000000000000  RDI: 0000000000000000
    RBP: ffff88021f007be0   R8: 0000000000000004   R9: 0000000000000008
    R10: ffffffff816fed00  R11: 0000000000000004  R12: 0000000000000000
    R13: ffffffff8156cc63  R14: 0000000000000000  R15: ffff8802222a0000
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #8 [ffff88021f007be8] sysfs_get_dirent at ffffffff81178c07
 #9 [ffff88021f007c18] sysfs_remove_group at ffffffff8117ac27
#10 [ffff88021f007c48] netdev_queue_update_kobjects at ffffffff813178f9
#11 [ffff88021f007c88] netif_set_real_num_tx_queues at ffffffff81303e38
#12 [ffff88021f007cc8] ixgbe_set_num_queues at ffffffffa0249763 [ixgbe]
#13 [ffff88021f007cf8] ixgbe_init_interrupt_scheme at ffffffffa024ea89 [ixgbe]
#14 [ffff88021f007d48] ixgbe_fcoe_disable at ffffffffa0267113 [ixgbe]
#15 [ffff88021f007d68] vlan_dev_fcoe_disable at ffffffffa014fef5 [8021q]
#16 [ffff88021f007d78] fcoe_interface_cleanup at ffffffffa02b7dfd [fcoe]
#17 [ffff88021f007df8] fcoe_destroy_work at ffffffffa02b7f08 [fcoe]
#18 [ffff88021f007e18] process_one_work at ffffffff8105d7ca
#19 [ffff88021f007e68] worker_thread at ffffffff81060513
#20 [ffff88021f007ee8] kthread at ffffffff810648b6
#21 [ffff88021f007f48] kernel_thread_helper at ffffffff813c40f4

Signed-off-by: Yi Zou <yi.zou@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Fix broken dependency on MAX_SKB_FRAGS being related to page size
Alexander Duyck [Thu, 10 Nov 2011 09:09:17 +0000 (09:09 +0000)]
ixgbe: Fix broken dependency on MAX_SKB_FRAGS being related to page size

This patch fixes an issue in which RSC will generate corrupted frames when
PAGE_SIZE is larger than 8K.  Specifically it looks like that in 2.6.39 a
change was made so that GRO would always have at least 16 frags available
for coalescing, but the ixgbe RSC logic was not updated.  As such the RSC
feature would generate a frame larger than 64K and then overflow the value
in the IP length field.

To correct that I am now basing things on the PAGE_SIZE.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Fix case of Tx Hang in PF with 32 VFs
Greg Rose [Wed, 25 Jan 2012 07:59:37 +0000 (07:59 +0000)]
ixgbe: Fix case of Tx Hang in PF with 32 VFs

A check for the number of VFs allocated should have used a greater than
equal operator instead of just greater than.  This caused allocation of
exactly 32 VFs to not enable the PF transmit and receive enables.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Tested-by: Robert E Garrett <robertX.e.garrett@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: fix vf lookup
Greg Rose [Fri, 3 Feb 2012 00:54:13 +0000 (00:54 +0000)]
ixgbe: fix vf lookup

Recent addition of code to find already allocated VFs failed to take
account that systems with 2 or more multi-port SR-IOV capable controllers
might have already enabled VFs.  Make sure that the VFs the function is
finding are actually subordinate to the particular instance of the adapter
that is looking for them and not subordinate to some device that has
previously enabled SR-IOV.

This bug exists in 3.2 stable as well as 3.3 release candidates.

CC: stable@vger.kernel.org
Reported-by: David Ahern <daahern@cisco.com>
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Tested-by: Robert E Garrett <robertX.e.garrett@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoigb: fix vf lookup
Greg Rose [Thu, 2 Feb 2012 23:51:43 +0000 (23:51 +0000)]
igb: fix vf lookup

Recent addition of code to find already allocated VFs failed to take
account that systems with 2 or more multi-port SR-IOV capable controllers
might have already enabled VFs.  Make sure that the VFs the function is
finding are actually subordinate to the particular instance of the adapter
that is looking for them and not subordinate to some device that has
previously enabled SR-IOV.

This is applicable to 3.2+ kernels.

CC: stable@vger.kernel.org
Reported-by: David Ahern <daahern@cisco.com>
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Tested-by: Robert E Garrett <robertX.e.garrett@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoe1000: add dropped DMA receive enable back in for WoL
Dean Nelson [Thu, 19 Jan 2012 17:47:24 +0000 (17:47 +0000)]
e1000: add dropped DMA receive enable back in for WoL

Commit d5bc77a223b0e9b9dfb002048d2b34a79e7d0b48 broke Wake-on-LAN by
inadvertently dropping the enabling of DMA receives.

Restore the enabling of DMA receives for WoL.

This is applicable to 3.1+ stable trees.

CC: stable@vger.stable.org
Reported-by: Tobias Klausmann <klausman@schwarzvogel.de>
Signed-off-by: Dean Nelson <dnelson@redhat.com>
Tested-by: Tobias Klausmann <klausman@schwarzvogel.de>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agohwmon: (f75375s) Fix bit shifting in f75375_write16
Nikolaus Schulz [Wed, 8 Feb 2012 17:56:10 +0000 (18:56 +0100)]
hwmon: (f75375s) Fix bit shifting in f75375_write16

In order to extract the high byte of the 16-bit word, shift the word to
the right, not to the left.

Signed-off-by: Nikolaus Schulz <mail@microschulz.de>
Cc: stable@kernel.org # 2.6.32+
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
12 years agohwmon: (f75375s) Fix reading of wrong register when initializing the F75387
Nikolaus Schulz [Wed, 8 Feb 2012 17:56:09 +0000 (18:56 +0100)]
hwmon: (f75375s) Fix reading of wrong register when initializing the F75387

Unlike the other chips supported by this driver, the F75387 stores the
pwm_mode in register F75375_REG_FAN_TIMER, not F75375_REG_CONFIG1.

Signed-off-by: Nikolaus Schulz <mail@microschulz.de>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
12 years agohwmon: (f75375s) Fix automatic pwm mode setting for F75373 & F75375
Nikolaus Schulz [Wed, 8 Feb 2012 17:56:08 +0000 (18:56 +0100)]
hwmon: (f75375s) Fix automatic pwm mode setting for F75373 & F75375

In order to enable temperature mode aka automatic mode for the F75373 and
F75375 chips, the two FANx_MODE bits in the fan configuration register
need be set to 01, not 10.

Signed-off-by: Nikolaus Schulz <mail@microschulz.de>
Cc: stable@kernel.org # 2.6.32+
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
12 years agohwmon: (w83627ehf) Remove duplicate code
Guenter Roeck [Tue, 7 Feb 2012 09:46:42 +0000 (01:46 -0800)]
hwmon: (w83627ehf) Remove duplicate code

Commit ec3e5a16446fef1891611fe3bdfa5954d1ddf5e4 slipped in some duplicate code.
Remove it.

Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
Acked-by: Jean Delvare <khali@linux-fr.org>
12 years agoLinux 3.3-rc3 v3.3-rc3
Linus Torvalds [Thu, 9 Feb 2012 03:21:53 +0000 (19:21 -0800)]
Linux 3.3-rc3

12 years agoMerge branch 'iommu/fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/joro...
Linus Torvalds [Thu, 9 Feb 2012 03:11:00 +0000 (19:11 -0800)]
Merge branch 'iommu/fixes' of git://git./linux/kernel/git/joro/iommu

One patch fixes an bug in the ARM/MSM IOMMU code which returned sucess
in the unmap function even when an error occured and the other patch
adds a workaround into the AMD IOMMU driver to better handle broken IVRS
ACPI tables (this patch fixes the case when a device is not listed in
the table but actually translated by the iommu).

* 'iommu/fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
  iommu/msm: Fix error handling in msm_iommu_unmap()
  iommu/amd: Work around broken IVRS tables

12 years agoMerge branch '3.3-rc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/nab...
Linus Torvalds [Thu, 9 Feb 2012 03:09:25 +0000 (19:09 -0800)]
Merge branch '3.3-rc-fixes' of git://git./linux/kernel/git/nab/target-pending

This series contains pending target bug-fixes and cleanups for v3.3-rc3
that have been addressed the past weeks in lio-core.git.

Some of the highlights include:

 - Fix handling for control CDBs with data greater than PAGE_SIZE (andy)
 - Use IP_FREEBIND for iscsi-target to address network portal creation
   issues with systemd (dax)
 - Allow PERSISTENT RESERVE IN for non-reservation holder (marco)
 - Fix iblock se_dev_attrib.unmap_granularity (marco)
 - Fix unsupported WRITE_SAME sense payload handling (martin)
 - Add workaround for zero-length control CDB handling (nab)
 - Fix discovery with INADDR_ANY and IN6ADDR_ANY_INIT (nab)
 - Fix target_submit_cmd() exception handling (nab)
 - Return correct ASC for unimplemented VPD pages (roland)
 - Don't zero pages used for data buffers (roland)
 - Fix return code of core_tpg_.*_lun (sebastian)

* '3.3-rc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending: (26 commits)
  target: Fix unsupported WRITE_SAME sense payload
  iscsi: use IP_FREEBIND socket option
  iblock: fix handling of large requests
  target: handle empty string writes in sysfs
  iscsi_target: in_aton needs linux/inet.h
  target: Fix iblock se_dev_attrib.unmap_granularity
  target: Fix target_submit_cmd() exception handling
  target: Change target_submit_cmd() to return void
  target: accept REQUEST_SENSE with 18bytes
  target: Fail INQUIRY commands with EVPD==0 but PAGE CODE!=0
  target: Return correct ASC for unimplemented VPD pages
  iscsi-target: Fix discovery with INADDR_ANY and IN6ADDR_ANY_INIT
  target: Allow control CDBs with data > 1 page
  iscsi-target: Fix up a few assignments
  iscsi-target: make one-bit bitfields unsigned
  iscsi-target: Fix double list_add with iscsit_alloc_buffs reject
  iscsi-target: Fix reject release handling in iscsit_free_cmd()
  target: fix return code of core_tpg_.*_lun
  target: use save/restore lock primitive in core_dec_lacl_count()
  target: avoid multiple outputs in scsi_dump_inquiry()
  ...

12 years agoMerge tag 'md-3.3-fixes' of git://neil.brown.name/md
Linus Torvalds [Thu, 9 Feb 2012 03:06:30 +0000 (19:06 -0800)]
Merge tag 'md-3.3-fixes' of git://neil.brown.name/md

Some simple md-related fixes.

1/ two small fixes to ensure we handle an interrupted resync properly.
2/ avoid loading the bitmap multiple times in dm-raid

* tag 'md-3.3-fixes' of git://neil.brown.name/md:
  md: two small fixes to handling interrupt resync.
  Prevent DM RAID from loading bitmap twice.

12 years agoMerge tag 'spi-for-linus' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Thu, 9 Feb 2012 03:05:47 +0000 (19:05 -0800)]
Merge tag 'spi-for-linus' of git://git.secretlab.ca/git/linux-2.6

SPI bug fixes for v3.3-rc2

Minor SPI device driver changes.  A rename of the pch_spi_pcidev symbol
that merely eliminates a modpost warning, and a Kconfig change to allow
the Samsung spi driver to build on EXYNOS.

* tag 'spi-for-linus' of git://git.secretlab.ca/git/linux-2.6:
  spi-topcliff-pch: rename pch_spi_pcidev to pch_spi_pcidev_driver
  spi: Add spi-s3c64xx driver dependency on ARCH_EXYNOS4

12 years agoMerge branch 'akpm' (Andrew's tree)
Linus Torvalds [Thu, 9 Feb 2012 03:04:47 +0000 (19:04 -0800)]
Merge branch 'akpm' (Andrew's tree)

Five fixes

* branch 'akpm':
  pcmcia: fix socket refcount decrementing on each resume
  mm: fix UP THP spin_is_locked BUGs
  drivers/leds/leds-lm3530.c: fix setting pltfm->als_vmax
  mm: compaction: check for overlapping nodes during isolation for migration
  nilfs2: avoid overflowing segment numbers in nilfs_ioctl_clean_segments()

12 years agopcmcia: fix socket refcount decrementing on each resume
Russell King [Thu, 9 Feb 2012 01:13:41 +0000 (17:13 -0800)]
pcmcia: fix socket refcount decrementing on each resume

This fixes a memory-corrupting bug: not only does it cause the warning,
but as a result of dropping the refcount to zero, it causes the
pcmcia_socket0 device structure to be freed while it still has
references, causing slab caches corruption.  A fatal oops quickly
follows this warning - often even just a 'dmesg' following the warning
causes the kernel to oops.

While testing suspend/resume on an ARM device with PCMCIA support, and a
CF card inserted, I found that after five suspend and resumes, the
kernel would complain, and shortly die after with slab corruption.

  WARNING: at include/linux/kref.h:41 kobject_get+0x28/0x50()

As the message doesn't give a clue about which kobject, and the built-in
debugging in drivers/base/power/main.c happens too late, this was added
right before each get_device():

  printk("%s: %p [%s] %u\n", __func__, dev, kobject_name(&dev->kobj), atomic_read(&dev->kobj.kref.refcount));

and on the 3rd s2ram cycle, the following behaviour observed:

On the 3rd suspend/resume cycle:

  dpm_prepare: c1a0d998 [pcmcia_socket0] 3
  dpm_suspend: c1a0d998 [pcmcia_socket0] 3
  dpm_suspend_noirq: c1a0d998 [pcmcia_socket0] 3
  dpm_resume_noirq: c1a0d998 [pcmcia_socket0] 3
  dpm_resume: c1a0d998 [pcmcia_socket0] 3
  dpm_complete: c1a0d998 [pcmcia_socket0] 2

4th:

  dpm_prepare: c1a0d998 [pcmcia_socket0] 2
  dpm_suspend: c1a0d998 [pcmcia_socket0] 2
  dpm_suspend_noirq: c1a0d998 [pcmcia_socket0] 2
  dpm_resume_noirq: c1a0d998 [pcmcia_socket0] 2
  dpm_resume: c1a0d998 [pcmcia_socket0] 2
  dpm_complete: c1a0d998 [pcmcia_socket0] 1

5th:

  dpm_prepare: c1a0d998 [pcmcia_socket0] 1
  dpm_suspend: c1a0d998 [pcmcia_socket0] 1
  dpm_suspend_noirq: c1a0d998 [pcmcia_socket0] 1
  dpm_resume_noirq: c1a0d998 [pcmcia_socket0] 1
  dpm_resume: c1a0d998 [pcmcia_socket0] 1
  dpm_complete: c1a0d998 [pcmcia_socket0] 0
  ------------[ cut here ]------------
  WARNING: at include/linux/kref.h:41 kobject_get+0x28/0x50()
  Modules linked in: ucb1x00_core
  Backtrace:
  [<c0212090>] (dump_backtrace+0x0/0x110) from [<c04799dc>] (dump_stack+0x18/0x1c)
  [<c04799c4>] (dump_stack+0x0/0x1c) from [<c021cba0>] (warn_slowpath_common+0x50/0x68)
  [<c021cb50>] (warn_slowpath_common+0x0/0x68) from [<c021cbdc>] (warn_slowpath_null+0x24/0x28)
  [<c021cbb8>] (warn_slowpath_null+0x0/0x28) from [<c0335374>] (kobject_get+0x28/0x50)
  [<c033534c>] (kobject_get+0x0/0x50) from [<c03804f4>] (get_device+0x1c/0x24)
  [<c0388c90>] (dpm_complete+0x0/0x1a0) from [<c0389cc0>] (dpm_resume_end+0x1c/0x20)
  ...

Looking at commit 7b24e7988263 ("pcmcia: split up central event handler"),
the following change was made to cs.c:

                return 0;
        }
 #endif
-
-       send_event(skt, CS_EVENT_PM_RESUME, CS_EVENT_PRI_LOW);
+       if (!(skt->state & SOCKET_CARDBUS) && (skt->callback))
+               skt->callback->early_resume(skt);
        return 0;
 }

And the corresponding change in ds.c is from:

-static int ds_event(struct pcmcia_socket *skt, event_t event, int priority)
-{
-       struct pcmcia_socket *s = pcmcia_get_socket(skt);
...
-       switch (event) {
...
-       case CS_EVENT_PM_RESUME:
-               if (verify_cis_cache(skt) != 0) {
-                       dev_dbg(&skt->dev, "cis mismatch - different card\n");
-                       /* first, remove the card */
-                       ds_event(skt, CS_EVENT_CARD_REMOVAL, CS_EVENT_PRI_HIGH);
-                       mutex_lock(&s->ops_mutex);
-                       destroy_cis_cache(skt);
-                       kfree(skt->fake_cis);
-                       skt->fake_cis = NULL;
-                       s->functions = 0;
-                       mutex_unlock(&s->ops_mutex);
-                       /* now, add the new card */
-                       ds_event(skt, CS_EVENT_CARD_INSERTION,
-                                CS_EVENT_PRI_LOW);
-               }
-               break;
...
-    }

-    pcmcia_put_socket(s);

-    return 0;
-} /* ds_event */

to:

+static int pcmcia_bus_early_resume(struct pcmcia_socket *skt)
+{
+       if (!verify_cis_cache(skt)) {
+               pcmcia_put_socket(skt);
+               return 0;
+       }

+       dev_dbg(&skt->dev, "cis mismatch - different card\n");

+       /* first, remove the card */
+       pcmcia_bus_remove(skt);
+       mutex_lock(&skt->ops_mutex);
+       destroy_cis_cache(skt);
+       kfree(skt->fake_cis);
+       skt->fake_cis = NULL;
+       skt->functions = 0;
+       mutex_unlock(&skt->ops_mutex);

+       /* now, add the new card */
+       pcmcia_bus_add(skt);
+       return 0;
+}

As can be seen, the original function called pcmcia_get_socket() and
pcmcia_put_socket() around the guts, whereas the replacement code
calls pcmcia_put_socket() only in one path.  This creates an imbalance
in the refcounting.

Testing with pcmcia_put_socket() put removed shows that the bug is gone:

  dpm_suspend: c1a10998 [pcmcia_socket0] 5
  dpm_suspend_noirq: c1a10998 [pcmcia_socket0] 5
  dpm_resume_noirq: c1a10998 [pcmcia_socket0] 5
  dpm_resume: c1a10998 [pcmcia_socket0] 5
  dpm_complete: c1a10998 [pcmcia_socket0] 5

Tested-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agomm: fix UP THP spin_is_locked BUGs
Hugh Dickins [Thu, 9 Feb 2012 01:13:40 +0000 (17:13 -0800)]
mm: fix UP THP spin_is_locked BUGs

Fix CONFIG_TRANSPARENT_HUGEPAGE=y CONFIG_SMP=n CONFIG_DEBUG_VM=y
CONFIG_DEBUG_SPINLOCK=n kernel: spin_is_locked() is then always false,
and so triggers some BUGs in Transparent HugePage codepaths.

asm-generic/bug.h mentions this problem, and provides a WARN_ON_SMP(x);
but being too lazy to add VM_BUG_ON_SMP, BUG_ON_SMP, WARN_ON_SMP_ONCE,
VM_WARN_ON_SMP_ONCE, just test NR_CPUS != 1 in the existing VM_BUG_ONs.

Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agodrivers/leds/leds-lm3530.c: fix setting pltfm->als_vmax
Axel Lin [Thu, 9 Feb 2012 01:13:39 +0000 (17:13 -0800)]
drivers/leds/leds-lm3530.c: fix setting pltfm->als_vmax

In current code, pltfm->als_vmin is set to LM3530_ALS_WINDOW_mV and
pltfm->als_vmax is 0.  This does not make sense.  I think what we want
here is setting pltfm->als_vmax to LM3530_ALS_WINDOW_mV.

Both als_vmin and als_vmax local variables will be set to
pltfm->als_vmin and pltfm->als_vmax by a few lines latter.  Thus also
remove a redundant assignment for als_vmin and als_vmax in this patch.

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Cc: Shreshtha Kumar Sahu <shreshthakumar.sahu@stericsson.com>
Acked-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Tested-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agomm: compaction: check for overlapping nodes during isolation for migration
Mel Gorman [Thu, 9 Feb 2012 01:13:38 +0000 (17:13 -0800)]
mm: compaction: check for overlapping nodes during isolation for migration

When isolating pages for migration, migration starts at the start of a
zone while the free scanner starts at the end of the zone.  Migration
avoids entering a new zone by never going beyond the free scanned.

Unfortunately, in very rare cases nodes can overlap.  When this happens,
migration isolates pages without the LRU lock held, corrupting lists
which will trigger errors in reclaim or during page free such as in the
following oops

  BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
  IP: [<ffffffff810f795c>] free_pcppages_bulk+0xcc/0x450
  PGD 1dda554067 PUD 1e1cb58067 PMD 0
  Oops: 0000 [#1] SMP
  CPU 37
  Pid: 17088, comm: memcg_process_s Tainted: G            X
  RIP: free_pcppages_bulk+0xcc/0x450
  Process memcg_process_s (pid: 17088, threadinfo ffff881c2926e000, task ffff881c2926c0c0)
  Call Trace:
    free_hot_cold_page+0x17e/0x1f0
    __pagevec_free+0x90/0xb0
    release_pages+0x22a/0x260
    pagevec_lru_move_fn+0xf3/0x110
    putback_lru_page+0x66/0xe0
    unmap_and_move+0x156/0x180
    migrate_pages+0x9e/0x1b0
    compact_zone+0x1f3/0x2f0
    compact_zone_order+0xa2/0xe0
    try_to_compact_pages+0xdf/0x110
    __alloc_pages_direct_compact+0xee/0x1c0
    __alloc_pages_slowpath+0x370/0x830
    __alloc_pages_nodemask+0x1b1/0x1c0
    alloc_pages_vma+0x9b/0x160
    do_huge_pmd_anonymous_page+0x160/0x270
    do_page_fault+0x207/0x4c0
    page_fault+0x25/0x30

The "X" in the taint flag means that external modules were loaded but but
is unrelated to the bug triggering.  The real problem was because the PFN
layout looks like this

  Zone PFN ranges:
    DMA      0x00000010 -> 0x00001000
    DMA32    0x00001000 -> 0x00100000
    Normal   0x00100000 -> 0x01e80000
  Movable zone start PFN for each node
  early_node_map[14] active PFN ranges
      0: 0x00000010 -> 0x0000009b
      0: 0x00000100 -> 0x0007a1ec
      0: 0x0007a354 -> 0x0007a379
      0: 0x0007f7ff -> 0x0007f800
      0: 0x00100000 -> 0x00680000
      1: 0x00680000 -> 0x00e80000
      0: 0x00e80000 -> 0x01080000
      1: 0x01080000 -> 0x01280000
      0: 0x01280000 -> 0x01480000
      1: 0x01480000 -> 0x01680000
      0: 0x01680000 -> 0x01880000
      1: 0x01880000 -> 0x01a80000
      0: 0x01a80000 -> 0x01c80000
      1: 0x01c80000 -> 0x01e80000

The fix is straight-forward.  isolate_migratepages() has to make a
similar check to isolate_freepage to ensure that it never isolates pages
from a zone it does not hold the LRU lock for.

This was discovered in a 3.0-based kernel but it affects 3.1.x, 3.2.x
and current mainline.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Michal Nazarewicz <mina86@mina86.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agonilfs2: avoid overflowing segment numbers in nilfs_ioctl_clean_segments()
Xi Wang [Thu, 9 Feb 2012 01:13:37 +0000 (17:13 -0800)]
nilfs2: avoid overflowing segment numbers in nilfs_ioctl_clean_segments()

nsegs is read from userspace.  Limit its value and avoid overflowing nsegs
* sizeof(__u64) in the subsequent call to memdup_user().

This patch complements 481fe17e973fb9 ("nilfs2: potential integer overflow
in nilfs_ioctl_clean_segments()").

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Cc: Haogang Chen <haogangchen@gmail.com>
Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agostaging: pohmelfs: remove drivers/staging/pohmelfs
Evgeniy Polyakov [Wed, 8 Feb 2012 23:44:50 +0000 (02:44 +0300)]
staging: pohmelfs: remove drivers/staging/pohmelfs

New pohmelfs is coming, and it is time to remove deadly old design
https://lkml.org/lkml/2012/2/8/293

Signed-off-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoMerge branch 'for-davem' of git://git.kernel.org/pub/scm/linux/kernel/git/linville...
David S. Miller [Thu, 9 Feb 2012 00:12:12 +0000 (19:12 -0500)]
Merge branch 'for-davem' of git://git./linux/kernel/git/linville/wireless

12 years agogro: more generic L2 header check
Eric Dumazet [Wed, 8 Feb 2012 08:51:50 +0000 (08:51 +0000)]
gro: more generic L2 header check

Shlomo Pongratz reported GRO L2 header check was suited for Ethernet
only, and failed on IB/ipoib traffic.

He provided a patch faking a zeroed header to let GRO aggregates frames.

Roland Dreier, Herbert Xu, and others suggested we change GRO L2 header
check to be more generic, ie not assuming L2 header is 14 bytes, but
taking into account hard_header_len.

__napi_gro_receive() has special handling for the common case (Ethernet)
to avoid a memcmp() call and use an inline optimized function instead.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reported-by: Shlomo Pongratz <shlomop@mellanox.com>
Cc: Roland Dreier <roland@kernel.org>
Cc: Or Gerlitz <ogerlitz@mellanox.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Tested-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoIPoIB: Stop lying about hard_header_len and use skb->cb to stash LL addresses
Roland Dreier [Tue, 7 Feb 2012 14:51:21 +0000 (14:51 +0000)]
IPoIB: Stop lying about hard_header_len and use skb->cb to stash LL addresses

Commit a0417fa3a18a ("net: Make qdisc_skb_cb upper size bound
explicit.") made it possible for a netdev driver to use skb->cb
between its header_ops.create method and its .ndo_start_xmit
method.  Use this in ipoib_hard_header() to stash away the LL address
(GID + QPN), instead of the "ipoib_pseudoheader" hack.  This allows
IPoIB to stop lying about its hard_header_len, which will let us fix
the L2 check for GRO.

Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agodriver-core: cpu: fix kobject warning when hotplugging a cpu
Greg Kroah-Hartman [Wed, 8 Feb 2012 23:11:17 +0000 (15:11 -0800)]
driver-core: cpu: fix kobject warning when hotplugging a cpu

Due to the sysdev conversion to struct device, the cpu objects get
reused when adding a cpu after offlining it, which causes a big warning
that the kobject portion is not properly initialized.

So clear out the object before we register it again, so all is quiet.

Reported-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Tested-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoACPI: remove duplicated lines of merging problems with acpi_processor_add
Yinghai Lu [Tue, 7 Feb 2012 16:20:37 +0000 (08:20 -0800)]
ACPI: remove duplicated lines of merging problems with acpi_processor_add

Those lines have two copies.

Not sure how it could happen, looks like git problem.

Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Renninger <trenn@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agom32r: relocate drivers back out of 8250 dir
Paul Gortmaker [Wed, 8 Feb 2012 14:05:42 +0000 (09:05 -0500)]
m32r: relocate drivers back out of 8250 dir

Commit 9bef3d4197379a995fa80f81950bbbf8d32e9e8b

"serial: group all the 8250 related code together"

inadvertently swept up the m32r driver in the move, because
it had comments mentioning 8250 registers within it.  However
these are only there by nature of the driver being based off
the 8250 source code -- the hardware itself does not actually
have any relation to the original 8250 style UARTs.

Reported-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoMerge tag 'sound-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Linus Torvalds [Wed, 8 Feb 2012 22:56:39 +0000 (14:56 -0800)]
Merge tag 'sound-fixes' of git://git./linux/kernel/git/tiwai/sound

sound fixes #2 for 3.3-rc3

A collection of small fixes, mostly for regressions.
In addition, a few ASoC wm8994 updates are included, too.

* tag 'sound-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ASoC: wm8994: Disable line output discharge prior to ramping VMID
  ASoC: wm8994: Fix typo in VMID ramp setting
  ALSA: oxygen, virtuoso: fix exchanged L/R volumes of aux and CD inputs
  ALSA: usb-audio: add Edirol UM-3G support
  ALSA: hda - add support for Uniwill ECS M31EI notebook
  ALSA: hda - Fix error handling in patch_ca0132.c
  ASoC: wm8994: Enabling VMID should take a runtime PM reference
  ALSA: hda/realtek - Fix a wrong condition
  ALSA: emu8000: Remove duplicate linux/moduleparam.h include from emu8000_patch.c
  ALSA: hda/realtek - Add missing Bass and CLFE as vmaster slaves
  ASoC: wm_hubs: Correct line input to line output 2 paths
  ASoC: cs42l73: Fix Output [X|A|V]SP_SCLK Sourcing Mode setting for master mode
  ASoC: wm8962: Fix word length configuration
  ASoC: core: Better support for idle_bias_off suspend ignores
  ASoC: wm8994: Remove ASoC level register cache sync
  ASoC: wm_hubs: Fix routing of input PGAs to line output mixer

12 years agostaging: android/ram_console: Don't build on arches w/o ioremap
Anton Vorontsov [Tue, 7 Feb 2012 05:13:27 +0000 (09:13 +0400)]
staging: android/ram_console: Don't build on arches w/o ioremap

This patch fixes UML build:

    CC      drivers/staging/android/ram_console.o
  drivers/staging/android/ram_console.c: In function
  'ram_console_driver_probe':
  drivers/staging/android/ram_console.c:358:2: error: implicit declaration
  of function 'ioremap' [-Werror=implicit-function-declaration]
  cc1: some warnings being treated as errors
  make[3]: *** [drivers/staging/android/ram_console.o] Error 1

Signed-off-by: Anton Vorontsov <anton.vorontsov@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: r8712u: Use asynchronous firmware loading
Larry Finger [Mon, 6 Feb 2012 03:12:26 +0000 (21:12 -0600)]
staging: r8712u: Use asynchronous firmware loading

In https://bugs.archlinux.org/task/27996, failure of driver r8712u is
reported, with a timeout during module loading due to synchronous loading
of the firmware. The code now uses request_firmware_nowait().

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: usbip: fix to prevent potentially using uninitialized spinlock
Bart Westgeest [Wed, 25 Jan 2012 18:46:32 +0000 (13:46 -0500)]
staging: usbip: fix to prevent potentially using uninitialized spinlock

The stub_probe function can be called as soon as the stub_driver is
registered. This can lead to the busid_table_lock being used before it
is initialized. Moved calling the init_busid_table function (which
initalizes this spinlock) to be called earlier in the init function to
prevent this from happening.

Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: r8712u: Fix problem when CONFIG_R8712_AP is set
Larry Finger [Fri, 3 Feb 2012 22:05:13 +0000 (16:05 -0600)]
staging: r8712u: Fix problem when CONFIG_R8712_AP is set

When this driver was upgraded to the vendor 20100831 version in
commit 93c55dda092c7 et al,, one listhead initialization was missed.
This broke complete operation of the driver whenever AP mode was
enabled.

This patch fixes https://bugs.archlinux.org/task/27996.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.stable.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: tidspbridge: fix incorrect free to drv_datap
Omar Ramirez Luna [Tue, 31 Jan 2012 01:20:18 +0000 (19:20 -0600)]
staging: tidspbridge: fix incorrect free to drv_datap

This structure is still used after it has been freed, since it
is being allocated in probe, calls to free it have been moved to
module's remove routine.

This should fix the follwoing messages when attempting to remove the
module:
 drv_get_first_dev_extension: Failed to retrieve the object handle
 drv_get_first_dev_extension: Failed to retrieve the object handle
 drv_destroy: Failed to store DRV object
 mgr_destroy: Failed to store MGR object

Signed-off-by: Omar Ramirez Luna <omar.ramirez@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: tidspbridge: fix bridge_open memory leaks
Omar Ramirez Luna [Tue, 31 Jan 2012 01:20:17 +0000 (19:20 -0600)]
staging: tidspbridge: fix bridge_open memory leaks

There are two members of pr_ctxt allocated during bridge_open that
are never freed resulting in memory leaks, these are stream_id and
node_id, they are now freed on release of the handle (bridge_release)
right before freeing pr_ctxt.

Error path for bridge_open was also fixed since the same variables
could result in memory leaking due to missing handling of failure
scenarios. While at it, the indentation changes were introduced to
avoid interleaved goto statements inside big if blocks.

Signed-off-by: Omar Ramirez Luna <omar.ramirez@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: android: lowmemorykiller: Don't wait more than one second for a process...
Arve Hjønnevåg [Fri, 13 Jan 2012 18:21:25 +0000 (22:21 +0400)]
staging: android: lowmemorykiller: Don't wait more than one second for a process to die

If a process forked and the child process was killed by the
lowmemorykiller, the lowmemory killer would be disabled until
the parent process reaped the child or it died itself.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
Signed-off-by: Anton Vorontsov <anton.vorontsov@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoMAINTAINERS: staging: iio: add iio information
Ameya Palande [Tue, 10 Jan 2012 17:00:58 +0000 (09:00 -0800)]
MAINTAINERS: staging: iio: add iio information

Signed-off-by: Ameya Palande <ameya.palande@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: zcache: fix serialization bug in zv stats
Seth Jennings [Fri, 30 Dec 2011 16:42:15 +0000 (10:42 -0600)]
staging: zcache: fix serialization bug in zv stats

In a multithreaded workload, the zv_curr_dist_counts
and zv_cumul_dist_counts statistics are being corrupted
because the increments and decrements in zv_create
and zv_free are not atomic.

This patch converts these statistics and their corresponding
increments/decrements/reads to atomic operations.

Signed-off-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Acked-by: Dan Magenheimer <dan.magenheimer@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agostaging: fix go7007-usb license
Randy Dunlap [Thu, 2 Feb 2012 03:04:02 +0000 (19:04 -0800)]
staging: fix go7007-usb license

Add MODULE_LICENSE() for this source module to fix build warning:

WARNING: modpost: missing MODULE_LICENSE() in drivers/staging/media/go7007/go7007-usb.o
see include/linux/module.h for more information

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Cc: Ross Cohen <rcohen@snurgle.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoStaging: android: binder: Fix crashes when sharing a binder file between processes
Arve Hjønnevåg [Wed, 1 Feb 2012 23:29:13 +0000 (15:29 -0800)]
Staging: android: binder: Fix crashes when sharing a binder file between processes

Opening the binder driver and sharing the file returned with
other processes (e.g. by calling fork) can crash the kernel.
Prevent these crashes with the following changes:
- Add a mutex to protect against two processes mmapping the
  same binder_proc.
- After locking mmap_sem, check that the vma we want to access
  (still) points to the same mm_struct.
- Use proc->tsk instead of current to get the files struct since
  this is where we get the rlimit from.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
12 years agoStaging: android: Remove pmem driver
Shuah Khan [Fri, 27 Jan 2012 18:40:10 +0000 (11:40 -0700)]
Staging: android: Remove pmem driver

Addroid pmem driver is no longer used in any of the Android products.
This patch removes pmem driver from Android staging area

Reference: https://lkml.org/lkml/2012/1/23/183

Signed-off-by: Shuah Khan <shuahkhan@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoStaging: asus_oled: fix NULL-ptr crash on unloading
Pekka Paalanen [Sun, 22 Jan 2012 14:33:47 +0000 (16:33 +0200)]
Staging: asus_oled: fix NULL-ptr crash on unloading

Asus_oled triggers the following bug on module unloading:

 usbcore: deregistering interface driver asus-oled
 BUG: unable to handle kernel NULL pointer dereference at 0000000000000038
 IP: [<ffffffff8111292b>] sysfs_delete_link+0x30/0x66

 Call Trace:
  [<ffffffff81225373>] device_remove_class_symlinks+0x6b/0x70
  [<ffffffff812256a8>] device_del+0x9f/0x1ab
  [<ffffffff812257c5>] device_unregister+0x11/0x1e
  [<ffffffffa000cb82>] asus_oled_disconnect+0x4f/0x9e [asus_oled]
  [<ffffffff81277430>] usb_unbind_interface+0x54/0x103
  [<ffffffff812276c4>] __device_release_driver+0xa2/0xeb
  [<ffffffff81227794>] driver_detach+0x87/0xad
  [<ffffffff812269e9>] bus_remove_driver+0x91/0xc1
  [<ffffffff81227fb4>] driver_unregister+0x66/0x6e
  [<ffffffff812771ed>] usb_deregister+0xbb/0xc4
  [<ffffffffa000ce87>] asus_oled_exit+0x2f/0x31 [asus_oled]
  [<ffffffff81068365>] sys_delete_module+0x1b8/0x21b
  [<ffffffff810ae3de>] ? do_munmap+0x2ef/0x313
  [<ffffffff813699bb>] system_call_fastpath+0x16/0x1b

This is due to an incorrect destruction sequence in asus_oled_exit().

Fix the order, fixes the bug. Tested on an Asus G50V laptop only.

Cc: Jakub Schmidtke <sjakub@gmail.com>
Signed-off-by: Pekka Paalanen <pq@iki.fi>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoStaging: asus_oled: fix image processing
Pekka Paalanen [Sun, 22 Jan 2012 14:33:46 +0000 (16:33 +0200)]
Staging: asus_oled: fix image processing

Programming an image was broken, because odev->buf_offs was not advanced
for val == 0 in append_values(). This regression was introduced in:

 commit 1ff12a4aa354bed093a0240d5e6347b1e27601bc
 Author: Kevin A. Granade <kevin.granade@gmail.com>
 Date:   Sat Sep 5 01:03:39 2009 -0500

     Staging: asus_oled: Cleaned up checkpatch issues.

Fix the image processing by special-casing val == 0.

I have tested this change on an Asus G50V laptop only.

Cc: Jakub Schmidtke <sjakub@gmail.com>
Cc: Kevin A. Granade <kevin.granade@gmail.com>
Signed-off-by: Pekka Paalanen <pq@iki.fi>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoStaging: android: binder: Don't call dump_stack in binder_vma_open
Arve Hjønnevåg [Sat, 21 Jan 2012 03:56:21 +0000 (19:56 -0800)]
Staging: android: binder: Don't call dump_stack in binder_vma_open

If user-space partially unmaps the driver, binder_vma_open
would dump the kernel stack. This is not a kernel bug however
and will be treated as if the whole area was unmapped once
binder_vma_close gets called.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: r8712u: Add new Sitecom UsB ID
Larry Finger [Sat, 7 Jan 2012 16:07:03 +0000 (10:07 -0600)]
staging: r8712u: Add new Sitecom UsB ID

Add USB ID for SITECOM WLA-1000 V1 001 WLAN

Reported-and-tested-by: Roland Gruber <post@rolandgruber.de>
Reported-and-tested-by: Dario Lucia <dario.lucia@gmail.com>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agozcache: Set SWIZ_BITS to 8 to reduce tmem bucket lock contention.
Dan Magenheimer [Mon, 23 Jan 2012 21:52:20 +0000 (16:52 -0500)]
zcache: Set SWIZ_BITS to 8 to reduce tmem bucket lock contention.

SWIZ_BITS > 8 results in a much larger number of "tmem_obj"
allocations, likely one per page-placed-in-frontswap.  The
tmem_obj is not huge (roughly 100 bytes), but it is large
enough to add a not-insignificant memory overhead to zcache.

The SWIZ_BITS=8  will get roughly the same lock contention
without the space wastage.

The effect of SWIZ_BITS can be thought of as "2^SWIZ_BITS is
the number of unique oids that be generated" (This concept is
limited to frontswap's use of tmem).

Acked-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agozcache: fix deadlock condition
Dan Magenheimer [Wed, 25 Jan 2012 22:32:51 +0000 (14:32 -0800)]
zcache: fix deadlock condition

I discovered this deadlock condition awhile ago working on RAMster
but it affects zcache as well.  The list spinlock must be
locked prior to the page spinlock and released after.  As
a result, the page copy must also be done while the locks are held.

Applies to 3.2.  Konrad, please push (via GregKH?)...
this is definitely a bug fix so need not be pushed during
a -rc0 window.

Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com>
Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: fix locking issue
Rob Clark [Thu, 19 Jan 2012 00:33:02 +0000 (18:33 -0600)]
staging: drm/omap: fix locking issue

The create/free mmap offset code must be synchronized.  Yet only some
callers of omap_gem_mmap_offset() held struct_mutex.  Leading to various
crashes around drm_mm_insert_helper_range().  (In the free-object path,
which is currently the only place we drm_gem_free_mmap_offset(),
struct_mutex is already held.)

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agodrivers/staging/omapdrm/omap_fbdev.c: move free after uses
Julia Lawall [Mon, 9 Jan 2012 07:51:03 +0000 (08:51 +0100)]
drivers/staging/omapdrm/omap_fbdev.c: move free after uses

Move the free after the final uses.

The semantic patch that makes this report is available
in scripts/coccinelle/free/kfree.cocci.

More information about semantic patching is available at
http://coccinelle.lip6.fr/

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: fix minimum width/height
Rob Clark [Mon, 9 Jan 2012 01:37:37 +0000 (19:37 -0600)]
staging: drm/omap: fix minimum width/height

Fix minimum width/height so planes could be used to implement
hw mouse cursor.

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: updates for DSS fifomerge API changes
Rob Clark [Mon, 16 Jan 2012 18:51:18 +0000 (12:51 -0600)]
staging: drm/omap: updates for DSS fifomerge API changes

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: multiplanar and YUV support
Rob Clark [Mon, 16 Jan 2012 18:51:17 +0000 (12:51 -0600)]
staging: drm/omap: multiplanar and YUV support

Add support in framebuffer objects for other color formats and multi-
planar YUV (NV12).  Since this requires changing the API between the
plane and fb for getting scanout information (paddr, etc), take
advantage of the opportunity and put in place a way to allow fb's to
be unpinned when they are not being scanned out.  Now, before start
of scanout the plane calls omap_framebuffer_pin() which takes care
to pin all the backing bo's, then omap_framebuffer_update_scanout()
however many times to update the scanout address(es), etc, and then
when finished omap_framebuffer_unpin().

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: add drm_plane support
Rob Clark [Mon, 16 Jan 2012 18:51:16 +0000 (12:51 -0600)]
staging: drm/omap: add drm_plane support

Because framebuffer layer and overlay scanout video pipes are basically
thing in OMAP display subsystem (the only difference being that the first
video pipe does not support scaling or YUV formats), much of the CRTC
code is pulled into the plane implementation, and a private plane object
is used by the CRTC object.  This avoids code duplication between the
plane and CRTC.

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: drm API update: addfb2
Rob Clark [Mon, 16 Jan 2012 18:51:15 +0000 (12:51 -0600)]
staging: drm/omap: drm API update: addfb2

Update to reflect changes in:
"drm: add an fb creation ioctl that takes a pixel format v5"

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: drm/omap: drm API update: make fops struct const
Rob Clark [Mon, 16 Jan 2012 18:51:14 +0000 (12:51 -0600)]
staging: drm/omap: drm API update: make fops struct const

Update to reflect changes in:
"Make the per-driver file_operations struct const"

Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: delete gma500 driver
Greg Kroah-Hartman [Mon, 30 Jan 2012 22:12:59 +0000 (14:12 -0800)]
staging: delete gma500 driver

It's now "properly" merged into the DRM tree in the kernel, so delete
the staging version of the driver as it is far obsolete and broken.

Requested-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agodrm/i915: fixup interlaced bits clearing in PIPECONF on PCH_SPLIT (v2)
Keith Packard [Wed, 8 Feb 2012 21:53:38 +0000 (13:53 -0800)]
drm/i915: fixup interlaced bits clearing in PIPECONF on PCH_SPLIT (v2)

An identical patch has been merged for i9xx_crtc_mode_set:

Commit 59df7b1771c150163e522f33c638096ab0efbf42
Author: Christian Schmidt <schmidt@digadd.de>
Date:   Mon Dec 19 20:03:33 2011 +0100

    drm/intel: Fix initialization if startup happens in interlaced mode [v2]

But that one neglected to fix up the ironlake+ path.

This should fix the issue reported by Alfonso Fiore where booting with
only a HDMI cable connected to his TV failed to display anything. The
issue is that the bios set up things for 1080i and used the pannel
fitter to scale up the lower progressive resolutions. We failed to
clear the interlace bit in the PIPEACONF register, resulting in havoc.

v2: Be more paranoid and just unconditionally clear the field before
setting new values.

Cc: Peter Ross <pross@xvid.org>
Cc: Alfonso Fiore <alfonso.fiore@gmail.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agoloop: zero fill bio instead of return -EIO for partial read
Dave Young [Wed, 8 Feb 2012 21:07:19 +0000 (22:07 +0100)]
loop: zero fill bio instead of return -EIO for partial read

commit 8268f5a741 ("deny partial write for loop dev fd") tried to fix the
loop device partial read information leak problem.  But it changed the
semantics of read behavior.  When we read beyond the end of the device we
should get 0 bytes, which is normal behavior, we should not just return
-EIO

Instead of returning -EIO, zero out the bio to avoid information leak in
case of partail read.

Signed-off-by: Dave Young <dyoung@redhat.com>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Tested-by: Jeff Moyer <jmoyer@redhat.com>
Cc: Dmitry Monakhov <dmonakhov@sw.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agobio: don't overflow in bio_get_nr_vecs()
Kent Overstreet [Wed, 8 Feb 2012 21:07:18 +0000 (22:07 +0100)]
bio: don't overflow in bio_get_nr_vecs()

There were two places bio_get_nr_vecs() could overflow:

First, it did a left shift to convert from sectors to bytes immediately
before dividing by PAGE_SIZE.  If PAGE_SIZE ever was less than 512 a great
many things would break, so dividing by PAGE_SIZE >> 9 is safe and will
generate smaller code too.

The nastier overflow was in the DIV_ROUND_UP() (that's what the code was
effectively doing, anyways).  If n + d overflowed, the whole thing would
return 0 which breaks things rather effectively.

bio_get_nr_vecs() doesn't claim to give an exact value anyways, so the
DIV_ROUND_UP() is silly; we could do a straight divide except if a
device's queue_max_sectors was less than PAGE_SIZE we'd return 0.  So we
just add 1; this should always be safe - things will break badly if
bio_get_nr_vecs() returns > BIO_MAX_PAGES (bio_alloc() will suddenly start
failing) but it's queue_max_segments that must guard against this, if
queue_max_sectors is preventing this from happen things are going to
explode on architectures with different PAGE_SIZE.

Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Tejun Heo <tj@kernel.org>
Acked-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agoMerge tag 'asoc-3.3' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound...
Takashi Iwai [Wed, 8 Feb 2012 20:29:38 +0000 (21:29 +0100)]
Merge tag 'asoc-3.3' of git://git./linux/kernel/git/broonie/sound into for-linus

A few small WM8994 updates to go on top of the previous lot of things
that were sent.  They collide with some -next work so I'd really like to
get them into 3.3-rc3 if possible to merge back up into the -next code.
All driver specific and unexciting in the grand scheme of things.

12 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wirel...
John W. Linville [Wed, 8 Feb 2012 19:53:43 +0000 (14:53 -0500)]
Merge branch 'master' of git://git./linux/kernel/git/linville/wireless into for-davem

12 years agoASoC: wm8994: Disable line output discharge prior to ramping VMID
Mark Brown [Tue, 7 Feb 2012 14:18:29 +0000 (14:18 +0000)]
ASoC: wm8994: Disable line output discharge prior to ramping VMID

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
12 years agoASoC: wm8994: Fix typo in VMID ramp setting
Mark Brown [Tue, 7 Feb 2012 17:24:19 +0000 (17:24 +0000)]
ASoC: wm8994: Fix typo in VMID ramp setting

The VMID ramp rate is supposed to be 0x3, not 11b. Fix that.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Cc: stable@vger.kernel.org
12 years agofloppy: Fix a crash during rmmod
Vivek Goyal [Wed, 8 Feb 2012 19:03:39 +0000 (20:03 +0100)]
floppy: Fix a crash during rmmod

floppy driver does not call add_disk() on all the drives hence we don't take
gendisk reference on request queue for these drives. Don't call put_disk()
with disk->queue set, otherwise we try to put the reference we never took.

Reported-and-tested-by: Dirk Gouders <gouders@et.bocholt.fh-gelsenkirchen.de>
Signed-off-by: Vivek Goyal<vgoyal@redhat.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agofloppy: Cleanup disk->queue before caling put_disk() if add_disk() was never called
Vivek Goyal [Wed, 8 Feb 2012 19:03:38 +0000 (20:03 +0100)]
floppy: Cleanup disk->queue before caling put_disk() if add_disk() was never called

add_disk() takes gendisk reference on request queue. If driver failed during
initialization and never called add_disk() then that extra reference is not
taken. That reference is put in put_disk(). floppy driver allocates the
disk, allocates queue, sets disk->queue and then relizes that floppy
controller is not present. It tries to tear down everything and tries to
put a reference down in put_disk() which was never taken.

In such error cases cleanup disk->queue before calling put_disk() so that
we never try to put down a reference which was never taken in first place.

Reported-and-tested-by: Suresh Jayaraman <sjayaraman@suse.com>
Tested-by: Dirk Gouders <gouders@et.bocholt.fh-gelsenkirchen.de>
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agocdrom: move shared static to cdrom_device_info
Paolo Bonzini [Wed, 8 Feb 2012 19:03:14 +0000 (20:03 +0100)]
cdrom: move shared static to cdrom_device_info

The keeplocked variable in the cdrom driver is shared across multiple
drives, but set in per-device ioctls.  Move it to the per-device struct,
avoiding that the setting on one drive affects the driver's behavior
when closing another.

[ Impact: limit udev's confusion to one drive when a CD burning program
  unlocks the CD door at the end of burning. ]

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agobsg: fix sysfs link remove warning
Stanislaw Gruszka [Wed, 8 Feb 2012 19:02:03 +0000 (20:02 +0100)]
bsg: fix sysfs link remove warning

We create "bsg" link if q->kobj.sd is not NULL, so remove it only
when the same condition is true.

Fixes:

WARNING: at fs/sysfs/inode.c:323 sysfs_hash_and_remove+0x2b/0x77()
sysfs: can not remove 'bsg', no directory
Call Trace:
  [<c0429683>] warn_slowpath_common+0x6a/0x7f
  [<c0537a68>] ? sysfs_hash_and_remove+0x2b/0x77
  [<c042970b>] warn_slowpath_fmt+0x2b/0x2f
  [<c0537a68>] sysfs_hash_and_remove+0x2b/0x77
  [<c053969a>] sysfs_remove_link+0x20/0x23
  [<c05d88f1>] bsg_unregister_queue+0x40/0x6d
  [<c0692263>] __scsi_remove_device+0x31/0x9d
  [<c069149f>] scsi_forget_host+0x41/0x52
  [<c0689fa9>] scsi_remove_host+0x71/0xe0
  [<f7de5945>] quiesce_and_remove_host+0x51/0x83 [usb_storage]
  [<f7de5a1e>] usb_stor_disconnect+0x18/0x22 [usb_storage]
  [<c06c29de>] usb_unbind_interface+0x4e/0x109
  [<c067a80f>] __device_release_driver+0x6b/0xa6
  [<c067a861>] device_release_driver+0x17/0x22
  [<c067a46a>] bus_remove_device+0xd6/0xe6
  [<c06785e2>] device_del+0xf2/0x137
  [<c06c101f>] usb_disable_device+0x94/0x1a0

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agodrm/i915: no lvds quirk for AOpen MP45
Daniel Vetter [Wed, 8 Feb 2012 15:42:52 +0000 (16:42 +0100)]
drm/i915: no lvds quirk for AOpen MP45

According to a bug report, it doesn't have one.

Cc: stable@kernel.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44263
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agoALSA: oxygen, virtuoso: fix exchanged L/R volumes of aux and CD inputs
Clemens Ladisch [Sat, 4 Feb 2012 19:56:47 +0000 (20:56 +0100)]
ALSA: oxygen, virtuoso: fix exchanged L/R volumes of aux and CD inputs

The driver accidentally exchanged the left/right fields for stereo AC'97
mixer registers.  This affected only the aux and CD inputs because the
line input bypasses the AC'97 codec and the mic input is mono; cards
without AC'97 (Xonar DS/DG/HDAV Slim, HG2PCI, HiFier) were not affected.

Reported-and-tested-by: Abby Cedar <abbycedar@yahoo.com.au>
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: 2.6.31+ <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
12 years agoALSA: usb-audio: add Edirol UM-3G support
Clemens Ladisch [Sat, 4 Feb 2012 19:51:43 +0000 (20:51 +0100)]
ALSA: usb-audio: add Edirol UM-3G support

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
12 years agoblock: don't call elevator callbacks for plug merges
Tejun Heo [Wed, 8 Feb 2012 08:19:42 +0000 (09:19 +0100)]
block: don't call elevator callbacks for plug merges

Plug merge calls two elevator callbacks outside queue lock -
elevator_allow_merge_fn() and elevator_bio_merged_fn().  Although
attempt_plug_merge() suggests that elevator is guaranteed to be there
through the existing request on the plug list, nothing prevents plug
merge from calling into dying or initializing elevator.

For regular merges, bypass ensures elvpriv count to reach zero, which
in turn prevents merges as all !ELVPRIV requests get REQ_SOFTBARRIER
from forced back insertion.  Plug merge doesn't check ELVPRIV, and, as
the requests haven't gone through elevator insertion yet, it doesn't
have SOFTBARRIER set allowing merges on a bypassed queue.

This, for example, leads to the following crash during elevator
switch.

 BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
 IP: [<ffffffff813b34e9>] cfq_allow_merge+0x49/0xa0
 PGD 112cbc067 PUD 115d5c067 PMD 0
 Oops: 0000 [#1] PREEMPT SMP
 CPU 1
 Modules linked in: deadline_iosched

 Pid: 819, comm: dd Not tainted 3.3.0-rc2-work+ #76 Bochs Bochs
 RIP: 0010:[<ffffffff813b34e9>]  [<ffffffff813b34e9>] cfq_allow_merge+0x49/0xa0
 RSP: 0018:ffff8801143a38f8  EFLAGS: 00010297
 RAX: 0000000000000000 RBX: ffff88011817ce28 RCX: ffff880116eb6cc0
 RDX: 0000000000000000 RSI: ffff880118056e20 RDI: ffff8801199512f8
 RBP: ffff8801143a3908 R08: 0000000000000000 R09: 0000000000000000
 R10: 0000000000000001 R11: 0000000000000000 R12: ffff880118195708
 R13: ffff880118052aa0 R14: ffff8801143a3d50 R15: ffff880118195708
 FS:  00007f19f82cb700(0000) GS:ffff88011fc80000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
 CR2: 0000000000000008 CR3: 0000000112c6a000 CR4: 00000000000006e0
 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
 Process dd (pid: 819, threadinfo ffff8801143a2000, task ffff880116eb6cc0)
 Stack:
  ffff88011817ce28 ffff880118195708 ffff8801143a3928 ffffffff81391bba
  ffff88011817ce28 ffff880118195708 ffff8801143a3948 ffffffff81391bf1
  ffff88011817ce28 0000000000000000 ffff8801143a39a8 ffffffff81398e3e
 Call Trace:
  [<ffffffff81391bba>] elv_rq_merge_ok+0x4a/0x60
  [<ffffffff81391bf1>] elv_try_merge+0x21/0x40
  [<ffffffff81398e3e>] blk_queue_bio+0x8e/0x390
  [<ffffffff81396a5a>] generic_make_request+0xca/0x100
  [<ffffffff81396b04>] submit_bio+0x74/0x100
  [<ffffffff811d45c2>] __blockdev_direct_IO+0x1ce2/0x3450
  [<ffffffff811d0dc7>] blkdev_direct_IO+0x57/0x60
  [<ffffffff811460b5>] generic_file_aio_read+0x6d5/0x760
  [<ffffffff811986b2>] do_sync_read+0xe2/0x120
  [<ffffffff81199345>] vfs_read+0xc5/0x180
  [<ffffffff81199501>] sys_read+0x51/0x90
  [<ffffffff81aeac12>] system_call_fastpath+0x16/0x1b

There are multiple ways to fix this including making plug merge check
ELVPRIV; however,

* Calling into elevator outside queue lock is confusing and
  error-prone.

* Requests on plug list aren't known to the elevator.  They aren't on
  the elevator yet, so there's no elevator specific state to update.

* Given the nature of plug merges - collecting bio's for the same
  purpose from the same issuer - elevator specific restrictions aren't
  applicable.

So, simply don't call into elevator methods from plug merge by moving
elv_bio_merged() from bio_attempt_*_merge() to blk_queue_bio(), and
using blk_try_merge() in attempt_plug_merge().

This is based on Jens' patch to skip elevator_allow_merge_fn() from
plug merge.

Note that this makes per-cgroup merged stats skip plug merging.

Signed-off-by: Tejun Heo <tj@kernel.org>
LKML-Reference: <4F16F3CA.90904@kernel.dk>
Original-patch-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Jens Axboe <axboe@kernel.dk>