sfrench/cifs-2.6.git
10 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-next-2.6
Linus Torvalds [Tue, 16 Jun 2009 18:52:41 +0000 (11:52 -0700)]
Merge git://git./linux/kernel/git/davem/sparc-next-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-next-2.6: (30 commits)
  sparc64: Update defconfig.
  sparc: Wire up sys_rt_tgsigqueueinfo().
  openprom: Squelch useless GCC warning.
  sparc: replace uses of CPU_MASK_ALL_PTR
  sparc64: Add proper dynamic ftrace support.
  sparc: Simplify code using is_power_of_2() routine.
  sparc: move of_device common code to of_device_common
  sparc: remove dma-mapping_{32|64}.h
  sparc: use dma_map_page instead of dma_map_single
  sparc: add sync_single_for_device and sync_sg_for_device to struct dma_ops
  sparc: move the duplication in dma-mapping_{32|64}.h to dma-mapping.h
  p9100: use standard fields for framebuffer physical address and length
  leo: use standard fields for framebuffer physical address and length
  cg6: use standard fields for framebuffer physical address and length
  cg3: use standard fields for framebuffer physical address and length
  cg14: use standard fields for framebuffer physical address and length
  bw2: use standard fields for framebuffer physical address and length
  sparc64: fix and optimize irq distribution
  sparc64: Use new dynamic per-cpu allocator.
  sparc64: Only allocate per-cpu areas for possible cpus.
  ...

10 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/vapier...
Linus Torvalds [Tue, 16 Jun 2009 18:49:58 +0000 (11:49 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/vapier/blackfin

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/vapier/blackfin: (27 commits)
  Blackfin: hook up new rt_tgsigqueueinfo syscall
  Blackfin: improve CLKIN_HZ config default
  Blackfin: initial support for ftrace grapher
  Blackfin: initial support for ftrace
  Blackfin: enable support for LOCKDEP
  Blackfin: add preliminary support for STACKTRACE
  Blackfin: move custom sections into sections.h
  Blackfin: punt unused/wrong mutex-dec.h
  Blackfin: add support for irqflags
  Blackfin: add support for bzip2/lzma compressed kernel images
  Blackfin: convert Kconfig style to def_bool
  Blackfin: bf548-ezkit: update smsc911x resources
  Blackfin: update aedos-ipipe code to upstream 1.10-00
  Blackfin: bf537-stamp: update ADP5520 resources
  Blackfin: bf518f-ezbrd: fix SPI CS for SPI flash
  Blackfin: define SPI IRQ in board resources
  Blackfin: do not configure the UART early if on wrong processor
  Blackfin: fix deadlock in SMP IPI handler
  Blackfin: fix flag storage for irq funcs
  Blackfin: push down exception oops checking
  ...

10 years agoMerge branch 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6
Linus Torvalds [Tue, 16 Jun 2009 18:48:13 +0000 (11:48 -0700)]
Merge branch 'for-linus' of git://git390.marist.edu/linux-2.6

* 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6: (33 commits)
  [S390] s390: hibernation support for s390
  [S390] pm: dcssblk power management callbacks.
  [S390] pm: monreader power management callbacks.
  [S390] pm: monwriter power management callbacks.
  [S390] pm: memory hotplug power management callbacks
  [S390] pm: con3270 power management callbacks.
  [S390] pm: smsgiucv power management callbacks.
  [S390] pm: hvc_iucv power management callbacks
  [S390] PM: af_iucv power management callbacks.
  [S390] pm: netiucv power management callbacks.
  [S390] pm: iucv power management callbacks.
  [S390] iucv: establish reboot notifier
  [S390] pm: power management support for SCLP drivers.
  [S390] pm: tape power management callbacks
  [S390] pm: vmlogrdr power management callbacks
  [S390] pm: vmur driver power management callbacks
  [S390] pm: appldata power management callbacks
  [S390] pm: vmwatchdog power management callbacks.
  [S390] pm: zfcp driver power management callbacks
  [S390] pm: claw driver power management callbacks
  ...

10 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block
Linus Torvalds [Tue, 16 Jun 2009 18:46:45 +0000 (11:46 -0700)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block

* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
  block: remove some includings of blktrace_api.h
  mg_disk: seperate mg_disk.h again
  block: Introduce helper to reset queue limits to default values
  cfq: remove extraneous '\n' in blktrace output
  ubifs: register backing_dev_info
  btrfs: properly register fs backing device
  block: don't overwrite bdi->state after bdi_init() has been run
  cfq: cleanup for last_end_request in cfq_data

10 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Tue, 16 Jun 2009 18:30:37 +0000 (11:30 -0700)]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (38 commits)
  ps3flash: Always read chunks of 256 KiB, and cache them
  ps3flash: Cache the last accessed FLASH chunk
  ps3: Replace direct file operations by callback
  ps3: Switch ps3_os_area_[gs]et_rtc_diff to EXPORT_SYMBOL_GPL()
  ps3: Correct debug message in dma_ioc0_map_pages()
  drivers/ps3: Add missing annotations
  ps3fb: Use ps3_system_bus_[gs]et_drvdata() instead of direct access
  ps3flash: Use ps3_system_bus_[gs]et_drvdata() instead of direct access
  ps3: shorten ps3_system_bus_[gs]et_driver_data to ps3_system_bus_[gs]et_drvdata
  ps3: Use dev_[gs]et_drvdata() instead of direct access for system bus devices
  block/ps3: remove driver_data direct access of struct device
  ps3vram: Make ps3vram_priv.reports a void *
  ps3vram: Remove no longer used ps3vram_priv.ddr_base
  ps3vram: Replace mutex by spinlock + bio_list
  block: Add bio_list_peek()
  powerpc: Use generic atomic64_t implementation on 32-bit processors
  lib: Provide generic atomic64_t implementation
  powerpc: Add compiler memory barrier to mtmsr macro
  powerpc/iseries: Mark signal_vsp_instruction() as maybe unused
  powerpc/iseries: Fix unused function warning in iSeries DT code
  ...

10 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable
Linus Torvalds [Tue, 16 Jun 2009 18:30:16 +0000 (11:30 -0700)]
Merge git://git./linux/kernel/git/mason/btrfs-unstable

* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
  Btrfs: always update root items for fs trees at commit time

10 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/hirofumi/fatfs-2.6
Linus Torvalds [Tue, 16 Jun 2009 18:29:44 +0000 (11:29 -0700)]
Merge git://git./linux/kernel/git/hirofumi/fatfs-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/hirofumi/fatfs-2.6:
  fat: split fat_generic_ioctl
  FAT: add 'errors' mount option

10 years agoMerge branch 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvar...
Linus Torvalds [Tue, 16 Jun 2009 18:29:17 +0000 (11:29 -0700)]
Merge branch 'i2c-for-linus' of git://git./linux/kernel/git/jdelvare/staging

* 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  therm_windtunnel: Convert to a new-style i2c driver
  therm_adt746x: Convert to a new-style i2c driver
  windfarm: Convert to new-style i2c drivers
  therm_pm72: Convert to a new-style i2c driver
  i2c-viapro: Add new PCI device ID for VX855
  i2c/chips: Move max6875 to drivers/misc/eeprom
  i2c: Do not give adapters a default parent
  i2c: Do not probe for TV chips on Voodoo3 adapters
  i2c: Retry automatically on arbitration loss
  i2c: Remove void casts

10 years agoMerge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelv...
Linus Torvalds [Tue, 16 Jun 2009 18:28:50 +0000 (11:28 -0700)]
Merge branch 'hwmon-for-linus' of git://git./linux/kernel/git/jdelvare/staging

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  hwmon: (max6650) Add support for alarms
  hwmon: (f71882fg) Add support for the F71858F
  hwmon: (f71882fg) Add temp#_fault sysfs attr for f8000
  hwmon: (f71882fg) Sanity check f8000 pwm settings
  hwmon: (f71882fg) Cleanup f8000 pwm handling
  hwmon: PCI quirk for hwmon access on MSI MS-7031 board
  hwmon: (w83627ehf) Add W83627DHG-P support
  hwmon: (tmp401) Add documentation
  hwmon: (tmp401) Add support for TI's TMP411 sensors chip
  hwmon: (tmp401) Add support for TI's TMP401 sensor chip
  hwmon: (ibmaem) Automatically load on HC10 blade
  hwmon: Fix more __devexit_p glitches

10 years agoMerge branch 'acpica' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux...
Linus Torvalds [Tue, 16 Jun 2009 18:24:23 +0000 (11:24 -0700)]
Merge branch 'acpica' of git://git./linux/kernel/git/lenb/linux-acpi-2.6

* 'acpica' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6: (27 commits)
  ACPICA: Update version to 20090521.
  ACPICA: Disable preservation of SCI enable bit (SCI_EN)
  ACPICA: Region deletion: Ensure region object is removed from handler list
  ACPICA: Eliminate extra call to NsGetParentNode
  ACPICA: Simplify internal operation region interface
  ACPICA: Update Load() to use operation region interfaces
  ACPICA: New: AcpiInstallMethod - install a single control method
  ACPICA: Invalidate DdbHandle after table unload
  ACPICA: Fix reference count issues for DdbHandle object
  ACPICA: Simplify and optimize NsGetNextNode function
  ACPICA: Additional validation of _PRT packages (resource mgr)
  ACPICA: Fix DebugObject output for DdbHandle objects
  ACPICA: Fix allowable release order for ASL mutex objects
  ACPICA: Mutex support: Fix release ordering issue and current sync level
  ACPICA: Update version to 20090422.
  ACPICA: Linux OSL: cleanup/update/merge
  ACPICA: Fix implementation of AML BreakPoint operator (break to debugger)
  ACPICA: Fix miscellaneous warnings under gcc 4+
  ACPICA: Miscellaneous lint changes
  ACPICA: Fix possible dereference of null pointer
  ...

10 years agotty: Fix leaks introduced by the shift to separate ldisc objects
Alan Cox [Mon, 15 Jun 2009 15:28:29 +0000 (16:28 +0100)]
tty: Fix leaks introduced by the shift to separate ldisc objects

Gold star for the kmemleak detector.

Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agopty: Narrow the race on ldisc locking
Alan Cox [Mon, 15 Jun 2009 15:27:29 +0000 (16:27 +0100)]
pty: Narrow the race on ldisc locking

The pty code has always been buggy on its ldisc handling. The recent
changes made the window for the race much bigger. Pending fixing it
properly which is not at all trivial, at least make the race small again so
we don't disrupt other dev work.

Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agoprintk: add KERN_DEFAULT loglevel to print_modules()
Linus Torvalds [Tue, 16 Jun 2009 18:07:14 +0000 (11:07 -0700)]
printk: add KERN_DEFAULT loglevel to print_modules()

Several WARN_ON() messages omit the '\n' at the end of the string, which
is a simple (and understandable) error.  The next line printed after
that warning line is usually the current module list, and that printk
does not have a log-level marker - resulting in one long mixed-up line.

Adding this loglevel marker will now avoid this unreadable mess.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agoprintk: Add KERN_DEFAULT printk log-level
Linus Torvalds [Tue, 16 Jun 2009 18:02:28 +0000 (11:02 -0700)]
printk: Add KERN_DEFAULT printk log-level

This adds a KERN_DEFAULT loglevel marker, for when you cannot decide
which loglevel you want, and just want to keep an existing printk
with the default loglevel.

The difference between having KERN_DEFAULT and having no log-level
marker at all is two-fold:

 - having the log-level marker will now force a new-line if the
   previous printout had not added one (perhaps because it forgot,
   but perhaps because it expected a continuation)

 - having a log-level marker is required if you are printing out a
   message that otherwise itself could perhaps otherwise be mistaken
   for a log-level.

Signed-of-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agoprintk: clean up handling of log-levels and newlines
Linus Torvalds [Tue, 16 Jun 2009 17:57:02 +0000 (10:57 -0700)]
printk: clean up handling of log-levels and newlines

It used to be that we would only look at the log-level in a printk()
after explicit newlines, which can cause annoying problems when the
previous printk() did not end with a '\n'. In that case, the log-level
marker would be just printed out in the middle of the line, and be
seen as just noise rather than change the logging level.

This changes things to always look at the log-level in the first
bytes of the printout. If a log level marker is found, it is always
used as the log-level. Additionally, if no newline existed, one is
added (unless the log-level is the explicit KERN_CONT marker, to
explicitly show that it's a continuation of a previous line).

Acked-by: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agosparc64: Update defconfig.
David S. Miller [Tue, 16 Jun 2009 11:59:53 +0000 (04:59 -0700)]
sparc64: Update defconfig.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: Wire up sys_rt_tgsigqueueinfo().
David S. Miller [Tue, 16 Jun 2009 11:17:58 +0000 (04:17 -0700)]
sparc: Wire up sys_rt_tgsigqueueinfo().

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoopenprom: Squelch useless GCC warning.
David S. Miller [Tue, 16 Jun 2009 10:46:54 +0000 (03:46 -0700)]
openprom: Squelch useless GCC warning.

drivers/sbus/char/openprom.c: In function ‘openprom_sunos_ioctl’:
drivers/sbus/char/openprom.c:306: warning: ‘opp’ may be used uninitialized in this function

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: replace uses of CPU_MASK_ALL_PTR
Stephen Rothwell [Mon, 15 Jun 2009 10:06:18 +0000 (03:06 -0700)]
sparc: replace uses of CPU_MASK_ALL_PTR

CPU_MASK_ALL is the (deprecated) "all bits set" cpumask, defined as so:

#define CPU_MASK_ALL (cpumask_t) { { ... } }

Taking the address of such a temporary is questionable at best,
unfortunately 321a8e9d (cpumask: add CPU_MASK_ALL_PTR macro) added
CPU_MASK_ALL_PTR:

#define CPU_MASK_ALL_PTR (&CPU_MASK_ALL)

Which formalizes this practice.  One day gcc could bite us over this
usage (though we seem to have gotten away with it so far).

[Description by Rusty Russell]
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Add proper dynamic ftrace support.
David S. Miller [Sat, 13 Jun 2009 08:03:24 +0000 (01:03 -0700)]
sparc64: Add proper dynamic ftrace support.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Ingo Molnar <mingo@elte.hu>
10 years agosparc: Simplify code using is_power_of_2() routine.
Robert P. J. Day [Fri, 24 Apr 2009 03:58:24 +0000 (03:58 +0000)]
sparc: Simplify code using is_power_of_2() routine.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: move of_device common code to of_device_common
Robert Reif [Thu, 4 Jun 2009 09:00:02 +0000 (02:00 -0700)]
sparc: move of_device common code to of_device_common

This patch moves code common to of_device_32.c and of_device_64.c into
of_device_common.h and of_device_common.c.

The only functional difference is in sparc32 where of_bus_default_map is
used in place of of_bus_sbus_map because they are equivelent.

There is still room for further code consolidation with some minor
refactoring.

Boot tested on sparc32 and compile tested on sparc64.

Signed-off-by: Robert Reif <reif@earthlink.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: remove dma-mapping_{32|64}.h
FUJITA Tomonori [Thu, 14 May 2009 16:23:11 +0000 (16:23 +0000)]
sparc: remove dma-mapping_{32|64}.h

This modifies SPARC32 to use struct dma_map ops. It means that we can
remove dma-mapping_{32|64}.h.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Tested-by: Robert Reif <reif@earthlink.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: use dma_map_page instead of dma_map_single
FUJITA Tomonori [Thu, 14 May 2009 16:23:10 +0000 (16:23 +0000)]
sparc: use dma_map_page instead of dma_map_single

This patch converts dma_map_single and dma_unmap_single to use
map_page and unmap_page respectively and removes unnecessary
map_single and unmap_single. map_page can be used to implement
map_single but the opposite is impossible. Having only dma_map_page in
struct dma_ops is enough.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Tested-by: Robert Reif <reif@earthlink.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: add sync_single_for_device and sync_sg_for_device to struct dma_ops
FUJITA Tomonori [Thu, 14 May 2009 16:23:09 +0000 (16:23 +0000)]
sparc: add sync_single_for_device and sync_sg_for_device to struct dma_ops

This adds sync_single_for_device() and sync_sg_for_device() to struct
dma_ops in order to unify dma-mpping_{32|64}.h. dma-mpping_32.h needs them though dma-mpping_64.h doesn't.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Tested-by: Robert Reif <reif@earthlink.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: move the duplication in dma-mapping_{32|64}.h to dma-mapping.h
FUJITA Tomonori [Thu, 14 May 2009 16:23:08 +0000 (16:23 +0000)]
sparc: move the duplication in dma-mapping_{32|64}.h to dma-mapping.h

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Tested-by: Robert Reif <reif@earthlink.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agop9100: use standard fields for framebuffer physical address and length
Krzysztof Helt [Mon, 4 May 2009 10:42:57 +0000 (10:42 +0000)]
p9100: use standard fields for framebuffer physical address and length

Use standard fields fbinfo.fix.smem_start and fbinfo.fix.smem_len
for physical address and length of framebuffer.

This also fixes output of the 'fbset -i' command - address and length
of the framebuffer are displayed correctly.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoleo: use standard fields for framebuffer physical address and length
Krzysztof Helt [Mon, 4 May 2009 10:41:56 +0000 (10:41 +0000)]
leo: use standard fields for framebuffer physical address and length

Use standard fields fbinfo.fix.smem_start and fbinfo.fix.smem_len
for physical address and length of framebuffer.

This also fixes output of the 'fbset -i' command - address and length
of the framebuffer are displayed correctly.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agocg6: use standard fields for framebuffer physical address and length
Krzysztof Helt [Mon, 4 May 2009 10:39:45 +0000 (10:39 +0000)]
cg6: use standard fields for framebuffer physical address and length

Use standard fields fbinfo.fix.smem_start and fbinfo.fix.smem_len
for physical address and length of framebuffer.

This also fixes output of the 'fbset -i' command - address and length
of the framebuffer are displayed correctly.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agocg3: use standard fields for framebuffer physical address and length
Krzysztof Helt [Mon, 4 May 2009 10:37:46 +0000 (10:37 +0000)]
cg3: use standard fields for framebuffer physical address and length

Use standard fields fbinfo.fix.smem_start and fbinfo.fix.smem_len
for physical address and length of framebuffer.

This also fixes output of the 'fbset -i' command - address and length
of the framebuffer is displayed correctly.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agocg14: use standard fields for framebuffer physical address and length
Krzysztof Helt [Mon, 4 May 2009 10:36:50 +0000 (10:36 +0000)]
cg14: use standard fields for framebuffer physical address and length

Use standard fields fbinfo.fix.smem_start and fbinfo.fix.smem_len
for physical address and length of framebuffer.

This also fixes output of the 'fbset -i' command - address and length
of the framebuffer is displayed correctly.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agobw2: use standard fields for framebuffer physical address and length
Krzysztof Helt [Mon, 4 May 2009 10:35:36 +0000 (10:35 +0000)]
bw2: use standard fields for framebuffer physical address and length

Use standard fields fbinfo.fix.smem_start and fbinfo.fix.smem_len
for physical address and length of framebuffer.

This also fixes output of the 'fbset -i' command - address and length
of the framebuffer is displayed correctly.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: fix and optimize irq distribution
Hong H. Pham [Thu, 4 Jun 2009 09:10:11 +0000 (02:10 -0700)]
sparc64: fix and optimize irq distribution

irq_choose_cpu() should compare the affinity mask against cpu_online_map
rather than CPU_MASK_ALL, since irq_select_affinity() sets the interrupt's
affinity mask to cpu_online_map "and" CPU_MASK_ALL (which ends up being
just cpu_online_map).  The mask comparison in irq_choose_cpu() will always
fail since the two masks are not the same.  So the CPU chosen is the first CPU
in the intersection of cpu_online_map and CPU_MASK_ALL, which is always CPU0.
That means all interrupts are reassigned to CPU0...

Distributing interrupts to CPUs in a linearly increasing round robin fashion
is not optimal for the UltraSPARC T1/T2.  Also, the irq_rover in
irq_choose_cpu() causes an interrupt to be assigned to a different
processor each time the interrupt is allocated and released.  This may lead
to an unbalanced distribution over time.

A static mapping of interrupts to processors is done to optimize and balance
interrupt distribution.  For the T1/T2, interrupts are spread to different
cores first, and then to strands within a core.

The following is some benchmarks showing the effects of interrupt
distribution on a T2.  The test was done with iperf using a pair of T5220
boxes, each with a 10GBe NIU (XAUI) connected back to back.

  TCP     | Stock       Linear RR IRQ  Optimized IRQ
  Streams | 2.6.30-rc5  Distribution   Distribution
          | GBits/sec   GBits/sec      GBits/sec
  --------+-----------------------------------------
    1       0.839       0.862          0.868
    8       1.16        4.96           5.88
   16       1.15        6.40           8.04
  100       1.09        7.28           8.68

Signed-off-by: Hong H. Pham <hong.pham@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Use new dynamic per-cpu allocator.
David S. Miller [Thu, 9 Apr 2009 03:32:02 +0000 (20:32 -0700)]
sparc64: Use new dynamic per-cpu allocator.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Only allocate per-cpu areas for possible cpus.
David S. Miller [Wed, 1 Apr 2009 23:22:27 +0000 (16:22 -0700)]
sparc64: Only allocate per-cpu areas for possible cpus.

This gets us real close to the generic implementation of
setup_per_cpu_areas() except:

1) We store the per-cpu offset into the trap_block[], whereas
   the generic code has it's own static array.

2) We have to initialize the %g5 register to hold the boot cpu's
   per-cpu area offset.

3) The OBP/MDESC cpu info scan is performed at the end.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Get rid of real_setup_per_cpu_areas().
David S. Miller [Wed, 1 Apr 2009 23:15:20 +0000 (16:15 -0700)]
sparc64: Get rid of real_setup_per_cpu_areas().

Now that we defer the cpu_data() initializations to the end of per-cpu
setup, we can get rid of this local hack we had to setup the per-cpu
areas eary.

This is a necessary step in order to support HAVE_DYNAMIC_PER_CPU_AREA
since the per-cpu setup must run when page structs are available.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Defer cpu_data() setup until end of per-cpu data initialization.
David S. Miller [Wed, 27 May 2009 05:37:25 +0000 (22:37 -0700)]
sparc64: Defer cpu_data() setup until end of per-cpu data initialization.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Make mdesc_fill_in_cpu_data take a cpumask_t pointer.
David S. Miller [Wed, 1 Apr 2009 10:15:11 +0000 (03:15 -0700)]
sparc64: Make mdesc_fill_in_cpu_data take a cpumask_t pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc: Call OF and MD cpu scanning explicitly from paging_init()
David S. Miller [Wed, 1 Apr 2009 10:13:15 +0000 (03:13 -0700)]
sparc: Call OF and MD cpu scanning explicitly from paging_init()

We need to split up the cpu present mask setup from the cpu_data
initialization, and this is a first step towards that.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Refactor MDESC cpu scanning code using an iterator.
David S. Miller [Thu, 9 Apr 2009 02:55:47 +0000 (19:55 -0700)]
sparc64: Refactor MDESC cpu scanning code using an iterator.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Refactor OBP cpu scanning code using an iterator.
David S. Miller [Wed, 27 May 2009 04:46:14 +0000 (21:46 -0700)]
sparc64: Refactor OBP cpu scanning code using an iterator.

With feedback from Sam Ravnborg.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Use BUILD_BUG_ON() in trap_init().
David S. Miller [Wed, 1 Apr 2009 08:57:36 +0000 (01:57 -0700)]
sparc64: Use BUILD_BUG_ON() in trap_init().

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Store per-cpu offset in trap_block[]
David S. Miller [Wed, 1 Apr 2009 08:47:10 +0000 (01:47 -0700)]
sparc64: Store per-cpu offset in trap_block[]

Surprisingly this actually makes LOAD_PER_CPU_BASE() a little
more efficient.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Move trap_block[] definitions into a new header file.
David S. Miller [Wed, 1 Apr 2009 08:26:43 +0000 (01:26 -0700)]
sparc64: Move trap_block[] definitions into a new header file.

Later we're going to want to get at these definitions from
asm/percpu.h and that's not possible via cpudata.h because
of the set of dependencies the non-trap_block[] stuff has.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agosparc64: Reclaim trap_block[]->hdesc
David S. Miller [Wed, 1 Apr 2009 00:15:40 +0000 (17:15 -0700)]
sparc64: Reclaim trap_block[]->hdesc

This really isn't necessary at all, a local variable suits the
job just fine.

This frees up 8 bytes in the trap_block[] that we can use later
to store the per-cpu base addresses.

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoblock: remove some includings of blktrace_api.h
Li Zefan [Tue, 16 Jun 2009 09:19:36 +0000 (11:19 +0200)]
block: remove some includings of blktrace_api.h

When porting blktrace to tracepoints, we changed to trace/block.h
for trace prober declarations.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years ago[S390] s390: hibernation support for s390
Hans-Joachim Picht [Tue, 16 Jun 2009 08:30:52 +0000 (10:30 +0200)]
[S390] s390: hibernation support for s390

This patch introduces the hibernation backend support to the
s390 architecture. Now it is possible to suspend a mainframe Linux
guest using the following command:

echo disk > /sys/power/state

Signed-off-by: Hans-Joachim Picht <hans@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: dcssblk power management callbacks.
Gerald Schaefer [Tue, 16 Jun 2009 08:30:51 +0000 (10:30 +0200)]
[S390] pm: dcssblk power management callbacks.

Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: monreader power management callbacks.
Gerald Schaefer [Tue, 16 Jun 2009 08:30:50 +0000 (10:30 +0200)]
[S390] pm: monreader power management callbacks.

Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: monwriter power management callbacks.
Gerald Schaefer [Tue, 16 Jun 2009 08:30:49 +0000 (10:30 +0200)]
[S390] pm: monwriter power management callbacks.

Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: memory hotplug power management callbacks
Gerald Schaefer [Tue, 16 Jun 2009 08:30:48 +0000 (10:30 +0200)]
[S390] pm: memory hotplug power management callbacks

Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: con3270 power management callbacks.
Martin Schwidefsky [Tue, 16 Jun 2009 08:30:47 +0000 (10:30 +0200)]
[S390] pm: con3270 power management callbacks.

Introduce the power management callbacks to the 3270 driver. On suspend
the current 3270 view is deactivated and for non-console 3270 device
the release callback is done. This disconnects the current tty /
fullscreen application from the 3270 device. On resume the current
view is reactivated, on the tty you get a fresh login.
If the system panics before the 3270 device has been resumed, the ccw
device for the 3270 console is reactivated with ccw_device_force_console.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: smsgiucv power management callbacks.
Martin Schwidefsky [Tue, 16 Jun 2009 08:30:46 +0000 (10:30 +0200)]
[S390] pm: smsgiucv power management callbacks.

Create dummy iucv-device to get control when the system is suspended
and resumed. Server the smsg iucv path on suspend, reestablish the
path on resume.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: hvc_iucv power management callbacks
Hendrik Brueckner [Tue, 16 Jun 2009 08:30:45 +0000 (10:30 +0200)]
[S390] pm: hvc_iucv power management callbacks

The patch adds supporting for suspending and resuming IUCV HVC terminal
devices from disk. The obligatory Linux device driver interfaces has
been added by registering a device driver on the IUCV bus.
For each IUCV HVC terminal device the driver creates a respective device
on the IUCV bus.

To support suspend and resume, the PM freeze callback severs any established
IUCV communication path and triggers a HVC tty hang-up when the system image
is restored.
IUCV communication path are no longer valid when the z/VM guest is halted.

The device driver initialization has been updated to register devices and
the a new routine has been extracted to facilitate the hang-up of IUCV HVC
terminal devices.

Signed-off-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] PM: af_iucv power management callbacks.
Ursula Braun [Tue, 16 Jun 2009 08:30:44 +0000 (10:30 +0200)]
[S390] PM: af_iucv power management callbacks.

Patch establishes a dummy afiucv-device to make sure af_iucv is
notified as iucv-bus device about suspend/resume.

The PM freeze callback severs all iucv pathes of connected af_iucv sockets.
The PM thaw/restore callback switches the state of all previously connected
sockets to IUCV_DISCONN.

Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: netiucv power management callbacks.
Ursula Braun [Tue, 16 Jun 2009 08:30:43 +0000 (10:30 +0200)]
[S390] pm: netiucv power management callbacks.

Patch establishes a dummy netiucv device to make sure iucv is notified
about suspend/resume even if netiucv is the only loaded iucv-exploting
module without any real net_device defined.

The PM freeze callback closes all open netiucv connections. Thus the
corresponding iucv path is removed.
The PM thaw/restore callback re-opens previously closed netiucv
connections.

Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: iucv power management callbacks.
Ursula Braun [Tue, 16 Jun 2009 08:30:42 +0000 (10:30 +0200)]
[S390] pm: iucv power management callbacks.

Patch calls the PM callback functions of iucv-bus devices, which are
responsible for removal of their established iucv pathes.

The PM freeze callback for the first iucv-bus device disables all iucv
interrupts except the connection severed interrupt.
The PM freeze callback for the last iucv-bus device shuts down iucv.

The PM thaw callback for the first iucv-bus device re-enables iucv
if it has been shut down during freeze. If freezing has been interrupted,
it re-enables iucv interrupts according to the needs of iucv-exploiters.

The PM restore callback for the first iucv-bus device re-enables iucv.

Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] iucv: establish reboot notifier
Ursula Braun [Tue, 16 Jun 2009 08:30:41 +0000 (10:30 +0200)]
[S390] iucv: establish reboot notifier

To guarantee a proper cleanup, patch adds a reboot notifier to
the iucv base code, which disables iucv interrupts, shuts down
established iucv pathes, and removes iucv declarations for z/VM.

Checks have to be added to the iucv-API functions, whether
iucv-buffers removed at reboot time are still declared.

Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: power management support for SCLP drivers.
Michael Holzheu [Tue, 16 Jun 2009 08:30:40 +0000 (10:30 +0200)]
[S390] pm: power management support for SCLP drivers.

The SCLP base driver defines a new notifier call back for all upper level SCLP
drivers, like the SCLP console, etc. This guarantees that in suspend first the
upper level drivers are suspended and afterwards the SCLP base driver. For
resume it is the other way round. The SCLP base driver itself registers a
new platform device at the platform bus and gets PM notifications via
the dev_pm_ops.

In suspend, the SCLP base driver switches off the receiver and sender mask
This is done in sclp_deactivate(). After suspend all new requests will be
rejected with -EIO and no more interrupts will be received, because the masks
are switched off. For resume the sender and receiver masks are reset in
the sclp_reactivate() function.

When the SCLP console is suspended, all new messages are cached in the
sclp console buffers. In resume, all the cached messages are written to the
console. In addition to that we have an early resume function that removes
the cached messages from the suspend image.

Signed-off-by: Michael Holzheu <holzheu@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: tape power management callbacks
Frank Munzert [Tue, 16 Jun 2009 08:30:39 +0000 (10:30 +0200)]
[S390] pm: tape power management callbacks

Signed-off-by: Frank Munzert <munzert@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: vmlogrdr power management callbacks
Stefan Weinhuber [Tue, 16 Jun 2009 08:30:38 +0000 (10:30 +0200)]
[S390] pm: vmlogrdr power management callbacks

Signed-off-by: Stefan Weinhuber <wein@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: vmur driver power management callbacks
Frank Munzert [Tue, 16 Jun 2009 08:30:37 +0000 (10:30 +0200)]
[S390] pm: vmur driver power management callbacks

Signed-off-by: Frank Munzert <munzert@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: appldata power management callbacks
Gerald Schaefer [Tue, 16 Jun 2009 08:30:36 +0000 (10:30 +0200)]
[S390] pm: appldata power management callbacks

Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: vmwatchdog power management callbacks.
Christian Borntraeger [Tue, 16 Jun 2009 08:30:35 +0000 (10:30 +0200)]
[S390] pm: vmwatchdog power management callbacks.

This patch implements suspend/hibernation for the vmwatchdog driver. The
pm_notifier_callchain is used to get control on PM events. Since watchdog
operation and suspend cannot work together in a reliable fashion, the open
flag is also used to prevent suspend and open from happening at the same
time.
The watchdog can also be active with no open file descriptor. This patch
adds another flag which is only changed in vmwdt_keep_alive and
vmwdt_disable.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: zfcp driver power management callbacks
Martin Petermann [Tue, 16 Jun 2009 08:30:34 +0000 (10:30 +0200)]
[S390] pm: zfcp driver power management callbacks

Signed-off-by: Martin Petermann <martin.petermann@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: claw driver power management callbacks
Frank Blaschka [Tue, 16 Jun 2009 08:30:33 +0000 (10:30 +0200)]
[S390] pm: claw driver power management callbacks

Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: ctcm driver power management callbacks
Frank Blaschka [Tue, 16 Jun 2009 08:30:32 +0000 (10:30 +0200)]
[S390] pm: ctcm driver power management callbacks

Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: qeth driver power management callbacks
Frank Blaschka [Tue, 16 Jun 2009 08:30:31 +0000 (10:30 +0200)]
[S390] pm: qeth driver power management callbacks

Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: lcs driver power management callbacks
Klaus-Dieter Wacket [Tue, 16 Jun 2009 08:30:30 +0000 (10:30 +0200)]
[S390] pm: lcs driver power management callbacks

Signed-off-by: Klaus-Dieter Wacker <kdwacker@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: con3215 power management callbacks
Martin Schwidefsky [Tue, 16 Jun 2009 08:30:29 +0000 (10:30 +0200)]
[S390] pm: con3215 power management callbacks

Introduce the power management callbacks to the 3215 console. On suspend
the console buffer is flushed to the 3215 device to have an empty console
buffer. Printks done while the 3215 device is suspended are buffered in
the 64K buffer of the 3215 device. If the buffer is full new messages will
push out the oldest messages to make room for the most recent message.
On resume the buffered messages are printed. If the system panics before
the 3215 device is resumed ccw_device_force_console is used to get the
console working again.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] cio: force console function
Martin Schwidefsky [Tue, 16 Jun 2009 08:30:28 +0000 (10:30 +0200)]
[S390] cio: force console function

If something goes wrong in a suspend / resume cycle a ccw based console
if very likely in the suspended state and cannot print anything.
Introduce ccw_device_force_console to force the wake up of the console
device to be able to print the oops message. The console device drivers
should use this function only if the system paniced.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: xpram driver power management callbacks
Michael Holzheu [Tue, 16 Jun 2009 08:30:27 +0000 (10:30 +0200)]
[S390] pm: xpram driver power management callbacks

Signed-off-by: Michael Holzheu <holzheu@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: add kernel_page_present
Hans-Joachim Picht [Tue, 16 Jun 2009 08:30:26 +0000 (10:30 +0200)]
[S390] pm: add kernel_page_present

Fix the following build failure caused by make allyesconfig using
CONFIG_HIBERNATION and CONFIG_DEBUG_PAGEALLOC

kernel/built-in.o: In function `saveable_page':
kernel/power/snapshot.c:897: undefined reference to `kernel_page_present'
kernel/built-in.o: In function `safe_copy_page':
kernel/power/snapshot.c:948: undefined reference to `kernel_page_present'
make: *** [.tmp_vmlinux1] Error 1

Signed-off-by: Hans-Joachim Picht <hans@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: dasd power management callbacks.
Stefan Haberland [Tue, 16 Jun 2009 08:30:25 +0000 (10:30 +0200)]
[S390] pm: dasd power management callbacks.

Introduce the power management callbacks to the dasd driver. On suspend
the dasd devices are stopped and removed from the focus of alias
management.
On resume they are reinitialized by rereading the device characteristics
and adding the device to the alias management.
In case the device has gone away during suspend it will caught in the
suspend state with stopped flag set to UNRESUMED. After it appears again
the restore function is called again.

Signed-off-by: Stefan Haberland <stefan.haberland@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: chsc subchannel driver power management callbacks
Cornelia Huck [Tue, 16 Jun 2009 08:30:24 +0000 (10:30 +0200)]
[S390] pm: chsc subchannel driver power management callbacks

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: io subchannel driver power management callbacks
Cornelia Huck [Tue, 16 Jun 2009 08:30:23 +0000 (10:30 +0200)]
[S390] pm: io subchannel driver power management callbacks

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: css bus power management callbacks
Sebastian Ott [Tue, 16 Jun 2009 08:30:22 +0000 (10:30 +0200)]
[S390] pm: css bus power management callbacks

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: ccwgroup bus power management callbacks
Sebastian Ott [Tue, 16 Jun 2009 08:30:21 +0000 (10:30 +0200)]
[S390] pm: ccwgroup bus power management callbacks

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years ago[S390] pm: ccw bus power management callbacks
Sebastian Ott [Tue, 16 Jun 2009 08:30:20 +0000 (10:30 +0200)]
[S390] pm: ccw bus power management callbacks

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
10 years agomg_disk: seperate mg_disk.h again
unsik Kim [Tue, 16 Jun 2009 06:40:20 +0000 (08:40 +0200)]
mg_disk: seperate mg_disk.h again

eec9462088a26c046d4db3100796a340a50890b8 fold mg_disk.h into mg_disk.c,
but mg_disk platform driver needs private data for operation. This also
make mg_disk.c as machine independent. Seperate only needed structure and
defines to mg_disk.h

Signed-off-by: unsik Kim <donari75@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agoblock: Introduce helper to reset queue limits to default values
Martin K. Petersen [Tue, 16 Jun 2009 06:23:52 +0000 (08:23 +0200)]
block: Introduce helper to reset queue limits to default values

DM reuses the request queue when swapping in a new device table
Introduce blk_set_default_limits() which can be used to reset the the
queue_limits prior to stacking devices.

Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Acked-by: Alasdair G Kergon <agk@redhat.com>
Acked-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agocfq: remove extraneous '\n' in blktrace output
Jeff Moyer [Fri, 12 Jun 2009 13:29:30 +0000 (15:29 +0200)]
cfq: remove extraneous '\n' in blktrace output

I noticed a blank line in blktrace output.  This patch fixes that.

Signed-off-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agoubifs: register backing_dev_info
Jens Axboe [Fri, 29 May 2009 07:19:23 +0000 (09:19 +0200)]
ubifs: register backing_dev_info

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agobtrfs: properly register fs backing device
Jens Axboe [Fri, 12 Jun 2009 12:43:40 +0000 (14:43 +0200)]
btrfs: properly register fs backing device

btrfs assigns this bdi to all inodes on that file system, so make
sure it's registered. This isn't really important now, but will be
when we put dirty inodes there. Even now, we miss the stats when the
bdi isn't visible.

Also fixes failure to check bdi_init() return value, and bad inherit of
->capabilities flags from the default bdi.

Acked-by: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agoblock: don't overwrite bdi->state after bdi_init() has been run
Jens Axboe [Fri, 12 Jun 2009 12:42:56 +0000 (14:42 +0200)]
block: don't overwrite bdi->state after bdi_init() has been run

Move the defaults to where we do the init of the backing_dev_info.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agocfq: cleanup for last_end_request in cfq_data
Gui Jianfeng [Fri, 12 Jun 2009 13:27:50 +0000 (15:27 +0200)]
cfq: cleanup for last_end_request in cfq_data

Actually, last_end_request in cfq_data isn't used now. So lets
just remove it.

Signed-off-by: Gui Jianfeng <guijianfeng@cn.fujitsu.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
10 years agoBtrfs: always update root items for fs trees at commit time
Yan Zheng [Tue, 16 Jun 2009 00:01:02 +0000 (20:01 -0400)]
Btrfs: always update root items for fs trees at commit time

commit_fs_roots skips updating root items for fs trees that aren't modified.
This is unsafe now that relocation code modifies root item's last_snapshot
field without modifying corresponding fs tree.

Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
10 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Mon, 15 Jun 2009 17:38:06 +0000 (10:38 -0700)]
Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] fix compile error in arch/ia64/mm/extable.c

10 years agoMerge branch 'timers-for-linus-migration' of git://git.kernel.org/pub/scm/linux/kerne...
Linus Torvalds [Mon, 15 Jun 2009 17:06:19 +0000 (10:06 -0700)]
Merge branch 'timers-for-linus-migration' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-for-linus-migration' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  timers: Logic to move non pinned timers
  timers: /proc/sys sysctl hook to enable timer migration
  timers: Identifying the existing pinned timers
  timers: Framework for identifying pinned timers
  timers: allow deferrable timers for intervals tv2-tv5 to be deferred

Fix up conflicts in kernel/sched.c and kernel/timer.c manually

10 years agoMerge branch 'timers-for-linus-clockevents' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Mon, 15 Jun 2009 16:58:50 +0000 (09:58 -0700)]
Merge branch 'timers-for-linus-clockevents' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-for-linus-clockevents' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  clockevent: export register_device and delta2ns
  clockevents: tick_broadcast_device can become static

10 years agoMerge branch 'timers-for-linus-clocksource' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Mon, 15 Jun 2009 16:58:33 +0000 (09:58 -0700)]
Merge branch 'timers-for-linus-clocksource' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-for-linus-clocksource' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  clocksource: prevent selection of low resolution clocksourse also for nohz=on
  clocksource: sanity check sysfs clocksource changes

10 years agoMerge branch 'timers-for-linus-ntp' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 15 Jun 2009 16:43:24 +0000 (09:43 -0700)]
Merge branch 'timers-for-linus-ntp' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-for-linus-ntp' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  ntp: fix comment typos
  ntp: adjust SHIFT_PLL to improve NTP convergence

10 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6
Linus Torvalds [Mon, 15 Jun 2009 16:40:05 +0000 (09:40 -0700)]
Merge git://git./linux/kernel/git/davem/net-next-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6: (1244 commits)
  pkt_sched: Rename PSCHED_US2NS and PSCHED_NS2US
  ipv4: Fix fib_trie rebalancing
  Bluetooth: Fix issue with uninitialized nsh.type in DTL-1 driver
  Bluetooth: Fix Kconfig issue with RFKILL integration
  PIM-SM: namespace changes
  ipv4: update ARPD help text
  net: use a deferred timer in rt_check_expire
  ieee802154: fix kconfig bool/tristate muckup
  bonding: initialization rework
  bonding: use is_zero_ether_addr
  bonding: network device names are case sensative
  bonding: elminate bad refcount code
  bonding: fix style issues
  bonding: fix destructor
  bonding: remove bonding read/write semaphore
  bonding: initialize before registration
  bonding: bond_create always called with default parameters
  x_tables: Convert printk to pr_err
  netfilter: conntrack: optional reliable conntrack event delivery
  list_nulls: add hlist_nulls_add_head and hlist_nulls_del
  ...

10 years agohwmon: (max6650) Add support for alarms
Christian Engelmayer [Mon, 15 Jun 2009 16:39:52 +0000 (18:39 +0200)]
hwmon: (max6650) Add support for alarms

Export the alarm flags provided by the MAX6650/MAX6651 fan-speed regulator
and monitor chips via sysfs.

Signed-off-by: Christian Engelmayer <christian.engelmayer@frequentis.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
10 years agohwmon: (f71882fg) Add support for the F71858F
Hans de Goede [Mon, 15 Jun 2009 16:39:52 +0000 (18:39 +0200)]
hwmon: (f71882fg) Add support for the F71858F

Add support for the hwmon part of the Fintek F71858FG superio IC to the
f71882fg driver. Many thanks to Jelle de Jong for lending me a motherboard
with this superio on it.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
10 years agohwmon: (f71882fg) Add temp#_fault sysfs attr for f8000
Hans de Goede [Mon, 15 Jun 2009 16:39:51 +0000 (18:39 +0200)]
hwmon: (f71882fg) Add temp#_fault sysfs attr for f8000

While working on f71852fg support I noticed that the f8000 sysfs attr
table was missing entries for temp#_fault, which the f8000 does have and
which we were already reading.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
10 years agohwmon: (f71882fg) Sanity check f8000 pwm settings
Hans de Goede [Mon, 15 Jun 2009 16:39:51 +0000 (18:39 +0200)]
hwmon: (f71882fg) Sanity check f8000 pwm settings

We depend up on the pwm_enable register (0x96) not containing any reserved
settings in various places. We were already checking to make sure there
were no reserved settings in the register for the f71862fg, this patch adds
the same checking for the f8000, while at it it also moves the code to
a more apropriate place so we don't need to check if the fan/pwm part
of the IC is enabled twice.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
10 years agohwmon: (f71882fg) Cleanup f8000 pwm handling
Hans de Goede [Mon, 15 Jun 2009 16:39:50 +0000 (18:39 +0200)]
hwmon: (f71882fg) Cleanup f8000 pwm handling

Currently we are using separate per model sysfs attr for the 3th pwm, because
the 3th pwm of the f8000 only has automatic mode and not manual mode. Doing
things this way was getting in the way for adding f71858fg support, so this
patch makes the pwm attr identical for all models, and instead adds a check
to store_pwm_enable() disallowing setting the 3th pwm to manual mode
on a f8000 IC.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
10 years agohwmon: PCI quirk for hwmon access on MSI MS-7031 board
Jean Delvare [Mon, 15 Jun 2009 16:39:50 +0000 (18:39 +0200)]
hwmon: PCI quirk for hwmon access on MSI MS-7031 board

The MSI MS-7031 is based on an ATI IXP300 south bridge. On this south
bridge, accessible I/O ports must be enabled explicitly. Unfortunately
the BIOS forgets to enable access to the hardware monitoring chip I/O
ports, so hardware monitoring fails.

Add a quirk enabling access to the required ports (0x295-0x296). This
is exactly what MSI's own hardware monitoring application is doing, so
it has to be the right way.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
10 years agohwmon: (w83627ehf) Add W83627DHG-P support
Jean Delvare [Mon, 15 Jun 2009 16:39:50 +0000 (18:39 +0200)]
hwmon: (w83627ehf) Add W83627DHG-P support

Add support for the new incarnation of the Winbond/Nuvoton W83627DHG
chip known as W83627DHG-P. It is basically the same as the original
W83627DHG with an additional automatic can speed control mode (not
supported by the driver yet.)

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tested-by: Madhu <madhu.chinakonda@gmail.com>