12 years agoMerge branch 'sched/urgent' into sched/core
Ingo Molnar [Mon, 8 Dec 2008 12:52:00 +0000]
Merge branch 'sched/urgent' into sched/core

12 years agosched: add uid information to sched_debug for CONFIG_USER_SCHED
Arun R Bharadwaj [Mon, 1 Dec 2008 15:19:05 +0000]
sched: add uid information to sched_debug for CONFIG_USER_SCHED

Impact: extend information in /proc/sched_debug

This patch adds uid information in sched_debug for CONFIG_USER_SCHED

Signed-off-by: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agosched: don't export sched_mc_power_savings in laptops
Mahesh Salgaonkar [Sat, 29 Nov 2008 11:20:12 +0000]
sched: don't export sched_mc_power_savings in laptops

Impact: do not expose a control that has no effect

Fix to prevent sched_mc_power_saving from being exported through sysfs
on single-socket systems. (Say multicore single socket (Laptop))

CPU core map of the boot cpu should be equal to possible number
of cpus for single socket system.

This fix has been developed at FOSS.in kernel workout.

Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agoMerge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied...
Linus Torvalds [Mon, 1 Dec 2008 00:45:13 +0000]
Merge branch 'drm-fixes' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
  drm/i915: Fix copy'n'pasteo that broke VT switch if flushing was non-empty.

12 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc
Linus Torvalds [Mon, 1 Dec 2008 00:44:18 +0000]
Merge branch 'merge' of git://git./linux/kernel/git/paulus/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc:
  powerpc: Fix system calls on Cell entered with XER.SO=1
  powerpc/cell: Fix GDB watchpoints, again
  powerpc/mpic: Don't reset affinity for secondary MPIC on boot
  powerpc/cell/axon-msi: Retry on missing interrupt
  powerpc: Fix boot freeze on machine with empty memory node
  powerpc: Fix IRQ assignment for some PCIe devices
  powerpc/spufs: Fix spinning in spufs_ps_fault on signal
  powerpc/mpc832x_rdb: fix swapped ethernet ids
  powerpc: Use generic PHY driver for Marvell 88E1111 PHY on GE Fanuc SBC610
  powerpc/85xx: L2 cache size wrong in 8572DS dts
  powerpc/virtex: Update defconfigs
  powerpc/52xx: update defconfigs
  xsysace: Fix driver to use resource_size_t instead of unsigned long
  powerpc/virtex: fix various format/casting printk mismatches
  powerpc/mpc5200: fix bestcomm Kconfig dependencies
  powerpc/44x: Fix 460EX/460GT machine check handling
  powerpc/40x: Limit allocable DRAM during early mapping

12 years agoMerge master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Mon, 1 Dec 2008 00:39:06 +0000]
Merge master.kernel.org:/home/rmk/linux-2.6-arm

* master.kernel.org:/home/rmk/linux-2.6-arm:
  Allow architectures to override copy_user_highpage()
  [ARM] pxa/palmtx: misc fixes to use generic GPIO API
  ARM: OMAP: Fixes for suspend / resume GPIO wake-up handling
  [ARM] pxa/corgi: update default config to exclude tosa from being built
  [ARM] pxa/pcm990: use negative number for an invalid GPIO in camera data
  ARM: OMAP: Typo fix for clock_allow_idle
  ARM: OMAP: Remove broken LCD driver for SX1
  [ARM] 5335/1: pxa25x_udc: Fix is_vbus_present to return 1 or 0
  [ARM] pxa/MioA701: bluetooth resume fix
  [ARM] pxa/MioA701: fix memory corruption.

12 years agodrm/i915: Fix copy'n'pasteo that broke VT switch if flushing was non-empty.
Eric Anholt [Mon, 1 Dec 2008 00:23:21 +0000]
drm/i915: Fix copy'n'pasteo that broke VT switch if flushing was non-empty.

Introduced in the "Avoid BUG_ONs on VT switch" commit.

Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>

12 years agopowerpc: Fix system calls on Cell entered with XER.SO=1
Paul Mackerras [Sun, 30 Nov 2008 11:49:45 +0000]
powerpc: Fix system calls on Cell entered with XER.SO=1

It turns out that on Cell, on a kernel with CONFIG_VIRT_CPU_ACCOUNTING
= y, if a program sets the SO (summary overflow) bit in the XER and
then does a system call, the SO bit in CR0 will be set on return
regardless of whether the system call detected an error.  Since CR0.SO
is used as the error indication from the system call, this means that
all system calls appear to fail.

The reason is that the workaround for the timebase bug on Cell uses a
compare instruction.  With CONFIG_VIRT_CPU_ACCOUNTING = y, the
ACCOUNT_CPU_USER_ENTRY macro reads the timebase, so we end up doing a
compare instruction, which copies XER.SO to CR0.SO.  Since we were
doing this in the system call entry patch after clearing CR0.SO but
before saving the CR, this meant that the saved CR image had CR0.SO
set if XER.SO was set on entry.

This fixes it by moving the clearing of CR0.SO to after the
ACCOUNT_CPU_USER_ENTRY call in the system call entry path.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

12 years agopowerpc/cell: Fix GDB watchpoints, again
Arnd Bergmann [Fri, 28 Nov 2008 09:51:24 +0000]
powerpc/cell: Fix GDB watchpoints, again

An earlier patch from Jens Osterkamp attempted to fix GDB
watchpoints by enabling the DABRX register at boot time.
Unfortunately, this did not work on SMP setups, where
secondary CPUs were still using the power-on DABRX value.

This introduces the same change for secondary CPUs on cell
as well.

Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
Tested-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>

12 years agopowerpc/mpic: Don't reset affinity for secondary MPIC on boot
Arnd Bergmann [Fri, 28 Nov 2008 09:51:23 +0000]
powerpc/mpic: Don't reset affinity for secondary MPIC on boot

Kexec/kdump currently fails on the IBM QS2x blades when the kexec happens
on a CPU other than the initial boot CPU.  It turns out that this is the
result of mpic_init trying to set affinity of each interrupt vector to the
current boot CPU.

As far as I can tell,  the same problem is likely to exist on any
secondary MPIC, because they have to deliver interrupts to the first
output all the time. There are two potential solutions for this: either
not set up affinity at all for secondary MPICs, or assume that a single
CPU output is connected to the upstream interrupt controller and hardcode
affinity to that per architecture.

This patch implements the second approach, defaulting to the first output.
Currently, all known secondary MPICs are routed to their upstream port
using the first destination, so we hardcode that.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>

12 years agopowerpc/cell/axon-msi: Retry on missing interrupt
Arnd Bergmann [Fri, 28 Nov 2008 09:51:22 +0000]
powerpc/cell/axon-msi: Retry on missing interrupt

The MSI capture logic for the axon bridge can sometimes
lose interrupts in case of high DMA and interrupt load,
when it signals an MSI interrupt to the MPIC interrupt
controller while we are already handling another MSI.

Each MSI vector gets written into a FIFO buffer in main
memory using DMA, and that DMA access is normally flushed
by the actual interrupt packet on the IOIF.  An MMIO
register in the MSIC holds the position of the last
entry in the FIFO buffer that was written.  However,
reading that position does not flush the DMA, so that
we can observe stale data in the buffer.

In a stress test, we have observed the DMA to arrive
up to 14 microseconds after reading the register.

This patch works around this problem by retrying the
access to the FIFO buffer.

We can reliably detect the conditioning by writing
an invalid MSI vector into the FIFO buffer after
reading from it, assuming that all MSIs we get
are valid.  After detecting an invalid MSI vector,
we udelay(1) in the interrupt cascade for up to
100 times before giving up.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>

12 years agopowerpc: Fix boot freeze on machine with empty memory node
Dave Hansen [Mon, 24 Nov 2008 12:02:35 +0000]
powerpc: Fix boot freeze on machine with empty memory node

I got a bug report about a distro kernel not booting on a particular
machine.  It would freeze during boot:

> ...
> Could not find start_pfn for node 1
> [boot]0015 Setup Done
> Built 2 zonelists in Node order, mobility grouping on.  Total pages: 123783
> Policy zone: DMA
> Kernel command line:
> [boot]0020 XICS Init
> [boot]0021 XICS Done
> PID hash table entries: 4096 (order: 12, 32768 bytes)
> clocksource: timebase mult[7d0000] shift[22] registered
> Console: colour dummy device 80x25
> console handover: boot [udbg0] -> real [hvc0]
> Dentry cache hash table entries: 1048576 (order: 7, 8388608 bytes)
> Inode-cache hash table entries: 524288 (order: 6, 4194304 bytes)
> freeing bootmem node 0

I've reproduced this on 2.6.27.7.  It is caused by commit
8f64e1f2d1e09267ac926e15090fd505c1c0cbcb ("powerpc: Reserve in bootmem
lmb reserved regions that cross NUMA nodes").

The problem is that Jon took a loop which was (in pseudocode):

for_each_node(nid)
NODE_DATA(nid) = careful_alloc(nid);
setup_bootmem(nid);
reserve_node_bootmem(nid);

and broke it up into:

for_each_node(nid)
NODE_DATA(nid) = careful_alloc(nid);
setup_bootmem(nid);
for_each_node(nid)
reserve_node_bootmem(nid);

The issue comes in when the 'careful_alloc()' is called on a node with
no memory.  It falls back to using bootmem from a previously-initialized
node.  But, bootmem has not yet been reserved when Jon's patch is
applied.  It gives back bogus memory (0xc000000000000000) and pukes
later in boot.

The following patch collapses the loop back together.  It also breaks
the mark_reserved_regions_for_nid() code out into a function and adds
some comments.  I think a huge part of introducing this bug is because
for loop was too long and hard to read.

The actual bug fix here is the:

+ if (end_pfn <= node->node_start_pfn ||
+     start_pfn >= node_end_pfn)
+ continue;

Signed-off-by: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>

12 years agopowerpc: Fix IRQ assignment for some PCIe devices
Adhemerval Zanella [Wed, 19 Nov 2008 03:55:35 +0000]
powerpc: Fix IRQ assignment for some PCIe devices

Currently, some PCIe devices on POWER6 machines do not get interrupts
assigned correctly.  The problem is that OF doesn't create an
"interrupt" property for them.  The fix is for of_irq_map_pci to fall
back to using the value in the PCI interrupt-pin register in config
space, as we do when there is no OF device-tree node for the device.

I have verified that this works fine with a pair of Squib-E SAS
adapter on a P6-570.

Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: Paul Mackerras <paulus@samba.org>

12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/kyle/parisc-2.6
Linus Torvalds [Sun, 30 Nov 2008 22:04:31 +0000]
Merge git://git./linux/kernel/git/kyle/parisc-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/kyle/parisc-2.6:
  parisc: struct device - replace bus_id with dev_name(), dev_set_name()
  parisc: fix kernel crash when unwinding a userspace process
  parisc: __kernel_time_t is always long

12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6
Linus Torvalds [Sun, 30 Nov 2008 22:04:02 +0000]
Merge git://git./linux/kernel/git/sfrench/cifs-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
  [CIFS] fix regression in cifs_write_begin/cifs_write_end

12 years agoMerge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 30 Nov 2008 21:06:47 +0000]
Merge branch 'sched-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  sched: prevent divide by zero error in cpu_avg_load_per_task, update
  sched, cpusets: fix warning in kernel/cpuset.c
  sched: prevent divide by zero error in cpu_avg_load_per_task

12 years agoMerge branch 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 30 Nov 2008 21:06:20 +0000]
Merge branch 'irq-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  irq.h: fix missing/extra kernel-doc
  genirq: __irq_set_trigger: change pr_warning to pr_debug
  irq: fix typo
  x86: apic honour irq affinity which was set in early boot
  genirq: fix the affinity setting in setup_irq
  genirq: keep affinities set from userspace across free/request_irq()

12 years agoMerge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 30 Nov 2008 21:05:46 +0000]
Merge branch 'core-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  lockdep: consistent alignement for lockdep info

12 years agoMerge branch 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 30 Nov 2008 21:05:31 +0000]
Merge branch 'tracing-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  ftrace: prevent recursion
  tracing, doc: update mmiotrace documentation
  x86, mmiotrace: fix buffer overrun detection
  function tracing: fix wrong position computing of stack_trace

12 years agoMerge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 30 Nov 2008 21:01:04 +0000]
Merge branch 'x86-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86: always define DECLARE_PCI_UNMAP* macros
  x86: fixup config space size of CPU functions for AMD family 11h
  x86, bts: fix wrmsr and spinlock over kmalloc
  x86, pebs: fix PEBS record size configuration
  x86, bts: turn macro into static inline function
  x86, bts: exclude ds.c from build when disabled
  arch/x86/kernel/pci-calgary_64.c: change simple_strtol to simple_strtoul
  x86: use limited register constraint for setnz
  xen: pin correct PGD on suspend
  x86: revert irq number limitation
  x86: fixing __cpuinit/__init tangle, xsave_cntxt_init()
  x86: fix __cpuinit/__init tangle in init_thread_xstate()
  oprofile: fix an overflow in ppro code

12 years agoMerge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied...
Linus Torvalds [Sun, 30 Nov 2008 21:00:21 +0000]
Merge branch 'drm-fixes' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
  drm/i915: Save/restore HWS_PGA on suspend/resume
  drm: move drm vblank initialization/cleanup to driver load/unload
  drm/i915: execbuffer pins objects, no need to ensure they're still in the GTT
  drm/i915: Always read pipestat in irq_handler
  drm/i915: Subtract total pinned bytes from available aperture size
  drm/i915: Avoid BUG_ONs on VT switch with a wedged chipset.
  drm/i915: Remove IMR masking during interrupt handler, and restart it if needed.
  drm/i915: Manage PIPESTAT to control vblank interrupts instead of IMR.

12 years agoMerge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux...
Linus Torvalds [Sun, 30 Nov 2008 20:34:22 +0000]
Merge branch 'for_linus' of git://git./linux/kernel/git/jack/linux-udf-2.6

* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-udf-2.6:
  udf: Fix BUG_ON() in destroy_inode()

12 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/davej/cpufreq
Linus Torvalds [Sun, 30 Nov 2008 19:43:41 +0000]
Merge branch 'fixes' of git://git./linux/kernel/git/davej/cpufreq

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/davej/cpufreq:
  [CPUFREQ] powernow-k8: ignore out-of-range PstateStatus value
  [CPUFREQ] Documentation: Add Blackfin to list of supported processors

12 years agoTrivial Documentation/filesystems/ramfs-rootfs-initramfs.txt fix
frans [Sat, 22 Nov 2008 14:39:06 +0000]
Trivial Documentation/filesystems/ramfs-rootfs-initramfs.txt fix

A very minor patch on ramfs-rootfs-initramfs.txt: update the location
where CONFIG_INITRAMFS_SOURCE lives in menuconfig

Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
Acked-by: Rob Landley <rob@landley.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Sun, 30 Nov 2008 19:36:57 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Check model for Dell 92HD73xx laptops
  ALSA: hda - mark Dell studio 1535 quirk
  ALSA: hda - No 'Headphone as Line-out' swich without line-outs
  ALSA: hda - Fix AFG power management on IDT 92HD* codecs
  ALSA: hda - Fix caching of SPDIF status bits
  ALSA: hda - Add a quirk for Dell Studio 15
  ALSA: hda: Add STAC_DELL_M4_3 quirk
  sound/sound_core: Fix sparse warnings
  ALSA: hda: STAC_DELL_M6 EAPD

12 years agoMerge branch 'linux-next' of git://git.infradead.org/~dedekind/ubi-2.6
Linus Torvalds [Sun, 30 Nov 2008 19:34:17 +0000]
Merge branch 'linux-next' of git://git.infradead.org/~dedekind/ubi-2.6

* 'linux-next' of git://git.infradead.org/~dedekind/ubi-2.6:
  UBI: Don't exit from ubi_thread until kthread_should_stop() is true
  UBI: fix EBADMSG handling

12 years agoMerge branch 'i2c-for-linus' of git://jdelvare.pck.nerim.net/jdelvare-2.6
Linus Torvalds [Sun, 30 Nov 2008 19:21:43 +0000]
Merge branch 'i2c-for-linus' of git://jdelvare.pck.nerim.net/jdelvare-2.6

* 'i2c-for-linus' of git://jdelvare.pck.nerim.net/jdelvare-2.6:
  i2c-parport: Fix misplaced parport_release call
  i2c: Remove i2c clients in reverse order
  i2c/isp1301_omap: Build fixes

12 years agoMN10300: Tighten up the code using case ranges
Robert P. J. Day [Fri, 28 Nov 2008 11:48:37 +0000]
MN10300: Tighten up the code using case ranges

Compress a set of consecutive switch cases into a case-range.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoMerge branch 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6
Linus Torvalds [Sun, 30 Nov 2008 19:07:16 +0000]
Merge branch 'for-linus' of git://git390.osdl.marist.edu/linux-2.6

* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6:
  [S390] Update default configuration.
  [S390] Fix alignment of initial kernel stack.
  [S390] pgtable.h: Fix oops in unmap_vmas for KVM processes
  [S390] fix/cleanup sched_clock
  [S390] fix system call parameter functions.

12 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux...
Linus Torvalds [Sun, 30 Nov 2008 19:06:40 +0000]
Merge branch 'release' of git://git./linux/kernel/git/lenb/linux-acpi-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6:
  toshiba_acpi: close race in toshiba_acpi driver
  ACPICA: disable _BIF warning
  ACPI: delete OSI(Linux) DMI dmesg spam
  ACPICA: Allow _WAK method to return an Integer
  ACPI: thinkpad-acpi: fix fan sleep/resume path
  sony-laptop: printk tweak
  sony-laptop: brightness regression fix
  Revert "ACPI: don't enable control method power button as wakeup device when Fixed Power button is used"
  ACPI suspend: Blacklist boxes that require us to set SCI_EN directly on resume
  ACPI: scheduling in atomic via acpi_evaluate_integer ()
  ACPI: battery: Convert discharge energy rate to current properly
  ACPI: EC: count interrupts only if called from interrupt handler.

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
Linus Torvalds [Sun, 30 Nov 2008 19:05:21 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/dtor/input

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: wacom - add support for new USB Tablet PCs
  Input: replace spin_lock_bh with spin_lock_irqsave in ml_ff_playback
  Input: i8042 - add Compal Hel80 laptop to nomux blacklist
  Input: cm109 - add keymap for ATCom AU-100 phone
  Input: fix the example of an input device driver
  Input: psmouse - fix incorrect validate_byte check in OLPC protocol
  Input: atkbd - cancel delayed work before freeing its structure
  Input: atkbd - add keymap quirk for Inventec Symphony systems
  Input: i8042 - add Dell XPS M1530 to nomux list
  Input: elo - fix format string in elo driver

12 years agoremove __ARCH_WANT_COMPAT_SYS_PTRACE
Christoph Hellwig [Tue, 25 Nov 2008 07:10:03 +0000]
remove __ARCH_WANT_COMPAT_SYS_PTRACE

All architectures now use the generic compat_sys_ptrace, as should every
new architecture that needs 32bit compat (if we'll ever get another).

Remove the now superflous __ARCH_WANT_COMPAT_SYS_PTRACE define, and also
kill a comment about __ARCH_SYS_PTRACE that was added after
__ARCH_SYS_PTRACE was already gone.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoMerge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Sun, 30 Nov 2008 18:38:22 +0000]
Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  MIPS: Make BUG() __noreturn.

12 years agoinput: replace spin_lock_bh with spin_lock_irqsave in ml_ff_playback
Arjan van de Ven [Mon, 24 Nov 2008 00:57:36 +0000]
input: replace spin_lock_bh with spin_lock_irqsave in ml_ff_playback

ml_ff_playback() uses spin_(un)lock_bh. However this function is called
with interrupts disabled from erase_effect() in drivers/input/ff-core.c:196.

This is not permitted, and will result in a WARN_ON in the bottom half handling code.
This patch changes this function to just use spin_lock_irqsave() instead, solving
the problem and simplifying the locking logic.

This was reported as entry #106559 in kerneloops.org

Reported-by: kerneloops.org
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agohotplug_memory_notifier section annotation
Al Viro [Sat, 22 Nov 2008 17:38:34 +0000]
hotplug_memory_notifier section annotation

Same as for hotplug_cpu - we want static notifier_block in there in meminitdata,
to avoid false positives whenever it's used.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agosn_pci_controller_fixup() should be __init
Al Viro [Sat, 22 Nov 2008 17:38:24 +0000]
sn_pci_controller_fixup() should be __init

called only from __init

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoget xenbus_driver ->probe() "recognized" by modpost
Al Viro [Sat, 22 Nov 2008 17:38:14 +0000]
get xenbus_driver ->probe() "recognized" by modpost

... by giving the instances' names magic suffix recognized by modpost ;-/
Their ->probe() is __devinit

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoxen_play_dead() is __cpuinit
Al Viro [Sat, 22 Nov 2008 17:38:04 +0000]
xen_play_dead() is __cpuinit

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoxen_setup_vcpu_info_placement() is not init on x86
Al Viro [Sat, 22 Nov 2008 17:37:54 +0000]
xen_setup_vcpu_info_placement() is not init on x86

... so get xen-ops.h in agreement with xen/smp.c

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agokvm_setup_secondary_clock() is cpuinit
Al Viro [Sat, 22 Nov 2008 17:37:44 +0000]
kvm_setup_secondary_clock() is cpuinit

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoenable_IR_x2apic() needs to be __init
Al Viro [Sat, 22 Nov 2008 17:37:34 +0000]
enable_IR_x2apic() needs to be __init

calls __init, called only from __init

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agopci_setup() is init, not devinit
Al Viro [Sat, 22 Nov 2008 17:37:14 +0000]
pci_setup() is init, not devinit

for fsck sake, it's used only when parsing kernel command line...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoalpha: pcibios_resource_to_bus() is callable from normal code
Al Viro [Sat, 22 Nov 2008 17:37:04 +0000]
alpha: pcibios_resource_to_bus() is callable from normal code

pci_enable_rom(), specifically.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agotricky one: hisax sections
Al Viro [Sat, 22 Nov 2008 17:36:54 +0000]
tricky one: hisax sections

a) hisax_init_pcmcia() needs to be defined only if we have
   CONFIG_HOTPLUG (no PCMCIA support otherwise) and can be declared
   __devinit.

b) HiSax_inithardware() can go __init

c) hisax_register() is passing to checkcard() full-blown hisax_cs_setup_card():
checkcard(i, id, NULL, hisax_d_if->owner, hisax_cs_setup_card);
   The problem with it is that
* hisax_cs_setup_card() is __devinit
* hisax_register() is not
* hisax_cs_setup_card() is a switch from hell, calling a lot of
  setup_some_weirdcard() depending on card->typ.  _These_ are also
  __devinit.

   However, in hisax_register() we have card->typ equal to
   ISDN_CTYPE_DYNAMIC, which reduces hisax_cs_setup_card() to "nevermind
   all that crap, just do nothing and return 2".  So we add a
   trimmed-down callback doing just that and passed to checkcard() by
   hisax_register().  _This_ is non-init (we can stand the impact on
   .text size).

Voila - no section warnings from drivers/isdn

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agocpuinit fixes in kernel/*
Al Viro [Sat, 22 Nov 2008 17:36:44 +0000]
cpuinit fixes in kernel/*

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agouninorth-agp section mess
Al Viro [Sat, 22 Nov 2008 17:36:34 +0000]
uninorth-agp section mess

'aperture' is declared devinitdata (the whole word of it) and
is used from ->fetch_size() which can, AFAICS, be used on
!HOTPLUG after init time.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agorapidio section noise
Al Viro [Sat, 22 Nov 2008 17:36:24 +0000]
rapidio section noise

functions calling devinit and called only from devinit

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agosection errors in smc911x/smc91x
Al Viro [Sat, 22 Nov 2008 17:36:14 +0000]
section errors in smc911x/smc91x

a) ->probe() can be __devinit; no need to put it into .text
 b) calling __init stuff from it, OTOH, is wrong
 c) ->remove() is __devexit fodder

Acked-by: rmk+kernel@arm.linux.org.uk
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agofix the section noise in sparc head.S
Al Viro [Sat, 22 Nov 2008 17:36:04 +0000]
fix the section noise in sparc head.S

usual .text.head trick

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agom32r: section noise in head.S
Al Viro [Sat, 22 Nov 2008 17:35:54 +0000]
m32r: section noise in head.S

usual "introduce .text.head, put it in front of TEXT_TEXT in vmlinux.lds.S,
make the stuff up to jump to start_kernel live in it", same as on other
targets.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agosection misannotation in ibmtr_cs
Al Viro [Sat, 22 Nov 2008 17:35:44 +0000]
section misannotation in ibmtr_cs

ibmtr_resume() is calling ibmtr_probe(), which is devinit.  Whether
that's the right thing to do there is a separate question, but
since it's PCMCIA and thus will never compile without HOTPLUG...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoixgbe section fixes
Al Viro [Sat, 22 Nov 2008 17:35:34 +0000]
ixgbe section fixes

ixgbe_init_interrupt_scheme() is called from ixgbe_resume().  Build that
with CONFIG_PM and without CONFIG_HOTPLUG and you've got a problem.
Several helpers called by it also are misannotated __devinit.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agorackmeter section fixes
Al Viro [Sat, 22 Nov 2008 17:35:24 +0000]
rackmeter section fixes

* rackmeter_remove() reference needs devexit_p
 * rackmeter_setup() is calls devinit and is called only from devinit

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agogdth section fixes
Al Viro [Sat, 22 Nov 2008 17:35:14 +0000]
gdth section fixes

PCI side of driver should be devinit, not init

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoof_platform_driver noise on sparce
Al Viro [Sat, 22 Nov 2008 17:35:04 +0000]
of_platform_driver noise on sparce

switch to __init for those; unlike powerpc sparc has no hotplug support
for that stuff and their ->probe() tends to call __init functions while
being declared __devinit.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoadvansys fix on ISA-less configs
Al Viro [Sat, 22 Nov 2008 17:34:54 +0000]
advansys fix on ISA-less configs

The code

        if (shost->dma_channel != NO_ISA_DMA)
                free_dma(shost->dma_channel);

in there is triggerable only if we have CONFIG_ISA (we only set ->dma_channel to
something other than NO_ISA_DMA under #ifdef CONFIG_ISA).  OTOH, free_dma() is
not guaranteed to be there in absense of CONFIG_ISA.  IOW, driver runs into
undefined symbols on PCI-but-not-ISA configs (e.g. on frv) and it's a false
positive.

Fix: put the entire if () under #ifdef CONFIG_ISA; behaviour doesn't change and
dependency on free_dma() disappears for !CONFIG_ISA.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoW1_MASTER_DS1WM should depend on HAVE_CLK
Al Viro [Sat, 22 Nov 2008 17:34:44 +0000]
W1_MASTER_DS1WM should depend on HAVE_CLK

Uses clk_...() a lot

Acked-by: rmk+kernel@arm.linux.org.uk
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoicside section warnings
Al Viro [Sat, 22 Nov 2008 17:34:34 +0000]
icside section warnings

icside_register_v[56] is called from (__devinit) icside_probe

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agofix talitos
Al Viro [Sat, 22 Nov 2008 17:34:24 +0000]
fix talitos

talitos_remove() can be called from talitos_probe() on failure
exit path, so it can't be __devexit.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agoistallion section warnings
Al Viro [Sat, 22 Nov 2008 17:34:14 +0000]
istallion section warnings

stli_findeisabrds() and stli_initbrds() are using __init and called only
from __init.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agosparc64 trivial section misannotations
Al Viro [Sat, 22 Nov 2008 17:34:04 +0000]
sparc64 trivial section misannotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agosparc32 cpuinit flase positives
Al Viro [Sat, 22 Nov 2008 17:33:54 +0000]
sparc32 cpuinit flase positives

All noise since we don't have CPU hotplug there.  However, they
did expose something very odd-looking in there - poke_viking()
does a bunch of identical btfixup each time it's called (i.e.
for each CPU).  That one is left alone for now; just the trivial
misannotation fixes.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agopowerpc set_huge_psize() false positive
Al Viro [Sat, 22 Nov 2008 17:33:44 +0000]
powerpc set_huge_psize() false positive

called only from __init, calls __init.  Incidentally, it ought to be static
in file.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agofalse __cpuinit positives on alpha
Al Viro [Sat, 22 Nov 2008 17:33:34 +0000]
false __cpuinit positives on alpha

pure noise - alpha doesn't have CPU hotplug

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agomeminit section warnings
Al Viro [Sat, 22 Nov 2008 17:33:24 +0000]
meminit section warnings

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

12 years agosched: prevent divide by zero error in cpu_avg_load_per_task, update
Ingo Molnar [Sat, 29 Nov 2008 19:45:15 +0000]
sched: prevent divide by zero error in cpu_avg_load_per_task, update

Regarding the bug addressed in:

  4cd4262: sched: prevent divide by zero error in cpu_avg_load_per_task

Linus points out that the fix is not complete:

> There's nothing that keeps gcc from deciding not to reload
> rq->nr_running.
>
> Of course, in _practice_, I don't think gcc ever will (if it decides
> that it will spill, gcc is likely going to decide that it will
> literally spill the local variable to the stack rather than decide to
> reload off the pointer), but it's a valid compiler optimization, and
> it even has a name (rematerialization).
>
> So I suspect that your patch does fix the bug, but it still leaves the
> fairly unlikely _potential_ for it to re-appear at some point.
>
> We have ACCESS_ONCE() as a macro to guarantee that the compiler
> doesn't rematerialize a pointer access. That also would clarify
> the fact that we access something unsafe outside a lock.

So make sure our nr_running value is immutable and cannot change
after we check it for nonzero.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agosched, cpusets: fix warning in kernel/cpuset.c
Ingo Molnar [Tue, 25 Nov 2008 09:27:49 +0000]
sched, cpusets: fix warning in kernel/cpuset.c

this warning:

  kernel/cpuset.c: In function ‘generate_sched_domains’:
  kernel/cpuset.c:588: warning: ‘ndoms’ may be used uninitialized in this function

triggers because GCC does not recognize that ndoms stays uninitialized
only if doms is NULL - but that flow is covered at the end of
generate_sched_domains().

Help out GCC by initializing this variable to 0. (that's prudent anyway)

Also, this function needs a splitup and code flow simplification:
with 160 lines length it's clearly too long.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agotoshiba_acpi: close race in toshiba_acpi driver
Arjan van de Ven [Fri, 28 Nov 2008 16:19:09 +0000]
toshiba_acpi: close race in toshiba_acpi driver

the toshiba ACPI driver will, in a failure case, free the rfkill state
before stopping the polling timer that would use this state. More interesting,
in the same failure case handling, it calls the exit function, which also
frees the rfkill state, but after stopping the polling.

If the race happens, a NULL pointer is passed to rfkill_force_state()
which then causes a nice dereference.

Fix the race by just not doing the too-early freeing of the rfkill state.

This appears to be the cause of a hot issue on kerneloops.org; while I
have no solid evidence of that this patch will fix the issue, the race
appears rather real.

Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agosched: move double_unlock_balance() higher
Alexey Dobriyan [Fri, 28 Nov 2008 19:08:00 +0000]
sched: move double_unlock_balance() higher

Move double_lock_balance()/double_unlock_balance() higher to fix the following
with gcc-3.4.6:

CC kernel/sched.o
 In file included from kernel/sched.c:1605:
 kernel/sched_rt.c: In function `find_lock_lowest_rq':
 kernel/sched_rt.c:914: sorry, unimplemented: inlining failed in call to 'double_unlock_balance': function body not available
 kernel/sched_rt.c:1077: sorry, unimplemented: called from here
 make[2]: *** [kernel/sched.o] Error 1

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agoMerge branch 'sched/urgent' into sched/core
Ingo Molnar [Fri, 28 Nov 2008 19:11:05 +0000]
Merge branch 'sched/urgent' into sched/core

12 years agoi2c-parport: Fix misplaced parport_release call
Jean Delvare [Fri, 28 Nov 2008 14:24:39 +0000]
i2c-parport: Fix misplaced parport_release call

We shouldn't release the parallel port until we are actually done with
it.

Signed-off-by: Jean Delvare <khali@linux-fr.org>

12 years agoi2c: Remove i2c clients in reverse order
Jean Delvare [Fri, 28 Nov 2008 14:24:38 +0000]
i2c: Remove i2c clients in reverse order

i2c clients should be removed in reverse order compared to the probe
(actually: bind) order. This matters when several clients depend on
each other.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tested-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>

12 years agoi2c/isp1301_omap: Build fixes
David Brownell [Fri, 28 Nov 2008 14:24:38 +0000]
i2c/isp1301_omap: Build fixes

Build fixes for isp1301_omap; no behavior changes:

  - fix incorrect probe() signature (it changed many months ago)
  - provide missing functions on H3 and H4 boards
  - "sparse" fixes (static, NULL-vs-0)

The H3 build bits subset some of the stuff that was previously in
the OMAP tree but never went to mainline.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>

12 years agoMerge branch 'for-rmk' of git://git.kernel.org/pub/scm/linux/kernel/git/ycmiao/pxa...
Russell King [Thu, 27 Nov 2008 23:50:31 +0000]
Merge branch 'for-rmk' of git://git./linux/kernel/git/ycmiao/pxa-linux-2.6

12 years agoAllow architectures to override copy_user_highpage()
Russell King [Thu, 27 Nov 2008 11:13:58 +0000]
Allow architectures to override copy_user_highpage()

With aliasing VIPT cache support, the ARM implementation of
clear_user_page() and copy_user_page() sets up a temporary kernel space
mapping such that we have the same cache colour as the userspace page.
This avoids having to consider any userspace aliases from this operation.

However, when highmem is enabled, kmap_atomic() have to setup mappings.
The copy_user_highpage() and clear_user_highpage() call these functions
before delegating the copies to copy_user_page() and clear_user_page().

The effect of this is that each of the *_user_highpage() functions setup
their own kmap mapping, followed by the *_user_page() functions setting
up another mapping.  This is rather wasteful.

Thankfully, copy_user_highpage() can be overriden by architectures by
defining __HAVE_ARCH_COPY_USER_HIGHPAGE.  However, replacement of
clear_user_highpage() is more difficult because its inline definition
is not conditional.  It seems that you're expected to define
__HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE and provide a replacement
__alloc_zeroed_user_highpage() implementation instead.

The allocation itself is fine, so we don't want to override that.  What
we really want to do is to override clear_user_highpage() with our own
version which doesn't kmap_atomic() unnecessarily.

Other VIPT architectures (PARISC and SH) would also like to override
this function as well.

Acked-by: Hugh Dickins <hugh@veritas.com>
Acked-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Acked-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

12 years agoudf: Fix BUG_ON() in destroy_inode()
Jan Kara [Tue, 23 Sep 2008 16:24:08 +0000]
udf: Fix BUG_ON() in destroy_inode()

udf_clear_inode() can leave behind buffers on mapping's i_private list (when
we truncated preallocation). Call invalidate_inode_buffers() so that the list
is properly cleaned-up before we return from udf_clear_inode(). This is ugly
and suggest that we should cleanup preallocation earlier than in clear_inode()
but currently there's no such call available since drop_inode() is called under
inode lock and thus is unusable for disk operations.

Signed-off-by: Jan Kara <jack@suse.cz>

12 years ago[ARM] pxa/palmtx: misc fixes to use generic GPIO API
Marek Vasut [Thu, 20 Nov 2008 16:34:57 +0000]
[ARM] pxa/palmtx: misc fixes to use generic GPIO API

Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Signed-off-by: Eric Miao <eric.miao@marvell.com>

12 years agox86: always define DECLARE_PCI_UNMAP* macros
Joerg Roedel [Thu, 20 Nov 2008 19:49:56 +0000]
x86: always define DECLARE_PCI_UNMAP* macros

Impact: fix boot crash on AMD IOMMU if CONFIG_GART_IOMMU is off

Currently these macros evaluate to a no-op except the kernel is compiled
with GART or Calgary support. But we also need these macros when we have
SWIOTLB, VT-d or AMD IOMMU in the kernel. Since we always compile at
least with SWIOTLB we can define these macros always.

This patch is also for stable backport for the same reason the SWIOTLB
default selection patch is.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Cc: <stable@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agoMerge branch 'omap-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind...
Russell King [Thu, 27 Nov 2008 11:13:10 +0000]
Merge branch 'omap-fixes' of git://git./linux/kernel/git/tmlind/linux-omap-2.6

12 years ago[S390] Update default configuration.
Martin Schwidefsky [Thu, 27 Nov 2008 10:05:59 +0000]
[S390] Update default configuration.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

12 years ago[S390] Fix alignment of initial kernel stack.
Heiko Carstens [Thu, 27 Nov 2008 10:05:58 +0000]
[S390] Fix alignment of initial kernel stack.

We need an alignment of 16384 bytes for the initial kernel stack if
the kernel is configured for 16384 bytes stacks but the linker script
currently guarantees only an alignment of 8192 bytes.

So fix this and simply use THREAD_SIZE as alignment value which will
always do the right thing.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

12 years ago[S390] pgtable.h: Fix oops in unmap_vmas for KVM processes
Christian Borntraeger [Thu, 27 Nov 2008 10:05:57 +0000]
[S390] pgtable.h: Fix oops in unmap_vmas for KVM processes

When running several kvm processes with lots of memory overcommitment,
we have seen an oops during process shutdown:
------------[ cut here ]------------
Kernel BUG at 0000000000193434 [verbose debug info unavailable]
addressing exception: 0005 [#1] PREEMPT SMP
Modules linked in: kvm sunrpc qeth_l2 dm_mod qeth ccwgroup
CPU: 10 Not tainted 2.6.28-rc4-kvm-bigiron-00521-g0ccca08-dirty #8
Process kuli (pid: 14460, task: 0000000149822338, ksp: 0000000024f57650)
Krnl PSW : 0704e00180000000 0000000000193434 (unmap_vmas+0x884/0xf10)
R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:2 PM:0 EA:3
Krnl GPRS: 0000000000000002 0000000000000000 000000051008d000 000003e05e6034e0
           00000000001933f6 00000000000001e9 0000000407259e0a 00000002be88c400
           00000200001c1000 0000000407259608 0000000407259e08 0000000024f577f0
           0000000407259e09 0000000000445fa8 00000000001933f6 0000000024f577f0
Krnl Code: 0000000000193426: eb22000c000d sllg %r2,%r2,12
           000000000019342c: a7180000 lhi %r1,0
           0000000000193430: b2290012 iske %r1,%r2
          >0000000000193434: a7110002 tmll %r1,2
           0000000000193438: a7840006 brc 8,193444
           000000000019343c: 9602c000 oi 0(%r12),2
           0000000000193440: 96806000 oi 0(%r6),128
           0000000000193444: a7110004 tmll %r1,4
Call Trace:
([<00000000001933f6>] unmap_vmas+0x846/0xf10)
[<0000000000199680>] exit_mmap+0x210/0x458
[<000000000012a8f8>] mmput+0x54/0xfc
[<000000000012f714>] exit_mm+0x134/0x144
[<000000000013120c>] do_exit+0x240/0x878
[<00000000001318dc>] do_group_exit+0x98/0xc8
[<000000000013e6b0>] get_signal_to_deliver+0x30c/0x358
[<000000000010bee0>] do_signal+0xec/0x860
[<0000000000112e30>] sysc_sigpending+0xe/0x22
[<000002000013198a>] 0x2000013198a
INFO: lockdep is turned off.
Last Breaking-Event-Address:
[<00000000001a68d0>] free_swap_and_cache+0x1a0/0x1a4
<4>---[ end trace bc19f1d51ac9db7c ]---

The faulting instruction is the storage key operation (iske) in
ptep_rcp_copy (called by pte_clear, called by unmap_vmas). iske
reads dirty and reference bit information for a physical page and
requires a valid physical address. Since we are in pte_clear, we
cannot rely on the pte containing a valid address. Fortunately we
dont need these information in pte_clear - after all there is no
mapping. The best fix is to remove the needless call to ptep_rcp_copy
that contains the iske.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

12 years ago[S390] fix/cleanup sched_clock
Christian Borntraeger [Thu, 27 Nov 2008 10:05:56 +0000]
[S390] fix/cleanup sched_clock

CONFIG_PRINTK_TIME reveals that sched_clock has a wrong offset during boot:
..
[    0.000000]   Movable zone: 0 pages used for memmap
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 775679
[    0.000000] Kernel command line: dasd=4b6c root=/dev/dasda1 ro noinitrd
[    0.000000] PID hash table entries: 4096 (order: 12, 32768 bytes)
[6920575.975232] console [ttyS0] enabled
[6920575.987586] Dentry cache hash table entries: 524288 (order: 10, 4194304 bytes)
[6920575.991404] Inode-cache hash table entries: 262144 (order: 9, 2097152 bytes)
..

The s390 implementation of sched_clock uses the store clock instruction and
subtracts jiffies_timer_cc.
jiffies_timer_cc is a local variable in arch/s390/kernel/time.c and only used
for sched_clock and monotonic clock. For historical reasons there is an offset
on that value. With todays code this offset is unnecessary. By removing that
offset we can get a sched_clock which returns the nanoseconds after time_init.
This improves CONFIG_PRINTK_TIME.

Since sched_clock is the only user, I have also renamed jiffies_timer_cc to
sched_clock_base_cc. In addition, the local variable init_timer_cc is redundant
and can be romved as well.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

12 years ago[S390] fix system call parameter functions.
Martin Schwidefsky [Thu, 27 Nov 2008 10:05:55 +0000]
[S390] fix system call parameter functions.

syscall_get_nr() currently returns a valid result only if the call
chain of the traced process includes do_syscall_trace_enter(). But
collect_syscall() can be called for any sleeping task, the result of
syscall_get_nr() in general is completely bogus.

To make syscall_get_nr() work for any sleeping task the traps field
in pt_regs is replace with svcnr - the system call number the process
is executing. If svcnr == 0 the process is not on a system call path.

The syscall_get_arguments and syscall_set_arguments use regs->gprs[2]
for the first system call parameter. This is incorrect since gprs[2]
may have been overwritten with the system call number if the call
chain includes do_syscall_trace_enter. Use regs->orig_gprs2 instead.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

12 years agosched: prevent divide by zero error in cpu_avg_load_per_task
Steven Rostedt [Thu, 27 Nov 2008 02:04:24 +0000]
sched: prevent divide by zero error in cpu_avg_load_per_task

Impact: fix divide by zero crash in scheduler rebalance irq

While testing the branch profiler, I hit this crash:

divide error: 0000 [#1] PREEMPT SMP
[...]
RIP: 0010:[<ffffffff8024a008>]  [<ffffffff8024a008>] cpu_avg_load_per_task+0x50/0x7f
[...]
Call Trace:
 <IRQ> <0> [<ffffffff8024fd43>] find_busiest_group+0x3e5/0xcaa
 [<ffffffff8025da75>] rebalance_domains+0x2da/0xa21
 [<ffffffff80478769>] ? find_next_bit+0x1b2/0x1e6
 [<ffffffff8025e2ce>] run_rebalance_domains+0x112/0x19f
 [<ffffffff8026d7c2>] __do_softirq+0xa8/0x232
 [<ffffffff8020ea7c>] call_softirq+0x1c/0x3e
 [<ffffffff8021047a>] do_softirq+0x94/0x1cd
 [<ffffffff8026d5eb>] irq_exit+0x6b/0x10e
 [<ffffffff8022e6ec>] smp_apic_timer_interrupt+0xd3/0xff
 [<ffffffff8020e4b3>] apic_timer_interrupt+0x13/0x20

The code for cpu_avg_load_per_task has:

if (rq->nr_running)
rq->avg_load_per_task = rq->load.weight / rq->nr_running;

The runqueue lock is not held here, and there is nothing that prevents
the rq->nr_running from going to zero after it passes the if condition.

The branch profiler simply made the race window bigger.

This patch saves off the rq->nr_running to a local variable and uses that
for both the condition and the division.

Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agoftrace: prevent recursion
Lai Jiangshan [Thu, 27 Nov 2008 02:21:46 +0000]
ftrace: prevent recursion

Impact: prevent unnecessary stack recursion

if the resched flag was set before we entered, then don't reschedule.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

12 years agoACPICA: disable _BIF warning
Lin Ming [Thu, 27 Nov 2008 06:42:30 +0000]
ACPICA: disable _BIF warning

A generic work-around from ACPICA is in the queue,
but since Linux has a work-around in its battery
driver, we can disable this warning now.

Allow _BIF method to return an Package with Buffer elements

http://bugzilla.kernel.org/show_bug.cgi?id=11822

Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoACPI: delete OSI(Linux) DMI dmesg spam
Len Brown [Sat, 8 Nov 2008 06:21:10 +0000]
ACPI: delete OSI(Linux) DMI dmesg spam

Linux will continue to ignore OSI(Linux),
except for a white-list containing a few systems.

So delete the black-list,
and stop soliciting user-feedback on the console.

Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoACPICA: Allow _WAK method to return an Integer
Bob Moore [Thu, 13 Nov 2008 03:01:34 +0000]
ACPICA: Allow _WAK method to return an Integer

This can happen if the _WAK method returns nothing (as per ACPI
1.0) but does return an integer if the implicit return mechanism
is enabled.  This is the only method that has this problem,
since it is also defined to return a package of two integers
(ACPI 1.0b+). In all other cases, if a method returns an object
when one was not expected, no warning is issued.

Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoACPI: thinkpad-acpi: fix fan sleep/resume path
Henrique de Moraes Holschuh [Sun, 9 Nov 2008 12:54:02 +0000]
ACPI: thinkpad-acpi: fix fan sleep/resume path

This fixes a regression from v2.6.27, caused by commit
5814f737e1cd2cfa2893badd62189acae3e1e1fd, "ACPI: thinkpad-acpi:
attempt to preserve fan state on resume".

It is possible for fan_suspend() to fail to properly initialize
fan_control_desired_level as required by fan_resume(), resulting on
the fan always being set to level 7 on resume if the user didn't
touch the fan controller.

In order to get fan sleep/resume handling to work right:

1. Fix the fan_suspend handling of the T43 firmware quirk. If it is
still undefined, we didn't touch the fan yet and that means we have no
business doing it on resume.

2. Store the fan level on its own variable to avoid any possible
issues with hijacking fan_control_desired_level (which isn't supposed
to have anything other than 0-7 in it, anyway).

3. Change the fan_resume code to me more straightforward to understand
(although we DO optimize the boolean logic there, otherwise it looks
disgusting).

4. Add comments to help understand what the code is supposed to be
doing.

5. Change fan_set_level to be less strict about how auto and
full-speed modes are requested.

http://bugzilla.kernel.org/show_bug.cgi?id=11982

Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Reported-by: Tino Keitel <tino.keitel@tikei.de>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agosony-laptop: printk tweak
Alessandro Guido [Wed, 12 Nov 2008 22:13:35 +0000]
sony-laptop: printk tweak

There's no need to print "Sony: " just after "sony-laptop: " (DRV_PFX).

Signed-off-by: Alessandro Guido <ag@alessandroguido.name>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agosony-laptop: brightness regression fix
Alessandro Guido [Wed, 12 Nov 2008 22:03:28 +0000]
sony-laptop: brightness regression fix

After commit 540b8bb9c33935183ceb5bed466a42ad72b2af56:

  sony-laptop: fingers off backlight if video.ko is serving this functionality

I can't set brightness on my sony laptop (nothing in /sys/class/backlight).
dmesg says "sony-laptop: Sony: Brightness ignored, must be controlled by ACPI
video driver".

The function acpi_video_backlight_support returns 0 if we should use the
vendor-specific backlight support, while non-0 if the ACPI generic should
be used. Because of this, the check introduced by the said commit appears
reversed.

Signed-off-by: Alessandro Guido <ag@alessandroguido.name>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoRevert "ACPI: don't enable control method power button as wakeup device when Fixed...
Len Brown [Wed, 26 Nov 2008 22:55:15 +0000]
Revert "ACPI: don't enable control method power button as wakeup device when Fixed Power button is used"

This reverts commit faee816b1502385dc9bc5abf2960d1cc645844d1.

http://bugzilla.kernel.org/show_bug.cgi?id=12091

Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoACPI suspend: Blacklist boxes that require us to set SCI_EN directly on resume
Rafael J. Wysocki [Wed, 26 Nov 2008 22:53:13 +0000]
ACPI suspend: Blacklist boxes that require us to set SCI_EN directly on resume

Some Apple boxes evidently require us to set SCI_EN on resume
directly, because if we don't do that, they hung somewhere in the
resume code path.  Moreover, on these boxes it is not sufficient to
use acpi_enable() to turn ACPI on during resume.  All of this is
against the ACPI specification which states that (1) the BIOS is
supposed to return from the S3 sleep state with ACPI enabled
(SCI_EN set) and (2) the SCI_EN bit is owned by the hardware and we
are not supposed to change it.

For this reason, blacklist the affected systems so that the SCI_EN
bit is set during resume on them.

[NOTE: Unconditional setting SCI_EN for all system on resume doesn't
 work, because it makes some other systems crash (that's to be
 expected).  Also, it is not entirely clear right now if all of the
 Apple boxes require this workaround.]

This patch fixes the recent regression tracked as
http://bugzilla.kernel.org/show_bug.cgi?id=12038

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Tested-by: Tino Keitel <tino.keitel@gmx.de>
Tested-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoACPI: scheduling in atomic via acpi_evaluate_integer ()
Pavel Machek [Tue, 25 Nov 2008 11:05:08 +0000]
ACPI: scheduling in atomic via acpi_evaluate_integer ()

Now I know why I had strange "scheduling in atomic" problems:
acpi_evaluate_integer() does malloc(..., irqs_disabled() ? GFP_ATOMIC
: GFP_KERNEL)... which is (of course) broken.

There's no way to reliably tell if we need GFP_ATOMIC or not from
code, this one for example fails to detect spinlocks held.

Fortunately, allocation seems small enough to be done on stack.

Signed-off-by: Pavel Machek <pavel@suse.cz>
Acked-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoARM: OMAP: Fixes for suspend / resume GPIO wake-up handling
Tero Kristo [Wed, 26 Nov 2008 22:35:16 +0000]
ARM: OMAP: Fixes for suspend / resume GPIO wake-up handling

Use the correct wake-up enable register, and make it
work with 34xx also.

Signed-off-by: Tero Kristo <tero.kristo@nokia.com>
Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

12 years agoACPI: battery: Convert discharge energy rate to current properly
Alexey Starikovskiy [Fri, 21 Nov 2008 14:41:01 +0000]
ACPI: battery: Convert discharge energy rate to current properly

ACPI battery interface reports its state either in mW or in mA, and
discharge rate in your case is reported in mW. power_supply interface
does not have such a parameter, so current_now parameter is used
for all cases. But in case of mW, reported discharge should
be converted into mA.

Signed-off-by: Alexey Starikovskiy <astarikovskiy@suse.de>
Tested-by: Ferenc Wagner <wferi@niif.hu>
Signed-off-by: Len Brown <len.brown@intel.com>

12 years agoparisc: struct device - replace bus_id with dev_name(), dev_set_name()
Kay Sievers [Fri, 7 Nov 2008 00:42:46 +0000]
parisc: struct device - replace bus_id with dev_name(), dev_set_name()

(I did not compile or test it, please let me know, or help fixing
 it, if something is wrong with the conversion)

This patch is part of a larger patch series which will remove
the "char bus_id[20]" name string from struct device. The device
name is managed in the kobject anyway, and without any size
limitation, and just needlessly copied into "struct device".

To set and read the device name dev_name(dev) and dev_set_name(dev)
must be used. If your code uses static kobjects, which it shouldn't
do, "const char *init_name" can be used to statically provide the
name the registered device should have. At registration time, the
init_name field is cleared, to enforce the use of dev_name(dev) to
access the device name at a later time.

We need to get rid of all occurrences of bus_id in the entire tree
to be able to enable the new interface. Please apply this patch,
and possibly convert any remaining remaining occurrences of bus_id.

We want to submit a patch to -next, which will remove bus_id from
"struct device", to find the remaining pieces to convert, and finally
switch over to the new api, which will remove the 20 bytes array
and does no longer have a size limitation.

Thanks,
Kay

Cc: Matthew Wilcox <matthew@wil.cx>
Cc: linux-parisc@vger.kernel.org
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>

12 years agoparisc: fix kernel crash when unwinding a userspace process
Helge Deller [Wed, 26 Nov 2008 20:46:22 +0000]
parisc: fix kernel crash when unwinding a userspace process

Any user on existing parisc 32- and 64bit-kernels can easily crash
the kernel and as such enforce a DSO.
A simple testcase is available here:
        http://gsyprf10.external.hp.com/~deller/crash.tgz

The problem is introduced by the fact, that the handle_interruption()
crash handler calls the show_regs() function, which in turn tries to
unwind the stack by calling parisc_show_stack().  Since the stack contains
userspace addresses, a try to unwind the stack is dangerous and useless
and leads to the crash.

The fix is trivial: For userspace processes
a) avoid to unwind the stack, and
b) avoid to resolve userspace addresses to kernel symbol names.

While touching this code, I converted print_symbol() to %pS
printk formats and made parisc_show_stack() static.

An initial patch for this was written by Kyle McMartin back in August:
http://marc.info/?l=linux-parisc&m=121805168830283&w=2

Compile and run-tested with a 64bit parisc kernel.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: Grant Grundler <grundler@parisc-linux.org>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x, 2.6.27.x, earlier...]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>

12 years agoparisc: __kernel_time_t is always long
Geert Uytterhoeven [Sun, 16 Nov 2008 11:04:13 +0000]
parisc: __kernel_time_t is always long

__kernel_time_t is always long on PA-RISC, irrespective of CONFIG_64BIT,
hence move it out of the #ifdef CONFIG_64BIT / #else / #endif block.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>