[MIPS] Update mpc30x_defconfig

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Update e55_defconfig

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Update capcella_defconfig

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] IP22: Fix modpost warning.

  MODPOST vmlinux.o
WARNING: vmlinux.o(.text+0xc70): Section mismatch: reference to .init.text:add_memory_region (between 'probe_memory' and 'enable_local0_irq')

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Gcc 3.3 build fixes.

Work around gcc 3.3's unability to evaluate that certain expressions indeed
are constant.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] MT: Use kallsyms in CPU state dump

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] MIPSsim: Delete old file that survived moving around in the tree.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] unwind_stack should return a value ...

And gcc 3.4 doesn't even warn out this, grrr.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SMTC: Fix crash on bootup with idebus= command line argument.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Fix build error if CONFIG_KALLSYMS is undefined.

  CC      arch/mips/kernel/traps.o
arch/mips/kernel/traps.c: In function 'show_backtrace':
arch/mips/kernel/traps.c:110: warning: unused variable 'ra'

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Excite: disable 64-bit kernel support.

  CC      arch/mips/basler/excite/excite_prom.o
arch/mips/basler/excite/excite_prom.c:136:3: #error 64 bit support not implemented

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Kconfig: Fix configuration warning by hardwiring HOTPLUG_CPU to n.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Update a few defconfigs.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SMTC: Move MIPS_CPU_IPI_IRQ definition into header.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Malta: Include <linux/irq.h for cp0_compare_irq / cp0_perfcount_irq.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SMP: Scatter __cpuinit over the code as needed.

MIPS doesn't do CPU hotplugging yet but since many of the functions don't
even have an __init let's fix this right.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SEAD: Don't mark as experimental.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] remove unused marvell.h

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] remove unused gt64240.h

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SNI: remove unused pcimt_scache.c

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SNI: Remove unneeded sni_machine_halt

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Remove dead code from irq_txx9.c

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Do not discard extra debugging sections.

Leaving these sections is useful to some tools that look at the image, and
none of them are loaded into memory.  The .mdebug.abi64 section, in
particular, lets GDB recognize vmlinux.32 as an N64 program instead of
guessing that it is O32.

Signed-off-by: Daniel Jacobowitz <dan@codesourcery.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Update defconfigs for TX39/TX49

Update defconfigs, disabling CONFIG_EXPERIMENTAL.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Use -Werror on TX39/TX49 boards

Now these directories can be built cleanly.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Drop unneeded config options for RBTX4938

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Cleanup TX39/TX49 irq code

Cleanup jmr3927, tx4927 and tx4938 irq codes, using common IRQ_CPU,
I8259 and IRQ_TXX9 irq routines.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] The irq_chip for TX39/TX49 SoCs

Add generic irq_chip for TX39/TX49 SoCs.  This can be replace
jmr3927_irq_irc, tx4927_irq_pic_type and tx4938_irq_pic_type.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Sibyte: CONFIG_SIBYTE_SB1250_DUART -> CONFIG_SERIAL_SB1250_DUART

This is needed since the Sibyte serial driver was exchanged.

Issue report by Imre Kaloz <kaloz@openwrt.org>.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Fix au1xxx_gpio_direction_* return value

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Remove unused pnx8550 Kconfig

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Kconfig: Enable 64BIT_PHYS_ADDR only through select.

The user should not have to have any clue about this setting.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] SMTC: Fix secondary VPE interrupt mask initialization.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Compute PGD_ORDER from the select page size.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Extract {PGD,PMD,PTE}_ORDER into asm-offset.h

For debugging purposes only.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Fix computation of {PGD,PMD,PTE}_T_LOG2.

For the generation of asm-offset.h to work these need to be evaulatable
by gcc as a constant expression.  This issue did exist for a while but
didn't bite because they're only in asm-offset.h for debugging purposes.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

[MIPS] Fix computation of PGDIR_SHIFT for 16K pagesize on 32-bit kernels.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

sched: s/sched_latency/sched_min_granularity

runtime limit and wakeup granularity used to be a function of
granularity and that was incorrect changed to sched_latency.

Fix this to make wakeup granularity a function of min-granularity,
and the runtime limit equal to latency.

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

Merge git://git./linux/kernel/git/mingo/linux-2.6-sched

* git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-sched:
  sched: cleanup, sched_granularity -> sched_min_granularity
  sched: adaptive scheduler granularity
  sched: fix CONFIG_SCHED_DEBUG dependency of lockdep sysctls

Merge branch 'for-linus' of git://git./linux/kernel/git/ieee1394/linux1394-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
  firewire: Add ref-counting for sbp2 orbs (fix command abortion)
  firewire: fix unloading of fw-ohci while devices are attached
  ieee1394: sbp2: fix sbp2_remove_device for error cases

sched: cleanup, sched_granularity -> sched_min_granularity

due to adaptive granularity scheduling the role of sched_granularity
has changed to "minimum granularity", so rename the variable (and the
tunable) accordingly.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>

sched: adaptive scheduler granularity

Instead of specifying the preemption granularity, specify the wanted
latency. By fixing the granlarity to a constany the wakeup latency
it a function of the number of running tasks on the rq.

Invert this relation.

sysctl_sched_granularity becomes a minimum for the dynamic granularity
computed from the new sysctl_sched_latency.

Then use this latency to do more intelligent granularity decisions: if
there are fewer tasks running then we can schedule coarser. This helps
performance while still always keeping the latency target.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

sched: fix CONFIG_SCHED_DEBUG dependency of lockdep sysctls

Make the lockdep sysctls not depend on CONFIG_SCHED_DEBUG.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

firewire: Add ref-counting for sbp2 orbs (fix command abortion)

This handles the case where we get the status write before getting the
complete_transaction callback ("status write for unknown orb").  In
this case, we just assume that the initial orb pointer transaction
succeeded and finish the orb.  To prevent the transaction callback
from touching freed memory, we ref-count the orb structures.

Signed-off-by: Kristian Høgsberg <krh@redhat.com>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>

firewire: fix unloading of fw-ohci while devices are attached

Fix panic in run_timer_softirq right after "modprobe -r firewire-ohci"
if a FireWire disk was attached and firewire-sbp2 loaded.

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>

ieee1394: sbp2: fix sbp2_remove_device for error cases

Bug found by Olaf Hering <olh@suse.de>:
sbp2util_remove_command_orb_pool requires a valid lu->hi pointer.

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>

Merge branch 'agp-patches' of git://git./linux/kernel/git/airlied/agp-2.6

* 'agp-patches' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/agp-2.6:
  agp: balance ioremap checks
  agp: Add device id for P4M900 to via-agp module
  efficeon-agp leaks 'struct agp_bridge_data' in error paths of agp_efficeon_probe()

Merge branch 'drm-patches' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-patches' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
  drm: ioremap return value checks
  drm/via: Fix dmablit when blit queue is full
  drm_rmmap_ioctl(): remove dead code

Merge branch 'upstream-linus' of /linux/kernel/git/jgarzik/netdev-2.6

* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6:
  myri10ge: update driver version to 1.3.2-1.269
  myri10ge: use pcie_get/set_readrq
  ehea: fix queue destructor
  ehea: fix module parameter description
  ehea: fix interface to DLPAR tools
  sgiseeq: Fix return type of sgiseeq_remove
  sky2 1.17
  sky2: only bring up watchdog if link is active
  sky2: clear PCI power control reg at startup
  DM9000: fix interface hang under load
  phy layer: fix genphy_setup_forced (don't reset)
  Don't use GFP_DMA for zone allocation.
  fix realtek phy id in forcedeth

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: (30 commits)
  ACPI: work around duplicate name "VID" problem on T61
  acpiphp_ibm: add missing '\n' to error message
  ACPI: add dump_stack() to trace acpi_format_exception programming errors
  make drivers/acpi/scan.c:create_modalias() static
  ACPI: Fix a warning of discarding qualifiers from pointer target type
  ACPI: "ACPI handle has no context!" should be KERN_DEBUG
  ACPI video hotkey: export missing ACPI video hotkey events via input layer
  ACPI: Validate XSDT, use RSDT if XSDT fails
  ACPI: /proc/acpi/thermal_zone trip points are now read-only, mark them as such
  ACPI: fix ia64 allnoconfig build
  PNP: remove null pointer checks
  PNP: remove MODULE infrastructure
  ISAPNP: removed unused isapnp_detected and ISAPNP_DEBUG
  PNPACPI: remove unnecessary casts of "void *"
  PNPACPI: simplify irq_flags()
  PNP: fix up after Lindent
  ACPI: enable GPEs before calling _WAK on resume
  asus-laptop: Fix rmmod of asus_laptop
  sony-laptop: call sonypi_compat_init earlier
  sony-laptop: enable Vaio FZ events
  ...

drm: ioremap return value checks

Signed-off-by: Scott Thompson <postfail <at> hushmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>

agp: balance ioremap checks

patchset against 2.6.23-rc3.
corrects missing ioremap return checks and balancing on iounmap calls, integrated changes per list
recommendations on the original set of patches..

Signed-off-by: Scott Thompson <postfail <at> hushmail.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>

agp: Add device id for P4M900 to via-agp module

Signed-off-by: Dave Airlie <airlied@linux.ie>

myri10ge: update driver version to 1.3.2-1.269

Update myri10ge driver version to 1.3.2-1.269.

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

myri10ge: use pcie_get/set_readrq

Based on a patch from Peter Oruba, convert myri10ge to use pcie_get_readrq()
and pcie_set_readrq() instead of our own PCI calls and arithmetics.

These driver changes incorporate the proposed PCI-X / PCI-Express read byte
count interface.  Reading and setting those values doesn't take place
"manually", instead wrapping functions are called to allow quirks for some
PCI bridges.

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off by: Peter Oruba <peter.oruba@amd.com>
Based on work by Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

ehea: fix queue destructor

Includes hcp_epas_dtor in eq/cq/qp destructors to unmap
HW register.

Signed-off-by: Jan-Bernd Themann <themann@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

ehea: fix module parameter description

Update the module parameter description of "use_mcs" to
show correct default value

Signed-off-by: Jan-Bernd Themann <themann@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

ehea: fix interface to DLPAR tools

Userspace DLPAR tool expects decimal numbers to be written to
and read from sysfs entries.

Signed-off-by: Jan-Bernd Themann <themann@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

sgiseeq: Fix return type of sgiseeq_remove

The driver remove method needs to return an int not void.  This was just
never noticed because usually this driver is not being built as a module.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

sky2 1.17

Mark new version to track if current driver is in use.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

sky2: only bring up watchdog if link is active

This fixes the extra timer overhead that people were whining about
as a 2.6.23 regression.

Running the watchdog timer all the time is unneeded. Change it
to run only if link is up, and reduce frequency to save power.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

sky2: clear PCI power control reg at startup

Make sure PCI register for PHY power gets cleared on boot, and make
sure to avoid any PCI posting problems.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

DM9000: fix interface hang under load

When transferring data at full speed, the DM9000 network interface
sometimes stops sending/receiving data. Worse, ksoftirqd consumes
100% cpu and the net tx watchdog never triggers.
Fix by spin_lock_irqsave() in dm9000_start_xmit() to prevent the
interrupt handler from interfering.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

phy layer: fix genphy_setup_forced (don't reset)

Writing BMCR_RESET bit will reset MII_BMCR to default values. This is
clearly not what we want.

Signed-off-by: Domen Puncer <domen.puncer@telargo.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

Don't use GFP_DMA for zone allocation.

IP32 doesn't even have a ZONE_DMA so no point in using GFP_DMA in any
IP32-specific device driver.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

fix realtek phy id in forcedeth

As noticed by Chuck Ebbert, commit c5e3ae8823693b260ce1f217adca8add1bc0b3de
introduced a copy-paste typo, as realtek phy is 0x732 and not 0x1c1. Obvious
fix below suggested by Ayaz Abdulla.

Signed-off-by: Willy Tarreau <w@1wt.eu>
Cc: Ayaz Abdulla <aabdulla@nvidia.com>
Cc: Chuck Ebbert <cebbert@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

Pull bugzilla-8798 into release branch

ACPI: work around duplicate name "VID" problem on T61

This can only fix the problem that more than one video bus device
have the same AML name "VID".
ie. the proc I/F for the second "VID" video bus device is located under
/proc/acpi/video/VID1/...

As this is really rare and the ACPI proc I/F is a legacy feature that
we are planning to remove.
We won't provide a generic solution for this problem.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

Pull events into release branch

Conflicts:

drivers/acpi/video.c

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

Pull misc into release branch

acpiphp_ibm: add missing '\n' to error message

Add missing \n to error in ibm_find_acpi_device.

Signed-off-by: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Cc: Adrian Bunk <trivial@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Len Brown <len.brown@intel.com>

ACPI: add dump_stack() to trace acpi_format_exception programming errors

Dump the stack so we can find the secretive caller to acpi_format_exception().

Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Len Brown <len.brown@intel.com>

make drivers/acpi/scan.c:create_modalias() static

This patch makes the needlessly global create_modalias() static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Len Brown <len.brown@intel.com>

ACPI: Fix a warning of discarding qualifiers from pointer target type

drivers/acpi/ec.c: In function `acpi_ec_ecdt_probe':
drivers/acpi/ec.c:873: warning: passing arg 1 of `acpi_get_devices' discards qualifiers from pointer target type

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Len Brown <len.brown@intel.com>

ACPI: "ACPI handle has no context!" should be KERN_DEBUG

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

Pull suspend into release branch

Pull video into release branch

Pull thermal into release branch

ACPI video hotkey: export missing ACPI video hotkey events via input layer

Signed-off-by: Yu Luming <luming.yu@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

Pull sony into release branch

Pull pnp into release branch

Pull ec into release branch

Pull bugzilla-8630 into release branch

Pull bugzilla-1641 into release branch

Pull battery into release branch

Pull asus into release branch

Merge git://git./linux/kernel/git/mingo/linux-2.6-sched

* git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-sched:
  sched: fix startup penalty calculation
  sched: simplify bonus calculation #2
  sched: simplify bonus calculation #1
  sched: tidy up and simplify the bonus balance
  sched: optimize task_tick_rt() a bit
  sched: simplify can_migrate_task()
  sched: remove HZ dependency from the granularity default
  sched: CONFIG_SCHED_GROUP_FAIR=y fixlet

Merge branch 'for-linus' of git://git./linux/kernel/git/hpa/linux-2.6-x86setup

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/hpa/linux-2.6-x86setup:
  [x86 setup] Make sure AH=00h when setting a video mode
  [x86 setup] Volatilize asm() statements

DM_MULTIPATH_RDAC: "scsi_normalize_sense" undefined

DM_MULTIPATH_RDAC uses SCSI API(s) and is for a SCSI device,
so add SCSI to its depends on to prevent build errors.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
[ Tested and Verified by Chandra Seetharaman ]
Acked-by: Chandra Seetharaman <sekharan@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ACPI: Validate XSDT, use RSDT if XSDT fails

ACPI 1.0 used an RSDT with 32-bit physical addresses.
ACPI 2.0 adds an XSDT with 32-bit physical addresses.
An ACPI 2.0 aware OS is supposed to use the XSDT
(when present) instead of the RSDT.

However, several systems have failed because the XSDT
contains NULL entries -- while it is missing pointers
to needed tables, such as SSDTs.

When we find an XSDT with NULL entries, discard it
and use the ACPI 1.0 RSDT instead.

http://bugzilla.kernel.org/show_bug.cgi?id=8630

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

ACPI: /proc/acpi/thermal_zone trip points are now read-only, mark them as such

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

sched: fix startup penalty calculation

fix task startup penalty miscalculation: sysctl_sched_granularity is
unsigned int and wait_runtime is long so we first have to convert it
to long before turning it negative ...

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

sched: simplify bonus calculation #2

current code:

 delta = calc_delta_mine(delta_exec, curr->load.weight, lw);
 delta = min((u64)delta, cfs_rq->sleeper_bonus);

Notice that this calc_delta_mine() line is exactly delta_mine, which
gives:

 delta = min((u64)delta_mine, cfs_rq->sleeper_bonus);

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

sched: simplify bonus calculation #1

current code:

 delta = min(cfs_rq->sleeper_bonus, (u64)delta_exec);
 delta = calc_delta_mine(delta, curr->load.weight, lw);
 delta = min((u64)delta, cfs_rq->sleeper_bonus);

drop the first min(), because we clip against sleeper_bonus in the 3rd line
again. That gives:

 delta = calc_delta_mine(delta_exec, curr->load.weight, lw);
 delta = min((u64)delta, cfs_rq->sleeper_bonus);

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

sched: tidy up and simplify the bonus balance

make the bonus balance more consistent: do not hand out a bonus if
there's too much in flight already, and only deduct as much from a
runner as it has the capacity. This makes the bonus engine a zero-sum
game (as intended).

this also simplifies the code:

   text    data     bss     dec     hex filename
  34770    2998      24   37792    93a0 sched.o.before
  34749    2998      24   37771    938b sched.o.after

and it also avoids overscheduling in sleep-happy workloads like
hackbench.c.

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

sched: optimize task_tick_rt() a bit

Mitchell Erblich suggested a quality-of-implementation change to
not requeue SCHED_RR tasks if there's only a single task on the
runqueue, by checking for rq->nr_running == 1.

provide a more efficient implementation of that, to check that
particular RT priority-queue only.

[ From: mingo@elte.hu ]

Also first requeue the task then set need_resched - results in slightly
better machine-instruction ordering. Also clean up the code a bit.

Signed-off-by: Dmitry Adamushko <dmitry.adamushko@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

sched: simplify can_migrate_task()

Remove trivial conditional branch in Linux scheduler's
can_migrate_task() function.

   text    data     bss     dec     hex filename
   34770    2998      24   37792    93a0 sched.o.before
   34757    2998      24   37779    9393 sched.o.after

Signed-off-by: Sven-Thorsten Dietrich <sven@thebigcorporation.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

sched: remove HZ dependency from the granularity default

remove HZ dependency from the granularity default. Use 10 msec for
the base granularity, 1 msec for wakeup granularity and 25 msec for
batch wakeup granularity. (These defaults are close to the values
that the default HZ=250 setting got previously, and thus it's the
most common setting.)

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

sched: CONFIG_SCHED_GROUP_FAIR=y fixlet

when I built with CONFIG_FAIR_GROUP_SCHED=y, I need the following change
to make things right.

[ From: mingo@elte.hu ]

this config option is not upstream-configurable right now but lets fix
this for completeness.

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

ACPI: fix ia64 allnoconfig build

drivers/acpi/event.c:238: error: conflicting types for ‘acpi_bus_generate_netlink_event’
include/acpi/acpi_bus.h:324: error: previous declaration of ‘acpi_bus_generate_netlink_event’ was here

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