platform/kernel/linux-rpi.git
17 years ago[CPUFREQ] ondemand: fix tickless accounting and software coordination bug
Venki Pallipadi [Wed, 20 Jun 2007 21:26:24 +0000 (14:26 -0700)]
[CPUFREQ] ondemand: fix tickless accounting and software coordination bug

With tickless kernel and software coordination os P-states, ondemand
can look at wrong idle statistics. This can happen when ondemand sampling
is happening on CPU 0 and due to software coordination sampling also looks at
utilization of CPU 1. If CPU 1 is in tickless state at that moment, its idle
statistics will not be uptodate and CPU 0 thinks CPU 1 is idle for less
amount of time than it actually is.

This can be resolved by looking at all the busy times of CPUs, which is
accurate, even with tickless, and use that to determine idle time in a
round about way (total time - busy time).

Thanks to Arjan for originally reporting the ondemand bug on
Lenovo T61.

Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] ondemand: add a check to avoid negative load calculation
Venki Pallipadi [Wed, 20 Jun 2007 21:24:52 +0000 (14:24 -0700)]
[CPUFREQ] ondemand: add a check to avoid negative load calculation

Due to rounding and inexact jiffy accounting, idle_ticks can sometimes
be higher than total_ticks. Make sure those cases are handled as
zero load case.

Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Keep userspace governor quiet when it is not being used
Venki Pallipadi [Wed, 20 Jun 2007 21:24:00 +0000 (14:24 -0700)]
[CPUFREQ] Keep userspace governor quiet when it is not being used

Userspace governor registers a frequency change notifier at init time, even
when no CPU is set to userspace governor. Make it register only when
atleast one CPU is using userspace.

Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Proper register access
Rafał Bilski [Thu, 7 Jun 2007 20:31:24 +0000 (22:31 +0200)]
[CPUFREQ] Longhaul - Proper register access

In previous commit I used u32 for u16 register.
This code will work only when ACPI block address is set.
For now it is only for VT8235 and VT8237.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Kconfig powernow-k8 driver should depend on ACPI P-States driver
Joshua Hoblitt [Tue, 22 May 2007 02:47:43 +0000 (16:47 -1000)]
[CPUFREQ] Kconfig powernow-k8 driver should depend on ACPI P-States driver

powernow-k8 really needs to use ACPI to function on SMP systems.
The current Kconfig allows us to build kernels which fail mysteriously
for some users due to us trying to automatically enable this, and
getting it wrong.  It's easier to just present this as an option
to the user.

Signed-off-by: Joshua Hoblitt <jhoblitt@ifa.hawaii.edu>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Replace ACPI functions with direct I/O
Rafał Bilski [Tue, 5 Jun 2007 20:08:50 +0000 (22:08 +0200)]
[CPUFREQ] Longhaul - Replace ACPI functions with direct I/O

Current version of "bm status" bit test works as long as
no USB device is in use. When USB device is plugged in ACPI
function in this context is always returning 1. Until reboot.
Direct I/O is working fine even when many USB devices are
connected.
Change bm_timeout value to less annoying. 1000 is still much
more then worst case observed and it is much better when status
bit gets stuck.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Remove duplicate multipliers
Rafał Bilski [Mon, 28 May 2007 19:58:09 +0000 (21:58 +0200)]
[CPUFREQ] Longhaul - Remove duplicate multipliers

Remove duplicate multipliers in clock_ratio table. On 1,4GHz
Nehemiah two frequencies are present twice in table. It isn't
fatal, but with voltage scaling enabled each will be set twice.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Embedded "conservative"
Rafał Bilski [Mon, 28 May 2007 19:56:19 +0000 (21:56 +0200)]
[CPUFREQ] Longhaul - Embedded "conservative"

Longhaul with voltage scaling enabled works great on Ezra
CPU (Longhaul ver. 2). As long as "conservative" governor is
used. Both "ondemand" and "userspace" can change voltage
from min to max at once. Motherboard unfortunatly turns off
when vid difference is big. Longhaul was printing warning
message, but it is not enough. Now driver will have
"conservative" governor built in and will split bigger
changes to smaller ones.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] acpi-cpufreq: Proper ReadModifyWrite of PERF_CTL MSR
Venki Pallipadi [Wed, 23 May 2007 22:42:13 +0000 (15:42 -0700)]
[CPUFREQ] acpi-cpufreq: Proper ReadModifyWrite of PERF_CTL MSR

During recent acpi-cpufreq changes, writing to PERF_CTL msr
changed from RMW of entire 64 bit to RMW of low 32 bit and clearing of
upper 32 bit. Fix it back to do a proper RMW of the MSR.

Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] check return value of sysfs_create_file
Thomas Renninger [Mon, 21 May 2007 12:20:04 +0000 (07:20 -0500)]
[CPUFREQ] check return value of sysfs_create_file

Eliminate build warning (sysfs_create_file return value must be checked)

Signed-off-by: Thomas Renninger <trenn@suse.de>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Check ACPI "BM DMA in progress" bit
Rafał Bilski [Thu, 17 May 2007 20:39:02 +0000 (22:39 +0200)]
[CPUFREQ] Longhaul - Check ACPI "BM DMA in progress" bit

It is good idea to wait for PCI bus to become idle before
frequency change. Thanks to ACPI it is possible. It makes
sense only when northbridge support is in use because it is
only case in which we can disable arbiter after check if PCI
bus is busy.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Move old_ratio to correct place
Rafał Bilski [Thu, 17 May 2007 20:36:42 +0000 (22:36 +0200)]
[CPUFREQ] Longhaul - Move old_ratio to correct place

Move one line where it should be. After first transition
Longhaul will skip frequency transition if destination
frequency is already set.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - VT8237 support
Rafał Bilski [Thu, 17 May 2007 20:35:29 +0000 (22:35 +0200)]
[CPUFREQ] Longhaul - VT8237 support

Looks like VT8237 has the same bits which VT8235 has.
Poke registers if it is found.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Longhaul - Use all kinds of support
Rafał Bilski [Thu, 17 May 2007 20:33:46 +0000 (22:33 +0200)]
[CPUFREQ] Longhaul - Use all kinds of support

This patch is removing southbridge support as separate
kind of support. Instead it is used to make other kinds
of support more stable. Also northbridge and ACPI C3
support both will be used if both are available.

Signed-off-by: Rafal Bilski <rafalbilski@interia.pl>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] powernow-k8: clarify number of cores.
Dave Jones [Fri, 18 May 2007 17:22:28 +0000 (13:22 -0400)]
[CPUFREQ] powernow-k8: clarify number of cores.

Indicate number of processors and cores more cleanly
in startup messages.

Signed-off-by: Mark Langsdorf <mark.langsdorf@amd.com>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Correct revision mask for powernow-k8
Dave Jones [Mon, 14 May 2007 22:27:29 +0000 (18:27 -0400)]
[CPUFREQ] Correct revision mask for powernow-k8

Mark Langsdorf points out that the correct define for this
revision bump is 0x80000.  Also to save us having to keep
renaming the #define, give it a more meaningful name.

Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] powernow-k7: fix MHz rounding issue with perflib
Daniel Drake [Wed, 2 May 2007 22:19:05 +0000 (23:19 +0100)]
[CPUFREQ] powernow-k7: fix MHz rounding issue with perflib

When the PST tables are broken, powernow-k7 uses ACPI's processor_perflib to
deduce the available frequency multipliers from the _PSS tables.

Upon frequency change, processor_perflib performs some verification on the
frequency (checks that it's within allowable bounds).

powernow-k7 deals with absolute frequencies in KHz, whereas perflib only
deals with MHz values. When performing the above verification, perflib
multiplies the MHz values by 1000 to obtain the KHz value.

We then end up with situations like the following:
 - powernow-k7 multiplies the multiplier by the FSB, and obtains a value
   such as 1266768 KHz
 - perflib belives the same state has frequency of 1266 MHz
 - acpi_processor_ppc_notifier calls cpufreq_verify_within_limits to verify
   that 1266768 is in the allowable range of 0 to 1266000 (i.e. 1266 * 1000)
 - it's not, so that frequency is rejected
 - the maximum CPU frequency is not reachable

This patch solves the problem by rounding up the MHz values stored in perflib's
tables. Additionally it corrects a broken URL.

It also fixes http://bugzilla.kernel.org/show_bug.cgi?id=8255 although this
case is a bit different: the frequencies in the _PSS tables are wildly wrong,
but we get better results if we force ACPI to respect the fsb * multiplier
calculations (even though it seems that the multiplier values aren't entirely
correct either).

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years ago[CPUFREQ] Support rev H AMD64s in powernow-k8
Dave Jones [Sun, 13 May 2007 15:55:14 +0000 (11:55 -0400)]
[CPUFREQ] Support rev H AMD64s in powernow-k8

Reported-by: Calvin Dodge <caldodge@gmail.com>
Signed-off-by: Dave Jones <davej@redhat.com>
17 years agoLinux 2.6.22-rc1
Linus Torvalds [Sun, 13 May 2007 01:45:56 +0000 (18:45 -0700)]
Linux 2.6.22-rc1

.. close the merge window

17 years agoMerge branch 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Sun, 13 May 2007 01:11:33 +0000 (18:11 -0700)]
Merge branch 'for-linus' of /home/rmk/linux-2.6-arm

* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm: (30 commits)
  [ARM] Use new get_irqnr_preamble
  [ARM] Ensure machine class menu is sorted alphabetically
  [ARM] 4333/2: KS8695: Micrel Development board
  [ARM] 4332/2: KS8695: Serial driver
  [ARM] 4331/3: Support for Micrel/Kendin KS8695 processor
  [ARM] 4371/1: AT91: Support for Atmel AT91SAM9RL-EK development board
  [ARM] 4372/1: Define byte sizes in asm-arm/sizes.h
  [ARM] 4370/3: AT91: Support for Atmel AT91SAM9RL processors.
  [ARM] Update mach-types
  [ARM] export symbol csum_partial_copy_from_user
  [ARM] iop13xx: msi support
  [ARM] stacktrace fix
  [ARM] Spinlock initializer cleanup
  [ARM] remove useless config option GENERIC_BUST_SPINLOCK
  [ARM] 4303/3: base kernel support for TI DaVinci
  [ARM] 4369/1: AT91: Fix circular dependency in header files
  [ARM] 4368/1: S3C24xx: build fix
  [ARM] 4364/1: AT91: LEDS on AT91SAM9261-EK
  [ARM] Fix iop32x/iop33x build
  [ARM] EBSA110: fix build errors caused by missing "const"
  ...

17 years agoinclude/linux: trivial repair whitespace damage
Daniel Walker [Sat, 12 May 2007 23:28:35 +0000 (16:28 -0700)]
include/linux: trivial repair whitespace damage

Adding tabs where spaces currently are.

Signed-off-by: Daniel Walker <dwalker@mvista.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agofix the dynamic allocation and probe in loop.c
Al Viro [Sat, 12 May 2007 20:23:15 +0000 (16:23 -0400)]
fix the dynamic allocation and probe in loop.c

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Ken Chen <kenchen@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoSLUB: i386 support
Christoph Lameter [Sat, 12 May 2007 18:15:24 +0000 (11:15 -0700)]
SLUB: i386 support

SLUB cannot run on i386 at this point because i386 uses the page->private and
page->index field of slab pages for the pgd cache.

Make SLUB run on i386 by replacing the pgd slab cache with a quicklist.
Limit the changes as much as possible. Leave the improvised linked list in place
etc etc. This has been working here for a couple of weeks now.

Acked-by: William Lee Irwin III <wli@holomorphy.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agocompat signalfd and timerfd are cond syscalls
Heiko Carstens [Sat, 12 May 2007 17:37:02 +0000 (10:37 -0700)]
compat signalfd and timerfd are cond syscalls

Add missing cond_syscall statements for compat_sys_signalfd and
compat_sys_timerfd.

Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoIPMI: Add PPC openfirmware unregister
Christian Krafft [Sat, 12 May 2007 17:37:01 +0000 (10:37 -0700)]
IPMI: Add PPC openfirmware unregister

When trying to load the ipmi_si module on a powerpc with no BMC (baseboard
management controller) the driver failes to load correctly, but doesn't
unregister itself from of_platform.  So, on a second modprobe the kernel
crashes.  This patch adds the missing unregister call.

Signed-off-by: Christian Krafft <krafft@de.ibm.com>
Signed-off-by: Corey Minyard <minyard@acm.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoIPMI: fix SI address space settings
Corey Minyard [Sat, 12 May 2007 17:36:58 +0000 (10:36 -0700)]
IPMI: fix SI address space settings

Fix a rather obvious error that Patrick found in the setup routines.  Need to
set the proper address space in the ACPI case.

Signed-off-by: Corey Minyard <minyard@acm.org>
Cc: Patrick Schoeller <Patrick.Schoeller@hp.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoMAINTAINERS: remove invalid list address for TPM
Stephen Rothwell [Sat, 12 May 2007 17:36:57 +0000 (10:36 -0700)]
MAINTAINERS: remove invalid list address for TPM

This address bounces with "550 Unknown user".

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Kylene Hall <kjhall@us.ibm.com>
Cc: Marcel Selhorst <tpm@selhorst.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoMissing include file in tpm_atmel.h
Stephen Rothwell [Sat, 12 May 2007 17:36:56 +0000 (10:36 -0700)]
Missing include file in tpm_atmel.h

On PPC64, we need to include asm/prom.h for function definitions.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agotty: flush flip buffer on ldisc input queue flush
Paul Fulghum [Sat, 12 May 2007 17:36:55 +0000 (10:36 -0700)]
tty: flush flip buffer on ldisc input queue flush

Flush the tty flip buffer when the line discipline input queue is flushed,
including the user call tcflush(TCIFLUSH/TCIOFLUSH).  This prevents
unexpected stale data after a user application calls tcflush().

Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Antonino Ingargiola <tritemio@gmail.com>
Signed-off-by: Paul Fulghum <paulkf@microgate.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agontfs: use zero_user_page
Nate Diller [Sat, 12 May 2007 17:36:54 +0000 (10:36 -0700)]
ntfs: use zero_user_page

Use zero_user_page() instead of open-coding it.

[akpm@linux-foundation.org: kmap-type fixes]
Signed-off-by: Nate Diller <nate.diller@gmail.com>
Acked-by: Anton Altaparmakov <aia21@cantab.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agofault injection: disable stacktrace filter for x86-64
Akinobu Mita [Sat, 12 May 2007 17:36:53 +0000 (10:36 -0700)]
fault injection: disable stacktrace filter for x86-64

Disable stacktrace filter support for x86-64 for now.  Will be enable when we
can get the dwarf2 unwinder back.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6
Linus Torvalds [Sat, 12 May 2007 16:47:47 +0000 (09:47 -0700)]
Merge /pub/scm/linux/kernel/git/davem/sparc-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6:
  [SPARC]: Wire up signalfd/timerfd/eventfd syscalls.
  [SPARC64]: Add support for bq4802 TOD chip, as found on ultra45.
  [SPARC64]: Correct FIRE_IOMMU_FLUSHINV register offset.
  [SPARC64]: envctrl.c needs asm/io.h
  [SPARC64]: Update defconfig.
  [TTY]: Export proc_clear_tty() to modulea.
  [SPARC64]: pci_resource_adjust() cannot be __init.
  [SPARC64]: Spelling fixes.
  [SPARC]: Spelling fixes.
  [SPARC64]: Kill LARGE_ALLOCS and update defconfig.

17 years agox86_64: use signalfd and timerfd compat syscalls
Heiko Carstens [Sat, 12 May 2007 15:27:41 +0000 (17:27 +0200)]
x86_64: use signalfd and timerfd compat syscalls

Looks like these two are wired up in a wrong way.

Cc: Davide Libenzi <davidel@xmailserver.org>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agox86_64: Add asm/mtrr.h include for some builds
Andi Kleen [Sat, 12 May 2007 15:27:40 +0000 (17:27 +0200)]
x86_64: Add asm/mtrr.h include for some builds

The earlier change to call the bp mtrr init from bugs.c broke
on some configurations due to missing includes.  Noticed
by "Avuton Olrich" <avuton@gmail.com>

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agopm3fb: Fix compile error if module
Antonino A. Daplas [Sat, 12 May 2007 14:03:48 +0000 (22:03 +0800)]
pm3fb: Fix compile error if module

pm3fb_setup() should only compiled if pm3fb is built statically in the kernel.

Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agoMerge branch 'fixes' into devel
Russell King [Sat, 12 May 2007 15:12:12 +0000 (16:12 +0100)]
Merge branch 'fixes' into devel

17 years agoMerge branch 'omap-fixes' into fixes
Russell King [Sat, 12 May 2007 15:10:03 +0000 (16:10 +0100)]
Merge branch 'omap-fixes' into fixes

17 years ago[ARM] Use new get_irqnr_preamble
Russell King [Sat, 12 May 2007 14:25:50 +0000 (15:25 +0100)]
[ARM] Use new get_irqnr_preamble

Use the new get_irqnr_preamble macro to move the address of the IRQ
controller outside the IRQ handling loop.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[ARM] Ensure machine class menu is sorted alphabetically
Russell King [Sat, 12 May 2007 10:25:44 +0000 (11:25 +0100)]
[ARM] Ensure machine class menu is sorted alphabetically

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
17 years ago[SPARC]: Wire up signalfd/timerfd/eventfd syscalls.
David S. Miller [Sat, 12 May 2007 05:06:51 +0000 (22:06 -0700)]
[SPARC]: Wire up signalfd/timerfd/eventfd syscalls.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: Add support for bq4802 TOD chip, as found on ultra45.
David S. Miller [Sat, 12 May 2007 04:18:50 +0000 (21:18 -0700)]
[SPARC64]: Add support for bq4802 TOD chip, as found on ultra45.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: Correct FIRE_IOMMU_FLUSHINV register offset.
David S. Miller [Sat, 12 May 2007 04:02:09 +0000 (21:02 -0700)]
[SPARC64]: Correct FIRE_IOMMU_FLUSHINV register offset.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: envctrl.c needs asm/io.h
David S. Miller [Fri, 11 May 2007 23:34:35 +0000 (16:34 -0700)]
[SPARC64]: envctrl.c needs asm/io.h

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: Update defconfig.
David S. Miller [Fri, 11 May 2007 21:32:07 +0000 (14:32 -0700)]
[SPARC64]: Update defconfig.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[TTY]: Export proc_clear_tty() to modulea.
David S. Miller [Fri, 11 May 2007 21:30:57 +0000 (14:30 -0700)]
[TTY]: Export proc_clear_tty() to modulea.

A use was added to the solaris syscall module which
can be built modular.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: pci_resource_adjust() cannot be __init.
David S. Miller [Fri, 11 May 2007 21:29:43 +0000 (14:29 -0700)]
[SPARC64]: pci_resource_adjust() cannot be __init.

Noticed by Meelis Roos.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: Spelling fixes.
Simon Arlott [Fri, 11 May 2007 20:52:08 +0000 (13:52 -0700)]
[SPARC64]: Spelling fixes.

Spelling fixes in arch/sparc64/.

Signed-off-by: Simon Arlott <simon@fire.lp0.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC]: Spelling fixes.
Simon Arlott [Fri, 11 May 2007 20:51:23 +0000 (13:51 -0700)]
[SPARC]: Spelling fixes.

Spelling fixes in arch/sparc/.

Signed-off-by: Simon Arlott <simon@fire.lp0.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 years ago[SPARC64]: Kill LARGE_ALLOCS and update defconfig.
David S. Miller [Fri, 11 May 2007 00:30:50 +0000 (17:30 -0700)]
[SPARC64]: Kill LARGE_ALLOCS and update defconfig.

Let's use SLUB, since it works now, in order to get it
tested a bit.

Signed-off-by: David S. Miller <davem@davemloft.net>
17 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc
Linus Torvalds [Sat, 12 May 2007 03:12:57 +0000 (20:12 -0700)]
Merge branch 'master' of git://git./linux/kernel/git/paulus/powerpc

* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc: (23 commits)
  [POWERPC] Add arch/powerpc support for the Motorola PrPMC2800
  [POWERPC] Add bootwrapper support for Motorola PrPMC2800 platform
  [POWERPC] Add DTS file for the Motorola PrPMC2800 platform
  [POWERPC] Check cache coherency of kernel vs firmware
  [POWERPC] Add Marvell mv64x60 PCI bridge support
  [POWERPC] Create Marvell mv64x60 I2C platform_data
  [POWERPC] Create Marvell mv64x60 ethernet platform_data
  [POWERPC] Create Marvell mv64x60 MPSC (serial) platform_data
  [POWERPC] Add interrupt support for Marvell mv64x60 chips
  [POWERPC] Add bootwrapper support for Marvell/mv64x60 I2C
  [POWERPC] Add bootwrapper support for Marvell MPSC
  [POWERPC] Add bootwrapper support for Marvell/mv64x60 hostbridge
  [POWERPC] Add Makefile rules to wrap dts file in zImage
  [POWERPC] Spelling fixes: arch/ppc/
  [POWERPC] U-boot passes the initrd as start/end, not start/size.
  [POWERPC] PS3: Update ps3_defconfig
  [POWERPC] PS3: Fix request_irq warning
  [POWERPC] Don't complain if size-cells == 0 in prom_parse()
  [POWERPC] Simplify smp_space_timers
  [POWERPC] Trivial ps3 warning fixes
  ...

17 years ago[POWERPC] Add arch/powerpc support for the Motorola PrPMC2800
Dale Farnsworth [Sat, 12 May 2007 00:58:43 +0000 (10:58 +1000)]
[POWERPC] Add arch/powerpc support for the Motorola PrPMC2800

This finally adds the PPC_PRPMC2800 Kconfig option, the board setup
code (the setup and reset functions) and the defconfig, to support the
Motorola PrPMC2800 platform.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add bootwrapper support for Motorola PrPMC2800 platform
Mark A. Greer [Sat, 12 May 2007 00:58:18 +0000 (10:58 +1000)]
[POWERPC] Add bootwrapper support for Motorola PrPMC2800 platform

The Motorola PrPMC280 and PrPMC2800 processor modules sit on an F101 or
PrPMC2800 baseboard, respectively.  There are several variants of each
type of processor module which can have different amounts of memory,
amounts of FLASH, cpu frequencies, and an mv64360 or an mv64362.

The bootwrapper code for that platform reads VPD from an I2C EEPROM
to determine the processor module variant.  From the variant, the
amount of memory, etc. is determined and the device tree is updated
accordingly.  If the variant cannot be determined (e.g., corrupted
VPD or a previously unknown variant), the property values already
in the device tree are used.

Also, the firmware for those platforms does not completely configure
the mv64x60 host bridge so that configuration is done here.

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add DTS file for the Motorola PrPMC2800 platform
Mark A. Greer [Sat, 12 May 2007 00:57:58 +0000 (10:57 +1000)]
[POWERPC] Add DTS file for the Motorola PrPMC2800 platform

Add the device tree source file for the prpmc2800 line of processor PMCs.
Several of the property values are updated by the bootwrapper but sane
defaults have been chosen in case the bootwrapper can't determine the
exact processor board variant.  The defaults should allow the kernel
to boot despite having non-optimal device tree property values.

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Check cache coherency of kernel vs firmware
Dale Farnsworth [Sat, 12 May 2007 00:57:35 +0000 (10:57 +1000)]
[POWERPC] Check cache coherency of kernel vs firmware

check_cache_coherency() verifies that the cache coherency setting of
the kernel (CONFIG_NOT_COHERENT_CACHE) matches that left by the firmware,
as indicated by coherency-off device tree property.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add Marvell mv64x60 PCI bridge support
Dale Farnsworth [Sat, 12 May 2007 00:57:12 +0000 (10:57 +1000)]
[POWERPC] Add Marvell mv64x60 PCI bridge support

This patch adds PCI bridge support for the Marvell mv64x60 chip.
We also provide the ability to read/write the mv64x60 hotswap
register via sysfs if the hs_reg_valid property is set in the
device tree.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Create Marvell mv64x60 I2C platform_data
Dale Farnsworth [Sat, 12 May 2007 00:56:47 +0000 (10:56 +1000)]
[POWERPC] Create Marvell mv64x60 I2C platform_data

This patch creates platform_device entries for the Marvell mv64x60
I2C ports, based on information contained in device tree.

This driver (like the other mv64x60 drivers) are unusual in that it
works on both the MIPS and PowerPC architectures.  Because of that,
the drivers do not support the normal PowerPC of_platform_bus_type.
They support platform_bus_type instead.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Create Marvell mv64x60 ethernet platform_data
Dale Farnsworth [Sat, 12 May 2007 00:56:24 +0000 (10:56 +1000)]
[POWERPC] Create Marvell mv64x60 ethernet platform_data

This patch creates platform_device entries for the Marvell mv64x60
ethernet controller ports, based on information contained in the
device tree.

This driver (like the other mv64x60 drivers) are unusual in that it
works on both the MIPS and PowerPC architectures.  Because of that,
the drivers do not support the normal PowerPC of_platform_bus_type.
They support platform_bus_type instead.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Create Marvell mv64x60 MPSC (serial) platform_data
Dale Farnsworth [Sat, 12 May 2007 00:55:53 +0000 (10:55 +1000)]
[POWERPC] Create Marvell mv64x60 MPSC (serial) platform_data

This patch creates platform_device entries for the Marvell mv64x60
MPSC (multi-protocol serial controller) ports, based on information
contained in the device tree.

This driver (like the other mv64x60 drivers) are unusual in that it
works on both the MIPS and PowerPC architectures.  Because of that,
the drivers do not support the normal PowerPC of_platform_bus_type.
They support platform_bus_type instead.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add interrupt support for Marvell mv64x60 chips
Dale Farnsworth [Sat, 12 May 2007 00:55:24 +0000 (10:55 +1000)]
[POWERPC] Add interrupt support for Marvell mv64x60 chips

There are 3 interrupt groups each with its own status/mask registers.
We use a separate struct irq_chip for each interrupt group and handle
interrupts in two stages or levels: level 1 selects the appropriate
struct irq_chip, and level 2 selects individual interrupts within
that irq_chip.

Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add bootwrapper support for Marvell/mv64x60 I2C
Mark A. Greer [Sat, 12 May 2007 00:54:53 +0000 (10:54 +1000)]
[POWERPC] Add bootwrapper support for Marvell/mv64x60 I2C

Some platforms support a variety processor modules with no method of
determining which exact processor module is being used except by
examining Vital Product Data (VPD).  The modules may have different
amounts of memory, clock frequencies, etc. so reading the VPD becomes
necessary to correctly set properties in the device tree before its
passed to the kernel.

Often the VPD is stored in I2C EEPROMs so an I2C driver becomes necessary.
This I2C driver is for the I2C controller that's embedded on the Marvel
mv64x60 line of host bridges.

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add bootwrapper support for Marvell MPSC
Mark A. Greer [Sat, 12 May 2007 00:54:31 +0000 (10:54 +1000)]
[POWERPC] Add bootwrapper support for Marvell MPSC

The bootwrapper requires a serial driver to allow cmdline editing
and information reporting on the console.  This driver is required
by platforms that boot a zImage and use the MPSC for the console.

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add bootwrapper support for Marvell/mv64x60 hostbridge
Mark A. Greer [Sat, 12 May 2007 00:54:05 +0000 (10:54 +1000)]
[POWERPC] Add bootwrapper support for Marvell/mv64x60 hostbridge

The mv64x60 host bridge has many windows between its various components
(cpu, system memory, ethernet ctlr, MPSC, DMA ctlr, PCI MEM, PCI I/O).
Unfortunately, the firmware on some of mv64x60-based platforms do not
properly or completely configure those windows (e.g., MPSC->system memory
windows not configured or CPU->PCI MEM space not configured).

So, the missing configuration needs to be done in either the bootwrapper
or in the kernel.  To keep the kernel as clean as possible, it is done
in the bootwrapper.  Note that I/O controller configuration is NOT being
done, its only the windows to allow the I/O controllers and other components
to access memory, etc. that is being done--drivers assume that their
controllers can already access system memory).

Table of routines and the windows they configure:
mv64x60_config_ctlr_windows() ENET->System Memory
MPSC->System Memory
IDMA->System Memory

mv64x60_config_pci_windows() PCI MEM->System Memory
PCI I/O->Bridge's Registers

mv64x60_config_cpu2pci_window() CPU->PCI MEM
CPU->PCI I/O

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add Makefile rules to wrap dts file in zImage
Mark A. Greer [Sat, 12 May 2007 00:53:36 +0000 (10:53 +1000)]
[POWERPC] Add Makefile rules to wrap dts file in zImage

Add 'zImage.dts' and 'zImage.dts_initrd' build rules that automatically
compile and wrap a dts file from arch/powerpc/boot/dts into the zImage file.
The resulting zImage will be arch/powerpc/boot/zImage.dts.<platform> and
arch/powerpc/boot/zImage.dts_initrd.<platform>, respectively.

Having separate rules allows the user to choose whether to include a device
tree--and which device tree--at build time.  This is useful when one Makefile
target builds a zImage that runs on several platforms except for differing
device trees.  By just setting CONFIG_DEVICE_TREE and running "make zImage.dts"
the exact zImage you want is built without Makefile bloat or manually running
the wrapper script.

The dts file is expected to be arch/powerpc/boot/dts/$(CONFIG_DEVICE_TREE)

Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Spelling fixes: arch/ppc/
Simon Arlott [Fri, 11 May 2007 19:42:54 +0000 (05:42 +1000)]
[POWERPC] Spelling fixes: arch/ppc/

Spelling fixes in arch/ppc/.

Signed-off-by: Simon Arlott <simon@fire.lp0.eu>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] U-boot passes the initrd as start/end, not start/size.
Scott Wood [Fri, 11 May 2007 17:52:03 +0000 (03:52 +1000)]
[POWERPC] U-boot passes the initrd as start/end, not start/size.

The boot wrapper platform init code on 83xx and 85xx using the cuboot
platform type was incorrectly assuming that u-boot supplied the size
of the initrd, whereas it actually supplies the end address.  This
fixes it.

Signed-off-by: Scott Wood <scottwood@freescale.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] PS3: Update ps3_defconfig
Geoff Levand [Fri, 11 May 2007 17:42:03 +0000 (03:42 +1000)]
[POWERPC] PS3: Update ps3_defconfig

Update ps3_defconfig.

Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] PS3: Fix request_irq warning
Geoff Levand [Fri, 11 May 2007 17:41:59 +0000 (03:41 +1000)]
[POWERPC] PS3: Fix request_irq warning

Fix compiler warning:

ps3/smp.c:122: warning: ignoring return value of 'request_irq'

Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Don't complain if size-cells == 0 in prom_parse()
Kim Phillips [Fri, 11 May 2007 17:19:38 +0000 (03:19 +1000)]
[POWERPC] Don't complain if size-cells == 0 in prom_parse()

An mdio bus scan was added with ucc_geth phylib
migration patches, now machines complain on boot, saying:

prom_parse: Bad cell count for /qe@e0100000/mdio@2120/ethernet-phy@00
prom_parse: Bad cell count for /qe@e0100000/mdio@2120/ethernet-phy@01

since size-cells can indeed be 0, this patch fixes the check.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Simplify smp_space_timers
will schmidt [Fri, 11 May 2007 13:34:16 +0000 (23:34 +1000)]
[POWERPC] Simplify smp_space_timers

Greatly simplify the function smp_space_timers.

The stolen time calculation (per comment within the code) doesn't need the
half-jiffy stagger any more.  There isn't an issue with bouncing off global
locks, so we really shouldn't need any sort of staggering at all.

However, the last_jiffy value still needs to be set.   This removes the
extra stagger logic, and just sets the values.

This change should benefit applications that rely on barrier
synchronization, and will help cut down OS jitter.

Boot tested across the board (G5,power3,power4,power5,970mp blade).

Signed-off-by: Will Schmidt <will_schmidt@vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Trivial ps3 warning fixes
Stephen Rothwell [Fri, 11 May 2007 05:42:44 +0000 (15:42 +1000)]
[POWERPC] Trivial ps3 warning fixes

Fixes warnings:

arch/powerpc/platforms/ps3/system-bus.c: In function 'ps3_map_sg':
arch/powerpc/platforms/ps3/system-bus.c:278: warning: unused variable 'i'
arch/powerpc/platforms/ps3/system-bus.c:277: warning: unused variable 'dev'
arch/powerpc/platforms/ps3/setup.c:103: warning: 'prealloc' defined but not used

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Geoff Levand <geoffrey.levand@am.sony.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Add missed include
Stephen Rothwell [Fri, 11 May 2007 05:40:36 +0000 (15:40 +1000)]
[POWERPC] Add missed include

fixes:
arch/powerpc/platforms/pasemi/idle.c: In function 'pasemi_system_reset_exception':
arch/powerpc/platforms/pasemi/idle.c:55: warning: implicit declaration of function 'do_IRQ'

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Remove unused variable in hpte_decode()
Stephen Rothwell [Fri, 11 May 2007 05:38:34 +0000 (15:38 +1000)]
[POWERPC] Remove unused variable in hpte_decode()

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years ago[POWERPC] Assign correct variable in hpte_decode()
Stephen Rothwell [Fri, 11 May 2007 05:37:38 +0000 (15:37 +1000)]
[POWERPC] Assign correct variable in hpte_decode()

This case will never be hit, but it should be corrected anyway.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
17 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog
Linus Torvalds [Fri, 11 May 2007 23:05:43 +0000 (16:05 -0700)]
Merge /linux/kernel/git/wim/linux-2.6-watchdog

* master.kernel.org:/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog:
  [WATCHDOG] iTCO_wdt.c - fix ACPI Base register

17 years agoMerge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik...
Linus Torvalds [Fri, 11 May 2007 23:03:03 +0000 (16:03 -0700)]
Merge branch 'upstream-linus' of /linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  pata_platform: don't use generic ata_port_start
  Use menuconfig objects: libata
  add the ATI SB700 SATA controller device id to AHCI pci table
  Add the combined mode for ATI SB700
  pata_pcmcia: recognize 2GB CompactFlash from Transcend
  git-libata-all: sata_via build fix
  libata-acpi: clean up parameters and misc stuff
  libata-acpi: s/CONFIG_SATA_ACPI/CONFIG_ATA_ACPI/
  libata: give devices one last chance even if recovery failed with -EINVAL
  libata: fallback to the other IDENTIFY on device error, take#2
  libata: ignore EH scheduling during initialization
  libata: clean up SFF init mess
  libata: implement libata.spindown_compat
  libata: reimplement suspend/resume support using sdev->manage_start_stop

17 years agoMerge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik...
Linus Torvalds [Fri, 11 May 2007 23:02:05 +0000 (16:02 -0700)]
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: (35 commits)
  Add support for the Davicom DM9161A PHY
  sky2: only disable 88e8056 on some boards
  sky2: 88e8071 support not ready
  skge: crash on shutdown/suspend
  sky2: fix oops on shutdown
  mlx4: don't use deprecated IRQ flags
  netxen_nic_main don't use deprecated irq flags
  Use menuconfig objects II - netdev/wan
  Use menuconfig objects II - netdev/pcmcia
  Use menuconfig objects II - netdev/atm
  Use menuconfig objects: netdev
  Use menuconfig objects: PHY
  spidernet: remove unnecessary accesses to phy
  S2IO: Statistics for link up/down and memory allocated/freed
  S2IO: statistics for memory allocation failuers
  S2IO: getringparam ethtool option
  [PATCH] libertas: 64-bit cleanups
  [PATCH] libertas: sparse fixes
  [PATCH] libertas: fix missing unlock in TX error path
  [PATCH] libertas: make debugfs.c sparse-clean
  ...

17 years agoFix assertion failure with MSI on sparc64
David Miller [Fri, 11 May 2007 20:26:44 +0000 (13:26 -0700)]
Fix assertion failure with MSI on sparc64

Today's find is a triggered assertion in msi_free_irqs() when the system
doesn't support MSI, in which case arch_setup_msi_irqs() always returns
an error.

The problem is that when this happens we branch into msi_free_irqs(), to
which you added the following assertion loop:

list_for_each_entry(entry, &dev->msi_list, list)
BUG_ON(irq_has_action(entry->irq));

Well, if arch_setup_msi_irqs() fails, entry->irq will be zero and
although that's never assigned to any normal devices we use that IRQ
number for the timer interrupt on sparc64 so this assertion triggers.

Better to test for zero before doing the irq_has_action() assertion
thing.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 years agopata_platform: don't use generic ata_port_start
Olof Johansson [Thu, 10 May 2007 04:00:20 +0000 (23:00 -0500)]
pata_platform: don't use generic ata_port_start

I have a system where I have a simple IDE controller that sits on a
local bus without bus master dma capability, and thus no dma_mapping
ops defined for the device/bus.

pata_platform works great for me, with the exception of using the generic
ata_port_start which tries to do a dmam_alloc_coherent.

Looks like it doesn't need to allocate a prd table at all, so replace it
with a dummy function instead.

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoUse menuconfig objects: libata
Jan Engelhardt [Fri, 11 May 2007 05:48:54 +0000 (22:48 -0700)]
Use menuconfig objects: libata

Use menuconfigs instead of menus, so the whole menu can be disabled at once
instead of going through all options.

Signed-off-by: Jan Engelhardt <jengelh@gmx.de>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoadd the ATI SB700 SATA controller device id to AHCI pci table
Henry Su [Fri, 11 May 2007 05:48:51 +0000 (22:48 -0700)]
add the ATI SB700 SATA controller device id to AHCI pci table

Add the device ID to AHCI pci table for ATI SB700 SATA controller, the
subsequent chipset of SB600.

Signed-off-by: henry su<henry.su@amd.com>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Tejun Heo <htejun@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoAdd the combined mode for ATI SB700
Henry Su [Fri, 11 May 2007 05:48:50 +0000 (22:48 -0700)]
Add the combined mode for ATI SB700

Besides those modes in ATI SB600 SATA controller, ATI SB700 supports one
more mode:the combined mode.

The combined mode is a Legacy IDE mode used for compatibility with some old
OS without AHCI driver, but now it is not necessary for Linux since the
kernel has supported AHCI.

Signed-off-by: Luugi Marsan <luugi.marsan@amd.com>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Tejun Heo <htejun@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agopata_pcmcia: recognize 2GB CompactFlash from Transcend
Aeschbacher, Fabrice [Fri, 11 May 2007 05:48:48 +0000 (22:48 -0700)]
pata_pcmcia: recognize 2GB CompactFlash from Transcend

Allow the pata_pcmcia driver to automatically detect 2GB CompactFlash cards
from Transcend.

Signed-off-by: Fabrice Aeschbacher <fabrice.aeschbacher@siemens.com>
Cc: "Peter Stuge" <stuge-linux-pcmcia@cdy.org>
Acked-by: Alan Cox <alan@redhat.com>
Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agogit-libata-all: sata_via build fix
Andrew Morton [Fri, 11 May 2007 05:48:43 +0000 (22:48 -0700)]
git-libata-all: sata_via build fix

Cc: Tejun Heo <htejun@gmail.com>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata-acpi: clean up parameters and misc stuff
Tejun Heo [Sat, 5 May 2007 14:50:38 +0000 (23:50 +0900)]
libata-acpi: clean up parameters and misc stuff

This patch cleans up libata-acpi such that it looks similar to other
libata files.  This patch doesn't introuce any behavior changes.

* make libata-acpi functions take ata_device instead of ata_port +
  device index
* s/atadev/dev/
* de-indent local variable declarations

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata-acpi: s/CONFIG_SATA_ACPI/CONFIG_ATA_ACPI/
Tejun Heo [Sat, 5 May 2007 14:50:38 +0000 (23:50 +0900)]
libata-acpi: s/CONFIG_SATA_ACPI/CONFIG_ATA_ACPI/

ACPI applies to both SATA and PATA.  Drop the 'S' from the config
variable.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: give devices one last chance even if recovery failed with -EINVAL
Tejun Heo [Fri, 11 May 2007 12:46:51 +0000 (14:46 +0200)]
libata: give devices one last chance even if recovery failed with -EINVAL

After certain errors, some devices report complete garbage on
IDENTIFY.  This can cause ata_dev_read_id() to fail with -EINVAL
resulting in immediate disabling of the device.  Give the device one
last chance after -EINVAL to allow recovery from such situations.  As
-EINVAL is triggered very rarely, this shouldn't cause any noticeable
affect on more common error paths.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Harald Dunkel <harald.dunkel@t-online.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: fallback to the other IDENTIFY on device error, take#2
Tejun Heo [Fri, 11 May 2007 12:35:29 +0000 (14:35 +0200)]
libata: fallback to the other IDENTIFY on device error, take#2

It seems the world isn't as frank as we thought and some devices lie
about who they are.  Fallback to the other IDENTIFY if IDENTIFY is
aborted by the device.  As this is the strategy used by IDE for a long
time, it shouldn't cause too much problem.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: William Thompson <wt@electro-mechanical.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: ignore EH scheduling during initialization
Tejun Heo [Tue, 1 May 2007 09:50:15 +0000 (11:50 +0200)]
libata: ignore EH scheduling during initialization

libata enables SCSI host during ATA host activation which happens
after IRQ handler is registered and IRQ is enabled.  All ATA ports are
in frozen state when IRQ is enabled but frozen ports may raise limited
number of IRQs after being frozen - IOW, ->freeze() is not responsible
for clearing pending IRQs.  During normal operation, the IRQ handler
is responsible for clearing spurious IRQs on frozen ports and it
usually doesn't require any extra code.

Unfortunately, during host initialization, the IRQ handler can end up
scheduling EH for a port whose SCSI host isn't initialized yet.  This
results in OOPS in the SCSI midlayer.  This is relatively short window
and scheduling EH for probing is the first thing libata does after
initialization, so ignoring EH scheduling until initialization is
complete solves the problem nicely.

This problem was spotted by Berck E. Nash in the following thread.

  http://thread.gmane.org/gmane.linux.kernel/519412

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Berck E. Nash <flyboy@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: clean up SFF init mess
Tejun Heo [Fri, 4 May 2007 10:43:58 +0000 (12:43 +0200)]
libata: clean up SFF init mess

The intention of using port_mask in SFF init helpers was to eventually
support exoctic configurations such as combination of legacy and
native port on the same controller.  This never became actually
necessary and the related code always has been subtly broken one way
or the other.  Now that new init model is in place, there is no reason
to make common helpers capable of handling all corner cases.  Exotic
cases can simply dealt within LLDs as necessary.

This patch removes port_mask handling in SFF init helpers.  SFF init
helpers don't take n_ports argument and interpret it into port_mask
anymore.  All information is carried via port_info.  n_ports argument
is dropped and always two ports are allocated.  LLD can tell SFF to
skip certain port by marking it dummy.  Note that SFF code has been
treating unuvailable ports this way for a long time until recent
breakage fix from Linus and is consistent with how other drivers
handle with unavailable ports.

This fixes 1-port legacy host handling still broken after the recent
native mode fix and simplifies SFF init logic.  The following changes
are made...

* ata_pci_init_native_host() and ata_init_legacy_host() both now try
  to initialized whatever they can and mark failed ports dummy.  They
  return 0 if any port is successfully initialized.

* ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't
  take n_ports argument.  All info should be specified via port_info
  array.  Always two ports are allocated.

* ata_pci_init_bmdma() exported to be used by LLDs in exotic cases.

* port_info handling in all LLDs are standardized - all port_info
  arrays are const stack variable named ppi.  Unless the second port
  is different from the first, its port_info is specified as NULL
  (tells libata that it's identical to the last non-NULL port_info).

* pata_hpt37x/hpt3x2n: don't modify static variable directly.  Make an
  on-stack copy instead as ata_piix does.

* pata_uli: It has 4 ports instead of 2.  Don't use
  ata_pci_prepare_native_host().  Allocate the host explicitly and use
  init helpers.  It's simple enough.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: implement libata.spindown_compat
Tejun Heo [Fri, 4 May 2007 19:28:48 +0000 (21:28 +0200)]
libata: implement libata.spindown_compat

Now that libata uses sd->manage_start_stop, libata spins down disk on
shutdown.  In an attempt to compensate libata's previous shortcoming,
some distros sync and spin down disks attached via libata in their
shutdown(8).  Some disks spin back up just to spin down again on
STANDBYNOW1 if the command is issued when the disk is spun down, so
this double spinning down causes problem.

This patch implements module parameter libata.spindown_compat which,
when set to one (default value), prevents libata from spinning down
disks on shutdown thus avoiding double spinning down.  Note that
libata spins down disks for suspend to mem and disk, so with
libata.spindown_compat set to one, disks should be properly spun down
in all cases without modifying shutdown(8).

shutdown(8) should be fixed eventually.  Some drive do spin up on
SYNCHRONZE_CACHE even when their cache is clean.  Those disks
currently spin up briefly when sd tries to shutdown the device and
then the machine powers off immediately, which can't be good for the
head.  We can't skip SYNCHRONIZE_CACHE during shudown as it can be
dangerous data integrity-wise.

So, this spindown_compat parameter is already scheduled for removal by
the end of the next year and here's what shutdown(8) should do.

  * Check whether /sys/modules/libata/parameters/spindown_compat
    exists.  If it does, write 0 to it.

  * For each libata harddisk {
* Check whether /sys/class/scsi_disk/h:c:i:l/manage_start_stop
  exists.  Iff it doesn't, synchronize cache and spin the disk
  down as before.
  }

The above procedure will make shutdown(8) work properly with kernels
before this change, ones with this workaround and later ones without
it.

To accelerate shutdown(8) updates, if the compat mode is in use, this
patch prints BIG FAT warning for five seconds during shutdown (the
optimal interval to annoy the user just the right amount discovered by
hours of tireless usability testing).

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agolibata: reimplement suspend/resume support using sdev->manage_start_stop
Tejun Heo [Fri, 4 May 2007 19:27:47 +0000 (21:27 +0200)]
libata: reimplement suspend/resume support using sdev->manage_start_stop

Reimplement suspend/resume support using sdev->manage_start_stop.

* Device suspend/resume is now SCSI layer's responsibility and the
  code is simplified a lot.

* DPM is dropped.  This also simplifies code a lot.  Suspend/resume
  status is port-wide now.

* ata_scsi_device_suspend/resume() and ata_dev_ready() removed.

* Resume now has to wait for disk to spin up before proceeding.  I
  couldn't find easy way out as libata is in EH waiting for the
  disk to be ready and sd is waiting for EH to complete to issue
  START_STOP.

* sdev->manage_start_stop is set to 1 in ata_scsi_slave_config().
  This fixes spindown on shutdown and suspend-to-disk.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoMerge branch 'libertas' of git://git.kernel.org/pub/scm/linux/kernel/git/linville...
Jeff Garzik [Fri, 11 May 2007 21:54:19 +0000 (17:54 -0400)]
Merge branch 'libertas' of git://git./linux/kernel/git/linville/wireless-2.6 into upstream

17 years agoAdd support for the Davicom DM9161A PHY
Kim Phillips [Thu, 10 May 2007 20:16:04 +0000 (15:16 -0500)]
Add support for the Davicom DM9161A PHY

Distinguish between the Davicom DM9161A PHY and the DM9161E.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agosky2: only disable 88e8056 on some boards
Stephen Hemminger [Fri, 11 May 2007 18:21:47 +0000 (11:21 -0700)]
sky2: only disable 88e8056 on some boards

Use DMI to add a blacklist of broken boards (so far only one).
Hopefully, the problems will be solved later, and the the whole
blacklist can disappear.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agosky2: 88e8071 support not ready
Stephen Hemminger [Fri, 11 May 2007 18:21:46 +0000 (11:21 -0700)]
sky2: 88e8071 support not ready

The driver is not ready to support 88e8071 chip.
If this chip is present, system will hang on boot.

So remove it from PCI device id's for now.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoskge: crash on shutdown/suspend
Stephen Hemminger [Fri, 11 May 2007 18:21:45 +0000 (11:21 -0700)]
skge: crash on shutdown/suspend

If device is not fails during module startup (like unsupported chip
version) then driver would crash dereferencing a null pointer, on shutdown
or suspend/resume.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agosky2: fix oops on shutdown
Stephen Hemminger [Fri, 11 May 2007 18:21:44 +0000 (11:21 -0700)]
sky2: fix oops on shutdown

If the device is fails during module startup for some reason like
unsupported chip version then the driver would crash dereferencing a
null pointer, on shutdown or suspend/resume.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agomlx4: don't use deprecated IRQ flags
Andrew Morton [Fri, 11 May 2007 05:53:01 +0000 (22:53 -0700)]
mlx4: don't use deprecated IRQ flags

Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agonetxen_nic_main don't use deprecated irq flags
Andrew Morton [Fri, 11 May 2007 05:53:00 +0000 (22:53 -0700)]
netxen_nic_main don't use deprecated irq flags

drivers/net/netxen/netxen_nic_main.c: In function 'netxen_nic_open':
drivers/net/netxen/netxen_nic_main.c:738: warning: 'deprecated_irq_flag' is deprecated (declared at include/linux/interrupt.h:66)
drivers/net/netxen/netxen_nic_main.c:738: warning: 'deprecated_irq_flag' is deprecated (declared at include/linux/interrupt.h:66)

Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
17 years agoUse menuconfig objects II - netdev/wan
Jan Engelhardt [Fri, 11 May 2007 05:52:59 +0000 (22:52 -0700)]
Use menuconfig objects II - netdev/wan

Change Kconfig objects from "menu, config" into "menuconfig" so
that the user can disable the whole feature without having to
enter the menu first.

Signed-off-by: Jan Engelhardt <jengelh@gmx.de>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Krzysztof Halasa <khc@pm.waw.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>