initial_commit

36 files changed, 16089 insertions, 0 deletions

diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig
new file mode 100644
index 00000000..19470885
--- /dev/null
+++ b/drivers/cpufreq/Kconfig
@@ -0,0 +1,215 @@
+menu "CPU Frequency scaling"
+
+config CPU_FREQ
+	bool "CPU Frequency scaling"
+	help
+	  CPU Frequency scaling allows you to change the clock speed of 
+	  CPUs on the fly. This is a nice method to save power, because 
+	  the lower the CPU clock speed, the less power the CPU consumes.
+
+	  Note that this driver doesn't automatically change the CPU
+	  clock speed, you need to either enable a dynamic cpufreq governor
+	  (see below) after boot, or use a userspace tool.
+
+	  For details, take a look at <file:Documentation/cpu-freq>.
+
+	  If in doubt, say N.
+
+if CPU_FREQ
+
+config CPU_FREQ_TABLE
+	tristate
+
+config CPU_FREQ_STAT
+	tristate "CPU frequency translation statistics"
+	select CPU_FREQ_TABLE
+	default y
+	help
+	  This driver exports CPU frequency statistics information through sysfs
+	  file system.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_stats.
+
+	  If in doubt, say N.
+
+config CPU_FREQ_STAT_DETAILS
+	bool "CPU frequency translation statistics details"
+	depends on CPU_FREQ_STAT
+	help
+	  This will show detail CPU frequency translation table in sysfs file
+	  system.
+
+	  If in doubt, say N.
+
+choice
+	prompt "Default CPUFreq governor"
+	default CPU_FREQ_DEFAULT_GOV_USERSPACE if CPU_FREQ_SA1100 || CPU_FREQ_SA1110
+	default CPU_FREQ_DEFAULT_GOV_PERFORMANCE
+	help
+	  This option sets which CPUFreq governor shall be loaded at
+	  startup. If in doubt, select 'performance'.
+
+config CPU_FREQ_DEFAULT_GOV_PERFORMANCE
+	bool "performance"
+	select CPU_FREQ_GOV_PERFORMANCE
+	help
+	  Use the CPUFreq governor 'performance' as default. This sets
+	  the frequency statically to the highest frequency supported by
+	  the CPU.
+
+config CPU_FREQ_DEFAULT_GOV_POWERSAVE
+	bool "powersave"
+	depends on EXPERT
+	select CPU_FREQ_GOV_POWERSAVE
+	help
+	  Use the CPUFreq governor 'powersave' as default. This sets
+	  the frequency statically to the lowest frequency supported by
+	  the CPU.
+
+config CPU_FREQ_DEFAULT_GOV_USERSPACE
+	bool "userspace"
+	select CPU_FREQ_GOV_USERSPACE
+	help
+	  Use the CPUFreq governor 'userspace' as default. This allows
+	  you to set the CPU frequency manually or when a userspace 
+	  program shall be able to set the CPU dynamically without having
+	  to enable the userspace governor manually.
+
+config CPU_FREQ_DEFAULT_GOV_ONDEMAND
+	bool "ondemand"
+	select CPU_FREQ_GOV_ONDEMAND
+	select CPU_FREQ_GOV_PERFORMANCE
+	help
+	  Use the CPUFreq governor 'ondemand' as default. This allows
+	  you to get a full dynamic frequency capable system by simply
+	  loading your cpufreq low-level hardware driver.
+	  Be aware that not all cpufreq drivers support the ondemand
+	  governor. If unsure have a look at the help section of the
+	  driver. Fallback governor will be the performance governor.
+
+config CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
+	bool "conservative"
+	select CPU_FREQ_GOV_CONSERVATIVE
+	select CPU_FREQ_GOV_PERFORMANCE
+	help
+	  Use the CPUFreq governor 'conservative' as default. This allows
+	  you to get a full dynamic frequency capable system by simply
+	  loading your cpufreq low-level hardware driver.
+	  Be aware that not all cpufreq drivers support the conservative
+	  governor. If unsure have a look at the help section of the
+	  driver. Fallback governor will be the performance governor.
+
+config CPU_FREQ_DEFAULT_GOV_INTERACTIVE
+	bool "interactive"
+	select CPU_FREQ_GOV_INTERACTIVE
+	help
+	  Use the CPUFreq governor 'interactive' as default. This allows
+	  you to get a full dynamic cpu frequency capable system by simply
+	  loading your cpufreq low-level hardware driver, using the
+	  'interactive' governor for latency-sensitive workloads.
+
+endchoice
+
+config CPU_FREQ_GOV_PERFORMANCE
+	tristate "'performance' governor"
+	help
+	  This cpufreq governor sets the frequency statically to the
+	  highest available CPU frequency.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_performance.
+
+	  If in doubt, say Y.
+
+config CPU_FREQ_GOV_POWERSAVE
+	tristate "'powersave' governor"
+	help
+	  This cpufreq governor sets the frequency statically to the
+	  lowest available CPU frequency.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_powersave.
+
+	  If in doubt, say Y.
+
+config CPU_FREQ_GOV_USERSPACE
+	tristate "'userspace' governor for userspace frequency scaling"
+	help
+	  Enable this cpufreq governor when you either want to set the
+	  CPU frequency manually or when a userspace program shall
+	  be able to set the CPU dynamically, like on LART 
+	  <http://www.lartmaker.nl/>.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_userspace.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say Y.
+
+config CPU_FREQ_GOV_ONDEMAND
+	tristate "'ondemand' cpufreq policy governor"
+	select CPU_FREQ_TABLE
+	help
+	  'ondemand' - This driver adds a dynamic cpufreq policy governor.
+	  The governor does a periodic polling and 
+	  changes frequency based on the CPU utilization.
+	  The support for this governor depends on CPU capability to
+	  do fast frequency switching (i.e, very low latency frequency
+	  transitions). 
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_ondemand.
+
+	  For details, take a look at linux/Documentation/cpu-freq.
+
+	  If in doubt, say N.
+
+config CPU_FREQ_GOV_INTERACTIVE
+	tristate "'interactive' cpufreq policy governor"
+	help
+	  'interactive' - This driver adds a dynamic cpufreq policy governor
+	  designed for latency-sensitive workloads.
+
+	  This governor attempts to reduce the latency of clock
+	  increases so that the system is more responsive to
+	  interactive workloads.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_interactive.
+
+	  For details, take a look at linux/Documentation/cpu-freq.
+
+	  If in doubt, say N.
+
+config CPU_FREQ_GOV_CONSERVATIVE
+	tristate "'conservative' cpufreq governor"
+	depends on CPU_FREQ
+	help
+	  'conservative' - this driver is rather similar to the 'ondemand'
+	  governor both in its source code and its purpose, the difference is
+	  its optimisation for better suitability in a battery powered
+	  environment.  The frequency is gracefully increased and decreased
+	  rather than jumping to 100% when speed is required.
+
+	  If you have a desktop machine then you should really be considering
+	  the 'ondemand' governor instead, however if you are using a laptop,
+	  PDA or even an AMD64 based computer (due to the unacceptable
+	  step-by-step latency issues between the minimum and maximum frequency
+	  transitions in the CPU) you will probably want to use this governor.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called cpufreq_conservative.
+
+	  For details, take a look at linux/Documentation/cpu-freq.
+
+	  If in doubt, say N.
+
+menu "x86 CPU frequency scaling drivers"
+depends on X86
+source "drivers/cpufreq/Kconfig.x86"
+endmenu
+
+endif
+endmenu
diff --git a/drivers/cpufreq/Kconfig.x86 b/drivers/cpufreq/Kconfig.x86
new file mode 100644
index 00000000..78ff7ee4
--- /dev/null
+++ b/drivers/cpufreq/Kconfig.x86
@@ -0,0 +1,255 @@
+#
+# x86 CPU Frequency scaling drivers
+#
+
+config X86_PCC_CPUFREQ
+	tristate "Processor Clocking Control interface driver"
+	depends on ACPI && ACPI_PROCESSOR
+	help
+	  This driver adds support for the PCC interface.
+
+	  For details, take a look at:
+	  <file:Documentation/cpu-freq/pcc-cpufreq.txt>.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called pcc-cpufreq.
+
+	  If in doubt, say N.
+
+config X86_ACPI_CPUFREQ
+	tristate "ACPI Processor P-States driver"
+	select CPU_FREQ_TABLE
+	depends on ACPI_PROCESSOR
+	help
+	  This driver adds a CPUFreq driver which utilizes the ACPI
+	  Processor Performance States.
+	  This driver also supports Intel Enhanced Speedstep.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called acpi-cpufreq.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config ELAN_CPUFREQ
+	tristate "AMD Elan SC400 and SC410"
+	select CPU_FREQ_TABLE
+	depends on MELAN
+	---help---
+	  This adds the CPUFreq driver for AMD Elan SC400 and SC410
+	  processors.
+
+	  You need to specify the processor maximum speed as boot
+	  parameter: elanfreq=maxspeed (in kHz) or as module
+	  parameter "max_freq".
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config SC520_CPUFREQ
+	tristate "AMD Elan SC520"
+	select CPU_FREQ_TABLE
+	depends on MELAN
+	---help---
+	  This adds the CPUFreq driver for AMD Elan SC520 processor.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+
+config X86_POWERNOW_K6
+	tristate "AMD Mobile K6-2/K6-3 PowerNow!"
+	select CPU_FREQ_TABLE
+	depends on X86_32
+	help
+	  This adds the CPUFreq driver for mobile AMD K6-2+ and mobile
+	  AMD K6-3+ processors.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_POWERNOW_K7
+	tristate "AMD Mobile Athlon/Duron PowerNow!"
+	select CPU_FREQ_TABLE
+	depends on X86_32
+	help
+	  This adds the CPUFreq driver for mobile AMD K7 mobile processors.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_POWERNOW_K7_ACPI
+	bool
+	depends on X86_POWERNOW_K7 && ACPI_PROCESSOR
+	depends on !(X86_POWERNOW_K7 = y && ACPI_PROCESSOR = m)
+	depends on X86_32
+	default y
+
+config X86_POWERNOW_K8
+	tristate "AMD Opteron/Athlon64 PowerNow!"
+	select CPU_FREQ_TABLE
+	depends on ACPI && ACPI_PROCESSOR
+	help
+	  This adds the CPUFreq driver for K8/K10 Opteron/Athlon64 processors.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called powernow-k8.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+config X86_GX_SUSPMOD
+	tristate "Cyrix MediaGX/NatSemi Geode Suspend Modulation"
+	depends on X86_32 && PCI
+	help
+	 This add the CPUFreq driver for NatSemi Geode processors which
+	 support suspend modulation.
+
+	 For details, take a look at <file:Documentation/cpu-freq/>.
+
+	 If in doubt, say N.
+
+config X86_SPEEDSTEP_CENTRINO
+	tristate "Intel Enhanced SpeedStep (deprecated)"
+	select CPU_FREQ_TABLE
+	select X86_SPEEDSTEP_CENTRINO_TABLE if X86_32
+	depends on X86_32 || (X86_64 && ACPI_PROCESSOR)
+	help
+	  This is deprecated and this functionality is now merged into
+	  acpi_cpufreq (X86_ACPI_CPUFREQ). Use that driver instead of
+	  speedstep_centrino.
+	  This adds the CPUFreq driver for Enhanced SpeedStep enabled
+	  mobile CPUs.  This means Intel Pentium M (Centrino) CPUs
+	  or 64bit enabled Intel Xeons.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called speedstep-centrino.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_SPEEDSTEP_CENTRINO_TABLE
+	bool "Built-in tables for Banias CPUs"
+	depends on X86_32 && X86_SPEEDSTEP_CENTRINO
+	default y
+	help
+	  Use built-in tables for Banias CPUs if ACPI encoding
+	  is not available.
+
+	  If in doubt, say N.
+
+config X86_SPEEDSTEP_ICH
+	tristate "Intel Speedstep on ICH-M chipsets (ioport interface)"
+	select CPU_FREQ_TABLE
+	depends on X86_32
+	help
+	  This adds the CPUFreq driver for certain mobile Intel Pentium III
+	  (Coppermine), all mobile Intel Pentium III-M (Tualatin) and all
+	  mobile Intel Pentium 4 P4-M on systems which have an Intel ICH2,
+	  ICH3 or ICH4 southbridge.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_SPEEDSTEP_SMI
+	tristate "Intel SpeedStep on 440BX/ZX/MX chipsets (SMI interface)"
+	select CPU_FREQ_TABLE
+	depends on X86_32 && EXPERIMENTAL
+	help
+	  This adds the CPUFreq driver for certain mobile Intel Pentium III
+	  (Coppermine), all mobile Intel Pentium III-M (Tualatin)
+	  on systems which have an Intel 440BX/ZX/MX southbridge.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_P4_CLOCKMOD
+	tristate "Intel Pentium 4 clock modulation"
+	select CPU_FREQ_TABLE
+	help
+	  This adds the CPUFreq driver for Intel Pentium 4 / XEON
+	  processors.  When enabled it will lower CPU temperature by skipping
+	  clocks.
+
+	  This driver should be only used in exceptional
+	  circumstances when very low power is needed because it causes severe
+	  slowdowns and noticeable latencies.  Normally Speedstep should be used
+	  instead.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called p4-clockmod.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  Unless you are absolutely sure say N.
+
+config X86_CPUFREQ_NFORCE2
+	tristate "nVidia nForce2 FSB changing"
+	depends on X86_32 && EXPERIMENTAL
+	help
+	  This adds the CPUFreq driver for FSB changing on nVidia nForce2
+	  platforms.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_LONGRUN
+	tristate "Transmeta LongRun"
+	depends on X86_32
+	help
+	  This adds the CPUFreq driver for Transmeta Crusoe and Efficeon processors
+	  which support LongRun.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_LONGHAUL
+	tristate "VIA Cyrix III Longhaul"
+	select CPU_FREQ_TABLE
+	depends on X86_32 && ACPI_PROCESSOR
+	help
+	  This adds the CPUFreq driver for VIA Samuel/CyrixIII,
+	  VIA Cyrix Samuel/C3, VIA Cyrix Ezra and VIA Cyrix Ezra-T
+	  processors.
+
+	  For details, take a look at <file:Documentation/cpu-freq/>.
+
+	  If in doubt, say N.
+
+config X86_E_POWERSAVER
+	tristate "VIA C7 Enhanced PowerSaver (DANGEROUS)"
+	select CPU_FREQ_TABLE
+	depends on X86_32 && EXPERIMENTAL
+	help
+	  This adds the CPUFreq driver for VIA C7 processors.  However, this driver
+	  does not have any safeguards to prevent operating the CPU out of spec
+	  and is thus considered dangerous.  Please use the regular ACPI cpufreq
+	  driver, enabled by CONFIG_X86_ACPI_CPUFREQ.
+
+	  If in doubt, say N.
+
+comment "shared options"
+
+config X86_SPEEDSTEP_LIB
+	tristate
+	default (X86_SPEEDSTEP_ICH || X86_SPEEDSTEP_SMI || X86_P4_CLOCKMOD)
+
+config X86_SPEEDSTEP_RELAXED_CAP_CHECK
+	bool "Relaxed speedstep capability checks"
+	depends on X86_32 && (X86_SPEEDSTEP_SMI || X86_SPEEDSTEP_ICH)
+	help
+	  Don't perform all checks for a speedstep capable system which would
+	  normally be done. Some ancient or strange systems, though speedstep
+	  capable, don't always indicate that they are speedstep capable. This
+	  option lets the probing code bypass some of those checks if the
+	  parameter "relaxed_check=1" is passed to the module.
+
diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile
new file mode 100644
index 00000000..c371bb46
--- /dev/null
+++ b/drivers/cpufreq/Makefile
@@ -0,0 +1,44 @@
+# CPUfreq core
+obj-$(CONFIG_CPU_FREQ)			+= cpufreq.o
+# CPUfreq stats
+obj-$(CONFIG_CPU_FREQ_STAT)             += cpufreq_stats.o
+
+# CPUfreq governors 
+obj-$(CONFIG_CPU_FREQ_GOV_PERFORMANCE)	+= cpufreq_performance.o
+obj-$(CONFIG_CPU_FREQ_GOV_POWERSAVE)	+= cpufreq_powersave.o
+obj-$(CONFIG_CPU_FREQ_GOV_USERSPACE)	+= cpufreq_userspace.o
+obj-$(CONFIG_CPU_FREQ_GOV_ONDEMAND)	+= cpufreq_ondemand.o
+obj-$(CONFIG_CPU_FREQ_GOV_CONSERVATIVE)	+= cpufreq_conservative.o
+obj-$(CONFIG_CPU_FREQ_GOV_INTERACTIVE)  += cpufreq_interactive.o
+
+# CPUfreq cross-arch helpers
+obj-$(CONFIG_CPU_FREQ_TABLE)		+= freq_table.o
+
+##################################################################################d
+# x86 drivers.
+# Link order matters. K8 is preferred to ACPI because of firmware bugs in early
+# K8 systems. ACPI is preferred to all other hardware-specific drivers.
+# speedstep-* is preferred over p4-clockmod.
+
+obj-$(CONFIG_X86_POWERNOW_K8)		+= powernow-k8.o mperf.o
+obj-$(CONFIG_X86_ACPI_CPUFREQ)		+= acpi-cpufreq.o mperf.o
+obj-$(CONFIG_X86_PCC_CPUFREQ)		+= pcc-cpufreq.o
+obj-$(CONFIG_X86_POWERNOW_K6)		+= powernow-k6.o
+obj-$(CONFIG_X86_POWERNOW_K7)		+= powernow-k7.o
+obj-$(CONFIG_X86_LONGHAUL)		+= longhaul.o
+obj-$(CONFIG_X86_E_POWERSAVER)		+= e_powersaver.o
+obj-$(CONFIG_ELAN_CPUFREQ)		+= elanfreq.o
+obj-$(CONFIG_SC520_CPUFREQ)		+= sc520_freq.o
+obj-$(CONFIG_X86_LONGRUN)		+= longrun.o
+obj-$(CONFIG_X86_GX_SUSPMOD)		+= gx-suspmod.o
+obj-$(CONFIG_X86_SPEEDSTEP_ICH)		+= speedstep-ich.o
+obj-$(CONFIG_X86_SPEEDSTEP_LIB)		+= speedstep-lib.o
+obj-$(CONFIG_X86_SPEEDSTEP_SMI)		+= speedstep-smi.o
+obj-$(CONFIG_X86_SPEEDSTEP_CENTRINO)	+= speedstep-centrino.o
+obj-$(CONFIG_X86_P4_CLOCKMOD)		+= p4-clockmod.o
+obj-$(CONFIG_X86_CPUFREQ_NFORCE2)	+= cpufreq-nforce2.o
+
+##################################################################################d
+
+# ARM SoC drivers
+obj-$(CONFIG_UX500_SOC_DB8500)		+= db8500-cpufreq.o
diff --git a/drivers/cpufreq/acpi-cpufreq.c b/drivers/cpufreq/acpi-cpufreq.c
new file mode 100644
index 00000000..596d5dd3
--- /dev/null
+++ b/drivers/cpufreq/acpi-cpufreq.c
@@ -0,0 +1,773 @@
+/*
+ * acpi-cpufreq.c - ACPI Processor P-States Driver
+ *
+ *  Copyright (C) 2001, 2002 Andy Grover <andrew.grover@intel.com>
+ *  Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh@intel.com>
+ *  Copyright (C) 2002 - 2004 Dominik Brodowski <linux@brodo.de>
+ *  Copyright (C) 2006       Denis Sadykov <denis.m.sadykov@intel.com>
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or (at
+ *  your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful, but
+ *  WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ *  General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License along
+ *  with this program; if not, write to the Free Software Foundation, Inc.,
+ *  59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/sched.h>
+#include <linux/cpufreq.h>
+#include <linux/compiler.h>
+#include <linux/dmi.h>
+#include <linux/slab.h>
+
+#include <linux/acpi.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/uaccess.h>
+
+#include <acpi/processor.h>
+
+#include <asm/msr.h>
+#include <asm/processor.h>
+#include <asm/cpufeature.h>
+#include "mperf.h"
+
+MODULE_AUTHOR("Paul Diefenbaugh, Dominik Brodowski");
+MODULE_DESCRIPTION("ACPI Processor P-States Driver");
+MODULE_LICENSE("GPL");
+
+enum {
+	UNDEFINED_CAPABLE = 0,
+	SYSTEM_INTEL_MSR_CAPABLE,
+	SYSTEM_IO_CAPABLE,
+};
+
+#define INTEL_MSR_RANGE		(0xffff)
+
+struct acpi_cpufreq_data {
+	struct acpi_processor_performance *acpi_data;
+	struct cpufreq_frequency_table *freq_table;
+	unsigned int resume;
+	unsigned int cpu_feature;
+};
+
+static DEFINE_PER_CPU(struct acpi_cpufreq_data *, acfreq_data);
+
+/* acpi_perf_data is a pointer to percpu data. */
+static struct acpi_processor_performance __percpu *acpi_perf_data;
+
+static struct cpufreq_driver acpi_cpufreq_driver;
+
+static unsigned int acpi_pstate_strict;
+
+static int check_est_cpu(unsigned int cpuid)
+{
+	struct cpuinfo_x86 *cpu = &cpu_data(cpuid);
+
+	return cpu_has(cpu, X86_FEATURE_EST);
+}
+
+static unsigned extract_io(u32 value, struct acpi_cpufreq_data *data)
+{
+	struct acpi_processor_performance *perf;
+	int i;
+
+	perf = data->acpi_data;
+
+	for (i = 0; i < perf->state_count; i++) {
+		if (value == perf->states[i].status)
+			return data->freq_table[i].frequency;
+	}
+	return 0;
+}
+
+static unsigned extract_msr(u32 msr, struct acpi_cpufreq_data *data)
+{
+	int i;
+	struct acpi_processor_performance *perf;
+
+	msr &= INTEL_MSR_RANGE;
+	perf = data->acpi_data;
+
+	for (i = 0; data->freq_table[i].frequency != CPUFREQ_TABLE_END; i++) {
+		if (msr == perf->states[data->freq_table[i].index].status)
+			return data->freq_table[i].frequency;
+	}
+	return data->freq_table[0].frequency;
+}
+
+static unsigned extract_freq(u32 val, struct acpi_cpufreq_data *data)
+{
+	switch (data->cpu_feature) {
+	case SYSTEM_INTEL_MSR_CAPABLE:
+		return extract_msr(val, data);
+	case SYSTEM_IO_CAPABLE:
+		return extract_io(val, data);
+	default:
+		return 0;
+	}
+}
+
+struct msr_addr {
+	u32 reg;
+};
+
+struct io_addr {
+	u16 port;
+	u8 bit_width;
+};
+
+struct drv_cmd {
+	unsigned int type;
+	const struct cpumask *mask;
+	union {
+		struct msr_addr msr;
+		struct io_addr io;
+	} addr;
+	u32 val;
+};
+
+/* Called via smp_call_function_single(), on the target CPU */
+static void do_drv_read(void *_cmd)
+{
+	struct drv_cmd *cmd = _cmd;
+	u32 h;
+
+	switch (cmd->type) {
+	case SYSTEM_INTEL_MSR_CAPABLE:
+		rdmsr(cmd->addr.msr.reg, cmd->val, h);
+		break;
+	case SYSTEM_IO_CAPABLE:
+		acpi_os_read_port((acpi_io_address)cmd->addr.io.port,
+				&cmd->val,
+				(u32)cmd->addr.io.bit_width);
+		break;
+	default:
+		break;
+	}
+}
+
+/* Called via smp_call_function_many(), on the target CPUs */
+static void do_drv_write(void *_cmd)
+{
+	struct drv_cmd *cmd = _cmd;
+	u32 lo, hi;
+
+	switch (cmd->type) {
+	case SYSTEM_INTEL_MSR_CAPABLE:
+		rdmsr(cmd->addr.msr.reg, lo, hi);
+		lo = (lo & ~INTEL_MSR_RANGE) | (cmd->val & INTEL_MSR_RANGE);
+		wrmsr(cmd->addr.msr.reg, lo, hi);
+		break;
+	case SYSTEM_IO_CAPABLE:
+		acpi_os_write_port((acpi_io_address)cmd->addr.io.port,
+				cmd->val,
+				(u32)cmd->addr.io.bit_width);
+		break;
+	default:
+		break;
+	}
+}
+
+static void drv_read(struct drv_cmd *cmd)
+{
+	int err;
+	cmd->val = 0;
+
+	err = smp_call_function_any(cmd->mask, do_drv_read, cmd, 1);
+	WARN_ON_ONCE(err);	/* smp_call_function_any() was buggy? */
+}
+
+static void drv_write(struct drv_cmd *cmd)
+{
+	int this_cpu;
+
+	this_cpu = get_cpu();
+	if (cpumask_test_cpu(this_cpu, cmd->mask))
+		do_drv_write(cmd);
+	smp_call_function_many(cmd->mask, do_drv_write, cmd, 1);
+	put_cpu();
+}
+
+static u32 get_cur_val(const struct cpumask *mask)
+{
+	struct acpi_processor_performance *perf;
+	struct drv_cmd cmd;
+
+	if (unlikely(cpumask_empty(mask)))
+		return 0;
+
+	switch (per_cpu(acfreq_data, cpumask_first(mask))->cpu_feature) {
+	case SYSTEM_INTEL_MSR_CAPABLE:
+		cmd.type = SYSTEM_INTEL_MSR_CAPABLE;
+		cmd.addr.msr.reg = MSR_IA32_PERF_STATUS;
+		break;
+	case SYSTEM_IO_CAPABLE:
+		cmd.type = SYSTEM_IO_CAPABLE;
+		perf = per_cpu(acfreq_data, cpumask_first(mask))->acpi_data;
+		cmd.addr.io.port = perf->control_register.address;
+		cmd.addr.io.bit_width = perf->control_register.bit_width;
+		break;
+	default:
+		return 0;
+	}
+
+	cmd.mask = mask;
+	drv_read(&cmd);
+
+	pr_debug("get_cur_val = %u\n", cmd.val);
+
+	return cmd.val;
+}
+
+static unsigned int get_cur_freq_on_cpu(unsigned int cpu)
+{
+	struct acpi_cpufreq_data *data = per_cpu(acfreq_data, cpu);
+	unsigned int freq;
+	unsigned int cached_freq;
+
+	pr_debug("get_cur_freq_on_cpu (%d)\n", cpu);
+
+	if (unlikely(data == NULL ||
+		     data->acpi_data == NULL || data->freq_table == NULL)) {
+		return 0;
+	}
+
+	cached_freq = data->freq_table[data->acpi_data->state].frequency;
+	freq = extract_freq(get_cur_val(cpumask_of(cpu)), data);
+	if (freq != cached_freq) {
+		/*
+		 * The dreaded BIOS frequency change behind our back.
+		 * Force set the frequency on next target call.
+		 */
+		data->resume = 1;
+	}
+
+	pr_debug("cur freq = %u\n", freq);
+
+	return freq;
+}
+
+static unsigned int check_freqs(const struct cpumask *mask, unsigned int freq,
+				struct acpi_cpufreq_data *data)
+{
+	unsigned int cur_freq;
+	unsigned int i;
+
+	for (i = 0; i < 100; i++) {
+		cur_freq = extract_freq(get_cur_val(mask), data);
+		if (cur_freq == freq)
+			return 1;
+		udelay(10);
+	}
+	return 0;
+}
+
+static int acpi_cpufreq_target(struct cpufreq_policy *policy,
+			       unsigned int target_freq, unsigned int relation)
+{
+	struct acpi_cpufreq_data *data = per_cpu(acfreq_data, policy->cpu);
+	struct acpi_processor_performance *perf;
+	struct cpufreq_freqs freqs;
+	struct drv_cmd cmd;
+	unsigned int next_state = 0; /* Index into freq_table */
+	unsigned int next_perf_state = 0; /* Index into perf table */
+	unsigned int i;
+	int result = 0;
+
+	pr_debug("acpi_cpufreq_target %d (%d)\n", target_freq, policy->cpu);
+
+	if (unlikely(data == NULL ||
+	     data->acpi_data == NULL || data->freq_table == NULL)) {
+		return -ENODEV;
+	}
+
+	perf = data->acpi_data;
+	result = cpufreq_frequency_table_target(policy,
+						data->freq_table,
+						target_freq,
+						relation, &next_state);
+	if (unlikely(result)) {
+		result = -ENODEV;
+		goto out;
+	}
+
+	next_perf_state = data->freq_table[next_state].index;
+	if (perf->state == next_perf_state) {
+		if (unlikely(data->resume)) {
+			pr_debug("Called after resume, resetting to P%d\n",
+				next_perf_state);
+			data->resume = 0;
+		} else {
+			pr_debug("Already at target state (P%d)\n",
+				next_perf_state);
+			goto out;
+		}
+	}
+
+	switch (data->cpu_feature) {
+	case SYSTEM_INTEL_MSR_CAPABLE:
+		cmd.type = SYSTEM_INTEL_MSR_CAPABLE;
+		cmd.addr.msr.reg = MSR_IA32_PERF_CTL;
+		cmd.val = (u32) perf->states[next_perf_state].control;
+		break;
+	case SYSTEM_IO_CAPABLE:
+		cmd.type = SYSTEM_IO_CAPABLE;
+		cmd.addr.io.port = perf->control_register.address;
+		cmd.addr.io.bit_width = perf->control_register.bit_width;
+		cmd.val = (u32) perf->states[next_perf_state].control;
+		break;
+	default:
+		result = -ENODEV;
+		goto out;
+	}
+
+	/* cpufreq holds the hotplug lock, so we are safe from here on */
+	if (policy->shared_type != CPUFREQ_SHARED_TYPE_ANY)
+		cmd.mask = policy->cpus;
+	else
+		cmd.mask = cpumask_of(policy->cpu);
+
+	freqs.old = perf->states[perf->state].core_frequency * 1000;
+	freqs.new = data->freq_table[next_state].frequency;
+	for_each_cpu(i, policy->cpus) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	}
+
+	drv_write(&cmd);
+
+	if (acpi_pstate_strict) {
+		if (!check_freqs(cmd.mask, freqs.new, data)) {
+			pr_debug("acpi_cpufreq_target failed (%d)\n",
+				policy->cpu);
+			result = -EAGAIN;
+			goto out;
+		}
+	}
+
+	for_each_cpu(i, policy->cpus) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	}
+	perf->state = next_perf_state;
+
+out:
+	return result;
+}
+
+static int acpi_cpufreq_verify(struct cpufreq_policy *policy)
+{
+	struct acpi_cpufreq_data *data = per_cpu(acfreq_data, policy->cpu);
+
+	pr_debug("acpi_cpufreq_verify\n");
+
+	return cpufreq_frequency_table_verify(policy, data->freq_table);
+}
+
+static unsigned long
+acpi_cpufreq_guess_freq(struct acpi_cpufreq_data *data, unsigned int cpu)
+{
+	struct acpi_processor_performance *perf = data->acpi_data;
+
+	if (cpu_khz) {
+		/* search the closest match to cpu_khz */
+		unsigned int i;
+		unsigned long freq;
+		unsigned long freqn = perf->states[0].core_frequency * 1000;
+
+		for (i = 0; i < (perf->state_count-1); i++) {
+			freq = freqn;
+			freqn = perf->states[i+1].core_frequency * 1000;
+			if ((2 * cpu_khz) > (freqn + freq)) {
+				perf->state = i;
+				return freq;
+			}
+		}
+		perf->state = perf->state_count-1;
+		return freqn;
+	} else {
+		/* assume CPU is at P0... */
+		perf->state = 0;
+		return perf->states[0].core_frequency * 1000;
+	}
+}
+
+static void free_acpi_perf_data(void)
+{
+	unsigned int i;
+
+	/* Freeing a NULL pointer is OK, and alloc_percpu zeroes. */
+	for_each_possible_cpu(i)
+		free_cpumask_var(per_cpu_ptr(acpi_perf_data, i)
+				 ->shared_cpu_map);
+	free_percpu(acpi_perf_data);
+}
+
+/*
+ * acpi_cpufreq_early_init - initialize ACPI P-States library
+ *
+ * Initialize the ACPI P-States library (drivers/acpi/processor_perflib.c)
+ * in order to determine correct frequency and voltage pairings. We can
+ * do _PDC and _PSD and find out the processor dependency for the
+ * actual init that will happen later...
+ */
+static int __init acpi_cpufreq_early_init(void)
+{
+	unsigned int i;
+	pr_debug("acpi_cpufreq_early_init\n");
+
+	acpi_perf_data = alloc_percpu(struct acpi_processor_performance);
+	if (!acpi_perf_data) {
+		pr_debug("Memory allocation error for acpi_perf_data.\n");
+		return -ENOMEM;
+	}
+	for_each_possible_cpu(i) {
+		if (!zalloc_cpumask_var_node(
+			&per_cpu_ptr(acpi_perf_data, i)->shared_cpu_map,
+			GFP_KERNEL, cpu_to_node(i))) {
+
+			/* Freeing a NULL pointer is OK: alloc_percpu zeroes. */
+			free_acpi_perf_data();
+			return -ENOMEM;
+		}
+	}
+
+	/* Do initialization in ACPI core */
+	acpi_processor_preregister_performance(acpi_perf_data);
+	return 0;
+}
+
+#ifdef CONFIG_SMP
+/*
+ * Some BIOSes do SW_ANY coordination internally, either set it up in hw
+ * or do it in BIOS firmware and won't inform about it to OS. If not
+ * detected, this has a side effect of making CPU run at a different speed
+ * than OS intended it to run at. Detect it and handle it cleanly.
+ */
+static int bios_with_sw_any_bug;
+
+static int sw_any_bug_found(const struct dmi_system_id *d)
+{
+	bios_with_sw_any_bug = 1;
+	return 0;
+}
+
+static const struct dmi_system_id sw_any_bug_dmi_table[] = {
+	{
+		.callback = sw_any_bug_found,
+		.ident = "Supermicro Server X6DLP",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Supermicro"),
+			DMI_MATCH(DMI_BIOS_VERSION, "080010"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "X6DLP"),
+		},
+	},
+	{ }
+};
+
+static int acpi_cpufreq_blacklist(struct cpuinfo_x86 *c)
+{
+	/* Intel Xeon Processor 7100 Series Specification Update
+	 * http://www.intel.com/Assets/PDF/specupdate/314554.pdf
+	 * AL30: A Machine Check Exception (MCE) Occurring during an
+	 * Enhanced Intel SpeedStep Technology Ratio Change May Cause
+	 * Both Processor Cores to Lock Up. */
+	if (c->x86_vendor == X86_VENDOR_INTEL) {
+		if ((c->x86 == 15) &&
+		    (c->x86_model == 6) &&
+		    (c->x86_mask == 8)) {
+			printk(KERN_INFO "acpi-cpufreq: Intel(R) "
+			    "Xeon(R) 7100 Errata AL30, processors may "
+			    "lock up on frequency changes: disabling "
+			    "acpi-cpufreq.\n");
+			return -ENODEV;
+		    }
+		}
+	return 0;
+}
+#endif
+
+static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy)
+{
+	unsigned int i;
+	unsigned int valid_states = 0;
+	unsigned int cpu = policy->cpu;
+	struct acpi_cpufreq_data *data;
+	unsigned int result = 0;
+	struct cpuinfo_x86 *c = &cpu_data(policy->cpu);
+	struct acpi_processor_performance *perf;
+#ifdef CONFIG_SMP
+	static int blacklisted;
+#endif
+
+	pr_debug("acpi_cpufreq_cpu_init\n");
+
+#ifdef CONFIG_SMP
+	if (blacklisted)
+		return blacklisted;
+	blacklisted = acpi_cpufreq_blacklist(c);
+	if (blacklisted)
+		return blacklisted;
+#endif
+
+	data = kzalloc(sizeof(struct acpi_cpufreq_data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	data->acpi_data = per_cpu_ptr(acpi_perf_data, cpu);
+	per_cpu(acfreq_data, cpu) = data;
+
+	if (cpu_has(c, X86_FEATURE_CONSTANT_TSC))
+		acpi_cpufreq_driver.flags |= CPUFREQ_CONST_LOOPS;
+
+	result = acpi_processor_register_performance(data->acpi_data, cpu);
+	if (result)
+		goto err_free;
+
+	perf = data->acpi_data;
+	policy->shared_type = perf->shared_type;
+
+	/*
+	 * Will let policy->cpus know about dependency only when software
+	 * coordination is required.
+	 */
+	if (policy->shared_type == CPUFREQ_SHARED_TYPE_ALL ||
+	    policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) {
+		cpumask_copy(policy->cpus, perf->shared_cpu_map);
+	}
+	cpumask_copy(policy->related_cpus, perf->shared_cpu_map);
+
+#ifdef CONFIG_SMP
+	dmi_check_system(sw_any_bug_dmi_table);
+	if (bios_with_sw_any_bug && cpumask_weight(policy->cpus) == 1) {
+		policy->shared_type = CPUFREQ_SHARED_TYPE_ALL;
+		cpumask_copy(policy->cpus, cpu_core_mask(cpu));
+	}
+#endif
+
+	/* capability check */
+	if (perf->state_count <= 1) {
+		pr_debug("No P-States\n");
+		result = -ENODEV;
+		goto err_unreg;
+	}
+
+	if (perf->control_register.space_id != perf->status_register.space_id) {
+		result = -ENODEV;
+		goto err_unreg;
+	}
+
+	switch (perf->control_register.space_id) {
+	case ACPI_ADR_SPACE_SYSTEM_IO:
+		pr_debug("SYSTEM IO addr space\n");
+		data->cpu_feature = SYSTEM_IO_CAPABLE;
+		break;
+	case ACPI_ADR_SPACE_FIXED_HARDWARE:
+		pr_debug("HARDWARE addr space\n");
+		if (!check_est_cpu(cpu)) {
+			result = -ENODEV;
+			goto err_unreg;
+		}
+		data->cpu_feature = SYSTEM_INTEL_MSR_CAPABLE;
+		break;
+	default:
+		pr_debug("Unknown addr space %d\n",
+			(u32) (perf->control_register.space_id));
+		result = -ENODEV;
+		goto err_unreg;
+	}
+
+	data->freq_table = kmalloc(sizeof(struct cpufreq_frequency_table) *
+		    (perf->state_count+1), GFP_KERNEL);
+	if (!data->freq_table) {
+		result = -ENOMEM;
+		goto err_unreg;
+	}
+
+	/* detect transition latency */
+	policy->cpuinfo.transition_latency = 0;
+	for (i = 0; i < perf->state_count; i++) {
+		if ((perf->states[i].transition_latency * 1000) >
+		    policy->cpuinfo.transition_latency)
+			policy->cpuinfo.transition_latency =
+			    perf->states[i].transition_latency * 1000;
+	}
+
+	/* Check for high latency (>20uS) from buggy BIOSes, like on T42 */
+	if (perf->control_register.space_id == ACPI_ADR_SPACE_FIXED_HARDWARE &&
+	    policy->cpuinfo.transition_latency > 20 * 1000) {
+		policy->cpuinfo.transition_latency = 20 * 1000;
+		printk_once(KERN_INFO
+			    "P-state transition latency capped at 20 uS\n");
+	}
+
+	/* table init */
+	for (i = 0; i < perf->state_count; i++) {
+		if (i > 0 && perf->states[i].core_frequency >=
+		    data->freq_table[valid_states-1].frequency / 1000)
+			continue;
+
+		data->freq_table[valid_states].index = i;
+		data->freq_table[valid_states].frequency =
+		    perf->states[i].core_frequency * 1000;
+		valid_states++;
+	}
+	data->freq_table[valid_states].frequency = CPUFREQ_TABLE_END;
+	perf->state = 0;
+
+	result = cpufreq_frequency_table_cpuinfo(policy, data->freq_table);
+	if (result)
+		goto err_freqfree;
+
+	if (perf->states[0].core_frequency * 1000 != policy->cpuinfo.max_freq)
+		printk(KERN_WARNING FW_WARN "P-state 0 is not max freq\n");
+
+	switch (perf->control_register.space_id) {
+	case ACPI_ADR_SPACE_SYSTEM_IO:
+		/* Current speed is unknown and not detectable by IO port */
+		policy->cur = acpi_cpufreq_guess_freq(data, policy->cpu);
+		break;
+	case ACPI_ADR_SPACE_FIXED_HARDWARE:
+		acpi_cpufreq_driver.get = get_cur_freq_on_cpu;
+		policy->cur = get_cur_freq_on_cpu(cpu);
+		break;
+	default:
+		break;
+	}
+
+	/* notify BIOS that we exist */
+	acpi_processor_notify_smm(THIS_MODULE);
+
+	/* Check for APERF/MPERF support in hardware */
+	if (cpu_has(c, X86_FEATURE_APERFMPERF))
+		acpi_cpufreq_driver.getavg = cpufreq_get_measured_perf;
+
+	pr_debug("CPU%u - ACPI performance management activated.\n", cpu);
+	for (i = 0; i < perf->state_count; i++)
+		pr_debug("     %cP%d: %d MHz, %d mW, %d uS\n",
+			(i == perf->state ? '*' : ' '), i,
+			(u32) perf->states[i].core_frequency,
+			(u32) perf->states[i].power,
+			(u32) perf->states[i].transition_latency);
+
+	cpufreq_frequency_table_get_attr(data->freq_table, policy->cpu);
+
+	/*
+	 * the first call to ->target() should result in us actually
+	 * writing something to the appropriate registers.
+	 */
+	data->resume = 1;
+
+	return result;
+
+err_freqfree:
+	kfree(data->freq_table);
+err_unreg:
+	acpi_processor_unregister_performance(perf, cpu);
+err_free:
+	kfree(data);
+	per_cpu(acfreq_data, cpu) = NULL;
+
+	return result;
+}
+
+static int acpi_cpufreq_cpu_exit(struct cpufreq_policy *policy)
+{
+	struct acpi_cpufreq_data *data = per_cpu(acfreq_data, policy->cpu);
+
+	pr_debug("acpi_cpufreq_cpu_exit\n");
+
+	if (data) {
+		cpufreq_frequency_table_put_attr(policy->cpu);
+		per_cpu(acfreq_data, policy->cpu) = NULL;
+		acpi_processor_unregister_performance(data->acpi_data,
+						      policy->cpu);
+		kfree(data->freq_table);
+		kfree(data);
+	}
+
+	return 0;
+}
+
+static int acpi_cpufreq_resume(struct cpufreq_policy *policy)
+{
+	struct acpi_cpufreq_data *data = per_cpu(acfreq_data, policy->cpu);
+
+	pr_debug("acpi_cpufreq_resume\n");
+
+	data->resume = 1;
+
+	return 0;
+}
+
+static struct freq_attr *acpi_cpufreq_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver acpi_cpufreq_driver = {
+	.verify		= acpi_cpufreq_verify,
+	.target		= acpi_cpufreq_target,
+	.bios_limit	= acpi_processor_get_bios_limit,
+	.init		= acpi_cpufreq_cpu_init,
+	.exit		= acpi_cpufreq_cpu_exit,
+	.resume		= acpi_cpufreq_resume,
+	.name		= "acpi-cpufreq",
+	.owner		= THIS_MODULE,
+	.attr		= acpi_cpufreq_attr,
+};
+
+static int __init acpi_cpufreq_init(void)
+{
+	int ret;
+
+	if (acpi_disabled)
+		return 0;
+
+	pr_debug("acpi_cpufreq_init\n");
+
+	ret = acpi_cpufreq_early_init();
+	if (ret)
+		return ret;
+
+	ret = cpufreq_register_driver(&acpi_cpufreq_driver);
+	if (ret)
+		free_acpi_perf_data();
+
+	return ret;
+}
+
+static void __exit acpi_cpufreq_exit(void)
+{
+	pr_debug("acpi_cpufreq_exit\n");
+
+	cpufreq_unregister_driver(&acpi_cpufreq_driver);
+
+	free_acpi_perf_data();
+}
+
+module_param(acpi_pstate_strict, uint, 0644);
+MODULE_PARM_DESC(acpi_pstate_strict,
+	"value 0 or non-zero. non-zero -> strict ACPI checks are "
+	"performed during frequency changes.");
+
+late_initcall(acpi_cpufreq_init);
+module_exit(acpi_cpufreq_exit);
+
+MODULE_ALIAS("acpi");
diff --git a/drivers/cpufreq/cpufreq-nforce2.c b/drivers/cpufreq/cpufreq-nforce2.c
new file mode 100644
index 00000000..7bac8088
--- /dev/null
+++ b/drivers/cpufreq/cpufreq-nforce2.c
@@ -0,0 +1,444 @@
+/*
+ * (C) 2004-2006  Sebastian Witt <se.witt@gmx.net>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *  Based upon reverse engineered information
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+
+#define NFORCE2_XTAL 25
+#define NFORCE2_BOOTFSB 0x48
+#define NFORCE2_PLLENABLE 0xa8
+#define NFORCE2_PLLREG 0xa4
+#define NFORCE2_PLLADR 0xa0
+#define NFORCE2_PLL(mul, div) (0x100000 | (mul << 8) | div)
+
+#define NFORCE2_MIN_FSB 50
+#define NFORCE2_SAFE_DISTANCE 50
+
+/* Delay in ms between FSB changes */
+/* #define NFORCE2_DELAY 10 */
+
+/*
+ * nforce2_chipset:
+ * FSB is changed using the chipset
+ */
+static struct pci_dev *nforce2_dev;
+
+/* fid:
+ * multiplier * 10
+ */
+static int fid;
+
+/* min_fsb, max_fsb:
+ * minimum and maximum FSB (= FSB at boot time)
+ */
+static int min_fsb;
+static int max_fsb;
+
+MODULE_AUTHOR("Sebastian Witt <se.witt@gmx.net>");
+MODULE_DESCRIPTION("nForce2 FSB changing cpufreq driver");
+MODULE_LICENSE("GPL");
+
+module_param(fid, int, 0444);
+module_param(min_fsb, int, 0444);
+
+MODULE_PARM_DESC(fid, "CPU multiplier to use (11.5 = 115)");
+MODULE_PARM_DESC(min_fsb,
+		"Minimum FSB to use, if not defined: current FSB - 50");
+
+#define PFX "cpufreq-nforce2: "
+
+/**
+ * nforce2_calc_fsb - calculate FSB
+ * @pll: PLL value
+ *
+ *   Calculates FSB from PLL value
+ */
+static int nforce2_calc_fsb(int pll)
+{
+	unsigned char mul, div;
+
+	mul = (pll >> 8) & 0xff;
+	div = pll & 0xff;
+
+	if (div > 0)
+		return NFORCE2_XTAL * mul / div;
+
+	return 0;
+}
+
+/**
+ * nforce2_calc_pll - calculate PLL value
+ * @fsb: FSB
+ *
+ *   Calculate PLL value for given FSB
+ */
+static int nforce2_calc_pll(unsigned int fsb)
+{
+	unsigned char xmul, xdiv;
+	unsigned char mul = 0, div = 0;
+	int tried = 0;
+
+	/* Try to calculate multiplier and divider up to 4 times */
+	while (((mul == 0) || (div == 0)) && (tried <= 3)) {
+		for (xdiv = 2; xdiv <= 0x80; xdiv++)
+			for (xmul = 1; xmul <= 0xfe; xmul++)
+				if (nforce2_calc_fsb(NFORCE2_PLL(xmul, xdiv)) ==
+				    fsb + tried) {
+					mul = xmul;
+					div = xdiv;
+				}
+		tried++;
+	}
+
+	if ((mul == 0) || (div == 0))
+		return -1;
+
+	return NFORCE2_PLL(mul, div);
+}
+
+/**
+ * nforce2_write_pll - write PLL value to chipset
+ * @pll: PLL value
+ *
+ *   Writes new FSB PLL value to chipset
+ */
+static void nforce2_write_pll(int pll)
+{
+	int temp;
+
+	/* Set the pll addr. to 0x00 */
+	pci_write_config_dword(nforce2_dev, NFORCE2_PLLADR, 0);
+
+	/* Now write the value in all 64 registers */
+	for (temp = 0; temp <= 0x3f; temp++)
+		pci_write_config_dword(nforce2_dev, NFORCE2_PLLREG, pll);
+
+	return;
+}
+
+/**
+ * nforce2_fsb_read - Read FSB
+ *
+ *   Read FSB from chipset
+ *   If bootfsb != 0, return FSB at boot-time
+ */
+static unsigned int nforce2_fsb_read(int bootfsb)
+{
+	struct pci_dev *nforce2_sub5;
+	u32 fsb, temp = 0;
+
+	/* Get chipset boot FSB from subdevice 5 (FSB at boot-time) */
+	nforce2_sub5 = pci_get_subsys(PCI_VENDOR_ID_NVIDIA, 0x01EF,
+				PCI_ANY_ID, PCI_ANY_ID, NULL);
+	if (!nforce2_sub5)
+		return 0;
+
+	pci_read_config_dword(nforce2_sub5, NFORCE2_BOOTFSB, &fsb);
+	fsb /= 1000000;
+
+	/* Check if PLL register is already set */
+	pci_read_config_byte(nforce2_dev, NFORCE2_PLLENABLE, (u8 *)&temp);
+
+	if (bootfsb || !temp)
+		return fsb;
+
+	/* Use PLL register FSB value */
+	pci_read_config_dword(nforce2_dev, NFORCE2_PLLREG, &temp);
+	fsb = nforce2_calc_fsb(temp);
+
+	return fsb;
+}
+
+/**
+ * nforce2_set_fsb - set new FSB
+ * @fsb: New FSB
+ *
+ *   Sets new FSB
+ */
+static int nforce2_set_fsb(unsigned int fsb)
+{
+	u32 temp = 0;
+	unsigned int tfsb;
+	int diff;
+	int pll = 0;
+
+	if ((fsb > max_fsb) || (fsb < NFORCE2_MIN_FSB)) {
+		printk(KERN_ERR PFX "FSB %d is out of range!\n", fsb);
+		return -EINVAL;
+	}
+
+	tfsb = nforce2_fsb_read(0);
+	if (!tfsb) {
+		printk(KERN_ERR PFX "Error while reading the FSB\n");
+		return -EINVAL;
+	}
+
+	/* First write? Then set actual value */
+	pci_read_config_byte(nforce2_dev, NFORCE2_PLLENABLE, (u8 *)&temp);
+	if (!temp) {
+		pll = nforce2_calc_pll(tfsb);
+
+		if (pll < 0)
+			return -EINVAL;
+
+		nforce2_write_pll(pll);
+	}
+
+	/* Enable write access */
+	temp = 0x01;
+	pci_write_config_byte(nforce2_dev, NFORCE2_PLLENABLE, (u8)temp);
+
+	diff = tfsb - fsb;
+
+	if (!diff)
+		return 0;
+
+	while ((tfsb != fsb) && (tfsb <= max_fsb) && (tfsb >= min_fsb)) {
+		if (diff < 0)
+			tfsb++;
+		else
+			tfsb--;
+
+		/* Calculate the PLL reg. value */
+		pll = nforce2_calc_pll(tfsb);
+		if (pll == -1)
+			return -EINVAL;
+
+		nforce2_write_pll(pll);
+#ifdef NFORCE2_DELAY
+		mdelay(NFORCE2_DELAY);
+#endif
+	}
+
+	temp = 0x40;
+	pci_write_config_byte(nforce2_dev, NFORCE2_PLLADR, (u8)temp);
+
+	return 0;
+}
+
+/**
+ * nforce2_get - get the CPU frequency
+ * @cpu: CPU number
+ *
+ * Returns the CPU frequency
+ */
+static unsigned int nforce2_get(unsigned int cpu)
+{
+	if (cpu)
+		return 0;
+	return nforce2_fsb_read(0) * fid * 100;
+}
+
+/**
+ * nforce2_target - set a new CPUFreq policy
+ * @policy: new policy
+ * @target_freq: the target frequency
+ * @relation: how that frequency relates to achieved frequency
+ *  (CPUFREQ_RELATION_L or CPUFREQ_RELATION_H)
+ *
+ * Sets a new CPUFreq policy.
+ */
+static int nforce2_target(struct cpufreq_policy *policy,
+			  unsigned int target_freq, unsigned int relation)
+{
+/*        unsigned long         flags; */
+	struct cpufreq_freqs freqs;
+	unsigned int target_fsb;
+
+	if ((target_freq > policy->max) || (target_freq < policy->min))
+		return -EINVAL;
+
+	target_fsb = target_freq / (fid * 100);
+
+	freqs.old = nforce2_get(policy->cpu);
+	freqs.new = target_fsb * fid * 100;
+	freqs.cpu = 0;		/* Only one CPU on nForce2 platforms */
+
+	if (freqs.old == freqs.new)
+		return 0;
+
+	pr_debug("Old CPU frequency %d kHz, new %d kHz\n",
+	       freqs.old, freqs.new);
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	/* Disable IRQs */
+	/* local_irq_save(flags); */
+
+	if (nforce2_set_fsb(target_fsb) < 0)
+		printk(KERN_ERR PFX "Changing FSB to %d failed\n",
+			target_fsb);
+	else
+		pr_debug("Changed FSB successfully to %d\n",
+			target_fsb);
+
+	/* Enable IRQs */
+	/* local_irq_restore(flags); */
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+
+	return 0;
+}
+
+/**
+ * nforce2_verify - verifies a new CPUFreq policy
+ * @policy: new policy
+ */
+static int nforce2_verify(struct cpufreq_policy *policy)
+{
+	unsigned int fsb_pol_max;
+
+	fsb_pol_max = policy->max / (fid * 100);
+
+	if (policy->min < (fsb_pol_max * fid * 100))
+		policy->max = (fsb_pol_max + 1) * fid * 100;
+
+	cpufreq_verify_within_limits(policy,
+				     policy->cpuinfo.min_freq,
+				     policy->cpuinfo.max_freq);
+	return 0;
+}
+
+static int nforce2_cpu_init(struct cpufreq_policy *policy)
+{
+	unsigned int fsb;
+	unsigned int rfid;
+
+	/* capability check */
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	/* Get current FSB */
+	fsb = nforce2_fsb_read(0);
+
+	if (!fsb)
+		return -EIO;
+
+	/* FIX: Get FID from CPU */
+	if (!fid) {
+		if (!cpu_khz) {
+			printk(KERN_WARNING PFX
+			"cpu_khz not set, can't calculate multiplier!\n");
+			return -ENODEV;
+		}
+
+		fid = cpu_khz / (fsb * 100);
+		rfid = fid % 5;
+
+		if (rfid) {
+			if (rfid > 2)
+				fid += 5 - rfid;
+			else
+				fid -= rfid;
+		}
+	}
+
+	printk(KERN_INFO PFX "FSB currently at %i MHz, FID %d.%d\n", fsb,
+	       fid / 10, fid % 10);
+
+	/* Set maximum FSB to FSB at boot time */
+	max_fsb = nforce2_fsb_read(1);
+
+	if (!max_fsb)
+		return -EIO;
+
+	if (!min_fsb)
+		min_fsb = max_fsb - NFORCE2_SAFE_DISTANCE;
+
+	if (min_fsb < NFORCE2_MIN_FSB)
+		min_fsb = NFORCE2_MIN_FSB;
+
+	/* cpuinfo and default policy values */
+	policy->cpuinfo.min_freq = min_fsb * fid * 100;
+	policy->cpuinfo.max_freq = max_fsb * fid * 100;
+	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
+	policy->cur = nforce2_get(policy->cpu);
+	policy->min = policy->cpuinfo.min_freq;
+	policy->max = policy->cpuinfo.max_freq;
+
+	return 0;
+}
+
+static int nforce2_cpu_exit(struct cpufreq_policy *policy)
+{
+	return 0;
+}
+
+static struct cpufreq_driver nforce2_driver = {
+	.name = "nforce2",
+	.verify = nforce2_verify,
+	.target = nforce2_target,
+	.get = nforce2_get,
+	.init = nforce2_cpu_init,
+	.exit = nforce2_cpu_exit,
+	.owner = THIS_MODULE,
+};
+
+/**
+ * nforce2_detect_chipset - detect the Southbridge which contains FSB PLL logic
+ *
+ * Detects nForce2 A2 and C1 stepping
+ *
+ */
+static int nforce2_detect_chipset(void)
+{
+	nforce2_dev = pci_get_subsys(PCI_VENDOR_ID_NVIDIA,
+					PCI_DEVICE_ID_NVIDIA_NFORCE2,
+					PCI_ANY_ID, PCI_ANY_ID, NULL);
+
+	if (nforce2_dev == NULL)
+		return -ENODEV;
+
+	printk(KERN_INFO PFX "Detected nForce2 chipset revision %X\n",
+	       nforce2_dev->revision);
+	printk(KERN_INFO PFX
+	       "FSB changing is maybe unstable and can lead to "
+	       "crashes and data loss.\n");
+
+	return 0;
+}
+
+/**
+ * nforce2_init - initializes the nForce2 CPUFreq driver
+ *
+ * Initializes the nForce2 FSB support. Returns -ENODEV on unsupported
+ * devices, -EINVAL on problems during initiatization, and zero on
+ * success.
+ */
+static int __init nforce2_init(void)
+{
+	/* TODO: do we need to detect the processor? */
+
+	/* detect chipset */
+	if (nforce2_detect_chipset()) {
+		printk(KERN_INFO PFX "No nForce2 chipset.\n");
+		return -ENODEV;
+	}
+
+	return cpufreq_register_driver(&nforce2_driver);
+}
+
+/**
+ * nforce2_exit - unregisters cpufreq module
+ *
+ *   Unregisters nForce2 FSB change support.
+ */
+static void __exit nforce2_exit(void)
+{
+	cpufreq_unregister_driver(&nforce2_driver);
+}
+
+module_init(nforce2_init);
+module_exit(nforce2_exit);
+
diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
new file mode 100644
index 00000000..0a5bea9e
--- /dev/null
+++ b/drivers/cpufreq/cpufreq.c
@@ -0,0 +1,1897 @@
+/*
+ *  linux/drivers/cpufreq/cpufreq.c
+ *
+ *  Copyright (C) 2001 Russell King
+ *            (C) 2002 - 2003 Dominik Brodowski <linux@brodo.de>
+ *
+ *  Oct 2005 - Ashok Raj <ashok.raj@intel.com>
+ *	Added handling for CPU hotplug
+ *  Feb 2006 - Jacob Shin <jacob.shin@amd.com>
+ *	Fix handling for CPU hotplug -- affected CPUs
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/notifier.h>
+#include <linux/cpufreq.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/device.h>
+#include <linux/slab.h>
+#include <linux/cpu.h>
+#include <linux/completion.h>
+#include <linux/mutex.h>
+#include <linux/syscore_ops.h>
+
+#include <trace/events/power.h>
+
+/**
+ * The "cpufreq driver" - the arch- or hardware-dependent low
+ * level driver of CPUFreq support, and its spinlock. This lock
+ * also protects the cpufreq_cpu_data array.
+ */
+static struct cpufreq_driver *cpufreq_driver;
+static DEFINE_PER_CPU(struct cpufreq_policy *, cpufreq_cpu_data);
+#ifdef CONFIG_HOTPLUG_CPU
+/* This one keeps track of the previously set governor of a removed CPU */
+static DEFINE_PER_CPU(char[CPUFREQ_NAME_LEN], cpufreq_cpu_governor);
+#endif
+static DEFINE_SPINLOCK(cpufreq_driver_lock);
+
+/*
+ * cpu_policy_rwsem is a per CPU reader-writer semaphore designed to cure
+ * all cpufreq/hotplug/workqueue/etc related lock issues.
+ *
+ * The rules for this semaphore:
+ * - Any routine that wants to read from the policy structure will
+ *   do a down_read on this semaphore.
+ * - Any routine that will write to the policy structure and/or may take away
+ *   the policy altogether (eg. CPU hotplug), will hold this lock in write
+ *   mode before doing so.
+ *
+ * Additional rules:
+ * - All holders of the lock should check to make sure that the CPU they
+ *   are concerned with are online after they get the lock.
+ * - Governor routines that can be called in cpufreq hotplug path should not
+ *   take this sem as top level hotplug notifier handler takes this.
+ * - Lock should not be held across
+ *     __cpufreq_governor(data, CPUFREQ_GOV_STOP);
+ */
+static DEFINE_PER_CPU(int, cpufreq_policy_cpu);
+static DEFINE_PER_CPU(struct rw_semaphore, cpu_policy_rwsem);
+
+#define lock_policy_rwsem(mode, cpu)					\
+static int lock_policy_rwsem_##mode					\
+(int cpu)								\
+{									\
+	int policy_cpu = per_cpu(cpufreq_policy_cpu, cpu);		\
+	BUG_ON(policy_cpu == -1);					\
+	down_##mode(&per_cpu(cpu_policy_rwsem, policy_cpu));		\
+	if (unlikely(!cpu_online(cpu))) {				\
+		up_##mode(&per_cpu(cpu_policy_rwsem, policy_cpu));	\
+		return -1;						\
+	}								\
+									\
+	return 0;							\
+}
+
+lock_policy_rwsem(read, cpu);
+
+lock_policy_rwsem(write, cpu);
+
+static void unlock_policy_rwsem_read(int cpu)
+{
+	int policy_cpu = per_cpu(cpufreq_policy_cpu, cpu);
+	BUG_ON(policy_cpu == -1);
+	up_read(&per_cpu(cpu_policy_rwsem, policy_cpu));
+}
+
+static void unlock_policy_rwsem_write(int cpu)
+{
+	int policy_cpu = per_cpu(cpufreq_policy_cpu, cpu);
+	BUG_ON(policy_cpu == -1);
+	up_write(&per_cpu(cpu_policy_rwsem, policy_cpu));
+}
+
+
+/* internal prototypes */
+static int __cpufreq_governor(struct cpufreq_policy *policy,
+		unsigned int event);
+static unsigned int __cpufreq_get(unsigned int cpu);
+static void handle_update(struct work_struct *work);
+
+/**
+ * Two notifier lists: the "policy" list is involved in the
+ * validation process for a new CPU frequency policy; the
+ * "transition" list for kernel code that needs to handle
+ * changes to devices when the CPU clock speed changes.
+ * The mutex locks both lists.
+ */
+static BLOCKING_NOTIFIER_HEAD(cpufreq_policy_notifier_list);
+static struct srcu_notifier_head cpufreq_transition_notifier_list;
+
+static bool init_cpufreq_transition_notifier_list_called;
+static int __init init_cpufreq_transition_notifier_list(void)
+{
+	srcu_init_notifier_head(&cpufreq_transition_notifier_list);
+	init_cpufreq_transition_notifier_list_called = true;
+	return 0;
+}
+pure_initcall(init_cpufreq_transition_notifier_list);
+
+static LIST_HEAD(cpufreq_governor_list);
+static DEFINE_MUTEX(cpufreq_governor_mutex);
+
+struct cpufreq_policy *cpufreq_cpu_get(unsigned int cpu)
+{
+	struct cpufreq_policy *data;
+	unsigned long flags;
+
+	if (cpu >= nr_cpu_ids)
+		goto err_out;
+
+	/* get the cpufreq driver */
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+
+	if (!cpufreq_driver)
+		goto err_out_unlock;
+
+	if (!try_module_get(cpufreq_driver->owner))
+		goto err_out_unlock;
+
+
+	/* get the CPU */
+	data = per_cpu(cpufreq_cpu_data, cpu);
+
+	if (!data)
+		goto err_out_put_module;
+
+	if (!kobject_get(&data->kobj))
+		goto err_out_put_module;
+
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+	return data;
+
+err_out_put_module:
+	module_put(cpufreq_driver->owner);
+err_out_unlock:
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+err_out:
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(cpufreq_cpu_get);
+
+
+void cpufreq_cpu_put(struct cpufreq_policy *data)
+{
+	kobject_put(&data->kobj);
+	module_put(cpufreq_driver->owner);
+}
+EXPORT_SYMBOL_GPL(cpufreq_cpu_put);
+
+
+/*********************************************************************
+ *            EXTERNALLY AFFECTING FREQUENCY CHANGES                 *
+ *********************************************************************/
+
+/**
+ * adjust_jiffies - adjust the system "loops_per_jiffy"
+ *
+ * This function alters the system "loops_per_jiffy" for the clock
+ * speed change. Note that loops_per_jiffy cannot be updated on SMP
+ * systems as each CPU might be scaled differently. So, use the arch
+ * per-CPU loops_per_jiffy value wherever possible.
+ */
+#ifndef CONFIG_SMP
+static unsigned long l_p_j_ref;
+static unsigned int  l_p_j_ref_freq;
+
+static void adjust_jiffies(unsigned long val, struct cpufreq_freqs *ci)
+{
+	if (ci->flags & CPUFREQ_CONST_LOOPS)
+		return;
+
+	if (!l_p_j_ref_freq) {
+		l_p_j_ref = loops_per_jiffy;
+		l_p_j_ref_freq = ci->old;
+		pr_debug("saving %lu as reference value for loops_per_jiffy; "
+			"freq is %u kHz\n", l_p_j_ref, l_p_j_ref_freq);
+	}
+	if ((val == CPUFREQ_PRECHANGE  && ci->old < ci->new) ||
+	    (val == CPUFREQ_POSTCHANGE && ci->old > ci->new) ||
+	    (val == CPUFREQ_RESUMECHANGE || val == CPUFREQ_SUSPENDCHANGE)) {
+		loops_per_jiffy = cpufreq_scale(l_p_j_ref, l_p_j_ref_freq,
+								ci->new);
+		pr_debug("scaling loops_per_jiffy to %lu "
+			"for frequency %u kHz\n", loops_per_jiffy, ci->new);
+	}
+}
+#else
+static inline void adjust_jiffies(unsigned long val, struct cpufreq_freqs *ci)
+{
+	return;
+}
+#endif
+
+
+/**
+ * cpufreq_notify_transition - call notifier chain and adjust_jiffies
+ * on frequency transition.
+ *
+ * This function calls the transition notifiers and the "adjust_jiffies"
+ * function. It is called twice on all CPU frequency changes that have
+ * external effects.
+ */
+void cpufreq_notify_transition(struct cpufreq_freqs *freqs, unsigned int state)
+{
+	struct cpufreq_policy *policy;
+
+	BUG_ON(irqs_disabled());
+
+	freqs->flags = cpufreq_driver->flags;
+	pr_debug("notification %u of frequency transition to %u kHz\n",
+		state, freqs->new);
+
+	policy = per_cpu(cpufreq_cpu_data, freqs->cpu);
+	switch (state) {
+
+	case CPUFREQ_PRECHANGE:
+		/* detect if the driver reported a value as "old frequency"
+		 * which is not equal to what the cpufreq core thinks is
+		 * "old frequency".
+		 */
+		if (!(cpufreq_driver->flags & CPUFREQ_CONST_LOOPS)) {
+			if ((policy) && (policy->cpu == freqs->cpu) &&
+			    (policy->cur) && (policy->cur != freqs->old)) {
+				pr_debug("Warning: CPU frequency is"
+					" %u, cpufreq assumed %u kHz.\n",
+					freqs->old, policy->cur);
+				freqs->old = policy->cur;
+			}
+		}
+		srcu_notifier_call_chain(&cpufreq_transition_notifier_list,
+				CPUFREQ_PRECHANGE, freqs);
+		adjust_jiffies(CPUFREQ_PRECHANGE, freqs);
+		break;
+
+	case CPUFREQ_POSTCHANGE:
+		adjust_jiffies(CPUFREQ_POSTCHANGE, freqs);
+		pr_debug("FREQ: %lu - CPU: %lu", (unsigned long)freqs->new,
+			(unsigned long)freqs->cpu);
+		trace_power_frequency(POWER_PSTATE, freqs->new, freqs->cpu);
+		trace_cpu_frequency(freqs->new, freqs->cpu);
+		srcu_notifier_call_chain(&cpufreq_transition_notifier_list,
+				CPUFREQ_POSTCHANGE, freqs);
+		if (likely(policy) && likely(policy->cpu == freqs->cpu))
+			policy->cur = freqs->new;
+		break;
+	}
+}
+EXPORT_SYMBOL_GPL(cpufreq_notify_transition);
+
+
+
+/*********************************************************************
+ *                          SYSFS INTERFACE                          *
+ *********************************************************************/
+
+static struct cpufreq_governor *__find_governor(const char *str_governor)
+{
+	struct cpufreq_governor *t;
+
+	list_for_each_entry(t, &cpufreq_governor_list, governor_list)
+		if (!strnicmp(str_governor, t->name, CPUFREQ_NAME_LEN))
+			return t;
+
+	return NULL;
+}
+
+/**
+ * cpufreq_parse_governor - parse a governor string
+ */
+static int cpufreq_parse_governor(char *str_governor, unsigned int *policy,
+				struct cpufreq_governor **governor)
+{
+	int err = -EINVAL;
+
+	if (!cpufreq_driver)
+		goto out;
+
+	if (cpufreq_driver->setpolicy) {
+		if (!strnicmp(str_governor, "performance", CPUFREQ_NAME_LEN)) {
+			*policy = CPUFREQ_POLICY_PERFORMANCE;
+			err = 0;
+		} else if (!strnicmp(str_governor, "powersave",
+						CPUFREQ_NAME_LEN)) {
+			*policy = CPUFREQ_POLICY_POWERSAVE;
+			err = 0;
+		}
+	} else if (cpufreq_driver->target) {
+		struct cpufreq_governor *t;
+
+		mutex_lock(&cpufreq_governor_mutex);
+
+		t = __find_governor(str_governor);
+
+		if (t == NULL) {
+			int ret;
+
+			mutex_unlock(&cpufreq_governor_mutex);
+			ret = request_module("cpufreq_%s", str_governor);
+			mutex_lock(&cpufreq_governor_mutex);
+
+			if (ret == 0)
+				t = __find_governor(str_governor);
+		}
+
+		if (t != NULL) {
+			*governor = t;
+			err = 0;
+		}
+
+		mutex_unlock(&cpufreq_governor_mutex);
+	}
+out:
+	return err;
+}
+
+
+/**
+ * cpufreq_per_cpu_attr_read() / show_##file_name() -
+ * print out cpufreq information
+ *
+ * Write out information from cpufreq_driver->policy[cpu]; object must be
+ * "unsigned int".
+ */
+
+#define show_one(file_name, object)			\
+static ssize_t show_##file_name				\
+(struct cpufreq_policy *policy, char *buf)		\
+{							\
+	return sprintf(buf, "%u\n", policy->object);	\
+}
+
+show_one(cpuinfo_min_freq, cpuinfo.min_freq);
+show_one(cpuinfo_max_freq, cpuinfo.max_freq);
+show_one(cpuinfo_transition_latency, cpuinfo.transition_latency);
+show_one(scaling_min_freq, min);
+show_one(scaling_max_freq, max);
+show_one(scaling_cur_freq, cur);
+
+static int __cpufreq_set_policy(struct cpufreq_policy *data,
+				struct cpufreq_policy *policy);
+
+/**
+ * cpufreq_per_cpu_attr_write() / store_##file_name() - sysfs write access
+ */
+#define store_one(file_name, object)			\
+static ssize_t store_##file_name					\
+(struct cpufreq_policy *policy, const char *buf, size_t count)		\
+{									\
+	unsigned int ret = -EINVAL;					\
+	struct cpufreq_policy new_policy;				\
+									\
+	ret = cpufreq_get_policy(&new_policy, policy->cpu);		\
+	if (ret)							\
+		return -EINVAL;						\
+									\
+	ret = sscanf(buf, "%u", &new_policy.object);			\
+	if (ret != 1)							\
+		return -EINVAL;						\
+									\
+	ret = __cpufreq_set_policy(policy, &new_policy);		\
+	policy->user_policy.object = policy->object;			\
+									\
+	return ret ? ret : count;					\
+}
+
+store_one(scaling_min_freq, min);
+store_one(scaling_max_freq, max);
+
+/**
+ * show_cpuinfo_cur_freq - current CPU frequency as detected by hardware
+ */
+static ssize_t show_cpuinfo_cur_freq(struct cpufreq_policy *policy,
+					char *buf)
+{
+	unsigned int cur_freq = __cpufreq_get(policy->cpu);
+	if (!cur_freq)
+		return sprintf(buf, "<unknown>");
+	return sprintf(buf, "%u\n", cur_freq);
+}
+
+
+/**
+ * show_scaling_governor - show the current policy for the specified CPU
+ */
+static ssize_t show_scaling_governor(struct cpufreq_policy *policy, char *buf)
+{
+	if (policy->policy == CPUFREQ_POLICY_POWERSAVE)
+		return sprintf(buf, "powersave\n");
+	else if (policy->policy == CPUFREQ_POLICY_PERFORMANCE)
+		return sprintf(buf, "performance\n");
+	else if (policy->governor)
+		return scnprintf(buf, CPUFREQ_NAME_LEN, "%s\n",
+				policy->governor->name);
+	return -EINVAL;
+}
+
+
+/**
+ * store_scaling_governor - store policy for the specified CPU
+ */
+static ssize_t store_scaling_governor(struct cpufreq_policy *policy,
+					const char *buf, size_t count)
+{
+	unsigned int ret = -EINVAL;
+	char	str_governor[16];
+	struct cpufreq_policy new_policy;
+
+	ret = cpufreq_get_policy(&new_policy, policy->cpu);
+	if (ret)
+		return ret;
+
+	ret = sscanf(buf, "%15s", str_governor);
+	if (ret != 1)
+		return -EINVAL;
+
+	if (cpufreq_parse_governor(str_governor, &new_policy.policy,
+						&new_policy.governor))
+		return -EINVAL;
+
+	/* Do not use cpufreq_set_policy here or the user_policy.max
+	   will be wrongly overridden */
+	ret = __cpufreq_set_policy(policy, &new_policy);
+
+	policy->user_policy.policy = policy->policy;
+	policy->user_policy.governor = policy->governor;
+
+	if (ret)
+		return ret;
+	else
+		return count;
+}
+
+/**
+ * show_scaling_driver - show the cpufreq driver currently loaded
+ */
+static ssize_t show_scaling_driver(struct cpufreq_policy *policy, char *buf)
+{
+	return scnprintf(buf, CPUFREQ_NAME_LEN, "%s\n", cpufreq_driver->name);
+}
+
+/**
+ * show_scaling_available_governors - show the available CPUfreq governors
+ */
+static ssize_t show_scaling_available_governors(struct cpufreq_policy *policy,
+						char *buf)
+{
+	ssize_t i = 0;
+	struct cpufreq_governor *t;
+
+	if (!cpufreq_driver->target) {
+		i += sprintf(buf, "performance powersave");
+		goto out;
+	}
+
+	list_for_each_entry(t, &cpufreq_governor_list, governor_list) {
+		if (i >= (ssize_t) ((PAGE_SIZE / sizeof(char))
+		    - (CPUFREQ_NAME_LEN + 2)))
+			goto out;
+		i += scnprintf(&buf[i], CPUFREQ_NAME_LEN, "%s ", t->name);
+	}
+out:
+	i += sprintf(&buf[i], "\n");
+	return i;
+}
+
+static ssize_t show_cpus(const struct cpumask *mask, char *buf)
+{
+	ssize_t i = 0;
+	unsigned int cpu;
+
+	for_each_cpu(cpu, mask) {
+		if (i)
+			i += scnprintf(&buf[i], (PAGE_SIZE - i - 2), " ");
+		i += scnprintf(&buf[i], (PAGE_SIZE - i - 2), "%u", cpu);
+		if (i >= (PAGE_SIZE - 5))
+			break;
+	}
+	i += sprintf(&buf[i], "\n");
+	return i;
+}
+
+/**
+ * show_related_cpus - show the CPUs affected by each transition even if
+ * hw coordination is in use
+ */
+static ssize_t show_related_cpus(struct cpufreq_policy *policy, char *buf)
+{
+	if (cpumask_empty(policy->related_cpus))
+		return show_cpus(policy->cpus, buf);
+	return show_cpus(policy->related_cpus, buf);
+}
+
+/**
+ * show_affected_cpus - show the CPUs affected by each transition
+ */
+static ssize_t show_affected_cpus(struct cpufreq_policy *policy, char *buf)
+{
+	return show_cpus(policy->cpus, buf);
+}
+
+static ssize_t store_scaling_setspeed(struct cpufreq_policy *policy,
+					const char *buf, size_t count)
+{
+	unsigned int freq = 0;
+	unsigned int ret;
+
+	if (!policy->governor || !policy->governor->store_setspeed)
+		return -EINVAL;
+
+	ret = sscanf(buf, "%u", &freq);
+	if (ret != 1)
+		return -EINVAL;
+
+	policy->governor->store_setspeed(policy, freq);
+
+	return count;
+}
+
+static ssize_t show_scaling_setspeed(struct cpufreq_policy *policy, char *buf)
+{
+	if (!policy->governor || !policy->governor->show_setspeed)
+		return sprintf(buf, "<unsupported>\n");
+
+	return policy->governor->show_setspeed(policy, buf);
+}
+
+/**
+ * show_scaling_driver - show the current cpufreq HW/BIOS limitation
+ */
+static ssize_t show_bios_limit(struct cpufreq_policy *policy, char *buf)
+{
+	unsigned int limit;
+	int ret;
+	if (cpufreq_driver->bios_limit) {
+		ret = cpufreq_driver->bios_limit(policy->cpu, &limit);
+		if (!ret)
+			return sprintf(buf, "%u\n", limit);
+	}
+	return sprintf(buf, "%u\n", policy->cpuinfo.max_freq);
+}
+
+cpufreq_freq_attr_ro_perm(cpuinfo_cur_freq, 0400);
+cpufreq_freq_attr_ro(cpuinfo_min_freq);
+cpufreq_freq_attr_ro(cpuinfo_max_freq);
+cpufreq_freq_attr_ro(cpuinfo_transition_latency);
+cpufreq_freq_attr_ro(scaling_available_governors);
+cpufreq_freq_attr_ro(scaling_driver);
+cpufreq_freq_attr_ro(scaling_cur_freq);
+cpufreq_freq_attr_ro(bios_limit);
+cpufreq_freq_attr_ro(related_cpus);
+cpufreq_freq_attr_ro(affected_cpus);
+cpufreq_freq_attr_rw(scaling_min_freq);
+cpufreq_freq_attr_rw(scaling_max_freq);
+cpufreq_freq_attr_rw(scaling_governor);
+cpufreq_freq_attr_rw(scaling_setspeed);
+
+static struct attribute *default_attrs[] = {
+	&cpuinfo_min_freq.attr,
+	&cpuinfo_max_freq.attr,
+	&cpuinfo_transition_latency.attr,
+	&scaling_min_freq.attr,
+	&scaling_max_freq.attr,
+	&affected_cpus.attr,
+	&related_cpus.attr,
+	&scaling_governor.attr,
+	&scaling_driver.attr,
+	&scaling_available_governors.attr,
+	&scaling_setspeed.attr,
+	NULL
+};
+
+struct kobject *cpufreq_global_kobject;
+EXPORT_SYMBOL(cpufreq_global_kobject);
+
+#define to_policy(k) container_of(k, struct cpufreq_policy, kobj)
+#define to_attr(a) container_of(a, struct freq_attr, attr)
+
+static ssize_t show(struct kobject *kobj, struct attribute *attr, char *buf)
+{
+	struct cpufreq_policy *policy = to_policy(kobj);
+	struct freq_attr *fattr = to_attr(attr);
+	ssize_t ret = -EINVAL;
+	policy = cpufreq_cpu_get(policy->cpu);
+	if (!policy)
+		goto no_policy;
+
+	if (lock_policy_rwsem_read(policy->cpu) < 0)
+		goto fail;
+
+	if (fattr->show)
+		ret = fattr->show(policy, buf);
+	else
+		ret = -EIO;
+
+	unlock_policy_rwsem_read(policy->cpu);
+fail:
+	cpufreq_cpu_put(policy);
+no_policy:
+	return ret;
+}
+
+static ssize_t store(struct kobject *kobj, struct attribute *attr,
+		     const char *buf, size_t count)
+{
+	struct cpufreq_policy *policy = to_policy(kobj);
+	struct freq_attr *fattr = to_attr(attr);
+	ssize_t ret = -EINVAL;
+	policy = cpufreq_cpu_get(policy->cpu);
+	if (!policy)
+		goto no_policy;
+
+	if (lock_policy_rwsem_write(policy->cpu) < 0)
+		goto fail;
+
+	if (fattr->store)
+		ret = fattr->store(policy, buf, count);
+	else
+		ret = -EIO;
+
+	unlock_policy_rwsem_write(policy->cpu);
+fail:
+	cpufreq_cpu_put(policy);
+no_policy:
+	return ret;
+}
+
+static void cpufreq_sysfs_release(struct kobject *kobj)
+{
+	struct cpufreq_policy *policy = to_policy(kobj);
+	pr_debug("last reference is dropped\n");
+	complete(&policy->kobj_unregister);
+}
+
+static const struct sysfs_ops sysfs_ops = {
+	.show	= show,
+	.store	= store,
+};
+
+static struct kobj_type ktype_cpufreq = {
+	.sysfs_ops	= &sysfs_ops,
+	.default_attrs	= default_attrs,
+	.release	= cpufreq_sysfs_release,
+};
+
+/*
+ * Returns:
+ *   Negative: Failure
+ *   0:        Success
+ *   Positive: When we have a managed CPU and the sysfs got symlinked
+ */
+static int cpufreq_add_dev_policy(unsigned int cpu,
+				  struct cpufreq_policy *policy,
+				  struct sys_device *sys_dev)
+{
+	int ret = 0;
+#ifdef CONFIG_SMP
+	unsigned long flags;
+	unsigned int j;
+#ifdef CONFIG_HOTPLUG_CPU
+	struct cpufreq_governor *gov;
+
+	gov = __find_governor(per_cpu(cpufreq_cpu_governor, cpu));
+	if (gov) {
+		policy->governor = gov;
+		pr_debug("Restoring governor %s for cpu %d\n",
+		       policy->governor->name, cpu);
+	}
+#endif
+
+	for_each_cpu(j, policy->cpus) {
+		struct cpufreq_policy *managed_policy;
+
+		if (cpu == j)
+			continue;
+
+		/* Check for existing affected CPUs.
+		 * They may not be aware of it due to CPU Hotplug.
+		 * cpufreq_cpu_put is called when the device is removed
+		 * in __cpufreq_remove_dev()
+		 */
+		managed_policy = cpufreq_cpu_get(j);
+		if (unlikely(managed_policy)) {
+
+			/* Set proper policy_cpu */
+			unlock_policy_rwsem_write(cpu);
+			per_cpu(cpufreq_policy_cpu, cpu) = managed_policy->cpu;
+
+			if (lock_policy_rwsem_write(cpu) < 0) {
+				/* Should not go through policy unlock path */
+				if (cpufreq_driver->exit)
+					cpufreq_driver->exit(policy);
+				cpufreq_cpu_put(managed_policy);
+				return -EBUSY;
+			}
+
+			spin_lock_irqsave(&cpufreq_driver_lock, flags);
+			cpumask_copy(managed_policy->cpus, policy->cpus);
+			per_cpu(cpufreq_cpu_data, cpu) = managed_policy;
+			spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+
+			pr_debug("CPU already managed, adding link\n");
+			ret = sysfs_create_link(&sys_dev->kobj,
+						&managed_policy->kobj,
+						"cpufreq");
+			if (ret)
+				cpufreq_cpu_put(managed_policy);
+			/*
+			 * Success. We only needed to be added to the mask.
+			 * Call driver->exit() because only the cpu parent of
+			 * the kobj needed to call init().
+			 */
+			if (cpufreq_driver->exit)
+				cpufreq_driver->exit(policy);
+
+			if (!ret)
+				return 1;
+			else
+				return ret;
+		}
+	}
+#endif
+	return ret;
+}
+
+
+/* symlink affected CPUs */
+static int cpufreq_add_dev_symlink(unsigned int cpu,
+				   struct cpufreq_policy *policy)
+{
+	unsigned int j;
+	int ret = 0;
+
+	for_each_cpu(j, policy->cpus) {
+		struct cpufreq_policy *managed_policy;
+		struct sys_device *cpu_sys_dev;
+
+		if (j == cpu)
+			continue;
+		if (!cpu_online(j))
+			continue;
+
+		pr_debug("CPU %u already managed, adding link\n", j);
+		managed_policy = cpufreq_cpu_get(cpu);
+		cpu_sys_dev = get_cpu_sysdev(j);
+		ret = sysfs_create_link(&cpu_sys_dev->kobj, &policy->kobj,
+					"cpufreq");
+		if (ret) {
+			cpufreq_cpu_put(managed_policy);
+			return ret;
+		}
+	}
+	return ret;
+}
+
+static int cpufreq_add_dev_interface(unsigned int cpu,
+				     struct cpufreq_policy *policy,
+				     struct sys_device *sys_dev)
+{
+	struct cpufreq_policy new_policy;
+	struct freq_attr **drv_attr;
+	unsigned long flags;
+	int ret = 0;
+	unsigned int j;
+
+	/* prepare interface data */
+	ret = kobject_init_and_add(&policy->kobj, &ktype_cpufreq,
+				   &sys_dev->kobj, "cpufreq");
+	if (ret)
+		return ret;
+
+	/* set up files for this cpu device */
+	drv_attr = cpufreq_driver->attr;
+	while ((drv_attr) && (*drv_attr)) {
+		ret = sysfs_create_file(&policy->kobj, &((*drv_attr)->attr));
+		if (ret)
+			goto err_out_kobj_put;
+		drv_attr++;
+	}
+	if (cpufreq_driver->get) {
+		ret = sysfs_create_file(&policy->kobj, &cpuinfo_cur_freq.attr);
+		if (ret)
+			goto err_out_kobj_put;
+	}
+	if (cpufreq_driver->target) {
+		ret = sysfs_create_file(&policy->kobj, &scaling_cur_freq.attr);
+		if (ret)
+			goto err_out_kobj_put;
+	}
+	if (cpufreq_driver->bios_limit) {
+		ret = sysfs_create_file(&policy->kobj, &bios_limit.attr);
+		if (ret)
+			goto err_out_kobj_put;
+	}
+
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+	for_each_cpu(j, policy->cpus) {
+		if (!cpu_online(j))
+			continue;
+		per_cpu(cpufreq_cpu_data, j) = policy;
+		per_cpu(cpufreq_policy_cpu, j) = policy->cpu;
+	}
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+
+	ret = cpufreq_add_dev_symlink(cpu, policy);
+	if (ret)
+		goto err_out_kobj_put;
+
+	memcpy(&new_policy, policy, sizeof(struct cpufreq_policy));
+	/* assure that the starting sequence is run in __cpufreq_set_policy */
+	policy->governor = NULL;
+
+	/* set default policy */
+	ret = __cpufreq_set_policy(policy, &new_policy);
+	policy->user_policy.policy = policy->policy;
+	policy->user_policy.governor = policy->governor;
+
+	if (ret) {
+		pr_debug("setting policy failed\n");
+		if (cpufreq_driver->exit)
+			cpufreq_driver->exit(policy);
+	}
+	return ret;
+
+err_out_kobj_put:
+	kobject_put(&policy->kobj);
+	wait_for_completion(&policy->kobj_unregister);
+	return ret;
+}
+
+
+/**
+ * cpufreq_add_dev - add a CPU device
+ *
+ * Adds the cpufreq interface for a CPU device.
+ *
+ * The Oracle says: try running cpufreq registration/unregistration concurrently
+ * with with cpu hotplugging and all hell will break loose. Tried to clean this
+ * mess up, but more thorough testing is needed. - Mathieu
+ */
+static int cpufreq_add_dev(struct sys_device *sys_dev)
+{
+	unsigned int cpu = sys_dev->id;
+	int ret = 0, found = 0;
+	struct cpufreq_policy *policy;
+	unsigned long flags;
+	unsigned int j;
+#ifdef CONFIG_HOTPLUG_CPU
+	int sibling;
+#endif
+
+	if (cpu_is_offline(cpu))
+		return 0;
+
+	pr_debug("adding CPU %u\n", cpu);
+
+#ifdef CONFIG_SMP
+	/* check whether a different CPU already registered this
+	 * CPU because it is in the same boat. */
+	policy = cpufreq_cpu_get(cpu);
+	if (unlikely(policy)) {
+		cpufreq_cpu_put(policy);
+		return 0;
+	}
+#endif
+
+	if (!try_module_get(cpufreq_driver->owner)) {
+		ret = -EINVAL;
+		goto module_out;
+	}
+
+	ret = -ENOMEM;
+	policy = kzalloc(sizeof(struct cpufreq_policy), GFP_KERNEL);
+	if (!policy)
+		goto nomem_out;
+
+	if (!alloc_cpumask_var(&policy->cpus, GFP_KERNEL))
+		goto err_free_policy;
+
+	if (!zalloc_cpumask_var(&policy->related_cpus, GFP_KERNEL))
+		goto err_free_cpumask;
+
+	policy->cpu = cpu;
+	cpumask_copy(policy->cpus, cpumask_of(cpu));
+
+	/* Initially set CPU itself as the policy_cpu */
+	per_cpu(cpufreq_policy_cpu, cpu) = cpu;
+	ret = (lock_policy_rwsem_write(cpu) < 0);
+	WARN_ON(ret);
+
+	init_completion(&policy->kobj_unregister);
+	INIT_WORK(&policy->update, handle_update);
+
+	/* Set governor before ->init, so that driver could check it */
+#ifdef CONFIG_HOTPLUG_CPU
+	for_each_online_cpu(sibling) {
+		struct cpufreq_policy *cp = per_cpu(cpufreq_cpu_data, sibling);
+		if (cp && cp->governor &&
+		    (cpumask_test_cpu(cpu, cp->related_cpus))) {
+			policy->governor = cp->governor;
+			found = 1;
+			break;
+		}
+	}
+#endif
+	if (!found)
+		policy->governor = CPUFREQ_DEFAULT_GOVERNOR;
+	/* call driver. From then on the cpufreq must be able
+	 * to accept all calls to ->verify and ->setpolicy for this CPU
+	 */
+	ret = cpufreq_driver->init(policy);
+	if (ret) {
+		pr_debug("initialization failed\n");
+		goto err_unlock_policy;
+	}
+	policy->user_policy.min = policy->min;
+	policy->user_policy.max = policy->max;
+
+	blocking_notifier_call_chain(&cpufreq_policy_notifier_list,
+				     CPUFREQ_START, policy);
+
+	ret = cpufreq_add_dev_policy(cpu, policy, sys_dev);
+	if (ret) {
+		if (ret > 0)
+			/* This is a managed cpu, symlink created,
+			   exit with 0 */
+			ret = 0;
+		goto err_unlock_policy;
+	}
+
+	ret = cpufreq_add_dev_interface(cpu, policy, sys_dev);
+	if (ret)
+		goto err_out_unregister;
+
+	unlock_policy_rwsem_write(cpu);
+
+	kobject_uevent(&policy->kobj, KOBJ_ADD);
+	module_put(cpufreq_driver->owner);
+	pr_debug("initialization complete\n");
+
+	return 0;
+
+
+err_out_unregister:
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+	for_each_cpu(j, policy->cpus)
+		per_cpu(cpufreq_cpu_data, j) = NULL;
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+
+	kobject_put(&policy->kobj);
+	wait_for_completion(&policy->kobj_unregister);
+
+err_unlock_policy:
+	unlock_policy_rwsem_write(cpu);
+	free_cpumask_var(policy->related_cpus);
+err_free_cpumask:
+	free_cpumask_var(policy->cpus);
+err_free_policy:
+	kfree(policy);
+nomem_out:
+	module_put(cpufreq_driver->owner);
+module_out:
+	return ret;
+}
+
+
+/**
+ * __cpufreq_remove_dev - remove a CPU device
+ *
+ * Removes the cpufreq interface for a CPU device.
+ * Caller should already have policy_rwsem in write mode for this CPU.
+ * This routine frees the rwsem before returning.
+ */
+static int __cpufreq_remove_dev(struct sys_device *sys_dev)
+{
+	unsigned int cpu = sys_dev->id;
+	unsigned long flags;
+	struct cpufreq_policy *data;
+	struct kobject *kobj;
+	struct completion *cmp;
+#ifdef CONFIG_SMP
+	struct sys_device *cpu_sys_dev;
+	unsigned int j;
+#endif
+
+	pr_debug("unregistering CPU %u\n", cpu);
+
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+	data = per_cpu(cpufreq_cpu_data, cpu);
+
+	if (!data) {
+		spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+		unlock_policy_rwsem_write(cpu);
+		return -EINVAL;
+	}
+	per_cpu(cpufreq_cpu_data, cpu) = NULL;
+
+
+#ifdef CONFIG_SMP
+	/* if this isn't the CPU which is the parent of the kobj, we
+	 * only need to unlink, put and exit
+	 */
+	if (unlikely(cpu != data->cpu)) {
+		pr_debug("removing link\n");
+		cpumask_clear_cpu(cpu, data->cpus);
+		spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+		kobj = &sys_dev->kobj;
+		cpufreq_cpu_put(data);
+		unlock_policy_rwsem_write(cpu);
+		sysfs_remove_link(kobj, "cpufreq");
+		return 0;
+	}
+#endif
+
+#ifdef CONFIG_SMP
+
+#ifdef CONFIG_HOTPLUG_CPU
+	strncpy(per_cpu(cpufreq_cpu_governor, cpu), data->governor->name,
+			CPUFREQ_NAME_LEN);
+#endif
+
+	/* if we have other CPUs still registered, we need to unlink them,
+	 * or else wait_for_completion below will lock up. Clean the
+	 * per_cpu(cpufreq_cpu_data) while holding the lock, and remove
+	 * the sysfs links afterwards.
+	 */
+	if (unlikely(cpumask_weight(data->cpus) > 1)) {
+		for_each_cpu(j, data->cpus) {
+			if (j == cpu)
+				continue;
+			per_cpu(cpufreq_cpu_data, j) = NULL;
+		}
+	}
+
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+
+	if (unlikely(cpumask_weight(data->cpus) > 1)) {
+		for_each_cpu(j, data->cpus) {
+			if (j == cpu)
+				continue;
+			pr_debug("removing link for cpu %u\n", j);
+#ifdef CONFIG_HOTPLUG_CPU
+			strncpy(per_cpu(cpufreq_cpu_governor, j),
+				data->governor->name, CPUFREQ_NAME_LEN);
+#endif
+			cpu_sys_dev = get_cpu_sysdev(j);
+			kobj = &cpu_sys_dev->kobj;
+			unlock_policy_rwsem_write(cpu);
+			sysfs_remove_link(kobj, "cpufreq");
+			lock_policy_rwsem_write(cpu);
+			cpufreq_cpu_put(data);
+		}
+	}
+#else
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+#endif
+
+	if (cpufreq_driver->target)
+		__cpufreq_governor(data, CPUFREQ_GOV_STOP);
+
+	kobj = &data->kobj;
+	cmp = &data->kobj_unregister;
+	unlock_policy_rwsem_write(cpu);
+	kobject_put(kobj);
+
+	/* we need to make sure that the underlying kobj is actually
+	 * not referenced anymore by anybody before we proceed with
+	 * unloading.
+	 */
+	pr_debug("waiting for dropping of refcount\n");
+	wait_for_completion(cmp);
+	pr_debug("wait complete\n");
+
+	lock_policy_rwsem_write(cpu);
+	if (cpufreq_driver->exit)
+		cpufreq_driver->exit(data);
+	unlock_policy_rwsem_write(cpu);
+
+#ifdef CONFIG_HOTPLUG_CPU
+	/* when the CPU which is the parent of the kobj is hotplugged
+	 * offline, check for siblings, and create cpufreq sysfs interface
+	 * and symlinks
+	 */
+	if (unlikely(cpumask_weight(data->cpus) > 1)) {
+		/* first sibling now owns the new sysfs dir */
+		cpumask_clear_cpu(cpu, data->cpus);
+		cpufreq_add_dev(get_cpu_sysdev(cpumask_first(data->cpus)));
+
+		/* finally remove our own symlink */
+		lock_policy_rwsem_write(cpu);
+		__cpufreq_remove_dev(sys_dev);
+	}
+#endif
+
+	free_cpumask_var(data->related_cpus);
+	free_cpumask_var(data->cpus);
+	kfree(data);
+
+	return 0;
+}
+
+
+static int cpufreq_remove_dev(struct sys_device *sys_dev)
+{
+	unsigned int cpu = sys_dev->id;
+	int retval;
+
+	if (cpu_is_offline(cpu))
+		return 0;
+
+	if (unlikely(lock_policy_rwsem_write(cpu)))
+		BUG();
+
+	retval = __cpufreq_remove_dev(sys_dev);
+	return retval;
+}
+
+
+static void handle_update(struct work_struct *work)
+{
+	struct cpufreq_policy *policy =
+		container_of(work, struct cpufreq_policy, update);
+	unsigned int cpu = policy->cpu;
+	pr_debug("handle_update for cpu %u called\n", cpu);
+	cpufreq_update_policy(cpu);
+}
+
+/**
+ *	cpufreq_out_of_sync - If actual and saved CPU frequency differs, we're in deep trouble.
+ *	@cpu: cpu number
+ *	@old_freq: CPU frequency the kernel thinks the CPU runs at
+ *	@new_freq: CPU frequency the CPU actually runs at
+ *
+ *	We adjust to current frequency first, and need to clean up later.
+ *	So either call to cpufreq_update_policy() or schedule handle_update()).
+ */
+static void cpufreq_out_of_sync(unsigned int cpu, unsigned int old_freq,
+				unsigned int new_freq)
+{
+	struct cpufreq_freqs freqs;
+
+	pr_debug("Warning: CPU frequency out of sync: cpufreq and timing "
+	       "core thinks of %u, is %u kHz.\n", old_freq, new_freq);
+
+	freqs.cpu = cpu;
+	freqs.old = old_freq;
+	freqs.new = new_freq;
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+}
+
+
+/**
+ * cpufreq_quick_get - get the CPU frequency (in kHz) from policy->cur
+ * @cpu: CPU number
+ *
+ * This is the last known freq, without actually getting it from the driver.
+ * Return value will be same as what is shown in scaling_cur_freq in sysfs.
+ */
+unsigned int cpufreq_quick_get(unsigned int cpu)
+{
+	struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
+	unsigned int ret_freq = 0;
+
+	if (policy) {
+		ret_freq = policy->cur;
+		cpufreq_cpu_put(policy);
+	}
+
+	return ret_freq;
+}
+EXPORT_SYMBOL(cpufreq_quick_get);
+
+
+static unsigned int __cpufreq_get(unsigned int cpu)
+{
+	struct cpufreq_policy *policy = per_cpu(cpufreq_cpu_data, cpu);
+	unsigned int ret_freq = 0;
+
+	if (!cpufreq_driver->get)
+		return ret_freq;
+
+	ret_freq = cpufreq_driver->get(cpu);
+
+	if (ret_freq && policy->cur &&
+		!(cpufreq_driver->flags & CPUFREQ_CONST_LOOPS)) {
+		/* verify no discrepancy between actual and
+					saved value exists */
+		if (unlikely(ret_freq != policy->cur)) {
+			cpufreq_out_of_sync(cpu, policy->cur, ret_freq);
+			schedule_work(&policy->update);
+		}
+	}
+
+	return ret_freq;
+}
+
+/**
+ * cpufreq_get - get the current CPU frequency (in kHz)
+ * @cpu: CPU number
+ *
+ * Get the CPU current (static) CPU frequency
+ */
+unsigned int cpufreq_get(unsigned int cpu)
+{
+	unsigned int ret_freq = 0;
+	struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
+
+	if (!policy)
+		goto out;
+
+	if (unlikely(lock_policy_rwsem_read(cpu)))
+		goto out_policy;
+
+	ret_freq = __cpufreq_get(cpu);
+
+	unlock_policy_rwsem_read(cpu);
+
+out_policy:
+	cpufreq_cpu_put(policy);
+out:
+	return ret_freq;
+}
+EXPORT_SYMBOL(cpufreq_get);
+
+static struct sysdev_driver cpufreq_sysdev_driver = {
+	.add		= cpufreq_add_dev,
+	.remove		= cpufreq_remove_dev,
+};
+
+
+/**
+ * cpufreq_bp_suspend - Prepare the boot CPU for system suspend.
+ *
+ * This function is only executed for the boot processor.  The other CPUs
+ * have been put offline by means of CPU hotplug.
+ */
+static int cpufreq_bp_suspend(void)
+{
+	int ret = 0;
+
+	int cpu = smp_processor_id();
+	struct cpufreq_policy *cpu_policy;
+
+	pr_debug("suspending cpu %u\n", cpu);
+
+	/* If there's no policy for the boot CPU, we have nothing to do. */
+	cpu_policy = cpufreq_cpu_get(cpu);
+	if (!cpu_policy)
+		return 0;
+
+	if (cpufreq_driver->suspend) {
+		ret = cpufreq_driver->suspend(cpu_policy);
+		if (ret)
+			printk(KERN_ERR "cpufreq: suspend failed in ->suspend "
+					"step on CPU %u\n", cpu_policy->cpu);
+	}
+
+	cpufreq_cpu_put(cpu_policy);
+	return ret;
+}
+
+/**
+ * cpufreq_bp_resume - Restore proper frequency handling of the boot CPU.
+ *
+ *	1.) resume CPUfreq hardware support (cpufreq_driver->resume())
+ *	2.) schedule call cpufreq_update_policy() ASAP as interrupts are
+ *	    restored. It will verify that the current freq is in sync with
+ *	    what we believe it to be. This is a bit later than when it
+ *	    should be, but nonethteless it's better than calling
+ *	    cpufreq_driver->get() here which might re-enable interrupts...
+ *
+ * This function is only executed for the boot CPU.  The other CPUs have not
+ * been turned on yet.
+ */
+static void cpufreq_bp_resume(void)
+{
+	int ret = 0;
+
+	int cpu = smp_processor_id();
+	struct cpufreq_policy *cpu_policy;
+
+	pr_debug("resuming cpu %u\n", cpu);
+
+	/* If there's no policy for the boot CPU, we have nothing to do. */
+	cpu_policy = cpufreq_cpu_get(cpu);
+	if (!cpu_policy)
+		return;
+
+	if (cpufreq_driver->resume) {
+		ret = cpufreq_driver->resume(cpu_policy);
+		if (ret) {
+			printk(KERN_ERR "cpufreq: resume failed in ->resume "
+					"step on CPU %u\n", cpu_policy->cpu);
+			goto fail;
+		}
+	}
+
+	schedule_work(&cpu_policy->update);
+
+fail:
+	cpufreq_cpu_put(cpu_policy);
+}
+
+static struct syscore_ops cpufreq_syscore_ops = {
+	.suspend	= cpufreq_bp_suspend,
+	.resume		= cpufreq_bp_resume,
+};
+
+
+/*********************************************************************
+ *                     NOTIFIER LISTS INTERFACE                      *
+ *********************************************************************/
+
+/**
+ *	cpufreq_register_notifier - register a driver with cpufreq
+ *	@nb: notifier function to register
+ *      @list: CPUFREQ_TRANSITION_NOTIFIER or CPUFREQ_POLICY_NOTIFIER
+ *
+ *	Add a driver to one of two lists: either a list of drivers that
+ *      are notified about clock rate changes (once before and once after
+ *      the transition), or a list of drivers that are notified about
+ *      changes in cpufreq policy.
+ *
+ *	This function may sleep, and has the same return conditions as
+ *	blocking_notifier_chain_register.
+ */
+int cpufreq_register_notifier(struct notifier_block *nb, unsigned int list)
+{
+	int ret;
+
+	WARN_ON(!init_cpufreq_transition_notifier_list_called);
+
+	switch (list) {
+	case CPUFREQ_TRANSITION_NOTIFIER:
+		ret = srcu_notifier_chain_register(
+				&cpufreq_transition_notifier_list, nb);
+		break;
+	case CPUFREQ_POLICY_NOTIFIER:
+		ret = blocking_notifier_chain_register(
+				&cpufreq_policy_notifier_list, nb);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(cpufreq_register_notifier);
+
+
+/**
+ *	cpufreq_unregister_notifier - unregister a driver with cpufreq
+ *	@nb: notifier block to be unregistered
+ *      @list: CPUFREQ_TRANSITION_NOTIFIER or CPUFREQ_POLICY_NOTIFIER
+ *
+ *	Remove a driver from the CPU frequency notifier list.
+ *
+ *	This function may sleep, and has the same return conditions as
+ *	blocking_notifier_chain_unregister.
+ */
+int cpufreq_unregister_notifier(struct notifier_block *nb, unsigned int list)
+{
+	int ret;
+
+	switch (list) {
+	case CPUFREQ_TRANSITION_NOTIFIER:
+		ret = srcu_notifier_chain_unregister(
+				&cpufreq_transition_notifier_list, nb);
+		break;
+	case CPUFREQ_POLICY_NOTIFIER:
+		ret = blocking_notifier_chain_unregister(
+				&cpufreq_policy_notifier_list, nb);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(cpufreq_unregister_notifier);
+
+
+/*********************************************************************
+ *                              GOVERNORS                            *
+ *********************************************************************/
+
+
+int __cpufreq_driver_target(struct cpufreq_policy *policy,
+			    unsigned int target_freq,
+			    unsigned int relation)
+{
+	int retval = -EINVAL;
+
+	pr_debug("target for CPU %u: %u kHz, relation %u\n", policy->cpu,
+		target_freq, relation);
+	if (cpu_online(policy->cpu) && cpufreq_driver->target)
+		retval = cpufreq_driver->target(policy, target_freq, relation);
+
+	return retval;
+}
+EXPORT_SYMBOL_GPL(__cpufreq_driver_target);
+
+int cpufreq_driver_target(struct cpufreq_policy *policy,
+			  unsigned int target_freq,
+			  unsigned int relation)
+{
+	int ret = -EINVAL;
+
+	policy = cpufreq_cpu_get(policy->cpu);
+	if (!policy)
+		goto no_policy;
+
+	if (unlikely(lock_policy_rwsem_write(policy->cpu)))
+		goto fail;
+
+	ret = __cpufreq_driver_target(policy, target_freq, relation);
+
+	unlock_policy_rwsem_write(policy->cpu);
+
+fail:
+	cpufreq_cpu_put(policy);
+no_policy:
+	return ret;
+}
+EXPORT_SYMBOL_GPL(cpufreq_driver_target);
+
+int __cpufreq_driver_getavg(struct cpufreq_policy *policy, unsigned int cpu)
+{
+	int ret = 0;
+
+	policy = cpufreq_cpu_get(policy->cpu);
+	if (!policy)
+		return -EINVAL;
+
+	if (cpu_online(cpu) && cpufreq_driver->getavg)
+		ret = cpufreq_driver->getavg(policy, cpu);
+
+	cpufreq_cpu_put(policy);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(__cpufreq_driver_getavg);
+
+/*
+ * when "event" is CPUFREQ_GOV_LIMITS
+ */
+
+static int __cpufreq_governor(struct cpufreq_policy *policy,
+					unsigned int event)
+{
+	int ret;
+
+	/* Only must be defined when default governor is known to have latency
+	   restrictions, like e.g. conservative or ondemand.
+	   That this is the case is already ensured in Kconfig
+	*/
+#ifdef CONFIG_CPU_FREQ_GOV_PERFORMANCE
+	struct cpufreq_governor *gov = &cpufreq_gov_performance;
+#else
+	struct cpufreq_governor *gov = NULL;
+#endif
+
+	if (policy->governor->max_transition_latency &&
+	    policy->cpuinfo.transition_latency >
+	    policy->governor->max_transition_latency) {
+		if (!gov)
+			return -EINVAL;
+		else {
+			printk(KERN_WARNING "%s governor failed, too long"
+			       " transition latency of HW, fallback"
+			       " to %s governor\n",
+			       policy->governor->name,
+			       gov->name);
+			policy->governor = gov;
+		}
+	}
+
+	if (!try_module_get(policy->governor->owner))
+		return -EINVAL;
+
+	pr_debug("__cpufreq_governor for CPU %u, event %u\n",
+						policy->cpu, event);
+	ret = policy->governor->governor(policy, event);
+
+	/* we keep one module reference alive for
+			each CPU governed by this CPU */
+	if ((event != CPUFREQ_GOV_START) || ret)
+		module_put(policy->governor->owner);
+	if ((event == CPUFREQ_GOV_STOP) && !ret)
+		module_put(policy->governor->owner);
+
+	return ret;
+}
+
+
+int cpufreq_register_governor(struct cpufreq_governor *governor)
+{
+	int err;
+
+	if (!governor)
+		return -EINVAL;
+
+	mutex_lock(&cpufreq_governor_mutex);
+
+	err = -EBUSY;
+	if (__find_governor(governor->name) == NULL) {
+		err = 0;
+		list_add(&governor->governor_list, &cpufreq_governor_list);
+	}
+
+	mutex_unlock(&cpufreq_governor_mutex);
+	return err;
+}
+EXPORT_SYMBOL_GPL(cpufreq_register_governor);
+
+
+void cpufreq_unregister_governor(struct cpufreq_governor *governor)
+{
+#ifdef CONFIG_HOTPLUG_CPU
+	int cpu;
+#endif
+
+	if (!governor)
+		return;
+
+#ifdef CONFIG_HOTPLUG_CPU
+	for_each_present_cpu(cpu) {
+		if (cpu_online(cpu))
+			continue;
+		if (!strcmp(per_cpu(cpufreq_cpu_governor, cpu), governor->name))
+			strcpy(per_cpu(cpufreq_cpu_governor, cpu), "\0");
+	}
+#endif
+
+	mutex_lock(&cpufreq_governor_mutex);
+	list_del(&governor->governor_list);
+	mutex_unlock(&cpufreq_governor_mutex);
+	return;
+}
+EXPORT_SYMBOL_GPL(cpufreq_unregister_governor);
+
+
+
+/*********************************************************************
+ *                          POLICY INTERFACE                         *
+ *********************************************************************/
+
+/**
+ * cpufreq_get_policy - get the current cpufreq_policy
+ * @policy: struct cpufreq_policy into which the current cpufreq_policy
+ *	is written
+ *
+ * Reads the current cpufreq policy.
+ */
+int cpufreq_get_policy(struct cpufreq_policy *policy, unsigned int cpu)
+{
+	struct cpufreq_policy *cpu_policy;
+	if (!policy)
+		return -EINVAL;
+
+	cpu_policy = cpufreq_cpu_get(cpu);
+	if (!cpu_policy)
+		return -EINVAL;
+
+	memcpy(policy, cpu_policy, sizeof(struct cpufreq_policy));
+
+	cpufreq_cpu_put(cpu_policy);
+	return 0;
+}
+EXPORT_SYMBOL(cpufreq_get_policy);
+
+
+/*
+ * data   : current policy.
+ * policy : policy to be set.
+ */
+static int __cpufreq_set_policy(struct cpufreq_policy *data,
+				struct cpufreq_policy *policy)
+{
+	int ret = 0;
+
+	pr_debug("setting new policy for CPU %u: %u - %u kHz\n", policy->cpu,
+		policy->min, policy->max);
+
+	memcpy(&policy->cpuinfo, &data->cpuinfo,
+				sizeof(struct cpufreq_cpuinfo));
+
+	if (policy->min > data->max || policy->max < data->min) {
+		ret = -EINVAL;
+		goto error_out;
+	}
+
+	/* verify the cpu speed can be set within this limit */
+	ret = cpufreq_driver->verify(policy);
+	if (ret)
+		goto error_out;
+
+	/* adjust if necessary - all reasons */
+	blocking_notifier_call_chain(&cpufreq_policy_notifier_list,
+			CPUFREQ_ADJUST, policy);
+
+	/* adjust if necessary - hardware incompatibility*/
+	blocking_notifier_call_chain(&cpufreq_policy_notifier_list,
+			CPUFREQ_INCOMPATIBLE, policy);
+
+	/* verify the cpu speed can be set within this limit,
+	   which might be different to the first one */
+	ret = cpufreq_driver->verify(policy);
+	if (ret)
+		goto error_out;
+
+	/* notification of the new policy */
+	blocking_notifier_call_chain(&cpufreq_policy_notifier_list,
+			CPUFREQ_NOTIFY, policy);
+
+	data->min = policy->min;
+	data->max = policy->max;
+
+	pr_debug("new min and max freqs are %u - %u kHz\n",
+					data->min, data->max);
+
+	if (cpufreq_driver->setpolicy) {
+		data->policy = policy->policy;
+		pr_debug("setting range\n");
+		ret = cpufreq_driver->setpolicy(policy);
+	} else {
+		if (policy->governor != data->governor) {
+			/* save old, working values */
+			struct cpufreq_governor *old_gov = data->governor;
+
+			pr_debug("governor switch\n");
+
+			/* end old governor */
+			if (data->governor)
+				__cpufreq_governor(data, CPUFREQ_GOV_STOP);
+
+			/* start new governor */
+			data->governor = policy->governor;
+			if (__cpufreq_governor(data, CPUFREQ_GOV_START)) {
+				/* new governor failed, so re-start old one */
+				pr_debug("starting governor %s failed\n",
+							data->governor->name);
+				if (old_gov) {
+					data->governor = old_gov;
+					__cpufreq_governor(data,
+							   CPUFREQ_GOV_START);
+				}
+				ret = -EINVAL;
+				goto error_out;
+			}
+			/* might be a policy change, too, so fall through */
+		}
+		pr_debug("governor: change or update limits\n");
+		__cpufreq_governor(data, CPUFREQ_GOV_LIMITS);
+	}
+
+error_out:
+	return ret;
+}
+
+/**
+ *	cpufreq_update_policy - re-evaluate an existing cpufreq policy
+ *	@cpu: CPU which shall be re-evaluated
+ *
+ *	Useful for policy notifiers which have different necessities
+ *	at different times.
+ */
+int cpufreq_update_policy(unsigned int cpu)
+{
+	struct cpufreq_policy *data = cpufreq_cpu_get(cpu);
+	struct cpufreq_policy policy;
+	int ret;
+
+	if (!data) {
+		ret = -ENODEV;
+		goto no_policy;
+	}
+
+	if (unlikely(lock_policy_rwsem_write(cpu))) {
+		ret = -EINVAL;
+		goto fail;
+	}
+
+	pr_debug("updating policy for CPU %u\n", cpu);
+	memcpy(&policy, data, sizeof(struct cpufreq_policy));
+	policy.min = data->user_policy.min;
+	policy.max = data->user_policy.max;
+	policy.policy = data->user_policy.policy;
+	policy.governor = data->user_policy.governor;
+
+	/* BIOS might change freq behind our back
+	  -> ask driver for current freq and notify governors about a change */
+	if (cpufreq_driver->get) {
+		policy.cur = cpufreq_driver->get(cpu);
+		if (!data->cur) {
+			pr_debug("Driver did not initialize current freq");
+			data->cur = policy.cur;
+		} else {
+			if (data->cur != policy.cur)
+				cpufreq_out_of_sync(cpu, data->cur,
+								policy.cur);
+		}
+	}
+
+	ret = __cpufreq_set_policy(data, &policy);
+
+	unlock_policy_rwsem_write(cpu);
+
+fail:
+	cpufreq_cpu_put(data);
+no_policy:
+	return ret;
+}
+EXPORT_SYMBOL(cpufreq_update_policy);
+
+static int __cpuinit cpufreq_cpu_callback(struct notifier_block *nfb,
+					unsigned long action, void *hcpu)
+{
+	unsigned int cpu = (unsigned long)hcpu;
+	struct sys_device *sys_dev;
+
+	sys_dev = get_cpu_sysdev(cpu);
+	if (sys_dev) {
+		switch (action) {
+		case CPU_ONLINE:
+		case CPU_ONLINE_FROZEN:
+			cpufreq_add_dev(sys_dev);
+			break;
+		case CPU_DOWN_PREPARE:
+		case CPU_DOWN_PREPARE_FROZEN:
+			if (unlikely(lock_policy_rwsem_write(cpu)))
+				BUG();
+
+			__cpufreq_remove_dev(sys_dev);
+			break;
+		case CPU_DOWN_FAILED:
+		case CPU_DOWN_FAILED_FROZEN:
+			cpufreq_add_dev(sys_dev);
+			break;
+		}
+	}
+	return NOTIFY_OK;
+}
+
+static struct notifier_block __refdata cpufreq_cpu_notifier = {
+    .notifier_call = cpufreq_cpu_callback,
+};
+
+/*********************************************************************
+ *               REGISTER / UNREGISTER CPUFREQ DRIVER                *
+ *********************************************************************/
+
+/**
+ * cpufreq_register_driver - register a CPU Frequency driver
+ * @driver_data: A struct cpufreq_driver containing the values#
+ * submitted by the CPU Frequency driver.
+ *
+ *   Registers a CPU Frequency driver to this core code. This code
+ * returns zero on success, -EBUSY when another driver got here first
+ * (and isn't unregistered in the meantime).
+ *
+ */
+int cpufreq_register_driver(struct cpufreq_driver *driver_data)
+{
+	unsigned long flags;
+	int ret;
+
+	if (!driver_data || !driver_data->verify || !driver_data->init ||
+	    ((!driver_data->setpolicy) && (!driver_data->target)))
+		return -EINVAL;
+
+	pr_debug("trying to register driver %s\n", driver_data->name);
+
+	if (driver_data->setpolicy)
+		driver_data->flags |= CPUFREQ_CONST_LOOPS;
+
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+	if (cpufreq_driver) {
+		spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+		return -EBUSY;
+	}
+	cpufreq_driver = driver_data;
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+
+	ret = sysdev_driver_register(&cpu_sysdev_class,
+					&cpufreq_sysdev_driver);
+	if (ret)
+		goto err_null_driver;
+
+	if (!(cpufreq_driver->flags & CPUFREQ_STICKY)) {
+		int i;
+		ret = -ENODEV;
+
+		/* check for at least one working CPU */
+		for (i = 0; i < nr_cpu_ids; i++)
+			if (cpu_possible(i) && per_cpu(cpufreq_cpu_data, i)) {
+				ret = 0;
+				break;
+			}
+
+		/* if all ->init() calls failed, unregister */
+		if (ret) {
+			pr_debug("no CPU initialized for driver %s\n",
+							driver_data->name);
+			goto err_sysdev_unreg;
+		}
+	}
+
+	register_hotcpu_notifier(&cpufreq_cpu_notifier);
+	pr_debug("driver %s up and running\n", driver_data->name);
+
+	return 0;
+err_sysdev_unreg:
+	sysdev_driver_unregister(&cpu_sysdev_class,
+			&cpufreq_sysdev_driver);
+err_null_driver:
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+	cpufreq_driver = NULL;
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(cpufreq_register_driver);
+
+
+/**
+ * cpufreq_unregister_driver - unregister the current CPUFreq driver
+ *
+ *    Unregister the current CPUFreq driver. Only call this if you have
+ * the right to do so, i.e. if you have succeeded in initialising before!
+ * Returns zero if successful, and -EINVAL if the cpufreq_driver is
+ * currently not initialised.
+ */
+int cpufreq_unregister_driver(struct cpufreq_driver *driver)
+{
+	unsigned long flags;
+
+	if (!cpufreq_driver || (driver != cpufreq_driver))
+		return -EINVAL;
+
+	pr_debug("unregistering driver %s\n", driver->name);
+
+	sysdev_driver_unregister(&cpu_sysdev_class, &cpufreq_sysdev_driver);
+	unregister_hotcpu_notifier(&cpufreq_cpu_notifier);
+
+	spin_lock_irqsave(&cpufreq_driver_lock, flags);
+	cpufreq_driver = NULL;
+	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(cpufreq_unregister_driver);
+
+static int __init cpufreq_core_init(void)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu) {
+		per_cpu(cpufreq_policy_cpu, cpu) = -1;
+		init_rwsem(&per_cpu(cpu_policy_rwsem, cpu));
+	}
+
+	cpufreq_global_kobject = kobject_create_and_add("cpufreq",
+						&cpu_sysdev_class.kset.kobj);
+	BUG_ON(!cpufreq_global_kobject);
+	register_syscore_ops(&cpufreq_syscore_ops);
+
+	return 0;
+}
+core_initcall(cpufreq_core_init);
diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
new file mode 100644
index 00000000..fbd15688
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_conservative.c
@@ -0,0 +1,630 @@
+/*
+ *  drivers/cpufreq/cpufreq_conservative.c
+ *
+ *  Copyright (C)  2001 Russell King
+ *            (C)  2003 Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>.
+ *                      Jun Nakajima <jun.nakajima@intel.com>
+ *            (C)  2009 Alexander Clouter <alex@digriz.org.uk>
+ *  Copyright (C) 2012 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/cpu.h>
+#include <linux/jiffies.h>
+#include <linux/kernel_stat.h>
+#include <linux/mutex.h>
+#include <linux/hrtimer.h>
+#include <linux/tick.h>
+#include <linux/ktime.h>
+#include <linux/sched.h>
+
+/*
+ * dbs is used in this file as a shortform for demandbased switching
+ * It helps to keep variable names smaller, simpler
+ */
+
+#define DEF_FREQUENCY_UP_THRESHOLD		(80)
+#define DEF_FREQUENCY_DOWN_THRESHOLD		(20)
+
+/*
+ * The polling frequency of this governor depends on the capability of
+ * the processor. Default polling frequency is 1000 times the transition
+ * latency of the processor. The governor will work on any processor with
+ * transition latency <= 10mS, using appropriate sampling
+ * rate.
+ * For CPUs with transition latency > 10mS (mostly drivers with CPUFREQ_ETERNAL)
+ * this governor will not work.
+ * All times here are in uS.
+ */
+#define MIN_SAMPLING_RATE_RATIO			(2)
+
+static unsigned int min_sampling_rate;
+
+#define LATENCY_MULTIPLIER			(1000)
+#define MIN_LATENCY_MULTIPLIER			(100)
+#define DEF_SAMPLING_DOWN_FACTOR		(1)
+#define MAX_SAMPLING_DOWN_FACTOR		(10)
+#define TRANSITION_LATENCY_LIMIT		(10 * 1000 * 1000)
+
+static void do_dbs_timer(struct work_struct *work);
+
+struct cpu_dbs_info_s {
+	cputime64_t prev_cpu_idle;
+	cputime64_t prev_cpu_wall;
+	cputime64_t prev_cpu_nice;
+	struct cpufreq_policy *cur_policy;
+	struct delayed_work work;
+	unsigned int down_skip;
+	unsigned int requested_freq;
+	int cpu;
+	unsigned int enable:1;
+	/*
+	 * percpu mutex that serializes governor limit change with
+	 * do_dbs_timer invocation. We do not want do_dbs_timer to run
+	 * when user is changing the governor or limits.
+	 */
+	struct mutex timer_mutex;
+};
+static DEFINE_PER_CPU(struct cpu_dbs_info_s, cs_cpu_dbs_info);
+
+static unsigned int dbs_enable;	/* number of CPUs using this policy */
+
+/*
+ * dbs_mutex protects dbs_enable in governor start/stop.
+ */
+static DEFINE_MUTEX(dbs_mutex);
+
+static struct dbs_tuners {
+	unsigned int sampling_rate;
+	unsigned int sampling_down_factor;
+	unsigned int up_threshold;
+	unsigned int down_threshold;
+	unsigned int ignore_nice;
+	unsigned int freq_step;
+} dbs_tuners_ins = {
+	.up_threshold = DEF_FREQUENCY_UP_THRESHOLD,
+	.down_threshold = DEF_FREQUENCY_DOWN_THRESHOLD,
+	.sampling_down_factor = DEF_SAMPLING_DOWN_FACTOR,
+	.ignore_nice = 0,
+	.freq_step = 5,
+};
+
+static inline cputime64_t get_cpu_idle_time_jiffy(unsigned int cpu,
+							cputime64_t *wall)
+{
+	cputime64_t idle_time;
+	cputime64_t cur_wall_time;
+	cputime64_t busy_time;
+
+	cur_wall_time = jiffies64_to_cputime64(get_jiffies_64());
+	busy_time = cputime64_add(kstat_cpu(cpu).cpustat.user,
+			kstat_cpu(cpu).cpustat.system);
+
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.irq);
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.softirq);
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.steal);
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.nice);
+
+	idle_time = cputime64_sub(cur_wall_time, busy_time);
+	if (wall)
+		*wall = (cputime64_t)jiffies_to_usecs(cur_wall_time);
+
+	return (cputime64_t)jiffies_to_usecs(idle_time);
+}
+
+static inline cputime64_t get_cpu_idle_time(unsigned int cpu, cputime64_t *wall)
+{
+	u64 idle_time = get_cpu_idle_time_us(cpu, wall);
+
+	if (idle_time == -1ULL)
+		return get_cpu_idle_time_jiffy(cpu, wall);
+
+	return idle_time;
+}
+
+/* keep track of frequency transitions */
+static int
+dbs_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
+		     void *data)
+{
+	struct cpufreq_freqs *freq = data;
+	struct cpu_dbs_info_s *this_dbs_info = &per_cpu(cs_cpu_dbs_info,
+							freq->cpu);
+
+	struct cpufreq_policy *policy;
+
+	if (!this_dbs_info->enable)
+		return 0;
+
+	policy = this_dbs_info->cur_policy;
+
+	/*
+	 * we only care if our internally tracked freq moves outside
+	 * the 'valid' ranges of freqency available to us otherwise
+	 * we do not change it
+	*/
+	if (this_dbs_info->requested_freq > policy->max
+			|| this_dbs_info->requested_freq < policy->min)
+		this_dbs_info->requested_freq = freq->new;
+
+	return 0;
+}
+
+static struct notifier_block dbs_cpufreq_notifier_block = {
+	.notifier_call = dbs_cpufreq_notifier
+};
+
+/************************** sysfs interface ************************/
+static ssize_t show_sampling_rate_min(struct kobject *kobj,
+				      struct attribute *attr, char *buf)
+{
+	return sprintf(buf, "%u\n", min_sampling_rate);
+}
+
+define_one_global_ro(sampling_rate_min);
+
+/* cpufreq_conservative Governor Tunables */
+#define show_one(file_name, object)					\
+static ssize_t show_##file_name						\
+(struct kobject *kobj, struct attribute *attr, char *buf)		\
+{									\
+	return sprintf(buf, "%u\n", dbs_tuners_ins.object);		\
+}
+show_one(sampling_rate, sampling_rate);
+show_one(sampling_down_factor, sampling_down_factor);
+show_one(up_threshold, up_threshold);
+show_one(down_threshold, down_threshold);
+show_one(ignore_nice_load, ignore_nice);
+show_one(freq_step, freq_step);
+
+static ssize_t store_sampling_down_factor(struct kobject *a,
+					  struct attribute *b,
+					  const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1 || input > MAX_SAMPLING_DOWN_FACTOR || input < 1)
+		return -EINVAL;
+
+	dbs_tuners_ins.sampling_down_factor = input;
+	return count;
+}
+
+static ssize_t store_sampling_rate(struct kobject *a, struct attribute *b,
+				   const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1)
+		return -EINVAL;
+
+	dbs_tuners_ins.sampling_rate = max(input, min_sampling_rate);
+	return count;
+}
+
+static ssize_t store_up_threshold(struct kobject *a, struct attribute *b,
+				  const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1 || input > 100 ||
+			input <= dbs_tuners_ins.down_threshold)
+		return -EINVAL;
+
+	dbs_tuners_ins.up_threshold = input;
+	return count;
+}
+
+static ssize_t store_down_threshold(struct kobject *a, struct attribute *b,
+				    const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	/* cannot be lower than 11 otherwise freq will not fall */
+	if (ret != 1 || input < 11 || input > 100 ||
+			input >= dbs_tuners_ins.up_threshold)
+		return -EINVAL;
+
+	dbs_tuners_ins.down_threshold = input;
+	return count;
+}
+
+static ssize_t store_ignore_nice_load(struct kobject *a, struct attribute *b,
+				      const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+
+	unsigned int j;
+
+	ret = sscanf(buf, "%u", &input);
+	if (ret != 1)
+		return -EINVAL;
+
+	if (input > 1)
+		input = 1;
+
+	if (input == dbs_tuners_ins.ignore_nice) /* nothing to do */
+		return count;
+
+	dbs_tuners_ins.ignore_nice = input;
+
+	/* we need to re-evaluate prev_cpu_idle */
+	for_each_online_cpu(j) {
+		struct cpu_dbs_info_s *dbs_info;
+		dbs_info = &per_cpu(cs_cpu_dbs_info, j);
+		dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
+						&dbs_info->prev_cpu_wall);
+		if (dbs_tuners_ins.ignore_nice)
+			dbs_info->prev_cpu_nice = kstat_cpu(j).cpustat.nice;
+	}
+	return count;
+}
+
+static ssize_t store_freq_step(struct kobject *a, struct attribute *b,
+			       const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1)
+		return -EINVAL;
+
+	if (input > 100)
+		input = 100;
+
+	/* no need to test here if freq_step is zero as the user might actually
+	 * want this, they would be crazy though :) */
+	dbs_tuners_ins.freq_step = input;
+	return count;
+}
+
+define_one_global_rw(sampling_rate);
+define_one_global_rw(sampling_down_factor);
+define_one_global_rw(up_threshold);
+define_one_global_rw(down_threshold);
+define_one_global_rw(ignore_nice_load);
+define_one_global_rw(freq_step);
+
+static struct attribute *dbs_attributes[] = {
+	&sampling_rate_min.attr,
+	&sampling_rate.attr,
+	&sampling_down_factor.attr,
+	&up_threshold.attr,
+	&down_threshold.attr,
+	&ignore_nice_load.attr,
+	&freq_step.attr,
+	NULL
+};
+
+static struct attribute_group dbs_attr_group = {
+	.attrs = dbs_attributes,
+	.name = "conservative",
+};
+
+/************************** sysfs end ************************/
+
+static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
+{
+	unsigned int load = 0;
+	unsigned int max_load = 0;
+	unsigned int freq_target;
+
+	struct cpufreq_policy *policy;
+	unsigned int j;
+
+	policy = this_dbs_info->cur_policy;
+
+	/*
+	 * Every sampling_rate, we check, if current idle time is less
+	 * than 20% (default), then we try to increase frequency
+	 * Every sampling_rate*sampling_down_factor, we check, if current
+	 * idle time is more than 80%, then we try to decrease frequency
+	 *
+	 * Any frequency increase takes it to the maximum frequency.
+	 * Frequency reduction happens at minimum steps of
+	 * 5% (default) of maximum frequency
+	 */
+
+	/* Get Absolute Load */
+	for_each_online_cpu(j) {
+		struct cpu_dbs_info_s *j_dbs_info;
+		cputime64_t cur_wall_time, cur_idle_time;
+		unsigned int idle_time, wall_time;
+
+		j_dbs_info = &per_cpu(cs_cpu_dbs_info, j);
+
+		cur_idle_time = get_cpu_idle_time(j, &cur_wall_time);
+
+		wall_time = (unsigned int) cputime64_sub(cur_wall_time,
+				j_dbs_info->prev_cpu_wall);
+		j_dbs_info->prev_cpu_wall = cur_wall_time;
+
+		idle_time = (unsigned int) cputime64_sub(cur_idle_time,
+				j_dbs_info->prev_cpu_idle);
+		j_dbs_info->prev_cpu_idle = cur_idle_time;
+
+		if (dbs_tuners_ins.ignore_nice) {
+			cputime64_t cur_nice;
+			unsigned long cur_nice_jiffies;
+
+			cur_nice = cputime64_sub(kstat_cpu(j).cpustat.nice,
+					 j_dbs_info->prev_cpu_nice);
+			/*
+			 * Assumption: nice time between sampling periods will
+			 * be less than 2^32 jiffies for 32 bit sys
+			 */
+			cur_nice_jiffies = (unsigned long)
+					cputime64_to_jiffies64(cur_nice);
+
+			j_dbs_info->prev_cpu_nice = kstat_cpu(j).cpustat.nice;
+			idle_time += jiffies_to_usecs(cur_nice_jiffies);
+		}
+
+		if (unlikely(!wall_time || wall_time < idle_time))
+			continue;
+
+		load = 100 * (wall_time - idle_time) / wall_time;
+
+		if (load > max_load)
+			max_load = load;
+	}
+
+	/*
+	 * break out if we 'cannot' reduce the speed as the user might
+	 * want freq_step to be zero
+	 */
+	if (dbs_tuners_ins.freq_step == 0)
+		return;
+
+	/* Check for frequency increase */
+	if (max_load > dbs_tuners_ins.up_threshold) {
+		this_dbs_info->down_skip = 0;
+
+		/* if we are already at full speed then break out early */
+		if (this_dbs_info->requested_freq == policy->max)
+			return;
+
+#if 1
+		this_dbs_info->requested_freq = policy->max;
+#else
+		freq_target = (dbs_tuners_ins.freq_step * policy->max) / 100;
+
+		/* max freq cannot be less than 100. But who knows.... */
+		if (unlikely(freq_target == 0))
+			freq_target = 5;
+
+		this_dbs_info->requested_freq += freq_target;
+		if (this_dbs_info->requested_freq > policy->max)
+			this_dbs_info->requested_freq = policy->max;
+#endif
+
+		__cpufreq_driver_target(policy, this_dbs_info->requested_freq,
+			CPUFREQ_RELATION_H);
+		return;
+	}
+
+	/*
+	 * The optimal frequency is the frequency that is the lowest that
+	 * can support the current CPU usage without triggering the up
+	 * policy. To be safe, we focus 10 points under the threshold.
+	 */
+	if (max_load < (dbs_tuners_ins.down_threshold - 10)) {
+		freq_target = (dbs_tuners_ins.freq_step * policy->max) / 100;
+
+		this_dbs_info->requested_freq -= freq_target;
+		if (this_dbs_info->requested_freq < policy->min)
+			this_dbs_info->requested_freq = policy->min;
+
+		/*
+		 * if we cannot reduce the frequency anymore, break out early
+		 */
+		if (policy->cur == policy->min)
+			return;
+
+		__cpufreq_driver_target(policy, this_dbs_info->requested_freq,
+				CPUFREQ_RELATION_H);
+		return;
+	}
+}
+
+static void do_dbs_timer(struct work_struct *work)
+{
+	struct cpu_dbs_info_s *dbs_info =
+		container_of(work, struct cpu_dbs_info_s, work.work);
+	unsigned int cpu = dbs_info->cpu;
+
+	/* We want all CPUs to do sampling nearly on same jiffy */
+	int delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
+
+	delay -= jiffies % delay;
+
+	mutex_lock(&dbs_info->timer_mutex);
+
+	dbs_check_cpu(dbs_info);
+
+	schedule_delayed_work_on(cpu, &dbs_info->work, delay);
+	mutex_unlock(&dbs_info->timer_mutex);
+}
+
+static inline void dbs_timer_init(struct cpu_dbs_info_s *dbs_info)
+{
+	/* We want all CPUs to do sampling nearly on same jiffy */
+	int delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
+	delay -= jiffies % delay;
+
+	dbs_info->enable = 1;
+	INIT_DELAYED_WORK_DEFERRABLE(&dbs_info->work, do_dbs_timer);
+	schedule_delayed_work_on(dbs_info->cpu, &dbs_info->work, delay);
+}
+
+static inline void dbs_timer_exit(struct cpu_dbs_info_s *dbs_info)
+{
+	dbs_info->enable = 0;
+	cancel_delayed_work_sync(&dbs_info->work);
+}
+
+static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
+				   unsigned int event)
+{
+	unsigned int cpu = policy->cpu;
+	struct cpu_dbs_info_s *this_dbs_info;
+	unsigned int j;
+	int rc;
+
+	this_dbs_info = &per_cpu(cs_cpu_dbs_info, cpu);
+
+	switch (event) {
+	case CPUFREQ_GOV_START:
+		if ((!cpu_online(cpu)) || (!policy->cur))
+			return -EINVAL;
+
+		mutex_lock(&dbs_mutex);
+
+		for_each_online_cpu(j) {
+			struct cpu_dbs_info_s *j_dbs_info;
+			j_dbs_info = &per_cpu(cs_cpu_dbs_info, j);
+			j_dbs_info->cur_policy = policy;
+
+			j_dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
+						&j_dbs_info->prev_cpu_wall);
+			if (dbs_tuners_ins.ignore_nice) {
+				j_dbs_info->prev_cpu_nice =
+						kstat_cpu(j).cpustat.nice;
+			}
+		}
+		this_dbs_info->down_skip = 0;
+		this_dbs_info->requested_freq = policy->cur;
+
+		mutex_init(&this_dbs_info->timer_mutex);
+		dbs_enable++;
+		/*
+		 * Start the timerschedule work, when this governor
+		 * is used for first time
+		 */
+		if (dbs_enable == 1) {
+			unsigned int latency;
+			/* policy latency is in nS. Convert it to uS first */
+			latency = policy->cpuinfo.transition_latency / 1000;
+			if (latency == 0)
+				latency = 1;
+
+			rc = sysfs_create_group(cpufreq_global_kobject,
+						&dbs_attr_group);
+			if (rc) {
+				mutex_unlock(&dbs_mutex);
+				return rc;
+			}
+
+			/*
+			 * conservative does not implement micro like ondemand
+			 * governor, thus we are bound to jiffes/HZ
+			 */
+			min_sampling_rate =
+				MIN_SAMPLING_RATE_RATIO * jiffies_to_usecs(10);
+			/* Bring kernel and HW constraints together */
+			min_sampling_rate = max(min_sampling_rate,
+					MIN_LATENCY_MULTIPLIER * latency);
+			dbs_tuners_ins.sampling_rate =
+				max(min_sampling_rate,
+				    latency * LATENCY_MULTIPLIER);
+
+			cpufreq_register_notifier(
+					&dbs_cpufreq_notifier_block,
+					CPUFREQ_TRANSITION_NOTIFIER);
+		}
+		mutex_unlock(&dbs_mutex);
+
+		dbs_timer_init(this_dbs_info);
+
+		break;
+
+	case CPUFREQ_GOV_STOP:
+		dbs_timer_exit(this_dbs_info);
+
+		mutex_lock(&dbs_mutex);
+		dbs_enable--;
+		mutex_destroy(&this_dbs_info->timer_mutex);
+
+		/*
+		 * Stop the timerschedule work, when this governor
+		 * is used for first time
+		 */
+		if (dbs_enable == 0)
+			cpufreq_unregister_notifier(
+					&dbs_cpufreq_notifier_block,
+					CPUFREQ_TRANSITION_NOTIFIER);
+
+		mutex_unlock(&dbs_mutex);
+		if (!dbs_enable)
+			sysfs_remove_group(cpufreq_global_kobject,
+					   &dbs_attr_group);
+
+		break;
+
+	case CPUFREQ_GOV_LIMITS:
+		mutex_lock(&this_dbs_info->timer_mutex);
+		if (policy->max < this_dbs_info->cur_policy->cur)
+			__cpufreq_driver_target(
+					this_dbs_info->cur_policy,
+					policy->max, CPUFREQ_RELATION_H);
+		else if (policy->min > this_dbs_info->cur_policy->cur)
+			__cpufreq_driver_target(
+					this_dbs_info->cur_policy,
+					policy->min, CPUFREQ_RELATION_L);
+		mutex_unlock(&this_dbs_info->timer_mutex);
+
+		break;
+	}
+	return 0;
+}
+
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
+static
+#endif
+struct cpufreq_governor cpufreq_gov_conservative = {
+	.name			= "conservative",
+	.governor		= cpufreq_governor_dbs,
+	.max_transition_latency	= TRANSITION_LATENCY_LIMIT,
+	.owner			= THIS_MODULE,
+};
+
+static int __init cpufreq_gov_dbs_init(void)
+{
+	return cpufreq_register_governor(&cpufreq_gov_conservative);
+}
+
+static void __exit cpufreq_gov_dbs_exit(void)
+{
+	cpufreq_unregister_governor(&cpufreq_gov_conservative);
+}
+
+
+MODULE_AUTHOR("Alexander Clouter <alex@digriz.org.uk>");
+MODULE_DESCRIPTION("'cpufreq_conservative' - A dynamic cpufreq governor for "
+		"Low Latency Frequency Transition capable processors "
+		"optimised for use in a battery environment");
+MODULE_LICENSE("GPL");
+
+#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
+fs_initcall(cpufreq_gov_dbs_init);
+#else
+module_init(cpufreq_gov_dbs_init);
+#endif
+module_exit(cpufreq_gov_dbs_exit);
diff --git a/drivers/cpufreq/cpufreq_interactive.c b/drivers/cpufreq/cpufreq_interactive.c
new file mode 100644
index 00000000..2db752c1
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_interactive.c
@@ -0,0 +1,832 @@
+/*
+ * drivers/cpufreq/cpufreq_interactive.c
+ *
+ * Copyright (C) 2010 Google, Inc.
+ * Copyright (C) 2012 Freescale Semiconductor, Inc.
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * Author: Mike Chan (mike@android.com)
+ *
+ */
+
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/cpufreq.h>
+#include <linux/mutex.h>
+#include <linux/sched.h>
+#include <linux/tick.h>
+#include <linux/time.h>
+#include <linux/timer.h>
+#include <linux/workqueue.h>
+#include <linux/kthread.h>
+#include <linux/mutex.h>
+#include <linux/kernel_stat.h>
+
+#include <asm/cputime.h>
+
+static atomic_t active_count = ATOMIC_INIT(0);
+
+struct cpufreq_interactive_cpuinfo {
+	struct timer_list cpu_timer;
+	int timer_idlecancel;
+	u64 time_in_idle;
+	u64 idle_exit_time;
+	u64 timer_run_time;
+	int idling;
+	u64 freq_change_time;
+	u64 freq_change_time_in_idle;
+	struct cpufreq_policy *policy;
+	struct cpufreq_frequency_table *freq_table;
+	unsigned int target_freq;
+	int governor_enabled;
+};
+
+static DEFINE_PER_CPU(struct cpufreq_interactive_cpuinfo, cpuinfo);
+
+/* Workqueues handle frequency scaling */
+static struct task_struct *up_task;
+static struct workqueue_struct *down_wq;
+static struct work_struct freq_scale_down_work;
+static cpumask_t up_cpumask;
+static spinlock_t up_cpumask_lock;
+static cpumask_t down_cpumask;
+static spinlock_t down_cpumask_lock;
+static struct mutex set_speed_lock;
+
+/* Hi speed to bump to from lo speed when load burst (default max) */
+static u64 hispeed_freq;
+
+/* Go to hi speed when CPU load at or above this value. */
+#define DEFAULT_GO_HISPEED_LOAD 95
+static unsigned long go_hispeed_load;
+
+/*
+ * The minimum amount of time to spend at a frequency before we can ramp down.
+ */
+#define DEFAULT_MIN_SAMPLE_TIME (20 * USEC_PER_MSEC)
+static unsigned long min_sample_time;
+
+/*
+ * The sample rate of the timer used to increase frequency
+ */
+#define DEFAULT_TIMER_RATE (50 * USEC_PER_MSEC)
+#define CPUFREQ_IRQ_LEN 60
+#define CPUFREQ_NOTE_LEN 120
+static unsigned long timer_rate;
+
+static int cpufreq_governor_interactive(struct cpufreq_policy *policy,
+		unsigned int event);
+
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_INTERACTIVE
+static
+#endif
+struct cpufreq_governor cpufreq_gov_interactive = {
+	.name = "interactive",
+	.governor = cpufreq_governor_interactive,
+	.max_transition_latency = 10000000,
+	.owner = THIS_MODULE,
+};
+
+static struct irq_tuner irq_tuner_ins[MAX_CPUFREQ_IRQ_NUMBER];
+static struct irq_desc *cpufreq_irq_desc[MAX_CPUFREQ_IRQ_NUMBER];
+
+static bool cpufreq_interactive_check_irq(void)
+{
+	bool val;
+	unsigned int irq_count = 0, i;
+	static unsigned int irq_count_start[MAX_CPUFREQ_IRQ_NUMBER];
+	static unsigned int irq_count_end[MAX_CPUFREQ_IRQ_NUMBER];
+
+	val = false;
+	for (i = 0; i < MAX_CPUFREQ_IRQ_NUMBER; i++) {
+		if (irq_tuner_ins[i].irq_number == 0)
+			break;
+		if (!irq_tuner_ins[i].enable)
+			continue;
+		if (irq_count_start[i] == 0)
+			irq_count_start[i] = cpufreq_irq_desc[i] &&
+			cpufreq_irq_desc[i]->kstat_irqs ?
+			*per_cpu_ptr(cpufreq_irq_desc[i]->kstat_irqs, 0) : 0;
+		else if (irq_count_end[i] == 0)
+			irq_count_end[i] = cpufreq_irq_desc[i] &&
+			cpufreq_irq_desc[i]->kstat_irqs ?
+			*per_cpu_ptr(cpufreq_irq_desc[i]->kstat_irqs, 0) : 0;
+		else {
+			irq_count = irq_count_end[i] - irq_count_start[i];
+			irq_count_start[i] = irq_count_end[i];
+			irq_count_end[i] = 0;
+		}
+		if (irq_count > irq_tuner_ins[i].up_threshold) {
+			irq_count = 0;
+			val = true;
+			break;
+		}
+	}
+
+	return val;
+}
+static void cpufreq_interactive_timer(unsigned long data)
+{
+	unsigned int delta_idle;
+	unsigned int delta_time;
+	int cpu_load;
+	int load_since_change;
+	u64 time_in_idle;
+	u64 idle_exit_time;
+	struct cpufreq_interactive_cpuinfo *pcpu =
+		&per_cpu(cpuinfo, data);
+	u64 now_idle;
+	unsigned int new_freq;
+	unsigned int index;
+	unsigned long flags;
+	bool irq_load;
+
+	smp_rmb();
+
+	if (!pcpu->governor_enabled)
+		goto exit;
+
+	/*
+	 * Once pcpu->timer_run_time is updated to >= pcpu->idle_exit_time,
+	 * this lets idle exit know the current idle time sample has
+	 * been processed, and idle exit can generate a new sample and
+	 * re-arm the timer.  This prevents a concurrent idle
+	 * exit on that CPU from writing a new set of info at the same time
+	 * the timer function runs (the timer function can't use that info
+	 * until more time passes).
+	 */
+	time_in_idle = pcpu->time_in_idle;
+	idle_exit_time = pcpu->idle_exit_time;
+	now_idle = get_cpu_idle_time_us(data, &pcpu->timer_run_time);
+	smp_wmb();
+
+	/* If we raced with cancelling a timer, skip. */
+	if (!idle_exit_time)
+		goto exit;
+
+	delta_idle = (unsigned int) cputime64_sub(now_idle, time_in_idle);
+	delta_time = (unsigned int) cputime64_sub(pcpu->timer_run_time,
+						  idle_exit_time);
+
+	/*
+	 * If timer ran less than 1ms after short-term sample started, retry.
+	 */
+	if (delta_time < 1000)
+		goto rearm;
+
+	if (delta_idle > delta_time)
+		cpu_load = 0;
+	else
+		cpu_load = 100 * (delta_time - delta_idle) / delta_time;
+
+	delta_idle = (unsigned int) cputime64_sub(now_idle,
+						pcpu->freq_change_time_in_idle);
+	delta_time = (unsigned int) cputime64_sub(pcpu->timer_run_time,
+						  pcpu->freq_change_time);
+
+	if ((delta_time == 0) || (delta_idle > delta_time))
+		load_since_change = 0;
+	else
+		load_since_change =
+			100 * (delta_time - delta_idle) / delta_time;
+
+	/*
+	 * Choose greater of short-term load (since last idle timer
+	 * started or timer function re-armed itself) or long-term load
+	 * (since last frequency change).
+	 */
+	if (load_since_change > cpu_load)
+		cpu_load = load_since_change;
+
+	irq_load = cpufreq_interactive_check_irq();
+	if (cpu_load >= go_hispeed_load || irq_load) {
+		if (pcpu->policy->cur == pcpu->policy->min)
+			new_freq = hispeed_freq;
+		else
+			new_freq = pcpu->policy->max * cpu_load / 100;
+		if (irq_load)
+			new_freq = hispeed_freq;
+	} else {
+		new_freq = pcpu->policy->cur * cpu_load / 100;
+	}
+
+	if (cpufreq_frequency_table_target(pcpu->policy, pcpu->freq_table,
+					   new_freq, CPUFREQ_RELATION_H,
+					   &index)) {
+		pr_warn_once("timer %d: cpufreq_frequency_table_target error\n",
+			     (int) data);
+		goto rearm;
+	}
+
+	new_freq = pcpu->freq_table[index].frequency;
+	if (pcpu->target_freq == new_freq)
+		goto rearm_if_notmax;
+
+	/*
+	 * Do not scale down unless we have been at this frequency for the
+	 * minimum sample time.
+	 */
+	if (new_freq < pcpu->target_freq) {
+		if (cputime64_sub(pcpu->timer_run_time, pcpu->freq_change_time)
+		    < min_sample_time)
+			goto rearm;
+	}
+
+	if (new_freq < pcpu->target_freq) {
+		pcpu->target_freq = new_freq;
+		spin_lock_irqsave(&down_cpumask_lock, flags);
+		cpumask_set_cpu(data, &down_cpumask);
+		spin_unlock_irqrestore(&down_cpumask_lock, flags);
+		queue_work(down_wq, &freq_scale_down_work);
+	} else {
+		pcpu->target_freq = new_freq;
+		spin_lock_irqsave(&up_cpumask_lock, flags);
+		cpumask_set_cpu(data, &up_cpumask);
+		spin_unlock_irqrestore(&up_cpumask_lock, flags);
+		wake_up_process(up_task);
+	}
+
+rearm_if_notmax:
+	/*
+	 * Already set max speed and don't see a need to change that,
+	 * wait until next idle to re-evaluate, don't need timer.
+	 */
+	if (pcpu->target_freq == pcpu->policy->max)
+		goto exit;
+
+rearm:
+	if (!timer_pending(&pcpu->cpu_timer)) {
+		/*
+		 * If already at min: if that CPU is idle, don't set timer.
+		 * Else cancel the timer if that CPU goes idle.  We don't
+		 * need to re-evaluate speed until the next idle exit.
+		 */
+		if (pcpu->target_freq == pcpu->policy->min) {
+			smp_rmb();
+
+			if (pcpu->idling)
+				goto exit;
+
+			pcpu->timer_idlecancel = 1;
+		}
+
+		pcpu->time_in_idle = get_cpu_idle_time_us(
+			data, &pcpu->idle_exit_time);
+		mod_timer(&pcpu->cpu_timer,
+			  jiffies + usecs_to_jiffies(timer_rate));
+	}
+
+exit:
+	return;
+}
+
+static void cpufreq_interactive_idle_start(void)
+{
+	struct cpufreq_interactive_cpuinfo *pcpu =
+		&per_cpu(cpuinfo, smp_processor_id());
+	int pending;
+
+	pcpu->idling = 1;
+	smp_wmb();
+	if (!pcpu->governor_enabled)
+		return;
+	pending = timer_pending(&pcpu->cpu_timer);
+
+	if (pcpu->target_freq != pcpu->policy->min) {
+#ifdef CONFIG_SMP
+		/*
+		 * Entering idle while not at lowest speed.  On some
+		 * platforms this can hold the other CPU(s) at that speed
+		 * even though the CPU is idle. Set a timer to re-evaluate
+		 * speed so this idle CPU doesn't hold the other CPUs above
+		 * min indefinitely.  This should probably be a quirk of
+		 * the CPUFreq driver.
+		 */
+		if (!pending) {
+			pcpu->time_in_idle = get_cpu_idle_time_us(
+				smp_processor_id(), &pcpu->idle_exit_time);
+			pcpu->timer_idlecancel = 0;
+			mod_timer(&pcpu->cpu_timer,
+				  jiffies + usecs_to_jiffies(timer_rate));
+		}
+#endif
+	} else {
+		/*
+		 * If at min speed and entering idle after load has
+		 * already been evaluated, and a timer has been set just in
+		 * case the CPU suddenly goes busy, cancel that timer.  The
+		 * CPU didn't go busy; we'll recheck things upon idle exit.
+		 */
+		if (pending && pcpu->timer_idlecancel) {
+			del_timer(&pcpu->cpu_timer);
+			/*
+			 * Ensure last timer run time is after current idle
+			 * sample start time, so next idle exit will always
+			 * start a new idle sampling period.
+			 */
+			pcpu->idle_exit_time = 0;
+			pcpu->timer_idlecancel = 0;
+		}
+	}
+
+}
+
+static void cpufreq_interactive_idle_end(void)
+{
+	struct cpufreq_interactive_cpuinfo *pcpu =
+		&per_cpu(cpuinfo, smp_processor_id());
+
+	pcpu->idling = 0;
+	smp_wmb();
+
+	/*
+	 * Arm the timer for 1-2 ticks later if not already, and if the timer
+	 * function has already processed the previous load sampling
+	 * interval.  (If the timer is not pending but has not processed
+	 * the previous interval, it is probably racing with us on another
+	 * CPU.  Let it compute load based on the previous sample and then
+	 * re-arm the timer for another interval when it's done, rather
+	 * than updating the interval start time to be "now", which doesn't
+	 * give the timer function enough time to make a decision on this
+	 * run.)
+	 */
+	if (timer_pending(&pcpu->cpu_timer) == 0 &&
+	    pcpu->timer_run_time >= pcpu->idle_exit_time &&
+	    pcpu->governor_enabled) {
+		pcpu->time_in_idle =
+			get_cpu_idle_time_us(smp_processor_id(),
+					     &pcpu->idle_exit_time);
+		pcpu->timer_idlecancel = 0;
+		mod_timer(&pcpu->cpu_timer,
+			  jiffies + usecs_to_jiffies(timer_rate));
+	}
+
+}
+
+static int cpufreq_interactive_up_task(void *data)
+{
+	unsigned int cpu;
+	unsigned long flags;
+	struct cpufreq_interactive_cpuinfo *pcpu;
+
+	while (1) {
+		set_current_state(TASK_INTERRUPTIBLE);
+		spin_lock_irqsave(&up_cpumask_lock, flags);
+
+		if (cpumask_empty(&up_cpumask)) {
+			spin_unlock_irqrestore(&up_cpumask_lock, flags);
+			schedule();
+
+			if (kthread_should_stop())
+				break;
+
+			spin_lock_irqsave(&up_cpumask_lock, flags);
+		}
+
+		set_current_state(TASK_RUNNING);
+		cpumask_clear(&up_cpumask);
+		spin_unlock_irqrestore(&up_cpumask_lock, flags);
+
+		for_each_online_cpu(cpu) {
+			unsigned int j;
+			unsigned int max_freq = 0;
+
+			pcpu = &per_cpu(cpuinfo, cpu);
+			smp_rmb();
+
+			if (!pcpu->governor_enabled)
+				continue;
+
+			mutex_lock(&set_speed_lock);
+
+			for_each_online_cpu(j) {
+				struct cpufreq_interactive_cpuinfo *pjcpu =
+					&per_cpu(cpuinfo, j);
+				if (pjcpu->target_freq > max_freq)
+					max_freq = pjcpu->target_freq;
+			}
+			if (max_freq != pcpu->policy->cur)
+				__cpufreq_driver_target(pcpu->policy,
+							max_freq,
+							CPUFREQ_RELATION_H);
+			mutex_unlock(&set_speed_lock);
+
+			pcpu->freq_change_time_in_idle =
+				get_cpu_idle_time_us(cpu,
+						     &pcpu->freq_change_time);
+		}
+	}
+
+	return 0;
+}
+
+static void cpufreq_interactive_freq_down(struct work_struct *work)
+{
+	unsigned int cpu;
+	unsigned long flags;
+	struct cpufreq_interactive_cpuinfo *pcpu;
+
+	spin_lock_irqsave(&down_cpumask_lock, flags);
+	cpumask_clear(&down_cpumask);
+	spin_unlock_irqrestore(&down_cpumask_lock, flags);
+
+	for_each_online_cpu(cpu) {
+		unsigned int j;
+		unsigned int max_freq = 0;
+
+		pcpu = &per_cpu(cpuinfo, cpu);
+		smp_rmb();
+
+		if (!pcpu->governor_enabled)
+			continue;
+
+		mutex_lock(&set_speed_lock);
+
+		for_each_online_cpu(j) {
+			struct cpufreq_interactive_cpuinfo *pjcpu =
+				&per_cpu(cpuinfo, j);
+
+			if (pjcpu->target_freq > max_freq)
+				max_freq = pjcpu->target_freq;
+		}
+
+		if (max_freq != pcpu->policy->cur)
+			__cpufreq_driver_target(pcpu->policy, max_freq,
+						CPUFREQ_RELATION_H);
+
+		mutex_unlock(&set_speed_lock);
+		pcpu->freq_change_time_in_idle =
+			get_cpu_idle_time_us(cpu,
+					     &pcpu->freq_change_time);
+	}
+}
+
+static ssize_t show_hispeed_freq(struct kobject *kobj,
+				 struct attribute *attr, char *buf)
+{
+	return sprintf(buf, "%llu\n", hispeed_freq);
+}
+
+static ssize_t store_hispeed_freq(struct kobject *kobj,
+				  struct attribute *attr, const char *buf,
+				  size_t count)
+{
+	int ret;
+	u64 val;
+
+	ret = strict_strtoull(buf, 0, &val);
+	if (ret < 0)
+		return ret;
+	hispeed_freq = val;
+	return count;
+}
+
+static struct global_attr hispeed_freq_attr = __ATTR(hispeed_freq, 0644,
+		show_hispeed_freq, store_hispeed_freq);
+
+
+static ssize_t show_go_hispeed_load(struct kobject *kobj,
+				     struct attribute *attr, char *buf)
+{
+	return sprintf(buf, "%lu\n", go_hispeed_load);
+}
+
+static ssize_t store_go_hispeed_load(struct kobject *kobj,
+			struct attribute *attr, const char *buf, size_t count)
+{
+	int ret;
+	unsigned long val;
+
+	ret = strict_strtoul(buf, 0, &val);
+	if (ret < 0)
+		return ret;
+	go_hispeed_load = val;
+	return count;
+}
+
+static struct global_attr go_hispeed_load_attr = __ATTR(go_hispeed_load, 0644,
+		show_go_hispeed_load, store_go_hispeed_load);
+
+static ssize_t show_min_sample_time(struct kobject *kobj,
+				struct attribute *attr, char *buf)
+{
+	return sprintf(buf, "%lu\n", min_sample_time);
+}
+
+static ssize_t store_min_sample_time(struct kobject *kobj,
+			struct attribute *attr, const char *buf, size_t count)
+{
+	int ret;
+	unsigned long val;
+
+	ret = strict_strtoul(buf, 0, &val);
+	if (ret < 0)
+		return ret;
+	min_sample_time = val;
+	return count;
+}
+
+static struct global_attr min_sample_time_attr = __ATTR(min_sample_time, 0644,
+		show_min_sample_time, store_min_sample_time);
+
+static ssize_t show_timer_rate(struct kobject *kobj,
+			struct attribute *attr, char *buf)
+{
+	return sprintf(buf, "%lu\n", timer_rate);
+}
+
+static ssize_t store_timer_rate(struct kobject *kobj,
+			struct attribute *attr, const char *buf, size_t count)
+{
+	int ret;
+	unsigned long val;
+
+	ret = strict_strtoul(buf, 0, &val);
+	if (ret < 0)
+		return ret;
+	timer_rate = val;
+	return count;
+}
+
+static struct global_attr timer_rate_attr = __ATTR(timer_rate, 0644,
+		show_timer_rate, store_timer_rate);
+
+
+static ssize_t show_irq_param(struct kobject *kobj,
+			struct attribute *attr, char *buf)
+{
+	int i, j = 0;
+	j += scnprintf(&buf[j], CPUFREQ_NOTE_LEN, "Change irq setting by echo a data, format: 0xAABBBC, AA:irq number, BBB:up_threshold, C:enable\n");
+	for (i = 0; i < MAX_CPUFREQ_IRQ_NUMBER; i++) {
+		if (irq_tuner_ins[i].irq_number != 0)
+			j += scnprintf(&buf[j], CPUFREQ_IRQ_LEN, "irq number: 0x%x, up_threshold 0x%x, %s\n", irq_tuner_ins[i].irq_number, irq_tuner_ins[i].up_threshold, irq_tuner_ins[i].enable ? "enabled" : "disabled");
+	}
+
+	return j;
+}
+
+static ssize_t store_irq_param(struct kobject *kobj,
+			struct attribute *attr, const char *buf, size_t count)
+{
+	int ret, i;
+	unsigned long val;
+
+	ret = strict_strtoul(buf, 0, &val);
+	if (ret < 0)
+		return ret;
+
+	for (i = 0; i < MAX_CPUFREQ_IRQ_NUMBER; i++) {
+		if (irq_tuner_ins[i].irq_number == val >> 16)
+			break;
+	}
+
+	if (i >= MAX_CPUFREQ_IRQ_NUMBER) {
+		printk(KERN_WARNING "Invalid irq number!\n");
+		return -EINVAL;
+	}
+	irq_tuner_ins[i].irq_number = val >> 16;
+	irq_tuner_ins[i].up_threshold = (val & 0xFFF0) >> 4;
+	irq_tuner_ins[i].enable = (val & 0xF) ? true : false;
+
+	return count;
+}
+
+
+static struct global_attr irq_param_attr = __ATTR(irq_scaling, 0644,
+		show_irq_param, store_irq_param);
+
+static struct attribute *interactive_attributes[] = {
+	&hispeed_freq_attr.attr,
+	&go_hispeed_load_attr.attr,
+	&min_sample_time_attr.attr,
+	&timer_rate_attr.attr,
+	&irq_param_attr.attr,
+	NULL,
+};
+
+static struct attribute_group interactive_attr_group = {
+	.attrs = interactive_attributes,
+	.name = "interactive",
+};
+
+static int cpufreq_governor_interactive(struct cpufreq_policy *policy,
+		unsigned int event)
+{
+	int rc;
+	unsigned int j;
+	struct cpufreq_interactive_cpuinfo *pcpu;
+	struct cpufreq_frequency_table *freq_table;
+
+	switch (event) {
+	case CPUFREQ_GOV_START:
+		if (!cpu_online(policy->cpu))
+			return -EINVAL;
+
+		freq_table =
+			cpufreq_frequency_get_table(policy->cpu);
+
+		for_each_cpu(j, policy->cpus) {
+			pcpu = &per_cpu(cpuinfo, j);
+			pcpu->policy = policy;
+			if (pcpu->idling)
+				pcpu->target_freq = policy->min;
+			else
+				pcpu->target_freq = policy->cur;
+
+			pcpu->freq_table = freq_table;
+			pcpu->freq_change_time_in_idle =
+				get_cpu_idle_time_us(j,
+					     &pcpu->freq_change_time);
+			pcpu->governor_enabled = 1;
+			smp_wmb();
+		}
+
+		if (!hispeed_freq)
+			hispeed_freq = policy->max;
+
+		/*
+		 * Do not register the idle hook and create sysfs
+		 * entries if we have already done so.
+		 */
+		if (atomic_inc_return(&active_count) > 1)
+			return 0;
+
+		rc = sysfs_create_group(cpufreq_global_kobject,
+				&interactive_attr_group);
+		if (rc)
+			return rc;
+
+		break;
+
+	case CPUFREQ_GOV_STOP:
+		for_each_cpu(j, policy->cpus) {
+			pcpu = &per_cpu(cpuinfo, j);
+			pcpu->governor_enabled = 0;
+			smp_wmb();
+			del_timer_sync(&pcpu->cpu_timer);
+
+			/*
+			 * Reset idle exit time since we may cancel the timer
+			 * before it can run after the last idle exit time,
+			 * to avoid tripping the check in idle exit for a timer
+			 * that is trying to run.
+			 */
+			pcpu->idle_exit_time = 0;
+		}
+
+		flush_work(&freq_scale_down_work);
+		if (atomic_dec_return(&active_count) > 0)
+			return 0;
+
+		sysfs_remove_group(cpufreq_global_kobject,
+				&interactive_attr_group);
+
+		break;
+
+	case CPUFREQ_GOV_LIMITS:
+		if (policy->max < policy->cur)
+			__cpufreq_driver_target(policy,
+					policy->max, CPUFREQ_RELATION_H);
+		else if (policy->min > policy->cur)
+			__cpufreq_driver_target(policy,
+					policy->min, CPUFREQ_RELATION_L);
+		break;
+	}
+	return 0;
+}
+
+static int cpufreq_interactive_idle_notifier(struct notifier_block *nb,
+					     unsigned long val,
+					     void *data)
+{
+	switch (val) {
+	case IDLE_START:
+		cpufreq_interactive_idle_start();
+		break;
+	case IDLE_END:
+		cpufreq_interactive_idle_end();
+		break;
+	}
+
+	return 0;
+}
+
+static struct notifier_block cpufreq_interactive_idle_nb = {
+	.notifier_call = cpufreq_interactive_idle_notifier,
+};
+
+static int __init cpufreq_interactive_init(void)
+{
+	unsigned int i;
+	struct cpufreq_interactive_cpuinfo *pcpu;
+	struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 };
+
+	go_hispeed_load = DEFAULT_GO_HISPEED_LOAD;
+	min_sample_time = DEFAULT_MIN_SAMPLE_TIME;
+	timer_rate = DEFAULT_TIMER_RATE;
+
+	/* Initalize per-cpu timers */
+	for_each_possible_cpu(i) {
+		pcpu = &per_cpu(cpuinfo, i);
+		init_timer(&pcpu->cpu_timer);
+		pcpu->cpu_timer.function = cpufreq_interactive_timer;
+		pcpu->cpu_timer.data = i;
+	}
+
+	up_task = kthread_create(cpufreq_interactive_up_task, NULL,
+				 "kinteractiveup");
+	if (IS_ERR(up_task))
+		return PTR_ERR(up_task);
+
+	sched_setscheduler_nocheck(up_task, SCHED_FIFO, &param);
+	get_task_struct(up_task);
+
+	/* No rescuer thread, bind to CPU queuing the work for possibly
+	   warm cache (probably doesn't matter much). */
+	down_wq = alloc_workqueue("kinteractive_down", 0, 1);
+
+	if (!down_wq)
+		goto err_freeuptask;
+
+	INIT_WORK(&freq_scale_down_work,
+		  cpufreq_interactive_freq_down);
+
+	spin_lock_init(&up_cpumask_lock);
+	spin_lock_init(&down_cpumask_lock);
+	mutex_init(&set_speed_lock);
+
+	idle_notifier_register(&cpufreq_interactive_idle_nb);
+
+	return cpufreq_register_governor(&cpufreq_gov_interactive);
+
+err_freeuptask:
+	put_task_struct(up_task);
+	return -ENOMEM;
+}
+
+int cpufreq_gov_irq_tuner_register(struct irq_tuner dbs_irq_tuner)
+{
+	int i, ret = 0;
+	static bool init_flag;
+
+	/* Init the global irq_tuner_ins structure */
+	if (!init_flag) {
+		for (i = 0; i < MAX_CPUFREQ_IRQ_NUMBER; i++) {
+			irq_tuner_ins[i].irq_number = 0;
+			irq_tuner_ins[i].up_threshold = 0;
+			irq_tuner_ins[i].enable = 0;
+		}
+		init_flag = true;
+	}
+
+	if (dbs_irq_tuner.irq_number == 0)
+		return -EINVAL;
+	/* Find an unused struct */
+	for (i = 0; i < MAX_CPUFREQ_IRQ_NUMBER; i++) {
+		if (irq_tuner_ins[i].irq_number != 0)
+			continue;
+		else
+			break;
+	}
+	/* Check index */
+	if (i >= MAX_CPUFREQ_IRQ_NUMBER) {
+		printk(KERN_WARNING "Too many irq number requested!\n");
+		return -EINVAL;
+	}
+
+	irq_tuner_ins[i].irq_number = dbs_irq_tuner.irq_number;
+	irq_tuner_ins[i].up_threshold = dbs_irq_tuner.up_threshold;
+	irq_tuner_ins[i].enable = dbs_irq_tuner.enable;
+	cpufreq_irq_desc[i] = irq_to_desc(irq_tuner_ins[i].irq_number);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(cpufreq_gov_irq_tuner_register);
+#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_INTERACTIVE
+late_initcall(cpufreq_interactive_init);
+#else
+module_init(cpufreq_interactive_init);
+#endif
+
+static void __exit cpufreq_interactive_exit(void)
+{
+	cpufreq_unregister_governor(&cpufreq_gov_interactive);
+	kthread_stop(up_task);
+	put_task_struct(up_task);
+	destroy_workqueue(down_wq);
+}
+
+module_exit(cpufreq_interactive_exit);
+
+MODULE_AUTHOR("Mike Chan <mike@android.com>");
+MODULE_DESCRIPTION("'cpufreq_interactive' - A cpufreq governor for "
+	"Latency sensitive workloads");
+MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
new file mode 100644
index 00000000..891360ed
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_ondemand.c
@@ -0,0 +1,759 @@
+/*
+ *  drivers/cpufreq/cpufreq_ondemand.c
+ *
+ *  Copyright (C)  2001 Russell King
+ *            (C)  2003 Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>.
+ *                      Jun Nakajima <jun.nakajima@intel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/cpu.h>
+#include <linux/jiffies.h>
+#include <linux/kernel_stat.h>
+#include <linux/mutex.h>
+#include <linux/hrtimer.h>
+#include <linux/tick.h>
+#include <linux/ktime.h>
+#include <linux/sched.h>
+
+/*
+ * dbs is used in this file as a shortform for demandbased switching
+ * It helps to keep variable names smaller, simpler
+ */
+
+#define DEF_FREQUENCY_DOWN_DIFFERENTIAL		(10)
+#define DEF_FREQUENCY_UP_THRESHOLD		(80)
+#define DEF_SAMPLING_DOWN_FACTOR		(1)
+#define MAX_SAMPLING_DOWN_FACTOR		(100000)
+#define MICRO_FREQUENCY_DOWN_DIFFERENTIAL	(3)
+#define MICRO_FREQUENCY_UP_THRESHOLD		(95)
+#define MICRO_FREQUENCY_MIN_SAMPLE_RATE		(10000)
+#define MIN_FREQUENCY_UP_THRESHOLD		(11)
+#define MAX_FREQUENCY_UP_THRESHOLD		(100)
+
+/*
+ * The polling frequency of this governor depends on the capability of
+ * the processor. Default polling frequency is 1000 times the transition
+ * latency of the processor. The governor will work on any processor with
+ * transition latency <= 10mS, using appropriate sampling
+ * rate.
+ * For CPUs with transition latency > 10mS (mostly drivers with CPUFREQ_ETERNAL)
+ * this governor will not work.
+ * All times here are in uS.
+ */
+#define MIN_SAMPLING_RATE_RATIO			(2)
+
+static unsigned int min_sampling_rate;
+
+#define LATENCY_MULTIPLIER			(1000)
+#define MIN_LATENCY_MULTIPLIER			(100)
+#define TRANSITION_LATENCY_LIMIT		(10 * 1000 * 1000)
+
+static void do_dbs_timer(struct work_struct *work);
+static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
+				unsigned int event);
+
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND
+static
+#endif
+struct cpufreq_governor cpufreq_gov_ondemand = {
+       .name                   = "ondemand",
+       .governor               = cpufreq_governor_dbs,
+       .max_transition_latency = TRANSITION_LATENCY_LIMIT,
+       .owner                  = THIS_MODULE,
+};
+
+/* Sampling types */
+enum {DBS_NORMAL_SAMPLE, DBS_SUB_SAMPLE};
+
+struct cpu_dbs_info_s {
+	cputime64_t prev_cpu_idle;
+	cputime64_t prev_cpu_iowait;
+	cputime64_t prev_cpu_wall;
+	cputime64_t prev_cpu_nice;
+	struct cpufreq_policy *cur_policy;
+	struct delayed_work work;
+	struct cpufreq_frequency_table *freq_table;
+	unsigned int freq_lo;
+	unsigned int freq_lo_jiffies;
+	unsigned int freq_hi_jiffies;
+	unsigned int rate_mult;
+	int cpu;
+	unsigned int sample_type:1;
+	/*
+	 * percpu mutex that serializes governor limit change with
+	 * do_dbs_timer invocation. We do not want do_dbs_timer to run
+	 * when user is changing the governor or limits.
+	 */
+	struct mutex timer_mutex;
+};
+static DEFINE_PER_CPU(struct cpu_dbs_info_s, od_cpu_dbs_info);
+
+static unsigned int dbs_enable;	/* number of CPUs using this policy */
+
+/*
+ * dbs_mutex protects dbs_enable in governor start/stop.
+ */
+static DEFINE_MUTEX(dbs_mutex);
+
+static struct dbs_tuners {
+	unsigned int sampling_rate;
+	unsigned int up_threshold;
+	unsigned int down_differential;
+	unsigned int ignore_nice;
+	unsigned int sampling_down_factor;
+	unsigned int powersave_bias;
+	unsigned int io_is_busy;
+} dbs_tuners_ins = {
+	.up_threshold = DEF_FREQUENCY_UP_THRESHOLD,
+	.sampling_down_factor = DEF_SAMPLING_DOWN_FACTOR,
+	.down_differential = DEF_FREQUENCY_DOWN_DIFFERENTIAL,
+	.ignore_nice = 0,
+	.powersave_bias = 0,
+};
+
+static inline cputime64_t get_cpu_idle_time_jiffy(unsigned int cpu,
+							cputime64_t *wall)
+{
+	cputime64_t idle_time;
+	cputime64_t cur_wall_time;
+	cputime64_t busy_time;
+
+	cur_wall_time = jiffies64_to_cputime64(get_jiffies_64());
+	busy_time = cputime64_add(kstat_cpu(cpu).cpustat.user,
+			kstat_cpu(cpu).cpustat.system);
+
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.irq);
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.softirq);
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.steal);
+	busy_time = cputime64_add(busy_time, kstat_cpu(cpu).cpustat.nice);
+
+	idle_time = cputime64_sub(cur_wall_time, busy_time);
+	if (wall)
+		*wall = (cputime64_t)jiffies_to_usecs(cur_wall_time);
+
+	return (cputime64_t)jiffies_to_usecs(idle_time);
+}
+
+static inline cputime64_t get_cpu_idle_time(unsigned int cpu, cputime64_t *wall)
+{
+	u64 idle_time = get_cpu_idle_time_us(cpu, wall);
+
+	if (idle_time == -1ULL)
+		return get_cpu_idle_time_jiffy(cpu, wall);
+
+	return idle_time;
+}
+
+static inline cputime64_t get_cpu_iowait_time(unsigned int cpu, cputime64_t *wall)
+{
+	u64 iowait_time = get_cpu_iowait_time_us(cpu, wall);
+
+	if (iowait_time == -1ULL)
+		return 0;
+
+	return iowait_time;
+}
+
+/*
+ * Find right freq to be set now with powersave_bias on.
+ * Returns the freq_hi to be used right now and will set freq_hi_jiffies,
+ * freq_lo, and freq_lo_jiffies in percpu area for averaging freqs.
+ */
+static unsigned int powersave_bias_target(struct cpufreq_policy *policy,
+					  unsigned int freq_next,
+					  unsigned int relation)
+{
+	unsigned int freq_req, freq_reduc, freq_avg;
+	unsigned int freq_hi, freq_lo;
+	unsigned int index = 0;
+	unsigned int jiffies_total, jiffies_hi, jiffies_lo;
+	struct cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info,
+						   policy->cpu);
+
+	if (!dbs_info->freq_table) {
+		dbs_info->freq_lo = 0;
+		dbs_info->freq_lo_jiffies = 0;
+		return freq_next;
+	}
+
+	cpufreq_frequency_table_target(policy, dbs_info->freq_table, freq_next,
+			relation, &index);
+	freq_req = dbs_info->freq_table[index].frequency;
+	freq_reduc = freq_req * dbs_tuners_ins.powersave_bias / 1000;
+	freq_avg = freq_req - freq_reduc;
+
+	/* Find freq bounds for freq_avg in freq_table */
+	index = 0;
+	cpufreq_frequency_table_target(policy, dbs_info->freq_table, freq_avg,
+			CPUFREQ_RELATION_H, &index);
+	freq_lo = dbs_info->freq_table[index].frequency;
+	index = 0;
+	cpufreq_frequency_table_target(policy, dbs_info->freq_table, freq_avg,
+			CPUFREQ_RELATION_L, &index);
+	freq_hi = dbs_info->freq_table[index].frequency;
+
+	/* Find out how long we have to be in hi and lo freqs */
+	if (freq_hi == freq_lo) {
+		dbs_info->freq_lo = 0;
+		dbs_info->freq_lo_jiffies = 0;
+		return freq_lo;
+	}
+	jiffies_total = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
+	jiffies_hi = (freq_avg - freq_lo) * jiffies_total;
+	jiffies_hi += ((freq_hi - freq_lo) / 2);
+	jiffies_hi /= (freq_hi - freq_lo);
+	jiffies_lo = jiffies_total - jiffies_hi;
+	dbs_info->freq_lo = freq_lo;
+	dbs_info->freq_lo_jiffies = jiffies_lo;
+	dbs_info->freq_hi_jiffies = jiffies_hi;
+	return freq_hi;
+}
+
+static void ondemand_powersave_bias_init_cpu(int cpu)
+{
+	struct cpu_dbs_info_s *dbs_info = &per_cpu(od_cpu_dbs_info, cpu);
+	dbs_info->freq_table = cpufreq_frequency_get_table(cpu);
+	dbs_info->freq_lo = 0;
+}
+
+static void ondemand_powersave_bias_init(void)
+{
+	int i;
+	for_each_online_cpu(i) {
+		ondemand_powersave_bias_init_cpu(i);
+	}
+}
+
+/************************** sysfs interface ************************/
+
+static ssize_t show_sampling_rate_min(struct kobject *kobj,
+				      struct attribute *attr, char *buf)
+{
+	return sprintf(buf, "%u\n", min_sampling_rate);
+}
+
+define_one_global_ro(sampling_rate_min);
+
+/* cpufreq_ondemand Governor Tunables */
+#define show_one(file_name, object)					\
+static ssize_t show_##file_name						\
+(struct kobject *kobj, struct attribute *attr, char *buf)              \
+{									\
+	return sprintf(buf, "%u\n", dbs_tuners_ins.object);		\
+}
+show_one(sampling_rate, sampling_rate);
+show_one(io_is_busy, io_is_busy);
+show_one(up_threshold, up_threshold);
+show_one(sampling_down_factor, sampling_down_factor);
+show_one(ignore_nice_load, ignore_nice);
+show_one(powersave_bias, powersave_bias);
+
+static ssize_t store_sampling_rate(struct kobject *a, struct attribute *b,
+				   const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+	if (ret != 1)
+		return -EINVAL;
+	dbs_tuners_ins.sampling_rate = max(input, min_sampling_rate);
+	return count;
+}
+
+static ssize_t store_io_is_busy(struct kobject *a, struct attribute *b,
+				   const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+
+	ret = sscanf(buf, "%u", &input);
+	if (ret != 1)
+		return -EINVAL;
+	dbs_tuners_ins.io_is_busy = !!input;
+	return count;
+}
+
+static ssize_t store_up_threshold(struct kobject *a, struct attribute *b,
+				  const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1 || input > MAX_FREQUENCY_UP_THRESHOLD ||
+			input < MIN_FREQUENCY_UP_THRESHOLD) {
+		return -EINVAL;
+	}
+	dbs_tuners_ins.up_threshold = input;
+	return count;
+}
+
+static ssize_t store_sampling_down_factor(struct kobject *a,
+			struct attribute *b, const char *buf, size_t count)
+{
+	unsigned int input, j;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1 || input > MAX_SAMPLING_DOWN_FACTOR || input < 1)
+		return -EINVAL;
+	dbs_tuners_ins.sampling_down_factor = input;
+
+	/* Reset down sampling multiplier in case it was active */
+	for_each_online_cpu(j) {
+		struct cpu_dbs_info_s *dbs_info;
+		dbs_info = &per_cpu(od_cpu_dbs_info, j);
+		dbs_info->rate_mult = 1;
+	}
+	return count;
+}
+
+static ssize_t store_ignore_nice_load(struct kobject *a, struct attribute *b,
+				      const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+
+	unsigned int j;
+
+	ret = sscanf(buf, "%u", &input);
+	if (ret != 1)
+		return -EINVAL;
+
+	if (input > 1)
+		input = 1;
+
+	if (input == dbs_tuners_ins.ignore_nice) { /* nothing to do */
+		return count;
+	}
+	dbs_tuners_ins.ignore_nice = input;
+
+	/* we need to re-evaluate prev_cpu_idle */
+	for_each_online_cpu(j) {
+		struct cpu_dbs_info_s *dbs_info;
+		dbs_info = &per_cpu(od_cpu_dbs_info, j);
+		dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
+						&dbs_info->prev_cpu_wall);
+		if (dbs_tuners_ins.ignore_nice)
+			dbs_info->prev_cpu_nice = kstat_cpu(j).cpustat.nice;
+
+	}
+	return count;
+}
+
+static ssize_t store_powersave_bias(struct kobject *a, struct attribute *b,
+				    const char *buf, size_t count)
+{
+	unsigned int input;
+	int ret;
+	ret = sscanf(buf, "%u", &input);
+
+	if (ret != 1)
+		return -EINVAL;
+
+	if (input > 1000)
+		input = 1000;
+
+	dbs_tuners_ins.powersave_bias = input;
+	ondemand_powersave_bias_init();
+	return count;
+}
+
+define_one_global_rw(sampling_rate);
+define_one_global_rw(io_is_busy);
+define_one_global_rw(up_threshold);
+define_one_global_rw(sampling_down_factor);
+define_one_global_rw(ignore_nice_load);
+define_one_global_rw(powersave_bias);
+
+static struct attribute *dbs_attributes[] = {
+	&sampling_rate_min.attr,
+	&sampling_rate.attr,
+	&up_threshold.attr,
+	&sampling_down_factor.attr,
+	&ignore_nice_load.attr,
+	&powersave_bias.attr,
+	&io_is_busy.attr,
+	NULL
+};
+
+static struct attribute_group dbs_attr_group = {
+	.attrs = dbs_attributes,
+	.name = "ondemand",
+};
+
+/************************** sysfs end ************************/
+
+static void dbs_freq_increase(struct cpufreq_policy *p, unsigned int freq)
+{
+	if (dbs_tuners_ins.powersave_bias)
+		freq = powersave_bias_target(p, freq, CPUFREQ_RELATION_H);
+	else if (p->cur == p->max)
+		return;
+
+	__cpufreq_driver_target(p, freq, dbs_tuners_ins.powersave_bias ?
+			CPUFREQ_RELATION_L : CPUFREQ_RELATION_H);
+}
+
+static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
+{
+	unsigned int max_load_freq;
+
+	struct cpufreq_policy *policy;
+	unsigned int j;
+
+	this_dbs_info->freq_lo = 0;
+	policy = this_dbs_info->cur_policy;
+
+	/*
+	 * Every sampling_rate, we check, if current idle time is less
+	 * than 20% (default), then we try to increase frequency
+	 * Every sampling_rate, we look for a the lowest
+	 * frequency which can sustain the load while keeping idle time over
+	 * 30%. If such a frequency exist, we try to decrease to this frequency.
+	 *
+	 * Any frequency increase takes it to the maximum frequency.
+	 * Frequency reduction happens at minimum steps of
+	 * 5% (default) of current frequency
+	 */
+
+	/* Get Absolute Load - in terms of freq */
+	max_load_freq = 0;
+
+	for_each_cpu(j, policy->cpus) {
+		struct cpu_dbs_info_s *j_dbs_info;
+		cputime64_t cur_wall_time, cur_idle_time, cur_iowait_time;
+		unsigned int idle_time, wall_time, iowait_time;
+		unsigned int load, load_freq;
+		int freq_avg;
+
+		j_dbs_info = &per_cpu(od_cpu_dbs_info, j);
+
+		cur_idle_time = get_cpu_idle_time(j, &cur_wall_time);
+		cur_iowait_time = get_cpu_iowait_time(j, &cur_wall_time);
+
+		wall_time = (unsigned int) cputime64_sub(cur_wall_time,
+				j_dbs_info->prev_cpu_wall);
+		j_dbs_info->prev_cpu_wall = cur_wall_time;
+
+		idle_time = (unsigned int) cputime64_sub(cur_idle_time,
+				j_dbs_info->prev_cpu_idle);
+		j_dbs_info->prev_cpu_idle = cur_idle_time;
+
+		iowait_time = (unsigned int) cputime64_sub(cur_iowait_time,
+				j_dbs_info->prev_cpu_iowait);
+		j_dbs_info->prev_cpu_iowait = cur_iowait_time;
+
+		if (dbs_tuners_ins.ignore_nice) {
+			cputime64_t cur_nice;
+			unsigned long cur_nice_jiffies;
+
+			cur_nice = cputime64_sub(kstat_cpu(j).cpustat.nice,
+					 j_dbs_info->prev_cpu_nice);
+			/*
+			 * Assumption: nice time between sampling periods will
+			 * be less than 2^32 jiffies for 32 bit sys
+			 */
+			cur_nice_jiffies = (unsigned long)
+					cputime64_to_jiffies64(cur_nice);
+
+			j_dbs_info->prev_cpu_nice = kstat_cpu(j).cpustat.nice;
+			idle_time += jiffies_to_usecs(cur_nice_jiffies);
+		}
+
+		/*
+		 * For the purpose of ondemand, waiting for disk IO is an
+		 * indication that you're performance critical, and not that
+		 * the system is actually idle. So subtract the iowait time
+		 * from the cpu idle time.
+		 */
+
+		if (dbs_tuners_ins.io_is_busy && idle_time >= iowait_time)
+			idle_time -= iowait_time;
+
+		if (unlikely(!wall_time || wall_time < idle_time))
+			continue;
+
+		load = 100 * (wall_time - idle_time) / wall_time;
+
+		freq_avg = __cpufreq_driver_getavg(policy, j);
+		if (freq_avg <= 0)
+			freq_avg = policy->cur;
+
+		load_freq = load * freq_avg;
+		if (load_freq > max_load_freq)
+			max_load_freq = load_freq;
+	}
+
+	/* Check for frequency increase */
+	if (max_load_freq > dbs_tuners_ins.up_threshold * policy->cur) {
+		/* If switching to max speed, apply sampling_down_factor */
+		if (policy->cur < policy->max)
+			this_dbs_info->rate_mult =
+				dbs_tuners_ins.sampling_down_factor;
+		dbs_freq_increase(policy, policy->max);
+		return;
+	}
+
+	/* Check for frequency decrease */
+	/* if we cannot reduce the frequency anymore, break out early */
+	if (policy->cur == policy->min)
+		return;
+
+	/*
+	 * The optimal frequency is the frequency that is the lowest that
+	 * can support the current CPU usage without triggering the up
+	 * policy. To be safe, we focus 10 points under the threshold.
+	 */
+	if (max_load_freq <
+	    (dbs_tuners_ins.up_threshold - dbs_tuners_ins.down_differential) *
+	     policy->cur) {
+		unsigned int freq_next;
+		freq_next = max_load_freq /
+				(dbs_tuners_ins.up_threshold -
+				 dbs_tuners_ins.down_differential);
+
+		/* No longer fully busy, reset rate_mult */
+		this_dbs_info->rate_mult = 1;
+
+		if (freq_next < policy->min)
+			freq_next = policy->min;
+
+		if (!dbs_tuners_ins.powersave_bias) {
+			__cpufreq_driver_target(policy, freq_next,
+					CPUFREQ_RELATION_L);
+		} else {
+			int freq = powersave_bias_target(policy, freq_next,
+					CPUFREQ_RELATION_L);
+			__cpufreq_driver_target(policy, freq,
+				CPUFREQ_RELATION_L);
+		}
+	}
+}
+
+static void do_dbs_timer(struct work_struct *work)
+{
+	struct cpu_dbs_info_s *dbs_info =
+		container_of(work, struct cpu_dbs_info_s, work.work);
+	unsigned int cpu = dbs_info->cpu;
+	int sample_type = dbs_info->sample_type;
+
+	int delay;
+
+	mutex_lock(&dbs_info->timer_mutex);
+
+	/* Common NORMAL_SAMPLE setup */
+	dbs_info->sample_type = DBS_NORMAL_SAMPLE;
+	if (!dbs_tuners_ins.powersave_bias ||
+	    sample_type == DBS_NORMAL_SAMPLE) {
+		dbs_check_cpu(dbs_info);
+		if (dbs_info->freq_lo) {
+			/* Setup timer for SUB_SAMPLE */
+			dbs_info->sample_type = DBS_SUB_SAMPLE;
+			delay = dbs_info->freq_hi_jiffies;
+		} else {
+			/* We want all CPUs to do sampling nearly on
+			 * same jiffy
+			 */
+			delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate
+				* dbs_info->rate_mult);
+
+			if (num_online_cpus() > 1)
+				delay -= jiffies % delay;
+		}
+	} else {
+		__cpufreq_driver_target(dbs_info->cur_policy,
+			dbs_info->freq_lo, CPUFREQ_RELATION_H);
+		delay = dbs_info->freq_lo_jiffies;
+	}
+	schedule_delayed_work_on(cpu, &dbs_info->work, delay);
+	mutex_unlock(&dbs_info->timer_mutex);
+}
+
+static inline void dbs_timer_init(struct cpu_dbs_info_s *dbs_info)
+{
+	/* We want all CPUs to do sampling nearly on same jiffy */
+	int delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
+
+	if (num_online_cpus() > 1)
+		delay -= jiffies % delay;
+
+	dbs_info->sample_type = DBS_NORMAL_SAMPLE;
+	INIT_DELAYED_WORK_DEFERRABLE(&dbs_info->work, do_dbs_timer);
+	schedule_delayed_work_on(dbs_info->cpu, &dbs_info->work, delay);
+}
+
+static inline void dbs_timer_exit(struct cpu_dbs_info_s *dbs_info)
+{
+	cancel_delayed_work_sync(&dbs_info->work);
+}
+
+/*
+ * Not all CPUs want IO time to be accounted as busy; this dependson how
+ * efficient idling at a higher frequency/voltage is.
+ * Pavel Machek says this is not so for various generations of AMD and old
+ * Intel systems.
+ * Mike Chan (androidlcom) calis this is also not true for ARM.
+ * Because of this, whitelist specific known (series) of CPUs by default, and
+ * leave all others up to the user.
+ */
+static int should_io_be_busy(void)
+{
+#if defined(CONFIG_X86)
+	/*
+	 * For Intel, Core 2 (model 15) andl later have an efficient idle.
+	 */
+	if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL &&
+	    boot_cpu_data.x86 == 6 &&
+	    boot_cpu_data.x86_model >= 15)
+		return 1;
+#endif
+	return 0;
+}
+
+static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
+				   unsigned int event)
+{
+	unsigned int cpu = policy->cpu;
+	struct cpu_dbs_info_s *this_dbs_info;
+	unsigned int j;
+	int rc;
+
+	this_dbs_info = &per_cpu(od_cpu_dbs_info, cpu);
+
+	switch (event) {
+	case CPUFREQ_GOV_START:
+		if ((!cpu_online(cpu)) || (!policy->cur))
+			return -EINVAL;
+
+		mutex_lock(&dbs_mutex);
+
+		dbs_enable++;
+		for_each_cpu(j, policy->cpus) {
+			struct cpu_dbs_info_s *j_dbs_info;
+			j_dbs_info = &per_cpu(od_cpu_dbs_info, j);
+			j_dbs_info->cur_policy = policy;
+
+			j_dbs_info->prev_cpu_idle = get_cpu_idle_time(j,
+						&j_dbs_info->prev_cpu_wall);
+			if (dbs_tuners_ins.ignore_nice) {
+				j_dbs_info->prev_cpu_nice =
+						kstat_cpu(j).cpustat.nice;
+			}
+		}
+		this_dbs_info->cpu = cpu;
+		this_dbs_info->rate_mult = 1;
+		ondemand_powersave_bias_init_cpu(cpu);
+		/*
+		 * Start the timerschedule work, when this governor
+		 * is used for first time
+		 */
+		if (dbs_enable == 1) {
+			unsigned int latency;
+
+			rc = sysfs_create_group(cpufreq_global_kobject,
+						&dbs_attr_group);
+			if (rc) {
+				mutex_unlock(&dbs_mutex);
+				return rc;
+			}
+
+			/* policy latency is in nS. Convert it to uS first */
+			latency = policy->cpuinfo.transition_latency / 1000;
+			if (latency == 0)
+				latency = 1;
+			/* Bring kernel and HW constraints together */
+			min_sampling_rate = max(min_sampling_rate,
+					MIN_LATENCY_MULTIPLIER * latency);
+			dbs_tuners_ins.sampling_rate =
+				max(min_sampling_rate,
+				    latency * LATENCY_MULTIPLIER);
+			dbs_tuners_ins.io_is_busy = should_io_be_busy();
+		}
+		mutex_unlock(&dbs_mutex);
+
+		mutex_init(&this_dbs_info->timer_mutex);
+		dbs_timer_init(this_dbs_info);
+		break;
+
+	case CPUFREQ_GOV_STOP:
+		dbs_timer_exit(this_dbs_info);
+
+		mutex_lock(&dbs_mutex);
+		mutex_destroy(&this_dbs_info->timer_mutex);
+		dbs_enable--;
+		mutex_unlock(&dbs_mutex);
+		if (!dbs_enable)
+			sysfs_remove_group(cpufreq_global_kobject,
+					   &dbs_attr_group);
+
+		break;
+
+	case CPUFREQ_GOV_LIMITS:
+		mutex_lock(&this_dbs_info->timer_mutex);
+		if (policy->max < this_dbs_info->cur_policy->cur)
+			__cpufreq_driver_target(this_dbs_info->cur_policy,
+				policy->max, CPUFREQ_RELATION_H);
+		else if (policy->min > this_dbs_info->cur_policy->cur)
+			__cpufreq_driver_target(this_dbs_info->cur_policy,
+				policy->min, CPUFREQ_RELATION_L);
+		mutex_unlock(&this_dbs_info->timer_mutex);
+		break;
+	}
+	return 0;
+}
+
+static int __init cpufreq_gov_dbs_init(void)
+{
+	cputime64_t wall;
+	u64 idle_time;
+	int cpu = get_cpu();
+
+	idle_time = get_cpu_idle_time_us(cpu, &wall);
+	put_cpu();
+	if (idle_time != -1ULL) {
+		/* Idle micro accounting is supported. Use finer thresholds */
+		dbs_tuners_ins.up_threshold = MICRO_FREQUENCY_UP_THRESHOLD;
+		dbs_tuners_ins.down_differential =
+					MICRO_FREQUENCY_DOWN_DIFFERENTIAL;
+		/*
+		 * In no_hz/micro accounting case we set the minimum frequency
+		 * not depending on HZ, but fixed (very low). The deferred
+		 * timer might skip some samples if idle/sleeping as needed.
+		*/
+		min_sampling_rate = MICRO_FREQUENCY_MIN_SAMPLE_RATE;
+	} else {
+		/* For correct statistics, we need 10 ticks for each measure */
+		min_sampling_rate =
+			MIN_SAMPLING_RATE_RATIO * jiffies_to_usecs(10);
+	}
+
+	return cpufreq_register_governor(&cpufreq_gov_ondemand);
+}
+
+static void __exit cpufreq_gov_dbs_exit(void)
+{
+	cpufreq_unregister_governor(&cpufreq_gov_ondemand);
+}
+
+
+MODULE_AUTHOR("Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>");
+MODULE_AUTHOR("Alexey Starikovskiy <alexey.y.starikovskiy@intel.com>");
+MODULE_DESCRIPTION("'cpufreq_ondemand' - A dynamic cpufreq governor for "
+	"Low Latency Frequency Transition capable processors");
+MODULE_LICENSE("GPL");
+
+#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND
+fs_initcall(cpufreq_gov_dbs_init);
+#else
+module_init(cpufreq_gov_dbs_init);
+#endif
+module_exit(cpufreq_gov_dbs_exit);
diff --git a/drivers/cpufreq/cpufreq_performance.c b/drivers/cpufreq/cpufreq_performance.c
new file mode 100644
index 00000000..f13a8a9a
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_performance.c
@@ -0,0 +1,63 @@
+/*
+ *  linux/drivers/cpufreq/cpufreq_performance.c
+ *
+ *  Copyright (C) 2002 - 2003 Dominik Brodowski <linux@brodo.de>
+ *
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/cpufreq.h>
+#include <linux/init.h>
+
+
+static int cpufreq_governor_performance(struct cpufreq_policy *policy,
+					unsigned int event)
+{
+	switch (event) {
+	case CPUFREQ_GOV_START:
+	case CPUFREQ_GOV_LIMITS:
+		pr_debug("setting to %u kHz because of event %u\n",
+						policy->max, event);
+		__cpufreq_driver_target(policy, policy->max,
+						CPUFREQ_RELATION_H);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+#ifdef CONFIG_CPU_FREQ_GOV_PERFORMANCE_MODULE
+static
+#endif
+struct cpufreq_governor cpufreq_gov_performance = {
+	.name		= "performance",
+	.governor	= cpufreq_governor_performance,
+	.owner		= THIS_MODULE,
+};
+
+
+static int __init cpufreq_gov_performance_init(void)
+{
+	return cpufreq_register_governor(&cpufreq_gov_performance);
+}
+
+
+static void __exit cpufreq_gov_performance_exit(void)
+{
+	cpufreq_unregister_governor(&cpufreq_gov_performance);
+}
+
+
+MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("CPUfreq policy governor 'performance'");
+MODULE_LICENSE("GPL");
+
+fs_initcall(cpufreq_gov_performance_init);
+module_exit(cpufreq_gov_performance_exit);
diff --git a/drivers/cpufreq/cpufreq_powersave.c b/drivers/cpufreq/cpufreq_powersave.c
new file mode 100644
index 00000000..4c2eb512
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_powersave.c
@@ -0,0 +1,65 @@
+/*
+ *  linux/drivers/cpufreq/cpufreq_powersave.c
+ *
+ *  Copyright (C) 2002 - 2003 Dominik Brodowski <linux@brodo.de>
+ *
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/cpufreq.h>
+#include <linux/init.h>
+
+static int cpufreq_governor_powersave(struct cpufreq_policy *policy,
+					unsigned int event)
+{
+	switch (event) {
+	case CPUFREQ_GOV_START:
+	case CPUFREQ_GOV_LIMITS:
+		pr_debug("setting to %u kHz because of event %u\n",
+							policy->min, event);
+		__cpufreq_driver_target(policy, policy->min,
+						CPUFREQ_RELATION_L);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_POWERSAVE
+static
+#endif
+struct cpufreq_governor cpufreq_gov_powersave = {
+	.name		= "powersave",
+	.governor	= cpufreq_governor_powersave,
+	.owner		= THIS_MODULE,
+};
+
+static int __init cpufreq_gov_powersave_init(void)
+{
+	return cpufreq_register_governor(&cpufreq_gov_powersave);
+}
+
+
+static void __exit cpufreq_gov_powersave_exit(void)
+{
+	cpufreq_unregister_governor(&cpufreq_gov_powersave);
+}
+
+
+MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("CPUfreq policy governor 'powersave'");
+MODULE_LICENSE("GPL");
+
+#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_POWERSAVE
+fs_initcall(cpufreq_gov_powersave_init);
+#else
+module_init(cpufreq_gov_powersave_init);
+#endif
+module_exit(cpufreq_gov_powersave_exit);
diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
new file mode 100644
index 00000000..c315ec9d
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_stats.c
@@ -0,0 +1,427 @@
+/*
+ *  drivers/cpufreq/cpufreq_stats.c
+ *
+ *  Copyright (C) 2003-2004 Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>.
+ *  (C) 2004 Zou Nan hai <nanhai.zou@intel.com>.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/sysdev.h>
+#include <linux/cpu.h>
+#include <linux/sysfs.h>
+#include <linux/cpufreq.h>
+#include <linux/jiffies.h>
+#include <linux/percpu.h>
+#include <linux/kobject.h>
+#include <linux/spinlock.h>
+#include <linux/notifier.h>
+#include <asm/cputime.h>
+
+static spinlock_t cpufreq_stats_lock;
+
+#define CPUFREQ_STATDEVICE_ATTR(_name, _mode, _show) \
+static struct freq_attr _attr_##_name = {\
+	.attr = {.name = __stringify(_name), .mode = _mode, }, \
+	.show = _show,\
+};
+
+struct cpufreq_stats {
+	unsigned int cpu;
+	unsigned int total_trans;
+	unsigned long long  last_time;
+	unsigned int max_state;
+	unsigned int state_num;
+	unsigned int last_index;
+	cputime64_t *time_in_state;
+	unsigned int *freq_table;
+#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+	unsigned int *trans_table;
+#endif
+};
+
+static DEFINE_PER_CPU(struct cpufreq_stats *, cpufreq_stats_table);
+
+struct cpufreq_stats_attribute {
+	struct attribute attr;
+	ssize_t(*show) (struct cpufreq_stats *, char *);
+};
+
+static int cpufreq_stats_update(unsigned int cpu)
+{
+	struct cpufreq_stats *stat;
+	unsigned long long cur_time;
+
+	cur_time = get_jiffies_64();
+	spin_lock(&cpufreq_stats_lock);
+	stat = per_cpu(cpufreq_stats_table, cpu);
+	if (stat->time_in_state)
+		stat->time_in_state[stat->last_index] =
+			cputime64_add(stat->time_in_state[stat->last_index],
+				      cputime_sub(cur_time, stat->last_time));
+	stat->last_time = cur_time;
+	spin_unlock(&cpufreq_stats_lock);
+	return 0;
+}
+
+static ssize_t show_total_trans(struct cpufreq_policy *policy, char *buf)
+{
+	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
+	if (!stat)
+		return 0;
+	return sprintf(buf, "%d\n",
+			per_cpu(cpufreq_stats_table, stat->cpu)->total_trans);
+}
+
+static ssize_t show_time_in_state(struct cpufreq_policy *policy, char *buf)
+{
+	ssize_t len = 0;
+	int i;
+	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
+	if (!stat)
+		return 0;
+	cpufreq_stats_update(stat->cpu);
+	for (i = 0; i < stat->state_num; i++) {
+		len += sprintf(buf + len, "%u %llu\n", stat->freq_table[i],
+			(unsigned long long)
+			cputime64_to_clock_t(stat->time_in_state[i]));
+	}
+	return len;
+}
+
+#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+static ssize_t show_trans_table(struct cpufreq_policy *policy, char *buf)
+{
+	ssize_t len = 0;
+	int i, j;
+
+	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, policy->cpu);
+	if (!stat)
+		return 0;
+	cpufreq_stats_update(stat->cpu);
+	len += snprintf(buf + len, PAGE_SIZE - len, "   From  :    To\n");
+	len += snprintf(buf + len, PAGE_SIZE - len, "         : ");
+	for (i = 0; i < stat->state_num; i++) {
+		if (len >= PAGE_SIZE)
+			break;
+		len += snprintf(buf + len, PAGE_SIZE - len, "%9u ",
+				stat->freq_table[i]);
+	}
+	if (len >= PAGE_SIZE)
+		return PAGE_SIZE;
+
+	len += snprintf(buf + len, PAGE_SIZE - len, "\n");
+
+	for (i = 0; i < stat->state_num; i++) {
+		if (len >= PAGE_SIZE)
+			break;
+
+		len += snprintf(buf + len, PAGE_SIZE - len, "%9u: ",
+				stat->freq_table[i]);
+
+		for (j = 0; j < stat->state_num; j++)   {
+			if (len >= PAGE_SIZE)
+				break;
+			len += snprintf(buf + len, PAGE_SIZE - len, "%9u ",
+					stat->trans_table[i*stat->max_state+j]);
+		}
+		if (len >= PAGE_SIZE)
+			break;
+		len += snprintf(buf + len, PAGE_SIZE - len, "\n");
+	}
+	if (len >= PAGE_SIZE)
+		return PAGE_SIZE;
+	return len;
+}
+CPUFREQ_STATDEVICE_ATTR(trans_table, 0444, show_trans_table);
+#endif
+
+CPUFREQ_STATDEVICE_ATTR(total_trans, 0444, show_total_trans);
+CPUFREQ_STATDEVICE_ATTR(time_in_state, 0444, show_time_in_state);
+
+static struct attribute *default_attrs[] = {
+	&_attr_total_trans.attr,
+	&_attr_time_in_state.attr,
+#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+	&_attr_trans_table.attr,
+#endif
+	NULL
+};
+static struct attribute_group stats_attr_group = {
+	.attrs = default_attrs,
+	.name = "stats"
+};
+
+static int freq_table_get_index(struct cpufreq_stats *stat, unsigned int freq)
+{
+	int index;
+	for (index = 0; index < stat->max_state; index++)
+		if (stat->freq_table[index] == freq)
+			return index;
+	return -1;
+}
+
+/* should be called late in the CPU removal sequence so that the stats
+ * memory is still available in case someone tries to use it.
+ */
+static void cpufreq_stats_free_table(unsigned int cpu)
+{
+	struct cpufreq_stats *stat = per_cpu(cpufreq_stats_table, cpu);
+	if (stat) {
+		kfree(stat->time_in_state);
+		kfree(stat);
+	}
+	per_cpu(cpufreq_stats_table, cpu) = NULL;
+}
+
+/* must be called early in the CPU removal sequence (before
+ * cpufreq_remove_dev) so that policy is still valid.
+ */
+static void cpufreq_stats_free_sysfs(unsigned int cpu)
+{
+	struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
+	if (policy && policy->cpu == cpu)
+		sysfs_remove_group(&policy->kobj, &stats_attr_group);
+	if (policy)
+		cpufreq_cpu_put(policy);
+}
+
+static int cpufreq_stats_create_table(struct cpufreq_policy *policy,
+		struct cpufreq_frequency_table *table)
+{
+	unsigned int i, j, count = 0, ret = 0;
+	struct cpufreq_stats *stat;
+	struct cpufreq_policy *data;
+	unsigned int alloc_size;
+	unsigned int cpu = policy->cpu;
+	if (per_cpu(cpufreq_stats_table, cpu))
+		return -EBUSY;
+	stat = kzalloc(sizeof(struct cpufreq_stats), GFP_KERNEL);
+	if ((stat) == NULL)
+		return -ENOMEM;
+
+	data = cpufreq_cpu_get(cpu);
+	if (data == NULL) {
+		ret = -EINVAL;
+		goto error_get_fail;
+	}
+
+	ret = sysfs_create_group(&data->kobj, &stats_attr_group);
+	if (ret)
+		goto error_out;
+
+	stat->cpu = cpu;
+	per_cpu(cpufreq_stats_table, cpu) = stat;
+
+	for (i = 0; table[i].frequency != CPUFREQ_TABLE_END; i++) {
+		unsigned int freq = table[i].frequency;
+		if (freq == CPUFREQ_ENTRY_INVALID)
+			continue;
+		count++;
+	}
+
+	alloc_size = count * sizeof(int) + count * sizeof(cputime64_t);
+
+#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+	alloc_size += count * count * sizeof(int);
+#endif
+	stat->max_state = count;
+	stat->time_in_state = kzalloc(alloc_size, GFP_KERNEL);
+	if (!stat->time_in_state) {
+		ret = -ENOMEM;
+		goto error_out;
+	}
+	stat->freq_table = (unsigned int *)(stat->time_in_state + count);
+
+#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+	stat->trans_table = stat->freq_table + count;
+#endif
+	j = 0;
+	for (i = 0; table[i].frequency != CPUFREQ_TABLE_END; i++) {
+		unsigned int freq = table[i].frequency;
+		if (freq == CPUFREQ_ENTRY_INVALID)
+			continue;
+		if (freq_table_get_index(stat, freq) == -1)
+			stat->freq_table[j++] = freq;
+	}
+	stat->state_num = j;
+	spin_lock(&cpufreq_stats_lock);
+	stat->last_time = get_jiffies_64();
+	stat->last_index = freq_table_get_index(stat, policy->cur);
+	spin_unlock(&cpufreq_stats_lock);
+	cpufreq_cpu_put(data);
+	return 0;
+error_out:
+	cpufreq_cpu_put(data);
+error_get_fail:
+	kfree(stat);
+	per_cpu(cpufreq_stats_table, cpu) = NULL;
+	return ret;
+}
+
+static int cpufreq_stat_notifier_policy(struct notifier_block *nb,
+		unsigned long val, void *data)
+{
+	int ret;
+	struct cpufreq_policy *policy = data;
+	struct cpufreq_frequency_table *table;
+	unsigned int cpu = policy->cpu;
+	if (val != CPUFREQ_NOTIFY)
+		return 0;
+	table = cpufreq_frequency_get_table(cpu);
+	if (!table)
+		return 0;
+	ret = cpufreq_stats_create_table(policy, table);
+	if (ret)
+		return ret;
+	return 0;
+}
+
+static int cpufreq_stat_notifier_trans(struct notifier_block *nb,
+		unsigned long val, void *data)
+{
+	struct cpufreq_freqs *freq = data;
+	struct cpufreq_stats *stat;
+	int old_index, new_index;
+
+	if (val != CPUFREQ_POSTCHANGE)
+		return 0;
+
+	stat = per_cpu(cpufreq_stats_table, freq->cpu);
+	if (!stat)
+		return 0;
+
+	old_index = stat->last_index;
+	new_index = freq_table_get_index(stat, freq->new);
+
+	/* We can't do stat->time_in_state[-1]= .. */
+	if (old_index == -1 || new_index == -1)
+		return 0;
+
+	cpufreq_stats_update(freq->cpu);
+
+	if (old_index == new_index)
+		return 0;
+
+	spin_lock(&cpufreq_stats_lock);
+	stat->last_index = new_index;
+#ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+	stat->trans_table[old_index * stat->max_state + new_index]++;
+#endif
+	stat->total_trans++;
+	spin_unlock(&cpufreq_stats_lock);
+	return 0;
+}
+
+static int cpufreq_stats_create_table_cpu(unsigned int cpu)
+{
+	struct cpufreq_policy *policy;
+	struct cpufreq_frequency_table *table;
+	int ret = -ENODEV;
+
+	policy = cpufreq_cpu_get(cpu);
+	if (!policy)
+		return -ENODEV;
+
+	table = cpufreq_frequency_get_table(cpu);
+	if (!table)
+		goto out;
+
+	ret = cpufreq_stats_create_table(policy, table);
+
+out:
+	cpufreq_cpu_put(policy);
+	return ret;
+}
+
+static int __cpuinit cpufreq_stat_cpu_callback(struct notifier_block *nfb,
+					       unsigned long action,
+					       void *hcpu)
+{
+	unsigned int cpu = (unsigned long)hcpu;
+
+	switch (action) {
+	case CPU_ONLINE:
+	case CPU_ONLINE_FROZEN:
+		cpufreq_update_policy(cpu);
+		break;
+	case CPU_DOWN_PREPARE:
+		cpufreq_stats_free_sysfs(cpu);
+		break;
+	case CPU_DEAD:
+	case CPU_DEAD_FROZEN:
+		cpufreq_stats_free_table(cpu);
+		break;
+	case CPU_DOWN_FAILED:
+	case CPU_DOWN_FAILED_FROZEN:
+		cpufreq_stats_create_table_cpu(cpu);
+		break;
+	}
+	return NOTIFY_OK;
+}
+
+/* priority=1 so this will get called before cpufreq_remove_dev */
+static struct notifier_block cpufreq_stat_cpu_notifier __refdata = {
+	.notifier_call = cpufreq_stat_cpu_callback,
+	.priority = 1,
+};
+
+static struct notifier_block notifier_policy_block = {
+	.notifier_call = cpufreq_stat_notifier_policy
+};
+
+static struct notifier_block notifier_trans_block = {
+	.notifier_call = cpufreq_stat_notifier_trans
+};
+
+static int __init cpufreq_stats_init(void)
+{
+	int ret;
+	unsigned int cpu;
+
+	spin_lock_init(&cpufreq_stats_lock);
+	ret = cpufreq_register_notifier(&notifier_policy_block,
+				CPUFREQ_POLICY_NOTIFIER);
+	if (ret)
+		return ret;
+
+	ret = cpufreq_register_notifier(&notifier_trans_block,
+				CPUFREQ_TRANSITION_NOTIFIER);
+	if (ret) {
+		cpufreq_unregister_notifier(&notifier_policy_block,
+				CPUFREQ_POLICY_NOTIFIER);
+		return ret;
+	}
+
+	register_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
+	for_each_online_cpu(cpu) {
+		cpufreq_update_policy(cpu);
+	}
+	return 0;
+}
+static void __exit cpufreq_stats_exit(void)
+{
+	unsigned int cpu;
+
+	cpufreq_unregister_notifier(&notifier_policy_block,
+			CPUFREQ_POLICY_NOTIFIER);
+	cpufreq_unregister_notifier(&notifier_trans_block,
+			CPUFREQ_TRANSITION_NOTIFIER);
+	unregister_hotcpu_notifier(&cpufreq_stat_cpu_notifier);
+	for_each_online_cpu(cpu) {
+		cpufreq_stats_free_table(cpu);
+		cpufreq_stats_free_sysfs(cpu);
+	}
+}
+
+MODULE_AUTHOR("Zou Nan hai <nanhai.zou@intel.com>");
+MODULE_DESCRIPTION("'cpufreq_stats' - A driver to export cpufreq stats "
+				"through sysfs filesystem");
+MODULE_LICENSE("GPL");
+
+module_init(cpufreq_stats_init);
+module_exit(cpufreq_stats_exit);
diff --git a/drivers/cpufreq/cpufreq_userspace.c b/drivers/cpufreq/cpufreq_userspace.c
new file mode 100644
index 00000000..f2310159
--- /dev/null
+++ b/drivers/cpufreq/cpufreq_userspace.c
@@ -0,0 +1,220 @@
+
+/*
+ *  linux/drivers/cpufreq/cpufreq_userspace.c
+ *
+ *  Copyright (C)  2001 Russell King
+ *            (C)  2002 - 2004 Dominik Brodowski <linux@brodo.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/smp.h>
+#include <linux/init.h>
+#include <linux/spinlock.h>
+#include <linux/interrupt.h>
+#include <linux/cpufreq.h>
+#include <linux/cpu.h>
+#include <linux/types.h>
+#include <linux/fs.h>
+#include <linux/sysfs.h>
+#include <linux/mutex.h>
+
+/**
+ * A few values needed by the userspace governor
+ */
+static DEFINE_PER_CPU(unsigned int, cpu_max_freq);
+static DEFINE_PER_CPU(unsigned int, cpu_min_freq);
+static DEFINE_PER_CPU(unsigned int, cpu_cur_freq); /* current CPU freq */
+static DEFINE_PER_CPU(unsigned int, cpu_set_freq); /* CPU freq desired by
+							userspace */
+static DEFINE_PER_CPU(unsigned int, cpu_is_managed);
+
+static DEFINE_MUTEX(userspace_mutex);
+static int cpus_using_userspace_governor;
+
+/* keep track of frequency transitions */
+static int
+userspace_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
+	void *data)
+{
+	struct cpufreq_freqs *freq = data;
+
+	if (!per_cpu(cpu_is_managed, freq->cpu))
+		return 0;
+
+	pr_debug("saving cpu_cur_freq of cpu %u to be %u kHz\n",
+			freq->cpu, freq->new);
+	per_cpu(cpu_cur_freq, freq->cpu) = freq->new;
+
+	return 0;
+}
+
+static struct notifier_block userspace_cpufreq_notifier_block = {
+	.notifier_call  = userspace_cpufreq_notifier
+};
+
+
+/**
+ * cpufreq_set - set the CPU frequency
+ * @policy: pointer to policy struct where freq is being set
+ * @freq: target frequency in kHz
+ *
+ * Sets the CPU frequency to freq.
+ */
+static int cpufreq_set(struct cpufreq_policy *policy, unsigned int freq)
+{
+	int ret = -EINVAL;
+
+	pr_debug("cpufreq_set for cpu %u, freq %u kHz\n", policy->cpu, freq);
+
+	mutex_lock(&userspace_mutex);
+	if (!per_cpu(cpu_is_managed, policy->cpu))
+		goto err;
+
+	per_cpu(cpu_set_freq, policy->cpu) = freq;
+
+	if (freq < per_cpu(cpu_min_freq, policy->cpu))
+		freq = per_cpu(cpu_min_freq, policy->cpu);
+	if (freq > per_cpu(cpu_max_freq, policy->cpu))
+		freq = per_cpu(cpu_max_freq, policy->cpu);
+
+	/*
+	 * We're safe from concurrent calls to ->target() here
+	 * as we hold the userspace_mutex lock. If we were calling
+	 * cpufreq_driver_target, a deadlock situation might occur:
+	 * A: cpufreq_set (lock userspace_mutex) ->
+	 *      cpufreq_driver_target(lock policy->lock)
+	 * B: cpufreq_set_policy(lock policy->lock) ->
+	 *      __cpufreq_governor ->
+	 *         cpufreq_governor_userspace (lock userspace_mutex)
+	 */
+	ret = __cpufreq_driver_target(policy, freq, CPUFREQ_RELATION_L);
+
+ err:
+	mutex_unlock(&userspace_mutex);
+	return ret;
+}
+
+
+static ssize_t show_speed(struct cpufreq_policy *policy, char *buf)
+{
+	return sprintf(buf, "%u\n", per_cpu(cpu_cur_freq, policy->cpu));
+}
+
+static int cpufreq_governor_userspace(struct cpufreq_policy *policy,
+				   unsigned int event)
+{
+	unsigned int cpu = policy->cpu;
+	int rc = 0;
+
+	switch (event) {
+	case CPUFREQ_GOV_START:
+		if (!cpu_online(cpu))
+			return -EINVAL;
+		BUG_ON(!policy->cur);
+		mutex_lock(&userspace_mutex);
+
+		if (cpus_using_userspace_governor == 0) {
+			cpufreq_register_notifier(
+					&userspace_cpufreq_notifier_block,
+					CPUFREQ_TRANSITION_NOTIFIER);
+		}
+		cpus_using_userspace_governor++;
+
+		per_cpu(cpu_is_managed, cpu) = 1;
+		per_cpu(cpu_min_freq, cpu) = policy->min;
+		per_cpu(cpu_max_freq, cpu) = policy->max;
+		per_cpu(cpu_cur_freq, cpu) = policy->cur;
+		per_cpu(cpu_set_freq, cpu) = policy->cur;
+		pr_debug("managing cpu %u started "
+			"(%u - %u kHz, currently %u kHz)\n",
+				cpu,
+				per_cpu(cpu_min_freq, cpu),
+				per_cpu(cpu_max_freq, cpu),
+				per_cpu(cpu_cur_freq, cpu));
+
+		mutex_unlock(&userspace_mutex);
+		break;
+	case CPUFREQ_GOV_STOP:
+		mutex_lock(&userspace_mutex);
+		cpus_using_userspace_governor--;
+		if (cpus_using_userspace_governor == 0) {
+			cpufreq_unregister_notifier(
+					&userspace_cpufreq_notifier_block,
+					CPUFREQ_TRANSITION_NOTIFIER);
+		}
+
+		per_cpu(cpu_is_managed, cpu) = 0;
+		per_cpu(cpu_min_freq, cpu) = 0;
+		per_cpu(cpu_max_freq, cpu) = 0;
+		per_cpu(cpu_set_freq, cpu) = 0;
+		pr_debug("managing cpu %u stopped\n", cpu);
+		mutex_unlock(&userspace_mutex);
+		break;
+	case CPUFREQ_GOV_LIMITS:
+		mutex_lock(&userspace_mutex);
+		pr_debug("limit event for cpu %u: %u - %u kHz, "
+			"currently %u kHz, last set to %u kHz\n",
+			cpu, policy->min, policy->max,
+			per_cpu(cpu_cur_freq, cpu),
+			per_cpu(cpu_set_freq, cpu));
+		if (policy->max < per_cpu(cpu_set_freq, cpu)) {
+			__cpufreq_driver_target(policy, policy->max,
+						CPUFREQ_RELATION_H);
+		} else if (policy->min > per_cpu(cpu_set_freq, cpu)) {
+			__cpufreq_driver_target(policy, policy->min,
+						CPUFREQ_RELATION_L);
+		} else {
+			__cpufreq_driver_target(policy,
+						per_cpu(cpu_set_freq, cpu),
+						CPUFREQ_RELATION_L);
+		}
+		per_cpu(cpu_min_freq, cpu) = policy->min;
+		per_cpu(cpu_max_freq, cpu) = policy->max;
+		per_cpu(cpu_cur_freq, cpu) = policy->cur;
+		mutex_unlock(&userspace_mutex);
+		break;
+	}
+	return rc;
+}
+
+
+#ifndef CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE
+static
+#endif
+struct cpufreq_governor cpufreq_gov_userspace = {
+	.name		= "userspace",
+	.governor	= cpufreq_governor_userspace,
+	.store_setspeed	= cpufreq_set,
+	.show_setspeed	= show_speed,
+	.owner		= THIS_MODULE,
+};
+
+static int __init cpufreq_gov_userspace_init(void)
+{
+	return cpufreq_register_governor(&cpufreq_gov_userspace);
+}
+
+
+static void __exit cpufreq_gov_userspace_exit(void)
+{
+	cpufreq_unregister_governor(&cpufreq_gov_userspace);
+}
+
+
+MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>, "
+		"Russell King <rmk@arm.linux.org.uk>");
+MODULE_DESCRIPTION("CPUfreq policy governor 'userspace'");
+MODULE_LICENSE("GPL");
+
+#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE
+fs_initcall(cpufreq_gov_userspace_init);
+#else
+module_init(cpufreq_gov_userspace_init);
+#endif
+module_exit(cpufreq_gov_userspace_exit);
diff --git a/drivers/cpufreq/db8500-cpufreq.c b/drivers/cpufreq/db8500-cpufreq.c
new file mode 100644
index 00000000..d90456a8
--- /dev/null
+++ b/drivers/cpufreq/db8500-cpufreq.c
@@ -0,0 +1,169 @@
+/*
+ * Copyright (C) STMicroelectronics 2009
+ * Copyright (C) ST-Ericsson SA 2010
+ *
+ * License Terms: GNU General Public License v2
+ * Author: Sundar Iyer <sundar.iyer@stericsson.com>
+ * Author: Martin Persson <martin.persson@stericsson.com>
+ * Author: Jonas Aaberg <jonas.aberg@stericsson.com>
+ *
+ */
+#include <linux/kernel.h>
+#include <linux/cpufreq.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/mfd/db8500-prcmu.h>
+#include <mach/id.h>
+
+static struct cpufreq_frequency_table freq_table[] = {
+	[0] = {
+		.index = 0,
+		.frequency = 300000,
+	},
+	[1] = {
+		.index = 1,
+		.frequency = 600000,
+	},
+	[2] = {
+		/* Used for MAX_OPP, if available */
+		.index = 2,
+		.frequency = CPUFREQ_TABLE_END,
+	},
+	[3] = {
+		.index = 3,
+		.frequency = CPUFREQ_TABLE_END,
+	},
+};
+
+static enum arm_opp idx2opp[] = {
+	ARM_50_OPP,
+	ARM_100_OPP,
+	ARM_MAX_OPP
+};
+
+static struct freq_attr *db8500_cpufreq_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static int db8500_cpufreq_verify_speed(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, freq_table);
+}
+
+static int db8500_cpufreq_target(struct cpufreq_policy *policy,
+				unsigned int target_freq,
+				unsigned int relation)
+{
+	struct cpufreq_freqs freqs;
+	unsigned int idx;
+
+	/* scale the target frequency to one of the extremes supported */
+	if (target_freq < policy->cpuinfo.min_freq)
+		target_freq = policy->cpuinfo.min_freq;
+	if (target_freq > policy->cpuinfo.max_freq)
+		target_freq = policy->cpuinfo.max_freq;
+
+	/* Lookup the next frequency */
+	if (cpufreq_frequency_table_target
+	    (policy, freq_table, target_freq, relation, &idx)) {
+		return -EINVAL;
+	}
+
+	freqs.old = policy->cur;
+	freqs.new = freq_table[idx].frequency;
+	freqs.cpu = policy->cpu;
+
+	if (freqs.old == freqs.new)
+		return 0;
+
+	/* pre-change notification */
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	/* request the PRCM unit for opp change */
+	if (prcmu_set_arm_opp(idx2opp[idx])) {
+		pr_err("db8500-cpufreq:  Failed to set OPP level\n");
+		return -EINVAL;
+	}
+
+	/* post change notification */
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+
+	return 0;
+}
+
+static unsigned int db8500_cpufreq_getspeed(unsigned int cpu)
+{
+	int i;
+	/* request the prcm to get the current ARM opp */
+	for (i = 0; prcmu_get_arm_opp() != idx2opp[i]; i++)
+		;
+	return freq_table[i].frequency;
+}
+
+static int __cpuinit db8500_cpufreq_init(struct cpufreq_policy *policy)
+{
+	int res;
+	int i;
+
+	BUILD_BUG_ON(ARRAY_SIZE(idx2opp) + 1 != ARRAY_SIZE(freq_table));
+
+	if (cpu_is_u8500v2() && !prcmu_is_u8400()) {
+		freq_table[0].frequency = 400000;
+		freq_table[1].frequency = 800000;
+		if (prcmu_has_arm_maxopp())
+			freq_table[2].frequency = 1000000;
+	}
+
+	/* get policy fields based on the table */
+	res = cpufreq_frequency_table_cpuinfo(policy, freq_table);
+	if (!res)
+		cpufreq_frequency_table_get_attr(freq_table, policy->cpu);
+	else {
+		pr_err("db8500-cpufreq : Failed to read policy table\n");
+		return res;
+	}
+
+	policy->min = policy->cpuinfo.min_freq;
+	policy->max = policy->cpuinfo.max_freq;
+	policy->cur = db8500_cpufreq_getspeed(policy->cpu);
+
+	for (i = 0; freq_table[i].frequency != policy->cur; i++)
+		;
+
+	policy->governor = CPUFREQ_DEFAULT_GOVERNOR;
+
+	/*
+	 * FIXME : Need to take time measurement across the target()
+	 *	   function with no/some/all drivers in the notification
+	 *	   list.
+	 */
+	policy->cpuinfo.transition_latency = 20 * 1000; /* in ns */
+
+	/* policy sharing between dual CPUs */
+	cpumask_copy(policy->cpus, &cpu_present_map);
+
+	policy->shared_type = CPUFREQ_SHARED_TYPE_ALL;
+
+	return 0;
+}
+
+static struct cpufreq_driver db8500_cpufreq_driver = {
+	.flags  = CPUFREQ_STICKY,
+	.verify = db8500_cpufreq_verify_speed,
+	.target = db8500_cpufreq_target,
+	.get    = db8500_cpufreq_getspeed,
+	.init   = db8500_cpufreq_init,
+	.name   = "DB8500",
+	.attr   = db8500_cpufreq_attr,
+};
+
+static int __init db8500_cpufreq_register(void)
+{
+	if (!cpu_is_u8500v20_or_later())
+		return -ENODEV;
+
+	pr_info("cpufreq for DB8500 started\n");
+	return cpufreq_register_driver(&db8500_cpufreq_driver);
+}
+device_initcall(db8500_cpufreq_register);
diff --git a/drivers/cpufreq/e_powersaver.c b/drivers/cpufreq/e_powersaver.c
new file mode 100644
index 00000000..35a257dd
--- /dev/null
+++ b/drivers/cpufreq/e_powersaver.c
@@ -0,0 +1,367 @@
+/*
+ *  Based on documentation provided by Dave Jones. Thanks!
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/ioport.h>
+#include <linux/slab.h>
+#include <linux/timex.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+
+#include <asm/msr.h>
+#include <asm/tsc.h>
+
+#define EPS_BRAND_C7M	0
+#define EPS_BRAND_C7	1
+#define EPS_BRAND_EDEN	2
+#define EPS_BRAND_C3	3
+#define EPS_BRAND_C7D	4
+
+struct eps_cpu_data {
+	u32 fsb;
+	struct cpufreq_frequency_table freq_table[];
+};
+
+static struct eps_cpu_data *eps_cpu[NR_CPUS];
+
+
+static unsigned int eps_get(unsigned int cpu)
+{
+	struct eps_cpu_data *centaur;
+	u32 lo, hi;
+
+	if (cpu)
+		return 0;
+	centaur = eps_cpu[cpu];
+	if (centaur == NULL)
+		return 0;
+
+	/* Return current frequency */
+	rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+	return centaur->fsb * ((lo >> 8) & 0xff);
+}
+
+static int eps_set_state(struct eps_cpu_data *centaur,
+			 unsigned int cpu,
+			 u32 dest_state)
+{
+	struct cpufreq_freqs freqs;
+	u32 lo, hi;
+	int err = 0;
+	int i;
+
+	freqs.old = eps_get(cpu);
+	freqs.new = centaur->fsb * ((dest_state >> 8) & 0xff);
+	freqs.cpu = cpu;
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	/* Wait while CPU is busy */
+	rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+	i = 0;
+	while (lo & ((1 << 16) | (1 << 17))) {
+		udelay(16);
+		rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+		i++;
+		if (unlikely(i > 64)) {
+			err = -ENODEV;
+			goto postchange;
+		}
+	}
+	/* Set new multiplier and voltage */
+	wrmsr(MSR_IA32_PERF_CTL, dest_state & 0xffff, 0);
+	/* Wait until transition end */
+	i = 0;
+	do {
+		udelay(16);
+		rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+		i++;
+		if (unlikely(i > 64)) {
+			err = -ENODEV;
+			goto postchange;
+		}
+	} while (lo & ((1 << 16) | (1 << 17)));
+
+	/* Return current frequency */
+postchange:
+	rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+	freqs.new = centaur->fsb * ((lo >> 8) & 0xff);
+
+#ifdef DEBUG
+	{
+	u8 current_multiplier, current_voltage;
+
+	/* Print voltage and multiplier */
+	rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+	current_voltage = lo & 0xff;
+	printk(KERN_INFO "eps: Current voltage = %dmV\n",
+		current_voltage * 16 + 700);
+	current_multiplier = (lo >> 8) & 0xff;
+	printk(KERN_INFO "eps: Current multiplier = %d\n",
+		current_multiplier);
+	}
+#endif
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	return err;
+}
+
+static int eps_target(struct cpufreq_policy *policy,
+			       unsigned int target_freq,
+			       unsigned int relation)
+{
+	struct eps_cpu_data *centaur;
+	unsigned int newstate = 0;
+	unsigned int cpu = policy->cpu;
+	unsigned int dest_state;
+	int ret;
+
+	if (unlikely(eps_cpu[cpu] == NULL))
+		return -ENODEV;
+	centaur = eps_cpu[cpu];
+
+	if (unlikely(cpufreq_frequency_table_target(policy,
+			&eps_cpu[cpu]->freq_table[0],
+			target_freq,
+			relation,
+			&newstate))) {
+		return -EINVAL;
+	}
+
+	/* Make frequency transition */
+	dest_state = centaur->freq_table[newstate].index & 0xffff;
+	ret = eps_set_state(centaur, cpu, dest_state);
+	if (ret)
+		printk(KERN_ERR "eps: Timeout!\n");
+	return ret;
+}
+
+static int eps_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy,
+			&eps_cpu[policy->cpu]->freq_table[0]);
+}
+
+static int eps_cpu_init(struct cpufreq_policy *policy)
+{
+	unsigned int i;
+	u32 lo, hi;
+	u64 val;
+	u8 current_multiplier, current_voltage;
+	u8 max_multiplier, max_voltage;
+	u8 min_multiplier, min_voltage;
+	u8 brand = 0;
+	u32 fsb;
+	struct eps_cpu_data *centaur;
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	struct cpufreq_frequency_table *f_table;
+	int k, step, voltage;
+	int ret;
+	int states;
+
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	/* Check brand */
+	printk(KERN_INFO "eps: Detected VIA ");
+
+	switch (c->x86_model) {
+	case 10:
+		rdmsr(0x1153, lo, hi);
+		brand = (((lo >> 2) ^ lo) >> 18) & 3;
+		printk(KERN_CONT "Model A ");
+		break;
+	case 13:
+		rdmsr(0x1154, lo, hi);
+		brand = (((lo >> 4) ^ (lo >> 2))) & 0x000000ff;
+		printk(KERN_CONT "Model D ");
+		break;
+	}
+
+	switch (brand) {
+	case EPS_BRAND_C7M:
+		printk(KERN_CONT "C7-M\n");
+		break;
+	case EPS_BRAND_C7:
+		printk(KERN_CONT "C7\n");
+		break;
+	case EPS_BRAND_EDEN:
+		printk(KERN_CONT "Eden\n");
+		break;
+	case EPS_BRAND_C7D:
+		printk(KERN_CONT "C7-D\n");
+		break;
+	case EPS_BRAND_C3:
+		printk(KERN_CONT "C3\n");
+		return -ENODEV;
+		break;
+	}
+	/* Enable Enhanced PowerSaver */
+	rdmsrl(MSR_IA32_MISC_ENABLE, val);
+	if (!(val & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
+		val |= MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP;
+		wrmsrl(MSR_IA32_MISC_ENABLE, val);
+		/* Can be locked at 0 */
+		rdmsrl(MSR_IA32_MISC_ENABLE, val);
+		if (!(val & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
+			printk(KERN_INFO "eps: Can't enable Enhanced PowerSaver\n");
+			return -ENODEV;
+		}
+	}
+
+	/* Print voltage and multiplier */
+	rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+	current_voltage = lo & 0xff;
+	printk(KERN_INFO "eps: Current voltage = %dmV\n",
+			current_voltage * 16 + 700);
+	current_multiplier = (lo >> 8) & 0xff;
+	printk(KERN_INFO "eps: Current multiplier = %d\n", current_multiplier);
+
+	/* Print limits */
+	max_voltage = hi & 0xff;
+	printk(KERN_INFO "eps: Highest voltage = %dmV\n",
+			max_voltage * 16 + 700);
+	max_multiplier = (hi >> 8) & 0xff;
+	printk(KERN_INFO "eps: Highest multiplier = %d\n", max_multiplier);
+	min_voltage = (hi >> 16) & 0xff;
+	printk(KERN_INFO "eps: Lowest voltage = %dmV\n",
+			min_voltage * 16 + 700);
+	min_multiplier = (hi >> 24) & 0xff;
+	printk(KERN_INFO "eps: Lowest multiplier = %d\n", min_multiplier);
+
+	/* Sanity checks */
+	if (current_multiplier == 0 || max_multiplier == 0
+	    || min_multiplier == 0)
+		return -EINVAL;
+	if (current_multiplier > max_multiplier
+	    || max_multiplier <= min_multiplier)
+		return -EINVAL;
+	if (current_voltage > 0x1f || max_voltage > 0x1f)
+		return -EINVAL;
+	if (max_voltage < min_voltage)
+		return -EINVAL;
+
+	/* Calc FSB speed */
+	fsb = cpu_khz / current_multiplier;
+	/* Calc number of p-states supported */
+	if (brand == EPS_BRAND_C7M)
+		states = max_multiplier - min_multiplier + 1;
+	else
+		states = 2;
+
+	/* Allocate private data and frequency table for current cpu */
+	centaur = kzalloc(sizeof(struct eps_cpu_data)
+		    + (states + 1) * sizeof(struct cpufreq_frequency_table),
+		    GFP_KERNEL);
+	if (!centaur)
+		return -ENOMEM;
+	eps_cpu[0] = centaur;
+
+	/* Copy basic values */
+	centaur->fsb = fsb;
+
+	/* Fill frequency and MSR value table */
+	f_table = &centaur->freq_table[0];
+	if (brand != EPS_BRAND_C7M) {
+		f_table[0].frequency = fsb * min_multiplier;
+		f_table[0].index = (min_multiplier << 8) | min_voltage;
+		f_table[1].frequency = fsb * max_multiplier;
+		f_table[1].index = (max_multiplier << 8) | max_voltage;
+		f_table[2].frequency = CPUFREQ_TABLE_END;
+	} else {
+		k = 0;
+		step = ((max_voltage - min_voltage) * 256)
+			/ (max_multiplier - min_multiplier);
+		for (i = min_multiplier; i <= max_multiplier; i++) {
+			voltage = (k * step) / 256 + min_voltage;
+			f_table[k].frequency = fsb * i;
+			f_table[k].index = (i << 8) | voltage;
+			k++;
+		}
+		f_table[k].frequency = CPUFREQ_TABLE_END;
+	}
+
+	policy->cpuinfo.transition_latency = 140000; /* 844mV -> 700mV in ns */
+	policy->cur = fsb * current_multiplier;
+
+	ret = cpufreq_frequency_table_cpuinfo(policy, &centaur->freq_table[0]);
+	if (ret) {
+		kfree(centaur);
+		return ret;
+	}
+
+	cpufreq_frequency_table_get_attr(&centaur->freq_table[0], policy->cpu);
+	return 0;
+}
+
+static int eps_cpu_exit(struct cpufreq_policy *policy)
+{
+	unsigned int cpu = policy->cpu;
+	struct eps_cpu_data *centaur;
+	u32 lo, hi;
+
+	if (eps_cpu[cpu] == NULL)
+		return -ENODEV;
+	centaur = eps_cpu[cpu];
+
+	/* Get max frequency */
+	rdmsr(MSR_IA32_PERF_STATUS, lo, hi);
+	/* Set max frequency */
+	eps_set_state(centaur, cpu, hi & 0xffff);
+	/* Bye */
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	kfree(eps_cpu[cpu]);
+	eps_cpu[cpu] = NULL;
+	return 0;
+}
+
+static struct freq_attr *eps_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver eps_driver = {
+	.verify		= eps_verify,
+	.target		= eps_target,
+	.init		= eps_cpu_init,
+	.exit		= eps_cpu_exit,
+	.get		= eps_get,
+	.name		= "e_powersaver",
+	.owner		= THIS_MODULE,
+	.attr		= eps_attr,
+};
+
+static int __init eps_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	/* This driver will work only on Centaur C7 processors with
+	 * Enhanced SpeedStep/PowerSaver registers */
+	if (c->x86_vendor != X86_VENDOR_CENTAUR
+	    || c->x86 != 6 || c->x86_model < 10)
+		return -ENODEV;
+	if (!cpu_has(c, X86_FEATURE_EST))
+		return -ENODEV;
+
+	if (cpufreq_register_driver(&eps_driver))
+		return -EINVAL;
+	return 0;
+}
+
+static void __exit eps_exit(void)
+{
+	cpufreq_unregister_driver(&eps_driver);
+}
+
+MODULE_AUTHOR("Rafal Bilski <rafalbilski@interia.pl>");
+MODULE_DESCRIPTION("Enhanced PowerSaver driver for VIA C7 CPU's.");
+MODULE_LICENSE("GPL");
+
+module_init(eps_init);
+module_exit(eps_exit);
diff --git a/drivers/cpufreq/elanfreq.c b/drivers/cpufreq/elanfreq.c
new file mode 100644
index 00000000..c587db47
--- /dev/null
+++ b/drivers/cpufreq/elanfreq.c
@@ -0,0 +1,309 @@
+/*
+ *	elanfreq:	cpufreq driver for the AMD ELAN family
+ *
+ *	(c) Copyright 2002 Robert Schwebel <r.schwebel@pengutronix.de>
+ *
+ *	Parts of this code are (c) Sven Geggus <sven@geggus.net>
+ *
+ *      All Rights Reserved.
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ *
+ *	2002-02-13: - initial revision for 2.4.18-pre9 by Robert Schwebel
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <linux/delay.h>
+#include <linux/cpufreq.h>
+
+#include <asm/msr.h>
+#include <linux/timex.h>
+#include <linux/io.h>
+
+#define REG_CSCIR 0x22		/* Chip Setup and Control Index Register    */
+#define REG_CSCDR 0x23		/* Chip Setup and Control Data  Register    */
+
+/* Module parameter */
+static int max_freq;
+
+struct s_elan_multiplier {
+	int clock;		/* frequency in kHz                         */
+	int val40h;		/* PMU Force Mode register                  */
+	int val80h;		/* CPU Clock Speed Register                 */
+};
+
+/*
+ * It is important that the frequencies
+ * are listed in ascending order here!
+ */
+static struct s_elan_multiplier elan_multiplier[] = {
+	{1000,	0x02,	0x18},
+	{2000,	0x02,	0x10},
+	{4000,	0x02,	0x08},
+	{8000,	0x00,	0x00},
+	{16000,	0x00,	0x02},
+	{33000,	0x00,	0x04},
+	{66000,	0x01,	0x04},
+	{99000,	0x01,	0x05}
+};
+
+static struct cpufreq_frequency_table elanfreq_table[] = {
+	{0,	1000},
+	{1,	2000},
+	{2,	4000},
+	{3,	8000},
+	{4,	16000},
+	{5,	33000},
+	{6,	66000},
+	{7,	99000},
+	{0,	CPUFREQ_TABLE_END},
+};
+
+
+/**
+ *	elanfreq_get_cpu_frequency: determine current cpu speed
+ *
+ *	Finds out at which frequency the CPU of the Elan SOC runs
+ *	at the moment. Frequencies from 1 to 33 MHz are generated
+ *	the normal way, 66 and 99 MHz are called "Hyperspeed Mode"
+ *	and have the rest of the chip running with 33 MHz.
+ */
+
+static unsigned int elanfreq_get_cpu_frequency(unsigned int cpu)
+{
+	u8 clockspeed_reg;    /* Clock Speed Register */
+
+	local_irq_disable();
+	outb_p(0x80, REG_CSCIR);
+	clockspeed_reg = inb_p(REG_CSCDR);
+	local_irq_enable();
+
+	if ((clockspeed_reg & 0xE0) == 0xE0)
+		return 0;
+
+	/* Are we in CPU clock multiplied mode (66/99 MHz)? */
+	if ((clockspeed_reg & 0xE0) == 0xC0) {
+		if ((clockspeed_reg & 0x01) == 0)
+			return 66000;
+		else
+			return 99000;
+	}
+
+	/* 33 MHz is not 32 MHz... */
+	if ((clockspeed_reg & 0xE0) == 0xA0)
+		return 33000;
+
+	return (1<<((clockspeed_reg & 0xE0) >> 5)) * 1000;
+}
+
+
+/**
+ *	elanfreq_set_cpu_frequency: Change the CPU core frequency
+ *	@cpu: cpu number
+ *	@freq: frequency in kHz
+ *
+ *	This function takes a frequency value and changes the CPU frequency
+ *	according to this. Note that the frequency has to be checked by
+ *	elanfreq_validatespeed() for correctness!
+ *
+ *	There is no return value.
+ */
+
+static void elanfreq_set_cpu_state(unsigned int state)
+{
+	struct cpufreq_freqs    freqs;
+
+	freqs.old = elanfreq_get_cpu_frequency(0);
+	freqs.new = elan_multiplier[state].clock;
+	freqs.cpu = 0; /* elanfreq.c is UP only driver */
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	printk(KERN_INFO "elanfreq: attempting to set frequency to %i kHz\n",
+			elan_multiplier[state].clock);
+
+
+	/*
+	 * Access to the Elan's internal registers is indexed via
+	 * 0x22: Chip Setup & Control Register Index Register (CSCI)
+	 * 0x23: Chip Setup & Control Register Data  Register (CSCD)
+	 *
+	 */
+
+	/*
+	 * 0x40 is the Power Management Unit's Force Mode Register.
+	 * Bit 6 enables Hyperspeed Mode (66/100 MHz core frequency)
+	 */
+
+	local_irq_disable();
+	outb_p(0x40, REG_CSCIR);		/* Disable hyperspeed mode */
+	outb_p(0x00, REG_CSCDR);
+	local_irq_enable();		/* wait till internal pipelines and */
+	udelay(1000);			/* buffers have cleaned up          */
+
+	local_irq_disable();
+
+	/* now, set the CPU clock speed register (0x80) */
+	outb_p(0x80, REG_CSCIR);
+	outb_p(elan_multiplier[state].val80h, REG_CSCDR);
+
+	/* now, the hyperspeed bit in PMU Force Mode Register (0x40) */
+	outb_p(0x40, REG_CSCIR);
+	outb_p(elan_multiplier[state].val40h, REG_CSCDR);
+	udelay(10000);
+	local_irq_enable();
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+};
+
+
+/**
+ *	elanfreq_validatespeed: test if frequency range is valid
+ *	@policy: the policy to validate
+ *
+ *	This function checks if a given frequency range in kHz is valid
+ *	for the hardware supported by the driver.
+ */
+
+static int elanfreq_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, &elanfreq_table[0]);
+}
+
+static int elanfreq_target(struct cpufreq_policy *policy,
+			    unsigned int target_freq,
+			    unsigned int relation)
+{
+	unsigned int newstate = 0;
+
+	if (cpufreq_frequency_table_target(policy, &elanfreq_table[0],
+				target_freq, relation, &newstate))
+		return -EINVAL;
+
+	elanfreq_set_cpu_state(newstate);
+
+	return 0;
+}
+
+
+/*
+ *	Module init and exit code
+ */
+
+static int elanfreq_cpu_init(struct cpufreq_policy *policy)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	unsigned int i;
+	int result;
+
+	/* capability check */
+	if ((c->x86_vendor != X86_VENDOR_AMD) ||
+	    (c->x86 != 4) || (c->x86_model != 10))
+		return -ENODEV;
+
+	/* max freq */
+	if (!max_freq)
+		max_freq = elanfreq_get_cpu_frequency(0);
+
+	/* table init */
+	for (i = 0; (elanfreq_table[i].frequency != CPUFREQ_TABLE_END); i++) {
+		if (elanfreq_table[i].frequency > max_freq)
+			elanfreq_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+	}
+
+	/* cpuinfo and default policy values */
+	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
+	policy->cur = elanfreq_get_cpu_frequency(0);
+
+	result = cpufreq_frequency_table_cpuinfo(policy, elanfreq_table);
+	if (result)
+		return result;
+
+	cpufreq_frequency_table_get_attr(elanfreq_table, policy->cpu);
+	return 0;
+}
+
+
+static int elanfreq_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+
+#ifndef MODULE
+/**
+ * elanfreq_setup - elanfreq command line parameter parsing
+ *
+ * elanfreq command line parameter.  Use:
+ *  elanfreq=66000
+ * to set the maximum CPU frequency to 66 MHz. Note that in
+ * case you do not give this boot parameter, the maximum
+ * frequency will fall back to _current_ CPU frequency which
+ * might be lower. If you build this as a module, use the
+ * max_freq module parameter instead.
+ */
+static int __init elanfreq_setup(char *str)
+{
+	max_freq = simple_strtoul(str, &str, 0);
+	printk(KERN_WARNING "You're using the deprecated elanfreq command line option. Use elanfreq.max_freq instead, please!\n");
+	return 1;
+}
+__setup("elanfreq=", elanfreq_setup);
+#endif
+
+
+static struct freq_attr *elanfreq_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+
+static struct cpufreq_driver elanfreq_driver = {
+	.get		= elanfreq_get_cpu_frequency,
+	.verify		= elanfreq_verify,
+	.target		= elanfreq_target,
+	.init		= elanfreq_cpu_init,
+	.exit		= elanfreq_cpu_exit,
+	.name		= "elanfreq",
+	.owner		= THIS_MODULE,
+	.attr		= elanfreq_attr,
+};
+
+
+static int __init elanfreq_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	/* Test if we have the right hardware */
+	if ((c->x86_vendor != X86_VENDOR_AMD) ||
+		(c->x86 != 4) || (c->x86_model != 10)) {
+		printk(KERN_INFO "elanfreq: error: no Elan processor found!\n");
+		return -ENODEV;
+	}
+	return cpufreq_register_driver(&elanfreq_driver);
+}
+
+
+static void __exit elanfreq_exit(void)
+{
+	cpufreq_unregister_driver(&elanfreq_driver);
+}
+
+
+module_param(max_freq, int, 0444);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Robert Schwebel <r.schwebel@pengutronix.de>, "
+		"Sven Geggus <sven@geggus.net>");
+MODULE_DESCRIPTION("cpufreq driver for AMD's Elan CPUs");
+
+module_init(elanfreq_init);
+module_exit(elanfreq_exit);
diff --git a/drivers/cpufreq/freq_table.c b/drivers/cpufreq/freq_table.c
new file mode 100644
index 00000000..90431cb9
--- /dev/null
+++ b/drivers/cpufreq/freq_table.c
@@ -0,0 +1,236 @@
+/*
+ * linux/drivers/cpufreq/freq_table.c
+ *
+ * Copyright (C) 2002 - 2003 Dominik Brodowski
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+
+/*********************************************************************
+ *                     FREQUENCY TABLE HELPERS                       *
+ *********************************************************************/
+
+int cpufreq_frequency_table_cpuinfo(struct cpufreq_policy *policy,
+				    struct cpufreq_frequency_table *table)
+{
+	unsigned int min_freq = ~0;
+	unsigned int max_freq = 0;
+	unsigned int i;
+
+	for (i = 0; (table[i].frequency != CPUFREQ_TABLE_END); i++) {
+		unsigned int freq = table[i].frequency;
+		if (freq == CPUFREQ_ENTRY_INVALID) {
+			pr_debug("table entry %u is invalid, skipping\n", i);
+
+			continue;
+		}
+		pr_debug("table entry %u: %u kHz, %u index\n",
+					i, freq, table[i].index);
+		if (freq < min_freq)
+			min_freq = freq;
+		if (freq > max_freq)
+			max_freq = freq;
+	}
+
+	policy->min = policy->cpuinfo.min_freq = min_freq;
+	policy->max = policy->cpuinfo.max_freq = max_freq;
+
+	if (policy->min == ~0)
+		return -EINVAL;
+	else
+		return 0;
+}
+EXPORT_SYMBOL_GPL(cpufreq_frequency_table_cpuinfo);
+
+
+int cpufreq_frequency_table_verify(struct cpufreq_policy *policy,
+				   struct cpufreq_frequency_table *table)
+{
+	unsigned int next_larger = ~0;
+	unsigned int i;
+	unsigned int count = 0;
+
+	pr_debug("request for verification of policy (%u - %u kHz) for cpu %u\n",
+					policy->min, policy->max, policy->cpu);
+
+	if (!cpu_online(policy->cpu))
+		return -EINVAL;
+
+	cpufreq_verify_within_limits(policy, policy->cpuinfo.min_freq,
+				     policy->cpuinfo.max_freq);
+
+	for (i = 0; (table[i].frequency != CPUFREQ_TABLE_END); i++) {
+		unsigned int freq = table[i].frequency;
+		if (freq == CPUFREQ_ENTRY_INVALID)
+			continue;
+		if ((freq >= policy->min) && (freq <= policy->max))
+			count++;
+		else if ((next_larger > freq) && (freq > policy->max))
+			next_larger = freq;
+	}
+
+	if (!count)
+		policy->max = next_larger;
+
+	cpufreq_verify_within_limits(policy, policy->cpuinfo.min_freq,
+				     policy->cpuinfo.max_freq);
+
+	pr_debug("verification lead to (%u - %u kHz) for cpu %u\n",
+				policy->min, policy->max, policy->cpu);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(cpufreq_frequency_table_verify);
+
+
+int cpufreq_frequency_table_target(struct cpufreq_policy *policy,
+				   struct cpufreq_frequency_table *table,
+				   unsigned int target_freq,
+				   unsigned int relation,
+				   unsigned int *index)
+{
+	struct cpufreq_frequency_table optimal = {
+		.index = ~0,
+		.frequency = 0,
+	};
+	struct cpufreq_frequency_table suboptimal = {
+		.index = ~0,
+		.frequency = 0,
+	};
+	unsigned int i;
+
+	pr_debug("request for target %u kHz (relation: %u) for cpu %u\n",
+					target_freq, relation, policy->cpu);
+
+	switch (relation) {
+	case CPUFREQ_RELATION_H:
+		suboptimal.frequency = ~0;
+		break;
+	case CPUFREQ_RELATION_L:
+		optimal.frequency = ~0;
+		break;
+	}
+
+	if (!cpu_online(policy->cpu))
+		return -EINVAL;
+
+	for (i = 0; (table[i].frequency != CPUFREQ_TABLE_END); i++) {
+		unsigned int freq = table[i].frequency;
+		if (freq == CPUFREQ_ENTRY_INVALID)
+			continue;
+		if ((freq < policy->min) || (freq > policy->max))
+			continue;
+		switch (relation) {
+		case CPUFREQ_RELATION_H:
+			if (freq <= target_freq) {
+				if (freq >= optimal.frequency) {
+					optimal.frequency = freq;
+					optimal.index = i;
+				}
+			} else {
+				if (freq <= suboptimal.frequency) {
+					suboptimal.frequency = freq;
+					suboptimal.index = i;
+				}
+			}
+			break;
+		case CPUFREQ_RELATION_L:
+			if (freq >= target_freq) {
+				if (freq <= optimal.frequency) {
+					optimal.frequency = freq;
+					optimal.index = i;
+				}
+			} else {
+				if (freq >= suboptimal.frequency) {
+					suboptimal.frequency = freq;
+					suboptimal.index = i;
+				}
+			}
+			break;
+		}
+	}
+	if (optimal.index > i) {
+		if (suboptimal.index > i)
+			return -EINVAL;
+		*index = suboptimal.index;
+	} else
+		*index = optimal.index;
+
+	pr_debug("target is %u (%u kHz, %u)\n", *index, table[*index].frequency,
+		table[*index].index);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(cpufreq_frequency_table_target);
+
+static DEFINE_PER_CPU(struct cpufreq_frequency_table *, cpufreq_show_table);
+/**
+ * show_available_freqs - show available frequencies for the specified CPU
+ */
+static ssize_t show_available_freqs(struct cpufreq_policy *policy, char *buf)
+{
+	unsigned int i = 0;
+	unsigned int cpu = policy->cpu;
+	ssize_t count = 0;
+	struct cpufreq_frequency_table *table;
+
+	if (!per_cpu(cpufreq_show_table, cpu))
+		return -ENODEV;
+
+	table = per_cpu(cpufreq_show_table, cpu);
+
+	for (i = 0; (table[i].frequency != CPUFREQ_TABLE_END); i++) {
+		if (table[i].frequency == CPUFREQ_ENTRY_INVALID)
+			continue;
+		count += sprintf(&buf[count], "%d ", table[i].frequency);
+	}
+	count += sprintf(&buf[count], "\n");
+
+	return count;
+
+}
+
+struct freq_attr cpufreq_freq_attr_scaling_available_freqs = {
+	.attr = { .name = "scaling_available_frequencies",
+		  .mode = 0444,
+		},
+	.show = show_available_freqs,
+};
+EXPORT_SYMBOL_GPL(cpufreq_freq_attr_scaling_available_freqs);
+
+/*
+ * if you use these, you must assure that the frequency table is valid
+ * all the time between get_attr and put_attr!
+ */
+void cpufreq_frequency_table_get_attr(struct cpufreq_frequency_table *table,
+				      unsigned int cpu)
+{
+	pr_debug("setting show_table for cpu %u to %p\n", cpu, table);
+	per_cpu(cpufreq_show_table, cpu) = table;
+}
+EXPORT_SYMBOL_GPL(cpufreq_frequency_table_get_attr);
+
+void cpufreq_frequency_table_put_attr(unsigned int cpu)
+{
+	pr_debug("clearing show_table for cpu %u\n", cpu);
+	per_cpu(cpufreq_show_table, cpu) = NULL;
+}
+EXPORT_SYMBOL_GPL(cpufreq_frequency_table_put_attr);
+
+struct cpufreq_frequency_table *cpufreq_frequency_get_table(unsigned int cpu)
+{
+	return per_cpu(cpufreq_show_table, cpu);
+}
+EXPORT_SYMBOL_GPL(cpufreq_frequency_get_table);
+
+MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("CPUfreq frequency table helpers");
+MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/gx-suspmod.c b/drivers/cpufreq/gx-suspmod.c
new file mode 100644
index 00000000..ffe1f2c9
--- /dev/null
+++ b/drivers/cpufreq/gx-suspmod.c
@@ -0,0 +1,514 @@
+/*
+ *	Cyrix MediaGX and NatSemi Geode Suspend Modulation
+ *	(C) 2002 Zwane Mwaikambo <zwane@commfireservices.com>
+ *	(C) 2002 Hiroshi Miura   <miura@da-cha.org>
+ *	All Rights Reserved
+ *
+ *	This program is free software; you can redistribute it and/or
+ *      modify it under the terms of the GNU General Public License
+ *      version 2 as published by the Free Software Foundation
+ *
+ *      The author(s) of this software shall not be held liable for damages
+ *      of any nature resulting due to the use of this software. This
+ *      software is provided AS-IS with no warranties.
+ *
+ * Theoretical note:
+ *
+ *	(see Geode(tm) CS5530 manual (rev.4.1) page.56)
+ *
+ *	CPU frequency control on NatSemi Geode GX1/GXLV processor and CS55x0
+ *	are based on Suspend Modulation.
+ *
+ *	Suspend Modulation works by asserting and de-asserting the SUSP# pin
+ *	to CPU(GX1/GXLV) for configurable durations. When asserting SUSP#
+ *	the CPU enters an idle state. GX1 stops its core clock when SUSP# is
+ *	asserted then power consumption is reduced.
+ *
+ *	Suspend Modulation's OFF/ON duration are configurable
+ *	with 'Suspend Modulation OFF Count Register'
+ *	and 'Suspend Modulation ON Count Register'.
+ *	These registers are 8bit counters that represent the number of
+ *	32us intervals which the SUSP# pin is asserted(ON)/de-asserted(OFF)
+ *	to the processor.
+ *
+ *	These counters define a ratio which is the effective frequency
+ *	of operation of the system.
+ *
+ *			       OFF Count
+ *	F_eff = Fgx * ----------------------
+ *	                OFF Count + ON Count
+ *
+ *	0 <= On Count, Off Count <= 255
+ *
+ *	From these limits, we can get register values
+ *
+ *	off_duration + on_duration <= MAX_DURATION
+ *	on_duration = off_duration * (stock_freq - freq) / freq
+ *
+ *      off_duration  =  (freq * DURATION) / stock_freq
+ *      on_duration = DURATION - off_duration
+ *
+ *
+ *---------------------------------------------------------------------------
+ *
+ * ChangeLog:
+ *	Dec. 12, 2003	Hiroshi Miura <miura@da-cha.org>
+ *		- fix on/off register mistake
+ *		- fix cpu_khz calc when it stops cpu modulation.
+ *
+ *	Dec. 11, 2002	Hiroshi Miura <miura@da-cha.org>
+ *		- rewrite for Cyrix MediaGX Cx5510/5520 and
+ *		  NatSemi Geode Cs5530(A).
+ *
+ *	Jul. ??, 2002  Zwane Mwaikambo <zwane@commfireservices.com>
+ *		- cs5530_mod patch for 2.4.19-rc1.
+ *
+ *---------------------------------------------------------------------------
+ *
+ * Todo
+ *	Test on machines with 5510, 5530, 5530A
+ */
+
+/************************************************************************
+ *			Suspend Modulation - Definitions		*
+ ************************************************************************/
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/cpufreq.h>
+#include <linux/pci.h>
+#include <linux/errno.h>
+#include <linux/slab.h>
+
+#include <asm/processor-cyrix.h>
+
+/* PCI config registers, all at F0 */
+#define PCI_PMER1	0x80	/* power management enable register 1 */
+#define PCI_PMER2	0x81	/* power management enable register 2 */
+#define PCI_PMER3	0x82	/* power management enable register 3 */
+#define PCI_IRQTC	0x8c	/* irq speedup timer counter register:typical 2 to 4ms */
+#define PCI_VIDTC	0x8d	/* video speedup timer counter register: typical 50 to 100ms */
+#define PCI_MODOFF	0x94	/* suspend modulation OFF counter register, 1 = 32us */
+#define PCI_MODON	0x95	/* suspend modulation ON counter register */
+#define PCI_SUSCFG	0x96	/* suspend configuration register */
+
+/* PMER1 bits */
+#define GPM		(1<<0)	/* global power management */
+#define GIT		(1<<1)	/* globally enable PM device idle timers */
+#define GTR		(1<<2)	/* globally enable IO traps */
+#define IRQ_SPDUP	(1<<3)	/* disable clock throttle during interrupt handling */
+#define VID_SPDUP	(1<<4)	/* disable clock throttle during vga video handling */
+
+/* SUSCFG bits */
+#define SUSMOD		(1<<0)	/* enable/disable suspend modulation */
+/* the below is supported only with cs5530 (after rev.1.2)/cs5530A */
+#define SMISPDUP	(1<<1)	/* select how SMI re-enable suspend modulation: */
+				/* IRQTC timer or read SMI speedup disable reg.(F1BAR[08-09h]) */
+#define SUSCFG		(1<<2)	/* enable powering down a GXLV processor. "Special 3Volt Suspend" mode */
+/* the below is supported only with cs5530A */
+#define PWRSVE_ISA	(1<<3)	/* stop ISA clock  */
+#define PWRSVE		(1<<4)	/* active idle */
+
+struct gxfreq_params {
+	u8 on_duration;
+	u8 off_duration;
+	u8 pci_suscfg;
+	u8 pci_pmer1;
+	u8 pci_pmer2;
+	struct pci_dev *cs55x0;
+};
+
+static struct gxfreq_params *gx_params;
+static int stock_freq;
+
+/* PCI bus clock - defaults to 30.000 if cpu_khz is not available */
+static int pci_busclk;
+module_param(pci_busclk, int, 0444);
+
+/* maximum duration for which the cpu may be suspended
+ * (32us * MAX_DURATION). If no parameter is given, this defaults
+ * to 255.
+ * Note that this leads to a maximum of 8 ms(!) where the CPU clock
+ * is suspended -- processing power is just 0.39% of what it used to be,
+ * though. 781.25 kHz(!) for a 200 MHz processor -- wow. */
+static int max_duration = 255;
+module_param(max_duration, int, 0444);
+
+/* For the default policy, we want at least some processing power
+ * - let's say 5%. (min = maxfreq / POLICY_MIN_DIV)
+ */
+#define POLICY_MIN_DIV 20
+
+
+/**
+ * we can detect a core multipiler from dir0_lsb
+ * from GX1 datasheet p.56,
+ *	MULT[3:0]:
+ *	0000 = SYSCLK multiplied by 4 (test only)
+ *	0001 = SYSCLK multiplied by 10
+ *	0010 = SYSCLK multiplied by 4
+ *	0011 = SYSCLK multiplied by 6
+ *	0100 = SYSCLK multiplied by 9
+ *	0101 = SYSCLK multiplied by 5
+ *	0110 = SYSCLK multiplied by 7
+ *	0111 = SYSCLK multiplied by 8
+ *              of 33.3MHz
+ **/
+static int gx_freq_mult[16] = {
+		4, 10, 4, 6, 9, 5, 7, 8,
+		0, 0, 0, 0, 0, 0, 0, 0
+};
+
+
+/****************************************************************
+ *	Low Level chipset interface				*
+ ****************************************************************/
+static struct pci_device_id gx_chipset_tbl[] __initdata = {
+	{ PCI_VDEVICE(CYRIX, PCI_DEVICE_ID_CYRIX_5530_LEGACY), },
+	{ PCI_VDEVICE(CYRIX, PCI_DEVICE_ID_CYRIX_5520), },
+	{ PCI_VDEVICE(CYRIX, PCI_DEVICE_ID_CYRIX_5510), },
+	{ 0, },
+};
+
+static void gx_write_byte(int reg, int value)
+{
+	pci_write_config_byte(gx_params->cs55x0, reg, value);
+}
+
+/**
+ * gx_detect_chipset:
+ *
+ **/
+static __init struct pci_dev *gx_detect_chipset(void)
+{
+	struct pci_dev *gx_pci = NULL;
+
+	/* check if CPU is a MediaGX or a Geode. */
+	if ((boot_cpu_data.x86_vendor != X86_VENDOR_NSC) &&
+	    (boot_cpu_data.x86_vendor != X86_VENDOR_CYRIX)) {
+		pr_debug("error: no MediaGX/Geode processor found!\n");
+		return NULL;
+	}
+
+	/* detect which companion chip is used */
+	for_each_pci_dev(gx_pci) {
+		if ((pci_match_id(gx_chipset_tbl, gx_pci)) != NULL)
+			return gx_pci;
+	}
+
+	pr_debug("error: no supported chipset found!\n");
+	return NULL;
+}
+
+/**
+ * gx_get_cpuspeed:
+ *
+ * Finds out at which efficient frequency the Cyrix MediaGX/NatSemi
+ * Geode CPU runs.
+ */
+static unsigned int gx_get_cpuspeed(unsigned int cpu)
+{
+	if ((gx_params->pci_suscfg & SUSMOD) == 0)
+		return stock_freq;
+
+	return (stock_freq * gx_params->off_duration)
+		/ (gx_params->on_duration + gx_params->off_duration);
+}
+
+/**
+ *      gx_validate_speed:
+ *      determine current cpu speed
+ *
+ **/
+
+static unsigned int gx_validate_speed(unsigned int khz, u8 *on_duration,
+		u8 *off_duration)
+{
+	unsigned int i;
+	u8 tmp_on, tmp_off;
+	int old_tmp_freq = stock_freq;
+	int tmp_freq;
+
+	*off_duration = 1;
+	*on_duration = 0;
+
+	for (i = max_duration; i > 0; i--) {
+		tmp_off = ((khz * i) / stock_freq) & 0xff;
+		tmp_on = i - tmp_off;
+		tmp_freq = (stock_freq * tmp_off) / i;
+		/* if this relation is closer to khz, use this. If it's equal,
+		 * prefer it, too - lower latency */
+		if (abs(tmp_freq - khz) <= abs(old_tmp_freq - khz)) {
+			*on_duration = tmp_on;
+			*off_duration = tmp_off;
+			old_tmp_freq = tmp_freq;
+		}
+	}
+
+	return old_tmp_freq;
+}
+
+
+/**
+ * gx_set_cpuspeed:
+ * set cpu speed in khz.
+ **/
+
+static void gx_set_cpuspeed(unsigned int khz)
+{
+	u8 suscfg, pmer1;
+	unsigned int new_khz;
+	unsigned long flags;
+	struct cpufreq_freqs freqs;
+
+	freqs.cpu = 0;
+	freqs.old = gx_get_cpuspeed(0);
+
+	new_khz = gx_validate_speed(khz, &gx_params->on_duration,
+			&gx_params->off_duration);
+
+	freqs.new = new_khz;
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	local_irq_save(flags);
+
+
+
+	if (new_khz != stock_freq) {
+		/* if new khz == 100% of CPU speed, it is special case */
+		switch (gx_params->cs55x0->device) {
+		case PCI_DEVICE_ID_CYRIX_5530_LEGACY:
+			pmer1 = gx_params->pci_pmer1 | IRQ_SPDUP | VID_SPDUP;
+			/* FIXME: need to test other values -- Zwane,Miura */
+			/* typical 2 to 4ms */
+			gx_write_byte(PCI_IRQTC, 4);
+			/* typical 50 to 100ms */
+			gx_write_byte(PCI_VIDTC, 100);
+			gx_write_byte(PCI_PMER1, pmer1);
+
+			if (gx_params->cs55x0->revision < 0x10) {
+				/* CS5530(rev 1.2, 1.3) */
+				suscfg = gx_params->pci_suscfg|SUSMOD;
+			} else {
+				/* CS5530A,B.. */
+				suscfg = gx_params->pci_suscfg|SUSMOD|PWRSVE;
+			}
+			break;
+		case PCI_DEVICE_ID_CYRIX_5520:
+		case PCI_DEVICE_ID_CYRIX_5510:
+			suscfg = gx_params->pci_suscfg | SUSMOD;
+			break;
+		default:
+			local_irq_restore(flags);
+			pr_debug("fatal: try to set unknown chipset.\n");
+			return;
+		}
+	} else {
+		suscfg = gx_params->pci_suscfg & ~(SUSMOD);
+		gx_params->off_duration = 0;
+		gx_params->on_duration = 0;
+		pr_debug("suspend modulation disabled: cpu runs 100%% speed.\n");
+	}
+
+	gx_write_byte(PCI_MODOFF, gx_params->off_duration);
+	gx_write_byte(PCI_MODON, gx_params->on_duration);
+
+	gx_write_byte(PCI_SUSCFG, suscfg);
+	pci_read_config_byte(gx_params->cs55x0, PCI_SUSCFG, &suscfg);
+
+	local_irq_restore(flags);
+
+	gx_params->pci_suscfg = suscfg;
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+
+	pr_debug("suspend modulation w/ duration of ON:%d us, OFF:%d us\n",
+		gx_params->on_duration * 32, gx_params->off_duration * 32);
+	pr_debug("suspend modulation w/ clock speed: %d kHz.\n", freqs.new);
+}
+
+/****************************************************************
+ *             High level functions                             *
+ ****************************************************************/
+
+/*
+ *	cpufreq_gx_verify: test if frequency range is valid
+ *
+ *	This function checks if a given frequency range in kHz is valid
+ *      for the hardware supported by the driver.
+ */
+
+static int cpufreq_gx_verify(struct cpufreq_policy *policy)
+{
+	unsigned int tmp_freq = 0;
+	u8 tmp1, tmp2;
+
+	if (!stock_freq || !policy)
+		return -EINVAL;
+
+	policy->cpu = 0;
+	cpufreq_verify_within_limits(policy, (stock_freq / max_duration),
+			stock_freq);
+
+	/* it needs to be assured that at least one supported frequency is
+	 * within policy->min and policy->max. If it is not, policy->max
+	 * needs to be increased until one freuqency is supported.
+	 * policy->min may not be decreased, though. This way we guarantee a
+	 * specific processing capacity.
+	 */
+	tmp_freq = gx_validate_speed(policy->min, &tmp1, &tmp2);
+	if (tmp_freq < policy->min)
+		tmp_freq += stock_freq / max_duration;
+	policy->min = tmp_freq;
+	if (policy->min > policy->max)
+		policy->max = tmp_freq;
+	tmp_freq = gx_validate_speed(policy->max, &tmp1, &tmp2);
+	if (tmp_freq > policy->max)
+		tmp_freq -= stock_freq / max_duration;
+	policy->max = tmp_freq;
+	if (policy->max < policy->min)
+		policy->max = policy->min;
+	cpufreq_verify_within_limits(policy, (stock_freq / max_duration),
+			stock_freq);
+
+	return 0;
+}
+
+/*
+ *      cpufreq_gx_target:
+ *
+ */
+static int cpufreq_gx_target(struct cpufreq_policy *policy,
+			     unsigned int target_freq,
+			     unsigned int relation)
+{
+	u8 tmp1, tmp2;
+	unsigned int tmp_freq;
+
+	if (!stock_freq || !policy)
+		return -EINVAL;
+
+	policy->cpu = 0;
+
+	tmp_freq = gx_validate_speed(target_freq, &tmp1, &tmp2);
+	while (tmp_freq < policy->min) {
+		tmp_freq += stock_freq / max_duration;
+		tmp_freq = gx_validate_speed(tmp_freq, &tmp1, &tmp2);
+	}
+	while (tmp_freq > policy->max) {
+		tmp_freq -= stock_freq / max_duration;
+		tmp_freq = gx_validate_speed(tmp_freq, &tmp1, &tmp2);
+	}
+
+	gx_set_cpuspeed(tmp_freq);
+
+	return 0;
+}
+
+static int cpufreq_gx_cpu_init(struct cpufreq_policy *policy)
+{
+	unsigned int maxfreq, curfreq;
+
+	if (!policy || policy->cpu != 0)
+		return -ENODEV;
+
+	/* determine maximum frequency */
+	if (pci_busclk)
+		maxfreq = pci_busclk * gx_freq_mult[getCx86(CX86_DIR1) & 0x0f];
+	else if (cpu_khz)
+		maxfreq = cpu_khz;
+	else
+		maxfreq = 30000 * gx_freq_mult[getCx86(CX86_DIR1) & 0x0f];
+
+	stock_freq = maxfreq;
+	curfreq = gx_get_cpuspeed(0);
+
+	pr_debug("cpu max frequency is %d.\n", maxfreq);
+	pr_debug("cpu current frequency is %dkHz.\n", curfreq);
+
+	/* setup basic struct for cpufreq API */
+	policy->cpu = 0;
+
+	if (max_duration < POLICY_MIN_DIV)
+		policy->min = maxfreq / max_duration;
+	else
+		policy->min = maxfreq / POLICY_MIN_DIV;
+	policy->max = maxfreq;
+	policy->cur = curfreq;
+	policy->cpuinfo.min_freq = maxfreq / max_duration;
+	policy->cpuinfo.max_freq = maxfreq;
+	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
+
+	return 0;
+}
+
+/*
+ * cpufreq_gx_init:
+ *   MediaGX/Geode GX initialize cpufreq driver
+ */
+static struct cpufreq_driver gx_suspmod_driver = {
+	.get		= gx_get_cpuspeed,
+	.verify		= cpufreq_gx_verify,
+	.target		= cpufreq_gx_target,
+	.init		= cpufreq_gx_cpu_init,
+	.name		= "gx-suspmod",
+	.owner		= THIS_MODULE,
+};
+
+static int __init cpufreq_gx_init(void)
+{
+	int ret;
+	struct gxfreq_params *params;
+	struct pci_dev *gx_pci;
+
+	/* Test if we have the right hardware */
+	gx_pci = gx_detect_chipset();
+	if (gx_pci == NULL)
+		return -ENODEV;
+
+	/* check whether module parameters are sane */
+	if (max_duration > 0xff)
+		max_duration = 0xff;
+
+	pr_debug("geode suspend modulation available.\n");
+
+	params = kzalloc(sizeof(struct gxfreq_params), GFP_KERNEL);
+	if (params == NULL)
+		return -ENOMEM;
+
+	params->cs55x0 = gx_pci;
+	gx_params = params;
+
+	/* keep cs55x0 configurations */
+	pci_read_config_byte(params->cs55x0, PCI_SUSCFG, &(params->pci_suscfg));
+	pci_read_config_byte(params->cs55x0, PCI_PMER1, &(params->pci_pmer1));
+	pci_read_config_byte(params->cs55x0, PCI_PMER2, &(params->pci_pmer2));
+	pci_read_config_byte(params->cs55x0, PCI_MODON, &(params->on_duration));
+	pci_read_config_byte(params->cs55x0, PCI_MODOFF,
+			&(params->off_duration));
+
+	ret = cpufreq_register_driver(&gx_suspmod_driver);
+	if (ret) {
+		kfree(params);
+		return ret;                   /* register error! */
+	}
+
+	return 0;
+}
+
+static void __exit cpufreq_gx_exit(void)
+{
+	cpufreq_unregister_driver(&gx_suspmod_driver);
+	pci_dev_put(gx_params->cs55x0);
+	kfree(gx_params);
+}
+
+MODULE_AUTHOR("Hiroshi Miura <miura@da-cha.org>");
+MODULE_DESCRIPTION("Cpufreq driver for Cyrix MediaGX and NatSemi Geode");
+MODULE_LICENSE("GPL");
+
+module_init(cpufreq_gx_init);
+module_exit(cpufreq_gx_exit);
+
diff --git a/drivers/cpufreq/longhaul.c b/drivers/cpufreq/longhaul.c
new file mode 100644
index 00000000..f47d26e2
--- /dev/null
+++ b/drivers/cpufreq/longhaul.c
@@ -0,0 +1,1024 @@
+/*
+ *  (C) 2001-2004  Dave Jones. <davej@redhat.com>
+ *  (C) 2002  Padraig Brady. <padraig@antefacto.com>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *  Based upon datasheets & sample CPUs kindly provided by VIA.
+ *
+ *  VIA have currently 3 different versions of Longhaul.
+ *  Version 1 (Longhaul) uses the BCR2 MSR at 0x1147.
+ *   It is present only in Samuel 1 (C5A), Samuel 2 (C5B) stepping 0.
+ *  Version 2 of longhaul is backward compatible with v1, but adds
+ *   LONGHAUL MSR for purpose of both frequency and voltage scaling.
+ *   Present in Samuel 2 (steppings 1-7 only) (C5B), and Ezra (C5C).
+ *  Version 3 of longhaul got renamed to Powersaver and redesigned
+ *   to use only the POWERSAVER MSR at 0x110a.
+ *   It is present in Ezra-T (C5M), Nehemiah (C5X) and above.
+ *   It's pretty much the same feature wise to longhaul v2, though
+ *   there is provision for scaling FSB too, but this doesn't work
+ *   too well in practice so we don't even try to use this.
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/delay.h>
+#include <linux/timex.h>
+#include <linux/io.h>
+#include <linux/acpi.h>
+
+#include <asm/msr.h>
+#include <acpi/processor.h>
+
+#include "longhaul.h"
+
+#define PFX "longhaul: "
+
+#define TYPE_LONGHAUL_V1	1
+#define TYPE_LONGHAUL_V2	2
+#define TYPE_POWERSAVER		3
+
+#define	CPU_SAMUEL	1
+#define	CPU_SAMUEL2	2
+#define	CPU_EZRA	3
+#define	CPU_EZRA_T	4
+#define	CPU_NEHEMIAH	5
+#define	CPU_NEHEMIAH_C	6
+
+/* Flags */
+#define USE_ACPI_C3		(1 << 1)
+#define USE_NORTHBRIDGE		(1 << 2)
+
+static int cpu_model;
+static unsigned int numscales = 16;
+static unsigned int fsb;
+
+static const struct mV_pos *vrm_mV_table;
+static const unsigned char *mV_vrm_table;
+
+static unsigned int highest_speed, lowest_speed; /* kHz */
+static unsigned int minmult, maxmult;
+static int can_scale_voltage;
+static struct acpi_processor *pr;
+static struct acpi_processor_cx *cx;
+static u32 acpi_regs_addr;
+static u8 longhaul_flags;
+static unsigned int longhaul_index;
+
+/* Module parameters */
+static int scale_voltage;
+static int disable_acpi_c3;
+static int revid_errata;
+
+
+/* Clock ratios multiplied by 10 */
+static int mults[32];
+static int eblcr[32];
+static int longhaul_version;
+static struct cpufreq_frequency_table *longhaul_table;
+
+static char speedbuffer[8];
+
+static char *print_speed(int speed)
+{
+	if (speed < 1000) {
+		snprintf(speedbuffer, sizeof(speedbuffer), "%dMHz", speed);
+		return speedbuffer;
+	}
+
+	if (speed%1000 == 0)
+		snprintf(speedbuffer, sizeof(speedbuffer),
+			"%dGHz", speed/1000);
+	else
+		snprintf(speedbuffer, sizeof(speedbuffer),
+			"%d.%dGHz", speed/1000, (speed%1000)/100);
+
+	return speedbuffer;
+}
+
+
+static unsigned int calc_speed(int mult)
+{
+	int khz;
+	khz = (mult/10)*fsb;
+	if (mult%10)
+		khz += fsb/2;
+	khz *= 1000;
+	return khz;
+}
+
+
+static int longhaul_get_cpu_mult(void)
+{
+	unsigned long invalue = 0, lo, hi;
+
+	rdmsr(MSR_IA32_EBL_CR_POWERON, lo, hi);
+	invalue = (lo & (1<<22|1<<23|1<<24|1<<25))>>22;
+	if (longhaul_version == TYPE_LONGHAUL_V2 ||
+	    longhaul_version == TYPE_POWERSAVER) {
+		if (lo & (1<<27))
+			invalue += 16;
+	}
+	return eblcr[invalue];
+}
+
+/* For processor with BCR2 MSR */
+
+static void do_longhaul1(unsigned int mults_index)
+{
+	union msr_bcr2 bcr2;
+
+	rdmsrl(MSR_VIA_BCR2, bcr2.val);
+	/* Enable software clock multiplier */
+	bcr2.bits.ESOFTBF = 1;
+	bcr2.bits.CLOCKMUL = mults_index & 0xff;
+
+	/* Sync to timer tick */
+	safe_halt();
+	/* Change frequency on next halt or sleep */
+	wrmsrl(MSR_VIA_BCR2, bcr2.val);
+	/* Invoke transition */
+	ACPI_FLUSH_CPU_CACHE();
+	halt();
+
+	/* Disable software clock multiplier */
+	local_irq_disable();
+	rdmsrl(MSR_VIA_BCR2, bcr2.val);
+	bcr2.bits.ESOFTBF = 0;
+	wrmsrl(MSR_VIA_BCR2, bcr2.val);
+}
+
+/* For processor with Longhaul MSR */
+
+static void do_powersaver(int cx_address, unsigned int mults_index,
+			  unsigned int dir)
+{
+	union msr_longhaul longhaul;
+	u32 t;
+
+	rdmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+	/* Setup new frequency */
+	if (!revid_errata)
+		longhaul.bits.RevisionKey = longhaul.bits.RevisionID;
+	else
+		longhaul.bits.RevisionKey = 0;
+	longhaul.bits.SoftBusRatio = mults_index & 0xf;
+	longhaul.bits.SoftBusRatio4 = (mults_index & 0x10) >> 4;
+	/* Setup new voltage */
+	if (can_scale_voltage)
+		longhaul.bits.SoftVID = (mults_index >> 8) & 0x1f;
+	/* Sync to timer tick */
+	safe_halt();
+	/* Raise voltage if necessary */
+	if (can_scale_voltage && dir) {
+		longhaul.bits.EnableSoftVID = 1;
+		wrmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+		/* Change voltage */
+		if (!cx_address) {
+			ACPI_FLUSH_CPU_CACHE();
+			halt();
+		} else {
+			ACPI_FLUSH_CPU_CACHE();
+			/* Invoke C3 */
+			inb(cx_address);
+			/* Dummy op - must do something useless after P_LVL3
+			 * read */
+			t = inl(acpi_gbl_FADT.xpm_timer_block.address);
+		}
+		longhaul.bits.EnableSoftVID = 0;
+		wrmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+	}
+
+	/* Change frequency on next halt or sleep */
+	longhaul.bits.EnableSoftBusRatio = 1;
+	wrmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+	if (!cx_address) {
+		ACPI_FLUSH_CPU_CACHE();
+		halt();
+	} else {
+		ACPI_FLUSH_CPU_CACHE();
+		/* Invoke C3 */
+		inb(cx_address);
+		/* Dummy op - must do something useless after P_LVL3 read */
+		t = inl(acpi_gbl_FADT.xpm_timer_block.address);
+	}
+	/* Disable bus ratio bit */
+	longhaul.bits.EnableSoftBusRatio = 0;
+	wrmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+
+	/* Reduce voltage if necessary */
+	if (can_scale_voltage && !dir) {
+		longhaul.bits.EnableSoftVID = 1;
+		wrmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+		/* Change voltage */
+		if (!cx_address) {
+			ACPI_FLUSH_CPU_CACHE();
+			halt();
+		} else {
+			ACPI_FLUSH_CPU_CACHE();
+			/* Invoke C3 */
+			inb(cx_address);
+			/* Dummy op - must do something useless after P_LVL3
+			 * read */
+			t = inl(acpi_gbl_FADT.xpm_timer_block.address);
+		}
+		longhaul.bits.EnableSoftVID = 0;
+		wrmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+	}
+}
+
+/**
+ * longhaul_set_cpu_frequency()
+ * @mults_index : bitpattern of the new multiplier.
+ *
+ * Sets a new clock ratio.
+ */
+
+static void longhaul_setstate(unsigned int table_index)
+{
+	unsigned int mults_index;
+	int speed, mult;
+	struct cpufreq_freqs freqs;
+	unsigned long flags;
+	unsigned int pic1_mask, pic2_mask;
+	u16 bm_status = 0;
+	u32 bm_timeout = 1000;
+	unsigned int dir = 0;
+
+	mults_index = longhaul_table[table_index].index;
+	/* Safety precautions */
+	mult = mults[mults_index & 0x1f];
+	if (mult == -1)
+		return;
+	speed = calc_speed(mult);
+	if ((speed > highest_speed) || (speed < lowest_speed))
+		return;
+	/* Voltage transition before frequency transition? */
+	if (can_scale_voltage && longhaul_index < table_index)
+		dir = 1;
+
+	freqs.old = calc_speed(longhaul_get_cpu_mult());
+	freqs.new = speed;
+	freqs.cpu = 0; /* longhaul.c is UP only driver */
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	pr_debug("Setting to FSB:%dMHz Mult:%d.%dx (%s)\n",
+			fsb, mult/10, mult%10, print_speed(speed/1000));
+retry_loop:
+	preempt_disable();
+	local_irq_save(flags);
+
+	pic2_mask = inb(0xA1);
+	pic1_mask = inb(0x21);	/* works on C3. save mask. */
+	outb(0xFF, 0xA1);	/* Overkill */
+	outb(0xFE, 0x21);	/* TMR0 only */
+
+	/* Wait while PCI bus is busy. */
+	if (acpi_regs_addr && (longhaul_flags & USE_NORTHBRIDGE
+	    || ((pr != NULL) && pr->flags.bm_control))) {
+		bm_status = inw(acpi_regs_addr);
+		bm_status &= 1 << 4;
+		while (bm_status && bm_timeout) {
+			outw(1 << 4, acpi_regs_addr);
+			bm_timeout--;
+			bm_status = inw(acpi_regs_addr);
+			bm_status &= 1 << 4;
+		}
+	}
+
+	if (longhaul_flags & USE_NORTHBRIDGE) {
+		/* Disable AGP and PCI arbiters */
+		outb(3, 0x22);
+	} else if ((pr != NULL) && pr->flags.bm_control) {
+		/* Disable bus master arbitration */
+		acpi_write_bit_register(ACPI_BITREG_ARB_DISABLE, 1);
+	}
+	switch (longhaul_version) {
+
+	/*
+	 * Longhaul v1. (Samuel[C5A] and Samuel2 stepping 0[C5B])
+	 * Software controlled multipliers only.
+	 */
+	case TYPE_LONGHAUL_V1:
+		do_longhaul1(mults_index);
+		break;
+
+	/*
+	 * Longhaul v2 appears in Samuel2 Steppings 1->7 [C5B] and Ezra [C5C]
+	 *
+	 * Longhaul v3 (aka Powersaver). (Ezra-T [C5M] & Nehemiah [C5N])
+	 * Nehemiah can do FSB scaling too, but this has never been proven
+	 * to work in practice.
+	 */
+	case TYPE_LONGHAUL_V2:
+	case TYPE_POWERSAVER:
+		if (longhaul_flags & USE_ACPI_C3) {
+			/* Don't allow wakeup */
+			acpi_write_bit_register(ACPI_BITREG_BUS_MASTER_RLD, 0);
+			do_powersaver(cx->address, mults_index, dir);
+		} else {
+			do_powersaver(0, mults_index, dir);
+		}
+		break;
+	}
+
+	if (longhaul_flags & USE_NORTHBRIDGE) {
+		/* Enable arbiters */
+		outb(0, 0x22);
+	} else if ((pr != NULL) && pr->flags.bm_control) {
+		/* Enable bus master arbitration */
+		acpi_write_bit_register(ACPI_BITREG_ARB_DISABLE, 0);
+	}
+	outb(pic2_mask, 0xA1);	/* restore mask */
+	outb(pic1_mask, 0x21);
+
+	local_irq_restore(flags);
+	preempt_enable();
+
+	freqs.new = calc_speed(longhaul_get_cpu_mult());
+	/* Check if requested frequency is set. */
+	if (unlikely(freqs.new != speed)) {
+		printk(KERN_INFO PFX "Failed to set requested frequency!\n");
+		/* Revision ID = 1 but processor is expecting revision key
+		 * equal to 0. Jumpers at the bottom of processor will change
+		 * multiplier and FSB, but will not change bits in Longhaul
+		 * MSR nor enable voltage scaling. */
+		if (!revid_errata) {
+			printk(KERN_INFO PFX "Enabling \"Ignore Revision ID\" "
+						"option.\n");
+			revid_errata = 1;
+			msleep(200);
+			goto retry_loop;
+		}
+		/* Why ACPI C3 sometimes doesn't work is a mystery for me.
+		 * But it does happen. Processor is entering ACPI C3 state,
+		 * but it doesn't change frequency. I tried poking various
+		 * bits in northbridge registers, but without success. */
+		if (longhaul_flags & USE_ACPI_C3) {
+			printk(KERN_INFO PFX "Disabling ACPI C3 support.\n");
+			longhaul_flags &= ~USE_ACPI_C3;
+			if (revid_errata) {
+				printk(KERN_INFO PFX "Disabling \"Ignore "
+						"Revision ID\" option.\n");
+				revid_errata = 0;
+			}
+			msleep(200);
+			goto retry_loop;
+		}
+		/* This shouldn't happen. Longhaul ver. 2 was reported not
+		 * working on processors without voltage scaling, but with
+		 * RevID = 1. RevID errata will make things right. Just
+		 * to be 100% sure. */
+		if (longhaul_version == TYPE_LONGHAUL_V2) {
+			printk(KERN_INFO PFX "Switching to Longhaul ver. 1\n");
+			longhaul_version = TYPE_LONGHAUL_V1;
+			msleep(200);
+			goto retry_loop;
+		}
+	}
+	/* Report true CPU frequency */
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+
+	if (!bm_timeout)
+		printk(KERN_INFO PFX "Warning: Timeout while waiting for "
+				"idle PCI bus.\n");
+}
+
+/*
+ * Centaur decided to make life a little more tricky.
+ * Only longhaul v1 is allowed to read EBLCR BSEL[0:1].
+ * Samuel2 and above have to try and guess what the FSB is.
+ * We do this by assuming we booted at maximum multiplier, and interpolate
+ * between that value multiplied by possible FSBs and cpu_mhz which
+ * was calculated at boot time. Really ugly, but no other way to do this.
+ */
+
+#define ROUNDING	0xf
+
+static int guess_fsb(int mult)
+{
+	int speed = cpu_khz / 1000;
+	int i;
+	int speeds[] = { 666, 1000, 1333, 2000 };
+	int f_max, f_min;
+
+	for (i = 0; i < 4; i++) {
+		f_max = ((speeds[i] * mult) + 50) / 100;
+		f_max += (ROUNDING / 2);
+		f_min = f_max - ROUNDING;
+		if ((speed <= f_max) && (speed >= f_min))
+			return speeds[i] / 10;
+	}
+	return 0;
+}
+
+
+static int __cpuinit longhaul_get_ranges(void)
+{
+	unsigned int i, j, k = 0;
+	unsigned int ratio;
+	int mult;
+
+	/* Get current frequency */
+	mult = longhaul_get_cpu_mult();
+	if (mult == -1) {
+		printk(KERN_INFO PFX "Invalid (reserved) multiplier!\n");
+		return -EINVAL;
+	}
+	fsb = guess_fsb(mult);
+	if (fsb == 0) {
+		printk(KERN_INFO PFX "Invalid (reserved) FSB!\n");
+		return -EINVAL;
+	}
+	/* Get max multiplier - as we always did.
+	 * Longhaul MSR is useful only when voltage scaling is enabled.
+	 * C3 is booting at max anyway. */
+	maxmult = mult;
+	/* Get min multiplier */
+	switch (cpu_model) {
+	case CPU_NEHEMIAH:
+		minmult = 50;
+		break;
+	case CPU_NEHEMIAH_C:
+		minmult = 40;
+		break;
+	default:
+		minmult = 30;
+		break;
+	}
+
+	pr_debug("MinMult:%d.%dx MaxMult:%d.%dx\n",
+		 minmult/10, minmult%10, maxmult/10, maxmult%10);
+
+	highest_speed = calc_speed(maxmult);
+	lowest_speed = calc_speed(minmult);
+	pr_debug("FSB:%dMHz  Lowest speed: %s   Highest speed:%s\n", fsb,
+		 print_speed(lowest_speed/1000),
+		 print_speed(highest_speed/1000));
+
+	if (lowest_speed == highest_speed) {
+		printk(KERN_INFO PFX "highestspeed == lowest, aborting.\n");
+		return -EINVAL;
+	}
+	if (lowest_speed > highest_speed) {
+		printk(KERN_INFO PFX "nonsense! lowest (%d > %d) !\n",
+			lowest_speed, highest_speed);
+		return -EINVAL;
+	}
+
+	longhaul_table = kmalloc((numscales + 1) * sizeof(*longhaul_table),
+			GFP_KERNEL);
+	if (!longhaul_table)
+		return -ENOMEM;
+
+	for (j = 0; j < numscales; j++) {
+		ratio = mults[j];
+		if (ratio == -1)
+			continue;
+		if (ratio > maxmult || ratio < minmult)
+			continue;
+		longhaul_table[k].frequency = calc_speed(ratio);
+		longhaul_table[k].index	= j;
+		k++;
+	}
+	if (k <= 1) {
+		kfree(longhaul_table);
+		return -ENODEV;
+	}
+	/* Sort */
+	for (j = 0; j < k - 1; j++) {
+		unsigned int min_f, min_i;
+		min_f = longhaul_table[j].frequency;
+		min_i = j;
+		for (i = j + 1; i < k; i++) {
+			if (longhaul_table[i].frequency < min_f) {
+				min_f = longhaul_table[i].frequency;
+				min_i = i;
+			}
+		}
+		if (min_i != j) {
+			swap(longhaul_table[j].frequency,
+			     longhaul_table[min_i].frequency);
+			swap(longhaul_table[j].index,
+			     longhaul_table[min_i].index);
+		}
+	}
+
+	longhaul_table[k].frequency = CPUFREQ_TABLE_END;
+
+	/* Find index we are running on */
+	for (j = 0; j < k; j++) {
+		if (mults[longhaul_table[j].index & 0x1f] == mult) {
+			longhaul_index = j;
+			break;
+		}
+	}
+	return 0;
+}
+
+
+static void __cpuinit longhaul_setup_voltagescaling(void)
+{
+	union msr_longhaul longhaul;
+	struct mV_pos minvid, maxvid, vid;
+	unsigned int j, speed, pos, kHz_step, numvscales;
+	int min_vid_speed;
+
+	rdmsrl(MSR_VIA_LONGHAUL, longhaul.val);
+	if (!(longhaul.bits.RevisionID & 1)) {
+		printk(KERN_INFO PFX "Voltage scaling not supported by CPU.\n");
+		return;
+	}
+
+	if (!longhaul.bits.VRMRev) {
+		printk(KERN_INFO PFX "VRM 8.5\n");
+		vrm_mV_table = &vrm85_mV[0];
+		mV_vrm_table = &mV_vrm85[0];
+	} else {
+		printk(KERN_INFO PFX "Mobile VRM\n");
+		if (cpu_model < CPU_NEHEMIAH)
+			return;
+		vrm_mV_table = &mobilevrm_mV[0];
+		mV_vrm_table = &mV_mobilevrm[0];
+	}
+
+	minvid = vrm_mV_table[longhaul.bits.MinimumVID];
+	maxvid = vrm_mV_table[longhaul.bits.MaximumVID];
+
+	if (minvid.mV == 0 || maxvid.mV == 0 || minvid.mV > maxvid.mV) {
+		printk(KERN_INFO PFX "Bogus values Min:%d.%03d Max:%d.%03d. "
+					"Voltage scaling disabled.\n",
+					minvid.mV/1000, minvid.mV%1000,
+					maxvid.mV/1000, maxvid.mV%1000);
+		return;
+	}
+
+	if (minvid.mV == maxvid.mV) {
+		printk(KERN_INFO PFX "Claims to support voltage scaling but "
+				"min & max are both %d.%03d. "
+				"Voltage scaling disabled\n",
+				maxvid.mV/1000, maxvid.mV%1000);
+		return;
+	}
+
+	/* How many voltage steps*/
+	numvscales = maxvid.pos - minvid.pos + 1;
+	printk(KERN_INFO PFX
+		"Max VID=%d.%03d  "
+		"Min VID=%d.%03d, "
+		"%d possible voltage scales\n",
+		maxvid.mV/1000, maxvid.mV%1000,
+		minvid.mV/1000, minvid.mV%1000,
+		numvscales);
+
+	/* Calculate max frequency at min voltage */
+	j = longhaul.bits.MinMHzBR;
+	if (longhaul.bits.MinMHzBR4)
+		j += 16;
+	min_vid_speed = eblcr[j];
+	if (min_vid_speed == -1)
+		return;
+	switch (longhaul.bits.MinMHzFSB) {
+	case 0:
+		min_vid_speed *= 13333;
+		break;
+	case 1:
+		min_vid_speed *= 10000;
+		break;
+	case 3:
+		min_vid_speed *= 6666;
+		break;
+	default:
+		return;
+		break;
+	}
+	if (min_vid_speed >= highest_speed)
+		return;
+	/* Calculate kHz for one voltage step */
+	kHz_step = (highest_speed - min_vid_speed) / numvscales;
+
+	j = 0;
+	while (longhaul_table[j].frequency != CPUFREQ_TABLE_END) {
+		speed = longhaul_table[j].frequency;
+		if (speed > min_vid_speed)
+			pos = (speed - min_vid_speed) / kHz_step + minvid.pos;
+		else
+			pos = minvid.pos;
+		longhaul_table[j].index |= mV_vrm_table[pos] << 8;
+		vid = vrm_mV_table[mV_vrm_table[pos]];
+		printk(KERN_INFO PFX "f: %d kHz, index: %d, vid: %d mV\n",
+				speed, j, vid.mV);
+		j++;
+	}
+
+	can_scale_voltage = 1;
+	printk(KERN_INFO PFX "Voltage scaling enabled.\n");
+}
+
+
+static int longhaul_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, longhaul_table);
+}
+
+
+static int longhaul_target(struct cpufreq_policy *policy,
+			    unsigned int target_freq, unsigned int relation)
+{
+	unsigned int table_index = 0;
+	unsigned int i;
+	unsigned int dir = 0;
+	u8 vid, current_vid;
+
+	if (cpufreq_frequency_table_target(policy, longhaul_table, target_freq,
+				relation, &table_index))
+		return -EINVAL;
+
+	/* Don't set same frequency again */
+	if (longhaul_index == table_index)
+		return 0;
+
+	if (!can_scale_voltage)
+		longhaul_setstate(table_index);
+	else {
+		/* On test system voltage transitions exceeding single
+		 * step up or down were turning motherboard off. Both
+		 * "ondemand" and "userspace" are unsafe. C7 is doing
+		 * this in hardware, C3 is old and we need to do this
+		 * in software. */
+		i = longhaul_index;
+		current_vid = (longhaul_table[longhaul_index].index >> 8);
+		current_vid &= 0x1f;
+		if (table_index > longhaul_index)
+			dir = 1;
+		while (i != table_index) {
+			vid = (longhaul_table[i].index >> 8) & 0x1f;
+			if (vid != current_vid) {
+				longhaul_setstate(i);
+				current_vid = vid;
+				msleep(200);
+			}
+			if (dir)
+				i++;
+			else
+				i--;
+		}
+		longhaul_setstate(table_index);
+	}
+	longhaul_index = table_index;
+	return 0;
+}
+
+
+static unsigned int longhaul_get(unsigned int cpu)
+{
+	if (cpu)
+		return 0;
+	return calc_speed(longhaul_get_cpu_mult());
+}
+
+static acpi_status longhaul_walk_callback(acpi_handle obj_handle,
+					  u32 nesting_level,
+					  void *context, void **return_value)
+{
+	struct acpi_device *d;
+
+	if (acpi_bus_get_device(obj_handle, &d))
+		return 0;
+
+	*return_value = acpi_driver_data(d);
+	return 1;
+}
+
+/* VIA don't support PM2 reg, but have something similar */
+static int enable_arbiter_disable(void)
+{
+	struct pci_dev *dev;
+	int status = 1;
+	int reg;
+	u8 pci_cmd;
+
+	/* Find PLE133 host bridge */
+	reg = 0x78;
+	dev = pci_get_device(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8601_0,
+			     NULL);
+	/* Find PM133/VT8605 host bridge */
+	if (dev == NULL)
+		dev = pci_get_device(PCI_VENDOR_ID_VIA,
+				     PCI_DEVICE_ID_VIA_8605_0, NULL);
+	/* Find CLE266 host bridge */
+	if (dev == NULL) {
+		reg = 0x76;
+		dev = pci_get_device(PCI_VENDOR_ID_VIA,
+				     PCI_DEVICE_ID_VIA_862X_0, NULL);
+		/* Find CN400 V-Link host bridge */
+		if (dev == NULL)
+			dev = pci_get_device(PCI_VENDOR_ID_VIA, 0x7259, NULL);
+	}
+	if (dev != NULL) {
+		/* Enable access to port 0x22 */
+		pci_read_config_byte(dev, reg, &pci_cmd);
+		if (!(pci_cmd & 1<<7)) {
+			pci_cmd |= 1<<7;
+			pci_write_config_byte(dev, reg, pci_cmd);
+			pci_read_config_byte(dev, reg, &pci_cmd);
+			if (!(pci_cmd & 1<<7)) {
+				printk(KERN_ERR PFX
+					"Can't enable access to port 0x22.\n");
+				status = 0;
+			}
+		}
+		pci_dev_put(dev);
+		return status;
+	}
+	return 0;
+}
+
+static int longhaul_setup_southbridge(void)
+{
+	struct pci_dev *dev;
+	u8 pci_cmd;
+
+	/* Find VT8235 southbridge */
+	dev = pci_get_device(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8235, NULL);
+	if (dev == NULL)
+		/* Find VT8237 southbridge */
+		dev = pci_get_device(PCI_VENDOR_ID_VIA,
+				     PCI_DEVICE_ID_VIA_8237, NULL);
+	if (dev != NULL) {
+		/* Set transition time to max */
+		pci_read_config_byte(dev, 0xec, &pci_cmd);
+		pci_cmd &= ~(1 << 2);
+		pci_write_config_byte(dev, 0xec, pci_cmd);
+		pci_read_config_byte(dev, 0xe4, &pci_cmd);
+		pci_cmd &= ~(1 << 7);
+		pci_write_config_byte(dev, 0xe4, pci_cmd);
+		pci_read_config_byte(dev, 0xe5, &pci_cmd);
+		pci_cmd |= 1 << 7;
+		pci_write_config_byte(dev, 0xe5, pci_cmd);
+		/* Get address of ACPI registers block*/
+		pci_read_config_byte(dev, 0x81, &pci_cmd);
+		if (pci_cmd & 1 << 7) {
+			pci_read_config_dword(dev, 0x88, &acpi_regs_addr);
+			acpi_regs_addr &= 0xff00;
+			printk(KERN_INFO PFX "ACPI I/O at 0x%x\n",
+					acpi_regs_addr);
+		}
+
+		pci_dev_put(dev);
+		return 1;
+	}
+	return 0;
+}
+
+static int __cpuinit longhaul_cpu_init(struct cpufreq_policy *policy)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	char *cpuname = NULL;
+	int ret;
+	u32 lo, hi;
+
+	/* Check what we have on this motherboard */
+	switch (c->x86_model) {
+	case 6:
+		cpu_model = CPU_SAMUEL;
+		cpuname = "C3 'Samuel' [C5A]";
+		longhaul_version = TYPE_LONGHAUL_V1;
+		memcpy(mults, samuel1_mults, sizeof(samuel1_mults));
+		memcpy(eblcr, samuel1_eblcr, sizeof(samuel1_eblcr));
+		break;
+
+	case 7:
+		switch (c->x86_mask) {
+		case 0:
+			longhaul_version = TYPE_LONGHAUL_V1;
+			cpu_model = CPU_SAMUEL2;
+			cpuname = "C3 'Samuel 2' [C5B]";
+			/* Note, this is not a typo, early Samuel2's had
+			 * Samuel1 ratios. */
+			memcpy(mults, samuel1_mults, sizeof(samuel1_mults));
+			memcpy(eblcr, samuel2_eblcr, sizeof(samuel2_eblcr));
+			break;
+		case 1 ... 15:
+			longhaul_version = TYPE_LONGHAUL_V2;
+			if (c->x86_mask < 8) {
+				cpu_model = CPU_SAMUEL2;
+				cpuname = "C3 'Samuel 2' [C5B]";
+			} else {
+				cpu_model = CPU_EZRA;
+				cpuname = "C3 'Ezra' [C5C]";
+			}
+			memcpy(mults, ezra_mults, sizeof(ezra_mults));
+			memcpy(eblcr, ezra_eblcr, sizeof(ezra_eblcr));
+			break;
+		}
+		break;
+
+	case 8:
+		cpu_model = CPU_EZRA_T;
+		cpuname = "C3 'Ezra-T' [C5M]";
+		longhaul_version = TYPE_POWERSAVER;
+		numscales = 32;
+		memcpy(mults, ezrat_mults, sizeof(ezrat_mults));
+		memcpy(eblcr, ezrat_eblcr, sizeof(ezrat_eblcr));
+		break;
+
+	case 9:
+		longhaul_version = TYPE_POWERSAVER;
+		numscales = 32;
+		memcpy(mults, nehemiah_mults, sizeof(nehemiah_mults));
+		memcpy(eblcr, nehemiah_eblcr, sizeof(nehemiah_eblcr));
+		switch (c->x86_mask) {
+		case 0 ... 1:
+			cpu_model = CPU_NEHEMIAH;
+			cpuname = "C3 'Nehemiah A' [C5XLOE]";
+			break;
+		case 2 ... 4:
+			cpu_model = CPU_NEHEMIAH;
+			cpuname = "C3 'Nehemiah B' [C5XLOH]";
+			break;
+		case 5 ... 15:
+			cpu_model = CPU_NEHEMIAH_C;
+			cpuname = "C3 'Nehemiah C' [C5P]";
+			break;
+		}
+		break;
+
+	default:
+		cpuname = "Unknown";
+		break;
+	}
+	/* Check Longhaul ver. 2 */
+	if (longhaul_version == TYPE_LONGHAUL_V2) {
+		rdmsr(MSR_VIA_LONGHAUL, lo, hi);
+		if (lo == 0 && hi == 0)
+			/* Looks like MSR isn't present */
+			longhaul_version = TYPE_LONGHAUL_V1;
+	}
+
+	printk(KERN_INFO PFX "VIA %s CPU detected.  ", cpuname);
+	switch (longhaul_version) {
+	case TYPE_LONGHAUL_V1:
+	case TYPE_LONGHAUL_V2:
+		printk(KERN_CONT "Longhaul v%d supported.\n", longhaul_version);
+		break;
+	case TYPE_POWERSAVER:
+		printk(KERN_CONT "Powersaver supported.\n");
+		break;
+	};
+
+	/* Doesn't hurt */
+	longhaul_setup_southbridge();
+
+	/* Find ACPI data for processor */
+	acpi_walk_namespace(ACPI_TYPE_PROCESSOR, ACPI_ROOT_OBJECT,
+				ACPI_UINT32_MAX, &longhaul_walk_callback, NULL,
+				NULL, (void *)&pr);
+
+	/* Check ACPI support for C3 state */
+	if (pr != NULL && longhaul_version == TYPE_POWERSAVER) {
+		cx = &pr->power.states[ACPI_STATE_C3];
+		if (cx->address > 0 && cx->latency <= 1000)
+			longhaul_flags |= USE_ACPI_C3;
+	}
+	/* Disable if it isn't working */
+	if (disable_acpi_c3)
+		longhaul_flags &= ~USE_ACPI_C3;
+	/* Check if northbridge is friendly */
+	if (enable_arbiter_disable())
+		longhaul_flags |= USE_NORTHBRIDGE;
+
+	/* Check ACPI support for bus master arbiter disable */
+	if (!(longhaul_flags & USE_ACPI_C3
+	     || longhaul_flags & USE_NORTHBRIDGE)
+	    && ((pr == NULL) || !(pr->flags.bm_control))) {
+		printk(KERN_ERR PFX
+			"No ACPI support. Unsupported northbridge.\n");
+		return -ENODEV;
+	}
+
+	if (longhaul_flags & USE_NORTHBRIDGE)
+		printk(KERN_INFO PFX "Using northbridge support.\n");
+	if (longhaul_flags & USE_ACPI_C3)
+		printk(KERN_INFO PFX "Using ACPI support.\n");
+
+	ret = longhaul_get_ranges();
+	if (ret != 0)
+		return ret;
+
+	if ((longhaul_version != TYPE_LONGHAUL_V1) && (scale_voltage != 0))
+		longhaul_setup_voltagescaling();
+
+	policy->cpuinfo.transition_latency = 200000;	/* nsec */
+	policy->cur = calc_speed(longhaul_get_cpu_mult());
+
+	ret = cpufreq_frequency_table_cpuinfo(policy, longhaul_table);
+	if (ret)
+		return ret;
+
+	cpufreq_frequency_table_get_attr(longhaul_table, policy->cpu);
+
+	return 0;
+}
+
+static int __devexit longhaul_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+static struct freq_attr *longhaul_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver longhaul_driver = {
+	.verify	= longhaul_verify,
+	.target	= longhaul_target,
+	.get	= longhaul_get,
+	.init	= longhaul_cpu_init,
+	.exit	= __devexit_p(longhaul_cpu_exit),
+	.name	= "longhaul",
+	.owner	= THIS_MODULE,
+	.attr	= longhaul_attr,
+};
+
+
+static int __init longhaul_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	if (c->x86_vendor != X86_VENDOR_CENTAUR || c->x86 != 6)
+		return -ENODEV;
+
+#ifdef CONFIG_SMP
+	if (num_online_cpus() > 1) {
+		printk(KERN_ERR PFX "More than 1 CPU detected, "
+				"longhaul disabled.\n");
+		return -ENODEV;
+	}
+#endif
+#ifdef CONFIG_X86_IO_APIC
+	if (cpu_has_apic) {
+		printk(KERN_ERR PFX "APIC detected. Longhaul is currently "
+				"broken in this configuration.\n");
+		return -ENODEV;
+	}
+#endif
+	switch (c->x86_model) {
+	case 6 ... 9:
+		return cpufreq_register_driver(&longhaul_driver);
+	case 10:
+		printk(KERN_ERR PFX "Use acpi-cpufreq driver for VIA C7\n");
+	default:
+		;
+	}
+
+	return -ENODEV;
+}
+
+
+static void __exit longhaul_exit(void)
+{
+	int i;
+
+	for (i = 0; i < numscales; i++) {
+		if (mults[i] == maxmult) {
+			longhaul_setstate(i);
+			break;
+		}
+	}
+
+	cpufreq_unregister_driver(&longhaul_driver);
+	kfree(longhaul_table);
+}
+
+/* Even if BIOS is exporting ACPI C3 state, and it is used
+ * with success when CPU is idle, this state doesn't
+ * trigger frequency transition in some cases. */
+module_param(disable_acpi_c3, int, 0644);
+MODULE_PARM_DESC(disable_acpi_c3, "Don't use ACPI C3 support");
+/* Change CPU voltage with frequency. Very useful to save
+ * power, but most VIA C3 processors aren't supporting it. */
+module_param(scale_voltage, int, 0644);
+MODULE_PARM_DESC(scale_voltage, "Scale voltage of processor");
+/* Force revision key to 0 for processors which doesn't
+ * support voltage scaling, but are introducing itself as
+ * such. */
+module_param(revid_errata, int, 0644);
+MODULE_PARM_DESC(revid_errata, "Ignore CPU Revision ID");
+
+MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_DESCRIPTION("Longhaul driver for VIA Cyrix processors.");
+MODULE_LICENSE("GPL");
+
+late_initcall(longhaul_init);
+module_exit(longhaul_exit);
diff --git a/drivers/cpufreq/longhaul.h b/drivers/cpufreq/longhaul.h
new file mode 100644
index 00000000..cbf48fbc
--- /dev/null
+++ b/drivers/cpufreq/longhaul.h
@@ -0,0 +1,353 @@
+/*
+ *  longhaul.h
+ *  (C) 2003 Dave Jones.
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  VIA-specific information
+ */
+
+union msr_bcr2 {
+	struct {
+		unsigned Reseved:19,	// 18:0
+		ESOFTBF:1,		// 19
+		Reserved2:3,		// 22:20
+		CLOCKMUL:4,		// 26:23
+		Reserved3:5;		// 31:27
+	} bits;
+	unsigned long val;
+};
+
+union msr_longhaul {
+	struct {
+		unsigned RevisionID:4,	// 3:0
+		RevisionKey:4,		// 7:4
+		EnableSoftBusRatio:1,	// 8
+		EnableSoftVID:1,	// 9
+		EnableSoftBSEL:1,	// 10
+		Reserved:3,		// 11:13
+		SoftBusRatio4:1,	// 14
+		VRMRev:1,		// 15
+		SoftBusRatio:4,		// 19:16
+		SoftVID:5,		// 24:20
+		Reserved2:3,		// 27:25
+		SoftBSEL:2,		// 29:28
+		Reserved3:2,		// 31:30
+		MaxMHzBR:4,		// 35:32
+		MaximumVID:5,		// 40:36
+		MaxMHzFSB:2,		// 42:41
+		MaxMHzBR4:1,		// 43
+		Reserved4:4,		// 47:44
+		MinMHzBR:4,		// 51:48
+		MinimumVID:5,		// 56:52
+		MinMHzFSB:2,		// 58:57
+		MinMHzBR4:1,		// 59
+		Reserved5:4;		// 63:60
+	} bits;
+	unsigned long long val;
+};
+
+/*
+ * Clock ratio tables. Div/Mod by 10 to get ratio.
+ * The eblcr values specify the ratio read from the CPU.
+ * The mults values specify what to write to the CPU.
+ */
+
+/*
+ * VIA C3 Samuel 1  & Samuel 2 (stepping 0)
+ */
+static const int __cpuinitdata samuel1_mults[16] = {
+	-1, /* 0000 -> RESERVED */
+	30, /* 0001 ->  3.0x */
+	40, /* 0010 ->  4.0x */
+	-1, /* 0011 -> RESERVED */
+	-1, /* 0100 -> RESERVED */
+	35, /* 0101 ->  3.5x */
+	45, /* 0110 ->  4.5x */
+	55, /* 0111 ->  5.5x */
+	60, /* 1000 ->  6.0x */
+	70, /* 1001 ->  7.0x */
+	80, /* 1010 ->  8.0x */
+	50, /* 1011 ->  5.0x */
+	65, /* 1100 ->  6.5x */
+	75, /* 1101 ->  7.5x */
+	-1, /* 1110 -> RESERVED */
+	-1, /* 1111 -> RESERVED */
+};
+
+static const int __cpuinitdata samuel1_eblcr[16] = {
+	50, /* 0000 -> RESERVED */
+	30, /* 0001 ->  3.0x */
+	40, /* 0010 ->  4.0x */
+	-1, /* 0011 -> RESERVED */
+	55, /* 0100 ->  5.5x */
+	35, /* 0101 ->  3.5x */
+	45, /* 0110 ->  4.5x */
+	-1, /* 0111 -> RESERVED */
+	-1, /* 1000 -> RESERVED */
+	70, /* 1001 ->  7.0x */
+	80, /* 1010 ->  8.0x */
+	60, /* 1011 ->  6.0x */
+	-1, /* 1100 -> RESERVED */
+	75, /* 1101 ->  7.5x */
+	-1, /* 1110 -> RESERVED */
+	65, /* 1111 ->  6.5x */
+};
+
+/*
+ * VIA C3 Samuel2 Stepping 1->15
+ */
+static const int __cpuinitdata samuel2_eblcr[16] = {
+	50,  /* 0000 ->  5.0x */
+	30,  /* 0001 ->  3.0x */
+	40,  /* 0010 ->  4.0x */
+	100, /* 0011 -> 10.0x */
+	55,  /* 0100 ->  5.5x */
+	35,  /* 0101 ->  3.5x */
+	45,  /* 0110 ->  4.5x */
+	110, /* 0111 -> 11.0x */
+	90,  /* 1000 ->  9.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	60,  /* 1011 ->  6.0x */
+	120, /* 1100 -> 12.0x */
+	75,  /* 1101 ->  7.5x */
+	130, /* 1110 -> 13.0x */
+	65,  /* 1111 ->  6.5x */
+};
+
+/*
+ * VIA C3 Ezra
+ */
+static const int __cpuinitdata ezra_mults[16] = {
+	100, /* 0000 -> 10.0x */
+	30,  /* 0001 ->  3.0x */
+	40,  /* 0010 ->  4.0x */
+	90,  /* 0011 ->  9.0x */
+	95,  /* 0100 ->  9.5x */
+	35,  /* 0101 ->  3.5x */
+	45,  /* 0110 ->  4.5x */
+	55,  /* 0111 ->  5.5x */
+	60,  /* 1000 ->  6.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	50,  /* 1011 ->  5.0x */
+	65,  /* 1100 ->  6.5x */
+	75,  /* 1101 ->  7.5x */
+	85,  /* 1110 ->  8.5x */
+	120, /* 1111 -> 12.0x */
+};
+
+static const int __cpuinitdata ezra_eblcr[16] = {
+	50,  /* 0000 ->  5.0x */
+	30,  /* 0001 ->  3.0x */
+	40,  /* 0010 ->  4.0x */
+	100, /* 0011 -> 10.0x */
+	55,  /* 0100 ->  5.5x */
+	35,  /* 0101 ->  3.5x */
+	45,  /* 0110 ->  4.5x */
+	95,  /* 0111 ->  9.5x */
+	90,  /* 1000 ->  9.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	60,  /* 1011 ->  6.0x */
+	120, /* 1100 -> 12.0x */
+	75,  /* 1101 ->  7.5x */
+	85,  /* 1110 ->  8.5x */
+	65,  /* 1111 ->  6.5x */
+};
+
+/*
+ * VIA C3 (Ezra-T) [C5M].
+ */
+static const int __cpuinitdata ezrat_mults[32] = {
+	100, /* 0000 -> 10.0x */
+	30,  /* 0001 ->  3.0x */
+	40,  /* 0010 ->  4.0x */
+	90,  /* 0011 ->  9.0x */
+	95,  /* 0100 ->  9.5x */
+	35,  /* 0101 ->  3.5x */
+	45,  /* 0110 ->  4.5x */
+	55,  /* 0111 ->  5.5x */
+	60,  /* 1000 ->  6.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	50,  /* 1011 ->  5.0x */
+	65,  /* 1100 ->  6.5x */
+	75,  /* 1101 ->  7.5x */
+	85,  /* 1110 ->  8.5x */
+	120, /* 1111 ->  12.0x */
+
+	-1,  /* 0000 -> RESERVED (10.0x) */
+	110, /* 0001 -> 11.0x */
+	-1, /* 0010 -> 12.0x */
+	-1,  /* 0011 -> RESERVED (9.0x)*/
+	105, /* 0100 -> 10.5x */
+	115, /* 0101 -> 11.5x */
+	125, /* 0110 -> 12.5x */
+	135, /* 0111 -> 13.5x */
+	140, /* 1000 -> 14.0x */
+	150, /* 1001 -> 15.0x */
+	160, /* 1010 -> 16.0x */
+	130, /* 1011 -> 13.0x */
+	145, /* 1100 -> 14.5x */
+	155, /* 1101 -> 15.5x */
+	-1,  /* 1110 -> RESERVED (13.0x) */
+	-1,  /* 1111 -> RESERVED (12.0x) */
+};
+
+static const int __cpuinitdata ezrat_eblcr[32] = {
+	50,  /* 0000 ->  5.0x */
+	30,  /* 0001 ->  3.0x */
+	40,  /* 0010 ->  4.0x */
+	100, /* 0011 -> 10.0x */
+	55,  /* 0100 ->  5.5x */
+	35,  /* 0101 ->  3.5x */
+	45,  /* 0110 ->  4.5x */
+	95,  /* 0111 ->  9.5x */
+	90,  /* 1000 ->  9.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	60,  /* 1011 ->  6.0x */
+	120, /* 1100 -> 12.0x */
+	75,  /* 1101 ->  7.5x */
+	85,  /* 1110 ->  8.5x */
+	65,  /* 1111 ->  6.5x */
+
+	-1,  /* 0000 -> RESERVED (9.0x) */
+	110, /* 0001 -> 11.0x */
+	120, /* 0010 -> 12.0x */
+	-1,  /* 0011 -> RESERVED (10.0x)*/
+	135, /* 0100 -> 13.5x */
+	115, /* 0101 -> 11.5x */
+	125, /* 0110 -> 12.5x */
+	105, /* 0111 -> 10.5x */
+	130, /* 1000 -> 13.0x */
+	150, /* 1001 -> 15.0x */
+	160, /* 1010 -> 16.0x */
+	140, /* 1011 -> 14.0x */
+	-1,  /* 1100 -> RESERVED (12.0x) */
+	155, /* 1101 -> 15.5x */
+	-1,  /* 1110 -> RESERVED (13.0x) */
+	145, /* 1111 -> 14.5x */
+};
+
+/*
+ * VIA C3 Nehemiah */
+
+static const int __cpuinitdata nehemiah_mults[32] = {
+	100, /* 0000 -> 10.0x */
+	-1, /* 0001 -> 16.0x */
+	40,  /* 0010 ->  4.0x */
+	90,  /* 0011 ->  9.0x */
+	95,  /* 0100 ->  9.5x */
+	-1,  /* 0101 ->  RESERVED */
+	45,  /* 0110 ->  4.5x */
+	55,  /* 0111 ->  5.5x */
+	60,  /* 1000 ->  6.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	50,  /* 1011 ->  5.0x */
+	65,  /* 1100 ->  6.5x */
+	75,  /* 1101 ->  7.5x */
+	85,  /* 1110 ->  8.5x */
+	120, /* 1111 -> 12.0x */
+	-1, /* 0000 -> 10.0x */
+	110, /* 0001 -> 11.0x */
+	-1, /* 0010 -> 12.0x */
+	-1,  /* 0011 ->  9.0x */
+	105, /* 0100 -> 10.5x */
+	115, /* 0101 -> 11.5x */
+	125, /* 0110 -> 12.5x */
+	135, /* 0111 -> 13.5x */
+	140, /* 1000 -> 14.0x */
+	150, /* 1001 -> 15.0x */
+	160, /* 1010 -> 16.0x */
+	130, /* 1011 -> 13.0x */
+	145, /* 1100 -> 14.5x */
+	155, /* 1101 -> 15.5x */
+	-1,  /* 1110 -> RESERVED (13.0x) */
+	-1, /* 1111 -> 12.0x */
+};
+
+static const int __cpuinitdata nehemiah_eblcr[32] = {
+	50,  /* 0000 ->  5.0x */
+	160, /* 0001 -> 16.0x */
+	40,  /* 0010 ->  4.0x */
+	100, /* 0011 -> 10.0x */
+	55,  /* 0100 ->  5.5x */
+	-1,  /* 0101 ->  RESERVED */
+	45,  /* 0110 ->  4.5x */
+	95,  /* 0111 ->  9.5x */
+	90,  /* 1000 ->  9.0x */
+	70,  /* 1001 ->  7.0x */
+	80,  /* 1010 ->  8.0x */
+	60,  /* 1011 ->  6.0x */
+	120, /* 1100 -> 12.0x */
+	75,  /* 1101 ->  7.5x */
+	85,  /* 1110 ->  8.5x */
+	65,  /* 1111 ->  6.5x */
+	90,  /* 0000 ->  9.0x */
+	110, /* 0001 -> 11.0x */
+	120, /* 0010 -> 12.0x */
+	100, /* 0011 -> 10.0x */
+	135, /* 0100 -> 13.5x */
+	115, /* 0101 -> 11.5x */
+	125, /* 0110 -> 12.5x */
+	105, /* 0111 -> 10.5x */
+	130, /* 1000 -> 13.0x */
+	150, /* 1001 -> 15.0x */
+	160, /* 1010 -> 16.0x */
+	140, /* 1011 -> 14.0x */
+	120, /* 1100 -> 12.0x */
+	155, /* 1101 -> 15.5x */
+	-1,  /* 1110 -> RESERVED (13.0x) */
+	145 /* 1111 -> 14.5x */
+};
+
+/*
+ * Voltage scales. Div/Mod by 1000 to get actual voltage.
+ * Which scale to use depends on the VRM type in use.
+ */
+
+struct mV_pos {
+	unsigned short mV;
+	unsigned short pos;
+};
+
+static const struct mV_pos __cpuinitdata vrm85_mV[32] = {
+	{1250, 8},	{1200, 6},	{1150, 4},	{1100, 2},
+	{1050, 0},	{1800, 30},	{1750, 28},	{1700, 26},
+	{1650, 24},	{1600, 22},	{1550, 20},	{1500, 18},
+	{1450, 16},	{1400, 14},	{1350, 12},	{1300, 10},
+	{1275, 9},	{1225, 7},	{1175, 5},	{1125, 3},
+	{1075, 1},	{1825, 31},	{1775, 29},	{1725, 27},
+	{1675, 25},	{1625, 23},	{1575, 21},	{1525, 19},
+	{1475, 17},	{1425, 15},	{1375, 13},	{1325, 11}
+};
+
+static const unsigned char __cpuinitdata mV_vrm85[32] = {
+	0x04,	0x14,	0x03,	0x13,	0x02,	0x12,	0x01,	0x11,
+	0x00,	0x10,	0x0f,	0x1f,	0x0e,	0x1e,	0x0d,	0x1d,
+	0x0c,	0x1c,	0x0b,	0x1b,	0x0a,	0x1a,	0x09,	0x19,
+	0x08,	0x18,	0x07,	0x17,	0x06,	0x16,	0x05,	0x15
+};
+
+static const struct mV_pos __cpuinitdata mobilevrm_mV[32] = {
+	{1750, 31},	{1700, 30},	{1650, 29},	{1600, 28},
+	{1550, 27},	{1500, 26},	{1450, 25},	{1400, 24},
+	{1350, 23},	{1300, 22},	{1250, 21},	{1200, 20},
+	{1150, 19},	{1100, 18},	{1050, 17},	{1000, 16},
+	{975, 15},	{950, 14},	{925, 13},	{900, 12},
+	{875, 11},	{850, 10},	{825, 9},	{800, 8},
+	{775, 7},	{750, 6},	{725, 5},	{700, 4},
+	{675, 3},	{650, 2},	{625, 1},	{600, 0}
+};
+
+static const unsigned char __cpuinitdata mV_mobilevrm[32] = {
+	0x1f,	0x1e,	0x1d,	0x1c,	0x1b,	0x1a,	0x19,	0x18,
+	0x17,	0x16,	0x15,	0x14,	0x13,	0x12,	0x11,	0x10,
+	0x0f,	0x0e,	0x0d,	0x0c,	0x0b,	0x0a,	0x09,	0x08,
+	0x07,	0x06,	0x05,	0x04,	0x03,	0x02,	0x01,	0x00
+};
+
diff --git a/drivers/cpufreq/longrun.c b/drivers/cpufreq/longrun.c
new file mode 100644
index 00000000..34ea359b
--- /dev/null
+++ b/drivers/cpufreq/longrun.c
@@ -0,0 +1,324 @@
+/*
+ * (C) 2002 - 2003  Dominik Brodowski <linux@brodo.de>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/timex.h>
+
+#include <asm/msr.h>
+#include <asm/processor.h>
+
+static struct cpufreq_driver	longrun_driver;
+
+/**
+ * longrun_{low,high}_freq is needed for the conversion of cpufreq kHz
+ * values into per cent values. In TMTA microcode, the following is valid:
+ * performance_pctg = (current_freq - low_freq)/(high_freq - low_freq)
+ */
+static unsigned int longrun_low_freq, longrun_high_freq;
+
+
+/**
+ * longrun_get_policy - get the current LongRun policy
+ * @policy: struct cpufreq_policy where current policy is written into
+ *
+ * Reads the current LongRun policy by access to MSR_TMTA_LONGRUN_FLAGS
+ * and MSR_TMTA_LONGRUN_CTRL
+ */
+static void __cpuinit longrun_get_policy(struct cpufreq_policy *policy)
+{
+	u32 msr_lo, msr_hi;
+
+	rdmsr(MSR_TMTA_LONGRUN_FLAGS, msr_lo, msr_hi);
+	pr_debug("longrun flags are %x - %x\n", msr_lo, msr_hi);
+	if (msr_lo & 0x01)
+		policy->policy = CPUFREQ_POLICY_PERFORMANCE;
+	else
+		policy->policy = CPUFREQ_POLICY_POWERSAVE;
+
+	rdmsr(MSR_TMTA_LONGRUN_CTRL, msr_lo, msr_hi);
+	pr_debug("longrun ctrl is %x - %x\n", msr_lo, msr_hi);
+	msr_lo &= 0x0000007F;
+	msr_hi &= 0x0000007F;
+
+	if (longrun_high_freq <= longrun_low_freq) {
+		/* Assume degenerate Longrun table */
+		policy->min = policy->max = longrun_high_freq;
+	} else {
+		policy->min = longrun_low_freq + msr_lo *
+			((longrun_high_freq - longrun_low_freq) / 100);
+		policy->max = longrun_low_freq + msr_hi *
+			((longrun_high_freq - longrun_low_freq) / 100);
+	}
+	policy->cpu = 0;
+}
+
+
+/**
+ * longrun_set_policy - sets a new CPUFreq policy
+ * @policy: new policy
+ *
+ * Sets a new CPUFreq policy on LongRun-capable processors. This function
+ * has to be called with cpufreq_driver locked.
+ */
+static int longrun_set_policy(struct cpufreq_policy *policy)
+{
+	u32 msr_lo, msr_hi;
+	u32 pctg_lo, pctg_hi;
+
+	if (!policy)
+		return -EINVAL;
+
+	if (longrun_high_freq <= longrun_low_freq) {
+		/* Assume degenerate Longrun table */
+		pctg_lo = pctg_hi = 100;
+	} else {
+		pctg_lo = (policy->min - longrun_low_freq) /
+			((longrun_high_freq - longrun_low_freq) / 100);
+		pctg_hi = (policy->max - longrun_low_freq) /
+			((longrun_high_freq - longrun_low_freq) / 100);
+	}
+
+	if (pctg_hi > 100)
+		pctg_hi = 100;
+	if (pctg_lo > pctg_hi)
+		pctg_lo = pctg_hi;
+
+	/* performance or economy mode */
+	rdmsr(MSR_TMTA_LONGRUN_FLAGS, msr_lo, msr_hi);
+	msr_lo &= 0xFFFFFFFE;
+	switch (policy->policy) {
+	case CPUFREQ_POLICY_PERFORMANCE:
+		msr_lo |= 0x00000001;
+		break;
+	case CPUFREQ_POLICY_POWERSAVE:
+		break;
+	}
+	wrmsr(MSR_TMTA_LONGRUN_FLAGS, msr_lo, msr_hi);
+
+	/* lower and upper boundary */
+	rdmsr(MSR_TMTA_LONGRUN_CTRL, msr_lo, msr_hi);
+	msr_lo &= 0xFFFFFF80;
+	msr_hi &= 0xFFFFFF80;
+	msr_lo |= pctg_lo;
+	msr_hi |= pctg_hi;
+	wrmsr(MSR_TMTA_LONGRUN_CTRL, msr_lo, msr_hi);
+
+	return 0;
+}
+
+
+/**
+ * longrun_verify_poliy - verifies a new CPUFreq policy
+ * @policy: the policy to verify
+ *
+ * Validates a new CPUFreq policy. This function has to be called with
+ * cpufreq_driver locked.
+ */
+static int longrun_verify_policy(struct cpufreq_policy *policy)
+{
+	if (!policy)
+		return -EINVAL;
+
+	policy->cpu = 0;
+	cpufreq_verify_within_limits(policy,
+		policy->cpuinfo.min_freq,
+		policy->cpuinfo.max_freq);
+
+	if ((policy->policy != CPUFREQ_POLICY_POWERSAVE) &&
+	    (policy->policy != CPUFREQ_POLICY_PERFORMANCE))
+		return -EINVAL;
+
+	return 0;
+}
+
+static unsigned int longrun_get(unsigned int cpu)
+{
+	u32 eax, ebx, ecx, edx;
+
+	if (cpu)
+		return 0;
+
+	cpuid(0x80860007, &eax, &ebx, &ecx, &edx);
+	pr_debug("cpuid eax is %u\n", eax);
+
+	return eax * 1000;
+}
+
+/**
+ * longrun_determine_freqs - determines the lowest and highest possible core frequency
+ * @low_freq: an int to put the lowest frequency into
+ * @high_freq: an int to put the highest frequency into
+ *
+ * Determines the lowest and highest possible core frequencies on this CPU.
+ * This is necessary to calculate the performance percentage according to
+ * TMTA rules:
+ * performance_pctg = (target_freq - low_freq)/(high_freq - low_freq)
+ */
+static int __cpuinit longrun_determine_freqs(unsigned int *low_freq,
+						      unsigned int *high_freq)
+{
+	u32 msr_lo, msr_hi;
+	u32 save_lo, save_hi;
+	u32 eax, ebx, ecx, edx;
+	u32 try_hi;
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	if (!low_freq || !high_freq)
+		return -EINVAL;
+
+	if (cpu_has(c, X86_FEATURE_LRTI)) {
+		/* if the LongRun Table Interface is present, the
+		 * detection is a bit easier:
+		 * For minimum frequency, read out the maximum
+		 * level (msr_hi), write that into "currently
+		 * selected level", and read out the frequency.
+		 * For maximum frequency, read out level zero.
+		 */
+		/* minimum */
+		rdmsr(MSR_TMTA_LRTI_READOUT, msr_lo, msr_hi);
+		wrmsr(MSR_TMTA_LRTI_READOUT, msr_hi, msr_hi);
+		rdmsr(MSR_TMTA_LRTI_VOLT_MHZ, msr_lo, msr_hi);
+		*low_freq = msr_lo * 1000; /* to kHz */
+
+		/* maximum */
+		wrmsr(MSR_TMTA_LRTI_READOUT, 0, msr_hi);
+		rdmsr(MSR_TMTA_LRTI_VOLT_MHZ, msr_lo, msr_hi);
+		*high_freq = msr_lo * 1000; /* to kHz */
+
+		pr_debug("longrun table interface told %u - %u kHz\n",
+				*low_freq, *high_freq);
+
+		if (*low_freq > *high_freq)
+			*low_freq = *high_freq;
+		return 0;
+	}
+
+	/* set the upper border to the value determined during TSC init */
+	*high_freq = (cpu_khz / 1000);
+	*high_freq = *high_freq * 1000;
+	pr_debug("high frequency is %u kHz\n", *high_freq);
+
+	/* get current borders */
+	rdmsr(MSR_TMTA_LONGRUN_CTRL, msr_lo, msr_hi);
+	save_lo = msr_lo & 0x0000007F;
+	save_hi = msr_hi & 0x0000007F;
+
+	/* if current perf_pctg is larger than 90%, we need to decrease the
+	 * upper limit to make the calculation more accurate.
+	 */
+	cpuid(0x80860007, &eax, &ebx, &ecx, &edx);
+	/* try decreasing in 10% steps, some processors react only
+	 * on some barrier values */
+	for (try_hi = 80; try_hi > 0 && ecx > 90; try_hi -= 10) {
+		/* set to 0 to try_hi perf_pctg */
+		msr_lo &= 0xFFFFFF80;
+		msr_hi &= 0xFFFFFF80;
+		msr_hi |= try_hi;
+		wrmsr(MSR_TMTA_LONGRUN_CTRL, msr_lo, msr_hi);
+
+		/* read out current core MHz and current perf_pctg */
+		cpuid(0x80860007, &eax, &ebx, &ecx, &edx);
+
+		/* restore values */
+		wrmsr(MSR_TMTA_LONGRUN_CTRL, save_lo, save_hi);
+	}
+	pr_debug("percentage is %u %%, freq is %u MHz\n", ecx, eax);
+
+	/* performance_pctg = (current_freq - low_freq)/(high_freq - low_freq)
+	 * eqals
+	 * low_freq * (1 - perf_pctg) = (cur_freq - high_freq * perf_pctg)
+	 *
+	 * high_freq * perf_pctg is stored tempoarily into "ebx".
+	 */
+	ebx = (((cpu_khz / 1000) * ecx) / 100); /* to MHz */
+
+	if ((ecx > 95) || (ecx == 0) || (eax < ebx))
+		return -EIO;
+
+	edx = ((eax - ebx) * 100) / (100 - ecx);
+	*low_freq = edx * 1000; /* back to kHz */
+
+	pr_debug("low frequency is %u kHz\n", *low_freq);
+
+	if (*low_freq > *high_freq)
+		*low_freq = *high_freq;
+
+	return 0;
+}
+
+
+static int __cpuinit longrun_cpu_init(struct cpufreq_policy *policy)
+{
+	int result = 0;
+
+	/* capability check */
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	/* detect low and high frequency */
+	result = longrun_determine_freqs(&longrun_low_freq, &longrun_high_freq);
+	if (result)
+		return result;
+
+	/* cpuinfo and default policy values */
+	policy->cpuinfo.min_freq = longrun_low_freq;
+	policy->cpuinfo.max_freq = longrun_high_freq;
+	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
+	longrun_get_policy(policy);
+
+	return 0;
+}
+
+
+static struct cpufreq_driver longrun_driver = {
+	.flags		= CPUFREQ_CONST_LOOPS,
+	.verify		= longrun_verify_policy,
+	.setpolicy	= longrun_set_policy,
+	.get		= longrun_get,
+	.init		= longrun_cpu_init,
+	.name		= "longrun",
+	.owner		= THIS_MODULE,
+};
+
+
+/**
+ * longrun_init - initializes the Transmeta Crusoe LongRun CPUFreq driver
+ *
+ * Initializes the LongRun support.
+ */
+static int __init longrun_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	if (c->x86_vendor != X86_VENDOR_TRANSMETA ||
+	    !cpu_has(c, X86_FEATURE_LONGRUN))
+		return -ENODEV;
+
+	return cpufreq_register_driver(&longrun_driver);
+}
+
+
+/**
+ * longrun_exit - unregisters LongRun support
+ */
+static void __exit longrun_exit(void)
+{
+	cpufreq_unregister_driver(&longrun_driver);
+}
+
+
+MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("LongRun driver for Transmeta Crusoe and "
+		"Efficeon processors.");
+MODULE_LICENSE("GPL");
+
+module_init(longrun_init);
+module_exit(longrun_exit);
diff --git a/drivers/cpufreq/mperf.c b/drivers/cpufreq/mperf.c
new file mode 100644
index 00000000..911e1930
--- /dev/null
+++ b/drivers/cpufreq/mperf.c
@@ -0,0 +1,51 @@
+#include <linux/kernel.h>
+#include <linux/smp.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/slab.h>
+
+#include "mperf.h"
+
+static DEFINE_PER_CPU(struct aperfmperf, acfreq_old_perf);
+
+/* Called via smp_call_function_single(), on the target CPU */
+static void read_measured_perf_ctrs(void *_cur)
+{
+	struct aperfmperf *am = _cur;
+
+	get_aperfmperf(am);
+}
+
+/*
+ * Return the measured active (C0) frequency on this CPU since last call
+ * to this function.
+ * Input: cpu number
+ * Return: Average CPU frequency in terms of max frequency (zero on error)
+ *
+ * We use IA32_MPERF and IA32_APERF MSRs to get the measured performance
+ * over a period of time, while CPU is in C0 state.
+ * IA32_MPERF counts at the rate of max advertised frequency
+ * IA32_APERF counts at the rate of actual CPU frequency
+ * Only IA32_APERF/IA32_MPERF ratio is architecturally defined and
+ * no meaning should be associated with absolute values of these MSRs.
+ */
+unsigned int cpufreq_get_measured_perf(struct cpufreq_policy *policy,
+					unsigned int cpu)
+{
+	struct aperfmperf perf;
+	unsigned long ratio;
+	unsigned int retval;
+
+	if (smp_call_function_single(cpu, read_measured_perf_ctrs, &perf, 1))
+		return 0;
+
+	ratio = calc_aperfmperf_ratio(&per_cpu(acfreq_old_perf, cpu), &perf);
+	per_cpu(acfreq_old_perf, cpu) = perf;
+
+	retval = (policy->cpuinfo.max_freq * ratio) >> APERFMPERF_SHIFT;
+
+	return retval;
+}
+EXPORT_SYMBOL_GPL(cpufreq_get_measured_perf);
+MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/mperf.h b/drivers/cpufreq/mperf.h
new file mode 100644
index 00000000..5dbf2950
--- /dev/null
+++ b/drivers/cpufreq/mperf.h
@@ -0,0 +1,9 @@
+/*
+ *  (c) 2010 Advanced Micro Devices, Inc.
+ *  Your use of this code is subject to the terms and conditions of the
+ *  GNU general public license version 2. See "COPYING" or
+ *  http://www.gnu.org/licenses/gpl.html
+ */
+
+unsigned int cpufreq_get_measured_perf(struct cpufreq_policy *policy,
+					unsigned int cpu);
diff --git a/drivers/cpufreq/p4-clockmod.c b/drivers/cpufreq/p4-clockmod.c
new file mode 100644
index 00000000..6be3e076
--- /dev/null
+++ b/drivers/cpufreq/p4-clockmod.c
@@ -0,0 +1,329 @@
+/*
+ *	Pentium 4/Xeon CPU on demand clock modulation/speed scaling
+ *	(C) 2002 - 2003 Dominik Brodowski <linux@brodo.de>
+ *	(C) 2002 Zwane Mwaikambo <zwane@commfireservices.com>
+ *	(C) 2002 Arjan van de Ven <arjanv@redhat.com>
+ *	(C) 2002 Tora T. Engstad
+ *	All Rights Reserved
+ *
+ *	This program is free software; you can redistribute it and/or
+ *      modify it under the terms of the GNU General Public License
+ *      as published by the Free Software Foundation; either version
+ *      2 of the License, or (at your option) any later version.
+ *
+ *      The author(s) of this software shall not be held liable for damages
+ *      of any nature resulting due to the use of this software. This
+ *      software is provided AS-IS with no warranties.
+ *
+ *	Date		Errata			Description
+ *	20020525	N44, O17	12.5% or 25% DC causes lockup
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/cpufreq.h>
+#include <linux/cpumask.h>
+#include <linux/timex.h>
+
+#include <asm/processor.h>
+#include <asm/msr.h>
+#include <asm/timer.h>
+
+#include "speedstep-lib.h"
+
+#define PFX	"p4-clockmod: "
+
+/*
+ * Duty Cycle (3bits), note DC_DISABLE is not specified in
+ * intel docs i just use it to mean disable
+ */
+enum {
+	DC_RESV, DC_DFLT, DC_25PT, DC_38PT, DC_50PT,
+	DC_64PT, DC_75PT, DC_88PT, DC_DISABLE
+};
+
+#define DC_ENTRIES	8
+
+
+static int has_N44_O17_errata[NR_CPUS];
+static unsigned int stock_freq;
+static struct cpufreq_driver p4clockmod_driver;
+static unsigned int cpufreq_p4_get(unsigned int cpu);
+
+static int cpufreq_p4_setdc(unsigned int cpu, unsigned int newstate)
+{
+	u32 l, h;
+
+	if (!cpu_online(cpu) ||
+	    (newstate > DC_DISABLE) || (newstate == DC_RESV))
+		return -EINVAL;
+
+	rdmsr_on_cpu(cpu, MSR_IA32_THERM_STATUS, &l, &h);
+
+	if (l & 0x01)
+		pr_debug("CPU#%d currently thermal throttled\n", cpu);
+
+	if (has_N44_O17_errata[cpu] &&
+	    (newstate == DC_25PT || newstate == DC_DFLT))
+		newstate = DC_38PT;
+
+	rdmsr_on_cpu(cpu, MSR_IA32_THERM_CONTROL, &l, &h);
+	if (newstate == DC_DISABLE) {
+		pr_debug("CPU#%d disabling modulation\n", cpu);
+		wrmsr_on_cpu(cpu, MSR_IA32_THERM_CONTROL, l & ~(1<<4), h);
+	} else {
+		pr_debug("CPU#%d setting duty cycle to %d%%\n",
+			cpu, ((125 * newstate) / 10));
+		/* bits 63 - 5	: reserved
+		 * bit  4	: enable/disable
+		 * bits 3-1	: duty cycle
+		 * bit  0	: reserved
+		 */
+		l = (l & ~14);
+		l = l | (1<<4) | ((newstate & 0x7)<<1);
+		wrmsr_on_cpu(cpu, MSR_IA32_THERM_CONTROL, l, h);
+	}
+
+	return 0;
+}
+
+
+static struct cpufreq_frequency_table p4clockmod_table[] = {
+	{DC_RESV, CPUFREQ_ENTRY_INVALID},
+	{DC_DFLT, 0},
+	{DC_25PT, 0},
+	{DC_38PT, 0},
+	{DC_50PT, 0},
+	{DC_64PT, 0},
+	{DC_75PT, 0},
+	{DC_88PT, 0},
+	{DC_DISABLE, 0},
+	{DC_RESV, CPUFREQ_TABLE_END},
+};
+
+
+static int cpufreq_p4_target(struct cpufreq_policy *policy,
+			     unsigned int target_freq,
+			     unsigned int relation)
+{
+	unsigned int    newstate = DC_RESV;
+	struct cpufreq_freqs freqs;
+	int i;
+
+	if (cpufreq_frequency_table_target(policy, &p4clockmod_table[0],
+				target_freq, relation, &newstate))
+		return -EINVAL;
+
+	freqs.old = cpufreq_p4_get(policy->cpu);
+	freqs.new = stock_freq * p4clockmod_table[newstate].index / 8;
+
+	if (freqs.new == freqs.old)
+		return 0;
+
+	/* notifiers */
+	for_each_cpu(i, policy->cpus) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	}
+
+	/* run on each logical CPU,
+	 * see section 13.15.3 of IA32 Intel Architecture Software
+	 * Developer's Manual, Volume 3
+	 */
+	for_each_cpu(i, policy->cpus)
+		cpufreq_p4_setdc(i, p4clockmod_table[newstate].index);
+
+	/* notifiers */
+	for_each_cpu(i, policy->cpus) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	}
+
+	return 0;
+}
+
+
+static int cpufreq_p4_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, &p4clockmod_table[0]);
+}
+
+
+static unsigned int cpufreq_p4_get_frequency(struct cpuinfo_x86 *c)
+{
+	if (c->x86 == 0x06) {
+		if (cpu_has(c, X86_FEATURE_EST))
+			printk_once(KERN_WARNING PFX "Warning: EST-capable "
+			       "CPU detected. The acpi-cpufreq module offers "
+			       "voltage scaling in addition to frequency "
+			       "scaling. You should use that instead of "
+			       "p4-clockmod, if possible.\n");
+		switch (c->x86_model) {
+		case 0x0E: /* Core */
+		case 0x0F: /* Core Duo */
+		case 0x16: /* Celeron Core */
+		case 0x1C: /* Atom */
+			p4clockmod_driver.flags |= CPUFREQ_CONST_LOOPS;
+			return speedstep_get_frequency(SPEEDSTEP_CPU_PCORE);
+		case 0x0D: /* Pentium M (Dothan) */
+			p4clockmod_driver.flags |= CPUFREQ_CONST_LOOPS;
+			/* fall through */
+		case 0x09: /* Pentium M (Banias) */
+			return speedstep_get_frequency(SPEEDSTEP_CPU_PM);
+		}
+	}
+
+	if (c->x86 != 0xF)
+		return 0;
+
+	/* on P-4s, the TSC runs with constant frequency independent whether
+	 * throttling is active or not. */
+	p4clockmod_driver.flags |= CPUFREQ_CONST_LOOPS;
+
+	if (speedstep_detect_processor() == SPEEDSTEP_CPU_P4M) {
+		printk(KERN_WARNING PFX "Warning: Pentium 4-M detected. "
+		       "The speedstep-ich or acpi cpufreq modules offer "
+		       "voltage scaling in addition of frequency scaling. "
+		       "You should use either one instead of p4-clockmod, "
+		       "if possible.\n");
+		return speedstep_get_frequency(SPEEDSTEP_CPU_P4M);
+	}
+
+	return speedstep_get_frequency(SPEEDSTEP_CPU_P4D);
+}
+
+
+
+static int cpufreq_p4_cpu_init(struct cpufreq_policy *policy)
+{
+	struct cpuinfo_x86 *c = &cpu_data(policy->cpu);
+	int cpuid = 0;
+	unsigned int i;
+
+#ifdef CONFIG_SMP
+	cpumask_copy(policy->cpus, cpu_sibling_mask(policy->cpu));
+#endif
+
+	/* Errata workaround */
+	cpuid = (c->x86 << 8) | (c->x86_model << 4) | c->x86_mask;
+	switch (cpuid) {
+	case 0x0f07:
+	case 0x0f0a:
+	case 0x0f11:
+	case 0x0f12:
+		has_N44_O17_errata[policy->cpu] = 1;
+		pr_debug("has errata -- disabling low frequencies\n");
+	}
+
+	if (speedstep_detect_processor() == SPEEDSTEP_CPU_P4D &&
+	    c->x86_model < 2) {
+		/* switch to maximum frequency and measure result */
+		cpufreq_p4_setdc(policy->cpu, DC_DISABLE);
+		recalibrate_cpu_khz();
+	}
+	/* get max frequency */
+	stock_freq = cpufreq_p4_get_frequency(c);
+	if (!stock_freq)
+		return -EINVAL;
+
+	/* table init */
+	for (i = 1; (p4clockmod_table[i].frequency != CPUFREQ_TABLE_END); i++) {
+		if ((i < 2) && (has_N44_O17_errata[policy->cpu]))
+			p4clockmod_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+		else
+			p4clockmod_table[i].frequency = (stock_freq * i)/8;
+	}
+	cpufreq_frequency_table_get_attr(p4clockmod_table, policy->cpu);
+
+	/* cpuinfo and default policy values */
+
+	/* the transition latency is set to be 1 higher than the maximum
+	 * transition latency of the ondemand governor */
+	policy->cpuinfo.transition_latency = 10000001;
+	policy->cur = stock_freq;
+
+	return cpufreq_frequency_table_cpuinfo(policy, &p4clockmod_table[0]);
+}
+
+
+static int cpufreq_p4_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+static unsigned int cpufreq_p4_get(unsigned int cpu)
+{
+	u32 l, h;
+
+	rdmsr_on_cpu(cpu, MSR_IA32_THERM_CONTROL, &l, &h);
+
+	if (l & 0x10) {
+		l = l >> 1;
+		l &= 0x7;
+	} else
+		l = DC_DISABLE;
+
+	if (l != DC_DISABLE)
+		return stock_freq * l / 8;
+
+	return stock_freq;
+}
+
+static struct freq_attr *p4clockmod_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver p4clockmod_driver = {
+	.verify		= cpufreq_p4_verify,
+	.target		= cpufreq_p4_target,
+	.init		= cpufreq_p4_cpu_init,
+	.exit		= cpufreq_p4_cpu_exit,
+	.get		= cpufreq_p4_get,
+	.name		= "p4-clockmod",
+	.owner		= THIS_MODULE,
+	.attr		= p4clockmod_attr,
+};
+
+
+static int __init cpufreq_p4_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	int ret;
+
+	/*
+	 * THERM_CONTROL is architectural for IA32 now, so
+	 * we can rely on the capability checks
+	 */
+	if (c->x86_vendor != X86_VENDOR_INTEL)
+		return -ENODEV;
+
+	if (!test_cpu_cap(c, X86_FEATURE_ACPI) ||
+				!test_cpu_cap(c, X86_FEATURE_ACC))
+		return -ENODEV;
+
+	ret = cpufreq_register_driver(&p4clockmod_driver);
+	if (!ret)
+		printk(KERN_INFO PFX "P4/Xeon(TM) CPU On-Demand Clock "
+				"Modulation available\n");
+
+	return ret;
+}
+
+
+static void __exit cpufreq_p4_exit(void)
+{
+	cpufreq_unregister_driver(&p4clockmod_driver);
+}
+
+
+MODULE_AUTHOR("Zwane Mwaikambo <zwane@commfireservices.com>");
+MODULE_DESCRIPTION("cpufreq driver for Pentium(TM) 4/Xeon(TM)");
+MODULE_LICENSE("GPL");
+
+late_initcall(cpufreq_p4_init);
+module_exit(cpufreq_p4_exit);
diff --git a/drivers/cpufreq/pcc-cpufreq.c b/drivers/cpufreq/pcc-cpufreq.c
new file mode 100644
index 00000000..cdc02ac8
--- /dev/null
+++ b/drivers/cpufreq/pcc-cpufreq.c
@@ -0,0 +1,624 @@
+/*
+ *  pcc-cpufreq.c - Processor Clocking Control firmware cpufreq interface
+ *
+ *  Copyright (C) 2009 Red Hat, Matthew Garrett <mjg@redhat.com>
+ *  Copyright (C) 2009 Hewlett-Packard Development Company, L.P.
+ *	Nagananda Chumbalkar <nagananda.chumbalkar@hp.com>
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; version 2 of the License.
+ *
+ *  This program is distributed in the hope that it will be useful, but
+ *  WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or NON
+ *  INFRINGEMENT. See the GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License along
+ *  with this program; if not, write to the Free Software Foundation, Inc.,
+ *  675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/sched.h>
+#include <linux/cpufreq.h>
+#include <linux/compiler.h>
+#include <linux/slab.h>
+
+#include <linux/acpi.h>
+#include <linux/io.h>
+#include <linux/spinlock.h>
+#include <linux/uaccess.h>
+
+#include <acpi/processor.h>
+
+#define PCC_VERSION	"1.10.00"
+#define POLL_LOOPS 	300
+
+#define CMD_COMPLETE 	0x1
+#define CMD_GET_FREQ 	0x0
+#define CMD_SET_FREQ 	0x1
+
+#define BUF_SZ		4
+
+struct pcc_register_resource {
+	u8 descriptor;
+	u16 length;
+	u8 space_id;
+	u8 bit_width;
+	u8 bit_offset;
+	u8 access_size;
+	u64 address;
+} __attribute__ ((packed));
+
+struct pcc_memory_resource {
+	u8 descriptor;
+	u16 length;
+	u8 space_id;
+	u8 resource_usage;
+	u8 type_specific;
+	u64 granularity;
+	u64 minimum;
+	u64 maximum;
+	u64 translation_offset;
+	u64 address_length;
+} __attribute__ ((packed));
+
+static struct cpufreq_driver pcc_cpufreq_driver;
+
+struct pcc_header {
+	u32 signature;
+	u16 length;
+	u8 major;
+	u8 minor;
+	u32 features;
+	u16 command;
+	u16 status;
+	u32 latency;
+	u32 minimum_time;
+	u32 maximum_time;
+	u32 nominal;
+	u32 throttled_frequency;
+	u32 minimum_frequency;
+};
+
+static void __iomem *pcch_virt_addr;
+static struct pcc_header __iomem *pcch_hdr;
+
+static DEFINE_SPINLOCK(pcc_lock);
+
+static struct acpi_generic_address doorbell;
+
+static u64 doorbell_preserve;
+static u64 doorbell_write;
+
+static u8 OSC_UUID[16] = {0x9F, 0x2C, 0x9B, 0x63, 0x91, 0x70, 0x1f, 0x49,
+			  0xBB, 0x4F, 0xA5, 0x98, 0x2F, 0xA1, 0xB5, 0x46};
+
+struct pcc_cpu {
+	u32 input_offset;
+	u32 output_offset;
+};
+
+static struct pcc_cpu __percpu *pcc_cpu_info;
+
+static int pcc_cpufreq_verify(struct cpufreq_policy *policy)
+{
+	cpufreq_verify_within_limits(policy, policy->cpuinfo.min_freq,
+				     policy->cpuinfo.max_freq);
+	return 0;
+}
+
+static inline void pcc_cmd(void)
+{
+	u64 doorbell_value;
+	int i;
+
+	acpi_read(&doorbell_value, &doorbell);
+	acpi_write((doorbell_value & doorbell_preserve) | doorbell_write,
+		   &doorbell);
+
+	for (i = 0; i < POLL_LOOPS; i++) {
+		if (ioread16(&pcch_hdr->status) & CMD_COMPLETE)
+			break;
+	}
+}
+
+static inline void pcc_clear_mapping(void)
+{
+	if (pcch_virt_addr)
+		iounmap(pcch_virt_addr);
+	pcch_virt_addr = NULL;
+}
+
+static unsigned int pcc_get_freq(unsigned int cpu)
+{
+	struct pcc_cpu *pcc_cpu_data;
+	unsigned int curr_freq;
+	unsigned int freq_limit;
+	u16 status;
+	u32 input_buffer;
+	u32 output_buffer;
+
+	spin_lock(&pcc_lock);
+
+	pr_debug("get: get_freq for CPU %d\n", cpu);
+	pcc_cpu_data = per_cpu_ptr(pcc_cpu_info, cpu);
+
+	input_buffer = 0x1;
+	iowrite32(input_buffer,
+			(pcch_virt_addr + pcc_cpu_data->input_offset));
+	iowrite16(CMD_GET_FREQ, &pcch_hdr->command);
+
+	pcc_cmd();
+
+	output_buffer =
+		ioread32(pcch_virt_addr + pcc_cpu_data->output_offset);
+
+	/* Clear the input buffer - we are done with the current command */
+	memset_io((pcch_virt_addr + pcc_cpu_data->input_offset), 0, BUF_SZ);
+
+	status = ioread16(&pcch_hdr->status);
+	if (status != CMD_COMPLETE) {
+		pr_debug("get: FAILED: for CPU %d, status is %d\n",
+			cpu, status);
+		goto cmd_incomplete;
+	}
+	iowrite16(0, &pcch_hdr->status);
+	curr_freq = (((ioread32(&pcch_hdr->nominal) * (output_buffer & 0xff))
+			/ 100) * 1000);
+
+	pr_debug("get: SUCCESS: (virtual) output_offset for cpu %d is "
+		"0x%p, contains a value of: 0x%x. Speed is: %d MHz\n",
+		cpu, (pcch_virt_addr + pcc_cpu_data->output_offset),
+		output_buffer, curr_freq);
+
+	freq_limit = (output_buffer >> 8) & 0xff;
+	if (freq_limit != 0xff) {
+		pr_debug("get: frequency for cpu %d is being temporarily"
+			" capped at %d\n", cpu, curr_freq);
+	}
+
+	spin_unlock(&pcc_lock);
+	return curr_freq;
+
+cmd_incomplete:
+	iowrite16(0, &pcch_hdr->status);
+	spin_unlock(&pcc_lock);
+	return 0;
+}
+
+static int pcc_cpufreq_target(struct cpufreq_policy *policy,
+			      unsigned int target_freq,
+			      unsigned int relation)
+{
+	struct pcc_cpu *pcc_cpu_data;
+	struct cpufreq_freqs freqs;
+	u16 status;
+	u32 input_buffer;
+	int cpu;
+
+	spin_lock(&pcc_lock);
+	cpu = policy->cpu;
+	pcc_cpu_data = per_cpu_ptr(pcc_cpu_info, cpu);
+
+	pr_debug("target: CPU %d should go to target freq: %d "
+		"(virtual) input_offset is 0x%p\n",
+		cpu, target_freq,
+		(pcch_virt_addr + pcc_cpu_data->input_offset));
+
+	freqs.new = target_freq;
+	freqs.cpu = cpu;
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	input_buffer = 0x1 | (((target_freq * 100)
+			       / (ioread32(&pcch_hdr->nominal) * 1000)) << 8);
+	iowrite32(input_buffer,
+			(pcch_virt_addr + pcc_cpu_data->input_offset));
+	iowrite16(CMD_SET_FREQ, &pcch_hdr->command);
+
+	pcc_cmd();
+
+	/* Clear the input buffer - we are done with the current command */
+	memset_io((pcch_virt_addr + pcc_cpu_data->input_offset), 0, BUF_SZ);
+
+	status = ioread16(&pcch_hdr->status);
+	if (status != CMD_COMPLETE) {
+		pr_debug("target: FAILED for cpu %d, with status: 0x%x\n",
+			cpu, status);
+		goto cmd_incomplete;
+	}
+	iowrite16(0, &pcch_hdr->status);
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	pr_debug("target: was SUCCESSFUL for cpu %d\n", cpu);
+	spin_unlock(&pcc_lock);
+
+	return 0;
+
+cmd_incomplete:
+	iowrite16(0, &pcch_hdr->status);
+	spin_unlock(&pcc_lock);
+	return -EINVAL;
+}
+
+static int pcc_get_offset(int cpu)
+{
+	acpi_status status;
+	struct acpi_buffer buffer = {ACPI_ALLOCATE_BUFFER, NULL};
+	union acpi_object *pccp, *offset;
+	struct pcc_cpu *pcc_cpu_data;
+	struct acpi_processor *pr;
+	int ret = 0;
+
+	pr = per_cpu(processors, cpu);
+	pcc_cpu_data = per_cpu_ptr(pcc_cpu_info, cpu);
+
+	if (!pr)
+		return -ENODEV;
+
+	status = acpi_evaluate_object(pr->handle, "PCCP", NULL, &buffer);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	pccp = buffer.pointer;
+	if (!pccp || pccp->type != ACPI_TYPE_PACKAGE) {
+		ret = -ENODEV;
+		goto out_free;
+	};
+
+	offset = &(pccp->package.elements[0]);
+	if (!offset || offset->type != ACPI_TYPE_INTEGER) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	pcc_cpu_data->input_offset = offset->integer.value;
+
+	offset = &(pccp->package.elements[1]);
+	if (!offset || offset->type != ACPI_TYPE_INTEGER) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	pcc_cpu_data->output_offset = offset->integer.value;
+
+	memset_io((pcch_virt_addr + pcc_cpu_data->input_offset), 0, BUF_SZ);
+	memset_io((pcch_virt_addr + pcc_cpu_data->output_offset), 0, BUF_SZ);
+
+	pr_debug("pcc_get_offset: for CPU %d: pcc_cpu_data "
+		"input_offset: 0x%x, pcc_cpu_data output_offset: 0x%x\n",
+		cpu, pcc_cpu_data->input_offset, pcc_cpu_data->output_offset);
+out_free:
+	kfree(buffer.pointer);
+	return ret;
+}
+
+static int __init pcc_cpufreq_do_osc(acpi_handle *handle)
+{
+	acpi_status status;
+	struct acpi_object_list input;
+	struct acpi_buffer output = {ACPI_ALLOCATE_BUFFER, NULL};
+	union acpi_object in_params[4];
+	union acpi_object *out_obj;
+	u32 capabilities[2];
+	u32 errors;
+	u32 supported;
+	int ret = 0;
+
+	input.count = 4;
+	input.pointer = in_params;
+	in_params[0].type               = ACPI_TYPE_BUFFER;
+	in_params[0].buffer.length      = 16;
+	in_params[0].buffer.pointer     = OSC_UUID;
+	in_params[1].type               = ACPI_TYPE_INTEGER;
+	in_params[1].integer.value      = 1;
+	in_params[2].type               = ACPI_TYPE_INTEGER;
+	in_params[2].integer.value      = 2;
+	in_params[3].type               = ACPI_TYPE_BUFFER;
+	in_params[3].buffer.length      = 8;
+	in_params[3].buffer.pointer     = (u8 *)&capabilities;
+
+	capabilities[0] = OSC_QUERY_ENABLE;
+	capabilities[1] = 0x1;
+
+	status = acpi_evaluate_object(*handle, "_OSC", &input, &output);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	if (!output.length)
+		return -ENODEV;
+
+	out_obj = output.pointer;
+	if (out_obj->type != ACPI_TYPE_BUFFER) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	errors = *((u32 *)out_obj->buffer.pointer) & ~(1 << 0);
+	if (errors) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	supported = *((u32 *)(out_obj->buffer.pointer + 4));
+	if (!(supported & 0x1)) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	kfree(output.pointer);
+	capabilities[0] = 0x0;
+	capabilities[1] = 0x1;
+
+	status = acpi_evaluate_object(*handle, "_OSC", &input, &output);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	if (!output.length)
+		return -ENODEV;
+
+	out_obj = output.pointer;
+	if (out_obj->type != ACPI_TYPE_BUFFER) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	errors = *((u32 *)out_obj->buffer.pointer) & ~(1 << 0);
+	if (errors) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	supported = *((u32 *)(out_obj->buffer.pointer + 4));
+	if (!(supported & 0x1)) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+out_free:
+	kfree(output.pointer);
+	return ret;
+}
+
+static int __init pcc_cpufreq_probe(void)
+{
+	acpi_status status;
+	struct acpi_buffer output = {ACPI_ALLOCATE_BUFFER, NULL};
+	struct pcc_memory_resource *mem_resource;
+	struct pcc_register_resource *reg_resource;
+	union acpi_object *out_obj, *member;
+	acpi_handle handle, osc_handle, pcch_handle;
+	int ret = 0;
+
+	status = acpi_get_handle(NULL, "\\_SB", &handle);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	status = acpi_get_handle(handle, "PCCH", &pcch_handle);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	status = acpi_get_handle(handle, "_OSC", &osc_handle);
+	if (ACPI_SUCCESS(status)) {
+		ret = pcc_cpufreq_do_osc(&osc_handle);
+		if (ret)
+			pr_debug("probe: _OSC evaluation did not succeed\n");
+		/* Firmware's use of _OSC is optional */
+		ret = 0;
+	}
+
+	status = acpi_evaluate_object(handle, "PCCH", NULL, &output);
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	out_obj = output.pointer;
+	if (out_obj->type != ACPI_TYPE_PACKAGE) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	member = &out_obj->package.elements[0];
+	if (member->type != ACPI_TYPE_BUFFER) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	mem_resource = (struct pcc_memory_resource *)member->buffer.pointer;
+
+	pr_debug("probe: mem_resource descriptor: 0x%x,"
+		" length: %d, space_id: %d, resource_usage: %d,"
+		" type_specific: %d, granularity: 0x%llx,"
+		" minimum: 0x%llx, maximum: 0x%llx,"
+		" translation_offset: 0x%llx, address_length: 0x%llx\n",
+		mem_resource->descriptor, mem_resource->length,
+		mem_resource->space_id, mem_resource->resource_usage,
+		mem_resource->type_specific, mem_resource->granularity,
+		mem_resource->minimum, mem_resource->maximum,
+		mem_resource->translation_offset,
+		mem_resource->address_length);
+
+	if (mem_resource->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+
+	pcch_virt_addr = ioremap_nocache(mem_resource->minimum,
+					mem_resource->address_length);
+	if (pcch_virt_addr == NULL) {
+		pr_debug("probe: could not map shared mem region\n");
+		goto out_free;
+	}
+	pcch_hdr = pcch_virt_addr;
+
+	pr_debug("probe: PCCH header (virtual) addr: 0x%p\n", pcch_hdr);
+	pr_debug("probe: PCCH header is at physical address: 0x%llx,"
+		" signature: 0x%x, length: %d bytes, major: %d, minor: %d,"
+		" supported features: 0x%x, command field: 0x%x,"
+		" status field: 0x%x, nominal latency: %d us\n",
+		mem_resource->minimum, ioread32(&pcch_hdr->signature),
+		ioread16(&pcch_hdr->length), ioread8(&pcch_hdr->major),
+		ioread8(&pcch_hdr->minor), ioread32(&pcch_hdr->features),
+		ioread16(&pcch_hdr->command), ioread16(&pcch_hdr->status),
+		ioread32(&pcch_hdr->latency));
+
+	pr_debug("probe: min time between commands: %d us,"
+		" max time between commands: %d us,"
+		" nominal CPU frequency: %d MHz,"
+		" minimum CPU frequency: %d MHz,"
+		" minimum CPU frequency without throttling: %d MHz\n",
+		ioread32(&pcch_hdr->minimum_time),
+		ioread32(&pcch_hdr->maximum_time),
+		ioread32(&pcch_hdr->nominal),
+		ioread32(&pcch_hdr->throttled_frequency),
+		ioread32(&pcch_hdr->minimum_frequency));
+
+	member = &out_obj->package.elements[1];
+	if (member->type != ACPI_TYPE_BUFFER) {
+		ret = -ENODEV;
+		goto pcch_free;
+	}
+
+	reg_resource = (struct pcc_register_resource *)member->buffer.pointer;
+
+	doorbell.space_id = reg_resource->space_id;
+	doorbell.bit_width = reg_resource->bit_width;
+	doorbell.bit_offset = reg_resource->bit_offset;
+	doorbell.access_width = 64;
+	doorbell.address = reg_resource->address;
+
+	pr_debug("probe: doorbell: space_id is %d, bit_width is %d, "
+		"bit_offset is %d, access_width is %d, address is 0x%llx\n",
+		doorbell.space_id, doorbell.bit_width, doorbell.bit_offset,
+		doorbell.access_width, reg_resource->address);
+
+	member = &out_obj->package.elements[2];
+	if (member->type != ACPI_TYPE_INTEGER) {
+		ret = -ENODEV;
+		goto pcch_free;
+	}
+
+	doorbell_preserve = member->integer.value;
+
+	member = &out_obj->package.elements[3];
+	if (member->type != ACPI_TYPE_INTEGER) {
+		ret = -ENODEV;
+		goto pcch_free;
+	}
+
+	doorbell_write = member->integer.value;
+
+	pr_debug("probe: doorbell_preserve: 0x%llx,"
+		" doorbell_write: 0x%llx\n",
+		doorbell_preserve, doorbell_write);
+
+	pcc_cpu_info = alloc_percpu(struct pcc_cpu);
+	if (!pcc_cpu_info) {
+		ret = -ENOMEM;
+		goto pcch_free;
+	}
+
+	printk(KERN_DEBUG "pcc-cpufreq: (v%s) driver loaded with frequency"
+	       " limits: %d MHz, %d MHz\n", PCC_VERSION,
+	       ioread32(&pcch_hdr->minimum_frequency),
+	       ioread32(&pcch_hdr->nominal));
+	kfree(output.pointer);
+	return ret;
+pcch_free:
+	pcc_clear_mapping();
+out_free:
+	kfree(output.pointer);
+	return ret;
+}
+
+static int pcc_cpufreq_cpu_init(struct cpufreq_policy *policy)
+{
+	unsigned int cpu = policy->cpu;
+	unsigned int result = 0;
+
+	if (!pcch_virt_addr) {
+		result = -1;
+		goto out;
+	}
+
+	result = pcc_get_offset(cpu);
+	if (result) {
+		pr_debug("init: PCCP evaluation failed\n");
+		goto out;
+	}
+
+	policy->max = policy->cpuinfo.max_freq =
+		ioread32(&pcch_hdr->nominal) * 1000;
+	policy->min = policy->cpuinfo.min_freq =
+		ioread32(&pcch_hdr->minimum_frequency) * 1000;
+	policy->cur = pcc_get_freq(cpu);
+
+	if (!policy->cur) {
+		pr_debug("init: Unable to get current CPU frequency\n");
+		result = -EINVAL;
+		goto out;
+	}
+
+	pr_debug("init: policy->max is %d, policy->min is %d\n",
+		policy->max, policy->min);
+out:
+	return result;
+}
+
+static int pcc_cpufreq_cpu_exit(struct cpufreq_policy *policy)
+{
+	return 0;
+}
+
+static struct cpufreq_driver pcc_cpufreq_driver = {
+	.flags = CPUFREQ_CONST_LOOPS,
+	.get = pcc_get_freq,
+	.verify = pcc_cpufreq_verify,
+	.target = pcc_cpufreq_target,
+	.init = pcc_cpufreq_cpu_init,
+	.exit = pcc_cpufreq_cpu_exit,
+	.name = "pcc-cpufreq",
+	.owner = THIS_MODULE,
+};
+
+static int __init pcc_cpufreq_init(void)
+{
+	int ret;
+
+	if (acpi_disabled)
+		return 0;
+
+	ret = pcc_cpufreq_probe();
+	if (ret) {
+		pr_debug("pcc_cpufreq_init: PCCH evaluation failed\n");
+		return ret;
+	}
+
+	ret = cpufreq_register_driver(&pcc_cpufreq_driver);
+
+	return ret;
+}
+
+static void __exit pcc_cpufreq_exit(void)
+{
+	cpufreq_unregister_driver(&pcc_cpufreq_driver);
+
+	pcc_clear_mapping();
+
+	free_percpu(pcc_cpu_info);
+}
+
+MODULE_AUTHOR("Matthew Garrett, Naga Chumbalkar");
+MODULE_VERSION(PCC_VERSION);
+MODULE_DESCRIPTION("Processor Clocking Control interface driver");
+MODULE_LICENSE("GPL");
+
+late_initcall(pcc_cpufreq_init);
+module_exit(pcc_cpufreq_exit);
diff --git a/drivers/cpufreq/powernow-k6.c b/drivers/cpufreq/powernow-k6.c
new file mode 100644
index 00000000..b3379d6a
--- /dev/null
+++ b/drivers/cpufreq/powernow-k6.c
@@ -0,0 +1,261 @@
+/*
+ *  This file was based upon code in Powertweak Linux (http://powertweak.sf.net)
+ *  (C) 2000-2003  Dave Jones, Arjan van de Ven, Janne Pänkälä,
+ *                 Dominik Brodowski.
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/ioport.h>
+#include <linux/timex.h>
+#include <linux/io.h>
+
+#include <asm/msr.h>
+
+#define POWERNOW_IOPORT 0xfff0          /* it doesn't matter where, as long
+					   as it is unused */
+
+#define PFX "powernow-k6: "
+static unsigned int                     busfreq;   /* FSB, in 10 kHz */
+static unsigned int                     max_multiplier;
+
+
+/* Clock ratio multiplied by 10 - see table 27 in AMD#23446 */
+static struct cpufreq_frequency_table clock_ratio[] = {
+	{45,  /* 000 -> 4.5x */ 0},
+	{50,  /* 001 -> 5.0x */ 0},
+	{40,  /* 010 -> 4.0x */ 0},
+	{55,  /* 011 -> 5.5x */ 0},
+	{20,  /* 100 -> 2.0x */ 0},
+	{30,  /* 101 -> 3.0x */ 0},
+	{60,  /* 110 -> 6.0x */ 0},
+	{35,  /* 111 -> 3.5x */ 0},
+	{0, CPUFREQ_TABLE_END}
+};
+
+
+/**
+ * powernow_k6_get_cpu_multiplier - returns the current FSB multiplier
+ *
+ *   Returns the current setting of the frequency multiplier. Core clock
+ * speed is frequency of the Front-Side Bus multiplied with this value.
+ */
+static int powernow_k6_get_cpu_multiplier(void)
+{
+	u64 invalue = 0;
+	u32 msrval;
+
+	msrval = POWERNOW_IOPORT + 0x1;
+	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
+	invalue = inl(POWERNOW_IOPORT + 0x8);
+	msrval = POWERNOW_IOPORT + 0x0;
+	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
+
+	return clock_ratio[(invalue >> 5)&7].index;
+}
+
+
+/**
+ * powernow_k6_set_state - set the PowerNow! multiplier
+ * @best_i: clock_ratio[best_i] is the target multiplier
+ *
+ *   Tries to change the PowerNow! multiplier
+ */
+static void powernow_k6_set_state(unsigned int best_i)
+{
+	unsigned long outvalue = 0, invalue = 0;
+	unsigned long msrval;
+	struct cpufreq_freqs freqs;
+
+	if (clock_ratio[best_i].index > max_multiplier) {
+		printk(KERN_ERR PFX "invalid target frequency\n");
+		return;
+	}
+
+	freqs.old = busfreq * powernow_k6_get_cpu_multiplier();
+	freqs.new = busfreq * clock_ratio[best_i].index;
+	freqs.cpu = 0; /* powernow-k6.c is UP only driver */
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	/* we now need to transform best_i to the BVC format, see AMD#23446 */
+
+	outvalue = (1<<12) | (1<<10) | (1<<9) | (best_i<<5);
+
+	msrval = POWERNOW_IOPORT + 0x1;
+	wrmsr(MSR_K6_EPMR, msrval, 0); /* enable the PowerNow port */
+	invalue = inl(POWERNOW_IOPORT + 0x8);
+	invalue = invalue & 0xf;
+	outvalue = outvalue | invalue;
+	outl(outvalue , (POWERNOW_IOPORT + 0x8));
+	msrval = POWERNOW_IOPORT + 0x0;
+	wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+
+	return;
+}
+
+
+/**
+ * powernow_k6_verify - verifies a new CPUfreq policy
+ * @policy: new policy
+ *
+ * Policy must be within lowest and highest possible CPU Frequency,
+ * and at least one possible state must be within min and max.
+ */
+static int powernow_k6_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, &clock_ratio[0]);
+}
+
+
+/**
+ * powernow_k6_setpolicy - sets a new CPUFreq policy
+ * @policy: new policy
+ * @target_freq: the target frequency
+ * @relation: how that frequency relates to achieved frequency
+ *  (CPUFREQ_RELATION_L or CPUFREQ_RELATION_H)
+ *
+ * sets a new CPUFreq policy
+ */
+static int powernow_k6_target(struct cpufreq_policy *policy,
+			       unsigned int target_freq,
+			       unsigned int relation)
+{
+	unsigned int newstate = 0;
+
+	if (cpufreq_frequency_table_target(policy, &clock_ratio[0],
+				target_freq, relation, &newstate))
+		return -EINVAL;
+
+	powernow_k6_set_state(newstate);
+
+	return 0;
+}
+
+
+static int powernow_k6_cpu_init(struct cpufreq_policy *policy)
+{
+	unsigned int i, f;
+	int result;
+
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	/* get frequencies */
+	max_multiplier = powernow_k6_get_cpu_multiplier();
+	busfreq = cpu_khz / max_multiplier;
+
+	/* table init */
+	for (i = 0; (clock_ratio[i].frequency != CPUFREQ_TABLE_END); i++) {
+		f = clock_ratio[i].index;
+		if (f > max_multiplier)
+			clock_ratio[i].frequency = CPUFREQ_ENTRY_INVALID;
+		else
+			clock_ratio[i].frequency = busfreq * f;
+	}
+
+	/* cpuinfo and default policy values */
+	policy->cpuinfo.transition_latency = 200000;
+	policy->cur = busfreq * max_multiplier;
+
+	result = cpufreq_frequency_table_cpuinfo(policy, clock_ratio);
+	if (result)
+		return result;
+
+	cpufreq_frequency_table_get_attr(clock_ratio, policy->cpu);
+
+	return 0;
+}
+
+
+static int powernow_k6_cpu_exit(struct cpufreq_policy *policy)
+{
+	unsigned int i;
+	for (i = 0; i < 8; i++) {
+		if (i == max_multiplier)
+			powernow_k6_set_state(i);
+	}
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+static unsigned int powernow_k6_get(unsigned int cpu)
+{
+	unsigned int ret;
+	ret = (busfreq * powernow_k6_get_cpu_multiplier());
+	return ret;
+}
+
+static struct freq_attr *powernow_k6_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver powernow_k6_driver = {
+	.verify		= powernow_k6_verify,
+	.target		= powernow_k6_target,
+	.init		= powernow_k6_cpu_init,
+	.exit		= powernow_k6_cpu_exit,
+	.get		= powernow_k6_get,
+	.name		= "powernow-k6",
+	.owner		= THIS_MODULE,
+	.attr		= powernow_k6_attr,
+};
+
+
+/**
+ * powernow_k6_init - initializes the k6 PowerNow! CPUFreq driver
+ *
+ *   Initializes the K6 PowerNow! support. Returns -ENODEV on unsupported
+ * devices, -EINVAL or -ENOMEM on problems during initiatization, and zero
+ * on success.
+ */
+static int __init powernow_k6_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	if ((c->x86_vendor != X86_VENDOR_AMD) || (c->x86 != 5) ||
+		((c->x86_model != 12) && (c->x86_model != 13)))
+		return -ENODEV;
+
+	if (!request_region(POWERNOW_IOPORT, 16, "PowerNow!")) {
+		printk(KERN_INFO PFX "PowerNow IOPORT region already used.\n");
+		return -EIO;
+	}
+
+	if (cpufreq_register_driver(&powernow_k6_driver)) {
+		release_region(POWERNOW_IOPORT, 16);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+
+/**
+ * powernow_k6_exit - unregisters AMD K6-2+/3+ PowerNow! support
+ *
+ *   Unregisters AMD K6-2+ / K6-3+ PowerNow! support.
+ */
+static void __exit powernow_k6_exit(void)
+{
+	cpufreq_unregister_driver(&powernow_k6_driver);
+	release_region(POWERNOW_IOPORT, 16);
+}
+
+
+MODULE_AUTHOR("Arjan van de Ven, Dave Jones <davej@redhat.com>, "
+		"Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("PowerNow! driver for AMD K6-2+ / K6-3+ processors.");
+MODULE_LICENSE("GPL");
+
+module_init(powernow_k6_init);
+module_exit(powernow_k6_exit);
diff --git a/drivers/cpufreq/powernow-k7.c b/drivers/cpufreq/powernow-k7.c
new file mode 100644
index 00000000..d71d9f37
--- /dev/null
+++ b/drivers/cpufreq/powernow-k7.c
@@ -0,0 +1,747 @@
+/*
+ *  AMD K7 Powernow driver.
+ *  (C) 2003 Dave Jones on behalf of SuSE Labs.
+ *  (C) 2003-2004 Dave Jones <davej@redhat.com>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *  Based upon datasheets & sample CPUs kindly provided by AMD.
+ *
+ * Errata 5:
+ *  CPU may fail to execute a FID/VID change in presence of interrupt.
+ *  - We cli/sti on stepping A0 CPUs around the FID/VID transition.
+ * Errata 15:
+ *  CPU with half frequency multipliers may hang upon wakeup from disconnect.
+ *  - We disable half multipliers if ACPI is used on A0 stepping CPUs.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/dmi.h>
+#include <linux/timex.h>
+#include <linux/io.h>
+
+#include <asm/timer.h>		/* Needed for recalibrate_cpu_khz() */
+#include <asm/msr.h>
+#include <asm/system.h>
+
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+#include <linux/acpi.h>
+#include <acpi/processor.h>
+#endif
+
+#include "powernow-k7.h"
+
+#define PFX "powernow: "
+
+
+struct psb_s {
+	u8 signature[10];
+	u8 tableversion;
+	u8 flags;
+	u16 settlingtime;
+	u8 reserved1;
+	u8 numpst;
+};
+
+struct pst_s {
+	u32 cpuid;
+	u8 fsbspeed;
+	u8 maxfid;
+	u8 startvid;
+	u8 numpstates;
+};
+
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+union powernow_acpi_control_t {
+	struct {
+		unsigned long fid:5,
+			vid:5,
+			sgtc:20,
+			res1:2;
+	} bits;
+	unsigned long val;
+};
+#endif
+
+/* divide by 1000 to get VCore voltage in V. */
+static const int mobile_vid_table[32] = {
+    2000, 1950, 1900, 1850, 1800, 1750, 1700, 1650,
+    1600, 1550, 1500, 1450, 1400, 1350, 1300, 0,
+    1275, 1250, 1225, 1200, 1175, 1150, 1125, 1100,
+    1075, 1050, 1025, 1000, 975, 950, 925, 0,
+};
+
+/* divide by 10 to get FID. */
+static const int fid_codes[32] = {
+    110, 115, 120, 125, 50, 55, 60, 65,
+    70, 75, 80, 85, 90, 95, 100, 105,
+    30, 190, 40, 200, 130, 135, 140, 210,
+    150, 225, 160, 165, 170, 180, -1, -1,
+};
+
+/* This parameter is used in order to force ACPI instead of legacy method for
+ * configuration purpose.
+ */
+
+static int acpi_force;
+
+static struct cpufreq_frequency_table *powernow_table;
+
+static unsigned int can_scale_bus;
+static unsigned int can_scale_vid;
+static unsigned int minimum_speed = -1;
+static unsigned int maximum_speed;
+static unsigned int number_scales;
+static unsigned int fsb;
+static unsigned int latency;
+static char have_a0;
+
+static int check_fsb(unsigned int fsbspeed)
+{
+	int delta;
+	unsigned int f = fsb / 1000;
+
+	delta = (fsbspeed > f) ? fsbspeed - f : f - fsbspeed;
+	return delta < 5;
+}
+
+static int check_powernow(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	unsigned int maxei, eax, ebx, ecx, edx;
+
+	if ((c->x86_vendor != X86_VENDOR_AMD) || (c->x86 != 6)) {
+#ifdef MODULE
+		printk(KERN_INFO PFX "This module only works with "
+				"AMD K7 CPUs\n");
+#endif
+		return 0;
+	}
+
+	/* Get maximum capabilities */
+	maxei = cpuid_eax(0x80000000);
+	if (maxei < 0x80000007) {	/* Any powernow info ? */
+#ifdef MODULE
+		printk(KERN_INFO PFX "No powernow capabilities detected\n");
+#endif
+		return 0;
+	}
+
+	if ((c->x86_model == 6) && (c->x86_mask == 0)) {
+		printk(KERN_INFO PFX "K7 660[A0] core detected, "
+				"enabling errata workarounds\n");
+		have_a0 = 1;
+	}
+
+	cpuid(0x80000007, &eax, &ebx, &ecx, &edx);
+
+	/* Check we can actually do something before we say anything.*/
+	if (!(edx & (1 << 1 | 1 << 2)))
+		return 0;
+
+	printk(KERN_INFO PFX "PowerNOW! Technology present. Can scale: ");
+
+	if (edx & 1 << 1) {
+		printk("frequency");
+		can_scale_bus = 1;
+	}
+
+	if ((edx & (1 << 1 | 1 << 2)) == 0x6)
+		printk(" and ");
+
+	if (edx & 1 << 2) {
+		printk("voltage");
+		can_scale_vid = 1;
+	}
+
+	printk(".\n");
+	return 1;
+}
+
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+static void invalidate_entry(unsigned int entry)
+{
+	powernow_table[entry].frequency = CPUFREQ_ENTRY_INVALID;
+}
+#endif
+
+static int get_ranges(unsigned char *pst)
+{
+	unsigned int j;
+	unsigned int speed;
+	u8 fid, vid;
+
+	powernow_table = kzalloc((sizeof(struct cpufreq_frequency_table) *
+				(number_scales + 1)), GFP_KERNEL);
+	if (!powernow_table)
+		return -ENOMEM;
+
+	for (j = 0 ; j < number_scales; j++) {
+		fid = *pst++;
+
+		powernow_table[j].frequency = (fsb * fid_codes[fid]) / 10;
+		powernow_table[j].index = fid; /* lower 8 bits */
+
+		speed = powernow_table[j].frequency;
+
+		if ((fid_codes[fid] % 10) == 5) {
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+			if (have_a0 == 1)
+				invalidate_entry(j);
+#endif
+		}
+
+		if (speed < minimum_speed)
+			minimum_speed = speed;
+		if (speed > maximum_speed)
+			maximum_speed = speed;
+
+		vid = *pst++;
+		powernow_table[j].index |= (vid << 8); /* upper 8 bits */
+
+		pr_debug("   FID: 0x%x (%d.%dx [%dMHz])  "
+			 "VID: 0x%x (%d.%03dV)\n", fid, fid_codes[fid] / 10,
+			 fid_codes[fid] % 10, speed/1000, vid,
+			 mobile_vid_table[vid]/1000,
+			 mobile_vid_table[vid]%1000);
+	}
+	powernow_table[number_scales].frequency = CPUFREQ_TABLE_END;
+	powernow_table[number_scales].index = 0;
+
+	return 0;
+}
+
+
+static void change_FID(int fid)
+{
+	union msr_fidvidctl fidvidctl;
+
+	rdmsrl(MSR_K7_FID_VID_CTL, fidvidctl.val);
+	if (fidvidctl.bits.FID != fid) {
+		fidvidctl.bits.SGTC = latency;
+		fidvidctl.bits.FID = fid;
+		fidvidctl.bits.VIDC = 0;
+		fidvidctl.bits.FIDC = 1;
+		wrmsrl(MSR_K7_FID_VID_CTL, fidvidctl.val);
+	}
+}
+
+
+static void change_VID(int vid)
+{
+	union msr_fidvidctl fidvidctl;
+
+	rdmsrl(MSR_K7_FID_VID_CTL, fidvidctl.val);
+	if (fidvidctl.bits.VID != vid) {
+		fidvidctl.bits.SGTC = latency;
+		fidvidctl.bits.VID = vid;
+		fidvidctl.bits.FIDC = 0;
+		fidvidctl.bits.VIDC = 1;
+		wrmsrl(MSR_K7_FID_VID_CTL, fidvidctl.val);
+	}
+}
+
+
+static void change_speed(unsigned int index)
+{
+	u8 fid, vid;
+	struct cpufreq_freqs freqs;
+	union msr_fidvidstatus fidvidstatus;
+	int cfid;
+
+	/* fid are the lower 8 bits of the index we stored into
+	 * the cpufreq frequency table in powernow_decode_bios,
+	 * vid are the upper 8 bits.
+	 */
+
+	fid = powernow_table[index].index & 0xFF;
+	vid = (powernow_table[index].index & 0xFF00) >> 8;
+
+	freqs.cpu = 0;
+
+	rdmsrl(MSR_K7_FID_VID_STATUS, fidvidstatus.val);
+	cfid = fidvidstatus.bits.CFID;
+	freqs.old = fsb * fid_codes[cfid] / 10;
+
+	freqs.new = powernow_table[index].frequency;
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	/* Now do the magic poking into the MSRs.  */
+
+	if (have_a0 == 1)	/* A0 errata 5 */
+		local_irq_disable();
+
+	if (freqs.old > freqs.new) {
+		/* Going down, so change FID first */
+		change_FID(fid);
+		change_VID(vid);
+	} else {
+		/* Going up, so change VID first */
+		change_VID(vid);
+		change_FID(fid);
+	}
+
+
+	if (have_a0 == 1)
+		local_irq_enable();
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+}
+
+
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+
+static struct acpi_processor_performance *acpi_processor_perf;
+
+static int powernow_acpi_init(void)
+{
+	int i;
+	int retval = 0;
+	union powernow_acpi_control_t pc;
+
+	if (acpi_processor_perf != NULL && powernow_table != NULL) {
+		retval = -EINVAL;
+		goto err0;
+	}
+
+	acpi_processor_perf = kzalloc(sizeof(struct acpi_processor_performance),
+				      GFP_KERNEL);
+	if (!acpi_processor_perf) {
+		retval = -ENOMEM;
+		goto err0;
+	}
+
+	if (!zalloc_cpumask_var(&acpi_processor_perf->shared_cpu_map,
+								GFP_KERNEL)) {
+		retval = -ENOMEM;
+		goto err05;
+	}
+
+	if (acpi_processor_register_performance(acpi_processor_perf, 0)) {
+		retval = -EIO;
+		goto err1;
+	}
+
+	if (acpi_processor_perf->control_register.space_id !=
+			ACPI_ADR_SPACE_FIXED_HARDWARE) {
+		retval = -ENODEV;
+		goto err2;
+	}
+
+	if (acpi_processor_perf->status_register.space_id !=
+			ACPI_ADR_SPACE_FIXED_HARDWARE) {
+		retval = -ENODEV;
+		goto err2;
+	}
+
+	number_scales = acpi_processor_perf->state_count;
+
+	if (number_scales < 2) {
+		retval = -ENODEV;
+		goto err2;
+	}
+
+	powernow_table = kzalloc((sizeof(struct cpufreq_frequency_table) *
+				(number_scales + 1)), GFP_KERNEL);
+	if (!powernow_table) {
+		retval = -ENOMEM;
+		goto err2;
+	}
+
+	pc.val = (unsigned long) acpi_processor_perf->states[0].control;
+	for (i = 0; i < number_scales; i++) {
+		u8 fid, vid;
+		struct acpi_processor_px *state =
+			&acpi_processor_perf->states[i];
+		unsigned int speed, speed_mhz;
+
+		pc.val = (unsigned long) state->control;
+		pr_debug("acpi:  P%d: %d MHz %d mW %d uS control %08x SGTC %d\n",
+			 i,
+			 (u32) state->core_frequency,
+			 (u32) state->power,
+			 (u32) state->transition_latency,
+			 (u32) state->control,
+			 pc.bits.sgtc);
+
+		vid = pc.bits.vid;
+		fid = pc.bits.fid;
+
+		powernow_table[i].frequency = fsb * fid_codes[fid] / 10;
+		powernow_table[i].index = fid; /* lower 8 bits */
+		powernow_table[i].index |= (vid << 8); /* upper 8 bits */
+
+		speed = powernow_table[i].frequency;
+		speed_mhz = speed / 1000;
+
+		/* processor_perflib will multiply the MHz value by 1000 to
+		 * get a KHz value (e.g. 1266000). However, powernow-k7 works
+		 * with true KHz values (e.g. 1266768). To ensure that all
+		 * powernow frequencies are available, we must ensure that
+		 * ACPI doesn't restrict them, so we round up the MHz value
+		 * to ensure that perflib's computed KHz value is greater than
+		 * or equal to powernow's KHz value.
+		 */
+		if (speed % 1000 > 0)
+			speed_mhz++;
+
+		if ((fid_codes[fid] % 10) == 5) {
+			if (have_a0 == 1)
+				invalidate_entry(i);
+		}
+
+		pr_debug("   FID: 0x%x (%d.%dx [%dMHz])  "
+			 "VID: 0x%x (%d.%03dV)\n", fid, fid_codes[fid] / 10,
+			 fid_codes[fid] % 10, speed_mhz, vid,
+			 mobile_vid_table[vid]/1000,
+			 mobile_vid_table[vid]%1000);
+
+		if (state->core_frequency != speed_mhz) {
+			state->core_frequency = speed_mhz;
+			pr_debug("   Corrected ACPI frequency to %d\n",
+				speed_mhz);
+		}
+
+		if (latency < pc.bits.sgtc)
+			latency = pc.bits.sgtc;
+
+		if (speed < minimum_speed)
+			minimum_speed = speed;
+		if (speed > maximum_speed)
+			maximum_speed = speed;
+	}
+
+	powernow_table[i].frequency = CPUFREQ_TABLE_END;
+	powernow_table[i].index = 0;
+
+	/* notify BIOS that we exist */
+	acpi_processor_notify_smm(THIS_MODULE);
+
+	return 0;
+
+err2:
+	acpi_processor_unregister_performance(acpi_processor_perf, 0);
+err1:
+	free_cpumask_var(acpi_processor_perf->shared_cpu_map);
+err05:
+	kfree(acpi_processor_perf);
+err0:
+	printk(KERN_WARNING PFX "ACPI perflib can not be used on "
+			"this platform\n");
+	acpi_processor_perf = NULL;
+	return retval;
+}
+#else
+static int powernow_acpi_init(void)
+{
+	printk(KERN_INFO PFX "no support for ACPI processor found."
+	       "  Please recompile your kernel with ACPI processor\n");
+	return -EINVAL;
+}
+#endif
+
+static void print_pst_entry(struct pst_s *pst, unsigned int j)
+{
+	pr_debug("PST:%d (@%p)\n", j, pst);
+	pr_debug(" cpuid: 0x%x  fsb: %d  maxFID: 0x%x  startvid: 0x%x\n",
+		pst->cpuid, pst->fsbspeed, pst->maxfid, pst->startvid);
+}
+
+static int powernow_decode_bios(int maxfid, int startvid)
+{
+	struct psb_s *psb;
+	struct pst_s *pst;
+	unsigned int i, j;
+	unsigned char *p;
+	unsigned int etuple;
+	unsigned int ret;
+
+	etuple = cpuid_eax(0x80000001);
+
+	for (i = 0xC0000; i < 0xffff0 ; i += 16) {
+
+		p = phys_to_virt(i);
+
+		if (memcmp(p, "AMDK7PNOW!",  10) == 0) {
+			pr_debug("Found PSB header at %p\n", p);
+			psb = (struct psb_s *) p;
+			pr_debug("Table version: 0x%x\n", psb->tableversion);
+			if (psb->tableversion != 0x12) {
+				printk(KERN_INFO PFX "Sorry, only v1.2 tables"
+						" supported right now\n");
+				return -ENODEV;
+			}
+
+			pr_debug("Flags: 0x%x\n", psb->flags);
+			if ((psb->flags & 1) == 0)
+				pr_debug("Mobile voltage regulator\n");
+			else
+				pr_debug("Desktop voltage regulator\n");
+
+			latency = psb->settlingtime;
+			if (latency < 100) {
+				printk(KERN_INFO PFX "BIOS set settling time "
+						"to %d microseconds. "
+						"Should be at least 100. "
+						"Correcting.\n", latency);
+				latency = 100;
+			}
+			pr_debug("Settling Time: %d microseconds.\n",
+					psb->settlingtime);
+			pr_debug("Has %d PST tables. (Only dumping ones "
+					"relevant to this CPU).\n",
+					psb->numpst);
+
+			p += sizeof(struct psb_s);
+
+			pst = (struct pst_s *) p;
+
+			for (j = 0; j < psb->numpst; j++) {
+				pst = (struct pst_s *) p;
+				number_scales = pst->numpstates;
+
+				if ((etuple == pst->cpuid) &&
+				    check_fsb(pst->fsbspeed) &&
+				    (maxfid == pst->maxfid) &&
+				    (startvid == pst->startvid)) {
+					print_pst_entry(pst, j);
+					p = (char *)pst + sizeof(struct pst_s);
+					ret = get_ranges(p);
+					return ret;
+				} else {
+					unsigned int k;
+					p = (char *)pst + sizeof(struct pst_s);
+					for (k = 0; k < number_scales; k++)
+						p += 2;
+				}
+			}
+			printk(KERN_INFO PFX "No PST tables match this cpuid "
+					"(0x%x)\n", etuple);
+			printk(KERN_INFO PFX "This is indicative of a broken "
+					"BIOS.\n");
+
+			return -EINVAL;
+		}
+		p++;
+	}
+
+	return -ENODEV;
+}
+
+
+static int powernow_target(struct cpufreq_policy *policy,
+			    unsigned int target_freq,
+			    unsigned int relation)
+{
+	unsigned int newstate;
+
+	if (cpufreq_frequency_table_target(policy, powernow_table, target_freq,
+				relation, &newstate))
+		return -EINVAL;
+
+	change_speed(newstate);
+
+	return 0;
+}
+
+
+static int powernow_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, powernow_table);
+}
+
+/*
+ * We use the fact that the bus frequency is somehow
+ * a multiple of 100000/3 khz, then we compute sgtc according
+ * to this multiple.
+ * That way, we match more how AMD thinks all of that work.
+ * We will then get the same kind of behaviour already tested under
+ * the "well-known" other OS.
+ */
+static int __cpuinit fixup_sgtc(void)
+{
+	unsigned int sgtc;
+	unsigned int m;
+
+	m = fsb / 3333;
+	if ((m % 10) >= 5)
+		m += 5;
+
+	m /= 10;
+
+	sgtc = 100 * m * latency;
+	sgtc = sgtc / 3;
+	if (sgtc > 0xfffff) {
+		printk(KERN_WARNING PFX "SGTC too large %d\n", sgtc);
+		sgtc = 0xfffff;
+	}
+	return sgtc;
+}
+
+static unsigned int powernow_get(unsigned int cpu)
+{
+	union msr_fidvidstatus fidvidstatus;
+	unsigned int cfid;
+
+	if (cpu)
+		return 0;
+	rdmsrl(MSR_K7_FID_VID_STATUS, fidvidstatus.val);
+	cfid = fidvidstatus.bits.CFID;
+
+	return fsb * fid_codes[cfid] / 10;
+}
+
+
+static int __cpuinit acer_cpufreq_pst(const struct dmi_system_id *d)
+{
+	printk(KERN_WARNING PFX
+		"%s laptop with broken PST tables in BIOS detected.\n",
+		d->ident);
+	printk(KERN_WARNING PFX
+		"You need to downgrade to 3A21 (09/09/2002), or try a newer "
+		"BIOS than 3A71 (01/20/2003)\n");
+	printk(KERN_WARNING PFX
+		"cpufreq scaling has been disabled as a result of this.\n");
+	return 0;
+}
+
+/*
+ * Some Athlon laptops have really fucked PST tables.
+ * A BIOS update is all that can save them.
+ * Mention this, and disable cpufreq.
+ */
+static struct dmi_system_id __cpuinitdata powernow_dmi_table[] = {
+	{
+		.callback = acer_cpufreq_pst,
+		.ident = "Acer Aspire",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Insyde Software"),
+			DMI_MATCH(DMI_BIOS_VERSION, "3A71"),
+		},
+	},
+	{ }
+};
+
+static int __cpuinit powernow_cpu_init(struct cpufreq_policy *policy)
+{
+	union msr_fidvidstatus fidvidstatus;
+	int result;
+
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	rdmsrl(MSR_K7_FID_VID_STATUS, fidvidstatus.val);
+
+	recalibrate_cpu_khz();
+
+	fsb = (10 * cpu_khz) / fid_codes[fidvidstatus.bits.CFID];
+	if (!fsb) {
+		printk(KERN_WARNING PFX "can not determine bus frequency\n");
+		return -EINVAL;
+	}
+	pr_debug("FSB: %3dMHz\n", fsb/1000);
+
+	if (dmi_check_system(powernow_dmi_table) || acpi_force) {
+		printk(KERN_INFO PFX "PSB/PST known to be broken.  "
+				"Trying ACPI instead\n");
+		result = powernow_acpi_init();
+	} else {
+		result = powernow_decode_bios(fidvidstatus.bits.MFID,
+				fidvidstatus.bits.SVID);
+		if (result) {
+			printk(KERN_INFO PFX "Trying ACPI perflib\n");
+			maximum_speed = 0;
+			minimum_speed = -1;
+			latency = 0;
+			result = powernow_acpi_init();
+			if (result) {
+				printk(KERN_INFO PFX
+					"ACPI and legacy methods failed\n");
+			}
+		} else {
+			/* SGTC use the bus clock as timer */
+			latency = fixup_sgtc();
+			printk(KERN_INFO PFX "SGTC: %d\n", latency);
+		}
+	}
+
+	if (result)
+		return result;
+
+	printk(KERN_INFO PFX "Minimum speed %d MHz. Maximum speed %d MHz.\n",
+				minimum_speed/1000, maximum_speed/1000);
+
+	policy->cpuinfo.transition_latency =
+		cpufreq_scale(2000000UL, fsb, latency);
+
+	policy->cur = powernow_get(0);
+
+	cpufreq_frequency_table_get_attr(powernow_table, policy->cpu);
+
+	return cpufreq_frequency_table_cpuinfo(policy, powernow_table);
+}
+
+static int powernow_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+	if (acpi_processor_perf) {
+		acpi_processor_unregister_performance(acpi_processor_perf, 0);
+		free_cpumask_var(acpi_processor_perf->shared_cpu_map);
+		kfree(acpi_processor_perf);
+	}
+#endif
+
+	kfree(powernow_table);
+	return 0;
+}
+
+static struct freq_attr *powernow_table_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver powernow_driver = {
+	.verify		= powernow_verify,
+	.target		= powernow_target,
+	.get		= powernow_get,
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+	.bios_limit	= acpi_processor_get_bios_limit,
+#endif
+	.init		= powernow_cpu_init,
+	.exit		= powernow_cpu_exit,
+	.name		= "powernow-k7",
+	.owner		= THIS_MODULE,
+	.attr		= powernow_table_attr,
+};
+
+static int __init powernow_init(void)
+{
+	if (check_powernow() == 0)
+		return -ENODEV;
+	return cpufreq_register_driver(&powernow_driver);
+}
+
+
+static void __exit powernow_exit(void)
+{
+	cpufreq_unregister_driver(&powernow_driver);
+}
+
+module_param(acpi_force,  int, 0444);
+MODULE_PARM_DESC(acpi_force, "Force ACPI to be used.");
+
+MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_DESCRIPTION("Powernow driver for AMD K7 processors.");
+MODULE_LICENSE("GPL");
+
+late_initcall(powernow_init);
+module_exit(powernow_exit);
+
diff --git a/drivers/cpufreq/powernow-k7.h b/drivers/cpufreq/powernow-k7.h
new file mode 100644
index 00000000..35fb4eaf
--- /dev/null
+++ b/drivers/cpufreq/powernow-k7.h
@@ -0,0 +1,43 @@
+/*
+ *  (C) 2003 Dave Jones.
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  AMD-specific information
+ *
+ */
+
+union msr_fidvidctl {
+	struct {
+		unsigned FID:5,			// 4:0
+		reserved1:3,	// 7:5
+		VID:5,			// 12:8
+		reserved2:3,	// 15:13
+		FIDC:1,			// 16
+		VIDC:1,			// 17
+		reserved3:2,	// 19:18
+		FIDCHGRATIO:1,	// 20
+		reserved4:11,	// 31-21
+		SGTC:20,		// 32:51
+		reserved5:12;	// 63:52
+	} bits;
+	unsigned long long val;
+};
+
+union msr_fidvidstatus {
+	struct {
+		unsigned CFID:5,			// 4:0
+		reserved1:3,	// 7:5
+		SFID:5,			// 12:8
+		reserved2:3,	// 15:13
+		MFID:5,			// 20:16
+		reserved3:11,	// 31:21
+		CVID:5,			// 36:32
+		reserved4:3,	// 39:37
+		SVID:5,			// 44:40
+		reserved5:3,	// 47:45
+		MVID:5,			// 52:48
+		reserved6:11;	// 63:53
+	} bits;
+	unsigned long long val;
+};
diff --git a/drivers/cpufreq/powernow-k8.c b/drivers/cpufreq/powernow-k8.c
new file mode 100644
index 00000000..ad683ec2
--- /dev/null
+++ b/drivers/cpufreq/powernow-k8.c
@@ -0,0 +1,1619 @@
+/*
+ *   (c) 2003-2010 Advanced Micro Devices, Inc.
+ *  Your use of this code is subject to the terms and conditions of the
+ *  GNU general public license version 2. See "COPYING" or
+ *  http://www.gnu.org/licenses/gpl.html
+ *
+ *  Support : mark.langsdorf@amd.com
+ *
+ *  Based on the powernow-k7.c module written by Dave Jones.
+ *  (C) 2003 Dave Jones on behalf of SuSE Labs
+ *  (C) 2004 Dominik Brodowski <linux@brodo.de>
+ *  (C) 2004 Pavel Machek <pavel@ucw.cz>
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *  Based upon datasheets & sample CPUs kindly provided by AMD.
+ *
+ *  Valuable input gratefully received from Dave Jones, Pavel Machek,
+ *  Dominik Brodowski, Jacob Shin, and others.
+ *  Originally developed by Paul Devriendt.
+ *  Processor information obtained from Chapter 9 (Power and Thermal Management)
+ *  of the "BIOS and Kernel Developer's Guide for the AMD Athlon 64 and AMD
+ *  Opteron Processors" available for download from www.amd.com
+ *
+ *  Tables for specific CPUs can be inferred from
+ *     http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/30430.pdf
+ */
+
+#include <linux/kernel.h>
+#include <linux/smp.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/cpumask.h>
+#include <linux/sched.h>	/* for current / set_cpus_allowed() */
+#include <linux/io.h>
+#include <linux/delay.h>
+
+#include <asm/msr.h>
+
+#include <linux/acpi.h>
+#include <linux/mutex.h>
+#include <acpi/processor.h>
+
+#define PFX "powernow-k8: "
+#define VERSION "version 2.20.00"
+#include "powernow-k8.h"
+#include "mperf.h"
+
+/* serialize freq changes  */
+static DEFINE_MUTEX(fidvid_mutex);
+
+static DEFINE_PER_CPU(struct powernow_k8_data *, powernow_data);
+
+static int cpu_family = CPU_OPTERON;
+
+/* array to map SW pstate number to acpi state */
+static u32 ps_to_as[8];
+
+/* core performance boost */
+static bool cpb_capable, cpb_enabled;
+static struct msr __percpu *msrs;
+
+static struct cpufreq_driver cpufreq_amd64_driver;
+
+#ifndef CONFIG_SMP
+static inline const struct cpumask *cpu_core_mask(int cpu)
+{
+	return cpumask_of(0);
+}
+#endif
+
+/* Return a frequency in MHz, given an input fid */
+static u32 find_freq_from_fid(u32 fid)
+{
+	return 800 + (fid * 100);
+}
+
+/* Return a frequency in KHz, given an input fid */
+static u32 find_khz_freq_from_fid(u32 fid)
+{
+	return 1000 * find_freq_from_fid(fid);
+}
+
+static u32 find_khz_freq_from_pstate(struct cpufreq_frequency_table *data,
+				     u32 pstate)
+{
+	return data[ps_to_as[pstate]].frequency;
+}
+
+/* Return the vco fid for an input fid
+ *
+ * Each "low" fid has corresponding "high" fid, and you can get to "low" fids
+ * only from corresponding high fids. This returns "high" fid corresponding to
+ * "low" one.
+ */
+static u32 convert_fid_to_vco_fid(u32 fid)
+{
+	if (fid < HI_FID_TABLE_BOTTOM)
+		return 8 + (2 * fid);
+	else
+		return fid;
+}
+
+/*
+ * Return 1 if the pending bit is set. Unless we just instructed the processor
+ * to transition to a new state, seeing this bit set is really bad news.
+ */
+static int pending_bit_stuck(void)
+{
+	u32 lo, hi;
+
+	if (cpu_family == CPU_HW_PSTATE)
+		return 0;
+
+	rdmsr(MSR_FIDVID_STATUS, lo, hi);
+	return lo & MSR_S_LO_CHANGE_PENDING ? 1 : 0;
+}
+
+/*
+ * Update the global current fid / vid values from the status msr.
+ * Returns 1 on error.
+ */
+static int query_current_values_with_pending_wait(struct powernow_k8_data *data)
+{
+	u32 lo, hi;
+	u32 i = 0;
+
+	if (cpu_family == CPU_HW_PSTATE) {
+		rdmsr(MSR_PSTATE_STATUS, lo, hi);
+		i = lo & HW_PSTATE_MASK;
+		data->currpstate = i;
+
+		/*
+		 * a workaround for family 11h erratum 311 might cause
+		 * an "out-of-range Pstate if the core is in Pstate-0
+		 */
+		if ((boot_cpu_data.x86 == 0x11) && (i >= data->numps))
+			data->currpstate = HW_PSTATE_0;
+
+		return 0;
+	}
+	do {
+		if (i++ > 10000) {
+			pr_debug("detected change pending stuck\n");
+			return 1;
+		}
+		rdmsr(MSR_FIDVID_STATUS, lo, hi);
+	} while (lo & MSR_S_LO_CHANGE_PENDING);
+
+	data->currvid = hi & MSR_S_HI_CURRENT_VID;
+	data->currfid = lo & MSR_S_LO_CURRENT_FID;
+
+	return 0;
+}
+
+/* the isochronous relief time */
+static void count_off_irt(struct powernow_k8_data *data)
+{
+	udelay((1 << data->irt) * 10);
+	return;
+}
+
+/* the voltage stabilization time */
+static void count_off_vst(struct powernow_k8_data *data)
+{
+	udelay(data->vstable * VST_UNITS_20US);
+	return;
+}
+
+/* need to init the control msr to a safe value (for each cpu) */
+static void fidvid_msr_init(void)
+{
+	u32 lo, hi;
+	u8 fid, vid;
+
+	rdmsr(MSR_FIDVID_STATUS, lo, hi);
+	vid = hi & MSR_S_HI_CURRENT_VID;
+	fid = lo & MSR_S_LO_CURRENT_FID;
+	lo = fid | (vid << MSR_C_LO_VID_SHIFT);
+	hi = MSR_C_HI_STP_GNT_BENIGN;
+	pr_debug("cpu%d, init lo 0x%x, hi 0x%x\n", smp_processor_id(), lo, hi);
+	wrmsr(MSR_FIDVID_CTL, lo, hi);
+}
+
+/* write the new fid value along with the other control fields to the msr */
+static int write_new_fid(struct powernow_k8_data *data, u32 fid)
+{
+	u32 lo;
+	u32 savevid = data->currvid;
+	u32 i = 0;
+
+	if ((fid & INVALID_FID_MASK) || (data->currvid & INVALID_VID_MASK)) {
+		printk(KERN_ERR PFX "internal error - overflow on fid write\n");
+		return 1;
+	}
+
+	lo = fid;
+	lo |= (data->currvid << MSR_C_LO_VID_SHIFT);
+	lo |= MSR_C_LO_INIT_FID_VID;
+
+	pr_debug("writing fid 0x%x, lo 0x%x, hi 0x%x\n",
+		fid, lo, data->plllock * PLL_LOCK_CONVERSION);
+
+	do {
+		wrmsr(MSR_FIDVID_CTL, lo, data->plllock * PLL_LOCK_CONVERSION);
+		if (i++ > 100) {
+			printk(KERN_ERR PFX
+				"Hardware error - pending bit very stuck - "
+				"no further pstate changes possible\n");
+			return 1;
+		}
+	} while (query_current_values_with_pending_wait(data));
+
+	count_off_irt(data);
+
+	if (savevid != data->currvid) {
+		printk(KERN_ERR PFX
+			"vid change on fid trans, old 0x%x, new 0x%x\n",
+			savevid, data->currvid);
+		return 1;
+	}
+
+	if (fid != data->currfid) {
+		printk(KERN_ERR PFX
+			"fid trans failed, fid 0x%x, curr 0x%x\n", fid,
+			data->currfid);
+		return 1;
+	}
+
+	return 0;
+}
+
+/* Write a new vid to the hardware */
+static int write_new_vid(struct powernow_k8_data *data, u32 vid)
+{
+	u32 lo;
+	u32 savefid = data->currfid;
+	int i = 0;
+
+	if ((data->currfid & INVALID_FID_MASK) || (vid & INVALID_VID_MASK)) {
+		printk(KERN_ERR PFX "internal error - overflow on vid write\n");
+		return 1;
+	}
+
+	lo = data->currfid;
+	lo |= (vid << MSR_C_LO_VID_SHIFT);
+	lo |= MSR_C_LO_INIT_FID_VID;
+
+	pr_debug("writing vid 0x%x, lo 0x%x, hi 0x%x\n",
+		vid, lo, STOP_GRANT_5NS);
+
+	do {
+		wrmsr(MSR_FIDVID_CTL, lo, STOP_GRANT_5NS);
+		if (i++ > 100) {
+			printk(KERN_ERR PFX "internal error - pending bit "
+					"very stuck - no further pstate "
+					"changes possible\n");
+			return 1;
+		}
+	} while (query_current_values_with_pending_wait(data));
+
+	if (savefid != data->currfid) {
+		printk(KERN_ERR PFX "fid changed on vid trans, old "
+			"0x%x new 0x%x\n",
+		       savefid, data->currfid);
+		return 1;
+	}
+
+	if (vid != data->currvid) {
+		printk(KERN_ERR PFX "vid trans failed, vid 0x%x, "
+				"curr 0x%x\n",
+				vid, data->currvid);
+		return 1;
+	}
+
+	return 0;
+}
+
+/*
+ * Reduce the vid by the max of step or reqvid.
+ * Decreasing vid codes represent increasing voltages:
+ * vid of 0 is 1.550V, vid of 0x1e is 0.800V, vid of VID_OFF is off.
+ */
+static int decrease_vid_code_by_step(struct powernow_k8_data *data,
+		u32 reqvid, u32 step)
+{
+	if ((data->currvid - reqvid) > step)
+		reqvid = data->currvid - step;
+
+	if (write_new_vid(data, reqvid))
+		return 1;
+
+	count_off_vst(data);
+
+	return 0;
+}
+
+/* Change hardware pstate by single MSR write */
+static int transition_pstate(struct powernow_k8_data *data, u32 pstate)
+{
+	wrmsr(MSR_PSTATE_CTRL, pstate, 0);
+	data->currpstate = pstate;
+	return 0;
+}
+
+/* Change Opteron/Athlon64 fid and vid, by the 3 phases. */
+static int transition_fid_vid(struct powernow_k8_data *data,
+		u32 reqfid, u32 reqvid)
+{
+	if (core_voltage_pre_transition(data, reqvid, reqfid))
+		return 1;
+
+	if (core_frequency_transition(data, reqfid))
+		return 1;
+
+	if (core_voltage_post_transition(data, reqvid))
+		return 1;
+
+	if (query_current_values_with_pending_wait(data))
+		return 1;
+
+	if ((reqfid != data->currfid) || (reqvid != data->currvid)) {
+		printk(KERN_ERR PFX "failed (cpu%d): req 0x%x 0x%x, "
+				"curr 0x%x 0x%x\n",
+				smp_processor_id(),
+				reqfid, reqvid, data->currfid, data->currvid);
+		return 1;
+	}
+
+	pr_debug("transitioned (cpu%d): new fid 0x%x, vid 0x%x\n",
+		smp_processor_id(), data->currfid, data->currvid);
+
+	return 0;
+}
+
+/* Phase 1 - core voltage transition ... setup voltage */
+static int core_voltage_pre_transition(struct powernow_k8_data *data,
+		u32 reqvid, u32 reqfid)
+{
+	u32 rvosteps = data->rvo;
+	u32 savefid = data->currfid;
+	u32 maxvid, lo, rvomult = 1;
+
+	pr_debug("ph1 (cpu%d): start, currfid 0x%x, currvid 0x%x, "
+		"reqvid 0x%x, rvo 0x%x\n",
+		smp_processor_id(),
+		data->currfid, data->currvid, reqvid, data->rvo);
+
+	if ((savefid < LO_FID_TABLE_TOP) && (reqfid < LO_FID_TABLE_TOP))
+		rvomult = 2;
+	rvosteps *= rvomult;
+	rdmsr(MSR_FIDVID_STATUS, lo, maxvid);
+	maxvid = 0x1f & (maxvid >> 16);
+	pr_debug("ph1 maxvid=0x%x\n", maxvid);
+	if (reqvid < maxvid) /* lower numbers are higher voltages */
+		reqvid = maxvid;
+
+	while (data->currvid > reqvid) {
+		pr_debug("ph1: curr 0x%x, req vid 0x%x\n",
+			data->currvid, reqvid);
+		if (decrease_vid_code_by_step(data, reqvid, data->vidmvs))
+			return 1;
+	}
+
+	while ((rvosteps > 0) &&
+			((rvomult * data->rvo + data->currvid) > reqvid)) {
+		if (data->currvid == maxvid) {
+			rvosteps = 0;
+		} else {
+			pr_debug("ph1: changing vid for rvo, req 0x%x\n",
+				data->currvid - 1);
+			if (decrease_vid_code_by_step(data, data->currvid-1, 1))
+				return 1;
+			rvosteps--;
+		}
+	}
+
+	if (query_current_values_with_pending_wait(data))
+		return 1;
+
+	if (savefid != data->currfid) {
+		printk(KERN_ERR PFX "ph1 err, currfid changed 0x%x\n",
+				data->currfid);
+		return 1;
+	}
+
+	pr_debug("ph1 complete, currfid 0x%x, currvid 0x%x\n",
+		data->currfid, data->currvid);
+
+	return 0;
+}
+
+/* Phase 2 - core frequency transition */
+static int core_frequency_transition(struct powernow_k8_data *data, u32 reqfid)
+{
+	u32 vcoreqfid, vcocurrfid, vcofiddiff;
+	u32 fid_interval, savevid = data->currvid;
+
+	if (data->currfid == reqfid) {
+		printk(KERN_ERR PFX "ph2 null fid transition 0x%x\n",
+				data->currfid);
+		return 0;
+	}
+
+	pr_debug("ph2 (cpu%d): starting, currfid 0x%x, currvid 0x%x, "
+		"reqfid 0x%x\n",
+		smp_processor_id(),
+		data->currfid, data->currvid, reqfid);
+
+	vcoreqfid = convert_fid_to_vco_fid(reqfid);
+	vcocurrfid = convert_fid_to_vco_fid(data->currfid);
+	vcofiddiff = vcocurrfid > vcoreqfid ? vcocurrfid - vcoreqfid
+	    : vcoreqfid - vcocurrfid;
+
+	if ((reqfid <= LO_FID_TABLE_TOP) && (data->currfid <= LO_FID_TABLE_TOP))
+		vcofiddiff = 0;
+
+	while (vcofiddiff > 2) {
+		(data->currfid & 1) ? (fid_interval = 1) : (fid_interval = 2);
+
+		if (reqfid > data->currfid) {
+			if (data->currfid > LO_FID_TABLE_TOP) {
+				if (write_new_fid(data,
+						data->currfid + fid_interval))
+					return 1;
+			} else {
+				if (write_new_fid
+				    (data,
+				     2 + convert_fid_to_vco_fid(data->currfid)))
+					return 1;
+			}
+		} else {
+			if (write_new_fid(data, data->currfid - fid_interval))
+				return 1;
+		}
+
+		vcocurrfid = convert_fid_to_vco_fid(data->currfid);
+		vcofiddiff = vcocurrfid > vcoreqfid ? vcocurrfid - vcoreqfid
+		    : vcoreqfid - vcocurrfid;
+	}
+
+	if (write_new_fid(data, reqfid))
+		return 1;
+
+	if (query_current_values_with_pending_wait(data))
+		return 1;
+
+	if (data->currfid != reqfid) {
+		printk(KERN_ERR PFX
+			"ph2: mismatch, failed fid transition, "
+			"curr 0x%x, req 0x%x\n",
+			data->currfid, reqfid);
+		return 1;
+	}
+
+	if (savevid != data->currvid) {
+		printk(KERN_ERR PFX "ph2: vid changed, save 0x%x, curr 0x%x\n",
+			savevid, data->currvid);
+		return 1;
+	}
+
+	pr_debug("ph2 complete, currfid 0x%x, currvid 0x%x\n",
+		data->currfid, data->currvid);
+
+	return 0;
+}
+
+/* Phase 3 - core voltage transition flow ... jump to the final vid. */
+static int core_voltage_post_transition(struct powernow_k8_data *data,
+		u32 reqvid)
+{
+	u32 savefid = data->currfid;
+	u32 savereqvid = reqvid;
+
+	pr_debug("ph3 (cpu%d): starting, currfid 0x%x, currvid 0x%x\n",
+		smp_processor_id(),
+		data->currfid, data->currvid);
+
+	if (reqvid != data->currvid) {
+		if (write_new_vid(data, reqvid))
+			return 1;
+
+		if (savefid != data->currfid) {
+			printk(KERN_ERR PFX
+			       "ph3: bad fid change, save 0x%x, curr 0x%x\n",
+			       savefid, data->currfid);
+			return 1;
+		}
+
+		if (data->currvid != reqvid) {
+			printk(KERN_ERR PFX
+			       "ph3: failed vid transition\n, "
+			       "req 0x%x, curr 0x%x",
+			       reqvid, data->currvid);
+			return 1;
+		}
+	}
+
+	if (query_current_values_with_pending_wait(data))
+		return 1;
+
+	if (savereqvid != data->currvid) {
+		pr_debug("ph3 failed, currvid 0x%x\n", data->currvid);
+		return 1;
+	}
+
+	if (savefid != data->currfid) {
+		pr_debug("ph3 failed, currfid changed 0x%x\n",
+			data->currfid);
+		return 1;
+	}
+
+	pr_debug("ph3 complete, currfid 0x%x, currvid 0x%x\n",
+		data->currfid, data->currvid);
+
+	return 0;
+}
+
+static void check_supported_cpu(void *_rc)
+{
+	u32 eax, ebx, ecx, edx;
+	int *rc = _rc;
+
+	*rc = -ENODEV;
+
+	if (__this_cpu_read(cpu_info.x86_vendor) != X86_VENDOR_AMD)
+		return;
+
+	eax = cpuid_eax(CPUID_PROCESSOR_SIGNATURE);
+	if (((eax & CPUID_XFAM) != CPUID_XFAM_K8) &&
+	    ((eax & CPUID_XFAM) < CPUID_XFAM_10H))
+		return;
+
+	if ((eax & CPUID_XFAM) == CPUID_XFAM_K8) {
+		if (((eax & CPUID_USE_XFAM_XMOD) != CPUID_USE_XFAM_XMOD) ||
+		    ((eax & CPUID_XMOD) > CPUID_XMOD_REV_MASK)) {
+			printk(KERN_INFO PFX
+				"Processor cpuid %x not supported\n", eax);
+			return;
+		}
+
+		eax = cpuid_eax(CPUID_GET_MAX_CAPABILITIES);
+		if (eax < CPUID_FREQ_VOLT_CAPABILITIES) {
+			printk(KERN_INFO PFX
+			       "No frequency change capabilities detected\n");
+			return;
+		}
+
+		cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx);
+		if ((edx & P_STATE_TRANSITION_CAPABLE)
+			!= P_STATE_TRANSITION_CAPABLE) {
+			printk(KERN_INFO PFX
+				"Power state transitions not supported\n");
+			return;
+		}
+	} else { /* must be a HW Pstate capable processor */
+		cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx);
+		if ((edx & USE_HW_PSTATE) == USE_HW_PSTATE)
+			cpu_family = CPU_HW_PSTATE;
+		else
+			return;
+	}
+
+	*rc = 0;
+}
+
+static int check_pst_table(struct powernow_k8_data *data, struct pst_s *pst,
+		u8 maxvid)
+{
+	unsigned int j;
+	u8 lastfid = 0xff;
+
+	for (j = 0; j < data->numps; j++) {
+		if (pst[j].vid > LEAST_VID) {
+			printk(KERN_ERR FW_BUG PFX "vid %d invalid : 0x%x\n",
+			       j, pst[j].vid);
+			return -EINVAL;
+		}
+		if (pst[j].vid < data->rvo) {
+			/* vid + rvo >= 0 */
+			printk(KERN_ERR FW_BUG PFX "0 vid exceeded with pstate"
+			       " %d\n", j);
+			return -ENODEV;
+		}
+		if (pst[j].vid < maxvid + data->rvo) {
+			/* vid + rvo >= maxvid */
+			printk(KERN_ERR FW_BUG PFX "maxvid exceeded with pstate"
+			       " %d\n", j);
+			return -ENODEV;
+		}
+		if (pst[j].fid > MAX_FID) {
+			printk(KERN_ERR FW_BUG PFX "maxfid exceeded with pstate"
+			       " %d\n", j);
+			return -ENODEV;
+		}
+		if (j && (pst[j].fid < HI_FID_TABLE_BOTTOM)) {
+			/* Only first fid is allowed to be in "low" range */
+			printk(KERN_ERR FW_BUG PFX "two low fids - %d : "
+			       "0x%x\n", j, pst[j].fid);
+			return -EINVAL;
+		}
+		if (pst[j].fid < lastfid)
+			lastfid = pst[j].fid;
+	}
+	if (lastfid & 1) {
+		printk(KERN_ERR FW_BUG PFX "lastfid invalid\n");
+		return -EINVAL;
+	}
+	if (lastfid > LO_FID_TABLE_TOP)
+		printk(KERN_INFO FW_BUG PFX
+			"first fid not from lo freq table\n");
+
+	return 0;
+}
+
+static void invalidate_entry(struct cpufreq_frequency_table *powernow_table,
+		unsigned int entry)
+{
+	powernow_table[entry].frequency = CPUFREQ_ENTRY_INVALID;
+}
+
+static void print_basics(struct powernow_k8_data *data)
+{
+	int j;
+	for (j = 0; j < data->numps; j++) {
+		if (data->powernow_table[j].frequency !=
+				CPUFREQ_ENTRY_INVALID) {
+			if (cpu_family == CPU_HW_PSTATE) {
+				printk(KERN_INFO PFX
+					"   %d : pstate %d (%d MHz)\n", j,
+					data->powernow_table[j].index,
+					data->powernow_table[j].frequency/1000);
+			} else {
+				printk(KERN_INFO PFX
+					"fid 0x%x (%d MHz), vid 0x%x\n",
+					data->powernow_table[j].index & 0xff,
+					data->powernow_table[j].frequency/1000,
+					data->powernow_table[j].index >> 8);
+			}
+		}
+	}
+	if (data->batps)
+		printk(KERN_INFO PFX "Only %d pstates on battery\n",
+				data->batps);
+}
+
+static u32 freq_from_fid_did(u32 fid, u32 did)
+{
+	u32 mhz = 0;
+
+	if (boot_cpu_data.x86 == 0x10)
+		mhz = (100 * (fid + 0x10)) >> did;
+	else if (boot_cpu_data.x86 == 0x11)
+		mhz = (100 * (fid + 8)) >> did;
+	else
+		BUG();
+
+	return mhz * 1000;
+}
+
+static int fill_powernow_table(struct powernow_k8_data *data,
+		struct pst_s *pst, u8 maxvid)
+{
+	struct cpufreq_frequency_table *powernow_table;
+	unsigned int j;
+
+	if (data->batps) {
+		/* use ACPI support to get full speed on mains power */
+		printk(KERN_WARNING PFX
+			"Only %d pstates usable (use ACPI driver for full "
+			"range\n", data->batps);
+		data->numps = data->batps;
+	}
+
+	for (j = 1; j < data->numps; j++) {
+		if (pst[j-1].fid >= pst[j].fid) {
+			printk(KERN_ERR PFX "PST out of sequence\n");
+			return -EINVAL;
+		}
+	}
+
+	if (data->numps < 2) {
+		printk(KERN_ERR PFX "no p states to transition\n");
+		return -ENODEV;
+	}
+
+	if (check_pst_table(data, pst, maxvid))
+		return -EINVAL;
+
+	powernow_table = kmalloc((sizeof(struct cpufreq_frequency_table)
+		* (data->numps + 1)), GFP_KERNEL);
+	if (!powernow_table) {
+		printk(KERN_ERR PFX "powernow_table memory alloc failure\n");
+		return -ENOMEM;
+	}
+
+	for (j = 0; j < data->numps; j++) {
+		int freq;
+		powernow_table[j].index = pst[j].fid; /* lower 8 bits */
+		powernow_table[j].index |= (pst[j].vid << 8); /* upper 8 bits */
+		freq = find_khz_freq_from_fid(pst[j].fid);
+		powernow_table[j].frequency = freq;
+	}
+	powernow_table[data->numps].frequency = CPUFREQ_TABLE_END;
+	powernow_table[data->numps].index = 0;
+
+	if (query_current_values_with_pending_wait(data)) {
+		kfree(powernow_table);
+		return -EIO;
+	}
+
+	pr_debug("cfid 0x%x, cvid 0x%x\n", data->currfid, data->currvid);
+	data->powernow_table = powernow_table;
+	if (cpumask_first(cpu_core_mask(data->cpu)) == data->cpu)
+		print_basics(data);
+
+	for (j = 0; j < data->numps; j++)
+		if ((pst[j].fid == data->currfid) &&
+		    (pst[j].vid == data->currvid))
+			return 0;
+
+	pr_debug("currfid/vid do not match PST, ignoring\n");
+	return 0;
+}
+
+/* Find and validate the PSB/PST table in BIOS. */
+static int find_psb_table(struct powernow_k8_data *data)
+{
+	struct psb_s *psb;
+	unsigned int i;
+	u32 mvs;
+	u8 maxvid;
+	u32 cpst = 0;
+	u32 thiscpuid;
+
+	for (i = 0xc0000; i < 0xffff0; i += 0x10) {
+		/* Scan BIOS looking for the signature. */
+		/* It can not be at ffff0 - it is too big. */
+
+		psb = phys_to_virt(i);
+		if (memcmp(psb, PSB_ID_STRING, PSB_ID_STRING_LEN) != 0)
+			continue;
+
+		pr_debug("found PSB header at 0x%p\n", psb);
+
+		pr_debug("table vers: 0x%x\n", psb->tableversion);
+		if (psb->tableversion != PSB_VERSION_1_4) {
+			printk(KERN_ERR FW_BUG PFX "PSB table is not v1.4\n");
+			return -ENODEV;
+		}
+
+		pr_debug("flags: 0x%x\n", psb->flags1);
+		if (psb->flags1) {
+			printk(KERN_ERR FW_BUG PFX "unknown flags\n");
+			return -ENODEV;
+		}
+
+		data->vstable = psb->vstable;
+		pr_debug("voltage stabilization time: %d(*20us)\n",
+				data->vstable);
+
+		pr_debug("flags2: 0x%x\n", psb->flags2);
+		data->rvo = psb->flags2 & 3;
+		data->irt = ((psb->flags2) >> 2) & 3;
+		mvs = ((psb->flags2) >> 4) & 3;
+		data->vidmvs = 1 << mvs;
+		data->batps = ((psb->flags2) >> 6) & 3;
+
+		pr_debug("ramp voltage offset: %d\n", data->rvo);
+		pr_debug("isochronous relief time: %d\n", data->irt);
+		pr_debug("maximum voltage step: %d - 0x%x\n", mvs, data->vidmvs);
+
+		pr_debug("numpst: 0x%x\n", psb->num_tables);
+		cpst = psb->num_tables;
+		if ((psb->cpuid == 0x00000fc0) ||
+		    (psb->cpuid == 0x00000fe0)) {
+			thiscpuid = cpuid_eax(CPUID_PROCESSOR_SIGNATURE);
+			if ((thiscpuid == 0x00000fc0) ||
+			    (thiscpuid == 0x00000fe0))
+				cpst = 1;
+		}
+		if (cpst != 1) {
+			printk(KERN_ERR FW_BUG PFX "numpst must be 1\n");
+			return -ENODEV;
+		}
+
+		data->plllock = psb->plllocktime;
+		pr_debug("plllocktime: 0x%x (units 1us)\n", psb->plllocktime);
+		pr_debug("maxfid: 0x%x\n", psb->maxfid);
+		pr_debug("maxvid: 0x%x\n", psb->maxvid);
+		maxvid = psb->maxvid;
+
+		data->numps = psb->numps;
+		pr_debug("numpstates: 0x%x\n", data->numps);
+		return fill_powernow_table(data,
+				(struct pst_s *)(psb+1), maxvid);
+	}
+	/*
+	 * If you see this message, complain to BIOS manufacturer. If
+	 * he tells you "we do not support Linux" or some similar
+	 * nonsense, remember that Windows 2000 uses the same legacy
+	 * mechanism that the old Linux PSB driver uses. Tell them it
+	 * is broken with Windows 2000.
+	 *
+	 * The reference to the AMD documentation is chapter 9 in the
+	 * BIOS and Kernel Developer's Guide, which is available on
+	 * www.amd.com
+	 */
+	printk(KERN_ERR FW_BUG PFX "No PSB or ACPI _PSS objects\n");
+	printk(KERN_ERR PFX "Make sure that your BIOS is up to date"
+		" and Cool'N'Quiet support is enabled in BIOS setup\n");
+	return -ENODEV;
+}
+
+static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data,
+		unsigned int index)
+{
+	u64 control;
+
+	if (!data->acpi_data.state_count || (cpu_family == CPU_HW_PSTATE))
+		return;
+
+	control = data->acpi_data.states[index].control;
+	data->irt = (control >> IRT_SHIFT) & IRT_MASK;
+	data->rvo = (control >> RVO_SHIFT) & RVO_MASK;
+	data->exttype = (control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK;
+	data->plllock = (control >> PLL_L_SHIFT) & PLL_L_MASK;
+	data->vidmvs = 1 << ((control >> MVS_SHIFT) & MVS_MASK);
+	data->vstable = (control >> VST_SHIFT) & VST_MASK;
+}
+
+static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
+{
+	struct cpufreq_frequency_table *powernow_table;
+	int ret_val = -ENODEV;
+	u64 control, status;
+
+	if (acpi_processor_register_performance(&data->acpi_data, data->cpu)) {
+		pr_debug("register performance failed: bad ACPI data\n");
+		return -EIO;
+	}
+
+	/* verify the data contained in the ACPI structures */
+	if (data->acpi_data.state_count <= 1) {
+		pr_debug("No ACPI P-States\n");
+		goto err_out;
+	}
+
+	control = data->acpi_data.control_register.space_id;
+	status = data->acpi_data.status_register.space_id;
+
+	if ((control != ACPI_ADR_SPACE_FIXED_HARDWARE) ||
+	    (status != ACPI_ADR_SPACE_FIXED_HARDWARE)) {
+		pr_debug("Invalid control/status registers (%llx - %llx)\n",
+			control, status);
+		goto err_out;
+	}
+
+	/* fill in data->powernow_table */
+	powernow_table = kmalloc((sizeof(struct cpufreq_frequency_table)
+		* (data->acpi_data.state_count + 1)), GFP_KERNEL);
+	if (!powernow_table) {
+		pr_debug("powernow_table memory alloc failure\n");
+		goto err_out;
+	}
+
+	/* fill in data */
+	data->numps = data->acpi_data.state_count;
+	powernow_k8_acpi_pst_values(data, 0);
+
+	if (cpu_family == CPU_HW_PSTATE)
+		ret_val = fill_powernow_table_pstate(data, powernow_table);
+	else
+		ret_val = fill_powernow_table_fidvid(data, powernow_table);
+	if (ret_val)
+		goto err_out_mem;
+
+	powernow_table[data->acpi_data.state_count].frequency =
+		CPUFREQ_TABLE_END;
+	powernow_table[data->acpi_data.state_count].index = 0;
+	data->powernow_table = powernow_table;
+
+	if (cpumask_first(cpu_core_mask(data->cpu)) == data->cpu)
+		print_basics(data);
+
+	/* notify BIOS that we exist */
+	acpi_processor_notify_smm(THIS_MODULE);
+
+	if (!zalloc_cpumask_var(&data->acpi_data.shared_cpu_map, GFP_KERNEL)) {
+		printk(KERN_ERR PFX
+				"unable to alloc powernow_k8_data cpumask\n");
+		ret_val = -ENOMEM;
+		goto err_out_mem;
+	}
+
+	return 0;
+
+err_out_mem:
+	kfree(powernow_table);
+
+err_out:
+	acpi_processor_unregister_performance(&data->acpi_data, data->cpu);
+
+	/* data->acpi_data.state_count informs us at ->exit()
+	 * whether ACPI was used */
+	data->acpi_data.state_count = 0;
+
+	return ret_val;
+}
+
+static int fill_powernow_table_pstate(struct powernow_k8_data *data,
+		struct cpufreq_frequency_table *powernow_table)
+{
+	int i;
+	u32 hi = 0, lo = 0;
+	rdmsr(MSR_PSTATE_CUR_LIMIT, lo, hi);
+	data->max_hw_pstate = (lo & HW_PSTATE_MAX_MASK) >> HW_PSTATE_MAX_SHIFT;
+
+	for (i = 0; i < data->acpi_data.state_count; i++) {
+		u32 index;
+
+		index = data->acpi_data.states[i].control & HW_PSTATE_MASK;
+		if (index > data->max_hw_pstate) {
+			printk(KERN_ERR PFX "invalid pstate %d - "
+					"bad value %d.\n", i, index);
+			printk(KERN_ERR PFX "Please report to BIOS "
+					"manufacturer\n");
+			invalidate_entry(powernow_table, i);
+			continue;
+		}
+
+		ps_to_as[index] = i;
+
+		/* Frequency may be rounded for these */
+		if ((boot_cpu_data.x86 == 0x10 && boot_cpu_data.x86_model < 10)
+				 || boot_cpu_data.x86 == 0x11) {
+
+			rdmsr(MSR_PSTATE_DEF_BASE + index, lo, hi);
+			if (!(hi & HW_PSTATE_VALID_MASK)) {
+				pr_debug("invalid pstate %d, ignoring\n", index);
+				invalidate_entry(powernow_table, i);
+				continue;
+			}
+
+			powernow_table[i].frequency =
+				freq_from_fid_did(lo & 0x3f, (lo >> 6) & 7);
+		} else
+			powernow_table[i].frequency =
+				data->acpi_data.states[i].core_frequency * 1000;
+
+		powernow_table[i].index = index;
+	}
+	return 0;
+}
+
+static int fill_powernow_table_fidvid(struct powernow_k8_data *data,
+		struct cpufreq_frequency_table *powernow_table)
+{
+	int i;
+
+	for (i = 0; i < data->acpi_data.state_count; i++) {
+		u32 fid;
+		u32 vid;
+		u32 freq, index;
+		u64 status, control;
+
+		if (data->exttype) {
+			status =  data->acpi_data.states[i].status;
+			fid = status & EXT_FID_MASK;
+			vid = (status >> VID_SHIFT) & EXT_VID_MASK;
+		} else {
+			control =  data->acpi_data.states[i].control;
+			fid = control & FID_MASK;
+			vid = (control >> VID_SHIFT) & VID_MASK;
+		}
+
+		pr_debug("   %d : fid 0x%x, vid 0x%x\n", i, fid, vid);
+
+		index = fid | (vid<<8);
+		powernow_table[i].index = index;
+
+		freq = find_khz_freq_from_fid(fid);
+		powernow_table[i].frequency = freq;
+
+		/* verify frequency is OK */
+		if ((freq > (MAX_FREQ * 1000)) || (freq < (MIN_FREQ * 1000))) {
+			pr_debug("invalid freq %u kHz, ignoring\n", freq);
+			invalidate_entry(powernow_table, i);
+			continue;
+		}
+
+		/* verify voltage is OK -
+		 * BIOSs are using "off" to indicate invalid */
+		if (vid == VID_OFF) {
+			pr_debug("invalid vid %u, ignoring\n", vid);
+			invalidate_entry(powernow_table, i);
+			continue;
+		}
+
+		if (freq != (data->acpi_data.states[i].core_frequency * 1000)) {
+			printk(KERN_INFO PFX "invalid freq entries "
+				"%u kHz vs. %u kHz\n", freq,
+				(unsigned int)
+				(data->acpi_data.states[i].core_frequency
+				 * 1000));
+			invalidate_entry(powernow_table, i);
+			continue;
+		}
+	}
+	return 0;
+}
+
+static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data)
+{
+	if (data->acpi_data.state_count)
+		acpi_processor_unregister_performance(&data->acpi_data,
+				data->cpu);
+	free_cpumask_var(data->acpi_data.shared_cpu_map);
+}
+
+static int get_transition_latency(struct powernow_k8_data *data)
+{
+	int max_latency = 0;
+	int i;
+	for (i = 0; i < data->acpi_data.state_count; i++) {
+		int cur_latency = data->acpi_data.states[i].transition_latency
+			+ data->acpi_data.states[i].bus_master_latency;
+		if (cur_latency > max_latency)
+			max_latency = cur_latency;
+	}
+	if (max_latency == 0) {
+		/*
+		 * Fam 11h and later may return 0 as transition latency. This
+		 * is intended and means "very fast". While cpufreq core and
+		 * governors currently can handle that gracefully, better set it
+		 * to 1 to avoid problems in the future.
+		 */
+		if (boot_cpu_data.x86 < 0x11)
+			printk(KERN_ERR FW_WARN PFX "Invalid zero transition "
+				"latency\n");
+		max_latency = 1;
+	}
+	/* value in usecs, needs to be in nanoseconds */
+	return 1000 * max_latency;
+}
+
+/* Take a frequency, and issue the fid/vid transition command */
+static int transition_frequency_fidvid(struct powernow_k8_data *data,
+		unsigned int index)
+{
+	u32 fid = 0;
+	u32 vid = 0;
+	int res, i;
+	struct cpufreq_freqs freqs;
+
+	pr_debug("cpu %d transition to index %u\n", smp_processor_id(), index);
+
+	/* fid/vid correctness check for k8 */
+	/* fid are the lower 8 bits of the index we stored into
+	 * the cpufreq frequency table in find_psb_table, vid
+	 * are the upper 8 bits.
+	 */
+	fid = data->powernow_table[index].index & 0xFF;
+	vid = (data->powernow_table[index].index & 0xFF00) >> 8;
+
+	pr_debug("table matched fid 0x%x, giving vid 0x%x\n", fid, vid);
+
+	if (query_current_values_with_pending_wait(data))
+		return 1;
+
+	if ((data->currvid == vid) && (data->currfid == fid)) {
+		pr_debug("target matches current values (fid 0x%x, vid 0x%x)\n",
+			fid, vid);
+		return 0;
+	}
+
+	pr_debug("cpu %d, changing to fid 0x%x, vid 0x%x\n",
+		smp_processor_id(), fid, vid);
+	freqs.old = find_khz_freq_from_fid(data->currfid);
+	freqs.new = find_khz_freq_from_fid(fid);
+
+	for_each_cpu(i, data->available_cores) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	}
+
+	res = transition_fid_vid(data, fid, vid);
+	if (res)
+		return res;
+
+	freqs.new = find_khz_freq_from_fid(data->currfid);
+
+	for_each_cpu(i, data->available_cores) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	}
+	return res;
+}
+
+/* Take a frequency, and issue the hardware pstate transition command */
+static int transition_frequency_pstate(struct powernow_k8_data *data,
+		unsigned int index)
+{
+	u32 pstate = 0;
+	int res, i;
+	struct cpufreq_freqs freqs;
+
+	pr_debug("cpu %d transition to index %u\n", smp_processor_id(), index);
+
+	/* get MSR index for hardware pstate transition */
+	pstate = index & HW_PSTATE_MASK;
+	if (pstate > data->max_hw_pstate)
+		return -EINVAL;
+
+	freqs.old = find_khz_freq_from_pstate(data->powernow_table,
+			data->currpstate);
+	freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate);
+
+	for_each_cpu(i, data->available_cores) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	}
+
+	res = transition_pstate(data, pstate);
+	freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate);
+
+	for_each_cpu(i, data->available_cores) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	}
+	return res;
+}
+
+/* Driver entry point to switch to the target frequency */
+static int powernowk8_target(struct cpufreq_policy *pol,
+		unsigned targfreq, unsigned relation)
+{
+	cpumask_var_t oldmask;
+	struct powernow_k8_data *data = per_cpu(powernow_data, pol->cpu);
+	u32 checkfid;
+	u32 checkvid;
+	unsigned int newstate;
+	int ret = -EIO;
+
+	if (!data)
+		return -EINVAL;
+
+	checkfid = data->currfid;
+	checkvid = data->currvid;
+
+	/* only run on specific CPU from here on. */
+	/* This is poor form: use a workqueue or smp_call_function_single */
+	if (!alloc_cpumask_var(&oldmask, GFP_KERNEL))
+		return -ENOMEM;
+
+	cpumask_copy(oldmask, tsk_cpus_allowed(current));
+	set_cpus_allowed_ptr(current, cpumask_of(pol->cpu));
+
+	if (smp_processor_id() != pol->cpu) {
+		printk(KERN_ERR PFX "limiting to cpu %u failed\n", pol->cpu);
+		goto err_out;
+	}
+
+	if (pending_bit_stuck()) {
+		printk(KERN_ERR PFX "failing targ, change pending bit set\n");
+		goto err_out;
+	}
+
+	pr_debug("targ: cpu %d, %d kHz, min %d, max %d, relation %d\n",
+		pol->cpu, targfreq, pol->min, pol->max, relation);
+
+	if (query_current_values_with_pending_wait(data))
+		goto err_out;
+
+	if (cpu_family != CPU_HW_PSTATE) {
+		pr_debug("targ: curr fid 0x%x, vid 0x%x\n",
+		data->currfid, data->currvid);
+
+		if ((checkvid != data->currvid) ||
+		    (checkfid != data->currfid)) {
+			printk(KERN_INFO PFX
+				"error - out of sync, fix 0x%x 0x%x, "
+				"vid 0x%x 0x%x\n",
+				checkfid, data->currfid,
+				checkvid, data->currvid);
+		}
+	}
+
+	if (cpufreq_frequency_table_target(pol, data->powernow_table,
+				targfreq, relation, &newstate))
+		goto err_out;
+
+	mutex_lock(&fidvid_mutex);
+
+	powernow_k8_acpi_pst_values(data, newstate);
+
+	if (cpu_family == CPU_HW_PSTATE)
+		ret = transition_frequency_pstate(data,
+			data->powernow_table[newstate].index);
+	else
+		ret = transition_frequency_fidvid(data, newstate);
+	if (ret) {
+		printk(KERN_ERR PFX "transition frequency failed\n");
+		ret = 1;
+		mutex_unlock(&fidvid_mutex);
+		goto err_out;
+	}
+	mutex_unlock(&fidvid_mutex);
+
+	if (cpu_family == CPU_HW_PSTATE)
+		pol->cur = find_khz_freq_from_pstate(data->powernow_table,
+				data->powernow_table[newstate].index);
+	else
+		pol->cur = find_khz_freq_from_fid(data->currfid);
+	ret = 0;
+
+err_out:
+	set_cpus_allowed_ptr(current, oldmask);
+	free_cpumask_var(oldmask);
+	return ret;
+}
+
+/* Driver entry point to verify the policy and range of frequencies */
+static int powernowk8_verify(struct cpufreq_policy *pol)
+{
+	struct powernow_k8_data *data = per_cpu(powernow_data, pol->cpu);
+
+	if (!data)
+		return -EINVAL;
+
+	return cpufreq_frequency_table_verify(pol, data->powernow_table);
+}
+
+struct init_on_cpu {
+	struct powernow_k8_data *data;
+	int rc;
+};
+
+static void __cpuinit powernowk8_cpu_init_on_cpu(void *_init_on_cpu)
+{
+	struct init_on_cpu *init_on_cpu = _init_on_cpu;
+
+	if (pending_bit_stuck()) {
+		printk(KERN_ERR PFX "failing init, change pending bit set\n");
+		init_on_cpu->rc = -ENODEV;
+		return;
+	}
+
+	if (query_current_values_with_pending_wait(init_on_cpu->data)) {
+		init_on_cpu->rc = -ENODEV;
+		return;
+	}
+
+	if (cpu_family == CPU_OPTERON)
+		fidvid_msr_init();
+
+	init_on_cpu->rc = 0;
+}
+
+/* per CPU init entry point to the driver */
+static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
+{
+	static const char ACPI_PSS_BIOS_BUG_MSG[] =
+		KERN_ERR FW_BUG PFX "No compatible ACPI _PSS objects found.\n"
+		FW_BUG PFX "Try again with latest BIOS.\n";
+	struct powernow_k8_data *data;
+	struct init_on_cpu init_on_cpu;
+	int rc;
+	struct cpuinfo_x86 *c = &cpu_data(pol->cpu);
+
+	if (!cpu_online(pol->cpu))
+		return -ENODEV;
+
+	smp_call_function_single(pol->cpu, check_supported_cpu, &rc, 1);
+	if (rc)
+		return -ENODEV;
+
+	data = kzalloc(sizeof(struct powernow_k8_data), GFP_KERNEL);
+	if (!data) {
+		printk(KERN_ERR PFX "unable to alloc powernow_k8_data");
+		return -ENOMEM;
+	}
+
+	data->cpu = pol->cpu;
+	data->currpstate = HW_PSTATE_INVALID;
+
+	if (powernow_k8_cpu_init_acpi(data)) {
+		/*
+		 * Use the PSB BIOS structure. This is only available on
+		 * an UP version, and is deprecated by AMD.
+		 */
+		if (num_online_cpus() != 1) {
+			printk_once(ACPI_PSS_BIOS_BUG_MSG);
+			goto err_out;
+		}
+		if (pol->cpu != 0) {
+			printk(KERN_ERR FW_BUG PFX "No ACPI _PSS objects for "
+			       "CPU other than CPU0. Complain to your BIOS "
+			       "vendor.\n");
+			goto err_out;
+		}
+		rc = find_psb_table(data);
+		if (rc)
+			goto err_out;
+
+		/* Take a crude guess here.
+		 * That guess was in microseconds, so multiply with 1000 */
+		pol->cpuinfo.transition_latency = (
+			 ((data->rvo + 8) * data->vstable * VST_UNITS_20US) +
+			 ((1 << data->irt) * 30)) * 1000;
+	} else /* ACPI _PSS objects available */
+		pol->cpuinfo.transition_latency = get_transition_latency(data);
+
+	/* only run on specific CPU from here on */
+	init_on_cpu.data = data;
+	smp_call_function_single(data->cpu, powernowk8_cpu_init_on_cpu,
+				 &init_on_cpu, 1);
+	rc = init_on_cpu.rc;
+	if (rc != 0)
+		goto err_out_exit_acpi;
+
+	if (cpu_family == CPU_HW_PSTATE)
+		cpumask_copy(pol->cpus, cpumask_of(pol->cpu));
+	else
+		cpumask_copy(pol->cpus, cpu_core_mask(pol->cpu));
+	data->available_cores = pol->cpus;
+
+	if (cpu_family == CPU_HW_PSTATE)
+		pol->cur = find_khz_freq_from_pstate(data->powernow_table,
+				data->currpstate);
+	else
+		pol->cur = find_khz_freq_from_fid(data->currfid);
+	pr_debug("policy current frequency %d kHz\n", pol->cur);
+
+	/* min/max the cpu is capable of */
+	if (cpufreq_frequency_table_cpuinfo(pol, data->powernow_table)) {
+		printk(KERN_ERR FW_BUG PFX "invalid powernow_table\n");
+		powernow_k8_cpu_exit_acpi(data);
+		kfree(data->powernow_table);
+		kfree(data);
+		return -EINVAL;
+	}
+
+	/* Check for APERF/MPERF support in hardware */
+	if (cpu_has(c, X86_FEATURE_APERFMPERF))
+		cpufreq_amd64_driver.getavg = cpufreq_get_measured_perf;
+
+	cpufreq_frequency_table_get_attr(data->powernow_table, pol->cpu);
+
+	if (cpu_family == CPU_HW_PSTATE)
+		pr_debug("cpu_init done, current pstate 0x%x\n",
+				data->currpstate);
+	else
+		pr_debug("cpu_init done, current fid 0x%x, vid 0x%x\n",
+			data->currfid, data->currvid);
+
+	per_cpu(powernow_data, pol->cpu) = data;
+
+	return 0;
+
+err_out_exit_acpi:
+	powernow_k8_cpu_exit_acpi(data);
+
+err_out:
+	kfree(data);
+	return -ENODEV;
+}
+
+static int __devexit powernowk8_cpu_exit(struct cpufreq_policy *pol)
+{
+	struct powernow_k8_data *data = per_cpu(powernow_data, pol->cpu);
+
+	if (!data)
+		return -EINVAL;
+
+	powernow_k8_cpu_exit_acpi(data);
+
+	cpufreq_frequency_table_put_attr(pol->cpu);
+
+	kfree(data->powernow_table);
+	kfree(data);
+	per_cpu(powernow_data, pol->cpu) = NULL;
+
+	return 0;
+}
+
+static void query_values_on_cpu(void *_err)
+{
+	int *err = _err;
+	struct powernow_k8_data *data = __this_cpu_read(powernow_data);
+
+	*err = query_current_values_with_pending_wait(data);
+}
+
+static unsigned int powernowk8_get(unsigned int cpu)
+{
+	struct powernow_k8_data *data = per_cpu(powernow_data, cpu);
+	unsigned int khz = 0;
+	int err;
+
+	if (!data)
+		return 0;
+
+	smp_call_function_single(cpu, query_values_on_cpu, &err, true);
+	if (err)
+		goto out;
+
+	if (cpu_family == CPU_HW_PSTATE)
+		khz = find_khz_freq_from_pstate(data->powernow_table,
+						data->currpstate);
+	else
+		khz = find_khz_freq_from_fid(data->currfid);
+
+
+out:
+	return khz;
+}
+
+static void _cpb_toggle_msrs(bool t)
+{
+	int cpu;
+
+	get_online_cpus();
+
+	rdmsr_on_cpus(cpu_online_mask, MSR_K7_HWCR, msrs);
+
+	for_each_cpu(cpu, cpu_online_mask) {
+		struct msr *reg = per_cpu_ptr(msrs, cpu);
+		if (t)
+			reg->l &= ~BIT(25);
+		else
+			reg->l |= BIT(25);
+	}
+	wrmsr_on_cpus(cpu_online_mask, MSR_K7_HWCR, msrs);
+
+	put_online_cpus();
+}
+
+/*
+ * Switch on/off core performance boosting.
+ *
+ * 0=disable
+ * 1=enable.
+ */
+static void cpb_toggle(bool t)
+{
+	if (!cpb_capable)
+		return;
+
+	if (t && !cpb_enabled) {
+		cpb_enabled = true;
+		_cpb_toggle_msrs(t);
+		printk(KERN_INFO PFX "Core Boosting enabled.\n");
+	} else if (!t && cpb_enabled) {
+		cpb_enabled = false;
+		_cpb_toggle_msrs(t);
+		printk(KERN_INFO PFX "Core Boosting disabled.\n");
+	}
+}
+
+static ssize_t store_cpb(struct cpufreq_policy *policy, const char *buf,
+				 size_t count)
+{
+	int ret = -EINVAL;
+	unsigned long val = 0;
+
+	ret = strict_strtoul(buf, 10, &val);
+	if (!ret && (val == 0 || val == 1) && cpb_capable)
+		cpb_toggle(val);
+	else
+		return -EINVAL;
+
+	return count;
+}
+
+static ssize_t show_cpb(struct cpufreq_policy *policy, char *buf)
+{
+	return sprintf(buf, "%u\n", cpb_enabled);
+}
+
+#define define_one_rw(_name) \
+static struct freq_attr _name = \
+__ATTR(_name, 0644, show_##_name, store_##_name)
+
+define_one_rw(cpb);
+
+static struct freq_attr *powernow_k8_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	&cpb,
+	NULL,
+};
+
+static struct cpufreq_driver cpufreq_amd64_driver = {
+	.verify		= powernowk8_verify,
+	.target		= powernowk8_target,
+	.bios_limit	= acpi_processor_get_bios_limit,
+	.init		= powernowk8_cpu_init,
+	.exit		= __devexit_p(powernowk8_cpu_exit),
+	.get		= powernowk8_get,
+	.name		= "powernow-k8",
+	.owner		= THIS_MODULE,
+	.attr		= powernow_k8_attr,
+};
+
+/*
+ * Clear the boost-disable flag on the CPU_DOWN path so that this cpu
+ * cannot block the remaining ones from boosting. On the CPU_UP path we
+ * simply keep the boost-disable flag in sync with the current global
+ * state.
+ */
+static int cpb_notify(struct notifier_block *nb, unsigned long action,
+		      void *hcpu)
+{
+	unsigned cpu = (long)hcpu;
+	u32 lo, hi;
+
+	switch (action) {
+	case CPU_UP_PREPARE:
+	case CPU_UP_PREPARE_FROZEN:
+
+		if (!cpb_enabled) {
+			rdmsr_on_cpu(cpu, MSR_K7_HWCR, &lo, &hi);
+			lo |= BIT(25);
+			wrmsr_on_cpu(cpu, MSR_K7_HWCR, lo, hi);
+		}
+		break;
+
+	case CPU_DOWN_PREPARE:
+	case CPU_DOWN_PREPARE_FROZEN:
+		rdmsr_on_cpu(cpu, MSR_K7_HWCR, &lo, &hi);
+		lo &= ~BIT(25);
+		wrmsr_on_cpu(cpu, MSR_K7_HWCR, lo, hi);
+		break;
+
+	default:
+		break;
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block cpb_nb = {
+	.notifier_call		= cpb_notify,
+};
+
+/* driver entry point for init */
+static int __cpuinit powernowk8_init(void)
+{
+	unsigned int i, supported_cpus = 0, cpu;
+	int rv;
+
+	for_each_online_cpu(i) {
+		int rc;
+		smp_call_function_single(i, check_supported_cpu, &rc, 1);
+		if (rc == 0)
+			supported_cpus++;
+	}
+
+	if (supported_cpus != num_online_cpus())
+		return -ENODEV;
+
+	printk(KERN_INFO PFX "Found %d %s (%d cpu cores) (" VERSION ")\n",
+		num_online_nodes(), boot_cpu_data.x86_model_id, supported_cpus);
+
+	if (boot_cpu_has(X86_FEATURE_CPB)) {
+
+		cpb_capable = true;
+
+		msrs = msrs_alloc();
+		if (!msrs) {
+			printk(KERN_ERR "%s: Error allocating msrs!\n", __func__);
+			return -ENOMEM;
+		}
+
+		register_cpu_notifier(&cpb_nb);
+
+		rdmsr_on_cpus(cpu_online_mask, MSR_K7_HWCR, msrs);
+
+		for_each_cpu(cpu, cpu_online_mask) {
+			struct msr *reg = per_cpu_ptr(msrs, cpu);
+			cpb_enabled |= !(!!(reg->l & BIT(25)));
+		}
+
+		printk(KERN_INFO PFX "Core Performance Boosting: %s.\n",
+			(cpb_enabled ? "on" : "off"));
+	}
+
+	rv = cpufreq_register_driver(&cpufreq_amd64_driver);
+	if (rv < 0 && boot_cpu_has(X86_FEATURE_CPB)) {
+		unregister_cpu_notifier(&cpb_nb);
+		msrs_free(msrs);
+		msrs = NULL;
+	}
+	return rv;
+}
+
+/* driver entry point for term */
+static void __exit powernowk8_exit(void)
+{
+	pr_debug("exit\n");
+
+	if (boot_cpu_has(X86_FEATURE_CPB)) {
+		msrs_free(msrs);
+		msrs = NULL;
+
+		unregister_cpu_notifier(&cpb_nb);
+	}
+
+	cpufreq_unregister_driver(&cpufreq_amd64_driver);
+}
+
+MODULE_AUTHOR("Paul Devriendt <paul.devriendt@amd.com> and "
+		"Mark Langsdorf <mark.langsdorf@amd.com>");
+MODULE_DESCRIPTION("AMD Athlon 64 and Opteron processor frequency driver.");
+MODULE_LICENSE("GPL");
+
+late_initcall(powernowk8_init);
+module_exit(powernowk8_exit);
diff --git a/drivers/cpufreq/powernow-k8.h b/drivers/cpufreq/powernow-k8.h
new file mode 100644
index 00000000..3744d26c
--- /dev/null
+++ b/drivers/cpufreq/powernow-k8.h
@@ -0,0 +1,222 @@
+/*
+ *  (c) 2003-2006 Advanced Micro Devices, Inc.
+ *  Your use of this code is subject to the terms and conditions of the
+ *  GNU general public license version 2. See "COPYING" or
+ *  http://www.gnu.org/licenses/gpl.html
+ */
+
+enum pstate {
+	HW_PSTATE_INVALID = 0xff,
+	HW_PSTATE_0 = 0,
+	HW_PSTATE_1 = 1,
+	HW_PSTATE_2 = 2,
+	HW_PSTATE_3 = 3,
+	HW_PSTATE_4 = 4,
+	HW_PSTATE_5 = 5,
+	HW_PSTATE_6 = 6,
+	HW_PSTATE_7 = 7,
+};
+
+struct powernow_k8_data {
+	unsigned int cpu;
+
+	u32 numps;  /* number of p-states */
+	u32 batps;  /* number of p-states supported on battery */
+	u32 max_hw_pstate; /* maximum legal hardware pstate */
+
+	/* these values are constant when the PSB is used to determine
+	 * vid/fid pairings, but are modified during the ->target() call
+	 * when ACPI is used */
+	u32 rvo;     /* ramp voltage offset */
+	u32 irt;     /* isochronous relief time */
+	u32 vidmvs;  /* usable value calculated from mvs */
+	u32 vstable; /* voltage stabilization time, units 20 us */
+	u32 plllock; /* pll lock time, units 1 us */
+        u32 exttype; /* extended interface = 1 */
+
+	/* keep track of the current fid / vid or pstate */
+	u32 currvid;
+	u32 currfid;
+	enum pstate currpstate;
+
+	/* the powernow_table includes all frequency and vid/fid pairings:
+	 * fid are the lower 8 bits of the index, vid are the upper 8 bits.
+	 * frequency is in kHz */
+	struct cpufreq_frequency_table  *powernow_table;
+
+	/* the acpi table needs to be kept. it's only available if ACPI was
+	 * used to determine valid frequency/vid/fid states */
+	struct acpi_processor_performance acpi_data;
+
+	/* we need to keep track of associated cores, but let cpufreq
+	 * handle hotplug events - so just point at cpufreq pol->cpus
+	 * structure */
+	struct cpumask *available_cores;
+};
+
+/* processor's cpuid instruction support */
+#define CPUID_PROCESSOR_SIGNATURE	1	/* function 1 */
+#define CPUID_XFAM			0x0ff00000	/* extended family */
+#define CPUID_XFAM_K8			0
+#define CPUID_XMOD			0x000f0000	/* extended model */
+#define CPUID_XMOD_REV_MASK		0x000c0000
+#define CPUID_XFAM_10H			0x00100000	/* family 0x10 */
+#define CPUID_USE_XFAM_XMOD		0x00000f00
+#define CPUID_GET_MAX_CAPABILITIES	0x80000000
+#define CPUID_FREQ_VOLT_CAPABILITIES	0x80000007
+#define P_STATE_TRANSITION_CAPABLE	6
+
+/* Model Specific Registers for p-state transitions. MSRs are 64-bit. For     */
+/* writes (wrmsr - opcode 0f 30), the register number is placed in ecx, and   */
+/* the value to write is placed in edx:eax. For reads (rdmsr - opcode 0f 32), */
+/* the register number is placed in ecx, and the data is returned in edx:eax. */
+
+#define MSR_FIDVID_CTL      0xc0010041
+#define MSR_FIDVID_STATUS   0xc0010042
+
+/* Field definitions within the FID VID Low Control MSR : */
+#define MSR_C_LO_INIT_FID_VID     0x00010000
+#define MSR_C_LO_NEW_VID          0x00003f00
+#define MSR_C_LO_NEW_FID          0x0000003f
+#define MSR_C_LO_VID_SHIFT        8
+
+/* Field definitions within the FID VID High Control MSR : */
+#define MSR_C_HI_STP_GNT_TO	  0x000fffff
+
+/* Field definitions within the FID VID Low Status MSR : */
+#define MSR_S_LO_CHANGE_PENDING   0x80000000   /* cleared when completed */
+#define MSR_S_LO_MAX_RAMP_VID     0x3f000000
+#define MSR_S_LO_MAX_FID          0x003f0000
+#define MSR_S_LO_START_FID        0x00003f00
+#define MSR_S_LO_CURRENT_FID      0x0000003f
+
+/* Field definitions within the FID VID High Status MSR : */
+#define MSR_S_HI_MIN_WORKING_VID  0x3f000000
+#define MSR_S_HI_MAX_WORKING_VID  0x003f0000
+#define MSR_S_HI_START_VID        0x00003f00
+#define MSR_S_HI_CURRENT_VID      0x0000003f
+#define MSR_C_HI_STP_GNT_BENIGN	  0x00000001
+
+
+/* Hardware Pstate _PSS and MSR definitions */
+#define USE_HW_PSTATE		0x00000080
+#define HW_PSTATE_MASK 		0x00000007
+#define HW_PSTATE_VALID_MASK 	0x80000000
+#define HW_PSTATE_MAX_MASK	0x000000f0
+#define HW_PSTATE_MAX_SHIFT	4
+#define MSR_PSTATE_DEF_BASE 	0xc0010064 /* base of Pstate MSRs */
+#define MSR_PSTATE_STATUS 	0xc0010063 /* Pstate Status MSR */
+#define MSR_PSTATE_CTRL 	0xc0010062 /* Pstate control MSR */
+#define MSR_PSTATE_CUR_LIMIT	0xc0010061 /* pstate current limit MSR */
+
+/* define the two driver architectures */
+#define CPU_OPTERON 0
+#define CPU_HW_PSTATE 1
+
+
+/*
+ * There are restrictions frequencies have to follow:
+ * - only 1 entry in the low fid table ( <=1.4GHz )
+ * - lowest entry in the high fid table must be >= 2 * the entry in the
+ *   low fid table
+ * - lowest entry in the high fid table must be a <= 200MHz + 2 * the entry
+ *   in the low fid table
+ * - the parts can only step at <= 200 MHz intervals, odd fid values are
+ *   supported in revision G and later revisions.
+ * - lowest frequency must be >= interprocessor hypertransport link speed
+ *   (only applies to MP systems obviously)
+ */
+
+/* fids (frequency identifiers) are arranged in 2 tables - lo and hi */
+#define LO_FID_TABLE_TOP     7	/* fid values marking the boundary    */
+#define HI_FID_TABLE_BOTTOM  8	/* between the low and high tables    */
+
+#define LO_VCOFREQ_TABLE_TOP    1400	/* corresponding vco frequency values */
+#define HI_VCOFREQ_TABLE_BOTTOM 1600
+
+#define MIN_FREQ_RESOLUTION  200 /* fids jump by 2 matching freq jumps by 200 */
+
+#define MAX_FID 0x2a	/* Spec only gives FID values as far as 5 GHz */
+#define LEAST_VID 0x3e	/* Lowest (numerically highest) useful vid value */
+
+#define MIN_FREQ 800	/* Min and max freqs, per spec */
+#define MAX_FREQ 5000
+
+#define INVALID_FID_MASK 0xffffffc0  /* not a valid fid if these bits are set */
+#define INVALID_VID_MASK 0xffffffc0  /* not a valid vid if these bits are set */
+
+#define VID_OFF 0x3f
+
+#define STOP_GRANT_5NS 1 /* min poss memory access latency for voltage change */
+
+#define PLL_LOCK_CONVERSION (1000/5) /* ms to ns, then divide by clock period */
+
+#define MAXIMUM_VID_STEPS 1  /* Current cpus only allow a single step of 25mV */
+#define VST_UNITS_20US 20   /* Voltage Stabilization Time is in units of 20us */
+
+/*
+ * Most values of interest are encoded in a single field of the _PSS
+ * entries: the "control" value.
+ */
+
+#define IRT_SHIFT      30
+#define RVO_SHIFT      28
+#define EXT_TYPE_SHIFT 27
+#define PLL_L_SHIFT    20
+#define MVS_SHIFT      18
+#define VST_SHIFT      11
+#define VID_SHIFT       6
+#define IRT_MASK        3
+#define RVO_MASK        3
+#define EXT_TYPE_MASK   1
+#define PLL_L_MASK   0x7f
+#define MVS_MASK        3
+#define VST_MASK     0x7f
+#define VID_MASK     0x1f
+#define FID_MASK     0x1f
+#define EXT_VID_MASK 0x3f
+#define EXT_FID_MASK 0x3f
+
+
+/*
+ * Version 1.4 of the PSB table. This table is constructed by BIOS and is
+ * to tell the OS's power management driver which VIDs and FIDs are
+ * supported by this particular processor.
+ * If the data in the PSB / PST is wrong, then this driver will program the
+ * wrong values into hardware, which is very likely to lead to a crash.
+ */
+
+#define PSB_ID_STRING      "AMDK7PNOW!"
+#define PSB_ID_STRING_LEN  10
+
+#define PSB_VERSION_1_4  0x14
+
+struct psb_s {
+	u8 signature[10];
+	u8 tableversion;
+	u8 flags1;
+	u16 vstable;
+	u8 flags2;
+	u8 num_tables;
+	u32 cpuid;
+	u8 plllocktime;
+	u8 maxfid;
+	u8 maxvid;
+	u8 numps;
+};
+
+/* Pairs of fid/vid values are appended to the version 1.4 PSB table. */
+struct pst_s {
+	u8 fid;
+	u8 vid;
+};
+
+static int core_voltage_pre_transition(struct powernow_k8_data *data,
+	u32 reqvid, u32 regfid);
+static int core_voltage_post_transition(struct powernow_k8_data *data, u32 reqvid);
+static int core_frequency_transition(struct powernow_k8_data *data, u32 reqfid);
+
+static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index);
+
+static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table);
+static int fill_powernow_table_fidvid(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table);
diff --git a/drivers/cpufreq/sc520_freq.c b/drivers/cpufreq/sc520_freq.c
new file mode 100644
index 00000000..1e205e6b
--- /dev/null
+++ b/drivers/cpufreq/sc520_freq.c
@@ -0,0 +1,192 @@
+/*
+ *	sc520_freq.c: cpufreq driver for the AMD Elan sc520
+ *
+ *	Copyright (C) 2005 Sean Young <sean@mess.org>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ *
+ *	Based on elanfreq.c
+ *
+ *	2005-03-30: - initial revision
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <linux/delay.h>
+#include <linux/cpufreq.h>
+#include <linux/timex.h>
+#include <linux/io.h>
+
+#include <asm/msr.h>
+
+#define MMCR_BASE	0xfffef000	/* The default base address */
+#define OFFS_CPUCTL	0x2   /* CPU Control Register */
+
+static __u8 __iomem *cpuctl;
+
+#define PFX "sc520_freq: "
+
+static struct cpufreq_frequency_table sc520_freq_table[] = {
+	{0x01,	100000},
+	{0x02,	133000},
+	{0,	CPUFREQ_TABLE_END},
+};
+
+static unsigned int sc520_freq_get_cpu_frequency(unsigned int cpu)
+{
+	u8 clockspeed_reg = *cpuctl;
+
+	switch (clockspeed_reg & 0x03) {
+	default:
+		printk(KERN_ERR PFX "error: cpuctl register has unexpected "
+				"value %02x\n", clockspeed_reg);
+	case 0x01:
+		return 100000;
+	case 0x02:
+		return 133000;
+	}
+}
+
+static void sc520_freq_set_cpu_state(unsigned int state)
+{
+
+	struct cpufreq_freqs	freqs;
+	u8 clockspeed_reg;
+
+	freqs.old = sc520_freq_get_cpu_frequency(0);
+	freqs.new = sc520_freq_table[state].frequency;
+	freqs.cpu = 0; /* AMD Elan is UP */
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+	pr_debug("attempting to set frequency to %i kHz\n",
+			sc520_freq_table[state].frequency);
+
+	local_irq_disable();
+
+	clockspeed_reg = *cpuctl & ~0x03;
+	*cpuctl = clockspeed_reg | sc520_freq_table[state].index;
+
+	local_irq_enable();
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+};
+
+static int sc520_freq_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, &sc520_freq_table[0]);
+}
+
+static int sc520_freq_target(struct cpufreq_policy *policy,
+			    unsigned int target_freq,
+			    unsigned int relation)
+{
+	unsigned int newstate = 0;
+
+	if (cpufreq_frequency_table_target(policy, sc520_freq_table,
+				target_freq, relation, &newstate))
+		return -EINVAL;
+
+	sc520_freq_set_cpu_state(newstate);
+
+	return 0;
+}
+
+
+/*
+ *	Module init and exit code
+ */
+
+static int sc520_freq_cpu_init(struct cpufreq_policy *policy)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	int result;
+
+	/* capability check */
+	if (c->x86_vendor != X86_VENDOR_AMD ||
+	    c->x86 != 4 || c->x86_model != 9)
+		return -ENODEV;
+
+	/* cpuinfo and default policy values */
+	policy->cpuinfo.transition_latency = 1000000; /* 1ms */
+	policy->cur = sc520_freq_get_cpu_frequency(0);
+
+	result = cpufreq_frequency_table_cpuinfo(policy, sc520_freq_table);
+	if (result)
+		return result;
+
+	cpufreq_frequency_table_get_attr(sc520_freq_table, policy->cpu);
+
+	return 0;
+}
+
+
+static int sc520_freq_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+
+static struct freq_attr *sc520_freq_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+
+static struct cpufreq_driver sc520_freq_driver = {
+	.get	= sc520_freq_get_cpu_frequency,
+	.verify	= sc520_freq_verify,
+	.target	= sc520_freq_target,
+	.init	= sc520_freq_cpu_init,
+	.exit	= sc520_freq_cpu_exit,
+	.name	= "sc520_freq",
+	.owner	= THIS_MODULE,
+	.attr	= sc520_freq_attr,
+};
+
+
+static int __init sc520_freq_init(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	int err;
+
+	/* Test if we have the right hardware */
+	if (c->x86_vendor != X86_VENDOR_AMD ||
+	    c->x86 != 4 || c->x86_model != 9) {
+		pr_debug("no Elan SC520 processor found!\n");
+		return -ENODEV;
+	}
+	cpuctl = ioremap((unsigned long)(MMCR_BASE + OFFS_CPUCTL), 1);
+	if (!cpuctl) {
+		printk(KERN_ERR "sc520_freq: error: failed to remap memory\n");
+		return -ENOMEM;
+	}
+
+	err = cpufreq_register_driver(&sc520_freq_driver);
+	if (err)
+		iounmap(cpuctl);
+
+	return err;
+}
+
+
+static void __exit sc520_freq_exit(void)
+{
+	cpufreq_unregister_driver(&sc520_freq_driver);
+	iounmap(cpuctl);
+}
+
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Sean Young <sean@mess.org>");
+MODULE_DESCRIPTION("cpufreq driver for AMD's Elan sc520 CPU");
+
+module_init(sc520_freq_init);
+module_exit(sc520_freq_exit);
+
diff --git a/drivers/cpufreq/speedstep-centrino.c b/drivers/cpufreq/speedstep-centrino.c
new file mode 100644
index 00000000..6ea3455d
--- /dev/null
+++ b/drivers/cpufreq/speedstep-centrino.c
@@ -0,0 +1,633 @@
+/*
+ * cpufreq driver for Enhanced SpeedStep, as found in Intel's Pentium
+ * M (part of the Centrino chipset).
+ *
+ * Since the original Pentium M, most new Intel CPUs support Enhanced
+ * SpeedStep.
+ *
+ * Despite the "SpeedStep" in the name, this is almost entirely unlike
+ * traditional SpeedStep.
+ *
+ * Modelled on speedstep.c
+ *
+ * Copyright (C) 2003 Jeremy Fitzhardinge <jeremy@goop.org>
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/sched.h>	/* current */
+#include <linux/delay.h>
+#include <linux/compiler.h>
+#include <linux/gfp.h>
+
+#include <asm/msr.h>
+#include <asm/processor.h>
+#include <asm/cpufeature.h>
+
+#define PFX		"speedstep-centrino: "
+#define MAINTAINER	"cpufreq@vger.kernel.org"
+
+#define INTEL_MSR_RANGE	(0xffff)
+
+struct cpu_id
+{
+	__u8	x86;            /* CPU family */
+	__u8	x86_model;	/* model */
+	__u8	x86_mask;	/* stepping */
+};
+
+enum {
+	CPU_BANIAS,
+	CPU_DOTHAN_A1,
+	CPU_DOTHAN_A2,
+	CPU_DOTHAN_B0,
+	CPU_MP4HT_D0,
+	CPU_MP4HT_E0,
+};
+
+static const struct cpu_id cpu_ids[] = {
+	[CPU_BANIAS]	= { 6,  9, 5 },
+	[CPU_DOTHAN_A1]	= { 6, 13, 1 },
+	[CPU_DOTHAN_A2]	= { 6, 13, 2 },
+	[CPU_DOTHAN_B0]	= { 6, 13, 6 },
+	[CPU_MP4HT_D0]	= {15,  3, 4 },
+	[CPU_MP4HT_E0]	= {15,  4, 1 },
+};
+#define N_IDS	ARRAY_SIZE(cpu_ids)
+
+struct cpu_model
+{
+	const struct cpu_id *cpu_id;
+	const char	*model_name;
+	unsigned	max_freq; /* max clock in kHz */
+
+	struct cpufreq_frequency_table *op_points; /* clock/voltage pairs */
+};
+static int centrino_verify_cpu_id(const struct cpuinfo_x86 *c,
+				  const struct cpu_id *x);
+
+/* Operating points for current CPU */
+static DEFINE_PER_CPU(struct cpu_model *, centrino_model);
+static DEFINE_PER_CPU(const struct cpu_id *, centrino_cpu);
+
+static struct cpufreq_driver centrino_driver;
+
+#ifdef CONFIG_X86_SPEEDSTEP_CENTRINO_TABLE
+
+/* Computes the correct form for IA32_PERF_CTL MSR for a particular
+   frequency/voltage operating point; frequency in MHz, volts in mV.
+   This is stored as "index" in the structure. */
+#define OP(mhz, mv)							\
+	{								\
+		.frequency = (mhz) * 1000,				\
+		.index = (((mhz)/100) << 8) | ((mv - 700) / 16)		\
+	}
+
+/*
+ * These voltage tables were derived from the Intel Pentium M
+ * datasheet, document 25261202.pdf, Table 5.  I have verified they
+ * are consistent with my IBM ThinkPad X31, which has a 1.3GHz Pentium
+ * M.
+ */
+
+/* Ultra Low Voltage Intel Pentium M processor 900MHz (Banias) */
+static struct cpufreq_frequency_table banias_900[] =
+{
+	OP(600,  844),
+	OP(800,  988),
+	OP(900, 1004),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Ultra Low Voltage Intel Pentium M processor 1000MHz (Banias) */
+static struct cpufreq_frequency_table banias_1000[] =
+{
+	OP(600,   844),
+	OP(800,   972),
+	OP(900,   988),
+	OP(1000, 1004),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Low Voltage Intel Pentium M processor 1.10GHz (Banias) */
+static struct cpufreq_frequency_table banias_1100[] =
+{
+	OP( 600,  956),
+	OP( 800, 1020),
+	OP( 900, 1100),
+	OP(1000, 1164),
+	OP(1100, 1180),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+
+/* Low Voltage Intel Pentium M processor 1.20GHz (Banias) */
+static struct cpufreq_frequency_table banias_1200[] =
+{
+	OP( 600,  956),
+	OP( 800, 1004),
+	OP( 900, 1020),
+	OP(1000, 1100),
+	OP(1100, 1164),
+	OP(1200, 1180),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Intel Pentium M processor 1.30GHz (Banias) */
+static struct cpufreq_frequency_table banias_1300[] =
+{
+	OP( 600,  956),
+	OP( 800, 1260),
+	OP(1000, 1292),
+	OP(1200, 1356),
+	OP(1300, 1388),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Intel Pentium M processor 1.40GHz (Banias) */
+static struct cpufreq_frequency_table banias_1400[] =
+{
+	OP( 600,  956),
+	OP( 800, 1180),
+	OP(1000, 1308),
+	OP(1200, 1436),
+	OP(1400, 1484),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Intel Pentium M processor 1.50GHz (Banias) */
+static struct cpufreq_frequency_table banias_1500[] =
+{
+	OP( 600,  956),
+	OP( 800, 1116),
+	OP(1000, 1228),
+	OP(1200, 1356),
+	OP(1400, 1452),
+	OP(1500, 1484),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Intel Pentium M processor 1.60GHz (Banias) */
+static struct cpufreq_frequency_table banias_1600[] =
+{
+	OP( 600,  956),
+	OP( 800, 1036),
+	OP(1000, 1164),
+	OP(1200, 1276),
+	OP(1400, 1420),
+	OP(1600, 1484),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+
+/* Intel Pentium M processor 1.70GHz (Banias) */
+static struct cpufreq_frequency_table banias_1700[] =
+{
+	OP( 600,  956),
+	OP( 800, 1004),
+	OP(1000, 1116),
+	OP(1200, 1228),
+	OP(1400, 1308),
+	OP(1700, 1484),
+	{ .frequency = CPUFREQ_TABLE_END }
+};
+#undef OP
+
+#define _BANIAS(cpuid, max, name)	\
+{	.cpu_id		= cpuid,	\
+	.model_name	= "Intel(R) Pentium(R) M processor " name "MHz", \
+	.max_freq	= (max)*1000,	\
+	.op_points	= banias_##max,	\
+}
+#define BANIAS(max)	_BANIAS(&cpu_ids[CPU_BANIAS], max, #max)
+
+/* CPU models, their operating frequency range, and freq/voltage
+   operating points */
+static struct cpu_model models[] =
+{
+	_BANIAS(&cpu_ids[CPU_BANIAS], 900, " 900"),
+	BANIAS(1000),
+	BANIAS(1100),
+	BANIAS(1200),
+	BANIAS(1300),
+	BANIAS(1400),
+	BANIAS(1500),
+	BANIAS(1600),
+	BANIAS(1700),
+
+	/* NULL model_name is a wildcard */
+	{ &cpu_ids[CPU_DOTHAN_A1], NULL, 0, NULL },
+	{ &cpu_ids[CPU_DOTHAN_A2], NULL, 0, NULL },
+	{ &cpu_ids[CPU_DOTHAN_B0], NULL, 0, NULL },
+	{ &cpu_ids[CPU_MP4HT_D0], NULL, 0, NULL },
+	{ &cpu_ids[CPU_MP4HT_E0], NULL, 0, NULL },
+
+	{ NULL, }
+};
+#undef _BANIAS
+#undef BANIAS
+
+static int centrino_cpu_init_table(struct cpufreq_policy *policy)
+{
+	struct cpuinfo_x86 *cpu = &cpu_data(policy->cpu);
+	struct cpu_model *model;
+
+	for(model = models; model->cpu_id != NULL; model++)
+		if (centrino_verify_cpu_id(cpu, model->cpu_id) &&
+		    (model->model_name == NULL ||
+		     strcmp(cpu->x86_model_id, model->model_name) == 0))
+			break;
+
+	if (model->cpu_id == NULL) {
+		/* No match at all */
+		pr_debug("no support for CPU model \"%s\": "
+		       "send /proc/cpuinfo to " MAINTAINER "\n",
+		       cpu->x86_model_id);
+		return -ENOENT;
+	}
+
+	if (model->op_points == NULL) {
+		/* Matched a non-match */
+		pr_debug("no table support for CPU model \"%s\"\n",
+		       cpu->x86_model_id);
+		pr_debug("try using the acpi-cpufreq driver\n");
+		return -ENOENT;
+	}
+
+	per_cpu(centrino_model, policy->cpu) = model;
+
+	pr_debug("found \"%s\": max frequency: %dkHz\n",
+	       model->model_name, model->max_freq);
+
+	return 0;
+}
+
+#else
+static inline int centrino_cpu_init_table(struct cpufreq_policy *policy)
+{
+	return -ENODEV;
+}
+#endif /* CONFIG_X86_SPEEDSTEP_CENTRINO_TABLE */
+
+static int centrino_verify_cpu_id(const struct cpuinfo_x86 *c,
+				  const struct cpu_id *x)
+{
+	if ((c->x86 == x->x86) &&
+	    (c->x86_model == x->x86_model) &&
+	    (c->x86_mask == x->x86_mask))
+		return 1;
+	return 0;
+}
+
+/* To be called only after centrino_model is initialized */
+static unsigned extract_clock(unsigned msr, unsigned int cpu, int failsafe)
+{
+	int i;
+
+	/*
+	 * Extract clock in kHz from PERF_CTL value
+	 * for centrino, as some DSDTs are buggy.
+	 * Ideally, this can be done using the acpi_data structure.
+	 */
+	if ((per_cpu(centrino_cpu, cpu) == &cpu_ids[CPU_BANIAS]) ||
+	    (per_cpu(centrino_cpu, cpu) == &cpu_ids[CPU_DOTHAN_A1]) ||
+	    (per_cpu(centrino_cpu, cpu) == &cpu_ids[CPU_DOTHAN_B0])) {
+		msr = (msr >> 8) & 0xff;
+		return msr * 100000;
+	}
+
+	if ((!per_cpu(centrino_model, cpu)) ||
+	    (!per_cpu(centrino_model, cpu)->op_points))
+		return 0;
+
+	msr &= 0xffff;
+	for (i = 0;
+		per_cpu(centrino_model, cpu)->op_points[i].frequency
+							!= CPUFREQ_TABLE_END;
+	     i++) {
+		if (msr == per_cpu(centrino_model, cpu)->op_points[i].index)
+			return per_cpu(centrino_model, cpu)->
+							op_points[i].frequency;
+	}
+	if (failsafe)
+		return per_cpu(centrino_model, cpu)->op_points[i-1].frequency;
+	else
+		return 0;
+}
+
+/* Return the current CPU frequency in kHz */
+static unsigned int get_cur_freq(unsigned int cpu)
+{
+	unsigned l, h;
+	unsigned clock_freq;
+
+	rdmsr_on_cpu(cpu, MSR_IA32_PERF_STATUS, &l, &h);
+	clock_freq = extract_clock(l, cpu, 0);
+
+	if (unlikely(clock_freq == 0)) {
+		/*
+		 * On some CPUs, we can see transient MSR values (which are
+		 * not present in _PSS), while CPU is doing some automatic
+		 * P-state transition (like TM2). Get the last freq set 
+		 * in PERF_CTL.
+		 */
+		rdmsr_on_cpu(cpu, MSR_IA32_PERF_CTL, &l, &h);
+		clock_freq = extract_clock(l, cpu, 1);
+	}
+	return clock_freq;
+}
+
+
+static int centrino_cpu_init(struct cpufreq_policy *policy)
+{
+	struct cpuinfo_x86 *cpu = &cpu_data(policy->cpu);
+	unsigned freq;
+	unsigned l, h;
+	int ret;
+	int i;
+
+	/* Only Intel makes Enhanced Speedstep-capable CPUs */
+	if (cpu->x86_vendor != X86_VENDOR_INTEL ||
+	    !cpu_has(cpu, X86_FEATURE_EST))
+		return -ENODEV;
+
+	if (cpu_has(cpu, X86_FEATURE_CONSTANT_TSC))
+		centrino_driver.flags |= CPUFREQ_CONST_LOOPS;
+
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	for (i = 0; i < N_IDS; i++)
+		if (centrino_verify_cpu_id(cpu, &cpu_ids[i]))
+			break;
+
+	if (i != N_IDS)
+		per_cpu(centrino_cpu, policy->cpu) = &cpu_ids[i];
+
+	if (!per_cpu(centrino_cpu, policy->cpu)) {
+		pr_debug("found unsupported CPU with "
+		"Enhanced SpeedStep: send /proc/cpuinfo to "
+		MAINTAINER "\n");
+		return -ENODEV;
+	}
+
+	if (centrino_cpu_init_table(policy)) {
+		return -ENODEV;
+	}
+
+	/* Check to see if Enhanced SpeedStep is enabled, and try to
+	   enable it if not. */
+	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
+
+	if (!(l & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
+		l |= MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP;
+		pr_debug("trying to enable Enhanced SpeedStep (%x)\n", l);
+		wrmsr(MSR_IA32_MISC_ENABLE, l, h);
+
+		/* check to see if it stuck */
+		rdmsr(MSR_IA32_MISC_ENABLE, l, h);
+		if (!(l & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP)) {
+			printk(KERN_INFO PFX
+				"couldn't enable Enhanced SpeedStep\n");
+			return -ENODEV;
+		}
+	}
+
+	freq = get_cur_freq(policy->cpu);
+	policy->cpuinfo.transition_latency = 10000;
+						/* 10uS transition latency */
+	policy->cur = freq;
+
+	pr_debug("centrino_cpu_init: cur=%dkHz\n", policy->cur);
+
+	ret = cpufreq_frequency_table_cpuinfo(policy,
+		per_cpu(centrino_model, policy->cpu)->op_points);
+	if (ret)
+		return (ret);
+
+	cpufreq_frequency_table_get_attr(
+		per_cpu(centrino_model, policy->cpu)->op_points, policy->cpu);
+
+	return 0;
+}
+
+static int centrino_cpu_exit(struct cpufreq_policy *policy)
+{
+	unsigned int cpu = policy->cpu;
+
+	if (!per_cpu(centrino_model, cpu))
+		return -ENODEV;
+
+	cpufreq_frequency_table_put_attr(cpu);
+
+	per_cpu(centrino_model, cpu) = NULL;
+
+	return 0;
+}
+
+/**
+ * centrino_verify - verifies a new CPUFreq policy
+ * @policy: new policy
+ *
+ * Limit must be within this model's frequency range at least one
+ * border included.
+ */
+static int centrino_verify (struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy,
+			per_cpu(centrino_model, policy->cpu)->op_points);
+}
+
+/**
+ * centrino_setpolicy - set a new CPUFreq policy
+ * @policy: new policy
+ * @target_freq: the target frequency
+ * @relation: how that frequency relates to achieved frequency
+ *	(CPUFREQ_RELATION_L or CPUFREQ_RELATION_H)
+ *
+ * Sets a new CPUFreq policy.
+ */
+static int centrino_target (struct cpufreq_policy *policy,
+			    unsigned int target_freq,
+			    unsigned int relation)
+{
+	unsigned int    newstate = 0;
+	unsigned int	msr, oldmsr = 0, h = 0, cpu = policy->cpu;
+	struct cpufreq_freqs	freqs;
+	int			retval = 0;
+	unsigned int		j, k, first_cpu, tmp;
+	cpumask_var_t covered_cpus;
+
+	if (unlikely(!zalloc_cpumask_var(&covered_cpus, GFP_KERNEL)))
+		return -ENOMEM;
+
+	if (unlikely(per_cpu(centrino_model, cpu) == NULL)) {
+		retval = -ENODEV;
+		goto out;
+	}
+
+	if (unlikely(cpufreq_frequency_table_target(policy,
+			per_cpu(centrino_model, cpu)->op_points,
+			target_freq,
+			relation,
+			&newstate))) {
+		retval = -EINVAL;
+		goto out;
+	}
+
+	first_cpu = 1;
+	for_each_cpu(j, policy->cpus) {
+		int good_cpu;
+
+		/* cpufreq holds the hotplug lock, so we are safe here */
+		if (!cpu_online(j))
+			continue;
+
+		/*
+		 * Support for SMP systems.
+		 * Make sure we are running on CPU that wants to change freq
+		 */
+		if (policy->shared_type == CPUFREQ_SHARED_TYPE_ANY)
+			good_cpu = cpumask_any_and(policy->cpus,
+						   cpu_online_mask);
+		else
+			good_cpu = j;
+
+		if (good_cpu >= nr_cpu_ids) {
+			pr_debug("couldn't limit to CPUs in this domain\n");
+			retval = -EAGAIN;
+			if (first_cpu) {
+				/* We haven't started the transition yet. */
+				goto out;
+			}
+			break;
+		}
+
+		msr = per_cpu(centrino_model, cpu)->op_points[newstate].index;
+
+		if (first_cpu) {
+			rdmsr_on_cpu(good_cpu, MSR_IA32_PERF_CTL, &oldmsr, &h);
+			if (msr == (oldmsr & 0xffff)) {
+				pr_debug("no change needed - msr was and needs "
+					"to be %x\n", oldmsr);
+				retval = 0;
+				goto out;
+			}
+
+			freqs.old = extract_clock(oldmsr, cpu, 0);
+			freqs.new = extract_clock(msr, cpu, 0);
+
+			pr_debug("target=%dkHz old=%d new=%d msr=%04x\n",
+				target_freq, freqs.old, freqs.new, msr);
+
+			for_each_cpu(k, policy->cpus) {
+				if (!cpu_online(k))
+					continue;
+				freqs.cpu = k;
+				cpufreq_notify_transition(&freqs,
+					CPUFREQ_PRECHANGE);
+			}
+
+			first_cpu = 0;
+			/* all but 16 LSB are reserved, treat them with care */
+			oldmsr &= ~0xffff;
+			msr &= 0xffff;
+			oldmsr |= msr;
+		}
+
+		wrmsr_on_cpu(good_cpu, MSR_IA32_PERF_CTL, oldmsr, h);
+		if (policy->shared_type == CPUFREQ_SHARED_TYPE_ANY)
+			break;
+
+		cpumask_set_cpu(j, covered_cpus);
+	}
+
+	for_each_cpu(k, policy->cpus) {
+		if (!cpu_online(k))
+			continue;
+		freqs.cpu = k;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	}
+
+	if (unlikely(retval)) {
+		/*
+		 * We have failed halfway through the frequency change.
+		 * We have sent callbacks to policy->cpus and
+		 * MSRs have already been written on coverd_cpus.
+		 * Best effort undo..
+		 */
+
+		for_each_cpu(j, covered_cpus)
+			wrmsr_on_cpu(j, MSR_IA32_PERF_CTL, oldmsr, h);
+
+		tmp = freqs.new;
+		freqs.new = freqs.old;
+		freqs.old = tmp;
+		for_each_cpu(j, policy->cpus) {
+			if (!cpu_online(j))
+				continue;
+			cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+			cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+		}
+	}
+	retval = 0;
+
+out:
+	free_cpumask_var(covered_cpus);
+	return retval;
+}
+
+static struct freq_attr* centrino_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver centrino_driver = {
+	.name		= "centrino", /* should be speedstep-centrino,
+					 but there's a 16 char limit */
+	.init		= centrino_cpu_init,
+	.exit		= centrino_cpu_exit,
+	.verify		= centrino_verify,
+	.target		= centrino_target,
+	.get		= get_cur_freq,
+	.attr           = centrino_attr,
+	.owner		= THIS_MODULE,
+};
+
+
+/**
+ * centrino_init - initializes the Enhanced SpeedStep CPUFreq driver
+ *
+ * Initializes the Enhanced SpeedStep support. Returns -ENODEV on
+ * unsupported devices, -ENOENT if there's no voltage table for this
+ * particular CPU model, -EINVAL on problems during initiatization,
+ * and zero on success.
+ *
+ * This is quite picky.  Not only does the CPU have to advertise the
+ * "est" flag in the cpuid capability flags, we look for a specific
+ * CPU model and stepping, and we need to have the exact model name in
+ * our voltage tables.  That is, be paranoid about not releasing
+ * someone's valuable magic smoke.
+ */
+static int __init centrino_init(void)
+{
+	struct cpuinfo_x86 *cpu = &cpu_data(0);
+
+	if (!cpu_has(cpu, X86_FEATURE_EST))
+		return -ENODEV;
+
+	return cpufreq_register_driver(&centrino_driver);
+}
+
+static void __exit centrino_exit(void)
+{
+	cpufreq_unregister_driver(&centrino_driver);
+}
+
+MODULE_AUTHOR ("Jeremy Fitzhardinge <jeremy@goop.org>");
+MODULE_DESCRIPTION ("Enhanced SpeedStep driver for Intel Pentium M processors.");
+MODULE_LICENSE ("GPL");
+
+late_initcall(centrino_init);
+module_exit(centrino_exit);
diff --git a/drivers/cpufreq/speedstep-ich.c b/drivers/cpufreq/speedstep-ich.c
new file mode 100644
index 00000000..a748ce78
--- /dev/null
+++ b/drivers/cpufreq/speedstep-ich.c
@@ -0,0 +1,448 @@
+/*
+ * (C) 2001  Dave Jones, Arjan van de ven.
+ * (C) 2002 - 2003  Dominik Brodowski <linux@brodo.de>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *  Based upon reverse engineered information, and on Intel documentation
+ *  for chipsets ICH2-M and ICH3-M.
+ *
+ *  Many thanks to Ducrot Bruno for finding and fixing the last
+ *  "missing link" for ICH2-M/ICH3-M support, and to Thomas Winkler
+ *  for extensive testing.
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+
+/*********************************************************************
+ *                        SPEEDSTEP - DEFINITIONS                    *
+ *********************************************************************/
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/pci.h>
+#include <linux/sched.h>
+
+#include "speedstep-lib.h"
+
+
+/* speedstep_chipset:
+ *   It is necessary to know which chipset is used. As accesses to
+ * this device occur at various places in this module, we need a
+ * static struct pci_dev * pointing to that device.
+ */
+static struct pci_dev *speedstep_chipset_dev;
+
+
+/* speedstep_processor
+ */
+static enum speedstep_processor speedstep_processor;
+
+static u32 pmbase;
+
+/*
+ *   There are only two frequency states for each processor. Values
+ * are in kHz for the time being.
+ */
+static struct cpufreq_frequency_table speedstep_freqs[] = {
+	{SPEEDSTEP_HIGH,	0},
+	{SPEEDSTEP_LOW,		0},
+	{0,			CPUFREQ_TABLE_END},
+};
+
+
+/**
+ * speedstep_find_register - read the PMBASE address
+ *
+ * Returns: -ENODEV if no register could be found
+ */
+static int speedstep_find_register(void)
+{
+	if (!speedstep_chipset_dev)
+		return -ENODEV;
+
+	/* get PMBASE */
+	pci_read_config_dword(speedstep_chipset_dev, 0x40, &pmbase);
+	if (!(pmbase & 0x01)) {
+		printk(KERN_ERR "speedstep-ich: could not find speedstep register\n");
+		return -ENODEV;
+	}
+
+	pmbase &= 0xFFFFFFFE;
+	if (!pmbase) {
+		printk(KERN_ERR "speedstep-ich: could not find speedstep register\n");
+		return -ENODEV;
+	}
+
+	pr_debug("pmbase is 0x%x\n", pmbase);
+	return 0;
+}
+
+/**
+ * speedstep_set_state - set the SpeedStep state
+ * @state: new processor frequency state (SPEEDSTEP_LOW or SPEEDSTEP_HIGH)
+ *
+ *   Tries to change the SpeedStep state.  Can be called from
+ *   smp_call_function_single.
+ */
+static void speedstep_set_state(unsigned int state)
+{
+	u8 pm2_blk;
+	u8 value;
+	unsigned long flags;
+
+	if (state > 0x1)
+		return;
+
+	/* Disable IRQs */
+	local_irq_save(flags);
+
+	/* read state */
+	value = inb(pmbase + 0x50);
+
+	pr_debug("read at pmbase 0x%x + 0x50 returned 0x%x\n", pmbase, value);
+
+	/* write new state */
+	value &= 0xFE;
+	value |= state;
+
+	pr_debug("writing 0x%x to pmbase 0x%x + 0x50\n", value, pmbase);
+
+	/* Disable bus master arbitration */
+	pm2_blk = inb(pmbase + 0x20);
+	pm2_blk |= 0x01;
+	outb(pm2_blk, (pmbase + 0x20));
+
+	/* Actual transition */
+	outb(value, (pmbase + 0x50));
+
+	/* Restore bus master arbitration */
+	pm2_blk &= 0xfe;
+	outb(pm2_blk, (pmbase + 0x20));
+
+	/* check if transition was successful */
+	value = inb(pmbase + 0x50);
+
+	/* Enable IRQs */
+	local_irq_restore(flags);
+
+	pr_debug("read at pmbase 0x%x + 0x50 returned 0x%x\n", pmbase, value);
+
+	if (state == (value & 0x1))
+		pr_debug("change to %u MHz succeeded\n",
+			speedstep_get_frequency(speedstep_processor) / 1000);
+	else
+		printk(KERN_ERR "cpufreq: change failed - I/O error\n");
+
+	return;
+}
+
+/* Wrapper for smp_call_function_single. */
+static void _speedstep_set_state(void *_state)
+{
+	speedstep_set_state(*(unsigned int *)_state);
+}
+
+/**
+ * speedstep_activate - activate SpeedStep control in the chipset
+ *
+ *   Tries to activate the SpeedStep status and control registers.
+ * Returns -EINVAL on an unsupported chipset, and zero on success.
+ */
+static int speedstep_activate(void)
+{
+	u16 value = 0;
+
+	if (!speedstep_chipset_dev)
+		return -EINVAL;
+
+	pci_read_config_word(speedstep_chipset_dev, 0x00A0, &value);
+	if (!(value & 0x08)) {
+		value |= 0x08;
+		pr_debug("activating SpeedStep (TM) registers\n");
+		pci_write_config_word(speedstep_chipset_dev, 0x00A0, value);
+	}
+
+	return 0;
+}
+
+
+/**
+ * speedstep_detect_chipset - detect the Southbridge which contains SpeedStep logic
+ *
+ *   Detects ICH2-M, ICH3-M and ICH4-M so far. The pci_dev points to
+ * the LPC bridge / PM module which contains all power-management
+ * functions. Returns the SPEEDSTEP_CHIPSET_-number for the detected
+ * chipset, or zero on failure.
+ */
+static unsigned int speedstep_detect_chipset(void)
+{
+	speedstep_chipset_dev = pci_get_subsys(PCI_VENDOR_ID_INTEL,
+			      PCI_DEVICE_ID_INTEL_82801DB_12,
+			      PCI_ANY_ID, PCI_ANY_ID,
+			      NULL);
+	if (speedstep_chipset_dev)
+		return 4; /* 4-M */
+
+	speedstep_chipset_dev = pci_get_subsys(PCI_VENDOR_ID_INTEL,
+			      PCI_DEVICE_ID_INTEL_82801CA_12,
+			      PCI_ANY_ID, PCI_ANY_ID,
+			      NULL);
+	if (speedstep_chipset_dev)
+		return 3; /* 3-M */
+
+
+	speedstep_chipset_dev = pci_get_subsys(PCI_VENDOR_ID_INTEL,
+			      PCI_DEVICE_ID_INTEL_82801BA_10,
+			      PCI_ANY_ID, PCI_ANY_ID,
+			      NULL);
+	if (speedstep_chipset_dev) {
+		/* speedstep.c causes lockups on Dell Inspirons 8000 and
+		 * 8100 which use a pretty old revision of the 82815
+		 * host brige. Abort on these systems.
+		 */
+		static struct pci_dev *hostbridge;
+
+		hostbridge  = pci_get_subsys(PCI_VENDOR_ID_INTEL,
+			      PCI_DEVICE_ID_INTEL_82815_MC,
+			      PCI_ANY_ID, PCI_ANY_ID,
+			      NULL);
+
+		if (!hostbridge)
+			return 2; /* 2-M */
+
+		if (hostbridge->revision < 5) {
+			pr_debug("hostbridge does not support speedstep\n");
+			speedstep_chipset_dev = NULL;
+			pci_dev_put(hostbridge);
+			return 0;
+		}
+
+		pci_dev_put(hostbridge);
+		return 2; /* 2-M */
+	}
+
+	return 0;
+}
+
+static void get_freq_data(void *_speed)
+{
+	unsigned int *speed = _speed;
+
+	*speed = speedstep_get_frequency(speedstep_processor);
+}
+
+static unsigned int speedstep_get(unsigned int cpu)
+{
+	unsigned int speed;
+
+	/* You're supposed to ensure CPU is online. */
+	if (smp_call_function_single(cpu, get_freq_data, &speed, 1) != 0)
+		BUG();
+
+	pr_debug("detected %u kHz as current frequency\n", speed);
+	return speed;
+}
+
+/**
+ * speedstep_target - set a new CPUFreq policy
+ * @policy: new policy
+ * @target_freq: the target frequency
+ * @relation: how that frequency relates to achieved frequency
+ *	(CPUFREQ_RELATION_L or CPUFREQ_RELATION_H)
+ *
+ * Sets a new CPUFreq policy.
+ */
+static int speedstep_target(struct cpufreq_policy *policy,
+			     unsigned int target_freq,
+			     unsigned int relation)
+{
+	unsigned int newstate = 0, policy_cpu;
+	struct cpufreq_freqs freqs;
+	int i;
+
+	if (cpufreq_frequency_table_target(policy, &speedstep_freqs[0],
+				target_freq, relation, &newstate))
+		return -EINVAL;
+
+	policy_cpu = cpumask_any_and(policy->cpus, cpu_online_mask);
+	freqs.old = speedstep_get(policy_cpu);
+	freqs.new = speedstep_freqs[newstate].frequency;
+	freqs.cpu = policy->cpu;
+
+	pr_debug("transiting from %u to %u kHz\n", freqs.old, freqs.new);
+
+	/* no transition necessary */
+	if (freqs.old == freqs.new)
+		return 0;
+
+	for_each_cpu(i, policy->cpus) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	}
+
+	smp_call_function_single(policy_cpu, _speedstep_set_state, &newstate,
+				 true);
+
+	for_each_cpu(i, policy->cpus) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+	}
+
+	return 0;
+}
+
+
+/**
+ * speedstep_verify - verifies a new CPUFreq policy
+ * @policy: new policy
+ *
+ * Limit must be within speedstep_low_freq and speedstep_high_freq, with
+ * at least one border included.
+ */
+static int speedstep_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, &speedstep_freqs[0]);
+}
+
+struct get_freqs {
+	struct cpufreq_policy *policy;
+	int ret;
+};
+
+static void get_freqs_on_cpu(void *_get_freqs)
+{
+	struct get_freqs *get_freqs = _get_freqs;
+
+	get_freqs->ret =
+		speedstep_get_freqs(speedstep_processor,
+			    &speedstep_freqs[SPEEDSTEP_LOW].frequency,
+			    &speedstep_freqs[SPEEDSTEP_HIGH].frequency,
+			    &get_freqs->policy->cpuinfo.transition_latency,
+			    &speedstep_set_state);
+}
+
+static int speedstep_cpu_init(struct cpufreq_policy *policy)
+{
+	int result;
+	unsigned int policy_cpu, speed;
+	struct get_freqs gf;
+
+	/* only run on CPU to be set, or on its sibling */
+#ifdef CONFIG_SMP
+	cpumask_copy(policy->cpus, cpu_sibling_mask(policy->cpu));
+#endif
+	policy_cpu = cpumask_any_and(policy->cpus, cpu_online_mask);
+
+	/* detect low and high frequency and transition latency */
+	gf.policy = policy;
+	smp_call_function_single(policy_cpu, get_freqs_on_cpu, &gf, 1);
+	if (gf.ret)
+		return gf.ret;
+
+	/* get current speed setting */
+	speed = speedstep_get(policy_cpu);
+	if (!speed)
+		return -EIO;
+
+	pr_debug("currently at %s speed setting - %i MHz\n",
+		(speed == speedstep_freqs[SPEEDSTEP_LOW].frequency)
+		? "low" : "high",
+		(speed / 1000));
+
+	/* cpuinfo and default policy values */
+	policy->cur = speed;
+
+	result = cpufreq_frequency_table_cpuinfo(policy, speedstep_freqs);
+	if (result)
+		return result;
+
+	cpufreq_frequency_table_get_attr(speedstep_freqs, policy->cpu);
+
+	return 0;
+}
+
+
+static int speedstep_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+static struct freq_attr *speedstep_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+
+static struct cpufreq_driver speedstep_driver = {
+	.name	= "speedstep-ich",
+	.verify	= speedstep_verify,
+	.target	= speedstep_target,
+	.init	= speedstep_cpu_init,
+	.exit	= speedstep_cpu_exit,
+	.get	= speedstep_get,
+	.owner	= THIS_MODULE,
+	.attr	= speedstep_attr,
+};
+
+
+/**
+ * speedstep_init - initializes the SpeedStep CPUFreq driver
+ *
+ *   Initializes the SpeedStep support. Returns -ENODEV on unsupported
+ * devices, -EINVAL on problems during initiatization, and zero on
+ * success.
+ */
+static int __init speedstep_init(void)
+{
+	/* detect processor */
+	speedstep_processor = speedstep_detect_processor();
+	if (!speedstep_processor) {
+		pr_debug("Intel(R) SpeedStep(TM) capable processor "
+				"not found\n");
+		return -ENODEV;
+	}
+
+	/* detect chipset */
+	if (!speedstep_detect_chipset()) {
+		pr_debug("Intel(R) SpeedStep(TM) for this chipset not "
+				"(yet) available.\n");
+		return -ENODEV;
+	}
+
+	/* activate speedstep support */
+	if (speedstep_activate()) {
+		pci_dev_put(speedstep_chipset_dev);
+		return -EINVAL;
+	}
+
+	if (speedstep_find_register())
+		return -ENODEV;
+
+	return cpufreq_register_driver(&speedstep_driver);
+}
+
+
+/**
+ * speedstep_exit - unregisters SpeedStep support
+ *
+ *   Unregisters SpeedStep support.
+ */
+static void __exit speedstep_exit(void)
+{
+	pci_dev_put(speedstep_chipset_dev);
+	cpufreq_unregister_driver(&speedstep_driver);
+}
+
+
+MODULE_AUTHOR("Dave Jones <davej@redhat.com>, "
+		"Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("Speedstep driver for Intel mobile processors on chipsets "
+		"with ICH-M southbridges.");
+MODULE_LICENSE("GPL");
+
+module_init(speedstep_init);
+module_exit(speedstep_exit);
diff --git a/drivers/cpufreq/speedstep-lib.c b/drivers/cpufreq/speedstep-lib.c
new file mode 100644
index 00000000..8af2d2fd
--- /dev/null
+++ b/drivers/cpufreq/speedstep-lib.c
@@ -0,0 +1,478 @@
+/*
+ * (C) 2002 - 2003 Dominik Brodowski <linux@brodo.de>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  Library for common functions for Intel SpeedStep v.1 and v.2 support
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+
+#include <asm/msr.h>
+#include <asm/tsc.h>
+#include "speedstep-lib.h"
+
+#define PFX "speedstep-lib: "
+
+#ifdef CONFIG_X86_SPEEDSTEP_RELAXED_CAP_CHECK
+static int relaxed_check;
+#else
+#define relaxed_check 0
+#endif
+
+/*********************************************************************
+ *                   GET PROCESSOR CORE SPEED IN KHZ                 *
+ *********************************************************************/
+
+static unsigned int pentium3_get_frequency(enum speedstep_processor processor)
+{
+	/* See table 14 of p3_ds.pdf and table 22 of 29834003.pdf */
+	struct {
+		unsigned int ratio;	/* Frequency Multiplier (x10) */
+		u8 bitmap;		/* power on configuration bits
+					[27, 25:22] (in MSR 0x2a) */
+	} msr_decode_mult[] = {
+		{ 30, 0x01 },
+		{ 35, 0x05 },
+		{ 40, 0x02 },
+		{ 45, 0x06 },
+		{ 50, 0x00 },
+		{ 55, 0x04 },
+		{ 60, 0x0b },
+		{ 65, 0x0f },
+		{ 70, 0x09 },
+		{ 75, 0x0d },
+		{ 80, 0x0a },
+		{ 85, 0x26 },
+		{ 90, 0x20 },
+		{ 100, 0x2b },
+		{ 0, 0xff }	/* error or unknown value */
+	};
+
+	/* PIII(-M) FSB settings: see table b1-b of 24547206.pdf */
+	struct {
+		unsigned int value;	/* Front Side Bus speed in MHz */
+		u8 bitmap;		/* power on configuration bits [18: 19]
+					(in MSR 0x2a) */
+	} msr_decode_fsb[] = {
+		{  66, 0x0 },
+		{ 100, 0x2 },
+		{ 133, 0x1 },
+		{   0, 0xff}
+	};
+
+	u32 msr_lo, msr_tmp;
+	int i = 0, j = 0;
+
+	/* read MSR 0x2a - we only need the low 32 bits */
+	rdmsr(MSR_IA32_EBL_CR_POWERON, msr_lo, msr_tmp);
+	pr_debug("P3 - MSR_IA32_EBL_CR_POWERON: 0x%x 0x%x\n", msr_lo, msr_tmp);
+	msr_tmp = msr_lo;
+
+	/* decode the FSB */
+	msr_tmp &= 0x00c0000;
+	msr_tmp >>= 18;
+	while (msr_tmp != msr_decode_fsb[i].bitmap) {
+		if (msr_decode_fsb[i].bitmap == 0xff)
+			return 0;
+		i++;
+	}
+
+	/* decode the multiplier */
+	if (processor == SPEEDSTEP_CPU_PIII_C_EARLY) {
+		pr_debug("workaround for early PIIIs\n");
+		msr_lo &= 0x03c00000;
+	} else
+		msr_lo &= 0x0bc00000;
+	msr_lo >>= 22;
+	while (msr_lo != msr_decode_mult[j].bitmap) {
+		if (msr_decode_mult[j].bitmap == 0xff)
+			return 0;
+		j++;
+	}
+
+	pr_debug("speed is %u\n",
+		(msr_decode_mult[j].ratio * msr_decode_fsb[i].value * 100));
+
+	return msr_decode_mult[j].ratio * msr_decode_fsb[i].value * 100;
+}
+
+
+static unsigned int pentiumM_get_frequency(void)
+{
+	u32 msr_lo, msr_tmp;
+
+	rdmsr(MSR_IA32_EBL_CR_POWERON, msr_lo, msr_tmp);
+	pr_debug("PM - MSR_IA32_EBL_CR_POWERON: 0x%x 0x%x\n", msr_lo, msr_tmp);
+
+	/* see table B-2 of 24547212.pdf */
+	if (msr_lo & 0x00040000) {
+		printk(KERN_DEBUG PFX "PM - invalid FSB: 0x%x 0x%x\n",
+				msr_lo, msr_tmp);
+		return 0;
+	}
+
+	msr_tmp = (msr_lo >> 22) & 0x1f;
+	pr_debug("bits 22-26 are 0x%x, speed is %u\n",
+			msr_tmp, (msr_tmp * 100 * 1000));
+
+	return msr_tmp * 100 * 1000;
+}
+
+static unsigned int pentium_core_get_frequency(void)
+{
+	u32 fsb = 0;
+	u32 msr_lo, msr_tmp;
+	int ret;
+
+	rdmsr(MSR_FSB_FREQ, msr_lo, msr_tmp);
+	/* see table B-2 of 25366920.pdf */
+	switch (msr_lo & 0x07) {
+	case 5:
+		fsb = 100000;
+		break;
+	case 1:
+		fsb = 133333;
+		break;
+	case 3:
+		fsb = 166667;
+		break;
+	case 2:
+		fsb = 200000;
+		break;
+	case 0:
+		fsb = 266667;
+		break;
+	case 4:
+		fsb = 333333;
+		break;
+	default:
+		printk(KERN_ERR "PCORE - MSR_FSB_FREQ undefined value");
+	}
+
+	rdmsr(MSR_IA32_EBL_CR_POWERON, msr_lo, msr_tmp);
+	pr_debug("PCORE - MSR_IA32_EBL_CR_POWERON: 0x%x 0x%x\n",
+			msr_lo, msr_tmp);
+
+	msr_tmp = (msr_lo >> 22) & 0x1f;
+	pr_debug("bits 22-26 are 0x%x, speed is %u\n",
+			msr_tmp, (msr_tmp * fsb));
+
+	ret = (msr_tmp * fsb);
+	return ret;
+}
+
+
+static unsigned int pentium4_get_frequency(void)
+{
+	struct cpuinfo_x86 *c = &boot_cpu_data;
+	u32 msr_lo, msr_hi, mult;
+	unsigned int fsb = 0;
+	unsigned int ret;
+	u8 fsb_code;
+
+	/* Pentium 4 Model 0 and 1 do not have the Core Clock Frequency
+	 * to System Bus Frequency Ratio Field in the Processor Frequency
+	 * Configuration Register of the MSR. Therefore the current
+	 * frequency cannot be calculated and has to be measured.
+	 */
+	if (c->x86_model < 2)
+		return cpu_khz;
+
+	rdmsr(0x2c, msr_lo, msr_hi);
+
+	pr_debug("P4 - MSR_EBC_FREQUENCY_ID: 0x%x 0x%x\n", msr_lo, msr_hi);
+
+	/* decode the FSB: see IA-32 Intel (C) Architecture Software
+	 * Developer's Manual, Volume 3: System Prgramming Guide,
+	 * revision #12 in Table B-1: MSRs in the Pentium 4 and
+	 * Intel Xeon Processors, on page B-4 and B-5.
+	 */
+	fsb_code = (msr_lo >> 16) & 0x7;
+	switch (fsb_code) {
+	case 0:
+		fsb = 100 * 1000;
+		break;
+	case 1:
+		fsb = 13333 * 10;
+		break;
+	case 2:
+		fsb = 200 * 1000;
+		break;
+	}
+
+	if (!fsb)
+		printk(KERN_DEBUG PFX "couldn't detect FSB speed. "
+				"Please send an e-mail to <linux@brodo.de>\n");
+
+	/* Multiplier. */
+	mult = msr_lo >> 24;
+
+	pr_debug("P4 - FSB %u kHz; Multiplier %u; Speed %u kHz\n",
+			fsb, mult, (fsb * mult));
+
+	ret = (fsb * mult);
+	return ret;
+}
+
+
+/* Warning: may get called from smp_call_function_single. */
+unsigned int speedstep_get_frequency(enum speedstep_processor processor)
+{
+	switch (processor) {
+	case SPEEDSTEP_CPU_PCORE:
+		return pentium_core_get_frequency();
+	case SPEEDSTEP_CPU_PM:
+		return pentiumM_get_frequency();
+	case SPEEDSTEP_CPU_P4D:
+	case SPEEDSTEP_CPU_P4M:
+		return pentium4_get_frequency();
+	case SPEEDSTEP_CPU_PIII_T:
+	case SPEEDSTEP_CPU_PIII_C:
+	case SPEEDSTEP_CPU_PIII_C_EARLY:
+		return pentium3_get_frequency(processor);
+	default:
+		return 0;
+	};
+	return 0;
+}
+EXPORT_SYMBOL_GPL(speedstep_get_frequency);
+
+
+/*********************************************************************
+ *                 DETECT SPEEDSTEP-CAPABLE PROCESSOR                *
+ *********************************************************************/
+
+unsigned int speedstep_detect_processor(void)
+{
+	struct cpuinfo_x86 *c = &cpu_data(0);
+	u32 ebx, msr_lo, msr_hi;
+
+	pr_debug("x86: %x, model: %x\n", c->x86, c->x86_model);
+
+	if ((c->x86_vendor != X86_VENDOR_INTEL) ||
+	    ((c->x86 != 6) && (c->x86 != 0xF)))
+		return 0;
+
+	if (c->x86 == 0xF) {
+		/* Intel Mobile Pentium 4-M
+		 * or Intel Mobile Pentium 4 with 533 MHz FSB */
+		if (c->x86_model != 2)
+			return 0;
+
+		ebx = cpuid_ebx(0x00000001);
+		ebx &= 0x000000FF;
+
+		pr_debug("ebx value is %x, x86_mask is %x\n", ebx, c->x86_mask);
+
+		switch (c->x86_mask) {
+		case 4:
+			/*
+			 * B-stepping [M-P4-M]
+			 * sample has ebx = 0x0f, production has 0x0e.
+			 */
+			if ((ebx == 0x0e) || (ebx == 0x0f))
+				return SPEEDSTEP_CPU_P4M;
+			break;
+		case 7:
+			/*
+			 * C-stepping [M-P4-M]
+			 * needs to have ebx=0x0e, else it's a celeron:
+			 * cf. 25130917.pdf / page 7, footnote 5 even
+			 * though 25072120.pdf / page 7 doesn't say
+			 * samples are only of B-stepping...
+			 */
+			if (ebx == 0x0e)
+				return SPEEDSTEP_CPU_P4M;
+			break;
+		case 9:
+			/*
+			 * D-stepping [M-P4-M or M-P4/533]
+			 *
+			 * this is totally strange: CPUID 0x0F29 is
+			 * used by M-P4-M, M-P4/533 and(!) Celeron CPUs.
+			 * The latter need to be sorted out as they don't
+			 * support speedstep.
+			 * Celerons with CPUID 0x0F29 may have either
+			 * ebx=0x8 or 0xf -- 25130917.pdf doesn't say anything
+			 * specific.
+			 * M-P4-Ms may have either ebx=0xe or 0xf [see above]
+			 * M-P4/533 have either ebx=0xe or 0xf. [25317607.pdf]
+			 * also, M-P4M HTs have ebx=0x8, too
+			 * For now, they are distinguished by the model_id
+			 * string
+			 */
+			if ((ebx == 0x0e) ||
+				(strstr(c->x86_model_id,
+				    "Mobile Intel(R) Pentium(R) 4") != NULL))
+				return SPEEDSTEP_CPU_P4M;
+			break;
+		default:
+			break;
+		}
+		return 0;
+	}
+
+	switch (c->x86_model) {
+	case 0x0B: /* Intel PIII [Tualatin] */
+		/* cpuid_ebx(1) is 0x04 for desktop PIII,
+		 * 0x06 for mobile PIII-M */
+		ebx = cpuid_ebx(0x00000001);
+		pr_debug("ebx is %x\n", ebx);
+
+		ebx &= 0x000000FF;
+
+		if (ebx != 0x06)
+			return 0;
+
+		/* So far all PIII-M processors support SpeedStep. See
+		 * Intel's 24540640.pdf of June 2003
+		 */
+		return SPEEDSTEP_CPU_PIII_T;
+
+	case 0x08: /* Intel PIII [Coppermine] */
+
+		/* all mobile PIII Coppermines have FSB 100 MHz
+		 * ==> sort out a few desktop PIIIs. */
+		rdmsr(MSR_IA32_EBL_CR_POWERON, msr_lo, msr_hi);
+		pr_debug("Coppermine: MSR_IA32_EBL_CR_POWERON is 0x%x, 0x%x\n",
+				msr_lo, msr_hi);
+		msr_lo &= 0x00c0000;
+		if (msr_lo != 0x0080000)
+			return 0;
+
+		/*
+		 * If the processor is a mobile version,
+		 * platform ID has bit 50 set
+		 * it has SpeedStep technology if either
+		 * bit 56 or 57 is set
+		 */
+		rdmsr(MSR_IA32_PLATFORM_ID, msr_lo, msr_hi);
+		pr_debug("Coppermine: MSR_IA32_PLATFORM ID is 0x%x, 0x%x\n",
+				msr_lo, msr_hi);
+		if ((msr_hi & (1<<18)) &&
+		    (relaxed_check ? 1 : (msr_hi & (3<<24)))) {
+			if (c->x86_mask == 0x01) {
+				pr_debug("early PIII version\n");
+				return SPEEDSTEP_CPU_PIII_C_EARLY;
+			} else
+				return SPEEDSTEP_CPU_PIII_C;
+		}
+
+	default:
+		return 0;
+	}
+}
+EXPORT_SYMBOL_GPL(speedstep_detect_processor);
+
+
+/*********************************************************************
+ *                     DETECT SPEEDSTEP SPEEDS                       *
+ *********************************************************************/
+
+unsigned int speedstep_get_freqs(enum speedstep_processor processor,
+				  unsigned int *low_speed,
+				  unsigned int *high_speed,
+				  unsigned int *transition_latency,
+				  void (*set_state) (unsigned int state))
+{
+	unsigned int prev_speed;
+	unsigned int ret = 0;
+	unsigned long flags;
+	struct timeval tv1, tv2;
+
+	if ((!processor) || (!low_speed) || (!high_speed) || (!set_state))
+		return -EINVAL;
+
+	pr_debug("trying to determine both speeds\n");
+
+	/* get current speed */
+	prev_speed = speedstep_get_frequency(processor);
+	if (!prev_speed)
+		return -EIO;
+
+	pr_debug("previous speed is %u\n", prev_speed);
+
+	local_irq_save(flags);
+
+	/* switch to low state */
+	set_state(SPEEDSTEP_LOW);
+	*low_speed = speedstep_get_frequency(processor);
+	if (!*low_speed) {
+		ret = -EIO;
+		goto out;
+	}
+
+	pr_debug("low speed is %u\n", *low_speed);
+
+	/* start latency measurement */
+	if (transition_latency)
+		do_gettimeofday(&tv1);
+
+	/* switch to high state */
+	set_state(SPEEDSTEP_HIGH);
+
+	/* end latency measurement */
+	if (transition_latency)
+		do_gettimeofday(&tv2);
+
+	*high_speed = speedstep_get_frequency(processor);
+	if (!*high_speed) {
+		ret = -EIO;
+		goto out;
+	}
+
+	pr_debug("high speed is %u\n", *high_speed);
+
+	if (*low_speed == *high_speed) {
+		ret = -ENODEV;
+		goto out;
+	}
+
+	/* switch to previous state, if necessary */
+	if (*high_speed != prev_speed)
+		set_state(SPEEDSTEP_LOW);
+
+	if (transition_latency) {
+		*transition_latency = (tv2.tv_sec - tv1.tv_sec) * USEC_PER_SEC +
+			tv2.tv_usec - tv1.tv_usec;
+		pr_debug("transition latency is %u uSec\n", *transition_latency);
+
+		/* convert uSec to nSec and add 20% for safety reasons */
+		*transition_latency *= 1200;
+
+		/* check if the latency measurement is too high or too low
+		 * and set it to a safe value (500uSec) in that case
+		 */
+		if (*transition_latency > 10000000 ||
+		    *transition_latency < 50000) {
+			printk(KERN_WARNING PFX "frequency transition "
+					"measured seems out of range (%u "
+					"nSec), falling back to a safe one of"
+					"%u nSec.\n",
+					*transition_latency, 500000);
+			*transition_latency = 500000;
+		}
+	}
+
+out:
+	local_irq_restore(flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(speedstep_get_freqs);
+
+#ifdef CONFIG_X86_SPEEDSTEP_RELAXED_CAP_CHECK
+module_param(relaxed_check, int, 0444);
+MODULE_PARM_DESC(relaxed_check,
+		"Don't do all checks for speedstep capability.");
+#endif
+
+MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>");
+MODULE_DESCRIPTION("Library for Intel SpeedStep 1 or 2 cpufreq drivers.");
+MODULE_LICENSE("GPL");
diff --git a/drivers/cpufreq/speedstep-lib.h b/drivers/cpufreq/speedstep-lib.h
new file mode 100644
index 00000000..70d9cea1
--- /dev/null
+++ b/drivers/cpufreq/speedstep-lib.h
@@ -0,0 +1,49 @@
+/*
+ * (C) 2002 - 2003 Dominik Brodowski <linux@brodo.de>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ *  Library for common functions for Intel SpeedStep v.1 and v.2 support
+ *
+ *  BIG FAT DISCLAIMER: Work in progress code. Possibly *dangerous*
+ */
+
+
+
+/* processors */
+enum speedstep_processor {
+	SPEEDSTEP_CPU_PIII_C_EARLY = 0x00000001,  /* Coppermine core */
+	SPEEDSTEP_CPU_PIII_C	   = 0x00000002,  /* Coppermine core */
+	SPEEDSTEP_CPU_PIII_T	   = 0x00000003,  /* Tualatin core */
+	SPEEDSTEP_CPU_P4M	   = 0x00000004,  /* P4-M  */
+/* the following processors are not speedstep-capable and are not auto-detected
+ * in speedstep_detect_processor(). However, their speed can be detected using
+ * the speedstep_get_frequency() call. */
+	SPEEDSTEP_CPU_PM	   = 0xFFFFFF03,  /* Pentium M  */
+	SPEEDSTEP_CPU_P4D	   = 0xFFFFFF04,  /* desktop P4  */
+	SPEEDSTEP_CPU_PCORE	   = 0xFFFFFF05,  /* Core */
+};
+
+/* speedstep states -- only two of them */
+
+#define SPEEDSTEP_HIGH	0x00000000
+#define SPEEDSTEP_LOW	0x00000001
+
+
+/* detect a speedstep-capable processor */
+extern enum speedstep_processor speedstep_detect_processor(void);
+
+/* detect the current speed (in khz) of the processor */
+extern unsigned int speedstep_get_frequency(enum speedstep_processor processor);
+
+
+/* detect the low and high speeds of the processor. The callback
+ * set_state"'s first argument is either SPEEDSTEP_HIGH or
+ * SPEEDSTEP_LOW; the second argument is zero so that no
+ * cpufreq_notify_transition calls are initiated.
+ */
+extern unsigned int speedstep_get_freqs(enum speedstep_processor processor,
+	unsigned int *low_speed,
+	unsigned int *high_speed,
+	unsigned int *transition_latency,
+	void (*set_state) (unsigned int state));
diff --git a/drivers/cpufreq/speedstep-smi.c b/drivers/cpufreq/speedstep-smi.c
new file mode 100644
index 00000000..c76ead34
--- /dev/null
+++ b/drivers/cpufreq/speedstep-smi.c
@@ -0,0 +1,464 @@
+/*
+ * Intel SpeedStep SMI driver.
+ *
+ * (C) 2003  Hiroshi Miura <miura@da-cha.org>
+ *
+ *  Licensed under the terms of the GNU GPL License version 2.
+ *
+ */
+
+
+/*********************************************************************
+ *                        SPEEDSTEP - DEFINITIONS                    *
+ *********************************************************************/
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/cpufreq.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <asm/ist.h>
+
+#include "speedstep-lib.h"
+
+/* speedstep system management interface port/command.
+ *
+ * These parameters are got from IST-SMI BIOS call.
+ * If user gives it, these are used.
+ *
+ */
+static int smi_port;
+static int smi_cmd;
+static unsigned int smi_sig;
+
+/* info about the processor */
+static enum speedstep_processor speedstep_processor;
+
+/*
+ * There are only two frequency states for each processor. Values
+ * are in kHz for the time being.
+ */
+static struct cpufreq_frequency_table speedstep_freqs[] = {
+	{SPEEDSTEP_HIGH,	0},
+	{SPEEDSTEP_LOW,		0},
+	{0,			CPUFREQ_TABLE_END},
+};
+
+#define GET_SPEEDSTEP_OWNER 0
+#define GET_SPEEDSTEP_STATE 1
+#define SET_SPEEDSTEP_STATE 2
+#define GET_SPEEDSTEP_FREQS 4
+
+/* how often shall the SMI call be tried if it failed, e.g. because
+ * of DMA activity going on? */
+#define SMI_TRIES 5
+
+/**
+ * speedstep_smi_ownership
+ */
+static int speedstep_smi_ownership(void)
+{
+	u32 command, result, magic, dummy;
+	u32 function = GET_SPEEDSTEP_OWNER;
+	unsigned char magic_data[] = "Copyright (c) 1999 Intel Corporation";
+
+	command = (smi_sig & 0xffffff00) | (smi_cmd & 0xff);
+	magic = virt_to_phys(magic_data);
+
+	pr_debug("trying to obtain ownership with command %x at port %x\n",
+			command, smi_port);
+
+	__asm__ __volatile__(
+		"push %%ebp\n"
+		"out %%al, (%%dx)\n"
+		"pop %%ebp\n"
+		: "=D" (result),
+		  "=a" (dummy), "=b" (dummy), "=c" (dummy), "=d" (dummy),
+		  "=S" (dummy)
+		: "a" (command), "b" (function), "c" (0), "d" (smi_port),
+		  "D" (0), "S" (magic)
+		: "memory"
+	);
+
+	pr_debug("result is %x\n", result);
+
+	return result;
+}
+
+/**
+ * speedstep_smi_get_freqs - get SpeedStep preferred & current freq.
+ * @low: the low frequency value is placed here
+ * @high: the high frequency value is placed here
+ *
+ * Only available on later SpeedStep-enabled systems, returns false results or
+ * even hangs [cf. bugme.osdl.org # 1422] on earlier systems. Empirical testing
+ * shows that the latter occurs if !(ist_info.event & 0xFFFF).
+ */
+static int speedstep_smi_get_freqs(unsigned int *low, unsigned int *high)
+{
+	u32 command, result = 0, edi, high_mhz, low_mhz, dummy;
+	u32 state = 0;
+	u32 function = GET_SPEEDSTEP_FREQS;
+
+	if (!(ist_info.event & 0xFFFF)) {
+		pr_debug("bug #1422 -- can't read freqs from BIOS\n");
+		return -ENODEV;
+	}
+
+	command = (smi_sig & 0xffffff00) | (smi_cmd & 0xff);
+
+	pr_debug("trying to determine frequencies with command %x at port %x\n",
+			command, smi_port);
+
+	__asm__ __volatile__(
+		"push %%ebp\n"
+		"out %%al, (%%dx)\n"
+		"pop %%ebp"
+		: "=a" (result),
+		  "=b" (high_mhz),
+		  "=c" (low_mhz),
+		  "=d" (state), "=D" (edi), "=S" (dummy)
+		: "a" (command),
+		  "b" (function),
+		  "c" (state),
+		  "d" (smi_port), "S" (0), "D" (0)
+	);
+
+	pr_debug("result %x, low_freq %u, high_freq %u\n",
+			result, low_mhz, high_mhz);
+
+	/* abort if results are obviously incorrect... */
+	if ((high_mhz + low_mhz) < 600)
+		return -EINVAL;
+
+	*high = high_mhz * 1000;
+	*low  = low_mhz  * 1000;
+
+	return result;
+}
+
+/**
+ * speedstep_get_state - set the SpeedStep state
+ * @state: processor frequency state (SPEEDSTEP_LOW or SPEEDSTEP_HIGH)
+ *
+ */
+static int speedstep_get_state(void)
+{
+	u32 function = GET_SPEEDSTEP_STATE;
+	u32 result, state, edi, command, dummy;
+
+	command = (smi_sig & 0xffffff00) | (smi_cmd & 0xff);
+
+	pr_debug("trying to determine current setting with command %x "
+		"at port %x\n", command, smi_port);
+
+	__asm__ __volatile__(
+		"push %%ebp\n"
+		"out %%al, (%%dx)\n"
+		"pop %%ebp\n"
+		: "=a" (result),
+		  "=b" (state), "=D" (edi),
+		  "=c" (dummy), "=d" (dummy), "=S" (dummy)
+		: "a" (command), "b" (function), "c" (0),
+		  "d" (smi_port), "S" (0), "D" (0)
+	);
+
+	pr_debug("state is %x, result is %x\n", state, result);
+
+	return state & 1;
+}
+
+
+/**
+ * speedstep_set_state - set the SpeedStep state
+ * @state: new processor frequency state (SPEEDSTEP_LOW or SPEEDSTEP_HIGH)
+ *
+ */
+static void speedstep_set_state(unsigned int state)
+{
+	unsigned int result = 0, command, new_state, dummy;
+	unsigned long flags;
+	unsigned int function = SET_SPEEDSTEP_STATE;
+	unsigned int retry = 0;
+
+	if (state > 0x1)
+		return;
+
+	/* Disable IRQs */
+	local_irq_save(flags);
+
+	command = (smi_sig & 0xffffff00) | (smi_cmd & 0xff);
+
+	pr_debug("trying to set frequency to state %u "
+		"with command %x at port %x\n",
+		state, command, smi_port);
+
+	do {
+		if (retry) {
+			pr_debug("retry %u, previous result %u, waiting...\n",
+					retry, result);
+			mdelay(retry * 50);
+		}
+		retry++;
+		__asm__ __volatile__(
+			"push %%ebp\n"
+			"out %%al, (%%dx)\n"
+			"pop %%ebp"
+			: "=b" (new_state), "=D" (result),
+			  "=c" (dummy), "=a" (dummy),
+			  "=d" (dummy), "=S" (dummy)
+			: "a" (command), "b" (function), "c" (state),
+			  "d" (smi_port), "S" (0), "D" (0)
+			);
+	} while ((new_state != state) && (retry <= SMI_TRIES));
+
+	/* enable IRQs */
+	local_irq_restore(flags);
+
+	if (new_state == state)
+		pr_debug("change to %u MHz succeeded after %u tries "
+			"with result %u\n",
+			(speedstep_freqs[new_state].frequency / 1000),
+			retry, result);
+	else
+		printk(KERN_ERR "cpufreq: change to state %u "
+			"failed with new_state %u and result %u\n",
+			state, new_state, result);
+
+	return;
+}
+
+
+/**
+ * speedstep_target - set a new CPUFreq policy
+ * @policy: new policy
+ * @target_freq: new freq
+ * @relation:
+ *
+ * Sets a new CPUFreq policy/freq.
+ */
+static int speedstep_target(struct cpufreq_policy *policy,
+			unsigned int target_freq, unsigned int relation)
+{
+	unsigned int newstate = 0;
+	struct cpufreq_freqs freqs;
+
+	if (cpufreq_frequency_table_target(policy, &speedstep_freqs[0],
+				target_freq, relation, &newstate))
+		return -EINVAL;
+
+	freqs.old = speedstep_freqs[speedstep_get_state()].frequency;
+	freqs.new = speedstep_freqs[newstate].frequency;
+	freqs.cpu = 0; /* speedstep.c is UP only driver */
+
+	if (freqs.old == freqs.new)
+		return 0;
+
+	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	speedstep_set_state(newstate);
+	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+
+	return 0;
+}
+
+
+/**
+ * speedstep_verify - verifies a new CPUFreq policy
+ * @policy: new policy
+ *
+ * Limit must be within speedstep_low_freq and speedstep_high_freq, with
+ * at least one border included.
+ */
+static int speedstep_verify(struct cpufreq_policy *policy)
+{
+	return cpufreq_frequency_table_verify(policy, &speedstep_freqs[0]);
+}
+
+
+static int speedstep_cpu_init(struct cpufreq_policy *policy)
+{
+	int result;
+	unsigned int speed, state;
+	unsigned int *low, *high;
+
+	/* capability check */
+	if (policy->cpu != 0)
+		return -ENODEV;
+
+	result = speedstep_smi_ownership();
+	if (result) {
+		pr_debug("fails in acquiring ownership of a SMI interface.\n");
+		return -EINVAL;
+	}
+
+	/* detect low and high frequency */
+	low = &speedstep_freqs[SPEEDSTEP_LOW].frequency;
+	high = &speedstep_freqs[SPEEDSTEP_HIGH].frequency;
+
+	result = speedstep_smi_get_freqs(low, high);
+	if (result) {
+		/* fall back to speedstep_lib.c dection mechanism:
+		 * try both states out */
+		pr_debug("could not detect low and high frequencies "
+				"by SMI call.\n");
+		result = speedstep_get_freqs(speedstep_processor,
+				low, high,
+				NULL,
+				&speedstep_set_state);
+
+		if (result) {
+			pr_debug("could not detect two different speeds"
+					" -- aborting.\n");
+			return result;
+		} else
+			pr_debug("workaround worked.\n");
+	}
+
+	/* get current speed setting */
+	state = speedstep_get_state();
+	speed = speedstep_freqs[state].frequency;
+
+	pr_debug("currently at %s speed setting - %i MHz\n",
+		(speed == speedstep_freqs[SPEEDSTEP_LOW].frequency)
+		? "low" : "high",
+		(speed / 1000));
+
+	/* cpuinfo and default policy values */
+	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
+	policy->cur = speed;
+
+	result = cpufreq_frequency_table_cpuinfo(policy, speedstep_freqs);
+	if (result)
+		return result;
+
+	cpufreq_frequency_table_get_attr(speedstep_freqs, policy->cpu);
+
+	return 0;
+}
+
+static int speedstep_cpu_exit(struct cpufreq_policy *policy)
+{
+	cpufreq_frequency_table_put_attr(policy->cpu);
+	return 0;
+}
+
+static unsigned int speedstep_get(unsigned int cpu)
+{
+	if (cpu)
+		return -ENODEV;
+	return speedstep_get_frequency(speedstep_processor);
+}
+
+
+static int speedstep_resume(struct cpufreq_policy *policy)
+{
+	int result = speedstep_smi_ownership();
+
+	if (result)
+		pr_debug("fails in re-acquiring ownership of a SMI interface.\n");
+
+	return result;
+}
+
+static struct freq_attr *speedstep_attr[] = {
+	&cpufreq_freq_attr_scaling_available_freqs,
+	NULL,
+};
+
+static struct cpufreq_driver speedstep_driver = {
+	.name		= "speedstep-smi",
+	.verify		= speedstep_verify,
+	.target		= speedstep_target,
+	.init		= speedstep_cpu_init,
+	.exit		= speedstep_cpu_exit,
+	.get		= speedstep_get,
+	.resume		= speedstep_resume,
+	.owner		= THIS_MODULE,
+	.attr		= speedstep_attr,
+};
+
+/**
+ * speedstep_init - initializes the SpeedStep CPUFreq driver
+ *
+ *   Initializes the SpeedStep support. Returns -ENODEV on unsupported
+ * BIOS, -EINVAL on problems during initiatization, and zero on
+ * success.
+ */
+static int __init speedstep_init(void)
+{
+	speedstep_processor = speedstep_detect_processor();
+
+	switch (speedstep_processor) {
+	case SPEEDSTEP_CPU_PIII_T:
+	case SPEEDSTEP_CPU_PIII_C:
+	case SPEEDSTEP_CPU_PIII_C_EARLY:
+		break;
+	default:
+		speedstep_processor = 0;
+	}
+
+	if (!speedstep_processor) {
+		pr_debug("No supported Intel CPU detected.\n");
+		return -ENODEV;
+	}
+
+	pr_debug("signature:0x%.8ulx, command:0x%.8ulx, "
+		"event:0x%.8ulx, perf_level:0x%.8ulx.\n",
+		ist_info.signature, ist_info.command,
+		ist_info.event, ist_info.perf_level);
+
+	/* Error if no IST-SMI BIOS or no PARM
+		 sig= 'ISGE' aka 'Intel Speedstep Gate E' */
+	if ((ist_info.signature !=  0x47534943) && (
+	    (smi_port == 0) || (smi_cmd == 0)))
+		return -ENODEV;
+
+	if (smi_sig == 1)
+		smi_sig = 0x47534943;
+	else
+		smi_sig = ist_info.signature;
+
+	/* setup smi_port from MODLULE_PARM or BIOS */
+	if ((smi_port > 0xff) || (smi_port < 0))
+		return -EINVAL;
+	else if (smi_port == 0)
+		smi_port = ist_info.command & 0xff;
+
+	if ((smi_cmd > 0xff) || (smi_cmd < 0))
+		return -EINVAL;
+	else if (smi_cmd == 0)
+		smi_cmd = (ist_info.command >> 16) & 0xff;
+
+	return cpufreq_register_driver(&speedstep_driver);
+}
+
+
+/**
+ * speedstep_exit - unregisters SpeedStep support
+ *
+ *   Unregisters SpeedStep support.
+ */
+static void __exit speedstep_exit(void)
+{
+	cpufreq_unregister_driver(&speedstep_driver);
+}
+
+module_param(smi_port, int, 0444);
+module_param(smi_cmd,  int, 0444);
+module_param(smi_sig, uint, 0444);
+
+MODULE_PARM_DESC(smi_port, "Override the BIOS-given IST port with this value "
+		"-- Intel's default setting is 0xb2");
+MODULE_PARM_DESC(smi_cmd, "Override the BIOS-given IST command with this value "
+		"-- Intel's default setting is 0x82");
+MODULE_PARM_DESC(smi_sig, "Set to 1 to fake the IST signature when using the "
+		"SMI interface.");
+
+MODULE_AUTHOR("Hiroshi Miura");
+MODULE_DESCRIPTION("Speedstep driver for IST applet SMI interface.");
+MODULE_LICENSE("GPL");
+
+module_init(speedstep_init);
+module_exit(speedstep_exit);