Automatic merge with /usr/src/ntfs-2.6.git

53 files changed, 911 insertions, 302 deletions

diff --git a/arch/h8300/kernel/process.c b/arch/h8300/kernel/process.c
index 134aec1c6d1..b5f83e9f04d 100644
--- a/arch/h8300/kernel/process.c
+++ b/arch/h8300/kernel/process.c
@@ -54,7 +54,7 @@ asmlinkage void ret_from_fork(void);
 void default_idle(void)
 {
 	while(1) {
-		if (need_resched()) {
+		if (!need_resched()) {
 			local_irq_enable();
 			__asm__("sleep");
 			local_irq_disable();
diff --git a/arch/i386/kernel/cpu/cpufreq/Kconfig b/arch/i386/kernel/cpu/cpufreq/Kconfig
index f25ffd74235..0f1eb507233 100644
--- a/arch/i386/kernel/cpu/cpufreq/Kconfig
+++ b/arch/i386/kernel/cpu/cpufreq/Kconfig
@@ -23,7 +23,7 @@ config X86_ACPI_CPUFREQ
 	  If in doubt, say N.
 
 config ELAN_CPUFREQ
-	tristate "AMD Elan"
+	tristate "AMD Elan SC400 and SC410"
 	select CPU_FREQ_TABLE
 	depends on X86_ELAN
 	---help---
@@ -38,6 +38,18 @@ config ELAN_CPUFREQ
 
 	  If in doubt, say N.
 
+config SC520_CPUFREQ
+	tristate "AMD Elan SC520"
+	select CPU_FREQ_TABLE
+	depends on X86_ELAN
+	---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
diff --git a/arch/i386/kernel/cpu/cpufreq/Makefile b/arch/i386/kernel/cpu/cpufreq/Makefile
index a922e97aeed..2e894f1c891 100644
--- a/arch/i386/kernel/cpu/cpufreq/Makefile
+++ b/arch/i386/kernel/cpu/cpufreq/Makefile
@@ -3,6 +3,7 @@ obj-$(CONFIG_X86_POWERNOW_K7)		+= powernow-k7.o
 obj-$(CONFIG_X86_POWERNOW_K8)		+= powernow-k8.o
 obj-$(CONFIG_X86_LONGHAUL)		+= longhaul.o
 obj-$(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
diff --git a/arch/i386/kernel/cpu/cpufreq/longhaul.c b/arch/i386/kernel/cpu/cpufreq/longhaul.c
index ab0f9f5aac1..04e3563da4f 100644
--- a/arch/i386/kernel/cpu/cpufreq/longhaul.c
+++ b/arch/i386/kernel/cpu/cpufreq/longhaul.c
@@ -29,6 +29,7 @@
 #include <linux/cpufreq.h>
 #include <linux/slab.h>
 #include <linux/string.h>
+#include <linux/pci.h>
 
 #include <asm/msr.h>
 #include <asm/timex.h>
@@ -119,7 +120,13 @@ static int longhaul_get_cpu_mult(void)
 static void do_powersaver(union msr_longhaul *longhaul,
 			unsigned int clock_ratio_index)
 {
+	struct pci_dev *dev;
+	unsigned long flags;
+	unsigned int tmp_mask;
 	int version;
+	int i;
+	u16 pci_cmd;
+	u16 cmd_state[64];
 
 	switch (cpu_model) {
 	case CPU_EZRA_T:
@@ -137,17 +144,58 @@ static void do_powersaver(union msr_longhaul *longhaul,
 	longhaul->bits.SoftBusRatio4 = (clock_ratio_index & 0x10) >> 4;
 	longhaul->bits.EnableSoftBusRatio = 1;
 	longhaul->bits.RevisionKey = 0;
-	local_irq_disable();
-	wrmsrl(MSR_VIA_LONGHAUL, longhaul->val);
+
+	preempt_disable();
+	local_irq_save(flags);
+
+	/*
+	 * get current pci bus master state for all devices
+	 * and clear bus master bit
+	 */
+	dev = NULL;
+	i = 0;
+	do {
+		dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev);
+		if (dev != NULL) {
+			pci_read_config_word(dev, PCI_COMMAND, &pci_cmd);
+			cmd_state[i++] = pci_cmd;
+			pci_cmd &= ~PCI_COMMAND_MASTER;
+			pci_write_config_word(dev, PCI_COMMAND, pci_cmd);
+		}
+	} while (dev != NULL);
+
+	tmp_mask=inb(0x21);	/* works on C3. save mask. */
+	outb(0xFE,0x21);	/* TMR0 only */
+	outb(0xFF,0x80);	/* delay */
+
 	local_irq_enable();
+
+	__hlt();
+	wrmsrl(MSR_VIA_LONGHAUL, longhaul->val);
 	__hlt();
 
+	local_irq_disable();
+
+	outb(tmp_mask,0x21);	/* restore mask */
+
+	/* restore pci bus master state for all devices */
+	dev = NULL;
+	i = 0;
+	do {
+		dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev);
+		if (dev != NULL) {
+			pci_cmd = cmd_state[i++];
+			pci_write_config_byte(dev, PCI_COMMAND, pci_cmd);
+		}
+	} while (dev != NULL);
+	local_irq_restore(flags);
+	preempt_enable();
+
+	/* disable bus ratio bit */
 	rdmsrl(MSR_VIA_LONGHAUL, longhaul->val);
 	longhaul->bits.EnableSoftBusRatio = 0;
 	longhaul->bits.RevisionKey = version;
-	local_irq_disable();
 	wrmsrl(MSR_VIA_LONGHAUL, longhaul->val);
-	local_irq_enable();
 }
 
 /**
@@ -578,7 +626,7 @@ static int __init longhaul_cpu_init(struct cpufreq_policy *policy)
 		longhaul_setup_voltagescaling();
 
 	policy->governor = CPUFREQ_DEFAULT_GOVERNOR;
-	policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
+	policy->cpuinfo.transition_latency = 200000;	/* nsec */
 	policy->cur = calc_speed(longhaul_get_cpu_mult());
 
 	ret = cpufreq_frequency_table_cpuinfo(policy, longhaul_table);
diff --git a/arch/i386/kernel/cpu/cpufreq/powernow-k7.c b/arch/i386/kernel/cpu/cpufreq/powernow-k7.c
index 913f652623d..5c530064eb7 100644
--- a/arch/i386/kernel/cpu/cpufreq/powernow-k7.c
+++ b/arch/i386/kernel/cpu/cpufreq/powernow-k7.c
@@ -23,6 +23,7 @@
 #include <linux/dmi.h>
 
 #include <asm/msr.h>
+#include <asm/timer.h>
 #include <asm/timex.h>
 #include <asm/io.h>
 #include <asm/system.h>
@@ -586,13 +587,17 @@ static int __init powernow_cpu_init (struct cpufreq_policy *policy)
 
 	rdmsrl (MSR_K7_FID_VID_STATUS, fidvidstatus.val);
 
-	/* A K7 with powernow technology is set to max frequency by BIOS */
-	fsb = (10 * cpu_khz) / fid_codes[fidvidstatus.bits.MFID];
+	/* recalibrate cpu_khz */
+	result = recalibrate_cpu_khz();
+	if (result)
+		return result;
+
+	fsb = (10 * cpu_khz) / fid_codes[fidvidstatus.bits.CFID];
 	if (!fsb) {
 		printk(KERN_WARNING PFX "can not determine bus frequency\n");
 		return -EINVAL;
 	}
-	dprintk("FSB: %3d.%03d MHz\n", fsb/1000, fsb%1000);
+	dprintk("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");
diff --git a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c
index a65ff7e32e5..10cc096c0ad 100644
--- a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c
+++ b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c
@@ -4,7 +4,7 @@
  *  GNU general public license version 2. See "COPYING" or
  *  http://www.gnu.org/licenses/gpl.html
  *
- *  Support : paul.devriendt@amd.com
+ *  Support : mark.langsdorf@amd.com
  *
  *  Based on the powernow-k7.c module written by Dave Jones.
  *  (C) 2003 Dave Jones <davej@codemonkey.org.uk> on behalf of SuSE Labs
@@ -15,12 +15,13 @@
  *
  *  Valuable input gratefully received from Dave Jones, Pavel Machek,
  *  Dominik Brodowski, 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 infrerred from
- *	http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/30430.pdf
+ *     http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/30430.pdf
  */
 
 #include <linux/kernel.h>
@@ -30,6 +31,7 @@
 #include <linux/cpufreq.h>
 #include <linux/slab.h>
 #include <linux/string.h>
+#include <linux/cpumask.h>
 
 #include <asm/msr.h>
 #include <asm/io.h>
@@ -42,7 +44,7 @@
 
 #define PFX "powernow-k8: "
 #define BFX PFX "BIOS error: "
-#define VERSION "version 1.00.09e"
+#define VERSION "version 1.40.2"
 #include "powernow-k8.h"
 
 /* serialize freq changes  */
@@ -50,6 +52,10 @@ static DECLARE_MUTEX(fidvid_sem);
 
 static struct powernow_k8_data *powernow_data[NR_CPUS];
 
+#ifndef CONFIG_SMP
+static cpumask_t cpu_core_map[1];
+#endif
+
 /* Return a frequency in MHz, given an input fid */
 static u32 find_freq_from_fid(u32 fid)
 {
@@ -274,11 +280,18 @@ static int core_voltage_pre_transition(struct powernow_k8_data *data, u32 reqvid
 {
 	u32 rvosteps = data->rvo;
 	u32 savefid = data->currfid;
+	u32 maxvid, lo;
 
 	dprintk("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);
 
+	rdmsr(MSR_FIDVID_STATUS, lo, maxvid);
+	maxvid = 0x1f & (maxvid >> 16);
+	dprintk("ph1 maxvid=0x%x\n", maxvid);
+	if (reqvid < maxvid) /* lower numbers are higher voltages */
+		reqvid = maxvid;
+
 	while (data->currvid > reqvid) {
 		dprintk("ph1: curr 0x%x, req vid 0x%x\n",
 			data->currvid, reqvid);
@@ -286,8 +299,8 @@ static int core_voltage_pre_transition(struct powernow_k8_data *data, u32 reqvid
 			return 1;
 	}
 
-	while ((rvosteps > 0)  && ((data->rvo + data->currvid) > reqvid)) {
-		if (data->currvid == 0) {
+	while ((rvosteps > 0) && ((data->rvo + data->currvid) > reqvid)) {
+		if (data->currvid == maxvid) {
 			rvosteps = 0;
 		} else {
 			dprintk("ph1: changing vid for rvo, req 0x%x\n",
@@ -671,7 +684,7 @@ static int find_psb_table(struct powernow_k8_data *data)
 	 * BIOS and Kernel Developer's Guide, which is available on
 	 * www.amd.com
 	 */
-	printk(KERN_ERR PFX "BIOS error - no PSB\n");
+	printk(KERN_INFO PFX "BIOS error - no PSB or ACPI _PSS objects\n");
 	return -ENODEV;
 }
 
@@ -695,7 +708,7 @@ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
 	struct cpufreq_frequency_table *powernow_table;
 
 	if (acpi_processor_register_performance(&data->acpi_data, data->cpu)) {
-		dprintk("register performance failed\n");
+		dprintk("register performance failed: bad ACPI data\n");
 		return -EIO;
 	}
 
@@ -746,22 +759,23 @@ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
 			continue;
 		}
 
- 		if (fid < HI_FID_TABLE_BOTTOM) {
- 			if (cntlofreq) {
- 				/* if both entries are the same, ignore this
- 				 * one... 
- 				 */
- 				if ((powernow_table[i].frequency != powernow_table[cntlofreq].frequency) ||
- 				    (powernow_table[i].index != powernow_table[cntlofreq].index)) {
- 					printk(KERN_ERR PFX "Too many lo freq table entries\n");
- 					goto err_out_mem;
- 				}
-				
- 				dprintk("double low frequency table entry, ignoring it.\n");
- 				powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
- 				continue;
- 			} else
- 				cntlofreq = i;
+		/* verify only 1 entry from the lo frequency table */
+		if (fid < HI_FID_TABLE_BOTTOM) {
+			if (cntlofreq) {
+				/* if both entries are the same, ignore this
+				 * one... 
+				 */
+				if ((powernow_table[i].frequency != powernow_table[cntlofreq].frequency) ||
+				    (powernow_table[i].index != powernow_table[cntlofreq].index)) {
+					printk(KERN_ERR PFX "Too many lo freq table entries\n");
+					goto err_out_mem;
+				}
+
+				dprintk("double low frequency table entry, ignoring it.\n");
+				powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+				continue;
+			} else
+				cntlofreq = i;
 		}
 
 		if (powernow_table[i].frequency != (data->acpi_data.states[i].core_frequency * 1000)) {
@@ -816,7 +830,7 @@ static int transition_frequency(struct powernow_k8_data *data, unsigned int inde
 {
 	u32 fid;
 	u32 vid;
-	int res;
+	int res, i;
 	struct cpufreq_freqs freqs;
 
 	dprintk("cpu %d transition to index %u\n", smp_processor_id(), index);
@@ -841,7 +855,8 @@ static int transition_frequency(struct powernow_k8_data *data, unsigned int inde
 	}
 
 	if ((fid < HI_FID_TABLE_BOTTOM) && (data->currfid < HI_FID_TABLE_BOTTOM)) {
-		printk("ignoring illegal change in lo freq table-%x to 0x%x\n",
+		printk(KERN_ERR PFX
+		       "ignoring illegal change in lo freq table-%x to 0x%x\n",
 		       data->currfid, fid);
 		return 1;
 	}
@@ -850,18 +865,20 @@ static int transition_frequency(struct powernow_k8_data *data, unsigned int inde
 		smp_processor_id(), fid, vid);
 
 	freqs.cpu = data->cpu;
-
 	freqs.old = find_khz_freq_from_fid(data->currfid);
 	freqs.new = find_khz_freq_from_fid(fid);
-	cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	for_each_cpu_mask(i, cpu_core_map[data->cpu]) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+	}
 
-	down(&fidvid_sem);
 	res = transition_fid_vid(data, fid, vid);
-	up(&fidvid_sem);
 
 	freqs.new = find_khz_freq_from_fid(data->currfid);
-	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
-
+	for_each_cpu_mask(i, cpu_core_map[data->cpu]) {
+		freqs.cpu = i;
+		cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+        }
 	return res;
 }
 
@@ -874,6 +891,7 @@ static int powernowk8_target(struct cpufreq_policy *pol, unsigned targfreq, unsi
 	u32 checkvid = data->currvid;
 	unsigned int newstate;
 	int ret = -EIO;
+	int i;
 
 	/* only run on specific CPU from here on */
 	oldmask = current->cpus_allowed;
@@ -902,22 +920,41 @@ static int powernowk8_target(struct cpufreq_policy *pol, unsigned targfreq, unsi
 		data->currfid, data->currvid);
 
 	if ((checkvid != data->currvid) || (checkfid != data->currfid)) {
-		printk(KERN_ERR PFX
-		       "error - out of sync, fid 0x%x 0x%x, vid 0x%x 0x%x\n",
-		       checkfid, data->currfid, checkvid, data->currvid);
+		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;
 
+	down(&fidvid_sem);
+
+	for_each_cpu_mask(i, cpu_core_map[pol->cpu]) {
+		/* make sure the sibling is initialized */
+		if (!powernow_data[i]) {
+                        ret = 0;
+                        up(&fidvid_sem);
+                        goto err_out;
+                }
+	}
+
 	powernow_k8_acpi_pst_values(data, newstate);
 
 	if (transition_frequency(data, newstate)) {
 		printk(KERN_ERR PFX "transition frequency failed\n");
 		ret = 1;
+		up(&fidvid_sem);
 		goto err_out;
 	}
 
+	/* Update all the fid/vids of our siblings */
+	for_each_cpu_mask(i, cpu_core_map[pol->cpu]) {
+		powernow_data[i]->currvid = data->currvid;
+		powernow_data[i]->currfid = data->currfid;
+	}	
+	up(&fidvid_sem);
+
 	pol->cur = find_khz_freq_from_fid(data->currfid);
 	ret = 0;
 
@@ -962,7 +999,7 @@ static int __init powernowk8_cpu_init(struct cpufreq_policy *pol)
 		 */
 
 		if ((num_online_cpus() != 1) || (num_possible_cpus() != 1)) {
-			printk(KERN_INFO PFX "MP systems not supported by PSB BIOS structure\n");
+			printk(KERN_ERR PFX "MP systems not supported by PSB BIOS structure\n");
 			kfree(data);
 			return -ENODEV;
 		}
@@ -1003,6 +1040,7 @@ static int __init powernowk8_cpu_init(struct cpufreq_policy *pol)
 	schedule();
 
 	pol->governor = CPUFREQ_DEFAULT_GOVERNOR;
+	pol->cpus = cpu_core_map[pol->cpu];
 
 	/* Take a crude guess here. 
 	 * That guess was in microseconds, so multiply with 1000 */
@@ -1069,7 +1107,7 @@ static unsigned int powernowk8_get (unsigned int cpu)
 		return 0;
 	}
 	preempt_disable();
-
+	
 	if (query_current_values_with_pending_wait(data))
 		goto out;
 
@@ -1127,9 +1165,10 @@ static void __exit powernowk8_exit(void)
 	cpufreq_unregister_driver(&cpufreq_amd64_driver);
 }
 
-MODULE_AUTHOR("Paul Devriendt <paul.devriendt@amd.com>");
+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/arch/i386/kernel/cpu/cpufreq/powernow-k8.h b/arch/i386/kernel/cpu/cpufreq/powernow-k8.h
index 63ebc8470f5..9ed5bf221cb 100644
--- a/arch/i386/kernel/cpu/cpufreq/powernow-k8.h
+++ b/arch/i386/kernel/cpu/cpufreq/powernow-k8.h
@@ -174,3 +174,18 @@ static int core_voltage_post_transition(struct powernow_k8_data *data, u32 reqvi
 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);
+
+#ifndef for_each_cpu_mask
+#define for_each_cpu_mask(i,mask) for (i=0;i<1;i++)
+#endif
+                                                                                
+#ifdef CONFIG_SMP
+static inline void define_siblings(int cpu, cpumask_t cpu_sharedcore_mask[])
+{
+}
+#else
+static inline void define_siblings(int cpu, cpumask_t cpu_sharedcore_mask[])
+{
+	cpu_set(0, cpu_sharedcore_mask[0]);
+}
+#endif
diff --git a/arch/i386/kernel/cpu/cpufreq/sc520_freq.c b/arch/i386/kernel/cpu/cpufreq/sc520_freq.c
new file mode 100644
index 00000000000..ef457d50f4a
--- /dev/null
+++ b/arch/i386/kernel/cpu/cpufreq/sc520_freq.c
@@ -0,0 +1,186 @@
+/*
+ *	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 <asm/msr.h>
+#include <asm/timex.h>
+#include <asm/io.h>
+
+#define MMCR_BASE	0xfffef000	/* The default base address */
+#define OFFS_CPUCTL	0x2   /* CPU Control Register */
+
+static __u8 __iomem *cpuctl;
+
+#define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, "sc520_freq", msg)
+
+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 "sc520_freq: 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);
+
+	dprintk("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;
+	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->governor = CPUFREQ_DEFAULT_GOVERNOR;
+	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;
+
+	/* Test if we have the right hardware */
+	if(c->x86_vendor != X86_VENDOR_AMD ||
+				c->x86 != 4 || c->x86_model != 9) {
+		dprintk("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;
+	}
+
+	return cpufreq_register_driver(&sc520_freq_driver);
+}
+
+
+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/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c b/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c
index 07d5612dc00..7dcbf70fc16 100644
--- a/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c
+++ b/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c
@@ -54,6 +54,8 @@ enum {
 	CPU_DOTHAN_A1,
 	CPU_DOTHAN_A2,
 	CPU_DOTHAN_B0,
+	CPU_MP4HT_D0,
+	CPU_MP4HT_E0,
 };
 
 static const struct cpu_id cpu_ids[] = {
@@ -61,6 +63,8 @@ static const struct cpu_id cpu_ids[] = {
 	[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	(sizeof(cpu_ids)/sizeof(cpu_ids[0]))
 
@@ -226,6 +230,8 @@ static struct cpu_model models[] =
 	{ &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, }
 };
diff --git a/arch/i386/kernel/cpu/cpufreq/speedstep-lib.c b/arch/i386/kernel/cpu/cpufreq/speedstep-lib.c
index 8ba430a9c3a..d368b3f5fce 100644
--- a/arch/i386/kernel/cpu/cpufreq/speedstep-lib.c
+++ b/arch/i386/kernel/cpu/cpufreq/speedstep-lib.c
@@ -336,7 +336,7 @@ unsigned int speedstep_get_freqs(unsigned int processor,
 	if (!prev_speed)
 		return -EIO;
 
-	dprintk("previous seped is %u\n", prev_speed);
+	dprintk("previous speed is %u\n", prev_speed);
 	
 	local_irq_save(flags);
 
@@ -348,7 +348,7 @@ unsigned int speedstep_get_freqs(unsigned int processor,
 		goto out;
 	}
 
-	dprintk("low seped is %u\n", *low_speed);
+	dprintk("low speed is %u\n", *low_speed);
 
 	/* switch to high state */
 	set_state(SPEEDSTEP_HIGH);
@@ -358,7 +358,7 @@ unsigned int speedstep_get_freqs(unsigned int processor,
 		goto out;
 	}
 
-	dprintk("high seped is %u\n", *high_speed);
+	dprintk("high speed is %u\n", *high_speed);
 
 	if (*low_speed == *high_speed) {
 		ret = -ENODEV;
diff --git a/arch/i386/kernel/cpu/cpufreq/speedstep-smi.c b/arch/i386/kernel/cpu/cpufreq/speedstep-smi.c
index 79440b3f087..b25fb6b635a 100644
--- a/arch/i386/kernel/cpu/cpufreq/speedstep-smi.c
+++ b/arch/i386/kernel/cpu/cpufreq/speedstep-smi.c
@@ -357,6 +357,9 @@ static int __init speedstep_init(void)
 	case SPEEDSTEP_PROCESSOR_PIII_C:
 	case SPEEDSTEP_PROCESSOR_PIII_C_EARLY:
 		break;
+	case SPEEDSTEP_PROCESSOR_P4M:
+		printk(KERN_INFO "speedstep-smi: you're trying to use this cpufreq driver on a Pentium 4-based CPU. Most likely it will not work.\n");
+		break;
 	default:
 		speedstep_processor = 0;
 	}
diff --git a/arch/i386/kernel/setup.c b/arch/i386/kernel/setup.c
index 945ec73163c..2bfbddebdbf 100644
--- a/arch/i386/kernel/setup.c
+++ b/arch/i386/kernel/setup.c
@@ -1502,11 +1502,13 @@ void __init setup_arch(char **cmdline_p)
 	if (efi_enabled)
 		efi_map_memmap();
 
+#ifdef CONFIG_ACPI_BOOT
 	/*
 	 * Parse the ACPI tables for possible boot-time SMP configuration.
 	 */
 	acpi_boot_table_init();
 	acpi_boot_init();
+#endif
 
 #ifdef CONFIG_X86_LOCAL_APIC
 	if (smp_found_config)
diff --git a/arch/i386/kernel/timers/common.c b/arch/i386/kernel/timers/common.c
index f7f90005e22..8e201219f52 100644
--- a/arch/i386/kernel/timers/common.c
+++ b/arch/i386/kernel/timers/common.c
@@ -6,6 +6,7 @@
 #include <linux/timex.h>
 #include <linux/errno.h>
 #include <linux/jiffies.h>
+#include <linux/module.h>
 
 #include <asm/io.h>
 #include <asm/timer.h>
@@ -24,7 +25,7 @@
 
 #define CALIBRATE_TIME	(5 * 1000020/HZ)
 
-unsigned long __init calibrate_tsc(void)
+unsigned long calibrate_tsc(void)
 {
 	mach_prepare_counter();
 
@@ -139,7 +140,7 @@ bad_calibration:
 #endif
 
 /* calculate cpu_khz */
-void __init init_cpu_khz(void)
+void init_cpu_khz(void)
 {
 	if (cpu_has_tsc) {
 		unsigned long tsc_quotient = calibrate_tsc();
@@ -158,3 +159,4 @@ void __init init_cpu_khz(void)
 		}
 	}
 }
+
diff --git a/arch/i386/kernel/timers/timer_tsc.c b/arch/i386/kernel/timers/timer_tsc.c
index 7926d967be0..180444d8782 100644
--- a/arch/i386/kernel/timers/timer_tsc.c
+++ b/arch/i386/kernel/timers/timer_tsc.c
@@ -320,6 +320,26 @@ core_initcall(cpufreq_tsc);
 static inline void cpufreq_delayed_get(void) { return; }
 #endif 
 
+int recalibrate_cpu_khz(void)
+{
+#ifndef CONFIG_SMP
+	unsigned long cpu_khz_old = cpu_khz;
+
+	if (cpu_has_tsc) {
+		init_cpu_khz();
+		cpu_data[0].loops_per_jiffy =
+		    cpufreq_scale(cpu_data[0].loops_per_jiffy,
+			          cpu_khz_old,
+				  cpu_khz);
+		return 0;
+	} else
+		return -ENODEV;
+#else
+	return -ENODEV;
+#endif
+}
+EXPORT_SYMBOL(recalibrate_cpu_khz);
+
 static void mark_offset_tsc(void)
 {
 	unsigned long lost,delay;
diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c
index 247a21c64ae..c1e20d65dd6 100644
--- a/arch/ia64/ia32/sys_ia32.c
+++ b/arch/ia64/ia32/sys_ia32.c
@@ -2427,7 +2427,7 @@ sys32_epoll_wait(int epfd, struct epoll_event32 __user * events, int maxevents,
 {
 	struct epoll_event *events64 = NULL;
 	mm_segment_t old_fs = get_fs();
-	int error, numevents, size;
+	int numevents, size;
 	int evt_idx;
 	int do_free_pages = 0;
 
diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S
index 81c45d44739..d99316c9be2 100644
--- a/arch/ia64/kernel/entry.S
+++ b/arch/ia64/kernel/entry.S
@@ -1182,7 +1182,7 @@ ENTRY(notify_resume_user)
 	;;
 (pNonSys) mov out2=0				// out2==0 => not a syscall
 	.fframe 16
-	.spillpsp ar.unat, 16			// (note that offset is relative to psp+0x10!)
+	.spillsp ar.unat, 16
 	st8 [sp]=r9,-16				// allocate space for ar.unat and save it
 	st8 [out1]=loc1,-8			// save ar.pfs, out1=&sigscratch
 	.body
@@ -1208,7 +1208,7 @@ GLOBAL_ENTRY(sys_rt_sigsuspend)
 	adds out2=8,sp				// out2=&sigscratch->ar_pfs
 	;;
 	.fframe 16
-	.spillpsp ar.unat, 16			// (note that offset is relative to psp+0x10!)
+	.spillsp ar.unat, 16
 	st8 [sp]=r9,-16				// allocate space for ar.unat and save it
 	st8 [out2]=loc1,-8			// save ar.pfs, out2=&sigscratch
 	.body
diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c
index 4d6c7b8f667..736e328b5e6 100644
--- a/arch/ia64/kernel/mca.c
+++ b/arch/ia64/kernel/mca.c
@@ -1103,8 +1103,6 @@ ia64_mca_cpe_int_caller(int cpe_irq, void *arg, struct pt_regs *ptregs)
 	return IRQ_HANDLED;
 }
 
-#endif /* CONFIG_ACPI */
-
 /*
  *  ia64_mca_cpe_poll
  *
@@ -1122,6 +1120,8 @@ ia64_mca_cpe_poll (unsigned long dummy)
 	platform_send_ipi(first_cpu(cpu_online_map), IA64_CPEP_VECTOR, IA64_IPI_DM_INT, 0);
 }
 
+#endif /* CONFIG_ACPI */
+
 /*
  * C portion of the OS INIT handler
  *
@@ -1390,8 +1390,7 @@ ia64_mca_init(void)
 	register_percpu_irq(IA64_MCA_WAKEUP_VECTOR, &mca_wkup_irqaction);
 
 #ifdef CONFIG_ACPI
-	/* Setup the CPEI/P vector and handler */
-	cpe_vector = acpi_request_vector(ACPI_INTERRUPT_CPEI);
+	/* Setup the CPEI/P handler */
 	register_percpu_irq(IA64_CPEP_VECTOR, &mca_cpep_irqaction);
 #endif
 
@@ -1436,6 +1435,7 @@ ia64_mca_late_init(void)
 
 #ifdef CONFIG_ACPI
 	/* Setup the CPEI/P vector and handler */
+	cpe_vector = acpi_request_vector(ACPI_INTERRUPT_CPEI);
 	init_timer(&cpe_poll_timer);
 	cpe_poll_timer.function = ia64_mca_cpe_poll;
 
diff --git a/arch/ia64/kernel/minstate.h b/arch/ia64/kernel/minstate.h
index 1dbc7b2497c..f6d8a010d99 100644
--- a/arch/ia64/kernel/minstate.h
+++ b/arch/ia64/kernel/minstate.h
@@ -41,7 +41,7 @@
 (pKStk) addl r3=THIS_CPU(ia64_mca_data),r3;;							\
 (pKStk) ld8 r3 = [r3];;										\
 (pKStk) addl r3=IA64_MCA_CPU_INIT_STACK_OFFSET,r3;;						\
-(pKStk) addl sp=IA64_STK_OFFSET-IA64_PT_REGS_SIZE,r3;						\
+(pKStk) addl r1=IA64_STK_OFFSET-IA64_PT_REGS_SIZE,r3;						\
 (pUStk)	mov ar.rsc=0;		/* set enforced lazy mode, pl 0, little-endian, loadrs=0 */	\
 (pUStk)	addl r22=IA64_RBS_OFFSET,r1;		/* compute base of register backing store */	\
 	;;											\
@@ -50,7 +50,6 @@
 (pUStk)	mov r23=ar.bspstore;				/* save ar.bspstore */			\
 (pUStk)	dep r22=-1,r22,61,3;			/* compute kernel virtual addr of RBS */	\
 	;;											\
-(pKStk) addl r1=-IA64_PT_REGS_SIZE,r1;		/* if in kernel mode, use sp (r12) */		\
 (pUStk)	mov ar.bspstore=r22;			/* switch to kernel RBS */			\
 	;;											\
 (pUStk)	mov r18=ar.bsp;										\
diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index 71c101601e3..6407bff6bfd 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -11,7 +11,7 @@
  * Version Perfmon-2.x is a rewrite of perfmon-1.x
  * by Stephane Eranian, Hewlett Packard Co.
  *
- * Copyright (C) 1999-2003, 2005  Hewlett Packard Co
+ * Copyright (C) 1999-2005  Hewlett Packard Co
  *               Stephane Eranian <eranian@hpl.hp.com>
  *               David Mosberger-Tang <davidm@hpl.hp.com>
  *
@@ -497,6 +497,9 @@ typedef struct {
 static pfm_stats_t		pfm_stats[NR_CPUS];
 static pfm_session_t		pfm_sessions;	/* global sessions information */
 
+static spinlock_t pfm_alt_install_check = SPIN_LOCK_UNLOCKED;
+static pfm_intr_handler_desc_t  *pfm_alt_intr_handler;
+
 static struct proc_dir_entry 	*perfmon_dir;
 static pfm_uuid_t		pfm_null_uuid = {0,};
 
@@ -606,6 +609,7 @@ DEFINE_PER_CPU(unsigned long, pfm_syst_info);
 DEFINE_PER_CPU(struct task_struct *, pmu_owner);
 DEFINE_PER_CPU(pfm_context_t  *, pmu_ctx);
 DEFINE_PER_CPU(unsigned long, pmu_activation_number);
+EXPORT_PER_CPU_SYMBOL_GPL(pfm_syst_info);
 
 
 /* forward declaration */
@@ -1325,7 +1329,7 @@ pfm_reserve_session(struct task_struct *task, int is_syswide, unsigned int cpu)
 error_conflict:
 	DPRINT(("system wide not possible, conflicting session [%d] on CPU%d\n",
   		pfm_sessions.pfs_sys_session[cpu]->pid,
-		smp_processor_id()));
+		cpu));
 abort:
 	UNLOCK_PFS(flags);
 
@@ -5555,26 +5559,32 @@ pfm_interrupt_handler(int irq, void *arg, struct pt_regs *regs)
 	int ret;
 
 	this_cpu = get_cpu();
-	min      = pfm_stats[this_cpu].pfm_ovfl_intr_cycles_min;
-	max      = pfm_stats[this_cpu].pfm_ovfl_intr_cycles_max;
+	if (likely(!pfm_alt_intr_handler)) {
+		min = pfm_stats[this_cpu].pfm_ovfl_intr_cycles_min;
+		max = pfm_stats[this_cpu].pfm_ovfl_intr_cycles_max;
 
-	start_cycles = ia64_get_itc();
+		start_cycles = ia64_get_itc();
 
-	ret = pfm_do_interrupt_handler(irq, arg, regs);
+		ret = pfm_do_interrupt_handler(irq, arg, regs);
 
-	total_cycles = ia64_get_itc();
+		total_cycles = ia64_get_itc();
 
-	/*
-	 * don't measure spurious interrupts
-	 */
-	if (likely(ret == 0)) {
-		total_cycles -= start_cycles;
+		/*
+		 * don't measure spurious interrupts
+		 */
+		if (likely(ret == 0)) {
+			total_cycles -= start_cycles;
 
-		if (total_cycles < min) pfm_stats[this_cpu].pfm_ovfl_intr_cycles_min = total_cycles;
-		if (total_cycles > max) pfm_stats[this_cpu].pfm_ovfl_intr_cycles_max = total_cycles;
+			if (total_cycles < min) pfm_stats[this_cpu].pfm_ovfl_intr_cycles_min = total_cycles;
+			if (total_cycles > max) pfm_stats[this_cpu].pfm_ovfl_intr_cycles_max = total_cycles;
 
-		pfm_stats[this_cpu].pfm_ovfl_intr_cycles += total_cycles;
+			pfm_stats[this_cpu].pfm_ovfl_intr_cycles += total_cycles;
+		}
+	}
+	else {
+		(*pfm_alt_intr_handler->handler)(irq, arg, regs);
 	}
+
 	put_cpu_no_resched();
 	return IRQ_HANDLED;
 }
@@ -6425,6 +6435,141 @@ static struct irqaction perfmon_irqaction = {
 	.name    = "perfmon"
 };
 
+static void
+pfm_alt_save_pmu_state(void *data)
+{
+	struct pt_regs *regs;
+
+	regs = ia64_task_regs(current);
+
+	DPRINT(("called\n"));
+
+	/*
+	 * should not be necessary but
+	 * let's take not risk
+	 */
+	pfm_clear_psr_up();
+	pfm_clear_psr_pp();
+	ia64_psr(regs)->pp = 0;
+
+	/*
+	 * This call is required
+	 * May cause a spurious interrupt on some processors
+	 */
+	pfm_freeze_pmu();
+
+	ia64_srlz_d();
+}
+
+void
+pfm_alt_restore_pmu_state(void *data)
+{
+	struct pt_regs *regs;
+
+	regs = ia64_task_regs(current);
+
+	DPRINT(("called\n"));
+
+	/*
+	 * put PMU back in state expected
+	 * by perfmon
+	 */
+	pfm_clear_psr_up();
+	pfm_clear_psr_pp();
+	ia64_psr(regs)->pp = 0;
+
+	/*
+	 * perfmon runs with PMU unfrozen at all times
+	 */
+	pfm_unfreeze_pmu();
+
+	ia64_srlz_d();
+}
+
+int
+pfm_install_alt_pmu_interrupt(pfm_intr_handler_desc_t *hdl)
+{
+	int ret, i;
+	int reserve_cpu;
+
+	/* some sanity checks */
+	if (hdl == NULL || hdl->handler == NULL) return -EINVAL;
+
+	/* do the easy test first */
+	if (pfm_alt_intr_handler) return -EBUSY;
+
+	/* one at a time in the install or remove, just fail the others */
+	if (!spin_trylock(&pfm_alt_install_check)) {
+		return -EBUSY;
+	}
+
+	/* reserve our session */
+	for_each_online_cpu(reserve_cpu) {
+		ret = pfm_reserve_session(NULL, 1, reserve_cpu);
+		if (ret) goto cleanup_reserve;
+	}
+
+	/* save the current system wide pmu states */
+	ret = on_each_cpu(pfm_alt_save_pmu_state, NULL, 0, 1);
+	if (ret) {
+		DPRINT(("on_each_cpu() failed: %d\n", ret));
+		goto cleanup_reserve;
+	}
+
+	/* officially change to the alternate interrupt handler */
+	pfm_alt_intr_handler = hdl;
+
+	spin_unlock(&pfm_alt_install_check);
+
+	return 0;
+
+cleanup_reserve:
+	for_each_online_cpu(i) {
+		/* don't unreserve more than we reserved */
+		if (i >= reserve_cpu) break;
+
+		pfm_unreserve_session(NULL, 1, i);
+	}
+
+	spin_unlock(&pfm_alt_install_check);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(pfm_install_alt_pmu_interrupt);
+
+int
+pfm_remove_alt_pmu_interrupt(pfm_intr_handler_desc_t *hdl)
+{
+	int i;
+	int ret;
+
+	if (hdl == NULL) return -EINVAL;
+
+	/* cannot remove someone else's handler! */
+	if (pfm_alt_intr_handler != hdl) return -EINVAL;
+
+	/* one at a time in the install or remove, just fail the others */
+	if (!spin_trylock(&pfm_alt_install_check)) {
+		return -EBUSY;
+	}
+
+	pfm_alt_intr_handler = NULL;
+
+	ret = on_each_cpu(pfm_alt_restore_pmu_state, NULL, 0, 1);
+	if (ret) {
+		DPRINT(("on_each_cpu() failed: %d\n", ret));
+	}
+
+	for_each_online_cpu(i) {
+		pfm_unreserve_session(NULL, 1, i);
+	}
+
+	spin_unlock(&pfm_alt_install_check);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(pfm_remove_alt_pmu_interrupt);
+
 /*
  * perfmon initialization routine, called from the initcall() table
  */
diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c
index 907464ee727..08c8a5eb25a 100644
--- a/arch/ia64/kernel/ptrace.c
+++ b/arch/ia64/kernel/ptrace.c
@@ -692,16 +692,30 @@ convert_to_non_syscall (struct task_struct *child, struct pt_regs  *pt,
 			unsigned long cfm)
 {
 	struct unw_frame_info info, prev_info;
-	unsigned long ip, pr;
+	unsigned long ip, sp, pr;
 
 	unw_init_from_blocked_task(&info, child);
 	while (1) {
 		prev_info = info;
 		if (unw_unwind(&info) < 0)
 			return;
-		if (unw_get_rp(&info, &ip) < 0)
+
+		unw_get_sp(&info, &sp);
+		if ((long)((unsigned long)child + IA64_STK_OFFSET - sp)
+		    < IA64_PT_REGS_SIZE) {
+			dprintk("ptrace.%s: ran off the top of the kernel "
+				"stack\n", __FUNCTION__);
+			return;
+		}
+		if (unw_get_pr (&prev_info, &pr) < 0) {
+			unw_get_rp(&prev_info, &ip);
+			dprintk("ptrace.%s: failed to read "
+				"predicate register (ip=0x%lx)\n",
+				__FUNCTION__, ip);
 			return;
-		if (ip < FIXADDR_USER_END)
+		}
+		if (unw_is_intr_frame(&info)
+		    && (pr & (1UL << PRED_USER_STACK)))
 			break;
 	}
 
diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c
index 0d5ee57c986..3865f088ffa 100644
--- a/arch/ia64/kernel/smpboot.c
+++ b/arch/ia64/kernel/smpboot.c
@@ -624,7 +624,7 @@ static struct {
 	__u16   thread_id;
 	__u16   proc_fixed_addr;
 	__u8    valid;
-}mt_info[NR_CPUS] __devinit;
+} mt_info[NR_CPUS] __devinitdata;
 
 #ifdef CONFIG_HOTPLUG_CPU
 static inline void
diff --git a/arch/ia64/kernel/sys_ia64.c b/arch/ia64/kernel/sys_ia64.c
index a8cf6d8a509..770fab37928 100644
--- a/arch/ia64/kernel/sys_ia64.c
+++ b/arch/ia64/kernel/sys_ia64.c
@@ -182,13 +182,6 @@ do_mmap2 (unsigned long addr, unsigned long len, int prot, int flags, int fd, un
 		}
 	}
 
-	/*
-	 * A zero mmap always succeeds in Linux, independent of whether or not the
-	 * remaining arguments are valid.
-	 */
-	if (len == 0)
-		goto out;
-
 	/* Careful about overflows.. */
 	len = PAGE_ALIGN(len);
 	if (!len || len > TASK_SIZE) {
diff --git a/arch/ia64/sn/kernel/setup.c b/arch/ia64/sn/kernel/setup.c
index 4fb44984afe..e64cb8175f7 100644
--- a/arch/ia64/sn/kernel/setup.c
+++ b/arch/ia64/sn/kernel/setup.c
@@ -271,6 +271,8 @@ void __init sn_setup(char **cmdline_p)
 	int major = sn_sal_rev_major(), minor = sn_sal_rev_minor();
 	extern void sn_cpu_init(void);
 
+	ia64_sn_plat_set_error_handling_features();
+
 	/*
 	 * If the generic code has enabled vga console support - lets
 	 * get rid of it again. This is a kludge for the fact that ACPI
diff --git a/arch/m68knommu/kernel/process.c b/arch/m68knommu/kernel/process.c
index 2b6c9d32b7a..c4a33f265dc 100644
--- a/arch/m68knommu/kernel/process.c
+++ b/arch/m68knommu/kernel/process.c
@@ -45,11 +45,13 @@ asmlinkage void ret_from_fork(void);
  */
 void default_idle(void)
 {
-	while(1) {
-		if (need_resched())
-			__asm__("stop #0x2000" : : : "cc");
-		schedule();
+	local_irq_disable();
+ 	while (!need_resched()) {
+		/* This stop will re-enable interrupts */
+ 		__asm__("stop #0x2000" : : : "cc");
+		local_irq_disable();
 	}
+	local_irq_enable();
 }
 
 void (*idle)(void) = default_idle;
@@ -63,7 +65,12 @@ void (*idle)(void) = default_idle;
 void cpu_idle(void)
 {
 	/* endless idle loop with no priority at all */
-	idle();
+	while (1) {
+		idle();
+		preempt_enable_no_resched();
+		schedule();
+		preempt_disable();
+	}
 }
 
 void machine_restart(char * __unused)
diff --git a/arch/ppc/kernel/cputable.c b/arch/ppc/kernel/cputable.c
index 8aa5e8c6900..d44b7dc5390 100644
--- a/arch/ppc/kernel/cputable.c
+++ b/arch/ppc/kernel/cputable.c
@@ -838,6 +838,17 @@ struct cpu_spec	cpu_specs[] = {
 		.icache_bsize		= 32,
 		.dcache_bsize		= 32,
 	},
+	{	/* 405EP */
+		.pvr_mask		= 0xffff0000,
+		.pvr_value		= 0x51210000,
+		.cpu_name		= "405EP",
+		.cpu_features		= CPU_FTR_SPLIT_ID_CACHE |
+			CPU_FTR_USE_TB,
+		.cpu_user_features	= PPC_FEATURE_32 |
+			PPC_FEATURE_HAS_MMU | PPC_FEATURE_HAS_4xxMAC,
+		.icache_bsize		= 32,
+		.dcache_bsize		= 32,
+	},
 
 #endif /* CONFIG_40x */
 #ifdef CONFIG_44x
diff --git a/arch/ppc/kernel/misc.S b/arch/ppc/kernel/misc.S
index e4f1615ec13..7329ef177a1 100644
--- a/arch/ppc/kernel/misc.S
+++ b/arch/ppc/kernel/misc.S
@@ -619,7 +619,7 @@ _GLOBAL(flush_instruction_cache)
 _GLOBAL(flush_icache_range)
 BEGIN_FTR_SECTION
 	blr				/* for 601, do nothing */
-END_FTR_SECTION_IFSET(PPC_FEATURE_UNIFIED_CACHE)
+END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE)
 	li	r5,L1_CACHE_LINE_SIZE-1
 	andc	r3,r3,r5
 	subf	r4,r3,r4
@@ -736,7 +736,7 @@ _GLOBAL(flush_dcache_all)
 _GLOBAL(__flush_dcache_icache)
 BEGIN_FTR_SECTION
 	blr					/* for 601, do nothing */
-END_FTR_SECTION_IFSET(PPC_FEATURE_UNIFIED_CACHE)
+END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE)
 	rlwinm	r3,r3,0,0,19			/* Get page base address */
 	li	r4,4096/L1_CACHE_LINE_SIZE	/* Number of lines in a page */
 	mtctr	r4
@@ -764,7 +764,7 @@ END_FTR_SECTION_IFSET(PPC_FEATURE_UNIFIED_CACHE)
 _GLOBAL(__flush_dcache_icache_phys)
 BEGIN_FTR_SECTION
 	blr					/* for 601, do nothing */
-END_FTR_SECTION_IFSET(PPC_FEATURE_UNIFIED_CACHE)
+END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE)
 	mfmsr	r10
 	rlwinm	r0,r10,0,28,26			/* clear DR */
 	mtmsr	r0
diff --git a/arch/ppc/syslib/prom_init.c b/arch/ppc/syslib/prom_init.c
index 2cee87137f2..7f15136830f 100644
--- a/arch/ppc/syslib/prom_init.c
+++ b/arch/ppc/syslib/prom_init.c
@@ -626,8 +626,18 @@ inspect_node(phandle node, struct device_node *dad,
 	l = call_prom("package-to-path", 3, 1, node,
 		      mem_start, mem_end - mem_start);
 	if (l >= 0) {
+		char *p, *ep;
+
 		np->full_name = PTRUNRELOC((char *) mem_start);
 		*(char *)(mem_start + l) = 0;
+		/* Fixup an Apple bug where they have bogus \0 chars in the
+		 * middle of the path in some properties
+		 */
+		for (p = (char *)mem_start, ep = p + l; p < ep; p++)
+			if ((*p) == '\0') {
+				memmove(p, p+1, ep - p);
+				ep--;
+			}
 		mem_start = ALIGNUL(mem_start + l + 1);
 	}
 
diff --git a/arch/ppc64/kernel/entry.S b/arch/ppc64/kernel/entry.S
index d3604056e1a..b61572eb2a7 100644
--- a/arch/ppc64/kernel/entry.S
+++ b/arch/ppc64/kernel/entry.S
@@ -436,15 +436,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 	REST_8GPRS(14, r1)
 	REST_10GPRS(22, r1)
 
-#ifdef CONFIG_PPC_ISERIES
-	clrrdi	r7,r1,THREAD_SHIFT	/* get current_thread_info() */
-	ld	r7,TI_FLAGS(r7)		/* Get run light flag */
-	mfspr	r9,CTRLF
-	srdi	r7,r7,TIF_RUN_LIGHT
-	insrdi	r9,r7,1,63		/* Insert run light into CTRL */
-	mtspr	CTRLT,r9
-#endif
-
 	/* convert old thread to its task_struct for return value */
 	addi	r3,r3,-THREAD
 	ld	r7,_NIP(r1)	/* Return to _switch caller in new task */
diff --git a/arch/ppc64/kernel/head.S b/arch/ppc64/kernel/head.S
index 92a744c31ab..346dbf606b5 100644
--- a/arch/ppc64/kernel/head.S
+++ b/arch/ppc64/kernel/head.S
@@ -626,10 +626,10 @@ system_reset_iSeries:
 	lhz	r24,PACAPACAINDEX(r13)	/* Get processor # */
 	cmpwi	0,r24,0			/* Are we processor 0? */
 	beq	.__start_initialization_iSeries	/* Start up the first processor */
-	mfspr	r4,CTRLF
-	li	r5,RUNLATCH		/* Turn off the run light */
+	mfspr	r4,SPRN_CTRLF
+	li	r5,CTRL_RUNLATCH	/* Turn off the run light */
 	andc	r4,r4,r5
-	mtspr	CTRLT,r4
+	mtspr	SPRN_CTRLT,r4
 
 1:
 	HMT_LOW
@@ -2082,9 +2082,9 @@ _GLOBAL(hmt_start_secondary)
 	mfspr	r4, HID0
 	ori	r4, r4, 0x1
 	mtspr	HID0, r4
-	mfspr	r4, CTRLF
+	mfspr	r4, SPRN_CTRLF
 	oris	r4, r4, 0x40
-	mtspr	CTRLT, r4
+	mtspr	SPRN_CTRLT, r4
 	blr
 #endif
 
diff --git a/arch/ppc64/kernel/iSeries_setup.c b/arch/ppc64/kernel/iSeries_setup.c
index da20120f226..6d06eb550a3 100644
--- a/arch/ppc64/kernel/iSeries_setup.c
+++ b/arch/ppc64/kernel/iSeries_setup.c
@@ -852,6 +852,28 @@ static int __init iSeries_src_init(void)
 
 late_initcall(iSeries_src_init);
 
+static int set_spread_lpevents(char *str)
+{
+	unsigned long i;
+	unsigned long val = simple_strtoul(str, NULL, 0);
+
+	/*
+	 * The parameter is the number of processors to share in processing
+	 * lp events.
+	 */
+	if (( val > 0) && (val <= NR_CPUS)) {
+		for (i = 1; i < val; ++i)
+			paca[i].lpqueue_ptr = paca[0].lpqueue_ptr;
+
+		printk("lpevent processing spread over %ld processors\n", val);
+	} else {
+		printk("invalid spread_lpevents %ld\n", val);
+	}
+
+	return 1;
+}
+__setup("spread_lpevents=", set_spread_lpevents);
+
 void __init iSeries_early_setup(void)
 {
 	iSeries_fixup_klimit();
diff --git a/arch/ppc64/kernel/idle.c b/arch/ppc64/kernel/idle.c
index 6abc621d3ba..f24ce2b8720 100644
--- a/arch/ppc64/kernel/idle.c
+++ b/arch/ppc64/kernel/idle.c
@@ -75,13 +75,9 @@ static int iSeries_idle(void)
 {
 	struct paca_struct *lpaca;
 	long oldval;
-	unsigned long CTRL;
 
 	/* ensure iSeries run light will be out when idle */
-	clear_thread_flag(TIF_RUN_LIGHT);
-	CTRL = mfspr(CTRLF);
-	CTRL &= ~RUNLATCH;
-	mtspr(CTRLT, CTRL);
+	ppc64_runlatch_off();
 
 	lpaca = get_paca();
 
@@ -111,7 +107,9 @@ static int iSeries_idle(void)
 			}
 		}
 
+		ppc64_runlatch_on();
 		schedule();
+		ppc64_runlatch_off();
 	}
 
 	return 0;
diff --git a/arch/ppc64/kernel/pSeries_reconfig.c b/arch/ppc64/kernel/pSeries_reconfig.c
index cb5443f2e49..dc2a69d412a 100644
--- a/arch/ppc64/kernel/pSeries_reconfig.c
+++ b/arch/ppc64/kernel/pSeries_reconfig.c
@@ -47,14 +47,6 @@ static void remove_node_proc_entries(struct device_node *np)
 		remove_proc_entry(pp->name, np->pde);
 		pp = pp->next;
 	}
-
-	/* Assuming that symlinks have the same parent directory as
-	 * np->pde.
-	 */
-	if (np->name_link)
-		remove_proc_entry(np->name_link->name, parent->pde);
-	if (np->addr_link)
-		remove_proc_entry(np->addr_link->name, parent->pde);
 	if (np->pde)
 		remove_proc_entry(np->pde->name, parent->pde);
 }
diff --git a/arch/ppc64/kernel/process.c b/arch/ppc64/kernel/process.c
index 8b068612273..cdfecbeb331 100644
--- a/arch/ppc64/kernel/process.c
+++ b/arch/ppc64/kernel/process.c
@@ -378,9 +378,6 @@ copy_thread(int nr, unsigned long clone_flags, unsigned long usp,
 		childregs->gpr[1] = sp + sizeof(struct pt_regs);
 		p->thread.regs = NULL;	/* no user register state */
 		clear_ti_thread_flag(p->thread_info, TIF_32BIT);
-#ifdef CONFIG_PPC_ISERIES
-		set_ti_thread_flag(p->thread_info, TIF_RUN_LIGHT);
-#endif
 	} else {
 		childregs->gpr[1] = usp;
 		p->thread.regs = childregs;
diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c
index 6f79b7b9b44..b7683abfbe6 100644
--- a/arch/ppc64/kernel/prom_init.c
+++ b/arch/ppc64/kernel/prom_init.c
@@ -211,13 +211,23 @@ struct {
  */
 #define ADDR(x)		(u32) ((unsigned long)(x) - offset)
 
+/*
+ * Error results ... some OF calls will return "-1" on error, some
+ * will return 0, some will return either. To simplify, here are
+ * macros to use with any ihandle or phandle return value to check if
+ * it is valid
+ */
+
+#define PROM_ERROR		(-1u)
+#define PHANDLE_VALID(p)	((p) != 0 && (p) != PROM_ERROR)
+#define IHANDLE_VALID(i)	((i) != 0 && (i) != PROM_ERROR)
+
+
 /* This is the one and *ONLY* place where we actually call open
  * firmware from, since we need to make sure we're running in 32b
  * mode when we do.  We switch back to 64b mode upon return.
  */
 
-#define PROM_ERROR	(-1)
-
 static int __init call_prom(const char *service, int nargs, int nret, ...)
 {
 	int i;
@@ -587,14 +597,13 @@ static void __init prom_send_capabilities(void)
 {
 	unsigned long offset = reloc_offset();
 	ihandle elfloader;
-	int ret;
 
 	elfloader = call_prom("open", 1, 1, ADDR("/packages/elf-loader"));
 	if (elfloader == 0) {
 		prom_printf("couldn't open /packages/elf-loader\n");
 		return;
 	}
-	ret = call_prom("call-method", 3, 1, ADDR("process-elf-header"),
+	call_prom("call-method", 3, 1, ADDR("process-elf-header"),
 			elfloader, ADDR(&fake_elf));
 	call_prom("close", 1, 0, elfloader);
 }
@@ -646,7 +655,7 @@ static unsigned long __init alloc_up(unsigned long size, unsigned long align)
 	    base = _ALIGN_UP(base + 0x100000, align)) {
 		prom_debug("    trying: 0x%x\n\r", base);
 		addr = (unsigned long)prom_claim(base, size, 0);
-		if ((int)addr != PROM_ERROR)
+		if (addr != PROM_ERROR)
 			break;
 		addr = 0;
 		if (align == 0)
@@ -708,7 +717,7 @@ static unsigned long __init alloc_down(unsigned long size, unsigned long align,
 	for(; base > RELOC(alloc_bottom); base = _ALIGN_DOWN(base - 0x100000, align))  {
 		prom_debug("    trying: 0x%x\n\r", base);
 		addr = (unsigned long)prom_claim(base, size, 0);
-		if ((int)addr != PROM_ERROR)
+		if (addr != PROM_ERROR)
 			break;
 		addr = 0;
 	}
@@ -902,18 +911,19 @@ static void __init prom_instantiate_rtas(void)
 {
 	unsigned long offset = reloc_offset();
 	struct prom_t *_prom = PTRRELOC(&prom);
-	phandle prom_rtas, rtas_node;
+	phandle rtas_node;
+	ihandle rtas_inst;
 	u32 base, entry = 0;
 	u32 size = 0;
 
 	prom_debug("prom_instantiate_rtas: start...\n");
 
-	prom_rtas = call_prom("finddevice", 1, 1, ADDR("/rtas"));
-	prom_debug("prom_rtas: %x\n", prom_rtas);
-	if (prom_rtas == (phandle) -1)
+	rtas_node = call_prom("finddevice", 1, 1, ADDR("/rtas"));
+	prom_debug("rtas_node: %x\n", rtas_node);
+	if (!PHANDLE_VALID(rtas_node))
 		return;
 
-	prom_getprop(prom_rtas, "rtas-size", &size, sizeof(size));
+	prom_getprop(rtas_node, "rtas-size", &size, sizeof(size));
 	if (size == 0)
 		return;
 
@@ -922,14 +932,18 @@ static void __init prom_instantiate_rtas(void)
 		prom_printf("RTAS allocation failed !\n");
 		return;
 	}
-	prom_printf("instantiating rtas at 0x%x", base);
 
-	rtas_node = call_prom("open", 1, 1, ADDR("/rtas"));
-	prom_printf("...");
+	rtas_inst = call_prom("open", 1, 1, ADDR("/rtas"));
+	if (!IHANDLE_VALID(rtas_inst)) {
+		prom_printf("opening rtas package failed");
+		return;
+	}
+
+	prom_printf("instantiating rtas at 0x%x ...", base);
 
 	if (call_prom("call-method", 3, 2,
 		      ADDR("instantiate-rtas"),
-		      rtas_node, base) != PROM_ERROR) {
+		      rtas_inst, base) != PROM_ERROR) {
 		entry = (long)_prom->args.rets[1];
 	}
 	if (entry == 0) {
@@ -940,8 +954,8 @@ static void __init prom_instantiate_rtas(void)
 
 	reserve_mem(base, size);
 
-	prom_setprop(prom_rtas, "linux,rtas-base", &base, sizeof(base));
-	prom_setprop(prom_rtas, "linux,rtas-entry", &entry, sizeof(entry));
+	prom_setprop(rtas_node, "linux,rtas-base", &base, sizeof(base));
+	prom_setprop(rtas_node, "linux,rtas-entry", &entry, sizeof(entry));
 
 	prom_debug("rtas base     = 0x%x\n", base);
 	prom_debug("rtas entry    = 0x%x\n", entry);
@@ -1062,7 +1076,7 @@ static void __init prom_initialize_tce_table(void)
 
 		prom_printf("opening PHB %s", path);
 		phb_node = call_prom("open", 1, 1, path);
-		if ( (long)phb_node <= 0)
+		if (phb_node == 0)
 			prom_printf("... failed\n");
 		else
 			prom_printf("... done\n");
@@ -1279,12 +1293,12 @@ static void __init prom_init_client_services(unsigned long pp)
 
 	/* get a handle for the stdout device */
 	_prom->chosen = call_prom("finddevice", 1, 1, ADDR("/chosen"));
-	if ((long)_prom->chosen <= 0)
+	if (!PHANDLE_VALID(_prom->chosen))
 		prom_panic("cannot find chosen"); /* msg won't be printed :( */
 
 	/* get device tree root */
 	_prom->root = call_prom("finddevice", 1, 1, ADDR("/"));
-	if ((long)_prom->root <= 0)
+	if (!PHANDLE_VALID(_prom->root))
 		prom_panic("cannot find device tree root"); /* msg won't be printed :( */
 }
 
@@ -1356,9 +1370,8 @@ static int __init prom_find_machine_type(void)
 	}
 	/* Default to pSeries. We need to know if we are running LPAR */
 	rtas = call_prom("finddevice", 1, 1, ADDR("/rtas"));
-	if (rtas != (phandle) -1) {
-		unsigned long x;
-		x = prom_getproplen(rtas, "ibm,hypertas-functions");
+	if (PHANDLE_VALID(rtas)) {
+		int x = prom_getproplen(rtas, "ibm,hypertas-functions");
 		if (x != PROM_ERROR) {
 			prom_printf("Hypertas detected, assuming LPAR !\n");
 			return PLATFORM_PSERIES_LPAR;
@@ -1426,12 +1439,13 @@ static void __init prom_check_displays(void)
 		 * leave some room at the end of the path for appending extra
 		 * arguments
 		 */
-		if (call_prom("package-to-path", 3, 1, node, path, PROM_SCRATCH_SIZE-10) < 0)
+		if (call_prom("package-to-path", 3, 1, node, path,
+			      PROM_SCRATCH_SIZE-10) == PROM_ERROR)
 			continue;
 		prom_printf("found display   : %s, opening ... ", path);
 		
 		ih = call_prom("open", 1, 1, path);
-		if (ih == (ihandle)0 || ih == (ihandle)-1) {
+		if (ih == 0) {
 			prom_printf("failed\n");
 			continue;
 		}
@@ -1514,6 +1528,12 @@ static unsigned long __init dt_find_string(char *str)
 	return 0;
 }
 
+/*
+ * The Open Firmware 1275 specification states properties must be 31 bytes or
+ * less, however not all firmwares obey this. Make it 64 bytes to be safe.
+ */
+#define MAX_PROPERTY_NAME 64
+
 static void __init scan_dt_build_strings(phandle node, unsigned long *mem_start,
 					 unsigned long *mem_end)
 {
@@ -1527,10 +1547,12 @@ static void __init scan_dt_build_strings(phandle node, unsigned long *mem_start,
 	/* get and store all property names */
 	prev_name = RELOC("");
 	for (;;) {
-		
-		/* 32 is max len of name including nul. */
-		namep = make_room(mem_start, mem_end, 32, 1);
-		if (call_prom("nextprop", 3, 1, node, prev_name, namep) <= 0) {
+		int rc;
+
+		/* 64 is max len of name including nul. */
+		namep = make_room(mem_start, mem_end, MAX_PROPERTY_NAME, 1);
+		rc = call_prom("nextprop", 3, 1, node, prev_name, namep);
+		if (rc != 1) {
 			/* No more nodes: unwind alloc */
 			*mem_start = (unsigned long)namep;
 			break;
@@ -1555,18 +1577,12 @@ static void __init scan_dt_build_strings(phandle node, unsigned long *mem_start,
 	}
 }
 
-/*
- * The Open Firmware 1275 specification states properties must be 31 bytes or
- * less, however not all firmwares obey this. Make it 64 bytes to be safe.
- */
-#define MAX_PROPERTY_NAME 64
-
 static void __init scan_dt_build_struct(phandle node, unsigned long *mem_start,
 					unsigned long *mem_end)
 {
 	int l, align;
 	phandle child;
-	char *namep, *prev_name, *sstart;
+	char *namep, *prev_name, *sstart, *p, *ep;
 	unsigned long soff;
 	unsigned char *valp;
 	unsigned long offset = reloc_offset();
@@ -1588,6 +1604,14 @@ static void __init scan_dt_build_struct(phandle node, unsigned long *mem_start,
 			call_prom("package-to-path", 3, 1, node, namep, l);
 		}
 		namep[l] = '\0';
+		/* Fixup an Apple bug where they have bogus \0 chars in the
+		 * middle of the path in some properties
+		 */
+		for (p = namep, ep = namep + l; p < ep; p++)
+			if (*p == '\0') {
+				memmove(p, p+1, ep - p);
+				ep--; l--;
+			}
 		*mem_start = _ALIGN(((unsigned long) namep) + strlen(namep) + 1, 4);
 	}
 
@@ -1599,7 +1623,10 @@ static void __init scan_dt_build_struct(phandle node, unsigned long *mem_start,
 	prev_name = RELOC("");
 	sstart = (char *)RELOC(dt_string_start);
 	for (;;) {
-		if (call_prom("nextprop", 3, 1, node, prev_name, pname) <= 0)
+		int rc;
+
+		rc = call_prom("nextprop", 3, 1, node, prev_name, pname);
+		if (rc != 1)
 			break;
 
 		/* find string offset */
@@ -1615,7 +1642,7 @@ static void __init scan_dt_build_struct(phandle node, unsigned long *mem_start,
 		l = call_prom("getproplen", 2, 1, node, pname);
 
 		/* sanity checks */
-		if (l < 0)
+		if (l == PROM_ERROR)
 			continue;
 		if (l > MAX_PROPERTY_LENGTH) {
 			prom_printf("WARNING: ignoring large property ");
@@ -1763,17 +1790,18 @@ static void __init fixup_device_tree(void)
 
 	/* Some G5s have a missing interrupt definition, fix it up here */
 	u3 = call_prom("finddevice", 1, 1, ADDR("/u3@0,f8000000"));
-	if ((long)u3 <= 0)
+	if (!PHANDLE_VALID(u3))
 		return;
 	i2c = call_prom("finddevice", 1, 1, ADDR("/u3@0,f8000000/i2c@f8001000"));
-	if ((long)i2c <= 0)
+	if (!PHANDLE_VALID(i2c))
 		return;
 	mpic = call_prom("finddevice", 1, 1, ADDR("/u3@0,f8000000/mpic@f8040000"));
-	if ((long)mpic <= 0)
+	if (!PHANDLE_VALID(mpic))
 		return;
 
 	/* check if proper rev of u3 */
-	if (prom_getprop(u3, "device-rev", &u3_rev, sizeof(u3_rev)) <= 0)
+	if (prom_getprop(u3, "device-rev", &u3_rev, sizeof(u3_rev))
+	    == PROM_ERROR)
 		return;
 	if (u3_rev != 0x35)
 		return;
@@ -1881,6 +1909,12 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, unsigned long
 		     &getprop_rval, sizeof(getprop_rval));
 
 	/*
+	 * On pSeries, inform the firmware about our capabilities
+	 */
+	if (RELOC(of_platform) & PLATFORM_PSERIES)
+		prom_send_capabilities();
+
+	/*
 	 * On pSeries, copy the CPU hold code
 	 */
        	if (RELOC(of_platform) & PLATFORM_PSERIES)
diff --git a/arch/ppc64/kernel/setup.c b/arch/ppc64/kernel/setup.c
index 21c57f539c2..dce198d3932 100644
--- a/arch/ppc64/kernel/setup.c
+++ b/arch/ppc64/kernel/setup.c
@@ -103,11 +103,6 @@ extern void unflatten_device_tree(void);
 
 extern void smp_release_cpus(void);
 
-unsigned long decr_overclock = 1;
-unsigned long decr_overclock_proc0 = 1;
-unsigned long decr_overclock_set = 0;
-unsigned long decr_overclock_proc0_set = 0;
-
 int have_of = 1;
 int boot_cpuid = 0;
 int boot_cpuid_phys = 0;
@@ -1120,64 +1115,15 @@ void ppc64_dump_msg(unsigned int src, const char *msg)
 	printk("[dump]%04x %s\n", src, msg);
 }
 
-int set_spread_lpevents( char * str )
-{
-	/* The parameter is the number of processors to share in processing lp events */
-	unsigned long i;
-	unsigned long val = simple_strtoul( str, NULL, 0 );
-	if ( ( val > 0 ) && ( val <= NR_CPUS ) ) {
-		for ( i=1; i<val; ++i )
-			paca[i].lpqueue_ptr = paca[0].lpqueue_ptr;
-		printk("lpevent processing spread over %ld processors\n", val);
-	}
-	else
-		printk("invalid spreaqd_lpevents %ld\n", val);
-	return 1;
-}	
-
 /* This should only be called on processor 0 during calibrate decr */
 void setup_default_decr(void)
 {
 	struct paca_struct *lpaca = get_paca();
 
-	if ( decr_overclock_set && !decr_overclock_proc0_set )
-		decr_overclock_proc0 = decr_overclock;
-
-	lpaca->default_decr = tb_ticks_per_jiffy / decr_overclock_proc0;	
+	lpaca->default_decr = tb_ticks_per_jiffy;
 	lpaca->next_jiffy_update_tb = get_tb() + tb_ticks_per_jiffy;
 }
 
-int set_decr_overclock_proc0( char * str )
-{
-	unsigned long val = simple_strtoul( str, NULL, 0 );
-	if ( ( val >= 1 ) && ( val <= 48 ) ) {
-		decr_overclock_proc0_set = 1;
-		decr_overclock_proc0 = val;
-		printk("proc 0 decrementer overclock factor of %ld\n", val);
-	}
-	else
-		printk("invalid proc 0 decrementer overclock factor of %ld\n", val);
-	return 1;
-}
-
-int set_decr_overclock( char * str )
-{
-	unsigned long val = simple_strtoul( str, NULL, 0 );
-	if ( ( val >= 1 ) && ( val <= 48 ) ) {
-		decr_overclock_set = 1;
-		decr_overclock = val;
-		printk("decrementer overclock factor of %ld\n", val);
-	}
-	else
-		printk("invalid decrementer overclock factor of %ld\n", val);
-	return 1;
-
-}
-
-__setup("spread_lpevents=", set_spread_lpevents );
-__setup("decr_overclock_proc0=", set_decr_overclock_proc0 );
-__setup("decr_overclock=", set_decr_overclock );
-
 #ifndef CONFIG_PPC_ISERIES
 /*
  * This function can be used by platforms to "find" legacy serial ports.
diff --git a/arch/ppc64/kernel/smp.c b/arch/ppc64/kernel/smp.c
index 3b906cd9403..9ef5d36d6b2 100644
--- a/arch/ppc64/kernel/smp.c
+++ b/arch/ppc64/kernel/smp.c
@@ -334,7 +334,6 @@ void smp_call_function_interrupt(void)
 	}
 }
 
-extern unsigned long decr_overclock;
 extern struct gettimeofday_struct do_gtod;
 
 struct thread_info *current_set[NR_CPUS];
@@ -491,7 +490,7 @@ int __devinit __cpu_up(unsigned int cpu)
 	if (smp_ops->cpu_bootable && !smp_ops->cpu_bootable(cpu))
 		return -EINVAL;
 
-	paca[cpu].default_decr = tb_ticks_per_jiffy / decr_overclock;
+	paca[cpu].default_decr = tb_ticks_per_jiffy;
 
 	if (!cpu_has_feature(CPU_FTR_SLB)) {
 		void *tmp;
diff --git a/arch/ppc64/kernel/sysfs.c b/arch/ppc64/kernel/sysfs.c
index 0925694c3ce..c8fa6569b2f 100644
--- a/arch/ppc64/kernel/sysfs.c
+++ b/arch/ppc64/kernel/sysfs.c
@@ -113,7 +113,6 @@ void ppc64_enable_pmcs(void)
 #ifdef CONFIG_PPC_PSERIES
 	unsigned long set, reset;
 	int ret;
-	unsigned int ctrl;
 #endif /* CONFIG_PPC_PSERIES */
 
 	/* Only need to enable them once */
@@ -167,11 +166,8 @@ void ppc64_enable_pmcs(void)
 	 * On SMT machines we have to set the run latch in the ctrl register
 	 * in order to make PMC6 spin.
 	 */
-	if (cpu_has_feature(CPU_FTR_SMT)) {
-		ctrl = mfspr(CTRLF);
-		ctrl |= RUNLATCH;
-		mtspr(CTRLT, ctrl);
-	}
+	if (cpu_has_feature(CPU_FTR_SMT))
+		ppc64_runlatch_on();
 #endif /* CONFIG_PPC_PSERIES */
 }
 
diff --git a/arch/ppc64/kernel/time.c b/arch/ppc64/kernel/time.c
index 3d54745108c..33364a7d2cd 100644
--- a/arch/ppc64/kernel/time.c
+++ b/arch/ppc64/kernel/time.c
@@ -325,9 +325,7 @@ int timer_interrupt(struct pt_regs * regs)
 
 	irq_enter();
 
-#ifndef CONFIG_PPC_ISERIES
 	profile_tick(CPU_PROFILING, regs);
-#endif
 
 	lpaca->lppaca.int_dword.fields.decr_int = 0;
 
diff --git a/arch/s390/appldata/appldata_base.c b/arch/s390/appldata/appldata_base.c
index 01ae1964c93..c067435bae4 100644
--- a/arch/s390/appldata/appldata_base.c
+++ b/arch/s390/appldata/appldata_base.c
@@ -28,6 +28,7 @@
 //#include <linux/kernel_stat.h>
 #include <linux/notifier.h>
 #include <linux/cpu.h>
+#include <linux/workqueue.h>
 
 #include "appldata.h"
 
@@ -133,9 +134,12 @@ static int appldata_interval = APPLDATA_CPU_INTERVAL;
 static int appldata_timer_active;
 
 /*
- * Tasklet
+ * Work queue
  */
-static struct tasklet_struct appldata_tasklet_struct;
+static struct workqueue_struct *appldata_wq;
+static void appldata_work_fn(void *data);
+static DECLARE_WORK(appldata_work, appldata_work_fn, NULL);
+
 
 /*
  * Ops list
@@ -144,11 +148,11 @@ static DEFINE_SPINLOCK(appldata_ops_lock);
 static LIST_HEAD(appldata_ops_list);
 
 
-/************************* timer, tasklet, DIAG ******************************/
+/*************************** timer, work, DIAG *******************************/
 /*
  * appldata_timer_function()
  *
- * schedule tasklet and reschedule timer
+ * schedule work and reschedule timer
  */
 static void appldata_timer_function(unsigned long data, struct pt_regs *regs)
 {
@@ -157,22 +161,22 @@ static void appldata_timer_function(unsigned long data, struct pt_regs *regs)
 		atomic_read(&appldata_expire_count));
 	if (atomic_dec_and_test(&appldata_expire_count)) {
 		atomic_set(&appldata_expire_count, num_online_cpus());
-		tasklet_schedule((struct tasklet_struct *) data);
+		queue_work(appldata_wq, (struct work_struct *) data);
 	}
 }
 
 /*
- * appldata_tasklet_function()
+ * appldata_work_fn()
  *
  * call data gathering function for each (active) module
  */
-static void appldata_tasklet_function(unsigned long data)
+static void appldata_work_fn(void *data)
 {
 	struct list_head *lh;
 	struct appldata_ops *ops;
 	int i;
 
-	P_DEBUG("  -= Tasklet =-\n");
+	P_DEBUG("  -= Work Queue =-\n");
 	i = 0;
 	spin_lock(&appldata_ops_lock);
 	list_for_each(lh, &appldata_ops_list) {
@@ -231,7 +235,7 @@ static int appldata_diag(char record_nr, u16 function, unsigned long buffer,
 			: "=d" (ry) : "d" (&(appldata_parameter_list)) : "cc");
 	return (int) ry;
 }
-/********************** timer, tasklet, DIAG <END> ***************************/
+/************************ timer, work, DIAG <END> ****************************/
 
 
 /****************************** /proc stuff **********************************/
@@ -411,7 +415,7 @@ appldata_generic_handler(ctl_table *ctl, int write, struct file *filp,
 	struct list_head *lh;
 
 	found = 0;
-	spin_lock_bh(&appldata_ops_lock);
+	spin_lock(&appldata_ops_lock);
 	list_for_each(lh, &appldata_ops_list) {
 		tmp_ops = list_entry(lh, struct appldata_ops, list);
 		if (&tmp_ops->ctl_table[2] == ctl) {
@@ -419,15 +423,15 @@ appldata_generic_handler(ctl_table *ctl, int write, struct file *filp,
 		}
 	}
 	if (!found) {
-		spin_unlock_bh(&appldata_ops_lock);
+		spin_unlock(&appldata_ops_lock);
 		return -ENODEV;
 	}
 	ops = ctl->data;
 	if (!try_module_get(ops->owner)) {	// protect this function
-		spin_unlock_bh(&appldata_ops_lock);
+		spin_unlock(&appldata_ops_lock);
 		return -ENODEV;
 	}
-	spin_unlock_bh(&appldata_ops_lock);
+	spin_unlock(&appldata_ops_lock);
 
 	if (!*lenp || *ppos) {
 		*lenp = 0;
@@ -451,10 +455,11 @@ appldata_generic_handler(ctl_table *ctl, int write, struct file *filp,
 		return -EFAULT;
 	}
 
-	spin_lock_bh(&appldata_ops_lock);
+	spin_lock(&appldata_ops_lock);
 	if ((buf[0] == '1') && (ops->active == 0)) {
-		if (!try_module_get(ops->owner)) {	// protect tasklet
-			spin_unlock_bh(&appldata_ops_lock);
+		// protect work queue callback
+		if (!try_module_get(ops->owner)) {
+			spin_unlock(&appldata_ops_lock);
 			module_put(ops->owner);
 			return -ENODEV;
 		}
@@ -485,7 +490,7 @@ appldata_generic_handler(ctl_table *ctl, int write, struct file *filp,
 		}
 		module_put(ops->owner);
 	}
-	spin_unlock_bh(&appldata_ops_lock);
+	spin_unlock(&appldata_ops_lock);
 out:
 	*lenp = len;
 	*ppos += len;
@@ -529,7 +534,7 @@ int appldata_register_ops(struct appldata_ops *ops)
 	}
 	memset(ops->ctl_table, 0, 4*sizeof(struct ctl_table));
 
-	spin_lock_bh(&appldata_ops_lock);
+	spin_lock(&appldata_ops_lock);
 	list_for_each(lh, &appldata_ops_list) {
 		tmp_ops = list_entry(lh, struct appldata_ops, list);
 		P_DEBUG("register_ops loop: %i) name = %s, ctl = %i\n",
@@ -541,18 +546,18 @@ int appldata_register_ops(struct appldata_ops *ops)
 				APPLDATA_PROC_NAME_LENGTH) == 0) {
 			P_ERROR("Name \"%s\" already registered!\n", ops->name);
 			kfree(ops->ctl_table);
-			spin_unlock_bh(&appldata_ops_lock);
+			spin_unlock(&appldata_ops_lock);
 			return -EBUSY;
 		}
 		if (tmp_ops->ctl_nr == ops->ctl_nr) {
 			P_ERROR("ctl_nr %i already registered!\n", ops->ctl_nr);
 			kfree(ops->ctl_table);
-			spin_unlock_bh(&appldata_ops_lock);
+			spin_unlock(&appldata_ops_lock);
 			return -EBUSY;
 		}
 	}
 	list_add(&ops->list, &appldata_ops_list);
-	spin_unlock_bh(&appldata_ops_lock);
+	spin_unlock(&appldata_ops_lock);
 
 	ops->ctl_table[0].ctl_name = CTL_APPLDATA;
 	ops->ctl_table[0].procname = appldata_proc_name;
@@ -583,12 +588,12 @@ int appldata_register_ops(struct appldata_ops *ops)
  */
 void appldata_unregister_ops(struct appldata_ops *ops)
 {
-	spin_lock_bh(&appldata_ops_lock);
+	spin_lock(&appldata_ops_lock);
 	unregister_sysctl_table(ops->sysctl_header);
 	list_del(&ops->list);
 	kfree(ops->ctl_table);
 	ops->ctl_table = NULL;
-	spin_unlock_bh(&appldata_ops_lock);
+	spin_unlock(&appldata_ops_lock);
 	P_INFO("%s-ops unregistered!\n", ops->name);
 }
 /********************** module-ops management <END> **************************/
@@ -602,7 +607,7 @@ appldata_online_cpu(int cpu)
 	init_virt_timer(&per_cpu(appldata_timer, cpu));
 	per_cpu(appldata_timer, cpu).function = appldata_timer_function;
 	per_cpu(appldata_timer, cpu).data = (unsigned long)
-		&appldata_tasklet_struct;
+		&appldata_work;
 	atomic_inc(&appldata_expire_count);
 	spin_lock(&appldata_timer_lock);
 	__appldata_vtimer_setup(APPLDATA_MOD_TIMER);
@@ -615,7 +620,7 @@ appldata_offline_cpu(int cpu)
 	del_virt_timer(&per_cpu(appldata_timer, cpu));
 	if (atomic_dec_and_test(&appldata_expire_count)) {
 		atomic_set(&appldata_expire_count, num_online_cpus());
-		tasklet_schedule(&appldata_tasklet_struct);
+		queue_work(appldata_wq, &appldata_work);
 	}
 	spin_lock(&appldata_timer_lock);
 	__appldata_vtimer_setup(APPLDATA_MOD_TIMER);
@@ -648,7 +653,7 @@ static struct notifier_block __devinitdata appldata_nb = {
 /*
  * appldata_init()
  *
- * init timer and tasklet, register /proc entries
+ * init timer, register /proc entries
  */
 static int __init appldata_init(void)
 {
@@ -657,6 +662,12 @@ static int __init appldata_init(void)
 	P_DEBUG("sizeof(parameter_list) = %lu\n",
 		sizeof(struct appldata_parameter_list));
 
+	appldata_wq = create_singlethread_workqueue("appldata");
+	if (!appldata_wq) {
+		P_ERROR("Could not create work queue\n");
+		return -ENOMEM;
+	}
+
 	for_each_online_cpu(i)
 		appldata_online_cpu(i);
 
@@ -670,7 +681,6 @@ static int __init appldata_init(void)
 	appldata_table[1].de->owner = THIS_MODULE;
 #endif
 
-	tasklet_init(&appldata_tasklet_struct, appldata_tasklet_function, 0);
 	P_DEBUG("Base interface initialized.\n");
 	return 0;
 }
@@ -678,7 +688,7 @@ static int __init appldata_init(void)
 /*
  * appldata_exit()
  *
- * stop timer and tasklet, unregister /proc entries
+ * stop timer, unregister /proc entries
  */
 static void __exit appldata_exit(void)
 {
@@ -690,7 +700,7 @@ static void __exit appldata_exit(void)
 	/*
 	 * ops list should be empty, but just in case something went wrong...
 	 */
-	spin_lock_bh(&appldata_ops_lock);
+	spin_lock(&appldata_ops_lock);
 	list_for_each(lh, &appldata_ops_list) {
 		ops = list_entry(lh, struct appldata_ops, list);
 		rc = appldata_diag(ops->record_nr, APPLDATA_STOP_REC,
@@ -700,7 +710,7 @@ static void __exit appldata_exit(void)
 				"return code: %d\n", ops->name, rc);
 		}
 	}
-	spin_unlock_bh(&appldata_ops_lock);
+	spin_unlock(&appldata_ops_lock);
 
 	for_each_online_cpu(i)
 		appldata_offline_cpu(i);
@@ -709,7 +719,7 @@ static void __exit appldata_exit(void)
 
 	unregister_sysctl_table(appldata_sysctl_header);
 
-	tasklet_kill(&appldata_tasklet_struct);
+	destroy_workqueue(appldata_wq);
 	P_DEBUG("... module unloaded!\n");
 }
 /**************************** init / exit <END> ******************************/
diff --git a/arch/s390/appldata/appldata_mem.c b/arch/s390/appldata/appldata_mem.c
index 462ee9a84e7..f0e2fbed3d4 100644
--- a/arch/s390/appldata/appldata_mem.c
+++ b/arch/s390/appldata/appldata_mem.c
@@ -68,7 +68,7 @@ struct appldata_mem_data {
 	u64 pgmajfault;		/* page faults (major only) */
 // <-- New in 2.6
 
-} appldata_mem_data;
+} __attribute__((packed)) appldata_mem_data;
 
 
 static inline void appldata_debug_print(struct appldata_mem_data *mem_data)
diff --git a/arch/s390/appldata/appldata_net_sum.c b/arch/s390/appldata/appldata_net_sum.c
index dd61638d302..2a4c7432db4 100644
--- a/arch/s390/appldata/appldata_net_sum.c
+++ b/arch/s390/appldata/appldata_net_sum.c
@@ -57,7 +57,7 @@ struct appldata_net_sum_data {
 	u64 rx_dropped;		/* no space in linux buffers     */
 	u64 tx_dropped;		/* no space available in linux   */
 	u64 collisions;		/* collisions while transmitting */
-} appldata_net_sum_data;
+} __attribute__((packed)) appldata_net_sum_data;
 
 
 static inline void appldata_print_debug(struct appldata_net_sum_data *net_data)
diff --git a/arch/s390/appldata/appldata_os.c b/arch/s390/appldata/appldata_os.c
index b83f0748455..e0a476bf4fd 100644
--- a/arch/s390/appldata/appldata_os.c
+++ b/arch/s390/appldata/appldata_os.c
@@ -49,7 +49,7 @@ struct appldata_os_per_cpu {
 	u32 per_cpu_softirq;	/* ... spent in softirqs            */
 	u32 per_cpu_iowait;	/* ... spent while waiting for I/O  */
 // <-- New in 2.6
-};
+} __attribute__((packed));
 
 struct appldata_os_data {
 	u64 timestamp;
@@ -75,7 +75,7 @@ struct appldata_os_data {
 
 	/* per cpu data */
 	struct appldata_os_per_cpu os_cpu[0];
-};
+} __attribute__((packed));
 
 static struct appldata_os_data *appldata_os_data;
 
diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c
index 26889366929..06afa3103ac 100644
--- a/arch/s390/kernel/ptrace.c
+++ b/arch/s390/kernel/ptrace.c
@@ -40,6 +40,7 @@
 #include <asm/pgalloc.h>
 #include <asm/system.h>
 #include <asm/uaccess.h>
+#include <asm/unistd.h>
 
 #ifdef CONFIG_S390_SUPPORT
 #include "compat_ptrace.h"
@@ -130,13 +131,19 @@ static int
 peek_user(struct task_struct *child, addr_t addr, addr_t data)
 {
 	struct user *dummy = NULL;
-	addr_t offset, tmp;
+	addr_t offset, tmp, mask;
 
 	/*
 	 * Stupid gdb peeks/pokes the access registers in 64 bit with
 	 * an alignment of 4. Programmers from hell...
 	 */
-	if ((addr & 3) || addr > sizeof(struct user) - __ADDR_MASK)
+	mask = __ADDR_MASK;
+#ifdef CONFIG_ARCH_S390X
+	if (addr >= (addr_t) &dummy->regs.acrs &&
+	    addr < (addr_t) &dummy->regs.orig_gpr2)
+		mask = 3;
+#endif
+	if ((addr & mask) || addr > sizeof(struct user) - __ADDR_MASK)
 		return -EIO;
 
 	if (addr < (addr_t) &dummy->regs.acrs) {
@@ -153,6 +160,16 @@ peek_user(struct task_struct *child, addr_t addr, addr_t data)
 		 * access registers are stored in the thread structure
 		 */
 		offset = addr - (addr_t) &dummy->regs.acrs;
+#ifdef CONFIG_ARCH_S390X
+		/*
+		 * Very special case: old & broken 64 bit gdb reading
+		 * from acrs[15]. Result is a 64 bit value. Read the
+		 * 32 bit acrs[15] value and shift it by 32. Sick...
+		 */
+		if (addr == (addr_t) &dummy->regs.acrs[15])
+			tmp = ((unsigned long) child->thread.acrs[15]) << 32;
+		else
+#endif
 		tmp = *(addr_t *)((addr_t) &child->thread.acrs + offset);
 
 	} else if (addr == (addr_t) &dummy->regs.orig_gpr2) {
@@ -167,6 +184,9 @@ peek_user(struct task_struct *child, addr_t addr, addr_t data)
 		 */
 		offset = addr - (addr_t) &dummy->regs.fp_regs;
 		tmp = *(addr_t *)((addr_t) &child->thread.fp_regs + offset);
+		if (addr == (addr_t) &dummy->regs.fp_regs.fpc)
+			tmp &= (unsigned long) FPC_VALID_MASK
+				<< (BITS_PER_LONG - 32);
 
 	} else if (addr < (addr_t) (&dummy->regs.per_info + 1)) {
 		/*
@@ -191,13 +211,19 @@ static int
 poke_user(struct task_struct *child, addr_t addr, addr_t data)
 {
 	struct user *dummy = NULL;
-	addr_t offset;
+	addr_t offset, mask;
 
 	/*
 	 * Stupid gdb peeks/pokes the access registers in 64 bit with
 	 * an alignment of 4. Programmers from hell indeed...
 	 */
-	if ((addr & 3) || addr > sizeof(struct user) - __ADDR_MASK)
+	mask = __ADDR_MASK;
+#ifdef CONFIG_ARCH_S390X
+	if (addr >= (addr_t) &dummy->regs.acrs &&
+	    addr < (addr_t) &dummy->regs.orig_gpr2)
+		mask = 3;
+#endif
+	if ((addr & mask) || addr > sizeof(struct user) - __ADDR_MASK)
 		return -EIO;
 
 	if (addr < (addr_t) &dummy->regs.acrs) {
@@ -224,6 +250,17 @@ poke_user(struct task_struct *child, addr_t addr, addr_t data)
 		 * access registers are stored in the thread structure
 		 */
 		offset = addr - (addr_t) &dummy->regs.acrs;
+#ifdef CONFIG_ARCH_S390X
+		/*
+		 * Very special case: old & broken 64 bit gdb writing
+		 * to acrs[15] with a 64 bit value. Ignore the lower
+		 * half of the value and write the upper 32 bit to
+		 * acrs[15]. Sick...
+		 */
+		if (addr == (addr_t) &dummy->regs.acrs[15])
+			child->thread.acrs[15] = (unsigned int) (data >> 32);
+		else
+#endif
 		*(addr_t *)((addr_t) &child->thread.acrs + offset) = data;
 
 	} else if (addr == (addr_t) &dummy->regs.orig_gpr2) {
@@ -237,7 +274,8 @@ poke_user(struct task_struct *child, addr_t addr, addr_t data)
 		 * floating point regs. are stored in the thread structure
 		 */
 		if (addr == (addr_t) &dummy->regs.fp_regs.fpc &&
-		    (data & ~FPC_VALID_MASK) != 0)
+		    (data & ~((unsigned long) FPC_VALID_MASK
+			      << (BITS_PER_LONG - 32))) != 0)
 			return -EINVAL;
 		offset = addr - (addr_t) &dummy->regs.fp_regs;
 		*(addr_t *)((addr_t) &child->thread.fp_regs + offset) = data;
@@ -723,6 +761,13 @@ syscall_trace(struct pt_regs *regs, int entryexit)
 				 ? 0x80 : 0));
 
 	/*
+	 * If the debuffer has set an invalid system call number,
+	 * we prepare to skip the system call restart handling.
+	 */
+	if (!entryexit && regs->gprs[2] >= NR_syscalls)
+		regs->trap = -1;
+
+	/*
 	 * this isn't the same as continuing with a signal, but it will do
 	 * for normal use.  strace only continues with a signal if the
 	 * stopping signal is not SIGTRAP.  -brl
diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
index 80306bc8c79..75fde949d12 100644
--- a/arch/s390/mm/fault.c
+++ b/arch/s390/mm/fault.c
@@ -207,7 +207,7 @@ do_exception(struct pt_regs *regs, unsigned long error_code, int is_protection)
 	 * we are not in an interrupt and that there is a 
 	 * user context.
 	 */
-        if (user_address == 0 || in_interrupt() || !mm)
+        if (user_address == 0 || in_atomic() || !mm)
                 goto no_context;
 
 	/*
diff --git a/arch/sparc64/kernel/pci_iommu.c b/arch/sparc64/kernel/pci_iommu.c
index 33ca56c90da..2803bc7c2c7 100644
--- a/arch/sparc64/kernel/pci_iommu.c
+++ b/arch/sparc64/kernel/pci_iommu.c
@@ -196,6 +196,34 @@ static iopte_t *alloc_consistent_cluster(struct pci_iommu *iommu, unsigned long
 	return NULL;
 }
 
+static int iommu_alloc_ctx(struct pci_iommu *iommu)
+{
+	int lowest = iommu->ctx_lowest_free;
+	int sz = IOMMU_NUM_CTXS - lowest;
+	int n = find_next_zero_bit(iommu->ctx_bitmap, sz, lowest);
+
+	if (unlikely(n == sz)) {
+		n = find_next_zero_bit(iommu->ctx_bitmap, lowest, 1);
+		if (unlikely(n == lowest)) {
+			printk(KERN_WARNING "IOMMU: Ran out of contexts.\n");
+			n = 0;
+		}
+	}
+	if (n)
+		__set_bit(n, iommu->ctx_bitmap);
+
+	return n;
+}
+
+static inline void iommu_free_ctx(struct pci_iommu *iommu, int ctx)
+{
+	if (likely(ctx)) {
+		__clear_bit(ctx, iommu->ctx_bitmap);
+		if (ctx < iommu->ctx_lowest_free)
+			iommu->ctx_lowest_free = ctx;
+	}
+}
+
 /* Allocate and map kernel buffer of size SIZE using consistent mode
  * DMA for PCI device PDEV.  Return non-NULL cpu-side address if
  * successful and set *DMA_ADDRP to the PCI side dma address.
@@ -236,7 +264,7 @@ void *pci_alloc_consistent(struct pci_dev *pdev, size_t size, dma_addr_t *dma_ad
 	npages = size >> IO_PAGE_SHIFT;
 	ctx = 0;
 	if (iommu->iommu_ctxflush)
-		ctx = iommu->iommu_cur_ctx++;
+		ctx = iommu_alloc_ctx(iommu);
 	first_page = __pa(first_page);
 	while (npages--) {
 		iopte_val(*iopte) = (IOPTE_CONSISTENT(ctx) |
@@ -317,6 +345,8 @@ void pci_free_consistent(struct pci_dev *pdev, size_t size, void *cpu, dma_addr_
 		}
 	}
 
+	iommu_free_ctx(iommu, ctx);
+
 	spin_unlock_irqrestore(&iommu->lock, flags);
 
 	order = get_order(size);
@@ -360,7 +390,7 @@ dma_addr_t pci_map_single(struct pci_dev *pdev, void *ptr, size_t sz, int direct
 	base_paddr = __pa(oaddr & IO_PAGE_MASK);
 	ctx = 0;
 	if (iommu->iommu_ctxflush)
-		ctx = iommu->iommu_cur_ctx++;
+		ctx = iommu_alloc_ctx(iommu);
 	if (strbuf->strbuf_enabled)
 		iopte_protection = IOPTE_STREAMING(ctx);
 	else
@@ -380,39 +410,53 @@ bad:
 	return PCI_DMA_ERROR_CODE;
 }
 
-static void pci_strbuf_flush(struct pci_strbuf *strbuf, struct pci_iommu *iommu, u32 vaddr, unsigned long ctx, unsigned long npages)
+static void pci_strbuf_flush(struct pci_strbuf *strbuf, struct pci_iommu *iommu, u32 vaddr, unsigned long ctx, unsigned long npages, int direction)
 {
 	int limit;
 
-	PCI_STC_FLUSHFLAG_INIT(strbuf);
 	if (strbuf->strbuf_ctxflush &&
 	    iommu->iommu_ctxflush) {
 		unsigned long matchreg, flushreg;
+		u64 val;
 
 		flushreg = strbuf->strbuf_ctxflush;
 		matchreg = PCI_STC_CTXMATCH_ADDR(strbuf, ctx);
 
-		limit = 100000;
 		pci_iommu_write(flushreg, ctx);
-		for(;;) {
-			if (((long)pci_iommu_read(matchreg)) >= 0L)
-				break;
-			limit--;
-			if (!limit)
-				break;
-			udelay(1);
+		val = pci_iommu_read(matchreg);
+		val &= 0xffff;
+		if (!val)
+			goto do_flush_sync;
+
+		while (val) {
+			if (val & 0x1)
+				pci_iommu_write(flushreg, ctx);
+			val >>= 1;
 		}
-		if (!limit)
+		val = pci_iommu_read(matchreg);
+		if (unlikely(val)) {
 			printk(KERN_WARNING "pci_strbuf_flush: ctx flush "
-			       "timeout vaddr[%08x] ctx[%lx]\n",
-			       vaddr, ctx);
+			       "timeout matchreg[%lx] ctx[%lx]\n",
+			       val, ctx);
+			goto do_page_flush;
+		}
 	} else {
 		unsigned long i;
 
+	do_page_flush:
 		for (i = 0; i < npages; i++, vaddr += IO_PAGE_SIZE)
 			pci_iommu_write(strbuf->strbuf_pflush, vaddr);
 	}
 
+do_flush_sync:
+	/* If the device could not have possibly put dirty data into
+	 * the streaming cache, no flush-flag synchronization needs
+	 * to be performed.
+	 */
+	if (direction == PCI_DMA_TODEVICE)
+		return;
+
+	PCI_STC_FLUSHFLAG_INIT(strbuf);
 	pci_iommu_write(strbuf->strbuf_fsync, strbuf->strbuf_flushflag_pa);
 	(void) pci_iommu_read(iommu->write_complete_reg);
 
@@ -466,7 +510,7 @@ void pci_unmap_single(struct pci_dev *pdev, dma_addr_t bus_addr, size_t sz, int
 
 	/* Step 1: Kick data out of streaming buffers if necessary. */
 	if (strbuf->strbuf_enabled)
-		pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages);
+		pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages, direction);
 
 	/* Step 2: Clear out first TSB entry. */
 	iopte_make_dummy(iommu, base);
@@ -474,6 +518,8 @@ void pci_unmap_single(struct pci_dev *pdev, dma_addr_t bus_addr, size_t sz, int
 	free_streaming_cluster(iommu, bus_addr - iommu->page_table_map_base,
 			       npages, ctx);
 
+	iommu_free_ctx(iommu, ctx);
+
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
 
@@ -613,7 +659,7 @@ int pci_map_sg(struct pci_dev *pdev, struct scatterlist *sglist, int nelems, int
 	/* Step 4: Choose a context if necessary. */
 	ctx = 0;
 	if (iommu->iommu_ctxflush)
-		ctx = iommu->iommu_cur_ctx++;
+		ctx = iommu_alloc_ctx(iommu);
 
 	/* Step 5: Create the mappings. */
 	if (strbuf->strbuf_enabled)
@@ -678,7 +724,7 @@ void pci_unmap_sg(struct pci_dev *pdev, struct scatterlist *sglist, int nelems,
 
 	/* Step 1: Kick data out of streaming buffers if necessary. */
 	if (strbuf->strbuf_enabled)
-		pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages);
+		pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages, direction);
 
 	/* Step 2: Clear out first TSB entry. */
 	iopte_make_dummy(iommu, base);
@@ -686,6 +732,8 @@ void pci_unmap_sg(struct pci_dev *pdev, struct scatterlist *sglist, int nelems,
 	free_streaming_cluster(iommu, bus_addr - iommu->page_table_map_base,
 			       npages, ctx);
 
+	iommu_free_ctx(iommu, ctx);
+
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
 
@@ -724,7 +772,7 @@ void pci_dma_sync_single_for_cpu(struct pci_dev *pdev, dma_addr_t bus_addr, size
 	}
 
 	/* Step 2: Kick data out of streaming buffers. */
-	pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages);
+	pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages, direction);
 
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
@@ -768,7 +816,7 @@ void pci_dma_sync_sg_for_cpu(struct pci_dev *pdev, struct scatterlist *sglist, i
 	i--;
 	npages = (IO_PAGE_ALIGN(sglist[i].dma_address + sglist[i].dma_length)
 		  - bus_addr) >> IO_PAGE_SHIFT;
-	pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages);
+	pci_strbuf_flush(strbuf, iommu, bus_addr, ctx, npages, direction);
 
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
diff --git a/arch/sparc64/kernel/pci_psycho.c b/arch/sparc64/kernel/pci_psycho.c
index 3567fa879e1..534320ef0db 100644
--- a/arch/sparc64/kernel/pci_psycho.c
+++ b/arch/sparc64/kernel/pci_psycho.c
@@ -1212,7 +1212,7 @@ static void __init psycho_iommu_init(struct pci_controller_info *p)
 
 	/* Setup initial software IOMMU state. */
 	spin_lock_init(&iommu->lock);
-	iommu->iommu_cur_ctx = 0;
+	iommu->ctx_lowest_free = 1;
 
 	/* Register addresses. */
 	iommu->iommu_control  = p->pbm_A.controller_regs + PSYCHO_IOMMU_CONTROL;
diff --git a/arch/sparc64/kernel/pci_sabre.c b/arch/sparc64/kernel/pci_sabre.c
index 5525d1ec4af..53d333b4a4e 100644
--- a/arch/sparc64/kernel/pci_sabre.c
+++ b/arch/sparc64/kernel/pci_sabre.c
@@ -1265,7 +1265,7 @@ static void __init sabre_iommu_init(struct pci_controller_info *p,
 
 	/* Setup initial software IOMMU state. */
 	spin_lock_init(&iommu->lock);
-	iommu->iommu_cur_ctx = 0;
+	iommu->ctx_lowest_free = 1;
 
 	/* Register addresses. */
 	iommu->iommu_control  = p->pbm_A.controller_regs + SABRE_IOMMU_CONTROL;
diff --git a/arch/sparc64/kernel/pci_schizo.c b/arch/sparc64/kernel/pci_schizo.c
index e93fcadc372..5753175b94e 100644
--- a/arch/sparc64/kernel/pci_schizo.c
+++ b/arch/sparc64/kernel/pci_schizo.c
@@ -1753,7 +1753,7 @@ static void schizo_pbm_iommu_init(struct pci_pbm_info *pbm)
 
 	/* Setup initial software IOMMU state. */
 	spin_lock_init(&iommu->lock);
-	iommu->iommu_cur_ctx = 0;
+	iommu->ctx_lowest_free = 1;
 
 	/* Register addresses, SCHIZO has iommu ctx flushing. */
 	iommu->iommu_control  = pbm->pbm_regs + SCHIZO_IOMMU_CONTROL;
diff --git a/arch/sparc64/kernel/sbus.c b/arch/sparc64/kernel/sbus.c
index 76ea6455433..89f5e019f24 100644
--- a/arch/sparc64/kernel/sbus.c
+++ b/arch/sparc64/kernel/sbus.c
@@ -117,17 +117,25 @@ static void iommu_flush(struct sbus_iommu *iommu, u32 base, unsigned long npages
 
 #define STRBUF_TAG_VALID	0x02UL
 
-static void sbus_strbuf_flush(struct sbus_iommu *iommu, u32 base, unsigned long npages)
+static void sbus_strbuf_flush(struct sbus_iommu *iommu, u32 base, unsigned long npages, int direction)
 {
 	unsigned long n;
 	int limit;
 
-	iommu->strbuf_flushflag = 0UL;
 	n = npages;
 	while (n--)
 		upa_writeq(base + (n << IO_PAGE_SHIFT),
 			   iommu->strbuf_regs + STRBUF_PFLUSH);
 
+	/* If the device could not have possibly put dirty data into
+	 * the streaming cache, no flush-flag synchronization needs
+	 * to be performed.
+	 */
+	if (direction == SBUS_DMA_TODEVICE)
+		return;
+
+	iommu->strbuf_flushflag = 0UL;
+
 	/* Whoopee cushion! */
 	upa_writeq(__pa(&iommu->strbuf_flushflag),
 		   iommu->strbuf_regs + STRBUF_FSYNC);
@@ -421,7 +429,7 @@ void sbus_unmap_single(struct sbus_dev *sdev, dma_addr_t dma_addr, size_t size,
 
 	spin_lock_irqsave(&iommu->lock, flags);
 	free_streaming_cluster(iommu, dma_base, size >> IO_PAGE_SHIFT);
-	sbus_strbuf_flush(iommu, dma_base, size >> IO_PAGE_SHIFT);
+	sbus_strbuf_flush(iommu, dma_base, size >> IO_PAGE_SHIFT, direction);
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
 
@@ -584,7 +592,7 @@ void sbus_unmap_sg(struct sbus_dev *sdev, struct scatterlist *sg, int nents, int
 	iommu = sdev->bus->iommu;
 	spin_lock_irqsave(&iommu->lock, flags);
 	free_streaming_cluster(iommu, dvma_base, size >> IO_PAGE_SHIFT);
-	sbus_strbuf_flush(iommu, dvma_base, size >> IO_PAGE_SHIFT);
+	sbus_strbuf_flush(iommu, dvma_base, size >> IO_PAGE_SHIFT, direction);
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
 
@@ -596,7 +604,7 @@ void sbus_dma_sync_single_for_cpu(struct sbus_dev *sdev, dma_addr_t base, size_t
 	size = (IO_PAGE_ALIGN(base + size) - (base & IO_PAGE_MASK));
 
 	spin_lock_irqsave(&iommu->lock, flags);
-	sbus_strbuf_flush(iommu, base & IO_PAGE_MASK, size >> IO_PAGE_SHIFT);
+	sbus_strbuf_flush(iommu, base & IO_PAGE_MASK, size >> IO_PAGE_SHIFT, direction);
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
 
@@ -620,7 +628,7 @@ void sbus_dma_sync_sg_for_cpu(struct sbus_dev *sdev, struct scatterlist *sg, int
 	size = IO_PAGE_ALIGN(sg[i].dma_address + sg[i].dma_length) - base;
 
 	spin_lock_irqsave(&iommu->lock, flags);
-	sbus_strbuf_flush(iommu, base, size >> IO_PAGE_SHIFT);
+	sbus_strbuf_flush(iommu, base, size >> IO_PAGE_SHIFT, direction);
 	spin_unlock_irqrestore(&iommu->lock, flags);
 }
 
diff --git a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig
index 0f430d9d363..289f448ac89 100644
--- a/arch/x86_64/Kconfig
+++ b/arch/x86_64/Kconfig
@@ -305,6 +305,7 @@ config HPET_TIMER
 
 config X86_PM_TIMER
 	bool "PM timer"
+	depends on ACPI
 	default y
 	help
 	  Support the ACPI PM timer for time keeping. This is slow,
diff --git a/arch/x86_64/kernel/io_apic.c b/arch/x86_64/kernel/io_apic.c
index 80e9b498c44..afd87e64d0a 100644
--- a/arch/x86_64/kernel/io_apic.c
+++ b/arch/x86_64/kernel/io_apic.c
@@ -37,6 +37,7 @@
 #include <asm/desc.h>
 #include <asm/proto.h>
 #include <asm/mach_apic.h>
+#include <asm/acpi.h>
 
 #define __apicdebuginit  __init
 
diff --git a/arch/x86_64/kernel/mpparse.c b/arch/x86_64/kernel/mpparse.c
index f86d9db94bf..61a63be6b29 100644
--- a/arch/x86_64/kernel/mpparse.c
+++ b/arch/x86_64/kernel/mpparse.c
@@ -30,6 +30,7 @@
 #include <asm/pgalloc.h>
 #include <asm/io_apic.h>
 #include <asm/proto.h>
+#include <asm/acpi.h>
 
 /* Have we found an MP table */
 int smp_found_config;
diff --git a/arch/x86_64/kernel/time.c b/arch/x86_64/kernel/time.c
index 735b6767c8e..fb8c809b4cd 100644
--- a/arch/x86_64/kernel/time.c
+++ b/arch/x86_64/kernel/time.c
@@ -27,7 +27,9 @@
 #include <linux/bcd.h>
 #include <linux/kallsyms.h>
 #include <linux/acpi.h>
+#ifdef CONFIG_ACPI
 #include <acpi/achware.h>	/* for PM timer frequency */
+#endif
 #include <asm/8253pit.h>
 #include <asm/pgtable.h>
 #include <asm/vsyscall.h>