Merge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (49 commits)
  powerpc: Fix build bug with binutils < 2.18 and GCC < 4.2
  powerpc/eeh: Don't panic when EEH_MAX_FAILS is exceeded
  fbdev: Teaches offb about palette on radeon r5xx/r6xx
  powerpc/cell/edac: Log a syndrome code in case of correctable error
  powerpc/cell: Add DMA_ATTR_WEAK_ORDERING dma attribute and use in Cell IOMMU code
  powerpc: Indicate which oprofile counters to use while in compat mode
  powerpc/boot: Change spaces to tabs
  powerpc: Remove duplicate 6xx option in Kconfig
  powerpc: Use PPC_LONG and PPC_LONG_ALIGN in lib/string.S
  powerpc: Use PPC_LONG_ALIGN in uaccess.h
  powerpc: Add a #define for aligning to a long-sized boundary
  powerpc: Fix OF parsing of 64 bits PCI addresses
  powerpc: Use WARN_ON(1) instead of __WARN()
  powerpc: Fix support for latencytop
  powerpc/ps3: Update ps3_defconfig
  powerpc/ps3: Add a sub-match id to ps3_system_bus
  powerpc: Add a 6xx defconfig
  powerpc/dma: Use the struct dma_attrs in iommu code
  powerpc/cell: Add support for power button of future IBM cell blades
  powerpc/cell: Cleanup sysreset_hack for IBM cell blades
  ...

8 files changed, 496 insertions, 33 deletions

diff --git a/arch/powerpc/platforms/cell/Kconfig b/arch/powerpc/platforms/cell/Kconfig
index 3959fcfe731..c14d7d8d96c 100644
--- a/arch/powerpc/platforms/cell/Kconfig
+++ b/arch/powerpc/platforms/cell/Kconfig
@@ -83,6 +83,22 @@ config CBE_RAS
 	depends on PPC_CELL_NATIVE
 	default y
 
+config PPC_IBM_CELL_RESETBUTTON
+	bool "IBM Cell Blade Pinhole reset button"
+	depends on CBE_RAS && PPC_IBM_CELL_BLADE
+	default y
+	help
+	  Support Pinhole Resetbutton on IBM Cell blades.
+	  This adds a method to trigger system reset via front panel pinhole button.
+
+config PPC_IBM_CELL_POWERBUTTON
+	tristate "IBM Cell Blade power button"
+	depends on PPC_IBM_CELL_BLADE && PPC_PMI && INPUT_EVDEV
+	default y
+	help
+	  Support Powerbutton on IBM Cell blades.
+	  This will enable the powerbutton as an input device.
+
 config CBE_THERM
 	tristate "CBE thermal support"
 	default m
@@ -107,6 +123,15 @@ config CBE_CPUFREQ_PMI
 	  processor will not only be able to run at lower speed,
 	  but also at lower core voltage.
 
+config CBE_CPUFREQ_SPU_GOVERNOR
+	tristate "CBE frequency scaling based on SPU usage"
+	depends on SPU_FS && CPU_FREQ
+	default m
+	help
+	  This governor checks for spu usage to adjust the cpu frequency.
+	  If no spu is running on a given cpu, that cpu will be throttled to
+	  the minimal possible frequency.
+
 endmenu
 
 config OPROFILE_CELL
diff --git a/arch/powerpc/platforms/cell/Makefile b/arch/powerpc/platforms/cell/Makefile
index c2a7e4e5ddf..7fd830872c4 100644
--- a/arch/powerpc/platforms/cell/Makefile
+++ b/arch/powerpc/platforms/cell/Makefile
@@ -8,6 +8,9 @@ obj-$(CONFIG_CBE_THERM)			+= cbe_thermal.o
 obj-$(CONFIG_CBE_CPUFREQ_PMI)		+= cbe_cpufreq_pmi.o
 obj-$(CONFIG_CBE_CPUFREQ)		+= cbe-cpufreq.o
 cbe-cpufreq-y				+= cbe_cpufreq_pervasive.o cbe_cpufreq.o
+obj-$(CONFIG_CBE_CPUFREQ_SPU_GOVERNOR)	+= cpufreq_spudemand.o
+
+obj-$(CONFIG_PPC_IBM_CELL_POWERBUTTON)	+= cbe_powerbutton.o
 
 ifeq ($(CONFIG_SMP),y)
 obj-$(CONFIG_PPC_CELL_NATIVE)		+= smp.o
diff --git a/arch/powerpc/platforms/cell/cbe_powerbutton.c b/arch/powerpc/platforms/cell/cbe_powerbutton.c
new file mode 100644
index 00000000000..dcddaa5fcb6
--- /dev/null
+++ b/arch/powerpc/platforms/cell/cbe_powerbutton.c
@@ -0,0 +1,117 @@
+/*
+ * driver for powerbutton on IBM cell blades
+ *
+ * (C) Copyright IBM Corp. 2005-2008
+ *
+ * Author: Christian Krafft <krafft@de.ibm.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, 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., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/input.h>
+#include <linux/platform_device.h>
+#include <asm/pmi.h>
+#include <asm/prom.h>
+
+static struct input_dev *button_dev;
+static struct platform_device *button_pdev;
+
+static void cbe_powerbutton_handle_pmi(pmi_message_t pmi_msg)
+{
+	BUG_ON(pmi_msg.type != PMI_TYPE_POWER_BUTTON);
+
+	input_report_key(button_dev, KEY_POWER, 1);
+	input_sync(button_dev);
+	input_report_key(button_dev, KEY_POWER, 0);
+	input_sync(button_dev);
+}
+
+static struct pmi_handler cbe_pmi_handler = {
+	.type			= PMI_TYPE_POWER_BUTTON,
+	.handle_pmi_message	= cbe_powerbutton_handle_pmi,
+};
+
+static int __init cbe_powerbutton_init(void)
+{
+	int ret = 0;
+	struct input_dev *dev;
+
+	if (!machine_is_compatible("IBM,CBPLUS-1.0")) {
+		printk(KERN_ERR "%s: Not a cell blade.\n", __func__);
+		ret = -ENODEV;
+		goto out;
+	}
+
+	dev = input_allocate_device();
+	if (!dev) {
+		ret = -ENOMEM;
+		printk(KERN_ERR "%s: Not enough memory.\n", __func__);
+		goto out;
+	}
+
+	set_bit(EV_KEY, dev->evbit);
+	set_bit(KEY_POWER, dev->keybit);
+
+	dev->name = "Power Button";
+	dev->id.bustype = BUS_HOST;
+
+	/* this makes the button look like an acpi power button
+	 * no clue whether anyone relies on that though */
+	dev->id.product = 0x02;
+	dev->phys = "LNXPWRBN/button/input0";
+
+	button_pdev = platform_device_register_simple("power_button", 0, NULL, 0);
+	if (IS_ERR(button_pdev)) {
+		ret = PTR_ERR(button_pdev);
+		goto out_free_input;
+	}
+
+	dev->dev.parent = &button_pdev->dev;
+	ret = input_register_device(dev);
+	if (ret) {
+		printk(KERN_ERR "%s: Failed to register device\n", __func__);
+		goto out_free_pdev;
+	}
+
+	button_dev = dev;
+
+	ret = pmi_register_handler(&cbe_pmi_handler);
+	if (ret) {
+		printk(KERN_ERR "%s: Failed to register with pmi.\n", __func__);
+		goto out_free_pdev;
+	}
+
+	goto out;
+
+out_free_pdev:
+	platform_device_unregister(button_pdev);
+out_free_input:
+	input_free_device(dev);
+out:
+	return ret;
+}
+
+static void __exit cbe_powerbutton_exit(void)
+{
+	pmi_unregister_handler(&cbe_pmi_handler);
+	platform_device_unregister(button_pdev);
+	input_free_device(button_dev);
+}
+
+module_init(cbe_powerbutton_init);
+module_exit(cbe_powerbutton_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Christian Krafft <krafft@de.ibm.com>");
diff --git a/arch/powerpc/platforms/cell/cpufreq_spudemand.c b/arch/powerpc/platforms/cell/cpufreq_spudemand.c
new file mode 100644
index 00000000000..a3c6c01bd6d
--- /dev/null
+++ b/arch/powerpc/platforms/cell/cpufreq_spudemand.c
@@ -0,0 +1,184 @@
+/*
+ * spu aware cpufreq governor for the cell processor
+ *
+ * © Copyright IBM Corporation 2006-2008
+ *
+ * Author: Christian Krafft <krafft@de.ibm.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, 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., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/cpufreq.h>
+#include <linux/sched.h>
+#include <linux/timer.h>
+#include <linux/workqueue.h>
+#include <asm/atomic.h>
+#include <asm/machdep.h>
+#include <asm/spu.h>
+
+#define POLL_TIME	100000		/* in µs */
+#define EXP		753		/* exp(-1) in fixed-point */
+
+struct spu_gov_info_struct {
+	unsigned long busy_spus;	/* fixed-point */
+	struct cpufreq_policy *policy;
+	struct delayed_work work;
+	unsigned int poll_int;		/* µs */
+};
+static DEFINE_PER_CPU(struct spu_gov_info_struct, spu_gov_info);
+
+static struct workqueue_struct *kspugov_wq;
+
+static int calc_freq(struct spu_gov_info_struct *info)
+{
+	int cpu;
+	int busy_spus;
+
+	cpu = info->policy->cpu;
+	busy_spus = atomic_read(&cbe_spu_info[cpu_to_node(cpu)].busy_spus);
+
+	CALC_LOAD(info->busy_spus, EXP, busy_spus * FIXED_1);
+	pr_debug("cpu %d: busy_spus=%d, info->busy_spus=%ld\n",
+			cpu, busy_spus, info->busy_spus);
+
+	return info->policy->max * info->busy_spus / FIXED_1;
+}
+
+static void spu_gov_work(struct work_struct *work)
+{
+	struct spu_gov_info_struct *info;
+	int delay;
+	unsigned long target_freq;
+
+	info = container_of(work, struct spu_gov_info_struct, work.work);
+
+	/* after cancel_delayed_work_sync we unset info->policy */
+	BUG_ON(info->policy == NULL);
+
+	target_freq = calc_freq(info);
+	__cpufreq_driver_target(info->policy, target_freq, CPUFREQ_RELATION_H);
+
+	delay = usecs_to_jiffies(info->poll_int);
+	queue_delayed_work_on(info->policy->cpu, kspugov_wq, &info->work, delay);
+}
+
+static void spu_gov_init_work(struct spu_gov_info_struct *info)
+{
+	int delay = usecs_to_jiffies(info->poll_int);
+	INIT_DELAYED_WORK_DEFERRABLE(&info->work, spu_gov_work);
+	queue_delayed_work_on(info->policy->cpu, kspugov_wq, &info->work, delay);
+}
+
+static void spu_gov_cancel_work(struct spu_gov_info_struct *info)
+{
+	cancel_delayed_work_sync(&info->work);
+}
+
+static int spu_gov_govern(struct cpufreq_policy *policy, unsigned int event)
+{
+	unsigned int cpu = policy->cpu;
+	struct spu_gov_info_struct *info, *affected_info;
+	int i;
+	int ret = 0;
+
+	info = &per_cpu(spu_gov_info, cpu);
+
+	switch (event) {
+	case CPUFREQ_GOV_START:
+		if (!cpu_online(cpu)) {
+			printk(KERN_ERR "cpu %d is not online\n", cpu);
+			ret = -EINVAL;
+			break;
+		}
+
+		if (!policy->cur) {
+			printk(KERN_ERR "no cpu specified in policy\n");
+			ret = -EINVAL;
+			break;
+		}
+
+		/* initialize spu_gov_info for all affected cpus */
+		for_each_cpu_mask(i, policy->cpus) {
+			affected_info = &per_cpu(spu_gov_info, i);
+			affected_info->policy = policy;
+		}
+
+		info->poll_int = POLL_TIME;
+
+		/* setup timer */
+		spu_gov_init_work(info);
+
+		break;
+
+	case CPUFREQ_GOV_STOP:
+		/* cancel timer */
+		spu_gov_cancel_work(info);
+
+		/* clean spu_gov_info for all affected cpus */
+		for_each_cpu_mask (i, policy->cpus) {
+			info = &per_cpu(spu_gov_info, i);
+			info->policy = NULL;
+		}
+
+		break;
+	}
+
+	return ret;
+}
+
+static struct cpufreq_governor spu_governor = {
+	.name = "spudemand",
+	.governor = spu_gov_govern,
+	.owner = THIS_MODULE,
+};
+
+/*
+ * module init and destoy
+ */
+
+static int __init spu_gov_init(void)
+{
+	int ret;
+
+	kspugov_wq = create_workqueue("kspugov");
+	if (!kspugov_wq) {
+		printk(KERN_ERR "creation of kspugov failed\n");
+		ret = -EFAULT;
+		goto out;
+	}
+
+	ret = cpufreq_register_governor(&spu_governor);
+	if (ret) {
+		printk(KERN_ERR "registration of governor failed\n");
+		destroy_workqueue(kspugov_wq);
+		goto out;
+	}
+out:
+	return ret;
+}
+
+static void __exit spu_gov_exit(void)
+{
+	cpufreq_unregister_governor(&spu_governor);
+	destroy_workqueue(kspugov_wq);
+}
+
+
+module_init(spu_gov_init);
+module_exit(spu_gov_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Christian Krafft <krafft@de.ibm.com>");
+
diff --git a/arch/powerpc/platforms/cell/iommu.c b/arch/powerpc/platforms/cell/iommu.c
index eeacb3a52ca..208005ca262 100644
--- a/arch/powerpc/platforms/cell/iommu.c
+++ b/arch/powerpc/platforms/cell/iommu.c
@@ -173,7 +173,8 @@ static void invalidate_tce_cache(struct cbe_iommu *iommu, unsigned long *pte,
 }
 
 static void tce_build_cell(struct iommu_table *tbl, long index, long npages,
-		unsigned long uaddr, enum dma_data_direction direction)
+		unsigned long uaddr, enum dma_data_direction direction,
+		struct dma_attrs *attrs)
 {
 	int i;
 	unsigned long *io_pte, base_pte;
@@ -198,6 +199,8 @@ static void tce_build_cell(struct iommu_table *tbl, long index, long npages,
 	base_pte = IOPTE_PP_W | IOPTE_PP_R | IOPTE_M | IOPTE_SO_RW |
 		(window->ioid & IOPTE_IOID_Mask);
 #endif
+	if (unlikely(dma_get_attr(DMA_ATTR_WEAK_ORDERING, attrs)))
+		base_pte &= ~IOPTE_SO_RW;
 
 	io_pte = (unsigned long *)tbl->it_base + (index - tbl->it_offset);
 
@@ -519,7 +522,7 @@ cell_iommu_setup_window(struct cbe_iommu *iommu, struct device_node *np,
 
 	__set_bit(0, window->table.it_map);
 	tce_build_cell(&window->table, window->table.it_offset, 1,
-		       (unsigned long)iommu->pad_page, DMA_TO_DEVICE);
+		       (unsigned long)iommu->pad_page, DMA_TO_DEVICE, NULL);
 	window->table.it_hint = window->table.it_blocksize;
 
 	return window;
@@ -538,7 +541,9 @@ static struct cbe_iommu *cell_iommu_for_node(int nid)
 static unsigned long cell_dma_direct_offset;
 
 static unsigned long dma_iommu_fixed_base;
-struct dma_mapping_ops dma_iommu_fixed_ops;
+
+/* iommu_fixed_is_weak is set if booted with iommu_fixed=weak */
+static int iommu_fixed_is_weak;
 
 static struct iommu_table *cell_get_iommu_table(struct device *dev)
 {
@@ -562,6 +567,98 @@ static struct iommu_table *cell_get_iommu_table(struct device *dev)
 	return &window->table;
 }
 
+/* A coherent allocation implies strong ordering */
+
+static void *dma_fixed_alloc_coherent(struct device *dev, size_t size,
+				      dma_addr_t *dma_handle, gfp_t flag)
+{
+	if (iommu_fixed_is_weak)
+		return iommu_alloc_coherent(dev, cell_get_iommu_table(dev),
+					    size, dma_handle,
+					    device_to_mask(dev), flag,
+					    dev->archdata.numa_node);
+	else
+		return dma_direct_ops.alloc_coherent(dev, size, dma_handle,
+						     flag);
+}
+
+static void dma_fixed_free_coherent(struct device *dev, size_t size,
+				    void *vaddr, dma_addr_t dma_handle)
+{
+	if (iommu_fixed_is_weak)
+		iommu_free_coherent(cell_get_iommu_table(dev), size, vaddr,
+				    dma_handle);
+	else
+		dma_direct_ops.free_coherent(dev, size, vaddr, dma_handle);
+}
+
+static dma_addr_t dma_fixed_map_single(struct device *dev, void *ptr,
+				       size_t size,
+				       enum dma_data_direction direction,
+				       struct dma_attrs *attrs)
+{
+	if (iommu_fixed_is_weak == dma_get_attr(DMA_ATTR_WEAK_ORDERING, attrs))
+		return dma_direct_ops.map_single(dev, ptr, size, direction,
+						 attrs);
+	else
+		return iommu_map_single(dev, cell_get_iommu_table(dev), ptr,
+					size, device_to_mask(dev), direction,
+					attrs);
+}
+
+static void dma_fixed_unmap_single(struct device *dev, dma_addr_t dma_addr,
+				   size_t size,
+				   enum dma_data_direction direction,
+				   struct dma_attrs *attrs)
+{
+	if (iommu_fixed_is_weak == dma_get_attr(DMA_ATTR_WEAK_ORDERING, attrs))
+		dma_direct_ops.unmap_single(dev, dma_addr, size, direction,
+					    attrs);
+	else
+		iommu_unmap_single(cell_get_iommu_table(dev), dma_addr, size,
+				   direction, attrs);
+}
+
+static int dma_fixed_map_sg(struct device *dev, struct scatterlist *sg,
+			   int nents, enum dma_data_direction direction,
+			   struct dma_attrs *attrs)
+{
+	if (iommu_fixed_is_weak == dma_get_attr(DMA_ATTR_WEAK_ORDERING, attrs))
+		return dma_direct_ops.map_sg(dev, sg, nents, direction, attrs);
+	else
+		return iommu_map_sg(dev, cell_get_iommu_table(dev), sg, nents,
+				    device_to_mask(dev), direction, attrs);
+}
+
+static void dma_fixed_unmap_sg(struct device *dev, struct scatterlist *sg,
+			       int nents, enum dma_data_direction direction,
+			       struct dma_attrs *attrs)
+{
+	if (iommu_fixed_is_weak == dma_get_attr(DMA_ATTR_WEAK_ORDERING, attrs))
+		dma_direct_ops.unmap_sg(dev, sg, nents, direction, attrs);
+	else
+		iommu_unmap_sg(cell_get_iommu_table(dev), sg, nents, direction,
+			       attrs);
+}
+
+static int dma_fixed_dma_supported(struct device *dev, u64 mask)
+{
+	return mask == DMA_64BIT_MASK;
+}
+
+static int dma_set_mask_and_switch(struct device *dev, u64 dma_mask);
+
+struct dma_mapping_ops dma_iommu_fixed_ops = {
+	.alloc_coherent = dma_fixed_alloc_coherent,
+	.free_coherent  = dma_fixed_free_coherent,
+	.map_single     = dma_fixed_map_single,
+	.unmap_single   = dma_fixed_unmap_single,
+	.map_sg         = dma_fixed_map_sg,
+	.unmap_sg       = dma_fixed_unmap_sg,
+	.dma_supported  = dma_fixed_dma_supported,
+	.set_dma_mask   = dma_set_mask_and_switch,
+};
+
 static void cell_dma_dev_setup_fixed(struct device *dev);
 
 static void cell_dma_dev_setup(struct device *dev)
@@ -918,9 +1015,16 @@ static void cell_iommu_setup_fixed_ptab(struct cbe_iommu *iommu,
 
 	pr_debug("iommu: mapping 0x%lx pages from 0x%lx\n", fsize, fbase);
 
-	base_pte = IOPTE_PP_W | IOPTE_PP_R | IOPTE_M | IOPTE_SO_RW
+	base_pte = IOPTE_PP_W | IOPTE_PP_R | IOPTE_M
 		    | (cell_iommu_get_ioid(np) & IOPTE_IOID_Mask);
 
+	if (iommu_fixed_is_weak)
+		pr_info("IOMMU: Using weak ordering for fixed mapping\n");
+	else {
+		pr_info("IOMMU: Using strong ordering for fixed mapping\n");
+		base_pte |= IOPTE_SO_RW;
+	}
+
 	for (uaddr = 0; uaddr < fsize; uaddr += (1 << 24)) {
 		/* Don't touch the dynamic region */
 		ioaddr = uaddr + fbase;
@@ -1036,9 +1140,6 @@ static int __init cell_iommu_fixed_mapping_init(void)
 		cell_iommu_setup_window(iommu, np, dbase, dsize, 0);
 	}
 
-	dma_iommu_fixed_ops = dma_direct_ops;
-	dma_iommu_fixed_ops.set_dma_mask = dma_set_mask_and_switch;
-
 	dma_iommu_ops.set_dma_mask = dma_set_mask_and_switch;
 	set_pci_dma_ops(&dma_iommu_ops);
 
@@ -1052,6 +1153,9 @@ static int __init setup_iommu_fixed(char *str)
 	if (strcmp(str, "off") == 0)
 		iommu_fixed_disabled = 1;
 
+	else if (strcmp(str, "weak") == 0)
+		iommu_fixed_is_weak = 1;
+
 	return 1;
 }
 __setup("iommu_fixed=", setup_iommu_fixed);
diff --git a/arch/powerpc/platforms/cell/pervasive.c b/arch/powerpc/platforms/cell/pervasive.c
index 8a3631ce912..efdacc82957 100644
--- a/arch/powerpc/platforms/cell/pervasive.c
+++ b/arch/powerpc/platforms/cell/pervasive.c
@@ -38,8 +38,6 @@
 
 #include "pervasive.h"
 
-static int sysreset_hack;
-
 static void cbe_power_save(void)
 {
 	unsigned long ctrl, thread_switch_control;
@@ -87,9 +85,6 @@ static void cbe_power_save(void)
 
 static int cbe_system_reset_exception(struct pt_regs *regs)
 {
-	int cpu;
-	struct cbe_pmd_regs __iomem *pmd;
-
 	switch (regs->msr & SRR1_WAKEMASK) {
 	case SRR1_WAKEEE:
 		do_IRQ(regs);
@@ -98,19 +93,7 @@ static int cbe_system_reset_exception(struct pt_regs *regs)
 		timer_interrupt(regs);
 		break;
 	case SRR1_WAKEMT:
-		/*
-		 * The BMC can inject user triggered system reset exceptions,
-		 * but cannot set the system reset reason in srr1,
-		 * so check an extra register here.
-		 */
-		if (sysreset_hack && (cpu = smp_processor_id()) == 0) {
-			pmd = cbe_get_cpu_pmd_regs(cpu);
-			if (in_be64(&pmd->ras_esc_0) & 0xffff) {
-				out_be64(&pmd->ras_esc_0, 0);
-				return 0;
-			}
-		}
-		break;
+		return cbe_sysreset_hack();
 #ifdef CONFIG_CBE_RAS
 	case SRR1_WAKESYSERR:
 		cbe_system_error_exception(regs);
@@ -134,8 +117,6 @@ void __init cbe_pervasive_init(void)
 	if (!cpu_has_feature(CPU_FTR_PAUSE_ZERO))
 		return;
 
-	sysreset_hack = machine_is_compatible("IBM,CBPLUS-1.0");
-
 	for_each_possible_cpu(cpu) {
 		struct cbe_pmd_regs __iomem *regs = cbe_get_cpu_pmd_regs(cpu);
 		if (!regs)
@@ -144,12 +125,6 @@ void __init cbe_pervasive_init(void)
 		 /* Enable Pause(0) control bit */
 		out_be64(&regs->pmcr, in_be64(&regs->pmcr) |
 					    CBE_PMD_PAUSE_ZERO_CONTROL);
-
-		/* Enable JTAG system-reset hack */
-		if (sysreset_hack)
-			out_be32(&regs->fir_mode_reg,
-				in_be32(&regs->fir_mode_reg) |
-				CBE_PMD_FIR_MODE_M8);
 	}
 
 	ppc_md.power_save = cbe_power_save;
diff --git a/arch/powerpc/platforms/cell/pervasive.h b/arch/powerpc/platforms/cell/pervasive.h
index 7b50947f804..fd4d7b7092b 100644
--- a/arch/powerpc/platforms/cell/pervasive.h
+++ b/arch/powerpc/platforms/cell/pervasive.h
@@ -30,4 +30,13 @@ extern void cbe_system_error_exception(struct pt_regs *regs);
 extern void cbe_maintenance_exception(struct pt_regs *regs);
 extern void cbe_thermal_exception(struct pt_regs *regs);
 
+#ifdef CONFIG_PPC_IBM_CELL_RESETBUTTON
+extern int cbe_sysreset_hack(void);
+#else
+static inline int cbe_sysreset_hack(void)
+{
+	return 1;
+}
+#endif /* CONFIG_PPC_IBM_CELL_RESETBUTTON */
+
 #endif
diff --git a/arch/powerpc/platforms/cell/ras.c b/arch/powerpc/platforms/cell/ras.c
index 505f9b9bdf0..2a14b052abc 100644
--- a/arch/powerpc/platforms/cell/ras.c
+++ b/arch/powerpc/platforms/cell/ras.c
@@ -236,6 +236,52 @@ static struct notifier_block cbe_ptcal_reboot_notifier = {
 	.notifier_call = cbe_ptcal_notify_reboot
 };
 
+#ifdef CONFIG_PPC_IBM_CELL_RESETBUTTON
+static int sysreset_hack;
+
+static int __init cbe_sysreset_init(void)
+{
+	struct cbe_pmd_regs __iomem *regs;
+
+	sysreset_hack = machine_is_compatible("IBM,CBPLUS-1.0");
+	if (!sysreset_hack)
+		return 0;
+
+	regs = cbe_get_cpu_pmd_regs(0);
+	if (!regs)
+		return 0;
+
+	/* Enable JTAG system-reset hack */
+	out_be32(&regs->fir_mode_reg,
+		in_be32(&regs->fir_mode_reg) |
+		CBE_PMD_FIR_MODE_M8);
+
+	return 0;
+}
+device_initcall(cbe_sysreset_init);
+
+int cbe_sysreset_hack(void)
+{
+	struct cbe_pmd_regs __iomem *regs;
+
+	/*
+	 * The BMC can inject user triggered system reset exceptions,
+	 * but cannot set the system reset reason in srr1,
+	 * so check an extra register here.
+	 */
+	if (sysreset_hack && (smp_processor_id() == 0)) {
+		regs = cbe_get_cpu_pmd_regs(0);
+		if (!regs)
+			return 0;
+		if (in_be64(&regs->ras_esc_0) & 0x0000ffff) {
+			out_be64(&regs->ras_esc_0, 0);
+			return 0;
+		}
+	}
+	return 1;
+}
+#endif /* CONFIG_PPC_IBM_CELL_RESETBUTTON */
+
 int __init cbe_ptcal_init(void)
 {
 	int ret;