powerpc: Merge rtas.c into arch/powerpc/kernel

This splits arch/ppc64/kernel/rtas.c into arch/powerpc/kernel/rtas.c,
which contains generic RTAS functions useful on any CHRP platform,
and arch/powerpc/platforms/pseries/rtas-fw.[ch], which contain
some pSeries-specific firmware flashing bits.  The parts of rtas.c
that are to do with pSeries-specific error logging are protected
by a new CONFIG_RTAS_ERROR_LOGGING symbol.  The inclusion of rtas.o
is controlled by the CONFIG_PPC_RTAS symbol, and the relevant
platforms select that.

Signed-off-by: Paul Mackerras <paulus@samba.org>

11 files changed, 850 insertions, 23 deletions

diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 9f279e0d84f2..ce5100cadd34 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -276,6 +276,8 @@ config PPC_PSERIES
 	depends on PPC_MULTIPLATFORM && PPC64
 	bool "  IBM pSeries & new (POWER5-based) iSeries"
 	select PPC_I8259
+	select PPC_RTAS
+	select RTAS_ERROR_LOGGING
 	default y
 
 config PPC_CHRP
@@ -283,6 +285,7 @@ config PPC_CHRP
 	depends on PPC_MULTIPLATFORM && PPC32
 	select PPC_I8259
 	select PPC_INDIRECT_PCI
+	select PPC_RTAS
 	default y
 
 config PPC_PMAC
@@ -317,6 +320,7 @@ config PPC_MAPLE
 config PPC_BPA
 	bool "  Broadband Processor Architecture"
 	depends on PPC_MULTIPLATFORM && PPC64
+	select PPC_RTAS
 
 config PPC_OF
 	bool
@@ -338,6 +342,15 @@ config MPIC
 	bool
 	default y
 
+config PPC_RTAS
+	bool
+	default n
+
+config RTAS_ERROR_LOGGING
+	bool
+	depends on PPC_RTAS
+	default n
+
 config MPIC_BROKEN_U3
 	bool
 	depends on PPC_MAPLE
diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile
index f5d8038fed52..eaedb7e63315 100644
--- a/arch/powerpc/kernel/Makefile
+++ b/arch/powerpc/kernel/Makefile
@@ -16,6 +16,7 @@ obj-$(CONFIG_PPC64)		+= binfmt_elf32.o sys_ppc32.o ptrace32.o
 obj-$(CONFIG_ALTIVEC)		+= vecemu.o vector.o
 obj-$(CONFIG_POWER4)		+= idle_power4.o
 obj-$(CONFIG_PPC_OF)		+= of_device.o
+obj-$(CONFIG_PPC_RTAS)		+= rtas.o
 obj-$(CONFIG_IBMVIO)		+= vio.o
 
 ifeq ($(CONFIG_PPC_MERGE),y)
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 1c83abd9f37c..710336a9f0fd 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -36,11 +36,11 @@
 #include <asm/processor.h>
 #include <asm/cputable.h>
 #include <asm/thread_info.h>
+#include <asm/rtas.h>
 #ifdef CONFIG_PPC64
 #include <asm/paca.h>
 #include <asm/lppaca.h>
 #include <asm/iSeries/HvLpEvent.h>
-#include <asm/rtas.h>
 #include <asm/cache.h>
 #include <asm/systemcfg.h>
 #include <asm/compat.h>
@@ -97,7 +97,7 @@ int main(void)
 	DEFINE(TI_TASK, offsetof(struct thread_info, task));
 	DEFINE(TI_EXECDOMAIN, offsetof(struct thread_info, exec_domain));
 	DEFINE(TI_CPU, offsetof(struct thread_info, cpu));
-#endif /* CONFIG_PPC64 */
+#endif /* CONFIG_PPC32 */
 
 #ifdef CONFIG_PPC64
 	DEFINE(DCACHEL1LINESIZE, offsetof(struct ppc64_caches, dline_size));
@@ -142,11 +142,11 @@ int main(void)
 	DEFINE(LPPACASRR1, offsetof(struct lppaca, saved_srr1));
 	DEFINE(LPPACAANYINT, offsetof(struct lppaca, int_dword.any_int));
 	DEFINE(LPPACADECRINT, offsetof(struct lppaca, int_dword.fields.decr_int));
+#endif /* CONFIG_PPC64 */
 
 	/* RTAS */
 	DEFINE(RTASBASE, offsetof(struct rtas_t, base));
 	DEFINE(RTASENTRY, offsetof(struct rtas_t, entry));
-#endif /* CONFIG_PPC64 */
 
 	/* Interrupt register frame */
 	DEFINE(STACK_FRAME_OVERHEAD, STACK_FRAME_OVERHEAD);
diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S
index 37b4396ca978..960da7bea043 100644
--- a/arch/powerpc/kernel/entry_32.S
+++ b/arch/powerpc/kernel/entry_32.S
@@ -954,7 +954,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_601)
  * here so it's easy to add arch-specific sections later.
  * -- Cort
  */
-#ifdef CONFIG_PPC_OF
+#ifdef CONFIG_PPC_RTAS
 /*
  * On CHRP, the Run-Time Abstraction Services (RTAS) have to be
  * called with the MMU off.
@@ -963,14 +963,13 @@ _GLOBAL(enter_rtas)
 	stwu	r1,-INT_FRAME_SIZE(r1)
 	mflr	r0
 	stw	r0,INT_FRAME_SIZE+4(r1)
-	lis	r4,rtas_data@ha
-	lwz	r4,rtas_data@l(r4)
+	LOADADDR(r4, rtas)
 	lis	r6,1f@ha	/* physical return address for rtas */
 	addi	r6,r6,1f@l
 	tophys(r6,r6)
 	tophys(r7,r1)
-	lis	r8,rtas_entry@ha
-	lwz	r8,rtas_entry@l(r8)
+	lwz	r8,RTASENTRY(r4)
+	lwz	r4,RTASBASE(r4)
 	mfmsr	r9
 	stw	r9,8(r1)
 	LOAD_MSR_KERNEL(r0,MSR_KERNEL)
@@ -978,7 +977,6 @@ _GLOBAL(enter_rtas)
 	MTMSRD(r0)		/* don't get trashed */
 	li	r9,MSR_KERNEL & ~(MSR_IR|MSR_DR)
 	mtlr	r6
-	CLR_TOP32(r7)
 	mtspr	SPRN_SPRG2,r7
 	mtspr	SPRN_SRR0,r8
 	mtspr	SPRN_SRR1,r9
@@ -999,4 +997,4 @@ machine_check_in_rtas:
 	twi	31,0,0
 	/* XXX load up BATs and panic */
 
-#endif /* CONFIG_PPC_OF */
+#endif /* CONFIG_PPC_RTAS */
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 8d0c78cbc0bc..16ac15e73b44 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -110,9 +110,6 @@ struct device_node *of_chosen;
 struct device_node *dflt_interrupt_controller;
 int num_interrupt_controllers;
 
-u32 rtas_data;
-u32 rtas_entry;
-
 /*
  * Wrapper for allocating memory for various data that needs to be
  * attached to device nodes as they are processed at boot or when
diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c
new file mode 100644
index 000000000000..4d22eeeeb91d
--- /dev/null
+++ b/arch/powerpc/kernel/rtas.c
@@ -0,0 +1,680 @@
+/*
+ *
+ * Procedures for interfacing to the RTAS on CHRP machines.
+ *
+ * Peter Bergner, IBM	March 2001.
+ * Copyright (C) 2001 IBM.
+ *
+ *      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.
+ */
+
+#include <stdarg.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <asm/prom.h>
+#include <asm/rtas.h>
+#include <asm/semaphore.h>
+#include <asm/machdep.h>
+#include <asm/page.h>
+#include <asm/param.h>
+#include <asm/system.h>
+#include <asm/delay.h>
+#include <asm/uaccess.h>
+#include <asm/lmb.h>
+#ifdef CONFIG_PPC64
+#include <asm/systemcfg.h>
+#endif
+
+struct rtas_t rtas = {
+	.lock = SPIN_LOCK_UNLOCKED
+};
+
+EXPORT_SYMBOL(rtas);
+
+DEFINE_SPINLOCK(rtas_data_buf_lock);
+char rtas_data_buf[RTAS_DATA_BUF_SIZE] __cacheline_aligned;
+unsigned long rtas_rmo_buf;
+
+/*
+ * call_rtas_display_status and call_rtas_display_status_delay
+ * are designed only for very early low-level debugging, which
+ * is why the token is hard-coded to 10.
+ */
+void call_rtas_display_status(unsigned char c)
+{
+	struct rtas_args *args = &rtas.args;
+	unsigned long s;
+
+	if (!rtas.base)
+		return;
+	spin_lock_irqsave(&rtas.lock, s);
+
+	args->token = 10;
+	args->nargs = 1;
+	args->nret  = 1;
+	args->rets  = (rtas_arg_t *)&(args->args[1]);
+	args->args[0] = (int)c;
+
+	enter_rtas(__pa(args));
+
+	spin_unlock_irqrestore(&rtas.lock, s);
+}
+
+void call_rtas_display_status_delay(unsigned char c)
+{
+	static int pending_newline = 0;  /* did last write end with unprinted newline? */
+	static int width = 16;
+
+	if (c == '\n') {	
+		while (width-- > 0)
+			call_rtas_display_status(' ');
+		width = 16;
+		udelay(500000);
+		pending_newline = 1;
+	} else {
+		if (pending_newline) {
+			call_rtas_display_status('\r');
+			call_rtas_display_status('\n');
+		} 
+		pending_newline = 0;
+		if (width--) {
+			call_rtas_display_status(c);
+			udelay(10000);
+		}
+	}
+}
+
+void rtas_progress(char *s, unsigned short hex)
+{
+	struct device_node *root;
+	int width, *p;
+	char *os;
+	static int display_character, set_indicator;
+	static int display_width, display_lines, *row_width, form_feed;
+	static DEFINE_SPINLOCK(progress_lock);
+	static int current_line;
+	static int pending_newline = 0;  /* did last write end with unprinted newline? */
+
+	if (!rtas.base)
+		return;
+
+	if (display_width == 0) {
+		display_width = 0x10;
+		if ((root = find_path_device("/rtas"))) {
+			if ((p = (unsigned int *)get_property(root,
+					"ibm,display-line-length", NULL)))
+				display_width = *p;
+			if ((p = (unsigned int *)get_property(root,
+					"ibm,form-feed", NULL)))
+				form_feed = *p;
+			if ((p = (unsigned int *)get_property(root,
+					"ibm,display-number-of-lines", NULL)))
+				display_lines = *p;
+			row_width = (unsigned int *)get_property(root,
+					"ibm,display-truncation-length", NULL);
+		}
+		display_character = rtas_token("display-character");
+		set_indicator = rtas_token("set-indicator");
+	}
+
+	if (display_character == RTAS_UNKNOWN_SERVICE) {
+		/* use hex display if available */
+		if (set_indicator != RTAS_UNKNOWN_SERVICE)
+			rtas_call(set_indicator, 3, 1, NULL, 6, 0, hex);
+		return;
+	}
+
+	spin_lock(&progress_lock);
+
+	/*
+	 * Last write ended with newline, but we didn't print it since
+	 * it would just clear the bottom line of output. Print it now
+	 * instead.
+	 *
+	 * If no newline is pending and form feed is supported, clear the
+	 * display with a form feed; otherwise, print a CR to start output
+	 * at the beginning of the line.
+	 */
+	if (pending_newline) {
+		rtas_call(display_character, 1, 1, NULL, '\r');
+		rtas_call(display_character, 1, 1, NULL, '\n');
+		pending_newline = 0;
+	} else {
+		current_line = 0;
+		if (form_feed)
+			rtas_call(display_character, 1, 1, NULL,
+				  (char)form_feed);
+		else
+			rtas_call(display_character, 1, 1, NULL, '\r');
+	}
+ 
+	if (row_width)
+		width = row_width[current_line];
+	else
+		width = display_width;
+	os = s;
+	while (*os) {
+		if (*os == '\n' || *os == '\r') {
+			/* If newline is the last character, save it
+			 * until next call to avoid bumping up the
+			 * display output.
+			 */
+			if (*os == '\n' && !os[1]) {
+				pending_newline = 1;
+				current_line++;
+				if (current_line > display_lines-1)
+					current_line = display_lines-1;
+				spin_unlock(&progress_lock);
+				return;
+			}
+ 
+			/* RTAS wants CR-LF, not just LF */
+ 
+			if (*os == '\n') {
+				rtas_call(display_character, 1, 1, NULL, '\r');
+				rtas_call(display_character, 1, 1, NULL, '\n');
+			} else {
+				/* CR might be used to re-draw a line, so we'll
+				 * leave it alone and not add LF.
+				 */
+				rtas_call(display_character, 1, 1, NULL, *os);
+			}
+ 
+			if (row_width)
+				width = row_width[current_line];
+			else
+				width = display_width;
+		} else {
+			width--;
+			rtas_call(display_character, 1, 1, NULL, *os);
+		}
+ 
+		os++;
+ 
+		/* if we overwrite the screen length */
+		if (width <= 0)
+			while ((*os != 0) && (*os != '\n') && (*os != '\r'))
+				os++;
+	}
+ 
+	spin_unlock(&progress_lock);
+}
+
+int rtas_token(const char *service)
+{
+	int *tokp;
+	if (rtas.dev == NULL)
+		return RTAS_UNKNOWN_SERVICE;
+	tokp = (int *) get_property(rtas.dev, service, NULL);
+	return tokp ? *tokp : RTAS_UNKNOWN_SERVICE;
+}
+
+#ifdef CONFIG_RTAS_ERROR_LOGGING
+/*
+ * Return the firmware-specified size of the error log buffer
+ *  for all rtas calls that require an error buffer argument.
+ *  This includes 'check-exception' and 'rtas-last-error'.
+ */
+int rtas_get_error_log_max(void)
+{
+	static int rtas_error_log_max;
+	if (rtas_error_log_max)
+		return rtas_error_log_max;
+
+	rtas_error_log_max = rtas_token ("rtas-error-log-max");
+	if ((rtas_error_log_max == RTAS_UNKNOWN_SERVICE) ||
+	    (rtas_error_log_max > RTAS_ERROR_LOG_MAX)) {
+		printk (KERN_WARNING "RTAS: bad log buffer size %d\n",
+			rtas_error_log_max);
+		rtas_error_log_max = RTAS_ERROR_LOG_MAX;
+	}
+	return rtas_error_log_max;
+}
+EXPORT_SYMBOL(rtas_get_error_log_max);
+
+
+char rtas_err_buf[RTAS_ERROR_LOG_MAX];
+int rtas_last_error_token;
+
+/** Return a copy of the detailed error text associated with the
+ *  most recent failed call to rtas.  Because the error text
+ *  might go stale if there are any other intervening rtas calls,
+ *  this routine must be called atomically with whatever produced
+ *  the error (i.e. with rtas.lock still held from the previous call).
+ */
+static char *__fetch_rtas_last_error(char *altbuf)
+{
+	struct rtas_args err_args, save_args;
+	u32 bufsz;
+	char *buf = NULL;
+
+	if (rtas_last_error_token == -1)
+		return NULL;
+
+	bufsz = rtas_get_error_log_max();
+
+	err_args.token = rtas_last_error_token;
+	err_args.nargs = 2;
+	err_args.nret = 1;
+	err_args.args[0] = (rtas_arg_t)__pa(rtas_err_buf);
+	err_args.args[1] = bufsz;
+	err_args.args[2] = 0;
+
+	save_args = rtas.args;
+	rtas.args = err_args;
+
+	enter_rtas(__pa(&rtas.args));
+
+	err_args = rtas.args;
+	rtas.args = save_args;
+
+	/* Log the error in the unlikely case that there was one. */
+	if (unlikely(err_args.args[2] == 0)) {
+		if (altbuf) {
+			buf = altbuf;
+		} else {
+			buf = rtas_err_buf;
+			if (mem_init_done)
+				buf = kmalloc(RTAS_ERROR_LOG_MAX, GFP_ATOMIC);
+		}
+		if (buf)
+			memcpy(buf, rtas_err_buf, RTAS_ERROR_LOG_MAX);
+	}
+
+	return buf;
+}
+
+#define get_errorlog_buffer()	kmalloc(RTAS_ERROR_LOG_MAX, GFP_KERNEL)
+
+#else /* CONFIG_RTAS_ERROR_LOGGING */
+#define __fetch_rtas_last_error(x)	NULL
+#define get_errorlog_buffer()		NULL
+#endif
+
+int rtas_call(int token, int nargs, int nret, int *outputs, ...)
+{
+	va_list list;
+	int i;
+	unsigned long s;
+	struct rtas_args *rtas_args;
+	char *buff_copy = NULL;
+	int ret;
+
+	if (token == RTAS_UNKNOWN_SERVICE)
+		return -1;
+
+	/* Gotta do something different here, use global lock for now... */
+	spin_lock_irqsave(&rtas.lock, s);
+	rtas_args = &rtas.args;
+
+	rtas_args->token = token;
+	rtas_args->nargs = nargs;
+	rtas_args->nret  = nret;
+	rtas_args->rets  = (rtas_arg_t *)&(rtas_args->args[nargs]);
+	va_start(list, outputs);
+	for (i = 0; i < nargs; ++i)
+		rtas_args->args[i] = va_arg(list, rtas_arg_t);
+	va_end(list);
+
+	for (i = 0; i < nret; ++i)
+		rtas_args->rets[i] = 0;
+
+	enter_rtas(__pa(rtas_args));
+
+	/* A -1 return code indicates that the last command couldn't
+	   be completed due to a hardware error. */
+	if (rtas_args->rets[0] == -1)
+		buff_copy = __fetch_rtas_last_error(NULL);
+
+	if (nret > 1 && outputs != NULL)
+		for (i = 0; i < nret-1; ++i)
+			outputs[i] = rtas_args->rets[i+1];
+	ret = (nret > 0)? rtas_args->rets[0]: 0;
+
+	/* Gotta do something different here, use global lock for now... */
+	spin_unlock_irqrestore(&rtas.lock, s);
+
+	if (buff_copy) {
+		log_error(buff_copy, ERR_TYPE_RTAS_LOG, 0);
+		if (mem_init_done)
+			kfree(buff_copy);
+	}
+	return ret;
+}
+
+/* Given an RTAS status code of 990n compute the hinted delay of 10^n
+ * (last digit) milliseconds.  For now we bound at n=5 (100 sec).
+ */
+unsigned int rtas_extended_busy_delay_time(int status)
+{
+	int order = status - 9900;
+	unsigned long ms;
+
+	if (order < 0)
+		order = 0;	/* RTC depends on this for -2 clock busy */
+	else if (order > 5)
+		order = 5;	/* bound */
+
+	/* Use microseconds for reasonable accuracy */
+	for (ms = 1; order > 0; order--)
+		ms *= 10;
+
+	return ms; 
+}
+
+int rtas_error_rc(int rtas_rc)
+{
+	int rc;
+
+	switch (rtas_rc) {
+		case -1: 		/* Hardware Error */
+			rc = -EIO;
+			break;
+		case -3:		/* Bad indicator/domain/etc */
+			rc = -EINVAL;
+			break;
+		case -9000:		/* Isolation error */
+			rc = -EFAULT;
+			break;
+		case -9001:		/* Outstanding TCE/PTE */
+			rc = -EEXIST;
+			break;
+		case -9002:		/* No usable slot */
+			rc = -ENODEV;
+			break;
+		default:
+			printk(KERN_ERR "%s: unexpected RTAS error %d\n",
+					__FUNCTION__, rtas_rc);
+			rc = -ERANGE;
+			break;
+	}
+	return rc;
+}
+
+int rtas_get_power_level(int powerdomain, int *level)
+{
+	int token = rtas_token("get-power-level");
+	int rc;
+
+	if (token == RTAS_UNKNOWN_SERVICE)
+		return -ENOENT;
+
+	while ((rc = rtas_call(token, 1, 2, level, powerdomain)) == RTAS_BUSY)
+		udelay(1);
+
+	if (rc < 0)
+		return rtas_error_rc(rc);
+	return rc;
+}
+
+int rtas_set_power_level(int powerdomain, int level, int *setlevel)
+{
+	int token = rtas_token("set-power-level");
+	unsigned int wait_time;
+	int rc;
+
+	if (token == RTAS_UNKNOWN_SERVICE)
+		return -ENOENT;
+
+	while (1) {
+		rc = rtas_call(token, 2, 2, setlevel, powerdomain, level);
+		if (rc == RTAS_BUSY)
+			udelay(1);
+		else if (rtas_is_extended_busy(rc)) {
+			wait_time = rtas_extended_busy_delay_time(rc);
+			udelay(wait_time * 1000);
+		} else
+			break;
+	}
+
+	if (rc < 0)
+		return rtas_error_rc(rc);
+	return rc;
+}
+
+int rtas_get_sensor(int sensor, int index, int *state)
+{
+	int token = rtas_token("get-sensor-state");
+	unsigned int wait_time;
+	int rc;
+
+	if (token == RTAS_UNKNOWN_SERVICE)
+		return -ENOENT;
+
+	while (1) {
+		rc = rtas_call(token, 2, 2, state, sensor, index);
+		if (rc == RTAS_BUSY)
+			udelay(1);
+		else if (rtas_is_extended_busy(rc)) {
+			wait_time = rtas_extended_busy_delay_time(rc);
+			udelay(wait_time * 1000);
+		} else
+			break;
+	}
+
+	if (rc < 0)
+		return rtas_error_rc(rc);
+	return rc;
+}
+
+int rtas_set_indicator(int indicator, int index, int new_value)
+{
+	int token = rtas_token("set-indicator");
+	unsigned int wait_time;
+	int rc;
+
+	if (token == RTAS_UNKNOWN_SERVICE)
+		return -ENOENT;
+
+	while (1) {
+		rc = rtas_call(token, 3, 1, NULL, indicator, index, new_value);
+		if (rc == RTAS_BUSY)
+			udelay(1);
+		else if (rtas_is_extended_busy(rc)) {
+			wait_time = rtas_extended_busy_delay_time(rc);
+			udelay(wait_time * 1000);
+		}
+		else
+			break;
+	}
+
+	if (rc < 0)
+		return rtas_error_rc(rc);
+	return rc;
+}
+
+void rtas_restart(char *cmd)
+{
+	printk("RTAS system-reboot returned %d\n",
+	       rtas_call(rtas_token("system-reboot"), 0, 1, NULL));
+	for (;;);
+}
+
+void rtas_power_off(void)
+{
+	/* allow power on only with power button press */
+	printk("RTAS power-off returned %d\n",
+	       rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1));
+	for (;;);
+}
+
+void rtas_halt(void)
+{
+	rtas_power_off();
+}
+
+/* Must be in the RMO region, so we place it here */
+static char rtas_os_term_buf[2048];
+
+void rtas_os_term(char *str)
+{
+	int status;
+
+	if (RTAS_UNKNOWN_SERVICE == rtas_token("ibm,os-term"))
+		return;
+
+	snprintf(rtas_os_term_buf, 2048, "OS panic: %s", str);
+
+	do {
+		status = rtas_call(rtas_token("ibm,os-term"), 1, 1, NULL,
+				   __pa(rtas_os_term_buf));
+
+		if (status == RTAS_BUSY)
+			udelay(1);
+		else if (status != 0)
+			printk(KERN_EMERG "ibm,os-term call failed %d\n",
+			       status);
+	} while (status == RTAS_BUSY);
+}
+
+
+asmlinkage int ppc_rtas(struct rtas_args __user *uargs)
+{
+	struct rtas_args args;
+	unsigned long flags;
+	char *buff_copy, *errbuf = NULL;
+	int nargs;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	if (copy_from_user(&args, uargs, 3 * sizeof(u32)) != 0)
+		return -EFAULT;
+
+	nargs = args.nargs;
+	if (nargs > ARRAY_SIZE(args.args)
+	    || args.nret > ARRAY_SIZE(args.args)
+	    || nargs + args.nret > ARRAY_SIZE(args.args))
+		return -EINVAL;
+
+	/* Copy in args. */
+	if (copy_from_user(args.args, uargs->args,
+			   nargs * sizeof(rtas_arg_t)) != 0)
+		return -EFAULT;
+
+	buff_copy = get_errorlog_buffer();
+
+	spin_lock_irqsave(&rtas.lock, flags);
+
+	rtas.args = args;
+	enter_rtas(__pa(&rtas.args));
+	args = rtas.args;
+
+	args.rets = &args.args[nargs];
+
+	/* A -1 return code indicates that the last command couldn't
+	   be completed due to a hardware error. */
+	if (args.rets[0] == -1)
+		errbuf = __fetch_rtas_last_error(buff_copy);
+
+	spin_unlock_irqrestore(&rtas.lock, flags);
+
+	if (buff_copy) {
+		if (errbuf)
+			log_error(errbuf, ERR_TYPE_RTAS_LOG, 0);
+		kfree(buff_copy);
+	}
+
+	/* Copy out args. */
+	if (copy_to_user(uargs->args + nargs,
+			 args.args + nargs,
+			 args.nret * sizeof(rtas_arg_t)) != 0)
+		return -EFAULT;
+
+	return 0;
+}
+
+#ifdef CONFIG_SMP
+/* This version can't take the spinlock, because it never returns */
+
+struct rtas_args rtas_stop_self_args = {
+	/* The token is initialized for real in setup_system() */
+	.token = RTAS_UNKNOWN_SERVICE,
+	.nargs = 0,
+	.nret = 1,
+	.rets = &rtas_stop_self_args.args[0],
+};
+
+void rtas_stop_self(void)
+{
+	struct rtas_args *rtas_args = &rtas_stop_self_args;
+
+	local_irq_disable();
+
+	BUG_ON(rtas_args->token == RTAS_UNKNOWN_SERVICE);
+
+	printk("cpu %u (hwid %u) Ready to die...\n",
+	       smp_processor_id(), hard_smp_processor_id());
+	enter_rtas(__pa(rtas_args));
+
+	panic("Alas, I survived.\n");
+}
+#endif
+
+/*
+ * Call early during boot, before mem init or bootmem, to retreive the RTAS
+ * informations from the device-tree and allocate the RMO buffer for userland
+ * accesses.
+ */
+void __init rtas_initialize(void)
+{
+	unsigned long rtas_region = RTAS_INSTANTIATE_MAX;
+
+	/* Get RTAS dev node and fill up our "rtas" structure with infos
+	 * about it.
+	 */
+	rtas.dev = of_find_node_by_name(NULL, "rtas");
+	if (rtas.dev) {
+		u32 *basep, *entryp;
+		u32 *sizep;
+
+		basep = (u32 *)get_property(rtas.dev, "linux,rtas-base", NULL);
+		sizep = (u32 *)get_property(rtas.dev, "rtas-size", NULL);
+		if (basep != NULL && sizep != NULL) {
+			rtas.base = *basep;
+			rtas.size = *sizep;
+			entryp = (u32 *)get_property(rtas.dev, "linux,rtas-entry", NULL);
+			if (entryp == NULL) /* Ugh */
+				rtas.entry = rtas.base;
+			else
+				rtas.entry = *entryp;
+		} else
+			rtas.dev = NULL;
+	}
+	if (!rtas.dev)
+		return;
+
+	/* If RTAS was found, allocate the RMO buffer for it and look for
+	 * the stop-self token if any
+	 */
+#ifdef CONFIG_PPC64
+	if (systemcfg->platform == PLATFORM_PSERIES_LPAR)
+		rtas_region = min(lmb.rmo_size, RTAS_INSTANTIATE_MAX);
+#endif
+	rtas_rmo_buf = lmb_alloc_base(RTAS_RMOBUF_MAX, PAGE_SIZE, rtas_region);
+
+#ifdef CONFIG_HOTPLUG_CPU
+	rtas_stop_self_args.token = rtas_token("stop-self");
+#endif /* CONFIG_HOTPLUG_CPU */
+#ifdef CONFIG_RTAS_ERROR_LOGGING
+	rtas_last_error_token = rtas_token("rtas-last-error");
+#endif
+}
+
+
+EXPORT_SYMBOL(rtas_token);
+EXPORT_SYMBOL(rtas_call);
+EXPORT_SYMBOL(rtas_data_buf);
+EXPORT_SYMBOL(rtas_data_buf_lock);
+EXPORT_SYMBOL(rtas_extended_busy_delay_time);
+EXPORT_SYMBOL(rtas_get_sensor);
+EXPORT_SYMBOL(rtas_get_power_level);
+EXPORT_SYMBOL(rtas_set_power_level);
+EXPORT_SYMBOL(rtas_set_indicator);
diff --git a/arch/powerpc/platforms/pseries/Kconfig b/arch/powerpc/platforms/pseries/Kconfig
index 7a3b6fc4d976..2d57f588151d 100644
--- a/arch/powerpc/platforms/pseries/Kconfig
+++ b/arch/powerpc/platforms/pseries/Kconfig
@@ -21,11 +21,6 @@ config EEH
 	depends on PPC_PSERIES
 	default y if !EMBEDDED
 
-config PPC_RTAS
-	bool
-	depends on PPC_PSERIES || PPC_BPA
-	default y
-
 config RTAS_PROC
 	bool "Proc interface to RTAS"
 	depends on PPC_RTAS
diff --git a/arch/powerpc/platforms/pseries/Makefile b/arch/powerpc/platforms/pseries/Makefile
index 4c817304fc21..d5c160b789e3 100644
--- a/arch/powerpc/platforms/pseries/Makefile
+++ b/arch/powerpc/platforms/pseries/Makefile
@@ -1,4 +1,4 @@
 obj-y			:= pci.o lpar.o hvCall.o nvram.o reconfig.o \
-			   setup.o iommu.o ras.o
+			   setup.o iommu.o rtas-fw.o ras.o
 obj-$(CONFIG_SMP)	+= smp.o
 obj-$(CONFIG_IBMVIO)	+= vio.o
diff --git a/arch/powerpc/platforms/pseries/rtas-fw.c b/arch/powerpc/platforms/pseries/rtas-fw.c
new file mode 100644
index 000000000000..15d81d758ca0
--- /dev/null
+++ b/arch/powerpc/platforms/pseries/rtas-fw.c
@@ -0,0 +1,138 @@
+/*
+ *
+ * Procedures for firmware flash updates on pSeries systems.
+ *
+ * Peter Bergner, IBM	March 2001.
+ * Copyright (C) 2001 IBM.
+ *
+ *      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.
+ */
+
+#include <stdarg.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <asm/prom.h>
+#include <asm/rtas.h>
+#include <asm/semaphore.h>
+#include <asm/machdep.h>
+#include <asm/page.h>
+#include <asm/param.h>
+#include <asm/system.h>
+#include <asm/abs_addr.h>
+#include <asm/udbg.h>
+#include <asm/delay.h>
+#include <asm/uaccess.h>
+#include <asm/systemcfg.h>
+
+#include "rtas-fw.h"
+
+struct flash_block_list_header rtas_firmware_flash_list = {0, NULL};
+
+#define FLASH_BLOCK_LIST_VERSION (1UL)
+
+static void rtas_flash_firmware(void)
+{
+	unsigned long image_size;
+	struct flash_block_list *f, *next, *flist;
+	unsigned long rtas_block_list;
+	int i, status, update_token;
+
+	update_token = rtas_token("ibm,update-flash-64-and-reboot");
+	if (update_token == RTAS_UNKNOWN_SERVICE) {
+		printk(KERN_ALERT "FLASH: ibm,update-flash-64-and-reboot is not available -- not a service partition?\n");
+		printk(KERN_ALERT "FLASH: firmware will not be flashed\n");
+		return;
+	}
+
+	/* NOTE: the "first" block list is a global var with no data
+	 * blocks in the kernel data segment.  We do this because
+	 * we want to ensure this block_list addr is under 4GB.
+	 */
+	rtas_firmware_flash_list.num_blocks = 0;
+	flist = (struct flash_block_list *)&rtas_firmware_flash_list;
+	rtas_block_list = virt_to_abs(flist);
+	if (rtas_block_list >= 4UL*1024*1024*1024) {
+		printk(KERN_ALERT "FLASH: kernel bug...flash list header addr above 4GB\n");
+		return;
+	}
+
+	printk(KERN_ALERT "FLASH: preparing saved firmware image for flash\n");
+	/* Update the block_list in place. */
+	image_size = 0;
+	for (f = flist; f; f = next) {
+		/* Translate data addrs to absolute */
+		for (i = 0; i < f->num_blocks; i++) {
+			f->blocks[i].data = (char *)virt_to_abs(f->blocks[i].data);
+			image_size += f->blocks[i].length;
+		}
+		next = f->next;
+		/* Don't translate NULL pointer for last entry */
+		if (f->next)
+			f->next = (struct flash_block_list *)virt_to_abs(f->next);
+		else
+			f->next = NULL;
+		/* make num_blocks into the version/length field */
+		f->num_blocks = (FLASH_BLOCK_LIST_VERSION << 56) | ((f->num_blocks+1)*16);
+	}
+
+	printk(KERN_ALERT "FLASH: flash image is %ld bytes\n", image_size);
+	printk(KERN_ALERT "FLASH: performing flash and reboot\n");
+	rtas_progress("Flashing        \n", 0x0);
+	rtas_progress("Please Wait...  ", 0x0);
+	printk(KERN_ALERT "FLASH: this will take several minutes.  Do not power off!\n");
+	status = rtas_call(update_token, 1, 1, NULL, rtas_block_list);
+	switch (status) {	/* should only get "bad" status */
+	    case 0:
+		printk(KERN_ALERT "FLASH: success\n");
+		break;
+	    case -1:
+		printk(KERN_ALERT "FLASH: hardware error.  Firmware may not be not flashed\n");
+		break;
+	    case -3:
+		printk(KERN_ALERT "FLASH: image is corrupt or not correct for this platform.  Firmware not flashed\n");
+		break;
+	    case -4:
+		printk(KERN_ALERT "FLASH: flash failed when partially complete.  System may not reboot\n");
+		break;
+	    default:
+		printk(KERN_ALERT "FLASH: unknown flash return code %d\n", status);
+		break;
+	}
+}
+
+void rtas_flash_bypass_warning(void)
+{
+	printk(KERN_ALERT "FLASH: firmware flash requires a reboot\n");
+	printk(KERN_ALERT "FLASH: the firmware image will NOT be flashed\n");
+}
+
+
+void rtas_fw_restart(char *cmd)
+{
+	if (rtas_firmware_flash_list.next)
+		rtas_flash_firmware();
+	rtas_restart(cmd);
+}
+
+void rtas_fw_power_off(void)
+{
+	if (rtas_firmware_flash_list.next)
+		rtas_flash_bypass_warning();
+	rtas_power_off();
+}
+
+void rtas_fw_halt(void)
+{
+	if (rtas_firmware_flash_list.next)
+		rtas_flash_bypass_warning();
+	rtas_halt();
+}
+
+EXPORT_SYMBOL(rtas_firmware_flash_list);
diff --git a/arch/powerpc/platforms/pseries/rtas-fw.h b/arch/powerpc/platforms/pseries/rtas-fw.h
new file mode 100644
index 000000000000..e70fa69974a3
--- /dev/null
+++ b/arch/powerpc/platforms/pseries/rtas-fw.h
@@ -0,0 +1,3 @@
+void rtas_fw_restart(char *cmd);
+void rtas_fw_power_off(void);
+void rtas_fw_halt(void);
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index 0fa5beae6d1b..7e7e556e6b48 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -1,5 +1,5 @@
 /*
- *  linux/arch/ppc/kernel/setup.c
+ *  64-bit pSeries and RS/6000 setup code.
  *
  *  Copyright (C) 1995  Linus Torvalds
  *  Adapted from 'alpha' version by Gary Thomas
@@ -67,6 +67,8 @@
 #include <asm/i8259.h>
 #include <asm/udbg.h>
 
+#include "rtas-fw.h"
+
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
 #else
@@ -589,9 +591,9 @@ struct machdep_calls __initdata pSeries_md = {
 	.pcibios_fixup		= pSeries_final_fixup,
 	.pci_probe_mode		= pSeries_pci_probe_mode,
 	.irq_bus_setup		= pSeries_irq_bus_setup,
-	.restart		= rtas_restart,
-	.power_off		= rtas_power_off,
-	.halt			= rtas_halt,
+	.restart		= rtas_fw_restart,
+	.power_off		= rtas_fw_power_off,
+	.halt			= rtas_fw_halt,
 	.panic			= rtas_os_term,
 	.cpu_die		= pSeries_mach_cpu_die,
 	.get_boot_time		= rtas_get_boot_time,