[deliverable/linux.git] / arch / powerpc / kernel / machine_kexec_64.c

/*
 * PPC64 code to handle Linux booting another kernel.
 *
 * Copyright (C) 2004-2005, IBM Corp.
 *
 * Created by: Milton D Miller II
 *
 * This source code is licensed under the GNU General Public License,
 * Version 2.  See the file COPYING for more details.
 */


#include <linux/kexec.h>
#include <linux/smp.h>
#include <linux/thread_info.h>
#include <linux/init_task.h>
#include <linux/errno.h>
#include <linux/kernel.h>
#include <linux/cpu.h>

#include <asm/page.h>
#include <asm/current.h>
#include <asm/machdep.h>
#include <asm/cacheflush.h>
#include <asm/paca.h>
#include <asm/mmu.h>
#include <asm/sections.h>	/* _end */
#include <asm/prom.h>
#include <asm/smp.h>
#include <asm/hw_breakpoint.h>

int default_machine_kexec_prepare(struct kimage *image)
{
	int i;
	unsigned long begin, end;	/* limits of segment */
	unsigned long low, high;	/* limits of blocked memory range */
	struct device_node *node;
	const unsigned long *basep;
	const unsigned int *sizep;

	if (!ppc_md.hpte_clear_all)
		return -ENOENT;

	/*
	 * Since we use the kernel fault handlers and paging code to
	 * handle the virtual mode, we must make sure no destination
	 * overlaps kernel static data or bss.
	 */
	for (i = 0; i < image->nr_segments; i++)
		if (image->segment[i].mem < __pa(_end))
			return -ETXTBSY;

	/*
	 * For non-LPAR, we absolutely can not overwrite the mmu hash
	 * table, since we are still using the bolted entries in it to
	 * do the copy.  Check that here.
	 *
	 * It is safe if the end is below the start of the blocked
	 * region (end <= low), or if the beginning is after the
	 * end of the blocked region (begin >= high).  Use the
	 * boolean identity !(a || b)  === (!a && !b).
	 */
	if (htab_address) {
		low = __pa(htab_address);
		high = low + htab_size_bytes;

		for (i = 0; i < image->nr_segments; i++) {
			begin = image->segment[i].mem;
			end = begin + image->segment[i].memsz;

			if ((begin < high) && (end > low))
				return -ETXTBSY;
		}
	}

	/* We also should not overwrite the tce tables */
	for (node = of_find_node_by_type(NULL, "pci"); node != NULL;
			node = of_find_node_by_type(node, "pci")) {
		basep = of_get_property(node, "linux,tce-base", NULL);
		sizep = of_get_property(node, "linux,tce-size", NULL);
		if (basep == NULL || sizep == NULL)
			continue;

		low = *basep;
		high = low + (*sizep);

		for (i = 0; i < image->nr_segments; i++) {
			begin = image->segment[i].mem;
			end = begin + image->segment[i].memsz;

			if ((begin < high) && (end > low))
				return -ETXTBSY;
		}
	}

	return 0;
}

#define IND_FLAGS (IND_DESTINATION | IND_INDIRECTION | IND_DONE | IND_SOURCE)

static void copy_segments(unsigned long ind)
{
	unsigned long entry;
	unsigned long *ptr;
	void *dest;
	void *addr;

	/*
	 * We rely on kexec_load to create a lists that properly
	 * initializes these pointers before they are used.
	 * We will still crash if the list is wrong, but at least
	 * the compiler will be quiet.
	 */
	ptr = NULL;
	dest = NULL;

	for (entry = ind; !(entry & IND_DONE); entry = *ptr++) {
		addr = __va(entry & PAGE_MASK);

		switch (entry & IND_FLAGS) {
		case IND_DESTINATION:
			dest = addr;
			break;
		case IND_INDIRECTION:
			ptr = addr;
			break;
		case IND_SOURCE:
			copy_page(dest, addr);
			dest += PAGE_SIZE;
		}
	}
}

void kexec_copy_flush(struct kimage *image)
{
	long i, nr_segments = image->nr_segments;
	struct  kexec_segment ranges[KEXEC_SEGMENT_MAX];

	/* save the ranges on the stack to efficiently flush the icache */
	memcpy(ranges, image->segment, sizeof(ranges));

	/*
	 * After this call we may not use anything allocated in dynamic
	 * memory, including *image.
	 *
	 * Only globals and the stack are allowed.
	 */
	copy_segments(image->head);

	/*
	 * we need to clear the icache for all dest pages sometime,
	 * including ones that were in place on the original copy
	 */
	for (i = 0; i < nr_segments; i++)
		flush_icache_range((unsigned long)__va(ranges[i].mem),
			(unsigned long)__va(ranges[i].mem + ranges[i].memsz));
}

#ifdef CONFIG_SMP

static int kexec_all_irq_disabled = 0;

static void kexec_smp_down(void *arg)
{
	local_irq_disable();
	mb(); /* make sure our irqs are disabled before we say they are */
	get_paca()->kexec_state = KEXEC_STATE_IRQS_OFF;
	while(kexec_all_irq_disabled == 0)
		cpu_relax();
	mb(); /* make sure all irqs are disabled before this */
	hw_breakpoint_disable();
	/*
	 * Now every CPU has IRQs off, we can clear out any pending
	 * IPIs and be sure that no more will come in after this.
	 */
	if (ppc_md.kexec_cpu_down)
		ppc_md.kexec_cpu_down(0, 1);

	kexec_smp_wait();
	/* NOTREACHED */
}

static void kexec_prepare_cpus_wait(int wait_state)
{
	int my_cpu, i, notified=-1;

	hw_breakpoint_disable();
	my_cpu = get_cpu();
	/* Make sure each CPU has at least made it to the state we need.
	 *
	 * FIXME: There is a (slim) chance of a problem if not all of the CPUs
	 * are correctly onlined.  If somehow we start a CPU on boot with RTAS
	 * start-cpu, but somehow that CPU doesn't write callin_cpu_map[] in
	 * time, the boot CPU will timeout.  If it does eventually execute
	 * stuff, the secondary will start up (paca[].cpu_start was written) and
	 * get into a peculiar state.  If the platform supports
	 * smp_ops->take_timebase(), the secondary CPU will probably be spinning
	 * in there.  If not (i.e. pseries), the secondary will continue on and
	 * try to online itself/idle/etc. If it survives that, we need to find
	 * these possible-but-not-online-but-should-be CPUs and chaperone them
	 * into kexec_smp_wait().
	 */
	for_each_online_cpu(i) {
		if (i == my_cpu)
			continue;

		while (paca[i].kexec_state < wait_state) {
			barrier();
			if (i != notified) {
				printk(KERN_INFO "kexec: waiting for cpu %d "
				       "(physical %d) to enter %i state\n",
				       i, paca[i].hw_cpu_id, wait_state);
				notified = i;
			}
		}
	}
	mb();
}

/*
 * We need to make sure each present CPU is online.  The next kernel will scan
 * the device tree and assume primary threads are online and query secondary
 * threads via RTAS to online them if required.  If we don't online primary
 * threads, they will be stuck.  However, we also online secondary threads as we
 * may be using 'cede offline'.  In this case RTAS doesn't see the secondary
 * threads as offline -- and again, these CPUs will be stuck.
 *
 * So, we online all CPUs that should be running, including secondary threads.
 */
static void wake_offline_cpus(void)
{
	int cpu = 0;

	for_each_present_cpu(cpu) {
		if (!cpu_online(cpu)) {
			printk(KERN_INFO "kexec: Waking offline cpu %d.\n",
			       cpu);
			cpu_up(cpu);
		}
	}
}

static void kexec_prepare_cpus(void)
{
	wake_offline_cpus();
	smp_call_function(kexec_smp_down, NULL, /* wait */0);
	local_irq_disable();
	mb(); /* make sure IRQs are disabled before we say they are */
	get_paca()->kexec_state = KEXEC_STATE_IRQS_OFF;

	kexec_prepare_cpus_wait(KEXEC_STATE_IRQS_OFF);
	/* we are sure every CPU has IRQs off at this point */
	kexec_all_irq_disabled = 1;

	/* after we tell the others to go down */
	if (ppc_md.kexec_cpu_down)
		ppc_md.kexec_cpu_down(0, 0);

	/*
	 * Before removing MMU mappings make sure all CPUs have entered real
	 * mode:
	 */
	kexec_prepare_cpus_wait(KEXEC_STATE_REAL_MODE);

	put_cpu();
}

#else /* ! SMP */

static void kexec_prepare_cpus(void)
{
	/*
	 * move the secondarys to us so that we can copy
	 * the new kernel 0-0x100 safely
	 *
	 * do this if kexec in setup.c ?
	 *
	 * We need to release the cpus if we are ever going from an
	 * UP to an SMP kernel.
	 */
	smp_release_cpus();
	if (ppc_md.kexec_cpu_down)
		ppc_md.kexec_cpu_down(0, 0);
	local_irq_disable();
}

#endif /* SMP */

/*
 * kexec thread structure and stack.
 *
 * We need to make sure that this is 16384-byte aligned due to the
 * way process stacks are handled.  It also must be statically allocated
 * or allocated as part of the kimage, because everything else may be
 * overwritten when we copy the kexec image.  We piggyback on the
 * "init_task" linker section here to statically allocate a stack.
 *
 * We could use a smaller stack if we don't care about anything using
 * current, but that audit has not been performed.
 */
static union thread_union kexec_stack __init_task_data =
	{ };

/*
 * For similar reasons to the stack above, the kexecing CPU needs to be on a
 * static PACA; we switch to kexec_paca.
 */
struct paca_struct kexec_paca;

/* Our assembly helper, in kexec_stub.S */
extern NORET_TYPE void kexec_sequence(void *newstack, unsigned long start,
					void *image, void *control,
					void (*clear_all)(void)) ATTRIB_NORET;

/* too late to fail here */
void default_machine_kexec(struct kimage *image)
{
	/* prepare control code if any */

	/*
        * If the kexec boot is the normal one, need to shutdown other cpus
        * into our wait loop and quiesce interrupts.
        * Otherwise, in the case of crashed mode (crashing_cpu >= 0),
        * stopping other CPUs and collecting their pt_regs is done before
        * using debugger IPI.
        */

	if (crashing_cpu == -1)
		kexec_prepare_cpus();

	pr_debug("kexec: Starting switchover sequence.\n");

	/* switch to a staticly allocated stack.  Based on irq stack code.
	 * XXX: the task struct will likely be invalid once we do the copy!
	 */
	kexec_stack.thread_info.task = current_thread_info()->task;
	kexec_stack.thread_info.flags = 0;

	/* We need a static PACA, too; copy this CPU's PACA over and switch to
	 * it.  Also poison per_cpu_offset to catch anyone using non-static
	 * data.
	 */
	memcpy(&kexec_paca, get_paca(), sizeof(struct paca_struct));
	kexec_paca.data_offset = 0xedeaddeadeeeeeeeUL;
	paca = (struct paca_struct *)RELOC_HIDE(&kexec_paca, 0) -
		kexec_paca.paca_index;
	setup_paca(&kexec_paca);

	/* XXX: If anyone does 'dynamic lppacas' this will also need to be
	 * switched to a static version!
	 */

	/* Some things are best done in assembly.  Finding globals with
	 * a toc is easier in C, so pass in what we can.
	 */
	kexec_sequence(&kexec_stack, image->start, image,
			page_address(image->control_code_page),
			ppc_md.hpte_clear_all);
	/* NOTREACHED */
}

/* Values we need to export to the second kernel via the device tree. */
static unsigned long htab_base;

static struct property htab_base_prop = {
	.name = "linux,htab-base",
	.length = sizeof(unsigned long),
	.value = &htab_base,
};

static struct property htab_size_prop = {
	.name = "linux,htab-size",
	.length = sizeof(unsigned long),
	.value = &htab_size_bytes,
};

static int __init export_htab_values(void)
{
	struct device_node *node;
	struct property *prop;

	/* On machines with no htab htab_address is NULL */
	if (!htab_address)
		return -ENODEV;

	node = of_find_node_by_path("/chosen");
	if (!node)
		return -ENODEV;

	/* remove any stale propertys so ours can be found */
	prop = of_find_property(node, htab_base_prop.name, NULL);
	if (prop)
		prom_remove_property(node, prop);
	prop = of_find_property(node, htab_size_prop.name, NULL);
	if (prop)
		prom_remove_property(node, prop);

	htab_base = __pa(htab_address);
	prom_add_property(node, &htab_base_prop);
	prom_add_property(node, &htab_size_prop);

	of_node_put(node);
	return 0;
}
late_initcall(export_htab_values);
Commit	Line	Data
fce0d574	1	/*
3d1229d6	2	* PPC64 code to handle Linux booting another kernel.
fce0d574 S	3	*
	4	* Copyright (C) 2004-2005, IBM Corp.
	5	*
	6	* Created by: Milton D Miller II
	7	*
	8	* This source code is licensed under the GNU General Public License,
	9	* Version 2. See the file COPYING for more details.
	10	*/
	11
	12
fce0d574 S	13	#include <linux/kexec.h>
	14	#include <linux/smp.h>
	15	#include <linux/thread_info.h>
d200c922	16	#include <linux/init_task.h>
fce0d574	17	#include <linux/errno.h>
e2f7f737	18	#include <linux/kernel.h>
e8e5c215	19	#include <linux/cpu.h>
fce0d574 S	20
	21	#include <asm/page.h>
	22	#include <asm/current.h>
	23	#include <asm/machdep.h>
	24	#include <asm/cacheflush.h>
	25	#include <asm/paca.h>
	26	#include <asm/mmu.h>
	27	#include <asm/sections.h> /* _end */
	28	#include <asm/prom.h>
2249ca9d	29	#include <asm/smp.h>
5aae8a53	30	#include <asm/hw_breakpoint.h>
fce0d574	31
3d1229d6	32	int default_machine_kexec_prepare(struct kimage *image)
fce0d574 S	33	{
	34	int i;
	35	unsigned long begin, end; /* limits of segment */
	36	unsigned long low, high; /* limits of blocked memory range */
	37	struct device_node *node;
a7f67bdf JK	38	const unsigned long *basep;
a7f67bdf JK	39	const unsigned int *sizep;
fce0d574 S	40
	41	if (!ppc_md.hpte_clear_all)
	42	return -ENOENT;
	43
	44	/*
	45	* Since we use the kernel fault handlers and paging code to
	46	* handle the virtual mode, we must make sure no destination
	47	* overlaps kernel static data or bss.
	48	*/
72414d3f	49	for (i = 0; i < image->nr_segments; i++)
fce0d574 S	50	if (image->segment[i].mem < __pa(_end))
	51	return -ETXTBSY;
	52
	53	/*
	54	* For non-LPAR, we absolutely can not overwrite the mmu hash
	55	* table, since we are still using the bolted entries in it to
	56	* do the copy. Check that here.
	57	*
	58	* It is safe if the end is below the start of the blocked
	59	* region (end <= low), or if the beginning is after the
	60	* end of the blocked region (begin >= high). Use the
	61	* boolean identity !(a \|\| b) === (!a && !b).
	62	*/
	63	if (htab_address) {
	64	low = __pa(htab_address);
337a7128	65	high = low + htab_size_bytes;
fce0d574	66
72414d3f	67	for (i = 0; i < image->nr_segments; i++) {
fce0d574 S	68	begin = image->segment[i].mem;
	69	end = begin + image->segment[i].memsz;
	70
	71	if ((begin < high) && (end > low))
	72	return -ETXTBSY;
	73	}
	74	}
	75
	76	/* We also should not overwrite the tce tables */
	77	for (node = of_find_node_by_type(NULL, "pci"); node != NULL;
	78	node = of_find_node_by_type(node, "pci")) {
e2eb6392 SR	79	basep = of_get_property(node, "linux,tce-base", NULL);
e2eb6392 SR	80	sizep = of_get_property(node, "linux,tce-size", NULL);
fce0d574 S	81	if (basep == NULL \|\| sizep == NULL)
	82	continue;
	83
	84	low = *basep;
	85	high = low + (*sizep);
	86
72414d3f	87	for (i = 0; i < image->nr_segments; i++) {
fce0d574 S	88	begin = image->segment[i].mem;
	89	end = begin + image->segment[i].memsz;
	90
	91	if ((begin < high) && (end > low))
	92	return -ETXTBSY;
	93	}
	94	}
	95
	96	return 0;
	97	}
	98
fce0d574 S	99	#define IND_FLAGS (IND_DESTINATION \| IND_INDIRECTION \| IND_DONE \| IND_SOURCE)
	100
	101	static void copy_segments(unsigned long ind)
	102	{
	103	unsigned long entry;
	104	unsigned long *ptr;
	105	void *dest;
	106	void *addr;
	107
	108	/*
	109	* We rely on kexec_load to create a lists that properly
	110	* initializes these pointers before they are used.
	111	* We will still crash if the list is wrong, but at least
	112	* the compiler will be quiet.
	113	*/
	114	ptr = NULL;
	115	dest = NULL;
	116
	117	for (entry = ind; !(entry & IND_DONE); entry = *ptr++) {
	118	addr = __va(entry & PAGE_MASK);
	119
	120	switch (entry & IND_FLAGS) {
	121	case IND_DESTINATION:
	122	dest = addr;
	123	break;
	124	case IND_INDIRECTION:
	125	ptr = addr;
	126	break;
	127	case IND_SOURCE:
	128	copy_page(dest, addr);
	129	dest += PAGE_SIZE;
	130	}
	131	}
	132	}
	133
	134	void kexec_copy_flush(struct kimage *image)
	135	{
	136	long i, nr_segments = image->nr_segments;
	137	struct kexec_segment ranges[KEXEC_SEGMENT_MAX];
	138
	139	/* save the ranges on the stack to efficiently flush the icache */
	140	memcpy(ranges, image->segment, sizeof(ranges));
	141
	142	/*
	143	* After this call we may not use anything allocated in dynamic
	144	* memory, including *image.
	145	*
	146	* Only globals and the stack are allowed.
	147	*/
	148	copy_segments(image->head);
	149
	150	/*
	151	* we need to clear the icache for all dest pages sometime,
	152	* including ones that were in place on the original copy
	153	*/
	154	for (i = 0; i < nr_segments; i++)
b5666f70 ME	155	flush_icache_range((unsigned long)__va(ranges[i].mem),
b5666f70 ME	156	(unsigned long)__va(ranges[i].mem + ranges[i].memsz));
fce0d574 S	157	}
	158
	159	#ifdef CONFIG_SMP
	160
1fc711f7 MN	161	static int kexec_all_irq_disabled = 0;
1fc711f7 MN	162
1c21a293	163	static void kexec_smp_down(void *arg)
fce0d574	164	{
1fc711f7 MN	165	local_irq_disable();
	166	mb(); /* make sure our irqs are disabled before we say they are */
	167	get_paca()->kexec_state = KEXEC_STATE_IRQS_OFF;
	168	while(kexec_all_irq_disabled == 0)
	169	cpu_relax();
	170	mb(); /* make sure all irqs are disabled before this */
5aae8a53	171	hw_breakpoint_disable();
1fc711f7 MN	172	/*
	173	* Now every CPU has IRQs off, we can clear out any pending
	174	* IPIs and be sure that no more will come in after this.
	175	*/
c5e24354 ME	176	if (ppc_md.kexec_cpu_down)
c5e24354 ME	177	ppc_md.kexec_cpu_down(0, 1);
fce0d574	178
fce0d574 S	179	kexec_smp_wait();
	180	/* NOTREACHED */
	181	}
	182
1fc711f7	183	static void kexec_prepare_cpus_wait(int wait_state)
fce0d574 S	184	{
	185	int my_cpu, i, notified=-1;
	186
5aae8a53	187	hw_breakpoint_disable();
fce0d574	188	my_cpu = get_cpu();
e2f7f737 ME	189	/* Make sure each CPU has at least made it to the state we need.
	190	*
	191	* FIXME: There is a (slim) chance of a problem if not all of the CPUs
	192	* are correctly onlined. If somehow we start a CPU on boot with RTAS
	193	* start-cpu, but somehow that CPU doesn't write callin_cpu_map[] in
	194	* time, the boot CPU will timeout. If it does eventually execute
	195	* stuff, the secondary will start up (paca[].cpu_start was written) and
	196	* get into a peculiar state. If the platform supports
	197	* smp_ops->take_timebase(), the secondary CPU will probably be spinning
	198	* in there. If not (i.e. pseries), the secondary will continue on and
	199	* try to online itself/idle/etc. If it survives that, we need to find
	200	* these possible-but-not-online-but-should-be CPUs and chaperone them
	201	* into kexec_smp_wait().
	202	*/
b636f137	203	for_each_online_cpu(i) {
fce0d574 S	204	if (i == my_cpu)
	205	continue;
	206
1fc711f7	207	while (paca[i].kexec_state < wait_state) {
b3ca8093	208	barrier();
fce0d574	209	if (i != notified) {
e2f7f737 ME	210	printk(KERN_INFO "kexec: waiting for cpu %d "
	211	"(physical %d) to enter %i state\n",
	212	i, paca[i].hw_cpu_id, wait_state);
fce0d574 S	213	notified = i;
	214	}
	215	}
	216	}
1fc711f7 MN	217	mb();
	218	}
	219
e8e5c215 ME	220	/*
	221	* We need to make sure each present CPU is online. The next kernel will scan
	222	* the device tree and assume primary threads are online and query secondary
	223	* threads via RTAS to online them if required. If we don't online primary
	224	* threads, they will be stuck. However, we also online secondary threads as we
	225	* may be using 'cede offline'. In this case RTAS doesn't see the secondary
	226	* threads as offline -- and again, these CPUs will be stuck.
	227	*
	228	* So, we online all CPUs that should be running, including secondary threads.
	229	*/
	230	static void wake_offline_cpus(void)
1fc711f7	231	{
e8e5c215 ME	232	int cpu = 0;
	233
	234	for_each_present_cpu(cpu) {
	235	if (!cpu_online(cpu)) {
	236	printk(KERN_INFO "kexec: Waking offline cpu %d.\n",
	237	cpu);
	238	cpu_up(cpu);
	239	}
	240	}
	241	}
1fc711f7	242
e8e5c215 ME	243	static void kexec_prepare_cpus(void)
	244	{
	245	wake_offline_cpus();
1fc711f7 MN	246	smp_call_function(kexec_smp_down, NULL, /* wait */0);
	247	local_irq_disable();
	248	mb(); /* make sure IRQs are disabled before we say they are */
	249	get_paca()->kexec_state = KEXEC_STATE_IRQS_OFF;
	250
	251	kexec_prepare_cpus_wait(KEXEC_STATE_IRQS_OFF);
	252	/* we are sure every CPU has IRQs off at this point */
	253	kexec_all_irq_disabled = 1;
fce0d574 S	254
fce0d574 S	255	/* after we tell the others to go down */
c5e24354 ME	256	if (ppc_md.kexec_cpu_down)
c5e24354 ME	257	ppc_md.kexec_cpu_down(0, 0);
fce0d574	258
e2f7f737 ME	259	/*
	260	* Before removing MMU mappings make sure all CPUs have entered real
	261	* mode:
	262	*/
1fc711f7	263	kexec_prepare_cpus_wait(KEXEC_STATE_REAL_MODE);
fce0d574	264
1fc711f7	265	put_cpu();
fce0d574 S	266	}
	267
	268	#else /* ! SMP */
	269
	270	static void kexec_prepare_cpus(void)
	271	{
	272	/*
	273	* move the secondarys to us so that we can copy
	274	* the new kernel 0-0x100 safely
	275	*
	276	* do this if kexec in setup.c ?
75eedfed OJ	277	*
	278	* We need to release the cpus if we are ever going from an
	279	* UP to an SMP kernel.
fce0d574	280	*/
75eedfed	281	smp_release_cpus();
c5e24354 ME	282	if (ppc_md.kexec_cpu_down)
c5e24354 ME	283	ppc_md.kexec_cpu_down(0, 0);
fce0d574 S	284	local_irq_disable();
	285	}
	286
	287	#endif /* SMP */
	288
	289	/*
	290	* kexec thread structure and stack.
	291	*
	292	* We need to make sure that this is 16384-byte aligned due to the
	293	* way process stacks are handled. It also must be statically allocated
	294	* or allocated as part of the kimage, because everything else may be
	295	* overwritten when we copy the kexec image. We piggyback on the
	296	* "init_task" linker section here to statically allocate a stack.
	297	*
	298	* We could use a smaller stack if we don't care about anything using
	299	* current, but that audit has not been performed.
	300	*/
d200c922 JP	301	static union thread_union kexec_stack __init_task_data =
d200c922 JP	302	{ };
fce0d574	303
fc53b420 ME	304	/*
	305	* For similar reasons to the stack above, the kexecing CPU needs to be on a
	306	* static PACA; we switch to kexec_paca.
	307	*/
	308	struct paca_struct kexec_paca;
	309
fce0d574 S	310	/* Our assembly helper, in kexec_stub.S */
fce0d574 S	311	extern NORET_TYPE void kexec_sequence(void *newstack, unsigned long start,
72414d3f	312	void image, void control,
1767c8f3	313	void (*clear_all)(void)) ATTRIB_NORET;
fce0d574 S	314
fce0d574 S	315	/* too late to fail here */
3d1229d6	316	void default_machine_kexec(struct kimage *image)
fce0d574	317	{
fce0d574 S	318	/* prepare control code if any */
fce0d574 S	319
cc532915 ME	320	/*
	321	* If the kexec boot is the normal one, need to shutdown other cpus
	322	* into our wait loop and quiesce interrupts.
	323	* Otherwise, in the case of crashed mode (crashing_cpu >= 0),
	324	* stopping other CPUs and collecting their pt_regs is done before
	325	* using debugger IPI.
	326	*/
	327
54622f10 MK	328	if (crashing_cpu == -1)
54622f10 MK	329	kexec_prepare_cpus();
fce0d574	330
e2f7f737 ME	331	pr_debug("kexec: Starting switchover sequence.\n");
e2f7f737 ME	332
fce0d574 S	333	/* switch to a staticly allocated stack. Based on irq stack code.
	334	* XXX: the task struct will likely be invalid once we do the copy!
	335	*/
	336	kexec_stack.thread_info.task = current_thread_info()->task;
	337	kexec_stack.thread_info.flags = 0;
	338
fc53b420 ME	339	/* We need a static PACA, too; copy this CPU's PACA over and switch to
	340	* it. Also poison per_cpu_offset to catch anyone using non-static
	341	* data.
	342	*/
	343	memcpy(&kexec_paca, get_paca(), sizeof(struct paca_struct));
	344	kexec_paca.data_offset = 0xedeaddeadeeeeeeeUL;
	345	paca = (struct paca_struct *)RELOC_HIDE(&kexec_paca, 0) -
	346	kexec_paca.paca_index;
	347	setup_paca(&kexec_paca);
	348
	349	/* XXX: If anyone does 'dynamic lppacas' this will also need to be
	350	* switched to a static version!
	351	*/
	352
fce0d574 S	353	/* Some things are best done in assembly. Finding globals with
	354	* a toc is easier in C, so pass in what we can.
	355	*/
	356	kexec_sequence(&kexec_stack, image->start, image,
	357	page_address(image->control_code_page),
1767c8f3	358	ppc_md.hpte_clear_all);
fce0d574 S	359	/* NOTREACHED */
fce0d574 S	360	}
593e537b ME	361
593e537b ME	362	/* Values we need to export to the second kernel via the device tree. */
2e8e4f5b	363	static unsigned long htab_base;
593e537b ME	364
	365	static struct property htab_base_prop = {
	366	.name = "linux,htab-base",
	367	.length = sizeof(unsigned long),
1a38147e	368	.value = &htab_base,
593e537b ME	369	};
	370
	371	static struct property htab_size_prop = {
	372	.name = "linux,htab-size",
	373	.length = sizeof(unsigned long),
1a38147e	374	.value = &htab_size_bytes,
593e537b ME	375	};
593e537b ME	376
6f29c329	377	static int __init export_htab_values(void)
593e537b ME	378	{
593e537b ME	379	struct device_node *node;
ed7b2144	380	struct property *prop;
593e537b	381
2e8e4f5b DF	382	/* On machines with no htab htab_address is NULL */
2e8e4f5b DF	383	if (!htab_address)
6f29c329	384	return -ENODEV;
2e8e4f5b	385
593e537b ME	386	node = of_find_node_by_path("/chosen");
593e537b ME	387	if (!node)
6f29c329	388	return -ENODEV;
593e537b	389
ed7b2144	390	/* remove any stale propertys so ours can be found */
ed7b2144 MM	391	prop = of_find_property(node, htab_base_prop.name, NULL);
	392	if (prop)
	393	prom_remove_property(node, prop);
	394	prop = of_find_property(node, htab_size_prop.name, NULL);
	395	if (prop)
	396	prom_remove_property(node, prop);
	397
593e537b ME	398	htab_base = __pa(htab_address);
593e537b ME	399	prom_add_property(node, &htab_base_prop);
593e537b ME	400	prom_add_property(node, &htab_size_prop);
593e537b ME	401
593e537b	402	of_node_put(node);
aa98c50d	403	return 0;
35dd5432	404	}
6f29c329	405	late_initcall(export_htab_values);