[net-next-2.6.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/cpumask.h>
#include <linux/kexec.h>
#include <linux/smp.h>
#include <linux/thread_info.h>
#include <linux/errno.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>

#define HASH_GROUP_SIZE 0x80	/* size of each hash group, asm/mmu.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;
	unsigned long *basep;
	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_hash_mask + 1) * HASH_GROUP_SIZE;

		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 = (unsigned long *)get_property(node, "linux,tce-base",
							NULL);
		sizep = (unsigned int *)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

/* FIXME: we should schedule this function to be called on all cpus based
 * on calling the interrupts, but we would like to call it off irq level
 * so that the interrupt controller is clean.
 */
void kexec_smp_down(void *arg)
{
	if (ppc_md.kexec_cpu_down)
		ppc_md.kexec_cpu_down(0, 1);

	local_irq_disable();
	kexec_smp_wait();
	/* NOTREACHED */
}

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

	smp_call_function(kexec_smp_down, NULL, 0, /* wait */0);
	my_cpu = get_cpu();

	/* check the others cpus are now down (via paca hw cpu id == -1) */
	for (i=0; i < NR_CPUS; i++) {
		if (i == my_cpu)
			continue;

		while (paca[i].hw_cpu_id != -1) {
			barrier();
			if (!cpu_possible(i)) {
				printk("kexec: cpu %d hw_cpu_id %d is not"
						" possible, ignoring\n",
						i, paca[i].hw_cpu_id);
				break;
			}
			if (!cpu_online(i)) {
				/* Fixme: this can be spinning in
				 * pSeries_secondary_wait with a paca
				 * waiting for it to go online.
				 */
				printk("kexec: cpu %d hw_cpu_id %d is not"
						" online, ignoring\n",
						i, paca[i].hw_cpu_id);
				break;
			}
			if (i != notified) {
				printk( "kexec: waiting for cpu %d (physical"
						" %d) to go down\n",
						i, paca[i].hw_cpu_id);
				notified = i;
			}
		}
	}

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

	put_cpu();

	local_irq_disable();
}

#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.
 */
union thread_union kexec_stack
	__attribute__((__section__(".data.init_task"))) = { };

/* 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 */

	/* shutdown other cpus into our wait loop and quiesce interrupts */
	kexec_prepare_cpus();

	/* 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;

	/* 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, htab_size, kernel_end;

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

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

static struct property kernel_end_prop = {
	.name = "linux,kernel-end",
	.length = sizeof(unsigned long),
	.value = (unsigned char *)&kernel_end,
};

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

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

	kernel_end = __pa(_end);
	prom_add_property(node, &kernel_end_prop);

	/* On machines with no htab htab_address is NULL */
	if (NULL == htab_address)
		goto out;

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

	htab_size = 1UL << ppc64_pft_size;
	prom_add_property(node, &htab_size_prop);

 out:
	of_node_put(node);
}

void __init kexec_setup(void)
{
	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
	13	#include <linux/cpumask.h>
	14	#include <linux/kexec.h>
	15	#include <linux/smp.h>
	16	#include <linux/thread_info.h>
	17	#include <linux/errno.h>
	18
	19	#include <asm/page.h>
	20	#include <asm/current.h>
	21	#include <asm/machdep.h>
	22	#include <asm/cacheflush.h>
	23	#include <asm/paca.h>
	24	#include <asm/mmu.h>
	25	#include <asm/sections.h> /* _end */
	26	#include <asm/prom.h>
2249ca9d	27	#include <asm/smp.h>
fce0d574 S	28
	29	#define HASH_GROUP_SIZE 0x80 /* size of each hash group, asm/mmu.h */
	30
3d1229d6	31	int default_machine_kexec_prepare(struct kimage *image)
fce0d574 S	32	{
	33	int i;
	34	unsigned long begin, end; /* limits of segment */
	35	unsigned long low, high; /* limits of blocked memory range */
	36	struct device_node *node;
	37	unsigned long *basep;
	38	unsigned int *sizep;
	39
	40	if (!ppc_md.hpte_clear_all)
	41	return -ENOENT;
	42
	43	/*
	44	* Since we use the kernel fault handlers and paging code to
	45	* handle the virtual mode, we must make sure no destination
	46	* overlaps kernel static data or bss.
	47	*/
72414d3f	48	for (i = 0; i < image->nr_segments; i++)
fce0d574 S	49	if (image->segment[i].mem < __pa(_end))
	50	return -ETXTBSY;
	51
	52	/*
	53	* For non-LPAR, we absolutely can not overwrite the mmu hash
	54	* table, since we are still using the bolted entries in it to
	55	* do the copy. Check that here.
	56	*
	57	* It is safe if the end is below the start of the blocked
	58	* region (end <= low), or if the beginning is after the
	59	* end of the blocked region (begin >= high). Use the
	60	* boolean identity !(a \|\| b) === (!a && !b).
	61	*/
	62	if (htab_address) {
	63	low = __pa(htab_address);
	64	high = low + (htab_hash_mask + 1) * HASH_GROUP_SIZE;
	65
72414d3f	66	for (i = 0; i < image->nr_segments; i++) {
fce0d574 S	67	begin = image->segment[i].mem;
	68	end = begin + image->segment[i].memsz;
	69
	70	if ((begin < high) && (end > low))
	71	return -ETXTBSY;
	72	}
	73	}
	74
	75	/* We also should not overwrite the tce tables */
	76	for (node = of_find_node_by_type(NULL, "pci"); node != NULL;
	77	node = of_find_node_by_type(node, "pci")) {
	78	basep = (unsigned long *)get_property(node, "linux,tce-base",
	79	NULL);
	80	sizep = (unsigned int *)get_property(node, "linux,tce-size",
	81	NULL);
	82	if (basep == NULL \|\| sizep == NULL)
	83	continue;
	84
	85	low = *basep;
	86	high = low + (*sizep);
	87
72414d3f	88	for (i = 0; i < image->nr_segments; i++) {
fce0d574 S	89	begin = image->segment[i].mem;
	90	end = begin + image->segment[i].memsz;
	91
	92	if ((begin < high) && (end > low))
	93	return -ETXTBSY;
	94	}
	95	}
	96
	97	return 0;
	98	}
	99
fce0d574 S	100	#define IND_FLAGS (IND_DESTINATION \| IND_INDIRECTION \| IND_DONE \| IND_SOURCE)
	101
	102	static void copy_segments(unsigned long ind)
	103	{
	104	unsigned long entry;
	105	unsigned long *ptr;
	106	void *dest;
	107	void *addr;
	108
	109	/*
	110	* We rely on kexec_load to create a lists that properly
	111	* initializes these pointers before they are used.
	112	* We will still crash if the list is wrong, but at least
	113	* the compiler will be quiet.
	114	*/
	115	ptr = NULL;
	116	dest = NULL;
	117
	118	for (entry = ind; !(entry & IND_DONE); entry = *ptr++) {
	119	addr = __va(entry & PAGE_MASK);
	120
	121	switch (entry & IND_FLAGS) {
	122	case IND_DESTINATION:
	123	dest = addr;
	124	break;
	125	case IND_INDIRECTION:
	126	ptr = addr;
	127	break;
	128	case IND_SOURCE:
	129	copy_page(dest, addr);
	130	dest += PAGE_SIZE;
	131	}
	132	}
	133	}
	134
	135	void kexec_copy_flush(struct kimage *image)
	136	{
	137	long i, nr_segments = image->nr_segments;
	138	struct kexec_segment ranges[KEXEC_SEGMENT_MAX];
	139
	140	/* save the ranges on the stack to efficiently flush the icache */
	141	memcpy(ranges, image->segment, sizeof(ranges));
	142
	143	/*
	144	* After this call we may not use anything allocated in dynamic
	145	* memory, including *image.
	146	*
	147	* Only globals and the stack are allowed.
	148	*/
	149	copy_segments(image->head);
	150
	151	/*
	152	* we need to clear the icache for all dest pages sometime,
	153	* including ones that were in place on the original copy
	154	*/
	155	for (i = 0; i < nr_segments; i++)
b5666f70 ME	156	flush_icache_range((unsigned long)__va(ranges[i].mem),
b5666f70 ME	157	(unsigned long)__va(ranges[i].mem + ranges[i].memsz));
fce0d574 S	158	}
	159
	160	#ifdef CONFIG_SMP
	161
	162	/* FIXME: we should schedule this function to be called on all cpus based
	163	* on calling the interrupts, but we would like to call it off irq level
	164	* so that the interrupt controller is clean.
	165	*/
	166	void kexec_smp_down(void *arg)
	167	{
c5e24354 ME	168	if (ppc_md.kexec_cpu_down)
c5e24354 ME	169	ppc_md.kexec_cpu_down(0, 1);
fce0d574 S	170
	171	local_irq_disable();
	172	kexec_smp_wait();
	173	/* NOTREACHED */
	174	}
	175
	176	static void kexec_prepare_cpus(void)
	177	{
	178	int my_cpu, i, notified=-1;
	179
	180	smp_call_function(kexec_smp_down, NULL, 0, /* wait */0);
	181	my_cpu = get_cpu();
	182
	183	/* check the others cpus are now down (via paca hw cpu id == -1) */
	184	for (i=0; i < NR_CPUS; i++) {
	185	if (i == my_cpu)
	186	continue;
	187
	188	while (paca[i].hw_cpu_id != -1) {
b3ca8093	189	barrier();
fce0d574 S	190	if (!cpu_possible(i)) {
	191	printk("kexec: cpu %d hw_cpu_id %d is not"
	192	" possible, ignoring\n",
	193	i, paca[i].hw_cpu_id);
	194	break;
	195	}
	196	if (!cpu_online(i)) {
	197	/* Fixme: this can be spinning in
	198	* pSeries_secondary_wait with a paca
	199	* waiting for it to go online.
	200	*/
	201	printk("kexec: cpu %d hw_cpu_id %d is not"
	202	" online, ignoring\n",
	203	i, paca[i].hw_cpu_id);
	204	break;
	205	}
	206	if (i != notified) {
	207	printk( "kexec: waiting for cpu %d (physical"
	208	" %d) to go down\n",
	209	i, paca[i].hw_cpu_id);
	210	notified = i;
	211	}
	212	}
	213	}
	214
	215	/* after we tell the others to go down */
c5e24354 ME	216	if (ppc_md.kexec_cpu_down)
c5e24354 ME	217	ppc_md.kexec_cpu_down(0, 0);
fce0d574 S	218
	219	put_cpu();
	220
	221	local_irq_disable();
	222	}
	223
	224	#else /* ! SMP */
	225
	226	static void kexec_prepare_cpus(void)
	227	{
	228	/*
	229	* move the secondarys to us so that we can copy
	230	* the new kernel 0-0x100 safely
	231	*
	232	* do this if kexec in setup.c ?
75eedfed OJ	233	*
	234	* We need to release the cpus if we are ever going from an
	235	* UP to an SMP kernel.
fce0d574	236	*/
75eedfed	237	smp_release_cpus();
c5e24354 ME	238	if (ppc_md.kexec_cpu_down)
c5e24354 ME	239	ppc_md.kexec_cpu_down(0, 0);
fce0d574 S	240	local_irq_disable();
	241	}
	242
	243	#endif /* SMP */
	244
	245	/*
	246	* kexec thread structure and stack.
	247	*
	248	* We need to make sure that this is 16384-byte aligned due to the
	249	* way process stacks are handled. It also must be statically allocated
	250	* or allocated as part of the kimage, because everything else may be
	251	* overwritten when we copy the kexec image. We piggyback on the
	252	* "init_task" linker section here to statically allocate a stack.
	253	*
	254	* We could use a smaller stack if we don't care about anything using
	255	* current, but that audit has not been performed.
	256	*/
	257	union thread_union kexec_stack
	258	__attribute__((__section__(".data.init_task"))) = { };
	259
	260	/* Our assembly helper, in kexec_stub.S */
	261	extern NORET_TYPE void kexec_sequence(void *newstack, unsigned long start,
72414d3f MS	262	void image, void control,
72414d3f MS	263	void (*clear_all)(void)) ATTRIB_NORET;
fce0d574 S	264
fce0d574 S	265	/* too late to fail here */
3d1229d6	266	void default_machine_kexec(struct kimage *image)
fce0d574 S	267	{
	268
	269	/* prepare control code if any */
	270
	271	/* shutdown other cpus into our wait loop and quiesce interrupts */
	272	kexec_prepare_cpus();
	273
	274	/* switch to a staticly allocated stack. Based on irq stack code.
	275	* XXX: the task struct will likely be invalid once we do the copy!
	276	*/
	277	kexec_stack.thread_info.task = current_thread_info()->task;
	278	kexec_stack.thread_info.flags = 0;
	279
	280	/* Some things are best done in assembly. Finding globals with
	281	* a toc is easier in C, so pass in what we can.
	282	*/
	283	kexec_sequence(&kexec_stack, image->start, image,
	284	page_address(image->control_code_page),
	285	ppc_md.hpte_clear_all);
	286	/* NOTREACHED */
	287	}
593e537b ME	288
	289	/* Values we need to export to the second kernel via the device tree. */
	290	static unsigned long htab_base, htab_size, kernel_end;
	291
	292	static struct property htab_base_prop = {
	293	.name = "linux,htab-base",
	294	.length = sizeof(unsigned long),
	295	.value = (unsigned char *)&htab_base,
	296	};
	297
	298	static struct property htab_size_prop = {
	299	.name = "linux,htab-size",
	300	.length = sizeof(unsigned long),
	301	.value = (unsigned char *)&htab_size,
	302	};
	303
	304	static struct property kernel_end_prop = {
	305	.name = "linux,kernel-end",
	306	.length = sizeof(unsigned long),
	307	.value = (unsigned char *)&kernel_end,
	308	};
	309
	310	static void __init export_htab_values(void)
	311	{
	312	struct device_node *node;
	313
	314	node = of_find_node_by_path("/chosen");
	315	if (!node)
	316	return;
	317
	318	kernel_end = __pa(_end);
	319	prom_add_property(node, &kernel_end_prop);
	320
	321	/* On machines with no htab htab_address is NULL */
	322	if (NULL == htab_address)
	323	goto out;
	324
	325	htab_base = __pa(htab_address);
	326	prom_add_property(node, &htab_base_prop);
	327
	328	htab_size = 1UL << ppc64_pft_size;
	329	prom_add_property(node, &htab_size_prop);
	330
	331	out:
	332	of_node_put(node);
	333	}
	334
	335	void __init kexec_setup(void)
	336	{
	337	export_htab_values();
	338	}