powerpc/kernel/fadump.c

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * Firmware Assisted dump: A robust mechanism to get reliable kernel crash
 * dump with assistance from firmware. This approach does not use kexec,
 * instead firmware assists in booting the kdump kernel while preserving
 * memory contents. The most of the code implementation has been adapted
 * from phyp assisted dump implementation written by Linas Vepstas and
 * Manish Ahuja
 *
 * Copyright 2011 IBM Corporation
 * Author: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
 */

#undef DEBUG
#define pr_fmt(fmt) "fadump: " fmt

#include <linux/string.h>
#include <linux/memblock.h>
#include <linux/delay.h>
#include <linux/seq_file.h>
#include <linux/crash_dump.h>
#include <linux/kobject.h>
#include <linux/sysfs.h>
#include <linux/slab.h>
#include <linux/cma.h>
#include <linux/hugetlb.h>

#include <asm/debugfs.h>
#include <asm/page.h>
#include <asm/prom.h>
#include <asm/fadump.h>
#include <asm/fadump-internal.h>
#include <asm/setup.h>

static struct fw_dump fw_dump;

static DEFINE_MUTEX(fadump_mutex);
struct fad_crash_memory_ranges *crash_memory_ranges;
int crash_memory_ranges_size;
int crash_mem_ranges;
int max_crash_mem_ranges;

#ifdef CONFIG_CMA
static struct cma *fadump_cma;

/*
 * fadump_cma_init() - Initialize CMA area from a fadump reserved memory
 *
 * This function initializes CMA area from fadump reserved memory.
 * The total size of fadump reserved memory covers for boot memory size
 * + cpu data size + hpte size and metadata.
 * Initialize only the area equivalent to boot memory size for CMA use.
 * The reamining portion of fadump reserved memory will be not given
 * to CMA and pages for thoes will stay reserved. boot memory size is
 * aligned per CMA requirement to satisy cma_init_reserved_mem() call.
 * But for some reason even if it fails we still have the memory reservation
 * with us and we can still continue doing fadump.
 */
int __init fadump_cma_init(void)
{
	unsigned long long base, size;
	int rc;

	if (!fw_dump.fadump_enabled)
		return 0;

	/*
	 * Do not use CMA if user has provided fadump=nocma kernel parameter.
	 * Return 1 to continue with fadump old behaviour.
	 */
	if (fw_dump.nocma)
		return 1;

	base = fw_dump.reserve_dump_area_start;
	size = fw_dump.boot_memory_size;

	if (!size)
		return 0;

	rc = cma_init_reserved_mem(base, size, 0, "fadump_cma", &fadump_cma);
	if (rc) {
		pr_err("Failed to init cma area for firmware-assisted dump,%d\n", rc);
		/*
		 * Though the CMA init has failed we still have memory
		 * reservation with us. The reserved memory will be
		 * blocked from production system usage.  Hence return 1,
		 * so that we can continue with fadump.
		 */
		return 1;
	}

	/*
	 * So we now have successfully initialized cma area for fadump.
	 */
	pr_info("Initialized 0x%lx bytes cma area at %ldMB from 0x%lx "
		"bytes of memory reserved for firmware-assisted dump\n",
		cma_get_size(fadump_cma),
		(unsigned long)cma_get_base(fadump_cma) >> 20,
		fw_dump.reserve_dump_area_size);
	return 1;
}
#else
static int __init fadump_cma_init(void) { return 1; }
#endif /* CONFIG_CMA */

/* Scan the Firmware Assisted dump configuration details. */
int __init early_init_dt_scan_fw_dump(unsigned long node, const char *uname,
				      int depth, void *data)
{
	if (depth != 1)
		return 0;

	if (strcmp(uname, "rtas") == 0) {
		rtas_fadump_dt_scan(&fw_dump, node);
		return 1;
	}

	if (strcmp(uname, "ibm,opal") == 0) {
		opal_fadump_dt_scan(&fw_dump, node);
		return 1;
	}

	return 0;
}

/*
 * If fadump is registered, check if the memory provided
 * falls within boot memory area and reserved memory area.
 */
int is_fadump_memory_area(u64 addr, ulong size)
{
	u64 d_start = fw_dump.reserve_dump_area_start;
	u64 d_end = d_start + fw_dump.reserve_dump_area_size;

	if (!fw_dump.dump_registered)
		return 0;

	if (((addr + size) > d_start) && (addr <= d_end))
		return 1;

	return (addr + size) > RMA_START && addr <= fw_dump.boot_memory_size;
}

int should_fadump_crash(void)
{
	if (!fw_dump.dump_registered || !fw_dump.fadumphdr_addr)
		return 0;
	return 1;
}

int is_fadump_active(void)
{
	return fw_dump.dump_active;
}

/*
 * Returns true, if there are no holes in memory area between d_start to d_end,
 * false otherwise.
 */
static bool is_fadump_mem_area_contiguous(u64 d_start, u64 d_end)
{
	struct memblock_region *reg;
	bool ret = false;
	u64 start, end;

	for_each_memblock(memory, reg) {
		start = max_t(u64, d_start, reg->base);
		end = min_t(u64, d_end, (reg->base + reg->size));
		if (d_start < end) {
			/* Memory hole from d_start to start */
			if (start > d_start)
				break;

			if (end == d_end) {
				ret = true;
				break;
			}

			d_start = end + 1;
		}
	}

	return ret;
}

/*
 * Returns true, if there are no holes in boot memory area,
 * false otherwise.
 */
bool is_fadump_boot_mem_contiguous(void)
{
	return is_fadump_mem_area_contiguous(0, fw_dump.boot_memory_size);
}

/*
 * Returns true, if there are no holes in reserved memory area,
 * false otherwise.
 */
bool is_fadump_reserved_mem_contiguous(void)
{
	u64 d_start, d_end;

	d_start	= fw_dump.reserve_dump_area_start;
	d_end	= d_start + fw_dump.reserve_dump_area_size;
	return is_fadump_mem_area_contiguous(d_start, d_end);
}

/* Print firmware assisted dump configurations for debugging purpose. */
static void fadump_show_config(void)
{
	pr_debug("Support for firmware-assisted dump (fadump): %s\n",
			(fw_dump.fadump_supported ? "present" : "no support"));

	if (!fw_dump.fadump_supported)
		return;

	pr_debug("Fadump enabled    : %s\n",
				(fw_dump.fadump_enabled ? "yes" : "no"));
	pr_debug("Dump Active       : %s\n",
				(fw_dump.dump_active ? "yes" : "no"));
	pr_debug("Dump section sizes:\n");
	pr_debug("    CPU state data size: %lx\n", fw_dump.cpu_state_data_size);
	pr_debug("    HPTE region size   : %lx\n", fw_dump.hpte_region_size);
	pr_debug("Boot memory size  : %lx\n", fw_dump.boot_memory_size);
}

/**
 * fadump_calculate_reserve_size(): reserve variable boot area 5% of System RAM
 *
 * Function to find the largest memory size we need to reserve during early
 * boot process. This will be the size of the memory that is required for a
 * kernel to boot successfully.
 *
 * This function has been taken from phyp-assisted dump feature implementation.
 *
 * returns larger of 256MB or 5% rounded down to multiples of 256MB.
 *
 * TODO: Come up with better approach to find out more accurate memory size
 * that is required for a kernel to boot successfully.
 *
 */
static inline unsigned long fadump_calculate_reserve_size(void)
{
	int ret;
	unsigned long long base, size;

	if (fw_dump.reserve_bootvar)
		pr_warn("'fadump_reserve_mem=' parameter is deprecated in favor of 'crashkernel=' parameter.\n");

	/*
	 * Check if the size is specified through crashkernel= cmdline
	 * option. If yes, then use that but ignore base as fadump reserves
	 * memory at a predefined offset.
	 */
	ret = parse_crashkernel(boot_command_line, memblock_phys_mem_size(),
				&size, &base);
	if (ret == 0 && size > 0) {
		unsigned long max_size;

		if (fw_dump.reserve_bootvar)
			pr_info("Using 'crashkernel=' parameter for memory reservation.\n");

		fw_dump.reserve_bootvar = (unsigned long)size;

		/*
		 * Adjust if the boot memory size specified is above
		 * the upper limit.
		 */
		max_size = memblock_phys_mem_size() / MAX_BOOT_MEM_RATIO;
		if (fw_dump.reserve_bootvar > max_size) {
			fw_dump.reserve_bootvar = max_size;
			pr_info("Adjusted boot memory size to %luMB\n",
				(fw_dump.reserve_bootvar >> 20));
		}

		return fw_dump.reserve_bootvar;
	} else if (fw_dump.reserve_bootvar) {
		/*
		 * 'fadump_reserve_mem=' is being used to reserve memory
		 * for firmware-assisted dump.
		 */
		return fw_dump.reserve_bootvar;
	}

	/* divide by 20 to get 5% of value */
	size = memblock_phys_mem_size() / 20;

	/* round it down in multiples of 256 */
	size = size & ~0x0FFFFFFFUL;

	/* Truncate to memory_limit. We don't want to over reserve the memory.*/
	if (memory_limit && size > memory_limit)
		size = memory_limit;

	return (size > MIN_BOOT_MEM ? size : MIN_BOOT_MEM);
}

/*
 * Calculate the total memory size required to be reserved for
 * firmware-assisted dump registration.
 */
static unsigned long get_fadump_area_size(void)
{
	unsigned long size = 0;

	size += fw_dump.cpu_state_data_size;
	size += fw_dump.hpte_region_size;
	size += fw_dump.boot_memory_size;
	size += sizeof(struct fadump_crash_info_header);
	size += sizeof(struct elfhdr); /* ELF core header.*/
	size += sizeof(struct elf_phdr); /* place holder for cpu notes */
	/* Program headers for crash memory regions. */
	size += sizeof(struct elf_phdr) * (memblock_num_regions(memory) + 2);

	size = PAGE_ALIGN(size);

	/* This is to hold kernel metadata on platforms that support it */
	size += (fw_dump.ops->fadump_get_metadata_size ?
		 fw_dump.ops->fadump_get_metadata_size() : 0);
	return size;
}

static void __init fadump_reserve_crash_area(unsigned long base,
					     unsigned long size)
{
	struct memblock_region *reg;
	unsigned long mstart, mend, msize;

	for_each_memblock(memory, reg) {
		mstart = max_t(unsigned long, base, reg->base);
		mend = reg->base + reg->size;
		mend = min(base + size, mend);

		if (mstart < mend) {
			msize = mend - mstart;
			memblock_reserve(mstart, msize);
			pr_info("Reserved %ldMB of memory at %#016lx for saving crash dump\n",
				(msize >> 20), mstart);
		}
	}
}

int __init fadump_reserve_mem(void)
{
	u64 base, size, mem_boundary;
	int ret = 1;

	if (!fw_dump.fadump_enabled)
		return 0;

	if (!fw_dump.fadump_supported) {
		pr_info("Firmware-Assisted Dump is not supported on this hardware\n");
		goto error_out;
	}

	/*
	 * Initialize boot memory size
	 * If dump is active then we have already calculated the size during
	 * first kernel.
	 */
	if (!fw_dump.dump_active) {
		fw_dump.boot_memory_size =
			PAGE_ALIGN(fadump_calculate_reserve_size());
#ifdef CONFIG_CMA
		if (!fw_dump.nocma)
			fw_dump.boot_memory_size =
				ALIGN(fw_dump.boot_memory_size,
							FADUMP_CMA_ALIGNMENT);
#endif
	}

	/*
	 * Calculate the memory boundary.
	 * If memory_limit is less than actual memory boundary then reserve
	 * the memory for fadump beyond the memory_limit and adjust the
	 * memory_limit accordingly, so that the running kernel can run with
	 * specified memory_limit.
	 */
	if (memory_limit && memory_limit < memblock_end_of_DRAM()) {
		size = get_fadump_area_size();
		if ((memory_limit + size) < memblock_end_of_DRAM())
			memory_limit += size;
		else
			memory_limit = memblock_end_of_DRAM();
		printk(KERN_INFO "Adjusted memory_limit for firmware-assisted"
				" dump, now %#016llx\n", memory_limit);
	}
	if (memory_limit)
		mem_boundary = memory_limit;
	else
		mem_boundary = memblock_end_of_DRAM();

	base = fw_dump.boot_memory_size;
	size = get_fadump_area_size();
	fw_dump.reserve_dump_area_size = size;
	if (fw_dump.dump_active) {
		pr_info("Firmware-assisted dump is active.\n");

#ifdef CONFIG_HUGETLB_PAGE
		/*
		 * FADump capture kernel doesn't care much about hugepages.
		 * In fact, handling hugepages in capture kernel is asking for
		 * trouble. So, disable HugeTLB support when fadump is active.
		 */
		hugetlb_disabled = true;
#endif
		/*
		 * If last boot has crashed then reserve all the memory
		 * above boot_memory_size so that we don't touch it until
		 * dump is written to disk by userspace tool. This memory
		 * will be released for general use once the dump is saved.
		 */
		size = mem_boundary - base;
		fadump_reserve_crash_area(base, size);

		pr_debug("fadumphdr_addr = %#016lx\n", fw_dump.fadumphdr_addr);
		pr_debug("Reserve dump area start address: 0x%lx\n",
			 fw_dump.reserve_dump_area_start);
	} else {
		/*
		 * Reserve memory at an offset closer to bottom of the RAM to
		 * minimize the impact of memory hot-remove operation. We can't
		 * use memblock_find_in_range() here since it doesn't allocate
		 * from bottom to top.
		 */
		while (base <= (mem_boundary - size)) {
			if (memblock_is_region_memory(base, size) &&
			    !memblock_is_region_reserved(base, size))
				break;

			base += size;
		}

		if (base > (mem_boundary - size)) {
			pr_err("Failed to find memory chunk for reservation!\n");
			goto error_out;
		}
		fw_dump.reserve_dump_area_start = base;

		/*
		 * Calculate the kernel metadata address and register it with
		 * f/w if the platform supports.
		 */
		if (fw_dump.ops->fadump_setup_metadata &&
		    (fw_dump.ops->fadump_setup_metadata(&fw_dump) < 0))
			goto error_out;

		if (memblock_reserve(base, size)) {
			pr_err("Failed to reserve memory!\n");
			goto error_out;
		}

		pr_info("Reserved %lldMB of memory at %#016llx (System RAM: %lldMB)\n",
			(size >> 20), base, (memblock_phys_mem_size() >> 20));

		ret = fadump_cma_init();
	}

	return ret;
error_out:
	fw_dump.fadump_enabled = 0;
	return 0;
}

unsigned long __init arch_reserved_kernel_pages(void)
{
	return memblock_reserved_size() / PAGE_SIZE;
}

/* Look for fadump= cmdline option. */
static int __init early_fadump_param(char *p)
{
	if (!p)
		return 1;

	if (strncmp(p, "on", 2) == 0)
		fw_dump.fadump_enabled = 1;
	else if (strncmp(p, "off", 3) == 0)
		fw_dump.fadump_enabled = 0;
	else if (strncmp(p, "nocma", 5) == 0) {
		fw_dump.fadump_enabled = 1;
		fw_dump.nocma = 1;
	}

	return 0;
}
early_param("fadump", early_fadump_param);

/*
 * Look for fadump_reserve_mem= cmdline option
 * TODO: Remove references to 'fadump_reserve_mem=' parameter,
 *       the sooner 'crashkernel=' parameter is accustomed to.
 */
static int __init early_fadump_reserve_mem(char *p)
{
	if (p)
		fw_dump.reserve_bootvar = memparse(p, &p);
	return 0;
}
early_param("fadump_reserve_mem", early_fadump_reserve_mem);

void crash_fadump(struct pt_regs *regs, const char *str)
{
	struct fadump_crash_info_header *fdh = NULL;
	int old_cpu, this_cpu;

	if (!should_fadump_crash())
		return;

	/*
	 * old_cpu == -1 means this is the first CPU which has come here,
	 * go ahead and trigger fadump.
	 *
	 * old_cpu != -1 means some other CPU has already on it's way
	 * to trigger fadump, just keep looping here.
	 */
	this_cpu = smp_processor_id();
	old_cpu = cmpxchg(&crashing_cpu, -1, this_cpu);

	if (old_cpu != -1) {
		/*
		 * We can't loop here indefinitely. Wait as long as fadump
		 * is in force. If we race with fadump un-registration this
		 * loop will break and then we go down to normal panic path
		 * and reboot. If fadump is in force the first crashing
		 * cpu will definitely trigger fadump.
		 */
		while (fw_dump.dump_registered)
			cpu_relax();
		return;
	}

	fdh = __va(fw_dump.fadumphdr_addr);
	fdh->crashing_cpu = crashing_cpu;
	crash_save_vmcoreinfo();

	if (regs)
		fdh->regs = *regs;
	else
		ppc_save_regs(&fdh->regs);

	fdh->online_mask = *cpu_online_mask;

	fw_dump.ops->fadump_trigger(fdh, str);
}

u32 *fadump_regs_to_elf_notes(u32 *buf, struct pt_regs *regs)
{
	struct elf_prstatus prstatus;

	memset(&prstatus, 0, sizeof(prstatus));
	/*
	 * FIXME: How do i get PID? Do I really need it?
	 * prstatus.pr_pid = ????
	 */
	elf_core_copy_kernel_regs(&prstatus.pr_reg, regs);
	buf = append_elf_note(buf, CRASH_CORE_NOTE_NAME, NT_PRSTATUS,
			      &prstatus, sizeof(prstatus));
	return buf;
}

void fadump_update_elfcore_header(char *bufp)
{
	struct elfhdr *elf;
	struct elf_phdr *phdr;

	elf = (struct elfhdr *)bufp;
	bufp += sizeof(struct elfhdr);

	/* First note is a place holder for cpu notes info. */
	phdr = (struct elf_phdr *)bufp;

	if (phdr->p_type == PT_NOTE) {
		phdr->p_paddr	= __pa(fw_dump.cpu_notes_buf_vaddr);
		phdr->p_offset	= phdr->p_paddr;
		phdr->p_filesz	= fw_dump.cpu_notes_buf_size;
		phdr->p_memsz = fw_dump.cpu_notes_buf_size;
	}
	return;
}

static void *fadump_alloc_buffer(unsigned long size)
{
	unsigned long count, i;
	struct page *page;
	void *vaddr;

	vaddr = alloc_pages_exact(size, GFP_KERNEL | __GFP_ZERO);
	if (!vaddr)
		return NULL;

	count = PAGE_ALIGN(size) / PAGE_SIZE;
	page = virt_to_page(vaddr);
	for (i = 0; i < count; i++)
		mark_page_reserved(page + i);
	return vaddr;
}

static void fadump_free_buffer(unsigned long vaddr, unsigned long size)
{
	free_reserved_area((void *)vaddr, (void *)(vaddr + size), -1, NULL);
}

s32 fadump_setup_cpu_notes_buf(u32 num_cpus)
{
	/* Allocate buffer to hold cpu crash notes. */
	fw_dump.cpu_notes_buf_size = num_cpus * sizeof(note_buf_t);
	fw_dump.cpu_notes_buf_size = PAGE_ALIGN(fw_dump.cpu_notes_buf_size);
	fw_dump.cpu_notes_buf_vaddr =
		(unsigned long)fadump_alloc_buffer(fw_dump.cpu_notes_buf_size);
	if (!fw_dump.cpu_notes_buf_vaddr) {
		pr_err("Failed to allocate %ld bytes for CPU notes buffer\n",
		       fw_dump.cpu_notes_buf_size);
		return -ENOMEM;
	}

	pr_debug("Allocated buffer for cpu notes of size %ld at 0x%lx\n",
		 fw_dump.cpu_notes_buf_size,
		 fw_dump.cpu_notes_buf_vaddr);
	return 0;
}

void fadump_free_cpu_notes_buf(void)
{
	if (!fw_dump.cpu_notes_buf_vaddr)
		return;

	fadump_free_buffer(fw_dump.cpu_notes_buf_vaddr,
			   fw_dump.cpu_notes_buf_size);
	fw_dump.cpu_notes_buf_vaddr = 0;
	fw_dump.cpu_notes_buf_size = 0;
}

static void free_crash_memory_ranges(void)
{
	kfree(crash_memory_ranges);
	crash_memory_ranges = NULL;
	crash_memory_ranges_size = 0;
	max_crash_mem_ranges = 0;
}

/*
 * Allocate or reallocate crash memory ranges array in incremental units
 * of PAGE_SIZE.
 */
static int allocate_crash_memory_ranges(void)
{
	struct fad_crash_memory_ranges *new_array;
	u64 new_size;

	new_size = crash_memory_ranges_size + PAGE_SIZE;
	pr_debug("Allocating %llu bytes of memory for crash memory ranges\n",
		 new_size);

	new_array = krealloc(crash_memory_ranges, new_size, GFP_KERNEL);
	if (new_array == NULL) {
		pr_err("Insufficient memory for setting up crash memory ranges\n");
		free_crash_memory_ranges();
		return -ENOMEM;
	}

	crash_memory_ranges = new_array;
	crash_memory_ranges_size = new_size;
	max_crash_mem_ranges = (new_size /
				sizeof(struct fad_crash_memory_ranges));
	return 0;
}

static inline int fadump_add_crash_memory(unsigned long long base,
					  unsigned long long end)
{
	u64  start, size;
	bool is_adjacent = false;

	if (base == end)
		return 0;

	/*
	 * Fold adjacent memory ranges to bring down the memory ranges/
	 * PT_LOAD segments count.
	 */
	if (crash_mem_ranges) {
		start = crash_memory_ranges[crash_mem_ranges - 1].base;
		size = crash_memory_ranges[crash_mem_ranges - 1].size;

		if ((start + size) == base)
			is_adjacent = true;
	}
	if (!is_adjacent) {
		/* resize the array on reaching the limit */
		if (crash_mem_ranges == max_crash_mem_ranges) {
			int ret;

			ret = allocate_crash_memory_ranges();
			if (ret)
				return ret;
		}

		start = base;
		crash_memory_ranges[crash_mem_ranges].base = start;
		crash_mem_ranges++;
	}

	crash_memory_ranges[crash_mem_ranges - 1].size = (end - start);
	pr_debug("crash_memory_range[%d] [%#016llx-%#016llx], %#llx bytes\n",
		(crash_mem_ranges - 1), start, end - 1, (end - start));
	return 0;
}

static int fadump_exclude_reserved_area(unsigned long long start,
					unsigned long long end)
{
	unsigned long long ra_start, ra_end;
	int ret = 0;

	ra_start = fw_dump.reserve_dump_area_start;
	ra_end = ra_start + fw_dump.reserve_dump_area_size;

	if ((ra_start < end) && (ra_end > start)) {
		if ((start < ra_start) && (end > ra_end)) {
			ret = fadump_add_crash_memory(start, ra_start);
			if (ret)
				return ret;

			ret = fadump_add_crash_memory(ra_end, end);
		} else if (start < ra_start) {
			ret = fadump_add_crash_memory(start, ra_start);
		} else if (ra_end < end) {
			ret = fadump_add_crash_memory(ra_end, end);
		}
	} else
		ret = fadump_add_crash_memory(start, end);

	return ret;
}

static int fadump_init_elfcore_header(char *bufp)
{
	struct elfhdr *elf;

	elf = (struct elfhdr *) bufp;
	bufp += sizeof(struct elfhdr);
	memcpy(elf->e_ident, ELFMAG, SELFMAG);
	elf->e_ident[EI_CLASS] = ELF_CLASS;
	elf->e_ident[EI_DATA] = ELF_DATA;
	elf->e_ident[EI_VERSION] = EV_CURRENT;
	elf->e_ident[EI_OSABI] = ELF_OSABI;
	memset(elf->e_ident+EI_PAD, 0, EI_NIDENT-EI_PAD);
	elf->e_type = ET_CORE;
	elf->e_machine = ELF_ARCH;
	elf->e_version = EV_CURRENT;
	elf->e_entry = 0;
	elf->e_phoff = sizeof(struct elfhdr);
	elf->e_shoff = 0;
#if defined(_CALL_ELF)
	elf->e_flags = _CALL_ELF;
#else
	elf->e_flags = 0;
#endif
	elf->e_ehsize = sizeof(struct elfhdr);
	elf->e_phentsize = sizeof(struct elf_phdr);
	elf->e_phnum = 0;
	elf->e_shentsize = 0;
	elf->e_shnum = 0;
	elf->e_shstrndx = 0;

	return 0;
}

/*
 * Traverse through memblock structure and setup crash memory ranges. These
 * ranges will be used create PT_LOAD program headers in elfcore header.
 */
static int fadump_setup_crash_memory_ranges(void)
{
	struct memblock_region *reg;
	unsigned long long start, end;
	int ret;

	pr_debug("Setup crash memory ranges.\n");
	crash_mem_ranges = 0;

	/*
	 * add the first memory chunk (RMA_START through boot_memory_size) as
	 * a separate memory chunk. The reason is, at the time crash firmware
	 * will move the content of this memory chunk to different location
	 * specified during fadump registration. We need to create a separate
	 * program header for this chunk with the correct offset.
	 */
	ret = fadump_add_crash_memory(RMA_START, fw_dump.boot_memory_size);
	if (ret)
		return ret;

	for_each_memblock(memory, reg) {
		start = (unsigned long long)reg->base;
		end = start + (unsigned long long)reg->size;

		/*
		 * skip the first memory chunk that is already added (RMA_START
		 * through boot_memory_size). This logic needs a relook if and
		 * when RMA_START changes to a non-zero value.
		 */
		BUILD_BUG_ON(RMA_START != 0);
		if (start < fw_dump.boot_memory_size) {
			if (end > fw_dump.boot_memory_size)
				start = fw_dump.boot_memory_size;
			else
				continue;
		}

		/* add this range excluding the reserved dump area. */
		ret = fadump_exclude_reserved_area(start, end);
		if (ret)
			return ret;
	}

	return 0;
}

/*
 * If the given physical address falls within the boot memory region then
 * return the relocated address that points to the dump region reserved
 * for saving initial boot memory contents.
 */
static inline unsigned long fadump_relocate(unsigned long paddr)
{
	if (paddr > RMA_START && paddr < fw_dump.boot_memory_size)
		return fw_dump.boot_mem_dest_addr + paddr;
	else
		return paddr;
}

static int fadump_create_elfcore_headers(char *bufp)
{
	struct elfhdr *elf;
	struct elf_phdr *phdr;
	int i;

	fadump_init_elfcore_header(bufp);
	elf = (struct elfhdr *)bufp;
	bufp += sizeof(struct elfhdr);

	/*
	 * setup ELF PT_NOTE, place holder for cpu notes info. The notes info
	 * will be populated during second kernel boot after crash. Hence
	 * this PT_NOTE will always be the first elf note.
	 *
	 * NOTE: Any new ELF note addition should be placed after this note.
	 */
	phdr = (struct elf_phdr *)bufp;
	bufp += sizeof(struct elf_phdr);
	phdr->p_type = PT_NOTE;
	phdr->p_flags = 0;
	phdr->p_vaddr = 0;
	phdr->p_align = 0;

	phdr->p_offset = 0;
	phdr->p_paddr = 0;
	phdr->p_filesz = 0;
	phdr->p_memsz = 0;

	(elf->e_phnum)++;

	/* setup ELF PT_NOTE for vmcoreinfo */
	phdr = (struct elf_phdr *)bufp;
	bufp += sizeof(struct elf_phdr);
	phdr->p_type	= PT_NOTE;
	phdr->p_flags	= 0;
	phdr->p_vaddr	= 0;
	phdr->p_align	= 0;

	phdr->p_paddr	= fadump_relocate(paddr_vmcoreinfo_note());
	phdr->p_offset	= phdr->p_paddr;
	phdr->p_memsz	= phdr->p_filesz = VMCOREINFO_NOTE_SIZE;

	/* Increment number of program headers. */
	(elf->e_phnum)++;

	/* setup PT_LOAD sections. */

	for (i = 0; i < crash_mem_ranges; i++) {
		unsigned long long mbase, msize;
		mbase = crash_memory_ranges[i].base;
		msize = crash_memory_ranges[i].size;

		if (!msize)
			continue;

		phdr = (struct elf_phdr *)bufp;
		bufp += sizeof(struct elf_phdr);
		phdr->p_type	= PT_LOAD;
		phdr->p_flags	= PF_R|PF_W|PF_X;
		phdr->p_offset	= mbase;

		if (mbase == RMA_START) {
			/*
			 * The entire RMA region will be moved by firmware
			 * to the specified destination_address. Hence set
			 * the correct offset.
			 */
			phdr->p_offset = fw_dump.boot_mem_dest_addr;
		}

		phdr->p_paddr = mbase;
		phdr->p_vaddr = (unsigned long)__va(mbase);
		phdr->p_filesz = msize;
		phdr->p_memsz = msize;
		phdr->p_align = 0;

		/* Increment number of program headers. */
		(elf->e_phnum)++;
	}
	return 0;
}

static unsigned long init_fadump_header(unsigned long addr)
{
	struct fadump_crash_info_header *fdh;

	if (!addr)
		return 0;

	fdh = __va(addr);
	addr += sizeof(struct fadump_crash_info_header);

	memset(fdh, 0, sizeof(struct fadump_crash_info_header));
	fdh->magic_number = FADUMP_CRASH_INFO_MAGIC;
	fdh->elfcorehdr_addr = addr;
	/* We will set the crashing cpu id in crash_fadump() during crash. */
	fdh->crashing_cpu = FADUMP_CPU_UNKNOWN;

	return addr;
}

static int register_fadump(void)
{
	unsigned long addr;
	void *vaddr;
	int ret;

	/*
	 * If no memory is reserved then we can not register for firmware-
	 * assisted dump.
	 */
	if (!fw_dump.reserve_dump_area_size)
		return -ENODEV;

	ret = fadump_setup_crash_memory_ranges();
	if (ret)
		return ret;

	addr = fw_dump.fadumphdr_addr;

	/* Initialize fadump crash info header. */
	addr = init_fadump_header(addr);
	vaddr = __va(addr);

	pr_debug("Creating ELF core headers at %#016lx\n", addr);
	fadump_create_elfcore_headers(vaddr);

	/* register the future kernel dump with firmware. */
	pr_debug("Registering for firmware-assisted kernel dump...\n");
	return fw_dump.ops->fadump_register(&fw_dump);
}

void fadump_cleanup(void)
{
	/* Invalidate the registration only if dump is active. */
	if (fw_dump.dump_active) {
		pr_debug("Invalidating firmware-assisted dump registration\n");
		fw_dump.ops->fadump_invalidate(&fw_dump);
	} else if (fw_dump.dump_registered) {
		/* Un-register Firmware-assisted dump if it was registered. */
		fw_dump.ops->fadump_unregister(&fw_dump);
		free_crash_memory_ranges();
	}
}

static void fadump_free_reserved_memory(unsigned long start_pfn,
					unsigned long end_pfn)
{
	unsigned long pfn;
	unsigned long time_limit = jiffies + HZ;

	pr_info("freeing reserved memory (0x%llx - 0x%llx)\n",
		PFN_PHYS(start_pfn), PFN_PHYS(end_pfn));

	for (pfn = start_pfn; pfn < end_pfn; pfn++) {
		free_reserved_page(pfn_to_page(pfn));

		if (time_after(jiffies, time_limit)) {
			cond_resched();
			time_limit = jiffies + HZ;
		}
	}
}

/*
 * Skip memory holes and free memory that was actually reserved.
 */
static void fadump_release_reserved_area(unsigned long start, unsigned long end)
{
	struct memblock_region *reg;
	unsigned long tstart, tend;
	unsigned long start_pfn = PHYS_PFN(start);
	unsigned long end_pfn = PHYS_PFN(end);

	for_each_memblock(memory, reg) {
		tstart = max(start_pfn, memblock_region_memory_base_pfn(reg));
		tend = min(end_pfn, memblock_region_memory_end_pfn(reg));
		if (tstart < tend) {
			fadump_free_reserved_memory(tstart, tend);

			if (tend == end_pfn)
				break;

			start_pfn = tend + 1;
		}
	}
}

/*
 * Release the memory that was reserved in early boot to preserve the memory
 * contents. The released memory will be available for general use.
 */
static void fadump_release_memory(unsigned long begin, unsigned long end)
{
	unsigned long ra_start, ra_end;

	ra_start = fw_dump.reserve_dump_area_start;
	ra_end = ra_start + fw_dump.reserve_dump_area_size;

	/*
	 * exclude the dump reserve area. Will reuse it for next
	 * fadump registration.
	 */
	if (begin < ra_end && end > ra_start) {
		if (begin < ra_start)
			fadump_release_reserved_area(begin, ra_start);
		if (end > ra_end)
			fadump_release_reserved_area(ra_end, end);
	} else
		fadump_release_reserved_area(begin, end);
}

static void fadump_invalidate_release_mem(void)
{
	mutex_lock(&fadump_mutex);
	if (!fw_dump.dump_active) {
		mutex_unlock(&fadump_mutex);
		return;
	}

	fadump_cleanup();
	mutex_unlock(&fadump_mutex);

	fadump_release_memory(fw_dump.boot_memory_size, memblock_end_of_DRAM());
	fadump_free_cpu_notes_buf();

	/* Initialize the kernel dump memory structure for FAD registration. */
	fw_dump.ops->fadump_init_mem_struct(&fw_dump);
}

static ssize_t fadump_release_memory_store(struct kobject *kobj,
					struct kobj_attribute *attr,
					const char *buf, size_t count)
{
	int input = -1;

	if (!fw_dump.dump_active)
		return -EPERM;

	if (kstrtoint(buf, 0, &input))
		return -EINVAL;

	if (input == 1) {
		/*
		 * Take away the '/proc/vmcore'. We are releasing the dump
		 * memory, hence it will not be valid anymore.
		 */
#ifdef CONFIG_PROC_VMCORE
		vmcore_cleanup();
#endif
		fadump_invalidate_release_mem();

	} else
		return -EINVAL;
	return count;
}

static ssize_t fadump_enabled_show(struct kobject *kobj,
					struct kobj_attribute *attr,
					char *buf)
{
	return sprintf(buf, "%d\n", fw_dump.fadump_enabled);
}

static ssize_t fadump_register_show(struct kobject *kobj,
					struct kobj_attribute *attr,
					char *buf)
{
	return sprintf(buf, "%d\n", fw_dump.dump_registered);
}

static ssize_t fadump_register_store(struct kobject *kobj,
					struct kobj_attribute *attr,
					const char *buf, size_t count)
{
	int ret = 0;
	int input = -1;

	if (!fw_dump.fadump_enabled || fw_dump.dump_active)
		return -EPERM;

	if (kstrtoint(buf, 0, &input))
		return -EINVAL;

	mutex_lock(&fadump_mutex);

	switch (input) {
	case 0:
		if (fw_dump.dump_registered == 0) {
			goto unlock_out;
		}

		/* Un-register Firmware-assisted dump */
		pr_debug("Un-register firmware-assisted dump\n");
		fw_dump.ops->fadump_unregister(&fw_dump);
		break;
	case 1:
		if (fw_dump.dump_registered == 1) {
			/* Un-register Firmware-assisted dump */
			fw_dump.ops->fadump_unregister(&fw_dump);
		}
		/* Register Firmware-assisted dump */
		ret = register_fadump();
		break;
	default:
		ret = -EINVAL;
		break;
	}

unlock_out:
	mutex_unlock(&fadump_mutex);
	return ret < 0 ? ret : count;
}

static int fadump_region_show(struct seq_file *m, void *private)
{
	if (!fw_dump.fadump_enabled)
		return 0;

	mutex_lock(&fadump_mutex);
	fw_dump.ops->fadump_region_show(&fw_dump, m);
	mutex_unlock(&fadump_mutex);
	return 0;
}

static struct kobj_attribute fadump_release_attr = __ATTR(fadump_release_mem,
						0200, NULL,
						fadump_release_memory_store);
static struct kobj_attribute fadump_attr = __ATTR(fadump_enabled,
						0444, fadump_enabled_show,
						NULL);
static struct kobj_attribute fadump_register_attr = __ATTR(fadump_registered,
						0644, fadump_register_show,
						fadump_register_store);

DEFINE_SHOW_ATTRIBUTE(fadump_region);

static void fadump_init_files(void)
{
	struct dentry *debugfs_file;
	int rc = 0;

	rc = sysfs_create_file(kernel_kobj, &fadump_attr.attr);
	if (rc)
		printk(KERN_ERR "fadump: unable to create sysfs file"
			" fadump_enabled (%d)\n", rc);

	rc = sysfs_create_file(kernel_kobj, &fadump_register_attr.attr);
	if (rc)
		printk(KERN_ERR "fadump: unable to create sysfs file"
			" fadump_registered (%d)\n", rc);

	debugfs_file = debugfs_create_file("fadump_region", 0444,
					powerpc_debugfs_root, NULL,
					&fadump_region_fops);
	if (!debugfs_file)
		printk(KERN_ERR "fadump: unable to create debugfs file"
				" fadump_region\n");

	if (fw_dump.dump_active) {
		rc = sysfs_create_file(kernel_kobj, &fadump_release_attr.attr);
		if (rc)
			printk(KERN_ERR "fadump: unable to create sysfs file"
				" fadump_release_mem (%d)\n", rc);
	}
	return;
}

/*
 * Prepare for firmware-assisted dump.
 */
int __init setup_fadump(void)
{
	if (!fw_dump.fadump_enabled)
		return 0;

	if (!fw_dump.fadump_supported) {
		printk(KERN_ERR "Firmware-assisted dump is not supported on"
			" this hardware\n");
		return 0;
	}

	fadump_show_config();
	/*
	 * If dump data is available then see if it is valid and prepare for
	 * saving it to the disk.
	 */
	if (fw_dump.dump_active) {
		/*
		 * if dump process fails then invalidate the registration
		 * and release memory before proceeding for re-registration.
		 */
		if (fw_dump.ops->fadump_process(&fw_dump) < 0)
			fadump_invalidate_release_mem();
	}
	/* Initialize the kernel dump memory structure for FAD registration. */
	else if (fw_dump.reserve_dump_area_size)
		fw_dump.ops->fadump_init_mem_struct(&fw_dump);

	fadump_init_files();

	return 1;
}
subsys_initcall(setup_fadump);