dump.c

/*
 * QEMU dump
 *
 * Copyright Fujitsu, Corp. 2011, 2012
 *
 * Authors:
 *     Wen Congyang <wency@cn.fujitsu.com>
 *
 * This work is licensed under the terms of the GNU GPL, version 2 or later.
 * See the COPYING file in the top-level directory.
 *
 */

#include "qemu/osdep.h"
#include "qemu/cutils.h"
#include "elf.h"
#include "cpu.h"
#include "exec/cpu-all.h"
#include "exec/hwaddr.h"
#include "monitor/monitor.h"
#include "sysemu/kvm.h"
#include "sysemu/dump.h"
#include "sysemu/sysemu.h"
#include "sysemu/memory_mapping.h"
#include "sysemu/cpus.h"
#include "qapi/qmp/qerror.h"
#include "qmp-commands.h"
#include "qapi-event.h"

#include <zlib.h>
#ifdef CONFIG_LZO
#include <lzo/lzo1x.h>
#endif
#ifdef CONFIG_SNAPPY
#include <snappy-c.h>
#endif
#ifndef ELF_MACHINE_UNAME
#define ELF_MACHINE_UNAME "Unknown"
#endif

uint16_t cpu_to_dump16(DumpState *s, uint16_t val)
{
    if (s->dump_info.d_endian == ELFDATA2LSB) {
        val = cpu_to_le16(val);
    } else {
        val = cpu_to_be16(val);
    }

    return val;
}

uint32_t cpu_to_dump32(DumpState *s, uint32_t val)
{
    if (s->dump_info.d_endian == ELFDATA2LSB) {
        val = cpu_to_le32(val);
    } else {
        val = cpu_to_be32(val);
    }

    return val;
}

uint64_t cpu_to_dump64(DumpState *s, uint64_t val)
{
    if (s->dump_info.d_endian == ELFDATA2LSB) {
        val = cpu_to_le64(val);
    } else {
        val = cpu_to_be64(val);
    }

    return val;
}

static int dump_cleanup(DumpState *s)
{
    guest_phys_blocks_free(&s->guest_phys_blocks);
    memory_mapping_list_free(&s->list);
    close(s->fd);
    if (s->resume) {
        vm_start();
    }

    return 0;
}

static int fd_write_vmcore(const void *buf, size_t size, void *opaque)
{
    DumpState *s = opaque;
    size_t written_size;

    written_size = qemu_write_full(s->fd, buf, size);
    if (written_size != size) {
        return -1;
    }

    return 0;
}

static void write_elf64_header(DumpState *s, Error **errp)
{
    Elf64_Ehdr elf_header;
    int ret;

    memset(&elf_header, 0, sizeof(Elf64_Ehdr));
    memcpy(&elf_header, ELFMAG, SELFMAG);
    elf_header.e_ident[EI_CLASS] = ELFCLASS64;
    elf_header.e_ident[EI_DATA] = s->dump_info.d_endian;
    elf_header.e_ident[EI_VERSION] = EV_CURRENT;
    elf_header.e_type = cpu_to_dump16(s, ET_CORE);
    elf_header.e_machine = cpu_to_dump16(s, s->dump_info.d_machine);
    elf_header.e_version = cpu_to_dump32(s, EV_CURRENT);
    elf_header.e_ehsize = cpu_to_dump16(s, sizeof(elf_header));
    elf_header.e_phoff = cpu_to_dump64(s, sizeof(Elf64_Ehdr));
    elf_header.e_phentsize = cpu_to_dump16(s, sizeof(Elf64_Phdr));
    elf_header.e_phnum = cpu_to_dump16(s, s->phdr_num);
    if (s->have_section) {
        uint64_t shoff = sizeof(Elf64_Ehdr) + sizeof(Elf64_Phdr) * s->sh_info;

        elf_header.e_shoff = cpu_to_dump64(s, shoff);
        elf_header.e_shentsize = cpu_to_dump16(s, sizeof(Elf64_Shdr));
        elf_header.e_shnum = cpu_to_dump16(s, 1);
    }

    ret = fd_write_vmcore(&elf_header, sizeof(elf_header), s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write elf header");
    }
}

static void write_elf32_header(DumpState *s, Error **errp)
{
    Elf32_Ehdr elf_header;
    int ret;

    memset(&elf_header, 0, sizeof(Elf32_Ehdr));
    memcpy(&elf_header, ELFMAG, SELFMAG);
    elf_header.e_ident[EI_CLASS] = ELFCLASS32;
    elf_header.e_ident[EI_DATA] = s->dump_info.d_endian;
    elf_header.e_ident[EI_VERSION] = EV_CURRENT;
    elf_header.e_type = cpu_to_dump16(s, ET_CORE);
    elf_header.e_machine = cpu_to_dump16(s, s->dump_info.d_machine);
    elf_header.e_version = cpu_to_dump32(s, EV_CURRENT);
    elf_header.e_ehsize = cpu_to_dump16(s, sizeof(elf_header));
    elf_header.e_phoff = cpu_to_dump32(s, sizeof(Elf32_Ehdr));
    elf_header.e_phentsize = cpu_to_dump16(s, sizeof(Elf32_Phdr));
    elf_header.e_phnum = cpu_to_dump16(s, s->phdr_num);
    if (s->have_section) {
        uint32_t shoff = sizeof(Elf32_Ehdr) + sizeof(Elf32_Phdr) * s->sh_info;

        elf_header.e_shoff = cpu_to_dump32(s, shoff);
        elf_header.e_shentsize = cpu_to_dump16(s, sizeof(Elf32_Shdr));
        elf_header.e_shnum = cpu_to_dump16(s, 1);
    }

    ret = fd_write_vmcore(&elf_header, sizeof(elf_header), s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write elf header");
    }
}

static void write_elf64_load(DumpState *s, MemoryMapping *memory_mapping,
                             int phdr_index, hwaddr offset,
                             hwaddr filesz, Error **errp)
{
    Elf64_Phdr phdr;
    int ret;

    memset(&phdr, 0, sizeof(Elf64_Phdr));
    phdr.p_type = cpu_to_dump32(s, PT_LOAD);
    phdr.p_offset = cpu_to_dump64(s, offset);
    phdr.p_paddr = cpu_to_dump64(s, memory_mapping->phys_addr);
    phdr.p_filesz = cpu_to_dump64(s, filesz);
    phdr.p_memsz = cpu_to_dump64(s, memory_mapping->length);
    phdr.p_vaddr = cpu_to_dump64(s, memory_mapping->virt_addr);

    assert(memory_mapping->length >= filesz);

    ret = fd_write_vmcore(&phdr, sizeof(Elf64_Phdr), s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write program header table");
    }
}

static void write_elf32_load(DumpState *s, MemoryMapping *memory_mapping,
                             int phdr_index, hwaddr offset,
                             hwaddr filesz, Error **errp)
{
    Elf32_Phdr phdr;
    int ret;

    memset(&phdr, 0, sizeof(Elf32_Phdr));
    phdr.p_type = cpu_to_dump32(s, PT_LOAD);
    phdr.p_offset = cpu_to_dump32(s, offset);
    phdr.p_paddr = cpu_to_dump32(s, memory_mapping->phys_addr);
    phdr.p_filesz = cpu_to_dump32(s, filesz);
    phdr.p_memsz = cpu_to_dump32(s, memory_mapping->length);
    phdr.p_vaddr = cpu_to_dump32(s, memory_mapping->virt_addr);

    assert(memory_mapping->length >= filesz);

    ret = fd_write_vmcore(&phdr, sizeof(Elf32_Phdr), s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write program header table");
    }
}

static void write_elf64_note(DumpState *s, Error **errp)
{
    Elf64_Phdr phdr;
    hwaddr begin = s->memory_offset - s->note_size;
    int ret;

    memset(&phdr, 0, sizeof(Elf64_Phdr));
    phdr.p_type = cpu_to_dump32(s, PT_NOTE);
    phdr.p_offset = cpu_to_dump64(s, begin);
    phdr.p_paddr = 0;
    phdr.p_filesz = cpu_to_dump64(s, s->note_size);
    phdr.p_memsz = cpu_to_dump64(s, s->note_size);
    phdr.p_vaddr = 0;

    ret = fd_write_vmcore(&phdr, sizeof(Elf64_Phdr), s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write program header table");
    }
}

static inline int cpu_index(CPUState *cpu)
{
    return cpu->cpu_index + 1;
}

static void write_elf64_notes(WriteCoreDumpFunction f, DumpState *s,
                              Error **errp)
{
    CPUState *cpu;
    int ret;
    int id;

    CPU_FOREACH(cpu) {
        id = cpu_index(cpu);
        ret = cpu_write_elf64_note(f, cpu, id, s);
        if (ret < 0) {
            error_setg(errp, "dump: failed to write elf notes");
            return;
        }
    }

    CPU_FOREACH(cpu) {
        ret = cpu_write_elf64_qemunote(f, cpu, s);
        if (ret < 0) {
            error_setg(errp, "dump: failed to write CPU status");
            return;
        }
    }
}

static void write_elf32_note(DumpState *s, Error **errp)
{
    hwaddr begin = s->memory_offset - s->note_size;
    Elf32_Phdr phdr;
    int ret;

    memset(&phdr, 0, sizeof(Elf32_Phdr));
    phdr.p_type = cpu_to_dump32(s, PT_NOTE);
    phdr.p_offset = cpu_to_dump32(s, begin);
    phdr.p_paddr = 0;
    phdr.p_filesz = cpu_to_dump32(s, s->note_size);
    phdr.p_memsz = cpu_to_dump32(s, s->note_size);
    phdr.p_vaddr = 0;

    ret = fd_write_vmcore(&phdr, sizeof(Elf32_Phdr), s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write program header table");
    }
}

static void write_elf32_notes(WriteCoreDumpFunction f, DumpState *s,
                              Error **errp)
{
    CPUState *cpu;
    int ret;
    int id;

    CPU_FOREACH(cpu) {
        id = cpu_index(cpu);
        ret = cpu_write_elf32_note(f, cpu, id, s);
        if (ret < 0) {
            error_setg(errp, "dump: failed to write elf notes");
            return;
        }
    }

    CPU_FOREACH(cpu) {
        ret = cpu_write_elf32_qemunote(f, cpu, s);
        if (ret < 0) {
            error_setg(errp, "dump: failed to write CPU status");
            return;
        }
    }
}

static void write_elf_section(DumpState *s, int type, Error **errp)
{
    Elf32_Shdr shdr32;
    Elf64_Shdr shdr64;
    int shdr_size;
    void *shdr;
    int ret;

    if (type == 0) {
        shdr_size = sizeof(Elf32_Shdr);
        memset(&shdr32, 0, shdr_size);
        shdr32.sh_info = cpu_to_dump32(s, s->sh_info);
        shdr = &shdr32;
    } else {
        shdr_size = sizeof(Elf64_Shdr);
        memset(&shdr64, 0, shdr_size);
        shdr64.sh_info = cpu_to_dump32(s, s->sh_info);
        shdr = &shdr64;
    }

    ret = fd_write_vmcore(&shdr, shdr_size, s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to write section header table");
    }
}

static void write_data(DumpState *s, void *buf, int length, Error **errp)
{
    int ret;

    ret = fd_write_vmcore(buf, length, s);
    if (ret < 0) {
        error_setg(errp, "dump: failed to save memory");
    } else {
        s->written_size += length;
    }
}

/* write the memory to vmcore. 1 page per I/O. */
static void write_memory(DumpState *s, GuestPhysBlock *block, ram_addr_t start,
                         int64_t size, Error **errp)
{
    int64_t i;
    Error *local_err = NULL;

    for (i = 0; i < size / s->dump_info.page_size; i++) {
        write_data(s, block->host_addr + start + i * s->dump_info.page_size,
                   s->dump_info.page_size, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }
    }

    if ((size % s->dump_info.page_size) != 0) {
        write_data(s, block->host_addr + start + i * s->dump_info.page_size,
                   size % s->dump_info.page_size, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }
    }
}

/* get the memory's offset and size in the vmcore */
static void get_offset_range(hwaddr phys_addr,
                             ram_addr_t mapping_length,
                             DumpState *s,
                             hwaddr *p_offset,
                             hwaddr *p_filesz)
{
    GuestPhysBlock *block;
    hwaddr offset = s->memory_offset;
    int64_t size_in_block, start;

    /* When the memory is not stored into vmcore, offset will be -1 */
    *p_offset = -1;
    *p_filesz = 0;

    if (s->has_filter) {
        if (phys_addr < s->begin || phys_addr >= s->begin + s->length) {
            return;
        }
    }

    QTAILQ_FOREACH(block, &s->guest_phys_blocks.head, next) {
        if (s->has_filter) {
            if (block->target_start >= s->begin + s->length ||
                block->target_end <= s->begin) {
                /* This block is out of the range */
                continue;
            }

            if (s->begin <= block->target_start) {
                start = block->target_start;
            } else {
                start = s->begin;
            }

            size_in_block = block->target_end - start;
            if (s->begin + s->length < block->target_end) {
                size_in_block -= block->target_end - (s->begin + s->length);
            }
        } else {
            start = block->target_start;
            size_in_block = block->target_end - block->target_start;
        }

        if (phys_addr >= start && phys_addr < start + size_in_block) {
            *p_offset = phys_addr - start + offset;

            /* The offset range mapped from the vmcore file must not spill over
             * the GuestPhysBlock, clamp it. The rest of the mapping will be
             * zero-filled in memory at load time; see
             * <http://refspecs.linuxbase.org/elf/gabi4+/ch5.pheader.html>.
             */
            *p_filesz = phys_addr + mapping_length <= start + size_in_block ?
                        mapping_length :
                        size_in_block - (phys_addr - start);
            return;
        }

        offset += size_in_block;
    }
}

static void write_elf_loads(DumpState *s, Error **errp)
{
    hwaddr offset, filesz;
    MemoryMapping *memory_mapping;
    uint32_t phdr_index = 1;
    uint32_t max_index;
    Error *local_err = NULL;

    if (s->have_section) {
        max_index = s->sh_info;
    } else {
        max_index = s->phdr_num;
    }

    QTAILQ_FOREACH(memory_mapping, &s->list.head, next) {
        get_offset_range(memory_mapping->phys_addr,
                         memory_mapping->length,
                         s, &offset, &filesz);
        if (s->dump_info.d_class == ELFCLASS64) {
            write_elf64_load(s, memory_mapping, phdr_index++, offset,
                             filesz, &local_err);
        } else {
            write_elf32_load(s, memory_mapping, phdr_index++, offset,
                             filesz, &local_err);
        }

        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }

        if (phdr_index >= max_index) {
            break;
        }
    }
}

/* write elf header, PT_NOTE and elf note to vmcore. */
static void dump_begin(DumpState *s, Error **errp)
{
    Error *local_err = NULL;

    /*
     * the vmcore's format is:
     *   --------------
     *   |  elf header |
     *   --------------
     *   |  PT_NOTE    |
     *   --------------
     *   |  PT_LOAD    |
     *   --------------
     *   |  ......     |
     *   --------------
     *   |  PT_LOAD    |
     *   --------------
     *   |  sec_hdr    |
     *   --------------
     *   |  elf note   |
     *   --------------
     *   |  memory     |
     *   --------------
     *
     * we only know where the memory is saved after we write elf note into
     * vmcore.
     */

    /* write elf header to vmcore */
    if (s->dump_info.d_class == ELFCLASS64) {
        write_elf64_header(s, &local_err);
    } else {
        write_elf32_header(s, &local_err);
    }
    if (local_err) {
        error_propagate(errp, local_err);
        return;
    }

    if (s->dump_info.d_class == ELFCLASS64) {
        /* write PT_NOTE to vmcore */
        write_elf64_note(s, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }

        /* write all PT_LOAD to vmcore */
        write_elf_loads(s, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }

        /* write section to vmcore */
        if (s->have_section) {
            write_elf_section(s, 1, &local_err);
            if (local_err) {
                error_propagate(errp, local_err);
                return;
            }
        }

        /* write notes to vmcore */
        write_elf64_notes(fd_write_vmcore, s, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }
    } else {
        /* write PT_NOTE to vmcore */
        write_elf32_note(s, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }

        /* write all PT_LOAD to vmcore */
        write_elf_loads(s, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }

        /* write section to vmcore */
        if (s->have_section) {
            write_elf_section(s, 0, &local_err);
            if (local_err) {
                error_propagate(errp, local_err);
                return;
            }
        }

        /* write notes to vmcore */
        write_elf32_notes(fd_write_vmcore, s, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }
    }
}

static int get_next_block(DumpState *s, GuestPhysBlock *block)
{
    while (1) {
        block = QTAILQ_NEXT(block, next);
        if (!block) {
            /* no more block */
            return 1;
        }

        s->start = 0;
        s->next_block = block;
        if (s->has_filter) {
            if (block->target_start >= s->begin + s->length ||
                block->target_end <= s->begin) {
                /* This block is out of the range */
                continue;
            }

            if (s->begin > block->target_start) {
                s->start = s->begin - block->target_start;
            }
        }

        return 0;
    }
}

/* write all memory to vmcore */
static void dump_iterate(DumpState *s, Error **errp)
{
    GuestPhysBlock *block;
    int64_t size;
    Error *local_err = NULL;

    do {
        block = s->next_block;

        size = block->target_end - block->target_start;
        if (s->has_filter) {
            size -= s->start;
            if (s->begin + s->length < block->target_end) {
                size -= block->target_end - (s->begin + s->length);
            }
        }
        write_memory(s, block, s->start, size, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            return;
        }

    } while (!get_next_block(s, block));
}

static void create_vmcore(DumpState *s, Error **errp)
{
    Error *local_err = NULL;

    dump_begin(s, &local_err);
    if (local_err) {
        error_propagate(errp, local_err);
        return;
    }

    dump_iterate(s, errp);
}

static int write_start_flat_header(int fd)
{
    MakedumpfileHeader *mh;
    int ret = 0;

    QEMU_BUILD_BUG_ON(sizeof *mh > MAX_SIZE_MDF_HEADER);
    mh = g_malloc0(MAX_SIZE_MDF_HEADER);

    memcpy(mh->signature, MAKEDUMPFILE_SIGNATURE,
           MIN(sizeof mh->signature, sizeof MAKEDUMPFILE_SIGNATURE));

    mh->type = cpu_to_be64(TYPE_FLAT_HEADER);
    mh->version = cpu_to_be64(VERSION_FLAT_HEADER);

    size_t written_size;
    written_size = qemu_write_full(fd, mh, MAX_SIZE_MDF_HEADER);
    if (written_size != MAX_SIZE_MDF_HEADER) {
        ret = -1;
    }

    g_free(mh);
    return ret;
}

static int write_end_flat_header(int fd)
{
    MakedumpfileDataHeader mdh;

    mdh.offset = END_FLAG_FLAT_HEADER;
    mdh.buf_size = END_FLAG_FLAT_HEADER;

    size_t written_size;
    written_size = qemu_write_full(fd, &mdh, sizeof(mdh));
    if (written_size != sizeof(mdh)) {
        return -1;
    }

    return 0;
}

static int write_buffer(int fd, off_t offset, const void *buf, size_t size)
{
    size_t written_size;
    MakedumpfileDataHeader mdh;

    mdh.offset = cpu_to_be64(offset);
    mdh.buf_size = cpu_to_be64(size);

    written_size = qemu_write_full(fd, &mdh, sizeof(mdh));
    if (written_size != sizeof(mdh)) {
        return -1;
    }

    written_size = qemu_write_full(fd, buf, size);
    if (written_size != size) {
        return -1;
    }

    return 0;
}

static int buf_write_note(const void *buf, size_t size, void *opaque)
{
    DumpState *s = opaque;

    /* note_buf is not enough */
    if (s->note_buf_offset + size > s->note_size) {
        return -1;
    }

    memcpy(s->note_buf + s->note_buf_offset, buf, size);

    s->note_buf_offset += size;

    return 0;
}

/* write common header, sub header and elf note to vmcore */
static void create_header32(DumpState *s, Error **errp)
{
    DiskDumpHeader32 *dh = NULL;
    KdumpSubHeader32 *kh = NULL;
    size_t size;
    uint32_t block_size;
    uint32_t sub_hdr_size;
    uint32_t bitmap_blocks;
    uint32_t status = 0;
    uint64_t offset_note;
    Error *local_err = NULL;

    /* write common header, the version of kdump-compressed format is 6th */
    size = sizeof(DiskDumpHeader32);
    dh = g_malloc0(size);

    strncpy(dh->signature, KDUMP_SIGNATURE, strlen(KDUMP_SIGNATURE));
    dh->header_version = cpu_to_dump32(s, 6);
    block_size = s->dump_info.page_size;
    dh->block_size = cpu_to_dump32(s, block_size);
    sub_hdr_size = sizeof(struct KdumpSubHeader32) + s->note_size;
    sub_hdr_size = DIV_ROUND_UP(sub_hdr_size, block_size);
    dh->sub_hdr_size = cpu_to_dump32(s, sub_hdr_size);
    /* dh->max_mapnr may be truncated, full 64bit is in kh.max_mapnr_64 */
    dh->max_mapnr = cpu_to_dump32(s, MIN(s->max_mapnr, UINT_MAX));
    dh->nr_cpus = cpu_to_dump32(s, s->nr_cpus);
    bitmap_blocks = DIV_ROUND_UP(s->len_dump_bitmap, block_size) * 2;
    dh->bitmap_blocks = cpu_to_dump32(s, bitmap_blocks);
    strncpy(dh->utsname.machine, ELF_MACHINE_UNAME, sizeof(dh->utsname.machine));

    if (s->flag_compress & DUMP_DH_COMPRESSED_ZLIB) {
        status |= DUMP_DH_COMPRESSED_ZLIB;
    }
#ifdef CONFIG_LZO
    if (s->flag_compress & DUMP_DH_COMPRESSED_LZO) {
        status |= DUMP_DH_COMPRESSED_LZO;
    }
#endif
#ifdef CONFIG_SNAPPY
    if (s->flag_compress & DUMP_DH_COMPRESSED_SNAPPY) {
        status |= DUMP_DH_COMPRESSED_SNAPPY;
    }
#endif
    dh->status = cpu_to_dump32(s, status);

    if (write_buffer(s->fd, 0, dh, size) < 0) {
        error_setg(errp, "dump: failed to write disk dump header");
        goto out;
    }

    /* write sub header */
    size = sizeof(KdumpSubHeader32);
    kh = g_malloc0(size);

    /* 64bit max_mapnr_64 */
    kh->max_mapnr_64 = cpu_to_dump64(s, s->max_mapnr);
    kh->phys_base = cpu_to_dump32(s, s->dump_info.phys_base);
    kh->dump_level = cpu_to_dump32(s, DUMP_LEVEL);

    offset_note = DISKDUMP_HEADER_BLOCKS * block_size + size;
    kh->offset_note = cpu_to_dump64(s, offset_note);
    kh->note_size = cpu_to_dump32(s, s->note_size);

    if (write_buffer(s->fd, DISKDUMP_HEADER_BLOCKS *
                     block_size, kh, size) < 0) {
        error_setg(errp, "dump: failed to write kdump sub header");
        goto out;
    }

    /* write note */
    s->note_buf = g_malloc0(s->note_size);
    s->note_buf_offset = 0;

    /* use s->note_buf to store notes temporarily */
    write_elf32_notes(buf_write_note, s, &local_err);
    if (local_err) {
        error_propagate(errp, local_err);
        goto out;
    }
    if (write_buffer(s->fd, offset_note, s->note_buf,
                     s->note_size) < 0) {
        error_setg(errp, "dump: failed to write notes");
        goto out;
    }

    /* get offset of dump_bitmap */
    s->offset_dump_bitmap = (DISKDUMP_HEADER_BLOCKS + sub_hdr_size) *
                             block_size;

    /* get offset of page */
    s->offset_page = (DISKDUMP_HEADER_BLOCKS + sub_hdr_size + bitmap_blocks) *
                     block_size;

out:
    g_free(dh);
    g_free(kh);
    g_free(s->note_buf);
}

/* write common header, sub header and elf note to vmcore */
static void create_header64(DumpState *s, Error **errp)
{
    DiskDumpHeader64 *dh = NULL;
    KdumpSubHeader64 *kh = NULL;
    size_t size;
    uint32_t block_size;
    uint32_t sub_hdr_size;
    uint32_t bitmap_blocks;
    uint32_t status = 0;
    uint64_t offset_note;
    Error *local_err = NULL;

    /* write common header, the version of kdump-compressed format is 6th */
    size = sizeof(DiskDumpHeader64);
    dh = g_malloc0(size);

    strncpy(dh->signature, KDUMP_SIGNATURE, strlen(KDUMP_SIGNATURE));
    dh->header_version = cpu_to_dump32(s, 6);
    block_size = s->dump_info.page_size;
    dh->block_size = cpu_to_dump32(s, block_size);
    sub_hdr_size = sizeof(struct KdumpSubHeader64) + s->note_size;
    sub_hdr_size = DIV_ROUND_UP(sub_hdr_size, block_size);
    dh->sub_hdr_size = cpu_to_dump32(s, sub_hdr_size);
    /* dh->max_mapnr may be truncated, full 64bit is in kh.max_mapnr_64 */
    dh->max_mapnr = cpu_to_dump32(s, MIN(s->max_mapnr, UINT_MAX));
    dh->nr_cpus = cpu_to_dump32(s, s->nr_cpus);
    bitmap_blocks = DIV_ROUND_UP(s->len_dump_bitmap, block_size) * 2;
    dh->bitmap_blocks = cpu_to_dump32(s, bitmap_blocks);
    strncpy(dh->utsname.machine, ELF_MACHINE_UNAME, sizeof(dh->utsname.machine));

    if (s->flag_compress & DUMP_DH_COMPRESSED_ZLIB) {
        status |= DUMP_DH_COMPRESSED_ZLIB;
    }
#ifdef CONFIG_LZO
    if (s->flag_compress & DUMP_DH_COMPRESSED_LZO) {
        status |= DUMP_DH_COMPRESSED_LZO;
    }
#endif
#ifdef CONFIG_SNAPPY
    if (s->flag_compress & DUMP_DH_COMPRESSED_SNAPPY) {
        status |= DUMP_DH_COMPRESSED_SNAPPY;
    }
#endif
    dh->status = cpu_to_dump32(s, status);

    if (write_buffer(s->fd, 0, dh, size) < 0) {
        error_setg(errp, "dump: failed to write disk dump header");
        goto out;
    }

    /* write sub header */
    size = sizeof(KdumpSubHeader64);
    kh = g_malloc0(size);

    /* 64bit max_mapnr_64 */
    kh->max_mapnr_64 = cpu_to_dump64(s, s->max_mapnr);
    kh->phys_base = cpu_to_dump64(s, s->dump_info.phys_base);
    kh->dump_level = cpu_to_dump32(s, DUMP_LEVEL);

    offset_note = DISKDUMP_HEADER_BLOCKS * block_size + size;
    kh->offset_note = cpu_to_dump64(s, offset_note);
    kh->note_size = cpu_to_dump64(s, s->note_size);

    if (write_buffer(s->fd, DISKDUMP_HEADER_BLOCKS *
                     block_size, kh, size) < 0) {
        error_setg(errp, "dump: failed to write kdump sub header");
        goto out;
    }

    /* write note */
    s->note_buf = g_malloc0(s->note_size);
    s->note_buf_offset = 0;

    /* use s->note_buf to store notes temporarily */
    write_elf64_notes(buf_write_note, s, &local_err);
    if (local_err) {
        error_propagate(errp, local_err);
        goto out;
    }

    if (write_buffer(s->fd, offset_note, s->note_buf,
                     s->note_size) < 0) {
        error_setg(errp, "dump: failed to write notes");
        goto out;
    }

    /* get offset of dump_bitmap */
    s->offset_dump_bitmap = (DISKDUMP_HEADER_BLOCKS + sub_hdr_size) *
                             block_size;

    /* get offset of page */
    s->offset_page = (DISKDUMP_HEADER_BLOCKS + sub_hdr_size + bitmap_blocks) *
                     block_size;

out:
    g_free(dh);
    g_free(kh);
    g_free(s->note_buf);
}

static void write_dump_header(DumpState *s, Error **errp)
{
     Error *local_err = NULL;

    if (s->dump_info.d_class == ELFCLASS32) {
        create_header32(s, &local_err);
    } else {
        create_header64(s, &local_err);
    }
    error_propagate(errp, local_err);
}

static size_t dump_bitmap_get_bufsize(DumpState *s)
{
    return s->dump_info.page_size;
}

/*
 * set dump_bitmap sequencely. the bit before last_pfn is not allowed to be
 * rewritten, so if need to set the first bit, set last_pfn and pfn to 0.
 * set_dump_bitmap will always leave the recently set bit un-sync. And setting
 * (last bit + sizeof(buf) * 8) to 0 will do flushing the content in buf into
 * vmcore, ie. synchronizing un-sync bit into vmcore.
 */
static int set_dump_bitmap(uint64_t last_pfn, uint64_t pfn, bool value,
                           uint8_t *buf, DumpState *s)
{
    off_t old_offset, new_offset;
    off_t offset_bitmap1, offset_bitmap2;
    uint32_t byte, bit;
    size_t bitmap_bufsize = dump_bitmap_get_bufsize(s);
    size_t bits_per_buf = bitmap_bufsize * CHAR_BIT;

    /* should not set the previous place */
    assert(last_pfn <= pfn);

    /*
     * if the bit needed to be set is not cached in buf, flush the data in buf
     * to vmcore firstly.
     * making new_offset be bigger than old_offset can also sync remained data
     * into vmcore.
     */
    old_offset = bitmap_bufsize * (last_pfn / bits_per_buf);
    new_offset = bitmap_bufsize * (pfn / bits_per_buf);

    while (old_offset < new_offset) {
        /* calculate the offset and write dump_bitmap */
        offset_bitmap1 = s->offset_dump_bitmap + old_offset;
        if (write_buffer(s->fd, offset_bitmap1, buf,
                         bitmap_bufsize) < 0) {
            return -1;
        }

        /* dump level 1 is chosen, so 1st and 2nd bitmap are same */
        offset_bitmap2 = s->offset_dump_bitmap + s->len_dump_bitmap +
                         old_offset;
        if (write_buffer(s->fd, offset_bitmap2, buf,
                         bitmap_bufsize) < 0) {
            return -1;
        }

        memset(buf, 0, bitmap_bufsize);
        old_offset += bitmap_bufsize;
    }

    /* get the exact place of the bit in the buf, and set it */
    byte = (pfn % bits_per_buf) / CHAR_BIT;
    bit = (pfn % bits_per_buf) % CHAR_BIT;
    if (value) {
        buf[byte] |= 1u << bit;
    } else {
        buf[byte] &= ~(1u << bit);
    }

    return 0;
}

static uint64_t dump_paddr_to_pfn(DumpState *s, uint64_t addr)
{
    int target_page_shift = ctz32(s->dump_info.page_size);

    return (addr >> target_page_shift) - ARCH_PFN_OFFSET;
}

static uint64_t dump_pfn_to_paddr(DumpState *s, uint64_t pfn)
{
    int target_page_shift = ctz32(s->dump_info.page_size);

    return (pfn + ARCH_PFN_OFFSET) << target_page_shift;