启动期间的内存管理之build_zonelists初始化备用内存域列表zonelists--Linux内存管理(十三)

bootmem_init
始化内存数据结构包括内存节点, 内存域和页帧page
|
|---->arm64_numa_init();
|     支持numa架构
|
|---->zone_sizes_init(min, max);
    来初始化节点和管理区的一些数据项
    |
    |---->free_area_init_node
    |   初始化内存节点
    |
        |---->free_area_init_core
            |   初始化zone
            |
            |---->memmap_init
            |   初始化page页面
|
|---->memblock_dump_all();
|   初始化完成, 显示memblock的保留的所有内存信息

asmlinkage __visible void __init start_kernel(void)
{

    setup_arch(&command_line);


    build_all_zonelists(NULL, NULL);
    page_alloc_init();


    /*
     * These use large bootmem allocations and must precede
     * mem_init();
     * kmem_cache_init();
     */
    mm_init();

    kmem_cache_init_late();

    kmemleak_init();
    setup_per_cpu_pageset();

    rest_init();
}

//  http://lxr.free-electrons.com/source/include/linux/mmzone.h?v=4.7#L626
typedef struct pglist_data
{
    struct zone node_zones[MAX_NR_ZONES];
    struct zonelist node_zonelists[MAX_ZONELISTS];
}

  build_all_zonelists(NULL, NULL);

/*
 * Called with zonelists_mutex held always
 * unless system_state == SYSTEM_BOOTING.
 *
 * __ref due to (1) call of __meminit annotated setup_zone_pageset
 * [we're only called with non-NULL zone through __meminit paths] and
 * (2) call of __init annotated helper build_all_zonelists_init
 * [protected by SYSTEM_BOOTING].
 */
void __ref build_all_zonelists(pg_data_t *pgdat, struct zone *zone)
{
    /*  设置zonelist中节点和内存域的组织形式
     *  current_zonelist_order变量标识了当前系统的内存组织形式
     *  zonelist_order_name以字符串存储了系统中内存组织形式的名称  */
    set_zonelist_order();

    if (system_state == SYSTEM_BOOTING) {
        build_all_zonelists_init();
    } else {
#ifdef CONFIG_MEMORY_HOTPLUG
        if (zone)
            setup_zone_pageset(zone);
#endif
        /* we have to stop all cpus to guarantee there is no user
           of zonelist */
        stop_machine(__build_all_zonelists, pgdat, NULL);
        /* cpuset refresh routine should be here */
    }
    vm_total_pages = nr_free_pagecache_pages();
    /*
     * Disable grouping by mobility if the number of pages in the
     * system is too low to allow the mechanism to work. It would be
     * more accurate, but expensive to check per-zone. This check is
     * made on memory-hotadd so a system can start with mobility
     * disabled and enable it later
     */
    if (vm_total_pages < (pageblock_nr_pages * MIGRATE_TYPES))
        page_group_by_mobility_disabled = 1;
    else
        page_group_by_mobility_disabled = 0;

    pr_info("Built %i zonelists in %s order, mobility grouping %s.  Total pages: %ldn",
        nr_online_nodes,
        zonelist_order_name[current_zonelist_order],
        page_group_by_mobility_disabled ? "off" : "on",
        vm_total_pages);
#ifdef CONFIG_NUMA
    pr_info("Policy zone: %sn", zone_names[policy_zone]);
#endif
}

void __ref build_all_zonelists(pg_data_t *pgdat, struct zone *zone)
{
    set_zonelist_order();
    /* .......  */
}

/*
 *
 * The zone_table array is used to look up the address of the
 * struct zone corresponding to a given zone number (ZONE_DMA,
 * ZONE_NORMAL, or ZONE_HIGHMEM).
 */
zone_t *zone_table[MAX_NR_ZONES*MAX_NR_NODES];
EXPORT_SYMBOL(zone_table);

// http://lxr.free-electrons.com/source/mm/page_alloc.c?v=4.7#L4551
/*
 *  zonelist_order:
 *  0 = automatic detection of better ordering.
 *  1 = order by ([node] distance, -zonetype)
 *  2 = order by (-zonetype, [node] distance)
 *
 *  If not NUMA, ZONELIST_ORDER_ZONE and ZONELIST_ORDER_NODE will create
 *  the same zonelist. So only NUMA can configure this param.
 */
#define ZONELIST_ORDER_DEFAULT  0 /* 智能选择Node或Zone方式 */

#define ZONELIST_ORDER_NODE     1 /* 对应Node方式 */

#define ZONELIST_ORDER_ZONE     2 /* 对应Zone方式 */

//  http://lxr.free-electrons.com/source/mm/page_alloc.c?v=4.7#L4564
/* zonelist order in the kernel.
 * set_zonelist_order() will set this to NODE or ZONE.
 */
static int current_zonelist_order = ZONELIST_ORDER_DEFAULT;
static char zonelist_order_name[3][8] = {"Default", "Node", "Zone"};

// http://lxr.free-electrons.com/source/mm/page_alloc.c?v=4.7#L4571
#ifdef CONFIG_NUMA
/* The value user specified ....changed by config */
static int user_zonelist_order = ZONELIST_ORDER_DEFAULT;
/* string for sysctl */
#define NUMA_ZONELIST_ORDER_LEN 16
char numa_zonelist_order[16] = "default";


//  http://lxr.free-electrons.com/source/mm/page_alloc.c#L4571
static void set_zonelist_order(void)
{
    if (user_zonelist_order == ZONELIST_ORDER_DEFAULT)
        current_zonelist_order = default_zonelist_order();
    else
        current_zonelist_order = user_zonelist_order;
}


#else   /* CONFIG_NUMA */

//  http://lxr.free-electrons.com/source/mm/page_alloc.c?v=4.7#L4892
static void set_zonelist_order(void)
{
    current_zonelist_order = ZONELIST_ORDER_ZONE;
}

#if defined(CONFIG_64BIT)
/*
 * Devices that require DMA32/DMA are relatively rare and do not justify a
 * penalty to every machine in case the specialised case applies. Default
 * to Node-ordering on 64-bit NUMA machines
 */
static int default_zonelist_order(void)
{
    return ZONELIST_ORDER_NODE;
}
#else
/*
 * On 32-bit, the Normal zone needs to be preserved for allocations accessible
 * by the kernel. If processes running on node 0 deplete the low memory zone
 * then reclaim will occur more frequency increasing stalls and potentially
 * be easier to OOM if a large percentage of the zone is under writeback or
 * dirty. The problem is significantly worse if CONFIG_HIGHPTE is not set.
 * Hence, default to zone ordering on 32-bit.
 */
static int default_zonelist_order(void)
{
    return ZONELIST_ORDER_ZONE;
}
#endif /* CONFIG_64BIT */

#ifdef CONFIG_NUMA
/* The value user specified ....changed by config */
static int user_zonelist_order = ZONELIST_ORDER_DEFAULT;
/* string for sysctl */
#define NUMA_ZONELIST_ORDER_LEN 16
char numa_zonelist_order[16] = "default";

#else
/* ......*/
#endif

/*
 * interface for configure zonelist ordering.
 * command line option "numa_zonelist_order"
 *      = "[dD]efault   - default, automatic configuration.
 *      = "[nN]ode      - order by node locality, then by zone within node
 *      = "[zZ]one      - order by zone, then by locality within zone
 */

static int __parse_numa_zonelist_order(char *s)
{
    if (*s == 'd' || *s == 'D') {
        user_zonelist_order = ZONELIST_ORDER_DEFAULT;
    } else if (*s == 'n' || *s == 'N') {
        user_zonelist_order = ZONELIST_ORDER_NODE;
    } else if (*s == 'z' || *s == 'Z') {
        user_zonelist_order = ZONELIST_ORDER_ZONE;
    } else {
        pr_warn("Ignoring invalid numa_zonelist_order value:  %sn", s);
        return -EINVAL;
    }
    return 0;
}

static __init int setup_numa_zonelist_order(char *s)
{
    int ret;

    if (!s)
        return 0;

    ret = __parse_numa_zonelist_order(s);
    if (ret == 0)
        strlcpy(numa_zonelist_order, s, NUMA_ZONELIST_ORDER_LEN);

    return ret;
}
early_param("numa_zonelist_order", setup_numa_zonelist_order);

build_all_zonelists(pg_data_t *pgdat, struct zone *zone)
{
    /*  设置zonelist中节点和内存域的组织形式
     *  current_zonelist_order变量标识了当前系统的内存组织形式
     *  zonelist_order_name以字符串存储了系统中内存组织形式的名称  */
    set_zonelist_order();

    if (system_state == SYSTEM_BOOTING) {
        build_all_zonelists_init();

/* Values used for system_state */
extern enum system_states
{
    SYSTEM_BOOTING,
    SYSTEM_RUNNING,
    SYSTEM_HALT,
    SYSTEM_POWER_OFF,
    SYSTEM_RESTART,
} system_state;

if (system_state == SYSTEM_BOOTING)
{
    build_all_zonelists_init();
}
else
{
#ifdef CONFIG_MEMORY_HOTPLUG
    if (zone)
        setup_zone_pageset(zone);
#endif

static noinline void __init
build_all_zonelists_init(void)
{
    __build_all_zonelists(NULL);
    mminit_verify_zonelist();
    cpuset_init_current_mems_allowed();
}

/* return values int ....just for stop_machine() */
static int __build_all_zonelists(void *data)
{
    int nid;
    int cpu;
    pg_data_t *self = data;

    /*  ......  */

    for_each_online_node(nid) {
        pg_data_t *pgdat = NODE_DATA(nid);

        build_zonelists(pgdat);
    }
    /*  ......  */
}

// http://lxr.free-electrons.com/source/mm/page_alloc.c?v=4.7
4571 #ifdef CONFIG_NUMA

4586 static int __parse_numa_zonelist_order(char *s)

4601 static __init int setup_numa_zonelist_order(char *s)

4619 int numa_zonelist_order_handler(struct ctl_table *table, int write,
4620                 void __user *buffer, size_t *length,

4678 static int find_next_best_node(int node, nodemask_t *used_node_mask)

4730 static void build_zonelists_in_node_order(pg_data_t *pgdat, int node)

4746 static void build_thisnode_zonelists(pg_data_t *pgdat)

4765 static void build_zonelists_in_zone_order(pg_data_t *pgdat, int nr_nodes)

4789 #if defined(CONFIG_64BIT)

4795 static int default_zonelist_order(void)
4799 #else
4808 static int default_zonelist_order(void)
4812 #endif /* CONFIG_64BIT */

4822 static void build_zonelists(pg_data_t *pgdat)

4872 #ifdef CONFIG_HAVE_MEMORYLESS_NODES
4879 int local_memory_node(int node)
4888 #endif

4890 #else   /* CONFIG_NUMA */

4897 static void build_zonelists(pg_data_t *pgdat)

4892 static void set_zonelist_order(void)

4931 #endif  /* CONFIG_NUMA */

/*
 * Builds allocation fallback zone lists.
 *
 * Add all populated zones of a node to the zonelist.
 */
static int build_zonelists_node(pg_data_t *pgdat, struct zonelist *zonelist, int nr_zones)
{
    struct zone *zone;
    enum zone_type zone_type = MAX_NR_ZONES;

    do {
        zone_type--;
        zone = pgdat->node_zones + zone_type;
        if (populated_zone(zone)) {
            zoneref_set_zone(zone,
                &zonelist->_zonerefs[nr_zones++]);
            check_highest_zone(zone_type);
        }
    } while (zone_type);

    return nr_zones;
}

zonelist->zones[0] = ZONE_HIGHMEM;
zonelist->zones[1] = ZONE_NORMAL;
zonelist->zones[2] = ZONE_DMA;

其中
A=（NUMA）结点0 0=DMA内存域
B=（NUMA）结点1 1=普通内存域
C=（NUMA）结点2 2=高端内存域
D=（NUMA）结点3

static void build_zonelists(pg_data_t *pgdat)
{
    int node, local_node;
    enum zone_type j;
    struct zonelist *zonelist;

    /*  ......  */

    for (node = local_node + 1; node < MAX_NUMNODES; node++) {
        if (!node_online(node))
            continue;
        j = build_zonelists_node(NODE_DATA(node), zonelist, j);
    }
    for (node = 0; node < local_node; node++) {
        if (!node_online(node))
            continue;
        j = build_zonelists_node(NODE_DATA(node), zonelist, j);
    }

    zonelist->_zonerefs[j].zone = NULL;
    zonelist->_zonerefs[j].zone_idx = 0;
}

static void setup_pageset(struct per_cpu_pageset *p, unsigned long batch)
{
    pageset_init(p);
    pageset_set_batch(p, batch);
}

static __meminit void zone_pcp_init(struct zone *zone)
{
    /*
     * per cpu subsystem is not up at this point. The following code
     * relies on the ability of the linker to provide the
     * offset of a (static) per cpu variable into the per cpu area.
     */
    zone->pageset = &boot_pageset;

    if (populated_zone(zone))
        printk(KERN_DEBUG "  %s zone: %lu pages, LIFO batch:%un",
            zone->name, zone->present_pages,
                     zone_batchsize(zone));
}

start_kernel()
    |---->page_address_init()
    |     考虑支持高端内存
    |     业务：初始化page_address_pool链表；
    |          将page_address_maps数组元素按索引降序插入
    |          page_address_pool链表; 
    |          初始化page_address_htable数组.
    | 
    |---->setup_arch(&command_line);
    |
    |---->setup_per_cpu_areas();
    |     为per-CPU变量分配空间
    |
    |---->build_all_zonelist()
    |     为系统中的zone建立后备zone的列表.
    |     所有zone的后备列表都在
    |     pglist_data->node_zonelists[0]中;
    |
    |     期间也对per-CPU变量boot_pageset做了初始化. 
    |
    |---->page_alloc_init()
         |---->hotcpu_notifier(page_alloc_cpu_notifier, 0);
         |     不考虑热插拔CPU 
         |
    |---->pidhash_init()
    |     详见下文.
    |     根据低端内存页数和散列度，分配hash空间，并赋予pid_hash
    |
    |---->vfs_caches_init_early()
          |---->dcache_init_early()
          |     dentry_hashtable空间，d_hash_shift, h_hash_mask赋值；
          |     同pidhash_init();
          |     区别:
          |         散列度变化了（13 - PAGE_SHIFT）;
          |         传入alloc_large_system_hash的最后参数值为0;
          |
          |---->inode_init_early()
          |     inode_hashtable空间，i_hash_shift, i_hash_mask赋值；
          |     同pidhash_init();
          |     区别:
          |         散列度变化了（14 - PAGE_SHIFT）;
          |         传入alloc_large_system_hash的最后参数值为0;
          |

void pidhash_init(void)
    |---->pid_hash = alloc_large_system_hash("PID", sizeof(*pid_hash), 
    |         0, 18, HASH_EARLY|HASH_SMALL, &pidhash_shift, NULL, 4096);
    |     根据nr_kernel_pages(低端内存的页数)，分配哈希数组，以及各个哈希
    |     数组元素下的哈希链表的空间，原理如下：
    |     number = nr_kernel_pages; 
    |     number >= (18 - PAGE_SHIFT) 根据散列度获得数组元素个数
    |     number = roundup_pow_of_two(number);
    |     pidhash_shift = max{x | 2**x <= number}
    |     size = number * sizeof(*pid_hash);
    |     使用位图分配器分配size空间，将返回值付给pid_hash;
    |
    |---->pidhash_size = 1 << pidhash_shift;
    |
    |---->for(i = 0; i < pidhash_size; i++)
    |         INIT_HLIST_HEAD(&pid_hash[i]);

void build_all_zonelists(void)
    |---->set_zonelist_order()
         |---->current_zonelist_order = ZONELIST_ORDER_ZONE;
    |
    |---->__build_all_zonelists(NULL);
    |    Memory不支持热插拔, 为每个zone建立后备的zone,
    |    每个zone及自己后备的zone，形成zonelist
        |
        |---->pg_data_t *pgdat = NULL;
        |     pgdat = &contig_page_data;(单node)
        |
        |---->build_zonelists(pgdat);
        |     为每个zone建立后备zone的列表
            |
            |---->struct zonelist *zonelist = NULL;
            |     enum zone_type j;
            |     zonelist = &pgdat->node_zonelists[0];
            |
            |---->j = build_zonelists_node(pddat, zonelist, 0, MAX_NR_ZONES - 1);
            |     为pgdat->node_zones[0]建立后备的zone，node_zones[0]后备的zone
            |     存储在node_zonelist[0]内，对于node_zone[0]的后备zone，其后备的zone
            |     链表如下(只考虑UMA体系，而且不考虑ZONE_DMA)：
            |     node_zonelist[0]._zonerefs[0].zone = &node_zones[2];
            |     node_zonelist[0]._zonerefs[0].zone_idx = 2;
            |     node_zonelist[0]._zonerefs[1].zone = &node_zones[1];
            |     node_zonelist[0]._zonerefs[1].zone_idx = 1;
            |     node_zonelist[0]._zonerefs[2].zone = &node_zones[0];
            |     node_zonelist[0]._zonerefs[2].zone_idx = 0;
            |
            |     zonelist->_zonerefs[3].zone = NULL;
            |     zonelist->_zonerefs[3].zone_idx = 0;
        |
        |---->build_zonelist_cache(pgdat);
              |---->pdat->node_zonelists[0].zlcache_ptr = NULL;
              |     UMA体系结构
              |
        |---->for_each_possible_cpu(cpu)
        |     setup_pageset(&per_cpu(boot_pageset, cpu), 0);
              |详见下文
    |---->vm_total_pages = nr_free_pagecache_pages();
    |    业务：获得所有zone中的present_pages总和.
    |
    |---->page_group_by_mobility_disabled = 0;
    |     对于代码中的判断条件一般不会成立，因为页数会最够多（内存较大）