【Linux内存管理】伙伴管理算法（2）

前面已经分析了linux内存管理算法（伙伴管理算法）的准备工作。

具体的算法初始化则回到start_kernel()函数接着往下走，下一个函数是mm_init()：

【file：/init/main.c】
/*
 * Set up kernel memory allocators
 */
static void __init mm_init(void)
{
    /*
     * page_cgroup requires contiguous pages,
     * bigger than MAX_ORDER unless SPARSEMEM.
     */
    page_cgroup_init_flatmem();
    mem_init();
    kmem_cache_init();
    percpu_init_late();
    pgtable_init();
    vmalloc_init();
}

乍看仅仅是几个函数的调用，实际上这里的事情远远没这么简单。其中page_cgroup_init_flatmem()与cgroup相关，而mem_init()则是管理伙伴管理算法的初始化，此外kmem_cache_init()是用于内核slub内存分配体系的初始化，而vmalloc_init()则是用于vmalloc的初始化。

当前主要分析伙伴管理算法，则仅对mem_init()做专门的分析，其余的暂且后面再分析。

伙伴管理算法的初始化函数入口是mem_init()，其实现：

【file：/arch/x86/mm/init_32.c】
void __init mem_init(void)
{
    pci_iommu_alloc();

#ifdef CONFIG_FLATMEM
    BUG_ON(!mem_map);
#endif
    /*
     * With CONFIG_DEBUG_PAGEALLOC initialization of highmem pages has to
     * be done before free_all_bootmem(). Memblock use free low memory for
     * temporary data (see find_range_array()) and for this purpose can use
     * pages that was already passed to the buddy allocator, hence marked as
     * not accessible in the page tables when compiled with
     * CONFIG_DEBUG_PAGEALLOC. Otherwise order of initialization is not
     * important here.
     */
    set_highmem_pages_init();

    /* this will put all low memory onto the freelists */
    free_all_bootmem();

    after_bootmem = 1;

    mem_init_print_info(NULL);
    printk(KERN_INFO "virtual kernel memory layout:\n"
        "    fixmap  : 0x%08lx - 0x%08lx   (%4ld kB)\n"
#ifdef CONFIG_HIGHMEM
        "    pkmap   : 0x%08lx - 0x%08lx   (%4ld kB)\n"
#endif
        "    vmalloc : 0x%08lx - 0x%08lx   (%4ld MB)\n"
        "    lowmem  : 0x%08lx - 0x%08lx   (%4ld MB)\n"
        "      .init : 0x%08lx - 0x%08lx   (%4ld kB)\n"
        "      .data : 0x%08lx - 0x%08lx   (%4ld kB)\n"
        "      .text : 0x%08lx - 0x%08lx   (%4ld kB)\n",
        FIXADDR_START, FIXADDR_TOP,
        (FIXADDR_TOP - FIXADDR_START) >> 10,

#ifdef CONFIG_HIGHMEM
        PKMAP_BASE, PKMAP_BASE+LAST_PKMAP*PAGE_SIZE,
        (LAST_PKMAP*PAGE_SIZE) >> 10,
#endif

        VMALLOC_START, VMALLOC_END,
        (VMALLOC_END - VMALLOC_START) >> 20,

        (unsigned long)__va(0), (unsigned long)high_memory,
        ((unsigned long)high_memory - (unsigned long)__va(0)) >> 20,

        (unsigned long)&__init_begin, (unsigned long)&__init_end,
        ((unsigned long)&__init_end -
         (unsigned long)&__init_begin) >> 10,

        (unsigned long)&_etext, (unsigned long)&_edata,
        ((unsigned long)&_edata - (unsigned long)&_etext) >> 10,

        (unsigned long)&_text, (unsigned long)&_etext,
        ((unsigned long)&_etext - (unsigned long)&_text) >> 10);

    /*
     * Check boundaries twice: Some fundamental inconsistencies can
     * be detected at build time already.
     */
#define __FIXADDR_TOP (-PAGE_SIZE)
#ifdef CONFIG_HIGHMEM
    BUILD_BUG_ON(PKMAP_BASE + LAST_PKMAP*PAGE_SIZE	> FIXADDR_START);
    BUILD_BUG_ON(VMALLOC_END			> PKMAP_BASE);
#endif
#define high_memory (-128UL << 20)
    BUILD_BUG_ON(VMALLOC_START			>= VMALLOC_END);
#undef high_memory
#undef __FIXADDR_TOP
#ifdef CONFIG_RANDOMIZE_BASE
    BUILD_BUG_ON(CONFIG_RANDOMIZE_BASE_MAX_OFFSET > KERNEL_IMAGE_SIZE);
#endif

#ifdef CONFIG_HIGHMEM
    BUG_ON(PKMAP_BASE + LAST_PKMAP*PAGE_SIZE	> FIXADDR_START);
    BUG_ON(VMALLOC_END				> PKMAP_BASE);
#endif
    BUG_ON(VMALLOC_START				>= VMALLOC_END);
    BUG_ON((unsigned long)high_memory		> VMALLOC_START);

    if (boot_cpu_data.wp_works_ok < 0)
        test_wp_bit();
}

其中pci_iommu_alloc()不是伙伴算法重点相关的函数，不过还是稍微记录一下：

【file:/arch/x86/kernel/pci-dma.c】
void __init pci_iommu_alloc(void)
{
    struct iommu_table_entry *p;

    sort_iommu_table(__iommu_table, __iommu_table_end);
    check_iommu_entries(__iommu_table, __iommu_table_end);

    for (p = __iommu_table; p < __iommu_table_end; p++) {
        if (p && p->detect && p->detect() > 0) {
            p->flags |= IOMMU_DETECTED;
            if (p->early_init)
                p->early_init();
            if (p->flags & IOMMU_FINISH_IF_DETECTED)
                break;
        }
    }
}

该函数主要是将iommu table先行排序检查，然后调用各个表项注册的函数进行初始化。

而接着的set_highmem_pages_init()则是伙伴算法的开始：

【file:/arch/x86/mm/highmem_32.c】
void __init set_highmem_pages_init(void)
{
    struct zone *zone;
    int nid;

    /*
     * Explicitly reset zone->managed_pages because set_highmem_pages_init()
     * is invoked before free_all_bootmem()
     */
    reset_all_zones_managed_pages();
    for_each_zone(zone) {
        unsigned long zone_start_pfn, zone_end_pfn;

        if (!is_highmem(zone))
            continue;

        zone_start_pfn = zone->zone_start_pfn;
        zone_end_pfn = zone_start_pfn + zone->spanned_pages;

        nid = zone_to_nid(zone);
        printk(KERN_INFO "Initializing %s for node %d (%08lx:%08lx)\n",
                zone->name, nid, zone_start_pfn, zone_end_pfn);

        add_highpages_with_active_regions(nid, zone_start_pfn,
                 zone_end_pfn);
    }
}

该函数中reset_all_zones_managed_pages()主要是将所有的内存管理区zone的页面管理数据进行清0重置。而接下来的for_each_zone(zone)循环体结合is_highmem(zone)判断则是用于遍历查找出高端内存的管理区，对查找到高端内存调则用add_highpages_with_active_regions()将其释放添加至伙伴管理算法中。

add_highpages_with_active_regions()具体实现：

【file：/arch/x86/mm/init_32.c】
void __init add_highpages_with_active_regions(int nid,
             unsigned long start_pfn, unsigned long end_pfn)
{
    phys_addr_t start, end;
    u64 i;

    for_each_free_mem_range(i, nid, &start, &end, NULL) {
        unsigned long pfn = clamp_t(unsigned long, PFN_UP(start),
                        start_pfn, end_pfn);
        unsigned long e_pfn = clamp_t(unsigned long, PFN_DOWN(end),
                          start_pfn, end_pfn);
        for ( ; pfn < e_pfn; pfn++)
            if (pfn_valid(pfn))
                free_highmem_page(pfn_to_page(pfn));
    }
}

其中for_each_free_mem_range(i, nid, &start, &end, NULL)用于遍历查找memblock算法中空闲的空间区域，然后通过clamp_t()对空间区域进行去除内存空洞调整。里面的for ( ; pfn < e_pfn; pfn++)则用于将空间区域的各页面通过free_highmem_page()进行释放处理，其中if (pfn_valid(pfn))用于判断页面的有效性，而pfn_to_page(pfn)则是将页框号转换为页面管理结构。

进一步分析free_highmem_page()实现：

【file:/mm/page_alloc.c】
void free_highmem_page(struct page *page)
{
    __free_reserved_page(page);
    totalram_pages++;
    page_zone(page)->managed_pages++;
    totalhigh_pages++;
}

其中totalram_pages用于记录内存的总页面数，page_zone(page)->managed_pages则是记录管理区的管理页面数，totalhigh_pages则是记录高端内存的页面总数；

具体看一下__free_reserved_page()：

【file:/include/linux/mm.h】
/* Free the reserved page into the buddy system, so it gets managed. */
static inline void __free_reserved_page(struct page *page)
{
    ClearPageReserved(page);
    init_page_count(page);
    __free_page(page);
}

其中ClearPageReserved定义在/include/linux/page-flags.h中：

#define CLEARPAGEFLAG(uname, lname)                 \

static inline void ClearPage##uname(struct page *page)          \

{ clear_bit(PG_##lname, &page->flags); }

用于清除页面的flag中的reserved标志位，表示页面属于动态内存。

接着的init_page_count()这是设置页面的_count引用计数，设置为1，用于为__free_page()释放页面到内存管理算法中做准备。最后的__free_page()，该函数既是初始化伙伴管理算法，同时也是伙伴管理算法释放页面的操作函数。暂且搁置分析__free_page()的实现，后面再详细深入。

接着回到mem_init ()里面下一个调用free_all_bootmem()：

【file:/mm/nobootmem.c】
unsigned long __init free_all_bootmem(void)
{
    unsigned long pages;

    reset_all_zones_managed_pages();

    /*
     * We need to use NUMA_NO_NODE instead of NODE_DATA(0)->node_id
     *  because in some case like Node0 doesn't have RAM installed
     *  low ram will be on Node1
     */
    pages = free_low_memory_core_early();
    totalram_pages += pages;

    return pages;
}

其中reset_all_zones_managed_pages()是用于重置管理区zone结构中的managed_pages成员数据，着重分析一下free_low_memory_core_early()实现：

【file:/mm/nobootmem.c】
static unsigned long __init free_low_memory_core_early(void)
{
    unsigned long count = 0;
    phys_addr_t start, end;
    u64 i;

    for_each_free_mem_range(i, NUMA_NO_NODE, &start, &end, NULL)
        count += __free_memory_core(start, end);

#ifdef CONFIG_ARCH_DISCARD_MEMBLOCK
    {
        phys_addr_t size;

        /* Free memblock.reserved array if it was allocated */
        size = get_allocated_memblock_reserved_regions_info(&start);
        if (size)
            count += __free_memory_core(start, start + size);

        /* Free memblock.memory array if it was allocated */
        size = get_allocated_memblock_memory_regions_info(&start);
        if (size)
            count += __free_memory_core(start, start + size);
    }
#endif

    return count;
}

该函数通过for_each_free_mem_range()遍历memblock算法中的空闲内存空间，并调用__free_memory_core()来释放；而后面的get_allocated_memblock_reserved_regions_info()和get_allocated_memblock_memory_regions_info()用于获取“通过申请而得的memblock管理算法空间”信息，然后释放，其中如果其算法管理空间是系统定义的memblock_reserved_init_regions和memblock_memory_init_regions则仍保留不予以释放。

最后着重分析一下__free_memory_core()的实现：

【file:/mm/nobootmem.c】
static void __init __free_pages_memory(unsigned long start, unsigned long end)
{
    int order;

    while (start < end) {
        order = min(MAX_ORDER - 1UL, __ffs(start));

        while (start + (1UL << order) > end)
            order--;

        __free_pages_bootmem(pfn_to_page(start), order);

        start += (1UL << order);
    }
}

可以看到其调用通过里面的while循环调整查找到合适的阶去释放内存，而外层的while循环则是确保整个被释放的内存空间都将会被归还到伙伴管理算法中。而最后其里面的__free_pages_bootmem()实现则：

【file:/mm/nobootmem.c】
void __init __free_pages_bootmem(struct page *page, unsigned int order)
{
    unsigned int nr_pages = 1 << order;
    struct page *p = page;
    unsigned int loop;

    prefetchw(p);
    for (loop = 0; loop < (nr_pages - 1); loop++, p++) {
        prefetchw(p + 1);
        __ClearPageReserved(p);
        set_page_count(p, 0);
    }
    __ClearPageReserved(p);
    set_page_count(p, 0);

    page_zone(page)->managed_pages += nr_pages;
    set_page_refcounted(page);
    __free_pages(page, order);
}

由此可以看到，其最终调用的还是__free_pages()将页面予以释放。该函数在后面集中进行分析。

至此，伙伴管理算法初始化完毕。