setup.c

来自「xen虚拟机源代码安装包」· C语言 代码 · 共 1,123 行 · 第 1/3 页

}

void init_done(void)
{
    extern char __init_begin[], __init_end[];

    /* Free (or page-protect) the init areas. */
#ifndef MEMORY_GUARD
    init_xenheap_pages(__pa(__init_begin), __pa(__init_end));
#endif
    memguard_guard_range(__init_begin, __init_end - __init_begin);
#if defined(CONFIG_X86_64)
    /* Also zap the mapping in the 1:1 area. */
    memguard_guard_range(__va(__pa(__init_begin)), __init_end - __init_begin);
#endif
    printk("Freed %ldkB init memory.\n", (long)(__init_end-__init_begin)>>10);

    startup_cpu_idle_loop();
}

static char * __init cmdline_cook(char *p)
{
    p = p ? : "";
    while ( *p == ' ' )
        p++;
    while ( (*p != ' ') && (*p != '\0') )
        p++;
    while ( *p == ' ' )
        p++;
    return p;
}

void __init __start_xen(unsigned long mbi_p)
{
    char *memmap_type = NULL;
    char *cmdline, *kextra;
    unsigned long _initrd_start = 0, _initrd_len = 0;
    unsigned int initrdidx = 1;
    multiboot_info_t *mbi = __va(mbi_p);
    module_t *mod = (module_t *)__va(mbi->mods_addr);
    unsigned long nr_pages, modules_length;
    int i, e820_warn = 0, bytes = 0;
    struct ns16550_defaults ns16550 = {
        .data_bits = 8,
        .parity    = 'n',
        .stop_bits = 1
    };

    extern void early_page_fault(void);
    set_intr_gate(TRAP_page_fault, &early_page_fault);

    /* Parse the command-line options. */
    cmdline = cmdline_cook((mbi->flags & MBI_CMDLINE) ?
                           __va(mbi->cmdline) : NULL);
    if ( (kextra = strstr(cmdline, " -- ")) != NULL )
    {
        /*
         * Options after ' -- ' separator belong to dom0.
         *  1. Orphan dom0's options from Xen's command line.
         *  2. Skip all but final leading space from dom0's options.
         */
        *kextra = '\0';
        kextra += 3;
        while ( kextra[1] == ' ' ) kextra++;
    }
    cmdline_parse(cmdline);

    parse_video_info();

    set_current((struct vcpu *)0xfffff000); /* debug sanity */
    idle_vcpu[0] = current;
    set_processor_id(0); /* needed early, for smp_processor_id() */
    if ( cpu_has_efer )
        rdmsrl(MSR_EFER, this_cpu(efer));
    asm volatile ( "mov %%cr4,%0" : "=r" (this_cpu(cr4)) );

    smp_prepare_boot_cpu();

    /* We initialise the serial devices very early so we can get debugging. */
    ns16550.io_base = 0x3f8;
    ns16550.irq     = 4;
    ns16550_init(0, &ns16550);
    ns16550.io_base = 0x2f8;
    ns16550.irq     = 3;
    ns16550_init(1, &ns16550);
    serial_init_preirq();

    init_console();

    printk("Command line: %s\n", cmdline);

    printk("Video information:\n");

    /* Print VGA display mode information. */
    switch ( vga_console_info.video_type )
    {
    case XEN_VGATYPE_TEXT_MODE_3:
        printk(" VGA is text mode %dx%d, font 8x%d\n",
               vga_console_info.u.text_mode_3.columns,
               vga_console_info.u.text_mode_3.rows,
               vga_console_info.u.text_mode_3.font_height);
        break;
    case XEN_VGATYPE_VESA_LFB:
        printk(" VGA is graphics mode %dx%d, %d bpp\n",
               vga_console_info.u.vesa_lfb.width,
               vga_console_info.u.vesa_lfb.height,
               vga_console_info.u.vesa_lfb.bits_per_pixel);
        break;
    default:
        printk(" No VGA detected\n");
        break;
    }

    /* Print VBE/DDC EDID information. */
    if ( bootsym(boot_edid_caps) != 0x1313 )
    {
        u16 caps = bootsym(boot_edid_caps);
        printk(" VBE/DDC methods:%s%s%s; ",
               (caps & 1) ? " V1" : "",
               (caps & 2) ? " V2" : "",
               !(caps & 3) ? " none" : "");
        printk("EDID transfer time: %d seconds\n", caps >> 8);
        if ( *(u32 *)bootsym(boot_edid_info) == 0x13131313 )
        {
            printk(" EDID info not retrieved because ");
            if ( !(caps & 3) )
                printk("no DDC retrieval method detected\n");
            else if ( (caps >> 8) > 5 )
                printk("takes longer than 5 seconds\n");
            else
                printk("of reasons unknown\n");
        }
    }

    printk("Disc information:\n");
    printk(" Found %d MBR signatures\n",
           bootsym(boot_mbr_signature_nr));
    printk(" Found %d EDD information structures\n",
           bootsym(boot_edd_info_nr));

    /* Check that we have at least one Multiboot module. */
    if ( !(mbi->flags & MBI_MODULES) || (mbi->mods_count == 0) )
        EARLY_FAIL("dom0 kernel not specified. "
                   "Check bootloader configuration.\n");

    if ( ((unsigned long)cpu0_stack & (STACK_SIZE-1)) != 0 )
        EARLY_FAIL("Misaligned CPU0 stack.\n");

    if ( e820_raw_nr != 0 )
    {
        memmap_type = "Xen-e820";
    }
    else if ( bootsym(lowmem_kb) )
    {
        memmap_type = "Xen-e801";
        e820_raw[0].addr = 0;
        e820_raw[0].size = bootsym(lowmem_kb) << 10;
        e820_raw[0].type = E820_RAM;
        e820_raw[1].addr = 0x100000;
        e820_raw[1].size = bootsym(highmem_kb) << 10;
        e820_raw[1].type = E820_RAM;
        e820_raw_nr = 2;
    }
    else if ( mbi->flags & MBI_MEMMAP )
    {
        memmap_type = "Multiboot-e820";
        while ( (bytes < mbi->mmap_length) && (e820_raw_nr < E820MAX) )
        {
            memory_map_t *map = __va(mbi->mmap_addr + bytes);

            /*
             * This is a gross workaround for a BIOS bug. Some bootloaders do
             * not write e820 map entries into pre-zeroed memory. This is
             * okay if the BIOS fills in all fields of the map entry, but
             * some broken BIOSes do not bother to write the high word of
             * the length field if the length is smaller than 4GB. We
             * detect and fix this by flagging sections below 4GB that
             * appear to be larger than 4GB in size.
             */
            if ( (map->base_addr_high == 0) && (map->length_high != 0) )
            {
                if ( !e820_warn )
                {
                    printk("WARNING: Buggy e820 map detected and fixed "
                           "(truncated length fields).\n");
                    e820_warn = 1;
                }
                map->length_high = 0;
            }

            e820_raw[e820_raw_nr].addr = 
                ((u64)map->base_addr_high << 32) | (u64)map->base_addr_low;
            e820_raw[e820_raw_nr].size = 
                ((u64)map->length_high << 32) | (u64)map->length_low;
            e820_raw[e820_raw_nr].type = map->type;
            e820_raw_nr++;

            bytes += map->size + 4;
        }
    }
    else if ( mbi->flags & MBI_MEMLIMITS )
    {
        memmap_type = "Multiboot-e801";
        e820_raw[0].addr = 0;
        e820_raw[0].size = mbi->mem_lower << 10;
        e820_raw[0].type = E820_RAM;
        e820_raw[1].addr = 0x100000;
        e820_raw[1].size = mbi->mem_upper << 10;
        e820_raw[1].type = E820_RAM;
        e820_raw_nr = 2;
    }
    else
    {
        EARLY_FAIL("Bootloader provided no memory information.\n");
    }

    /* Sanitise the raw E820 map to produce a final clean version. */
    max_page = init_e820(memmap_type, e820_raw, &e820_raw_nr);

#ifdef CONFIG_X86_64
    /*
     * On x86/64 we are able to account for the allocation bitmap
     * (allocated in common/page_alloc.c:init_boot_allocator()) stealing
     * from the Xen heap. Here we make the Xen heap appropriately larger.
     */
    opt_xenheap_megabytes += (max_page / 8) >> 20;
#endif

    /*
     * Since there are some stubs getting built on the stacks which use
     * direct calls/jumps, the heap must be confined to the lower 2G so
     * that those branches can reach their targets.
     */
    if ( opt_xenheap_megabytes > 2048 )
        opt_xenheap_megabytes = 2048;

    /* Create a temporary copy of the E820 map. */
    memcpy(&boot_e820, &e820, sizeof(e820));

    /* Early kexec reservation (explicit static start address). */
    kexec_reserve_area(&boot_e820);

    /*
     * Iterate backwards over all superpage-aligned RAM regions.
     * 
     * We require superpage alignment because the boot allocator is not yet
     * initialised. Hence we can only map superpages in the address range
     * 0 to BOOTSTRAP_DIRECTMAP_END, as this is guaranteed not to require
     * dynamic allocation of pagetables.
     * 
     * As well as mapping superpages in that range, in preparation for
     * initialising the boot allocator, we also look for a region to which
     * we can relocate the dom0 kernel and other multiboot modules. Also, on
     * x86/64, we relocate Xen to higher memory.
     */
    modules_length = mod[mbi->mods_count-1].mod_end - mod[0].mod_start;
    for ( i = boot_e820.nr_map-1; i >= 0; i-- )
    {
        uint64_t s, e, mask = (1UL << L2_PAGETABLE_SHIFT) - 1;

        /* Superpage-aligned chunks from 16MB to BOOTSTRAP_DIRECTMAP_END. */
        s = (boot_e820.map[i].addr + mask) & ~mask;
        e = (boot_e820.map[i].addr + boot_e820.map[i].size) & ~mask;
        s = max_t(uint64_t, s, 16 << 20);
        e = min_t(uint64_t, e, BOOTSTRAP_DIRECTMAP_END);
        if ( (boot_e820.map[i].type != E820_RAM) || (s >= e) )
            continue;

        /* Map the chunk. No memory will need to be allocated to do this. */
        map_pages_to_xen(
            (unsigned long)maddr_to_bootstrap_virt(s),
            s >> PAGE_SHIFT, (e-s) >> PAGE_SHIFT, PAGE_HYPERVISOR);

#if defined(CONFIG_X86_64)
        /* Is the region suitable for relocating Xen? */
        if ( !xen_phys_start && (((e-s) >> 20) >= opt_xenheap_megabytes) )
        {
            extern l2_pgentry_t l2_xenmap[];
            l4_pgentry_t *pl4e;
            l3_pgentry_t *pl3e;
            l2_pgentry_t *pl2e;
            int i, j, k;

            /* Select relocation address. */
            e = (e - (opt_xenheap_megabytes << 20)) & ~mask;
            xen_phys_start = e;
            bootsym(trampoline_xen_phys_start) = e;

            /*
             * Perform relocation to new physical address.
             * Before doing so we must sync static/global data with main memory
             * with a barrier(). After this we must *not* modify static/global
             * data until after we have switched to the relocated pagetables!
             */
            barrier();
            move_memory(e, 0, __pa(&_end) - xen_phys_start);

            /* Poison low 1MB to detect stray pointers to physical 0-1MB. */
            memset(maddr_to_bootstrap_virt(e), 0x55, 1U<<20);

            /* Walk initial pagetables, relocating page directory entries. */
            pl4e = __va(__pa(idle_pg_table));
            for ( i = 0 ; i < L4_PAGETABLE_ENTRIES; i++, pl4e++ )
            {
                if ( !(l4e_get_flags(*pl4e) & _PAGE_PRESENT) )
                    continue;
                *pl4e = l4e_from_intpte(l4e_get_intpte(*pl4e) +
                                        xen_phys_start);
                pl3e = l4e_to_l3e(*pl4e);
                for ( j = 0; j < L3_PAGETABLE_ENTRIES; j++, pl3e++ )
                {
                    /* Not present, 1GB mapping, or already relocated? */
                    if ( !(l3e_get_flags(*pl3e) & _PAGE_PRESENT) ||
                         (l3e_get_flags(*pl3e) & _PAGE_PSE) ||
                         (l3e_get_pfn(*pl3e) > 0x1000) )
                        continue;
                    *pl3e = l3e_from_intpte(l3e_get_intpte(*pl3e) +
                                            xen_phys_start);
                    pl2e = l3e_to_l2e(*pl3e);
                    for ( k = 0; k < L2_PAGETABLE_ENTRIES; k++, pl2e++ )
                    {
                        /* Not present, PSE, or already relocated? */
                        if ( !(l2e_get_flags(*pl2e) & _PAGE_PRESENT) ||
                             (l2e_get_flags(*pl2e) & _PAGE_PSE) ||
                             (l2e_get_pfn(*pl2e) > 0x1000) )
                            continue;
                        *pl2e = l2e_from_intpte(l2e_get_intpte(*pl2e) +
                                                xen_phys_start);
                    }
                }
            }

            /* The only data mappings to be relocated are in the Xen area. */
            pl2e = __va(__pa(l2_xenmap));
            *pl2e++ = l2e_from_pfn(xen_phys_start >> PAGE_SHIFT,
                                   PAGE_HYPERVISOR | _PAGE_PSE);
            for ( i = 1; i < L2_PAGETABLE_ENTRIES; i++, pl2e++ )
            {
                if ( !(l2e_get_flags(*pl2e) & _PAGE_PRESENT) )
                    continue;
                *pl2e = l2e_from_intpte(l2e_get_intpte(*pl2e) +
                                        xen_phys_start);
            }

            /* Re-sync the stack and then switch to relocated pagetables. */
            asm volatile (
                "rep movsb        ; " /* re-sync the stack */
                "movq %%cr4,%%rsi ; "
                "andb $0x7f,%%sil ; "
                "movq %%rsi,%%cr4 ; " /* CR4.PGE == 0 */
                "movq %0,%%cr3    ; " /* CR3 == new pagetables */
                "orb $0x80,%%sil  ; "
                "movq %%rsi,%%cr4   " /* CR4.PGE == 1 */
                : : "r" (__pa(idle_pg_table)), "S" (cpu0_stack),
                "D" (__va(__pa(cpu0_stack))), "c" (STACK_SIZE) : "memory" );
        }
#endif

        /* Is the region suitable for relocating the multiboot modules? */
        if ( !initial_images_start && (s < e) && ((e-s) >= modules_length) )
        {
            initial_images_end = e;
            e = (e - modules_length) & PAGE_MASK;
            initial_images_start = e;
            move_memory(initial_images_start, 
                        mod[0].mod_start, mod[mbi->mods_count-1].mod_end);
        }

        if ( !kexec_crash_area.start && (s < e) &&
             ((e-s) >= kexec_crash_area.size) )
        {
            e = (e - kexec_crash_area.size) & PAGE_MASK;
            kexec_crash_area.start = e;
        }
    }