p2m.c

来自「xen虚拟机源代码安装包」· C语言 代码 · 共 1,188 行 · 第 1/3 页

        }
        mfn = _mfn(l3e_get_pfn(*l3e));
        unmap_domain_page(l3e);
    }

    l2e = map_domain_page(mfn_x(mfn));
    l2e += l2_table_offset(addr);
    if ( (l2e_get_flags(*l2e) & _PAGE_PRESENT) == 0 )
    {
        unmap_domain_page(l2e);
        return _mfn(INVALID_MFN);
    }
    else if ( (l2e_get_flags(*l2e) & _PAGE_PSE) )
    {
        mfn = _mfn(l2e_get_pfn(*l2e) + l1_table_offset(addr));
        *t = p2m_flags_to_type(l2e_get_flags(*l2e));
        unmap_domain_page(l2e);
        
        ASSERT(mfn_valid(mfn) || !p2m_is_ram(*t));
        return (p2m_is_valid(*t)) ? mfn : _mfn(INVALID_MFN);
    }

    mfn = _mfn(l2e_get_pfn(*l2e));
    unmap_domain_page(l2e);

    l1e = map_domain_page(mfn_x(mfn));
    l1e += l1_table_offset(addr);
    if ( (l1e_get_flags(*l1e) & _PAGE_PRESENT) == 0 )
    {
        unmap_domain_page(l1e);
        return _mfn(INVALID_MFN);
    }
    mfn = _mfn(l1e_get_pfn(*l1e));
    *t = p2m_flags_to_type(l1e_get_flags(*l1e));
    unmap_domain_page(l1e);

    ASSERT(mfn_valid(mfn) || !p2m_is_ram(*t));
    return (p2m_is_valid(*t)) ? mfn : _mfn(INVALID_MFN);
}

/* Read the current domain's p2m table (through the linear mapping). */
static mfn_t p2m_gfn_to_mfn_current(unsigned long gfn, p2m_type_t *t)
{
    mfn_t mfn = _mfn(INVALID_MFN);
    p2m_type_t p2mt = p2m_mmio_dm;
    paddr_t addr = ((paddr_t)gfn) << PAGE_SHIFT;
    /* XXX This is for compatibility with the old model, where anything not 
     * XXX marked as RAM was considered to be emulated MMIO space.
     * XXX Once we start explicitly registering MMIO regions in the p2m 
     * XXX we will return p2m_invalid for unmapped gfns */

    if ( gfn <= current->domain->arch.p2m->max_mapped_pfn )
    {
        l1_pgentry_t l1e = l1e_empty();
        l2_pgentry_t l2e = l2e_empty();
        int ret;

        ASSERT(gfn < (RO_MPT_VIRT_END - RO_MPT_VIRT_START) 
               / sizeof(l1_pgentry_t));

        ret = __copy_from_user(&l2e,
                               &__linear_l1_table[l1_linear_offset(RO_MPT_VIRT_START) + l2_linear_offset(addr)],
                               sizeof(l2e));
        
        if ( (ret == 0) && (l2e_get_flags(l2e) & _PAGE_PRESENT) && 
             (l2e_get_flags(l2e) & _PAGE_PSE) ) 
        {
            p2mt = p2m_flags_to_type(l2e_get_flags(l2e));
            ASSERT(l2e_get_pfn(l2e) != INVALID_MFN || !p2m_is_ram(p2mt));
            if ( p2m_is_valid(p2mt) )
                mfn = _mfn(l2e_get_pfn(l2e) + l1_table_offset(addr));
            else
                p2mt = p2m_mmio_dm;
        }
        else
        {
        
            /* Need to __copy_from_user because the p2m is sparse and this
             * part might not exist */
            ret = __copy_from_user(&l1e,
                                   &phys_to_machine_mapping[gfn],
                                   sizeof(l1e));
            
            if ( ret == 0 ) {
                p2mt = p2m_flags_to_type(l1e_get_flags(l1e));
                ASSERT(l1e_get_pfn(l1e) != INVALID_MFN || !p2m_is_ram(p2mt));
                if ( p2m_is_valid(p2mt) )
                    mfn = _mfn(l1e_get_pfn(l1e));
                else 
                    /* XXX see above */
                    p2mt = p2m_mmio_dm;
            }
        }
    }

    *t = p2mt;
    return mfn;
}

/* Init the datastructures for later use by the p2m code */
int p2m_init(struct domain *d)
{
    struct p2m_domain *p2m;

    p2m = xmalloc(struct p2m_domain);
    if ( p2m == NULL )
        return -ENOMEM;

    d->arch.p2m = p2m;

    memset(p2m, 0, sizeof(*p2m));
    p2m_lock_init(p2m);
    INIT_LIST_HEAD(&p2m->pages);

    p2m->set_entry = p2m_set_entry;
    p2m->get_entry = p2m_gfn_to_mfn;
    p2m->get_entry_current = p2m_gfn_to_mfn_current;
    p2m->change_entry_type_global = p2m_change_type_global;

    if ( is_hvm_domain(d) && d->arch.hvm_domain.hap_enabled &&
         (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) )
        ept_p2m_init(d);

    return 0;
}

void p2m_change_entry_type_global(struct domain *d,
                                  p2m_type_t ot, p2m_type_t nt)
{
    struct p2m_domain *p2m = d->arch.p2m;

    p2m_lock(p2m);
    p2m->change_entry_type_global(d, ot, nt);
    p2m_unlock(p2m);
}

static
int set_p2m_entry(struct domain *d, unsigned long gfn, mfn_t mfn, 
                    unsigned int page_order, p2m_type_t p2mt)
{
    unsigned long todo = 1ul << page_order;
    unsigned int order;
    int rc = 0;

    while ( todo )
    {
        order = (((gfn | mfn_x(mfn) | todo) & ((1ul << 9) - 1)) == 0) ? 9 : 0;
        rc = d->arch.p2m->set_entry(d, gfn, mfn, order, p2mt);
        gfn += 1ul << order;
        if ( mfn_x(mfn) != INVALID_MFN )
            mfn = _mfn(mfn_x(mfn) + (1ul << order));
        todo -= 1ul << order;
    }

    return rc;
}

// Allocate a new p2m table for a domain.
//
// The structure of the p2m table is that of a pagetable for xen (i.e. it is
// controlled by CONFIG_PAGING_LEVELS).
//
// The alloc_page and free_page functions will be used to get memory to
// build the p2m, and to release it again at the end of day.
//
// Returns 0 for success or -errno.
//
int p2m_alloc_table(struct domain *d,
                    struct page_info * (*alloc_page)(struct domain *d),
                    void (*free_page)(struct domain *d, struct page_info *pg))

{
    mfn_t mfn = _mfn(INVALID_MFN);
    struct list_head *entry;
    struct page_info *page, *p2m_top;
    unsigned int page_count = 0;
    unsigned long gfn = -1UL;
    struct p2m_domain *p2m = d->arch.p2m;

    p2m_lock(p2m);

    if ( pagetable_get_pfn(d->arch.phys_table) != 0 )
    {
        P2M_ERROR("p2m already allocated for this domain\n");
        p2m_unlock(p2m);
        return -EINVAL;
    }

    P2M_PRINTK("allocating p2m table\n");

    p2m->alloc_page = alloc_page;
    p2m->free_page = free_page;

    p2m_top = p2m->alloc_page(d);
    if ( p2m_top == NULL )
    {
        p2m_unlock(p2m);
        return -ENOMEM;
    }
    list_add_tail(&p2m_top->list, &p2m->pages);

    p2m_top->count_info = 1;
    p2m_top->u.inuse.type_info =
#if CONFIG_PAGING_LEVELS == 4
        PGT_l4_page_table
#else
        PGT_l3_page_table
#endif
        | 1 | PGT_validated;

    d->arch.phys_table = pagetable_from_mfn(page_to_mfn(p2m_top));

    P2M_PRINTK("populating p2m table\n");

    /* Initialise physmap tables for slot zero. Other code assumes this. */
    if ( !set_p2m_entry(d, 0, _mfn(INVALID_MFN), 0,
                        p2m_invalid) )
        goto error;

    /* Copy all existing mappings from the page list and m2p */
    for ( entry = d->page_list.next;
          entry != &d->page_list;
          entry = entry->next )
    {
        page = list_entry(entry, struct page_info, list);
        mfn = page_to_mfn(page);
        gfn = get_gpfn_from_mfn(mfn_x(mfn));
        page_count++;
        if (
#ifdef __x86_64__
            (gfn != 0x5555555555555555L)
#else
            (gfn != 0x55555555L)
#endif
             && gfn != INVALID_M2P_ENTRY
            && !set_p2m_entry(d, gfn, mfn, 0, p2m_ram_rw) )
            goto error;
    }

    P2M_PRINTK("p2m table initialised (%u pages)\n", page_count);
    p2m_unlock(p2m);
    return 0;

 error:
    P2M_PRINTK("failed to initialize p2m table, gfn=%05lx, mfn=%"
               PRI_mfn "\n", gfn, mfn_x(mfn));
    p2m_unlock(p2m);
    return -ENOMEM;
}

void p2m_teardown(struct domain *d)
/* Return all the p2m pages to Xen.
 * We know we don't have any extra mappings to these pages */
{
    struct list_head *entry, *n;
    struct page_info *pg;
    struct p2m_domain *p2m = d->arch.p2m;

    p2m_lock(p2m);
    d->arch.phys_table = pagetable_null();

    list_for_each_safe(entry, n, &p2m->pages)
    {
        pg = list_entry(entry, struct page_info, list);
        list_del(entry);
        p2m->free_page(d, pg);
    }
    p2m_unlock(p2m);
}

void p2m_final_teardown(struct domain *d)
{
    xfree(d->arch.p2m);
    d->arch.p2m = NULL;
}

#if P2M_AUDIT
static void audit_p2m(struct domain *d)
{
    struct list_head *entry;
    struct page_info *page;
    struct domain *od;
    unsigned long mfn, gfn, m2pfn, lp2mfn = 0;
    mfn_t p2mfn;
    unsigned long orphans_d = 0, orphans_i = 0, mpbad = 0, pmbad = 0;
    int test_linear;
    p2m_type_t type;

    if ( !paging_mode_translate(d) )
        return;

    //P2M_PRINTK("p2m audit starts\n");

    test_linear = ( (d == current->domain)
                    && !pagetable_is_null(current->arch.monitor_table) );
    if ( test_linear )
        flush_tlb_local();

    /* Audit part one: walk the domain's page allocation list, checking
     * the m2p entries. */
    for ( entry = d->page_list.next;
          entry != &d->page_list;
          entry = entry->next )
    {
        page = list_entry(entry, struct page_info, list);
        mfn = mfn_x(page_to_mfn(page));

        // P2M_PRINTK("auditing guest page, mfn=%#lx\n", mfn);

        od = page_get_owner(page);

        if ( od != d )
        {
            P2M_PRINTK("wrong owner %#lx -> %p(%u) != %p(%u)\n",
                       mfn, od, (od?od->domain_id:-1), d, d->domain_id);
            continue;
        }

        gfn = get_gpfn_from_mfn(mfn);
        if ( gfn == INVALID_M2P_ENTRY )
        {
            orphans_i++;
            //P2M_PRINTK("orphaned guest page: mfn=%#lx has invalid gfn\n",
            //               mfn);
            continue;
        }

        if ( gfn == 0x55555555 )
        {
            orphans_d++;
            //P2M_PRINTK("orphaned guest page: mfn=%#lx has debug gfn\n",
            //               mfn);
            continue;
        }

        p2mfn = gfn_to_mfn_foreign(d, gfn, &type);
        if ( mfn_x(p2mfn) != mfn )
        {
            mpbad++;
            P2M_PRINTK("map mismatch mfn %#lx -> gfn %#lx -> mfn %#lx"
                       " (-> gfn %#lx)\n",
                       mfn, gfn, mfn_x(p2mfn),
                       (mfn_valid(p2mfn)
                        ? get_gpfn_from_mfn(mfn_x(p2mfn))
                        : -1u));
            /* This m2p entry is stale: the domain has another frame in
             * this physical slot.  No great disaster, but for neatness,
             * blow away the m2p entry. */
            set_gpfn_from_mfn(mfn, INVALID_M2P_ENTRY);
        }

        if ( test_linear && (gfn <= d->arch.p2m->max_mapped_pfn) )
        {
            lp2mfn = mfn_x(gfn_to_mfn_current(gfn, &type));
            if ( lp2mfn != mfn_x(p2mfn) )
            {
                P2M_PRINTK("linear mismatch gfn %#lx -> mfn %#lx "
                           "(!= mfn %#lx)\n", gfn, lp2mfn, mfn_x(p2mfn));
            }
        }

        // P2M_PRINTK("OK: mfn=%#lx, gfn=%#lx, p2mfn=%#lx, lp2mfn=%#lx\n",
        //                mfn, gfn, p2mfn, lp2mfn);
    }

    /* Audit part two: walk the domain's p2m table, checking the entries. */
    if ( pagetable_get_pfn(d->arch.phys_table) != 0 )
    {
        l2_pgentry_t *l2e;
        l1_pgentry_t *l1e;
        int i1, i2;

#if CONFIG_PAGING_LEVELS == 4
        l4_pgentry_t *l4e;
        l3_pgentry_t *l3e;
        int i3, i4;
        l4e = map_domain_page(mfn_x(pagetable_get_mfn(d->arch.phys_table)));
#else /* CONFIG_PAGING_LEVELS == 3 */
        l3_pgentry_t *l3e;
        int i3;
        l3e = map_domain_page(mfn_x(pagetable_get_mfn(d->arch.phys_table)));
#endif

        gfn = 0;
#if CONFIG_PAGING_LEVELS >= 4
        for ( i4 = 0; i4 < L4_PAGETABLE_ENTRIES; i4++ )
        {
            if ( !(l4e_get_flags(l4e[i4]) & _PAGE_PRESENT) )
            {
                gfn += 1 << (L4_PAGETABLE_SHIFT - PAGE_SHIFT);
                continue;
            }
            l3e = map_domain_page(mfn_x(_mfn(l4e_get_pfn(l4e[i4]))));
#endif
            for ( i3 = 0;
                  i3 < ((CONFIG_PAGING_LEVELS==4) ? L3_PAGETABLE_ENTRIES : 8);