[wimlib] / src / lookup_table.c

/*
 * lookup_table.c
 *
 * Lookup table, implemented as a hash table, that maps SHA1 message digests to
 * data streams; plus code to read and write the corresponding on-disk data.
 */

/*
 * Copyright (C) 2012, 2013, 2014 Eric Biggers
 *
 * This file is free software; you can redistribute it and/or modify it under
 * the terms of the GNU Lesser General Public License as published by the Free
 * Software Foundation; either version 3 of the License, or (at your option) any
 * later version.
 *
 * This file is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
 * FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
 * details.
 *
 * You should have received a copy of the GNU Lesser General Public License
 * along with this file; if not, see http://www.gnu.org/licenses/.
 */

#ifdef HAVE_CONFIG_H
#  include "config.h"
#endif

#include "wimlib/assert.h"
#include "wimlib/endianness.h"
#include "wimlib/error.h"
#include "wimlib/lookup_table.h"
#include "wimlib/metadata.h"
#include "wimlib/ntfs_3g.h"
#include "wimlib/resource.h"
#include "wimlib/util.h"
#include "wimlib/write.h"

#include <stdlib.h>
#include <string.h>
#include <unistd.h> /* for unlink()  */

/* WIM lookup table:
 *
 * This is a logical mapping from SHA1 message digests to the data streams
 * contained in a WIM.
 *
 * Here it is implemented as a hash table.
 *
 * Note: Everything will break horribly if there is a SHA1 collision.
 */
struct wim_lookup_table {
        struct hlist_head *array;
        size_t num_entries;
        size_t capacity;
};

struct wim_lookup_table *
new_lookup_table(size_t capacity)
{
        struct wim_lookup_table *table;
        struct hlist_head *array;

        table = MALLOC(sizeof(struct wim_lookup_table));
        if (table == NULL)
                goto oom;

        array = CALLOC(capacity, sizeof(array[0]));
        if (array == NULL) {
                FREE(table);
                goto oom;
        }

        table->num_entries = 0;
        table->capacity = capacity;
        table->array = array;
        return table;

oom:
        ERROR("Failed to allocate memory for lookup table "
              "with capacity %zu", capacity);
        return NULL;
}

static int
do_free_lookup_table_entry(struct wim_lookup_table_entry *entry, void *ignore)
{
        free_lookup_table_entry(entry);
        return 0;
}

void
free_lookup_table(struct wim_lookup_table *table)
{
        if (table) {
                for_lookup_table_entry(table, do_free_lookup_table_entry, NULL);
                FREE(table->array);
                FREE(table);
        }
}

struct wim_lookup_table_entry *
new_lookup_table_entry(void)
{
        struct wim_lookup_table_entry *lte;

        lte = CALLOC(1, sizeof(struct wim_lookup_table_entry));
        if (lte == NULL)
                return NULL;

        lte->refcnt = 1;

        /* lte->resource_location = RESOURCE_NONEXISTENT  */
        BUILD_BUG_ON(RESOURCE_NONEXISTENT != 0);

        return lte;
}

struct wim_lookup_table_entry *
clone_lookup_table_entry(const struct wim_lookup_table_entry *old)
{
        struct wim_lookup_table_entry *new;

        new = memdup(old, sizeof(struct wim_lookup_table_entry));
        if (new == NULL)
                return NULL;

        switch (new->resource_location) {
        case RESOURCE_IN_WIM:
                list_add(&new->rspec_node, &new->rspec->stream_list);
                break;

        case RESOURCE_IN_FILE_ON_DISK:
#ifdef __WIN32__
        case RESOURCE_IN_WINNT_FILE_ON_DISK:
        case RESOURCE_WIN32_ENCRYPTED:
#endif
#ifdef WITH_FUSE
        case RESOURCE_IN_STAGING_FILE:
                BUILD_BUG_ON((void*)&old->file_on_disk !=
                             (void*)&old->staging_file_name);
#endif
                new->file_on_disk = TSTRDUP(old->file_on_disk);
                if (new->file_on_disk == NULL)
                        goto out_free;
                break;
        case RESOURCE_IN_ATTACHED_BUFFER:
                new->attached_buffer = memdup(old->attached_buffer, old->size);
                if (new->attached_buffer == NULL)
                        goto out_free;
                break;
#ifdef WITH_NTFS_3G
        case RESOURCE_IN_NTFS_VOLUME:
                if (old->ntfs_loc) {
                        struct ntfs_location *loc;
                        loc = memdup(old->ntfs_loc, sizeof(struct ntfs_location));
                        if (loc == NULL)
                                goto out_free;
                        loc->path = NULL;
                        loc->stream_name = NULL;
                        new->ntfs_loc = loc;
                        loc->path = STRDUP(old->ntfs_loc->path);
                        if (loc->path == NULL)
                                goto out_free;
                        if (loc->stream_name_nchars != 0) {
                                loc->stream_name = memdup(old->ntfs_loc->stream_name,
                                                          loc->stream_name_nchars * 2);
                                if (loc->stream_name == NULL)
                                        goto out_free;
                        }
                }
                break;
#endif
        default:
                break;
        }
        return new;

out_free:
        free_lookup_table_entry(new);
        return NULL;
}

void
lte_put_resource(struct wim_lookup_table_entry *lte)
{
        switch (lte->resource_location) {
        case RESOURCE_IN_WIM:
                list_del(&lte->rspec_node);
                if (list_empty(&lte->rspec->stream_list))
                        FREE(lte->rspec);
                break;
        case RESOURCE_IN_FILE_ON_DISK:
#ifdef __WIN32__
        case RESOURCE_IN_WINNT_FILE_ON_DISK:
        case RESOURCE_WIN32_ENCRYPTED:
#endif
#ifdef WITH_FUSE
        case RESOURCE_IN_STAGING_FILE:
                BUILD_BUG_ON((void*)&lte->file_on_disk !=
                             (void*)&lte->staging_file_name);
#endif
        case RESOURCE_IN_ATTACHED_BUFFER:
                BUILD_BUG_ON((void*)&lte->file_on_disk !=
                             (void*)&lte->attached_buffer);
                FREE(lte->file_on_disk);
                break;
#ifdef WITH_NTFS_3G
        case RESOURCE_IN_NTFS_VOLUME:
                if (lte->ntfs_loc) {
                        FREE(lte->ntfs_loc->path);
                        FREE(lte->ntfs_loc->stream_name);
                        FREE(lte->ntfs_loc);
                }
                break;
#endif
        default:
                break;
        }
}

void
free_lookup_table_entry(struct wim_lookup_table_entry *lte)
{
        if (lte) {
                lte_put_resource(lte);
                FREE(lte);
        }
}

/* Should this stream be retained even if it has no references?  */
static bool
should_retain_lte(const struct wim_lookup_table_entry *lte)
{
        return lte->resource_location == RESOURCE_IN_WIM;
}

static void
finalize_lte(struct wim_lookup_table_entry *lte)
{
        if (!should_retain_lte(lte))
                free_lookup_table_entry(lte);
}

/*
 * Decrements the reference count of the single-instance stream @lte, which must
 * be inserted in the stream lookup table @table.
 *
 * If the stream's reference count reaches 0, we may unlink it from @table and
 * free it.  However, we retain streams with 0 reference count that originated
 * from WIM files (RESOURCE_IN_WIM).  We do this for two reasons:
 *
 * 1. This prevents information about valid streams in a WIM file --- streams
 *    which will continue to be present after appending to the WIM file --- from
 *    being lost merely because we dropped all references to them.
 *
 * 2. Stream reference counts we read from WIM files can't be trusted.  It's
 *    possible that a WIM has reference counts that are too low; WIMGAPI
 *    sometimes creates WIMs where this is the case.  It's also possible that
 *    streams have been referenced from an external WIM; those streams can
 *    potentially have any reference count at all, either lower or higher than
 *    would be expected for this WIM ("this WIM" meaning the owner of @table) if
 *    it were a standalone WIM.
 *
 * So we can't take the reference counts too seriously.  But at least, we do
 * recalculate by default when writing a new WIM file.
 */
void
lte_decrement_refcnt(struct wim_lookup_table_entry *lte,
                     struct wim_lookup_table *table)
{
        if (unlikely(lte->refcnt == 0))  /* See comment above  */
                return;

        if (--lte->refcnt == 0) {
                if (lte->unhashed) {
                        list_del(&lte->unhashed_list);
                #ifdef WITH_FUSE
                        /* If the stream has been extracted to a staging file
                         * for a FUSE mount, unlink the staging file.  (Note
                         * that there still may be open file descriptors to it.)
                         * */
                        if (lte->resource_location == RESOURCE_IN_STAGING_FILE)
                                unlinkat(lte->staging_dir_fd,
                                         lte->staging_file_name, 0);
                #endif
                } else {
                        if (!should_retain_lte(lte))
                                lookup_table_unlink(table, lte);
                }

                /* If FUSE mounts are enabled, we don't actually free the entry
                 * until the last file descriptor has been closed by
                 * lte_decrement_num_opened_fds().  */
#ifdef WITH_FUSE
                if (lte->num_opened_fds == 0)
#endif
                        finalize_lte(lte);
        }
}

#ifdef WITH_FUSE
void
lte_decrement_num_opened_fds(struct wim_lookup_table_entry *lte)
{
        wimlib_assert(lte->num_opened_fds != 0);

        if (--lte->num_opened_fds == 0 && lte->refcnt == 0)
                finalize_lte(lte);
}
#endif

static void
lookup_table_insert_raw(struct wim_lookup_table *table,
                        struct wim_lookup_table_entry *lte)
{
        size_t i = lte->hash_short % table->capacity;

        hlist_add_head(&lte->hash_list, &table->array[i]);
}

static void
enlarge_lookup_table(struct wim_lookup_table *table)
{
        size_t old_capacity, new_capacity;
        struct hlist_head *old_array, *new_array;
        struct wim_lookup_table_entry *lte;
        struct hlist_node *cur, *tmp;
        size_t i;

        old_capacity = table->capacity;
        new_capacity = old_capacity * 2;
        new_array = CALLOC(new_capacity, sizeof(struct hlist_head));
        if (new_array == NULL)
                return;
        old_array = table->array;
        table->array = new_array;
        table->capacity = new_capacity;

        for (i = 0; i < old_capacity; i++) {
                hlist_for_each_entry_safe(lte, cur, tmp, &old_array[i], hash_list) {
                        hlist_del(&lte->hash_list);
                        lookup_table_insert_raw(table, lte);
                }
        }
        FREE(old_array);
}

/* Inserts an entry into the lookup table.  */
void
lookup_table_insert(struct wim_lookup_table *table,
                    struct wim_lookup_table_entry *lte)
{
        lookup_table_insert_raw(table, lte);
        if (++table->num_entries > table->capacity)
                enlarge_lookup_table(table);
}

/* Unlinks a lookup table entry from the table; does not free it.  */
void
lookup_table_unlink(struct wim_lookup_table *table,
                    struct wim_lookup_table_entry *lte)
{
        wimlib_assert(!lte->unhashed);
        wimlib_assert(table->num_entries != 0);

        hlist_del(&lte->hash_list);
        table->num_entries--;
}

/* Given a SHA1 message digest, return the corresponding entry in the WIM's
 * lookup table, or NULL if there is none.  */
struct wim_lookup_table_entry *
lookup_stream(const struct wim_lookup_table *table, const u8 hash[])
{
        size_t i;
        struct wim_lookup_table_entry *lte;
        struct hlist_node *pos;

        i = *(size_t*)hash % table->capacity;
        hlist_for_each_entry(lte, pos, &table->array[i], hash_list)
                if (hashes_equal(hash, lte->hash))
                        return lte;
        return NULL;
}

/* Calls a function on all the entries in the WIM lookup table.  Stop early and
 * return nonzero if any call to the function returns nonzero. */
int
for_lookup_table_entry(struct wim_lookup_table *table,
                       int (*visitor)(struct wim_lookup_table_entry *, void *),
                       void *arg)
{
        struct wim_lookup_table_entry *lte;
        struct hlist_node *pos, *tmp;
        int ret;

        for (size_t i = 0; i < table->capacity; i++) {
                hlist_for_each_entry_safe(lte, pos, tmp, &table->array[i],
                                          hash_list)
                {
                        ret = visitor(lte, arg);
                        if (ret)
                                return ret;
                }
        }
        return 0;
}

/* qsort() callback that sorts streams (represented by `struct
 * wim_lookup_table_entry's) into an order optimized for reading.
 *
 * Sorting is done primarily by resource location, then secondarily by a
 * per-resource location order.  For example, resources in WIM files are sorted
 * primarily by part number, then secondarily by offset, as to implement optimal
 * reading of either a standalone or split WIM.  */
static int
cmp_streams_by_sequential_order(const void *p1, const void *p2)
{
        const struct wim_lookup_table_entry *lte1, *lte2;
        int v;
        WIMStruct *wim1, *wim2;

        lte1 = *(const struct wim_lookup_table_entry**)p1;
        lte2 = *(const struct wim_lookup_table_entry**)p2;

        v = (int)lte1->resource_location - (int)lte2->resource_location;

        /* Different resource locations?  */
        if (v)
                return v;

        switch (lte1->resource_location) {
        case RESOURCE_IN_WIM:
                wim1 = lte1->rspec->wim;
                wim2 = lte2->rspec->wim;

                /* Different (possibly split) WIMs?  */
                if (wim1 != wim2) {
                        v = memcmp(wim1->hdr.guid, wim2->hdr.guid, WIM_GUID_LEN);
                        if (v)
                                return v;
                }

                /* Different part numbers in the same WIM?  */
                v = (int)wim1->hdr.part_number - (int)wim2->hdr.part_number;
                if (v)
                        return v;

                if (lte1->rspec->offset_in_wim != lte2->rspec->offset_in_wim)
                        return cmp_u64(lte1->rspec->offset_in_wim,
                                       lte2->rspec->offset_in_wim);

                return cmp_u64(lte1->offset_in_res, lte2->offset_in_res);

        case RESOURCE_IN_FILE_ON_DISK:
#ifdef WITH_FUSE
        case RESOURCE_IN_STAGING_FILE:
#endif
#ifdef __WIN32__
        case RESOURCE_IN_WINNT_FILE_ON_DISK:
        case RESOURCE_WIN32_ENCRYPTED:
#endif
                /* Compare files by path: just a heuristic that will place files
                 * in the same directory next to each other.  */
                return tstrcmp(lte1->file_on_disk, lte2->file_on_disk);
#ifdef WITH_NTFS_3G
        case RESOURCE_IN_NTFS_VOLUME:
                return tstrcmp(lte1->ntfs_loc->path, lte2->ntfs_loc->path);
#endif
        default:
                /* No additional sorting order defined for this resource
                 * location (e.g. RESOURCE_IN_ATTACHED_BUFFER); simply compare
                 * everything equal to each other.  */
                return 0;
        }
}

int
sort_stream_list(struct list_head *stream_list,
                 size_t list_head_offset,
                 int (*compar)(const void *, const void*))
{
        struct list_head *cur;
        struct wim_lookup_table_entry **array;
        size_t i;
        size_t array_size;
        size_t num_streams = 0;

        list_for_each(cur, stream_list)
                num_streams++;

        if (num_streams <= 1)
                return 0;

        array_size = num_streams * sizeof(array[0]);
        array = MALLOC(array_size);
        if (array == NULL)
                return WIMLIB_ERR_NOMEM;

        cur = stream_list->next;
        for (i = 0; i < num_streams; i++) {
                array[i] = (struct wim_lookup_table_entry*)((u8*)cur -
                                                            list_head_offset);
                cur = cur->next;
        }

        qsort(array, num_streams, sizeof(array[0]), compar);

        INIT_LIST_HEAD(stream_list);
        for (i = 0; i < num_streams; i++) {
                list_add_tail((struct list_head*)
                               ((u8*)array[i] + list_head_offset),
                              stream_list);
        }
        FREE(array);
        return 0;
}

/* Sort the specified list of streams in an order optimized for reading.  */
int
sort_stream_list_by_sequential_order(struct list_head *stream_list,
                                     size_t list_head_offset)
{
        return sort_stream_list(stream_list, list_head_offset,
                                cmp_streams_by_sequential_order);
}


static int
add_lte_to_array(struct wim_lookup_table_entry *lte,
                 void *_pp)
{
        struct wim_lookup_table_entry ***pp = _pp;
        *(*pp)++ = lte;
        return 0;
}

/* Iterate through the lookup table entries, but first sort them by stream
 * offset in the WIM.  Caution: this is intended to be used when the stream
 * offset field has actually been set. */
int
for_lookup_table_entry_pos_sorted(struct wim_lookup_table *table,
                                  int (*visitor)(struct wim_lookup_table_entry *,
                                                 void *),
                                  void *arg)
{
        struct wim_lookup_table_entry **lte_array, **p;
        size_t num_streams = table->num_entries;
        int ret;

        lte_array = MALLOC(num_streams * sizeof(lte_array[0]));
        if (!lte_array)
                return WIMLIB_ERR_NOMEM;
        p = lte_array;
        for_lookup_table_entry(table, add_lte_to_array, &p);

        wimlib_assert(p == lte_array + num_streams);

        qsort(lte_array, num_streams, sizeof(lte_array[0]),
              cmp_streams_by_sequential_order);
        ret = 0;
        for (size_t i = 0; i < num_streams; i++) {
                ret = visitor(lte_array[i], arg);
                if (ret)
                        break;
        }
        FREE(lte_array);
        return ret;
}

/* On-disk format of a WIM lookup table entry (stream entry). */
struct wim_lookup_table_entry_disk {
        /* Size, offset, and flags of the stream.  */
        struct wim_reshdr_disk reshdr;

        /* Which part of the split WIM this stream is in; indexed from 1. */
        le16 part_number;

        /* Reference count of this stream over all WIM images.  (But see comment
         * above lte_decrement_refcnt().)  */
        le32 refcnt;

        /* SHA1 message digest of the uncompressed data of this stream, or
         * optionally all zeroes if this stream is of zero length. */
        u8 hash[SHA1_HASH_SIZE];
} _packed_attribute;

#define WIM_LOOKUP_TABLE_ENTRY_DISK_SIZE 50

/* Given a nonempty run of consecutive lookup table entries with the
 * PACKED_STREAMS flag set, count how many specify resources (as opposed to
 * streams within those resources).
 *
 * Returns the resulting count.  */
static size_t
count_subpacks(const struct wim_lookup_table_entry_disk *entries, size_t max)
{
        size_t count = 0;
        do {
                struct wim_reshdr reshdr;

                get_wim_reshdr(&(entries++)->reshdr, &reshdr);

                if (!(reshdr.flags & WIM_RESHDR_FLAG_PACKED_STREAMS)) {
                        /* Run was terminated by a stand-alone stream entry.  */
                        break;
                }

                if (reshdr.uncompressed_size == WIM_PACK_MAGIC_NUMBER) {
                        /* This is a resource entry.  */
                        count++;
                }
        } while (--max);
        return count;
}

/* Given a run of consecutive lookup table entries with the PACKED_STREAMS flag
 * set and having @num_subpacks resource entries, load resource information from
 * them into the resource specifications in the @subpacks array.
 *
 * Returns 0 on success, or a nonzero error code on failure.  */
static int
do_load_subpack_info(WIMStruct *wim, struct wim_resource_spec **subpacks,
                     size_t num_subpacks,
                     const struct wim_lookup_table_entry_disk *entries)
{
        for (size_t i = 0; i < num_subpacks; i++) {
                struct wim_reshdr reshdr;
                struct alt_chunk_table_header_disk hdr;
                struct wim_resource_spec *rspec;
                int ret;

                /* Advance to next resource entry.  */

                do {
                        get_wim_reshdr(&(entries++)->reshdr, &reshdr);
                } while (reshdr.uncompressed_size != WIM_PACK_MAGIC_NUMBER);

                rspec = subpacks[i];

                wim_res_hdr_to_spec(&reshdr, wim, rspec);

                /* For packed resources, the uncompressed size, compression
                 * type, and chunk size are stored in the resource itself, not
                 * in the lookup table.  */

                ret = full_pread(&wim->in_fd, &hdr,
                                 sizeof(hdr), reshdr.offset_in_wim);
                if (ret) {
                        ERROR("Failed to read header of packed resource "
                              "(offset_in_wim=%"PRIu64")",
                              reshdr.offset_in_wim);
                        return ret;
                }

                rspec->uncompressed_size = le64_to_cpu(hdr.res_usize);

                /* Compression format numbers must be the same as in
                 * WIMGAPI to be compatible here.  */
                BUILD_BUG_ON(WIMLIB_COMPRESSION_TYPE_NONE != 0);
                BUILD_BUG_ON(WIMLIB_COMPRESSION_TYPE_XPRESS != 1);
                BUILD_BUG_ON(WIMLIB_COMPRESSION_TYPE_LZX != 2);
                BUILD_BUG_ON(WIMLIB_COMPRESSION_TYPE_LZMS != 3);
                rspec->compression_type = le32_to_cpu(hdr.compression_format);

                rspec->chunk_size = le32_to_cpu(hdr.chunk_size);

                DEBUG("Subpack %zu/%zu: %"PRIu64" => %"PRIu64" "
                      "(%"TS"/%"PRIu32") @ +%"PRIu64"",
                      i + 1, num_subpacks,
                      rspec->uncompressed_size,
                      rspec->size_in_wim,
                      wimlib_get_compression_type_string(rspec->compression_type),
                      rspec->chunk_size,
                      rspec->offset_in_wim);

        }
        return 0;
}

/* Given a nonempty run of consecutive lookup table entries with the
 * PACKED_STREAMS flag set, allocate a 'struct wim_resource_spec' for each
 * resource within that run.
 *
 * Returns 0 on success, or a nonzero error code on failure.
 * Returns the pointers and count in *subpacks_ret and *num_subpacks_ret.
 */
static int
load_subpack_info(WIMStruct *wim,
                  const struct wim_lookup_table_entry_disk *entries,
                  size_t num_remaining_entries,
                  struct wim_resource_spec ***subpacks_ret,
                  size_t *num_subpacks_ret)
{
        size_t num_subpacks;
        struct wim_resource_spec **subpacks;
        size_t i;
        int ret;

        num_subpacks = count_subpacks(entries, num_remaining_entries);
        subpacks = CALLOC(num_subpacks, sizeof(subpacks[0]));
        if (!subpacks)
                return WIMLIB_ERR_NOMEM;

        for (i = 0; i < num_subpacks; i++) {
                subpacks[i] = MALLOC(sizeof(struct wim_resource_spec));
                if (!subpacks[i]) {
                        ret = WIMLIB_ERR_NOMEM;
                        goto out_free_subpacks;
                }
        }

        ret = do_load_subpack_info(wim, subpacks, num_subpacks, entries);
        if (ret)
                goto out_free_subpacks;

        *subpacks_ret = subpacks;
        *num_subpacks_ret = num_subpacks;
        return 0;

out_free_subpacks:
        for (i = 0; i < num_subpacks; i++)
                FREE(subpacks[i]);
        FREE(subpacks);
        return ret;
}

/* Given a 'struct wim_lookup_table_entry' allocated for a stream entry with
 * PACKED_STREAMS set, try to bind it to a subpack of the current PACKED_STREAMS
 * run.  */
static int
bind_stream_to_subpack(const struct wim_reshdr *reshdr,
                       struct wim_lookup_table_entry *stream,
                       struct wim_resource_spec **subpacks,
                       size_t num_subpacks)
{
        u64 offset = reshdr->offset_in_wim;

        /* XXX: This linear search will be slow in the degenerate case where the
         * number of subpacks is huge.  */
        stream->size = reshdr->size_in_wim;
        stream->flags = reshdr->flags;
        for (size_t i = 0; i < num_subpacks; i++) {
                if (offset + stream->size <= subpacks[i]->uncompressed_size) {
                        stream->offset_in_res = offset;
                        lte_bind_wim_resource_spec(stream, subpacks[i]);
                        return 0;
                }
                offset -= subpacks[i]->uncompressed_size;
        }
        ERROR("Packed stream could not be assigned to any resource");
        return WIMLIB_ERR_INVALID_LOOKUP_TABLE_ENTRY;
}

static void
free_subpack_info(struct wim_resource_spec **subpacks, size_t num_subpacks)
{
        if (subpacks) {
                for (size_t i = 0; i < num_subpacks; i++)
                        if (list_empty(&subpacks[i]->stream_list))
                                FREE(subpacks[i]);
                FREE(subpacks);
        }
}

static int
cmp_streams_by_offset_in_res(const void *p1, const void *p2)
{
        const struct wim_lookup_table_entry *lte1, *lte2;

        lte1 = *(const struct wim_lookup_table_entry**)p1;
        lte2 = *(const struct wim_lookup_table_entry**)p2;

        return cmp_u64(lte1->offset_in_res, lte2->offset_in_res);
}

/* Validate the size and location of a WIM resource.  */
static int
validate_resource(struct wim_resource_spec *rspec)
{
        struct wim_lookup_table_entry *lte;
        bool out_of_order;
        u64 expected_next_offset;
        int ret;

        /* Verify that the resource itself has a valid offset and size.  */
        if (rspec->offset_in_wim + rspec->size_in_wim < rspec->size_in_wim)
                goto invalid_due_to_overflow;

        /* Verify that each stream in the resource has a valid offset and size.
         */
        expected_next_offset = 0;
        out_of_order = false;
        list_for_each_entry(lte, &rspec->stream_list, rspec_node) {
                if (lte->offset_in_res + lte->size < lte->size ||
                    lte->offset_in_res + lte->size > rspec->uncompressed_size)
                        goto invalid_due_to_overflow;

                if (lte->offset_in_res >= expected_next_offset)
                        expected_next_offset = lte->offset_in_res + lte->size;
                else
                        out_of_order = true;
        }

        /* If the streams were not located at strictly increasing positions (not
         * allowing for overlap), sort them.  Then make sure that none overlap.
         */
        if (out_of_order) {
                ret = sort_stream_list(&rspec->stream_list,
                                       offsetof(struct wim_lookup_table_entry,
                                                rspec_node),
                                       cmp_streams_by_offset_in_res);
                if (ret)
                        return ret;

                expected_next_offset = 0;
                list_for_each_entry(lte, &rspec->stream_list, rspec_node) {
                        if (lte->offset_in_res >= expected_next_offset)
                                expected_next_offset = lte->offset_in_res + lte->size;
                        else
                                goto invalid_due_to_overlap;
                }
        }

        return 0;

invalid_due_to_overflow:
        ERROR("Invalid resource entry (offset overflow)");
        return WIMLIB_ERR_INVALID_LOOKUP_TABLE_ENTRY;

invalid_due_to_overlap:
        ERROR("Invalid resource entry (streams in packed resource overlap)");
        return WIMLIB_ERR_INVALID_LOOKUP_TABLE_ENTRY;
}

static int
finish_subpacks(struct wim_resource_spec **subpacks, size_t num_subpacks)
{
        int ret = 0;
        for (size_t i = 0; i < num_subpacks; i++) {
                ret = validate_resource(subpacks[i]);
                if (ret)
                        break;
        }
        free_subpack_info(subpacks, num_subpacks);
        return ret;
}

/*
 * Reads the lookup table from a WIM file.  Usually, each entry specifies a
 * stream that the WIM file contains, along with its location and SHA1 message
 * digest.
 *
 * Saves lookup table entries for non-metadata streams in a hash table (set to
 * wim->lookup_table), and saves the metadata entry for each image in a special
 * per-image location (the wim->image_metadata array).
 *
 * This works for both version WIM_VERSION_DEFAULT (68864) and version
 * WIM_VERSION_PACKED_STREAMS (3584) WIMs.  In the latter, a consecutive run of
 * lookup table entries that all have flag WIM_RESHDR_FLAG_PACKED_STREAMS (0x10)
 * set is a "packed run".  A packed run logically contains zero or more
 * resources, each of which logically contains zero or more streams.
 * Physically, in such a run, a "lookup table entry" with uncompressed size
 * WIM_PACK_MAGIC_NUMBER (0x100000000) specifies a resource, whereas any other
 * entry specifies a stream.  Within such a run, stream entries and resource
 * entries need not be in any particular order, except that the order of the
 * resource entries is important, as it affects how streams are assigned to
 * resources.  See the code for details.
 *
 * Possible return values:
 *      WIMLIB_ERR_SUCCESS (0)
 *      WIMLIB_ERR_INVALID_LOOKUP_TABLE_ENTRY
 *      WIMLIB_ERR_NOMEM
 *
 *      Or an error code caused by failure to read the lookup table from the WIM
 *      file.
 */
int
read_wim_lookup_table(WIMStruct *wim)
{
        int ret;
        size_t num_entries;
        void *buf = NULL;
        struct wim_lookup_table *table = NULL;
        struct wim_lookup_table_entry *cur_entry = NULL;
        size_t num_duplicate_entries = 0;
        size_t num_wrong_part_entries = 0;
        u32 image_index = 0;
        struct wim_resource_spec **cur_subpacks = NULL;
        size_t cur_num_subpacks = 0;

        DEBUG("Reading lookup table.");

        /* Sanity check: lookup table entries are 50 bytes each.  */
        BUILD_BUG_ON(sizeof(struct wim_lookup_table_entry_disk) !=
                     WIM_LOOKUP_TABLE_ENTRY_DISK_SIZE);

        /* Calculate the number of entries in the lookup table.  */
        num_entries = wim->hdr.lookup_table_reshdr.uncompressed_size /
                      sizeof(struct wim_lookup_table_entry_disk);

        /* Read the lookup table into a buffer.  */
        ret = wim_reshdr_to_data(&wim->hdr.lookup_table_reshdr, wim, &buf);
        if (ret)
                goto out;

        /* Allocate a hash table to map SHA1 message digests into stream
         * specifications.  This is the in-memory "lookup table".  */
        table = new_lookup_table(num_entries * 2 + 1);
        if (!table)
                goto oom;

        /* Allocate and initalize stream entries ('struct
         * wim_lookup_table_entry's) from the raw lookup table buffer.  Each of
         * these entries will point to a 'struct wim_resource_spec' that
         * describes the underlying resource.  In WIMs with version number
         * WIM_VERSION_PACKED_STREAMS, a resource may contain multiple streams.
         */
        for (size_t i = 0; i < num_entries; i++) {
                const struct wim_lookup_table_entry_disk *disk_entry =
                        &((const struct wim_lookup_table_entry_disk*)buf)[i];
                struct wim_reshdr reshdr;
                u16 part_number;

                /* Get the resource header  */
                get_wim_reshdr(&disk_entry->reshdr, &reshdr);

                DEBUG("reshdr: size_in_wim=%"PRIu64", "
                      "uncompressed_size=%"PRIu64", "
                      "offset_in_wim=%"PRIu64", "
                      "flags=0x%02x",
                      reshdr.size_in_wim, reshdr.uncompressed_size,
                      reshdr.offset_in_wim, reshdr.flags);

                /* Ignore PACKED_STREAMS flag if it isn't supposed to be used in
                 * this WIM version.  */
                if (wim->hdr.wim_version == WIM_VERSION_DEFAULT)
                        reshdr.flags &= ~WIM_RESHDR_FLAG_PACKED_STREAMS;

                /* Allocate a new 'struct wim_lookup_table_entry'.  */
                cur_entry = new_lookup_table_entry();
                if (!cur_entry)
                        goto oom;

                /* Get the part number, reference count, and hash.  */
                part_number = le16_to_cpu(disk_entry->part_number);
                cur_entry->refcnt = le32_to_cpu(disk_entry->refcnt);
                copy_hash(cur_entry->hash, disk_entry->hash);

                if (reshdr.flags & WIM_RESHDR_FLAG_PACKED_STREAMS) {

                        /* PACKED_STREAMS entry  */

                        if (!cur_subpacks) {
                                /* Starting new run  */
                                ret = load_subpack_info(wim, disk_entry,
                                                        num_entries - i,
                                                        &cur_subpacks,
                                                        &cur_num_subpacks);
                                if (ret)
                                        goto out;
                        }

                        if (reshdr.uncompressed_size == WIM_PACK_MAGIC_NUMBER) {
                                /* Resource entry, not stream entry  */
                                goto free_cur_entry_and_continue;
                        }

                        /* Stream entry  */

                        ret = bind_stream_to_subpack(&reshdr,
                                                     cur_entry,
                                                     cur_subpacks,
                                                     cur_num_subpacks);
                        if (ret)
                                goto out;

                } else {
                        /* Normal stream/resource entry; PACKED_STREAMS not set.
                         */

                        struct wim_resource_spec *rspec;

                        if (unlikely(cur_subpacks)) {
                                /* This entry terminated a packed run.  */
                                ret = finish_subpacks(cur_subpacks,
                                                      cur_num_subpacks);
                                cur_subpacks = NULL;
                                if (ret)
                                        goto out;
                        }

                        /* How to handle an uncompressed resource with its
                         * uncompressed size different from its compressed size?
                         *
                         * Based on a simple test, WIMGAPI seems to handle this
                         * as follows:
                         *
                         * if (size_in_wim > uncompressed_size) {
                         *      Ignore uncompressed_size; use size_in_wim
                         *      instead.
                         * } else {
                         *      Honor uncompressed_size, but treat the part of
                         *      the file data above size_in_wim as all zeros.
                         * }
                         *
                         * So we will do the same.  */
                        if (unlikely(!(reshdr.flags &
                                       WIM_RESHDR_FLAG_COMPRESSED) &&
                                     (reshdr.size_in_wim >
                                      reshdr.uncompressed_size)))
                        {
                                reshdr.uncompressed_size = reshdr.size_in_wim;
                        }

                        /* Set up a resource specification for this stream.  */

                        rspec = MALLOC(sizeof(struct wim_resource_spec));
                        if (!rspec)
                                goto oom;

                        wim_res_hdr_to_spec(&reshdr, wim, rspec);

                        cur_entry->offset_in_res = 0;
                        cur_entry->size = reshdr.uncompressed_size;
                        cur_entry->flags = reshdr.flags;

                        lte_bind_wim_resource_spec(cur_entry, rspec);
                }

                /* cur_entry is now a stream bound to a resource.  */

                /* Ignore entries with all zeroes in the hash field.  */
                if (is_zero_hash(cur_entry->hash))
                        goto free_cur_entry_and_continue;

                /* Verify that the part number matches that of the underlying
                 * WIM file.  */
                if (part_number != wim->hdr.part_number) {
                        num_wrong_part_entries++;
                        goto free_cur_entry_and_continue;
                }

                if (reshdr.flags & WIM_RESHDR_FLAG_METADATA) {

                        /* Lookup table entry for a metadata resource.  */

                        /* Metadata entries with no references must be ignored.
                         * See, for example, the WinPE WIMs from the WAIK v2.1.
                         */
                        if (cur_entry->refcnt == 0)
                                goto free_cur_entry_and_continue;

                        if (cur_entry->refcnt != 1) {
                                /* We don't currently support this case due to
                                 * the complications of multiple images sharing
                                 * the same metadata resource or a metadata
                                 * resource also being referenced by files.  */
                                ERROR("Found metadata resource with refcnt != 1");
                                ret = WIMLIB_ERR_INVALID_LOOKUP_TABLE_ENTRY;
                                goto out;
                        }

                        if (wim->hdr.part_number != 1) {
                                WARNING("Ignoring metadata resource found in a "
                                        "non-first part of the split WIM");
                                goto free_cur_entry_and_continue;
                        }

                        /* The number of entries in the lookup table with
                         * WIM_RESHDR_FLAG_METADATA set should be the same as
                         * the image_count field in the WIM header.  */
                        if (image_index == wim->hdr.image_count) {
                                WARNING("Found more metadata resources than images");
                                goto free_cur_entry_and_continue;
                        }

                        /* Notice very carefully:  We are assigning the metadata
                         * resources to images in the same order in which their
                         * lookup table entries occur on disk.  (This is also
                         * the behavior of Microsoft's software.)  In
                         * particular, this overrides the actual locations of
                         * the metadata resources themselves in the WIM file as
                         * well as any information written in the XML data.  */
                        DEBUG("Found metadata resource for image %"PRIu32" at "
                              "offset %"PRIu64".",
                              image_index + 1,
                              reshdr.offset_in_wim);

                        wim->image_metadata[image_index++]->metadata_lte = cur_entry;
                } else {
                        /* Lookup table entry for a non-metadata stream.  */

                        /* Ignore this stream if it's a duplicate.  */
                        if (lookup_stream(table, cur_entry->hash)) {
                                num_duplicate_entries++;
                                goto free_cur_entry_and_continue;
                        }

                        /* Insert the stream into the in-memory lookup table,
                         * keyed by its SHA1 message digest.  */
                        lookup_table_insert(table, cur_entry);
                }

                continue;

        free_cur_entry_and_continue:
                if (cur_subpacks &&
                    cur_entry->resource_location == RESOURCE_IN_WIM)
                        lte_unbind_wim_resource_spec(cur_entry);
                free_lookup_table_entry(cur_entry);
        }
        cur_entry = NULL;

        if (cur_subpacks) {
                /* End of lookup table terminated a packed run.  */
                ret = finish_subpacks(cur_subpacks, cur_num_subpacks);
                cur_subpacks = NULL;
                if (ret)
                        goto out;
        }

        if (wim->hdr.part_number == 1 && image_index != wim->hdr.image_count) {
                WARNING("Could not find metadata resources for all images");
                for (u32 i = image_index; i < wim->hdr.image_count; i++)
                        put_image_metadata(wim->image_metadata[i], NULL);
                wim->hdr.image_count = image_index;
        }

        if (num_duplicate_entries > 0) {
                WARNING("Ignoring %zu duplicate streams in the WIM lookup table",
                        num_duplicate_entries);
        }

        if (num_wrong_part_entries > 0) {
                WARNING("Ignoring %zu streams with wrong part number",
                        num_wrong_part_entries);
        }

        DEBUG("Done reading lookup table.");
        wim->lookup_table = table;
        ret = 0;
        goto out_free_buf;

oom:
        ERROR("Not enough memory to read lookup table!");
        ret = WIMLIB_ERR_NOMEM;
out:
        free_subpack_info(cur_subpacks, cur_num_subpacks);
        free_lookup_table_entry(cur_entry);
        free_lookup_table(table);
out_free_buf:
        FREE(buf);
        return ret;
}

static void
put_wim_lookup_table_entry(struct wim_lookup_table_entry_disk *disk_entry,
                           const struct wim_reshdr *out_reshdr,
                           u16 part_number, u32 refcnt, const u8 *hash)
{
        put_wim_reshdr(out_reshdr, &disk_entry->reshdr);
        disk_entry->part_number = cpu_to_le16(part_number);
        disk_entry->refcnt = cpu_to_le32(refcnt);
        copy_hash(disk_entry->hash, hash);
}

/* Note: the list of stream entries must be sorted so that all entries for the
 * same packed resource are consecutive.  In addition, entries with
 * WIM_RESHDR_FLAG_METADATA set must be in the same order as the indices of the
 * underlying images.  */
int
write_wim_lookup_table_from_stream_list(struct list_head *stream_list,
                                        struct filedes *out_fd,
                                        u16 part_number,
                                        struct wim_reshdr *out_reshdr,
                                        int write_resource_flags)
{
        size_t table_size;
        struct wim_lookup_table_entry *lte;
        struct wim_lookup_table_entry_disk *table_buf;
        struct wim_lookup_table_entry_disk *table_buf_ptr;
        int ret;
        u64 prev_res_offset_in_wim = ~0ULL;
        u64 prev_uncompressed_size;
        u64 logical_offset;

        table_size = 0;
        list_for_each_entry(lte, stream_list, lookup_table_list) {
                table_size += sizeof(struct wim_lookup_table_entry_disk);

                if (lte->out_reshdr.flags & WIM_RESHDR_FLAG_PACKED_STREAMS &&
                    lte->out_res_offset_in_wim != prev_res_offset_in_wim)
                {
                        table_size += sizeof(struct wim_lookup_table_entry_disk);
                        prev_res_offset_in_wim = lte->out_res_offset_in_wim;
                }
        }

        DEBUG("Writing WIM lookup table (size=%zu, offset=%"PRIu64")",
              table_size, out_fd->offset);

        table_buf = MALLOC(table_size);
        if (table_buf == NULL) {
                ERROR("Failed to allocate %zu bytes for temporary lookup table",
                      table_size);
                return WIMLIB_ERR_NOMEM;
        }
        table_buf_ptr = table_buf;

        prev_res_offset_in_wim = ~0ULL;
        prev_uncompressed_size = 0;
        logical_offset = 0;
        list_for_each_entry(lte, stream_list, lookup_table_list) {
                if (lte->out_reshdr.flags & WIM_RESHDR_FLAG_PACKED_STREAMS) {
                        struct wim_reshdr tmp_reshdr;

                        /* Eww.  When WIMGAPI sees multiple resource packs, it
                         * expects the offsets to be adjusted as if there were
                         * really only one pack.  */

                        if (lte->out_res_offset_in_wim != prev_res_offset_in_wim) {
                                /* Put the resource entry for pack  */
                                tmp_reshdr.offset_in_wim = lte->out_res_offset_in_wim;
                                tmp_reshdr.size_in_wim = lte->out_res_size_in_wim;
                                tmp_reshdr.uncompressed_size = WIM_PACK_MAGIC_NUMBER;
                                tmp_reshdr.flags = WIM_RESHDR_FLAG_PACKED_STREAMS;

                                put_wim_lookup_table_entry(table_buf_ptr++,
                                                           &tmp_reshdr,
                                                           part_number,
                                                           1, zero_hash);

                                logical_offset += prev_uncompressed_size;

                                prev_res_offset_in_wim = lte->out_res_offset_in_wim;
                                prev_uncompressed_size = lte->out_res_uncompressed_size;
                        }
                        tmp_reshdr = lte->out_reshdr;
                        tmp_reshdr.offset_in_wim += logical_offset;
                        put_wim_lookup_table_entry(table_buf_ptr++,
                                                   &tmp_reshdr,
                                                   part_number,
                                                   lte->out_refcnt,
                                                   lte->hash);
                } else {
                        put_wim_lookup_table_entry(table_buf_ptr++,
                                                   &lte->out_reshdr,
                                                   part_number,
                                                   lte->out_refcnt,
                                                   lte->hash);
                }

        }
        wimlib_assert((u8*)table_buf_ptr - (u8*)table_buf == table_size);

        /* Write the lookup table uncompressed.  Although wimlib can handle a
         * compressed lookup table, MS software cannot.  */
        ret = write_wim_resource_from_buffer(table_buf,
                                             table_size,
                                             WIM_RESHDR_FLAG_METADATA,
                                             out_fd,
                                             WIMLIB_COMPRESSION_TYPE_NONE,
                                             0,
                                             out_reshdr,
                                             NULL,
                                             write_resource_flags);
        FREE(table_buf);
        DEBUG("ret=%d", ret);
        return ret;
}

/* Allocate a stream entry for the contents of the buffer, or re-use an existing
 * entry in @lookup_table for the same stream.  */
struct wim_lookup_table_entry *
new_stream_from_data_buffer(const void *buffer, size_t size,
                            struct wim_lookup_table *lookup_table)
{
        u8 hash[SHA1_HASH_SIZE];
        struct wim_lookup_table_entry *lte, *existing_lte;

        sha1_buffer(buffer, size, hash);
        existing_lte = lookup_stream(lookup_table, hash);
        if (existing_lte) {
                wimlib_assert(existing_lte->size == size);
                lte = existing_lte;
                lte->refcnt++;
        } else {
                void *buffer_copy;
                lte = new_lookup_table_entry();
                if (lte == NULL)
                        return NULL;
                buffer_copy = memdup(buffer, size);
                if (buffer_copy == NULL) {
                        free_lookup_table_entry(lte);
                        return NULL;
                }
                lte->resource_location  = RESOURCE_IN_ATTACHED_BUFFER;
                lte->attached_buffer    = buffer_copy;
                lte->size               = size;
                copy_hash(lte->hash, hash);
                lookup_table_insert(lookup_table, lte);
        }
        return lte;
}

/* Calculate the SHA1 message digest of a stream and move it from the list of
 * unhashed streams to the stream lookup table, possibly joining it with an
 * existing lookup table entry for an identical stream.
 *
 * @lte:  An unhashed lookup table entry.
 * @lookup_table:  Lookup table for the WIM.
 * @lte_ret:  On success, write a pointer to the resulting lookup table
 *            entry to this location.  This will be the same as @lte
 *            if it was inserted into the lookup table, or different if
 *            a duplicate stream was found.
 *
 * Returns 0 on success; nonzero if there is an error reading the stream.
 */
int
hash_unhashed_stream(struct wim_lookup_table_entry *lte,
                     struct wim_lookup_table *lookup_table,
                     struct wim_lookup_table_entry **lte_ret)
{
        int ret;
        struct wim_lookup_table_entry *duplicate_lte;
        struct wim_lookup_table_entry **back_ptr;

        wimlib_assert(lte->unhashed);

        /* back_ptr must be saved because @back_inode and @back_stream_id are in
         * union with the SHA1 message digest and will no longer be valid once
         * the SHA1 has been calculated. */
        back_ptr = retrieve_lte_pointer(lte);

        ret = sha1_stream(lte);
        if (ret)
                return ret;

        /* Look for a duplicate stream */
        duplicate_lte = lookup_stream(lookup_table, lte->hash);
        list_del(&lte->unhashed_list);
        if (duplicate_lte) {
                /* We have a duplicate stream.  Transfer the reference counts
                 * from this stream to the duplicate and update the reference to
                 * this stream (in an inode or ads_entry) to point to the
                 * duplicate.  The caller is responsible for freeing @lte if
                 * needed.  */
                wimlib_assert(!(duplicate_lte->unhashed));
                wimlib_assert(duplicate_lte->size == lte->size);
                duplicate_lte->refcnt += lte->refcnt;
                lte->refcnt = 0;
                *back_ptr = duplicate_lte;
                lte = duplicate_lte;
        } else {
                /* No duplicate stream, so we need to insert this stream into
                 * the lookup table and treat it as a hashed stream. */
                lookup_table_insert(lookup_table, lte);
                lte->unhashed = 0;
        }
        *lte_ret = lte;
        return 0;
}

void
lte_to_wimlib_resource_entry(const struct wim_lookup_table_entry *lte,
                             struct wimlib_resource_entry *wentry)
{
        memset(wentry, 0, sizeof(*wentry));

        wentry->uncompressed_size = lte->size;
        if (lte->resource_location == RESOURCE_IN_WIM) {
                wentry->part_number = lte->rspec->wim->hdr.part_number;
                if (lte->flags & WIM_RESHDR_FLAG_PACKED_STREAMS) {
                        wentry->compressed_size = 0;
                        wentry->offset = lte->offset_in_res;
                } else {
                        wentry->compressed_size = lte->rspec->size_in_wim;
                        wentry->offset = lte->rspec->offset_in_wim;
                }
                wentry->raw_resource_offset_in_wim = lte->rspec->offset_in_wim;
                /*wentry->raw_resource_uncompressed_size = lte->rspec->uncompressed_size;*/
                wentry->raw_resource_compressed_size = lte->rspec->size_in_wim;
        }
        copy_hash(wentry->sha1_hash, lte->hash);
        wentry->reference_count = lte->refcnt;
        wentry->is_compressed = (lte->flags & WIM_RESHDR_FLAG_COMPRESSED) != 0;
        wentry->is_metadata = (lte->flags & WIM_RESHDR_FLAG_METADATA) != 0;
        wentry->is_free = (lte->flags & WIM_RESHDR_FLAG_FREE) != 0;
        wentry->is_spanned = (lte->flags & WIM_RESHDR_FLAG_SPANNED) != 0;
        wentry->packed = (lte->flags & WIM_RESHDR_FLAG_PACKED_STREAMS) != 0;
}

struct iterate_lte_context {
        wimlib_iterate_lookup_table_callback_t cb;
        void *user_ctx;
};

static int
do_iterate_lte(struct wim_lookup_table_entry *lte, void *_ctx)
{
        struct iterate_lte_context *ctx = _ctx;
        struct wimlib_resource_entry entry;

        lte_to_wimlib_resource_entry(lte, &entry);
        return (*ctx->cb)(&entry, ctx->user_ctx);
}

/* API function documented in wimlib.h  */
WIMLIBAPI int
wimlib_iterate_lookup_table(WIMStruct *wim, int flags,
                            wimlib_iterate_lookup_table_callback_t cb,
                            void *user_ctx)
{
        if (flags != 0)
                return WIMLIB_ERR_INVALID_PARAM;

        struct iterate_lte_context ctx = {
                .cb = cb,
                .user_ctx = user_ctx,
        };
        if (wim_has_metadata(wim)) {
                int ret;
                for (int i = 0; i < wim->hdr.image_count; i++) {
                        ret = do_iterate_lte(wim->image_metadata[i]->metadata_lte,
                                             &ctx);
                        if (ret)
                                return ret;
                }
        }
        return for_lookup_table_entry(wim->lookup_table, do_iterate_lte, &ctx);
}