/* * unix_apply.c - Code to apply files from a WIM image on UNIX. */ /* * Copyright (C) 2012, 2013, 2014 Eric Biggers * * This file is part of wimlib, a library for working with WIM files. * * wimlib is free software; you can redistribute it and/or modify it under the * terms of the GNU General Public License as published by the Free * Software Foundation; either version 3 of the License, or (at your option) * any later version. * * wimlib 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 General Public License for more * details. * * You should have received a copy of the GNU General Public License * along with wimlib; if not, see http://www.gnu.org/licenses/. */ #ifdef HAVE_CONFIG_H # include "config.h" #endif #include "wimlib/apply.h" #include "wimlib/dentry.h" #include "wimlib/error.h" #include "wimlib/file_io.h" #include "wimlib/reparse.h" #include "wimlib/timestamp.h" #include "wimlib/unix_data.h" #include #include #include #include #include #include #include #include /* We don't require O_NOFOLLOW, but the advantage of having it is that if we * need to extract a file to a location at which there exists a symbolic link, * open(..., O_NOFOLLOW | ...) recognizes the symbolic link rather than * following it and creating the file somewhere else. (Equivalent to * FILE_OPEN_REPARSE_POINT on Windows.) */ #ifndef O_NOFOLLOW # define O_NOFOLLOW 0 #endif static int unix_get_supported_features(const char *target, struct wim_features *supported_features) { supported_features->hard_links = 1; supported_features->symlink_reparse_points = 1; supported_features->unix_data = 1; supported_features->timestamps = 1; supported_features->case_sensitive_filenames = 1; return 0; } #define NUM_PATHBUFS 2 /* We need 2 when creating hard links */ #define MAX_OPEN_FDS 1000 /* TODO: Add special case for when the number of identical streams exceeds this number. */ struct unix_apply_ctx { /* Extract flags, the pointer to the WIMStruct, etc. */ struct apply_ctx common; /* Buffers for building extraction paths (allocated). */ char *pathbufs[NUM_PATHBUFS]; /* Index of next pathbuf to use */ unsigned which_pathbuf; /* Currently open file descriptors for extraction */ struct filedes open_fds[MAX_OPEN_FDS]; /* Number of currently open file descriptors in open_fds, starting from * the beginning of the array. */ unsigned num_open_fds; /* Buffer for reading reparse data streams into memory */ u8 reparse_data[REPARSE_DATA_MAX_SIZE]; /* Pointer to the next byte in @reparse_data to fill */ u8 *reparse_ptr; /* Absolute path to the target directory (allocated buffer). Only set * if needed for absolute symbolic link fixups. */ char *target_abspath; /* Number of characters in target_abspath. */ size_t target_abspath_nchars; /* Number of special files we couldn't create due to EPERM */ unsigned long num_special_files_ignored; }; /* Returns the number of characters needed to represent the path to the * specified @dentry when extracted, not including the null terminator or the * path to the target directory itself. */ static size_t unix_dentry_path_length(const struct wim_dentry *dentry) { size_t len = 0; const struct wim_dentry *d; d = dentry; do { len += d->d_extraction_name_nchars + 1; d = d->d_parent; } while (!dentry_is_root(d) && will_extract_dentry(d)); return len; } /* Returns the maximum number of characters needed to represent the path to any * dentry in @dentry_list when extracted, including the null terminator and the * path to the target directory itself. */ static size_t unix_compute_path_max(const struct list_head *dentry_list, const struct unix_apply_ctx *ctx) { size_t max = 0; size_t len; const struct wim_dentry *dentry; list_for_each_entry(dentry, dentry_list, d_extraction_list_node) { len = unix_dentry_path_length(dentry); if (len > max) max = len; } /* Account for target and null terminator. */ return ctx->common.target_nchars + max + 1; } /* Builds and returns the filesystem path to which to extract @dentry. * This cycles through NUM_PATHBUFS different buffers. */ static const char * unix_build_extraction_path(const struct wim_dentry *dentry, struct unix_apply_ctx *ctx) { char *pathbuf; char *p; const struct wim_dentry *d; pathbuf = ctx->pathbufs[ctx->which_pathbuf]; ctx->which_pathbuf = (ctx->which_pathbuf + 1) % NUM_PATHBUFS; p = &pathbuf[ctx->common.target_nchars + unix_dentry_path_length(dentry)]; *p = '\0'; d = dentry; do { p -= d->d_extraction_name_nchars; memcpy(p, d->d_extraction_name, d->d_extraction_name_nchars); *--p = '/'; d = d->d_parent; } while (!dentry_is_root(d) && will_extract_dentry(d)); return pathbuf; } /* This causes the next call to unix_build_extraction_path() to use the same * path buffer as the previous call. */ static void unix_reuse_pathbuf(struct unix_apply_ctx *ctx) { ctx->which_pathbuf = (ctx->which_pathbuf - 1) % NUM_PATHBUFS; } /* Builds and returns the filesystem path to which to extract an unspecified * alias of the @inode. This cycles through NUM_PATHBUFS different buffers. */ static const char * unix_build_inode_extraction_path(const struct wim_inode *inode, struct unix_apply_ctx *ctx) { return unix_build_extraction_path(inode_first_extraction_dentry(inode), ctx); } /* Sets the timestamps on a file being extracted. * * Either @fd or @path must be specified (not -1 and not NULL, respectively). */ static int unix_set_timestamps(int fd, const char *path, u64 atime, u64 mtime) { { struct timespec times[2]; times[0] = wim_timestamp_to_timespec(atime); times[1] = wim_timestamp_to_timespec(mtime); errno = ENOSYS; #ifdef HAVE_FUTIMENS if (fd >= 0 && !futimens(fd, times)) return 0; #endif #ifdef HAVE_UTIMENSAT if (fd < 0 && !utimensat(AT_FDCWD, path, times, AT_SYMLINK_NOFOLLOW)) return 0; #endif if (errno != ENOSYS) return WIMLIB_ERR_SET_TIMESTAMPS; } { struct timeval times[2]; times[0] = wim_timestamp_to_timeval(atime); times[1] = wim_timestamp_to_timeval(mtime); if (fd >= 0 && !futimes(fd, times)) return 0; if (fd < 0 && !lutimes(path, times)) return 0; return WIMLIB_ERR_SET_TIMESTAMPS; } } static int unix_set_owner_and_group(int fd, const char *path, uid_t uid, gid_t gid) { if (fd >= 0 && !fchown(fd, uid, gid)) return 0; if (fd < 0 && !lchown(path, uid, gid)) return 0; return WIMLIB_ERR_SET_SECURITY; } static int unix_set_mode(int fd, const char *path, mode_t mode) { if (fd >= 0 && !fchmod(fd, mode)) return 0; if (fd < 0 && !chmod(path, mode)) return 0; return WIMLIB_ERR_SET_SECURITY; } /* * Set metadata on an extracted file. * * @fd is an open file descriptor to the extracted file, or -1. @path is the * path to the extracted file, or NULL. If valid, this function uses @fd. * Otherwise, if valid, it uses @path. Otherwise, it calculates the path to one * alias of the extracted file and uses it. */ static int unix_set_metadata(int fd, const struct wim_inode *inode, const char *path, struct unix_apply_ctx *ctx) { int ret; struct wimlib_unix_data unix_data; if (fd < 0 && !path) path = unix_build_inode_extraction_path(inode, ctx); if ((ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_UNIX_DATA) && inode_get_unix_data(inode, &unix_data)) { u32 uid = unix_data.uid; u32 gid = unix_data.gid; u32 mode = unix_data.mode; ret = unix_set_owner_and_group(fd, path, uid, gid); if (ret) { if (!path) path = unix_build_inode_extraction_path(inode, ctx); if (ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_STRICT_ACLS) { ERROR_WITH_ERRNO("Can't set uid=%"PRIu32" and " "gid=%"PRIu32" on \"%s\"", uid, gid, path); return ret; } else { WARNING_WITH_ERRNO("Can't set uid=%"PRIu32" and " "gid=%"PRIu32" on \"%s\"", uid, gid, path); } } ret = 0; if (!inode_is_symlink(inode)) ret = unix_set_mode(fd, path, mode); if (ret) { if (!path) path = unix_build_inode_extraction_path(inode, ctx); if (ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_STRICT_ACLS) { ERROR_WITH_ERRNO("Can't set mode=0%"PRIo32" " "on \"%s\"", mode, path); return ret; } else { WARNING_WITH_ERRNO("Can't set mode=0%"PRIo32" " "on \"%s\"", mode, path); } } } ret = unix_set_timestamps(fd, path, inode->i_last_access_time, inode->i_last_write_time); if (ret) { if (!path) path = unix_build_inode_extraction_path(inode, ctx); if (ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_STRICT_TIMESTAMPS) { ERROR_WITH_ERRNO("Can't set timestamps on \"%s\"", path); return ret; } else { WARNING_WITH_ERRNO("Can't set timestamps on \"%s\"", path); } } return 0; } /* Extract all needed aliases of the @inode, where one alias, corresponding to * @first_dentry, has already been extracted to @first_path. */ static int unix_create_hardlinks(const struct wim_inode *inode, const struct wim_dentry *first_dentry, const char *first_path, struct unix_apply_ctx *ctx) { const struct wim_dentry *dentry; const char *newpath; list_for_each_entry(dentry, &inode->i_extraction_aliases, d_extraction_alias_node) { if (dentry == first_dentry) continue; newpath = unix_build_extraction_path(dentry, ctx); retry_link: if (link(first_path, newpath)) { if (errno == EEXIST && !unlink(newpath)) goto retry_link; ERROR_WITH_ERRNO("Can't create hard link " "\"%s\" => \"%s\"", newpath, first_path); return WIMLIB_ERR_LINK; } unix_reuse_pathbuf(ctx); } return 0; } /* If @dentry represents a directory, create it. */ static int unix_create_if_directory(const struct wim_dentry *dentry, struct unix_apply_ctx *ctx) { const char *path; struct stat stbuf; if (!dentry_is_directory(dentry)) return 0; path = unix_build_extraction_path(dentry, ctx); if (mkdir(path, 0755) && /* It's okay if the path already exists, as long as it's a * directory. */ !(errno == EEXIST && !lstat(path, &stbuf) && S_ISDIR(stbuf.st_mode))) { ERROR_WITH_ERRNO("Can't create directory \"%s\"", path); return WIMLIB_ERR_MKDIR; } return 0; } /* If @dentry represents an empty regular file or a special file, create it, set * its metadata, and create any needed hard links. */ static int unix_extract_if_empty_file(const struct wim_dentry *dentry, struct unix_apply_ctx *ctx) { const struct wim_inode *inode; struct wimlib_unix_data unix_data; const char *path; int ret; inode = dentry->d_inode; /* Extract all aliases only when the "first" comes up. */ if (dentry != inode_first_extraction_dentry(inode)) return 0; /* Is this a directory, a symbolic link, or any type of nonempty file? */ if (inode_is_directory(inode) || inode_is_symlink(inode) || inode_unnamed_lte_resolved(inode)) return 0; /* Recognize special files in UNIX_DATA mode */ if ((ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_UNIX_DATA) && inode_get_unix_data(inode, &unix_data) && !S_ISREG(unix_data.mode)) { path = unix_build_extraction_path(dentry, ctx); retry_mknod: if (mknod(path, unix_data.mode, unix_data.rdev)) { if (errno == EPERM) { WARNING_WITH_ERRNO("Can't create special " "file \"%s\"", path); ctx->num_special_files_ignored++; return 0; } if (errno == EEXIST && !unlink(path)) goto retry_mknod; ERROR_WITH_ERRNO("Can't create special file \"%s\"", path); return WIMLIB_ERR_MKNOD; } /* On special files, we can set timestamps immediately because * we don't need to write any data to them. */ ret = unix_set_metadata(-1, inode, path, ctx); } else { int fd; path = unix_build_extraction_path(dentry, ctx); retry_create: fd = open(path, O_TRUNC | O_CREAT | O_WRONLY | O_NOFOLLOW, 0644); if (fd < 0) { if (errno == EEXIST && !unlink(path)) goto retry_create; ERROR_WITH_ERRNO("Can't create regular file \"%s\"", path); return WIMLIB_ERR_OPEN; } /* On empty files, we can set timestamps immediately because we * don't need to write any data to them. */ ret = unix_set_metadata(fd, inode, path, ctx); if (close(fd) && !ret) { ERROR_WITH_ERRNO("Error closing \"%s\"", path); ret = WIMLIB_ERR_WRITE; } } if (ret) return ret; return unix_create_hardlinks(inode, dentry, path, ctx); } static int unix_create_dirs_and_empty_files(const struct list_head *dentry_list, struct unix_apply_ctx *ctx) { const struct wim_dentry *dentry; int ret; list_for_each_entry(dentry, dentry_list, d_extraction_list_node) { ret = unix_create_if_directory(dentry, ctx); if (ret) return ret; } list_for_each_entry(dentry, dentry_list, d_extraction_list_node) { ret = unix_extract_if_empty_file(dentry, ctx); if (ret) return ret; } return 0; } static int unix_create_symlink(const struct wim_inode *inode, const char *path, const u8 *rpdata, u16 rpdatalen, bool rpfix, const char *apply_dir, size_t apply_dir_nchars) { char link_target[REPARSE_DATA_MAX_SIZE]; int ret; struct wim_lookup_table_entry lte_override; lte_override.resource_location = RESOURCE_IN_ATTACHED_BUFFER; lte_override.attached_buffer = (void *)rpdata; lte_override.size = rpdatalen; ret = wim_inode_readlink(inode, link_target, sizeof(link_target) - 1, <e_override); if (ret < 0) { errno = -ret; return WIMLIB_ERR_READLINK; } link_target[ret] = 0; if (rpfix && link_target[0] == '/') { /* "Fix" the absolute symbolic link by prepending the absolute * path to the target directory. */ if (sizeof(link_target) - (ret + 1) < apply_dir_nchars) { errno = ENAMETOOLONG; return WIMLIB_ERR_REPARSE_POINT_FIXUP_FAILED; } memmove(link_target + apply_dir_nchars, link_target, ret + 1); memcpy(link_target, apply_dir, apply_dir_nchars); } retry_symlink: if (symlink(link_target, path)) { if (errno == EEXIST && !unlink(path)) goto retry_symlink; return WIMLIB_ERR_LINK; } return 0; } static void unix_cleanup_open_fds(struct unix_apply_ctx *ctx, unsigned offset) { for (unsigned i = offset; i < ctx->num_open_fds; i++) filedes_close(&ctx->open_fds[i]); ctx->num_open_fds = 0; } static int unix_begin_extract_stream_instance(const struct wim_lookup_table_entry *stream, const struct wim_inode *inode, struct unix_apply_ctx *ctx) { const struct wim_dentry *first_dentry; const char *first_path; int fd; if (inode_is_symlink(inode)) { /* On UNIX, symbolic links must be created with symlink(), which * requires that the full link target be available. */ if (stream->size > REPARSE_DATA_MAX_SIZE) { ERROR_WITH_ERRNO("Reparse data of \"%s\" has size " "%"PRIu64" bytes (exceeds %u bytes)", inode_first_full_path(inode), stream->size, REPARSE_DATA_MAX_SIZE); return WIMLIB_ERR_INVALID_REPARSE_DATA; } ctx->reparse_ptr = ctx->reparse_data; return 0; } if (ctx->num_open_fds == MAX_OPEN_FDS) { ERROR("Can't extract data: too many open files!"); return WIMLIB_ERR_UNSUPPORTED; } first_dentry = inode_first_extraction_dentry(inode); first_path = unix_build_extraction_path(first_dentry, ctx); retry_create: fd = open(first_path, O_TRUNC | O_CREAT | O_WRONLY | O_NOFOLLOW, 0644); if (fd < 0) { if (errno == EEXIST && !unlink(first_path)) goto retry_create; ERROR_WITH_ERRNO("Can't create regular file \"%s\"", first_path); return WIMLIB_ERR_OPEN; } filedes_init(&ctx->open_fds[ctx->num_open_fds++], fd); return unix_create_hardlinks(inode, first_dentry, first_path, ctx); } /* Called when starting to read a single-instance stream for extraction */ static int unix_begin_extract_stream(struct wim_lookup_table_entry *stream, u32 flags, void *_ctx) { struct unix_apply_ctx *ctx = _ctx; const struct stream_owner *owners = stream_owners(stream); int ret; for (u32 i = 0; i < stream->out_refcnt; i++) { const struct wim_inode *inode = owners[i].inode; ret = unix_begin_extract_stream_instance(stream, inode, ctx); if (ret) { ctx->reparse_ptr = NULL; unix_cleanup_open_fds(ctx, 0); return ret; } } return 0; } /* Called when the next chunk of a single-instance stream has been read for * extraction */ static int unix_extract_chunk(const void *chunk, size_t size, void *_ctx) { struct unix_apply_ctx *ctx = _ctx; int ret; for (unsigned i = 0; i < ctx->num_open_fds; i++) { ret = full_write(&ctx->open_fds[i], chunk, size); if (ret) { ERROR_WITH_ERRNO("Error writing data to filesystem"); return ret; } } if (ctx->reparse_ptr) ctx->reparse_ptr = mempcpy(ctx->reparse_ptr, chunk, size); return 0; } /* Called when a single-instance stream has been fully read for extraction */ static int unix_end_extract_stream(struct wim_lookup_table_entry *stream, int status, void *_ctx) { struct unix_apply_ctx *ctx = _ctx; int ret; unsigned j; const struct stream_owner *owners = stream_owners(stream); ctx->reparse_ptr = NULL; if (status) { unix_cleanup_open_fds(ctx, 0); return status; } j = 0; ret = 0; for (u32 i = 0; i < stream->out_refcnt; i++) { struct wim_inode *inode = owners[i].inode; if (inode_is_symlink(inode)) { /* We finally have the symlink data, so we can create * the symlink. */ const char *path; path = unix_build_inode_extraction_path(inode, ctx); ret = unix_create_symlink(inode, path, ctx->reparse_data, stream->size, (ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_RPFIX), ctx->target_abspath, ctx->target_abspath_nchars); if (ret) { ERROR_WITH_ERRNO("Can't create symbolic link " "\"%s\"", path); break; } ret = unix_set_metadata(-1, inode, path, ctx); if (ret) break; } else { /* Set metadata on regular file just before closing it. */ struct filedes *fd = &ctx->open_fds[j]; ret = unix_set_metadata(fd->fd, inode, NULL, ctx); if (ret) break; if (filedes_close(fd)) { ERROR_WITH_ERRNO("Error closing \"%s\"", unix_build_inode_extraction_path(inode, ctx)); ret = WIMLIB_ERR_WRITE; break; } j++; } } unix_cleanup_open_fds(ctx, j); return ret; } static int unix_set_dir_metadata(struct list_head *dentry_list, struct unix_apply_ctx *ctx) { const struct wim_dentry *dentry; int ret; list_for_each_entry_reverse(dentry, dentry_list, d_extraction_list_node) { if (dentry_is_directory(dentry)) { ret = unix_set_metadata(-1, dentry->d_inode, NULL, ctx); if (ret) return ret; } } return 0; } static int unix_extract(struct list_head *dentry_list, struct apply_ctx *_ctx) { int ret; struct unix_apply_ctx *ctx = (struct unix_apply_ctx *)_ctx; size_t path_max; /* Compute the maximum path length that will be needed, then allocate * some path buffers. */ path_max = unix_compute_path_max(dentry_list, ctx); for (unsigned i = 0; i < NUM_PATHBUFS; i++) { ctx->pathbufs[i] = MALLOC(path_max); if (!ctx->pathbufs[i]) { ret = WIMLIB_ERR_NOMEM; goto out; } /* Pre-fill the target in each path buffer. We'll just append * the rest of the paths after this. */ memcpy(ctx->pathbufs[i], ctx->common.target, ctx->common.target_nchars); } /* Extract directories and empty regular files. Directories are needed * because we can't extract any other files until their directories * exist. Empty files are needed because they don't have * representatives in the stream list. */ ret = unix_create_dirs_and_empty_files(dentry_list, ctx); if (ret) goto out; /* Get full path to target if needed for absolute symlink fixups. */ if ((ctx->common.extract_flags & WIMLIB_EXTRACT_FLAG_RPFIX) && ctx->common.required_features.symlink_reparse_points) { ctx->target_abspath = realpath(ctx->common.target, NULL); if (!ctx->target_abspath) { ret = WIMLIB_ERR_NOMEM; goto out; } ctx->target_abspath_nchars = strlen(ctx->target_abspath); } /* Extract nonempty regular files and symbolic links. */ struct read_stream_list_callbacks cbs = { .begin_stream = unix_begin_extract_stream, .begin_stream_ctx = ctx, .consume_chunk = unix_extract_chunk, .consume_chunk_ctx = ctx, .end_stream = unix_end_extract_stream, .end_stream_ctx = ctx, }; ret = extract_stream_list(&ctx->common, &cbs); if (ret) goto out; /* Set directory metadata. We do this last so that we get the right * directory timestamps. */ ret = unix_set_dir_metadata(dentry_list, ctx); if (ret) goto out; if (ctx->num_special_files_ignored) { WARNING("%lu special files were not extracted due to EPERM!", ctx->num_special_files_ignored); } out: for (unsigned i = 0; i < NUM_PATHBUFS; i++) FREE(ctx->pathbufs[i]); FREE(ctx->target_abspath); return ret; } const struct apply_operations unix_apply_ops = { .name = "UNIX", .get_supported_features = unix_get_supported_features, .extract = unix_extract, .context_size = sizeof(struct unix_apply_ctx), };