/* * Copyright (c) 2018-2021, Andreas Kling <kling@serenityos.org> * * SPDX-License-Identifier: BSD-2-Clause */ #pragma once #include <AK/Badge.h> #include <AK/Error.h> #include <AK/Function.h> #include <AK/HashMap.h> #include <AK/OwnPtr.h> #include <AK/RefPtr.h> #include <Kernel/FileSystem/FileBackedFileSystem.h> #include <Kernel/FileSystem/FileSystem.h> #include <Kernel/FileSystem/Initializer.h> #include <Kernel/FileSystem/InodeIdentifier.h> #include <Kernel/FileSystem/InodeMetadata.h> #include <Kernel/FileSystem/Mount.h> #include <Kernel/FileSystem/MountFile.h> #include <Kernel/FileSystem/UnveilNode.h> #include <Kernel/Forward.h> #include <Kernel/Locking/MutexProtected.h> #include <Kernel/Locking/SpinlockProtected.h> namespace Kernel { // Kernel internal options. #define O_NOFOLLOW_NOERROR (1 << 29) #define O_UNLINK_INTERNAL (1 << 30) struct UidAndGid { UserID uid; GroupID gid; }; enum class AccessFlags { None = 0, EffectiveAccess = 1 << 0, DoNotFollowSymlinks = 1 << 1, }; AK_ENUM_BITWISE_OPERATORS(AccessFlags); class VirtualFileSystem { public: // Required to be at least 8 by POSIX // https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/limits.h.html static constexpr int symlink_recursion_limit = 8; static void initialize(); static VirtualFileSystem& the(); static ErrorOr<FileSystemInitializer const*> find_filesystem_type_initializer(StringView fs_type); VirtualFileSystem(); ~VirtualFileSystem(); ErrorOr<void> mount_root(FileSystem&); ErrorOr<void> mount(MountFile&, OpenFileDescription*, Custody& mount_point, int flags); ErrorOr<void> bind_mount(Custody& source, Custody& mount_point, int flags); ErrorOr<void> remount(Custody& mount_point, int new_flags); ErrorOr<void> unmount(Custody& mount_point); ErrorOr<void> unmount(Inode& guest_inode, StringView custody_path); ErrorOr<NonnullRefPtr<OpenFileDescription>> open(Credentials const&, StringView path, int options, mode_t mode, Custody& base, Optional<UidAndGid> = {}); ErrorOr<NonnullRefPtr<OpenFileDescription>> open(Process const&, Credentials const&, StringView path, int options, mode_t mode, Custody& base, Optional<UidAndGid> = {}); ErrorOr<NonnullRefPtr<OpenFileDescription>> create(Credentials const&, StringView path, int options, mode_t mode, Custody& parent_custody, Optional<UidAndGid> = {}); ErrorOr<NonnullRefPtr<OpenFileDescription>> create(Process const&, Credentials const&, StringView path, int options, mode_t mode, Custody& parent_custody, Optional<UidAndGid> = {}); ErrorOr<void> mkdir(Credentials const&, StringView path, mode_t mode, Custody& base); ErrorOr<void> link(Credentials const&, StringView old_path, StringView new_path, Custody& base); ErrorOr<void> unlink(Credentials const&, StringView path, Custody& base); ErrorOr<void> symlink(Credentials const&, StringView target, StringView linkpath, Custody& base); ErrorOr<void> rmdir(Credentials const&, StringView path, Custody& base); ErrorOr<void> chmod(Credentials const&, StringView path, mode_t, Custody& base, int options = 0); ErrorOr<void> chmod(Credentials const&, Custody&, mode_t); ErrorOr<void> chown(Credentials const&, StringView path, UserID, GroupID, Custody& base, int options); ErrorOr<void> chown(Credentials const&, Custody&, UserID, GroupID); ErrorOr<void> access(Credentials const&, StringView path, int mode, Custody& base, AccessFlags); ErrorOr<InodeMetadata> lookup_metadata(Credentials const&, StringView path, Custody& base, int options = 0); ErrorOr<void> utime(Credentials const&, StringView path, Custody& base, time_t atime, time_t mtime); ErrorOr<void> utimensat(Credentials const&, StringView path, Custody& base, timespec const& atime, timespec const& mtime, int options = 0); ErrorOr<void> do_utimens(Credentials const& credentials, Custody& custody, timespec const& atime, timespec const& mtime); ErrorOr<void> rename(Credentials const&, Custody& old_base, StringView oldpath, Custody& new_base, StringView newpath); ErrorOr<void> mknod(Credentials const&, StringView path, mode_t, dev_t, Custody& base); ErrorOr<NonnullRefPtr<Custody>> open_directory(Credentials const&, StringView path, Custody& base); ErrorOr<void> for_each_mount(Function<ErrorOr<void>(Mount const&)>) const; InodeIdentifier root_inode_id() const; void sync_filesystems(); void lock_all_filesystems(); static void sync(); NonnullRefPtr<Custody> root_custody(); ErrorOr<NonnullRefPtr<Custody>> resolve_path(Credentials const&, StringView path, NonnullRefPtr<Custody> base, RefPtr<Custody>* out_parent = nullptr, int options = 0, int symlink_recursion_level = 0); ErrorOr<NonnullRefPtr<Custody>> resolve_path(Process const&, Credentials const&, StringView path, NonnullRefPtr<Custody> base, RefPtr<Custody>* out_parent = nullptr, int options = 0, int symlink_recursion_level = 0); ErrorOr<NonnullRefPtr<Custody>> resolve_path_without_veil(Credentials const&, StringView path, NonnullRefPtr<Custody> base, RefPtr<Custody>* out_parent = nullptr, int options = 0, int symlink_recursion_level = 0); private: friend class OpenFileDescription; UnveilNode const& find_matching_unveiled_path(Process const&, StringView path); ErrorOr<void> validate_path_against_process_veil(Process const&, StringView path, int options); ErrorOr<void> validate_path_against_process_veil(Process const& process, Custody const& custody, int options); ErrorOr<void> validate_path_against_process_veil(Custody const& path, int options); ErrorOr<void> validate_path_against_process_veil(StringView path, int options); ErrorOr<void> add_file_system_to_mount_table(FileSystem& file_system, Custody& mount_point, int flags); bool is_vfs_root(InodeIdentifier) const; ErrorOr<void> traverse_directory_inode(Inode&, Function<ErrorOr<void>(FileSystem::DirectoryEntryView const&)>); bool mount_point_exists_at_inode(InodeIdentifier inode); // FIXME: These functions are totally unsafe as someone could unmount the returned Mount underneath us. Mount* find_mount_for_host(InodeIdentifier); Mount* find_mount_for_guest(InodeIdentifier); RefPtr<Inode> m_root_inode; SpinlockProtected<RefPtr<Custody>, LockRank::None> m_root_custody {}; SpinlockProtected<IntrusiveList<&Mount::m_vfs_list_node>, LockRank::None> m_mounts {}; // NOTE: The FileBackedFileSystem list is protected by a mutex because we need to scan it // to search for existing filesystems for already used block devices and therefore when doing // that we could fail to find a filesystem so we need to create a new filesystem which might // need to do disk access (i.e. taking Mutexes in other places) and then register that new filesystem // in this list, to avoid TOCTOU bugs. MutexProtected<IntrusiveList<&FileBackedFileSystem::m_file_backed_file_system_node>> m_file_backed_file_systems_list {}; SpinlockProtected<IntrusiveList<&FileSystem::m_file_system_node>, LockRank::FileSystem> m_file_systems_list {}; }; }