ladybird/Kernel/FileSystem/VirtualFileSystem.cpp
Ali Mohammad Pur 90de1ded55 Kernel: Ensure that an unveil node with no permission is never accepted
Otherwise nodes inheriting from root may still be accessed with
`access(..., F_OK)`.
Also adds a test case to TestKernelUnveil about this behaviour.
2021-05-29 22:05:34 +02:00

1036 lines
34 KiB
C++

/*
* Copyright (c) 2018-2020, Andreas Kling <kling@serenityos.org>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include <AK/LexicalPath.h>
#include <AK/Singleton.h>
#include <AK/StringBuilder.h>
#include <Kernel/Debug.h>
#include <Kernel/Devices/BlockDevice.h>
#include <Kernel/FileSystem/Custody.h>
#include <Kernel/FileSystem/FileBackedFileSystem.h>
#include <Kernel/FileSystem/FileDescription.h>
#include <Kernel/FileSystem/FileSystem.h>
#include <Kernel/FileSystem/VirtualFileSystem.h>
#include <Kernel/KSyms.h>
#include <Kernel/Process.h>
#include <LibC/errno_numbers.h>
namespace Kernel {
static AK::Singleton<VFS> s_the;
static constexpr int symlink_recursion_limit { 5 }; // FIXME: increase?
static constexpr int root_mount_flags = MS_NODEV | MS_NOSUID | MS_RDONLY;
UNMAP_AFTER_INIT void VFS::initialize()
{
s_the.ensure_instance();
}
VFS& VFS::the()
{
return *s_the;
}
UNMAP_AFTER_INIT VFS::VFS()
{
}
UNMAP_AFTER_INIT VFS::~VFS()
{
}
InodeIdentifier VFS::root_inode_id() const
{
VERIFY(m_root_inode);
return m_root_inode->identifier();
}
KResult VFS::mount(FS& file_system, Custody& mount_point, int flags)
{
Locker locker(m_lock);
auto& inode = mount_point.inode();
dbgln("VFS: Mounting {} at {} (inode: {}) with flags {}",
file_system.class_name(),
mount_point.absolute_path(),
inode.identifier(),
flags);
// FIXME: check that this is not already a mount point
Mount mount { file_system, &mount_point, flags };
m_mounts.append(move(mount));
return KSuccess;
}
KResult VFS::bind_mount(Custody& source, Custody& mount_point, int flags)
{
Locker locker(m_lock);
dbgln("VFS: Bind-mounting {} at {}", source.absolute_path(), mount_point.absolute_path());
// FIXME: check that this is not already a mount point
Mount mount { source.inode(), mount_point, flags };
m_mounts.append(move(mount));
return KSuccess;
}
KResult VFS::remount(Custody& mount_point, int new_flags)
{
Locker locker(m_lock);
dbgln("VFS: Remounting {}", mount_point.absolute_path());
Mount* mount = find_mount_for_guest(mount_point.inode());
if (!mount)
return ENODEV;
mount->set_flags(new_flags);
return KSuccess;
}
KResult VFS::unmount(Inode& guest_inode)
{
Locker locker(m_lock);
dbgln("VFS: unmount called with inode {}", guest_inode.identifier());
for (size_t i = 0; i < m_mounts.size(); ++i) {
auto& mount = m_mounts.at(i);
if (&mount.guest() == &guest_inode) {
if (auto result = mount.guest_fs().prepare_to_unmount(); result.is_error()) {
dbgln("VFS: Failed to unmount!");
return result;
}
dbgln("VFS: found fs {} at mount index {}! Unmounting...", mount.guest_fs().fsid(), i);
m_mounts.unstable_take(i);
return KSuccess;
}
}
dbgln("VFS: Nothing mounted on inode {}", guest_inode.identifier());
return ENODEV;
}
bool VFS::mount_root(FS& file_system)
{
if (m_root_inode) {
dmesgln("VFS: mount_root can't mount another root");
return false;
}
Mount mount { file_system, nullptr, root_mount_flags };
auto root_inode = file_system.root_inode();
if (!root_inode->is_directory()) {
dmesgln("VFS: root inode ({}) for / is not a directory :(", root_inode->identifier());
return false;
}
m_root_inode = move(root_inode);
dmesgln("VFS: mounted root from {} ({})", file_system.class_name(), static_cast<FileBackedFS&>(file_system).file_description().absolute_path());
m_mounts.append(move(mount));
auto custody_or_error = Custody::try_create(nullptr, "", *m_root_inode, root_mount_flags);
if (custody_or_error.is_error())
return false;
m_root_custody = custody_or_error.release_value();
return true;
}
auto VFS::find_mount_for_host(Inode& inode) -> Mount*
{
for (auto& mount : m_mounts) {
if (mount.host() == &inode)
return &mount;
}
return nullptr;
}
auto VFS::find_mount_for_host(InodeIdentifier id) -> Mount*
{
for (auto& mount : m_mounts) {
if (mount.host() && mount.host()->identifier() == id)
return &mount;
}
return nullptr;
}
auto VFS::find_mount_for_guest(Inode& inode) -> Mount*
{
for (auto& mount : m_mounts) {
if (&mount.guest() == &inode)
return &mount;
}
return nullptr;
}
auto VFS::find_mount_for_guest(InodeIdentifier id) -> Mount*
{
for (auto& mount : m_mounts) {
if (mount.guest().identifier() == id)
return &mount;
}
return nullptr;
}
bool VFS::is_vfs_root(InodeIdentifier inode) const
{
return inode == root_inode_id();
}
KResult VFS::traverse_directory_inode(Inode& dir_inode, Function<bool(const FS::DirectoryEntryView&)> callback)
{
return dir_inode.traverse_as_directory([&](auto& entry) {
InodeIdentifier resolved_inode;
if (auto mount = find_mount_for_host(entry.inode))
resolved_inode = mount->guest().identifier();
else
resolved_inode = entry.inode;
// FIXME: This is now broken considering chroot and bind mounts.
bool is_root_inode = dir_inode.identifier() == dir_inode.fs().root_inode()->identifier();
if (is_root_inode && !is_vfs_root(dir_inode.identifier()) && entry.name == "..") {
auto mount = find_mount_for_guest(dir_inode);
VERIFY(mount);
VERIFY(mount->host());
resolved_inode = mount->host()->identifier();
}
callback({ entry.name, resolved_inode, entry.file_type });
return true;
});
}
KResult VFS::utime(StringView path, Custody& base, time_t atime, time_t mtime)
{
auto custody_or_error = VFS::the().resolve_path(move(path), base);
if (custody_or_error.is_error())
return custody_or_error.error();
auto& custody = *custody_or_error.value();
auto& inode = custody.inode();
auto current_process = Process::current();
if (!current_process->is_superuser() && inode.metadata().uid != current_process->euid())
return EACCES;
if (custody.is_readonly())
return EROFS;
if (auto result = inode.set_atime(atime); result.is_error())
return result;
if (auto result = inode.set_mtime(mtime); result.is_error())
return result;
return KSuccess;
}
KResultOr<InodeMetadata> VFS::lookup_metadata(StringView path, Custody& base, int options)
{
auto custody_or_error = resolve_path(path, base, nullptr, options);
if (custody_or_error.is_error())
return custody_or_error.error();
return custody_or_error.value()->inode().metadata();
}
KResultOr<NonnullRefPtr<FileDescription>> VFS::open(StringView path, int options, mode_t mode, Custody& base, Optional<UidAndGid> owner)
{
if ((options & O_CREAT) && (options & O_DIRECTORY))
return EINVAL;
RefPtr<Custody> parent_custody;
auto custody_or_error = resolve_path(path, base, &parent_custody, options);
if (custody_or_error.is_error()) {
// NOTE: ENOENT with a non-null parent custody signals us that the immediate parent
// of the file exists, but the file itself does not.
if ((options & O_CREAT) && custody_or_error.error() == -ENOENT && parent_custody)
return create(path, options, mode, *parent_custody, move(owner));
return custody_or_error.error();
}
if ((options & O_CREAT) && (options & O_EXCL))
return EEXIST;
auto& custody = *custody_or_error.value();
auto& inode = custody.inode();
auto metadata = inode.metadata();
if ((options & O_DIRECTORY) && !metadata.is_directory())
return ENOTDIR;
bool should_truncate_file = false;
auto current_process = Process::current();
if ((options & O_RDONLY) && !metadata.may_read(*current_process))
return EACCES;
if (options & O_WRONLY) {
if (!metadata.may_write(*current_process))
return EACCES;
if (metadata.is_directory())
return EISDIR;
should_truncate_file = options & O_TRUNC;
}
if (options & O_EXEC) {
if (!metadata.may_execute(*current_process) || (custody.mount_flags() & MS_NOEXEC))
return EACCES;
}
if (auto preopen_fd = inode.preopen_fd())
return *preopen_fd;
if (metadata.is_fifo()) {
auto fifo = inode.fifo();
if (options & O_WRONLY) {
auto open_result = fifo->open_direction_blocking(FIFO::Direction::Writer);
if (open_result.is_error())
return open_result.error();
auto& description = open_result.value();
description->set_rw_mode(options);
description->set_file_flags(options);
description->set_original_inode({}, inode);
return description;
} else if (options & O_RDONLY) {
auto open_result = fifo->open_direction_blocking(FIFO::Direction::Reader);
if (open_result.is_error())
return open_result.error();
auto& description = open_result.value();
description->set_rw_mode(options);
description->set_file_flags(options);
description->set_original_inode({}, inode);
return description;
}
return EINVAL;
}
if (metadata.is_device()) {
if (custody.mount_flags() & MS_NODEV)
return EACCES;
auto device = Device::get_device(metadata.major_device, metadata.minor_device);
if (device == nullptr) {
return ENODEV;
}
auto descriptor_or_error = device->open(options);
if (descriptor_or_error.is_error())
return descriptor_or_error.error();
descriptor_or_error.value()->set_original_inode({}, inode);
return descriptor_or_error;
}
// Check for read-only FS. Do this after handling preopen FD and devices,
// but before modifying the inode in any way.
if ((options & O_WRONLY) && custody.is_readonly())
return EROFS;
if (should_truncate_file) {
if (auto result = inode.truncate(0); result.is_error())
return result;
if (auto result = inode.set_mtime(kgettimeofday().to_truncated_seconds()); result.is_error())
return result;
}
auto description = FileDescription::create(custody);
if (!description.is_error()) {
description.value()->set_rw_mode(options);
description.value()->set_file_flags(options);
}
return description;
}
KResult VFS::mknod(StringView path, mode_t mode, dev_t dev, Custody& base)
{
if (!is_regular_file(mode) && !is_block_device(mode) && !is_character_device(mode) && !is_fifo(mode) && !is_socket(mode))
return EINVAL;
RefPtr<Custody> parent_custody;
auto existing_file_or_error = resolve_path(path, base, &parent_custody);
if (!existing_file_or_error.is_error())
return EEXIST;
if (!parent_custody)
return ENOENT;
if (existing_file_or_error.error() != -ENOENT)
return existing_file_or_error.error();
auto& parent_inode = parent_custody->inode();
auto current_process = Process::current();
if (!parent_inode.metadata().may_write(*current_process))
return EACCES;
if (parent_custody->is_readonly())
return EROFS;
LexicalPath p(path);
dbgln("VFS::mknod: '{}' mode={} dev={} in {}", p.basename(), mode, dev, parent_inode.identifier());
return parent_inode.create_child(p.basename(), mode, dev, current_process->euid(), current_process->egid()).result();
}
KResultOr<NonnullRefPtr<FileDescription>> VFS::create(StringView path, int options, mode_t mode, Custody& parent_custody, Optional<UidAndGid> owner)
{
LexicalPath p(path);
if (auto result = validate_path_against_process_veil(String::formatted("{}/{}", parent_custody.absolute_path(), p.basename()), options); result.is_error())
return result;
if (!is_socket(mode) && !is_fifo(mode) && !is_block_device(mode) && !is_character_device(mode)) {
// Turn it into a regular file. (This feels rather hackish.)
mode |= 0100000;
}
auto& parent_inode = parent_custody.inode();
auto current_process = Process::current();
if (!parent_inode.metadata().may_write(*current_process))
return EACCES;
if (parent_custody.is_readonly())
return EROFS;
dbgln_if(VFS_DEBUG, "VFS::create: '{}' in {}", p.basename(), parent_inode.identifier());
uid_t uid = owner.has_value() ? owner.value().uid : current_process->euid();
gid_t gid = owner.has_value() ? owner.value().gid : current_process->egid();
auto inode_or_error = parent_inode.create_child(p.basename(), mode, 0, uid, gid);
if (inode_or_error.is_error())
return inode_or_error.error();
auto new_custody_or_error = Custody::try_create(&parent_custody, p.basename(), inode_or_error.value(), parent_custody.mount_flags());
if (new_custody_or_error.is_error())
return new_custody_or_error.error();
auto description = FileDescription::create(*new_custody_or_error.release_value());
if (!description.is_error()) {
description.value()->set_rw_mode(options);
description.value()->set_file_flags(options);
}
return description;
}
KResult VFS::mkdir(StringView path, mode_t mode, Custody& base)
{
// Unlike in basically every other case, where it's only the last
// path component (the one being created) that is allowed not to
// exist, POSIX allows mkdir'ed path to have trailing slashes.
// Let's handle that case by trimming any trailing slashes.
while (path.length() > 1 && path.ends_with("/"))
path = path.substring_view(0, path.length() - 1);
RefPtr<Custody> parent_custody;
if (auto result = resolve_path(path, base, &parent_custody); !result.is_error())
return EEXIST;
else if (!parent_custody)
return ENOENT;
else if (result.error() != -ENOENT)
return result.error();
auto& parent_inode = parent_custody->inode();
auto current_process = Process::current();
if (!parent_inode.metadata().may_write(*current_process))
return EACCES;
if (parent_custody->is_readonly())
return EROFS;
LexicalPath p(path);
dbgln_if(VFS_DEBUG, "VFS::mkdir: '{}' in {}", p.basename(), parent_inode.identifier());
return parent_inode.create_child(p.basename(), S_IFDIR | mode, 0, current_process->euid(), current_process->egid()).result();
}
KResult VFS::access(StringView path, int mode, Custody& base)
{
auto custody_or_error = resolve_path(path, base);
if (custody_or_error.is_error())
return custody_or_error.error();
auto& custody = *custody_or_error.value();
auto& inode = custody.inode();
auto metadata = inode.metadata();
auto current_process = Process::current();
if (mode & R_OK) {
if (!metadata.may_read(*current_process))
return EACCES;
}
if (mode & W_OK) {
if (!metadata.may_write(*current_process))
return EACCES;
if (custody.is_readonly())
return EROFS;
}
if (mode & X_OK) {
if (!metadata.may_execute(*current_process))
return EACCES;
}
return KSuccess;
}
KResultOr<NonnullRefPtr<Custody>> VFS::open_directory(StringView path, Custody& base)
{
auto inode_or_error = resolve_path(path, base);
if (inode_or_error.is_error())
return inode_or_error.error();
auto& custody = *inode_or_error.value();
auto& inode = custody.inode();
if (!inode.is_directory())
return ENOTDIR;
if (!inode.metadata().may_execute(*Process::current()))
return EACCES;
return custody;
}
KResult VFS::chmod(Custody& custody, mode_t mode)
{
auto& inode = custody.inode();
auto current_process = Process::current();
if (current_process->euid() != inode.metadata().uid && !current_process->is_superuser())
return EPERM;
if (custody.is_readonly())
return EROFS;
// Only change the permission bits.
mode = (inode.mode() & ~07777u) | (mode & 07777u);
return inode.chmod(mode);
}
KResult VFS::chmod(StringView path, mode_t mode, Custody& base)
{
auto custody_or_error = resolve_path(path, base);
if (custody_or_error.is_error())
return custody_or_error.error();
auto& custody = *custody_or_error.value();
return chmod(custody, mode);
}
KResult VFS::rename(StringView old_path, StringView new_path, Custody& base)
{
RefPtr<Custody> old_parent_custody;
auto old_custody_or_error = resolve_path(old_path, base, &old_parent_custody, O_NOFOLLOW_NOERROR);
if (old_custody_or_error.is_error())
return old_custody_or_error.error();
auto& old_custody = *old_custody_or_error.value();
auto& old_inode = old_custody.inode();
RefPtr<Custody> new_parent_custody;
auto new_custody_or_error = resolve_path(new_path, base, &new_parent_custody);
if (new_custody_or_error.is_error()) {
if (new_custody_or_error.error() != -ENOENT || !new_parent_custody)
return new_custody_or_error.error();
}
if (!old_parent_custody || !new_parent_custody) {
return EPERM;
}
auto& old_parent_inode = old_parent_custody->inode();
auto& new_parent_inode = new_parent_custody->inode();
if (&old_parent_inode.fs() != &new_parent_inode.fs())
return EXDEV;
for (auto* new_ancestor = new_parent_custody.ptr(); new_ancestor; new_ancestor = new_ancestor->parent()) {
if (&old_inode == &new_ancestor->inode())
return EDIRINTOSELF;
}
auto current_process = Process::current();
if (!new_parent_inode.metadata().may_write(*current_process))
return EACCES;
if (!old_parent_inode.metadata().may_write(*current_process))
return EACCES;
if (old_parent_inode.metadata().is_sticky()) {
if (!current_process->is_superuser() && old_inode.metadata().uid != current_process->euid())
return EACCES;
}
if (old_parent_custody->is_readonly() || new_parent_custody->is_readonly())
return EROFS;
auto new_basename = LexicalPath(new_path).basename();
if (!new_custody_or_error.is_error()) {
auto& new_custody = *new_custody_or_error.value();
auto& new_inode = new_custody.inode();
// FIXME: Is this really correct? Check what other systems do.
if (&new_inode == &old_inode)
return KSuccess;
if (new_parent_inode.metadata().is_sticky()) {
if (!current_process->is_superuser() && new_inode.metadata().uid != current_process->euid())
return EACCES;
}
if (new_inode.is_directory() && !old_inode.is_directory())
return EISDIR;
if (auto result = new_parent_inode.remove_child(new_basename); result.is_error())
return result;
}
if (auto result = new_parent_inode.add_child(old_inode, new_basename, old_inode.mode()); result.is_error())
return result;
if (auto result = old_parent_inode.remove_child(LexicalPath(old_path).basename()); result.is_error())
return result;
return KSuccess;
}
KResult VFS::chown(Custody& custody, uid_t a_uid, gid_t a_gid)
{
auto& inode = custody.inode();
auto metadata = inode.metadata();
auto current_process = Process::current();
if (current_process->euid() != metadata.uid && !current_process->is_superuser())
return EPERM;
uid_t new_uid = metadata.uid;
gid_t new_gid = metadata.gid;
if (a_uid != (uid_t)-1) {
if (current_process->euid() != a_uid && !current_process->is_superuser())
return EPERM;
new_uid = a_uid;
}
if (a_gid != (gid_t)-1) {
if (!current_process->in_group(a_gid) && !current_process->is_superuser())
return EPERM;
new_gid = a_gid;
}
if (custody.is_readonly())
return EROFS;
dbgln_if(VFS_DEBUG, "VFS::chown(): inode {} <- uid={} gid={}", inode.identifier(), new_uid, new_gid);
if (metadata.is_setuid() || metadata.is_setgid()) {
dbgln_if(VFS_DEBUG, "VFS::chown(): Stripping SUID/SGID bits from {}", inode.identifier());
if (auto result = inode.chmod(metadata.mode & ~(04000 | 02000)); result.is_error())
return result;
}
return inode.chown(new_uid, new_gid);
}
KResult VFS::chown(StringView path, uid_t a_uid, gid_t a_gid, Custody& base)
{
auto custody_or_error = resolve_path(path, base);
if (custody_or_error.is_error())
return custody_or_error.error();
auto& custody = *custody_or_error.value();
return chown(custody, a_uid, a_gid);
}
static bool hard_link_allowed(const Inode& inode)
{
auto metadata = inode.metadata();
if (Process::current()->euid() == metadata.uid)
return true;
if (metadata.is_regular_file()
&& !metadata.is_setuid()
&& !(metadata.is_setgid() && metadata.mode & S_IXGRP)
&& metadata.may_write(*Process::current())) {
return true;
}
return false;
}
KResult VFS::link(StringView old_path, StringView new_path, Custody& base)
{
auto old_custody_or_error = resolve_path(old_path, base);
if (old_custody_or_error.is_error())
return old_custody_or_error.error();
auto& old_custody = *old_custody_or_error.value();
auto& old_inode = old_custody.inode();
RefPtr<Custody> parent_custody;
auto new_custody_or_error = resolve_path(new_path, base, &parent_custody);
if (!new_custody_or_error.is_error())
return EEXIST;
if (!parent_custody)
return ENOENT;
auto& parent_inode = parent_custody->inode();
if (parent_inode.fsid() != old_inode.fsid())
return EXDEV;
if (!parent_inode.metadata().may_write(*Process::current()))
return EACCES;
if (old_inode.is_directory())
return EPERM;
if (parent_custody->is_readonly())
return EROFS;
if (!hard_link_allowed(old_inode))
return EPERM;
return parent_inode.add_child(old_inode, LexicalPath(new_path).basename(), old_inode.mode());
}
KResult VFS::unlink(StringView path, Custody& base)
{
RefPtr<Custody> parent_custody;
auto custody_or_error = resolve_path(path, base, &parent_custody, O_NOFOLLOW_NOERROR | O_UNLINK_INTERNAL);
if (custody_or_error.is_error())
return custody_or_error.error();
auto& custody = *custody_or_error.value();
auto& inode = custody.inode();
if (inode.is_directory())
return EISDIR;
// We have just checked that the inode is not a directory, and thus it's not
// the root. So it should have a parent. Note that this would be invalidated
// if we were to support bind-mounting regular files on top of the root.
VERIFY(parent_custody);
auto& parent_inode = parent_custody->inode();
auto current_process = Process::current();
if (!parent_inode.metadata().may_write(*current_process))
return EACCES;
if (parent_inode.metadata().is_sticky()) {
if (!current_process->is_superuser() && inode.metadata().uid != current_process->euid())
return EACCES;
}
if (parent_custody->is_readonly())
return EROFS;
if (auto result = parent_inode.remove_child(LexicalPath(path).basename()); result.is_error())
return result;
return KSuccess;
}
KResult VFS::symlink(StringView target, StringView linkpath, Custody& base)
{
RefPtr<Custody> parent_custody;
auto existing_custody_or_error = resolve_path(linkpath, base, &parent_custody);
if (!existing_custody_or_error.is_error())
return EEXIST;
if (!parent_custody)
return ENOENT;
if (existing_custody_or_error.error() != -ENOENT)
return existing_custody_or_error.error();
auto& parent_inode = parent_custody->inode();
auto current_process = Process::current();
if (!parent_inode.metadata().may_write(*current_process))
return EACCES;
if (parent_custody->is_readonly())
return EROFS;
LexicalPath p(linkpath);
dbgln_if(VFS_DEBUG, "VFS::symlink: '{}' (-> '{}') in {}", p.basename(), target, parent_inode.identifier());
auto inode_or_error = parent_inode.create_child(p.basename(), S_IFLNK | 0644, 0, current_process->euid(), current_process->egid());
if (inode_or_error.is_error())
return inode_or_error.error();
auto& inode = inode_or_error.value();
auto target_buffer = UserOrKernelBuffer::for_kernel_buffer(const_cast<u8*>((const u8*)target.characters_without_null_termination()));
auto result = inode->write_bytes(0, target.length(), target_buffer, nullptr);
if (result.is_error())
return result.error();
return KSuccess;
}
KResult VFS::rmdir(StringView path, Custody& base)
{
RefPtr<Custody> parent_custody;
auto custody_or_error = resolve_path(path, base, &parent_custody);
if (custody_or_error.is_error())
return KResult(custody_or_error.error());
auto& custody = *custody_or_error.value();
auto& inode = custody.inode();
// FIXME: We should return EINVAL if the last component of the path is "."
// FIXME: We should return ENOTEMPTY if the last component of the path is ".."
if (!inode.is_directory())
return ENOTDIR;
if (!parent_custody)
return EBUSY;
auto& parent_inode = parent_custody->inode();
auto parent_metadata = parent_inode.metadata();
if (!parent_metadata.may_write(*Process::current()))
return EACCES;
if (parent_metadata.is_sticky()) {
if (!Process::current()->is_superuser() && inode.metadata().uid != Process::current()->euid())
return EACCES;
}
KResultOr<size_t> dir_count_result = inode.directory_entry_count();
if (dir_count_result.is_error())
return dir_count_result.result();
if (dir_count_result.value() != 2)
return ENOTEMPTY;
if (custody.is_readonly())
return EROFS;
if (auto result = inode.remove_child("."); result.is_error())
return result;
if (auto result = inode.remove_child(".."); result.is_error())
return result;
return parent_inode.remove_child(LexicalPath(path).basename());
}
VFS::Mount::Mount(FS& guest_fs, Custody* host_custody, int flags)
: m_guest(guest_fs.root_inode())
, m_guest_fs(guest_fs)
, m_host_custody(host_custody)
, m_flags(flags)
{
}
VFS::Mount::Mount(Inode& source, Custody& host_custody, int flags)
: m_guest(source)
, m_guest_fs(source.fs())
, m_host_custody(host_custody)
, m_flags(flags)
{
}
String VFS::Mount::absolute_path() const
{
if (!m_host_custody)
return "/";
return m_host_custody->absolute_path();
}
Inode* VFS::Mount::host()
{
if (!m_host_custody)
return nullptr;
return &m_host_custody->inode();
}
const Inode* VFS::Mount::host() const
{
if (!m_host_custody)
return nullptr;
return &m_host_custody->inode();
}
void VFS::for_each_mount(Function<void(const Mount&)> callback) const
{
for (auto& mount : m_mounts) {
callback(mount);
}
}
void VFS::sync()
{
FS::sync();
}
Custody& VFS::root_custody()
{
return *m_root_custody;
}
const UnveilNode* VFS::find_matching_unveiled_path(StringView path)
{
auto& unveil_root = Process::current()->unveiled_paths();
if (unveil_root.is_empty())
return nullptr;
LexicalPath lexical_path { path };
auto& path_parts = lexical_path.parts();
auto& last_matching_node = unveil_root.traverse_until_last_accessible_node(path_parts.begin(), path_parts.end());
return &last_matching_node;
}
KResult VFS::validate_path_against_process_veil(StringView path, int options)
{
if (Process::current()->veil_state() == VeilState::None)
return KSuccess;
if (path == "/usr/lib/Loader.so")
return KSuccess;
// FIXME: Figure out a nicer way to do this.
if (String(path).contains("/.."))
return EINVAL;
auto* unveiled_path = find_matching_unveiled_path(path);
if (!unveiled_path || unveiled_path->permissions() == UnveilAccess::None) {
dbgln("Rejecting path '{}' since it hasn't been unveiled.", path);
dump_backtrace();
return ENOENT;
}
if (options & O_CREAT) {
if (!(unveiled_path->permissions() & UnveilAccess::CreateOrRemove)) {
dbgln("Rejecting path '{}' since it hasn't been unveiled with 'c' permission.", path);
dump_backtrace();
return EACCES;
}
}
if (options & O_UNLINK_INTERNAL) {
if (!(unveiled_path->permissions() & UnveilAccess::CreateOrRemove)) {
dbgln("Rejecting path '{}' for unlink since it hasn't been unveiled with 'c' permission.", path);
dump_backtrace();
return EACCES;
}
return KSuccess;
}
if (options & O_RDONLY) {
if (options & O_DIRECTORY) {
if (!(unveiled_path->permissions() & (UnveilAccess::Read | UnveilAccess::Browse))) {
dbgln("Rejecting path '{}' since it hasn't been unveiled with 'r' or 'b' permissions.", path);
dump_backtrace();
return EACCES;
}
} else {
if (!(unveiled_path->permissions() & UnveilAccess::Read)) {
dbgln("Rejecting path '{}' since it hasn't been unveiled with 'r' permission.", path);
dump_backtrace();
return EACCES;
}
}
}
if (options & O_WRONLY) {
if (!(unveiled_path->permissions() & UnveilAccess::Write)) {
dbgln("Rejecting path '{}' since it hasn't been unveiled with 'w' permission.", path);
dump_backtrace();
return EACCES;
}
}
if (options & O_EXEC) {
if (!(unveiled_path->permissions() & UnveilAccess::Execute)) {
dbgln("Rejecting path '{}' since it hasn't been unveiled with 'x' permission.", path);
dump_backtrace();
return EACCES;
}
}
return KSuccess;
}
KResultOr<NonnullRefPtr<Custody>> VFS::resolve_path(StringView path, Custody& base, RefPtr<Custody>* out_parent, int options, int symlink_recursion_level)
{
auto custody_or_error = resolve_path_without_veil(path, base, out_parent, options, symlink_recursion_level);
if (custody_or_error.is_error())
return custody_or_error.error();
auto& custody = custody_or_error.value();
if (auto result = validate_path_against_process_veil(custody->absolute_path(), options); result.is_error())
return result;
return custody;
}
static bool safe_to_follow_symlink(const Inode& inode, const InodeMetadata& parent_metadata)
{
auto metadata = inode.metadata();
if (Process::current()->euid() == metadata.uid)
return true;
if (!(parent_metadata.is_sticky() && parent_metadata.mode & S_IWOTH))
return true;
if (metadata.uid == parent_metadata.uid)
return true;
return false;
}
KResultOr<NonnullRefPtr<Custody>> VFS::resolve_path_without_veil(StringView path, Custody& base, RefPtr<Custody>* out_parent, int options, int symlink_recursion_level)
{
if (symlink_recursion_level >= symlink_recursion_limit)
return ELOOP;
if (path.is_empty())
return EINVAL;
GenericLexer path_lexer(path);
auto current_process = Process::current();
auto& current_root = current_process->root_directory();
NonnullRefPtr<Custody> custody = path[0] == '/' ? current_root : base;
bool extra_iteration = path[path.length() - 1] == '/';
while (!path_lexer.is_eof() || extra_iteration) {
if (path_lexer.is_eof())
extra_iteration = false;
auto part = path_lexer.consume_until('/');
path_lexer.consume_specific('/');
Custody& parent = custody;
auto parent_metadata = parent.inode().metadata();
if (!parent_metadata.is_directory())
return ENOTDIR;
// Ensure the current user is allowed to resolve paths inside this directory.
if (!parent_metadata.may_execute(*current_process))
return EACCES;
bool have_more_parts = !path_lexer.is_eof() || extra_iteration;
if (part == "..") {
// If we encounter a "..", take a step back, but don't go beyond the root.
if (custody->parent())
custody = *custody->parent();
continue;
} else if (part == "." || part.is_empty()) {
continue;
}
// Okay, let's look up this part.
auto child_inode = parent.inode().lookup(part);
if (!child_inode) {
if (out_parent) {
// ENOENT with a non-null parent custody signals to caller that
// we found the immediate parent of the file, but the file itself
// does not exist yet.
*out_parent = have_more_parts ? nullptr : &parent;
}
return ENOENT;
}
int mount_flags_for_child = parent.mount_flags();
// See if there's something mounted on the child; in that case
// we would need to return the guest inode, not the host inode.
if (auto mount = find_mount_for_host(*child_inode)) {
child_inode = mount->guest();
mount_flags_for_child = mount->flags();
}
auto new_custody_or_error = Custody::try_create(&parent, part, *child_inode, mount_flags_for_child);
if (new_custody_or_error.is_error())
return new_custody_or_error.error();
custody = new_custody_or_error.release_value();
if (child_inode->metadata().is_symlink()) {
if (!have_more_parts) {
if (options & O_NOFOLLOW)
return ELOOP;
if (options & O_NOFOLLOW_NOERROR)
break;
}
if (!safe_to_follow_symlink(*child_inode, parent_metadata))
return EACCES;
if (auto result = validate_path_against_process_veil(custody->absolute_path(), options); result.is_error())
return result;
auto symlink_target = child_inode->resolve_as_link(parent, out_parent, options, symlink_recursion_level + 1);
if (symlink_target.is_error() || !have_more_parts)
return symlink_target;
// Now, resolve the remaining path relative to the symlink target.
// We prepend a "." to it to ensure that it's not empty and that
// any initial slashes it might have get interpreted properly.
StringBuilder remaining_path;
remaining_path.append('.');
remaining_path.append(path.substring_view_starting_after_substring(part));
return resolve_path_without_veil(remaining_path.to_string(), *symlink_target.value(), out_parent, options, symlink_recursion_level + 1);
}
}
if (out_parent)
*out_parent = custody->parent();
return custody;
}
}