Struct PermissionCheck

pub struct PermissionCheck<'a> { /* private fields */ }

Implements the has_permission() API, based on supplied Query and AccessVectorComputer implementations.

Implementations

impl<'a> PermissionCheck<'a>

pub fn has_permission<P: ClassPermission + Into<KernelPermission> + Clone + 'static>( &self, source_sid: SecurityId, target_sid: SecurityId, permission: P, ) -> PermissionCheckResult

Returns whether the source_sid has the specified permission on target_sid. The result indicates both whether permission is permitted, and whether the caller should audit log the query.

pub fn has_ioctl_permission<P: ClassPermission + Into<KernelPermission> + Clone + 'static>( &self, source_sid: SecurityId, target_sid: SecurityId, permission: P, ioctl: u16, ) -> PermissionCheckResult

Returns whether the source_sid has both the ioctl permission and the specified extended permission on target_sid, and whether the decision should be audited.

A request is allowed if the ioctl permission is allowed and either the numeric ioctl extended permission is allowxperm, or ioctl extended permissions are not filtered for this domain.

A granted request is audited if the ioctl permission is auditallow and the numeric ioctl extended permission is auditallowxperm.

A denied request is audited if the ioctl permission is dontaudit or the numeric ioctl extended permission is dontauditxperm.

pub fn security_server(&self) -> &SecurityServer

pub fn compute_new_fs_node_sid( &self, source_sid: SecurityId, target_sid: SecurityId, fs_node_class: FsNodeClass, fs_node_name: NullessByteStr<'_>, ) -> Result<SecurityId, Error>

Returns the SID with which to label a new file_class instance created by subject_sid, with target_sid as its parent, taking into account role & type transition rules, and filename-transition rules. If a filename-transition rule matches the fs_node_name then that will be used, otherwise the filename-independent computation will be applied.