|
199 | 199 | \sltitle{Access rights testing} |
200 | 200 | \setlength{\baselineskip}{0.9\baselineskip} |
201 | 201 | \begin{itemize} |
202 | | -\item user is identified with (\emsl{UID}) number and group numbers |
203 | | -for groups he belongs to (\emsl{primary GID}, \emsl{supplementary GIDs}). |
| 202 | +\item user is identified with a \emsl{UID} number and numbers for groups he |
| 203 | +belongs to (\emsl{primary GID}, \emsl{supplementary GIDs}) |
204 | 204 | \item this identification is inherited by each process |
205 | 205 | \item file $F$ has owner ($UID_F$) and group owner ($GID_F$). |
206 | 206 | \item algorithm for evaluation of access rights for process: |
|
230 | 230 | rights do. Similarly the others rights are not used if the group is equal. |
231 | 231 | \emsl{Therefore if a file owned by my user has the rights set to |
232 | 232 | \texttt{---rwxrwx}, I cannot read/write/execute it until I change the rights.} |
233 | | -\item More and more system diverge from the classic model where many processes |
234 | | -were running under a user with UID 0. Security vulnerability in such application |
235 | | -meant total control of the system. To thwart this, these systems employ models |
236 | | -like \emph{least privilege} in Solarisu or \emph{privilege separation} and |
237 | | -\emph{pledge} in OpenBSD. |
| 233 | +\item More and more systems diverge from the classic model where many processes |
| 234 | +were running under a user with UID 0. A security vulnerability in such an |
| 235 | +application meant total control of the system. To thwart this, these systems |
| 236 | +employ models like \emph{least privilege} in Solaris or \emph{privilege |
| 237 | +separation} and \emph{pledge} in OpenBSD. |
238 | 238 | \item \label{FILEDELETE} In order to delete a file, the user has to have a right |
239 | 239 | to write to the \emsl{directory} containing the file, because this is the |
240 | 240 | ``file'', that is being changed. \emsl{The rights of the file to be deleted are |
|
0 commit comments