| 1 |
Joerg Schilling wrote: |
| 2 |
> Dale <rdalek1967@×××××.com> wrote: |
| 3 |
> |
| 4 |
> |
| 5 |
>>> In Autumn 2004 this bug was not fixed but the SCSI Linux kernel interface |
| 6 |
>>> was changed in an incompatible way. Now _some_ SCSI commands work as non-root. |
| 7 |
>>> |
| 8 |
>>> SCSI is a try and error protocol and cdrecord checks which commands are working. |
| 9 |
>>> If it is possible to burn with the limuted non-root command set, it _may_ work |
| 10 |
>>> but there is a high risk for buffer underruns. In this case cdrecord believes |
| 11 |
>>> that you own a really dumb burner that does not support most of the nice |
| 12 |
>>> features.... |
| 13 |
>>> |
| 14 |
>>> Jörg |
| 15 |
>>> |
| 16 |
> |
| 17 |
> |
| 18 |
>> So that I have a better understanding of this, if he is not a member of |
| 19 |
>> the cd/cdrw group then the command would still work as a user? |
| 20 |
>> |
| 21 |
> |
| 22 |
> Since the original security bug hass not been fxed, you are able so send |
| 23 |
> a limit set of SCSI commands if you are able to open the device read-only. |
| 24 |
> Libscg opens the devices read/write, so the only effect of such a grup |
| 25 |
> membership is that cdrecord is able to open the device. |
| 26 |
> |
| 27 |
> Jörg |
| 28 |
> |
| 29 |
> |
| 30 |
|
| 31 |
Then what is the point of having a cd/cdrw group? It doesn't seem to |
| 32 |
have "secured" much of anything by having it. |
| 33 |
|
| 34 |
Dale aka confused. It's ok. It's normal for me. |
| 35 |
|
| 36 |
:-) :-) |
Archives