1 |
On Monday 04 Jul 2011 17:15:55 Joshua Murphy wrote: |
2 |
> On Mon, Jul 4, 2011 at 12:00 PM, Mick <michaelkintzios@×××××.com> wrote: |
3 |
> > On Monday 04 Jul 2011 15:48:06 Joshua Murphy wrote: |
4 |
> >> On Mon, Jul 4, 2011 at 9:15 AM, Neil Bothwick <neil@××××××××××.uk> wrote: |
5 |
> >> > On Mon, 4 Jul 2011 12:12:03 +0000, Alan Mackenzie wrote: |
6 |
> >> >> > > o - Do live CDs actually mount filesystems on HDDs? |
7 |
> >> >> > |
8 |
> >> >> > Only when you ask them to. |
9 |
> >> >> |
10 |
> >> >> I'm stupid. Of _course_ a live CD can't mount HDD filesystems at |
11 |
> >> >> boot. To do this it would need /etc/fstab, for which it would need |
12 |
> >> >> to be told the root partition. A live CD doesn't get this. |
13 |
> >> > |
14 |
> >> > A live CD can mount partitions automatically at boot, some do. all it |
15 |
> >> > needs to do is scan the disk partition tables, create the mount points |
16 |
> >> > and mount them. |
17 |
> >> > |
18 |
> >> > Knoppix has been doing the first two for years, and writing the |
19 |
> >> > details to /etc/fstab to allow the user to mount them easily. |
20 |
> >> > |
21 |
> >> > |
22 |
> >> > -- |
23 |
> >> > Neil Bothwick |
24 |
> >> > |
25 |
> >> > A computer without Microsoft is like a chocolate cake without mustard. |
26 |
> >> |
27 |
> >> And to further complicate it, many also use a similar technique for |
28 |
> >> finding themselves, mounting one filesystem after another until they |
29 |
> >> find some distinct marker file to identify where to find the rest of |
30 |
> >> their data. Others auto-mount and poke around for auto-loading of |
31 |
> >> extensions unless such features are disabled by a boot-time option. |
32 |
> > |
33 |
> > I've only come across LiveCDs which scan the drive and create mount |
34 |
> > points - but not mount any device unless explicitly asked to do so by |
35 |
> > the user. |
36 |
> > |
37 |
> > However, I wouldn't be surprised if some more recent installation CDs go |
38 |
> > further than that, as Joshua claims. |
39 |
> > |
40 |
> > Joshua, which LiveCDs behave in the way you describe by automounting |
41 |
> > partitions and searching fs? |
42 |
> > |
43 |
> > -- |
44 |
> > Regards, |
45 |
> > Mick |
46 |
> |
47 |
> I haven't seen any install cds that do that, but DSL and, if I recall, |
48 |
> TinyCore/MicroCore look for extensions in a default path on the local |
49 |
> filesystems. |
50 |
|
51 |
I had to look again at DSL because last time I used it a couple of years ago |
52 |
it definitely did not automount anything - unless ... you had set up a |
53 |
persistent /home or settings directory. In that case it would mount the |
54 |
device in which you saved your settings, but this would be something the user |
55 |
would set up and run consciously at boot time. |
56 |
|
57 |
|
58 |
> One thing I'm fairly sure on, though, is that without the |
59 |
> "-f" flag, mount won't take the risk on an unclean NTFS, and instead |
60 |
> just tosses an "are you sure?" message, which would make me presume |
61 |
> even those livecds that do look for extensions wouldn't risk the |
62 |
> damage there. |
63 |
|
64 |
From what I recall the Linux kernel NTFS driver will mount a unclean NTFS |
65 |
partition regardless (can't recall for sure though), but the ntfs-3g will |
66 |
behave as you describe above. |
67 |
|
68 |
So in answer to the OP questions, the only way I can think that a Linux LiveCD |
69 |
would corrupt a NTFS partition is to mount it with the Linux Kernel driver as |
70 |
rw and then create or edit a file. |
71 |
|
72 |
If this was not the case and fs corruption ensued, then it would be just a |
73 |
coincidence that the drive had some bad blocks and they decided to play up at |
74 |
the time the MSWindows fs was being booted into. |
75 |
-- |
76 |
Regards, |
77 |
Mick |