Knoppix V7.4.1

[Werner P. Schulz]

The script posted in the mailing list is equal to this within V7.4.1 (checked by md5sum).

[utu]

.
This new Distrowatch paragraph gives slightly more info about 7.4.1,
indluding a new, undefined, mkimage boot cheatcode:

http://distrowatch.com/?newsid=08603

[henk]

.
This new Distrowatch paragraph gives slightly more info about 7.4.1,
indluding a new, undefined, mkimage boot cheatcode:

http://distrowatch.com/?newsid=08603

http://www.knopper.net/knoppix/knoppix741-en.html

Interesting. I'd like to try the "knoppix mkimage" option. I wonder if I can just extract the .iso and copy over the KNOPPIX file to update the distribution or do I have to remaster the whole thing from scratch? I am using a usb install.

[utu]

a new, undefined, mkimage boot cheatcode

My error in calling this news in post #12. I didn't pick up on 'mkimage'
until I went back to re-read the earlier 'release notes' and found the term
there. The ten-minute forum limit on editing posted material preserved my
error for posterity. My bad.

I can understand using something called 'mkimage' to transform a read/write
persistence file to a compact read/only cloop file. I do that myself;
but, I don't see how this would be useful as a (boot) cheatcode.

I hope mkimage is a bash script, and if so, I hope someone will post it here
so we can study it. IMO, I think it would be even better if our fearless leader, KK,
would document some of his inventions a little (or a lot) better.

[Werner P. Schulz]

I hope mkimage is a bash script, and if so, I hope someone will post it here
so we can study it.

mountdata(){
 local img base ext m mods found=""
 for base in $home /mnt-system/"$knoppix_dir"/knoppix-data /mnt-system/knoppix; do
  for ext in aes img inf; do
   img="$base.$ext"
   [ -r "$img" ]        || continue
   [ -b /dev/loop0 ]    || mknod -m 755 /dev/loop0 b 7 0
   [ -d /KNOPPIX-DATA ] || mkdir -m 755 /KNOPPIX-DATA
   [ -d /dev/mapper ]   || mkdir -m 755 /dev/mapper
   found="true"
   case "$ext" in
    aes)
     mountaes "$img" /KNOPPIX-DATA ext2 && return 0
    ;;
    img)
     checkfs "$img" ext2
     message -e "\r${CRE}${GREEN}${USING} ${YELLOW}$img${NORMAL}"
     mount -t ext2 -o loop,rw "$img" /KNOPPIX-DATA && return 0
     losetup -d /dev/loop0 >/dev/null 2>&1
    ;;
    inf) # Contains partition information in fstab format
     local part="" mp="" fs="" opts=""
     while read part mp fs opts; do
      [ -n "$part" ] || continue
      case "$mp" in *[Hh][Oo][Mm][Ee]) mp="/KNOPPIX-DATA" ;; esac
      case "$opts" in
       *aes*) mountaes "${ROOTDEV%[0-9]*}${part#/dev/}" "${mp:-/KNOPPIX-DATA}" "${fs:-ext2}" && return 0 ;;
       *)     checkfs "${ROOTDEV%[0-9]*}${part#/dev/}" "$fs"
              message -e "\r${CRE}${GREEN}${USING} ${YELLOW}${ROOTDEV%[0-9]*}${part#/dev/}${NORMAL}"
              mount -t "${fs:-ext2}" "${ROOTDEV%[0-9]*}${part#/dev/}" "${mp:-/KNOPPIX-DATA}" && return 0 ;;
      esac
     done <"$img"
    ;;
   esac
  done
 done
 if [ -z "$found" -a "$RW" = "rw" ]; then
  if checkbootparam "mkimage"; then
   createdata && return 0
  fi
 fi
 [ -d /KNOPPIX-DATA ] && rmdir /KNOPPIX-DATA
 return 1
}

If you compare the new part of the init script with this old part (up to V7.4.0)

mountdata(){
......
done
 [ -z "$found" -a "$RW" = "rw" ] && createdata && return 0
 [ -d /KNOPPIX-DATA ] && rmdir /KNOPPIX-DATA
 return 1
}

you'll find only a little difference. Have a look at "Hint: 10"

[utu]

.
Thanks, Werner for posting this 7.4.1 material.
A number of comments; I'd appreciate your critique on these.

1. Providing for encryption does complicate things.

2. I was surprised that mkisofs does not appear as an element of
of the mkimage process. I see only cloops and mount processes.
This may be a semantic distinction, since I know Klaus K uses
mkisofs in handling overlays in some of his own applications.

3. I think I have learned that, all along, knoppix-data.img must have
been a read/write cloop condensation of /KNOPPIX-DATA at shutdown, taking
two seconds or less to save or update on the USB. Correct me if I'm
wrong on this. One soon learns that a corrupt, saved, read/write
knoppix-data.img is really bad news.

4. It would appear that the mkimage cheatcode converts current
/KNOPPIX-DATA to a read/only saved cloop, and destroys the current
/KNOPPIX_DATA directory. Presumably the option to re-create a new
'empty' /KNOPPIX data directory will occur again at boot until one
is actually requested.

5. It's not clear if the mkimage option can be self-imposed on the
LiveUSB used as the KNOPPIX system du jour or only on another USB,
not on itself.
I worry that if the chosen /KNOPPIX-DATA to be converted is corrupt
that it may, in turn, corrupt the LiveUSB used as the KNOPPIX du jour
if option 5 is viable.

6. A corrupt knoppix-data.img will certainly poison any USB that receives it.
And, I'm not aware there is any test for whether /KNOPPIX-DATA has
become corrupt other than the unpleasant surprise which awaits on reboot.

[Werner P. Schulz]

Hello!

I fear there is a little misunderstanding. I've posted only the function "mountdata () {...}" of the new init of V7.4.1; it's the only thing, which has been changed against the version of V7.4.0.

I've inserteda the new hint within Knoppix Wiki:
(10) If Knoppix isn't started in forensic mode and if no persistent image or partition is found, then Knoppix DVD version 7.4.1 and upward only offers the option to create persistent image or partition by this new cheatcode.

In earlier versions you has been asked for creation of persistent memory, if Knoppix couldn't found it in the startup; now you have to use the new cheatcode.

To examine the content of the init-file you can use this script within your Knoppix V7.4.0 (make it executable and use as root)

#! /bin/bash
if [ "$(id -u)" != 0 ]; then
  echo " You must be root to use this script."
  exit 1
fi
mkdir /home/knoppix/tmp
cd /home/knoppix/tmp
cp /mnt-system/boot/isolinux/minirt.gz /home/knoppix/tmp
zcat minirt.gz | cpio -imd --no-absolute-filenames

[utu]

.
Greetings, Werner.

I have in my root 7.4.0 filesystem a file called init.
Is this the same or different than what I'd get by using
your script. Thanks.

[Werner P. Schulz]

Yes, it is the same.

[utu]

.
My geany seg faults on 7.4.1 LiveCD and LiveUSB. With LiveUSB, syslog error reads:
Sep 20 19:38:37 Microknoppix kernel: [ 493.612638] geany[3278]: segfault at c6 ip 000000004c1a9b18 sp 00000000ffaf7640 error 4 in libglib-2.0.so.0.4000.0[4c158000+120000]
Failure using menu is silent. With command line, geany <enter> reports 'Segmentation fault'.

Both LiveCD & LiveUSB passed testdvd.
Problem persists with LiveUSB even with Synaptic reinstall.
My geany in a 7.4.0 LiveUSB works ok.

7.4.1 LiveUSB was made with 'retain personal data' (first option) using
what was another 7.4.0 LiveUSB.
7.4.1 LiveUSB preparation from LiveCD appears to have kept original partitioning
parameters in making the new LiveUSB.
7.4.1 LiveUSB otherwise apparently behaving quite well and predictably.

Nothing unusual about my 7.4.1 configuration not also in previous 7.4.0 configuration.

Is anyone else having this problem?