Bug#281530: [s390] Installation successful, some issues
I saw an offer for an s390 on ebay, bought it and set it up in my
hobby room. Uhm, scratch that.
I downloaded hercules 3.01 , compiled and installed it without any
problems on my Toshiba i386 laptop, and had a beautiful s390 emulator
(lots of kudos to the developers!).
 I got the upstream version as the current version in Debian does
not run on 2.6 kernels (#241064)
As I don't know anything about s390, I did a really basic setup and install.
Main goal was to test some changes in network-console and prebaseconfig
related to remote installation over SSH (#279090 and #279213).
(this is the most current version available)
uname -a: Linux (none) 2.4.27-1-s390 #1 SMP Tue Aug 17 12:18:33 CEST 2004 s390 unknown
Date: Sunday Nov 15 (most of the evening)
Method: ipl'ed generic d-i image; set up networking;
set up network-console; continued installation over SSH
Machine: hercules s390 emulator on i386/Sarge/2.6.8 machine
# cat cpuinfo
vendor_id : IBM/S390
# processors : 2
bogomips per cpu: 130.66
processor 0: version = FD, identification = 002623, machine = 3090
processor 1: version = FD, identification = 102623, machine = 3090
(Only one processor during installation; second was added afterwards)
Memory: MemTotal: 61360 kB
Root Device: dasd created with 'dasdinit dasd0 3390-1 root 1114'
Root Size/partition table:
15 tracks per cylinder,
12 blocks per track
4096 bytes per block
volume label: VOL1, volume identifier: LX0120
maximum partition number: 3
------------------------------- tracks -------------------------------
Device start end length Id System
/dev/dasda1 2 16000 15999 1 Linux native
/dev/dasda2 16001 16694 694 2 Linux swap
# cat /etc/fstab
/dev/dasda1 / ext3 defaults,errors=remount-ro 0 1
/dev/dasda2 none swap sw 0 0
/dev/floppy/0 /floppy auto rw,user,noauto 0 0
/dev/cdroms/cdrom0 /cdrom auto ro,user,noauto 0 0
proc /proc proc defaults 0 0
/dev/dasda1 on / type ext3 (rw,errors=remount-ro)
proc on /proc type proc (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
tmpfs on /dev/shm type tmpfs (rw)
Base System Installation Checklist:
[O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it
Initial boot worked: [O]
Configure network HW: [O]
Config network: [O]
Detect CD: [-]
Load installer modules: [O]
Detect hard drives: [O]
Partition hard drives: [E]
Create file systems: [O]
Mount partitions: [O]
Install base system: [O]
Install boot loader: [O]
1) Locale not set up properly
It looks like languagechooser, countrychooser and kbd-chooser are not
run at all (as they are not included in the initrd).
It also looks like no proper defaults for language, country and locale are
set. This causes several problems:
- /etc/environment is set up with LANG="" and LANGUAGE=""
It would be better if LANG was set to "C" and LANGUAGE maybe not set at all.
/etc/environment seems to be created in languagechooser's prebaseconfig.
(/me files a separate bugreport against languagechooser)
- The countrylist for mirror-chooser is shown as country codes and not
I'm not sure what the best place to set general defaults would be.
2) languagechooser does not handles LANG="" correctly
During package installation in tasksel, locales was installed.
During the preconfig of locales, the top of the list of locales to choose from
showed: [*] Usage: /usr/sbin/validlocale <locale>
It appears this line was added to /etc/localegen by languagechooser's
prebaseconfig, by calling 'validlocale $LANG' instead of 'validlocale "$LANG"'.
(Note that I understand that an empty LANG should be treated the same as LANG="C".)
(/me files a separate bugreport against languagechooser for this issue)
3) Selected unstable in 1st stage, but testing was selected in 2nd stage
During apt-config I selected http and the same mirror as in 1st stage.
After that tasksel broke down because of dependency problems. Investigation
showed that a source line for testing had been added in /etc/apt/sources.list.
Not sure where this problem comes from.
4) network-console: SSH session aborts when changing window size on client
Separate bug report filed. See #281435.
All in all the installation was successful (and a lot of fun).