Xubuntu 12.04.1 on FSC-Futro S400 Thinclient locally

Bug #1101684 reported by Andreas Glaeser
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
installation-report (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

FSC Futro thinclients can be used as nettops, because high speed CompactFlash cards have become reasonably cheap today. Operatingsystems can be installed locally onto ConpactFlash-memory. The card used provides about the same read-speed as gigabit ethernet would.
It was a chronical problem with this box, that intermittent boot-problems occured, either when booted from cold condition or at warm-reboot. The system hung somewhere near IRQ- and ACPI-initialisation. First I thought this might be due to incorrect BIOS-settings, replaced the RAM, now I think, that there are powersupply-incompatibilities between different models. There are two different power-supply models available and incompatibilities may occurbetween systems, please beware.
I chose the low-latency kernel variant for this box, i.e. installation was only possible in expert-mode (from USB), during the process I selected the second kernel-version offered (generic) and when it was done, it was possible to install a low-latency kernel and remove the other. With debian I experienced earlier, similar boot-problems with pae-kernels. So it is very positive, that Ubuntu offers a wide variety of kernels with and without PAE, even a realtime-version. There is not so much need to build a kernel-package yourself, that matches the system.
512 MB of RAM are sufficient for Xubuntu for nettop-use. Hopefully it is possible to upgrade to the next LTS-version 2014 without having to upgrade the RAM-amount. The energy-consumption is very low, at a maximum of 21 Watts, so it can also be used as a home-server, running most of the time, or as a client for testing network-setups.
256 MB of RAM are insufficient for desktop-use, because even when using light webbrowsers like midori and 'webbroser' epiphany, it is not possible to open more than two or three tabs at the same times without making the system using swap space, slowing it down to non-usability, also without swapping for light applications it is slower, because there is lack of disk-cache, so even lighter desktop-environments would be recommended.
In my opinion this privides avery good price-performance ratio, the boxes are available in abundance in used or refurbished condition and quite affordable.
I changed /etc/default/grub and attach this file.

ProblemType: Bug
DistroRelease: Ubuntu 12.04
Package: installation-report 2.46ubuntu1
ProcVersionSignature: Ubuntu 3.2.0-36.36-lowlatency 3.2.35
Uname: Linux 3.2.0-36-lowlatency i686
ApportVersion: 2.0.1-0ubuntu17.1
Architecture: i386
Date: Sat Jan 19 06:56:59 2013
Dependencies:

InstallationMedia: Xubuntu 12.04.1 LTS "Precise Pangolin" - Release i386 (20120822.1)
MarkForUpload: True
PackageArchitecture: all
ProcEnviron:
 LANGUAGE=de_DE:en
 TERM=xterm
 PATH=(custom, no user)
 LANG=de_DE.UTF-8
 SHELL=/bin/bash
SourcePackage: installation-report
UpgradeStatus: No upgrade log present (probably fresh install)

Revision history for this message
Andreas Glaeser (aglaeser1) wrote :
Revision history for this message
Ubuntu Foundations Team Bug Bot (crichton) wrote :

The attachment "grub-defaults to enable kernel-selection at boot-time and avoid non-displayable graphic mode at shutdown" of this bug report has been identified as being a patch. The ubuntu-reviewers team has been subscribed to the bug report so that they can review the patch. In the event that this is in fact not a patch you can resolve this situation by removing the tag 'patch' from the bug report and editing the attachment so that it is not flagged as a patch. Additionally, if you are member of the ubuntu-reviewers team please also unsubscribe the team from this bug report.

[This is an automated message performed by a Launchpad user owned by Brian Murray. Please contact him regarding any issues with the action taken in this bug report.]

tags: added: patch
Revision history for this message
Andreas Glaeser (aglaeser1) wrote :

The bootproblems are really quite tricky, sometimes there were no problems for a whole week, then they reoccurred. The box I am currently testing has the current and only available Q-BIOS version. Following settings were made:
'Power Management Setup'/'ACPI function [Enabled]'
'PnP/PCI Configurations'/'Resources Controlled By [Manual]'
'Integrated Peripherals'/'SIS OnChip IDE Device'/'Primary Slave UltraDMA [Disabled]',
'IDE Primary Slave PIO [Mode 0]',
'IDE Burst Mode [Disabled]',
'IDE HDD Block Mode [Disbled]'.
This works for now with the current Quantal-Kernel version 3.5.22, which has the CPU run in performance mode initially. Maybe this is sufficient and sustainable.

Revision history for this message
Andreas Glaeser (aglaeser1) wrote :

Untill yesterday I saw problems twice while trying to (re)boot. The first picture 'hss-0.jpg' was taken at cold boot, so maybe it was due to too low environmental temperature. This occured once only, then I reinstallaed and used the ext4 filesystem this time instead of ext3. No boot-problems occured since then.
The second picture 'hss-1.jpg' shows the hanging boot-process during IRQ-initialisation with a different BIOS-setting that mentioned above, this happened with 'Resources controlled By [Auto(ESCD)] actually.
So it seems, that either the BIOS is a bit buggy and has a sloppy PnP-ACPI initialisation routine, or current Linux kernels are.

Revision history for this message
Andreas Glaeser (aglaeser1) wrote :
Revision history for this message
Andreas Glaeser (aglaeser1) wrote :

My current BIOS-Settings are:

    'Power Management Setup'/'ACPI function [Enabled]'
    'PnP/PCI Configurations'/'Resources Controlled By [Auto(ESCD)]'
    'Integrated Peripherals'/'SIS OnChip IDE Device'/'Primary Master UltraDMA [Disabled]',
    /'Primary Slave UltraDMA [Disabled]'
    /'IDE DMA transfer access [Disabled]'
    'IDE Primary Slave PIO [Mode 0]',
    'IDE Burst Mode [Enabled]' and 'IDE HDD Block Mode [Enabled]'

It works for now, so my last and latest guess is this: IDE-DMA-mode interferes with IRQ-Initalisation at boot-time.
The device runs well enough, even without IDE-DMA, though it would be a nice to have feature, not trying to measure the difference in overall-performance.

Revision history for this message
Andreas Glaeser (aglaeser1) wrote :

First I tried this for a few day with Debian Wheezy beta RC, now again with XUbuntu, doing frequent reboots, no hanging boot process occured anymore. The issue has been fixed.
I would like to say thank-you to the person who did it, just marked this as 'Fix Released'.

Changed in installation-report (Ubuntu):
status: New → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.