Posts: 8
uranus
Joined: 22 Jan 2011
#1
Hello, I've been running antiX for many months, permanently installed on hard drive of an old IBM Pent 4.

I *HATE* distro-hopping. I really want to stick with antiX because it's by far the best distro I've yet found for old hardware. Just stick in on there, and it works! Ubuntu and all the other well-known distros caused me great stress, particularly on my even older IBM laptop.

But now I need to do some audio recording, multitrack - antiX maybe not the obvious choice for a DAW, but that's what I've got."Ardour" looks perfect for what I want in screen shots, but it's a broken package as far as antiX is concerned, dependency hell.

Can I make it work by doing my dist-upgrade from different (test or unstable) repositories? Or is there another package I should be using instead? I've tried that old favourite Audacity, but it seems surprisingly flaky on this machine.

inxi -F
System: Host ThinkCentre Kernel 2.6.27-1-mepis-smp i686 (32 bit) Distro antiX-M8.2 Tȟašúŋke Witkó - Crazy Horse 24 July 2009
CPU: Single core Intel Pentium 4 (UP) cache 512 KB flags (sse2) bmips 5317.07 clocked at 2657.797 MHz
Graphics: Card Intel 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device X.Org 1.4.2 Res: 1024x768@60.0hz
GLX Renderer Mesa GLX Indirect GLX Version 1.4 (2.1 Mesa 7.0.4) Direct Rendering No
Audio: Card Intel 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller driver Intel ICH at ports 1c00 18c0
Sound: Advanced Linux Sound Architecture Version 1.0.17

Any help or comments really appreciated, thanks. And this may be a stupid question, but can I change to M8.5 or some newer version without wiping my current config?
anticapitalista
Posts: 5,955
Site Admin
Joined: 11 Sep 2007
#2
Firstly I would check to see if a later kernel works ok on your box. You could try installing the latest MEPIS one or one from liquorix.

deb
========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://liquorix.net/debian/"
linktext was:"http://liquorix.net/debian/"
====================================
sid main
deb
========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"ftp://ftp.mepis.com/mepis/"
linktext was:"ftp://ftp.mepis.com/mepis/"
====================================
mepis-11.0 main

Don't delete the old one, just in case there is a problem with a later kernel.

If the kernel is ok, then you can dist-upgrade using testing repos and the MEPIS ones disabled. You may have some issues with wicd. You cannot dist-upgrade and use the original kernel because of incompatible udev.

You could try installing ardour from testing repo using aptitude rather than apt-get.
Posts: 8
uranus
Joined: 22 Jan 2011
#3
Making some progress . . got the new kernel in there:

uname -a
Linux ThinkCentre 2.6.37-0.dmz.5-liquorix-686 #1 ZEN SMP PREEMPT Sun Jan 23 03:53:56 UTC 2011 i686 GNU/Linux

inxi -F
System: Host ThinkCentre Kernel 2.6.37-0.dmz.5-liquorix-686 i686 (32 bit) Distro antiX-M8.2 Tȟašúŋke Witkó - Crazy Horse 24 July 2009
CPU: Single core Intel Pentium 4 (-UP-) cache 512 KB flags (sse sse2) bmips 5317.9 clocked at 2658.950 MHz
Graphics: Card Intel 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device X.org 1.7.7 Res: 168x56 Gfx Data: N/A for root
Audio: Card Intel 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller driver Intel ICH at ports 1c00 18c0 BusID: 00:1f.5
Sound: Advanced Linux Sound Architecture Version 1.0.23

- lots of messages whizz past on the boot console that I don't understand, udevd?? They're not in dmesg, so perhaps they don't matter?

That's probably enough playing around with the system for today, I shall wait a bit before messing it all up again by installing a new package __{{emoticon}}__
Posts: 8
uranus
Joined: 22 Jan 2011
#4
X.org 1.7.7 Res: 168x56 __{{emoticon}}__

I know my screen is old, but it's not that bad . . .
anticapitalista
Posts: 5,955
Site Admin
Joined: 11 Sep 2007
#5
uranus those messages are probably connected with telling you that all files in /etc/modprobe.d/ neeed to end with xxx.conf eg blacklist.conf
Posts: 8
uranus
Joined: 22 Jan 2011
#6
Well, I can run Ardour so that's good, though managing input sources is obviously going to be a learning curve.

However, there's something not right with Xorg / graphics drivers . . dragging windows is extremely slow and uses 100% CPU, and after a certain amount of time (somewhere around 30 minutes) it will crash - screen scrambled, only way out is to press the hardware reset.

I guess that's what happens when you select"unstable / testing" repos!

Undecided as to whether to try to troubleshoot X, or to try to find a way to revert to stable releases __{{emoticon}}__

Any advice welcome . . .
Posts: 89
Stuart O'mahony
Joined: 07 Mar 2010
#7
I'm running an IBM p4, and my browser windows are slow to redraw when running at 100% CPU, but I only ever hit 100% when running multiple intensive apps, no ordinarily, i.e. xorg related.

Type 'top' into a teminal and you can see how much in terms of percentage processes tax the CPU. If the problem is with xorg then I don't know how to fix that i'm afraid.
Posts: 8
uranus
Joined: 22 Jan 2011
#8
Hi Stuart, good idea, proves that indeed it is Xorg that is going ballistic when I drag any window.

Didn't happen with 1.4.2, is happening now with 1.7.7 - that's why I was wondering if I could somehow"downgrade"!
anticapitalista
Posts: 5,955
Site Admin
Joined: 11 Sep 2007
#9
You won't be able to downgrade xorg.

Is everything upgraded ok? Does apt-get update apt-get dist-upgrade give any errors?

What about the boot errors?

I think if they get fixed than maybe you'll have less issues with xorg if any.

I dist-upgraded antiX-M8.2 to present on a PIII Coppermine 700Mhz and it doesn't show any 100% usage.
Posts: 8
uranus
Joined: 22 Jan 2011
#10
(Let me know if I'm getting too off-topic, and you want me to start a new thread.)

I think this is a KMS / Frame Buffer issue; I don't know much about KMS but it seems like fundamentally a bad idea - why should control of video hardware live in the kernel?? Seems like KMS may be in all newer kernels?

Anyway, it worked fine with vesafb, and now it doesn't work so well with inteldrmfb, so I might find a way to revert to the older way. I'll have a look at xorg.conf too.
[liquorix kernel]


[ dmesg | grep intel ]
intel_rng: Firmware space is locked read-only. If you can't or
intel_rng: don't want to disable this in firmware setup, and if
intel_rng: you are certain that your system has a functional
intel_rng: RNG, try using the 'no_fwh_detect' option.
agpgart-intel 0000:00:00.0: Intel 830M Chipset
agpgart-intel 0000:00:00.0: detected gtt size: 131072K total, 131072K mappable
agpgart-intel 0000:00:00.0: detected 8192K stolen memory
agpgart-intel 0000:00:00.0: AGP aperture is 128M @ 0x88000000
intel8x0_measure_ac97_clock: measured 50437 usecs (2431 samples)
intel8x0: clocking to 48000
fb: conflicting fb hw usage inteldrmfb vs VESA VGA - removing generic driver
fbcon: inteldrmfb (fb0) is primary device
fb0: inteldrmfb frame buffer device

[ dmesg | grep fb ]
fb: conflicting fb hw usage inteldrmfb vs VESA VGA - removing generic driver
fbcon: inteldrmfb (fb0) is primary device
fb0: inteldrmfb frame buffer device

[2.6.27-1-mepis-smp]

[ dmesg | grep intel ]
agpgart-intel 0000:00:00.0: Intel 830M Chipset
agpgart-intel 0000:00:00.0: detected 8060K stolen memory
agpgart-intel 0000:00:00.0: AGP aperture is 128M @ 0x88000000
intel_rng: Firmware space is locked read-only. If you can't or
intel_rng: don't want to disable this in firmware setup, and if
intel_rng: you are certain that your system has a functional
intel_rng: RNG, try using the 'no_fwh_detect' option.
intel8x0_measure_ac97_clock: measured 50557 usecs
intel8x0: clocking to 48000

[ dmesg | grep fb ]
BIOS-e820: 000000001f6f0000 - 000000001f6fb000 (ACPI data)
BIOS-e820: 000000001f6fb000 - 000000001f700000 (ACPI NVS)
mapped APIC to ffffb000 (fee00000)
PM: Registered nosave memory: 000000001f6f0000 - 000000001f6fb000
PM: Registered nosave memory: 000000001f6fb000 - 000000001f700000
vesafb: framebuffer at 0x88000000, mapped to 0xe0080000, using 3072k, total 8000k
vesafb: mode is 1024x768x16, linelength=2048, pages=4
vesafb: scrolling: redraw
vesafb: Truecolor: size=0:5:6:5, shift=0:11:5:0
fb0: VESA VGA frame buffer device
e100: eth0: e100_probe: addr 0xc0100000, irq 20, MAC addr 00:0d:60:a6:7d:fb
Posts: 89
Stuart O'mahony
Joined: 07 Mar 2010
#11
From your output.

Code: Select all

fb: conflicting fb hw usage inteldrmfb vs VESA VGA - removing generic driver
See this:


========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://linux.koolsolutions.com/2009/11/02/howto-enabling-kernel-mode-setting-kms-in-debian-linux-kernel/"
linktext was:"http://linux.koolsolutions.com/2009/11/ ... ux-kernel/"
====================================


As per the article, you appear to have"fb0: inteldrmfb frame buffer device" listed in your dmesg output which would suggest you have KMS enabled.

Plus this:

"So couple of things to remember. Just by upgrading to 2.6.29 kernel or later and configuring it properly, you cannot have KMS. You also need to make sure that you have relatively new user space installed i.e. XOrg system. Both, the Linux kernel and XOrg system have to work cooperatively to get KMS working. So you need to make sure that you are using xserver-xorg-video-intel package version 2.5.0 or later. Unfortunately currently, Debian Lenny’s (Stable) has version 2.3.2. This means that even if we upgrade to 2.6.29 or later kernel, we won’t have KMSworking because we have an old user-space (2.3.2<2.5.0) which is typical of Debian considering it’s main criteria is stability. So we will try to get KMSworking on Debian Testing (Squeeze) which has 2.9.0 version for xserver-xorg-video-intel package."

You can check your xserver-xorg-video-intel package version by entering the following:

Code: Select all

apt-cache policy xserver-xorg-video-intel
I guess from here you can either take steps to enable (functionly!) or disable KMS. I'd try disabling it first just to see if this is causing the problem.
Posts: 8
uranus
Joined: 22 Jan 2011
#12
xserver-xorg-video-intel:
Installed: 2:2.13.0-5

So I suppose that's new enough.

According to:

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://wiki.debian.org/KernelModesetting#DisablingKMS"
linktext was:"http://wiki.debian.org/KernelModesetting#DisablingKMS"
====================================
>>To disable KMS for Intel and Radeon cards, either:
Boot with the nomodeset kernel command line parameter.
[or]
Edit /etc/modprobe.d/i915-kms.conf or /etc/modprobe.d/radeon-kms.conf accordingly.
However, Setting modeset=0 in the conf files means X will not start [fatal server error - no screens found]

I'll try starting the kernel with nomodeset, but I suspect I will just have to live with it until someone else sorts out the device drivers - beyond my ability!