Content-Type: multipart/alternative; boundary="------------060100050107030607090101" --------------060100050107030607090101 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit ok, attached please find dmesg dumps from both the gparted .iso and the 'stock' debian distro for colinux (the only distro I can keep running long enough to dump the dmesg) ... Some more on the slowness problem ... near as I can tell, actual processing speed is not affected; it may be a display/update issue on the desktop. Mouse speed remains normal ... programs seem to execute at normal speeds. You just only get a screen/display update about every 4-5 seconds. You can see this most graphically by enabling the second hand on the vista desktop clock; when the problem starts, the second hand starts 'jumping' every 4-5 seconds to the current time. Per the 'gurus' here at my company, nobody's been able to get colinux working specifically on Vista on machines w/AMD Turion cpus. thx! - Jonathan Henry Nestler wrote: > http://gparted.sourceforge.net/livecd.php > > is a very small live cdrom with USB mount support and a console and very > recent kernel 2.6.22. I use it with VESA mode for mostly all PC. > > Henry > > Jonathan Deitch wrote: > >> Henry - >> >> capturing dmesg from knoppix will have to wait a bit ... I didn't >> realize the distro had moved from CD to DVD, hence the download is >> taking a bit longer. >> >> - Jonathan >> >> Henry Nestler wrote: >> >>> Jonathan Deitch wrote: >>> >>> >>>> I'll run these tests tonight ... >>>> >>>> However - and this is the puzzling part - when the system goes into the >>>> slow-mode, there is no abnormal CPU load. Yep, that's right ... no high >>>> CPU in either task manager, or top. >>>> >>>> Both colinux and Vista are still usable - they're just very very slow >>>> with a 4-5 second leg between any action and system response. >>>> >>>> >>> Than I have an idea: The cpu has a buggy or a special timer chip? I have >>> read some about buggy TSC, and Windows has shipped an update for such >>> chips in a service pack. >>> >>> You says, the taskman shows na normal cpu load? Than the sytesm timer >>> was falling down into slow mode? Currently I don't know why coLinux >>> would reprogrammed the timer. Colinux does no change the timer >>> intervall. Not directly, but perhaps the new kernel has some hocks for >>> special cpu's? I will grep for it. >>> >>> The mouse is slow. ok. What is with a keyboard resonse time, for example >>> in a notepad, or in a command prompt? >>> >>> Please boot a knoppix, or from a distribution you have in a rescure mode >>> with text console. Somethere you can cat the native kernel boot messages >>> (dmesg) into a file and send me, please. I will grep for special >>> messages about your cpu. You also can send me the kernel messages from >>> coLinux-daemon.exe, as you can do it via mouse "copy all text" from >>> screen into clipboard and paste it into notepad. Or run >>> "colinux-daemon.exe ... > file.log" >>> >>> > > --------------060100050107030607090101 Content-Type: text/html; charset=ISO-8859-15 Content-Transfer-Encoding: 8bit ok, attached please find dmesg dumps from both the gparted .iso and the 'stock' debian distro for colinux (the only distro I can keep running long enough to dump the dmesg) ...

Some more on the slowness problem ... near as I can tell, actual processing speed is not affected; it may be a display/update issue on the desktop.

Mouse speed remains normal ... programs seem to execute at normal speeds.

You just only get a screen/display update about every 4-5 seconds.  You can see this most graphically by enabling the second hand on the vista desktop clock; when the problem starts, the second hand starts 'jumping' every 4-5 seconds to the current time.

Per the 'gurus' here at my company, nobody's been able to get colinux working specifically on Vista on machines w/AMD Turion cpus.

thx!

- Jonathan

Henry Nestler wrote:
http://gparted.sourceforge.net/livecd.php

is a very small live cdrom with USB mount support and a console and very
recent kernel 2.6.22. I use it with VESA mode for mostly all PC.

Henry

Jonathan Deitch wrote:
  
 Henry -

capturing dmesg from knoppix will have to wait a bit ... I didn't
realize the distro had moved from CD to DVD, hence the download is
taking a bit longer.

- Jonathan

Henry Nestler wrote:
    
Jonathan Deitch wrote:
  
      
 I'll run these tests tonight ...

However - and this is the puzzling part - when the system goes into the
slow-mode, there is no abnormal CPU load.  Yep, that's right ... no high
CPU in either task manager, or top.

Both colinux and Vista are still usable - they're just very very slow
with a 4-5 second leg between any action and system response.
    
        
Than I have an idea: The cpu has a buggy or a special timer chip? I have
read some about buggy TSC, and Windows has shipped an update for such
chips in a service pack.

You says, the taskman shows na normal cpu load? Than the sytesm timer
was falling down into slow mode? Currently I don't know why coLinux
would reprogrammed the timer. Colinux does no change the timer
intervall. Not directly, but perhaps the new kernel has some hocks for
special cpu's? I will grep for it.

The mouse is slow. ok. What is with a keyboard resonse time, for example
in a notepad, or in a command prompt?

Please boot a knoppix, or from a distribution you have in a rescure mode
with text console. Somethere you can cat the native kernel boot messages
(dmesg) into a file and send me, please. I will grep for special
messages about your cpu. You also can send me the kernel messages from
coLinux-daemon.exe, as you can do it via mouse "copy all text" from
screen into clipboard and paste it into notepad. Or run
"colinux-daemon.exe ... > file.log"

      

  
--------------060100050107030607090101--