colinux-users

Re: [coLinux-users] Colinux blue screen crashes

[Bob W. <bwh...@my...>]

Qnx4ever wrote:

> but I'm a bit annoyed by the tone of your messages.

You are easily annoyed. Maybe your skin is too thin.

I see no reason for you to be annoyed. You are not on

The list of colinux developers. So what is your relation

To the colinux project if any? If you don't work for colinux project then
your comments should be ignored.

 

>Maybe there's something wrong with your PC if no one >else had similar
issues ? 

This PC had no blue screen crashes for 5+ years of running Windows XP. The
crashes only occurred when I started running colinux and they don't occur on
the days

That I don't run colinux. Since many users may not frequently use up 95% or
greater of memory they may not encounter this problem.

 

>Does anyone care about this new Ubuntu release,

Maybe the thousands of downloaders of Portable Ubuntu do care.

 

>piss everyone off ? 

I doubt if you speak for everyone.  Your comments are no  help to anyone.

 

>>These things that you have asked me to try seem like a >>waste of time so
far.

>>It would be better if somebody that understands >>colinux well looked at
the patterns in memory where >>the corruption is detected to see if they
recognize >>anything related to what colinux is doing.

The operative word here is "seem".  Here I am suggesting  a more productive
path than the trial and error method that was suggested. I have spent 2
months so far with Kaspersky and Colinux trying to solve this problem.
Kaspersky has determined that it is not caused by them. The tests with ndis,
pcap, slirp indicate that it does not matter which networking mode is being
used as the crash occurs on all modes. Obvious, I find this frustrating that
it is taking so long to get this problem solved.

Re: [coLinux-users] Colinux blue screen crashes

[yin s. <sun...@gm...>]

I remember someone suggest to try another distribution, disable
certain features etc, to narrow down the scope. Maybe you can
summarize what have been tried, and try to analyze the dump yourselves
to give us some directions.
There is really no sure way as a productive path when processes stump
on each other.
Please try as much as you can.


Thanks,

On Tue, Jan 24, 2012 at 11:38 AM, Bob Wheater <bwh...@my...> wrote:
> Qnx4ever wrote:
>
>> but I'm a bit annoyed by the tone of your messages.
>
> You are easily annoyed. Maybe your skin is too thin.
>
> I see no reason for you to be annoyed. You are not on
>
> The list of colinux developers. So what is your relation
>
> To the colinux project if any? If you don’t work for colinux project then
> your comments should be ignored.
>
>
>
>>Maybe there's something wrong with your PC if no one >else had similar
>> issues ?
>
> This PC had no blue screen crashes for 5+ years of running Windows XP. The
> crashes only occurred when I started running colinux and they don’t occur on
> the days
>
> That I don’t run colinux. Since many users may not frequently use up 95% or
> greater of memory they may not encounter this problem.
>
>
>
>>Does anyone care about this new Ubuntu release,
>
> Maybe the thousands of downloaders of Portable Ubuntu do care.
>
>
>
>>piss everyone off ?
>
> I doubt if you speak for everyone.  Your comments are no  help to anyone.
>
>
>
>>>These things that you have asked me to try seem like a >>waste of time so
>>> far.
>
>>>It would be better if somebody that understands >>colinux well looked at
>>> the patterns in memory where >>the corruption is detected to see if they
>>> recognize >>anything related to what colinux is doing.
>
> The operative word here is “seem”.  Here I am suggesting  a more productive
> path than the trial and error method that was suggested. I have spent 2
> months so far with Kaspersky and Colinux trying to solve this problem.
> Kaspersky has determined that it is not caused by them. The tests with ndis,
> pcap, slirp indicate that it does not matter which networking mode is being
> used as the crash occurs on all modes. Obvious, I find this frustrating that
> it is taking so long to get this problem solved.
>
>
> ------------------------------------------------------------------------------
> Keep Your Developer Skills Current with LearnDevNow!
> The most comprehensive online learning library for Microsoft developers
> is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
> Metro Style Apps, more. Free future releases when you subscribe now!
> http://p.sf.net/sfu/learndevnow-d2d
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users
>

Re: [coLinux-users] Colinux blue screen crashes

[Henry N. <hen...@ar...>]

Exactly, that is the problem. In the logs we only can the the coruppted
memory, but not the task who changed it.

For the test:
I have started a minimal Debian 4 with 128 MB RAM and no swapfile and in
there I started this Script:

=== start memeater.sh ===
#!/bin/sh

eat()
{
    sleep 100
}

count=0
while [ $count -le 10000 ]
do
    count=$(( $count + 1 ))
    eat &
done                                                                                                                                                        

===end memeater.sh ===

Warning! This Script starts lot of sub shells. For my my system, it was
unstable near 700 running tasks. I saw kernel messages "oom-killer ...
sh killed" inside coLinux console. But no problems on Windows side. I
have a Single Core 2.1GHz Windows XP.
Run the "sync.exe" from
http://technet.microsoft.com/en-en/sysinternals/bb897438 to save your
unbuffered files, before you forces the blue screen.

An other simple memory test you can do with the default initrd.gz from
coLinux installation.
Save the script as file "memeater.sh" into your coLinux install path
(for example "c:\colinux")

Run coLinux:

    sync.exe
    cd c:\colinux
    colinux-daemon kernel=vmlinux mem=128 initrd=initrd.gz
root=/dev/ram0 cofs0=.

Inside Linux ramdisk run theses commands:

    mount -t cofs cofs0 /mnt
    sh
    cd /mnt
    sh ./memeater.sh

For me here is the same effect: coLinux kernel kills unused task "sleep
100" via oom-killer, and Windows runs very stable.
If you have more RAM for coLinux in use, the increase the total count of
task inside the script.

I remember for a very old bug inside some Intel CPUs from the Core 2
Duo. There was an internal problem with the CPU cache lines, because
coLinux switches to fast the complete Page-Tables without flush. Maybe
your CPU has similar (but other, because the oldy has a workaround) and
you can disable the CPU-Cache in your BIOS for a test cycle.

What is your exactly CPU type. Please use a tool from manufacture. Then
we can locate an existing errata for this.

-- 
Henry N.