colinux-users — coLinux users mailing list

RE: [coLinux-users] Troubles booting the Gentoo-2.6 image

[peter g. <plu...@bi...>]

yeah you picked the wrong image

the gentoo images (for some crazy reason) use devfs

the devfs structure of colinux changed in the development versions since the
last release

Gentoo-2.6.7-colinux-0.6.2 is only suitable for use with the development
snapshots of colinux atm


> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of
> si...@ne...
> Sent: 13 October 2004 17:01
> To: col...@li...
> Subject: [coLinux-users] Troubles booting the Gentoo-2.6 image
>
>
> Hi,
>
> I downloaded Gentoo-2.6.7-colinux-0.6.2 image file from sf last
> week as well as
> the coLinux base software. I wasn't able to get this image to
> boot - I have
> since moved on to using gentoo-i586-ext3-2g-deluxe which works just fine.
>
> The kernel boots and begins the init process - but then errors
> out like the
> (imaginary) disk has a bad superblock. I thought the image might
> be corrupted -
> so I re-downloaded and decompressed it (using WinRAR). Same issue.
>
> Is there a known problem with this image file? A search of the
> mailing list
> archives was unfruitful, though I did find the resolution for my
> emerge perl
> issue with the 2.4 image.
>
> Thanks...
>
> --andy kruse
>
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

[coLinux-users] Troubles booting the Gentoo-2.6 image

[<si...@ne...>]

Hi,

I downloaded Gentoo-2.6.7-colinux-0.6.2 image file from sf last week as well as
the coLinux base software. I wasn't able to get this image to boot - I have
since moved on to using gentoo-i586-ext3-2g-deluxe which works just fine.

The kernel boots and begins the init process - but then errors out like the
(imaginary) disk has a bad superblock. I thought the image might be corrupted -
so I re-downloaded and decompressed it (using WinRAR). Same issue.

Is there a known problem with this image file? A search of the mailing list
archives was unfruitful, though I did find the resolution for my emerge perl
issue with the 2.4 image.

Thanks...

--andy kruse

Re: [coLinux-users] reiserfs error during boot, but I have xfs?

[Nuno L. <ml-...@xp...>]

Brian Krusic, dando pulos de alegria, escreveu :
[...]
> FAT: bogus logical sector size 0
> UMSDOS: msdos_read_super failed, mount aborted.
> FAT: bogus logical sector size 0
> FAT: bogus logical sector size 0
> sh-2021: reiserfs_read_super: can not find reiserfs on cobd(117,7)
> Kernel panic: VFS: Unable to mount root fs on 75:07
> 
> My partition layout (being XFS fs);
[...]

The current linux kernel has XFS as a module. When it tries to open the
root partition the modules aren't loaded yet, so it fails (after trying
the file systems it knows about).

The solution would be to either use a non-XFS image as root or
create an initrd image that will allow the xfs module to be loaded.

Another option is compiling a new kernel with xfs linked into.

-- 
Regards,
~Nuno Lucas

[coLinux-users] reiserfs error during boot, but I have xfs?

[Brian K. <br...@kr...>]

Hi,

This is a snippet of my boot screen;

cobd5 size: 1020096 kb
cobd6 size: 522081 kb
cobd7 size: 5349613 kb
cobd: loaded (max 32 devices)
conet: loaded (max 16 devices)
conet0: initialized
FAT: bogus logical sector size 0
UMSDOS: msdos_read_super failed, mount aborted.
FAT: bogus logical sector size 0
FAT: bogus logical sector size 0
sh-2021: reiserfs_read_super: can not find reiserfs on cobd(117,7)
Kernel panic: VFS: Unable to mount root fs on 75:07

My partition layout (being XFS fs);

/dev/hda5 /boot = 1GB
/dev/hda6 swap = 500MB
/dev/hda7 / = 5GB

My xml;

<block_device index="5" path="\Device\Harddisk0\Partition4" enabled="true" 
/>
<block_device index="6" path="\Device\Harddisk0\Partition5" enabled="true" 
/>
<block_device index="7" path="\Device\Harddisk0\Partition6" enabled="true" 
/>
<bootparams>ro root=/dev/cobd7 colinux 4</bootparams>

Does any one have some ideas for me?
I'll buy you a beer.

Bri- 

RE: [coLinux-users] crossing the bridge

[peter g. <plu...@bi...>]

btw colinux supports multiple network adaptors nowadays

so you can use tap for host-colinux and winpcap for network-colinux

> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of Curtis,
> Craig
> Sent: 12 October 2004 20:16
> To: col...@li...
> Cc: peter green
> Subject: Re: [coLinux-users] crossing the bridge
>
>
> Yes.  This seems to be my problem.  I have an Intel Pro/1000 and I do
> not see a way to get to ChecksumRxTcp.  I found the other parameters in
> my registry, but ChecksumRxTcp does not appear to exist.  Could the
> networking docs have a typo?
>
> Otherwise, I may need to try a new network card...
>
> Thanks,
> Craig
>
> peter green wrote:
>
> >checksum offload on some network cards has been known to cause
> problems with
> >communication from host-guest
> >
> >if all else fails you can always buy a cheap realtek ;)
> >
> >
> >
> >>-----Original Message-----
> >>From: col...@li...
> >>[mailto:col...@li...]On Behalf Of Curtis,
> >>Craig
> >>Sent: 12 October 2004 12:33
> >>To: col...@li...
> >>Subject: [coLinux-users] crossing the bridge
> >>
> >>
> >>It seems that only pinging is supported as a network protocol between
> >>the Windows side and Colinux on a bridged network.  I can ping just
> >>fine, but FTP, smb, and other protocols do not work on the same PC, but
> >>from another PC they do work.
> >>
> >>Is there a configuration on the PC-TAP network driver or in Win2K that I
> >>can change somewhere to allow this to work?
> >>
> >>Thanks,
> >>Craig
> >>
> >>
> >>-------------------------------------------------------
> >>This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> >>Use IT products in your business? Tell us what you think of
> them. Give us
> >>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> >>find out more
> >>http://productguide.itmanagersjournal.com/guidepromo.tmpl
> >>_______________________________________________
> >>coLinux-users mailing list
> >>coL...@li...
> >>https://lists.sourceforge.net/lists/listinfo/colinux-users
> >>
> >>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

Re: [coLinux-users] crossing the bridge

[Brian K. <br...@kr...>]

Hi,

> not see a way to get to ChecksumRxTcp.  I found the other parameters in my 
> registry, but ChecksumRxTcp does not appear to exist.

Add the key and see what happens.

I've added keys before to manipulate tcp windows sizes, etc and it didn't 
break anything.

Bri- 

Re: [coLinux-users] crossing the bridge

[Curtis, C. <Cra...@ec...>]

Yes.  This seems to be my problem.  I have an Intel Pro/1000 and I do 
not see a way to get to ChecksumRxTcp.  I found the other parameters in 
my registry, but ChecksumRxTcp does not appear to exist.  Could the 
networking docs have a typo?

Otherwise, I may need to try a new network card...

Thanks,
Craig

peter green wrote:

>checksum offload on some network cards has been known to cause problems with
>communication from host-guest
>
>if all else fails you can always buy a cheap realtek ;)
>
>  
>
>>-----Original Message-----
>>From: col...@li...
>>[mailto:col...@li...]On Behalf Of Curtis,
>>Craig
>>Sent: 12 October 2004 12:33
>>To: col...@li...
>>Subject: [coLinux-users] crossing the bridge
>>
>>
>>It seems that only pinging is supported as a network protocol between
>>the Windows side and Colinux on a bridged network.  I can ping just
>>fine, but FTP, smb, and other protocols do not work on the same PC, but
>>from another PC they do work.
>>
>>Is there a configuration on the PC-TAP network driver or in Win2K that I
>>can change somewhere to allow this to work?
>>
>>Thanks,
>>Craig
>>
>>
>>-------------------------------------------------------
>>This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
>>Use IT products in your business? Tell us what you think of them. Give us
>>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
>>find out more
>>http://productguide.itmanagersjournal.com/guidepromo.tmpl
>>_______________________________________________
>>coLinux-users mailing list
>>coL...@li...
>>https://lists.sourceforge.net/lists/listinfo/colinux-users
>>    
>>

RE: [coLinux-users] Re: colinux w/fc2 parition

[peter g. <plu...@bi...>]

you could but you would have to customise it
personally i think having filesystem drivers as modules is a really shitty
idea though


> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of Martin
> Kanich
> Sent: 12 October 2004 07:21
> To: col...@li...
> Subject: [coLinux-users] Re: colinux w/fc2 parition
>
>
> A. Klingenstein wrote:
> > You might to have compile your own CoLinux kernel to boot it
> from an XFS
> > partition (I think the latest standard CoLinux kernel 2.6.8.1 loads XFS
> > as a module), but it's not a big deal
> You can't use that initrd for this?
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

RE: [coLinux-users] crossing the bridge

[peter g. <plu...@bi...>]

checksum offload on some network cards has been known to cause problems with
communication from host-guest

if all else fails you can always buy a cheap realtek ;)

> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of Curtis,
> Craig
> Sent: 12 October 2004 12:33
> To: col...@li...
> Subject: [coLinux-users] crossing the bridge
>
>
> It seems that only pinging is supported as a network protocol between
> the Windows side and Colinux on a bridged network.  I can ping just
> fine, but FTP, smb, and other protocols do not work on the same PC, but
> from another PC they do work.
>
> Is there a configuration on the PC-TAP network driver or in Win2K that I
> can change somewhere to allow this to work?
>
> Thanks,
> Craig
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

[coLinux-users] NFS instead of Samba?

[Chris D. <da...@ya...>]

I was thinking about the performance issues that are especially evident when
using samba shares between coLinux and Windows.  I was wondering if anyone had
tried out the NFS client and server in Microsoft's "Services for UNIX".
(http://www.microsoft.com/windows/sfu/)  Is this more efficient than samba?  I
would hope that it would be, but maybe it would be slower... I don't really
know how it is implemented.

Forgive me if this has already been talked about.  I didn't remember it, and
didn't see anything in a quick check of the archives. :o)

--Chris

=====
There is a very fine line between "hobby" and "mental illness." --Dave Barry
Linux DVDs: http://www.LinuxDVDs.com
personal pages: http://www.dahlweb.net

Re: [coLinux-users] crossing the bridge

[Mitch B. <wm...@ag...>]

Other protocols work just fine for me.  I have colinux running bridged 
on Win2K Pro, WinXP Pro, and WinXP home systems, all using the PCAP 
driver.  I haven't tried FTP, but SMB and SSH work just fine between 
coLinux and the Windows host systems, and other systems too.

Perhaps you have firewall software on the Windows side that is filtering 
out some ports?


Curtis, Craig wrote:

> It seems that only pinging is supported as a network protocol between 
> the Windows side and Colinux on a bridged network.  I can ping just 
> fine, but FTP, smb, and other protocols do not work on the same PC, 
> but from another PC they do work.
>
> Is there a configuration on the PC-TAP network driver or in Win2K that 
> I can change somewhere to allow this to work?
>
> Thanks,
> Craig
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out 
> more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

[coLinux-users] crossing the bridge

[Curtis, C. <Cra...@ec...>]

It seems that only pinging is supported as a network protocol between 
the Windows side and Colinux on a bridged network.  I can ping just 
fine, but FTP, smb, and other protocols do not work on the same PC, but 
from another PC they do work.

Is there a configuration on the PC-TAP network driver or in Win2K that I 
can change somewhere to allow this to work?

Thanks,
Craig

[coLinux-users] Re: colinux w/fc2 parition

[Martin K. <ka...@po...>]

A. Klingenstein wrote:
> You might to have compile your own CoLinux kernel to boot it from an XFS 
> partition (I think the latest standard CoLinux kernel 2.6.8.1 loads XFS 
> as a module), but it's not a big deal
You can't use that initrd for this?

Re: [coLinux-users] colinux w/fc2 parition

[A. K. <kli...@we...>]

Brian Krusic wrote:

> Hi,
>  
> Before I spend time on this as I'm under the gun, will this work;
>  
> - Run colinux to access my fedora core 2 XFS partition.
> - I've dual boot and desire access to the Linux partitions via Windows 
> w/o having to boot into Linux itself.
>  
> Anybodys reply is greatly appreciated.

Yes it will work. To make it seemless, you want to run CoLinux as a 
service and automatically start samba on the CoLinux side, then you 
mount it like any other share. I currently access my reiserfs and xfs 
partitions this way under Windows.
You might to have compile your own CoLinux kernel to boot it from an XFS 
partition (I think the latest standard CoLinux kernel 2.6.8.1 loads XFS 
as a module), but it's not a big deal

Alex
 

Re: [coLinux-users] colinux w/fc2 parition

[Brian K. <br...@kr...>]

Thanks for the replay.

Unfortunatly, this won't work with the XFS file system.

My real goal is to run Samba from within the Linux distro that boot from 
within XP to get bi-dir communication using virtual network connections.

Bri-
Network Consulting Services
www.krusic.com

Re: [coLinux-users] colinux w/fc2 parition

[Simon P. <SJ...@jr...>]

On Monday 11 Oct 2004 20:39, Brian Krusic wrote:
> Hi,
>
> Before I spend time on this as I'm under the gun, will this work;
>
> - Run colinux to access my fedora core 2 XFS partition.
> - I've dual boot and desire access to the Linux partitions via Windows w/o
> having to boot into Linux itself.
>
> Anybodys reply is greatly appreciated.
>
> I am looking over the mail archives but its a lot of stuff.
> Bri-


Not sure if this is what you need but try:
<http://uranus.it.swin.edu.au/~jn/linux/explore2fs.htm>

I use this to get at files on my Mandrake EXT3 partitions from WinXP.
Not sure if it works with XFS, but it might...

If it works it's faster to download and simpler than doing it with colinux.

Simon

[coLinux-users] colinux w/fc2 parition

[Brian K. <br...@kr...>]

Hi,

Before I spend time on this as I'm under the gun, will this work;

- Run colinux to access my fedora core 2 XFS partition.
- I've dual boot and desire access to the Linux partitions via Windows =
w/o having to boot into Linux itself.

Anybodys reply is greatly appreciated.

I am looking over the mail archives but its a lot of stuff.
Bri-

RE: [coLinux-users] TAP runs at 10Mb/s

[peter g. <plu...@bi...>]

the reported speed is meaninless anyway

tap is a bit on the slow side right now mind you especially for protocols
that need lots of round trips

> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of A.
> Klingenstein
> Sent: 11 October 2004 16:14
> To: col...@li...
> Subject: [coLinux-users] TAP runs at 10Mb/s
>
>
> Is there a way to make TAP run with Fast Ethernet speeds (100Mb/s). By
> default it seems it only runs at 10MBit/s and that is very slow for
> using samba :(
>
> Alex
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

[coLinux-users] TAP runs at 10Mb/s

[A. K. <kli...@we...>]

Is there a way to make TAP run with Fast Ethernet speeds (100Mb/s). By 
default it seems it only runs at 10MBit/s and that is very slow for 
using samba :(

Alex

FW: [coLinux-users] Re: Block device aliasing for dual-boot

[peter g. <plu...@bi...>]

sorry forgot to hit reply to all

-----Original Message-----
From: peter green [mailto:plu...@bi...]
Sent: 11 October 2004 15:40
To: Ian Abbott
Subject: RE: [coLinux-users] Re: Block device aliasing for dual-boot


i think the problem is that the person who added aliasing to colinux didn't
properly understand devfs

btw hd?? is not the proper name for the device when running natively with
devfs it is an alias created by a compatibility profile

> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of Ian
> Abbott
> Sent: 11 October 2004 11:54
> To: col...@li...
> Subject: [coLinux-users] Re: Block device aliasing for dual-boot
>
>
> On 07/10/2004 19:06, Shi, Jue wrote:
> > I just upgraded to snapshot (20041006) and found a commend in
> the default
> > XML file the new block device alsasing stuff.
> > Is there more detail info about it, especially how to use it
> for dual-boot
> > system and how it works?
> >
> > I have setup the colinux successfully to boot into my suse
> partion and got
> > it running. The problem is that the fstab and the network need
> to be changed
> > to get it work. Then these setting need to be changed back in
> order to boot
> > into suse (not through colinux).
> >
> > My question is: can the block device aliasing help in this situation?
>
> I tried block device aliasing for dual-booting Gentoo.  Like
> yourself, I wanted to avoid having two different versions of
> /etc/fstab.  However, in its current form, block device aliasing is
> pretty useless for this task.  For example, if you specify "hda1" as
> an alias name, the current version of the aliasing code creates a
> device called "hda/1" so it ends up with /dev/hda/1 instead of
> /dev/hda1.  Of course, this is incompatible with the normal device
> names when I boot Gentoo natively.
>
> I hope the aliasing code will change to use the standard device
> names.  In the meantime, I use some devfsd rules to create and
> destroy symlinks from the cobd/X device names to the hdaX device names:
>
> REGISTER ^cobd/([0-9]+)$ CFUNCTION GLOBAL mksymlink $devname hda\1
> UNREGISTER ^cobd/([0-9]+)$ CFUNCTION GLOBAL unlink hda\1
>
> The "\1" at the end of the rules gets substituted with the cobd
> device number, so in my colinux.xml file I make sure the
> block_device index values match the hda partition numbers.
> (Actually, I use hdc instead of hda on my system.)  An alternative
> would be to have separate rules for each partition.  Something like
> that would be necessary when the cobd devices are spread over more
> than one disk.
>
> On Gentoo, I put those rules into a file called "mycolinux" in the
> /dev/devfs.d directory.  The /dev/devfsd.conf file contains ends
> with an INCLUDE rule that picks up these custom rules:
>
> # Support additional config installed by packages ...
> INCLUDE /etc/devfs.d
>
> Then I just have to make sure I set the "root=" kernel option in my
> colinux.xml file to correspond with my root partition.  For example,
> if my root partition was /dev/hda2, I would set root=/dev/cobd/2.
>
> I have various Windows partitions in my /etc/fstab which have no
> corresponding cobd devices defined (for obvious reasons!).  These
> are normally mounted at boot time when I boot Gentoo natively and of
> course fail to mount when I boot through coLinux.  This does not
> stop my Gentoo system booting up though.
>
> I haven't set up networking yet.
>
> > Another question is regarding the initrd file. Is it possible
> to use it to
> > solve my problem above?
>
> I don't think so, but I'm not sure what's in the initrd.  I'll
> probably find out when I try and set up networking!
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

Re: [coLinux-users] Invalid checksums

[Paradise <par...@gm...>]

yes, i did try disabling it but same result

On Sun, 10 Oct 2004 19:16:44 +0100, peter green <plu...@bi...> wrote:
> did you try disabling checksum offload?
> 
> its the only thing that is currently known to cause problems with winpcap
> based bridging
> 
> 
> 
> > -----Original Message-----
> > From: col...@li...
> > [mailto:col...@li...]On Behalf Of Paradise
> > Sent: 10 October 2004 19:02
> > To: Xavier Brouckaert
> > Cc: col...@li...
> > Subject: Re: [coLinux-users] Invalid checksums
> >
> >
> > Hi, my case is same with you, with same network card and bridged
> > connection, cannot ssh to colinux from host but my other pc on the
> > network has no problem at all.
> >
> >
> > On Sat, 9 Oct 2004 20:31:33 +0200, Xavier Brouckaert
> > <xav...@gm...> wrote:
> > > Hi,
> > >
> > > I'm running CoLinux 0.6.2 on Windows 2000 SP4 to run a Debian image.
> > > I've got invalid checksums which makes impossible to run an ssh
> > > session between the host and colinux.  I'm using a bridged ethernet
> > > connection.
> > >
> > > My configuration is :
> > > <?xml version="1.0" encoding="UTF-8"?>
> > > <colinux>
> > >     <block_device index="0"
> > > path="\DosDevices\e:\colinux\Debian-20040605-mit.ext3.1610mb"
> > > enabled="true" />
> > >     <block_device index="1" path="\DosDevices\e:\colinux\swap_512Mb"
> > > enabled="true"></block_device>
> > >     <bootparams>root=/dev/cobd0</bootparams>
> > >     <image path="vmlinux" />
> > >     <memory size="128" />
> > >     <!--  <network index="0" type="tap" /> -->
> > >     <network index="0" type="bridged" name="Marvell" />
> > > </colinux>
> > >
> > > A sample tethereal trace (host=192.168.1.100, colinux=192.168.1.102)
> > > of a putty ssh connection:
> > > No.     Time        Source                Destination
> > Protocol Info
> > >       3 0.008588    192.168.1.100         192.168.1.102         TCP
> > >   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
> > >       4 0.008769    192.168.1.100         192.168.1.102         TCP
> > >   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
> > >       5 0.023884    192.168.1.102         192.168.1.100         TCP
> > >   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
> > >       6 0.023902    192.168.1.100         192.168.1.102         TCP
> > >   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
> > >       7 0.024081    192.168.1.100         192.168.1.102         TCP
> > >   [TCP Dup ACK 6#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
> > >       8 0.024373    192.168.1.102         192.168.1.100         TCP
> > >   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
> > >       9 0.024387    192.168.1.100         192.168.1.102         TCP
> > >   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
> > >      10 0.024564    192.168.1.100         192.168.1.102         TCP
> > >   [TCP Dup ACK 9#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
> > >
> > > As you can see, there are some packets with an invalid checksum and
> > > others with a valid checksum.
> > >
> > > The network card is a "Marvell Yukon Gigabit Ethernet
> > > 10/100/1000Base-T Adapter, Copper RJ-45" running at 100Mbps.
> > >
> > > It only happens between the host and colinux.  On colinux, I can
> > > flawlessly install packages and so on...
> > >
> > > Any idea why it happens ?
> > >
> > > Thank you,
> > > Xavier Brouckaert
> > >
> > > -------------------------------------------------------
> > > This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> > > Use IT products in your business? Tell us what you think of
> > them. Give us
> > > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> > find out more
> > > http://productguide.itmanagersjournal.com/guidepromo.tmpl
> > > _______________________________________________
> > > coLinux-users mailing list
> > > coL...@li...
> > > https://lists.sourceforge.net/lists/listinfo/colinux-users
> > >
> >
> >
> > --
> > Regards,
> > Paradise
> >
> >
> > -------------------------------------------------------
> 
> 
> > This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> > Use IT products in your business? Tell us what you think of them. Give us
> > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> > find out more
> > http://productguide.itmanagersjournal.com/guidepromo.tmpl
> > _______________________________________________
> > coLinux-users mailing list
> > coL...@li...
> > https://lists.sourceforge.net/lists/listinfo/colinux-users
> 
> 


-- 
Regards,
Paradise

[coLinux-users] Re: Block device aliasing for dual-boot

[Ian A. <ia...@ab...>]

On 07/10/2004 19:06, Shi, Jue wrote:
> I just upgraded to snapshot (20041006) and found a commend in the default
> XML file the new block device alsasing stuff.
> Is there more detail info about it, especially how to use it for dual-boot
> system and how it works?
> 
> I have setup the colinux successfully to boot into my suse partion and got
> it running. The problem is that the fstab and the network need to be changed
> to get it work. Then these setting need to be changed back in order to boot
> into suse (not through colinux). 
> 
> My question is: can the block device aliasing help in this situation?

I tried block device aliasing for dual-booting Gentoo.  Like 
yourself, I wanted to avoid having two different versions of 
/etc/fstab.  However, in its current form, block device aliasing is 
pretty useless for this task.  For example, if you specify "hda1" as 
an alias name, the current version of the aliasing code creates a 
device called "hda/1" so it ends up with /dev/hda/1 instead of 
/dev/hda1.  Of course, this is incompatible with the normal device 
names when I boot Gentoo natively.

I hope the aliasing code will change to use the standard device 
names.  In the meantime, I use some devfsd rules to create and 
destroy symlinks from the cobd/X device names to the hdaX device names:

REGISTER ^cobd/([0-9]+)$ CFUNCTION GLOBAL mksymlink $devname hda\1
UNREGISTER ^cobd/([0-9]+)$ CFUNCTION GLOBAL unlink hda\1

The "\1" at the end of the rules gets substituted with the cobd 
device number, so in my colinux.xml file I make sure the 
block_device index values match the hda partition numbers. 
(Actually, I use hdc instead of hda on my system.)  An alternative 
would be to have separate rules for each partition.  Something like 
that would be necessary when the cobd devices are spread over more 
than one disk.

On Gentoo, I put those rules into a file called "mycolinux" in the 
/dev/devfs.d directory.  The /dev/devfsd.conf file contains ends 
with an INCLUDE rule that picks up these custom rules:

# Support additional config installed by packages ...
INCLUDE /etc/devfs.d

Then I just have to make sure I set the "root=" kernel option in my 
colinux.xml file to correspond with my root partition.  For example, 
if my root partition was /dev/hda2, I would set root=/dev/cobd/2.

I have various Windows partitions in my /etc/fstab which have no 
corresponding cobd devices defined (for obvious reasons!).  These 
are normally mounted at boot time when I boot Gentoo natively and of 
course fail to mount when I boot through coLinux.  This does not 
stop my Gentoo system booting up though.

I haven't set up networking yet.

> Another question is regarding the initrd file. Is it possible to use it to
> solve my problem above?

I don't think so, but I'm not sure what's in the initrd.  I'll 
probably find out when I try and set up networking!

RE: [coLinux-users] Invalid checksums

[peter g. <plu...@bi...>]

did you try disabling checksum offload?

its the only thing that is currently known to cause problems with winpcap
based bridging

> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of Paradise
> Sent: 10 October 2004 19:02
> To: Xavier Brouckaert
> Cc: col...@li...
> Subject: Re: [coLinux-users] Invalid checksums
>
>
> Hi, my case is same with you, with same network card and bridged
> connection, cannot ssh to colinux from host but my other pc on the
> network has no problem at all.
>
>
> On Sat, 9 Oct 2004 20:31:33 +0200, Xavier Brouckaert
> <xav...@gm...> wrote:
> > Hi,
> >
> > I'm running CoLinux 0.6.2 on Windows 2000 SP4 to run a Debian image.
> > I've got invalid checksums which makes impossible to run an ssh
> > session between the host and colinux.  I'm using a bridged ethernet
> > connection.
> >
> > My configuration is :
> > <?xml version="1.0" encoding="UTF-8"?>
> > <colinux>
> >     <block_device index="0"
> > path="\DosDevices\e:\colinux\Debian-20040605-mit.ext3.1610mb"
> > enabled="true" />
> >     <block_device index="1" path="\DosDevices\e:\colinux\swap_512Mb"
> > enabled="true"></block_device>
> >     <bootparams>root=/dev/cobd0</bootparams>
> >     <image path="vmlinux" />
> >     <memory size="128" />
> >     <!--  <network index="0" type="tap" /> -->
> >     <network index="0" type="bridged" name="Marvell" />
> > </colinux>
> >
> > A sample tethereal trace (host=192.168.1.100, colinux=192.168.1.102)
> > of a putty ssh connection:
> > No.     Time        Source                Destination
> Protocol Info
> >       3 0.008588    192.168.1.100         192.168.1.102         TCP
> >   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
> >       4 0.008769    192.168.1.100         192.168.1.102         TCP
> >   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
> >       5 0.023884    192.168.1.102         192.168.1.100         TCP
> >   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
> >       6 0.023902    192.168.1.100         192.168.1.102         TCP
> >   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
> >       7 0.024081    192.168.1.100         192.168.1.102         TCP
> >   [TCP Dup ACK 6#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
> >       8 0.024373    192.168.1.102         192.168.1.100         TCP
> >   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
> >       9 0.024387    192.168.1.100         192.168.1.102         TCP
> >   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
> >      10 0.024564    192.168.1.100         192.168.1.102         TCP
> >   [TCP Dup ACK 9#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
> >
> > As you can see, there are some packets with an invalid checksum and
> > others with a valid checksum.
> >
> > The network card is a "Marvell Yukon Gigabit Ethernet
> > 10/100/1000Base-T Adapter, Copper RJ-45" running at 100Mbps.
> >
> > It only happens between the host and colinux.  On colinux, I can
> > flawlessly install packages and so on...
> >
> > Any idea why it happens ?
> >
> > Thank you,
> > Xavier Brouckaert
> >
> > -------------------------------------------------------
> > This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> > Use IT products in your business? Tell us what you think of
> them. Give us
> > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> > http://productguide.itmanagersjournal.com/guidepromo.tmpl
> > _______________________________________________
> > coLinux-users mailing list
> > coL...@li...
> > https://lists.sourceforge.net/lists/listinfo/colinux-users
> >
>
>
> --
> Regards,
> Paradise
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users

Re: [coLinux-users] Invalid checksums

[Paradise <par...@gm...>]

Hi, my case is same with you, with same network card and bridged
connection, cannot ssh to colinux from host but my other pc on the
network has no problem at all.


On Sat, 9 Oct 2004 20:31:33 +0200, Xavier Brouckaert
<xav...@gm...> wrote:
> Hi,
> 
> I'm running CoLinux 0.6.2 on Windows 2000 SP4 to run a Debian image.
> I've got invalid checksums which makes impossible to run an ssh
> session between the host and colinux.  I'm using a bridged ethernet
> connection.
> 
> My configuration is :
> <?xml version="1.0" encoding="UTF-8"?>
> <colinux>
>     <block_device index="0"
> path="\DosDevices\e:\colinux\Debian-20040605-mit.ext3.1610mb"
> enabled="true" />
>     <block_device index="1" path="\DosDevices\e:\colinux\swap_512Mb"
> enabled="true"></block_device>
>     <bootparams>root=/dev/cobd0</bootparams>
>     <image path="vmlinux" />
>     <memory size="128" />
>     <!--  <network index="0" type="tap" /> -->
>     <network index="0" type="bridged" name="Marvell" />
> </colinux>
> 
> A sample tethereal trace (host=192.168.1.100, colinux=192.168.1.102)
> of a putty ssh connection:
> No.     Time        Source                Destination           Protocol Info
>       3 0.008588    192.168.1.100         192.168.1.102         TCP
>   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
>       4 0.008769    192.168.1.100         192.168.1.102         TCP
>   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
>       5 0.023884    192.168.1.102         192.168.1.100         TCP
>   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
>       6 0.023902    192.168.1.100         192.168.1.102         TCP
>   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
>       7 0.024081    192.168.1.100         192.168.1.102         TCP
>   [TCP Dup ACK 6#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
>       8 0.024373    192.168.1.102         192.168.1.100         TCP
>   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
>       9 0.024387    192.168.1.100         192.168.1.102         TCP
>   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
>      10 0.024564    192.168.1.100         192.168.1.102         TCP
>   [TCP Dup ACK 9#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
> 
> As you can see, there are some packets with an invalid checksum and
> others with a valid checksum.
> 
> The network card is a "Marvell Yukon Gigabit Ethernet
> 10/100/1000Base-T Adapter, Copper RJ-45" running at 100Mbps.
> 
> It only happens between the host and colinux.  On colinux, I can
> flawlessly install packages and so on...
> 
> Any idea why it happens ?
> 
> Thank you,
> Xavier Brouckaert
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users
> 


-- 
Regards,
Paradise

RE: [coLinux-users] Invalid checksums

[peter g. <plu...@bi...>]

is there a way to disable hardware checksums for your card? 
if so does changing it make any difference


> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...]On Behalf Of Xavier
> Brouckaert
> Sent: 09 October 2004 19:32
> To: col...@li...
> Subject: [coLinux-users] Invalid checksums
> 
> 
> Hi,
> 
> I'm running CoLinux 0.6.2 on Windows 2000 SP4 to run a Debian image. 
> I've got invalid checksums which makes impossible to run an ssh
> session between the host and colinux.  I'm using a bridged ethernet
> connection.
> 
> My configuration is :
> <?xml version="1.0" encoding="UTF-8"?>
> <colinux>
>     <block_device index="0"
> path="\DosDevices\e:\colinux\Debian-20040605-mit.ext3.1610mb"
> enabled="true" />
>     <block_device index="1" path="\DosDevices\e:\colinux\swap_512Mb"
> enabled="true"></block_device>
>     <bootparams>root=/dev/cobd0</bootparams>
>     <image path="vmlinux" />
>     <memory size="128" />
>     <!--  <network index="0" type="tap" /> -->
>     <network index="0" type="bridged" name="Marvell" />
> </colinux>
> 
> A sample tethereal trace (host=192.168.1.100, colinux=192.168.1.102)
> of a putty ssh connection:
> No.     Time        Source                Destination           
> Protocol Info
>       3 0.008588    192.168.1.100         192.168.1.102         TCP   
>   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
>       4 0.008769    192.168.1.100         192.168.1.102         TCP   
>   1163 > 22 [SYN] Seq=0 Ack=0 Win=65535 Len=0 MSS=1460
>       5 0.023884    192.168.1.102         192.168.1.100         TCP   
>   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
>       6 0.023902    192.168.1.100         192.168.1.102         TCP   
>   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
>       7 0.024081    192.168.1.100         192.168.1.102         TCP   
>   [TCP Dup ACK 6#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
>       8 0.024373    192.168.1.102         192.168.1.100         TCP   
>   22 > 1163 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
>       9 0.024387    192.168.1.100         192.168.1.102         TCP   
>   1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 [CHECKSUM INCORRECT] Len=0
>      10 0.024564    192.168.1.100         192.168.1.102         TCP   
>   [TCP Dup ACK 9#1] 1163 > 22 [ACK] Seq=1 Ack=1 Win=65535 Len=0
> 
> As you can see, there are some packets with an invalid checksum and
> others with a valid checksum.
> 
> The network card is a "Marvell Yukon Gigabit Ethernet
> 10/100/1000Base-T Adapter, Copper RJ-45" running at 100Mbps.
> 
> It only happens between the host and colinux.  On colinux, I can
> flawlessly install packages and so on...
> 
> Any idea why it happens ?
> 
> Thank you,
> Xavier Brouckaert
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to 
> find out more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> coLinux-users mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-users