Re: [Unattended] unattended 4.3, vmware 4.5.2 -- no combination works

[Joseph D. <jd...@we...>]

Patrick J. LoPresti wrote:

>Is it actually hung, or just very slow?
>
>(So DOSEMU on Linux on VMware on Windows is slow.  Go figure...)
>  
>
It's extremely slow.. It takes more than 5 minutes to print the FreeDOS 
kernel banner (imagine 3-4 seconds between characters)

>It would be nice to know where the problem actually lies.  If you can
>convince anybody else to try DOSEMU on Linux on VMware on Windows,
>maybe the VMware guys could be convinced to take an interest.  Perhaps
>bring it up on some VMware list?
>
>The only idea I have is to edit z:\install\linuxaux\etc\dosemu.conf to
>change this line:
>
>    $_hogthreshold = (0)
>
>...and set it to (1) instead of (0).  This has something to do with
>how much of the CPU DOSEMU will "hog", but the documentation is not
>very detailed.
>
>  
>
I'll see if I can do some more diagnosis on this.. My hunch is that the 
VMware folks broke something, since it supposedly works ok with 3.x...

>That means the VMware SCSI BIOS is not implementing INT13 extensions,
>which I find surprising.  But not impossible.  And you are correct
>that it should not matter for a straight full-disk install.
>
>  
>
I'll hunt around a bit for some VMware INT13 info..maybe theres a 
semi-documented flag to turn them on.

>That already indicates a problem.  Setup should be able to finish
>after winnt.exe reboots.
>
>Did you allow Unattended to automatically partition ("whole disk C:")
>the drive, format it, and replace the MBR?
>
>  
>
Yes to all those questions... I haven't tried any other script options 
(i.e. partition by hand, etc) yet, but am willing to try if that might 
illuminate what's going on..

>This probably means Windows 2000 does not include a driver for that
>SCSI hardware, and you need to use TXTSETUP.OEM and friends to add a
>text-mode driver (http://support.microsoft.com/?kbid=288344).
>  
>
I'm planning to try this to see if I can get any further, but I haven't 
gotten to that point yet...

>I assume you mean VMware detected an access violation committed under
>FreeDOS, not that VMware itself generated an access violation?  If
>VMware itself crashed with a Windows access violation, that is
>unambiguously a bug in VMware and you should report it to them.
>  
>
Actually, it is a windows level access violation.. It doesn't kill any 
other running VMware windows, or even the parent of the one that dies, 
but I do get the whole windows Send bug report/Dont send dialog box.  I 
did throw an email at vmware, but who knows whether they are listening.  
I was mostly curious whether anyone else had seen this behavior.

>FreeDOS does still have some issues with memory management.  Try
>booting from virtual floppy (instead of CD).  Also try editing
>config.sys to use umbpci.sys instead of emm386.exe.  Also try
>commenting out the UMB provider entirely.
>
>  
>
I'll try these things and report back..

Thanks for the info... will keep you posted as to my results..

-- 
Joseph Dickson
Unix Administrator | WEYCO, INC.