Hi,
After the upgrade to rc1 I am not able to install Solaris 9 with the pcn network interfaces anymore. Even installing vmxnet from vmware tools (which are intended for Solaris 10, I know) do not work anymore .. but they used to work previously.
My Solaris 8 images do not boot anymore at all.
This remembers me to my initial experience with parallels which made me switch to vmware. Now I have exactly the same problem with vmware fusion.
(http://forum.parallels.com/archive/index.php/t-4888.html)
It seems I am lost now ..
....
rm_check: top_realmem = 0x0 bot_realmem = 0x99ef0 overlap = 630512
prom_panic: memory soon to be hosed
Entering boot debugger:
....
You know how I could get Solaris 8 and 9 running propery again ?
thanks
axel
Creating and running Solaris 8 update 7 with workstation 6 on windows still makes no problems. Creating a clone on windows (doesn't matter if version 6 or 5 compatibility) and trying to start it with Fusion RC1 leads to the problem I described above.
comments welcome.
regards
axel
WS6 is running on a WS1100 Sun Workstation with AMD Opteron CPU on Windows XP.
Thanks. I will file a bug.
\[edit]
done:
The Solaris 9 networking problems:
Your Support Request Report Number is 192037371.
The Solaris 8 install and run issue:
Your Support Request Report Number is 192052284.
Axel
Message was edited by:
axelklatt
WS6 is running on a WS1100 Sun Workstation with AMD
Opteron CPU on Windows XP.
Any chance you could try Solaris 8 on an Intel+WS6 to see if the boot problem is processor-specific? And to confirm, Solaris 8 was working for you at one point with Fusion (which beta)?
I wanted to update this thread on the Solaris 8 problem - the answer is to have the guestOS set as "solaris8", not solaris9.
I'm not sure about the Solaris 9 networking problem.
Thanks!
Changing guestOS to "solaris8" makes the boot error from Solaris 8 go away. But during my testing I was not able to use the pcn ethernet driver with Solaris 8 anymore. The behaviour looks identical to the problem I have with Solaris 9 since Fusion RC1.
During device identification Solaris 8 detects the PCI: AMD 79C970 PCnet Ethernet adapter but is not able to plumb the driver. Either by trying a new install or taking an older VM
Just curious is that I had \*one* Solaris 8 image (out of five) where the pcn driver was working after changing the guestOS value .. I have to look at that closer tomorrow and see what the difference is.
\[edit]
this working image also does not work anymore after reboot.
With guestOS = "solaris8" I can now boot. But I have networking
problems. I cannot plumb the pcn driver for the virtal AMD nic ..
trying to access the device with ifconfig returns ENXIO aka "no
such device or address"..
I tried to install Solaris 8u7 on a MacBook Pro and on a MacPro
==> same result.
i tried to run older images
==> same result
I tried to manually make the device accessible within Solaris
with devfsadm .. eiditing path_to_inst and driver_aliases but
no success. Solaris sees the hardware but the driver does not
wor anymore ...
Problem looks similar to the one I reported for Solaris 9 not
working with pcn innterface anymore. Regarding that Solaris 9
problem I have to look closer.
I just can tell that I had no such problems with the Fusion beta versions before Fusion RC1. And no problems with windows workstation 5 + 6 in combination with Solaris 8 + 9 ever.
best regards
axel
Message was edited by:
axelklatt