Here is a common problem seen after swapping a graphics card for a different model. X.Org fails to start with the message:
(EE) NVIDIA(0): Failed to initialize the NVIDIA graphics device! (EE) NVIDIA(0): *** Aborting *** (II) UnloadModule: "nvidia" (EE) Screen(s) found, but none have a usable configuration.
This is because Solaris "remembers" the previous model that was installed and increments the instance numbering. For example, if an NVS 280 was the original card and then it is replaced with an FX 1100, the configuration reboot will create the following device links:
/dev/fbs/nvidia0 ----> NVS 280 instance /dev/fbs/nvidia1 ----> FX 1100 instance
The easiest way to solve this is remove the "nvidia" line(s) from the file /etc/path_to_inst before doing the configuration reboot:
# reboot -- -r
at the boot prompt. If you have already done the configuration reboot, just modify the file and do the configuration reboot again.
Note that corrupting this file can stop your computer from booting properly, in which case a boot -a must be performed, causing the file to be recreated from scratch.
Please also note that if you upgraded to an NVIDIA graphics card
released more recently than the NVIDIA Accelerated Solaris Graphics
Driver Set installed on your system, its PCI device ID may not have
been added to the file
/etc/driver_aliases. Please see Chapter 7, Common
Problems for more information.