When I make changes to the driver I am developing (on win2k), I use the
following steps to update it.
1 - build the driver in its local directory
2 - from device manager, disable the driver
3 - copy the .sys to c:\winnt\system32\drivers
4 - from device manager, enable the driver
5 - run my app that opens the driver
My issue is that, sometimes, for no obvious reason, Win2k decides that the
system needs a reboot to either disable or enable the driver. Sometimes it
works fine with no reboot required. I can’t see any dependencies in my
driver that would cause this.
I expect to be reprimanded by someone out there for not using the “legal
way”;), but I really hate the Device Manager wizards, and obviously I need
to make sure the driver is unloaded before copying the new one into the
standard driver directory. The wizards don’t always do what I expect -
So, while any non-wizard suggestions to updating drivers would be
appreciated, I was mainly wondering if anyone knew the exact criterion that
the os uses to decide whether a reboot is required when disabling/enabling a
driver from Device Manager. (I make sure there are no current references to
the driver when I am doing this.)
thanks,
Jeff Larson