We've run into this too with 7.3.2. Even a cifs terminate does not always let you remove the share (it basically hangs). We've had to fail the head to get it to stop. I am not sure what the problem is but it seems to happen on our busiest volumes only (maybe 1k concurrent CIFS users).
I think I hit this when I upgraded a filer to 8.0.1. I installed the 8.0.1 (download took a long time) and when I ran halt -t 0 CIFS failed to terminate, so the filer would not reboot. I tried "cifs terminate -t 0" and got the same error (CIFS not terminating). I finally had to halt -d and when the filer came back up I had to wait a long time for the core file to be saved. And then I saw that it was still running 7.3. Not knowing how to boot the 8.0.1 kernel, I quickly ran cifs terminate -t 0 (which worked, thankfully) and installed 8.0.1 again (and sat through that long download again). Then I was able to reboot and come up running 8.0.1. Whew! But that downtime was a LOT longer than we expected. The core dump and second download took about an hour.
Steve Losen scl@virginia.edu phone: 434-924-0640
University of Virginia ITC Unix Support