Re: [ntdev] Re: [ntdev] Re: [ntdev] One instance of UMDF device

On Dec 4, 2014, at 3:53 PM, Marion Bond > wrote:

It does beg the question though of why we have all given up on rings except 0 and 3. There are of course good reasons to despise them, especially pert, but the whole objective of UM drivers seems to be emulation of this concept

The Control Data Cyber 180 mainframes supported 16 rings. The NOS/VE operating system that ran on them actually used 10 of them. There were very elaborate standards for deciding what ran where. User code ran in ring 13. If you wanted to cal into a subsystem (in ring 6), you had to go through a proxy API in ring 12. That managed the transition inward. Overall, it didn’t seem to be worth the trouble.

That is, however, where I first encountered the “inverted call” concept.

Tim Roberts, xxxxx@probo.commailto:xxxxx
Providenza & Boekelheide, Inc.</mailto:xxxxx>