451061@UOTTAWA.BITNET (Valentin Pepelea) (03/16/89)
Gee, I was fooling around with address translation tables on the 68851 when
suddently a blue Jeep struck me.
Using only about 20 page descriptors, I first initialized the translation table
and loaded the CRP appropriately. The used bit then of course got set on some
descriptors. The weird thing at this point is that when I changed those
descriptors with metascope back to their initial value, their used bit got set
once again. Now this weird because the descriptors are supposed to be in the
translation cache, and not modified any more. So the question is:
1. Does the 68851 realize that the address just accessed contains a page
descriptor and therefore flushes that descriptor from the ATC? The 68851
User s Manual does not say this happens.
OR
2. Does the Exec somehow cause the cache to flushed on task switches? I did not
see that happen in my Exec disassmebly.
Valentin
_________________________________________________________________________
"An operating system without Name: Valentin Pepelea
virtual memory is an operating Phone: (613) 233-1821
system without virtue." Bitnet: 451061@uottawa
Usenet: Look at the header
- Ancient Inca Proverb Planet: Ontario!Canada!Earth