logical regions vs page table granularity 
Author Message
 logical regions vs page table granularity


I am trying to understand some of the finer details with
regards to memory management.  I just wanted to make sure
my assumptions were correct.

For an XScale processor:

If CE is using "course" second level, page table entries
with "small" page references (assumption one) and CE
commits/allocates memory in 64KB "regions" does this mean
that bits 15:0 of the virtual address are zeroed and only
bits 16:19 are used to index the second level page table?  
Thus, only having a maximum of 2^4 = 16 unique entries?

Or does CE actually use "large" page table references?

I guess I'm trying to understand the correlation between
the logical 64KB "regions" and the actual physical page

Thanks in advance for your patience with the "fresh fish"

Chris Kavanagh

Symbol Technologies - Winnipeg, Canada

Sat, 28 May 2005 02:13:57 GMT  
 [ 1 post ] 

 Relevant Pages 

1. client region, frame region, windows region?

2. extern int table[] vs extern int *table;

3. Message Table vs String Table

4. Printing -Defining page region problem?

5. Printing -Defining page region problem?

6. Large nested if clauses vs. many logical ANDs

7. Any equivalent to #region for VS C++ ??

8. Question regarding apartments vs synchronized regions in CLR

9. Granularity of bit field accesses

10. malloc alignment granularity

11. System allocation granularity

12. PowerPC MMU Page Table Creation


Powered by phpBB® Forum Software