I've got an 8-bit Ether3, so I've discovered that the address mapping is
non-linear. Comment it here so I don't forget.
This commit is contained in:
parent
5dc1e8334c
commit
ddc6e8e6ad
|
@ -1,4 +1,4 @@
|
|||
/* $NetBSD: if_eareg.h,v 1.2 2001/03/25 01:08:56 bjh21 Exp $ */
|
||||
/* $NetBSD: if_eareg.h,v 1.3 2001/06/15 00:28:01 bjh21 Exp $ */
|
||||
|
||||
/*
|
||||
* Copyright (c) 1995 Mark Brinicombe
|
||||
|
@ -39,6 +39,17 @@
|
|||
* Atomwide Ether3 Register Definitions
|
||||
*/
|
||||
|
||||
/*
|
||||
* The 8005 (or 80C04 or 80C04A) in an Ether3 is wired up as follows:
|
||||
* A0 -- LA[2]
|
||||
* A1 -- LA[6]
|
||||
* A2 -- LA[7]
|
||||
* A3 -- LA[8]
|
||||
*
|
||||
* Happily, we can ignore A0 in 16-bit mode, but it'll come back to bite
|
||||
* us on 8-bit versions.
|
||||
*/
|
||||
|
||||
#define EA_8005_BASE 0x000
|
||||
#define EA_8005_SHIFT 5
|
||||
|
||||
|
|
Loading…
Reference in New Issue