John G Dorsey wrote:
>On Wed, 5 Jul 2000, Brad Parker wrote:
>> #define _PA_DDR _SA1111( 0x1000 )
>...
>>
>> ? All the other defines match up with my docs correct, but the GPIO
>> one maps to A12:9=8 and my docs say it should A12:9=7 (or 0x0e00).
>
>These came from section 10.3.6 ("Memory Map") of the SA-1111 developer's
>manual; this section doesn't seem to have been modified in the manual
>update. Since the block A GPIO drives the MAX1600 on Neponset, and that
>actually appears to work with the current definitions, I've felt
>comfortable believing that these macros were correct (at least in the
>patch I submitted to Nicolas; I haven't looked at the last few -np
>releases).
>
>Which part of the documentation seems in conflict?
Well, looking back at the docs now I'm totally confused.
Yes, you are correct, 10.3.6 does say it should be at 0x1000.
However, The SA-1111 developer's manual, pg 3-4 shows the GPIO at
0x1000-0x11ff (it also shows the OHCI controller at 0x400, which is
what I really wanted)
Page 4-2, however ("RAB Target Blocks/Address Assignment") shows a table
where "Address A<12:9>" is 2 for the USB (this makes 0x400, which seems right)
and 7 for the GPIO (which makes 0x0e00, *not* 0x1000). Further, it shows
8 (which makes 0x1000) as "reserved".
In the "Specification Update October 1999", page 21, Figure 4-1 is
updated and matches the "RAB Target Blocks/Address Assignment" table...
I'll go with 10.3.6 :-)
-brad
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
++ Please use [EMAIL PROTECTED] for ++
++ kernel-related discussions. ++
