By Thomas.Lange@corelatus.se 2004-Oct-05
|
----------------------------------------
|
DbAu1xx0 are development boards from AMD containing
|
an Alchemy AU1xx0 series cpu with mips32 core.
|
Existing cpu:s are Au1000, Au1100, Au1500 and Au1550
|
|
Limitations & comments
|
----------------------
|
Support was originally big endian only.
|
I have not tested, but several u-boot users report working
|
configurations in little endian mode.
|
|
I named the board dbau1x00, to allow
|
support for all three development boards
|
( dbau1000, dbau1100 and dbau1500 ).
|
Now there is a new board called dbau1550 also, which
|
should be supported RSN.
|
|
I only have a dbau1000, so my testing is limited
|
to this board.
|
|
The board has two different flash banks, that can
|
be selected via dip switch. This makes it possible
|
to test new bootloaders without thrashing the YAMON
|
boot loader delivered with board.
|
|
NOTE! When you switch between the two boot flashes, the
|
base addresses will be swapped.
|
Have this in mind when you compile u-boot. CONFIG_SYS_TEXT_BASE has
|
to match the address where u-boot is located when you
|
actually launch.
|
|
Ethernet only supported for mac0.
|
|
PCMCIA only supported for slot 0, only 3.3V.
|
|
PCMCIA IDE tested with Sandisk Compact Flash and
|
IBM microdrive.
|
|
###################################
|
######## NOTE!!!!!! #########
|
###################################
|
If you partition a disk on another system (e.g. laptop),
|
all bytes will be swapped on 16bit level when using
|
PCMCIA and running cpu in big endian mode!!!!
|
|
This is probably due to an error in Au1000 chip.
|
|
Solution:
|
|
a) Boot via network and partition disk directly from
|
dbau1x00. The endian will then be correct.
|
|
b) Partition disk on "laptop" and fill it with all files
|
you need. Then write a simple program that endian swaps
|
whole disk,
|
|
Example:
|
Original "laptop" byte order:
|
B0 B1 B2 B3 B4 B5 B6 B7 B8 B9...
|
|
Dbau1000 byte order will then be:
|
B1 B0 B3 B2 B5 B4 B7 B6 B9 B8...
|
