Bill: On Thu, 2004-06-03 at 20:14, Bill Hargen wrote:
I have been trying to get u-boot and Linux running on the IBM 440GX "Ocotea" reference board and I have been hitting some weird problems.
<SNIP>
The first problem is if I try to run u-boot from flash with a BDI2000 connected (i.e. do "reset" and "go" commands), then u-boot starts running but hangs in the SDRAM init. Specifically, it hangs waiting for the MRS Command Complete status bit to be set.
IIRC I ran into the same thing, all I needed to do was remove the sdram configuration from the config for the BDI.
And also, with the BDI2000 disconnected, the system reset switch doesn't work. It kills the system, but it doesn't restart - I have to power cycle. It did work when IBM's PIBS code was in flash.
There is a patch in Wolfgang's queue to fix this. Check the archives for "flash reset ocotea".
- It reports about 37 BogoMIPS, which seems way to low for a 440GX running at 533MHz.
I've seen this also. However the fix escapes me at the moment. I'll try to dig up my fix for this.
- It gets read timeouts trying to find the Ethernet PHYs. (And this works from u-boot - I'm booting from emac0!)
- It isn't getting the boot command line from u-boot. And this is
You're bootargs and settings are making it to the kernel.
I don't know if the BDI2000 quirk, the reset switch quirk, and the Linux strcpy(cmd_line... quirks are connected or not.
Not at all.
-travis