On Feb 10, 2008 1:57 PM, Wolfgang Denk wd@denx.de wrote:
is not in the best of shape). That being said, we boot a
There is a number of MIPS board that are working pretty well - Andrew: what exactly do you mean by "not in the best of shape"?
AFAIK the current low level code is pretty specific to one processor - some Infineon chip used on one of the ports. It kinda works on a few others. There are a number of problems extending that to 'all MIPS' or even just MIPS32 compatible parts: cache init/cache handling, CP0 handling, hardcoded checks for denx toolchain, and probably a few others that I've forgotten.
That said, I mean no disrespect to Shinya and other maintainers, I think Shinya is doing a great job. We've discussed these issues, and it'll just take time and effort to clean up what's there. U-boot is the bootloader for PPC, but for MIPS many people use RedBoot or YAMON or CFE or pmon, depending on what their eval board came with, so it's a smaller user base.
Seems to work for you?
It does, but we started from a patch for the Au1500 that made it a lot easier. I think you also know I've submitted a fair number of bug fixes and patches for stuff along the way.
As far as getting the job done, since he has a platform that already suits most of his needs, it seems like adding basic ATA support and FAT support to that in C code would be easier than trying to work through the low-level issues with the u-boot MIPS code on a proc that is not supported. (Particularly if he doesn't have access to good debugging tools, or is not well versed in the low-level arch details)