} For the MBX in particular, the support of the PC-style I/O has } finally been broken. The support of the IDE was always a problem } because of the desire to use PCMCIA flash, the QSpan never seemed } to pass interrupt correctly, and now generic kernel changes finally } broke the indirect access to the 8259 (and I am still puzzled how } it is supported on other PMacs.....).
For one, 8259 isn't supported on PMacs :) I'd found when I first did IDE on the MBX that there was a lot of contradictory information in the docs on how those interrupts were passed through. Matt Porter at Moto finally got me the correct info (not on Moto time though, I don't want to get him into trouble...). If you've found it's not working correctly please let me know. I don't claim my understanding or implementation is full or completely correct. Can you tell me the 8259 problems you're having? I'd like to fix those at least. } Anyway, those are some of the things I am working on. There are } many other little things, like building without PCI configured } doesn't work either. The last working 2.3.xx kernel I used and } updated was 2.3.18. } } I am actively working on all of these because I want to make } sure we have a solid 8xx configuration when 2.4.x hits. } } I will make an announcement when I get something stable again. Please keep me up-to-date with what you do. I'm trying to get things to a semi-stable state on all platforms for 2.4. ** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/
