Michael Lotz wrote (2007-12-11, 13:15:56 [+0100]): > On Tue, 11 Dec 2007 12:06:13 +0100 (CET), Marcus Overhagen wrote > > It is unnecessary complicated and stalls the entire kernel driver > > development process. > > Yes, I fully agree. I have written a it8211 IDE controller driver for my > "legacy free" mainboard that only contains a non-standard IDE controller. > Works fine under R5 and supports DMA (have to publish that to BeBits some > time). Getting it to work under R5 required some extra work, but was > after all pretty easy to do. I then tried to port it to Haiku, but I > absolutely couldn't find out how the whole stuff works. How and when the > disk driver is supposed to load the controller, how controllers/channels > communicate with each other, etc.. This is the second time this has > happened to me, the first time was when I tried to port the usb_scsi to > Haiku. This is still not done and I don't really blame Siarzhuk (who has > taken over the task) as I've been there asking myself how everything is > supposed to work. I cannot comment on the subject from own experience, but I trust you guys on this. To what you have said, I could add that the original author/maintainer of this code is sadly not arround anymore. I would like to hear some input from Axel and Ingo though, since they seem to be most familiar with the code of the device manager. Best regards, -Stephan