On Wednesday 03 November 2004 16:36, Alessandro Amici wrote: > > > [...] it looks like an irq problem, the fact that the d-i uses a > > > 32bit kernel on a 64bit can well be to be the trigger, > > > > A similar problem was fixed for PReP systems a while ago. Before > > that, I worked around it by setting CONFIG_SCSI_SYM53C8XX_IOMAPPED. > >
i recompiled a 2.6.8 power3 kernel with CONFIG_SCSI_SYM53C8XX_IOMAPPED=y and managed to load the modified sym53c8xx driver, without any change whatsoever to the (buggy) behavior :( since i had the chance to go on a bit with the hardware detection process before loading the SCSI driver, i noticed that the IDE interface fails claiming irq problems, as well. after that i tried, with the help of Sven Luther, to produce a 2.6.8 ppc64 kernel able to start the d-i. we finally succeeded and with the 64 bit kernel the problem disappears completely. according to a conversation with leighbb and benh on #ppc64 the irq handling code in ppc64 has been changed recently. so i'm inclined to think that the differences in the irq handling between ppc and ppc64 are the root of the problem. cheers, alessandro
