On 16-05-12 19:20, Theo de Raadt wrote: > A good way to get non-working machines to work, is to offer it to > someone who can make it work.
Of course, I did, but the people I contacted have no time or have probably no time. > The wrong way is to dangle it as a reward for whoever makes it work, > when they don't have the hardware. If they had the hardware, they'd > make it work. See original posting, we are willing to send a machine to a person with enough indications that this person might be able to fix within a not too long time frame, no problem. If for example you say "I presume I can fix this in about 3 weeks", please send me the street address and you will have the machine there within a few day's. > If they had the hardware, maybe they don't want another. See the original posting, when someone has it working already and is willing to send us working code, and that might well be the case, that person may choose where we should sent the machine to, or money, I don't care. +++chefren (In the mean time we decided we don't need the 32 bit implementation anymore, 64 bit only is OK for us.) >> On 16-05-12 18:48, Benny Lofgren wrote: >>> Just out of curiosity I'd like to see dmesg:s on the machine to see what >>> works and what doesn't. It would probably also make it easier for those >>> thinking about taking up the offer to assess its feasibility. >> >> In general a good question, but since the Intel i350 is not supported >> yet and the interrupt routing seems a mess this is not so appropriate. >> >> Boot from the (i386) install CD gives: >> >> ioapic0 at ... >> ioapic1 at ... >> acpimcfg0 at acpi0 addr 0x0...0, bus, 0-255 >> panic: pci_make_tag: bad request >> >> >> 'boot -c' with 'disable acpi' "works". >> >> >> Booting from AMD-64 gives about the same chaos. >> >> +++chefren
