On Fri, Apr 04, 2008 at 10:50:37AM +0000, Nicolas Legrand wrote: > Raimo Niskanen <[EMAIL PROTECTED]> writes: > > > On Thu, Apr 03, 2008 at 04:38:08PM +0000, Nicolas Legrand wrote: > >> Reyk Floeter <[EMAIL PROTECTED]> writes: > >> > >> > > >> > i have the same problems on a HP Compaq dc7600 Convertible Minitower. > >> > acpi doesn't work. the acpi developers like marco@ are aware of the > >> > problem and it is being worked on. > >> > >> Excellent news! I have the same problem on a dc7700, though I don't > >> disable the whole acpi, but acpiprt* on /bsd or acpiprt* and acpimadt0 > >> on /bsd.mp. > > > > That does not cut it for me (hp2510p ultra-portable laptop). > > > > I get a kernel panic in aml_evalnode in acpi_inidev in aml_find_node > > in acpi_attach. Any hint on what I should disable that is smaller > > than the whole acpi, or how to find out a smaller disable set? > > > > I have tried to disable all acpi... except acpi0 but that did > > not help. Is it acpi0 itself that is the problem or can > > there be some acpi... device(s) that it is sufficient do disable? > > It takes time to try all combinations, it would be nice to > > know if it is useless. > > As Reyk Floeter said it's just a workaround I've found for the HP > dc7700. I've found it by reading the kernel message as written in > description section of > > http://cvs.openbsd.org/cgi-bin/query-pr-wrapper?full=yes&numbers=5738 >
Thank you that was informative! I see that your trace goes from config_attach -> acpi_attach -> acpi_foundprt -> config_attach -> acpiprt_attach -> ... parsing ... _aml_die -> panic while mine goes config_attach -> acpi_attach -> acpi_inidev -> ... parsing ... _aml_die -> panic So it indeed seems like the problem is in acpi(4) itself, not in any attached device. I would be wasting my time trying to disable all combinations of acpi... devices. > the bsd.mp the kernel was hanging with acpiprt* disabled. So I had a > look at apropos acpi and look at the (4) section of the answers before > trying to disable acpiadmt0. > > I'm sorry, this is the best I can do, I'm not (yet I hope) skilled > enough to give solutions and _I don't fully understand the > consequences of what I do_. I just know I can now stop my machine > without being physicaly with it, and that's usefull. > Alas, I was hoping for something like that. : : -- / Raimo Niskanen, Erlang/OTP, Ericsson AB