Distribution: Gentoo Hardware Environment: see attachment Software Environment: Gentoo kernel 2.6.19 Problem Description: * recieving APIC error 40(40) (unless I use "noapic") * occasional overheating (not too severe though) * system hangs * inability to resume from a suspend ...which drove me that far that I tried what's wrong with my BIOS with the Linux-Ready Firmware Developer Kit. I'm attaching the full report of it (includes DSDT, dmesg, lcpci et al.)
Created attachment 11537 [details] Results of the Linux-Ready Firmware Developer Kit LiveCD Tarball of the full report (dmesg, dsdt, at al.) from the Linux-Ready Firmware Developer Kit LiveCD.
Created attachment 11538 [details] DSDT patch This is the DSDT patch (BIOS version 1.20) which I made myself.
I actually do not know anymore what is the cause of which of my problems - the deeper I dig, the less I know. I hope at least someone can benefit from my trouble.
>* recieving APIC error 40(40) (unless I use "noapic") It's not unusual that a up platform has a broken APIC table. You'd better keep on using "noapic" to keep your system stable. >* occasional overheating (not too severe though) >* system hangs These may be the same problem. Does this still happen with "noapic"? >* inability to resume from a suspend You probably mean suspend to memory, don't you? :) Could you please give a detailed description?
With 'noapic' I don't get system hang-ups anymore, but the overheating stays. Since I started using 'noapic' again, at one poing my laptop actually powered itself off to prevent overheating (there was that nice "toasted electronics" smell around it as well). I think I can at least partially blame the overheating on Acer's bad design of the cooling system though. If I lift my laptop just 2 cm of the table and use it in that position it doesn't overheat anymore (enough to make problems). So now I'm using it with a small piece of wood underneath it to make space for the air to flow. (yea, looks as stupid as it sounds) Primarily I meant suspend to ram, yes. Sorry that I wasn't specific enough, but at the point when I posted this bug I was so overworked by these problems that I just didn't bother to write it the n^th time in that week. Suspending to ram seems to work, but it won't wake up (blank screen, disk spinning, no progress anywhere). Suspending to disk doesn't even finish and restarts the system. But I don't really intend to suspend to disk anyway, so I didn't even bother testing it much.
>Suspending to ram seems to work, but it won't wake up (blank screen, disk >spinning, no progress anywhere). >Suspending to disk doesn't even finish and restarts the system. Are you using a SATA disk? Are there any BIOS options about SATA mode? Please attach the kernel config file. >at one poing my laptop actually powered itself off to prevent overheating Please attach the result of "#cat /proc/acpi/thermal_zone/*/*".
Created attachment 11757 [details] .config from 2.6.19 (Gentoo)
Created attachment 11758 [details] cat /proc/acpi/thermal_zone/*/*
OK, odd. Suspending to RAM and resuming worked just now perfectly. It seems like I did something crucial but didn't realise it in the time since I posted this bug. Scratch the suspend to RAM problem then.
So, we only have the overheating problem here, right? Can you try a recent kernel and see if the problem still exists?
I'm fairly happy with the the current kernels (>2.6.20), so I think this bug can be closed. Probably off topic: I keep getting kernel messages that I should report this: " PCI: Transparent bridge - 0000:00:14.4 PCI: Bus #07 (-#08) is hidden behind transparent bridge #06 (-#06) (try 'pci=assign-busses') Please report the result to linux-kernel to fix this permanently " Does this go under a separate bug? And if so, should the summary include the laptop model or should I just append my "I second to that" on an already existing bug?
Re: PCI: Bus #7 yes, different issue merits different bug report. In that case it would be against PCI, rather than ACPI. Re: overheating temperature: 61 C critical (S5): 102 C passive: 82 C: tc1=2 tc2=5 tsp=300 devices=0xffff81000183ea90 ACPI doesn't control the fan on this system. If you have an over-heating problem, please verify that you also have the same over-heating problem with acpi=off. Re: overheating if the fan is not spinning when the systems gets hot, then try upgrading the BIOS and resetting SETUP defaults. If the fan _is_ spinning when the system gets hot, then your fan grill is probably clogged and the dust needs to be blown out of it.
if you find that the system has problems only without acpi=off, then please re-open.