Created attachment 23680 [details] output of lspci -vvv after 8 days of uptime, i've notice that dmesg filled up with the following messages. ------------- ACPI Error (psparse-0537): Method parse/execution failed [\_SB_.PCI0.LPC_.SMBR] (Node f7013c48), AE_AML_INFINITE_LOOP ACPI Error (psparse-0537): Method parse/execution failed [\_SB_.PCI0.LPC_.INIT] (Node f7013c30), AE_AML_INFINITE_LOOP ACPI Error (psparse-0537): Method parse/execution failed [\_GPE._L00] (Node f70101b0), AE_AML_INFINITE_LOOP ACPI Exception: AE_AML_INFINITE_LOOP, while evaluating GPE method [_L00] 20090521 evgpe-568 ------------- (the above messages keep on repeating) CPU info: ----------- #/proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 28 model name : Intel(R) Atom(TM) CPU 230 @ 1.60GHz stepping : 2 cpu MHz : 1599.996 cache size : 512 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss tm pbe nx lm constant_tsc up arch_perfmon pebs bts pni dtes64 monitor ds_cpl tm2 ssse3 cx16 xtpr pdcm movbe lahf_lm bogomips : 3193.90 clflush size : 64 power management: ----------- the output of lspci -vvv has been attached. between, i didn't seem to notice any issue related to ACPI though.
PS. the kernel is a stock archlinux kernel
Ming, I think we already have a fix for this issue, don't we?
I don't remember there is a fix for this. Nelson, please attach the acpidump file.
Created attachment 23712 [details] acpidump
it may be expected, i want to say that same messages still appear on kernel 2.6.31.6
Created attachment 23796 [details] custom dsdt 00:1f.3 SMBus: Intel Corporation 82801G (ICH7 Family) SMBus Controller (rev 01) Region 4: I/O ports at 3000 [size=32] The SMBus port address is 0x3000 as above, but the AML code defined it to 0x2000 OperationRegion (SMB1, SystemIO, 0x2000, 0x17) Field (SMB1, AnyAcc, NoLock, Preserve) { HSTS, 8, Offset (0x02), HCTL, 8, HCMD, 8, TSA, 8, HD0, 8, Offset (0x0D), AUXC, 8 } So SMBR method loops infinitely at below While Method (SMBR, 2, NotSerialized) { .... While (LNot (And (HSTS, 0x1E, Local0))) { Store (0xEE, IO80) } .... } This is a BIOS bug, please try the attached custom DSDT. See http://www.lesswatts.org/projects/acpi/overridingDSDT.php for detail. You can just start from step 5.
actually, i wonder would there be any potential problem and what kind if i do nothing against the bug?
between, i have one IDE hard disk that if i plug it in on the motherboard, the system won't even pass POST and hang there forever. would this be about the same bios bug?
There maybe potential problem, but not sure what kind. From your lspci data, 0x2000 is the I/O ports of ethernet controller. IDE hard disk: this seems another bug.
Nelson, Could you help to confirm if the custom DSDT in comment #6 works? Thanks.
""" [root@hades ~]# dmesg|grep DSDT ACPI: Override [DSDT-D945GLF ], this is unsafe: tainting kernel ACPI: DSDT @ 0x3f6f7000 Table override, replaced with: ACPI: DSDT c140f780 0434F (v01 INTEL D945GLF 00000067 INTL 20091013) ACPI: EC: Look up EC in DSDT """ does this mean that i'm using the the custom DSDT?
Yes, and does it work?
so far it seems fine, but only 21 hours uptime. i will check again tomorrow and attach a dmesg
between, can this custom DSDT be used with 2.6.32 ?
Created attachment 24163 [details] dmesg output captured with the custom DSDT applied and with 2 days uptime
i don't see the error message so far. the custom DSDT is probably working
(In reply to comment #14) > between, can this custom DSDT be used with 2.6.32 ? Yes.
*** Bug 14823 has been marked as a duplicate of this bug. ***
after 8 days of uptime, dmesg is fine. i think we can assume that the custom DSDT works. thanks
Close. BIOS should fix it.
as the address is incorrect in the AML, this is not a Linux bug. closed as a documented BIOS bug.
I am having the exact same problem with Debian8/Linux 3.16 and the same type of motherboard with latest BIOS firmware. Is this DSDT "hack" still valid/legit/current?