Bugzilla – Bug 113099
ACPI - AML Alias command is broken
Last modified: 2006-02-03 07:38:31 UTC
With my laptop AVERATEC AV3360, I got the following error messages from ACPI during boot: exstoren-0142: *** Error: Store into Alias - should never happen psparse-1172: *** Error: Method execution failed [\_TZ_.RTMP] (Node df7ae528), AE_AML_INTERNAL psparse-1172: *** Error: Method execution failed [\_TZ_.THRM._TMP] (Node df7ae2e8), AE_AML_INTERNAL
Created attachment 47647 [details] Full kernel log
This seems to be related to (#75808). In this case the table seems to be detected as a duplicate, but it is not? Can you send acpidmp output to be sure, please.
Created attachment 48573 [details] acpidmp output
The machine is reachable as alsa3.suse.de (using standard root).
This is not related to (#75808 - avoid double loading of ACPI tables). In fact avoid loading of ACPI tables seem to work quite well (seems as if ppc_acpi module also triggers a table load which already has been loaded). As far as I can tell for now the alias ASL command is broken in kernel. Try to find a fix for the DSDT (there are a lot more, but not sever things broken) and then try to find a solution for the kernel. Unfortunately ACPI_DEBUG is off again -> I probably will need to install a Beta3 kernel for debugging ...
*** Bug 114692 has been marked as a duplicate of this bug. ***
Oops, still I got a problem with the fixed DSDT.aml. It doesn't boot when no power cable is plugged in at the boot time. It shuts down with critical temperature (255C).
Not sure whether this is still an error resulting from broken aliases, but it could and we should wait until the alias commmand is fixed. According kernel.org bugzilla: http://bugzilla.kernel.org/show_bug.cgi?id=5189
Status update: As described in Bug #114692, the thermal_zone appeared again when I reloaded the thermal module. Until rc4 (2.6.13-14), it worked in this way. But, on rc4 with 2.6.13-15 kernel, it reports always "0C", or fails. It seems that the acpi-temperature-poll patch broke something. The question is: Is it safe to use this kernel even with such a state? I'm currently useing 2.6.13-14 again.
Correction: 2.6.13-14 kernel also fails to show temperature sometimes. And it spits the error messages: ACPI-0142: *** Error: Store into Alias - should never happen ACPI-1174: *** Error: Method execution failed [\_TZ_.RTMP] (Node df7b7760), AE_AML_INTERNAL ACPI-1174: *** Error: Method execution failed [\_TZ_.THRM._TMP] (Node df7b7640), AE_AML_INTERNAL ACPI-0142: *** Error: Store into Alias - should never happen ACPI-1174: *** Error: Method execution failed [\_TZ_.RTMP] (Node df7b7760), AE_AML_INTERNAL ACPI-1174: *** Error: Method execution failed [\_TZ_.THRM._TMP] (Node df7b7640), AE_AML_INTERNAL ACPI-1048: *** Warning: Failed to acquire semaphore[df7b7f40|1|0], AE_TIME ACPI-1048: *** Warning: Failed to acquire semaphore[df7b7f40|1|0], AE_TIME
More status update: The situation gets complicated. I played with the given DSDT again, and it worked now with ec_burst=0. 2.6.13-15 still spits the warnings about to acuire semaphore at each second, so I downgraded 2.6.13-8 again. The only problem is that now the suspend-to-ram doesn't work, and the suspend-to-disk doesn't shut out the power after writing pages. Both worked well without DSDT...
There is a big bundle of ACPI patches coming into mainline after 2.6.16-rc1. I hope this one could fix it: commit 96db255c8f014ae3497507104e8df809785a619f Author: Bob Moore <robert.moore@intel.com> Date: Wed Nov 2 00:00:00 2005 -0500 [ACPI] ACPICA 20051102 ... Implemented support for "local" internal ACPI object types within the debugger "Object" command and the acpi_walk_namespace() external interfaces. These local types include RegionFields, BankFields, IndexFields, Alias, and reference objects. ... Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Len Brown <len.brown@intel.com> I tell you when the stuff is in and it's worth a test. Also search for a BIOS update before testing, or after testing and it is still broken...
I tried the kernel on ~trenn/Export/2.6.*, but the problem seems still remaining. I guess it's really a broken DSDT.
Both problems should be fixed in version ACPICA 20060127 -> see http://bugzilla.kernel.org/show_bug.cgi?id=5189 Those probably were not included yet..., ahhh so much to do, I will have a look whether a patch pops up in Len's ACPI test tree the next days..., give me a reminder at the end of the week if I haven't ansered...
Could you retry with the current kotd, please. According to Intel people it should be fixed now. Tell me if you still have this problem or if other problems pop up.