Bug 5283 - Incorrect detection of PM-Timer ioport on ASUS a8v
Summary: Incorrect detection of PM-Timer ioport on ASUS a8v
Alias: None
Product: ACPI
Classification: Unclassified
Component: Config-Tables (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: acpi_aml-interpreter
Depends on:
Reported: 2005-09-21 01:08 UTC by Jonas Oreland
Modified: 2005-12-05 14:00 UTC (History)
2 users (show)

See Also:
Kernel Version:
Tree: Mainline
Regression: ---

dmesg and lspci (15.65 KB, application/octet-stream)
2005-09-21 01:10 UTC, Jonas Oreland
patch for handling xpm_pmtmr 0 case (948 bytes, patch)
2005-09-21 01:35 UTC, Shaohua
Details | Diff

Description Jonas Oreland 2005-09-21 01:08:53 UTC
Most recent kernel where this bug did not occur:
Distribution: Gentoo
Hardware Environment: ASUS a8v deluxe
Software Environment:
Problem Description:

I running 2.6.12-gentoo-r10
I have a ASUS a8v motherboard with a AMD64 X2

This gives me problem with clock drift
(see http://bugzilla.kernel.org/show_bug.cgi?id=5105)

So I tried using notsc to force usage of pm timer.

The the acpi code find the fadt->xpm_tmr_blk.address to 0
Making the code yet again fall back to PIT/TSC

I booted my system and run
perch> grep PM_ /proc/ioports 
0808-080b : PM_TMR

And harded the address 0x808 into pmtmr_ioport
This works. The system is using pmtmr and I get no clock drift
(tested now for ~a day)

But why do ACPI find 0 as pmtmr_ioport?
Comment 1 Jonas Oreland 2005-09-21 01:10:29 UTC
Created attachment 6069 [details]
dmesg and lspci
Comment 2 Shaohua 2005-09-21 01:35:14 UTC
Created attachment 6070 [details]
patch for handling xpm_pmtmr 0 case
Comment 3 Len Brown 2005-11-30 19:17:27 UTC
applied to acpi-test tree
Comment 4 Len Brown 2005-12-05 14:00:31 UTC
shipped in linux-2.6.15-rc5 -- closing.

Note You need to log in before you can comment on or make changes to this bug.