https://bugzilla.novell.com/show_bug.cgi?id=748616
https://bugzilla.novell.com/show_bug.cgi?id=748616#c31
Bjorn Helgaas changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #484724|0 |1
is obsolete| |
--- Comment #31 from Bjorn Helgaas 2012-04-05 23:11:14 UTC ---
Created an attachment (id=485112)
--> (http://bugzilla.novell.com/attachment.cgi?id=485112)
debug patch
I transcribed this from the comment #30 video:
pci_assign_unassigned_resources: updating hardware
pci_bus 0000:00: pci_enable_bridges:
pci_bus 0000:00: pci_enable_bridges: enabling 0000:00:01.0
pci 0000:00:01.0: pci_set_power_state: state 0
pci 0000:00:01.0: pci_set_power_state: returning 0
pci_bus 0000:01: pci_enable_bridges:
pci_bus 0000:01: pci_enable_bridges: done
pci_bus 0000:00: pci_enable_bridges: enabling 0000:00:1e.0
pci 0000:00:1e.0: pci_set_power_state: state 0
pci 0000:00:1e.0: pci_set_power_state: returning 0
pci 0000:00:1e.0: setting latency timer to 64
pci_bus 0000:02: pci_enable_bridges:
pci_bus 0000:02: pci_enable_bridges: enabling 0000:02:04.0
pci 0000:02:04.0: pci_set_power_state: state 0
Switching to clocksource acpi_pm
initcall 1_acpi_pm_clocksource_async+0x0/0x162 returned 0 after 51641 usecs
<hang>
We set the bus 00 bridges (01.0 and 1e.0) to state D0. We start setting the
02:04.0 CardBus bridge to D0, but then we apparently hang before completing
pci_set_power_state(). I think the "clocksource acpi_pm" stuff must be
unrelated.
This is getting into the area Rafael knows much more about, but here's another
patch that adds more debug in the pci_set_power_state() path. Remove the first
patch ("patch -p1 -R < patch1") before applying this one. This is against
3.4-rc1 (a5149bf3fed to be exact). If it's easier for you, I put a
pre-compiled image here: http://helgaas.com/linux/suse748616/bzImage2
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.