2.6.13 introduced a change to drivers/acpi/scan.c along the lines of:
*
* TBD: Assumes LDM provides driver hot-plug capability.
*/
- result = acpi_bus_find_driver(device);
+ acpi_bus_find_driver(device);
end:
if (!result)
This was inside this commit:
http://www.kernel.org/git/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=3fb02738b0fd36f47710a2bf207129efd2f5daa2
Since then, various HP/Compaq laptops have not been able to boot, freezing
after these messages:
PCI: Using ACPI for IRQ routing
PCI: If a device does not work, try "pci=routeirq". If it helps, post a report
PCI: Cannot allocate resource region 7 of bridge 0000:00:1c.1
PCI: Cannot allocate resource region 8 of bridge 0000:00:1c.1
PCI: Cannot allocate resource region 9 of bridge 0000:00:1c.1
Rajesh Shah (the author of said commit) has since then acked that this small
change should probably be reverted:
> Looking at this closely now, checking for the result does appear
> to be wrong. Binding a driver for a device should be optional,
> and should not fail adding the device to the acpi list. I suspect
> a previous iteration through this code failed to find a driver
> match, returned failure to the caller and caused bad things to
> happen. So, your patch looks good to me.
The main diagnosis/discussion has been at:
https://bugzilla.novell.com/show_bug.cgi?id=116763
There are also some other reports elsewhere:
http://bugzilla.kernel.org/show_bug.cgi?id=5221
http://bugs.gentoo.org/112601
So far the investigation/fixing has been done by Thomas Renninger and others.
I'm just hoping to breathe some life into an issue which has been sitting
around for nearly 3 months.
If acceptable, please apply.