© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
bbfa235
(Get the code!)
> Anyway, I still don't get why upgrading the wifi-cards firmware to a
> more recent version fixes the problem.
> Would you mind to elaborate on this a bit more?
When the system is in the hardware sleep state it can be woken up from either a GPIO or a GPE. Any of those ACPI events I mentioned will make the EC trigger such a wakeup. When the kernel wakes up it will process the ACPI events and then confirm whether there are any remaining interrupts active.
* If no interrupts are active, the system will enter back into hardware sleep state.
* If any interrupts are still active then the kernel treats it as a genuine wakeup source that should exit the s2idle loop.
What happens with this bug is that the WLAN controller is connected to a design specific GPIO, and when the system leaves hardware sleep state that GPIO is also triggered.
This means that two interrupts are active: 1 for that GPIO and 1 for the EC SCI. The firmware update for the WLAN controller fixes it so that the GPIO is not active during resume.