Hi Colton, thanks for testing it! It looks like the kernel correctly receives and delivers the ACPI events, but then it fails to perform the actual suspend to mem.
The first setting simply increases the PM debugging level (...and hopefully we can get more info from dmesg -w).
The second one disables the asynchronous execution of the PM suspend and resume callbacks. If the problem is a race condition in some of those callbacks, this should prevent the race.
Hi Colton, thanks for testing it! It looks like the kernel correctly receives and delivers the ACPI events, but then it fails to perform the actual suspend to mem.
Can you try one more test?
echo 1 > /sys/power/ pm_debug_ messages
echo 0 > /sys/power/pm_async
echo mem > /sys/power/state
The first setting simply increases the PM debugging level (...and hopefully we can get more info from dmesg -w).
The second one disables the asynchronous execution of the PM suspend and resume callbacks. If the problem is a race condition in some of those callbacks, this should prevent the race.
Thanks.