I am attempting to debug a problem with suspend on a Panasonic
Toughbook CF-C1 (Intel 2520M, circa 2011). Similar generation hardware
to my Thinkpad X220, but with a PC-card slot. In all other respects to
suspend (and maybe resume) it runs nornally. shutdown -p also works
normally.
The issue is that suspend does not initiate with apm -z (S3), apm -Z
(S4), or apm -S (standby). The display goes off and network interfaces
go down, but the system continues to run with increasing fan speed,
until I do a forced power off.
The same happens with and without Xorg running. S3 suspend under Linux
works, however.
The 'suspend' pulsing of the power button LED doesn't occur, so the
system is getting stuck somewhere before that point in the sequence,
but after display off and network disabling. Nothing of interest appears
in the system logs. fsck runs on /var and /usr after restart, so some
files were open there.
Full details including dmesg and ACPI tables were included in a sendbug
report to openbsd-bugs at:
https://marc.info/?l=openbsd-bugs&m=172556466630461&w=2
This showed up on a 7.5-release standard install. I have since upgraded
to a recent snapshot with no change in behavour.
Reading the mailing list archives about the suspend implementation, one
of the first steps is calling wsdisplay_suspend to turn off the
display. I want to try temporarily disabling this so I can see if there
are any clues on the console to the system state, such as a crash
report.
However, I can't find any calls to wsdisplay_suspend in the code
in /sys/dev/acpi/* .
Can anyone give me some pointers please, as to where the present
suspend process is entered, and how to disable the display blanking?
--
Chris Billington
No comments:
Post a Comment