Mark Kettenis on Wed, Jun 11 2025:
>> Not sure if it is the same issue, but I have also experienced occasional
>> stuck (modifier) keys that get unstuck by switching to a text console
>> and back to X.
>
> That could very well be the case.
FYI, I retried it today with a new kernel (compiled yesterday), since
there were recent commits that seemed like they might touch relevant
bits. However the behavior is unchanged as of:
OpenBSD 7.7-current (GENERIC.MP) #2: Wed Jun 11 15:41:00 CEST 2025
[email protected]:/sys/arch/amd64/compile/GENERIC.MP
Specifically, that build included your ACPI GPIO changes, so I also
tried suspend on this laptop. Not sure if that might be related to the
issue at hand, but here's my experience in short:
Closing the lid and reopening it had previously worked withough hanging
the machine, but knocked out the USB ports in that they would afterward
all report "device error, disabling port". Today I booted the machine
with the above kernel in single user (i.e. no X), closed the lid and
briefly saw the kernel print "S3 unavailable, using S0". Reopened it and
nothing happened. Pressed power button. After a moment, the system came
on and the kernel spewed out a bunch of "USB detached" messages. Nothing
reattached afterward.
>> This is running with machdep.allowaperture=0. I will try if setting it
>> to 1 changes something.
>
> That won't help and you should leave it set to 0.
Thanks, that's good to know. I wasn't able to tell from the description
in xf86(4) which setting would be "expected" for this machine. FWIW, the
X server did make it sound like I should enable it:
[ 28.706] (WW) checkDevMem: failed to open /dev/mem and /dev/xf86
(Operation not permitted)
Check that you have set 'machdep.allowaperture=1'
in /etc/sysctl.conf and reboot your machine
refer to xf86(4) for details
