[Bug 218198] Suspend/Resume Regression with attached ATA devices

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=218198

--- Comment #2 from Dieter Mummenschanz (dmummenschanz@web.de) ---
Thanks for the swift reply.

I've applied your patch. Booted up my machine and waited until it transitions
into lower package states (pc8 at the lowest). After that I closed the laptop
LID and let the machine suspend to RAM (S3). After that I reopened the LID and
gave the machine 1-3 minutes time to transition to lower package states which
it now does.
I've attached the dmesg part including your patch when the machine enters
suspend. One thing is odd though:

[  109.424369] ata5.00: qc timeout after 5000 msecs (cmd 0xe0)
[  109.424397] ata5.00: STANDBY IMMEDIATE failed (err_mask=0x4)

this shouldn't be there, right?

Regarding automatic transitioning I'm not sure how this works. However even
though I've set CONFIG_SATA_MOBILE_LPM_POLICY=3 in the kernel config, I have to
call an init script explicitly forcing the scsi host to use low power when
idle:

for foo in /sys/class/scsi_host/host*/link_power_management_policy;
  do echo med_power_with_dipm > $foo;
done

Otherwise the machine is stuck at PC2 state always.

So we have a fix I guess?

Anyway if there is anything alse I can try or provide you with more debug
output please let me know.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.