Re: SSD being put to sleep during boot-up when laptop on battery

Re: SSD being put to sleep during boot-up when laptop on battery

[Alan Stern]

On Sun, 24 Dec 2017, Adam Bennett wrote:

> Alan,
> 
> I have a Dell Precision 7520 with an Samsung SSD in addition to an NVMe 
> drive.  I'm currently running 4.14.8 but have seen the below problem 
> since I moved from 4.9 to 4.13.
> 
> When I boot without AC plugged in, the SSD is stopped almost immediately 
> in the boot process, not allowing the boot to continue.
> 
> I can boot on fine on AC.  When I unplug the cord, I see the "Stopping 
> disk" message, and the system is unresponsive until I plug the AC back in.
> 
> I initially thought the problem was in user-space, but I set up 
> laptop-mode-tools to keep "control" as "on", and the problem still 
> persists.  Also, I don't see the issue in 4.9.

It's entirely possible that the problem does lie in userspace, during 
boot-up.  Don't forget that your initramfs image could be causing this; 
have you tried to rebuild it with the new laptop settings?

> I have temporarily worked around the problem by returning from 
> sd_suspend_runtime without calling sd_suspend_common in sd.c (obviously 
> not the true fix).
> 
> I have tried a number of searches, and couldn't find any bug reports of 
> this nature, but I'm not that in-tune with the linux development process 
> to have searched all the correct places.

This should be sent to the linux-pm and linux-scsi mailing lists 
(CC'ed).

> I'd like to help track this problem down, do you have any suggestions or 
> is there some additional details I could provide?

You can try bisecting between the 4.9 and 4.13 kernels to find the 
commit which first caused the problem.

Alan Stern

[Solved] Re: SSD being put to sleep during boot-up when laptop on battery

[Adam Bennett]

The problem was resolved upon upgrading laptop-mode-tools to 1.72.2.

Adam


On 12/25/2017 11:29 AM, Alan Stern wrote:
> On Sun, 24 Dec 2017, Adam Bennett wrote:
>
>> Alan,
>>
>> I have a Dell Precision 7520 with an Samsung SSD in addition to an NVMe
>> drive.  I'm currently running 4.14.8 but have seen the below problem
>> since I moved from 4.9 to 4.13.
>>
>> When I boot without AC plugged in, the SSD is stopped almost immediately
>> in the boot process, not allowing the boot to continue.
>>
>> I can boot on fine on AC.  When I unplug the cord, I see the "Stopping
>> disk" message, and the system is unresponsive until I plug the AC back in.
>>
>> I initially thought the problem was in user-space, but I set up
>> laptop-mode-tools to keep "control" as "on", and the problem still
>> persists.  Also, I don't see the issue in 4.9.
> It's entirely possible that the problem does lie in userspace, during
> boot-up.  Don't forget that your initramfs image could be causing this;
> have you tried to rebuild it with the new laptop settings?
>
>> I have temporarily worked around the problem by returning from
>> sd_suspend_runtime without calling sd_suspend_common in sd.c (obviously
>> not the true fix).
>>
>> I have tried a number of searches, and couldn't find any bug reports of
>> this nature, but I'm not that in-tune with the linux development process
>> to have searched all the correct places.
> This should be sent to the linux-pm and linux-scsi mailing lists
> (CC'ed).
>
>> I'd like to help track this problem down, do you have any suggestions or
>> is there some additional details I could provide?
> You can try bisecting between the 4.9 and 4.13 kernels to find the
> commit which first caused the problem.
>
> Alan Stern
>