Thinkpad T440p does not boot with clean install when power adapter is connected

Hi!

Brief description of the problem:

When booting Thinkpad T440p, I see the error shown at the attached photo. The error only occurs if the laptop is plugged into a power outlet using an adapter. No error occurs when running on battery power.
Thus, the system can only be booted when running on battery power.

Detailed description:
  • There are several t440p's different models: i5/i7 with or without Nvidia. These laptops were produced in 14/15 but are almost new.
  • Error occurs on all these laptops.
  • Tested with various genuine power adapters (65/90W) and various genuine batteries. All parts are original and near-new.
  • Latest BIOS firmware. Initially, the BIOS settings were set to default. Then we experimented with disabling "Secure boot", "Security chip", "I / O port access". But without results.
  • FreeBSD 13.0 installed with the default settings. This assumes that I just booted from a USB stick and pressed <Enter> almost all the time. In doing so, I chose "Auto ZFS".
  • After standard installation from a USB drive, no other settings were made. Just boot after clean install.
  • In addition, I took a system snapshot from my workstation ZFS pool (which has been used for many months) and sent it to t440p using zfs send/receive. The result is the same.
  • If the laptop boots (on battery), then everything works as expected.
  • I can avoid the error if the power adapter is disconnected during boot. After booting, I can connect it, and it works.
Questions:

1 - Any ideas?
2 - Your opinion, will it be possible to make ZFS mirror using system 2.5 hdd + 2.5 caddy hdd on t440p's? It was for this that I started everything with t440p, not at all expecting a catch at such an early stage.

Thank you!
 

Attachments

  • out.jpg
    out.jpg
    847.3 KB · Views: 86
I run & ran several FreeBSD ThinkPad laptops (currently two in active use, both with FreeBSD 13.0-RELEASE). (Un)fortunately, I never encountered an issue like this.

A question for completeness of the information provided: Is this using the proprietary Lenovo rectangular power plug or does powering occur via USB-C PD?
 
Is this using the proprietary Lenovo rectangular power plug
Yes, used exclusively genuine adapter (the one you get when you open the box with brand new laptop). 65W for t440p with i5 and 90W for model with i7. These power supplies came with laptops natively and did not migrate between machines.

Moreover, one of the laptops used yesterday stood in the same place, was connected to the same original adapter and worked successfully with Win 8.1 Pro. Another has worked successfully with Gentoo. So far, there is no reason to doubt their correct functioning.
 
That stacktrace point to the kernel's final initialization method mi_startup.

To quote the kernel documentation
System startup; initialize the world, create process 0, mount root
filesystem, and fork to create init and pagedaemon. Most of the
hard work is done in the lower-level initialization routines including
startup(), which does memory initialization and autoconfiguration.

I can't be certain, but apparently some device might be causing a fault condition when you're plugged in. You could try compiling the kernel with verbose sysinit output; also attempting to debug the produced core dump is a good option. There is pretty good documentation on how to do this:

Since you seem to be getting this on more than one device, there could also simply be a bug somewhere in some kernel module. It's hard to tell without further digging, I'm afraid.

If you don't know C or never have done any debugging - maybe someone volunteers checking out your core dump, if you manage to provide a download of your core file (should be in /var/dumps)
 
Yes, used exclusively genuine adapter (the one you get when you open the box with brand new laptop). 65W for t440p with i5 and 90W for model with i7. These power supplies came with laptops natively and did not migrate between machines.

Moreover, one of the laptops used yesterday stood in the same place, was connected to the same original adapter and worked successfully with Win 8.1 Pro. Another has worked successfully with Gentoo. So far, there is no reason to doubt their correct functioning.
It was not about doubting their functionality but rather the fact that a USB-C PD situation has some additional complexity to it. While there are usually chipsets specifically to handle these things at least in embedded situations the system has some additional communication to do. I do not know - and therefore not rule out - that powering via USB-C PD might have some "side-effects" on the kernel/system that you won't see with the proprietary "here's some DC power for you" solution.

Other than that, the idea is to collect as much information as possible so that people more knowledgable than me in this area might have an easier time providing useful help.
 
Back
Top