At some point in the future, people doing reviews or summaries, describing the history of certain series of laptop models will quote the 6th generation of X1 Carbon as this special one model that got everyone pissed because Lenovo has decided to experiment with suspend for no good reason and refused to make it work properly.

S3 is there - we know it's supported by hardware and firmware. It's just hidden from the user and the system. One simple fix to the BIOS and everyone will forgive you. People will stop returning their X1 6ths, getting refunds and moving to other vendors. They may even consider buying your hardware in the future.

Seriously, I love the looks of ThinkPads. I love their non-curvy design (every other brand tries to make their laptops look like MacBooks) and the fact that in most cases they use components that work well with Linux. But right now, once my current machine is old enough, I will be considering a new laptop. And it's not going to be Lenovo, since it seems to be so very hard for Lenovo to make their flagship laptop right. What is more, I am not going to recommend these laptops to anyone. In the company I work for, when selecting machines for developers, we will no longer consider Lenovo laptops. Which so far has been an obvious choice. We will be looking into other brands and choose one that causes us fewer issues. Unless of course this X1 Carbon 6th edition poopstorm gets sorted out.

I know Linux users are a tiny percentage of your users, so losing that demographic is not a huge deal. It might even be beneficial - you can concentrate on Windows users and their issues. This is business and I totally understand that. I'm just honestly telling you what I think and feel as a user, and what consequences all this has for me and the people around me. It makes me sad and a bit angry, but I have no power over you.

---

 wrote:

---

I know what computer brand I won't recommend in the future.. :catsurprised:

---

 

To be fair it's not Lenovo's fault that no Linux kernel supports Si03 yet. It's just that they're not actively producing a custom BIOS for Linux.

---

It's not only about Linux. 1) it doesn't work in Windows either if you look around the forum, 2) it doesn't even make sense on a laptop when you think about it. S3 may be old, but it works super well - saves a lot of energy, does not make components hot and wakes up fast. This mode may make sense in some weirdly designed laptops that have an external display (on the other side of the cover or they let the cover go around near 360 degrees) OR it makes sense as idle mode when you do not use your opened laptop for a longer period and it becomes idle like a phone or tablet, but still receives notifications and stuff. It has no place in a normal laptop that you close.

Hi.. I managed to the my trackpad working after resume now when using S3.. So the computer became a bit more friendly now... 

In /lib/systemd/system-sleep I have a file called trackpad containing:

#!/bin/sh
TID=$(xinput list | grep -iPo 'trackpoint.*id=\K\d+')
if [ "${1}" == "post" ]; then
  echo -n none > /sys/devices/platform/i8042/serio1/drvctl
  sleep 1
  echo -n reconnect > /sys/devices/platform/i8042/serio1/drvctl
  xinput disable "$TID"
  xinput enable "$TID"
fi 

---

 wrote:

Hi.. I managed to the my trackpad working after resume now when using S3.. So the computer became a bit more friendly now... 

 

In /lib/systemd/system-sleep I have a file called trackpad containing:

 

#!/bin/sh
TID=$(xinput list | grep -iPo 'trackpoint.*id=\K\d+')
if [ "${1}" == "post" ]; then
  echo -n none > /sys/devices/platform/i8042/serio1/drvctl
  sleep 1
  echo -n reconnect > /sys/devices/platform/i8042/serio1/drvctl
  xinput disable "$TID"
  xinput enable "$TID"
fi 

 

---

In my case this works well with S2idle. After modifying the DSDT tables and getting full S3, trackpoint and trackpad are not recognized by the system after wakeup (not listed in xinput at all).

Ok, now I see what you mean. This occurred randomly for me now as well.

I also added this:

#!/bin/sh
TID=$(xinput list | grep -iPo 'trackpoint.*id=\K\d+')
if [ "${1}" == "post" ]; then
  rmmod rmi_smbus
  modprobe rmi_smbus 
  xinput disable "$TID"
  xinput enable "$TID"
  echo -n none > /sys/devices/platform/i8042/serio1/drvctl
  sleep 1
  echo -n reconnect > /sys/devices/platform/i8042/serio1/drvctl
fi

I tried running rmmod rmi_smbus and validating that synaptics/trackpoint is removed/restored. Hope that this works for you and others! 

There seems to be a new BIOS version 1.28 online with "(Fix) Improve power control." in the changelog. Has anyone tried it yet and can tell what exactly this means?

I did install it yesterday via fwupdmgr (worked great!)

fwupdmgr refresh
fwupdmgr update

I see no difference related to sleep or I can't find any more detailed changelog. It is probably something related to Windows.. 

(Off topic, discard my earlier workarounds as they don't work anymore..) 

Hi,

Initial testing with kernel 4.19-RC2 is showing that the sleep states issue seems to be resolved. At least from the limited time I've been able to run this kernel now. This is running Linux Mint 19 and only machine Setup change is to enable legacy BIOS.