10 ms·

Ubuntu 17.10 corrupting BIOS of many Lenovo laptop models

151 points by muxator 9y ago

codemac 9y ago
EFI allows for OS's to modify probably too much, and the state of EFI implementations aren't defensive enough. I personally would blame Lenovo for releasing something so brick-able.
- mcny 9y ago
  I managed to brick my Asus N61JV-X2 by updating the bios using their official bios update software. (Specs if you're curious https://www.anandtech.com/show/2962/asus-n61j-x2-optimus-gt325m-meets-arrandale https://www.anandtech.com/show/2962/asus-n61j-x2-optimus-gt3... ) I also learned that backup bios chips were a thing but not usually available on a notebook PC
  - vetinari 9y ago
    I managed to brick Thinkpad X230, flashing BIOS update that came via TVSU. Now, that's an achievement :/
    - ct0 9y ago
      Here I was thinking that the 230 was indestructible
- discreditable 9y ago
  OSes tickling the BIOS is nothing new. Vendors have been providing Windows-based flashing utilities for a long time. Even most of the special bootable flashing disks run freedos.
- JdeBP 9y ago
  As pointed out several times in the headlined bug report, this is not an EFI problem. Several people reproduced this problem on systems either that were booting using the CSM or that did not have EFI firmwares in the first place.
- Crontab 9y ago
  EFI is a fucking soup sandwich. In my opinion, it suffers from complexity and too much ambition.
julienfr112 9y ago
How could an OS corrupt a BIOS ? BIOS is in a <strike>higher</strike> lower (see alex_duf comment) level than OS, and then, should no be modified by any program, except maybe by an update program, and only with a firmware signed by lenovo.
- nikbackm 9y ago
  Well, you could at one time brick your computer by by executing an ill-advised "rm -rf" from Linux. Made the news some years back I think.
  - gerdesj 9y ago
    You can do a similar thing in any OS - delete the vast majority of your file system by running a command as root/administrator and ignoring warnings. FYI rm -rf no longer recurses upwards (. and ..) and hasn't done so for a very, very long time. This problem is about breaking the BIOS in some way, something that you should not be able to do.
    - msl 9y ago
      You used to be able (maybe still are?) to "rm" the files on a mounted efivarfs. See the relevant systemd issue ([1]). [1] https://github.com/systemd/systemd/issues/2402 https://github.com/systemd/systemd/issues/2402
      - merb 9y ago
        well windows can get write access to these aswell. (just not by using rm or any other tool to remove files by default). I think this is a big difference and should probably be addressed. make /sys read-only and allow to write them differently with a interface where I need to call THIS_METHOD_CAN_BREAK_A_SYSTEM_IF_USED_INCORRECT_WRITE_SYS() command. but well, systemd guys will/won't do that.
        JdeBP 9y ago
ezoe 9y ago
It looks like Lenovo's problem.
- antoncohen 9y ago
  Yes, while Ubuntu has regressed, the root of the problem seems to be a Lenovo logic error where Lenovo resets the BIOS settings if the write protection bit is set to read/write. https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1734147/comments/173 https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1734147...
tcfunk 9y ago
My XPS 13 Developer edition had something similar happen to it shortly after upgrading to 17.10. However I think this was circumstance, and that the real culprit was a bad BIOS update from Dell which removed all entries from the boot list (or something to that effect).
- wakeywakeywakey 9y ago
  What did you do to recover?
  - cevn 9y ago
    (not the parent, but I had the same laptop with the same problem) I reinstalled Fedora. I could have probably recovered the grub bootloader manually, but I wasn't really keen on sitting there for a few hours. All my data was backed up anyway.
    - wakeywakeywakey 9y ago
      It still allowed you to boot from USB? How do you recover if you lose that option (other than sending in your laptop to be reflashed)?
  - owaislone 9y ago
    I had the exact same issue on 17.10 + XPS 13 + Dell firmware update. I booted in the BIOS setup and manually pointed the EFI to ubuntu's grub's EFI file. That immediately fixed the problem.
vegbrasil 9y ago
Is it safe to install on a Dell XPS 15?
- bproven 9y ago
  I upgraded my XPS 15 (9550) from Ubuntu 17.04 to 17.10 about a week ago and have had no issues so far. Actually, I find it a little less buggy on my XPS than 17.04 was - overall I am pretty happy with it. Of course YMMV - my specs are below ---------------------- Dell BIOS: 1.5.1 i7-6700HQ CPU @ 2.60GHz 16GB RAM Intel® HD Graphics 530 + NVidia 512GB SSD
  - vegbrasil 9y ago
    Did you manage to get both GPUs to work?
    - pxc 9y ago
      It's a bit of a pain, but quite doable on most distros, if you know your way around. I just did it on two distros for my ThinkPad 25: with openSUSE it was part of a weekend of playing around with it and setting it up, and on NixOS it took an hour or two to set up how I like. The real problem, imo, is that all of the solutions work badly (they either work at a performance penalty (as with Bumblebee) or require a separate X server for each GPU). This is entirely a result of NVIDIA's refusal to implement decent support. The Nouveau drivers offer proper PRIME support, which yields an experience more like that on Windows in terms of workflow and the amount of setup (virtually none) required, but they have their own performance problems. Hybrid graphics with AMD GPUs apparently works nicely with whatever drivers you like, granted that the software stack is sufficiently up to date on your system.
    - bproven 9y ago
      I didnt bother - I set it up to only use the Intel GPU b/c I wanted longer battery life. Besides I don't game on the laptop anyway. I dual boot Windows (since the license came with the box) and can use that if I really want to game. However if you really want to, everything pxc says below is spot on...
sandGorgon 9y ago
For those who dont remember, this has happened in the past with systemd - https://github.com/systemd/systemd/issues/2402 https://github.com/systemd/systemd/issues/2402 At the time systemd got a lot of serious heat and poettering got called names for something that was not systemd's fault to begin with. > To make this very clear: we actually write to the EFI fs in systemd. Specifically, when you issue "systemctl reboot --firmware" we'll set the appropriate EFI variable, to ask for booting into the EFI firmware setup. And because we need it writable we'll mount it writable for that.
- JdeBP 9y ago
  Already mentioned on this page at https://news.ycombinator.com/item?id=15971180 https://news.ycombinator.com/item?id=15971180 .
- dogecoinbase 9y ago
  It absolutely was systemd's fault -- systemd regularly half-implements or changes default behavior in ways that "should" be okay, but aren't okay, because they don't know why the fence was built (in the Chestertonian sense). Systems programmers should know the history of their discipline.
  - mjg59 9y ago
    I wrote the EFI variables filesystem. It was intended to be mounted read/write. Systemd did nothing wrong whatsoever in mounting it read/write, and mounting it read only would have broken all the existing userland that handles EFI variables.
    - gens 9y ago
      What userland is it ? And why is it a problem for that userland to mount it itself ? Why would you even write to efi ? Maybe at boot, i don't know, but certainly not while running ? Not all the time ? Oh i hope there isn't a single program that continuously writes to efi. In my opinion systemd is at fault, as they aim to be the "system" and yet every once in a while prove that they don't know shit. EDIT: A program that changes the boot order is all i can think of. (I'm not the one who knows best, on this topic)
1001101 9y ago
Might be a coincidence, but I haven't been able to get into my Toshiba's BIOS since moving to 17.10. Every time I try and enter the BIOS config menu, it reboots.
ericfrederich 9y ago
Good thing I saw this. I was thinking about installing Ubuntu or Mint on my X1 Yoga during the holidays. I'll hold off now or do some more research.
- chrissnell 9y ago
  For what it's worth, I haven't had this problem with Arch Linux on my ThinkPad T470s or P50 and I did a fresh install on the P50 this week. Arch has really nice ThinkPad support and documentation, too.
just1nn 9y ago
17.10 was the absolute most buggy software that I have installed from Canonical. I tried it on Dell hardware (+ Apple Hardware, + A VMware VM on X64 + VirtualBox on OSX, etc) that does just fine with Fedora, Arch, Gentoo, and it was constantly prompting me to send error reports, killing processes, etc. I just flipped from using the "latest" vs the LTS versions at home because of this version, specifically.
- gitgud 9y ago
  Not the worst Ubuntu version, but I had a lot of issues with the default protocol being Wayland [1]. This broke a few programs I use, [2] mainly the screenshot application "Shutter". Most users don't realise it became the default, so their apps just stopped working... bad decision in my opinion. [1] http://www.omgubuntu.co.uk/2017/08/ubuntu-confirm-wayland-default-17-10 http://www.omgubuntu.co.uk/2017/08/ubuntu-confirm-wayland-de... [2] https://askubuntu.com/questions/971273/screenshot-feature-of-shutter-doesnt-work-after-upgrading-to-ubuntu-17-10/971278 https://askubuntu.com/questions/971273/screenshot-feature-of...
chisleu 9y ago
It isn't Ubuntu doing it. It is Intel and Lenovo. Owner of a $3k Lenovo P50 here that was "Linux compatible" but a stake xeon that had been useless on Linux until the latest Fedora.
tinus_hn 9y ago
Just like all the other UEFI implementations, the Lenovo one is probably a giant piece of crap. The main problem is Windows because it allows manufacturers to get away with the 'it runs Windows so it's good enough to release' mindset.
- user5994461 9y ago
  The main problem is Linux. Linux cannot be tested, it designates hundreds of variable setups of distributions, kernels and tools. It's not possible to test anything against that. Lenovo supports redhat and ubuntu, that's already a lot. https://support.lenovo.com/gb/en/solutions/pd031426 https://support.lenovo.com/gb/en/solutions/pd031426
  - tinus_hn 9y ago
    They shouldn’t support RedHat, Ubuntu and Windows. They should provide a solid, compliant UEFI implementation and then Windows, RedHat Linux and Ubuntu Linux will support them.

151 points by muxator 9y ago

EFI allows for OS's to modify probably too much, and the state of EFI implementations aren't defensive enough. I personally would blame Lenovo for releasing something so brick-able.

mcny 9y ago
I managed to brick my Asus N61JV-X2 by updating the bios using their official bios update software. (Specs if you're curious https://www.anandtech.com/show/2962/asus-n61j-x2-optimus-gt325m-meets-arrandale https://www.anandtech.com/show/2962/asus-n61j-x2-optimus-gt3... ) I also learned that backup bios chips were a thing but not usually available on a notebook PC
- vetinari 9y ago
  I managed to brick Thinkpad X230, flashing BIOS update that came via TVSU. Now, that's an achievement :/
  - ct0 9y ago
    Here I was thinking that the 230 was indestructible
discreditable 9y ago
OSes tickling the BIOS is nothing new. Vendors have been providing Windows-based flashing utilities for a long time. Even most of the special bootable flashing disks run freedos.
JdeBP 9y ago
As pointed out several times in the headlined bug report, this is not an EFI problem. Several people reproduced this problem on systems either that were booting using the CSM or that did not have EFI firmwares in the first place.
Crontab 9y ago
EFI is a fucking soup sandwich. In my opinion, it suffers from complexity and too much ambition.

julienfr112 9y ago

How could an OS corrupt a BIOS ? BIOS is in a <strike>higher</strike> lower (see alex_duf comment) level than OS, and then, should no be modified by any program, except maybe by an update program, and only with a firmware signed by lenovo.

nikbackm 9y ago
Well, you could at one time brick your computer by by executing an ill-advised "rm -rf" from Linux. Made the news some years back I think.
- gerdesj 9y ago
  You can do a similar thing in any OS - delete the vast majority of your file system by running a command as root/administrator and ignoring warnings. FYI rm -rf no longer recurses upwards (. and ..) and hasn't done so for a very, very long time. This problem is about breaking the BIOS in some way, something that you should not be able to do.
  - msl 9y ago
    You used to be able (maybe still are?) to "rm" the files on a mounted efivarfs. See the relevant systemd issue ([1]). [1] https://github.com/systemd/systemd/issues/2402 https://github.com/systemd/systemd/issues/2402
    - merb 9y ago
      well windows can get write access to these aswell. (just not by using rm or any other tool to remove files by default). I think this is a big difference and should probably be addressed. make /sys read-only and allow to write them differently with a interface where I need to call THIS_METHOD_CAN_BREAK_A_SYSTEM_IF_USED_INCORRECT_WRITE_SYS() command. but well, systemd guys will/won't do that.
      - JdeBP 9y ago

ezoe 9y ago

It looks like Lenovo's problem.

antoncohen 9y ago
Yes, while Ubuntu has regressed, the root of the problem seems to be a Lenovo logic error where Lenovo resets the BIOS settings if the write protection bit is set to read/write. https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1734147/comments/173 https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1734147...

tcfunk 9y ago

My XPS 13 Developer edition had something similar happen to it shortly after upgrading to 17.10. However I think this was circumstance, and that the real culprit was a bad BIOS update from Dell which removed all entries from the boot list (or something to that effect).

wakeywakeywakey 9y ago
What did you do to recover?
- cevn 9y ago
  (not the parent, but I had the same laptop with the same problem) I reinstalled Fedora. I could have probably recovered the grub bootloader manually, but I wasn't really keen on sitting there for a few hours. All my data was backed up anyway.
  - wakeywakeywakey 9y ago
    It still allowed you to boot from USB? How do you recover if you lose that option (other than sending in your laptop to be reflashed)?
- owaislone 9y ago
  I had the exact same issue on 17.10 + XPS 13 + Dell firmware update. I booted in the BIOS setup and manually pointed the EFI to ubuntu's grub's EFI file. That immediately fixed the problem.

vegbrasil 9y ago

Is it safe to install on a Dell XPS 15?

bproven 9y ago
I upgraded my XPS 15 (9550) from Ubuntu 17.04 to 17.10 about a week ago and have had no issues so far. Actually, I find it a little less buggy on my XPS than 17.04 was - overall I am pretty happy with it. Of course YMMV - my specs are below ---------------------- Dell BIOS: 1.5.1 i7-6700HQ CPU @ 2.60GHz 16GB RAM Intel® HD Graphics 530 + NVidia 512GB SSD
- vegbrasil 9y ago
  Did you manage to get both GPUs to work?
  - pxc 9y ago
    It's a bit of a pain, but quite doable on most distros, if you know your way around. I just did it on two distros for my ThinkPad 25: with openSUSE it was part of a weekend of playing around with it and setting it up, and on NixOS it took an hour or two to set up how I like. The real problem, imo, is that all of the solutions work badly (they either work at a performance penalty (as with Bumblebee) or require a separate X server for each GPU). This is entirely a result of NVIDIA's refusal to implement decent support. The Nouveau drivers offer proper PRIME support, which yields an experience more like that on Windows in terms of workflow and the amount of setup (virtually none) required, but they have their own performance problems. Hybrid graphics with AMD GPUs apparently works nicely with whatever drivers you like, granted that the software stack is sufficiently up to date on your system.
  - bproven 9y ago
    I didnt bother - I set it up to only use the Intel GPU b/c I wanted longer battery life. Besides I don't game on the laptop anyway. I dual boot Windows (since the license came with the box) and can use that if I really want to game. However if you really want to, everything pxc says below is spot on...

sandGorgon 9y ago

For those who dont remember, this has happened in the past with systemd - https://github.com/systemd/systemd/issues/2402 https://github.com/systemd/systemd/issues/2402 At the time systemd got a lot of serious heat and poettering got called names for something that was not systemd's fault to begin with. > To make this very clear: we actually write to the EFI fs in systemd. Specifically, when you issue "systemctl reboot --firmware" we'll set the appropriate EFI variable, to ask for booting into the EFI firmware setup. And because we need it writable we'll mount it writable for that.

JdeBP 9y ago
Already mentioned on this page at https://news.ycombinator.com/item?id=15971180 https://news.ycombinator.com/item?id=15971180 .
dogecoinbase 9y ago
It absolutely was systemd's fault -- systemd regularly half-implements or changes default behavior in ways that "should" be okay, but aren't okay, because they don't know why the fence was built (in the Chestertonian sense). Systems programmers should know the history of their discipline.
- mjg59 9y ago
  I wrote the EFI variables filesystem. It was intended to be mounted read/write. Systemd did nothing wrong whatsoever in mounting it read/write, and mounting it read only would have broken all the existing userland that handles EFI variables.
  - gens 9y ago
    What userland is it ? And why is it a problem for that userland to mount it itself ? Why would you even write to efi ? Maybe at boot, i don't know, but certainly not while running ? Not all the time ? Oh i hope there isn't a single program that continuously writes to efi. In my opinion systemd is at fault, as they aim to be the "system" and yet every once in a while prove that they don't know shit. EDIT: A program that changes the boot order is all i can think of. (I'm not the one who knows best, on this topic)

1001101 9y ago

Might be a coincidence, but I haven't been able to get into my Toshiba's BIOS since moving to 17.10. Every time I try and enter the BIOS config menu, it reboots.

ericfrederich 9y ago

Good thing I saw this. I was thinking about installing Ubuntu or Mint on my X1 Yoga during the holidays. I'll hold off now or do some more research.

chrissnell 9y ago
For what it's worth, I haven't had this problem with Arch Linux on my ThinkPad T470s or P50 and I did a fresh install on the P50 this week. Arch has really nice ThinkPad support and documentation, too.

just1nn 9y ago

17.10 was the absolute most buggy software that I have installed from Canonical. I tried it on Dell hardware (+ Apple Hardware, + A VMware VM on X64 + VirtualBox on OSX, etc) that does just fine with Fedora, Arch, Gentoo, and it was constantly prompting me to send error reports, killing processes, etc. I just flipped from using the "latest" vs the LTS versions at home because of this version, specifically.

gitgud 9y ago
Not the worst Ubuntu version, but I had a lot of issues with the default protocol being Wayland [1]. This broke a few programs I use, [2] mainly the screenshot application "Shutter". Most users don't realise it became the default, so their apps just stopped working... bad decision in my opinion. [1] http://www.omgubuntu.co.uk/2017/08/ubuntu-confirm-wayland-default-17-10 http://www.omgubuntu.co.uk/2017/08/ubuntu-confirm-wayland-de... [2] https://askubuntu.com/questions/971273/screenshot-feature-of-shutter-doesnt-work-after-upgrading-to-ubuntu-17-10/971278 https://askubuntu.com/questions/971273/screenshot-feature-of...

chisleu 9y ago

It isn't Ubuntu doing it. It is Intel and Lenovo. Owner of a $3k Lenovo P50 here that was "Linux compatible" but a stake xeon that had been useless on Linux until the latest Fedora.

tinus_hn 9y ago

Just like all the other UEFI implementations, the Lenovo one is probably a giant piece of crap. The main problem is Windows because it allows manufacturers to get away with the 'it runs Windows so it's good enough to release' mindset.

user5994461 9y ago
The main problem is Linux. Linux cannot be tested, it designates hundreds of variable setups of distributions, kernels and tools. It's not possible to test anything against that. Lenovo supports redhat and ubuntu, that's already a lot. https://support.lenovo.com/gb/en/solutions/pd031426 https://support.lenovo.com/gb/en/solutions/pd031426
- tinus_hn 9y ago
  They shouldn’t support RedHat, Ubuntu and Windows. They should provide a solid, compliant UEFI implementation and then Windows, RedHat Linux and Ubuntu Linux will support them.