Can't use L5 after update!

Hi Kyle, I have to say that is BS.
Support do NOT know what is going on, and after biting my tongue for a long time I have totally lost faith in Purism. I was not going to mention how long it actually took for us users to finally get our Librem 5’s and now to give an update through the PureOS store that has now bricked our phones.
No one is support is giving a definitive answer to solve this issue.
At this stage the advice I am getting is to re-flash the device! I mean WTF? seriously?
That is what I expect from the PinePhone, not Purism and the Librem 5.
Plus I’d like to know what kind of guarantee do we have that this won’t happen again, on the next update or one after that?
So thanks for the brick guys, I guess I have a new paper weight.

What is BS? Every Linux Kernel is new beginning, but the linux-image-5.16.0-1-librem5 is the LTS or simply the great one (currently as 5.16.3pureos1), IMHO.

I’d like to add some critics:

Had the issue like 12h ago - librem5 is bricked after update, issue fits the description provided here.

The initial post is more than 48h ago.

Understanding of the problem seems to have existed before the initial post.

Why didn’t Purism stop (kernel-) updates until the issue is resolved to deliver the resolution as the first update to make sure that not more customers are affected?

If not already happened: Stop updates, please, until the next update ships the resolution of this fatal error first. @jeremiah, @joao.azevedo?

Instead of heaving everybody strucken by this error contact support it would be nice to have that information on the wiki.

1 Like

… or even just in this topic.

Support didn’t have resolution steps to begin with when the problem first surfaced. They aren’t all-knowing and can’t know the resolution to every problem when it first surfaces. Instead they (like anyone doing troubleshooting) work the issue, and work with other folks at Purism to figure out the best way to resolve it. In doing that, they first advised reflashing as a simple (if destructive) way to definitely get the phone back to normal (it also has the side benefit of increasing the size of /boot to my understanding). We now have a universal set of steps we are advising people to perform who run into this issue. Early on, yes, reflashing probably seemed like the easiest way to get someone back to a functioning phone, but now we have a better process in place.

We have been working to halt the kernel update, however the problem to my understanding isn’t universal and isn’t within the kernel update itself (ie it’s not a kernel bug), it happens if you happen to have too many past kernels taking up space in /boot.

Sure, and if you want we can post in this topic too, although I’d still prefer people work with support to perform the steps in case they have questions or run into any issues along the way.

1 Like

Steps to perform:

On your computer with GNU/Linux system:

  1. Install the package uuu, on PureOS and other Debian based systems, from a terminal run the command: sudo apt install uuu
  2. Open terminal and run command: mkdir jumpdrive && cd jumpdrive (creates new folder and changes location to it)
  3. Download: wget
  4. Unpack: tar -xf purism-librem5.tar.xz
  5. Put the Librem 5 in uuu mode:
  1. Ensure that the phone is switched off.
  2. Turn all Hardware-Kill-Switches off
  3. Remove battery
  4. Hold volume-up
  5. Insert the USB-c cable: (red light blinks, no green light)
  6. Reinsert the battery: (red light is constantly on, the script will continue)
  7. Release volume-up
  1. To confirm that the device is in uuu mode, in the terminal run the command lsusb and search the following for an entry like:
    001 Device 106: ID 1fc9:012b NXP Semiconductors i.MX 8M Dual/8M QuadLite/8M Quad Serial Downloader

  2. Run the script: sudo ./

  3. Once it is run the L5 will show up as an external drive

  4. There will be a 465 MiB drive and a 28.7 GiB drive: open the 465 MiB one

  5. Delete “initrd.img” and “vmlinuz”

  6. Rename “initrd.img.bak” to “initrd.img”, rename “vmlinuz.bak” to “vmlinuz”

  7. At that point you should be able to unmount the external drive and reboot normally

Once you get the device booted you should clean up the failed parts, so run these commands in terminal, one by one:

  1. sudo apt remove linux-image-5.13.0-1-librem5
  2. sudo flash-kernel

@guru @ChriChri @DrTechnical @Marts @LinuxNoob888 please check the above post ^^^


My Librem 5 no longer had the wifi and bluetooth icon active after the update. So I rebooted and I could no longer start the Librem 5. Then I followed the instructions above, but after step 5 "Run the script:“ ./ “” on the Debian PC it did not appear the Librem 5, here is from the terminal:
$ ./
./ 3: ./ uuu: not found
I mean I DON’T see a hard drive. My Librem 5 no longer had the wifi and bluetooth icon active after the update. So I rebooted and I could no longer start the Librem 5. Then I followed the instructions above, but after step 5 "Run the script:“ ./ “” on the Debian PC it did not appear the Librem 5, here is from the terminal:
$ ./
./ 3: ./ uuu: not found
I mean I DON’T see a hard drive. So I can’t complete the procedure. I plugged the USB-c cable into the Librem and the other end into the USB of the PC, is that correct?

1 Like

Looks like you need to install the uuu utility. @mladen we may need to revise the instructions to take that into account.

Done, the instructions are updated


After installing uuu, here’s what happened.
jumpdrive# ./
uuu (Universal Update Utility) for nxp imx chips – lib1.4.77
Success 1 Failure 0
2:1 3/ 3 [=================100%=================] SDPV: jump
3:1 10/10 [Done ] FB: Done
No device on the PC, but on the Librem 5 the image of the smartphone profile with the word error in red. And on the bottom: “/dev/mmcblk0 could not be opened, possible eMMc defect”.

So I unplugged the USB cable and magically the Librem 5 came back on and I was able to log in again. Now I will go to the boot directory and continue with the deletion and renaming. To turn the librem back on I have to use volume-up.

where is the file 5.13.0-1-librem5 ? it shows:
E: unable to locate package 5.13.0-1-librem5

If you are thinking of “5.13.0-1-librem5” as in the command “sudo apt remove 5.13.0-1-librem5” suggested by @mladen above, then I think it is supposed to be a package name, but I suspect that @mladen means “linux-image-5.13.0-1-librem5”, so my guess is that the command should be as follows:

sudo apt remove linux-image-5.13.0-1-librem5

You can use the “apt list” command to list available packages, for example like this to see installed kernels:

apt list | grep linux-image | grep installed

which should give a list of installed linux-image-* packages with different version numbers, then you can compare that to the output from “uname -a” to figure out which one of them is currently used.

Removing kernels can be fun but don’t get too carried away, it’s probably wise to keep at least one. :wink:


@ [Skalman]
thank you very much, now my Librem 5 boot again in regular fashion!


I did try to follow the procedure on my desktop (PopOS) just for testing purposes but I couldn’t do:

I had an error:
Failure open usb device,Try sudo uuu

To make it work without sudo I had to add some udev rules:
SUBSYSTEMS==“usb”, ATTRS{idVendor}==“1209”, ATTRS{idProduct}==“4291”, MODE=“0660”, GROUP=“plugdev”
SUBSYSTEMS==“usb”, ATTRS{idVendor}==“0525”, ATTRS{idProduct}==“a4a5”, MODE=“0660”, GROUP=“plugdev”
SUBSYSTEMS==“usb”, ATTRS{idVendor}==“0525”, ATTRS{idProduct}==“b4a4”, MODE=“0660”, GROUP=“plugdev”
SUBSYSTEMS==“usb”, ATTRS{idVendor}==“1fc9”, ATTRS{idProduct}==“012b”, MODE=“0660”, GROUP=“plugdev”

Add my user to the plugdev group and reload udev:
sudo udevadm control --reload-rules && sudo udevadm trigger

Is all this already done on PureOS?

Yes it is already done in PureOS. So that it works without sudo.


If that happens to anyone, just turn the phone off with long power button press and try again. Jumpdrive is using an older kernel which had a bug that made it sometimes not recognize internal storage, but it should work after at most a few tries.

Before launching that script, check lsusb for an entry like:

Bus 001 Device 106: ID 1fc9:012b NXP Semiconductors i.MX 8M Dual/8M QuadLite/8M Quad Serial Downloader

If it’s not there, it means that either your phone isn’t actually in flash mode, or there may be something wrong with USB cable/connection.


I’ve got it working I just wanted to make sure that no one else gets the same issue, do you or @joao.azevedo knows why I had to add 4 different devices to be able to launch the script without sudo?

For those interested: these steps above could actually be replaced with just “turn the phone on while holding volume-up button and plug usb cable in”, so you can try that first before doing the battery dance - however, the phone itself gives no indication that you actually succeeded and there may be some edge cases where it’s not so simple, so the official steps are slightly complicated to make the instructions more bullet-proof (at cost of being a bit annoying)

(just clarifying since I’ve seen people in the past asking about why is taking the battery out necessary :slight_smile: )


Yes. Out-of-the-box on e.g. Ubuntu you will need to do

sudo ./

A udev rule avoids that but for the small number of times I need to boot Jumpdrive …