Zenbook UX501VW-FJ024T and xubuntu 15.10 (attempt 1)

This is a like a log of my travels trying to get stuff to work on a 2016 model Asus Zenbook (UX501VW-FJ024T) on Xubuntu 15.10. At the time of writing, 16.04 and 16.04.1 are a no-go, as it doesn't even properly boot up. There are a number of things broken on this machine and I'd like to fix at least some of them.

(TLDR: for conclusion on how to do this read this post. Below are notes from a failed attempt.)

Main problems right now:

- trackpad is NOT working at all. It's not looking good on this one.
- some fn-keys do not work, others do. I have good hopes some of them can be fixed.
- fan control seems absent. Good hopes this can be fixed.
- backlight dims for no reason (at boot). Can be fixed through xbacklight but I hope a more solid fix can be found.
- no 4k support so default resolution starts at 1:1 UHD, which is super tiny. Once installed you can set this to HD (1920x1080) but login screen will still be tiny, just as terminals (alt-ctrl-fx). And obviously HD is not the same as "retina". Not sure if this is fixable though it should be a regular option in Ubuntu so probably fixable in xfce as well.
- sd card reader reported not to work

I'm not alone

Magic

The remainder of this post is unsuccessful. I gave up and reinstalled 15.10.

Continue reading for that installation on attempt 2.

Bios

Seems the stock bios is v206 while the latest version at the time of writing is v300. So in windows (just give up in this one), download the bios flasher and run the updater.

The "From network" option is screwed up for me. When you press "Update" you get "The format of the downloaded file is not valid". Well goddamn you. Probably hitting some page that no longer exists or some bullshit.

Instead you go to the asus.com site directly, download the bios image from their insecure download page. Start the flasher and select the "from computer" option (The right one). Select the image and go. When you exit the program the computer will reboot and the flasher will update the bios... For the sake of the flasher I booted to windows at first. You never know... but I don't think that was required as I don't see anything.

For what it's worth.. it made diddly squat difference. Can't see a difference in the bios, support seems to have same state as before. But it probably fixed something. I hope so, anyways.

Trackpad

No luck so far, does not work

It seems to be an FTE:1001:00 (supposedly "an Elan touchpad") which seems to be broken for many people, though clearly not all. This devices is _not_ listed in /proc/bus/input/devices for me (I got it through errors in dmesg | grep -i fte). The path /devices/pci0000:00/0000:00:15.1/i2c_designware.1/i2c-7/i2c-FTE1001:00/0018:0B05:0101.0001 does exist, except the last bit is different for me and it the trailing input/input12/ does not exist.

This comment suggests adding i915.preliminary_hw_support=1 to grub (GRUB_CMDLINE_LINUX_DEFAULT in /etc/default/grub which will require sudo to edit, don't forget to update-grub afterwards...). But after a reboot the touchpad did not activate and the messages (dmesg | grep -i fte) did not disappear.

fn-keys

No luck so far beyond what works out of the box

Out of the box state of affairs (on 15.10):

- fn+F1 (sleep): not working they work out of the box if you enable the button in (xubuntu) settings to do what you want. the sleep function has a small delay without visual queue but should work after a few seconds. at the moment this feels safer than just closing the lid (after updating the settings to make that work) because sometimes the suspend is prevented by something. Explicitly suspending before closing the lid proved very reliable in my brief experience.
- fn+F2 (airplane mode): not working
- fn+F3 (keyboard backlit down): works
- fn+F4 (keyboard backlit up): works
- fn+F5 (screen backlit down, I think): not working
- fn+F6 (screen backlit up, I think): not working
- fn+F7 (screen off, I think): not working
- fn+F8 (screen output config): opens display settings (that suffices)
- fn+F9 (toggle trackpad): not working (no UI response)
- fn+F10 (mute sound): works
- fn+F11 (sound up): works (doesn't unmute)
- fn+F12 (sound down): works (doesn't unmute)
- fn+del (ins ...): works (somebody hates linux... cli pasting works)
- keypad fn+return (calc): works (heh)
- fn+numlock (scroll lock): dunno
- fn+arrow keys (media fwd/bck/pauseplay/stop): dunno
- fn+keypad 1/3/7/9 (end/pgdn/end/pgup): works
- fn+keypad 2/4/6/8 (arrow keys): works
- keypad fn+. (alternative del): works
- keypad fn+0 (alternative ins): dunno

If fn+5 is supposed to print a euro sign that doesn't work. It may not be supposed to (would be a cute feature though).

windows 2012

Tried to disable "Windows 2012" to enable the brightness buttons by editing grub (sudo nano /etc/default/grub) and update the GRUB_CMDLINE_LINUX_DEFAULT line to include acpi_osi='!Windows 2012'. That did not work (yes I did sudo update-grub afterwards).

fan control

Did not complete this

It seems this is not active out of the box. Which means the fans spin up and down spin down, even after sleep/thaw. So that needs to be fixed.

Starting at this question. I tried installing lmsensors, which doesn't seem to exist. Then proceeded with installing fancontrol. Running it didn't work, claims missing /etc/fancontrol, so I installed psensor.

Then I double checked back on lmsensors and it turns out that's actually called lm-sensors, which was already exited. Okay :)

This is when I notice that the aforementioned link mentions nvidia and I realize the laptop isn't even using nvidia, so I went to "additional drivers" and installed it (352.63 proprietary tested) and also some proprietary "processor microcode firmware" which I suppose can't hurt much. Doubt this will affect fancontrol, though.

Running psensor shows a cute little UI with various sensors. It shows most values around 55°c or 65°c. I'm not sure what they ought to be at this point. It's pretty idle so it seems a little high to me, even with the fans spinning (at 3200rpm according to this ui). Note that the cli says nvidia information was not found. It's also complaining about a missing ~/.psensor/psensor.cfg file but I'm ignoring that.

Okay, lm-sensors doesn't work so how do you use it? Here is an answer: sudo sensors-detect. Press return on all the defaults except the last one; write the results to the config.

Interesting, after doing all that sensors works. It did not work before so maybe the module thing enabled that? The result is currently:

~$ sensors
acpitz-virtual-0
Adapter: Virtual device
temp1: +63.0°C (crit = +103.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Physical id 0: +64.0°C (high = +100.0°C, crit = +100.0°C)
Core 0: +62.0°C (high = +100.0°C, crit = +100.0°C)
Core 1: +54.0°C (high = +100.0°C, crit = +100.0°C)
Core 2: +52.0°C (high = +100.0°C, crit = +100.0°C)
Core 3: +51.0°C (high = +100.0°C, crit = +100.0°C)

asus-isa-0000
Adapter: ISA adapter
cpu_fan: 3200 RPM
temp1: +61.0°C

Alright. Moving on! Oh and you can do watch sensors to keep a running tab. Thanks to an answer in the last link. Pretty much the same as the psensor ui though, just in cli.

Note that installing the nvidia drivers did not fix the report from psensor that nvidia information could not be retrieved. Maybe a reboot is required for that, I'll check later.

The initial walkthrough claims switching to nvidia from xorg drivers fixes the fans but drains the battery. Can't really attest to the battery yet because it's powered, but the fans still don't work. Let's reboot first. ... No dice. The fans quiet down while in Grub but spin back up once logged in.

Trying to install laptop-mode-tools, supposedly a power savings package specifically for laptops. I'd imagine that being installed by default, but maybe not. ... It was not installed.

Running laptop_mode ... does nothing? It says "laptop mode enabled, not active", and "cannot find device eth0". At least that we can agree on.

to be continued.

Kernel 4.6.0

No luck so far, unable to boot a newer kernel beyond that of kernel 4.2 that comes with 15.10

(last tested kernel was 4.8.1rc4, did not work)

According to here the 4.6 linux kernel may also fix problems. Since the 4.4 did not work for me but reverting is trivial, I'm willing to give this a shot.

Downloaded the three .dev files at this page. Should be these:

linux-headers-4.6.0-040600_4.6.0-040600.201606100558_all.deb
linux-headers-4.6.0-040600-generic_4.6.0-040600.201606100558_amd64.deb
linux-image-4.6.0-040600-generic_4.6.0-040600.201606100558_amd64.deb

Then from the dir where you downloaded them execute sudo dpkg -i *.deb. That will install them, then reboot. I explicitly picked the advanced boot menu and 4.6 from there but I think it'll be the default.

While 4.4 (and 3.5) simply crashed while booting, this one drops me to a shell. To be complete, the tiny cli shows:

DMAR: DRHD: handling fault status req 2
DMAR: INTR_REMAP: Request device [[f0:1f.0] fault index 0
INTR_REMAP:[fault reason 37] Blocked a compatible format interrupt request
ALERT! /dev/disk/by-uuid/<some-uuid> does not exist. Dropping to a shell!

And now I'm in a "BusyBox" shell. Ooookay.

Searching for the "handling fault status req 2" lands me to this result mentions intel hardware. Something I've seen mentioned in the above links. This particular link says it solved the problem with ntel_iommu=igfx_off, though I'm not sure whether this is the fix so I'm not testing this yet. I have a feeling it's going to be this intel/nvidia bridge bullshit again. Yes, again. Have had similar problems on a previous laptop with the Optimus bridge before.

Okay this isn't helping. Doing a ctrl+alt+del and booting back to 4.2, which worked.

Booting to bios to check intel/nvidia options, but there aren't any. The only real option that might be related is to disable some virtualization, but I'm not so sure this is related at all. I don't remember this thing coming with an intel card, anyways. But that hasn't stopped me from getting them before.

Going to try the nouveau.modeset=0 override. Drop to Grub and press e to get into a cli config editor. MMMmmm but how to add. There is list of linux kernel boot parameters but no examples so I can't see the expected syntax. Ah, here is an example, so I guess just adding it to the same line and quoting that, so acpi_osi="nouveau.modeset=0". Add it to the end of the line and press f10 to boot (ctrl+x didn't work for me?). Perhaps I should also get rid of the splash screen there... Anyways, no dice, dropped to shell again. Reverting this change. Oh, it didn't save. I guess editing it this way is a one time thing? Alright!

Trying with just nouveau.modeset=0 appended this time. Just in case. Doesn't look to be working, though. Still not booting to 4.6, dropping me to shell instead. Let's remove the 4.6 for now.

Following steps here, installing synaptic, searching for the linux-image, and removing it. Just sort the results by "installed version" and remove the 4.6 one. I only had one 4.6 choice in my case. Then press "apply" and follow through. Reboot to confirm. Indeed, no more 4.6 in the grub advanced menu. Pfew, at least this works :)

FYI: the "quiet" keyword will limit the output while booting (source) and the "splash" keyword, kind of obviously, will cause a splash screen to be shown (source). Can probably edit /boot/grub/grub.cfg to permanently remove them (I like to see the linux crap while booting :)).

Okay, there's also a nomodeset parameter which should "disable kernel mode setting". So let's try that (grub, e) and boot to 4.2. ... This shows me a 800x600 res boot (which makes sense if no video drivers were loaded). Okay so at least modifying that thing works :)

The file to edit for grub is actually /etc/default/grub. In particular, for us, the GRUB_CMDLINE_LINUX_DEFAULT, which defaults to "quiet splash". I've removed those two and left it empty for now. AFter reboot ... I still get a splash screen. Damnit?

That link mentions to make sure "plymouth-theme-ubuntu-text" doesn't exist. So I'm bluntly removing it (sudo apt-get remove plymouth-theme-ubuntu-text), but it doesn't change anything either. Okay let's not care about that for now. Don't do this...

Actually, I just noticed the sudo update-grub2 line now, so let's quickly run that. Hey, my terminal looks different. I guess the package is related to that. I kind of like this new look, though, so maybe I'll keep it as is. But the mouse looks stupid so maybe not... :) Rebooting. Yeahhh, no more splash screen and the [OK] fuzzies, which obviously go too fast to read, but I like this over the splash screen that means nothing. Okay, progress. So to edit grub you need to modify /etc/default/grub and then run sudo update-grub2 to actually compile the update.

Reinstalling the plymouth package because half the mouse pointers are super big now. Reboot, no worky. Oops? This may have removed more than I wished for.

To fix this edit sudo nano /etc/X11/Xresources/x11-common and add Xcursor.size: 22 to that file. Save and reboot and the cursor is forced to be okay. Mostly. The hand pointer is still big. Blah, now it's all messed up :(

Kernel 4.7.1

Basically repeats the steps above but with the currently latest stable kernel v4.7.1. Download the relevant files from http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.7.1/ (the "all", and both "generic" files from amd64). Then run sudo dpkg -i linux*.deb where you downloaded them (may want to move them to their own dir if your download dir contains other .debs).
It complained about "probably" missing some drivers but I ignored that. Reboot. It does booth further than before, but drops me to a shell eventually. At this point it may be related to eufi boot or something (see below somewhere) so I'm not even sure whether it's still caused by the kernel incompat. But it sure as hell is annoying.

Actually, reading more closely, it may simply be missing the certain drivers it was mentioning on install? Damnit...? Ok, back to 4.2 (you can still select it from the advanced menu) and remove 4.7.1. But hey, we got further. It's no longer just crashing on boot? Still a little disappointing.

The missing firmware, btw, is "/lib/firmware/i915/kbl_dmc_ver.... No idea, I think that's wifi? If it just looks like wifi then I that's probably why I think that.

Having a little confidence in being able to quickly revert the kernels I also tried 4.8.0rc2, just in case. Same steps as above, except for a different file. Again reporting some crap on nvidia. I really hope it's not nvidia that's crapping out here. But 4.8rc2 led to the same dropping me to a shell.

I'm going to try and remove nvidia stuff, for now, since it's not really doing anything anyways, and make sure that's not causing my problem. I actually think it's some weird disk problem but I'm not sure how to confirm that one. sudo apt-get remove nvidia-* then sudo dpkg -i linux*.deb to reinstall 4.8 (I think that's relevant?"). Now it's complaining about possibly missing various nvidia drivers. Okay. *reboot*. Nope. Okay, removing and giving up, for now. May want to try this with a clean install, but at this point I don't really want to go there yet.

Video

Little luck so far. Can install the drivers, make it use them, but it still won't detect nvidia or show the card in nvidia's settings.

(FYI: in attempt 2 I did a clean install of nvidia and it works now. So I probably installed one or two packages too many.)

Even though I installed the nvidia drivers by now I don't have nvidia control panel stuff in my settings. I'm a little confused because running nvidia-redetector says "none", which I take to mean it can't detect an nvidia card. That's even more confusing because this model should have one? Okay, back to the bios. Remember, press or hold escape during the boot screen to proc the bios stuff.

Nothing in the bios to switch or whatever between nvidia and intel. No optimus stuff (phew) either. Just ... nothing really. The only graphics option is something about DVMT (never heard of it). Weird weird. Back to linux.

Installed nvidia again, since I have no "additional drivers" option anymore in configuration (wtf?) through console. Not sure what to pick so I went with sudo apt-get install nvidia-current-updates. That popped up a confirmation to kill eufi boot. Wait. No let's not do that. So I canceled that one. After a reboot it took for-ever to boot up. There was a long pause, then some more long pauses, and then it was waiting for three jobs to start (2x dev-disk-by, 1x dev-mapper-cryptswap1). Eventually it dropped me to an emergency mode cli. Shit... it seems the disk had a problem? And I'm seeing /boot/efi, ahw crap, did it fubar the boot anyways? :'(

It seems to be a hitch (or maybe something corrected/reinstated itself) because it booted fine after a reboot. PHEW. Okay the hunt is on.

Doing lspci gets you this list:

00:00.0 Host bridge: Intel Corporation Sky Lake Host Bridge/DRAM Registers (rev 07)
00:01.0 PCI bridge: Intel Corporation Sky Lake PCIe Controller (x16) (rev 07)
00:02.0 VGA compatible controller: Intel Corporation Device 191b (rev 06)
00:04.0 Signal processing controller: Intel Corporation Device 1903 (rev 07)
00:14.0 USB controller: Intel Corporation Sunrise Point-H USB 3.0 xHCI Controller (rev 31)
00:14.2 Signal processing controller: Intel Corporation Sunrise Point-H Thermal subsystem (rev 31)
00:15.0 Signal processing controller: Intel Corporation Sunrise Point-H LPSS I2C Controller #0 (rev 31)
00:15.1 Signal processing controller: Intel Corporation Sunrise Point-H LPSS I2C Controller #1 (rev 31)
00:16.0 Communication controller: Intel Corporation Sunrise Point-H CSME HECI #1 (rev 31)
00:17.0 SATA controller: Intel Corporation Sunrise Point-H SATA Controller [AHCI mode] (rev 31)
00:1c.0 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #2 (rev f1)
00:1c.2 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #3 (rev f1)
00:1c.4 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #5 (rev f1)
00:1d.0 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #9 (rev f1)
00:1f.0 ISA bridge: Intel Corporation Sunrise Point-H LPC Controller (rev 31)
00:1f.2 Memory controller: Intel Corporation Sunrise Point-H PMC (rev 31)
00:1f.3 Audio device: Intel Corporation Sunrise Point-H HD Audio (rev 31)
00:1f.4 SMBus: Intel Corporation Sunrise Point-H SMBus (rev 31)
01:00.0 3D controller: NVIDIA Corporation GM107M [GeForce GTX 960M] (rev a2)
02:00.0 Unassigned class [ff00]: Alcor Micro Device 6621
03:00.0 Network controller: Intel Corporation Wireless 7265 (rev 59)
3d:00.0 Non-Volatile memory controller: Samsung Electronics Co Ltd Device a802 (rev 01)

Which only lists the intel card. But then doing a super verbose output gets us (amongst many others)...

01:00.0 3D controller: NVIDIA Corporation GM107M [GeForce GTX 960M] (rev a2)
Subsystem: ASUSTeK Computer Inc. Device 1080
Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- Interrupt: pin A routed to IRQ 16
Region 0: Memory at dc000000 (32-bit, non-prefetchable) [size=16M]
Region 1: Memory at b0000000 (64-bit, prefetchable) [size=256M]
Region 3: Memory at c0000000 (64-bit, prefetchable) [size=32M]
Region 5: I/O ports at e000 [size=128]
Expansion ROM at dd000000 [disabled] [size=512K]
Capabilities:


Heeey.... what's going on here? Okay, so we do have that intel/nvidia crap going on. And probably optimus as well, just without the bios stuff. Well ain't that 30 levels of stupid.

Let's try this nvidia thing again, then. sudo apt-get install nvidia-352 aaand go. Aaaand stop. I don't want to disable secure boot. Meh.

Looking up the secure boot issue seems to be a temporary thing where you sign a package you compiled yourself and then reenabling it. Blabla, I'm not sure so I don't really want to touch it. Most of all I don't want to risk scrapping the windows boot somehow. Not sure I can get the OEM back safely.

So with xubuntu striking out itself, let's try to download something from the website directly. Download the script from their website. chmod +x nvidia<tab> and sudo ./nvidia*lt;tab. And we need to exit X? GRmbl.

Using instructions found here I went to a terminal (ctrl+alt+f1), logged in, killed the x server (sudo service lightdm stop), "entered runlevel 3" (whatever that means) sudo init 3, and run the install script again sudo nvida....

I ignored the "pre install errors" warning, because I'm getting a little fed up with it all. I pressed enter at the "register for kernel updates" screen, which I think said no but intended to say yes there but okay, said yes to signing the driver, got an error related to signed drivers not loading, bashed head against desk a few times, and got kicked out of the installer. This should be fun... *reboot* Nah it booted fine. Okay.

Next; installing "nvidia current" from the ubuntu software package manager thing. Who knows, right? It worked, no reported hitches, but I don't think it really changed anything. No nvidia options in the settings. No obvious changes at all. Grrr.

I tried bumblebee (sudo apt-get install bumblebee) but the sudo optimus ... commands don't seem to change anything.

(At this point I realized I should update the bios, and so I have, with no noticeable difference)

Since the settings don't show an option for the driver thing anymore (nice bug) you can still find it through the lens thing (you know, where your apps are). Just search for "additional drivers" and you'll get it.

I switched it to nvidia through that. Neither option I had (352.93 either "propriatary" or "propriatary,tested") seemed to have any results, though. They installed (I think). But running the nvidia server settings thing basically gave me no options, an indication it's not working. Not even after a reboot. Grrmbl.

(skip to a few weeks later)

Still trying to get stuff to run.

$ sudo nvidia-detector
none
$ sudo ubuntu-drivers devices
== cpu-microcode.py ==
driver : intel-microcode - distro non-free

== /sys/devices/pci0000:00/0000:00:01.0/0000:01:00.0 ==
model : GM107M [GeForce GTX 960M]
vendor : NVIDIA Corporation
modalias : pci:v000010DEd0000139Bsv00001043sd00001080bc03sc02i00
driver : xserver-xorg-video-nouveau - distro free builtin
driver : nvidia-352 - distro non-free recommended
driver : nvidia-352-updates - distro non-free

Okay.. still confused by this. Still no nvidia option in settings and manually running nvidia-settings (neither with nor without root) shows me a card, just the two default "settings" options.

Running a bit through this thread it suggests running lshw to show active display drivers. Okay...

$ sudo lshw -C display
*-display UNCLAIMED
description: 3D controller
product: GM107M [GeForce GTX 960M]
vendor: NVIDIA Corporation
physical id: 0
bus info: pci@0000:01:00.0
version: a2
width: 64 bits
clock: 33MHz
capabilities: pm msi pciexpress bus_master cap_list
configuration: latency=0
resources: memory:dc000000-dcffffff memory:b0000000-bfffffff memory:c0000000-c1ffffff ioport:e000(size=128) memory:dd000000-dd07ffff
*-display
description: VGA compatible controller
product: Intel Corporation
vendor: Intel Corporation
physical id: 2
bus info: pci@0000:00:02.0
version: 06
width: 64 bits
clock: 33MHz
capabilities: pciexpress msi pm vga_controller bus_master cap_list rom
configuration: driver=i915 latency=0
resources: irq:125 memory:db000000-dbffffff memory:70000000-7fffffff ioport:f000(size=64)

Soooo, regular desktop uses intel and 3d stuff will use nvidia. Super. Note that I can run glxgears at over 100fps at the moment, usually an indicator that _some_ drivers are doing their thing.

Let's try to purge all the intel stuff, hopefully forcing the system to move to nvidia...? At this point I've given up on the current installation and don't mind b0rking it. So I'm willing to run some more risky stuff to fluke a fix :)

sudo apt-get purge nvidia-*
// reboot. after boot things look the same, sans the drivers
sudo apt-get remove xserver-xorg-video-intel
// reboot. after boot i get 4k resolution sans, sans the upscaling (so everything is super tiny)

Okay looks let's see the output of lshw now. ... The same.

What about lspci -nnk? Nearly the same. The intel card still seems to use the i951, which I think is broken somehow. The nvidia "3D" card currently has no driver.

I've been looking to remove the i951 thing but can't seem to find anything about it. There's a good chance you were led to this page when looking for information yourself. I'm sorry for wasting your time, unless I managed to fix it eventually.

Onwards. with no driver in use I'm going to install nvidia, again. This time I'm going to do the uefi thing and see what happens. If it b0rks win10, so be it.

The "cli ui" asks to install uefi, then asks you to enter a password and repeat it. It will be asked on next boot. Okay...

On boot I get a new tiny blue popup as the first thing when booting (so on/before bios splash). It will boot in 5 seconds and I just let it. I think it would have allowed me to disable the uefi or something? Unsure...

Once booted the login screen background was properly scaled (the xfce mouse is now visible, it wasn't before). So something changed at least? Stuff is still at 4k (aka. minute) so I switched back to FD resolution.

I feel like I'm back at square one with intel still being the leading driver. Grrrrr. The lspci -nnk output did not change in the display bits. 'Far as I can tell.

So now I'm thinking the nvidia installation kind of failed because I didn't do the blue popup properly. So let's try that again. Purge, install, reboot. Noticing now that the nvidia installation screen indeed says the system will boot without the new drivers if I failed to disable the secure boot. Okay.

In the blue secure popup (which may be intended to be a much larger thing but due to 4k is just tiny) I went and disabled secure boot. You don't actually enter your password but instead it asks you questions about your password like 90's game DRM which asked questions about the manual. "Enter character 3 of your password" that sort of thing. Kind of weird, but okay I guess there's a reason for it.

It reboots again. And .... gets stuck. Stuck while spinning up the fans to max. That's a first. So _something_ changed, but now I'm not certain whether I've permanently screwed this installation (not worried about the machine, it's just a driver). Can not jump to a terminal (altctrlf1) so let's ... reboot again! :) (even keyboiard was stuck there so a hard power-button-reset was required).

The reboot ended in the same state. Okay let's try default recovery mode. I see a prompt but keyboard doesn't work and after a little while the fans go into lift-off mode again. mmmmmmm okay. I also still get the "booting in insecure mode" and I'm not sure how to disable (ok, re-enable) it.

So while the "main" kernel (4.2.0-42) no longer works, I had another kernel (4.2.0-16) still installed and that one did boot.

Since stuff is breaking down pretty hard the gloves are now off. Something seems to have budged with the video card, at least, so maybe the 4.7 kernel will work now? I'm not holding my breath.

No, dropped to primitive shell as usual. And the card doesn't go into turbo fan mode either (was thinking maybe it was a default thing).

Ok here's something. Right now pressing apply after selecting nvidia drivers in the "additional drivers" tab just bumps my selection back to the nouveau driver. Thanks for that great piece of UI design, by the way. But if I go the first tab "ubuntu software" and select the main server instead of the dutch server, then back to apply for an nvidia driver, it actually does work. That's ... well, that's just seven layers of messed up. But okay, I think it's actually using the nvidia drivers now...? :) As an aside, my guess is that 15.10 is no longer supported and the server just returns a 404 for some files. The UI is just super bad and so the error is suppressed. But that's my guess. HEY I WOULD BE ON 16.04 IF IT'D LET ME

After a reboot, no noticeable changes. No nvidia in settings. No card to select in nvidia-settings, still "none" for nvidia-detector. Feels like a shallow victory. Indeed, lspci -nnk still shows no sign of nvidia drivers, even though the additional drivers tab claims the nvidia drivers are currently in use.

The main thing that bothers me in this is that nvidia-detector (or anything) refuses to detect the video card at all. Even when various other tools do see the card, even have drivers installed for the card. What is this magic. Even glxinfo reports Mesa (glxgears currently runs at 60fps).

Following these steps because this post had some luck. The bumblebee package already exists, I installed xserver-xorg-video-intel because I think that xf package is for arch. Not even sure if I need it at all, but whatever. Following the steps there did not lead to a victory. You live to die another day, nvidia. Or actually, the other way around.