UNSOLVED
New XPS 9370 Linux - Very strange boot and suspend behaviours
I just got my 9370 with Ubuntu on it. After 48 hours I've had nothing but frustration with it and have now bricked it trying to fix things. I'm really stumped and now trying to install Debian on it and finding it's behaving really strangely.
First, with the Dell shipped and configured Ubuntu it seemed suspend didn't work. Every time I closed the lid and opened it the keyboard would light up and the two leds would flash on and off but the screen didn't come back. This was before I touched anything.
First I tried updating the BIOS in hopes this was fixed. But that didn't fix it.
Then I updated Ubuntu with apt-get upgrade and apt-get dist-upgrade and installed newer kernels (still from the same Ubuntu LTS 16.08 iirc). These kernel packages managed to mess up the EFI install and I had to switch to "Legacy" boot mode to get it to boot at all.
But then I rather alarmingly got a "Hard Drive not present" error. And indeed when I ran the "Diagnosis" option it said the same thing. Experimenting it seems the NVMe drive doesn't show up if the power adapter is plugged into the USB-C port on the right hand side of the device.
In the process of experimenting I managed to wipe the partition table which so now I'm trying to install Debian on it but not only do I have to make sure the power adapter is not on the right hand side, it seems the boot process only recognizes USB drives if they *are* on the right hand side USB-C port. If I plu;g it in to either of the left hand side USB-C ports the USB flash drive isn't recognized at boot time.
I think I've seen the "Drive Not Present" error at least once when the power adapter was not plugged in on the right side so I'm not sure I have the diagnosis exactly right.
Is this right? Is there a big difference in the boot process between the different usb-c ports even if you're not doing any thunderbolty things and just using power delivery and usb 2.0 flash drives? That would be really weird and it would be weirder still if it actually caused the internal NVMe drive to behave differently.
