HP Stream 11 x360 (HP 11-ab00x) Hangs after init.xz

[JJ Fullmer]

@Tom-Elliott One video coming up. It’s hanging after it claims init.xz is loaded but then never leaves that screen to enter FOS. Here’s a picture of where it hangs
And here’s a link to the video from boot to hang
https://photos.app.goo.gl/PRxVATZNfY1N89P56

@Sebastian-Roth It has worked before, however I have a lot of the older ipxe and have tried them and they’re no longer working. I did attempt to install android x86 on this thing a while back, which didn’t work, hopefully that didn’t have some sort of unforseen consequence.

We have 2 startech USB21000S2 usb to ethernet adapters. In the video I run load usb.efi in the efi shell which is a renamed version of the efi driver for this adapter. I have a link to it posted elsewhere in the forum. I used the exact same adapter to image an asus tablet device right before trying to reimage this one.

[Sebastian Roth]

@JJ-Fullmer said in HP Stream 11 x360 (HP 11-ab00x) Hangs after init.xz:

I used the exact same adapter to image an asus tablet device right before trying to reimage this one.

This doesn’t mean much. One particular adapter may play well with one tablet but not with a different one. This is because the tablet’s (or PC’s or machine’s in general) firmware needs to have a driver for that specific network adapter.

In the video we see that iPXE seems to get an IP and load the kernel properly so I guess we see a hang in the Linux Kernel. Try using the bzImage_rb kernel image from here: https://drive.google.com/drive/folders/0B-bOeHjoUmyMV095YVpsR3U5VFk

See if that helps.

[JJ Fullmer]

@Sebastian-Roth I suppose a kernel problem makes more sense than an efi boot file.
I gave your kernel a try, thanks for that, sadly it did not work and gave me a kernel panic. However it did get past that initial hang.
Here’s a video
https://photos.app.goo.gl/1QTpmhz5syBno2eE6

[JJ Fullmer]

Re watching the video I can see that you can’t really read any of the output as I set the resolution lower to be able to upload it faster. I’ll take some pictures of the output and post them shortly so you can actually read the error messages.

[JJ Fullmer]

@Sebastian-Roth

[Sebastian Roth]

@JJ-Fullmer Thanks for the pictures! I am fairly sure I know what is causing the hang on this device. Is this a HP x2 210 tablet??

I’ll compile a special kernel image for you that skips some PCI initialization code that is causing big troubles on some devices (namely that HP x2 210)!

[JJ Fullmer]

@Sebastian-Roth Here is the full inventory that fog got on the tabtop

System Manufacturer HP
System Product HP Convertible x360 11-ab0XX
System Version Type1ProductConfigId
System Serial Number CND6381606
System UUID 75d38b44-b570-e611-b27e-98e7f4db4d7f
System Type Type: Notebook
BIOS Vendor Insyde
BIOS Version F.14
Motherboard Manufacturer HP
Motherboard Product Name 82E1
Motherboard Version 13.16
Motherboard Serial Number PGDJPF21U41RCZ
Motherboard Asset Tag Type2 - Board Asset Tag
CPU Manufacturer Intel Corporation
CPU Version Intel Celeron CPU N3060 @ 1.60GHz
CPU Normal Speed Current Speed: 1600 MHz
CPU Max Speed Max Speed: 1660 MHz
Memory 3.69 GiB
Hard Disk Model
Hard Disk Firmware
Hard Disk Serial Number
Chassis Manufacturer Hewlett-Packard
Chassis Version
Chassis Serial Chassis Serial Number
Chassis Asset <BAD INDEX>

edited
So no it’s not an HP x2 210 tablet, but I wouldn’t doubt it’s got enough similarity to have the same problem being an HP x tablet.

[JJ Fullmer]

@Sebastian-Roth I also tried various old kernels from the fog kernel update (giving them different names of bzImage of course) non of which worked but had either similar errors or just hung.
So a special kernel is greatly appreciated.

[Sebastian Roth]

@JJ-Fullmer Ok, here we go, try this kernel: https://fogproject.org/kernels/Kernel.SR.SKIP_PCI.4.8.11.64

[JJ Fullmer]

@Sebastian-Roth Sadly it just hung as well. Thanks for giving it a try though, I know compiling a kernel isn’t a 2 second process.

[Sebastian Roth]

@JJ-Fullmer As the kernel code is changing and I never saved that dirty hack as a proper patch I need to make this up again from memory all the time. So possible my hack was just wrong the first time. I now added some debug output to see if I am doing things at the right place. Please download from the same URL again, I just updated the binary a minute ago: https://fogproject.org/kernels/Kernel.SR.SKIP_PCI.4.8.11.64

[JJ Fullmer]

@Sebastian-Roth Gotta love how reflective this computer’s screen is…Took a picture of the debug output
https://photos.app.goo.gl/mVL8zAGAi9PN2sNc7

[Sebastian Roth]

@JJ-Fullmer Ohhhh my …, must have lost track of this in the middle of a hectic months. Sorry! Do you still have those devices at hand to test?

[JJ Fullmer]

@Sebastian-Roth I do still have it on hand if you have any new ideas. I just ended up going through my image prep process with a fresh install of windows as if I was making the image on this computer. It was much much slower of course, but it worked. This one in particular has been deployed now, but if we have something promising I could probably pull it back from production and give the user something else in the interim.

[Sebastian Roth]

@JJ-Fullmer I’ll look through the code again in the next days and post new kernels to test if you are available and have a device to test.

[Sebastian Roth]

@JJ-Fullmer Sorry for the long delay again. I am back on track now to tackle this. Do you think you can get it back for some testing in the next couple of days? I am preparing debug kernels already. So I’ll be ready to go when you are.

[Sebastian Roth]

@JJ-Fullmer Ok, I just prepared a first one to test for you - download 1_bzImage and put in /var/www/html/fog/service/ipxe. Then edit this host’s settings and set this kernel as Host Kernel and see what you get.

I used a fairly recent kernel version 4.19.13 and so we might even see different behaviour than when we tested some months ago.