Bug ID 1153835
Summary [20190925] Tumbleweed Installer hangs on Newport GW6304 (USB)
Classification openSUSE
Product openSUSE Tumbleweed
Version Current
Hardware aarch64
OS openSUSE Factory
Status NEW
Severity Major
Priority P5 - None
Component Kernel
Assignee yousaf.kaukab@suse.com
Reporter afaerber@suse.com
QA Contact qa-bugs@suse.de
CC mbrugger@suse.com, oneukum@suse.com, support@gateworks.com
Found By ---
Blocker ---

Created attachment 821336 [details]
Boot log with 20190925 from USB

After flashing the latest firmware-newport.img 0:0x8000 for USB storage support
and inserting an empty microSD card and a USB stick with the DVD .iso dd'ed
onto:

http://download.opensuse.org/ports/aarch64/tumbleweed/iso/openSUSE-Tumbleweed-DVD-aarch64-Snapshot20190925-Media.iso
(note: file may disappear once newer versions get published)

GW6304-SP391-D2> usb start && load usb 0:1 $kernel_addr_r efi/boot/bootaa64.efi
GW6304-SP391-D2> bootefi $kernel_addr_r $fdtcontroladdr

grub> configfile (hd1,1)/efi/boot/grub.cfg

Edit "Installation" to append to the "linux" command line:
cma=64M console=ttyAMA0,115200n8 ssh=1 sshpassword=linux

This leads to a USB-related hang:

[   77.235917] udevd[241]: usb1: Worker [250] processing SEQNUM=2069 is taking
a long time
...
[  197.235966] udevd[241]: usb1: Worker [250] processing SEQNUM=2069 killed
[  197.242881] udevd[241]: Worker [250] terminated by signal 9 (KILL)
[  197.249067] udevd[241]: usb1: Worker [250] failed
[  258.235941] udevd[241]: usb1: Worker [243] processing SEQNUM=2072 is taking
a long time
[  378.235954] udevd[241]: usb1: Worker [243] processing SEQNUM=2072 killed
[  378.242814] udevd[241]: Worker [243] terminated by signal 9 (KILL)
[  378.248999] udevd[241]: usb1: Worker [243] failed
[  996.599992] INFO: task kworker/0:0:5 blocked for more than 491 seconds.
[  996.606607]       Tainted: G        W         5.2.14-1-default #1
...

I don't see my USB stick getting announced before, only XHCI hubs.

There's a number of other issues in the attached log:
* "ARM_SMCCC_ARCH_WORKAROUND_1 missing from firmware" - i.e., pre-Spectre TF-A?
* Multiple WARNINGs from GICv3 code
* Should "fsl-mc MSI" be initializing itself on Octeon TX platform?
* 'arm-smmu: deprecated "mmu-masters" DT property in use' & tons of DMA
warnings
* "irq: no irq domain found for gsc@20" leading to gpio_keys failing?
* "xhci_hcd 0000:00:10.0: cache line size of 64 is not supported"
* "nicvf 0000:05:00.1: Driver probe function unexpectedly returned 1"


You are receiving this mail because: