[iotg][tgl][tgl-aaeon] 20211006 image does not boot

Bug #1946223 reported by Doug Jacobs
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
intel
Fix Committed
Critical
Unassigned
Lookout-canyon-series
Fix Released
Critical
Unassigned
linux-intel-5.13 (Ubuntu)
Fix Released
Undecided
Unassigned
Focal
Fix Released
Undecided
Anthony Wong

Bug Description

The daily build containing the new kernel does not boot. It gets stuck about 2 seconds into the cycle with the message "Run /init as init process" (see attached image file for boot message sequence.)

System: TGL-Aaeon (CID: 202110-29509)

Steps to reproduce:
Downloaded focal-preinstalled-desktop-amd64+intel-iot.img.xz from https://cdimage.ubuntu.com/focal/daily-preinstalled/20211006/

Uncompressed file with unxz -k

Copied the .img file to my secondary USB drive.

Boot 20.04 LTS from primary USB drive.

Select Try Ubuntu

Wipe out /dev/sda (primary storage)

Use dd to copy the image from secondary USB drive to /dev/sda

Reboot, removing USB drives when prompted.

Expected Result:
System should boot Ubuntu Desktop with kernel 5.13-1006

Actual Result:
System hangs 2 seconds into the boot cycle with
"Run /init as init process"

Reproducibility: 100%

CVE References

Revision history for this message
Doug Jacobs (djacobs98) wrote :
Revision history for this message
Doug Jacobs (djacobs98) wrote :

Update: Apparently the system is not entirely hung. Although the mouse and keyboard are not responsive (mouse's light is off), the system just posted another bootup message:

[1606.334337] perf: interrupt took too long (2612 > 2500), lowering kernel.perf_event_max_sample_rate to 76500

(1600 seconds is nearly 30 minutes.)

Revision history for this message
Anthony Wong (anthonywong) wrote :

As a starter, can you add the PPA https://launchpad.net/~anthonywong/+archive/ubuntu/kernel-staging, then run 'apt install linux-intel-wip', you will get the linux-intel 5.13.0-1005 kernel. Reboot the machine and choose the 1005 kernel in grub.

Revision history for this message
Doug Jacobs (djacobs98) wrote :

I reinstalled 5.13.0-1003.

I was able to add the PPA, install 5.13.0-1005 and it boots.

I don't know what is different between installing 5.13.0-1005 directly and the method in comment #3.

Revision history for this message
Doug Jacobs (djacobs98) wrote :

I got the TGL and EHL boards backwards.

The build WORKS on the EHL-Aaeon, but NOT the TGL-Aaeon.

I've updated the tags and the initial description.

summary: - [iotg][ehl][ehl-aaeon] 20211006 image does not boot
+ [iotg][tgl][tgl-aaeon] 20211006 image does not boot
description: updated
Alex Hung (alexhung)
Changed in intel:
assignee: nobody → Alex Hung (alexhung)
Changed in intel:
importance: Undecided → Critical
Changed in intel:
status: New → In Progress
Jesse Sung (wenchien)
Changed in linux-intel-5.13 (Ubuntu Focal):
status: New → Fix Committed
Jesse Sung (wenchien)
Changed in linux-intel-5.13 (Ubuntu Focal):
assignee: nobody → Anthony Wong (anthonywong)
Changed in intel:
assignee: Alex Hung (alexhung) → nobody
status: In Progress → Fix Committed
Revision history for this message
Chao Qin (chaoqin) wrote :

According to https://bugs.launchpad.net/intel/+bug/1842239, please make sure the toolchain is CET enabled and rebuilt the userspace applications in the Ubuntu image (as I known CET is not enabled in 20.04).

This issue can be worked around by adding boot parameters no_user_ibt and no_user_shstk.

tags: added: tgl-aaeon
removed: ehl ehl-aaeon
tags: added: verification-done-focal
Revision history for this message
Launchpad Janitor (janitor) wrote :
Download full text (25.4 KiB)

This bug was fixed in the package linux-intel-5.13 - 5.13.0-1007.7

---------------
linux-intel-5.13 (5.13.0-1007.7) focal; urgency=medium

  * focal/linux-intel-5.13: 5.13.0-1007.7 -proposed tracker (LP: #1946503)

  * [iotg][tgl][tgl-aaeon] 20211006 image does not boot (LP: #1946223)
    - Revert CET patches

linux-intel-5.13 (5.13.0-1006.6) focal; urgency=medium

  * focal/linux-intel-5.13: 5.13.0-1006.6 -proposed tracker (LP: #1945669)

  * Packaging resync (LP: #1786013)
    - [Packaging] update variants

  * MEI (Intel Management Engine Interface) for sprint 2 (LP: #1945464)
    - Revert "mei: dal: add test module"
    - mei: backport fix from 5.12
    - Revert "UBUNTU: [Config] Disable INTEL_MEI_DAL and INTEL_MEI_VIRTIO"
    - [Config] Enable CONFIG_INTEL_MEI_DAL and CONFIG_INTEL_MEI_VIRTIO

  * [EHL] Intel ishtp VNIC driver (LP: #1943524)
    - net: Add support for Intel vnic driver
    - [Config] CONFIG_INTEL_ISHTP_VNIC=m

  * [EHL] Quadrature Encoder Peripheral support for sprint 2 (LP: #1945494)
    - counter: Add support for Intel Quadrature Encoder Peripheral
    - counter: intel-qep: Mark PM callbacks with __maybe_unused
    - counter: intel-qep: Use to_pci_dev() helper
    - [Config] CONFIG_INTEL_QEP=m

  * I225-IT Ethernet (8086:0d9f) does not work on AAEON's EHL Board
    (LP: #1945548)
    - igc: Remove _I_PHY_ID checking
    - igc: Remove phy->type checking

  * [EHL][TGL][ADL] Enable Time Coordinated Compute interface driver
    (LP: #1929903)
    - tcc: update RTCT table parser to support two versions
    - Enable support to read a few whitelisted registers.
    - Remove Clock_Cycles_VT from MHL entry.
    - Add new IOCTL to read error log buffer.
    - Display errlog buffer raw data in kernel log as requested once this driver
      is loaded.
    - Fix issue found in acrn uos when convert cacheid to apicid.

  * Integrated TSN controller for sprint 2 (LP: #1945461)
    - net: pcs: Introducing support for DWC xpcs Energy Efficient Ethernet
    - net: stmmac: Add callbacks for DWC xpcs Energy Efficient Ethernet
    - net: stmmac: enable platform specific safety features
    - net: phy: probe for C45 PHYs that return PHY ID of zero in C22 space
    - net: stmmac: Fix mixed enum type warning
    - net: stmmac: Fix unused values warnings
    - stmmac: intel: move definitions to dwmac-intel header file
    - stmmac: intel: fix wrong kernel-doc
    - stmmac: align RX buffers
    - net: stmmac: Fix mixed enum type
    - net: pcs: xpcs: delete shim definition for mdio_xpcs_get_ops()
    - net: pcs: xpcs: there is only one PHY ID
    - net: pcs: xpcs: make the checks related to the PHY interface mode
    - net: pcs: xpcs: export xpcs_validate
    - net: pcs: xpcs: export xpcs_config_eee
    - net: pcs: xpcs: export xpcs_probe
    - net: pcs: xpcs: use mdiobus_c45_addr in xpcs_{read,write}
    - net: pcs: xpcs: convert to mdio_device
    - net: pcs: xpcs: convert to phylink_pcs_ops
    - net: stmmac: split xPCS setup from mdio register
    - net: pcs: add 2500BASEX support for Intel mGbE controller
    - net: stmmac: enable Intel mGbE 2.5Gbps link speed
    - net: stmmac: fix NPD with phylink_set_pcs if there is no MDIO bus
    - n...

Changed in linux-intel-5.13 (Ubuntu Focal):
status: Fix Committed → Fix Released
Changed in linux-intel-5.13 (Ubuntu):
status: New → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.