Bug 1219717 - Installation on aarch64 cpu max on qemu fails with segfault (signal 11)
Summary: Installation on aarch64 cpu max on qemu fails with segfault (signal 11)
Status: RESOLVED DUPLICATE of bug 1223072
Alias: None
Product: openSUSE Tumbleweed
Classification: openSUSE
Component: YaST2 (show other bugs)
Version: Current
Hardware: aarch64 Other
: P5 - None : Normal (vote)
Target Milestone: ---
Assignee: E-mail List
QA Contact: Jiri Srain
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2024-02-08 08:21 UTC by Guillaume GARDET
Modified: 2024-04-24 07:23 UTC (History)
2 users (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
Screenshot (32.86 KB, image/png)
2024-02-08 08:21 UTC, Guillaume GARDET
Details
y2log (6.17 MB, text/plain)
2024-02-08 08:21 UTC, Guillaume GARDET
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Guillaume GARDET 2024-02-08 08:21:02 UTC
Created attachment 872571 [details]
Screenshot

Installation on aarch64 cpu max (no KVM) on qemu fails in openQA since snapshot 20240128 which includes the switch to ruby 3.3 [0]. 
This particular qemu configuration allows to test for PAC (Pointer Authentication) and BTI (Branch Target Identifier), so likely the origin of this failure.

To reproduce the failure locally, I used the following command line:
**********
qemu-system-aarch64 -m 2000 -cpu max,pauth-impdef=true -smp 2 -M virt -serial stdio -device virtio-net,netdev=hostnet0,mac=52:54:00:09:a4:37 -netdev user,id=hostnet0 -drive id=pflash-code-overlay0,if=pflash,file=pflash-code-overlay0,unit=0,readonly=on -drive id=pflash-vars-overlay0,if=pflash,file=pflash-vars-overlay0,unit=1 -drive if=none,file=$ISO,id=hd0  -device virtio-blk-device,drive=hd0 -drive if=none,file=$HDD,id=hd1  -device virtio-blk-device,drive=hd1 -device virtio-gpu-pci -device nec-usb-xhci -device usb-tablet -device usb-kbd -audiodev none,id=snd0 -device intel-hda  -device hda-output,audiodev=snd0
**********

You also need to set $ISO and $HDD and also create pflash* files with:
**********
PFLASH_CODE="/usr/share/qemu/aavmf-aarch64-opensuse-code.bin"
PFLASH_VARS="/usr/share/qemu/aavmf-aarch64-vars.bin"
/usr/bin/qemu-img create -f qcow2 -F raw -b $PFLASH_CODE pflash-code-overlay0 67108864
/usr/bin/qemu-img create -f qcow2 -F raw -b $PFLASH_VARS pflash-vars-overlay0 67108864
**********


[0]: https://lists.opensuse.org/archives/list/arm@lists.opensuse.org/thread/SLKI5O7L3IRN5K3YUMWZN6VB4JCE5ITS/
Comment 1 Guillaume GARDET 2024-02-08 08:21:24 UTC
Created attachment 872574 [details]
y2log
Comment 2 Stefan Hundhammer 2024-02-08 09:10:45 UTC
"Installation fails" is much too generic.

WHAT failed at WHAT point? What does the test do? How far does it get?
Comment 3 Stefan Hundhammer 2024-02-08 09:14:34 UTC
In the attached y2log (which is NOT what we need for debugging any YaST problem; we need a full y2logs tarball generated with the supplied 'save_y2logs' script; see also the Bug Reporting FAQ where I documented this 20+ years ago), I see:


2024-02-07 19:55:20 <5> install(3664) [zypp] ZYppFactory.cc(backtraceHandler):57 Error: signal 11


So libzypp segfaults; probably. Or is it an OOM? Or something else? Impossible to tell without the full y2logs tarball.

That would also answer questions such as "how far does it get?" without having to go through that lone y2log file line by line.

So please attach y2logs generated with the supplied 'save_y2logs' script.
Comment 5 Stefan Hundhammer 2024-02-13 09:20:57 UTC
See comment #3: Still waiting for y2logs.
Comment 6 Guillaume GARDET 2024-02-13 12:56:58 UTC
(In reply to Stefan Hundhammer from comment #5)
> See comment #3: Still waiting for y2logs.

Yes, but the machine I used to reproduce the issue broke and I need to fix it first. Sorry for the delay.
Comment 7 Stefan Hundhammer 2024-02-26 09:22:59 UTC
.
Comment 8 Guillaume GARDET 2024-03-08 15:21:14 UTC
I cannot reproduce anymore locally and in openQA either: https://openqa.opensuse.org/tests/3993648
Comment 9 Guillaume GARDET 2024-04-24 07:23:50 UTC

*** This bug has been marked as a duplicate of bug 1223072 ***