core dumps reported in latest fwts

[Jeff Lane ]

On a couple runs using the latest version of fwts, I have noticed that fwts is creating a core dump (though it still seems to exit cleanly. I'm not sure which test it is failing on, though the uefidump test was previously the culprit (see bug 1174947)

I have also attached the fwts output log hoping maybe the key is in there somewhere...

This was most recently seen on a server being certified: https://certification.canonical.com/hardware/201409-15514/

Aborted (core dumped)
High Failures: 6
WARNING: The following test cases were reported as high
level failures by fwts. Please review the log at
/root/.cache/plainbox/sessions/pbox-nis0on_a.session/CHECKBOX_DATA/fwts_results.log for more information.
 - uefirttime
 - mpcheck
 - method
 - securebootcert
 - klog
 - syntaxcheck
Medium Failures: 3
WARNING: The following test cases were reported as medium
level failures by fwts. Please review the log at
/root/.cache/plainbox/sessions/pbox-nis0on_a.session/CHECKBOX_DATA/fwts_results.log for more information.
 - fan
 - dmicheck
 - aspm
Passed: 27
 - oops
 - csm
 - hpet_check
 - apicinstance
 - acpiinfo
 - fadt
 - version
 - nx
 - pciirq
 - cstates
 - pcc
 - crs
 - mcfg
 - apicedge
 - uefirtvariable
 - maxreadreq
 - bios_info
 - checksum
 - dmar
 - acpitables
 - wmi
 - uefirtmisc
 - maxfreq
 - msr
 - os2gap
 - virt
 - pnp

[Colin Ian King (colin-king)]

Thanks for reporting this issue Jeff. Can you run the following:

sudo apt-get update && sudo apt-get install acpidump
sudo acpidump > acpidump.log

and also:

sudo fwts - > fwts.log

And attach both logs to the bug. Thanks!

[Colin Ian King (colin-king)]

It looks like the acpidump is failing, but I really require the ACPI tables to double check this and corner the bug.

[Jeff Lane  (bladernr)]

201407-15326/submission/100812/test-results/

here's another one showing the Aborted (Core dumped) message in the output. Likely it's the same thing, but I'm attaching the log here for reference.

[Colin Ian King (colin-king)]

@Jeff, a copy of the acpidump for these failures is always helpful so I can try and reproduce the bug from that raw ACPI data. Thanks!

[Samantha Jian-Pielak (samantha-jian)]

fwts on 201409-15514

[Samantha Jian-Pielak (samantha-jian)]

acpidump on 201409-15514

[Colin Ian King (colin-king)]

I've build a version of fwts containing a way of tracing core dumps. The .deb packages are available to download from here: http://kernel.ubuntu.com/~cking/lp-1376448/

Can the appropriate packages be installed and tested. I am hoping to capture output on stderr that contains a stack dump.

Thanks,

Colin

[Colin Ian King (colin-king)]

@Jeff, any chance of getting this debug version of fwts run on the machine again?

[Colin Ian King (colin-king)]

If we can't test this, I'll close the bug in 4 weeks and wait until it get picked up again in subsequent faults. As it stands, the latest fwts can catch segfaults and will report where it is occurring.

@Jeff, is it possible to run the debug version of fwts from http://kernel.ubuntu.com/~cking/lp-1376448/ on this kit, or shall we close this bug for now?