Bug 1954143

Summary: crc32c_arm64(): fio killed by SIGILL
Product: Red Hat Enterprise Linux 8 Reporter: Martin Hoyer <mhoyer>
Component: fioAssignee: Pavel Reichl <preichl>
Status: CLOSED ERRATA QA Contact: Samuel Petrovic <spetrovi>
Severity: low Docs Contact:
Priority: unspecified    
Version: 8.4CC: preichl, spetrovi, yizhan
Target Milestone: betaKeywords: Triaged
Target Release: ---Flags: pm-rhel: mirror+
Hardware: aarch64   
OS: Linux   
Whiteboard:
Fixed In Version: fio-3.19-4.el8 Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2023-05-16 08:28:12 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
abrt report none

Description Martin Hoyer 2021-04-27 16:36:36 UTC
Created attachment 1776086 [details]
abrt report

Description of problem:
SIGILL while running fio with --verify=crc32c

cmdline:        fio --bs=4k --direct=1 --ioengine=libaio --iodepth=16 --verify=crc32c --verify_fatal=1 --do_verify=1 --rw=randwrite --runtime=120 --filename=/dev/mapper/mpathb --name=fio_test --numjobs=1 --verify_backlog=1024

Version-Release number of selected component (if applicable):
fio-3.19-3.el8

How reproducible:
once

Steps to Reproduce:
1. establish local iscsi session, enable multipath
2. run fio command with crc32c verification

Actual results:
SIGILL

Expected results:
no errors

Additional info:
please see the attached abrt report

Comment 2 Eric Sandeen 2022-09-27 21:13:59 UTC
Sorry for losing track of this.

This is probably known issue https://github.com/axboe/fio/issues/1239 - lots of arm packages got burned by this, it seems.

It's probably fixed in fio-3.29.  I built a test rpm w/ the fix, would you like to give it a re-check?

http://download.eng.bos.redhat.com/brewroot/work/tasks/7416/47987416/fio-3.19-4.el8.test1.aarch64.rpm

Thanks,
-Eric

Comment 3 Martin Hoyer 2022-09-29 13:51:45 UTC
Hi Eric,
I've tried a few aarch64 systems, but was not able to reproduce this(with shipped fio).

@yizhan Any chance you have a reproducer?

Comment 6 Eric Sandeen 2022-10-06 20:26:17 UTC
*** Bug 1974306 has been marked as a duplicate of this bug. ***

Comment 7 Martin Hoyer 2022-10-07 08:07:13 UTC
I've just noticed my last comment is not here. Not sure what happened - apologies!
It went something like this:

The issue is reproducible on HPE Apollo 80 systems (Fujitsu A64FX) by running
# fio --crctest=crc32c

With test package from comment#2, it works as expected.

Comment 16 errata-xmlrpc 2023-05-16 08:28:12 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (fio bug fix and enhancement update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2023:2787