fio crashes when using --offset with a device

Bug #1891169 reported by Mark Lehrer
12
This bug affects 2 people
Affects Status Importance Assigned to Milestone
fio (Ubuntu)
Confirmed
Undecided
Unassigned

Bug Description

e.g. here is how I was running it:

fio --loops=1 --direct=1 --rw=write --ioengine=libaio --bs=4K --iodepth=16 --numjobs=1 --time_based --runtime=10 --group_reporting --size=10% --offset=10% --name=TEST --filename=/dev/nvme0n1

I built the 3.17 source and the problem didn't happen. If you can't bump the universe version to 3.21 or 3.17 and need a patch for 3.16, I can make one for you if necessary.

ProblemType: Bug
DistroRelease: Ubuntu 20.04
Package: fio 3.16-1
ProcVersionSignature: Ubuntu 5.4.0-37.41-generic 5.4.41
Uname: Linux 5.4.0-37-generic x86_64
ApportVersion: 2.20.11-0ubuntu27.6
Architecture: amd64
CasperMD5CheckResult: pass
Date: Tue Aug 11 13:14:24 2020
InstallationDate: Installed on 2020-05-19 (84 days ago)
InstallationMedia: Ubuntu-Server 20.04 LTS "Focal Fossa" - Release amd64 (20200423)
ProcEnviron:
 TERM=xterm-256color
 PATH=(custom, no user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: fio
UpgradeStatus: No upgrade log present (probably fresh install)

Revision history for this message
Mark Lehrer (pyite) wrote :
Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in fio (Ubuntu):
status: New → Confirmed
Revision history for this message
Joel Colledge (jcolledge) wrote :

I have also observed strange behavior with the fio 3.16 package on Focal. It occurs with the combination of "size" and "offset", and is affected by the presence of "time_based".

For example with this command:

fio --name=test --filename=/dev/loop9 --ioengine=libaio --rw=write --direct=1 --offset=8M --size=8M --time_based --runtime=2

I observe that the device is written from 8M to the end. I have tested with a range of other fio versions and they all write from 8M to 16M as expected.

It looks like the issue was introduced with https://github.com/axboe/fio/commit/ad46c5715ffc5d4e70d2e941225b1f504f23e409
This commit was reverted in https://github.com/axboe/fio/commit/0d0601479d593ca9e96e35f43e2c0586f8358f34
The only fio version containing the original commit but not the revert is 3.16.

I have not experienced this exact failure but it seems very likely to have the same root cause.

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.