Thanks, Julian. I have added some more information about the autopkgtest
changes to the bug report.
Verification. test.sh is the test plan in the bug report.
root@sg3-utils:~# apt policy sg3-utils
sg3-utils:
Installed: 1.48-3ubuntu3
Candidate: 1.48-3ubuntu3
Version table:
1.48-3ubuntu3.1 100
100 http://archive.ubuntu.com/ubuntu resolute-proposed/main amd64
Packages
*** 1.48-3ubuntu3 500
500 http://archive.ubuntu.com/ubuntu resolute/main amd64 Packages
100 /var/lib/dpkg/status
root@sg3-utils:~# bash test.sh
bad argument to '--contents='
bad argument to '--contents='
bad argument to '--contents='
bad argument to '--cfile='
root@sg3-utils:~# apt policy sg3-utils
sg3-utils:
Installed: 1.48-3ubuntu3.1
Candidate: 1.48-3ubuntu3.1
Version table:
*** 1.48-3ubuntu3.1 100
100 http://archive.ubuntu.com/ubuntu resolute-proposed/main amd64
Packages
100 /var/lib/dpkg/status
1.48-3ubuntu3 500
500 http://archive.ubuntu.com/ubuntu resolute/main amd64 Packages
root@sg3-utils:~# bash test.sh
root@sg3-utils:~#
** Description changed:
[ Impact ]
* sg_wr_mode rejects all arguments (even if valid) to --contents= or
--cfile=.
* This impacts every usage of sg_wr_mode with --contents= or --cfile.
* This used to work in Noble, and is a regression in Resolute.
* The upload fixes the bug by including a patch from upstream that corrects
the argument handling logic.
[ Test Plan ]
* Running the following script should output nothing.
sg_wr_mode --page=8 --contents=0,1 /dev/null 2>&1 | grep 'bad argument'
sg_wr_mode --page=8 --contents='08 12 01' /dev/null 2>&1 | grep 'bad argument'
sg_wr_mode -p 8 -c 08,12,01 /dev/null 2>&1 | grep 'bad argument'
printf '\x08\x12\x01' > /tmp/page.bin
- sg_wr_mode --page=8 --cfile=/tmp/page.bin /dev/null 2>&1 | grep 'bad argument'
+ sg_wr_mode --raw --page=8 --cfile=/tmp/page.bin /dev/null 2>&1 | grep 'bad
argument'
* Current output in Resolute:
bad argument to '--contents='
bad argument to '--contents='
bad argument to '--contents='
bad argument to '--cfile='
[ Where problems could occur ]
* Changes are isolated to argument parsing in sg_wr_mode.
* Any regression would show up as other valid arguments for sg_wr_mode being
misparsed or rejected.
[ Other Info ]
* This issue was fixed upstream in svn r1048, but there's no upstream release
yet.
- * The upload also fixes autopkgtests, which were failing for unrelated
+ * The upload also fixes autopkgtests, which were failing for unrelated
reasons. Tests were previously trying to install linux-modules-
extra-$(uname-r), which does not exist anymore. It also needed some
adjustment handle deleting IPv6 portals (previously only deleted IPv4).
[ Original Bug Description ]
Package: sg3-utils 1.48-3ubuntu3
Ubuntu: 26.04 (Resolute Raccoon)
Binary: sg_wr_mode version "1.30 20230623"
sg_wr_mode rejects every --contents= and --cfile= argument before
opening the target device, regardless of value. The rejection fires from
argv parsing — it's not device-specific.
Repro (any of these):
$ sg_wr_mode --page=8 --contents=0,1 /dev/null
bad argument to '--contents='
$ sg_wr_mode --page=8 --contents="08 12 01" /dev/null
bad argument to '--contents='
$ sg_wr_mode -p 8 -c 08,12,01 /dev/null
bad argument to '--contents='
$ printf '\x08\x12\x01' > /tmp/page.bin
$ sg_wr_mode --page=8 --cfile=/tmp/page.bin /dev/null
bad argument to '--cfile='
Exit code: 32 in every case.
strace confirms the failure is pre-I/O: libsgutils2-1.48.so.2 loads,
then the error is written to stderr and the process exits — the device
is never opened. So the bug is in sg_wr_mode's argv handler, not the
underlying SCSI layer.
Confirmed not device-related: same failure against /dev/sg1 (a QEMU CD-
ROM SCSI generic device). sg_inq and sg_modes against the same device
work normally — only sg_wr_mode's argument parser rejects --contents=.
The sg3-utils library and the device are both healthy; the bug is
isolated to sg_wr_mode.
Workaround: issue MODE SELECT(10) directly via sg_raw, e.g.: sg_raw -v
-s 28 -i payload.bin /dev/sgN 55 10 00 00 00 00 00 00 1C 00
Expected: --contents= should accept the documented "comma-separated
string of hex numbers (00..ff)" syntax (per the man page and --help
text).
Actual: every value is rejected.
ProblemType: Bug
DistroRelease: Ubuntu 26.04
Package: sg3-utils 1.48-3ubuntu3
ProcVersionSignature: Ubuntu 7.0.0-15.15-generic 7.0.0
Uname: Linux 7.0.0-15-generic x86_64
ApportVersion: 2.34.0-0ubuntu2
Architecture: amd64
CasperMD5CheckResult: pass
Date: Sat May 9 13:59:51 2026
InstallationDate: Installed on 2026-05-02 (7 days ago)
InstallationMedia: Ubuntu-Server 26.04 "Resolute Raccoon" - Release amd64
(20260420.1)
ProcEnviron:
LANG=en_US.UTF-8
PATH=(custom, no user)
SHELL=/bin/bash
TERM=xterm-256color
XDG_RUNTIME_DIR=<set>
SourcePackage: sg3-utils
UpgradeStatus: No upgrade log present (probably fresh install)
** Tags removed: verification-needed-resolute
** Tags added: verification-done-resolute
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2152092
Title:
sg_wr_mode rejects every --contents= and --cfile= argument with "bad
argument"
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/sg3-utils/+bug/2152092/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
