[MIR] zfs-linux

Bug #1532198 reported by Leann Ogasawara
36
This bug affects 3 people
Affects Status Importance Assigned to Milestone
zfs-linux (Ubuntu)
Fix Released
High
Unassigned

Bug Description

Following the process documented at https://wiki.ubuntu.com/MainInclusionProcess , the following template needs to be filled in to start the MIR for zfs-linux in 16.04

Below are my answers to the various main inclusion requirements, marked by a * prefix:

[Availability]:

  "The package must already be in the Ubuntu universe, and must
  build for the architectures it is designed to work on."

  * http://packages.ubuntu.com/xenial/admin/zfsutils-linux
  * Yes - built for 64 bit arches only, because ZFS is designed to run
    well only on 64 bit architectures.

[Rationale]:

  "There must be a certain level of demand for the package, for example:
  The package is useful for a large part of our user base."

  * Yes - there is a lot of interest in ZFS in the server space and for
    users wanting to use a file system that supports huge collections of
    disks with excellent reliable features such as checksummed raid, mirroring
    striping with easy configuration and also simple data sanity checking and
    fixing.
  * Being requested by Kiko

  "The package is a new build dependency or dependency of a package that we
  already support (additionally, the official image builder requires all
  used packages be in main)."

  * Yes, already in Wily as a technology demo.

  "The package helps meet a specific Blueprint goal."

  * No blueprint goal.

  "The package replaces another package we currently support and promises
  higher quality and/or better features, so that we can drop the old
  package from the supported set."

  * Not applicable

[Security]:
  "The security history and the current state of security issues in
  the package must allow us to support the package for at least 18 months
  without exposing its users to an inappropriate level of security risks.
  This requires checking of several things that are explained in detail in
  the subsection Security checks."

  "Check how many vulnerabilities the package had in the past and how they
  were handled by upstream and the Debian/Ubuntu package:"

  "http://cve.mitre.org/cve/cve.html: Search in the National Vulnerability
   Database using the package as a keyword"

  NO ZFS Linux CVEs found, here is the complete list from Mitre:

CVE-2015-1415
  The bsdinstall installer in FreeBSD 10.x before 10.1 p9, when configuring
  full disk encrypted ZFS, uses world-readable permissions for the GELI
  keyfile (/boot/encryption.key), which allows local users to obtain sensitive
  key information by reading the file.

CVE-2015-0448
  Unspecified vulnerability in Oracle Sun Solaris 11.2 allows local users to
  affect confidentiality, integrity, and availability via vectors related to
  ZFS File system.

CVE-2013-3266
  The nfsrvd_readdir function in sys/fs/nfsserver/nfs_nfsdport.c in the new
  NFS server in FreeBSD 8.0 through 9.1-RELEASE-p3 does not verify that a
  READDIR request is for a directory node, which allows remote attackers to
  cause a denial of service (memory corruption) or possibly execute arbitrary
  code by specifying a plain file instead of a directory.

CVE-2011-2313
  Unspecified vulnerability in Oracle Solaris 10 allows local users to affect
  availability, related to ZFS.

CVE-2011-2312
  Unspecified vulnerability in Oracle Solaris 10 allows local users to affect
  confidentiality, related to ZFS.

CVE-2011-2311
  Unspecified vulnerability in Oracle Solaris 10 allows local users to affect
  availability, related to ZFS.

CVE-2011-2286
  Unspecified vulnerability in Oracle Solaris 10 and 11 Express allows remote
  authenticated users to affect availability, related to ZFS.

CVE-2010-4458
  Unspecified vulnerability in Oracle Solaris 11 Express allows local users
  to affect availability, related to ZFS.

CVE-2010-3540
  Unspecified vulnerability in Oracle Solaris 10 and OpenSolaris allows local
  users to affect availability, related to ZFS.

CVE-2010-2392
  Unspecified vulnerability in Oracle Solaris 10 and OpenSolaris allows local
  users to affect integrity and availability, related to ZFS.

CVE-2010-0318
  The replay functionality for ZFS Intent Log (ZIL) in FreeBSD 7.1, 7.2,
  and 8.0, when creating files during replay of a setattr transaction, uses
  7777 permissions instead of the original permissions, which might allow
  local users to read or modify unauthorized files in opportunistic
  circumstances after a system crash or power failure.

CVE-2009-3706
  Unspecified vulnerability in the ZFS filesystem in Sun Solaris 10, and
  OpenSolaris snv_100 through snv_117, allows local users to bypass intended
  limitations of the file_chown_self privilege via certain uses of the chown
  system call.

  "http://secunia.com/advisories/search/: search for the package as a keyword"

  * No security advisories found

  Ubuntu CVE Tracker:

  http://people.ubuntu.com/~ubuntu-security/cve/main.html
  * No
  http://people.ubuntu.com/~ubuntu-security/cve/universe.html
  * No
  http://people.ubuntu.com/~ubuntu-security/cve/partner.html
  * No

  "Check for security relevant binaries. If any are present, this
  requires a more in-depth security review."

  "Executables which have the suid or sgid bit set."
  * Not applicable

  "Executables in /sbin, /usr/sbin."
  * Applicable. This requires security review

  "Packages which install daemons (/etc/init.d/*)"
  * Applicable. This requires security review

  "Packages which open privileged ports (ports < 1024)."
  * Not applicable

  "Add-ons and plugins to security-sensitive software (filters,
  scanners, UI skins, etc)"
  * Not applicable

[Quality assurance]
  "After installing the package it must be possible to make it working with a reasonable effort of configuration and documentation reading."

  * Will work "out-of-the-box" once zfsutils-linux installed with 4.4 kernel
  * Quick start ZFS reference guide written:
    https://wiki.ubuntu.com/Kernel/Reference/ZFS
  * Package contains main pages

  "The package must not ask debconf questions higher than medium if it is
  going to be installed by default. The debconf questions must have
  reasonable defaults."

  * Does not apply.

  "There are no long-term outstanding bugs which affect the usability of the program to a major degree. To support a package, we must be reasonably convinced that upstream supports and cares for the package."

  * We have good upstream support from ZFS maintainers, response to bugs
    file upstream is within 24 hours

  "The status of important bugs in Debian's, Ubuntu's, and upstream's bug
  tracking systems must be evaluated. Links to these bug trackers need to
  be provided in the MIR report. Important bugs must be pointed out and
  discussed in the MIR report."

  Upsteam bug tracking:
    ZFS - https://github.com/zfsonlinux/zfs/issues
    SPL - https://github.com/zfsonlinux/spl/issues
  Ubuntu bug tracking:
    https://bugs.launchpad.net/ubuntu/+source/zfs-linux

  Resolved bugs:
  LP#1521952 Add dependency on dh-systemd for zfs-linux
  LP#1513124 Fix FTBFSs on ppc64el and arm64

  "The package is maintained well in Debian/Ubuntu (check out the Debian PTS)"
    Maintained by Kernel team in sync with kernel

  Testing: We have several sets of ZFS specific regression tests in the
    kernel team autotest test infrastructure:

  * The ZFS test suite:
    http://kernel.ubuntu.com/git/ubuntu/autotest-client-tests.git/tree/ubuntu_zfs
  * fstest (Linux POSIX file system test suite)
    http://kernel.ubuntu.com/git/ubuntu/autotest-client-tests.git/tree/ubuntu_zfs_fstest
  * ZFS I/O stress tests:
    http://kernel.ubuntu.com/git/ubuntu/autotest-client-tests.git/tree/ubuntu_zfs_stress
  * XFS generic tests on ZFS:
    http://kernel.ubuntu.com/git/ubuntu/autotest-client-tests.git/tree/ubuntu_zfs_xfs_generic

  Note: currently working on a adt set of tests for ZFS to cover core features as
  as set of kernel team smoke tests.

[Dependencies]:
  All build and binary dependencies (including Recommends:) must be
  satisfyable in main (i. e. the preferred alternative must be in main).
  If not, these dependencies need a separate MIR report (this can be a
  separate bug or another task on the main MIR bug)

  zfs-linux:
  * autotools-dev - Yes
  * autoconf - Yes
  * autogen - Yes
  * automake - Yes
  * debhelper - Yes
  * dh-autoreconf - Yes
  * dh-systemd - Yes
  * dkms - Yes
  * libselinux1-dev - Yes
  * libtool - Yes
  * uuid-dev - Yes
  * zlib1g-dev - Yes

[Standards compliance]

  "Standards compliance: The package should meet the FHS and Debian Policy
  standards. Major violations should be documented and justified. Also, the
  source packaging should be reasonably easy to understand and maintain."

  Yes, I believe so.

[Maintenance]

  "The package must have an acceptable level of maintenance
  corresponding to its complexity:
  Simple packages (e.g. language bindings, simple Perl modules, small
  command-line programs, etc.) might not need very much maintenance effort,
  and if they are maintained well in Debian we can just keep them synced

  More complex packages will usually need a developer or team of
  developers paying attention to their bugs, whether that be in Ubuntu or
  elsewhere (often Debian). Packages that deliver major new headline
  features in Ubuntu need to have commitment from Ubuntu developers
  willing to spend substantial time on them."

  * Falls into the complex package category. Colin King will primarily
    maintain this package, with ownership owned and covered by the
    Canonical Kernel Team. We have already performed SRU on ZFS in
    Wily, showing we have the means to actively support this package.

  "All packages must have a designated "owning" team, regardless of
  complexity, which is set as a package bug contact."

  * Yes, Canononical Kernel Team
    https://launchpad.net/~canonical-kernel-team

[Background information]
  "The package descriptions should explain the general purpose and context
  of the package. Additional explanations/justifications should be done
  in the MIR report."

  * Yes, package description covers the scope of the package

  "If the package was renamed recently, or has a different upstream name,
  this needs to be explained in the MIR report."

  The ZFS on Linux provides ZFS packaged under the debian-zfs. Debian
  provides zfsutils for *BSD based kernels (kFreeBSD). The package name
  zfsutils-linux was chosen for Linux based arches.

CVE References

description: updated
description: updated
Michael Terry (mterry)
Changed in zfs-linux (Ubuntu):
assignee: nobody → Jamie Strandboge (jdstrand)
Tyler Hicks (tyhicks)
Changed in zfs-linux (Ubuntu):
assignee: Jamie Strandboge (jdstrand) → Ubuntu Security Team (ubuntu-security)
Changed in zfs-linux (Ubuntu):
importance: Undecided → High
Revision history for this message
Hajo Möller (dasjoe) wrote :

Just for completeness sake, a CVE was assigned to ZFS on Linux, CVE-2015-3400.
This is based on an issue that was exclusive to Debian.

See http://www.openwall.com/lists/oss-security/2015/04/22/4 for the assignment and https://github.com/zfsonlinux/zfs/issues/3319 for the details.

In short, ZoL's Debian then-maintainer (who is not Debian's current ZoL maintainer, if I am not mistaken) built the "official" ZoL Debian packages with additional patches.
Those patches were subject to above's CVE, the issue has since been fixed and affected Debian exclusively.

Revision history for this message
Colin Ian King (colin-king) wrote :

So the version in Ubuntu is derived from the 0.6.4.2 so does not contain the issue for CVE-2015-3400

Revision history for this message
Seth Arnold (seth-arnold) wrote :
Download full text (6.5 KiB)

I reviewed zfs-linux version 0.6.5.4-0ubuntu2 as checked into xenial.
This should not be considered a full security audit but rather a gauge of
maintainability.

- zfs is a storage system with extensive features
- Build-Depends: autotools-dev, autoconf, autogen, automake, debhelper,
  dh-autoreconf, dh-systemd, dkms, libselinux1-dev, libtool, uuid-dev,
  zlib1g-dev,
- Is the dkms build-depend still necessary?
- Many lintian errors, warnings:
  - package-name-doesnt-match-sonames
  - postinst-must-call-ldconfig
  - debian-changelog-line-too-long
  - init.d-script-not-marked-as-conffile
  - init.d-script-not-included-in-package
  - systemd-no-service-for-init-script
- Does not (yet) do encryption
- Does not itself do networking, though zfs send | zfs receive often
  operate over e.g. ssh
- Provides kernel modules, zed event daemon, utilities, test tools
- zed event daemon spawns 'zedlets' on specific events; care must be taken
  when writing eventlets to ensure they properly terminate.
- zed has nice daemonization routines to send error messages to the parent
  process
- pre/post inst/rm scripts are automatically generated
- init scripts / systemd units discover devices in storage pools, import
  pools, mount filesystems, create zvol block devices, start zed, export
  any zfs shared filesystems via nfs or cifs, etc.
- No dbus services
- No setuid executables
- programs in /sbin/:
  - zpios
  - zhack
  - zfs
  - zinject
  - fsck.zfs
  - ztest
  - mount.zfs
  - zdb
  - zstreamdump
  - zpool
  /usr/sbin/:
  - zed
- The /etc/sudoers.d/zfs file is entirely commented out; it suggests some
  rules that can be used to allow passwordless access to the read-only zfs
  and zpool subcommands. However, this file also includes some rules to
  allow members of adm, admin, or staff groups access to full root
  privileges. This is not what the comment at the top of the file suggests
  and I see no reason for this file to contain these rules, even if
  commented out.
- Extensive udev rules to create zvol block devices and /dev/disk/by-vdev/
  for more-complicated vdev configurations
- No tests run during build
- Good selection of kernel smoke tests provided in debian/tests/
- There are no provided cron jobs but system administrators would be wise
  to schedule pool scrubs frequently; typically weekly or monthly is
  recommended based on drive quality, level of redundancy in the pool,
  data read rate and working set size.
- More than the usual number of warnings from libtool

- Some process spawning: some via zed, some via the kernel, some via
  utilities. They looked careful.
- Memory management looked careful
- Not much file IO in the usual sense
- Logging looked careful, very extensive error reporting
- A lot of environment variable use, some could have unexpected
  consequences
- Extensive privileged commands; far too many to carefully audit
- No cryptography (yet)
- Does not itself do networking
- There are extensive privileged portions of code, including kernel
  modules; some privilege checks that make sense in OpenSolaris are
  stubbed out and ignored in Linux, we should investigate further.
- Minimal use of temp files, handled carefully
- Does not use W...

Read more...

Changed in zfs-linux (Ubuntu):
assignee: Ubuntu Security Team (ubuntu-security) → nobody
Revision history for this message
Michael Terry (mterry) wrote :

OK, so I gave it a look over for packaging and maintainability checklists.

- Needs a team bug subscriber in Ubuntu

- zfs-zed Replaces zfsutils-linux without a version constraint. Is that correct? I assumed a file moved from one package to another, in which case there should be a version.

- I like the dep8 tests

So I'd say add a bug subscriber, fix the Replaces line, address the init script lintian errors, and fix the sudoers file.

Changed in zfs-linux (Ubuntu):
status: New → Incomplete
Revision history for this message
Matthias Klose (doko) wrote :

zfs-linux depends on spl-linux, MIR needed

Changed in spl-linux (Ubuntu):
status: New → Incomplete
Michael Terry (mterry)
Changed in spl-linux (Ubuntu):
assignee: nobody → Jamie Strandboge (jdstrand)
Changed in spl-linux (Ubuntu):
assignee: Jamie Strandboge (jdstrand) → Ubuntu Security Team (ubuntu-security)
Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

fwiw - libvirt is stuck in -proposed blocked on this.

Revision history for this message
Seth Arnold (seth-arnold) wrote :

Can someone describe what exactly from spl-linux is needed in main?

- spl provides a splat test-driver, users shouldn't need this, and we're not interested in supporting arbitrary testing platforms

- spl-dkms provides an infrastructure to build spl.ko on end-user systems, but this module should be provided by linux-image-* packages.

- spl-dkms provides an infrastructure to build splat.ko on end-user systems. This module should not be used except for testing. (splat.ko is also provided by linux-image-* but this feels like a mistake.)

I have a feeling this is an atavism.

Thanks

Revision history for this message
Seth Arnold (seth-arnold) wrote :

splat.ko in linux-image-* packages is bug #1566074.

Revision history for this message
Richard Laager (rlaager) wrote :

The spl package creates /etc/hostid in order to guarantee a stable hostid. Retaining that may be important, because if the hostid changes, pools will not import without being forced (which is not something that should be encouraged). Moving the contents of spl.postinst into zfsutils-linux.postinst may be the right approach.

Other than that, you can probably just drop the zfs-dkms package from the zfs-linux source and then spl-linux should be unnecessary as you note.

Revision history for this message
Colin Ian King (colin-king) wrote :

"Please remove the %adm %admin %staff section of /etc/sudoers.d/zfs."

 - fixed for 0.6.5.6-0ubuntu6

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

This bug was fixed in the package zfs-linux - 0.6.5.6-0ubuntu6

---------------
zfs-linux (0.6.5.6-0ubuntu6) xenial; urgency=medium

  * Remove adm, admin, staff groups from /etc/sudoers.d/zfs
    (LP: #1532198)

 -- Colin Ian King <email address hidden> Fri, 8 Apr 2016 15:16:39 +0100

Changed in zfs-linux (Ubuntu):
status: Incomplete → Fix Released
Revision history for this message
Colin Ian King (colin-king) wrote :

Following up comments on #3

 the libraries install directly to /lib, is this intentional?
- efi_use_whole_disk() leaks &efi_label via VT_ENOSPC error return

0.6.5.6: no leak:

        if ((efi_label->efi_altern_lba == 1) || (efi_label->efi_altern_lba
            >= efi_label->efi_last_lba)) {
                if (efi_debug) {
                        (void) fprintf(stderr,
                            "efi_use_whole_disk: requested space not found\n");
                }
                efi_free(efi_label);
                return (VT_ENOSPC);
        }

- vdev_elevator_switch() broken strncmp() test
   - bug filed upstream #4507, from code clean up which caused a regression

- _finish_daemonize() if devnull is 0, 1, 2, the close(devnull) call will
  leave the program without that fd
  * upstream fix 048bb5bd4950b9cb5368ed93d273f0f36e439122
    ("Ensure zed _finish_daemonize() leaves fds 0-2 open")
    - will SRU this

- zfs_fuid_node_add() modifies zfs_fuid_info_t list without obvious
  locking -- are calls to this routine suitably locked elsewhere?
   * checked this out, agreed, can't see locking,
    -- reported upstream, #4508

- ddt_zap_update() allocates cbuf[sizeof(dde_phys)] on the stack -- which
  is an array DDT_PHYS_TYPES long of structs which themselves have an
  embedded array of structs with further embedded array of integers. How
  likely is it for this to blow out the stack depth?

dde_phys is an array of DDT_PHYS_TYPES ddt_phys_t types, and ddt_phys_t is:

typedef struct ddt_phys {
        dva_t ddp_dva[SPA_DVAS_PER_BP];
        uint64_t ddp_refcnt;
        uint64_t ddp_phys_birth;
} ddt_phys_t;

and dva_t is 2 x uint64_t, SPA_DVAS_PER_BP is 3, so ddt_phys_t is

(2 x 8 x 3) + 8 + 8, making 64 bytes per ddt_phys_t. DDT_PHYS_TYPES is 4, making the allocation (64 * 4) + 1 = 257 bytes, so no problem at all, it's an acceptable size.

- zpool_vdev_remove() includes misleading error message that top-level
  devices can be removed
   * minor error message,
    - bug filed upstream #4506

- zpool_vdev_remove() awkward error message "pool must be upgrade to
  support log removal"
   - fixed in 0.6.5.6

Revision history for this message
Richard Laager (rlaager) wrote :

We are getting very close to the final freeze. What's the plan to deal with spl? Is it going in, or will the /etc/hostid code be moved to ZFS, or something else?

Revision history for this message
Colin Ian King (colin-king) wrote :

With reference to comment #5, a MIR for SPL has been created, https://bugs.launchpad.net/ubuntu/+source/spl-linux/+bug/1569294

Revision history for this message
Seth Arnold (seth-arnold) wrote :

spl-linux acked by security team for promotion to main in 1569294.

Changed in spl-linux (Ubuntu):
assignee: Ubuntu Security Team (ubuntu-security) → nobody
Revision history for this message
Colin Ian King (colin-king) wrote :

Following up comments in #3

"- zfs_fuid_node_add() modifies zfs_fuid_info_t list without obvious
  locking -- are calls to this routine suitably locked elsewhere?"

  cf : https://github.com/zfsonlinux/zfs/issues/4508#issuecomment-209130720
  "This functionality, along with everything else wrapped with HAVE_KSID, is dormant under Linux. Additional work is still required needed to map this illumos functionality to its Linux counterpart."

Revision history for this message
Stéphane Graber (stgraber) wrote :

 - Removed the spl task as it's been split out to its own bug.
 - Moved status back to Incomplete.

Looks like we have a security team ack for this, but are missing final ack from the MIR team (assuming everything else looks good).

Changed in zfs-linux (Ubuntu):
status: Fix Released → Incomplete
no longer affects: spl-linux (Ubuntu)
Revision history for this message
Michael Terry (mterry) wrote :

stgraber, I wrote in comment #4 what was needed from a normal MIR perspective -- team bug subscriber and to fix the zfs-zed Replaces line. Other than that, it seems fine and with the security ACK, we'd be good to go. (modulo spl-linux of course)

Revision history for this message
Dustin Kirkland  (kirkland) wrote :

Bump?

Revision history for this message
Michael Terry (mterry) wrote :

I see that there's a team bug subscriber now. So the only issue on my original list is fixing the zfs-zed Replaces line for zfsutils-linux. It's such a tiny thing to hold up a MIR, but I guess since xenial deadline is passed and it's so trivial to fix, let's just fix it.

Revision history for this message
Richard Laager (rlaager) wrote :

That Replaces line dates back from the ZoL PPA. I could be missing something, but I don't see why it was even necessary originally. Here's the commit where it was added: https://github.com/zfsonlinux/pkg-zfs/commit/9da00045fb5be4c61bcd29daf0a59daa32c9a43d

Revision history for this message
Michael Terry (mterry) wrote :

Yeah... Looks like it could be dropped. Even if it did refer to a file that moved at one point, we're past the LTS now, so file migration during upgrades aren't even a concern.

Revision history for this message
Darik Horn (dajhorn) wrote :

@rlaager, zed first shipped in the zfsutils package, and the `Replaces` line was required for a smooth upgrade between releases. It can/should be dropped now.

Revision history for this message
Richard Laager (rlaager) wrote :

It turns out that this is still necessary. zfsutils-linux installs /usr/share/man/man8/zed.8.gz, which is also installed by zfs-zed. That needs to be fixed at the same time the Replaces is versioned. If I'm understanding correctly, Policy says a versioned Breaks should be added too.

I've made these changes (along with other bug fixes) here:
https://launchpad.net/~rlaager/+archive/ubuntu/zfs

If you do test from my PPA, please remove it from your APT sources when you're done testing. I don't want some future experiment I upload to break your system.

Revision history for this message
Colin Ian King (colin-king) wrote :

The issues in comment #4 have been addressed, please refer to bug 1579082 - @Michael Terry, let me know if that is satisfactory. Thanks!

Revision history for this message
Colin Ian King (colin-king) wrote :

Thanks Richard, I'll put your change into the package ASAP.

Revision history for this message
Colin Ian King (colin-king) wrote :

@Michael Terry, I've prepared the MIR fixes and uploaded it into:

https://launchpad.net/~colin-king/+archive/ubuntu/zfs-0.6.5.6-sru/+packages

..if that looks OK I'll get them uploaded into -prosed.

Revision history for this message
Michael Terry (mterry) wrote :

That debdiff doesn't look like it stops the double-installation of the zed man page from comment #24

Revision history for this message
Richard Laager (rlaager) wrote :

Indeed, you need my changes to debian/zfsutils-linux.install as well.

Changed in zfs-linux (Ubuntu):
assignee: nobody → Colin Ian King (colin-king)
Revision history for this message
Colin Ian King (colin-king) wrote :

Bah, I messed up (brown paper bag).

@Michael, do you mind checking again, I've uploaded it to:

https://launchpad.net/~colin-king/+archive/ubuntu/zfs-0.6.5.6-sru-v2/+packages

Revision history for this message
Michael Terry (mterry) wrote :

That looks good! With that on its way to yakkety, I'll mark this approved. Thank you!

Changed in zfs-linux (Ubuntu):
assignee: Colin Ian King (colin-king) → nobody
status: Incomplete → Fix Committed
Revision history for this message
Steve Langasek (vorlon) wrote :
Download full text (8.4 KiB)

Override component to main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety amd64: universe/libs/extra/100% -> main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety arm64: universe/libs/extra/100% -> main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety armhf: universe/libs/extra/100% -> main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety i386: universe/libs/extra/100% -> main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety powerpc: universe/libs/extra/100% -> main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety ppc64el: universe/libs/extra/100% -> main
libnvpair1linux 0.6.5.7-0ubuntu3 in yakkety s390x: universe/libs/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety amd64: universe/debug/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety arm64: universe/debug/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety armhf: universe/debug/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety i386: universe/debug/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety powerpc: universe/debug/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety ppc64el: universe/debug/extra/100% -> main
libnvpair1linux-dbg 0.6.5.7-0ubuntu3 in yakkety s390x: universe/debug/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety amd64: universe/libs/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety arm64: universe/libs/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety armhf: universe/libs/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety i386: universe/libs/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety powerpc: universe/libs/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety ppc64el: universe/libs/extra/100% -> main
libuutil1linux 0.6.5.7-0ubuntu3 in yakkety s390x: universe/libs/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety amd64: universe/debug/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety arm64: universe/debug/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety armhf: universe/debug/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety i386: universe/debug/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety powerpc: universe/debug/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety ppc64el: universe/debug/extra/100% -> main
libuutil1linux-dbg 0.6.5.7-0ubuntu3 in yakkety s390x: universe/debug/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety amd64: universe/libs/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety arm64: universe/libs/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety armhf: universe/libs/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety i386: universe/libs/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety powerpc: universe/libs/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety ppc64el: universe/libs/extra/100% -> main
libzfs2linux 0.6.5.7-0ubuntu3 in yakkety s390x: universe/libs/extra/100% -> main
libzfs2linux-dbg 0.6.5.7-0ubuntu3 in yakkety amd64: universe/debug/extra/100% -> main
libzfs2linux-dbg 0.6.5.7-0ubuntu3 in yakkety arm64: universe...

Read more...

Changed in zfs-linux (Ubuntu):
status: Fix Committed → Fix Released
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.