ostree.git
7 years agoMerge version 2018.6-3+rpi1 and 2018.7-2 to produce 2018.7-2+rpi1 archive/raspbian/2018.7-2+rpi1 raspbian/2018.7-2+rpi1
Raspbian forward pporter [Thu, 9 Aug 2018 15:34:14 +0000 (16:34 +0100)]
Merge version 2018.6-3+rpi1 and 2018.7-2 to produce 2018.7-2+rpi1

7 years agoMerge ostree (2018.7-2) import into refs/heads/workingbranch
Simon McVittie [Mon, 30 Jul 2018 15:52:35 +0000 (16:52 +0100)]
Merge ostree (2018.7-2) import into refs/heads/workingbranch

7 years agoSkip test-pull-repeated during CI
Simon McVittie [Mon, 30 Jul 2018 15:51:01 +0000 (16:51 +0100)]
Skip test-pull-repeated during CI

This test is expected to fail a small proportion of the time. During
the build of ostree 2018.7-1 in Debian, it seems we were unlucky on
s390x. Non-deterministic tests are also problematic for autopkgtest,
where they can gate migration of our dependencies like GLib, so skip
this test unless the caller has opted-in to non-deterministic tests.

It would be appropriate to enable this test in environments where
failures can easily be retried and are not disruptive to other
packages.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Topic debian
Gbp-Pq: Name Skip-test-pull-repeated-during-CI.patch

7 years agolib/pull: Fix minor memleak in error path
Colin Walters [Thu, 19 Jul 2018 12:46:23 +0000 (12:46 +0000)]
lib/pull: Fix minor memleak in error path

Spotted by a downstream Coverity build.

Bug: #1684
Reviewed-by: jlebon
Origin: upstream, 2018.8, commit:93da56842259f18bde23daf17dcefcf213b30b81

Gbp-Pq: Name lib-pull-Fix-minor-memleak-in-error-path.patch

7 years agoostree (2018.7-2) unstable; urgency=medium
Simon McVittie [Mon, 30 Jul 2018 15:52:35 +0000 (16:52 +0100)]
ostree (2018.7-2) unstable; urgency=medium

  * d/p/debian/Skip-test-pull-repeated-during-CI.patch:
    Skip a test that is non-deterministic and can spuriously fail,
    which is not suitable for build-time testing or gating migration in
    autopkgtest.

[dgit import unpatched ostree 2018.7-2]

7 years agoImport ostree_2018.7-2.debian.tar.xz
Simon McVittie [Mon, 30 Jul 2018 15:52:35 +0000 (16:52 +0100)]
Import ostree_2018.7-2.debian.tar.xz

[dgit import tarball ostree 2018.7-2 ostree_2018.7-2.debian.tar.xz]

7 years agoImport ostree_2018.7.orig.tar.xz
Simon McVittie [Wed, 25 Jul 2018 22:46:29 +0000 (23:46 +0100)]
Import ostree_2018.7.orig.tar.xz

[dgit import orig ostree_2018.7.orig.tar.xz]

7 years agoMerge version 2018.5-2+rpi1 and 2018.6-3 to produce 2018.6-3+rpi1 archive/raspbian/2018.6-3+rpi1 raspbian/2018.6-3+rpi1
Raspbian forward pporter [Wed, 27 Jun 2018 09:26:40 +0000 (10:26 +0100)]
Merge version 2018.5-2+rpi1 and 2018.6-3 to produce 2018.6-3+rpi1

7 years agoMerge ostree (2018.6-3) import into refs/heads/workingbranch
Simon McVittie [Sun, 24 Jun 2018 12:04:09 +0000 (13:04 +0100)]
Merge ostree (2018.6-3) import into refs/heads/workingbranch

7 years agoOstreeRepoFinderConfig: Fix guint/gsize confusion
Simon McVittie [Sun, 24 Jun 2018 11:56:49 +0000 (12:56 +0100)]
OstreeRepoFinderConfig: Fix guint/gsize confusion

If a function has a guint "out argument", passing a pointer to a gsize
is not, in general, valid. On an ILP64 platform there is no problem
since guint and gsize are identical, but on an LP64 platform it will
overwrite only the first word of the gsize, leaving the second word
unaffected. On little-endian machines, if the second word is
zero-initialized (as it is here), the result is numerically equal to
the guint, but on big-endian machines the result is around 4 billion
times what it should be, resulting in
ostree_repo_finder_config_resolve_async() reading past the end of
the array and causing undefined behaviour.

In practice this caused assertion failures (and consequently test
failures) on Debian's s390x (z/Architecture), ppc64 (64-bit PowerPC)
and sparc64 (64-bit SPARC) ports.

Bug: https://github.com/ostreedev/ostree/issues/1640
Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=902209
Forwarded: https://github.com/ostreedev/ostree/pull/1641
Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Name OstreeRepoFinderConfig-Fix-guint-gsize-confusion.patch

7 years agoavahi: Fail immediately if we can't talk to D-Bus or Avahi
Simon McVittie [Fri, 22 Jun 2018 00:10:55 +0000 (01:10 +0100)]
avahi: Fail immediately if we can't talk to D-Bus or Avahi

We special-case AVAHI_ERR_NO_DAEMON to not cause warnings, but if
we pass AVAHI_CLIENT_NO_FAIL to avahi_client_new, we never actually
see AVAHI_ERR_NO_DAEMON. Instead, we will get AVAHI_ERR_BAD_STATE
when we try to use the client.

Bug: https://github.com/ostreedev/ostree/issues/1618
Signed-off-by: Simon McVittie <smcv@debian.org>
Forwarded: https://github.com/ostreedev/ostree/pull/1639
Applied-upstream: 2018.7, commit:https://github.com/ostreedev/ostree/commit/e120a6b1198aaa785533c76316898f15a804dae1

Gbp-Pq: Name avahi-Fail-immediately-if-we-can-t-talk-to-D-Bus-or-Avahi.patch

7 years agolib/repo: Fix 32 bit format string error
Colin Walters [Thu, 21 Jun 2018 14:17:28 +0000 (14:17 +0000)]
lib/repo: Fix 32 bit format string error

Origin: upstream, 2018.7, commit:1174d9f5ba537562c67084caf0214544fbb14ffc

Gbp-Pq: Name lib-repo-Fix-32-bit-format-string-error.patch

7 years agoostree (2018.6-3) unstable; urgency=medium
Simon McVittie [Sun, 24 Jun 2018 12:04:09 +0000 (13:04 +0100)]
ostree (2018.6-3) unstable; urgency=medium

  * d/p/avahi-Fail-immediately-if-we-can-t-talk-to-D-Bus-or-Avahi.patch:
    Mark as applied upstream
  * d/p/OstreeRepoFinderConfig-Fix-guint-gsize-confusion.patch:
    Add patch to fix incorrect type aliasing that caused assertion
    failures on 64-bit big-endian platforms (Closes: #902209)

[dgit import unpatched ostree 2018.6-3]

7 years agoImport ostree_2018.6-3.debian.tar.xz
Simon McVittie [Sun, 24 Jun 2018 12:04:09 +0000 (13:04 +0100)]
Import ostree_2018.6-3.debian.tar.xz

[dgit import tarball ostree 2018.6-3 ostree_2018.6-3.debian.tar.xz]

7 years agoImport ostree_2018.6.orig.tar.xz
Simon McVittie [Fri, 22 Jun 2018 20:05:29 +0000 (21:05 +0100)]
Import ostree_2018.6.orig.tar.xz

[dgit import orig ostree_2018.6.orig.tar.xz]

7 years agoMerge version 2018.5-1+rpi1 and 2018.5-2 to produce 2018.5-2+rpi1 archive/raspbian/2018.5-2+rpi1 raspbian/2018.5-2+rpi1
Raspbian forward pporter [Wed, 20 Jun 2018 20:43:35 +0000 (21:43 +0100)]
Merge version 2018.5-1+rpi1 and 2018.5-2 to produce 2018.5-2+rpi1

7 years agoMerge ostree (2018.5-2) import into refs/heads/workingbranch
Simon McVittie [Sun, 10 Jun 2018 12:57:19 +0000 (13:57 +0100)]
Merge ostree (2018.5-2) import into refs/heads/workingbranch

7 years agoostree (2018.5-2) unstable; urgency=medium
Simon McVittie [Sun, 10 Jun 2018 12:57:19 +0000 (13:57 +0100)]
ostree (2018.5-2) unstable; urgency=medium

  * d/tests/gnome-desktop-testing: Skip libostree/test-concurrency.py.test
    during autopkgtest. It does not appear to be completely reliable.
    (See #901170)

[dgit import unpatched ostree 2018.5-2]

7 years agoImport ostree_2018.5-2.debian.tar.xz
Simon McVittie [Sun, 10 Jun 2018 12:57:19 +0000 (13:57 +0100)]
Import ostree_2018.5-2.debian.tar.xz

[dgit import tarball ostree 2018.5-2 ostree_2018.5-2.debian.tar.xz]

7 years agoMerge version 2018.4-2+rpi1 and 2018.5-1 to produce 2018.5-1+rpi1 archive/raspbian/2018.5-1+rpi1 raspbian/2018.5-1+rpi1
Raspbian forward pporter [Sun, 20 May 2018 21:16:19 +0000 (22:16 +0100)]
Merge version 2018.4-2+rpi1 and 2018.5-1 to produce 2018.5-1+rpi1

7 years agoMerge ostree (2018.5-1) import into refs/heads/workingbranch
Simon McVittie [Tue, 15 May 2018 10:30:03 +0000 (11:30 +0100)]
Merge ostree (2018.5-1) import into refs/heads/workingbranch

7 years agoImport ostree_2018.5.orig.tar.xz
Simon McVittie [Tue, 15 May 2018 10:30:03 +0000 (11:30 +0100)]
Import ostree_2018.5.orig.tar.xz

[dgit import orig ostree_2018.5.orig.tar.xz]

7 years agoostree (2018.5-1) unstable; urgency=medium
Simon McVittie [Tue, 15 May 2018 10:30:03 +0000 (11:30 +0100)]
ostree (2018.5-1) unstable; urgency=medium

  * New upstream release
    - d/copyright: Upstream clarified that only doc/ is CC-BY-SA-3.0 or
      GFDL-1.3-or-later, and doc/ isn't included in dist tarballs, so
      remove those licenses
    - d/libostree-1-1.symbols: Update (and sort)
    - d/p/Don-t-write-to-parent-repo.patch: Drop, applied upstream
    - d/ostree-boot.install: Add new systemd service
      lib/systemd/system/ostree-finalize-staged.service
    - d/rules: Remove lib/systemd/system/ostree-finalize-staged.service
      until we build ostree-boot
  * ostree-tests: Add Lintian override for library-not-linked-against-libc.
    libreaddir-rand.so genuinely doesn't use any libc ABIs directly, only
    via GLib.
  * Standards-Version: 4.1.4 (no changes required)

[dgit import unpatched ostree 2018.5-1]

7 years agoImport ostree_2018.5-1.debian.tar.xz
Simon McVittie [Tue, 15 May 2018 10:30:03 +0000 (11:30 +0100)]
Import ostree_2018.5-1.debian.tar.xz

[dgit import tarball ostree 2018.5-1 ostree_2018.5-1.debian.tar.xz]

8 years agoMerge version 2018.4-1+rpi1 and 2018.4-2 to produce 2018.4-2+rpi1 archive/raspbian/2018.4-2+rpi1 raspbian/2018.4-2+rpi1
Raspbian forward pporter [Thu, 26 Apr 2018 06:23:03 +0000 (07:23 +0100)]
Merge version 2018.4-1+rpi1 and 2018.4-2 to produce 2018.4-2+rpi1

8 years agoMerge ostree (2018.4-2) import into refs/heads/workingbranch
Simon McVittie [Tue, 17 Apr 2018 08:06:42 +0000 (09:06 +0100)]
Merge ostree (2018.4-2) import into refs/heads/workingbranch

8 years agoDon't write to parent repo
Alexander Larsson [Thu, 29 Mar 2018 13:19:33 +0000 (15:19 +0200)]
Don't write to parent repo

In _try_clone_from_payload_link, don't try to do the clone in the
parent repo, because we don't want to modify that. parent repos are
typically used when you want a shared, immutable base.

For example in flatpak, the parent repo is the system repo which you
don't have write access to, so any modification to it will fail with
EACCES, making it impossible to install via the system helper.

Closes: #1524
Approved by: cgwalters
Applied-upstream: 2018.5, commit:28c7bc6d0e153a0b07bdb82d25473a490765067f

Gbp-Pq: Name Don-t-write-to-parent-repo.patch

8 years agoostree (2018.4-2) unstable; urgency=medium
Simon McVittie [Tue, 17 Apr 2018 08:06:42 +0000 (09:06 +0100)]
ostree (2018.4-2) unstable; urgency=medium

  * d/p/Don-t-write-to-parent-repo.patch:
    Add patch from upstream to prevent trying to write to a parent
    repository, fixing installation of Flatpak apps and runtimes into
    the system-wide repository (Closes: #895883)

[dgit import unpatched ostree 2018.4-2]

8 years agoImport ostree_2018.4-2.debian.tar.xz
Simon McVittie [Tue, 17 Apr 2018 08:06:42 +0000 (09:06 +0100)]
Import ostree_2018.4-2.debian.tar.xz

[dgit import tarball ostree 2018.4-2 ostree_2018.4-2.debian.tar.xz]

8 years agoMerge version 2018.2-1+rpi1 and 2018.4-1 to produce 2018.4-1+rpi1 archive/raspbian/2018.4-1+rpi1 raspbian/2018.4-1+rpi1
Raspbian forward pporter [Fri, 30 Mar 2018 13:25:40 +0000 (14:25 +0100)]
Merge version 2018.2-1+rpi1 and 2018.4-1 to produce 2018.4-1+rpi1

8 years agoMerge ostree (2018.4-1) import into refs/heads/workingbranch
Simon McVittie [Sat, 24 Mar 2018 19:20:08 +0000 (19:20 +0000)]
Merge ostree (2018.4-1) import into refs/heads/workingbranch

8 years agoImport ostree_2018.4.orig.tar.xz
Simon McVittie [Sat, 24 Mar 2018 19:20:08 +0000 (19:20 +0000)]
Import ostree_2018.4.orig.tar.xz

[dgit import orig ostree_2018.4.orig.tar.xz]

8 years agoostree (2018.4-1) unstable; urgency=medium
Simon McVittie [Sat, 24 Mar 2018 19:20:08 +0000 (19:20 +0000)]
ostree (2018.4-1) unstable; urgency=medium

  * New upstream release
  * Update symbols file
  * Drop all patches, including one that was previously considered to
    be Debian-specific (moving to Python 3 for tests)
  * d/copyright: Update

[dgit import unpatched ostree 2018.4-1]

8 years agoImport ostree_2018.4-1.debian.tar.xz
Simon McVittie [Sat, 24 Mar 2018 19:20:08 +0000 (19:20 +0000)]
Import ostree_2018.4-1.debian.tar.xz

[dgit import tarball ostree 2018.4-1 ostree_2018.4-1.debian.tar.xz]

8 years agoMerge version 2018.1-1+rpi1 and 2018.2-1 to produce 2018.2-1+rpi1 archive/raspbian/2018.2-1+rpi1 raspbian/2018.2-1+rpi1
Raspbian forward pporter [Thu, 1 Mar 2018 05:46:13 +0000 (05:46 +0000)]
Merge version 2018.1-1+rpi1 and 2018.2-1 to produce 2018.2-1+rpi1

8 years agoMerge ostree (2018.2-1) import into refs/heads/workingbranch
Simon McVittie [Mon, 19 Feb 2018 09:01:47 +0000 (09:01 +0000)]
Merge ostree (2018.2-1) import into refs/heads/workingbranch

8 years agoUse Python 3 for tests
Simon McVittie [Wed, 17 Jan 2018 14:07:32 +0000 (14:07 +0000)]
Use Python 3 for tests

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Topic debian
Gbp-Pq: Name Use-Python-3-for-tests.patch

8 years agotests/bootloader-entries-crosscheck: Use Python 3-friendly sorting
Simon McVittie [Wed, 17 Jan 2018 15:19:12 +0000 (15:19 +0000)]
tests/bootloader-entries-crosscheck: Use Python 3-friendly sorting

This is a little clearer than a strcmp()-style negative/zero/positive
return, and also works in Python 2.

Signed-off-by: Simon McVittie <smcv@debian.org>
Forwarded: https://github.com/ostreedev/ostree/pull/1457

Gbp-Pq: Name tests-bootloader-entries-crosscheck-Use-Python-3-friendly.patch

8 years agotest-concurrency: Explicitly use floor division
Simon McVittie [Wed, 17 Jan 2018 15:03:59 +0000 (15:03 +0000)]
test-concurrency: Explicitly use floor division

Python 3 is pickier about this. Python 2.7 has Python 3-compatible
semantics for division when the division feature is imported from the
future.

Signed-off-by: Simon McVittie <smcv@debian.org>
Forwarded: https://github.com/ostreedev/ostree/pull/1457

Gbp-Pq: Name test-concurrency-Explicitly-use-floor-division.patch

8 years agotest-concurrency: Replace range with xrange
Simon McVittie [Wed, 17 Jan 2018 14:42:20 +0000 (14:42 +0000)]
test-concurrency: Replace range with xrange

range in Python 3 does what xrange did in Python 2. This still works in
Python 2, it just uses a bit more memory.

Signed-off-by: Simon McVittie <smcv@debian.org>
Forwarded: https://github.com/ostreedev/ostree/pull/1457

Gbp-Pq: Name test-concurrency-Replace-range-with-xrange.patch

8 years agotest-concurrency: Use Python 3 syntax for octal
Simon McVittie [Wed, 17 Jan 2018 14:25:26 +0000 (14:25 +0000)]
test-concurrency: Use Python 3 syntax for octal

This also works in Python 2.7, and is a little clearer.

Signed-off-by: Simon McVittie <smcv@debian.org>
Forwarded: https://github.com/ostreedev/ostree/pull/1457

Gbp-Pq: Name test-concurrency-Use-Python-3-syntax-for-octal.patch

8 years agoostree (2018.2-1) unstable; urgency=medium
Simon McVittie [Mon, 19 Feb 2018 09:01:47 +0000 (09:01 +0000)]
ostree (2018.2-1) unstable; urgency=medium

  * New upstream release
    - d/copyright: Update
    - d/libostree-1-1.symbols: Update
  * Mark patches as forwarded
  * Add gnupg to build-time test dependencies (and ostree-tests
    dependencies) as it is no longer transitively build-essential

[dgit import unpatched ostree 2018.2-1]

8 years agoImport ostree_2018.2.orig.tar.xz
Simon McVittie [Mon, 19 Feb 2018 09:01:47 +0000 (09:01 +0000)]
Import ostree_2018.2.orig.tar.xz

[dgit import orig ostree_2018.2.orig.tar.xz]

8 years agoImport ostree_2018.2-1.debian.tar.xz
Simon McVittie [Mon, 19 Feb 2018 09:01:47 +0000 (09:01 +0000)]
Import ostree_2018.2-1.debian.tar.xz

[dgit import tarball ostree 2018.2-1 ostree_2018.2-1.debian.tar.xz]

8 years agoMerge version 2017.15-2+rpi1 and 2018.1-1 to produce 2018.1-1+rpi1 archive/raspbian/2018.1-1+rpi1 raspbian/2018.1-1+rpi1
Raspbian forward pporter [Wed, 24 Jan 2018 08:28:24 +0000 (08:28 +0000)]
Merge version 2017.15-2+rpi1 and 2018.1-1 to produce 2018.1-1+rpi1

8 years agoMerge ostree (2018.1-1) import into refs/heads/workingbranch
Simon McVittie [Wed, 17 Jan 2018 15:34:46 +0000 (15:34 +0000)]
Merge ostree (2018.1-1) import into refs/heads/workingbranch

8 years agoUse Python 3 for tests
Simon McVittie [Wed, 17 Jan 2018 14:07:32 +0000 (14:07 +0000)]
Use Python 3 for tests

Forwarded: not-needed
Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Topic debian
Gbp-Pq: Name Use-Python-3-for-tests.patch

8 years agotests/bootloader-entries-crosscheck: Use Python 3-friendly sorting
Simon McVittie [Wed, 17 Jan 2018 15:19:12 +0000 (15:19 +0000)]
tests/bootloader-entries-crosscheck: Use Python 3-friendly sorting

This is a little clearer than a strcmp()-style negative/zero/positive
return, and also works in Python 2.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Name tests-bootloader-entries-crosscheck-Use-Python-3-friendly.patch

8 years agotest-concurrency: Explicitly use floor division
Simon McVittie [Wed, 17 Jan 2018 15:03:59 +0000 (15:03 +0000)]
test-concurrency: Explicitly use floor division

Python 3 is pickier about this. Python 2.7 has Python 3-compatible
semantics for division when the division feature is imported from the
future.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Name test-concurrency-Explicitly-use-floor-division.patch

8 years agotest-concurrency: Replace range with xrange
Simon McVittie [Wed, 17 Jan 2018 14:42:20 +0000 (14:42 +0000)]
test-concurrency: Replace range with xrange

range in Python 3 does what xrange did in Python 2. This still works in
Python 2, it just uses a bit more memory.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Name test-concurrency-Replace-range-with-xrange.patch

8 years agotest-concurrency: Use Python 3 syntax for octal
Simon McVittie [Wed, 17 Jan 2018 14:25:26 +0000 (14:25 +0000)]
test-concurrency: Use Python 3 syntax for octal

This also works in Python 2.7, and is a little clearer.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Name test-concurrency-Use-Python-3-syntax-for-octal.patch

8 years agoostree (2018.1-1) unstable; urgency=medium
Simon McVittie [Wed, 17 Jan 2018 15:34:46 +0000 (15:34 +0000)]
ostree (2018.1-1) unstable; urgency=medium

  * New upstream release
    - d/copyright: Update
    - d/patches: Remove, applied upstream
    - d/libostree-1-1.symbols: Update
  * Move Vcs-* to salsa.debian.org
    d/p/test-concurrency-Explicitly-use-floor-division.patch,
    d/p/tests-bootloader-entries-crosscheck-Use-Python-3-friendly.patch:
    Make tests compatible with Python 3
  * d/control, d/p/debian/Use-Python-3-for-tests.patch:
    Switch build-time tests and autopkgtests to Python 3

[dgit import unpatched ostree 2018.1-1]

8 years agoImport ostree_2018.1.orig.tar.xz
Simon McVittie [Wed, 17 Jan 2018 15:34:46 +0000 (15:34 +0000)]
Import ostree_2018.1.orig.tar.xz

[dgit import orig ostree_2018.1.orig.tar.xz]

8 years agoImport ostree_2018.1-1.debian.tar.xz
Simon McVittie [Wed, 17 Jan 2018 15:34:46 +0000 (15:34 +0000)]
Import ostree_2018.1-1.debian.tar.xz

[dgit import tarball ostree 2018.1-1 ostree_2018.1-1.debian.tar.xz]

8 years agoMerge version 2017.14-1+rpi1 and 2017.15-2 to produce 2017.15-2+rpi1 archive/raspbian/2017.15-2+rpi1 raspbian/2017.15-2+rpi1
Raspbian forward pporter [Thu, 11 Jan 2018 18:46:19 +0000 (18:46 +0000)]
Merge version 2017.14-1+rpi1 and 2017.15-2 to produce 2017.15-2+rpi1

8 years agoMerge ostree (2017.15-2) import into refs/heads/workingbranch
Simon McVittie [Thu, 4 Jan 2018 19:26:16 +0000 (19:26 +0000)]
Merge ostree (2017.15-2) import into refs/heads/workingbranch

8 years agotests: Assert that byte-order is swapped on LE but not BE CPUs
Simon McVittie [Wed, 3 Jan 2018 08:23:10 +0000 (08:23 +0000)]
tests: Assert that byte-order is swapped on LE but not BE CPUs

Closes: #1392
Signed-off-by: Simon McVittie <smcv@collabora.com>
Closes: #1393
Approved by: cgwalters
Applied-upstream: 2018.1, commit:https://github.com/ostreedev/ostree/commit/994cd66744e559c92644f36028c6c262605ad75a

Gbp-Pq: Topic 2018.1
Gbp-Pq: Name tests-Assert-that-byte-order-is-swapped-on-LE-but-not-BE-.patch

8 years agotests: Don't assume uid == primary gid
Simon McVittie [Sun, 10 Dec 2017 19:39:38 +0000 (19:39 +0000)]
tests: Don't assume uid == primary gid

Nothing guarantees that each user has a group containing only
themselves. Even if they do, nothing guarantees that its group ID
equals the user ID, particularly if another user earlier in the same
range was created without a corresponding group or vice versa.

Signed-off-by: Simon McVittie <smcv@collabora.com>
Forwarded: https://github.com/ostreedev/ostree/pull/1390
Applied-upstream: 2018.1, commit:https://github.com/ostreedev/ostree/commit/f63e62fbd272edbdd9695a10d841fc6b5eb69df8

Gbp-Pq: Topic 2018.1
Gbp-Pq: Name tests-Don-t-assume-uid-primary-gid.patch

8 years agoostree (2017.15-2) unstable; urgency=medium
Simon McVittie [Thu, 4 Jan 2018 19:26:16 +0000 (19:26 +0000)]
ostree (2017.15-2) unstable; urgency=medium

  * d/p/2018.1/tests-Don-t-assume-uid-primary-gid.patch: Mark as applied
    upstream in 2018.1
  * d/p/2018.1/tests-Assert-that-byte-order-[etc.].patch:
    Add patch to fix test failures on big-endian machines
    (Closes: #886218)
  * Temporarily disable gjs tests. gjs is not currently installable on
    buildds due to the glibc transition and a long dependency chain
    involving systemd-shim being preferred over systemd-sysv, which
    ends with libnih1 Depends: libc6 (<< 2.26).

[dgit import unpatched ostree 2017.15-2]

8 years agoImport ostree_2017.15-2.debian.tar.xz
Simon McVittie [Thu, 4 Jan 2018 19:26:16 +0000 (19:26 +0000)]
Import ostree_2017.15-2.debian.tar.xz

[dgit import tarball ostree 2017.15-2 ostree_2017.15-2.debian.tar.xz]

8 years agoImport ostree_2017.15.orig.tar.xz
Simon McVittie [Tue, 2 Jan 2018 14:13:08 +0000 (14:13 +0000)]
Import ostree_2017.15.orig.tar.xz

[dgit import orig ostree_2017.15.orig.tar.xz]

8 years agoManual merge of version 2017.12-1+rpi1 and 2017.14-1 to produce 2017.14-1+rpi1 archive/raspbian/2017.14-1+rpi1 raspbian/2017.14-1+rpi1
Peter Michael Green [Sat, 23 Dec 2017 00:09:03 +0000 (00:09 +0000)]
Manual merge of version 2017.12-1+rpi1 and 2017.14-1 to produce 2017.14-1+rpi1

8 years agoMerge ostree (2017.14-1) import into refs/heads/workingbranch
Simon McVittie [Sun, 10 Dec 2017 19:42:36 +0000 (19:42 +0000)]
Merge ostree (2017.14-1) import into refs/heads/workingbranch

8 years agotests: Don't assume uid == primary gid
Simon McVittie [Sun, 10 Dec 2017 19:39:38 +0000 (19:39 +0000)]
tests: Don't assume uid == primary gid

Nothing guarantees that each user has a group containing only
themselves. Even if they do, nothing guarantees that its group ID
equals the user ID, particularly if another user earlier in the same
range was created without a corresponding group or vice versa.

Signed-off-by: Simon McVittie <smcv@collabora.com>
Gbp-Pq: Name tests-Don-t-assume-uid-primary-gid.patch

8 years agoostree (2017.14-1) unstable; urgency=medium
Simon McVittie [Sun, 10 Dec 2017 19:42:36 +0000 (19:42 +0000)]
ostree (2017.14-1) unstable; urgency=medium

  * New upstream release
    - d/libostree-1-1.symbols: Update
  * Standards-Version: 4.1.2 (no changes required)
  * d/p/tests-Don-t-assume-uid-primary-gid.patch:
    Add patch to fix automated test failure when uid != primary gid

[dgit import unpatched ostree 2017.14-1]

8 years agoImport ostree_2017.14.orig.tar.xz
Simon McVittie [Sun, 10 Dec 2017 19:42:36 +0000 (19:42 +0000)]
Import ostree_2017.14.orig.tar.xz

[dgit import orig ostree_2017.14.orig.tar.xz]

8 years agoImport ostree_2017.14-1.debian.tar.xz
Simon McVittie [Sun, 10 Dec 2017 19:42:36 +0000 (19:42 +0000)]
Import ostree_2017.14-1.debian.tar.xz

[dgit import tarball ostree 2017.14-1 ostree_2017.14-1.debian.tar.xz]

8 years agoMerge ostree (2017.13-1) import into refs/heads/workingbranch
Simon McVittie [Sat, 4 Nov 2017 14:05:06 +0000 (14:05 +0000)]
Merge ostree (2017.13-1) import into refs/heads/workingbranch

8 years agoostree (2017.13-1) unstable; urgency=medium
Simon McVittie [Sat, 4 Nov 2017 14:05:06 +0000 (14:05 +0000)]
ostree (2017.13-1) unstable; urgency=medium

  * New upstream release
    - d/patches: Drop all patches
    - d/copyright: Update
    - Update symbols file
  * d/control: Require dh-exec 0.23~, for build-profile support.
    Strictly speaking we might only need 0.15, but I'm not going to
    test versions older than the jessie backport.
  * Set Rules-Requires-Root to no
  * Standards-Version: 4.1.1 (no changes required)

[dgit import unpatched ostree 2017.13-1]

8 years agoImport ostree_2017.13.orig.tar.xz
Simon McVittie [Sat, 4 Nov 2017 14:05:06 +0000 (14:05 +0000)]
Import ostree_2017.13.orig.tar.xz

[dgit import orig ostree_2017.13.orig.tar.xz]

8 years agoImport ostree_2017.13-1.debian.tar.xz
Simon McVittie [Sat, 4 Nov 2017 14:05:06 +0000 (14:05 +0000)]
Import ostree_2017.13-1.debian.tar.xz

[dgit import tarball ostree 2017.13-1 ostree_2017.13-1.debian.tar.xz]

8 years agoMerge ostree (2017.12-2) import into refs/heads/workingbranch
Simon McVittie [Thu, 26 Oct 2017 23:19:45 +0000 (00:19 +0100)]
Merge ostree (2017.12-2) import into refs/heads/workingbranch

8 years agolib/repo: Fix loading commitstate with parent repos
Colin Walters [Wed, 25 Oct 2017 17:13:17 +0000 (13:13 -0400)]
lib/repo: Fix loading commitstate with parent repos

This makes the code nicer too. Properly unit testing this though really wants
like a whole set of stuff around parent repos...but we do have coverage of the
non-parent path in the current pull tests.

Closes: https://github.com/ostreedev/ostree/issues/1306
Closes: #1308
Approved by: alexlarsson
Origin: upstream, 2017.13, commit:90ebd48f6aaf45c47b48c44354359f973dcf22a8

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-repo-Fix-loading-commitstate-with-parent-repos.patch

8 years agofdio: allow NULL for fstatat_allow_noent stbuf
Jonathan Lebon [Fri, 6 Oct 2017 21:26:41 +0000 (21:26 +0000)]
fdio: allow NULL for fstatat_allow_noent stbuf

Often, the caller doesn't actually care about the details of the stat
struct itself, but just whether the entry exists or not. It does work
to just pass `NULL` directly to glibc in a quick test, but given that
the argument is tagged as `__nonnull` and that the documentation does
not explicitly specify this is supported, let's do this safely.

Origin: upstream (submodule libglnx), 2017.13, commit:5362f6bc3ff3e30f379e767b203d15c9e56d6f08

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name fdio-allow-NULL-for-fstatat_allow_noent-stbuf.patch

8 years agolib/repo: Properly handle NULL homedir when signing commit
Dan Nicholson [Tue, 24 Oct 2017 19:27:17 +0000 (19:27 +0000)]
lib/repo: Properly handle NULL homedir when signing commit

Without this, ostree_repo_sign_commit throws a critical message when no
homedir is provided:

(ostree gpg-sign:5034): GLib-GIO-CRITICAL **: g_file_new_for_path: assertion 'path != NULL' failed

Closes: #1305
Approved by: cgwalters
Origin: upstream, 2017.13, commit:63ce86d5977ebfbedd2cdfba1e4f6bd400a3a1b8

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-repo-Properly-handle-NULL-homedir-when-signing-commit.patch

8 years agotests: Add test-pull-bareuseronly
Colin Walters [Fri, 13 Oct 2017 22:14:17 +0000 (18:14 -0400)]
tests: Add test-pull-bareuseronly

I was going to fix a bug in the static deltas code and I noticed
we were missing `pull-test.sh` coverage for bareuseronly.

Obviously fixing this requires duplicating some of the bits we have in
`basic-test.sh`; need to hoist that into `libtest.sh`. For now though let's get
the coverage.

Closes: #1270
Approved by: jlebon
Origin: upstream, 2017.13, commit:95afe2848d65a8062f4a76adb1ab80b6bcc4d79e

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name tests-Add-test-pull-bareuseronly.patch

8 years agolib/sysroot: Fix error handling when mounting overlayfs fails
Colin Walters [Mon, 16 Oct 2017 19:29:38 +0000 (15:29 -0400)]
lib/sysroot: Fix error handling when mounting overlayfs fails

This isn't perfect, but at least we fix an error-overwrite error, and in
practice `ostree admin unlock` isn't wrapped by `rpm-ostree` yet, so spew to
stderr is OK.

Closes: https://github.com/ostreedev/ostree/issues/1273
Closes: #1279
Approved by: guyshapiro
Origin: upstream, 2017.13, commit:464965e6b4897f9c6d4487ca10eb5bc60ad9a670

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-sysroot-Fix-error-handling-when-mounting-overlayfs-fa.patch

8 years agoCope with xattr syscalls raising EOPNOTSUPP
Simon McVittie [Mon, 16 Oct 2017 11:51:04 +0000 (12:51 +0100)]
Cope with xattr syscalls raising EOPNOTSUPP

ENOTSUP and EOPNOTSUPP are numerically equal on most Linux ports,
but inexplicably differ on PA-RISC (hppa) and possibly other
rare architectures.

Signed-off-by: Simon McVittie <smcv@collabora.com>
Closes: #1275
Approved by: cgwalters
Applied-upstream: 2017.13, commit:a4723dafed722008ed1ee3c952b7ff8e3d9b9a45

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name Cope-with-xattr-syscalls-raising-EOPNOTSUPP.patch

8 years agolib/utils: Check for invalid UTF-8 in filenames
Simon McVittie [Thu, 26 Oct 2017 11:08:15 +0000 (12:08 +0100)]
lib/utils: Check for invalid UTF-8 in filenames

In case a filename contains invalid UTF-8 characters, libostree will
pass it to g_variant_builder_add() in create_tree_variant_from_hashes()
anyway, which leads to a critical warning from glib and an invalid
commit. This commit makes ostree print a useful error and exit instead.

Closes: #1271
Approved by: cgwalters
Origin: upstream, 2017.13, commit:2a9c5efe1d10b79681b0ee638994ead375be6597

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-utils-Check-for-invalid-UTF-8-in-filenames.patch

8 years agolib/deltas: Check cancellable during processing
Colin Walters [Thu, 12 Oct 2017 14:46:25 +0000 (10:46 -0400)]
lib/deltas: Check cancellable during processing

Let's react to `Ctrl-C` faster here. Noticed while I was doing an update on my
desktop and playing with cancellation.

Closes: #1266
Approved by: jlebon
Origin: upstream, 2017.13, commit:a1986b1a8083ef4f840973f1c73d932a52e43644

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-deltas-Check-cancellable-during-processing.patch

8 years agolib/deploy: Use _exit() for FIFREEZE watchdog
Colin Walters [Wed, 11 Oct 2017 21:02:02 +0000 (17:02 -0400)]
lib/deploy: Use _exit() for FIFREEZE watchdog

This works around an (IMO) SpiderMonkey bug - it tries to
clean up in a shared library destructor, but doesn't install a
`pthread_atfork()` handler to unset its state.

Closes: https://github.com/ostreedev/ostree/issues/1262
Closes: #1264
Approved by: dbnicholson
Origin: upstream, 2017.13, commit:8f6ec62bfb149ec8dfb6076228dd64e5df27a76b

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-deploy-Use-_exit-for-FIFREEZE-watchdog.patch

8 years agolib/deploy: Ignore FIFREEZE/FITHAW errors when already in state
Dan Nicholson [Wed, 11 Oct 2017 14:03:10 +0000 (14:03 +0000)]
lib/deploy: Ignore FIFREEZE/FITHAW errors when already in state

If the filesystem is already frozen, FIFREEZE returns EBUSY, and if the
filesystem is already thawed, FITHAW returns EINVAL. It's very unlikely
these issues would arise on a real ostree system since the sysroot would
be locked during the freeze/thaw cycle.

However, when multiple fake sysroots are used during the test suite (run
as root), the tests could race to run the freeze/thaw cycle without
locking. Furthermore, there's no reason why an independent process might
be trying to freeze the filesystem while ostree was deploying. Ignore
but warn for these errors since there's not much ostree can do about it,
anyways.

Closes: #1260
Approved by: cgwalters
Origin: upstream, 2017.13, commit:a5b7660c940a200adac1a7d217e4a1cd72719021

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-deploy-Ignore-FIFREEZE-FITHAW-errors-when-already-in-.patch

8 years agolib/sysroot: Fix pointer going out of scope in unlock code
Colin Walters [Wed, 4 Oct 2017 14:24:21 +0000 (10:24 -0400)]
lib/sysroot: Fix pointer going out of scope in unlock code

Fixes Coverity CID #1457317

Closes: #1249
Approved by: jlebon
Origin: upstream, 2017.13, commit:351ffdb9778436b193ba9d2fbeebe2358e55004f

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-sysroot-Fix-pointer-going-out-of-scope-in-unlock-code.patch

8 years agolib/core: Init struct stat buffer
Colin Walters [Wed, 4 Oct 2017 14:22:05 +0000 (10:22 -0400)]
lib/core: Init struct stat buffer

Regression from d57410a7e62dcb89321807dcb2d91c85f9d26df7

Fixes Coverity CID #1457316

Closes: #1249
Approved by: jlebon
Origin: upstream, 2017.13, commit:e80efe0b0668a351361bb0a218a809434dd33d63

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-core-Init-struct-stat-buffer.patch

8 years agolib/repo-commit: Import detached metadata even if hardlink exists
Philip Withnall [Tue, 3 Oct 2017 14:45:34 +0000 (15:45 +0100)]
lib/repo-commit: Import detached metadata even if hardlink exists

Spotted while reading through the code, it looks like the
copy_detached_metadata() call is accidentally omitted if a hardlink
already exists for the .commit object.

Signed-off-by: Philip Withnall <withnall@endlessm.com>
Closes: #1242
Approved by: cgwalters
Origin: upstream, 2017.13, commit:86e072bdbe48a4f16efb05c00eb79114e5fdbf61

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-repo-commit-Import-detached-metadata-even-if-hardlink.patch

8 years agolib/pull: Fix regression with pull-local for nonexistent refs
Colin Walters [Mon, 2 Oct 2017 15:24:05 +0000 (11:24 -0400)]
lib/pull: Fix regression with pull-local for nonexistent refs

I was reading the pull code for the last release, and spotted
a bug in commit f923c2e1eaebe0c781f07d34ae1a03f94357bccd - in
the case where the ref doesn't exist, we don't set an error,
tripping an assertion in the main code.

The previous code wanted the ref to always exist, so just flip back the boolean
for "ignore noent". I moved the `g_strchomp()` just into the HTTP path - if a
local repo is corrupted in this way it's something to fix in that repo.

Closes: #1238
Approved by: pwithnall
Origin: upstream, 2017.13, commit:b8c15ae859de7a353b99c98c6266ee626cd94e7e

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-pull-Fix-regression-with-pull-local-for-nonexistent-r.patch

8 years agoAdd missing test-libglnx-shutil
Simon McVittie [Tue, 3 Oct 2017 20:30:36 +0000 (21:30 +0100)]
Add missing test-libglnx-shutil

Because we re-run autogen.sh, we'll regenerate Makefile-libglnx.am.inc
(which results in the version after this patch, including
test-libglnx-shutil, being present) and then try to build
test-libglnx-shutil; but its source code wasn't included in the upstream
tarball, because Makefile-libglnx.am.inc wasn't up to date at
"make dist" time.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Topic dist
Gbp-Pq: Name Add-missing-test-libglnx-shutil.patch

8 years agoostree (2017.12-2) unstable; urgency=medium
Simon McVittie [Thu, 26 Oct 2017 23:19:45 +0000 (00:19 +0100)]
ostree (2017.12-2) unstable; urgency=medium

  * Disable gtk-doc if we are not going to build libostree-doc,
    in particular for architecture-specific builds. Note that it remains
    in Build-Depends (not Build-Depends-Indep) because it is also needed
    for gtkdocize during dh_autoreconf.
    - In particular this might fix FTBFS on sparc64, where highlight(1)
      fails.
  * d/p/2017.13/lib-core-Init-struct-stat-buffer.patch,
    d/p/2017.13/lib-sysroot-Fix-pointer-going-out-of-scope-in-unlock-code.patch,
    d/p/2017.13/lib-deploy-Ignore-FIFREEZE-FITHAW-errors-when-already-in-.patch,
    d/p/2017.13/lib-deploy-Use-_exit-for-FIFREEZE-watchdog.patch,
    d/p/2017.13/lib-deltas-Check-cancellable-during-processing.patch,
    d/p/2017.13/lib-utils-Check-for-invalid-UTF-8-in-filenames.patch,
    d/p/2017.13/Cope-with-xattr-syscalls-raising-EOPNOTSUPP.patch,
    d/p/2017.13/lib-sysroot-Fix-error-handling-when-mounting-overlayfs-fa.patch,
    d/p/2017.13/lib-repo-Properly-handle-NULL-homedir-when-signing-commit.patch,
    d/p/2017.13/fdio-allow-NULL-for-fstatat_allow_noent-stbuf.patch,
    d/p/2017.13/lib-repo-Fix-loading-commitstate-with-parent-repos.patch:
    Add various bugfix patches from upstream
    - In particular, dealing with the possibility that EOPNOTSUPP != ENOTSUP
      should fix test failures on hppa.
  * d/p/2017.13/tests-Add-test-pull-bareuseronly.patch:
    Add more test coverage from upstream
    - d/rules: Make the new test executable
  * d/test.sh: Clean up ostree-trivial-httpd processes
  * d/test.sh: Don't repeat build-time tests if they fail once. They seem
    to be somewhat reliable now.

[dgit import unpatched ostree 2017.12-2]

8 years agoImport ostree_2017.12-2.debian.tar.xz
Simon McVittie [Thu, 26 Oct 2017 23:19:45 +0000 (00:19 +0100)]
Import ostree_2017.12-2.debian.tar.xz

[dgit import tarball ostree 2017.12-2 ostree_2017.12-2.debian.tar.xz]

8 years agoMerge version 2017.11-2+rpi1 and 2017.12-1 to produce 2017.12-1+rpi1 archive/raspbian/2017.12-1+rpi1 raspbian/2017.12-1+rpi1
Raspbian forward pporter [Thu, 12 Oct 2017 11:52:04 +0000 (12:52 +0100)]
Merge version 2017.11-2+rpi1 and 2017.12-1 to produce 2017.12-1+rpi1

8 years agoMerge ostree (2017.12-1) import into refs/heads/workingbranch
Simon McVittie [Tue, 3 Oct 2017 21:47:48 +0000 (22:47 +0100)]
Merge ostree (2017.12-1) import into refs/heads/workingbranch

8 years agoImport ostree_2017.12.orig.tar.xz
Simon McVittie [Tue, 3 Oct 2017 21:47:48 +0000 (22:47 +0100)]
Import ostree_2017.12.orig.tar.xz

[dgit import orig ostree_2017.12.orig.tar.xz]

8 years agoAdd missing test-libglnx-shutil
Simon McVittie [Tue, 3 Oct 2017 20:30:36 +0000 (21:30 +0100)]
Add missing test-libglnx-shutil

Because we re-run autogen.sh, we'll regenerate Makefile-libglnx.am.inc
(which results in the version after this patch, including
test-libglnx-shutil, being present) and then try to build
test-libglnx-shutil; but its source code wasn't included in the upstream
tarball, because Makefile-libglnx.am.inc wasn't up to date at
"make dist" time.

Signed-off-by: Simon McVittie <smcv@debian.org>
Gbp-Pq: Topic dist
Gbp-Pq: Name Add-missing-test-libglnx-shutil.patch

8 years agolib/repo-commit: Import detached metadata even if hardlink exists
Philip Withnall [Tue, 3 Oct 2017 14:45:34 +0000 (15:45 +0100)]
lib/repo-commit: Import detached metadata even if hardlink exists

Spotted while reading through the code, it looks like the
copy_detached_metadata() call is accidentally omitted if a hardlink
already exists for the .commit object.

Signed-off-by: Philip Withnall <withnall@endlessm.com>
Closes: #1242
Approved by: cgwalters
Origin: upstream, 2017.13, commit:86e072bdbe48a4f16efb05c00eb79114e5fdbf61

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-repo-commit-Import-detached-metadata-even-if-hardlink.patch

8 years agolib/pull: Fix regression with pull-local for nonexistent refs
Colin Walters [Mon, 2 Oct 2017 15:24:05 +0000 (11:24 -0400)]
lib/pull: Fix regression with pull-local for nonexistent refs

I was reading the pull code for the last release, and spotted
a bug in commit f923c2e1eaebe0c781f07d34ae1a03f94357bccd - in
the case where the ref doesn't exist, we don't set an error,
tripping an assertion in the main code.

The previous code wanted the ref to always exist, so just flip back the boolean
for "ignore noent". I moved the `g_strchomp()` just into the HTTP path - if a
local repo is corrupted in this way it's something to fix in that repo.

Closes: #1238
Approved by: pwithnall
Origin: upstream, 2017.13, commit:b8c15ae859de7a353b99c98c6266ee626cd94e7e

Gbp-Pq: Topic 2017.13
Gbp-Pq: Name lib-pull-Fix-regression-with-pull-local-for-nonexistent-r.patch

8 years agoostree (2017.12-1) unstable; urgency=medium
Simon McVittie [Tue, 3 Oct 2017 21:47:48 +0000 (22:47 +0100)]
ostree (2017.12-1) unstable; urgency=medium

  * New upstream release
    - Drop all current patches, applied upstream
    - Update symbols file
  * Add some post-release bug fix patches
  * Simplify autopkgtest now that test-local-pull seems to be stable
  * Unexport HTTP proxy variables in autopkgtest to work around lack of
    support for no_proxy, which breaks the tests on Ubuntu autopkgtest.
    We don't actually need Internet access, so this is OK.
  * debian/test.sh: Unexport HTTP proxy variables for build-time tests
    too
  * Make build-time test failures fatal if they fail at least twice
    out of 5 tries (previously they had to fail at least 3 times)
  * Add patch to reinstate test-libglnx-shutil.c, which was missed out
    of the upstream tarball

[dgit import unpatched ostree 2017.12-1]

8 years agoImport ostree_2017.12-1.debian.tar.xz
Simon McVittie [Tue, 3 Oct 2017 21:47:48 +0000 (22:47 +0100)]
Import ostree_2017.12-1.debian.tar.xz

[dgit import tarball ostree 2017.12-1 ostree_2017.12-1.debian.tar.xz]

8 years agoMerge ostree (2017.11-2+rpi1) import into refs/heads/workingbranch
Peter Michael Green [Thu, 28 Sep 2017 16:17:06 +0000 (17:17 +0100)]
Merge ostree (2017.11-2+rpi1) import into refs/heads/workingbranch

8 years agorofiles-fuse: also pass mode for O_RDONLY
Jonathan Lebon [Wed, 20 Sep 2017 18:38:16 +0000 (18:38 +0000)]
rofiles-fuse: also pass mode for O_RDONLY

In the `O_RDONLY` case, we were calling `openat` without a mode
argument. However, it's perfectly legal (albeit unusual) to do
`open(O_RDONLY|O_CREAT)`. One such application that makes use of this is
`flock(1)`.

This was actually caught by `_FORTIFY_SOURCE=2`, and once we run
`rofiles-fuse` with `-f`, the message is clear:

```
*** invalid openat64 call: O_CREAT or O_TMPFILE without mode ***:
rofiles-fuse terminated
======= Backtrace: =========
/lib64/libc.so.6(+0x7c8dc)[0x7f36d9f188dc]
/lib64/libc.so.6(__fortify_fail+0x37)[0x7f36d9fbfaa7]
/lib64/libc.so.6(+0x10019a)[0x7f36d9f9c19a]
rofiles-fuse[0x401768]
...
```

Without `_FORTIFY_SOURCE`, the file gets created, but its mode is
completely random.

I ran into this while investigating
https://github.com/projectatomic/rpm-ostree/pull/1003.

Closes: #1200
Approved by: cgwalters
Origin: upstream, 2017.12, commit:d4c7093e370843c57eab2f89f0c39ef449e6b32e

Gbp-Pq: Topic 2017.12
Gbp-Pq: Name rofiles-fuse-also-pass-mode-for-O_RDONLY.patch

8 years agotests: Reset umask to 022 while creating test repository
Simon McVittie [Tue, 19 Sep 2017 17:37:58 +0000 (18:37 +0100)]
tests: Reset umask to 022 while creating test repository

In test-basic-root.sh we make assertions about the permissions
of files like baz/cow, which were created without an explicit chmod.
We can't do that unless we control the permissions.

For some reason the "debomatic" autobuilder used to do some Debian
archive rebuilds does the entire build including build-time tests
as uid 0 with umask 002, which broke those assertions. This seems
a weird thing to do, and I've opened a bug, but it also seems
reasonable to fix this test.

This also lets us remove a couple of existing workarounds for the
same issue.

Bug-Debian: https://bugs.debian.org/876138
Signed-off-by: Simon McVittie <smcv@collabora.com>
Closes: #1192
Approved by: cgwalters
Forwarded: https://github.com/ostreedev/ostree/pull/1192
Applied-upstream: 2017.12, commit:e3c3ec5dd91492e82c79223052443d038c60f41c

Gbp-Pq: Topic 2017.12
Gbp-Pq: Name tests-Reset-umask-to-022-while-creating-test-repository.patch