ostree.git
23 months agoMerge version 2023.7-3+rpi1 and 2023.8-2 to produce 2023.8-2+rpi1 archive/raspbian/2023.8-2+rpi1 raspbian/2023.8-2+rpi1
Raspbian automatic forward porter [Wed, 10 Jan 2024 04:32:00 +0000 (04:32 +0000)]
Merge version 2023.7-3+rpi1 and 2023.8-2 to produce 2023.8-2+rpi1

23 months agoMake fast forward from 2023.8-1
Simon McVittie [Sun, 31 Dec 2023 15:34:37 +0000 (15:34 +0000)]
Make fast forward from 2023.8-1

[dgit --quilt=unapplied]

23 months agoSkip test-admin-deploy-uboot.sh on s390x
Simon McVittie [Thu, 26 Oct 2023 10:26:04 +0000 (11:26 +0100)]
Skip test-admin-deploy-uboot.sh on s390x

It fails on a porterbox. ostree hard-codes zipl to be used on s390x,
so it's reasonable that tests for other bootloaders might not work.

Bug: https://github.com/ostreedev/ostree/issues/3086
Forwarded: no

Gbp-Pq: Topic debian
Gbp-Pq: Name Skip-test-admin-deploy-uboot.sh-on-s390x.patch

23 months agotest-sysroot: Skip on s390x by default
Simon McVittie [Tue, 6 Dec 2022 10:59:33 +0000 (10:59 +0000)]
test-sysroot: Skip on s390x by default

This test regularly fails on the buildds, but I cannot reproduce the
failure on a porterbox.

Bug: https://github.com/ostreedev/ostree/issues/2527
Bug-Debian: https://bugs.debian.org/1025532
Forwarded: not-needed

Gbp-Pq: Topic debian
Gbp-Pq: Name test-sysroot-Skip-on-s390x-by-default.patch

23 months 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

23 months agod/rules: Fix autoreconf during nodoc builds by providing a stub gtkdocize
Simon McVittie [Sun, 31 Dec 2023 15:46:49 +0000 (15:46 +0000)]
d/rules: Fix autoreconf during nodoc builds by providing a stub gtkdocize

23 months agoRelease to unstable
Simon McVittie [Sun, 31 Dec 2023 15:34:50 +0000 (15:34 +0000)]
Release to unstable

23 months agod/control, d/rules: Enable cross-compilation
Simon McVittie [Sun, 31 Dec 2023 15:34:21 +0000 (15:34 +0000)]
d/control, d/rules: Enable cross-compilation

When cross-compiling, this requires a version of gobject-introspection
that is currently only available in experimental. When not
cross-compiling, the version in testing is sufficient.

2 years agoRelease to unstable
Simon McVittie [Fri, 8 Dec 2023 10:35:22 +0000 (10:35 +0000)]
Release to unstable

2 years agoMake fast forward from 2023.7-3
Simon McVittie [Fri, 8 Dec 2023 10:35:12 +0000 (10:35 +0000)]
Make fast forward from 2023.7-3

[dgit --quilt=unapplied]

2 years agoSkip test-admin-deploy-uboot.sh on s390x
Simon McVittie [Thu, 26 Oct 2023 10:26:04 +0000 (11:26 +0100)]
Skip test-admin-deploy-uboot.sh on s390x

It fails on a porterbox. ostree hard-codes zipl to be used on s390x,
so it's reasonable that tests for other bootloaders might not work.

Bug: https://github.com/ostreedev/ostree/issues/3086
Forwarded: no

Gbp-Pq: Topic debian
Gbp-Pq: Name Skip-test-admin-deploy-uboot.sh-on-s390x.patch

2 years agotest-sysroot: Skip on s390x by default
Simon McVittie [Tue, 6 Dec 2022 10:59:33 +0000 (10:59 +0000)]
test-sysroot: Skip on s390x by default

This test regularly fails on the buildds, but I cannot reproduce the
failure on a porterbox.

Bug: https://github.com/ostreedev/ostree/issues/2527
Bug-Debian: https://bugs.debian.org/1025532
Forwarded: not-needed

Gbp-Pq: Topic debian
Gbp-Pq: Name test-sysroot-Skip-on-s390x-by-default.patch

2 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

2 years agoUpdate changelog
Simon McVittie [Fri, 8 Dec 2023 10:34:21 +0000 (10:34 +0000)]
Update changelog

2 years agod/p/bootloader-zipl-No-op-if-run-as-non-root.patch: Drop patch, applied upstream
Simon McVittie [Fri, 8 Dec 2023 10:33:58 +0000 (10:33 +0000)]
d/p/bootloader-zipl-No-op-if-run-as-non-root.patch: Drop patch, applied upstream

2 years agod/libostree-1-1.symbols: Update
Simon McVittie [Fri, 8 Dec 2023 10:20:34 +0000 (10:20 +0000)]
d/libostree-1-1.symbols: Update

2 years agoNew upstream release
Simon McVittie [Fri, 8 Dec 2023 10:15:24 +0000 (10:15 +0000)]
New upstream release

2 years agoUpdate upstream source from tag 'upstream/2023.8'
Simon McVittie [Fri, 8 Dec 2023 10:14:41 +0000 (10:14 +0000)]
Update upstream source from tag 'upstream/2023.8'

Update to upstream version '2023.8'
with Debian dir d00b8908db1e3bfc0d025e9ae84885daed42ed22

2 years agoNew upstream version 2023.8
Simon McVittie [Fri, 8 Dec 2023 10:14:33 +0000 (10:14 +0000)]
New upstream version 2023.8

2 years agoRelease 2023.8
Colin Walters [Mon, 4 Dec 2023 23:28:32 +0000 (18:28 -0500)]
Release 2023.8

2 years agoMerge pull request #3106 from cgwalters/commit-reflink
Colin Walters [Tue, 5 Dec 2023 18:36:00 +0000 (13:36 -0500)]
Merge pull request #3106 from cgwalters/commit-reflink

commit: Try reflinks for local commits by default

2 years agocommit: Try reflinks for local commits by default
Colin Walters [Sat, 2 Dec 2023 20:48:12 +0000 (15:48 -0500)]
commit: Try reflinks for local commits by default

I think we originally used to do this, but at some point in a
code refactoring, this optimization got lost.

It's a quite important optimization for the case of writing content
generated by an external system into an ostree repository.

2 years agoMerge pull request #3105 from cgwalters/sysroot-borrow-fd
Joseph Marrero Corchado [Mon, 4 Dec 2023 23:22:50 +0000 (18:22 -0500)]
Merge pull request #3105 from cgwalters/sysroot-borrow-fd

sysroot: Add a method to borrow sysroot fd

2 years agoMerge pull request #3107 from cgwalters/drop-userxattr-test
Colin Walters [Mon, 4 Dec 2023 18:14:07 +0000 (13:14 -0500)]
Merge pull request #3107 from cgwalters/drop-userxattr-test

tests: Drop bare-user on tmpfs negative test

2 years agotests: Drop bare-user on tmpfs negative test
Colin Walters [Sun, 3 Dec 2023 14:32:36 +0000 (09:32 -0500)]
tests: Drop bare-user on tmpfs negative test

As of https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=2daf18a7884dc03d5164ab9c7dc3f2ea70638469
the restriction is lifted, FINALLY.

2 years agoMerge pull request #3104 from cgwalters/s390x-target
Colin Walters [Fri, 1 Dec 2023 21:21:34 +0000 (16:21 -0500)]
Merge pull request #3104 from cgwalters/s390x-target

bootloader/zipl: Run in target deployment as container if needed

2 years agobootloader/zipl: Run in target deployment as container if needed
Colin Walters [Thu, 30 Nov 2023 23:16:33 +0000 (18:16 -0500)]
bootloader/zipl: Run in target deployment as container if needed

xref https://issues.redhat.com/browse/MGMT-16303

Basically the OCP Assisted installer has now grown code
to try to do OS updates offline post-install, and this means
we need to handle the case of running zipl from the target
root.

2 years agosysroot: Expose deployment runner outside of selinux
Colin Walters [Fri, 1 Dec 2023 20:05:40 +0000 (15:05 -0500)]
sysroot: Expose deployment runner outside of selinux

We will use it in e.g. s390x zipl backend.

2 years agoMerge pull request #3103 from cgwalters/tmpfiles-factory-var
Joseph Marrero Corchado [Fri, 1 Dec 2023 19:46:39 +0000 (14:46 -0500)]
Merge pull request #3103 from cgwalters/tmpfiles-factory-var

tmpfiles: Copy `/usr/share/factory/var` to `/var`

2 years agosysroot: Add a method to borrow sysroot fd
Colin Walters [Fri, 1 Dec 2023 19:35:41 +0000 (14:35 -0500)]
sysroot: Add a method to borrow sysroot fd

For the same reason we have this method on `Repo`.

2 years agosysroot: Support specifying bwrap arguments
Colin Walters [Fri, 1 Dec 2023 13:14:15 +0000 (08:14 -0500)]
sysroot: Support specifying bwrap arguments

Prep for use with zipl.

2 years agosysroot: Expose deployment container executor internally
Colin Walters [Thu, 30 Nov 2023 16:12:55 +0000 (11:12 -0500)]
sysroot: Expose deployment container executor internally

Prep for using this for zipl.

2 years agotmpfiles: Copy `/usr/share/factory/var` to `/var`
Colin Walters [Thu, 30 Nov 2023 20:33:42 +0000 (15:33 -0500)]
tmpfiles: Copy `/usr/share/factory/var` to `/var`

This is a pattern we want to encourage.  It's honestly just
way simpler than what rpm-ostree is doing today in auto-synthesizing
individual tmpfiles.d snippets.

2 years agoMerge pull request #3102 from cgwalters/deploy-cleanup-run
Jonathan Lebon [Thu, 30 Nov 2023 22:34:29 +0000 (17:34 -0500)]
Merge pull request #3102 from cgwalters/deploy-cleanup-run

2 years agosysroot/deploy: Drop unnecessary g_ptr_array_free
Colin Walters [Thu, 30 Nov 2023 15:26:32 +0000 (10:26 -0500)]
sysroot/deploy: Drop unnecessary g_ptr_array_free

We can just reference the pdata, and autoptr the array.

2 years agodeploy: Use NULL terminated strv
Colin Walters [Thu, 30 Nov 2023 15:25:12 +0000 (10:25 -0500)]
deploy: Use NULL terminated strv

This makes the code simpler and avoids the need
to keep the string array and length in sync.  Prep for
further callers.

2 years agoMerge pull request #3100 from cgwalters/misc-finalization
Jonathan Lebon [Wed, 29 Nov 2023 19:10:22 +0000 (14:10 -0500)]
Merge pull request #3100 from cgwalters/misc-finalization

2 years agotests: Add a dedicated finalization test
Colin Walters [Mon, 27 Nov 2023 20:34:20 +0000 (15:34 -0500)]
tests: Add a dedicated finalization test

This one covers `admin lock-finalization --unlock`.

2 years agoMerge pull request #3101 from cgwalters/drop-redundant-includes
Joseph Marrero Corchado [Tue, 28 Nov 2023 02:28:43 +0000 (21:28 -0500)]
Merge pull request #3101 from cgwalters/drop-redundant-includes

main: Drop some redundant includes of `ot-main.h`

2 years agomain: Drop some redundant includes of `ot-main.h`
Colin Walters [Tue, 28 Nov 2023 01:08:43 +0000 (20:08 -0500)]
main: Drop some redundant includes of `ot-main.h`

When backporting a patch recently we hit a non-obvious
dependency on another fix for `ot-main.h` includes.  Clean
this up a bit by dropping the redundant includes.

2 years agoman: Two finalization fixes
Colin Walters [Mon, 27 Nov 2023 18:50:07 +0000 (13:50 -0500)]
man: Two finalization fixes

From PR review.

2 years agoMerge pull request #3090 from cgwalters/stabilize-locking
Colin Walters [Mon, 27 Nov 2023 18:48:59 +0000 (13:48 -0500)]
Merge pull request #3090 from cgwalters/stabilize-locking

sysroot: Stabilize deployment finalization, add API

2 years agosysroot: Stabilize deployment finalization, add API and CLI
Colin Walters [Wed, 8 Nov 2023 22:19:19 +0000 (17:19 -0500)]
sysroot: Stabilize deployment finalization, add API and CLI

It's about time we do this; deployment finalization locking
is a useful feature.  An absolutely key thing here is that
we've slowly been moving towards the deployments as the primary
"source of truth".

Specifically in bootc for example, we will GC container images
not referenced by a deployment.

This is then neecessary to support a "pull but don't apply automatically" model.

This stabilizes the existing `ostree admin deploy --lock-finalization`
CLI, and adds a new `ostree admin unlock-finalization`.

We still check the old lock file path, but there's a new boolean
value as part of the staged deployment data which is intended
to be the source of truth in the future.  At some point then we
can drop the rpm-ostree lockfile handling.

Closes: https://github.com/ostreedev/ostree/issues/3025
2 years agostatus: Add error prefix for gpg verification
Colin Walters [Mon, 27 Nov 2023 15:54:18 +0000 (10:54 -0500)]
status: Add error prefix for gpg verification

It's helpful to see which deployment has an error.

2 years agotests: Turn off gpg verification for dev builds
Colin Walters [Mon, 27 Nov 2023 15:59:04 +0000 (10:59 -0500)]
tests: Turn off gpg verification for dev builds

Right now `ostree admin status` errors out in this case, but
`rpm-ostree status` doesn't.  The former behavior is probably
more of a bug, work around it for now.

2 years agoMerge pull request #3099 from markmc/man-sysroot-readonly
Colin Walters [Mon, 27 Nov 2023 13:39:35 +0000 (08:39 -0500)]
Merge pull request #3099 from markmc/man-sysroot-readonly

man: improve sysroot.readonly docs

2 years agoman: improve sysroot.readonly docs
Mark McLoughlin [Fri, 24 Nov 2023 14:27:11 +0000 (14:27 +0000)]
man: improve sysroot.readonly docs

The explanation of sysroot.readonly is a little confusing - we say
that "everything else is mounted read-only" but it's perhaps clearer
to say /sysroot is mounted read-only.

Also note that read-only is the default with composefs.

Finally, document the option in ostree.repo-config even though it is
now considered legacy - as of commit 22b8e4f9 (#2930) - it is still
commonly seen in repo configs, so users will look to understand
what it means.

2 years agoMerge pull request #3098 from jlebon/pr/abort-concurrent
Jonathan Lebon [Tue, 21 Nov 2023 22:18:17 +0000 (17:18 -0500)]
Merge pull request #3098 from jlebon/pr/abort-concurrent

2 years agoci: cancel previous build on PR update
Jonathan Lebon [Tue, 21 Nov 2023 20:31:38 +0000 (15:31 -0500)]
ci: cancel previous build on PR update

This is an easy way to save CI resources; when a PR is updated, abort
any previous build for that PR to focus on testing the latest push.

2 years agoMerge pull request #3097 from cgwalters/symbol-cleanups
Colin Walters [Tue, 21 Nov 2023 13:34:32 +0000 (08:34 -0500)]
Merge pull request #3097 from cgwalters/symbol-cleanups

Symbol cleanups

2 years agolib: Don't hardcode year in sample symver section
Colin Walters [Mon, 20 Nov 2023 14:47:40 +0000 (09:47 -0500)]
lib: Don't hardcode year in sample symver section

The 2021 was misleading.

2 years agodevel: Fix symbol versioning number
Colin Walters [Mon, 20 Nov 2023 14:45:39 +0000 (09:45 -0500)]
devel: Fix symbol versioning number

The .11 was wrong.

2 years agoMerge pull request #3095 from alexlarsson/fix-post-copy-symlinks
Colin Walters [Tue, 14 Nov 2023 22:40:08 +0000 (17:40 -0500)]
Merge pull request #3095 from alexlarsson/fix-post-copy-symlinks

Fix admin post-copy handling of symlinks

2 years agoFix admin post-copy handling of symlinks
Alexander Larsson [Tue, 14 Nov 2023 21:36:58 +0000 (22:36 +0100)]
Fix admin post-copy handling of symlinks

The code to enable fs-verity on an object file was failing with ENOENT
for symlink objects.

2 years agoMerge pull request #3094 from alexlarsson/admin-deploy-post-copy
Colin Walters [Tue, 14 Nov 2023 14:01:41 +0000 (09:01 -0500)]
Merge pull request #3094 from alexlarsson/admin-deploy-post-copy

Add `ostree admin post-copy` command

2 years agoAdd `ostree admin post-copy` command
Alexander Larsson [Mon, 13 Nov 2023 16:33:57 +0000 (17:33 +0100)]
Add `ostree admin post-copy` command

This command will apply fs-verity on all objects that need it and
needs to be called when an ostree deployment has been copied on a
file-by-file basis, which would loose information such as fs-verity.

This is needed by osbuild which works by creating the final image in a
rootfs, and then separately copying that rootfs file-by-file to a
loopback mounted filesystem image.

2 years agoMerge version 2023.6-2+rpi1 and 2023.7-3 to produce 2023.7-3+rpi1 archive/raspbian/2023.7-3+rpi1 raspbian/2023.7-3+rpi1
Raspbian automatic forward porter [Fri, 10 Nov 2023 22:59:45 +0000 (22:59 +0000)]
Merge version 2023.6-2+rpi1 and 2023.7-3 to produce 2023.7-3+rpi1

2 years agoMerge pull request #3091 from cgwalters/fix-soup3
Eric Curtin [Fri, 10 Nov 2023 16:39:13 +0000 (16:39 +0000)]
Merge pull request #3091 from cgwalters/fix-soup3

build-sys: Enable libsoup3 by default if installed

2 years agounlock: Don't pass options again to overlayfs
Colin Walters [Fri, 10 Nov 2023 01:53:52 +0000 (20:53 -0500)]
unlock: Don't pass options again to overlayfs

There seems to be a tricky regression here with the util-linux
support for the new mount API, plus overlays support for it.

```
[2023-11-09T21:05:30.633Z] Nov 09 21:05:26 qemu0 kola-runext-unlock-transient.sh[2108]: + unshare -m -- /bin/sh -c 'mount -o remount,rw /usr && echo hello from transient unlock >/usr/share/writable-usr-test'
[2023-11-09T21:05:30.633Z] Nov 09 21:05:26 qemu0 kola-runext-unlock-transient.sh[2148]: mount: /usr: mount point not mounted or bad option.
[2023-11-09T21:05:30.633Z] Nov 09 21:05:26 qemu0 kola-runext-unlock-transient.sh[2148]:        dmesg(1) may have more information after failed mount system call.
```

OK this seems related to the new mount API support in util-linux and overlayfs.  From a strace:

```
2095  open_tree(AT_FDCWD, "/usr", OPEN_TREE_CLOEXEC) = 3
2095  mount_setattr(-1, NULL, 0, NULL, 0) = -1 EINVAL (Invalid argument)
...
2095  fspick(3, "", FSPICK_NO_AUTOMOUNT|FSPICK_EMPTY_PATH) = 4
2095  fsconfig(4, FSCONFIG_SET_FLAG, "seclabel", NULL, 0) = 0
2095  fsconfig(4, FSCONFIG_SET_STRING, "lowerdir", "usr", 0) = -1 EINVAL (Invalid argument)
```

I think the core problem here is it's trying to reconfigure the mount with existing options,
but in the new mount namespace we can't see the lowerdir.

Here we really really just want to remount writable.  Telling
util-linux to not pass existing options fixes it.

2 years agoMerge pull request #3093 from cgwalters/rust-alias-stateroot
Colin Walters [Thu, 9 Nov 2023 20:21:47 +0000 (15:21 -0500)]
Merge pull request #3093 from cgwalters/rust-alias-stateroot

rust: Add a `stateroot()` alias in the Rust bindings

2 years agotests: Work around systemd regression for boot id parsing
Colin Walters [Thu, 9 Nov 2023 19:08:41 +0000 (14:08 -0500)]
tests: Work around systemd regression for boot id parsing

cc https://github.com/systemd/systemd/issues/29275

2 years agorust: Add a `stateroot()` alias in the Rust bindings
Colin Walters [Thu, 9 Nov 2023 19:53:53 +0000 (14:53 -0500)]
rust: Add a `stateroot()` alias in the Rust bindings

Easy to do here, super annoying in C.

2 years agobuild-sys: Enable libsoup3 by default if installed
Colin Walters [Thu, 9 Nov 2023 00:58:27 +0000 (19:58 -0500)]
build-sys: Enable libsoup3 by default if installed

In f39 we switched to libsoup3 by default; this ensures our CI
picks that up automatically so we still have ostree-trivial-httpd.

2 years agoMerge pull request #3089 from cgwalters/fix-undeploy-parsing
Colin Walters [Thu, 9 Nov 2023 00:59:31 +0000 (19:59 -0500)]
Merge pull request #3089 from cgwalters/fix-undeploy-parsing

undeploy: Parse integer more properly

2 years agoundeploy: Parse integer more properly
Colin Walters [Wed, 8 Nov 2023 21:15:01 +0000 (16:15 -0500)]
undeploy: Parse integer more properly

`atoi` doesn't offer any error checking.

Closes: https://github.com/ostreedev/ostree/issues/3088
2 years agoMerge pull request #3080 from nikita-dubrovskii/s390x_secure_boot
Colin Walters [Tue, 7 Nov 2023 21:42:07 +0000 (16:42 -0500)]
Merge pull request #3080 from nikita-dubrovskii/s390x_secure_boot

lib/bootloader-zipl: Check for Secure Boot before zipl

2 years agolib/bootloader-zipl: Check for Secure Boot before zipl
Nikita Dubrovskii [Tue, 17 Oct 2023 14:10:34 +0000 (16:10 +0200)]
lib/bootloader-zipl: Check for Secure Boot before zipl

2 years agoRelease to unstable
Simon McVittie [Mon, 6 Nov 2023 09:45:50 +0000 (09:45 +0000)]
Release to unstable

2 years agoMake fast forward from 2023.7-2
Simon McVittie [Mon, 6 Nov 2023 09:45:08 +0000 (09:45 +0000)]
Make fast forward from 2023.7-2

[dgit --quilt=unapplied]

2 years agoSkip test-admin-deploy-uboot.sh on s390x
Simon McVittie [Thu, 26 Oct 2023 10:26:04 +0000 (11:26 +0100)]
Skip test-admin-deploy-uboot.sh on s390x

It fails on a porterbox. ostree hard-codes zipl to be used on s390x,
so it's reasonable that tests for other bootloaders might not work.

Bug: https://github.com/ostreedev/ostree/issues/3086
Forwarded: no

Gbp-Pq: Topic debian
Gbp-Pq: Name Skip-test-admin-deploy-uboot.sh-on-s390x.patch

2 years agotest-sysroot: Skip on s390x by default
Simon McVittie [Tue, 6 Dec 2022 10:59:33 +0000 (10:59 +0000)]
test-sysroot: Skip on s390x by default

This test regularly fails on the buildds, but I cannot reproduce the
failure on a porterbox.

Bug: https://github.com/ostreedev/ostree/issues/2527
Bug-Debian: https://bugs.debian.org/1025532
Forwarded: not-needed

Gbp-Pq: Topic debian
Gbp-Pq: Name test-sysroot-Skip-on-s390x-by-default.patch

2 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

2 years agobootloader/zipl: No-op if run as non-root
Colin Walters [Tue, 24 Oct 2023 14:40:51 +0000 (10:40 -0400)]
bootloader/zipl: No-op if run as non-root

Not the most elegant fix but should get the job done.

Bug: https://github.com/ostreedev/ostree/issues/3084
Origin: upstream, 2023.8, commit:75c7e51a8bedb5edeadb9f54e84ccfff2adcc755

Gbp-Pq: Name bootloader-zipl-No-op-if-run-as-non-root.patch

2 years agoReally skip test-admin-deploy-uboot.sh on s390x
Simon McVittie [Mon, 6 Nov 2023 09:44:29 +0000 (09:44 +0000)]
Really skip test-admin-deploy-uboot.sh on s390x

2 years agoRemove constraints unnecessary since bullseye (oldstable)
Simon McVittie [Thu, 2 Nov 2023 10:08:53 +0000 (10:08 +0000)]
Remove constraints unnecessary since bullseye (oldstable)

* Build-Depends: Drop versioned constraint on gobject-introspection (>= 1.51.5).
* Build-Depends: Drop versioned constraint on libavahi-client-dev (>= 0.6.31).
* Build-Depends: Drop versioned constraint on libavahi-glib-dev (>= 0.6.31).
* Build-Depends: Drop versioned constraint on libgirepository1.0-dev (>= 1.51.5).
* Build-Depends: Drop versioned constraint on libglib2.0-dev (>= 2.66.0).
* Build-Depends: Drop versioned constraint on libmount-dev (>= 2.23).
* libostree-1-1: Drop conflict with removed package flatpak (<< 0.8.7-2~) in Breaks.
* libostree-dev: Drop versioned constraint on libglib2.0-dev (>= 2.66.0) in Depends.
* Remove 2 maintscript entries from 1 files.

Changes-By: deb-scrub-obsolete
2 years agoRelease to unstable
Simon McVittie [Thu, 26 Oct 2023 10:34:08 +0000 (11:34 +0100)]
Release to unstable

2 years agoMake fast forward from 2023.7-1
Simon McVittie [Thu, 26 Oct 2023 10:33:28 +0000 (11:33 +0100)]
Make fast forward from 2023.7-1

[dgit --quilt=unapplied]

2 years agoSkip test-admin-deploy-uboot.sh on s390x
Simon McVittie [Thu, 26 Oct 2023 10:26:04 +0000 (11:26 +0100)]
Skip test-admin-deploy-uboot.sh on s390x

It fails on a porterbox. ostree hard-codes zipl to be used on s390x,
so it's reasonable that tests for other bootloaders might not work.

Bug: https://github.com/ostreedev/ostree/issues/3086
Forwarded: no

Gbp-Pq: Topic debian
Gbp-Pq: Name Skip-test-admin-deploy-uboot.sh-on-s390x.patch

2 years agotest-sysroot: Skip on s390x by default
Simon McVittie [Tue, 6 Dec 2022 10:59:33 +0000 (10:59 +0000)]
test-sysroot: Skip on s390x by default

This test regularly fails on the buildds, but I cannot reproduce the
failure on a porterbox.

Bug: https://github.com/ostreedev/ostree/issues/2527
Bug-Debian: https://bugs.debian.org/1025532
Forwarded: not-needed

Gbp-Pq: Topic debian
Gbp-Pq: Name test-sysroot-Skip-on-s390x-by-default.patch

2 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

2 years agobootloader/zipl: No-op if run as non-root
Colin Walters [Tue, 24 Oct 2023 14:40:51 +0000 (10:40 -0400)]
bootloader/zipl: No-op if run as non-root

Not the most elegant fix but should get the job done.

Bug: https://github.com/ostreedev/ostree/issues/3084
Origin: upstream, 2023.8, commit:75c7e51a8bedb5edeadb9f54e84ccfff2adcc755

Gbp-Pq: Name bootloader-zipl-No-op-if-run-as-non-root.patch

2 years agoAdd patch to avoid another test failure on s390x
Simon McVittie [Thu, 26 Oct 2023 10:26:25 +0000 (11:26 +0100)]
Add patch to avoid another test failure on s390x

2 years agolibostree-1-dev Provides gir1.2-ostree-1.0-dev
Simon McVittie [Thu, 26 Oct 2023 10:13:09 +0000 (11:13 +0100)]
libostree-1-dev Provides gir1.2-ostree-1.0-dev

This is likely to be useful in the migration path to being able to turn
off GObject-Introspection bindings when bootstrapping or cross-compiling.

2 years agoAdd patch from upstream to avoid build-time test failure on s390x
Simon McVittie [Thu, 26 Oct 2023 10:07:31 +0000 (11:07 +0100)]
Add patch from upstream to avoid build-time test failure on s390x

2 years agoMerge pull request #3085 from cgwalters/s390x-units
Jonathan Lebon [Tue, 24 Oct 2023 20:40:12 +0000 (16:40 -0400)]
Merge pull request #3085 from cgwalters/s390x-units

2 years agobootloader/zipl: No-op if run as non-root
Colin Walters [Tue, 24 Oct 2023 14:40:51 +0000 (10:40 -0400)]
bootloader/zipl: No-op if run as non-root

Closes: https://github.com/ostreedev/ostree/issues/3084
Not the most elegant fix but should get the job done.

2 years agoRelease to unstable
Simon McVittie [Mon, 23 Oct 2023 09:37:01 +0000 (10:37 +0100)]
Release to unstable

2 years agoMake fast forward from 2023.6-2
Simon McVittie [Mon, 23 Oct 2023 09:36:48 +0000 (10:36 +0100)]
Make fast forward from 2023.6-2

[dgit --quilt=unapplied]

2 years agotest-sysroot: Skip on s390x by default
Simon McVittie [Tue, 6 Dec 2022 10:59:33 +0000 (10:59 +0000)]
test-sysroot: Skip on s390x by default

This test regularly fails on the buildds, but I cannot reproduce the
failure on a porterbox.

Bug: https://github.com/ostreedev/ostree/issues/2527
Bug-Debian: https://bugs.debian.org/1025532
Forwarded: not-needed

Gbp-Pq: Topic debian
Gbp-Pq: Name test-sysroot-Skip-on-s390x-by-default.patch

2 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

2 years agoDepend on pkgconf in preference to pkg-config
Simon McVittie [Mon, 23 Oct 2023 09:36:27 +0000 (10:36 +0100)]
Depend on pkgconf in preference to pkg-config

2 years agoInstall systemd system units into /usr/lib/systemd/system
Simon McVittie [Mon, 23 Oct 2023 09:35:10 +0000 (10:35 +0100)]
Install systemd system units into /usr/lib/systemd/system

This was allowed by TC resolution #1053901. Build-depend on debhelper
13.11.6~ to ensure that the units are still picked up by
dh_installsystemd.

2 years agoNew upstream release
Simon McVittie [Mon, 23 Oct 2023 09:33:48 +0000 (10:33 +0100)]
New upstream release

2 years agoUpdate upstream source from tag 'upstream/2023.7'
Simon McVittie [Mon, 23 Oct 2023 09:23:34 +0000 (10:23 +0100)]
Update upstream source from tag 'upstream/2023.7'

Update to upstream version '2023.7'
with Debian dir 87efd50332ba01b4f3d82cb59fd1c034293ace78

2 years agoNew upstream version 2023.7
Simon McVittie [Mon, 23 Oct 2023 09:23:27 +0000 (10:23 +0100)]
New upstream version 2023.7

2 years agoMerge pull request #3083 from cgwalters/release
Colin Walters [Fri, 20 Oct 2023 14:56:55 +0000 (10:56 -0400)]
Merge pull request #3083 from cgwalters/release

Release 2023.7

2 years agoRelease 2023.7
Colin Walters [Fri, 20 Oct 2023 13:24:56 +0000 (09:24 -0400)]
Release 2023.7

2 years agoconfigure: post-release version bump
Colin Walters [Fri, 20 Oct 2023 13:26:09 +0000 (09:26 -0400)]
configure: post-release version bump

2 years agoMerge pull request #3081 from jlebon/pr/log-selinux-refresh
Colin Walters [Fri, 20 Oct 2023 13:36:10 +0000 (09:36 -0400)]
Merge pull request #3081 from jlebon/pr/log-selinux-refresh

lib/deploy: Log SELinux policy refresh

2 years agolib/deploy: Log SELinux policy refresh
Jonathan Lebon [Thu, 19 Oct 2023 15:35:25 +0000 (11:35 -0400)]
lib/deploy: Log SELinux policy refresh

I have a suspicion that the `semodule -N --refresh` we do here is
involved in https://github.com/coreos/fedora-coreos-tracker/issues/1597.
Let's log when we execute it and include its time.

While we're here, also log the time it takes to `syncfs()` and the
fsfreeze/thaw cycling. It's logged in a structured journal entry, but
that's less accessible than just having it in the journal message field.