b56786c415
Merge pull request #2496 from lucab/ups/repo-asserts
...
lib/repo: fix problematic invariant checks
2021-12-07 19:58:23 -05:00
490f515e18
Merge pull request #2494 from jmarrero/fsf-address
...
Update FSF license notices to use URL instead of address
2021-12-07 15:51:48 -05:00
58bdfb90c9
lib/repo: do no return a NULL on failure
...
This turns an existing check into an assert. The previously returned
NULL may result in confusing callers, as none of them is checking for
that.
2021-12-07 16:24:56 +00:00
4a0ebe507f
lib/repo: do no return an arbitrary mode on failure
...
This turns the existing check into an assert. Otherwise, the previous
code may return an arbitrary repo mode (bare) on failure.
2021-12-07 16:23:02 +00:00
f1e24945fa
lib/repo: assert that writable state and error agree
...
This adds an assertion to check that writable stable and error
are in sync. The subsequent logic uses them interchangeably.
2021-12-07 16:21:21 +00:00
581a58067b
Update FSF license notices to use URL instead of address
2021-12-07 08:34:25 -05:00
a07b8d6c00
Merge pull request #2493 from cgwalters/summary-lock-shared
...
repo: Change locking for summary regeneration to be shared
2021-12-06 08:19:12 -05:00
2c39bd88a9
repo: Change locking for summary regeneration to be shared
...
This is trying to address:
https://pagure.io/fedora-iot/issue/48
Basically we changed rpm-ostree to start doing a shared lock during
commit by default, but this broke because pungi is starting a process
doing a commit for each architecture, and then trying to regenerate
the summary after each one.
This patch is deleting a big comment with a rationale for why
summary regeneration should be exclusive. Point by point:
> This makes sure the commits and deltas don't get
> deleted while generating the summary.
But prune operations require an exclusive lock, which means that
data still can't be deleted when the summary grabs a shared lock.
> It also means we can be sure refs
> won't be created/updated/deleted during the operation, without having to
> add exclusive locks to those operations which would prevent concurrent
> commits from working.
First: The status quo *has* prevented concurrent commits from working!
There is no real locking solution to this problem. What we really
need to do here is regenerate the summary after each commit *or*
when the caller decides to do it and e.g. include deltas at the same
time.
It's OK if multiple threads race to regenerate the summary;
last-one-wins behavior here is totally fine.
2021-12-03 14:42:03 -05:00
267ca93da2
Merge pull request #2490 from lucab/ups/static-fixes
2021-11-26 10:55:15 -05:00
cefc1d7854
lib: misc static analysis fixes
...
This fixes a few warnings from coverity, none of which really
interesting.
2021-11-26 15:11:00 +00:00
2dc6ec70ec
Merge pull request #2489 from ostreedev/release-2021.6
...
Release 2021.6
2021-11-23 09:01:57 -05:00
281d310224
configure: post-release version bump
2021-11-23 10:15:19 +00:00
f1155c8d28
Release 2021.6
2021-11-23 10:13:32 +00:00
48dfff2754
Merge pull request #2486 from jlebon/pr/remount-ostree
...
app: Only remount /sysroot if needed
2021-11-22 13:35:28 -05:00
947acbf178
app: Only remount /sysroot if needed
...
We should only try to remount `/sysroot` if we're actually handling the
sysroot repo and the repo isn't writable. We already have public APIs to
check each of those, so let's use them.
Closes : #2485
2021-11-19 11:01:18 -05:00
f552e30c0b
Merge pull request #2484 from dbnicholson/prune-serialize
...
lib/prune: Avoid unnecessary object serialization
2021-11-18 14:49:15 -05:00
47d32d9ead
lib/prune: Avoid unnecessary object serialization
...
`repo_prune_internal` was deserializing each object and passing the
components to `maybe_prune_loose_object`, which promptly reserialized
it.
2021-11-18 11:04:06 -07:00
edbcf52134
Merge pull request #2481 from refi64/null-error
...
lib: Avoid dereferencing NULL error values
2021-11-12 15:19:29 +00:00
9c1fe55bbc
lib: Avoid dereferencing NULL error values
...
Otherwise, this will segfault when callers don't need any exact errors.
Signed-off-by: Ryan Gonzalez <ryan.gonzalez@collabora.com>
2021-11-12 08:29:05 -06:00
e39280ff21
Merge pull request #2479 from cgwalters/ci-capsh
...
ci: Require `libcap2-bin` for `capsh`
2021-11-09 10:44:08 -05:00
dff36697cd
Merge pull request #2477 from ratajs/patch-1
...
Add Fedora Kinoite link
2021-11-09 09:32:02 -05:00
1a86517d65
Merge pull request #2476 from valentindavid/valentindavid/glib-fix
...
lib: Fix a bad call to g_file_get_child
2021-11-09 09:27:18 -05:00
3b338df956
ci: Require `libcap2-bin` for `capsh`
...
This was previously pulled in indirectly, but it looks like we need
to require it explicitly in newer Ubuntu.
2021-11-09 09:22:43 -05:00
56428d5e0e
Merge branch 'ostreedev:main' into patch-1
2021-11-04 18:50:34 +01:00
e580eca214
Merge pull request #2187 from cgwalters/sysroot-ro-initramfs
...
prepare-root: Set up sysroot readonly in initramfs
2021-11-04 10:37:18 -04:00
adc097a2ed
lib: Fix a bad call to g_file_get_child
...
In Glib, since commit 3a6e8bc8876e149c36b6b14c6a25a718edb581ed,
`g_file_get_child` does not accept absolute path as paramater anymore.
The broken assertion was encountered during `ostree admin deploy`
command for the checkout of subpath `etc`.
Example of error log:
```
(ostree admin deploy:1640): GLib-GIO-CRITICAL **: 03:42:00.570: g_file_get_child: assertion '!g_path_is_absolute (name)' failed
(ostree admin deploy:1640): GLib-GIO-CRITICAL **: 03:42:00.570: g_file_query_info: assertion 'G_IS_FILE (file)' failed
**
OSTree:ERROR:src/ostree/ot-main.c:232:ostree_run: assertion failed: (success || error)
Bail out! OSTree:ERROR:src/ostree/ot-main.c:232:ostree_run: assertion failed: (success || error)
```
2021-11-03 21:40:35 +01:00
48f582b918
Added Fedora Kinoite link
2021-11-03 21:02:53 +01:00
c553b5c69a
prepare-root: Set up sysroot readonly in initramfs
...
Let's ensure things are right from the start in the initramfs;
this closes off various race conditions. Followup to
3564225917https://github.com/ostreedev/ostree/issues/2115
2021-11-03 16:37:20 +00:00
800289a8db
Merge pull request #2475 from lucab/ups/prepare-root-check-first
...
prepare-root: check for read-only sysroot status early on
2021-11-01 09:58:06 -04:00
63d0c4c781
prepare-root: check for read-only sysroot status early on
...
This moves read-only sysroot checks upfront, so that they are not
intermixed with mount operations.
It has no immediate side-effects, but allow these check to be
independent from the rest of the mounting logic (and future changes
to it).
2021-11-01 09:42:36 +00:00
b7efd16cc5
Merge pull request #2472 from lucab/ups/prepare-root-checked-printf
2021-10-26 15:14:21 -04:00
6e0165020a
Merge pull request #2473 from lucab/ups/prepare-root-less-global-mutable-state
2021-10-26 13:25:37 -04:00
7c17daad17
prepare-root: get rid of a global variable
...
This moves a global mutable variable to a smaller local scope,
as it is not really used outside of that.
2021-10-26 16:27:22 +00:00
ca84da679a
prepare-root: check return codes for errors when assembling paths
...
This adds checks around all `snprintf` calls in order to detect
failures and gracefully abort.
2021-10-26 16:16:34 +00:00
1e6077af03
Merge pull request #2471 from lucab/ups/prepare-root-silent-mounts
...
prepare-root: make all mount operations silent
2021-10-26 10:06:52 -04:00
848fe542af
prepare-root: make all mount operations silent
...
This adds a `MS_SILENT` flag to all `mount(2)` calls, reducing the
amount of kernel logs produced on each boot.
Those messages do not contain actionable details, and in the "mount
plus read-only remount" case they can easily become highly redundant.
2021-10-26 10:51:37 +00:00
9692437491
Merge pull request #2468 from lucab/ups/tests-var-mount
...
tests/var-mount: tweak test setup
2021-10-19 09:03:19 +00:00
f93d96620c
tests/var-mount: tweak test setup
...
This reworks the var-mount destructive test in order to properly use
the datadir for the current stateroot instead of a duplicated one.
In turn, it ensures that the resulting `var.mount` after reboot is
correctly pointing to the same location which hosted `/var` on the
previous boot.
2021-10-19 08:03:09 +00:00
5d045f3ae1
Merge pull request #2466 from cgwalters/ci-fanalyzer
...
ci: Enable -fanalyzer
2021-10-18 16:44:56 +00:00
58dc6a08b4
tests/rollsum: Use `g_malloc` not `malloc`
...
To pacify gcc's `-fanalyzer`.
2021-10-18 11:44:44 -04:00
fda41e8d24
ci: Enable -fanalyzer
...
Followup to https://github.com/ostreedev/ostree/pull/2463
One thing I noticed here is we lost usage of `build-check.sh`
which also invokes `clang`, which doesn't speak `-fanalyzer`
and would be broken by this if we try to enable `build-check.sh`
again. But that can come later.
2021-10-15 15:44:04 -04:00
ec8cf31e98
Merge pull request #2455 from cgwalters/packit
...
Attempt to update packit flow to build in COPR
2021-10-15 07:47:28 +00:00
09f4ad708c
Merge pull request #2467 from cgwalters/bump-libglnx-20211014
2021-10-14 15:33:33 -04:00
baa57ffe0d
libglnx: Bump to ef502aabf7d3a0d37f9c4d228f870ac93404447b
...
Various fixes there, including one for `gcc -fanalyzer`.
Update submodule: libglnx
2021-10-14 12:41:02 -04:00
7ba8dbf0cc
Attempt to update packit flow to build in COPR
...
No idea if this will really work, but at least `packit srpm`
does work now.
2021-10-14 12:36:38 -04:00
6b8295ada1
Merge pull request #2463 from cgwalters/fix-fanalyzer
...
Fix various mostly theoretical gcc `-fanalyzer` issues
2021-10-14 08:30:25 +00:00
029a9d56c3
variantutil: Fix gcc `-fanalyzer` warnin
...
Add some not-NULL assertions for return values from glib,
and upgrade some `g_return_if_fail` to `g_assert`.
2021-10-13 17:13:14 -04:00
54bf42c3e5
utils: Fix unreachable `NULL` deref by adding assertion
...
Again this one is just in theory, but let's add an assertion.
2021-10-13 17:13:14 -04:00
f355482e1f
static-delta: Fix probably not actually possible NULL deref
...
Flagged by `gcc -fanalyzer`. I didn't study this really deeply
but I think it's not actually reachable. Anyways, let's catch
it on general principle.
2021-10-13 17:13:14 -04:00
3159e04980
fetcher/soup: Fix gcc `-fanalyzer` warning
...
In general, we're probably going to need to change most of our
`g_return_if_fail` to `g_assert`. The analyzer flags that
the function can return `NULL`, but the caller isn't prepared for
this.
In practice, let's abort.
2021-10-13 17:13:14 -04:00
Colin Walters
Luca BRUNO
Joseph Marrero
Jonathan Lebon
Dan Nicholson
Ryan Gonzalez
Šimon (Simon) Rataj
Valentin David