Philip Withnall [Mon, 19 Jun 2017 14:26:50 +0000 (15:26 +0100)]
lib/pull: Don’t cache summary file until its signature is verified
This makes no difference to the validity of the code, since any summary
file loaded from the cache will be verified before being read anyway;
but it will make some upcoming changes a little simpler.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #961
Approved by: cgwalters
Philip Withnall [Fri, 23 Jun 2017 20:44:29 +0000 (21:44 +0100)]
tests: Fix incorrect `summary --update` usage in test-local-pull.sh
Previously the spurious ‘update’ atom was ignored; now that `ostree
summary --update` accepts extra arguments as additional metadata, it’s
causing an error. Drop it.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #961
Approved by: cgwalters
Colin Walters [Mon, 26 Jun 2017 19:10:57 +0000 (15:10 -0400)]
pull: Check free space when pulling deltas
Computing download/storage size for `archive` pulls is hard; there's
`OSTREE_REPO_COMMIT_MODIFIER_FLAGS_GENERATE_SIZES` which was from a
pre-static-deltas effort by Endless, but we aren't currently making use of this
much.
Static deltas were designed to solve this problem; we have the total
uncompressed size. Let's check free space before doing a delta pull.
Colin Walters [Sat, 24 Jun 2017 14:28:09 +0000 (10:28 -0400)]
lib: Use OtTmpFile for static delta processing
The `OstreeRepoContentBareCommit` struct was basically an `OtTmpFile`, so let's
make it one. I moved the "convert to `GOutputStream`" logic into the callers,
since that bit can't fail; it makes the implementation much simpler since we can
just return the result of `ot_open_tmpfile_linkable_at()`.
Colin Walters [Sat, 24 Jun 2017 13:30:02 +0000 (09:30 -0400)]
lib: Hoist unlinkat() cleanup API to fsutil, use in pull
The pull code also could make use of this in both the metadata and content
paths. I changed it to own the tempfile malloc (just like `GLnxTmpFile`), since
there's no reason to have different lifetimes for the filename and the file, and
that way we only have one variable rather than two.
The content path turns out to be a special case though, where
at least for mirroring archives, we directly pass the file *path*
down into `_ostree_repo_commit_loose_final()`.
Colin Walters [Fri, 23 Jun 2017 21:13:15 +0000 (17:13 -0400)]
lib/commit: Clean up commit file type handling variables
The variables here were duplicative; we don't need two booleans to distinguish
between symlinks and regular files. What we do need to handle is the "physical"
state versus the "object" state. Symlinks objects are stored as regular files in
`bare-user` and `archive`.
Philip Withnall [Fri, 23 Jun 2017 14:55:19 +0000 (15:55 +0100)]
lib/refs: Add runtime error checking for collection ID validity
Instead of treating it as a programming error — given that it’s user
input, that’s not really appropriate. This modifies write_ref() and
list_collection_refs() to implement validation.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 13 Jun 2017 17:55:53 +0000 (18:55 +0100)]
tests: Add integration tests for collections
Test various pieces of core and command line utility functionality
relating to the newly-introduced concept of collections. Mostly focussed
around the find-remotes utility, and around handling of collection–refs.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Fri, 9 Jun 2017 17:45:39 +0000 (18:45 +0100)]
ostree/builtins: Add support for collection–refs to a few utilities
These utilities were not needed for the initial port to support
OstreeCollectionRef, so have been delayed a bit and, in some cases, left
as FIXME comments for follow up later.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 13:53:58 +0000 (14:53 +0100)]
ostree/dump: Include collection IDs and mirrored refs in summary dumps
If a repository’s summary file includes a collection ID, output that. If
it includes refs from other collections (in the ‘collection map’), output
those and include the same metadata detail as for refs in the summary
file’s main refs map.
If collection IDs are specified in the summary file, this changes the
output format from `ostree summary -v` to use (collection ID, ref name)
tuples.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 13:51:58 +0000 (14:51 +0100)]
refs: Add a --collections argument to the built-in refs command
Rather than change the output format used by the existing refs command
to output collection IDs in addition to ref names, this functionality has
been hidden behind an --collections argument. If it’s not specified `ostree
refs` will output the same content as before for a given repository. If
it is specified, the collection ID for each ref will be included in the
output as (collection ID, ref name).
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 13:50:51 +0000 (14:50 +0100)]
remote-add: Add a --collection-id argument to the built-in add command
This allows newly configured remotes to have their collection ID specified,
so that refs from them can be downloaded from peers as well as the
upstream collection, using the remote’s configuration.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Thu, 4 May 2017 09:47:23 +0000 (10:47 +0100)]
find-remotes: Add pull support to the find-remotes built-in command
This will pull the remotes after finding them. This potentially needs to
go in its own pull-from-remotes built-in command, but it will be fine
here for now.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 25 Apr 2017 18:10:04 +0000 (19:10 +0100)]
find-remotes: Add a find-remotes built-in command
This is a wrapper around the new ostree_repo_find_remotes() method; it
tries to find available remotes which can serve updates for the
user-provided refs.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 18 Apr 2017 23:13:28 +0000 (00:13 +0100)]
lib/repo-finder: Add Avahi based OstreeRepoFinder implementation
This is a more complex implementation of OstreeRepoFinder which resolves
ref names to remote URIs by looking for refs advertised by peers on the
local network using DNS-SD records and mDNS (Avahi). The idea is to
allow OS and app updates to be propagated over local networks, without
the internet.
It requires an OSTree server and code to generate the DNS-SD adverts in
order to be fully functional — support for this will be added
separately.
Unit tests are included.
Includes fixes by Krzesimir Nowak <krzesimir@kinvolk.io>.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 18 Apr 2017 23:11:28 +0000 (00:11 +0100)]
lib/bloom: Add an internal bloom filter implementation
This will be used in an upcoming commit. It adds a basic bloom filter
implementation, using the SipHash family of hash functions.
The implementation (including its parameter choices and hash functions)
will become a protocol detail in future, so must not be changed so that
its output is bitwise incompatible between OSTree versions.
Unit tests are included.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 18 Apr 2017 23:07:51 +0000 (00:07 +0100)]
lib/repo-finder: Add mount based OstreeRepoFinder implementation
This is a basic implementation of OstreeRepoFinder which resolves ref
names to remote URIs by looking for them on any currently mounted
removable storage volumes. The idea is to support OS and app updates via
USB stick.
Unit tests are included.
This bumps libostree’s maximum GLib dependency from 2.44 to 2.50 for
g_drive_is_removable(). If GLib 2.50 is not available, the call which
needs it will be omitted and the OstreeRepoFinderMount implementation
will scan all volumes (not just removable ones); this is a performance
hit, but not a functionality hit.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 18 Apr 2017 23:05:06 +0000 (00:05 +0100)]
lib/repo-finder: Add config-file based OstreeRepoFinder implementation
This is a basic implementation of OstreeRepoFinder which resolves ref
names to remote URIs by looking their collection IDs up in the local
configuration of remotes who have their collection-id key set.
Unit tests are included.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Tue, 18 Apr 2017 22:59:33 +0000 (23:59 +0100)]
lib/repo-finder: Add basic support for finding remote URIs by ref name
Add an initial OstreeRepoFinder interface (but no implementations),
which will find remote URIs by ref names and collection IDs, the
combination of which is globally unique.
The new API is used in a new ostree_repo_find_updates() function, which
resolves a list of ref names to update into a set of remote URIs to pull
them from, which can be treated as mirrors. It is an attempt to
generalise resolution of the URIs to pull from, and to generalise
determination of the order and parallelisation which they should be
downloaded from in.
Includes fixes by Krzesimir Nowak <krzesimir@kinvolk.io>.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 13:42:22 +0000 (14:42 +0100)]
lib/pull: Add collection support to ostree_repo_pull_with_options()
This adds a new collection-refs option of type a(sss), giving a list of
(collection ID, ref name, checksum) tuples to pull from the given remote.
This option is intended to supersede the refs and override-commit-ids
options, so is mutually exclusive with them.
This includes support for resolving the refs from the remote’s summary
file, or from its refs/heads and refs/mirrors directories.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 13:25:59 +0000 (14:25 +0100)]
lib/refs: Add methods for setting/listing collection–refs
These are tuples of (collection ID, ref name) which are a globally-unique
form of local ref. They use OstreeCollectionRef as an identifier, and hence
need to be accessed using new API, as the existing API uses string
identifiers and sometimes accepts refspecs. Remote names are not
supported as part an OstreeCollectionRef.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 11:38:59 +0000 (12:38 +0100)]
lib/repo: Add collection ID support to OstreeRepo
Add {get,set}_collection_id() methods to OstreeRepo and some documentation
about the concept of a collection ID which globally identifies an
upstream repository. See the documentation for more details.
This will be used in future commits. For now, the new API is marked as
experimental (--enable-experimental-api).
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 14:27:10 +0000 (15:27 +0100)]
lib/ref: Add OstreeCollectionRef type for globally unique refs
This is a type representing the tuple (collection ID, ref name), which is
guaranteed to be globally unique. It will be used in upcoming commits.
It introduces the concept of a ‘collection’ which is a unique, curated
set of refs which lie in the same trust domain (i.e. all signed by the
same key and validated by the same developer). Flathub might be a
collection, for example; or the set of OS refs coming from a particular
OS vendor.
It includes a function for validating collection IDs.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #924
Approved by: cgwalters
Colin Walters [Fri, 23 Jun 2017 16:14:07 +0000 (12:14 -0400)]
lib/repo: More cleanup of load_file() internals
This is followon work from previous cleanups. Basically
`stat_bare_content_object()` was the `fstatat()` logic
and `ostree_repo_read_bare_fd()` was the `openat()` implementation;
they duplicated some bits to find the object in staging, recurse
into parent etc.
Further, I wanted an internal-only version of this API which didn't allocate
`GFileInfo`/`GInputStream` but used a plain `fd` and `struct stat` to avoid
mallocs.
The end version here I think looks a lot nicer, since we deduplicate the various
`open()` calls in the different cases for example.
Philip Withnall [Wed, 21 Jun 2017 11:53:48 +0000 (12:53 +0100)]
lib/sym: Fix symbol versions for 2017.7 experimental symbols
There was a typo in the group name. It should be OK to change the
version since this is all hidden behind the --enable-experimental-api
configure option.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #950
Approved by: cgwalters
Colin Walters [Tue, 20 Jun 2017 21:19:01 +0000 (17:19 -0400)]
ci: Enable -Werror=maybe-uninitialized
We don't want to inject this warning by default for every build
like the other ones in `configure.ac`, since it can be spruriously
wrong. But there's no reason not to have a more extended set of
warnings for well-known toolchains (e.g. f25).
Colin Walters [Mon, 19 Jun 2017 14:23:21 +0000 (10:23 -0400)]
build-sys: Add "release build" flag, use for symbol versioning
I was trying to do a release and move the symbols from `-devel.sym` into
`-release.sym`, but it turns out that at least GNU binutils `ld` treats an empty
version script as a syntax error. Fix this by adding a "release build"
flag, and only include `-devel` in non-release builds.
This would also make it easier to inject that flag into our `.pc` and
`ostree-version.h` and `ostree --version` metadata, but I didn't do that yet.
EDIT: Turns out a simpler fix is just to add an empty section. However I kept
this commit since it's a useful sanity check for whether we should include
`-devel.sym` in builds, and we may want to inject the metadata later.
David Shea [Sat, 17 Jun 2017 16:08:41 +0000 (12:08 -0400)]
lib/repo: Fix annotations for out parameters
Change the annotation of the out parameters on ostree_repo_load_file
from `(allow-none)` to `(optional) (nullable)`. `allow-none` is
ambiguous, since these parameters can be both NULL on input and set to
NULL on return.
Flatpak make check is failing when applying a static delta
to a bare-user-only repo due to an assert. The fix is to add
bare-user-only to the assert check.
pull: When mirroring, only replace summary if we're doing a full mirror
We're hitting this in flathub, where we have a bunch of local builds,
but we also mirror a few refs from the gnome runtime repo into it.
Its fixable by re-doing the summary, but for a short time the
wrong version is visible.
ostreee-version.h.in: Added Since: version annotations
This is especially interesting for the versioning symbols themselves,
as it is an indicator of when applications using introspection information
can start to use a symbol in the library to check if they have a
recent enough version of OSTree to use.
Colin Walters [Wed, 14 Jun 2017 21:55:11 +0000 (17:55 -0400)]
lib/pull: Extend BAREUSERONLY_FILES flag to HTTP requests
For the flatpak PR: https://github.com/flatpak/flatpak/pull/849
It's really more convenient if this works for HTTP pulls as well, since flatpak
does various types of pulling, and we can just set the flag everywhere.
Further, we might as well reject the content as early as possible.
Anton Gerasimov [Thu, 1 Jun 2017 10:43:50 +0000 (12:43 +0200)]
lib/sysroot: Add API to get pending/rollback for given stateroot
This imports a function that is used in rpm-ostree, and it's also intended for
use by https://github.com/advancedtelematic/aktualizr to display
what deployment we're going to boot next after the reboot.
Updated-by: Colin Walters <walters@verbum.org> Closes: #897
Approved by: OYTIS
This option suppresses mode bits outside of `0775` for directory
checkouts. I think most people should start doing this by default,
and use explicit overrides for e.g. `/tmp` if doing a recommit based
on a checkout.
This is an option which is intended mostly for flatpak;
see: https://github.com/flatpak/flatpak/issues/845
We're adding an option for pulling into *all*
repo modes that has an effect similar to the `bare-user-only`
change from https://github.com/ostreedev/ostree/pull/909
This way one can pull content into e.g. a root-owned `bare` repository and
ensure that there aren't any setuid or world-writable files.
Colin Walters [Mon, 12 Jun 2017 17:59:33 +0000 (13:59 -0400)]
lib/repo: Support hardlink conversions from bare-user to bu-only
Thinking about the problem of flatpak converting from `bare-user` to `bare-user-only`
"in place" by creating a new repo and doing a `pull-local`, I realized
that we can optimize this process by doing hardlinks for both metadata
and regular files. The repo formats are *almost* compatible, the
exception being symlinks.
An earlier patch caused us to do hardlinks for metadata, this patch takes things
to the next step and special cases this specific conversion. In this case we
need to parse the source object to determine whether or not it's a symlink.
Colin Walters [Mon, 12 Jun 2017 17:38:52 +0000 (13:38 -0400)]
lib/repo: Import metadata via hardlink even for distinct repo modes
Our previous logic for import-via-hardlink only tried if the repo modes match,
but we *can* hardlink metadata between e.g. `archive` and `bare-user` repos, and
that's quite useful thing to do. Our documentation encourages converting to/from
those repo modes locally for build systems.
Colin Walters [Mon, 12 Jun 2017 17:20:42 +0000 (13:20 -0400)]
lib/repo: Skip import via hardlink if repo owners don't match
Before this, if one had repos of matching mode but different owners,
which could happen if one e.g. makes a `bare` non-root repo in
`/ostree/deploy/$stateroot/var/tmp`, every time we tried to call `linkat()`
we'd get `EPERM` and fall back to a copy.
Fix this by saving the repo owner uid, and avoid trying to call `linkat()` if we
know it's going to fail. Of course most commonly in this scenario we'll
immediately fail trying to `chown` the files to `0`, but this is prep for a
future patch to improve `bare-user` → `bare-user-only` imports where we'll be a
bit more sophisticated.
Philip Withnall [Wed, 7 Jun 2017 13:29:55 +0000 (14:29 +0100)]
lib/pull: Fix construction of a refspec to use the correct separator
This code looks like it was supposed to build a refspec, but it used a
slash as a separator rather than a colon. The following code does
recover by supporting prefix matching with slashes, but it seems like
this was perhaps not the intention.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #912
Approved by: cgwalters
lib/repo: Always look in staging directory for objects
Its often the case that we want to look at objects inside a commit,
before the objects the transaction is finished. For instance:
https://github.com/flatpak/flatpak/pull/837
Which tries to verify the file permissions before committing the
transaction.
And:
https://github.com/flatpak/flatpak/commit/1e5ffa926a25acb655af7889b679b140bf44870b
Which collects the storage size of the objects so that we can
put the total download size in the commit metadata.
I tried to find all the places where we did reads from the
object directories, and in particular this fixes:
- `ostree_repo_load_file()` for `bare` repos (`archive` was already working).
- `ostree_repo_query_object_storage_size()`
- Applying deltas that reference not-yet-commited objects
Colin Walters [Mon, 12 Jun 2017 19:36:16 +0000 (15:36 -0400)]
lib/repo: Refactor object copy import function
This came up in: https://github.com/ostreedev/ostree/pull/881
Basically doing streaming for metadata is dumb. Split up the metadata/content
paths so we pass metadata around as `GVariant`. This drops the last internal
caller of `ostree_repo_write_metadata_stream_trusted()` which was the dumb
function mentioned.
Colin Walters [Wed, 7 Jun 2017 18:21:59 +0000 (14:21 -0400)]
lib/checkout: Ignore world-writable dirs for bare-user-only checkout
See https://github.com/ostreedev/ostree/pull/909 for more information on the
rationale. Basically there's no reason for flatpak (which uses `bare-user-only`)
to have world-writable dirs. Particularly with the presence of the system
helper.
An approach I considered instead was to parse and validate directory metadata
objects at commit time. We still may do that in addition; for file objects we *had*
to do it that way because the actual files would be laid down suid. But directories
live only as inert `.dirmeta` objects until we do a checkout (i.e. `mkdir()`), so
we can solve the problem at checkout time.
Colin Walters [Wed, 7 Jun 2017 17:56:47 +0000 (13:56 -0400)]
lib/commit: Drop some conditionals/clarify code in content path
Both callers of `commit_loose_object_trusted()` were passing
`OSTREE_OBJECT_TYPE_FILE`, so drop that parameter. This in turn
allows us to drop lots of checking of that inside the function.
Add a doc comment, and rename to `commit_loose_content_object()` for clarity.
Colin Walters [Wed, 7 Jun 2017 17:42:15 +0000 (13:42 -0400)]
lib/commit: Port final object writing function to new code style
I noticed my previous patches incorrectly started doing `return glnx_throw*`
inside a `goto out;` function. Fix this by porting forward consistently to new
style. We just do the error prefixing in the caller.
Colin Walters [Thu, 8 Jun 2017 17:47:20 +0000 (13:47 -0400)]
ci: Update to match current rpm-ostree
This copies the `ci/` directory from rpm-ostree, with much the same rationale;
among other things we don't want to depend on the Docker hub.
The specific reason I'm doing this is that I want to add a CentOS7 build, but
that means we can't use `projectatomic/ostree-tester`, and at that point we
might as well unwind it all.
repo: After renaming in all loose objects, ensure metadata is stable
When a transaction is finished and we have moved all the staged loose
objects into the repo we fsync all the object directory, to ensure the
filenames are stable before we update the refs files to point to the
new commits.
With out this an unclean shutdown after the transaction is finished
could result in a refs file that points to an incomplete commit.
Colin Walters [Wed, 7 Jun 2017 17:18:04 +0000 (13:18 -0400)]
repo/commit: Support group-writable files for bare-user-only
These exist in the wild for flatpak, and aren't really a problem. The canonical
permissions are still either `0755` or `0644`, we just support the additional
writable bit for the group (i.e. extend the set to include `0775` and `0664`)
now to avoid breaking some flatpak content.
Colin Walters [Mon, 5 Jun 2017 15:32:52 +0000 (11:32 -0400)]
lib/repo: For bare-user, mask content object modes with 0775
Having every object in a bare-user repo (and checkouts) be executable
is ugly. I can't think of a good reason to do that; they should only
be executable if their input is. This does
for `bare-user` what we did for `bare-user-only` in
https://github.com/ostreedev/ostree/pull/909
It's also a stronger version of what we do with `checkout -U` in suppressing
suid - here we also strip world-writable files and the sticky bit (even though
that's meaningless today, it might not be in the future).
Philip Withnall [Wed, 7 Jun 2017 14:41:17 +0000 (15:41 +0100)]
lib/fetcher: Add cleanup function for OstreeFetcher
This is only used internally (the header is not public), so it doesn’t
have to go in ostree-autocleanups.h. It will be used in some following
commits.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #911
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 13:32:29 +0000 (14:32 +0100)]
lib/repo: Omit deltas from the summary file if there are none
If there are no deltas to be listed in the summary file, don’t bother
including the key for them in the additional metadata section of the
file. This saves a few bytes in some cases.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #911
Approved by: cgwalters
Philip Withnall [Wed, 7 Jun 2017 11:03:57 +0000 (12:03 +0100)]
ostree/dump: Improve formatting for well-known commit metadata keys
This follows on from commit a946c3d4, which added formatting for
well-known summary metadata keys. This commit adds it for commits.
Currently, the only well-known commit metadata key is
ostree.commit.timestamp. Formatting this correctly is especially
important, since it’s a big-endian uint64, which is completely unusable
for mere mortals when presented as a number rather than a date.
Non-formatted output can still be retrieved using the OSTREE_DUMP_RAW
flag, and the non-formatted key name is always printed for clarity.
Signed-off-by: Philip Withnall <withnall@endlessm.com> Closes: #911
Approved by: cgwalters
Colin Walters [Tue, 6 Jun 2017 17:34:27 +0000 (13:34 -0400)]
Canonicalize bare-user-only perms with 0755 mask
For the flatpak use case where bare-user-only was introduced, we actually
don't want to support s{u,g} id files in particular.
Actually, I can't think of a reason to have anything outside of the
`0755 i.e. (u=rwx,g=rx,o=rx)` mask, so that's what we do here.
This will have the effect of treating existing `bare-user-only` repositories as
corrupted if they have files outside of that mask, but I think we should do this
now; most of the flatpak users will still be on `bare-user`, and we haven't
changed the semantics of that mode yet.
Note that in this patch we will also *reject* file content that doesn't
match this. This is somewhat asymmetric, since we aren't similarly rejecting
e.g. directory metadata. But, this will close off the biggest source
of the problem for flatpak (setuid binaries).
Owen W. Taylor [Wed, 7 Jun 2017 02:51:00 +0000 (22:51 -0400)]
lib/repo: Don't copy xattrs when manipulating the GPG keyring
Copying xattrs when manipulating the GPG keyring for a repository
causes errors when the underlying filesystem doesn't support writing
xattrs - overlayfs is a common example. It also causes the selinux
attributes of the keyring files to be copied from the temporary
location instead of properly inherited from the destination directory
(ending up, for example, as unconfined_u:object_r:user_tmp_t:s0, rather
than unconfined_u:object_r:data_home_t:s0)
Jonathan Lebon [Fri, 2 Jun 2017 17:41:33 +0000 (13:41 -0400)]
basic-test.sh: explicitly check for uncompressed objects
It's not enough to check that the dir exists, since that's done by
default when we open the repo. We want to actually check that
uncompressed objects were cached (i.e. the opposite of the earlier error
path).
Jonathan Lebon [Fri, 2 Jun 2017 14:09:23 +0000 (10:09 -0400)]
checkout: also chmod in the user checkout case
When falling back to copying, we previously would only chmod checked out
files in the non-user-checkout mode. Fix this by always doing chmod.
The file_mode was being prepared but never actually applied.
Add a basic test in the archive-z2 --> usermode checkout case in which
we're guaranteed to always fall back to copy mode.
projects / ostree.git / log