Commit Graph
663 Commits
Author SHA1 Message Date
Colin Walters 0d72168961 libostree: Make the format of file objects private
It's now isolated almost entirely to ostree-core.c, except
ostree-repo.c needs to know how to create archive-z2 file headers.  So
give it a private API for that.
2013-09-03 18:43:42 -04:00
Colin Walters 356de33b49 repo: Drop ostree_repo_load_variant_c API
Originally we had this to avoid forcing callers to malloc() if they
had a csum, but nowadays we have in-place conversion APIs that are
fast enough.
2013-09-03 18:23:11 -04:00
Colin Walters dd7d2f7b43 repo: Only apply setuid/xattrs after checksum validation
See the new comment in the source; basically if we're fetching content
over http, then someone with the capability to MITM the network could
create a transient setuid binary on disk with arbitrary content.  If
they also had a process running on the system (such as an application)
it could be escalated to root.

https://bugzilla.gnome.org/show_bug.cgi?id=707139
2013-09-02 15:31:55 -04:00
Tobias Hunger ed56908ccb Fix warnings about unused variables 2013-08-30 14:23:45 -04:00
Colin Walters d92eedac4a repo: Clean up tmpdir also on transaction abort
Pull the cleanup code to a helper function, and ensure we delete
leftover temporary files also when aborting a transaction.  Mainly
this will happen if a local 'ostree commit' fails.

While we're here, also change it to use gs_shutil_rm_rf() which also
handles directories, should we start using those.

Reviewed-by: Jeremy Whiting <jpwhiting@kde.org>
2013-08-27 11:35:45 -04:00
Jeremy Whiting d58a4c9f79 repo: Clean up temporaries after a transaction completes
Prevously, we were just leaving temporary files there forever if
a transaction was interrupted.

https://bugzilla.gnome.org/show_bug.cgi?id=706344
2013-08-27 11:35:36 -04:00
Colin Walters edfa76fad5 main: Code cleanup by passing OstreeRepo * directly to builtins
It turns out every builtin (with one special exception) that takes a
repo argument did the same thing; let's just centralize it.  The
special exception was "ostree init --repo=foo" where foo is expected
to *not* actually be a repo.  In that case, simply skip the
ostree_repo_check() invocation.

https://bugzilla.gnome.org/show_bug.cgi?id=706762
2013-08-25 15:28:23 -04:00
Colin Walters 6c61b19107 libostree: Improve commit filter API
Make the structure private, and document the flags.

https://bugzilla.gnome.org/show_bug.cgi?id=706214
2013-08-25 15:23:23 -04:00
Colin Walters b600ae5939 repo: Fix object storage size API to be 64 bit
This fixes a pointer size warning on 32-bit builds.

https://bugzilla.gnome.org/show_bug.cgi?id=706235
2013-08-25 15:19:54 -04:00
Colin Walters 35d7b9cdbf libostree: Increase gtk-doc coverage of ostree-repo.[ch] 2013-08-17 13:57:22 -04:00
Colin Walters 1fea88e64b libostree: Add gtk-doc section for ostree-repo.c 2013-08-17 13:22:58 -04:00
Colin Walters abe2320039 libostree: Fix many gtk-doc warnings 2013-08-17 08:41:31 -04:00
Colin Walters 1ec7c30408 core: Remove old "archive" mode
We'll always have "bare" mode for keeping files-as-hardlinks as root.
But "archive" was my second attempt at a format for non-root file
storage, used by the gnome-ostree buildsystem which runs as non-root.

It was really handy to have a "tar" like mode where I can create
tarballs as a user, that contain files owned by root for example.

The "archive" mode stored content files as two pieces in the
filesystem; ".file" contained metadata, and ".filecontent" was the
actual content, uncompressed.  The nice thing about this was that to
check out a tree as non-root, you could just hardlink into the repo.

However, archive was fairly bad for serving via HTTP; it required
*two* HTTP requests per content object, greatly magnifing the already
inefficient fetch process.  So "archive-z2" was introduced.

To allow gnome-ostree to still check out trees as a user, the
"uncompressed-object-cache" was introduced, and that's how things have
been working for a while.

So we should just be able to kill this code.  Specifically note just
how much better the stage_object() function became.

https://bugzilla.gnome.org/show_bug.cgi?id=706057
2013-08-16 22:56:42 -04:00
Colin Walters 11bdbe1fb8 repo: Add API to load any object as a stream
We have APIs to load metadata as variants, and files as parsed
content/info/xattrs, but for some cases such as static deltas, all we
want is to operate on all objects in their canonical representation.

https://bugzilla.gnome.org/show_bug.cgi?id=706031
2013-08-14 21:50:37 -04:00
Stef Walter 0111ec1307 Add some verbose logging to 'ostree commit'
This is so we can see near which file we fail when 'ostree commit'
falls over.

https://bugzilla.gnome.org/show_bug.cgi?id=705905
2013-08-13 15:38:51 +02:00
Colin Walters 71c4ea67f7 core: Clean up temporary file creation
Use new libgsystem API, don't expose the repo's tmpdir.
2013-08-06 23:58:29 +02:00
Colin Walters 9e480cd6a9 core: Finish making object path API private
This allows us to more easily change the internals later.
2013-07-31 03:48:33 -04:00
Colin Walters 0dec1b287c core: Reduce API surface for file access
This function was only used by internals, make it private.
2013-07-31 00:35:53 -04:00
Colin Walters 7c5c3f2af8 core: Fix all introspection warnings
Still lots more docs to write and API to cleanup, but this is better.
2013-07-27 10:13:30 -04:00
Colin Walters a5d43bb959 Install a shared library
This required a fair bit of surgery because previously ostree.h
included otutil.h, but that's supposed to be a private library.
2013-07-26 19:25:07 -04:00
Colin Walters 5aea9db3f2 Add API to retrieve statistics from transactions, use it in commit
It's just interesting to see =) This is also kind of prototyping out
some more "structured" output.
2013-07-25 19:23:25 -04:00
Colin Walters c9b61cbfee Drop support for related objects and metadata in commit objects
While the actual commit object format is presently the same, for a
number of reasons we'd like to change it fairly radically.  Among
other things, we need to drop our a{sv} types in objects, to protect
against GVariant changing format.

Since now gnome-ostree now longer uses related objects, and nothing
ever used metadata, just drop them both.
2013-07-24 12:59:27 -04:00
Colin Walters 8724565291 core: Drop a leftover unnecessary #define _GNU_SOURCE
We now use AC_USE_SYSTEM_EXTENSIONS.
2013-07-16 09:22:42 -04:00
Colin Walters 89181c4e8e Build without libsoup again 2013-07-09 20:35:15 -04:00
Colin Walters cb6b69616c libostree: Split off -refs.c
Continuing to break up ostree-repo.c.
2013-07-09 19:49:00 -04:00
Colin Walters 3b6a94d272 libostree: Split off libarchive code 2013-07-09 19:11:37 -04:00
Colin Walters dbffd31017 libostree: Start splitting up the monstrous ostree-repo.c
Checkout now lives in its own file.
2013-07-09 19:02:38 -04:00
Colin Walters 56f8584fe4 repo: Store transaction file, use it to optimize for pull resumes
If pull is interrupted, we may have downloaded an arbitrary subset of
the requested objects.  Previously, we handled this by scanning for
all objects each time.

However, there's an easy optimization - this patch creates a lock file
in the repo.  If we don't see that file when starting a pull, we know
we don't need to stat() every file; presence of a dirtree object for
example implies the existence of everything it references.
2013-07-08 09:05:34 -04:00
Colin Walters 4b170d656c Switch to libgsystem local allocation macros
And drop our compatibility wrapper.
2013-07-07 12:27:44 -04:00
Colin Walters dc0f3c3dcb Add support for deleting refs
The internal API will be used by admin, and "ostree refs --delete"
is handy for interactive management.
2013-07-01 15:41:27 -04:00
Colin Walters 5b3fca8426 Add "refs" builtin
This is just useful to look at before pruning, etc.
2013-06-29 14:51:08 -04:00
Colin Walters 1ba852ebaa core: Add "refspec" which is remote:refname
This allows an unambiguous reference; otherwise, it was too easy to
have confusion between local heads and remotes.
2013-06-29 12:51:29 -04:00
Colin Walters 7bb20b0533 core: Port more callsites to gs_file_enumerator_iterate()
See previous commit.
2013-06-14 19:45:40 -04:00
Colin Walters fbd9f17e39 core: Update libgsystem, port some uses of GFileEnumerator
This makes the code *so* much nicer.
2013-06-14 19:14:35 -04:00
Colin Walters f4327cc6a0 repo: Fix race condition in async checkout
When multiple threads need to uncompress an object, there was
a race condition where thread A could get EEXIST, unlink,
then thread B calls linkat(), then thread A tries to link() but
fails.

We can just loop in this case.
2013-04-30 19:34:20 -04:00
Colin Walters 5d1b0ec1b3 core: Fix object enumeration for archive-z2 repositories
This makes fsck work again, among other things.
2013-04-01 11:43:03 -04:00
Colin Walters 47bd290065 core: Drain fdatasync() API into libgsystem
Just code cleanup.
2013-01-04 17:25:32 -05:00
Colin Walters 1ac307e6b8 Switch to libgsystem file reading API
More draining common utilities into libgsystem.
2012-11-29 16:58:39 -05:00
Colin Walters f3d765c903 Switch to libgsystem directory creation API
More code being dropped into a common place.
2012-11-29 16:40:17 -05:00
Colin Walters 5761733218 Drop some bits from libotutil that are now merged in libgsystem 2012-11-29 16:24:49 -05:00
Colin Walters de1ce843f1 core: Drop old GLIB_CHECK_VERSION(2.32) bits
We hard require 2.34 now; if it doesn't exist on the system, that's
what embedded dependencies are for.
2012-11-18 15:44:21 -05:00
Colin Walters 5a51bee024 core: quiet a compiler warning 2012-10-17 20:54:52 -04:00
Colin Walters 0c3e901dae core: Check out files before directories
Otherwise we pretty easily hit maximum FD limits =(
2012-10-15 20:03:13 -04:00
Colin Walters bdbf76b6eb core: Ensure correct file data is synced in non-overwrite case 2012-10-15 18:44:55 -04:00
Colin Walters 1642310486 core: Use linkat() rather than link() in threaded checkouts
This seems to work around a likely Linux kernel VFS bug, where I
randomly see ENOENT on link() when we *definitely* called mkdir() at
an earlier point in time.
2012-10-15 10:49:48 -04:00
Colin Walters c0a6b5da71 core: Fix devino speedup for archive-z 2012-10-14 17:19:30 -04:00
Colin Walters 0d04738801 core: Rework archive-z mode to have header be uncompressed
This is an incompatible change to archive-z, thus it is now renamed to
archive-z2 and ostree will no longer parse archive-z.

I noticed in perf that we were spending some time zlib-decompressing
file headers, which is just inefficient.  Rather than do this, keep
the headers uncompressed, and just zlib-compress content.
2012-10-14 17:10:57 -04:00
Colin Walters 2b90e987bf core: Call fdatasync() before we rename()
Just noticed this from strace, doesn't really matter, but looks nicer.
2012-10-14 15:34:33 -04:00
Colin Walters 5b512d4c78 core: Fix case where we hardlinked successfully but thought we didn't
We need to return hardlink_supported even if we had to do an unlink
first.
2012-10-13 18:41:47 -04:00
Colin Walters ad26c7a80b core: A few more error-prefixings 2012-10-13 18:02:39 -04:00