| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
This requires disabling the feature that retains the original order of
fields in a morphlogy when it gets overwritten. The implementation relies
on features that are not available in Python 2.6. We need to support
Morph on Debian squeeze, for bootstrapping purposes, and therefore need
to have it work with Python 2.6. However, the morphology rewriting is
only relevant for system branching and merging, and that isn't needed
for bootstrapping, so we disable the affected tests on Python 2.6.
|
|
|
|
|
|
| |
Move this into a script which can be sourced by the 'setup' scripts
and the actual tests (this is needed as the environment in 'setup' is
not passed on to the tests).
|
|
|
|
|
|
|
|
|
| |
This gives us consistency with morphologies, where the triplets are
repo|ref|morphology, not repo|ref|filename
Anyone who runs 'morph build baserock:morphs master system.morph' will
now see an error ending with 'was looking for system.morph.morph', which
should make it clear where they have gone wrong.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
More importantly, this test verifies that after building once,
additional builds do not generate additional artifacts if nothing
has changed.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This new command does a lot. First of all, its command line interface
has now changed to 'morph build SYSTEM' and it needs to be run from
a system branch.
When called, the new 'build' command will identify the repositories
and morphologies involved in building the system from the system branch,
create a build ref behind the scenes based on the system branch and add a
commit with all uncommitted changes to this build branch for every repo
involved. It will then push those build branches to the repository server
and kick off a build of BRANCH_ROOT BUILD_BRANCH SYSTEM.morph.
After building has finished, the remote build branches will be
deleted again.
|
|
|
|
|
|
|
|
|
|
| |
This is needed because we are replacing the current build command
by a new one that builds from a local system branch. Since that will
be the more common scenario, we want the new implementaiton to go by
the name 'morph build'.
This commit therefore renames 'morph build' and updates all the
tests to use 'morph build-morphology' instead.
|
|
|
|
|
|
|
|
| |
This removes the requirement that all strata must be in the same
repo as the system morphology.
Both the system "strata" field and the stratum "build-depends" field
are affected.
|
|
|
|
| |
Rename "sources" field of stratum morphologies to "chunks".
|
|
|
|
|
| |
The Disk image builder can work on non-x86 and as such we should
let it test anywhere.
|
|
|
|
| |
This overrides the mbr search path to a file that doesn't exist.
|
| |
|
| |
|
|
|
|
| |
Previously, it was only being created on ARM.
|
| |
|
|
|
|
|
| |
We only do this building on architectures we support, so there's no
need for variation.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously it would extract the stratum tarballs, since it was
easier than mounting the disk image.
Then when strata became chunk lists, this created the tarball, then
extracted it.
Now systems may contain files that aren't in the component strata,
this causes problems for tbdiff-deploy, since it tries to alter a
file that isn't there.
So to fix this, it mounts the disk image, like it should have from
the beginning.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
This removes the code preventing the tests being run on a non-x86
system, and makes the test system it tries to build be the same
architecture as itself.
System Build assumes x86 if arch is not specified for backwards compatibility
right now.
|
|
|
|
|
| |
This fixes references to a system artifact without the -rootfs appended
and adds a unit test for resolving an arm system artifact.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
This will make a build fail if the morphology is missing explicit
build dependencies.
Also fix test causes so that ./check passes.
|
|
|
|
|
|
|
|
| |
Also, some bug fixes.
Also, when handling a BaseException, log the exception (with traceback)
that we're handling, in case there is an error while handling it, because
the second error will otherwise mask the first one.
|
| |
|
|
|
|
|
| |
make-patch expected a tarball
system-overlap was parsing logs rather than using a stable format
|
|\ |
|
| | |
|
| |
| |
| |
| | |
This happens on some machines, not others (squeeze vs baserock).
|
| |
| |
| |
| | |
It's too hard to test things on multiple machines at once before committing.
|
| | |
|
|/
|
|
|
| |
This only fixes the output so it is independent of the way various
tools change output on different systems (sqeeze vs baserock).
|
|
|
|
|
|
| |
output is properly sorted, and redundant data is removed
also the stratum in stratum-overlap-*.setup are unpacked in a different
order, so there is a symlink to stomp first
|
|
|
|
|
| |
It may be better to print it to the console, but that requires
getting self.msg set
|
| |
|
| |
|
|
|
|
|
|
| |
Make sure that this directory is actually created before trying
to save artifacts to it. Also adjust all tests to use this new
directory for looking up artifacts.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
tbdiff can traverse the directory tree in an unsorted order, so two
equivalent patches can make changes in any order
This is a problem because tbdiff prints out the commands it ran,
so the output is not always the same. For now, we can just discard
this output, but ideally we want a --quiet option for tbdiff
|