summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
...
* test suite: t-gbp-example-prep: Drop t-tstunt-parsechangelogIan Jackson2019-07-18
| | | | | | | And hoist it into all callers. In fact there is only one caller that doesn't have it already. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: Document dgit(1) archive-query docIan Jackson2019-07-17
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* dgit(1) document the presence of the aptget archive-query methodMatthew Vernon2019-07-17
| | | | | | | | | This is already used for debian-security, and might be useful for other archives that like madison or similar (e.g. the Ubuntu Cloud Archive). Closes: #932321 Signed-off-by: Matthew Vernon <mv3@sanger.ac.uk>
* git-debpush: refactor to avoid more than one push commandSean Whitton2019-07-17
| | | | | | No functional change. Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* changelog: update for tag-to-upload server-side fixesIan Jackson2019-07-17
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: update for git-debpush improvementsSean Whitton2019-07-17
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: check that upstream tag is ancestor of $branchSean Whitton2019-07-17
| | | | | | Closes: #932096 Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: gather some information earlierSean Whitton2019-07-17
| | | | | | | | For use by sanity checks. Pure code motion. Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: cope with spaces in user-supplied upstream tagSean Whitton2019-07-17
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: check for target suite change since last uploadSean Whitton2019-07-17
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: check for pushing the dgit view to the maintainer viewSean Whitton2019-07-17
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: factor out find_last_tag()Sean Whitton2019-07-17
| | | | | | | | Also determine the last debian/ tag earlier, for use by sanity checks. No functional change. Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: introduce and call fail_checkSean Whitton2019-07-17
| | | | | | | | | | | | | | Without this, passing --force to override an earlier sanity check might cause the user to miss the output of a later sanity check, which would never get run. We don't want to have multiple forcing options to override different sanity checks, as that is too much complexity for a script like this. So always run all checks, and error out afterwards if at least one of them failed. Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* test suite: tagupl-baredebian: New testIan Jackson2019-07-17
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: baredebian: Break up baredebian-test-coreIan Jackson2019-07-17
| | | | | | | For a baredebian tag-to-upload test, we're going to want to replace the push. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* Dgit.pm: printcmd: Print in one goIan Jackson2019-07-17
| | | | | | | | | | | | | | | | | | | | | | | Sometimes this message is used to report failure of `git fetch' etc. But when such operations fail, they may interleave output from the remote git, which typically prints fatal: The remote end hung up unexpectedly (referring to the local end as the `remote' end, from its point of view), after the the local end printed its message and failed. This sets up a race between Dgit.pm's printcmd and the message from the remote. If we're really unlucky, the remote's message interrupts the output from printcmd. We can avoid the within-line interleaving - at least, the interruption of our message - by printing the whole message in one go, so do that. (Strictly, the approach we use here only ensures that the message appears in a single write(2) call if it fits in the stdio buffer.) One of our test cases (tagupl) depends on this message not being split and has a small chance of failing without this fix. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: baredebian: Move the quilt fixup forwardIan Jackson2019-07-17
| | | | | | | | | We are going to split this function up so that our caller can do the push via tag-to-upload. No overall functional change. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* tag2upload: baredebian: Pass dgit refs/tags/, not the commit hashIan Jackson2019-07-17
| | | | | | | | | This will enable git to report the upstream tag in its messages. The tag name is safe to pass here since dgit only uses its referent, unless it is nice enough to print. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* Dgit.pm: resolve_upstream_version: Report tag, where sensibleIan Jackson2019-07-17
| | | | | | | | | | | | | This affect what is show in commit messages (including the parseable parts) made by dgit and git-debrebase. Now they prefer to refer to the tag name if the caller specified refs/tags/ and the tag name is a nice one. This is not expected to make much difference for human callers (who will probably not explicitly qualify their tag refs). But it will make a difference for scripts. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* tag2upload: Actually recognise and honour --quilt in tagIan Jackson2019-07-17
| | | | | | | | | Non-default quilt modes were completely broken before. We only didn't notice because the existing tests worked with the default. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* dgit-repos-server: parsetag: Print a bit more debug about junkIan Jackson2019-07-17
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* dgit-repos-server: Don't crash on --quilt=baredebianIan Jackson2019-07-17
| | | | | | | The die needs to be qualified. This little bit of fcode has never worked. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* Dgit.pm: reflog_cache_insert: Use ensurepath for the reflogIan Jackson2019-07-17
| | | | | | Otherwise quilt fixup can fail in a too-fresh git tree. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* Dgit.pm: Provide `ensurepath'Ian Jackson2019-07-17
| | | | | | | | | | Like `ensuredir' but also a bit like a more controlled `mkdir -p'. We use make_path from File::Path. It dies on failure. No caller yet. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: tagupl-*: Move t-pushed-good out of t-tagupl-testIan Jackson2019-07-17
| | | | | | | | We are going to want to do this separately in some cases. No functional change. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: Start 9.3Ian Jackson2019-07-09
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: Finalise 9.2 (no source changes)debian/9.2archive/debian/9.2Ian Jackson2019-07-09
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: start 9.2Ian Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: finalise 9.1debian/9.1archive/debian/9.1Ian Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: Minor adjustment to git-debpush discussionIan Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* changelog: Add a couple of missing entriesIan Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* dgit-repos-server: Remove comment saying it's a prototypeIan Jackson2019-07-07
| | | | | | | We have done everything to make it safe to use. Obviously it still needs some web service queue thing around it to invoke it. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* git-debpush(1): Tweak AUTHOR sectionSean Whitton2019-07-07
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: Don't pretend to support --baredebian+tarballSean Whitton2019-07-07
| | | | | | | | | Instead of just deleting the option from the argument parser, accept it, but print a message explicitly stating that the *quilt mode* is not supported by this workflow (i.e. it's not just git-debpush passing the quilt mode on to tag2upload that we don't support). Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* changelog: UpdateSean Whitton2019-07-07
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush(1): Fix POD formattingSean Whitton2019-07-07
| | | | Signed-off-by: Sean Whitton <spwhitton@spwhitton.name>
* test suite: unset CDPATHIan Jackson2019-07-07
| | | | | | This can cause tests to pass which should fail (!) Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: tagupl: Test git-debpush --upstream=Ian Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* git-debpush: Defend against git-deborig #931509Ian Jackson2019-07-07
| | | | | | | | | | | | | | | | | | | | | | | | Without this, when git-deborig fails and #931509 isn't fixed, git-debpush will silently exit with a nonzero exit status which is very unfriendly. With this change it does something like this: git-debpush: git-deborig failed; maybe try git-debpush --upstream=TAG fatal: Invalid object name 'refs/tags/couldn't find any of the following tags'. It then exits status 127. This is obviously daft, but making this any better is even more complex. When git-deborig is fixed it will do this: couldn't find any of the following tags: 1.0, v1.0, upstream/1.0 tell me a tag or branch head to make an orig.tar from: git deborig --just-print '--version=1.0-1' COMMITTISH git-debpush: git-deborig failed; maybe try git-debpush --upstream=TAG which is still not brilliant but I guess it will do. The downside, if it is one, is that we lose git-deborig's original exit status. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk> Acked-by: Sean Whitton <spwhitton@spwhitton.name>
* test suite: tagupl: Drop an unused variableIan Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* git-debpush: Support --upstream=TAG (and document it)Ian Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* git-debpush: Defend against unexpected git-deborig outputIan Jackson2019-07-07
| | | | | | | | | | | If git-deborig prints something other than a tag name, for example an error message (see #931509), this approach will bomb out, roughly appropriately. This will also become useful when we make upstream_tag overrideable. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk> Acked-by: Sean Whitton <spwhitton@spwhitton.name>
* git-debpush: Change comment headingIan Jackson2019-07-07
| | | | | | | | | IMO "Principles of Operation" would be a comprehensive reference detailing the model and functionality. I think "Design Principles" is the conventional name for semi-self-imposed choices/constraints. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk> Acked-by: Sean Whitton <spwhitton@spwhitton.name>
* test suite: t-tagupl-test: Check that salsa master was pushedIan Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: tagupl-native: Drop obsolete todo commentIan Jackson2019-07-07
| | | | | | We do now test all these, in tests/tests/tagupl. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: t-tagupl-test: Use "salsa" as the sourceIan Jackson2019-07-07
| | | | | | This checks that push pushed the tag(s) there. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: t-tagupl-run-drs: New source parameterIan Jackson2019-07-07
| | | | | | Change all callers. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: tagupl*: Drop use of --no-pushIan Jackson2019-07-07
| | | | | | Now we do actually push. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: t-tagupl-setting: Provide a "salsa" remoteIan Jackson2019-07-07
| | | | | | This will give git-debpush somewhere to push to. Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
* test suite: tagupl*: Run t-tagupl-settings in $pIan Jackson2019-07-07
| | | | Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
generated by cgit v1.2.3 (git 2.25.1) at 2024-04-19 13:14:27 +0000