Laurent Charignon <lcharignon@fb.com> [Tue, 16 Jun 2015 14:30:43 -0700] rev 1369
evolve: properly evolve stacked unstable with --rev
Before this patch, _singlesuccessor was not returning a revision number for
the case where the parent of the argument was not obsolete. This resulted in
bug when testing membership with the set of revisions. This patch fixes it and
adds a test.
Laurent Charignon <lcharignon@fb.com> [Tue, 16 Jun 2015 10:19:17 -0700] rev 1368
directaccess: disable directaccess for push and serve
Before this patch, push and serve were subject to directaccess. This patch
makes them throw error when trying to access hidden hashes.
Laurent Charignon <lcharignon@fb.com> [Tue, 16 Jun 2015 10:08:48 -0700] rev 1367
directaccess: change rule from opt-in to opt-out
Before this patch we would opt-in commands for direct access and the default
filter for new repository was 'visible'.
With this patch, the default filter for new repos is
'visible-directaccess-warn'. It means that by default all the commands have
directaccess with warnings.
Laurent Charignon <lcharignon@fb.com> [Tue, 16 Jun 2015 10:07:51 -0700] rev 1366
inhibit: move transaction wrapping outside of repo setup
Before this patch, transaction wrapping code was done in reposetup. It happened
to cause stackoverflows in repos with a lot of subreps.
This patch moves the wrapping to extsetup and avoids this problem.
Pierre-Yves David <pierre-yves.david@fb.com> [Mon, 15 Jun 2015 17:44:12 -0700] rev 1365
tests: ignores other core output in capability testing
This will avoid breaking the test anytime core changes.
Laurent Charignon <lcharignon@fb.com> [Wed, 20 May 2015 12:46:13 -0700] rev 1364
evolve: make uncommit respect allowunsable
Before this patch, the uncommit command was performing the same way regardless
of the value of experimental.evolution.
With this patch if the configuration does not allow unstability, uncommit won't
create instability.
Laurent Charignon <lcharignon@fb.com> [Sat, 13 Jun 2015 11:14:07 -0700] rev 1363
directaccess: use cached filteredrevs
Before this patch we were calling directly repoview.computehidden(repo) to
compute the revisions visible with direct access, without going through the
caching mechanism for the filtered revisions.
There was two issues with that:
(1) Performance: We were not leverating the cached values of the 'visible' revs
(2) Stability: If there were to be a cache inconsistency with the computation of
'visible' we would crash in the branchmap consistency check partial.validfor.
Consider the scenario of rebase with bookmarks:
- when we delete a bookmark on an obsolete changeset (like what rebase
does when moving the bookmark after rebasing the changesets)
- then this changes the value returned by repoview.computehidden(repo) as
bookmarks are used as dynamic blockers in repoview.computehidden(repo)
- as of now, we don't invalidate the cache in the case of bookmark change
- if we have a cached value from before the bookmark change,
repoview.filterrevs(repo, 'visible') considers the cached value correct and
returns something different than repoview.computehidden(repo)
- in turn, if we use repoview.computehidden(repo) in directaccess, the subset
relationship is broken and the cache consistency assertion (parial.validfor)
fails if branchmap.updatecache is called in this time window
This patch leverages the caching infrastructure in place to speed up the
computation of the filteredrevs for visible-directaccess-nowarn and
visible-directaccess-warn. Incidentally it prevents the bug discussed in (2)
from crashing when running a rebase with a bookmark. Note that there still
needs to be a fix in core for the case discussed in (2).
The test for this side of the fix (not core's fix for (2) is very hard to
implement without introducing a lot of dependencies and does not belong
here. It is much easier to have the test of the fix for the scenario (2) in
core along with the fix.