Summary:
Previously when the client has obsstore enabled, and the server has obsstore
disabled, clients won't get the obsmarkers, which is suboptimal. This diff
makes the server send obsmarkers in that case.
Practically, this means people will no longer need to run `strip` after landing
a diff in this repo.
Note: `test-pushrebase-manifests.t` is somehow flaky about the empty lines. I
guess they are not related to the change. But the test change is not 100%
reproducible.
Test Plan: Added a new test case
Reviewers: #mercurial, stash
Reviewed By: stash
Subscribers: rmcelroy, ikostia, stash, mjpieters
Differential Revision: https://phabricator.intern.facebook.com/D4865150
Signature: t1:4865150:1491895844:afbc3079a40a9a9fa9af1eab4eeaca91091e8d2d
Summary:
From diff:
```
This test does two things:
1/ Simulate a common condition of pushrebase under load. Normally pushrebase
caches data before acquiring the write lock (if lazy locking is enabled).
Under load, however, when a push has to wait for the lock more often than not,
much of this data becomes out of date and must be refetched once the lock is
acquired. This test simulates that particular case.
Specifically, we create two clients, client1 and client2, both with
nonconflicting changesets to push. client1's push is artificially blocked by a
`prepushrebase` hook (post-caching, pre-lock) that is only released after
client2's push succeeds.
2/ Checks how often we call manifest.read() inside the lock (and outside).
This way we can prevent regressions on manifest reads and test improvements.
manifest.read() is wrapped by an extension that prints a short trace. read calls
inside the lock are marked with a ":(".
```
This test is a bit of a doozy, simplification suggestions are encouraged.
Test Plan: this is, in fact, a test
Reviewers: #mercurial, durham
Reviewed By: durham
Subscribers: quark, rmcelroy, mjpieters
Differential Revision: https://phabricator.intern.facebook.com/D4785464
Signature: t1:4785464:1490737637:79d77f6b2e49a428d63b8e1b13d769c82149ee72
