perf: clear vfs audit_cache before each run
When generating a stream clone, we spend a large amount of time auditing path.
Before this changes, the first run was warming the vfs cache for the other
runs, leading to a large runtime difference and a "faulty" reported timing for
the operation.
We now clear this important cache between run to get a more realistic timing.
Below are some example of median time change when clearing these cases. The
maximum time for a run did not changed significantly.
### data-env-vars.name = mozilla-central-2018-08-01-zstd-sparse-revlog
# benchmark.name = hg.perf.exchange.stream.generate
# bin-env-vars.hg.flavor = default
# bin-env-vars.hg.py-re2-module = default
# benchmark.variants.version = latest
no-clearing: 17.289905
cache-clearing: 21.587965 (+24.86%, +4.30)
## data-env-vars.name = mozilla-central-2024-03-22-zstd-sparse-revlog
no-clearing: 32.670748
cache-clearing: 40.467095 (+23.86%, +7.80)
## data-env-vars.name = mozilla-try-2019-02-18-zstd-sparse-revlog
no-clearing: 37.838858
cache-clearing: 46.072749 (+21.76%, +8.23)
## data-env-vars.name = mozilla-unified-2024-03-22-zstd-sparse-revlog
no-clearing: 32.969395
cache-clearing: 39.646209 (+20.25%, +6.68)
In addition, this significantly reduce the timing difference between the
performance command, from the perf extensions and a `real `hg bundle` call
producing a stream bundle. Some significant differences remain especially on
the "mozilla-try" repositories, but they are now smaller.
Note that some of that difference will actually not be
attributable to the stream generation (like maybe phases or branch map
computation).
Below are some benchmarks done on a currently draft changeset fixing some
unrelated slowness in `hg bundle` (34a78972af409d1ff37c29e60f6ca811ad1a457d)
### data-env-vars.name = mozilla-central-2018-08-01-zstd-sparse-revlog
# bin-env-vars.hg.flavor = default
# bin-env-vars.hg.py-re2-module = default
hg.perf.exchange.stream.generate: 21.587965
hg.command.bundle: 24.301799 (+12.57%, +2.71)
## data-env-vars.name = mozilla-central-2024-03-22-zstd-sparse-revlog
hg.perf.exchange.stream.generate: 40.467095
hg.command.bundle: 44.831317 (+10.78%, +4.36)
## data-env-vars.name = mozilla-unified-2024-03-22-zstd-sparse-revlog
hg.perf.exchange.stream.generate: 39.646209
hg.command.bundle: 45.395258 (+14.50%, +5.75)
## data-env-vars.name = mozilla-try-2019-02-18-zstd-sparse-revlog
hg.perf.exchange.stream.generate: 46.072749
hg.command.bundle: 55.882608 (+21.29%, +9.81)
## data-env-vars.name = mozilla-try-2023-03-22-zlib-general-delta
hg.perf.exchange.stream.generate: 334.716708
hg.command.bundle: 377.856767 (+12.89%, +43.14)
## data-env-vars.name = mozilla-try-2023-03-22-zstd-sparse-revlog
hg.perf.exchange.stream.generate: 302.972301
hg.command.bundle: 326.098755 (+7.63%, +23.13)
== New Features ==
* `hg config` now has a `--source` option to show where each
configuration value comes from.
* Introduced a command (debug-repair-issue6528) to repair repositories affected
by issue6528 where certain files would show up as modified even if they were
clean due to an issue in the copy-tracing code.
== Default Format Change ==
These changes affect newly created repositories (or new clone) done with
Mercurial 5.9.
== New Experimental Features ==
* A `changelogv2` format has been introduced. It is not ready for use yet, but
will be used later to address some of the weaknesses of the current revlog
format.
* Initial experiment and support for `dirstatev2`, a new dirstate format that
addresses some of the weaknesses of the current dirstate format. Python + C
and Rust support are being implemented, but the Rust solution is the one
currently getting the attention for performance.
* Initial support for `rhg status`. `rhg` is the Rust wrapper executable for hg
that shortcuts some commands for faster execution speed.
== Bug Fixes ==
* Fixed committing empty files with `narrow`
* Allow overriding `pip`'s pep517 compliance to build C or Rust extensions
* Fixed regression on outgoing email when not specifying revisions
* Fixed a regression causing bookmarks to disappear when using Rust persistent nodemap
* Fixed a regression (in 5.9.1) introduced in 5.9 when cloning repos with
deep filenames
* Fixed detection of directories becoming symlinks, but only when using the
Rust extensions.
* Fixed ignore and include not composing in the Rust status
* `hg commit --interactive` now handles deselecting edits of a rename
* Fixed a case where `hg evolve` gives different results when interrupted
* Fixed a memory leak in phases computation
* `histedit` and `shelve` don't swallow errors when updating the working copy
anymore
* Improve error message when detecting content-divergence with a hidden
common predecessor
* No longer re-order parents in filelog, see issue6533
* Fix revisions affected by issue6533 on the fly during exchange
* Many Windows fixes for stability and py3 compatibility improvements
* Many other miscellaneous fixes
== Backwards Compatibility Changes ==
== Internal API Changes ==
The Dirstate API have been updated as the previous function leaked some
internal details and did not distinguish between two important cases: "We are
changing parent and need to adjust the dirstate" and "some command is changing
which file is tracked". To clarify the situation:
* the following functions have been deprecated,
- `dirstate.add`,
- `dirstate.normal`,
- `dirstate.normallookup`,
- `dirstate.merge`,
- `dirstate.otherparent`,
- `dirstate.remove`,
- `dirstate.drop`,
- `dirstateitem.__getitem__`,
* these new functions are added for the "adjusting parents" use-case:
- `dirstate.update_file`,
- `dirstate.update_file_p1`,
* these new function are added for the "adjusting wc file" use-case":
- `dirstate.set_tracked`,
- `dirstate.set_untracked`,
- `dirstate.set_clean`,
- `dirstate.set_possibly_dirty`,
See inline documentation of the new functions for details.
* Additionally, the following have been deprecated:
- `urlutil.getpath` function
- `localrepository.updatecaches`' `full` argument
* The following have been removed:
- `revlog.revlogio` has been removed