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)
$ testparseutil="$TESTDIR"/../contrib/testparseutil.py
Internal test by doctest
$ "$PYTHON" -m doctest "$testparseutil"
Tests for embedded python script
Typical cases
$ "$PYTHON" "$testparseutil" -v pyembedded <<NO_CHECK_EOF
> >>> for f in [1, 2, 3]:
> ... foo = 1
> >>> foo = 2
> $ echo "doctest is terminated by command, empty line, or comment"
> >>> foo = 31
> expected output of doctest fragment
> >>> foo = 32
>
> >>> foo = 33
>
> >>> foo = 34
> comment
> >>> foo = 35
>
> $ "\$PYTHON" <<EOF
> > foo = 4
> >
> > EOF
> $ cat > foo.py <<EOF
> > foo = 5
> > EOF
> $ cat >> foo.py <<EOF
> > foo = 6 # appended
> > EOF
>
> NO_CHECK_EOF limit mark makes parsing ignore corresponded fragment
> (this is useful to use bad code intentionally)
>
> $ "\$PYTHON" <<NO_CHECK_EOF
> > foo = 7 # this should be ignored at detection
> > NO_CHECK_EOF
> $ cat > foo.py <<NO_CHECK_EOF
> > foo = 8 # this should be ignored at detection
> > NO_CHECK_EOF
>
> doctest fragment ended by EOF
>
> >>> foo = 9
> NO_CHECK_EOF
<stdin>:1: <anonymous> starts
|for f in [1, 2, 3]:
| foo = 1
|foo = 2
<stdin>:4: <anonymous> ends
<stdin>:5: <anonymous> starts
|foo = 31
|
|foo = 32
|
|foo = 33
<stdin>:10: <anonymous> ends
<stdin>:11: <anonymous> starts
|foo = 34
<stdin>:12: <anonymous> ends
<stdin>:13: <anonymous> starts
|foo = 35
<stdin>:14: <anonymous> ends
<stdin>:16: <anonymous> starts
|foo = 4
|
<stdin>:18: <anonymous> ends
<stdin>:20: foo.py starts
|foo = 5
<stdin>:21: foo.py ends
<stdin>:23: foo.py starts
|foo = 6 # appended
<stdin>:24: foo.py ends
<stdin>:38: <anonymous> starts
|foo = 9
<stdin>:39: <anonymous> ends
Invalid test script
(similar test for shell script and hgrc configuration is omitted,
because this tests common base class of them)
$ "$PYTHON" "$testparseutil" -v pyembedded <<NO_CHECK_EOF > detected
> $ "\$PYTHON" <<EOF
> > foo = 1
>
> $ "\$PYTHON" <<EOF
> > foo = 2
> $ cat > bar.py <<EOF
> > bar = 2 # this fragment will be detected as expected
> > EOF
>
> $ cat > foo.py <<EOF
> > foo = 3
> NO_CHECK_EOF
<stdin>:3: unexpected line for "heredoc python invocation"
<stdin>:6: unexpected line for "heredoc python invocation"
<stdin>:11: unexpected end of file for "heredoc .py file"
[1]
$ cat detected
<stdin>:7: bar.py starts
|bar = 2 # this fragment will be detected as expected
<stdin>:8: bar.py ends
Tests for embedded shell script
$ "$PYTHON" "$testparseutil" -v shembedded <<NO_CHECK_EOF
> $ cat > foo.sh <<EOF
> > foo = 1
> >
> > foo = 2
> > EOF
> $ cat >> foo.sh <<EOF
> > foo = 3 # appended
> > EOF
>
> NO_CHECK_EOF limit mark makes parsing ignore corresponded fragment
> (this is useful to use bad code intentionally)
>
> $ cat > foo.sh <<NO_CHECK_EOF
> > # this should be ignored at detection
> > foo = 4
> > NO_CHECK_EOF
>
> NO_CHECK_EOF
<stdin>:2: foo.sh starts
|foo = 1
|
|foo = 2
<stdin>:5: foo.sh ends
<stdin>:7: foo.sh starts
|foo = 3 # appended
<stdin>:8: foo.sh ends
Tests for embedded hgrc configuration
$ "$PYTHON" "$testparseutil" -v hgrcembedded <<NO_CHECK_EOF
> $ cat > .hg/hgrc <<EOF
> > [ui]
> > verbose = true
> >
> > # end of local configuration
> > EOF
>
> $ cat > \$HGRCPATH <<EOF
> > [extensions]
> > rebase =
> > # end of global configuration
> > EOF
>
> $ cat >> \$HGRCPATH <<EOF
> > # appended
> > [extensions]
> > rebase =!
> > EOF
>
> NO_CHECK_EOF limit mark makes parsing ignore corresponded fragment
> (this is useful to use bad code intentionally)
>
> $ cat > .hg/hgrc <<NO_CHECK_EOF
> > # this local configuration should be ignored at detection
> > [ui]
> > username = foo bar
> > NO_CHECK_EOF
>
> $ cat > \$HGRCPATH <<NO_CHECK_EOF
> > # this global configuration should be ignored at detection
> > [extensions]
> > foobar =
> > NO_CHECK_EOF
> NO_CHECK_EOF
<stdin>:2: .hg/hgrc starts
|[ui]
|verbose = true
|
|# end of local configuration
<stdin>:6: .hg/hgrc ends
<stdin>:9: $HGRCPATH starts
|[extensions]
|rebase =
|# end of global configuration
<stdin>:12: $HGRCPATH ends
<stdin>:15: $HGRCPATH starts
|# appended
|[extensions]
|rebase =!
<stdin>:18: $HGRCPATH ends