push: rework the computation of fallbackheads to be correct
The previous computation tried to be smart but ended up being wrong. This was
caught by phase movement test while reworking the phase discovery logic to be
faster.
The previous logic was failing to catch case where the pushed set was not based
on a common heads (i.e. when the discovery seemed to have "over discovered"
content, outside the pushed set)
In the following graph, `e` is a common head and we `hg push -r f`. We need to
detect `c` as a fallback heads and we previous failed to do so::
e
|
d f
|/
c
|
b
|
a
The performance impact of the change seems minimal. On the most impacted
repository at hand (mozilla-try), the slowdown seems mostly mixed in the
overall noise `hg push` but seems to be in the hundred of milliseconds order of
magnitude. When using rust, we seems to be a bit faster, probably because we
leverage more accelaratd internals.
I added a couple of performance related common for further investigation later
on.
Tests if hgweb can run without touching sys.stdin, as is required
by the WSGI standard and strictly implemented by mod_wsgi.
$ hg init repo
$ cd repo
$ echo foo > bar
$ hg add bar
$ hg commit -m "test"
$ cat > request.py <<EOF
> import os
> import sys
> from mercurial import (
> dispatch,
> encoding,
> hg,
> ui as uimod,
> util,
> )
> from mercurial.utils import (
> procutil,
> )
> ui = uimod.ui
> from mercurial.hgweb import hgweb_mod
> stringio = util.stringio
>
> class FileLike(object):
> def __init__(self, real):
> self.real = real
> def fileno(self):
> print >> sys.__stdout__, 'FILENO'
> return self.real.fileno()
> def read(self):
> print >> sys.__stdout__, 'READ'
> return self.real.read()
> def readline(self):
> print >> sys.__stdout__, 'READLINE'
> return self.real.readline()
>
> sys.stdin = FileLike(sys.stdin)
> errors = stringio()
> input = stringio()
> output = stringio()
>
> def startrsp(status, headers):
> print('---- STATUS')
> print(status)
> print('---- HEADERS')
> print([i for i in headers if i[0] != 'ETag'])
> print('---- DATA')
> return output.write
>
> env = {
> 'wsgi.version': (1, 0),
> 'wsgi.url_scheme': 'http',
> 'wsgi.errors': errors,
> 'wsgi.input': input,
> 'wsgi.multithread': False,
> 'wsgi.multiprocess': False,
> 'wsgi.run_once': False,
> 'REQUEST_METHOD': 'GET',
> 'SCRIPT_NAME': '',
> 'PATH_INFO': '',
> 'QUERY_STRING': '',
> 'SERVER_NAME': '$LOCALIP',
> 'SERVER_PORT': os.environ['HGPORT'],
> 'SERVER_PROTOCOL': 'HTTP/1.0'
> }
>
> i = hgweb_mod.hgweb(b'.')
> for c in i(env, startrsp):
> pass
> sys.stdout.flush()
> procutil.stdout.write(b'---- ERRORS\n')
> procutil.stdout.write(b'%s\n' % errors.getvalue())
> print('---- OS.ENVIRON wsgi variables')
> print(sorted([x for x in os.environ if x.startswith('wsgi')]))
> print('---- request.ENVIRON wsgi variables')
> with i._obtainrepo() as repo:
> print(sorted([encoding.strfromlocal(x) for x in repo.ui.environ
> if x.startswith(b'wsgi')]))
> EOF
$ "$PYTHON" request.py
---- STATUS
200 Script output follows
---- HEADERS
[('Content-Type', 'text/html; charset=ascii')]
---- DATA
---- ERRORS
---- OS.ENVIRON wsgi variables
[]
---- request.ENVIRON wsgi variables
['wsgi.errors', 'wsgi.input', 'wsgi.multiprocess', 'wsgi.multithread', 'wsgi.run_once', 'wsgi.url_scheme', 'wsgi.version']
$ cd ..