localrepo: experimental support for non-zlib revlog compression
The final part of integrating the compression manager APIs into
revlog storage is the plumbing for repositories to advertise they
are using non-zlib storage and for revlogs to instantiate a non-zlib
compression engine.
The main intent of the compression manager work was to zstd all
of the things. Adding zstd to revlogs has proved to be more involved
than other places because revlogs are... special. Very small inputs
and the use of delta chains (which are themselves a form of
compression) are a completely different use case from streaming
compression, which bundles and the wire protocol employ. I've
conducted numerous experiments with zstd in revlogs and have yet
to formalize compression settings and a storage architecture that
I'm confident I won't regret later. In other words, I'm not yet
ready to commit to a new mechanism for using zstd - or any other
compression format - in revlogs.
That being said, having some support for zstd (and other compression
formats) in revlogs in core is beneficial. It can allow others to
conduct experiments.
This patch introduces *highly experimental* support for non-zlib
compression formats in revlogs. Introduced is a config option to
control which compression engine to use. Also introduced is a namespace
of "exp-compression-*" requirements to denote support for non-zlib
compression in revlogs. I've prefixed the namespace with "exp-"
(short for "experimental") because I'm not confident of the
requirements "schema" and in no way want to give the illusion of
supporting these requirements in the future. I fully intend to drop
support for these requirements once we figure out what we're doing
with zstd in revlogs.
A good portion of the patch is teaching the requirements system
about registered compression engines and passing the requested
compression engine as an opener option so revlogs can instantiate
the proper compression engine for new operations.
That's a verbose way of saying "we can now use zstd in revlogs!"
On an `hg pull` conversion of the mozilla-unified repo with no extra
redelta settings (like aggressivemergedeltas), we can see the impact
of zstd vs zlib in revlogs:
$ hg perfrevlogchunks -c
! chunk
! wall 2.032052 comb 2.040000 user 1.990000 sys 0.050000 (best of 5)
! wall 1.866360 comb 1.860000 user 1.820000 sys 0.040000 (best of 6)
! chunk batch
! wall 1.877261 comb 1.870000 user 1.860000 sys 0.010000 (best of 6)
! wall 1.705410 comb 1.710000 user 1.690000 sys 0.020000 (best of 6)
$ hg perfrevlogchunks -m
! chunk
! wall 2.721427 comb 2.720000 user 2.640000 sys 0.080000 (best of 4)
! wall 2.035076 comb 2.030000 user 1.950000 sys 0.080000 (best of 5)
! chunk batch
! wall 2.614561 comb 2.620000 user 2.580000 sys 0.040000 (best of 4)
! wall 1.910252 comb 1.910000 user 1.880000 sys 0.030000 (best of 6)
$ hg perfrevlog -c -d 1
! wall 4.812885 comb 4.820000 user 4.800000 sys 0.020000 (best of 3)
! wall 4.699621 comb 4.710000 user 4.700000 sys 0.010000 (best of 3)
$ hg perfrevlog -m -d 1000
! wall 34.252800 comb 34.250000 user 33.730000 sys 0.520000 (best of 3)
! wall 24.094999 comb 24.090000 user 23.320000 sys 0.770000 (best of 3)
Only modest wins for the changelog. But manifest reading is
significantly faster. What's going on?
One reason might be data volume. zstd decompresses faster. So given
more bytes, it will put more distance between it and zlib.
Another reason is size. In the current design, zstd revlogs are
*larger*:
debugcreatestreamclonebundle (size in bytes)
zlib: 1,638,852,492
zstd: 1,680,601,332
I haven't investigated this fully, but I reckon a significant cause of
larger revlogs is that the zstd frame/header has more bytes than
zlib's. For very small inputs or data that doesn't compress well, we'll
tend to store more uncompressed chunks than with zlib (because the
compressed size isn't smaller than original). This will make revlog
reading faster because it is doing less decompression.
Moving on to bundle performance:
$ hg bundle -a -t none-v2 (total CPU time)
zlib: 102.79s
zstd: 97.75s
So, marginal CPU decrease for reading all chunks in all revlogs
(this is somewhat disappointing).
$ hg bundle -a -t <engine>-v2 (total CPU time)
zlib: 191.59s
zstd: 115.36s
This last test effectively measures the difference between zlib->zlib
and zstd->zstd for revlogs to bundle. This is a rough approximation of
what a server does during `hg clone`.
There are some promising results for zstd. But not enough for me to
feel comfortable advertising it to users. We'll get there...
# Copyright (C) 2004, 2005 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, see <http://www.gnu.org/licenses/>.
from __future__ import absolute_import
import unittest
from mercurial import (
error,
simplemerge,
util,
)
TestCase = unittest.TestCase
# bzr compatible interface, for the tests
class Merge3(simplemerge.Merge3Text):
"""3-way merge of texts.
Given BASE, OTHER, THIS, tries to produce a combined text
incorporating the changes from both BASE->OTHER and BASE->THIS.
All three will typically be sequences of lines."""
def __init__(self, base, a, b):
basetext = '\n'.join([i.strip('\n') for i in base] + [''])
atext = '\n'.join([i.strip('\n') for i in a] + [''])
btext = '\n'.join([i.strip('\n') for i in b] + [''])
if util.binary(basetext) or util.binary(atext) or util.binary(btext):
raise error.Abort("don't know how to merge binary files")
simplemerge.Merge3Text.__init__(self, basetext, atext, btext,
base, a, b)
CantReprocessAndShowBase = simplemerge.CantReprocessAndShowBase
def split_lines(t):
return util.stringio(t).readlines()
############################################################
# test case data from the gnu diffutils manual
# common base
TZU = split_lines(""" The Nameless is the origin of Heaven and Earth;
The named is the mother of all things.
Therefore let there always be non-being,
so we may see their subtlety,
And let there always be being,
so we may see their outcome.
The two are the same,
But after they are produced,
they have different names.
They both may be called deep and profound.
Deeper and more profound,
The door of all subtleties!
""")
LAO = split_lines(""" The Way that can be told of is not the eternal Way;
The name that can be named is not the eternal name.
The Nameless is the origin of Heaven and Earth;
The Named is the mother of all things.
Therefore let there always be non-being,
so we may see their subtlety,
And let there always be being,
so we may see their outcome.
The two are the same,
But after they are produced,
they have different names.
""")
TAO = split_lines(""" The Way that can be told of is not the eternal Way;
The name that can be named is not the eternal name.
The Nameless is the origin of Heaven and Earth;
The named is the mother of all things.
Therefore let there always be non-being,
so we may see their subtlety,
And let there always be being,
so we may see their result.
The two are the same,
But after they are produced,
they have different names.
-- The Way of Lao-Tzu, tr. Wing-tsit Chan
""")
MERGED_RESULT = split_lines("""\
The Way that can be told of is not the eternal Way;
The name that can be named is not the eternal name.
The Nameless is the origin of Heaven and Earth;
The Named is the mother of all things.
Therefore let there always be non-being,
so we may see their subtlety,
And let there always be being,
so we may see their result.
The two are the same,
But after they are produced,
they have different names.
<<<<<<< LAO
=======
-- The Way of Lao-Tzu, tr. Wing-tsit Chan
>>>>>>> TAO
""")
class TestMerge3(TestCase):
def log(self, msg):
pass
def test_no_changes(self):
"""No conflicts because nothing changed"""
m3 = Merge3(['aaa', 'bbb'],
['aaa', 'bbb'],
['aaa', 'bbb'])
self.assertEquals(m3.find_unconflicted(),
[(0, 2)])
self.assertEquals(list(m3.find_sync_regions()),
[(0, 2,
0, 2,
0, 2),
(2, 2, 2, 2, 2, 2)])
self.assertEquals(list(m3.merge_regions()),
[('unchanged', 0, 2)])
self.assertEquals(list(m3.merge_groups()),
[('unchanged', ['aaa', 'bbb'])])
def test_front_insert(self):
m3 = Merge3(['zz'],
['aaa', 'bbb', 'zz'],
['zz'])
# todo: should use a sentinel at end as from get_matching_blocks
# to match without zz
self.assertEquals(list(m3.find_sync_regions()),
[(0, 1, 2, 3, 0, 1),
(1, 1, 3, 3, 1, 1)])
self.assertEquals(list(m3.merge_regions()),
[('a', 0, 2),
('unchanged', 0, 1)])
self.assertEquals(list(m3.merge_groups()),
[('a', ['aaa', 'bbb']),
('unchanged', ['zz'])])
def test_null_insert(self):
m3 = Merge3([],
['aaa', 'bbb'],
[])
# todo: should use a sentinel at end as from get_matching_blocks
# to match without zz
self.assertEquals(list(m3.find_sync_regions()),
[(0, 0, 2, 2, 0, 0)])
self.assertEquals(list(m3.merge_regions()),
[('a', 0, 2)])
self.assertEquals(list(m3.merge_lines()),
['aaa', 'bbb'])
def test_no_conflicts(self):
"""No conflicts because only one side changed"""
m3 = Merge3(['aaa', 'bbb'],
['aaa', '111', 'bbb'],
['aaa', 'bbb'])
self.assertEquals(m3.find_unconflicted(),
[(0, 1), (1, 2)])
self.assertEquals(list(m3.find_sync_regions()),
[(0, 1, 0, 1, 0, 1),
(1, 2, 2, 3, 1, 2),
(2, 2, 3, 3, 2, 2)])
self.assertEquals(list(m3.merge_regions()),
[('unchanged', 0, 1),
('a', 1, 2),
('unchanged', 1, 2)])
def test_append_a(self):
m3 = Merge3(['aaa\n', 'bbb\n'],
['aaa\n', 'bbb\n', '222\n'],
['aaa\n', 'bbb\n'])
self.assertEquals(''.join(m3.merge_lines()),
'aaa\nbbb\n222\n')
def test_append_b(self):
m3 = Merge3(['aaa\n', 'bbb\n'],
['aaa\n', 'bbb\n'],
['aaa\n', 'bbb\n', '222\n'])
self.assertEquals(''.join(m3.merge_lines()),
'aaa\nbbb\n222\n')
def test_append_agreement(self):
m3 = Merge3(['aaa\n', 'bbb\n'],
['aaa\n', 'bbb\n', '222\n'],
['aaa\n', 'bbb\n', '222\n'])
self.assertEquals(''.join(m3.merge_lines()),
'aaa\nbbb\n222\n')
def test_append_clash(self):
m3 = Merge3(['aaa\n', 'bbb\n'],
['aaa\n', 'bbb\n', '222\n'],
['aaa\n', 'bbb\n', '333\n'])
ml = m3.merge_lines(name_a='a',
name_b='b',
start_marker='<<',
mid_marker='--',
end_marker='>>')
self.assertEquals(''.join(ml),
'aaa\n'
'bbb\n'
'<< a\n'
'222\n'
'--\n'
'333\n'
'>> b\n'
)
def test_insert_agreement(self):
m3 = Merge3(['aaa\n', 'bbb\n'],
['aaa\n', '222\n', 'bbb\n'],
['aaa\n', '222\n', 'bbb\n'])
ml = m3.merge_lines(name_a='a',
name_b='b',
start_marker='<<',
mid_marker='--',
end_marker='>>')
self.assertEquals(''.join(ml), 'aaa\n222\nbbb\n')
def test_insert_clash(self):
"""Both try to insert lines in the same place."""
m3 = Merge3(['aaa\n', 'bbb\n'],
['aaa\n', '111\n', 'bbb\n'],
['aaa\n', '222\n', 'bbb\n'])
self.assertEquals(m3.find_unconflicted(),
[(0, 1), (1, 2)])
self.assertEquals(list(m3.find_sync_regions()),
[(0, 1, 0, 1, 0, 1),
(1, 2, 2, 3, 2, 3),
(2, 2, 3, 3, 3, 3)])
self.assertEquals(list(m3.merge_regions()),
[('unchanged', 0, 1),
('conflict', 1, 1, 1, 2, 1, 2),
('unchanged', 1, 2)])
self.assertEquals(list(m3.merge_groups()),
[('unchanged', ['aaa\n']),
('conflict', [], ['111\n'], ['222\n']),
('unchanged', ['bbb\n']),
])
ml = m3.merge_lines(name_a='a',
name_b='b',
start_marker='<<',
mid_marker='--',
end_marker='>>')
self.assertEquals(''.join(ml),
'''aaa
<< a
111
--
222
>> b
bbb
''')
def test_replace_clash(self):
"""Both try to insert lines in the same place."""
m3 = Merge3(['aaa', '000', 'bbb'],
['aaa', '111', 'bbb'],
['aaa', '222', 'bbb'])
self.assertEquals(m3.find_unconflicted(),
[(0, 1), (2, 3)])
self.assertEquals(list(m3.find_sync_regions()),
[(0, 1, 0, 1, 0, 1),
(2, 3, 2, 3, 2, 3),
(3, 3, 3, 3, 3, 3)])
def test_replace_multi(self):
"""Replacement with regions of different size."""
m3 = Merge3(['aaa', '000', '000', 'bbb'],
['aaa', '111', '111', '111', 'bbb'],
['aaa', '222', '222', '222', '222', 'bbb'])
self.assertEquals(m3.find_unconflicted(),
[(0, 1), (3, 4)])
self.assertEquals(list(m3.find_sync_regions()),
[(0, 1, 0, 1, 0, 1),
(3, 4, 4, 5, 5, 6),
(4, 4, 5, 5, 6, 6)])
def test_merge_poem(self):
"""Test case from diff3 manual"""
m3 = Merge3(TZU, LAO, TAO)
ml = list(m3.merge_lines('LAO', 'TAO'))
self.log('merge result:')
self.log(''.join(ml))
self.assertEquals(ml, MERGED_RESULT)
def test_binary(self):
self.assertRaises(error.Abort, Merge3, ['\x00'], ['a'], ['b'])
def test_dos_text(self):
base_text = 'a\r\n'
this_text = 'b\r\n'
other_text = 'c\r\n'
m3 = Merge3(base_text.splitlines(True), other_text.splitlines(True),
this_text.splitlines(True))
m_lines = m3.merge_lines('OTHER', 'THIS')
self.assertEqual('<<<<<<< OTHER\r\nc\r\n=======\r\nb\r\n'
'>>>>>>> THIS\r\n'.splitlines(True), list(m_lines))
def test_mac_text(self):
base_text = 'a\r'
this_text = 'b\r'
other_text = 'c\r'
m3 = Merge3(base_text.splitlines(True), other_text.splitlines(True),
this_text.splitlines(True))
m_lines = m3.merge_lines('OTHER', 'THIS')
self.assertEqual('<<<<<<< OTHER\rc\r=======\rb\r'
'>>>>>>> THIS\r'.splitlines(True), list(m_lines))
if __name__ == '__main__':
# hide the timer
import time
orig = time.time
try:
time.time = lambda: 0
unittest.main()
finally:
time.time = orig