Martin von Zweigbergk <martinvonz@google.com> [Fri, 04 Dec 2015 17:46:56 -0800] rev 27251
revlog: don't consider nullrev when choosing delta base
In the most complex case, we try using the incoming delta base, then
we try both parents, and then we try the previous revlog entry. If
none of these result in a good delta, we natually use the null
revision as base. However, we sometimes consider the nullrev before we
have exhausted our other options. Specifically, when both parents are
null, we use the nullrev as delta base if it produces a good delta
(according to _isgooddelta()), and we fail to try the previous revlog
entry as delta base. After
20a9226bdc8a (addrevision: use general
delta when the incoming base delta is bad, 2015-12-01), it can also
happen for non-merge commits when the incoming delta is not good.
The Firefox repo (from many months back) shrinks a tiny bit with this
patch: from 1.855GB to 1.830GB (1.4%). The hg repo itself shrinks even
less: by less than 0.1%. There may be repos that get larger instead.
This undoes the unexplained test change in
20a9226bdc8a.
Martin von Zweigbergk <martinvonz@google.com> [Fri, 04 Dec 2015 17:14:14 -0800] rev 27250
revlog: make calls to _isgooddelta() consistent
We always want to call _isgooddelta() before accepting a delta. We
mostly call the function right after building the delta, but not
always. Instead, we have an extra call at the end of the big code
block. Let's make it consistent, so we call _isgooddelta() right after
builddelta() and exactly once per delta. That also lets us rely on
"delta is None" to mean we didn't find a good delta.
Martin von Zweigbergk <martinvonz@google.com> [Fri, 04 Dec 2015 16:45:06 -0800] rev 27249
revlog: clarify which revision is added to 'tested' when using cached delta
The tested delta revisions are added to the 'tested' set. These are
the same revisions we pass to builddelta(). However, in one case, we
add builddelta(rev)[3] to the set intead of adding 'rev' itself. In
that particular case, that element is the same as the function's input
revision (because self._generaldelta is true), so the effect is the
same. Still, let's just add the function's input revision to avoid
confusing future readers.
Martin von Zweigbergk <martinvonz@google.com> [Fri, 04 Dec 2015 17:22:26 -0800] rev 27248
revlog: remove unused variable 'chainlen'
Matt Harbison <matt_harbison@yahoo.com> [Sat, 05 Dec 2015 22:19:48 -0500] rev 27247
commit: adjust the quoting in the examples to be Windows friendly
We should probably avoid strong quotes around command line args in the examples,
since cmd.exe doesn't recognize them, and it will surprise a user who cargo
cults them. I don't see a way to make a rule for this, since strong quoting is
OK inside command line args, like within revsets.
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 04 Dec 2015 15:12:11 -0800] rev 27246
wireproto: config options to disable bundle1
bundle2 is the new and preferred wire protocol format. For various
reasons, server operators may wish to force clients to use it.
One reason is performance. If a repository is stored in generaldelta,
the server must recompute deltas in order to produce the bundle1
changegroup. This can be extremely expensive. For mozilla-central,
bundle generation typically takes a few minutes. However, generating
a non-gd bundle from a generaldelta encoded mozilla-central requires
over 30 minutes of CPU! If a large repository like mozilla-central
were encoded in generaldelta and non-gd clients connected, they could
easily flood a server by cloning.
This patch gives server operators config knobs to control whether
bundle1 is allowed for push and pull operations. The default is to
support legacy bundle1 clients, making this patch backwards compatible.
Mathias De Maré <mathias.demare@gmail.com> [Mon, 07 Dec 2015 17:39:31 +0100] rev 27245
dockerlib: short form for non-unique uid/gid for CentOS 5 compat (
issue4977)
CentOS 5 does not support '--non-unique', but does support the short '-o'.
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 04 Dec 2015 13:31:01 -0800] rev 27244
exchange: standalone function to determine if bundle2 is requested
This will be used in a subsequent patch.
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 04 Dec 2015 13:15:14 -0800] rev 27243
wireproto: add docstring for wirepeer
Augie Fackler <augie@google.com> [Fri, 04 Dec 2015 15:24:05 -0500] rev 27242
commit: add some help examples (
issue4963)
Includes documenting --date now, which resolves the above issue.
Augie Fackler <augie@google.com> [Fri, 04 Dec 2015 11:39:03 -0500] rev 27241
changegroup: restate file linknode callback using generator expressions
I think this is slightly clearer, and it nicely avoids an extra nested
function.
Augie Fackler <augie@google.com> [Fri, 04 Dec 2015 11:38:02 -0500] rev 27240
changegroup: clean up file lookup function
One case is basically degenerate, so just extract it and make the
function clearer.
Augie Fackler <augie@google.com> [Fri, 04 Dec 2015 10:55:46 -0500] rev 27239
changegroup: remove one special case from lookupmflinknode
In the fastpathlinkrev case, lookupmflinknode was a very complicated
way of saying mfs.__getitem__, so let's just get that case out of our
way so it's easier to understand what's going on.
Augie Fackler <augie@google.com> [Fri, 04 Dec 2015 10:35:45 -0500] rev 27238
changegroup: drop 'if True' that made the previous change clearer
Augie Fackler <augie@google.com> [Fri, 04 Dec 2015 10:34:58 -0500] rev 27237
changegroup: avoid iterating the whole manifest
The old code gathered the list of all files that changed anywhere in
history and then gathered changed file nodes by walking the entirety
of each manifest to be sent in order to gather changed file
nodes. That's going to be unfortunate for narrowhg, and it's already
inefficient for medium-to-large repositories.
Timings for bundle --all on my hg repo, tested with hgperf:
Before:
! wall 23.442445 comb 23.440000 user 23.250000 sys 0.190000 (best of 3)
After:
! wall 20.272187 comb 20.270000 user 20.190000 sys 0.080000 (best of 3)