Augie Fackler <raf@durin42.com> [Mon, 03 Apr 2017 17:34:24 -0400] rev 30894
Added signature for changeset ed5b25874d99
Augie Fackler <raf@durin42.com> [Mon, 03 Apr 2017 17:34:22 -0400] rev 30893
Added tag 4.1.2 for changeset ed5b25874d99
Gregory Szorc <gregory.szorc@gmail.com> [Thu, 23 Mar 2017 19:54:59 -0700] rev 30892
changegroup: store old heads as a set
Previously, the "oldheads" variable was a list. On a repository at
Mozilla with 46,492 heads, profiling revealed that list membership
testing was dominating execution time of applying small changegroups.
This patch converts the list of old heads to a set. This makes
membership testing significantly faster. On the aforementioned
repository with 46,492 heads:
$ hg unbundle <file with 1 changeset>
before: 18.535s wall
after: 1.303s
Consumers of this variable only check for truthiness (`if oldheads`),
length (`len(oldheads)`), and (most importantly) item membership
(`h not in oldheads` - which occurs twice). So, the change to a set
should be safe and suitable for stable.
The practical effect of this change is that changegroup application
and related operations (like `hg push`) no longer exhibit an O(n^2)
CPU explosion as the number of heads grows.
Simon Farnsworth <simonfar@fb.com> [Mon, 20 Mar 2017 04:36:55 -0700] rev 30891
subrepo: move prompts out of the if (issue5505)
Prompts weren't available in the else clause
Gregory Szorc <gregory.szorc@gmail.com> [Thu, 16 Mar 2017 12:33:15 -0700] rev 30890
exchange: use v2 bundles for modern compression engines (issue5506)
Previously, `hg bundle zstd` on a non-generaldelta repo would
attempt to use a v1 bundle. This would fail because zstd is not
supported on v1 bundles.
This patch changes the behavior to automatically use a v2 bundle
when the user explicitly requests a bundlespec that is a compression
engine not supported on v1. If the bundlespec is <engine>-v1, it is
still explicitly rejected because that request cannot be fulfilled.
Gregory Szorc <gregory.szorc@gmail.com> [Thu, 16 Mar 2017 12:23:56 -0700] rev 30889
exchange: reject new compression engines for v1 bundles (issue5506)
Version 1 bundles only support a fixed set of compression engines.
Before this change, we would accept any compression engine for v1
bundles, even those that may not work on v1. This could lead to
an error.
We define a fixed set of compression engines known to work with v1
bundles and we add checking to ensure a newer engine (like zstd)
won't work with v1 bundles.
I also took the liberty of adding test coverage for unknown compression
names because I noticed we didn't have coverage of it before.
Augie Fackler <augie@google.com> [Tue, 07 Mar 2017 13:24:24 -0500] rev 30888
pycompat: verify sys.argv exists before forwarding it (issue5493)
ISAPI_WSGI doesn't set up sys.argv, so we have to look for the
attribute before assuming it exists.
Augie Fackler <raf@durin42.com> [Thu, 02 Mar 2017 20:19:45 -0500] rev 30887
Added signature for changeset 25703b624d27
Augie Fackler <raf@durin42.com> [Thu, 02 Mar 2017 20:19:43 -0500] rev 30886
Added tag 4.1.1 for changeset 25703b624d27
Augie Fackler <raf@durin42.com> [Thu, 02 Mar 2017 20:07:35 -0500] rev 30885
merge with i18n
Wagner Bruna <wbruna@softwareexpress.com.br> [Thu, 02 Mar 2017 11:08:20 -0300] rev 30884
i18n-pt_BR: synchronized with 7074589cf22a
FUJIWARA Katsunori <foozy@lares.dti.ne.jp> [Wed, 01 Mar 2017 16:43:22 +0900] rev 30883
i18n-ja: synchronized with 7074589cf22a
Yuya Nishihara <yuya@tcha.org> [Sat, 25 Feb 2017 12:48:50 +0900] rev 30882
worker: flush messages written by child processes before exit
I found some child outputs were lost while testing the previous patch. Since
os._exit() does nothing special, we need to do that explicitly.
Yuya Nishihara <yuya@tcha.org> [Sat, 25 Feb 2017 12:33:37 +0900] rev 30881
worker: add basic test to ensure child processes are managed well
This should catch the bug fixed by "worker: ignore meaningless exit status
indication returned by os.waitpid()."
Before, worker.py was untested since test repositories are relatively small.
eloimorlaas <eloimorlaas@fb.com> [Thu, 23 Feb 2017 11:23:17 -0800] rev 30880
record: update help to describe ui.interface
hg help record was not accurate if using another interface than the text one
Rishabh Madan <rishabhmadan96@gmail.com> [Sat, 25 Feb 2017 16:57:21 +0530] rev 30879
ui: replace obsolete default-push with default:pushurl (issue5485)
Default-push has been deprecated in favour of default:pushurl. But "hg clone" still
inserts this in every hgrc file it creates. This patch updates the message by replacing
default-push with default:pushurl and also makes the necessary changes to test files.
FUJIWARA Katsunori <foozy@lares.dti.ne.jp> [Sat, 25 Feb 2017 01:07:52 +0900] rev 30878
worker: ignore meaningless exit status indication returned by os.waitpid()
Before this patch, worker implementation assumes that os.waitpid()
with os.WNOHANG returns '(0, 0)' for still running child process. This
is explicitly specified as below in Python API document.
os.WNOHANG
The option for waitpid() to return immediately if no child
process status is available immediately. The function returns
(0, 0) in this case.
On the other hand, POSIX specification doesn't define the "stat_loc"
value returned by waitpid() with WNOHANG for such child process.
http://pubs.opengroup.org/onlinepubs/9699919799/functions/waitpid.html
CPython implementation for os.waitpid() on POSIX doesn't take any care
of this gap, and this may cause unexpected "exit status indication"
even on POSIX conformance platform.
For example, os.waitpid() with os.WNOHANG returns non-zero "exit
status indication" on FreeBSD. This implies os.kill() with own pid or
sys.exit() with non-zero exit code, even if no child process fails.
To ignore meaningless exit status indication returned by os.waitpid(),
this patch skips subsequent steps forcibly, if os.waitpid() returns 0
as pid.
This patch also arranges examination of 'p' value for readability.
FYI, there are some issues below about this behavior reported for
CPython.
https://bugs.python.org/issue21791
https://bugs.python.org/issue27808
Siddharth Agarwal <sid0@fb.com> [Mon, 13 Feb 2017 11:43:12 -0800] rev 30877
bundle2: fix assertion that 'compression' hasn't been set
`n.lower()` will return `compression`, not `Compression`.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Fri, 10 Feb 2017 18:20:58 +0100] rev 30876
wireproto: properly report server Abort during 'getbundle'
Previously Abort raised during 'getbundle' call poorly reported (HTTP-500 for
http, some scary messages for ssh). Abort error have been properly reported for
"push" for a long time, there is not reason to be different for 'getbundle'. We
properly catch such error and report them back the best way available. For
bundle, we issue a valid bundle2 reply (as expected by the client) with an
'error:abort' part. With bundle1 we do as best as we can depending of http or
ssh.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Fri, 10 Feb 2017 18:17:20 +0100] rev 30875
getbundle: cleanly handle remote abort during getbundle
bundle2 allow the server to report error explicitly. This was initially
implemented for push but there is not reason to not use it for pull too. This
changeset add logic similar to the one in 'unbundle' to the
client side of 'getbundle'. That logic make sure the error is properly reported
as "remote". This will allow the server side of getbundle to send clean "Abort"
message in the next changeset.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Fri, 10 Feb 2017 18:06:08 +0100] rev 30874
bundle1: fix bundle1-denied reporting for pull over ssh
Changeset b288fb2724bf introduced a config option to have the server deny pull
using bundle1. The original protocol has not really been design to allow that
kind of error reporting so some hack was used. It turned the hack only works on
HTTP and that ssh server hangs forever when this is used. After further
digging, there is no way to report the error in a unified way. Using `ooberror`
freeze ssh and raising 'Abort' makes HTTP return a HTTP-500 without further
details. So with sadness we implement a version that dispatch according to the
protocol used.
Now the error is properly reported, but we still have ungraceful abort after
that. The protocol do not allow anything better to happen using bundle1.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Fri, 10 Feb 2017 18:06:12 +0100] rev 30873
bundle-tests: operate from outside a repository
We are about to add a test for ssh pull/cloning being denied because of bundle1
usage. For this, it is cleaner to not operate from the clone using http. So we
update the test beforehand for clarity. This is more churns that what I'm happy
to see on stable, but the rests of the series is worth it in my opinion.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Fri, 10 Feb 2017 17:56:52 +0100] rev 30872
bundle1: display server abort hint during unbundle
The code was printing the abort message but not the hint. This is now fixed.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Fri, 10 Feb 2017 17:56:59 +0100] rev 30871
bundle1: fix bundle1-denied reporting for push over ssh
Changeset b288fb2724bf introduced a config option to have the server deny push
using bundle1. The original protocol has not really be design to allow such kind
of error reporting so some hack was used. It turned the hack only works on HTTP
and that ssh wire peer hangs forever when the same hack is used. After further
digging, there is no way to report the error in a unified way. Using 'ooberror'
freeze ssh and raising 'Abort' makes HTTP return a HTTP500 without further
details. So with sadness we implement a version that dispatch according to the
protocol used.
We also add a test for pushing over ssh to make sure we won't regress in the
future. That test show that the hint is missing, this is another bug fixed in
the next changeset.