Joey Hess [Tue, 4 Oct 2016 16:52:36 +0000 (12:52 -0400)]
Linux standalone: Add back the LOCPATH=/dev/null hack to avoid the system locale-archive being read.
Version mismatches between the system locale-archive and the glibc in the
bundle have been observed to cause git crashes.
Unfortunately, this causes locales to not be used in the linux standalone
bundle, as was the case until version 6.20160419.
glibc hardcodes the path to /usr/lib/locale/locale-archive and does not
let an environment variable cause a different locale-archive file to be used.
The only other option to include locales in the bundle would be to include
exploded locale definition directories in the bundle for a number of
locales, generated by localedef. But these take at least 300 kb per locale,
and there are a great many locales; it would be hundreds of megabytes to
include them all.
(Hmm, we could include localdef in the bundle, and check LANG in runshell
and compile the locale directories on the fly. This would need
/usr/share/i18n/ and /usr/lib/locale-archive to be included in the bundle.
It's.. doable.)
I know this is going to once again cause users of the bundle to complain
that eg, ls doesn't show their unicode filenames right. Better than strange
crashes though.
How to deal with files that change status from "precious, please keep n copies" to "junk, please delete it from everywhere you find it, now and forever".
allow multiple concurrent external special remote processes
Multiple external special remote processes for the same remote will be
started as needed when using -J.
This should not beak any existing external special remotes, because running
multiple git-annex commands at the same time could already start multiple
processes for the same external special remotes.
This sped up git annex find --not --in web from 6.64s to 5.69s.
The optimised parser is probably more like 50% faster than the general one
it replaced.
Optimisations to git-annex branch query and setting, avoiding repeated copies of the environment.
Speeds up commands like "git-annex find --in remote" by over 50%.
Profiling showed that adjustGitEnv was 21% of the time and 37% of the
allocations of that command. It copied the environment each time with
getEnvironment.
The only repeated use of adjustGitEnv is in withIndexFile, which tends to
be run at least once per file. So, it was optimised by keeping a cache of
the environment, which can be reused.
There could be other better ways to optimise this. Maybe get the while
environment once at startup. But, then it would have to be serialized back
out each time running a child process, so I doubt that would be a net win.
It might be better to cache a version of the environment that is
pre-modified to use .git-annex/index. But, profiling doesn't show that
modifying the enviroment is taking any significant time.
Optimisations to time it takes git-annex to walk working tree and find files to work on. Sped up by around 18%.
key2file and file2key were top cost centers according to profiling.
The repeated use of replace was not efficient. This new approach is quite a
lot more efficient.
This commit was sponsored by Denis Dzyubenko on Patreon.
This broke the i386ancient autobuilder due to its use of
--flag git-annex:XMPP --flag=git-annex:dbus
-- Failure when adding dependencies:
fdo-notify: needed ((>=0.3)), stack configuration has no specified version
(latest applicable is 0.3.1)
gnutls: needed ((>=0.1.4)), stack configuration has no specified version
(latest applicable is 0.2)
network-protocol-xmpp: needed (-any), stack configuration has no specified
version (latest applicable is 0.4.8)
OSX autobuilder also seems hosed by it, so too soon.
fix bugs in handing of deep branches with sync and adjusted branches
* sync: Previously, when run in a branch with a slash in its name,
such as "foo/bar", the sync branch was "synced/bar". That conflicted
with the sync branch used for branch "bar", so has been changed to
"synced/foo/bar".
* adjust: Previously, when adjusting a branch with a slash in its name,
such as "foo/bar", the adjusted branch was "adjusted/bar(unlocked)".
That conflicted with the adjusted branch used for branch "bar",
so has been changed to "adjusted/foo/bar(unlocked)"
* Also, running sync in an adjusted branch did not correctly sync
changes back to the parent branch when it had a slash in its name.
This bug has been fixed.
Eliminate use of Git.Ref.under and Git.Ref.basename; using
Git.Ref.underBase and Git.Ref.base make everything handle deep branches
correctly.
Probably noone was adjusting deep branches, and v6 is still experimental
anyway, so I'm not going to worry about the mess that was left by that bug.
In the case of git-annex sync, using a fixed git-annex with an old unfixed
one will mean they use different sync branches for a deep branch, and so
they may stop syncing until the old one is upgraded. However, that's only
a problem when syncing between repositories without going via a central
bare repository. Added a warning about this to the CHANGELOG, but it's
probably not going to affect many people at all.