Commit graph

536 commits

Author SHA1 Message Date
Joey Hess
10b5f79e2d
fix empty tree import when directory does not exist
Fix behavior when importing a tree from a directory remote when the
directory does not exist. An empty tree was imported, rather than the
import failing. Merging that tree would delete every file in the
branch, if those files had been exported to the directory before.

The problem was that dirContentsRecursive returned [] when the directory
did not exist. Better for it to throw an exception. But in commit
74f0d67aa3 back in 2012, I made it never
theow exceptions, because exceptions throw inside unsafeInterleaveIO become
untrappable when the list is being traversed.

So, changed it to list the contents of the directory before entering
unsafeInterleaveIO. So exceptions are thrown for the directory. But still
not if it's unable to list the contents of a subdirectory. That's less of a
problem, because the subdirectory does exist (or if not, it got removed
after being listed, and it's ok to not include it in the list). A
subdirectory that has permissions that don't allow listing it will have its
contents omitted from the list still.

(Might be better to have it return a type that includes indications of
errors listing contents of subdirectories?)

The rest of the changes are making callers of dirContentsRecursive
use emptyWhenDoesNotExist when they relied on the behavior of it not
throwing an exception when the directory does not exist. Note that
it's possible some callers of dirContentsRecursive that used to ignore
permissions problems listing a directory will now start throwing exceptions
on them.

The fix to the directory special remote consisted of not making its
call in listImportableContentsM use emptyWhenDoesNotExist. So it will
throw an exception as desired.

Sponsored-by: Joshua Antonishen on Patreon
2023-08-15 12:57:41 -04:00
Joey Hess
fb640bc2f4
support building with unix-compat 0.7
It removed System.PosixCompat.User.
2023-08-01 15:17:43 -04:00
Joey Hess
ff118cdfe8
fix formatting of \ continued shell examples in man pages
doc/git-annex-filter-branch.mdwn and one other file have these, and no
other files have \ at the end of line, so this should be safe
2023-07-13 19:55:38 -04:00
Joey Hess
96c372ac85
document how to include= a path with a space in it
POSIX character classes allowed in globs was a surprise, but just
happened to fall out of the implementation in a way that seems
to behave correctly.

mdwn2man has to be tweaked to render the example properly.
The line I modified is the one that strips ikiwiki wikilinks out of the
man page.

Sponsored-by: Graham Spencer on Patreon
2023-05-15 16:02:06 -04:00
Yaroslav Halchenko
84b0a3707a
Apply codespell -w throughout 2023-03-17 15:14:58 -04:00
Joey Hess
d8feda7a2f
added arm64-ancient build
Added arm64 build for ancient kernels, needed to support Android phones
whose kernels are too old to support kernels used by the current arm64
build.

Updated Android/git-annex-install to use it. (Also made it use i386-ancient
because that seems like a good idea.)

Sponsored-by: Noam Kremen on Patreon
2023-03-10 11:59:03 -04:00
Joey Hess
54ad1b4cfb
Windows: Support long filenames in more (possibly all) of the code
Works around this bug in unix-compat:
https://github.com/jacobstanley/unix-compat/issues/56
getFileStatus and other FilePath using functions in unix-compat do not do
UNC conversion on Windows.

Made Utility.RawFilePath use convertToWindowsNativeNamespace to do the
necessary conversion on windows to support long filenames.

Audited all imports of System.PosixCompat.Files to make sure that no
functions that operate on FilePath were imported from it. Instead, use
the equvilants from Utility.RawFilePath. In particular the
re-export of that module in Common had to be removed, which led to lots
of other changes throughout the code.

The changes to Build.Configure, Build.DesktopFile, and Build.TestConfig
make Utility.Directory not be needed to build setup. And so let it use
Utility.RawFilePath, which depends on unix, which cannot be in
setup-depends.

Sponsored-by: Dartmouth College's Datalad project
2023-03-01 15:55:58 -04:00
Joey Hess
d079fb0578
work around ldd crash on arm64 autobuilder
This does mean it has to run ldd once per executable, which is actually
quite a number of times, so will be a bit slower.

Sponsored-by: Luke Shumaker on Patreon
2023-01-28 14:26:01 -04:00
Joey Hess
bc980815ee
fix build of helper program
broken by ba7ecbc6a9

Sponsored-by: Svenne Krap on Patreon
2022-11-03 14:11:49 -04:00
Joey Hess
14f7a386f0
Make git-annex enable-tor work when using the linux standalone build
Clean the standalone environment before running the su command
to run "sh". Otherwise, PATH leaked through, causing it to run
git-annex.linux/bin/sh, but GIT_ANNEX_DIR was not set,
which caused that script to not work:

[2022-10-26 15:07:02.145466106] (Utility.Process) process [938146] call: pkexec ["sh","-c","cd '/home/joey/tmp/git-annex.linux/r' && '/home/joey/tmp/git-annex.linux/git-annex' 'enable-tor' '1000'"]
/home/joey/tmp/git-annex.linux/bin/sh: 4: exec: /exe/sh: not found

Changed programPath to not use GIT_ANNEX_PROGRAMPATH,
but instead run the scripts at the top of GIT_ANNEX_DIR.
That works both when the standalone environment is set up, and when it's
not.

Sponsored-by: Kevin Mueller on Patreon
2022-10-26 15:45:08 -04:00
Joey Hess
ba7ecbc6a9
avoid flushing keys db queue after each Annex action
The flush was only done Annex.run' to make sure that the queue was flushed
before git-annex exits. But, doing it there means that as soon as one
change gets queued, it gets flushed soon after, which contributes to
excessive writes to the database, slowing git-annex down.
(This does not yet speed git-annex up, but it is a stepping stone to
doing so.)

Database queues do not autoflush when garbage collected, so have to
be flushed explicitly. I don't think it's possible to make them
autoflush (except perhaps if git-annex sqitched to using ResourceT..).
The comment in Database.Keys.closeDb used to be accurate, since the
automatic flushing did mean that all writes reached the database even
when closeDb was not called. But now, closeDb or flushDb needs to be
called before stopping using an Annex state. So, removed that comment.

In Remote.Git, change to using quiesce everywhere that it used to use
stopCoProcesses. This means that uses on onLocal in there are just as
slow as before. I considered only calling closeDb on the local git remotes
when git-annex exits. But, the reason that Remote.Git calls stopCoProcesses
in each onLocal is so as not to leave git processes running that have files
open on the remote repo, when it's on removable media. So, it seemed to make
sense to also closeDb after each one, since sqlite may also keep files
open. Although that has not seemed to cause problems with removable
media so far. It was also just easier to quiesce in each onLocal than
once at the end. This does likely leave performance on the floor, so
could be revisited.

In Annex.Content.saveState, there was no reason to close the db,
flushing it is enough.

The rest of the changes are from auditing for Annex.new, and making
sure that quiesce is called, after any action that might possibly need
it.

After that audit, I'm pretty sure that the change to Annex.run' is
safe. The only concern might be that this does let more changes get
queued for write to the db, and if git-annex is interrupted, those will be
lost. But interrupting git-annex can obviously already prevent it from
writing the most recent change to the db, so it must recover from such
lost data... right?

Sponsored-by: Dartmouth College's Datalad project
2022-10-12 14:12:23 -04:00
Joey Hess
ff0490c7c6
fix build 2022-09-05 13:20:23 -04:00
Joey Hess
a93163d6f7
optimise linker in linux standalone tarballs
Trick the linker into not doing unncessary work searching for optimised
libraries that are not present, by symlinking the directories where
optimised libs would be to the main lib dir.

This reduces the ENOENT of git-annex init by about 1/2. The linker always
finds the files where it looks first time now. I have not looked at what
the wall clock speedup might be, it's probably rather small.

If a x86-64-v5 comes to be, the list will need to be extended. And there
may be other directories used on some machines that I have missed. Not done
for arm64 yet, or any uncommon architectures.

Sponsored-by: Dartmouth College's Datalad project
2022-08-30 15:20:04 -04:00
Joey Hess
a6ff335f0b
fix build 2022-07-25 14:10:30 -04:00
Joey Hess
77048906a4
assume @RPATH libs are present
Previous change caused them to be skipped. Probably when @RPATH is
included in the path, it's a path to a file that actually exists.

Sponsored-by: Dartmouth College's Datalad project
2022-01-03 15:05:15 -04:00
Joey Hess
abeb7dcdde
skip OSX libraries that do not exist in disk due to new cache
Addresses the build problem discussed here:
https://github.com/datalad/git-annex/issues/95

Sponsored-by: Dartmouth College's Datalad project
2022-01-03 13:19:33 -04:00
Joey Hess
5c105a8540
test borg special remote
Sponsored-by: Boyd Stephen Smith Jr. on Patreon
2021-09-03 13:18:07 -04:00
Joey Hess
35bbf4514c
fix build 2021-09-03 12:36:42 -04:00
Joey Hess
e17342b2a0
Run cp -a with --no-preserve=xattr, to avoid problems with copied xattrs
Including them breaking permissions setting on some NFS servers.

Sponsored-by: Dartmouth College's Datalad project
2021-08-27 13:09:34 -04:00
Joey Hess
5990942b6c
don't use changelog version in commit message
changelog may have a new unreleased version open already
2021-08-09 12:31:48 -04:00
Joey Hess
122f166f2a
update for OSX catalina build 2021-07-14 12:22:34 -04:00
Joey Hess
898a677bce
fix build on windows
Sponsored-by: Dartmouth College's Datalad project
2021-07-01 12:52:33 -04:00
Joey Hess
9cb5ed2446
avoid using cp --reflink=auto on windows
Sponsored-by: Dartmouth College's Datalad project
2021-06-28 12:51:40 -04:00
Joey Hess
a942ed4bb9
Windows: Correct the path to the html help file for 64 bit build. 2021-02-24 13:19:42 -04:00
Joey Hess
f36a0c1b13
deal with cabal unpack not preserving execute bit 2021-02-08 14:32:24 -04:00
Joey Hess
1b63132ca3
add searchPathContents
And rename related functions for consistency.
2021-02-02 19:06:15 -04:00
Joey Hess
fe4725d66e
switch to createrepo_c
createrepo used python2 and got removed from debian
2020-12-25 16:32:03 -04:00
Joey Hess
72ec7da41b
OSXMkLibs: skip @loader_path
This may or may not make the OSX build work on a newer version of OSX
than the one that's currently being used for release builds. I have not
been able to find good docs about how exactly to get back from such a
value to the actual path to the library that the linker would use.
2020-12-15 12:47:24 -04:00
Joey Hess
fac5ccb2e2
add USE_SYSTEM_LIBS=1 for osxapp build 2020-12-14 15:34:27 -04:00
Joey Hess
1a1d671dd2
fix build 2020-11-27 13:02:43 -04:00
Joey Hess
21ab496f6b
fix build 2020-11-16 09:56:03 -04:00
Joey Hess
dd52d8ebdc
update after RawFilePath transition 2020-11-09 12:12:25 -04:00
Joey Hess
e505c03bcc
more RawFilePath conversion
nukeFile replaced with removeWhenExistsWith removeLink, which allows
using RawFilePath. Utility.Directory cannot use RawFilePath since setup
does not depend on posix.

This commit was sponsored by Graham Spencer on Patreon.
2020-10-29 10:50:29 -04:00
Joey Hess
d6e94a6b2e
got configure working after Utility.Path ByteString conversion
Had to split out some modules because getWorkingDirectory needs unix,
which is not a build-dep of configure.

This commit was sponsored by Brock Spratlen on Patreon.
2020-10-28 15:01:19 -04:00
Joey Hess
e85be1f4fe
fix build and add build example 2020-10-26 13:45:38 -04:00
Joey Hess
3baef37caf
make windows installer builder include libmagic files when present, but skip otherwise 2020-10-26 13:42:43 -04:00
Joey Hess
a7d11aa3ab
include libmagic in windows installer
This is from https://raw.githubusercontent.com/datalad/datalad-extensions/master/resources/git-annex-magicBundle.patch
which was written by John Thorvald Wodder II.

I have converted line endings from windows, and fixed indentation to use
tabs.
2020-10-26 13:24:37 -04:00
Joey Hess
ae325018eb
fix build failure on i386-ancient
Standalone: tmp/git-annex.linux/usr/lib/i386-linux-gnu/i686: renameFile: inappropriate type (is a directory)
2020-08-10 16:49:09 -04:00
Joey Hess
6cb08876f0
make /usr/lib consolidation more robust
Directory enumeration could happen in either order, and also there could
be some other directories so the two to combine are not adjacent.
2020-08-10 12:20:32 -04:00
Joey Hess
5feaef773d
squelch build warnings on osx 2020-08-07 13:49:56 -04:00
Joey Hess
db45fb7048
osx build fix 2020-08-07 13:30:11 -04:00
Joey Hess
24e5bca2e0
osx build fix 2020-08-07 13:15:36 -04:00
Joey Hess
1c13745112
osx build fix 2020-08-07 12:59:47 -04:00
Joey Hess
93d0abaaa6
osx build fix 2020-08-07 12:39:59 -04:00
Joey Hess
b2eaca27a7
fix osx build 2020-08-07 12:14:02 -04:00
Joey Hess
8bfa7990b5
make symlinks not hardlinks
hardlinks happened to work because it was a hard link to a symlink, but
it would have failed if eg git-annex-shell symlink did not exist,
perhaps because the Makefile was not run to build it. Or something like
that. Also, it seems like it worked by accident and the Makefile used to
ln -s.
2020-08-05 11:58:22 -04:00
Joey Hess
ec4e742066
create missing directory 2020-08-05 11:37:03 -04:00
Joey Hess
bdfe6f7420
fix build in clean tree
removeDirectoryRecursive fails if the directory DNE.
2020-08-03 19:43:21 -04:00
Joey Hess
88e5ebcda7
runshell LD_HWCAP_MASK=0 optimisation 2020-08-03 14:34:15 -04:00
Joey Hess
e62817c00d
move standalone building code out of Makefile and into Build.Standalone
This includes making Build.Standalone run LinuxMkLibs or OSXMkLibs
rather than doing that separately. Which is groundwork for a later
optimisation.

Also it simplified the code some.
2020-08-03 14:15:03 -04:00