Commit Graph

51439 Commits

Author SHA1 Message Date
Stepan Kasal
a571340fbc tests: turn off git-daemon tests if FIFOs are not available
Signed-off-by: Stepan Kasal <kasal@ucw.cz>
2015-06-25 23:13:25 +02:00
Karsten Blees
579ce5fefa t800[12]: work around MSys limitation
MSys works very hard to convert Unix-style paths into DOS-style ones.
*Very* hard.

So hard, indeed, that

	git blame -L/hello/,/green/

is translated into something like

	git blame -LC:/msysgit/hello/,C:/msysgit/green/

As seen in msys_p2w in src\msys\msys\rt\src\winsup\cygwin\path.cc, line
3204ff:

	case '-':
	  //
	  // here we check for POSIX paths as attributes to a POSIX switch.
	  //
	...

seemingly absolute POSIX paths in single-letter options get expanded by
msys.dll unless they contain '=' or ';'.

So a quick and very dirty fix is to use '-L/;*evil/'. (Using an equal sign
works only when it is before a comma, so in the above example, /=*green/
would still be converted to a DOS-style path.)

Commit-message-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-25 23:13:25 +02:00
Pat Thoyts
7950162cd1 t0008: avoid absolute path on Windows as colon is used in the tests
The test separator char is a colon which means any absolute paths on windows
confuse the tests that use global_excludes.

Suggested-by: Karsten Blees <karsten.blees@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
2015-06-25 23:13:25 +02:00
Johannes Schindelin
ea846ce681 Work around a problem identified by BuildHive
Apparently the signal handling is not quite correct in the fsckobject
handling (most likely we rely on a side effect that lets us still output
some message after receiving a signal 13 but in the BuildHive setup this
fails intermittently).

As a consequence, the push in t5504 does fail as expected, but fails to
output anything (unexpected). Since this is good enough for now, let's
handle an empty output as success, too.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-25 23:13:25 +02:00
Sverre Rabbelier
60b175f96a t9350: point out that refs are not updated correctly
This happens only when the corresponding commits are not exported in
the current fast-export run. This can happen either when the relevant
commit is already marked, or when the commit is explicitly marked
as UNINTERESTING with a negative ref by another argument.

This breaks fast-export basec remote helpers.

Signed-off-by: Sverre Rabbelier <srabbelier@gmail.com>
2015-06-25 23:13:24 +02:00
Johannes Schindelin
3553dbae90 Handle new t1501 test case properly with MinGW
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-25 23:13:24 +02:00
Johannes Schindelin
26f3888bc2 Start the merging-rebase to v2.4.5
This commit starts the rebase of da509e1 to 3fa6dbe
2015-06-25 23:09:26 +02:00
Junio C Hamano
0df0541bf1 Git 2.4.5
Signed-off-by: Junio C Hamano <gitster@pobox.com>
v2.4.5
2015-06-25 11:03:05 -07:00
Junio C Hamano
ebf4fc90da Merge branch 'sg/merge-summary-config' into maint
Doc updates.

* sg/merge-summary-config:
  Documentation: include 'merge.branchdesc' for merge and config as well
2015-06-25 11:02:17 -07:00
Junio C Hamano
5fd72277d4 Merge branch 'jk/make-fix-dependencies' into maint
Build clean-up.

* jk/make-fix-dependencies:
  Makefile: silence perl/PM.stamp recipe
  Makefile: avoid timestamp updates to GIT-BUILD-OPTIONS
  Makefile: drop dependency between git-instaweb and gitweb
2015-06-25 11:02:16 -07:00
Junio C Hamano
68538cf954 Merge branch 'sb/pack-protocol-mention-smart-http' into maint
Doc updates.

* sb/pack-protocol-mention-smart-http:
  Documentation/technical/pack-protocol: mention http as possible protocol
2015-06-25 11:02:14 -07:00
Junio C Hamano
f2bd231414 Merge branch 'jk/die-on-bogus-worktree-late' into maint
The setup code used to die when core.bare and core.worktree are set
inconsistently, even for commands that do not need working tree.

* jk/die-on-bogus-worktree-late:
  setup_git_directory: delay core.bare/core.worktree errors
2015-06-25 11:02:13 -07:00
Junio C Hamano
6356003c80 Merge branch 'pt/pull-tags-error-diag' into maint
There was a dead code that used to handle "git pull --tags" and
show special-cased error message, which was made irrelevant when
the semantics of the option changed back in Git 1.9 days.

* pt/pull-tags-error-diag:
  pull: remove --tags error in no merge candidates case
2015-06-25 11:02:12 -07:00
Junio C Hamano
6998d890c7 Merge branch 'jk/color-diff-plain-is-context' into maint
"color.diff.plain" was a misnomer; give it 'color.diff.context' as
a more logical synonym.

* jk/color-diff-plain-is-context:
  diff.h: rename DIFF_PLAIN color slot to DIFF_CONTEXT
  diff: accept color.diff.context as a synonym for "plain"
2015-06-25 11:02:11 -07:00
Junio C Hamano
c5baf18a40 Merge branch 'jk/diagnose-config-mmap-failure' into maint
The configuration reader/writer uses mmap(2) interface to access
the files; when we find a directory, it barfed with "Out of memory?".

* jk/diagnose-config-mmap-failure:
  xmmap(): drop "Out of memory?"
  config.c: rewrite ENODEV into EISDIR when mmap fails
  config.c: avoid xmmap error messages
  config.c: fix mmap leak when writing config
  read-cache.c: drop PROT_WRITE from mmap of index
2015-06-25 11:02:11 -07:00
Junio C Hamano
c53312583b Merge branch 'jk/squelch-missing-link-warning-for-unreachable' into maint
Recent "git prune" traverses young unreachable objects to safekeep
old objects in the reachability chain from them, which sometimes
caused error messages that are unnecessarily alarming.

* jk/squelch-missing-link-warning-for-unreachable:
  suppress errors on missing UNINTERESTING links
  silence broken link warnings with revs->ignore_missing_links
  add quieter versions of parse_{tree,commit}
2015-06-25 11:02:10 -07:00
Junio C Hamano
f249409b6b Merge branch 'mm/rebase-i-post-rewrite-exec' into maint
"git rebase -i" fired post-rewrite hook when it shouldn't (namely,
when it was told to stop sequencing with 'exec' insn).

* mm/rebase-i-post-rewrite-exec:
  t5407: use <<- to align the expected output
  rebase -i: fix post-rewrite hook with failed exec command
  rebase -i: demonstrate incorrect behavior of post-rewrite
2015-06-25 11:02:09 -07:00
Johannes Schindelin
75cfeeadcf squash! mingw: keep trailing slashes when switching directories
Actually, we should always keep the trailing slashes as both users of
`normalize_ntpath()` require them: `_wchdir()` wants the slash in
particular when changing the directory to a drive root, and
`readlink()`'s semantics require a trailing slash for symbolic links
pointing to directories.

This fixes https://github.com/git-for-windows/git/issues/210

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-20 10:01:17 +00:00
Johannes Schindelin
381a90a597 mingw: keep trailing slashes when switching directories
This is needed so that `_wchdir()` can be used with drive root
directories, e.g. C:\ (`_wchdir("C:")` fails to switch the directory
to the root directory).

This fixes https://github.com/msysgit/git/issues/359 (in Git for Windows
2.x only, though).

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-19 21:00:46 +00:00
Johannes Schindelin
18e8d51113 git-wrapper: leave the working directory alone by default
The idea of `git-bash.exe` automatically running the Git Bash in the
home directory was to support the start menu item `Git Bash` (which
should not start in C:\Program Files\Git, but in $HOME), and to make
that behavior consistent with double-clicking in `git-bash.exe`
portable Git.

However, it turns out that one of the main use cases of portable Git is
to run the Git Bash in GitHub for Windows, and it should start in the
top-level directory of a given project. Therefore, the concern to keep
double-clicking `git-bash.exe` consistent with the start menu item was
actually unfounded.

As to the start menu item: it can easily be changed to launch
`git-bash.exe` with a command-line option. So let's introduce the
--cd-to-home option for that purpose.

As a bonus, the Git wrapper can now also serve as a drop-in redirector
/bin/bash.exe to provide backwards-compatibility of Git for Windows 2.x
with 1.x: some 3rd-party software expects to find that executable there,
and it also expects it to leave the working directory unchanged.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-19 20:20:42 +00:00
Johannes Schindelin
e96e074c25 Merge branch 'program-data-config'
This branch introduces support for reading the "Windows-wide" Git
configuration from `%PROGRAMDATA%\Git\config`. As these settings are
intended to be shared between *all* Git-related software, that config
file takes an even lower precedence than `$(prefix)/etc/gitconfig`.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:57 +02:00
Johannes Schindelin
8852f60193 Merge branch 'git-wrapper--command'
This topic branch adds the --command=<command> option that allows
starting the Git Bash (or Git CMD) with different terminal emulators
than the one encoded via embedded string resources.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:56 +02:00
Johannes Schindelin
ddbb590723 Windows: add support for a Windows-wide configuration
Between the libgit2 and the Git for Windows project, there has been a
discussion how we could share Git configuration to avoid duplication (or
worse: skew).

Earlier, libgit2 was nice enough to just re-use Git for Windows'

	C:\Program Files (x86)\Git\etc\gitconfig

but with the upcoming Git for Windows 2.x, there would be more paths to
search, as we will have 64-bit and 32-bit versions, and the
corresponding config files will be in %PROGRAMFILES%\Git\mingw64\etc and
...\mingw32\etc, respectively.

Worse: there are portable Git for Windows versions out there which live
in totally unrelated directories, still.

Therefore we came to a consensus to use `%PROGRAMDATA%\Git\config` as the
location for shared Git settings that are of wider interest than just Git
for Windows.

On XP, there is no %PROGRAMDATA%, therefore we need to use
"%ALLUSERSPROFILE%\Application Data\Git\config" in those setups.

Of course, the configuration in `%PROGRAMDATA%\Git\config` has the
widest reach, therefore it must take the lowest precedence, i.e. Git for
Windows can still override settings in its `etc/gitconfig` file.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:56 +02:00
Johannes Schindelin
b5e2c61b65 Merge pull request #159 from dscho/vagrant
Add Vagrant support (easy Linux VM setup)
2015-06-17 10:25:56 +02:00
Johannes Schindelin
f96fb70a8e git-wrapper: allow overriding the command to spawn via command-line args
By embedding string resources into the Git wrapper executable, it
can be configured to execute custom commands (after setting up the
environment in the way required for Git for Windows to work properly).
This feature is used e.g. for `git-bash.exe` which launches a Bash in
the configured terminal window.

Here, we introduce command-line options to override those string
resources. That way, a user can call `git-bash.exe` (which is a copy of
the Git wrapper with `usr\bin\bash.exe --login -i` embedded as string
resource) with command-line options that will override what command is
run.

ConEmu, for example, might want to call

	...\git-bash.exe --needs-console --no-hide --minimal-search-path ^
		--command=usr\\bin\\bash.exe --login -i

In particular, the following options are supported now:

--command=<command-line>::
	Executes `<command-line>` instead of the embedded string resource

--[no-]minimal-search-path::
	Ensures that only `/cmd/` is added to the `PATH` instead of
	`/mingw??/bin` and `/usr/bin/`, or not

--[no-]needs-console::
	Ensures that there is a Win32 console associated with the spawned
	process, or not

--[no-]hide::
	Hides the console window, or not

Helped-by: Eli Young <elyscape@gmail.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:55 +02:00
Karsten Blees
1df79b99f4 config.c: create missing parent directories when modifying config files
'git config' (--add / --unset etc.) automatically creates missing config
files. However, it fails with a misleading error message "could not lock
config file" if the parent directory doesn't exist.

Also create missing parent directories.

This is particularly important when calling

	git config -f /non/existing/directory/config ...

Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:55 +02:00
Johannes Schindelin
71690ca6f3 git wrapper: auto-grow buffer in expand_variables()
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:55 +02:00
Karsten Blees
692c7c7946 config: factor out repeated code
Factor out near identical per-file logic.

Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:55 +02:00
Johannes Schindelin
2ff0b3c1cd git wrapper: refactor @@VAR@@ expansion into its own function
We will enhance the function in the next commit to support @@VAR@@
expansion in the upcoming `--command=<command>` option.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:54 +02:00
Johannes Schindelin
c7205d0ad2 git wrapper: refactor extraction of 1st arg into its own function
This will be reused by the upcoming `--command=<command>` option.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:54 +02:00
Johannes Schindelin
54dc767232 Merge pull request #156 from kblees/kb/symlinks
Symlink support
2015-06-17 10:25:53 +02:00
Johannes Schindelin
6da5733d33 Support Vagrant: quick & easy Linux virtual machine setup
When developing Git for Windows, we always have to ensure that we do not
break any non-Windows platforms, e.g. by introducing Windows-specific code
into the platform-independent source code.

At other times, it is necessary to test whether a bug is Windows-specific
or not, in order to send the bug report to the correct place. Having
access to a Linux-based Git comes in really handy in such a situation.

Vagrant offers a painless way to install and use a defined Linux
development environment on Windows (and other Operating Systems). We offer
a Vagrantfile to that end for two reasons:

1) To allow Windows users to gain the full power of Linux' Git

2) To offer users an easy path to verify that the issue they are about
   to report is really a Windows-specific issue; otherwise they would
   need to report it to git@vger.kernel.org instead.

Using it is easy: Download and install https://www.virtualbox.org/, then
download and install https://www.vagrantup.com/, then direct your
command-line window to the Git source directory containing the Vagrantfile
and run the commands:

	vagrant up
	vagrant ssh

See https://github.com/git-for-windows/git/wiki/Vagrant for details.

As part of switching Git for Windows' development environment from msysGit
to the MSys2-based Git SDK, this Vagrantfile was copy-edited from msysGit:

	https://github.com/msysgit/msysgit/blob/0be8f2208/Vagrantfile

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:53 +02:00
Johannes Schindelin
3a67653bc9 Merge 'git-wrapper' into HEAD
Use msysGit's `git-wrapper` instead of the builtins. This works around
two issues:

- when the file system does not allow hard links, we would waste over
  800 megabyte by having 109 copies of a multi-megabyte executable

- even when the file system allows hard links, the Windows Explorer
  counts the disk usage as if it did not. Many users complained about
  Git for Windows using too much space (when it actually did not). We
  can easily avoid those user complaints by merging this branch.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:53 +02:00
Johannes Schindelin
2e1354290a Merge 'poll_inftim' into HEAD
This was originally 'pull request #330 from ethomson/poll_inftim' in
msysgit/git.

poll: honor the timeout on Win32

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:53 +02:00
Johannes Schindelin
664cc6840b Merge 'non-win-fixes' into HEAD 2015-06-17 10:25:52 +02:00
Johannes Schindelin
7dced585f4 Merge 'sideband-bug' into HEAD
This works around the push-over-git-protocol issues pointed out in
https://github.com/msysgit/git/issues/101.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:52 +02:00
Johannes Schindelin
f6278b4e00 Merge 'readme' into HEAD
Add a README.md for GitHub goodness.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:52 +02:00
Johannes Schindelin
a9624f25c5 Merge 'fix-is-exe' into HEAD 2015-06-17 10:25:52 +02:00
Johannes Schindelin
0d4ff8eee3 Merge 'fix-externals' into HEAD 2015-06-17 10:25:51 +02:00
Johannes Schindelin
4f35fd5f69 Merge 'remote-hg-prerequisites' into HEAD
These fixes were necessary for Sverre Rabbelier's remote-hg to work,
but for some magic reason they are not necessary for the current
remote-hg. Makes you wonder how that one gets away with it.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:51 +02:00
Johannes Schindelin
5c3f18af6f Merge 'win-tests-fixes' into HEAD 2015-06-17 10:25:51 +02:00
Johannes Schindelin
ead8c69481 Merge 'msys2' into HEAD
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2015-06-17 10:25:51 +02:00
Johannes Schindelin
f23091fc28 Merge 'pull-rebase-interactive' into HEAD 2015-06-17 10:25:50 +02:00
Johannes Schindelin
12aec15e1a Merge 'jberezanski/wincred-sso-r2' into HEAD 2015-06-17 10:25:50 +02:00
Johannes Schindelin
62959577f9 Merge 'gitk' into HEAD 2015-06-17 10:25:50 +02:00
Johannes Schindelin
f2fb7b4862 Merge 'git-gui' into HEAD 2015-06-17 10:25:50 +02:00
Johannes Schindelin
d60108b224 Merge 'criss-cross-merge' into HEAD 2015-06-17 10:25:49 +02:00
Johannes Schindelin
1f6864519e Merge 'hide-dotgit' into HEAD 2015-06-17 10:25:49 +02:00
Johannes Schindelin
71e3fd5847 Merge 'unicode' into HEAD 2015-06-17 10:25:49 +02:00
Karsten Blees
ce587acc3c t7800: configure $(pwd) for posix-paths on MINGW
In test #49, $(pwd) must match $(readlink), which is an MSys utility.

Signed-off-by: Karsten Blees <blees@dcon.de>
2015-06-17 10:25:48 +02:00