gitweb.git
i18n: index-pack: use plural string instead of normal oneVasco Almeida Fri, 8 Apr 2016 20:02:39 +0000 (20:02 +0000)

i18n: index-pack: use plural string instead of normal one

Git could output "completed with 1 local objects", but in this
case using "object" instead of "objects" is the correct form.

Use Q_() instead of _().

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

configure: remove checking for HMAC_CTX_cleanupKazuki Yamaguchi Fri, 8 Apr 2016 16:22:16 +0000 (01:22 +0900)

configure: remove checking for HMAC_CTX_cleanup

We don't need it, as we no longer use HMAC_CTX_cleanup() directly.

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

imap-send: avoid deprecated TLSv1_method()Kazuki Yamaguchi Fri, 8 Apr 2016 16:22:15 +0000 (01:22 +0900)

imap-send: avoid deprecated TLSv1_method()

Use SSLv23_method always and disable SSL if needed.

TLSv1_method() function is deprecated in OpenSSL 1.1.0 and the compiler
emits a warning.

SSLv23_method() is also deprecated, but the alternative, TLS_method(),
is new in OpenSSL 1.1.0 so requires checking by configure. Stick to
SSLv23_method() for now (this is aliased to TLS_method()).

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

imap-send: check NULL return of SSL_CTX_new()Kazuki Yamaguchi Fri, 8 Apr 2016 16:22:14 +0000 (01:22 +0900)

imap-send: check NULL return of SSL_CTX_new()

SSL_CTX_new() may fail with return value NULL.

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

imap-send: use HMAC() function provided by OpenSSLKazuki Yamaguchi Fri, 8 Apr 2016 16:22:13 +0000 (01:22 +0900)

imap-send: use HMAC() function provided by OpenSSL

Fix compile errors with OpenSSL 1.1.0.

HMAC_CTX is made opaque and HMAC_CTX_cleanup is removed in OpenSSL
1.1.0. But since we just want to calculate one HMAC, we can use HMAC()
here, which exists since OpenSSL 0.9.6 at least.

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

set_worktree_head_symref(): fix error messageKazuki Yamaguchi Fri, 8 Apr 2016 08:03:07 +0000 (17:03 +0900)

set_worktree_head_symref(): fix error message

Emit an informative error when failed to hold lock of HEAD.

2233066e (refs: add a new function set_worktree_head_symref,
2016-03-27) added set_worktree_head_symref(), but this is missing a
call to unable_to_lock_message() after hold_lock_file_for_update()
fails, so it emits an empty error message:

% git branch -m oldname newname
error:
error: HEAD of working tree /path/to/wt is not updated
fatal: Branch renamed to newname, but HEAD is not updated!

Thanks to Eric Sunshine for pointing this out.

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

commit: do not ignore an empty message given by -m ''Jeff King Thu, 7 Apr 2016 19:56:26 +0000 (12:56 -0700)

commit: do not ignore an empty message given by -m ''

When f9568530 (builtin-commit: resurrect behavior for multiple -m
options, 2007-11-11) converted a "char *message" to "struct strbuf
message" to hold the messages given with the "-m" option, it
incorrectly changed the checks "did we get a message with the -m
option?" to "is message.len 0?". Later, we noticed one breakage
from this change and corrected it with 25206778 (commit: don't start
editor if empty message is given with -m, 2013-05-25).

However, "we got a message with -m, even though an empty one, so we
shouldn't be launching an editor" was not the only breakage.

* "git commit --amend -m '' --allow-empty", even though it looks
strange, is a valid request to amend the commit to have no
message at all. Due to the misdetection of the presence of -m on
the command line, we ended up keeping the log messsage from the
original commit.

* "git commit -m "$msg" -F file" should be rejected whether $msg is
an empty string or not, but due to the same bug, was not rejected
when $msg is empty.

* "git -c template=file -m "$msg"" should ignore the template even
when $msg is empty, but it didn't and instead used the contents
from the template file.

Correct these by checking have_option_m, which the earlier 25206778
introduced to fix the same bug.

Reported-by: Adam Dinwoodie <adam@dinwoodie.org>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

commit: --amend -m '' silently fails to wipe messageAdam Dinwoodie Wed, 6 Apr 2016 17:15:03 +0000 (18:15 +0100)

commit: --amend -m '' silently fails to wipe message

`git commit --amend -m ''` seems to be an unambiguous request to blank a
commit message, but it actually leaves the commit message as-is. That's
the case regardless of whether `--allow-empty-message` is specified, and
doesn't so much as drop a non-zero return code.

Add failing tests to show this behaviour.

Signed-off-by: Adam Dinwoodie <adam@dinwoodie.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

setup.c: do not feed NULL to "%.*s" even with precision 0Junio C Hamano Thu, 7 Apr 2016 19:38:18 +0000 (12:38 -0700)

setup.c: do not feed NULL to "%.*s" even with precision 0

A recent update 75faa45a (replace trivial malloc + sprintf / strcpy
calls with xstrfmt, 2015-09-24) rewrote

prepare an empty buffer
if (len)
append the first len bytes of "prefix" to the buffer
append "path" to the buffer

that computed "path", optionally prefixed by "prefix", into

xstrfmt("%.*s%s", len, prefix, path);

However, passing a NULL pointer to the printf(3) family of functions
to format it with %s conversion, even with the precision set to 0,
i.e.

xstrfmt("%.*s", 0, NULL)

yields undefined results, at least on some platforms.

Avoid this problem by substituting prefix with "" when len==0, as
prefix can legally be NULL in that case. This would mimick the
intent of the original code better.

Reported-by: Tom G. Christensen <tgc@jupiterrise.com>
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

send-email: do not load Data::DumperEric Wong Wed, 6 Apr 2016 20:25:38 +0000 (20:25 +0000)

send-email: do not load Data::Dumper

We never used Data::Dumper in this script. The only reference
of it was always commented out and removed over a decade ago in
commit 4bc87a28be020a6bf7387161c65ea3d8e4a0228b
("send-email: Change from Mail::Sendmail to Net::SMTP")

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

send-email: more meaningful Message-IDEric Wong Wed, 6 Apr 2016 20:07:14 +0000 (20:07 +0000)

send-email: more meaningful Message-ID

Using a YYYYmmddHHMMSS date representation is more meaningful to
humans, especially when used for lookups on NNTP servers or linking
to archive sites via Message-ID (e.g. mid.gmane.org or
mid.mail-archive.com). This timestamp format more easily gives a
reader of the URL itself a rough date of a linked message compared
to having them calculate the seconds since the Unix epoch.

Furthermore, having the MUA name in the Message-ID seems to be a
rare oddity I haven't noticed outside of git-send-email. We
already have an optional X-Mailer header field to advertise for
us, so extending the Message-ID by 15 characters can make for
unpleasant Message-ID-based URLs to archive sites.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

correct blame for files commited with CRLFTorsten Bögershausen Tue, 5 Apr 2016 19:23:54 +0000 (21:23 +0200)

correct blame for files commited with CRLF

git blame reports lines as not "Not Committed Yet" when they have
CRLF in the index, CRLF in the worktree and core.autocrlf is true.

Since commit c4805393 (autocrlf: Make it work also for un-normalized
repositories, 2010-05-12), files that have CRLF in the index are not
normalized at commit when core.autocrl is set.

Add a call to read_cache() early in fake_working_tree_commit(),
before calling convert_to_git().

Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

api-trace.txt: fix typoElia Pinto Tue, 5 Apr 2016 10:05:23 +0000 (10:05 +0000)

api-trace.txt: fix typo

The correct api is trace_printf_key(), not trace_print_key().

Also do not throw a random string at printf(3)-like function;
instead, feed it as a parameter that is fed to a "%s" conversion
specifier.

Signed-off-by: Elia Pinto <gitter.spiros@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Makefile: fix misdirected redirectionsJunio C Hamano Tue, 5 Apr 2016 07:02:14 +0000 (00:02 -0700)

Makefile: fix misdirected redirections

In general "echo 2>&1 $msg" to redirect a possible error message
that comes from 'echo' itself into the same standard output stream
$msg is getting written to does not make any sense; it is not like
we are expecting to see any errors out of 'echo' in these statements,
and even if it were the case, there is no reason to prevent the
error messages from being sent to the standard error stream.

These are clearly meant to send the argument given to echo to the
standard error stream as error messages. Correctly redirect by
saying "send what is written to the standard output to the standard
error", i.e. "1>&2" aka ">&2".

Signed-off-by: Junio C Hamano <gitster@pobox.com>

branch: fix shortening of non-remote symrefsJeff King Sun, 3 Apr 2016 04:14:26 +0000 (00:14 -0400)

branch: fix shortening of non-remote symrefs

Commit aedcb7d (branch.c: use 'ref-filter' APIs, 2015-09-23)
adjusted the symref-printing code to look like this:

if (item->symref) {
skip_prefix(item->symref, "refs/remotes/", &desc);
strbuf_addf(&out, " -> %s", desc);
}

This has three bugs in it:

1. It always skips past "refs/remotes/", instead of
skipping past the prefix associated with the branch we
are showing (so commonly we see "refs/remotes/" for the
refs/remotes/origin/HEAD symref, but the previous code
would skip "refs/heads/" when showing a symref it found
in refs/heads/.

2. If skip_prefix() does not match, it leaves "desc"
untouched, and we show whatever happened to be in it
(which is the refname from a call to skip_prefix()
earlier in the function).

3. If we do match with skip_prefix(), we stomp on the
"desc" variable, which is later passed to
add_verbose_info(). We probably want to retain the
original refname there (though it likely doesn't matter
in practice, since after all, one points to the other).

The fix to match the original code is fairly easy: record
the prefix to strip based on item->kind, and use it here.
However, since we already have a local variable named "prefix",
let's give the two prefixes verbose names so we don't
confuse them.

Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Windows: shorten code by re-using convert_slashes()Johannes Sixt Sat, 2 Apr 2016 19:03:14 +0000 (21:03 +0200)

Windows: shorten code by re-using convert_slashes()

Make a few more spots more readable by using the recently introduced,
Windows-specific helper.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Acked-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

git-format-patch.txt: don't show -s as shorthand for... Eric Sunshine Sun, 27 Mar 2016 21:26:07 +0000 (17:26 -0400)

git-format-patch.txt: don't show -s as shorthand for multiple options

git-format-patch recognizes -s as shorthand only for --signoff, however,
its documentation shows -s as shorthand for both --signoff and
--no-patch. Resolve this confusion by suppressing the bogus -s shorthand
for --no-patch.

While here, also avoid showing the --no-patch option in git-format-patch
documentation since it doesn't make sense to ask to suppress the patch
while at the same time explicitly asking to format the patch (which,
after all, is the purpose of git-format-patch).

Reported-by: Kevin Brodsky <corax26@gmail.com>
Signed-off-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

branch -m: update all per-worktree HEADsKazuki Yamaguchi Sun, 27 Mar 2016 14:37:14 +0000 (23:37 +0900)

branch -m: update all per-worktree HEADs

When renaming a branch, currently only the HEAD of current working tree
is updated, but it must update HEADs of all working trees which point at
the old branch.

This is the current behavior, /path/to/wt's HEAD is not updated:

% git worktree list
/path/to 2c3c5f2 [master]
/path/to/wt 2c3c5f2 [oldname]
% git branch -m master master2
% git worktree list
/path/to 2c3c5f2 [master2]
/path/to/wt 2c3c5f2 [oldname]
% git branch -m oldname newname
% git worktree list
/path/to 2c3c5f2 [master2]
/path/to/wt 0000000 [oldname]

This patch fixes this issue by updating all relevant worktree HEADs
when renaming a branch.

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

refs: add a new function set_worktree_head_symrefKazuki Yamaguchi Sun, 27 Mar 2016 14:37:13 +0000 (23:37 +0900)

refs: add a new function set_worktree_head_symref

Add a new function set_worktree_head_symref, to update HEAD symref for
the specified worktree.

To update HEAD of a linked working tree,
create_symref("worktrees/$work_tree/HEAD", "refs/heads/$branch", msg)
could be used. However when it comes to updating HEAD of the main
working tree, it is unusable because it uses $GIT_DIR for
worktree-specific symrefs (HEAD).

The new function takes git_dir (real directory) as an argument, and
updates HEAD of the working tree. This function will be used when
renaming a branch.

Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Acked-by: David Turner <dturner@twopensource.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Git 2.8.1 v2.8.1Junio C Hamano Sun, 3 Apr 2016 17:11:35 +0000 (10:11 -0700)

Git 2.8.1

Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge branch 'mm/readme-markdown' into maintJunio C Hamano Sun, 3 Apr 2016 17:13:09 +0000 (10:13 -0700)

Merge branch 'mm/readme-markdown' into maint

* 'mm/readme-markdown':
git.spec.in: use README.md, not README

git.spec.in: use README.md, not READMEMatthieu Moy Fri, 1 Apr 2016 07:10:51 +0000 (09:10 +0200)

git.spec.in: use README.md, not README

The file was renamed in 4ad21f5 (README: use markdown syntax,
2016-02-25), but that commit forgot to update git.spec.in, which
caused the rpmbuild target in the Makefile to fail.

Reported-by: Ron Isaacson <isaacson.ljits@gmail.com>
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

ident: give "please tell me" message upon useConfigOnly... Marios Titas Wed, 30 Mar 2016 19:29:43 +0000 (22:29 +0300)

ident: give "please tell me" message upon useConfigOnly error

The env_hint message applies perfectly to the case when
user.useConfigOnly is set and at least one of the user.name and the
user.email are not provided.

Additionally, use a less descriptive error message to discourage
users from disabling user.useConfigOnly configuration variable to
work around this error condition. We want to encourage them to set
user.name or user.email instead.

Signed-off-by: Marios Titas <redneb@gmx.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

ident: check for useConfigOnly before auto-detection... Marios Titas Wed, 30 Mar 2016 19:29:42 +0000 (22:29 +0300)

ident: check for useConfigOnly before auto-detection of name/email

If user.useConfigOnly is set, it does not make sense to try to
auto-detect the name and/or the email. The auto-detection may
even result in a bogus name and trigger an error message.

Check if the use-config-only is set and die if no explicit name was
given, before attempting to auto-detect, to correct this.

Signed-off-by: Marios Titas <redneb@gmx.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule--helper, module_clone: catch fprintf failureStefan Beller Fri, 1 Apr 2016 00:17:29 +0000 (17:17 -0700)

submodule--helper, module_clone: catch fprintf failure

The return value of fprintf is unchecked, which may lead to
unreported errors. Use fprintf_or_die to report the error to the user.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule--helper: do not borrow absolute_path() result... Junio C Hamano Fri, 1 Apr 2016 19:23:16 +0000 (12:23 -0700)

submodule--helper: do not borrow absolute_path() result for too long

absolute_path() is designed to allow its callers to take a brief
peek of the result (typically, to be fed to functions like
strbuf_add() and relative_path() as a parameter) without having to
worry about freeing it, but the other side of the coin of that
memory model is that the caller shouldn't rely too much on the
result living forever--there may be a helper function the caller
subsequently calls that makes its own call to absolute_path(),
invalidating the earlier result.

Use xstrdup() to make our own copy, and free(3) it when we are done.
While at it, remove an unnecessary sm_gitdir_rel variable that was
only used to as a parameter to call absolute_path() and never used
again.

Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule--helper, module_clone: always operate on... Stefan Beller Fri, 1 Apr 2016 00:17:28 +0000 (17:17 -0700)

submodule--helper, module_clone: always operate on absolute paths

When giving relative paths to `relative_path` to compute a relative path
from one directory to another, this may fail in `relative_path`.
Make sure both arguments to `relative_path` are always absolute.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

credential-cache, send_request: close fd when doneStefan Beller Fri, 1 Apr 2016 00:35:46 +0000 (17:35 -0700)

credential-cache, send_request: close fd when done

No need to keep it open any further.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

bundle: don't leak an fd in case of early returnStefan Beller Fri, 1 Apr 2016 00:35:45 +0000 (17:35 -0700)

bundle: don't leak an fd in case of early return

In successful operation `write_pack_data` will close the `bundle_fd`,
but when we exit early, we need to take care of the file descriptor
as well as the lock file ourselves. The lock file may be deleted at the
end of running the program, but we are in library code, so we should
not rely on that.

Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

abbrev_sha1_in_line: don't leak memoryStefan Beller Fri, 1 Apr 2016 00:35:44 +0000 (17:35 -0700)

abbrev_sha1_in_line: don't leak memory

`split` is of type `struct strbuf **`, and currently we are leaking split
itself as well as each element in split[i]. We have a dedicated free
function for `struct strbuf **`, which takes care of freeing all
related memory.

Helped-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

notes: don't leak memory in git_config_get_notes_strategyStefan Beller Fri, 1 Apr 2016 00:35:43 +0000 (17:35 -0700)

notes: don't leak memory in git_config_get_notes_strategy

This function asks for the value of a configuration and after
using the value does not have to retain ownership of it.
git_config_get_string_const() however is a function to get a
copy of the value, but we forget to free it before we return.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

builtin/apply: free patch when parse_chunk() failsChristian Couder Wed, 16 Mar 2016 19:35:11 +0000 (20:35 +0100)

builtin/apply: free patch when parse_chunk() fails

When parse_chunk() fails it can return -1, for example
when find_header() doesn't find a patch header.

In this case it's better in apply_patch() to free the
"struct patch" that we just allocated instead of
leaking it.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

builtin/apply: handle parse_binary() failureChristian Couder Fri, 18 Mar 2016 12:30:41 +0000 (13:30 +0100)

builtin/apply: handle parse_binary() failure

In parse_binary() there is:

forward = parse_binary_hunk(&buffer, &size, &status, &used);
if (!forward && !status)
/* there has to be one hunk (forward hunk) */
return error(_("unrecognized binary patch at line %d"), linenr-1);

so parse_binary() can return -1, because that's what error() returns.

Also parse_binary_hunk() sets "status" to -1 in case of error and
parse_binary() does "if (status) return status;".

In this case parse_chunk() should not add -1 to the patchsize it computes.
It is better for future libification efforts to make it just return -1.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule--helper clone: create the submodule path... Stefan Beller Thu, 31 Mar 2016 21:04:38 +0000 (14:04 -0700)

submodule--helper clone: create the submodule path just once

We make sure that the parent directory of path exists (or create it
otherwise) and then do the same for path + "/.git".

That is equivalent to just making sure that the parent directory of
path + "/.git" exists (or create it otherwise).

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule--helper: fix potential NULL-dereferenceStefan Beller Thu, 31 Mar 2016 21:04:37 +0000 (14:04 -0700)

submodule--helper: fix potential NULL-dereference

Don't dereference NULL 'path' if it was never assigned. Also
protect against an empty --path argument.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

recursive submodules: test for relative pathsStefan Beller Thu, 31 Mar 2016 21:04:36 +0000 (14:04 -0700)

recursive submodules: test for relative paths

"git submodule update --init --recursive" uses full path to refer to
the true location of the repository in the "gitdir:" pointer for
nested submodules; the command used to use relative paths.

This was reported by Norio Nomura in $gmane/290280.

The root cause for that bug is in using recursive submodules as
their relative path handling was broken in ee8838d (2015-09-08,
submodule: rewrite `module_clone` shell function in C).

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

git-send-pack: fix --all option when used with directoryStanislav Kolotinskiy Thu, 31 Mar 2016 13:55:09 +0000 (16:55 +0300)

git-send-pack: fix --all option when used with directory

When using git send-pack with --all option
and a target repository specification ([<host>:]<directory>),
usage message is being displayed instead of performing
the actual transmission.

The reason for this issue is that destination and refspecs are being set
in the same conditional and are populated from argv. When a target
repository is passed, refspecs is being populated as well with its value.
This makes the check for refspecs not being NULL to always return true,
which, in conjunction with the check for --all or --mirror options,
is always true as well and returns usage message instead of proceeding.

This ensures that send-pack will stop execution only when --all
or --mirror switch is used in conjunction with any refspecs passed.

Signed-off-by: Stanislav Kolotinskiy <stanislav@assembla.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

for-each-ref: fix description of '--contains' in manpageSZEDER Gábor Wed, 30 Mar 2016 13:48:30 +0000 (15:48 +0200)

for-each-ref: fix description of '--contains' in manpage

'git for-each-ref's manpage says that '--contains' only lists tags,
but it lists all kinds of refs.

Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

diffcore: fix iteration order of identical files during... SZEDER Gábor Wed, 30 Mar 2016 08:35:07 +0000 (10:35 +0200)

diffcore: fix iteration order of identical files during rename detection

If the two paths 'dir/A/file' and 'dir/B/file' have identical content
and the parent directory is renamed, e.g. 'git mv dir other-dir', then
diffcore reports the following exact renames:

renamed: dir/B/file -> other-dir/A/file
renamed: dir/A/file -> other-dir/B/file

While technically not wrong, this is confusing not only for the user,
but also for git commands that make decisions based on rename
information, e.g. 'git log --follow other-dir/A/file' follows
'dir/B/file' past the rename.

This behavior is a side effect of commit v2.0.0-rc4~8^2~14
(diffcore-rename.c: simplify finding exact renames, 2013-11-14): the
hashmap storing sources returns entries from the same bucket, i.e.
sources matching the current destination, in LIFO order. Thus the
iteration first examines 'other-dir/A/file' and 'dir/B/file' and, upon
finding identical content and basename, reports an exact rename.

Other hashmap users are apparently happy with the current iteration
order over the entries of a bucket. Changing the iteration order
would risk upsetting other hashmap users and would increase the memory
footprint of each bucket by a pointer to the tail element.

Fill the hashmap with source entries in reverse order to restore the
original exact rename detection behavior.

Reported-by: Bill Okara <billokara@gmail.com>
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

t7407: make expectation as clear as possibleStefan Beller Wed, 30 Mar 2016 01:27:46 +0000 (18:27 -0700)

t7407: make expectation as clear as possible

Not everyone (including me) grasps the sed expression in a split second as
they would grasp the 4 lines printed as is.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule update: test recursive path reporting from... Stefan Beller Wed, 30 Mar 2016 01:27:45 +0000 (18:27 -0700)

submodule update: test recursive path reporting from subdirectory

This patch is just a test and fixes no bug as there is currently no bug
in the path handling of `submodule update`.

In `submodule update` we make a call to `submodule--helper list --prefix
"$wt_prefix"` which looks a bit brittle and likely to introduce a bug
for the path handling. It is not a bug as the prefix is ignored inside
the submodule helper for now. If this test breaks eventually, we want
to make sure the `wt_prefix` is passed correctly into recursive submodules.
Hint: In recursive submodules we expect `wt_prefix` to be empty.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule update: align reporting path for custom comma... Stefan Beller Wed, 30 Mar 2016 01:27:44 +0000 (18:27 -0700)

submodule update: align reporting path for custom command execution

In the predefined actions (merge, rebase, none, checkout), we use
the display path, which is relative to the current working directory.
Also use the display path when running a custom command.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule status: correct path handling in recursive... Stefan Beller Wed, 30 Mar 2016 01:27:43 +0000 (18:27 -0700)

submodule status: correct path handling in recursive submodules

The new test which is a replica of the previous test except
that it executes from a sub directory. Prior to this patch
the test failed by having too many '../' prefixed:

--- expect 2016-03-29 19:02:33.087336115 +0000
+++ actual 2016-03-29 19:02:33.359343311 +0000
@@ -1,7 +1,7 @@
b23f134787d96fae589a6b76da41f4db112fc8db ../nested1 (heads/master)
-+25d56d1ddfb35c3e91ff7d8f12331c2e53147dcc ../nested1/nested2 (file2)
- 5ec83512b76a0b8170b899f8e643913c3e9b72d9 ../nested1/nested2/nested3 (heads/master)
- 509f622a4f36a3e472affcf28fa959174f3dd5b5 ../nested1/nested2/nested3/submodule (heads/master)
++25d56d1ddfb35c3e91ff7d8f12331c2e53147dcc ../../nested1/nested2 (file2)
+ 5ec83512b76a0b8170b899f8e643913c3e9b72d9 ../../../nested1/nested2/nested3 (heads/master)
+ 509f622a4f36a3e472affcf28fa959174f3dd5b5 ../../../../nested1/nested2/nested3/submodule (heads/master)
0c90624ab7f1aaa301d3bb79f60dcfed1ec4897f ../sub1 (0c90624)
0c90624ab7f1aaa301d3bb79f60dcfed1ec4897f ../sub2 (0c90624)
509f622a4f36a3e472affcf28fa959174f3dd5b5 ../sub3 (heads/master)

The path code in question:
displaypath=$(relative_path "$prefix$sm_path")
prefix=$displaypath
if recursive:
eval cmd_status

That way we change `prefix` each iteration to contain another
'../', because of the the relative_path computation is done
on an already computed relative path.

We must call relative_path exactly once with `wt_prefix` non empty.
Further calls in recursive instances to to calculate the displaypath
already incorporate the correct prefix from before. Fix the issue by
clearing `wt_prefix` in recursive calls.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule update --init: correct path handling in recur... Stefan Beller Wed, 30 Mar 2016 01:27:42 +0000 (18:27 -0700)

submodule update --init: correct path handling in recursive submodules

When calling `git submodule init` from a recursive instance of
`git submodule update --recursive`, the reported path is wrong as it
skips the nested submodules.

The new test demonstrates a failure in the code prior to this patch.
Instead of getting the expected
Submodule 'submodule' (${pwd}/submodule) registered for path '../super/submodule'
the `super` directory is omitted and you get
Submodule 'submodule' (${pwd}/submodule) registered for path '../submodule'
instead.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule foreach: correct path display in recursive... Stefan Beller Wed, 30 Mar 2016 01:27:41 +0000 (18:27 -0700)

submodule foreach: correct path display in recursive submodules

The `prefix` was put in front of the display path unconditionally.
This is wrong as any relative path computation would need to be at
the front, so include the prefix into the display path.

The new test replicates the previous test with the difference of executing
from a sub directory. By executing from a sub directory all we would
expect all displayed paths to be prefixed by '../'.

Prior to this patch the test would report
Entering 'nested1/nested2/../nested3'
instead of the expected
Entering '../nested1/nested2/nested3'

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

MSVC: use shipped headers instead of fallback definitionsSven Strickroth Wed, 30 Mar 2016 11:37:36 +0000 (13:37 +0200)

MSVC: use shipped headers instead of fallback definitions

VS2010 comes with stdint.h [1]
VS2013 comes with inttypes.h [2]

[1] https://stackoverflow.com/a/2628014/3906760
[2] https://blogs.msdn.microsoft.com/vcblog/2013/07/19/c99-library-support-in-visual-studio-2013/

Signed-off-by: Sven Strickroth <sven@cs-ware.de>
Acked-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Acked-by: Sebastian Schuberth <sschuberth@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

MSVC: vsnprintf in Visual Studio 2015 doesn't need... Sven Strickroth Tue, 29 Mar 2016 16:25:28 +0000 (18:25 +0200)

MSVC: vsnprintf in Visual Studio 2015 doesn't need SNPRINTF_SIZE_CORR any more

In MSVC2015 the behavior of vsnprintf was changed.
W/o this fix there is one character missing at the end.

Signed-off-by: Sven Strickroth <sven@cs-ware.de>
Acked-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Acked-by: Sebastian Schuberth <sschuberth@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Documentation: git diff --check detects conflict markersOri Avtalion Tue, 29 Mar 2016 18:59:45 +0000 (21:59 +0300)

Documentation: git diff --check detects conflict markers

Signed-off-by: Ori Avtalion <ori@avtalion.name>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

branch -d: refuse deleting a branch which is currently... Kazuki Yamaguchi Tue, 29 Mar 2016 09:38:39 +0000 (18:38 +0900)

branch -d: refuse deleting a branch which is currently checked out

When a branch is checked out by current working tree, deleting the
branch is forbidden. However when the branch is checked out only by
other working trees, deleting incorrectly succeeds.
Use find_shared_symref() to check if the branch is in use, not just
comparing with the current working tree's HEAD.

Helped-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Kazuki Yamaguchi <k@rhe.jp>
Reviewed-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Git 2.8 v2.8.0Junio C Hamano Mon, 28 Mar 2016 19:19:45 +0000 (12:19 -0700)

Git 2.8

Signed-off-by: Junio C Hamano <gitster@pobox.com>

api-parse-options.txt: document OPT_CMDMODE()Pranit Bauva Fri, 25 Mar 2016 18:58:58 +0000 (18:58 +0000)

api-parse-options.txt: document OPT_CMDMODE()

OPT_CMDMODE mechanism was introduced in the release of 1.8.5 to actively
notice when multiple "operation mode" options that specify mutually
incompatible operation modes are given.

Signed-off-by: Pranit Bauva <pranit.bauva@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge pull request #9 from vascool/frJean-Noël Avila Fri, 25 Mar 2016 16:11:20 +0000 (17:11 +0100)

Merge pull request #9 from vascool/fr

Fix inconsistencies

l10n: fr: don't translate "merge" as a parameterVasco Almeida Tue, 15 Mar 2016 00:34:22 +0000 (23:34 -0100)

l10n: fr: don't translate "merge" as a parameter

At builtin/checkout.c:1154, merge is a parameter to --conflict=<style>
(git checkout --conflict=merge).

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>

l10n: fr: change "id de clé" to match "id-clé"Vasco Almeida Tue, 15 Mar 2016 00:34:22 +0000 (23:34 -0100)

l10n: fr: change "id de clé" to match "id-clé"

At builtin/tag.c:23 French message translation, "<key-id>" was
translated to "<id-clé>", but at builtin/tag.c:355 "key-id" was
translated to "id de clé", hence an inconsistency in git tag -h output.

Translate "key-id" to "id-clé".
Alternatively, both places could use "id de clé" instead of "id-clé".

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>

l10n: fr: fix wrongly translated option nameVasco Almeida Sun, 13 Mar 2016 12:56:03 +0000 (11:56 -0100)

l10n: fr: fix wrongly translated option name

In the original source, tags and heads refer to that options (--head and
--tags) for git show-ref.

Don't translate that terms, since they refer to actual option names.

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>

l10n: fr: fix transcation of "dir"Vasco Almeida Sun, 13 Mar 2016 12:56:03 +0000 (11:56 -0100)

l10n: fr: fix transcation of "dir"

"dir" was translated to the same string at builtin/log.c:1236,
but, also at that code line, "<dir>" was translate to "<répertoire>".

Before this commit, git format-patch -h would output "-o,
--output-directory <dir>" and <répertoire> in its description.

Use <répertoire> in both the parameter and description.

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>

Merge branch 'ls/p4-doc-markup'Junio C Hamano Thu, 24 Mar 2016 19:28:06 +0000 (12:28 -0700)

Merge branch 'ls/p4-doc-markup'

* ls/p4-doc-markup:
Documentation: fix git-p4 AsciiDoc formatting
Documentation: use ASCII quotation marks in git-p4

Merge branch 'js/mingw-tests-2.8'Junio C Hamano Thu, 24 Mar 2016 19:27:58 +0000 (12:27 -0700)

Merge branch 'js/mingw-tests-2.8'

* js/mingw-tests-2.8:
mingw: skip some tests in t9115 due to file name issues
t1300: fix the new --show-origin tests on Windows
t1300-repo-config: make it resilient to being run via 'sh -x'
config --show-origin: report paths with forward slashes

Merge branch 'sb/submodule-module-list-pathspec-fix'Junio C Hamano Thu, 24 Mar 2016 19:27:12 +0000 (12:27 -0700)

Merge branch 'sb/submodule-module-list-pathspec-fix'

A fix for a small regression in "module_list" helper that was
rewritten in C (also applies to 2.7.x).

* sb/submodule-module-list-pathspec-fix:
submodule: fix regression for deinit without submodules

git-apply.txt: mention the behavior inside a subdirNguyễn Thái Ngọc Duy Thu, 24 Mar 2016 11:56:17 +0000 (18:56 +0700)

git-apply.txt: mention the behavior inside a subdir

Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

git-apply.txt: remove a spaceNguyễn Thái Ngọc Duy Thu, 24 Mar 2016 11:56:16 +0000 (18:56 +0700)

git-apply.txt: remove a space

Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

submodule-config: use hashmap_iter_first()Alexander Kuleshov Wed, 16 Mar 2016 07:46:31 +0000 (13:46 +0600)

submodule-config: use hashmap_iter_first()

The hashmap API provides hashmap_iter_first() helper for initialion
and getting the first entry of a hashmap. Let's use it instead of
doing initialization manually and then get the first entry.

There are no functional changes, just cleanup.

Signed-off-by: Alexander Kuleshov <kuleshovmail@gmail.com>
Reviewed-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge branch 'master' of git://github.com/git-l10n... Junio C Hamano Wed, 23 Mar 2016 19:22:42 +0000 (12:22 -0700)

Merge branch 'master' of git://github.com/git-l10n/git-po

* 'master' of git://github.com/git-l10n/git-po:
l10n: pt_PT: Update and add new translations
l10n: ca.po: update translation
l10n: vi.po (2530t): Update translation

merge: fix NULL pointer dereference when merging nothin... Junio C Hamano Mon, 21 Mar 2016 19:01:43 +0000 (12:01 -0700)

merge: fix NULL pointer dereference when merging nothing into void

When we are on an unborn branch and merging only one foreign parent,
we allow "git merge" to fast-forward to that foreign parent commit.

This codepath incorrectly attempted to dereference the list of
parents that the merge is going to record even when the list is
empty. It must refuse to operate instead when there is no parent.

All other codepaths make sure the list is not empty before they
dereference it, and are safe.

Reported-by: Jose Ivan B. Vilarouca Filho
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Documentation: fix git-p4 AsciiDoc formattingLars Schneider Wed, 23 Mar 2016 10:59:01 +0000 (11:59 +0100)

Documentation: fix git-p4 AsciiDoc formatting

Noticed-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

mingw: skip some tests in t9115 due to file name issuesJohannes Schindelin Wed, 23 Mar 2016 10:55:20 +0000 (11:55 +0100)

mingw: skip some tests in t9115 due to file name issues

These two tests wanted to write file names which are incompatible with
Windows' file naming rules (even if they pass using Cygwin due to
Cygwin's magic path mangling).

While at it, skip the same tests also on MacOSX/HFS, as pointed out by
Torsten Bögershausen.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

t1300: fix the new --show-origin tests on WindowsJohannes Schindelin Wed, 23 Mar 2016 10:55:13 +0000 (11:55 +0100)

t1300: fix the new --show-origin tests on Windows

On Windows, we have that funny situation where the test script can refer
to POSIX paths because it runs in a shell that uses a POSIX emulation
layer ("MSYS2 runtime"). Yet, git.exe does *not* understand POSIX paths
at all but only pure Windows paths.

So let's just convert the POSIX paths to Windows paths before passing
them on to Git, using `pwd` (which is already modified on Windows to
output Windows paths).

While fixing the new tests on Windows, we also have to exclude the tests
that want to write a file with a name that is illegal on Windows
(unfortunately, there is more than one test trying to make use of that
file).

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

t1300-repo-config: make it resilient to being run via... Johannes Schindelin Wed, 23 Mar 2016 10:55:07 +0000 (11:55 +0100)

t1300-repo-config: make it resilient to being run via 'sh -x'

One way to diagnose broken regression tests is to run the test
script using 'sh -x t... -i -v' to find out which call actually
demonstrates the symptom.

Hence it is pretty counterproductive if the test script behaves
differently when being run via 'sh -x', in particular when using
test_cmp or test_i18ncmp on redirected stderr. A more recent way
"sh tXXXX -i -v -x" has the same issue.

So let's use test_i18ngrep (as suggested by Jonathan Nieder) instead of
test_cmp/test_i18ncmp to verify that stderr looks as expected.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

config --show-origin: report paths with forward slashesJohannes Schindelin Wed, 23 Mar 2016 10:55:00 +0000 (11:55 +0100)

config --show-origin: report paths with forward slashes

On Windows, the backslash is the native directory separator, but all
supported Windows versions also accept the forward slash in most
circumstances.

Our tests expect forward slashes.

Relative paths are generated by Git using forward slashes.

So let's try to be consistent and use forward slashes in the $HOME part
of the paths reported by `git config --show-origin`, too.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge branch 'master' of https://github.com/vnwildman/gitJiang Xin Wed, 23 Mar 2016 15:01:51 +0000 (23:01 +0800)

Merge branch 'master' of https://github.com/vnwildman/git

* 'master' of https://github.com/vnwildman/git:
l10n: vi.po (2530t): Update translation

Merge branch 'master' of git://github.com/alexhenrie... Jiang Xin Wed, 23 Mar 2016 14:48:14 +0000 (22:48 +0800)

Merge branch 'master' of git://github.com/alexhenrie/git-po

* 'master' of git://github.com/alexhenrie/git-po:
l10n: ca.po: update translation

submodule: fix regression for deinit without submodulesStefan Beller Tue, 22 Mar 2016 23:42:14 +0000 (16:42 -0700)

submodule: fix regression for deinit without submodules

Per Cederqvist wrote:
> It used to be possible to run
>
> git submodule deinit -f .
>
> to remove any submodules, no matter how many submodules you had. That
> is no longer possible in projects that don't have any submodules at
> all. The command will fail with:
>
> error: pathspec '.' did not match any file(s) known to git.

This regression was introduced in 74703a1e4dfc (submodule: rewrite
`module_list` shell function in C, 2015-09-02), as we changed the
order of checking in new module listing to first check whether it is
a gitlin before feeding it to match_pathspec(). It used to be that
a pathspec that does not match any path were diagnosed as an error,
but the new code complains for a pathspec that does not match any
submodule path.

Arguably the new behaviour may give us a better diagnosis, but that
is inconsistent with the suggestion "deinit" gives, and also this
was an unintended accident. The new behaviour hopefully can be
redesigned and implemented better in future releases, but for now,
switch these two checks to restore the same behavior as before. In
an empty repository, giving the pathspec '.' will still get the same
"did not match" error, but that is the same bug we had before 1.7.0.

Reported-by: Per Cederqvist <cederp@opera.com>
Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

apply: remove unused call to free() in gitdiff_{old... Junio C Hamano Tue, 22 Mar 2016 21:41:08 +0000 (14:41 -0700)

apply: remove unused call to free() in gitdiff_{old,new}name()

These two functions keep a copy of filename it was given, let
gitdiff_verify_name() to rewrite it to a new filename and then free
the original if they receive a newly minted filename.

However

(1) when the original name is NULL, gitdiff_verify_name() returns
either NULL or a newly minted value. Either case, we do not
have to worry about calling free() on the original NULL.

(2) when the original name is not NULL, gitdiff_verify_name()
either returns that as-is, or calls die() when it finds
inconsistency in the patch. When the function returns, we know
that "if ()" statement always is false.

Noticed by Christian Couder.

Signed-off-by: Junio C Hamano <gitster@pobox.com>

builtin/apply: get rid of useless 'name' variableChristian Couder Tue, 22 Mar 2016 20:58:40 +0000 (21:58 +0100)

builtin/apply: get rid of useless 'name' variable

While at it put an 'else' on the same line as the previous '}'.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

l10n: pt_PT: Update and add new translationsVasco Almeida Fri, 29 Jan 2016 20:06:29 +0000 (19:06 -0100)

l10n: pt_PT: Update and add new translations

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>

l10n: ca.po: update translationAlex Henrie Tue, 22 Mar 2016 05:04:22 +0000 (23:04 -0600)

l10n: ca.po: update translation

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>

commit: do not lose SQUASH_MSG contentsSven Strickroth Mon, 21 Mar 2016 22:29:40 +0000 (23:29 +0100)

commit: do not lose SQUASH_MSG contents

When concluding a conflicted "git merge --squash", the command
failed to read SQUASH_MSG that was prepared by "git merge", and
showed only the "# Conflicts:" list of conflicted paths.

Place the contents from SQUASH_MSG at the beginning, just like we
show the commit log skeleton first when concluding a normal merge,
and then show the "# Conflicts:" list, to help the user write the
log message for the resulting commit.

Test by Junio C Hamano <gitster@pobox.com>.

Signed-off-by: Sven Strickroth <sven@cs-ware.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Git 2.8-rc4 v2.8.0-rc4Junio C Hamano Mon, 21 Mar 2016 20:41:37 +0000 (13:41 -0700)

Git 2.8-rc4

Signed-off-by: Junio C Hamano <gitster@pobox.com>

Sync with maintJunio C Hamano Mon, 21 Mar 2016 20:32:42 +0000 (13:32 -0700)

Sync with maint

* maint:
Documentation: fix broken linkgit to git-config
git-compat-util: st_add4: work around gcc 4.2.x compiler crash

Merge branch 'mm/doc-hooks-linkgit-fix' into maintJunio C Hamano Mon, 21 Mar 2016 20:32:18 +0000 (13:32 -0700)

Merge branch 'mm/doc-hooks-linkgit-fix' into maint

* mm/doc-hooks-linkgit-fix:
Documentation: fix broken linkgit to git-config

Documentation: fix broken linkgit to git-configMatthieu Moy Mon, 21 Mar 2016 18:38:34 +0000 (19:38 +0100)

Documentation: fix broken linkgit to git-config

Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Documentation: use ASCII quotation marks in git-p4Lars Schneider Sun, 20 Mar 2016 18:39:21 +0000 (19:39 +0100)

Documentation: use ASCII quotation marks in git-p4

Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge branch 'tb/avoid-gcc-on-darwin-10-6'Junio C Hamano Mon, 21 Mar 2016 16:20:13 +0000 (09:20 -0700)

Merge branch 'tb/avoid-gcc-on-darwin-10-6'

* tb/avoid-gcc-on-darwin-10-6:
Revert "config.mak.uname: use clang for Mac OS X 10.6"

Revert "config.mak.uname: use clang for Mac OS X 10.6"Eric Sunshine Mon, 21 Mar 2016 04:35:58 +0000 (00:35 -0400)

Revert "config.mak.uname: use clang for Mac OS X 10.6"

This reverts commit 7b6daf8d2fee1a9866b1d4eddbfaa5dbc42c5dbb.

Now that st_add4() has been patched to work around the gcc 4.2.x
compiler crash, revert the sledge-hammer approach of forcing Mac OS X
10.6 to unconditionally use 'clang' rather than the default compiler
(gcc).

Signed-off-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge branch 'es/st-add4-gcc-4.2-workaround' into maintJunio C Hamano Mon, 21 Mar 2016 16:19:27 +0000 (09:19 -0700)

Merge branch 'es/st-add4-gcc-4.2-workaround' into maint

* es/st-add4-gcc-4.2-workaround:
git-compat-util: st_add4: work around gcc 4.2.x compiler crash

git-compat-util: st_add4: work around gcc 4.2.x compile... Eric Sunshine Mon, 21 Mar 2016 04:35:57 +0000 (00:35 -0400)

git-compat-util: st_add4: work around gcc 4.2.x compiler crash

Although changes by 5b442c4 (tree-diff: catch integer overflow in
combine_diff_path allocation, 2016-02-19) are perfectly valid, they
unfortunately trigger an internal compiler error in gcc 4.2.x:

combine-diff.c: In function 'diff_tree_combined':
combine-diff.c:1391: internal compiler error: Segmentation fault: 11

Experimentation reveals that changing st_add4()'s argument evaluation
order is sufficient to sidestep this problem.

Although st_add3() does not trigger the compiler bug, for style
consistency, change its argument evaluation order to match.

Signed-off-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Merge tag 'l10n-2.8.0-rnd3' of git://github.com/git... Junio C Hamano Mon, 21 Mar 2016 01:06:05 +0000 (18:06 -0700)

Merge tag 'l10n-2.8.0-rnd3' of git://github.com/git-l10n/git-po

l10n-2.8.0-rnd3

* tag 'l10n-2.8.0-rnd3' of git://github.com/git-l10n/git-po:
l10n: zh_CN: review for git v2.8.0 l10n round 2
l10n: de.po: add missing newlines
l10n: de.po: translate 22 new messages
l10n: ko.po: Update Korean translation
l10n: fr.po v2.8.0 round 3
l10n: sv.po: Update Swedish translation (2530t0f0u)
l10n: ru.po: update Russian translation

Merge branch 'master' of git://ozlabs.org/~paulus/gitkJunio C Hamano Mon, 21 Mar 2016 01:05:10 +0000 (18:05 -0700)

Merge branch 'master' of git://ozlabs.org/~paulus/gitk

* 'master' of git://ozlabs.org/~paulus/gitk:
gitk: Follow themed bgcolor in help dialogs
gitk: fr.po: Sync translations with git
gitk: Update French translation (311t)
gitk: Update German translation
gitk: Update Bulgarian translation (311t)

l10n: vi.po (2530t): Update translationTran Ngoc Quan Mon, 21 Mar 2016 00:21:04 +0000 (07:21 +0700)

l10n: vi.po (2530t): Update translation

Signed-off-by: Tran Ngoc Quan <vnwildman@gmail.com>

l10n: zh_CN: review for git v2.8.0 l10n round 2Ray Chen Wed, 16 Mar 2016 03:44:44 +0000 (11:44 +0800)

l10n: zh_CN: review for git v2.8.0 l10n round 2

Signed-off-by: Ray Chen <oldsharp@gmail.com>

gitk: Follow themed bgcolor in help dialogsGuillermo S. Romero Thu, 4 Feb 2016 02:32:19 +0000 (03:32 +0100)

gitk: Follow themed bgcolor in help dialogs

Make Help > About & Key bindings dialogs readable if theme
has changed font color to something incompatible with white.

Signed-off-by: Guillermo S. Romero <gsromero@infernal-iceberg.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>

gitk: fr.po: Sync translations with gitJean-Noel Avila Fri, 29 Jan 2016 20:21:35 +0000 (21:21 +0100)

gitk: fr.po: Sync translations with git

Signed-off-by: Jean-Noel Avila <jn.avila@free.fr>
Signed-off-by: Paul Mackerras <paulus@samba.org>

gitk: Update French translation (311t)Jean-Noel Avila Fri, 29 Jan 2016 20:21:34 +0000 (21:21 +0100)

gitk: Update French translation (311t)

Signed-off-by: Jean-Noel Avila <jn.avila@free.fr>
Signed-off-by: Paul Mackerras <paulus@samba.org>

gitk: Update German translationRalf Thielow Fri, 12 Feb 2016 18:40:39 +0000 (19:40 +0100)

gitk: Update German translation

Signed-off-by: Ralf Thielow <ralf.thielow@gmail.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>

gitk: Update Bulgarian translation (311t)Alexander Shopov Sat, 19 Dec 2015 22:36:09 +0000 (00:36 +0200)

gitk: Update Bulgarian translation (311t)

Signed-off-by: Alexander Shopov <ash@kambanaria.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>

t/lib-httpd: pass through GIT_CONFIG_NOSYSTEM envJeff King Wed, 16 Mar 2016 00:56:52 +0000 (20:56 -0400)

t/lib-httpd: pass through GIT_CONFIG_NOSYSTEM env

We set GIT_CONFIG_NOSYSTEM in our test scripts so that we do
not accidentally read /etc/gitconfig and have it influence
the outcome of the tests. But when running smart-http tests,
Apache will clean the environment, including this variable,
and the "server" side of our http operations will read it.

You can see this breakage by doing something like:

make
./git config --system http.getanyfile false
make test

which will cause t5561 to fail when it tests the
fallback-to-dumb operation.

We can fix this by instructing Apache to pass through the
variable. Unlike with other variables (e.g., 89c57ab3's
GIT_TRACE), we don't need to set a dummy value to prevent
warnings from Apache. test-lib.sh already makes sure that
GIT_CONFIG_NOSYSTEM is set and exported.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

credential-cache--daemon: clarify "exit" action semanticsJeff King Fri, 18 Mar 2016 06:12:01 +0000 (02:12 -0400)

credential-cache--daemon: clarify "exit" action semantics

When this code was originally written, there wasn't much
thought given to the timing between a client asking for
"exit", the daemon signaling that the action is done (with
EOF), and the actual cleanup of the socket.

However, we need to care about this so that our test scripts
do not end up racy (e.g., by asking for an exit and checking
that the socket was cleaned up). The code that is already
there happens to behave very reasonably; let's add a comment
to make it clear that any changes should retain the same
behavior.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

send-email: ignore trailing whitespace in mailrc alias... Jeff King Thu, 17 Mar 2016 23:58:22 +0000 (19:58 -0400)

send-email: ignore trailing whitespace in mailrc alias file

The regex for parsing mailrc considers everything after the
second whitespace to be the email address, up to the end of
the line. We have to include whitespace there, because you
may have multiple space-separated addresses, each with their
own internal quoting.

But if there is trailing whitespace, we include that, too.
This confuses quotewords() when we try to split the
individual addresses, and we end up storing "undef" in our
alias list. Later parts of the code then access that,
generating perl warnings.

Let's tweak our regex to throw away any trailing whitespace
on each line.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

RelNotes: remove the mention of !reinclusionJunio C Hamano Fri, 18 Mar 2016 18:10:53 +0000 (11:10 -0700)

RelNotes: remove the mention of !reinclusion

We will be postponing this to a later cycle.

Signed-off-by: Junio C Hamano <gitster@pobox.com>

Revert "Merge branch 'nd/exclusion-regression-fix'"Junio C Hamano Fri, 18 Mar 2016 18:06:15 +0000 (11:06 -0700)

Revert "Merge branch 'nd/exclusion-regression-fix'"

This reverts commit 5e57f9c3dfe7dd44a1b56bb5b3327d7a1356ec7c, reversing
changes made to e79112d21024beb997951381db21a70b087d459d.

We will be postponing nd/exclusion-regression-fix topic to later
cycle.