git
7 years agovcs-svn/fast_export: fix timestamp fmt specifiers
Mike Ralphson [Wed, 14 Sep 2016 06:40:57 +0000 (06:40 +0000)] 
vcs-svn/fast_export: fix timestamp fmt specifiers

Two instances of %ld being used for unsigned longs

Signed-off-by: Mike Ralphson <mike.ralphson@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agounpack-trees: pass checkout state explicitly to check_updates()
René Scharfe [Tue, 13 Sep 2016 17:37:06 +0000 (19:37 +0200)] 
unpack-trees: pass checkout state explicitly to check_updates()

Add a parameter for the struct checkout variable to check_updates()
instead of using a static global variable.  Passing it explicitly makes
object ownership and usage more easily apparent.  And we get rid of a
static variable; those can be problematic in library-like code.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agocheckout: constify parameters of checkout_stage() and checkout_merged()
René Scharfe [Tue, 13 Sep 2016 17:11:52 +0000 (19:11 +0200)] 
checkout: constify parameters of checkout_stage() and checkout_merged()

Document the fact that checkout_stage() and checkout_merged() don't
change the objects passed to them by adding the modifier const.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agopathspec: remove unnecessary function prototypes
Brandon Williams [Tue, 13 Sep 2016 16:52:51 +0000 (09:52 -0700)] 
pathspec: remove unnecessary function prototypes

A few functions were removed in 5a76aff1 ("add: convert to use
parse_pathspec", 2013-07-14), but we forgot to remove their external
declarations from pathspec.h while doing so.

Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agostrbuf: use valid pointer in strbuf_remove()
René Scharfe [Tue, 13 Sep 2016 16:40:22 +0000 (18:40 +0200)] 
strbuf: use valid pointer in strbuf_remove()

The fourth argument of strbuf_splice() is passed to memcpy(3), which is
not supposed to handle NULL pointers.  Let's be extra careful and use a
valid empty string instead.  It even shortens the source code. :)

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agohttp: always remove curl easy from curlm session on release
Eric Wong [Tue, 13 Sep 2016 00:25:57 +0000 (00:25 +0000)] 
http: always remove curl easy from curlm session on release

We must call curl_multi_remove_handle when releasing the slot to
prevent subsequent calls to curl_multi_add_handle from failing
with CURLM_ADDED_ALREADY (in curl 7.32.1+; older versions
returned CURLM_BAD_EASY_HANDLE)

Signed-off-by: Eric Wong <e@80x24.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agohttp: consolidate #ifdefs for curl_multi_remove_handle
Eric Wong [Tue, 13 Sep 2016 00:25:56 +0000 (00:25 +0000)] 
http: consolidate #ifdefs for curl_multi_remove_handle

I find #ifdefs makes code difficult-to-follow.

An early version of this patch had error checking for
curl_multi_remove_handle calls, but caused some tests (e.g.
t5541) to fail under curl 7.26.0 on old Debian wheezy.

Signed-off-by: Eric Wong <e@80x24.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agohttp: warn on curl_multi_add_handle failures
Eric Wong [Tue, 13 Sep 2016 00:25:55 +0000 (00:25 +0000)] 
http: warn on curl_multi_add_handle failures

This will be useful for tracking down curl usage errors.

Signed-off-by: Eric Wong <e@80x24.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoMerge tag 'l10n-2.10.0-rnd2.3' of git://github.com/git-l10n/git-po into maint
Junio C Hamano [Mon, 12 Sep 2016 22:23:42 +0000 (15:23 -0700)] 
Merge tag 'l10n-2.10.0-rnd2.3' of git://github.com/git-l10n/git-po into maint

l10n-2.10.0-rnd2.3

* tag 'l10n-2.10.0-rnd2.3' of git://github.com/git-l10n/git-po:
  l10n: zh_CN: review for git v2.10.0 l10n
  l10n: zh_CN: fixed some typos for git 2.10.0
  l10n: pt_PT: update Portuguese repository info
  l10n: pt_PT: update Portuguese translation

7 years agoadd: document the chmod option
Thomas Gummerer [Mon, 12 Sep 2016 21:08:15 +0000 (22:08 +0100)] 
add: document the chmod option

The git add --chmod option was introduced in 4e55ed3 ("add: add
--chmod=+x / --chmod=-x options", 2016-05-31), but was never
documented.  Document the feature.

Signed-off-by: Thomas Gummerer <t.gummerer@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agopatch-ids: refuse to compute patch-id for merge commit
Jeff King [Mon, 12 Sep 2016 17:56:41 +0000 (13:56 -0400)] 
patch-ids: refuse to compute patch-id for merge commit

The patch-id code which powers "log --cherry-pick" doesn't
look at whether each commit is a merge or not. It just feeds
the commit's first parent to the diff, and ignores any
additional parents.

In theory, this might be useful if you wanted to find
equivalence between, say, a merge commit and a squash-merge
that does the same thing.  But it also promotes a false
equivalence between distinct merges. For example, every
"merge -s ours" would look identical to an empty commit
(which is true in a sense, but presumably there was a value
in merging in the discarded history). Since patch-ids are
meant for throwing away duplicates, we should err on the
side of _not_ matching such merges.

Moreover, we may spend a lot of extra time computing these
merge diffs. In the case that inspired this patch, a "git
format-patch --cherry-pick" dropped from over 3 minutes to
less than 3 seconds.

This seems pretty drastic, but is easily explained. The
command was invoked by a "git rebase" of an older topic
branch; there had been tens of thousands of commits on the
upstream branch in the meantime. In addition, this project
used a topic-branch workflow with occasional "back-merges"
from "master" to each topic (to resolve conflicts on the
topics rather than in the merge commits). So there were not
only extra merges, but the diffs for these back-merges were
generally quite large (because they represented _everything_
that had been merged to master since the topic branched).

This patch treats a merge fed to commit_patch_id() or
add_commit_patch_id() as an error, and a lookup for such a
merge via has_commit_patch_id() will always return NULL.
An earlier version of the patch tried to distinguish between
"error" and "patch id for merges not defined", but that
becomes unnecessarily complicated. The only callers are:

  1. revision traversals which want to do --cherry-pick;
     they call add_commit_patch_id(), but do not care if it
     fails. They only want to add what we can, look it up
     later with has_commit_patch_id(), and err on the side
     of not-matching.

  2. format-patch --base, which calls commit_patch_id().
     This _does_ notice errors, but should never feed a
     merge in the first place (and if it were to do so
     accidentally, then this patch is a strict improvement;
     we notice the bug rather than generating a bogus
     patch-id).

So in both cases, this does the right thing.

Helped-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agol10n: de.po: fix translation of autostash
Ralf Thielow [Fri, 24 Jun 2016 18:03:22 +0000 (20:03 +0200)] 
l10n: de.po: fix translation of autostash

Signed-off-by: Ralf Thielow <ralf.thielow@gmail.com>
Acked-by: Matthias Rüster <matthias.ruester@gmail.com>
7 years agoMerge branch 'js/commit-gpgsign' of ../git-gui into js/git-gui-commit-gpgsign
Junio C Hamano [Sun, 11 Sep 2016 21:53:26 +0000 (14:53 -0700)] 
Merge branch 'js/commit-gpgsign' of ../git-gui into js/git-gui-commit-gpgsign

* 'js/commit-gpgsign' of ../git-gui:
  git-gui: respect commit.gpgsign again

7 years agogit-gui: respect commit.gpgsign again
Johannes Schindelin [Fri, 9 Sep 2016 12:28:24 +0000 (14:28 +0200)] 
git-gui: respect commit.gpgsign again

As of v2.9.0, `git commit-tree` no longer heeds the `commit.gpgsign`
config setting. This broke committing with GPG signature in Git GUI.

This fixes https://github.com/git-for-windows/git/issues/850

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agol10n: zh_CN: review for git v2.10.0 l10n
Ray Chen [Wed, 7 Sep 2016 10:03:04 +0000 (18:03 +0800)] 
l10n: zh_CN: review for git v2.10.0 l10n

Signed-off-by: Ray Chen <oldsharp@gmail.com>
7 years agol10n: zh_CN: fixed some typos for git 2.10.0
Jiang Xin [Mon, 5 Sep 2016 15:26:21 +0000 (23:26 +0800)] 
l10n: zh_CN: fixed some typos for git 2.10.0

Reviewed-by: Ray <tvvocold@163.com>
Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
7 years agopatch-ids: turn off rename detection
Jeff King [Fri, 9 Sep 2016 20:34:34 +0000 (16:34 -0400)] 
patch-ids: turn off rename detection

The patch-id code may be running inside another porcelain
like "git log" or "git format-patch", and therefore may have
set diff_detect_rename_default, either via the diff-ui
config, or by default since 5404c11 (diff: activate
diff.renames by default, 2016-02-25). This is the case even
if a command is run with `--no-renames`, as that is applied
only to the diff-options used by the command itself.

Rename detection doesn't help the patch-id results. It
_may_ actually hurt, as minor differences in the files that
would be overlooked by patch-id's canonicalization might
result in different renames (though I'd doubt that it ever
comes up in practice).

But mostly it is just a waste of CPU to compute these
renames.

Note that this does have one user-visible impact: the
prerequisite patches listed by "format-patch --base". There
may be some confusion between different versions of git as
older ones will enable renames, but newer ones will not.
However, this was already a problem, as people with
different settings for the "diff.renames" config would get
different results. After this patch, everyone should get the
same results, regardless of their config.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoStart maintenance track for 2.10.x series
Junio C Hamano [Fri, 9 Sep 2016 04:39:38 +0000 (21:39 -0700)] 
Start maintenance track for 2.10.x series

7 years agoPrepare for 2.9.4
Junio C Hamano [Fri, 9 Sep 2016 04:37:59 +0000 (21:37 -0700)] 
Prepare for 2.9.4

Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoMerge branch 'hv/doc-commit-reference-style' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:36:03 +0000 (21:36 -0700)] 
Merge branch 'hv/doc-commit-reference-style' into maint

A small doc update.

* hv/doc-commit-reference-style:
  SubmittingPatches: use gitk's "Copy commit summary" format
  SubmittingPatches: document how to reference previous commits

7 years agoMerge branch 'sg/reflog-past-root' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:36:02 +0000 (21:36 -0700)] 
Merge branch 'sg/reflog-past-root' into maint

A small test clean-up for a topic introduced in v2.9.1 and later.

* sg/reflog-past-root:
  t1410: remove superfluous 'git reflog' from the 'walk past root' test

7 years agoMerge branch 'rs/mailinfo-lib' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:36:01 +0000 (21:36 -0700)] 
Merge branch 'rs/mailinfo-lib' into maint

Small code clean-up.

* rs/mailinfo-lib:
  mailinfo: recycle strbuf in check_header()

7 years agoMerge branch 'jk/tighten-alloc' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:36:00 +0000 (21:36 -0700)] 
Merge branch 'jk/tighten-alloc' into maint

Small code and comment clean-up.

* jk/tighten-alloc:
  receive-pack: use FLEX_ALLOC_MEM in queue_command()
  correct FLEXPTR_* example in comment

7 years agoMerge branch 'rs/use-strbuf-add-unique-abbrev' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:59 +0000 (21:35 -0700)] 
Merge branch 'rs/use-strbuf-add-unique-abbrev' into maint

A small code clean-up.

* rs/use-strbuf-add-unique-abbrev:
  use strbuf_add_unique_abbrev() for adding short hashes

7 years agoMerge branch 'rs/merge-recursive-string-list-init' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:59 +0000 (21:35 -0700)] 
Merge branch 'rs/merge-recursive-string-list-init' into maint

A small code clean-up.

* rs/merge-recursive-string-list-init:
  merge-recursive: use STRING_LIST_INIT_NODUP

7 years agoMerge branch 'rs/merge-add-strategies-simplification' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:58 +0000 (21:35 -0700)] 
Merge branch 'rs/merge-add-strategies-simplification' into maint

A small code clean-up.

* rs/merge-add-strategies-simplification:
  merge: use string_list_split() in add_strategies()

7 years agoMerge branch 'ls/packet-line-protocol-doc-fix' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:57 +0000 (21:35 -0700)] 
Merge branch 'ls/packet-line-protocol-doc-fix' into maint

Correct an age-old calco (is that a typo-like word for calc)
in the documentation.

* ls/packet-line-protocol-doc-fix:
  pack-protocol: fix maximum pkt-line size

7 years agoMerge branch 'bw/mingw-avoid-inheriting-fd-to-lockfile' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:56 +0000 (21:35 -0700)] 
Merge branch 'bw/mingw-avoid-inheriting-fd-to-lockfile' into maint

The tempfile (hence its user lockfile) API lets the caller to open
a file descriptor to a temporary file, write into it and then
finalize it by first closing the filehandle and then either
removing or renaming the temporary file.  When the process spawns a
subprocess after obtaining the file descriptor, and if the
subprocess has not exited when the attempt to remove or rename is
made, the last step fails on Windows, because the subprocess has
the file descriptor still open.  Open tempfile with O_CLOEXEC flag
to avoid this (on Windows, this is mapped to O_NOINHERIT).

* bw/mingw-avoid-inheriting-fd-to-lockfile:
  mingw: ensure temporary file handles are not inherited by child processes
  t6026-merge-attr: child processes must not inherit index.lock handles

7 years agoMerge branch 'dg/document-git-c-in-git-config-doc' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:56 +0000 (21:35 -0700)] 
Merge branch 'dg/document-git-c-in-git-config-doc' into maint

The "git -c var[=val] cmd" facility to append a configuration
variable definition at the end of the search order was described in
git(1) manual page, but not in git-config(1), which was more likely
place for people to look for when they ask "can I make a one-shot
override, and if so how?"

* dg/document-git-c-in-git-config-doc:
  doc: mention `git -c` in git-config(1)

7 years agoMerge branch 'js/no-html-bypass-on-windows' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:55 +0000 (21:35 -0700)] 
Merge branch 'js/no-html-bypass-on-windows' into maint

On Windows, help.browser configuration variable used to be ignored,
which has been corrected.

* js/no-html-bypass-on-windows:
  Revert "display HTML in default browser using Windows' shell API"

7 years agoMerge branch 'jk/difftool-command-not-found' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:54 +0000 (21:35 -0700)] 
Merge branch 'jk/difftool-command-not-found' into maint

"git difftool" by default ignores the error exit from the backend
commands it spawns, because often they signal that they found
differences by exiting with a non-zero status code just like "diff"
does; the exit status codes 126 and above however are special in
that they are used to signal that the command is not executable,
does not exist, or killed by a signal.  "git difftool" has been
taught to notice these exit status codes.

* jk/difftool-command-not-found:
  difftool: always honor fatal error exit codes

7 years agoMerge branch 'sb/checkout-explit-detach-no-advice' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:54 +0000 (21:35 -0700)] 
Merge branch 'sb/checkout-explit-detach-no-advice' into maint

"git checkout --detach <branch>" used to give the same advice
message as that is issued when "git checkout <tag>" (or anything
that is not a branch name) is given, but asking with "--detach" is
an explicit enough sign that the user knows what is going on.  The
advice message has been squelched in this case.

* sb/checkout-explit-detach-no-advice:
  checkout: do not mention detach advice for explicit --detach option

7 years agoMerge branch 'rs/pull-signed-tag' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:54 +0000 (21:35 -0700)] 
Merge branch 'rs/pull-signed-tag' into maint

When "git merge-recursive" works on history with many criss-cross
merges in "verbose" mode, the names the command assigns to the
virtual merge bases could have overwritten each other by unintended
reuse of the same piece of memory.

* rs/pull-signed-tag:
  commit: use FLEX_ARRAY in struct merge_remote_desc
  merge-recursive: fix verbose output for multiple base trees
  commit: factor out set_merge_remote_desc()
  commit: use xstrdup() in get_merge_parent()

7 years agoMerge branch 'js/test-lint-pathname' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:54 +0000 (21:35 -0700)] 
Merge branch 'js/test-lint-pathname' into maint

The "t/" hierarchy is prone to get an unusual pathname; "make test"
has been taught to make sure they do not contain paths that cannot
be checked out on Windows (and the mechanism can be reusable to
catch pathnames that are not portable to other platforms as need
arises).

* js/test-lint-pathname:
  t/Makefile: ensure that paths are valid on platforms we care

7 years agoMerge branch 'js/mv-dir-to-new-directory' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:54 +0000 (21:35 -0700)] 
Merge branch 'js/mv-dir-to-new-directory' into maint

"git mv dir non-existing-dir/" did not work in some environments
the same way as existing mainstream platforms.  The code now moves
"dir" to "non-existing-dir", without relying on rename("A", "B/")
that strips the trailing slash of '/'.

* js/mv-dir-to-new-directory:
  git mv: do not keep slash in `git mv dir non-existing-dir/`

7 years agoMerge branch 'js/import-tars-hardlinks' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:54 +0000 (21:35 -0700)] 
Merge branch 'js/import-tars-hardlinks' into maint

"import-tars" fast-import script (in contrib/) used to ignore a
hardlink target and replaced it with an empty file, which has been
corrected to record the same blob as the other file the hardlink is
shared with.

* js/import-tars-hardlinks:
  import-tars: support hard links

7 years agoMerge branch 'ms/document-pack-window-memory-is-per-thread' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:53 +0000 (21:35 -0700)] 
Merge branch 'ms/document-pack-window-memory-is-per-thread' into maint

* ms/document-pack-window-memory-is-per-thread:
  document git-repack interaction of pack.threads and pack.windowMemory

7 years agoMerge branch 'jk/push-force-with-lease-creation' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:53 +0000 (21:35 -0700)] 
Merge branch 'jk/push-force-with-lease-creation' into maint

"git push --force-with-lease" already had enough logic to allow
ensuring that such a push results in creation of a ref (i.e. the
receiving end did not have another push from sideways that would be
discarded by our force-pushing), but didn't expose this possibility
to the users.  It does so now.

* jk/push-force-with-lease-creation:
  t5533: make it pass on case-sensitive filesystems
  push: allow pushing new branches with --force-with-lease
  push: add shorthand for --force-with-lease branch creation
  Documentation/git-push: fix placeholder formatting

7 years agoMerge branch 'jk/reflog-date' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:52 +0000 (21:35 -0700)] 
Merge branch 'jk/reflog-date' into maint

The reflog output format is documented better, and a new format
--date=unix to report the seconds-since-epoch (without timezone)
has been added.

* jk/reflog-date:
  date: clarify --date=raw description
  date: add "unix" format
  date: document and test "raw-local" mode
  doc/pretty-formats: explain shortening of %gd
  doc/pretty-formats: describe index/time formats for %gd
  doc/rev-list-options: explain "-g" output formats
  doc/rev-list-options: clarify "commit@{Nth}" for "-g" option

7 years agoMerge branch 'jc/renormalize-merge-kill-safer-crlf' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:51 +0000 (21:35 -0700)] 
Merge branch 'jc/renormalize-merge-kill-safer-crlf' into maint

"git merge" with renormalization did not work well with
merge-recursive, due to "safer crlf" conversion kicking in when it
shouldn't.

* jc/renormalize-merge-kill-safer-crlf:
  merge: avoid "safer crlf" during recording of merge results
  convert: unify the "auto" handling of CRLF

7 years agoMerge branch 'jk/common-main' into maint
Junio C Hamano [Fri, 9 Sep 2016 04:35:50 +0000 (21:35 -0700)] 
Merge branch 'jk/common-main' into maint

There are certain house-keeping tasks that need to be performed at
the very beginning of any Git program, and programs that are not
built-in commands had to do them exactly the same way as "git"
potty does.  It was easy to make mistakes in one-off standalone
programs (like test helpers).  A common "main()" function that
calls cmd_main() of individual program has been introduced to
make it harder to make mistakes.

* jk/common-main:
  mingw: declare main()'s argv as const
  common-main: call git_setup_gettext()
  common-main: call restore_sigpipe_to_default()
  common-main: call sanitize_stdfds()
  common-main: call git_extract_argv0_path()
  add an extra level of indirection to main()

7 years agodiff: remove dead code
Stefan Beller [Wed, 7 Sep 2016 23:36:48 +0000 (16:36 -0700)] 
diff: remove dead code

When `len < 1`, len has to be 0 or negative, emit_line will then remove the
first character and by then `len` would be negative. As this doesn't
happen, it is safe to assume it is dead code.

This continues to simplify the code, which was started in b8d9c1a66b
(2009-09-03,  diff.c: the builtin_diff() deals with only two-file
comparison).

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff: omit found pointer from emit_callback
Stefan Beller [Wed, 7 Sep 2016 23:36:46 +0000 (16:36 -0700)] 
diff: omit found pointer from emit_callback

We keep the actual data in the diff options, which are just as accessible.
Remove the pointer stored in struct emit_callback for readability.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff.c: use diff_options directly
Stefan Beller [Wed, 7 Sep 2016 23:36:44 +0000 (16:36 -0700)] 
diff.c: use diff_options directly

The value of `ecbdata->opt` is accessible via the short variable `o`
already, so let's use that instead.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agotransport: report missing submodule pushes consistently on stderr
Stefan Beller [Tue, 23 Aug 2016 21:40:08 +0000 (14:40 -0700)] 
transport: report missing submodule pushes consistently on stderr

The surrounding advice is printed to stderr, but the list of submodules
is not. Make the report consistent by reporting everything to stderr.

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agocheckout: add some spaces between code and comment
Nguyễn Thái Ngọc Duy [Wed, 7 Sep 2016 11:19:39 +0000 (18:19 +0700)] 
checkout: add some spaces between code and comment

Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agounpack-trees: do not capitalize "working"
Alex Henrie [Thu, 8 Sep 2016 04:34:53 +0000 (22:34 -0600)] 
unpack-trees: do not capitalize "working"

In English, only proper nouns are capitalized.

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-merge-octopus: do not capitalize "octopus"
Alex Henrie [Thu, 8 Sep 2016 04:34:40 +0000 (22:34 -0600)] 
git-merge-octopus: do not capitalize "octopus"

In English, only proper nouns are capitalized.

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-rebase--interactive: fix English grammar
Alex Henrie [Thu, 8 Sep 2016 04:34:17 +0000 (22:34 -0600)] 
git-rebase--interactive: fix English grammar

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agocat-file: put spaces around pipes in usage string
Alex Henrie [Thu, 8 Sep 2016 04:34:00 +0000 (22:34 -0600)] 
cat-file: put spaces around pipes in usage string

This makes the style a little more consistent with other usage strings,
and will resolve a warning at
https://www.softcatala.org/recursos/quality/git.html

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoam: put spaces around pipe in usage string
Alex Henrie [Thu, 8 Sep 2016 04:33:08 +0000 (22:33 -0600)] 
am: put spaces around pipe in usage string

This makes the style a little more consistent with other usage strings,
and will resolve a warning at
https://www.softcatala.org/recursos/quality/git.html

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoremote-curl: handle URLs without protocol
Jeff King [Wed, 7 Sep 2016 20:06:42 +0000 (16:06 -0400)] 
remote-curl: handle URLs without protocol

Generally remote-curl would never see a URL that did not
have "proto:" at the beginning, as that is what tells git to
run the "git-remote-proto" helper (and git-remote-http, etc,
are aliases for git-remote-curl).

However, the special syntax "proto::something" will run
git-remote-proto with only "something" as the URL. So a
malformed URL like:

  http::/example.com/repo.git

will feed the URL "/example.com/repo.git" to
git-remote-http. The resulting URL has no protocol, but the
code added by 372370f (http: use credential API to handle
proxy authentication, 2016-01-26) does not handle this case
and segfaults.

For the purposes of this code, we don't really care what the
exact protocol; only whether or not it is https. So let's
just assume that a missing protocol is not, and curl will
handle the real error (which is that the URL is nonsense).

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agorebase -i: improve advice on bad instruction lines
Ralf Thielow [Tue, 6 Sep 2016 18:59:18 +0000 (20:59 +0200)] 
rebase -i: improve advice on bad instruction lines

If we found bad instruction lines in the instruction sheet
of interactive rebase, we give the user advice on how to
fix it.  However, we don't tell the user what to do afterwards.
Give the user advice to run 'git rebase --continue' after
the fix.

Signed-off-by: Ralf Thielow <ralf.thielow@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agopack-objects: walk tag chains for --include-tag
Jeff King [Mon, 5 Sep 2016 21:52:26 +0000 (17:52 -0400)] 
pack-objects: walk tag chains for --include-tag

When pack-objects is given --include-tag, it peels each tag
ref down to a non-tag object, and if that non-tag object is
going to be packed, we include the tag, too. But what
happens if we have a chain of tags (e.g., tag "A" points to
tag "B", which points to commit "C")?

We'll peel down to "C" and realize that we want to include
tag "A", but we do not ever consider tag "B", leading to a
broken pack (assuming "B" was not otherwise selected).
Instead, we have to walk the whole chain, adding any tags we
find to the pack.

Interestingly, it doesn't seem possible to trigger this
problem with "git fetch", but you can with "git clone
--single-branch". The reason is that we generate the correct
pack when the client explicitly asks for "A" (because we do
a real reachability analysis there), and "fetch" is more
willing to do so. There are basically two cases:

  1. If "C" is already a ref tip, then the client can deduce
     that it needs "A" itself (via find_non_local_tags), and
     will ask for it explicitly rather than relying on the
     include-tag capability. Everything works.

  2. If "C" is not already a ref tip, then we hope for
     include-tag to send us the correct tag. But it doesn't;
     it generates a broken pack. However, the next step is
     to do a follow-up run of find_non_local_tags(),
     followed by fetch_refs() to backfill any tags we
     learned about.

     In the normal case, fetch_refs() calls quickfetch(),
     which does a connectivity check and sees we have no
     new objects to fetch. We just write the refs.

     But for the broken-pack case, the connectivity check
     fails, and quickfetch will follow-up with the remote,
     asking explicitly for each of the ref tips. This picks
     up the missing object in a new pack.

For a regular "git clone", we are similarly OK, because we
explicitly request all of the tag refs, and get a correct
pack. But with "--single-branch", we kick in tag
auto-following via "include-tag", but do _not_ do a
follow-up backfill. We just take whatever the server sent us
via include-tag and write out tag refs for any tag objects
we were sent. So prior to c6807a4 (clone: open a shortcut
for connectivity check, 2013-05-26), we actually claimed the
clone was a success, but the result was silently
corrupted!  Since c6807a4, index-pack's connectivity
check catches this case, and we correctly complain.

The included test directly checks that pack-objects does not
generate a broken pack, but also confirms that "clone
--single-branch" does not hit the bug.

Note that tag chains introduce another interesting question:
if we are packing the tag "B" but not the commit "C", should
"A" be included?

Both before and after this patch, we do not include "A",
because the initial peel_ref() check only knows about the
bottom-most level, "C". To realize that "B" is involved at
all, we would have to switch to an incremental peel, in
which we examine each tagged object, asking if it is being
packed (and including the outer tag if so).

But that runs contrary to the optimizations in peel_ref(),
which avoid accessing the objects at all, in favor of using
the value we pull from packed-refs. It's OK to walk the
whole chain once we know we're going to include the tag (we
have to access it anyway, so the effort is proportional to
the pack we're generating). But for the initial selection,
we have to look at every ref. If we're only packing a few
objects, we'd still have to parse every single referenced
tag object just to confirm that it isn't part of a tag
chain.

This could be addressed if packed-refs stored the complete
tag chain for each peeled ref (in most cases, this would be
the same cost as now, as each "chain" is only a single
link). But given the size of that project, it's out of scope
for this fix (and probably nobody cares enough anyway, as
it's such an obscure situation). This commit limits itself
to just avoiding the creation of a broken pack.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5305: simplify packname handling
Jeff King [Mon, 5 Sep 2016 21:52:22 +0000 (17:52 -0400)] 
t5305: simplify packname handling

We generate a series of packfiles test-1-$pack,
test-2-$pack, with different properties and then examine
them. However we always store the packname generated by
pack-objects in the variable packname_1. This probably was
meant to be packname_2 in the second test, but it turns out
that it doesn't matter: once we are done with the first
pack, we can just keep using the same $packname variable.

So let's drop the confusing "_1" parameter. At the same
time, let's give test-1 and test-2 more descriptive names,
which can help keep them straight (note that we _could_
likewise overwrite the packfiles in each test, but by using
separate filenames, we are sure that test 2 does not
accidentally use the packfile from test 1).

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5305: use "git -C"
Jeff King [Mon, 5 Sep 2016 21:52:17 +0000 (17:52 -0400)] 
t5305: use "git -C"

This test unpacks objects into a separate repository, and
accesses it by setting GIT_DIR in a subshell. We can do the
same thing these days by using "git init <repo>" and "git
-C". In most cases this is shorter, though when there are
multiple commands, we may end up repeating the "-C".

However, this repetition can actually be a good thing. This
patch also fixes a bug introduced by 512477b (tests: use
"env" to run commands with temporary env-var settings,
2014-03-18). That commit essentially converted:

   (GIT_DIR=...; export GIT_DIR
    cmd1 &&
    cmd2)

into:

   (GIT_DIR=... cmd1 &&
    cmd2)

which obviously loses the GIT_DIR setting for cmd2 (we never
noticed the bug because it simply runs "cmd2" in the parent
repo, which means we were simply failing to test anything
interesting). By using "git -C" rather than a subshell, it
becomes quite obvious where each command is supposed to be
running.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5305: drop "dry-run" of unpack-objects
Jeff King [Mon, 5 Sep 2016 21:52:14 +0000 (17:52 -0400)] 
t5305: drop "dry-run" of unpack-objects

For each test we do a dry-run of unpack-objects, followed by
a real run, followed by confirming that it contained the
objects we expected. The dry-run is telling us nothing, as
any errors it encounters would be found in the real run.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5305: move cleanup into test block
Jeff King [Mon, 5 Sep 2016 21:52:10 +0000 (17:52 -0400)] 
t5305: move cleanup into test block

We usually try to avoid doing any significant actions
outside of test blocks. Although "rm -rf" is unlikely to
either fail or to generate output, moving these to the
point of use makes it more clear that they are part of the
overall setup of "clone.git".

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5551-http-fetch-smart.sh: use the GIT_TRACE_CURL environment var
Elia Pinto [Mon, 5 Sep 2016 10:24:44 +0000 (10:24 +0000)] 
t5551-http-fetch-smart.sh: use the GIT_TRACE_CURL environment var

Use the new GIT_TRACE_CURL environment variable instead
of the deprecated GIT_CURL_VERBOSE.

Signed-off-by: Elia Pinto <gitter.spiros@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5550-http-fetch-dumb.sh: use the GIT_TRACE_CURL environment var
Elia Pinto [Mon, 5 Sep 2016 19:19:40 +0000 (19:19 +0000)] 
t5550-http-fetch-dumb.sh: use the GIT_TRACE_CURL environment var

Use the new GIT_TRACE_CURL environment variable instead
of the deprecated GIT_CURL_VERBOSE.

Signed-off-by: Elia Pinto <gitter.spiros@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agotest-lib.sh: preserve GIT_TRACE_CURL from the environment
Elia Pinto [Mon, 5 Sep 2016 10:24:42 +0000 (10:24 +0000)] 
test-lib.sh: preserve GIT_TRACE_CURL from the environment

Turning on this variable can be useful when debugging http
tests. It can break a few tests in t5541 if not set
to an absolute path but it is not a variable
that the user is likely to have enabled accidentally.

Signed-off-by: Elia Pinto <gitter.spiros@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot5541-http-push-smart.sh: use the GIT_TRACE_CURL environment var
Elia Pinto [Mon, 5 Sep 2016 10:24:41 +0000 (10:24 +0000)] 
t5541-http-push-smart.sh: use the GIT_TRACE_CURL environment var

Use the new GIT_TRACE_CURL environment variable instead
of the deprecated GIT_CURL_VERBOSE.

Signed-off-by: Elia Pinto <gitter.spiros@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot6026-merge-attr: clean up background process at end of test case
Johannes Sixt [Wed, 7 Sep 2016 06:10:41 +0000 (08:10 +0200)] 
t6026-merge-attr: clean up background process at end of test case

The process spawned in the hook uses the test's trash directory as CWD.
As long as it is alive, the directory cannot be removed on Windows.
Although the test succeeds, the 'test_done' that follows produces an
error message and leaves the trash directory around. Kill the process
before the test case advances.

Helped-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agot9903: fix broken && chain
Johannes Sixt [Mon, 5 Sep 2016 19:00:47 +0000 (21:00 +0200)] 
t9903: fix broken && chain

We might wonder why our && chain check does not catch this case:
The && chain check uses a strange exit code with the expectation that
the second or later part of a broken && chain would not exit with this
particular code.

This expectation does not work in this case because __git_ps1, being
the first command in the second part of the broken && chain, records
the current exit code, does its work, and finally returns to the caller
with the recorded exit code. This fools our && chain check.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Reviewed-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agointroduce hex2chr() for converting two hexadecimal digits to a character
René Scharfe [Sat, 3 Sep 2016 15:59:20 +0000 (17:59 +0200)] 
introduce hex2chr() for converting two hexadecimal digits to a character

Add and use a helper function that decodes the char value of two
hexadecimal digits.  It returns a negative number on error, avoids
running over the end of the given string and doesn't shift negative
values.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agocompat: move strdup(3) replacement to its own file
René Scharfe [Sat, 3 Sep 2016 15:59:15 +0000 (17:59 +0200)] 
compat: move strdup(3) replacement to its own file

Move our implementation of strdup(3) out of compat/nedmalloc/ and
allow it to be used independently from USE_NED_ALLOCATOR.  The
original nedmalloc doesn't come with strdup() and doesn't need it.
Only _users_ of nedmalloc need it, which was added when we imported
it to our compat/ hierarchy.

This reduces the difference of our copy of nedmalloc from the
original, making it easier to update, and allows for easier testing
and reusing of our version of strdup().

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoMerge branch 'sy/i18n' of git-gui
Junio C Hamano [Wed, 7 Sep 2016 17:23:51 +0000 (10:23 -0700)] 
Merge branch 'sy/i18n' of git-gui

* 'sy/i18n' of git-gui:
  git-gui: update Japanese information
  git-gui: update Japanese translation
  git-gui: add Japanese language code
  git-gui: apply po template to Japanese translation
  git-gui: consistently use the same word for "blame" in Japanese
  git-gui: consistently use the same word for "remote" in Japanese

7 years agogit-gui: update Japanese information
Satoshi Yasushima [Tue, 6 Sep 2016 16:02:21 +0000 (01:02 +0900)] 
git-gui: update Japanese information

Signed-off-by: Satoshi Yasushima <s.yasushima@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-gui: update Japanese translation
Satoshi Yasushima [Tue, 6 Sep 2016 16:02:20 +0000 (01:02 +0900)] 
git-gui: update Japanese translation

Signed-off-by: Satoshi Yasushima <s.yasushima@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-gui: add Japanese language code
Satoshi Yasushima [Tue, 6 Sep 2016 16:02:19 +0000 (01:02 +0900)] 
git-gui: add Japanese language code

Signed-off-by: Satoshi Yasushima <s.yasushima@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-gui: apply po template to Japanese translation
Satoshi Yasushima [Tue, 6 Sep 2016 16:02:18 +0000 (01:02 +0900)] 
git-gui: apply po template to Japanese translation

Signed-off-by: Satoshi Yasushima <s.yasushima@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-gui: consistently use the same word for "blame" in Japanese
Satoshi Yasushima [Tue, 6 Sep 2016 16:02:17 +0000 (01:02 +0900)] 
git-gui: consistently use the same word for "blame" in Japanese

Signed-off-by: Satoshi Yasushima <s.yasushima@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agogit-gui: consistently use the same word for "remote" in Japanese
Satoshi Yasushima [Tue, 6 Sep 2016 16:02:16 +0000 (01:02 +0900)] 
git-gui: consistently use the same word for "remote" in Japanese

Signed-off-by: Satoshi Yasushima <s.yasushima@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoxdiff: remove unneeded declarations
Stefan Beller [Sat, 3 Sep 2016 03:16:48 +0000 (20:16 -0700)] 
xdiff: remove unneeded declarations

Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agol10n: pt_PT: update Portuguese repository info
Vasco Almeida [Sat, 3 Sep 2016 12:10:32 +0000 (12:10 +0000)] 
l10n: pt_PT: update Portuguese repository info

Change Portuguese l10n leadership to Vasco Almeida.

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>
7 years agol10n: pt_PT: update Portuguese translation
Vasco Almeida [Sat, 3 Sep 2016 12:02:22 +0000 (12:02 +0000)] 
l10n: pt_PT: update Portuguese translation

Signed-off-by: Vasco Almeida <vascomalmeida@sapo.pt>
7 years agoGit 2.10 v2.10.0
Junio C Hamano [Fri, 2 Sep 2016 16:05:47 +0000 (09:05 -0700)] 
Git 2.10

Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agosymbolic-ref -d: do not allow removal of HEAD
Junio C Hamano [Thu, 1 Sep 2016 22:38:02 +0000 (15:38 -0700)] 
symbolic-ref -d: do not allow removal of HEAD

If you delete the symbolic-ref HEAD from a repository, Git no longer
considers the repository valid, and even "git symbolic-ref HEAD
refs/heads/master" would not be able to recover from that state
(although "git init" can, but that is a sure sign that you are
talking about a "broken" repository).

In the spirit similar to afe5d3d5 ("symbolic ref: refuse non-ref
targets in HEAD", 2009-01-29), forbid removal of HEAD to avoid
corrupting a repository.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoMerge tag 'l10n-2.10.0-rnd2.2' of git://github.com/git-l10n/git-po
Junio C Hamano [Fri, 2 Sep 2016 15:48:14 +0000 (08:48 -0700)] 
Merge tag 'l10n-2.10.0-rnd2.2' of git://github.com/git-l10n/git-po

l10n-2.10.0-rnd2.2

* tag 'l10n-2.10.0-rnd2.2' of git://github.com/git-l10n/git-po:
  l10n: Updated Vietnamese translation for v2.10.0-rc2 (2757t)

7 years agoMerge branch 'master' of https://github.com/vnwildman/git
Jiang Xin [Fri, 2 Sep 2016 13:29:48 +0000 (21:29 +0800)] 
Merge branch 'master' of https://github.com/vnwildman/git

* 'master' of https://github.com/vnwildman/git:
  l10n: Updated Vietnamese translation for v2.10.0-rc2 (2757t)

7 years agosubmodule: avoid auto-discovery in prepare_submodule_repo_env()
Junio C Hamano [Thu, 1 Sep 2016 20:51:48 +0000 (13:51 -0700)] 
submodule: avoid auto-discovery in prepare_submodule_repo_env()

The function is used to set up the environment variable used in a
subprocess we spawn in a submodule directory.  The callers set up a
child_process structure, find the working tree path of one submodule
and set .dir field to it, and then use start_command() API to spawn
the subprocess like "status", "fetch", etc.

When this happens, we expect that the ".git" (either a directory or
a gitfile that points at the real location) in the current working
directory of the subprocess MUST be the repository for the submodule.

If this ".git" thing is a corrupt repository, however, because
prepare_submodule_repo_env() unsets GIT_DIR and GIT_WORK_TREE, the
subprocess will see ".git", thinks it is not a repository, and
attempt to find one by going up, likely to end up in finding the
repository of the superproject.  In some codepaths, this will cause
a command run with the "--recurse-submodules" option to recurse
forever.

By exporting GIT_DIR=.git, disable the auto-discovery logic in the
subprocess, which would instead stop it and report an error.

The test illustrates existing problems in a few callsites of this
function.  Without this fix, "git fetch --recurse-submodules", "git
status" and "git diff" keep recursing forever.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agocolor_parse_mem: initialize "struct color" temporary
Jeff King [Wed, 31 Aug 2016 03:43:07 +0000 (23:43 -0400)] 
color_parse_mem: initialize "struct color" temporary

Compiling color.c with gcc 6.2.0 using -O3 produces some
-Wmaybe-uninitialized false positives:

    color.c: In function ‘color_parse_mem’:
    color.c:189:10: warning: ‘bg.blue’ may be used uninitialized in this function [-Wmaybe-uninitialized]
       out += xsnprintf(out, len, "%c8;2;%d;%d;%d", type,
              ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          c->red, c->green, c->blue);
          ~~~~~~~~~~~~~~~~~~~~~~~~~~
    color.c:208:15: note: ‘bg.blue’ was declared here
      struct color bg = { COLOR_UNSPECIFIED };
                   ^~
    [ditto for bg.green, bg.red, fg.blue, etc]

This is doubly confusing, because the declaration shows it
being initialized! Even though we do not explicitly
initialize the color components, an incomplete initializer
sets the unmentioned members to zero.

What the warning doesn't show is that we later do this:

  struct color c;
  if (!parse_color(&c, ...)) {
          if (fg.type == COLOR_UNSPECIFIED)
                fg = c;
          ...
  }

gcc is clever enough to realize that a struct assignment
from an uninitialized variable taints the destination. But
unfortunately it's _not_ clever enough to realize that we
only look at those members when type is set to COLOR_RGB, in
which case they are always initialized.

With -O2, gcc does not look into parse_color() and must
assume that "c" emerges fully initialized. With -O3, it
inlines parse_color(), and learns just enough to get
confused.

We can silence the false positive by initializing the
temporary "c". This also future-proofs us against
violating the type assumptions (the result would probably
still be buggy, but in a deterministic way).

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoerror_errno: use constant return similar to error()
Jeff King [Wed, 31 Aug 2016 03:41:22 +0000 (23:41 -0400)] 
error_errno: use constant return similar to error()

Commit e208f9c (make error()'s constant return value more
visible, 2012-12-15) introduced some macro trickery to make
the constant return from error() more visible to callers,
which in turn can help gcc produce better warnings (and
possibly even better code).

Later, fd1d672 (usage.c: add warning_errno() and
error_errno(), 2016-05-08) introduced another variant, and
subsequent commits converted some uses of error() to
error_errno(), losing the magic from e208f9c for those
sites.

As a result, compiling vcs-svn/svndiff.c with "gcc -O3"
produces -Wmaybe-uninitialized false positives (at least
with gcc 6.2.0). Let's give error_errno() the same
treatment, which silences these warnings.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoA few more fixes before the final 2.10
Junio C Hamano [Wed, 31 Aug 2016 17:21:05 +0000 (10:21 -0700)] 
A few more fixes before the final 2.10

Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoMerge tag 'l10n-2.10.0-rnd2' of git://github.com/git-l10n/git-po
Junio C Hamano [Wed, 31 Aug 2016 17:04:14 +0000 (10:04 -0700)] 
Merge tag 'l10n-2.10.0-rnd2' of git://github.com/git-l10n/git-po

l10n-2.10.0-rnd2

* tag 'l10n-2.10.0-rnd2' of git://github.com/git-l10n/git-po:
  l10n: zh_CN: for git v2.10.0 l10n round 2
  l10n: ca.po: update translation
  l10n: fr.po v2.10.0-rc2
  l10n: sv.po: Update Swedish translation (2757t0f0u)
  l10n: git.pot: v2.10.0 round 2 (12 new, 44 removed)
  l10n: Updated Vietnamese translation for v2.10.0 (2789t)
  l10n: pt_PT: update Portuguese translation
  l10n: pt_PT: merge git.pot
  l10n: ko.po: Update Korean translation
  l10n: git.pot: v2.10.0 round 1 (248 new, 56 removed)

7 years agoMerge branch 'ls/packet-line-protocol-doc-fix'
Junio C Hamano [Wed, 31 Aug 2016 17:03:51 +0000 (10:03 -0700)] 
Merge branch 'ls/packet-line-protocol-doc-fix'

Correct an age-old calco (is that a typo-like word for calc)
in the documentation.

* ls/packet-line-protocol-doc-fix:
  pack-protocol: fix maximum pkt-line size

7 years agoMerge branch 'mh/blame-worktree'
Junio C Hamano [Wed, 31 Aug 2016 17:03:50 +0000 (10:03 -0700)] 
Merge branch 'mh/blame-worktree'

* mh/blame-worktree:
  blame: fix segfault on untracked files

7 years agoMerge branch 'kw/patch-ids-optim'
Junio C Hamano [Wed, 31 Aug 2016 17:03:49 +0000 (10:03 -0700)] 
Merge branch 'kw/patch-ids-optim'

* kw/patch-ids-optim:
  p3400: make test script executable

7 years agodiff-highlight: avoid highlighting combined diffs
Jeff King [Wed, 31 Aug 2016 05:05:38 +0000 (01:05 -0400)] 
diff-highlight: avoid highlighting combined diffs

The algorithm in diff-highlight only understands how to look
at two sides of a diff; it cannot correctly handle combined
diffs with multiple preimages. Often highlighting does not
trigger at all for these diffs because the line counts do
not match up.  E.g., if we see:

  - ours
   -theirs
  ++resolved

we would not bother highlighting; it naively looks like a
single line went away, and then a separate hunk added
another single line.

But of course there are exceptions. E.g., if the other side
deleted the line, we might see:

  - ours
  ++resolved

which looks like we dropped " ours" and added "+resolved".
This is only a small highlighting glitch (we highlight the
space and the "+" along with the content), but it's also the
tip of the iceberg. Even if we learned to find the true
content here (by noticing we are in a 3-way combined diff
and marking _two_ characters from the front of the line as
uninteresting), there are other more complicated cases where
we really do need to handle a 3-way hunk.

Let's just punt for now; we can recognize combined diffs by
the presence of extra "@" symbols in the hunk header, and
treat them as non-diff content.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff-highlight: add multi-byte tests
Jeff King [Wed, 31 Aug 2016 05:03:10 +0000 (01:03 -0400)] 
diff-highlight: add multi-byte tests

Now that we have a test suite for diff highlight, we can
show off the improvements from 8d00662 (diff-highlight: do
not split multibyte characters, 2015-04-03).

While we're at it, we can also add another case that
_doesn't_ work: combining code points are treated as their
own unit, which means that we may stick colors between them
and the character they are modifying (with the result that
the color is not shown in an xterm, though it's possible
that other terminals err the other way, and show the color
but not the accent).  There's no fix here, but let's
document it as a failure.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff-highlight: ignore test cruft
Jeff King [Wed, 31 Aug 2016 05:02:53 +0000 (01:02 -0400)] 
diff-highlight: ignore test cruft

These are the same as in the normal t/.gitignore, with the
exception of ".prove", as our Makefile does not support it.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agotest-lib: drop PID from test-results/*.count
Jeff King [Tue, 30 Aug 2016 08:43:57 +0000 (04:43 -0400)] 
test-lib: drop PID from test-results/*.count

Each test run generates a "count" file in t/test-results
that stores the number of successful, failed, etc tests.
If you run "t1234-foo.sh", that file is named as
"t/test-results/t1234-foo-$$.count"

The addition of the PID there is serving no purpose, and
makes analysis of the count files harder.

The presence of the PID dates back to 2d84e9f (Modify
test-lib.sh to output stats to t/test-results/*,
2008-06-08), but no reasoning is given there. Looking at the
current code, we can see that other files we write to
test-results (like *.exit and *.out) do _not_ have the PID
included. So the presence of the PID does not meaningfully
allow one to store the results from multiple runs anyway.

Moreover, anybody wishing to read the *.count files to
aggregate results has to deal with the presence of multiple
files for a given test (and figure out which one is the most
recent based on their timestamps!). The only consumer of
these files is the aggregate.sh script, which arguably gets
this wrong. If a test is run multiple times, its counts will
appear multiple times in the total (I say arguably only
because the desired semantics aren't documented anywhere,
but I have trouble seeing how this behavior could be
useful).

So let's just drop the PID, which fixes aggregate.sh, and
will make new features based around the count files easier
to write.

Note that since the count-file may already exist (when
re-running a test), we also switch the "cat" from appending
to truncating. The use of append here was pointless in the
first place, as we expected to always write to a unique file.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agopack-protocol: fix maximum pkt-line size
Lars Schneider [Mon, 29 Aug 2016 17:55:09 +0000 (19:55 +0200)] 
pack-protocol: fix maximum pkt-line size

According to LARGE_PACKET_MAX in pkt-line.h the maximal length of a
pkt-line packet is 65520 bytes. The pkt-line header takes 4 bytes and
therefore the pkt-line data component must not exceed 65516 bytes.

Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agol10n: zh_CN: for git v2.10.0 l10n round 2
Jiang Xin [Sun, 28 Aug 2016 02:18:12 +0000 (10:18 +0800)] 
l10n: zh_CN: for git v2.10.0 l10n round 2

Update 215 translations (2757t0f0u) for git v2.10.0-rc2.

Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
7 years agop3400: make test script executable
René Scharfe [Sun, 28 Aug 2016 12:39:27 +0000 (14:39 +0200)] 
p3400: make test script executable

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff-highlight: add support for --graph output
Brian Henderson [Mon, 29 Aug 2016 17:33:47 +0000 (10:33 -0700)] 
diff-highlight: add support for --graph output

Signed-off-by: Brian Henderson <henderson.bj@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff-highlight: add failing test for handling --graph output
Brian Henderson [Mon, 29 Aug 2016 17:33:46 +0000 (10:33 -0700)] 
diff-highlight: add failing test for handling --graph output

Signed-off-by: Brian Henderson <henderson.bj@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agodiff-highlight: add some tests
Brian Henderson [Mon, 29 Aug 2016 17:33:45 +0000 (10:33 -0700)] 
diff-highlight: add some tests

Signed-off-by: Brian Henderson <henderson.bj@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agoblame: fix segfault on untracked files
Thomas Gummerer [Sat, 27 Aug 2016 20:01:50 +0000 (21:01 +0100)] 
blame: fix segfault on untracked files

Since 3b75ee9 ("blame: allow to blame paths freshly added to the index",
2016-07-16) git blame also looks at the index to determine if there is a
file that was freshly added to the index.

cache_name_pos returns -pos - 1 in case there is no match is found, or
if the name matches, but the entry has a stage other than 0.  As git
blame should work for unmerged files, it uses strcmp to determine
whether the name of the returned position matches, in which case the
file exists, but is merely unmerged, or if the file actually doesn't
exist in the index.

If the repository is empty, or if the file would lexicographically be
sorted as the last file in the repository, -cache_name_pos - 1 is
outside of the length of the active_cache array, causing git blame to
segfault.  Guard against that, and die() normally to restore the old
behaviour.

Reported-by: Simon Ruderich <simon@ruderich.org>
Signed-off-by: Thomas Gummerer <t.gummerer@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
7 years agol10n: ca.po: update translation
Alex Henrie [Sun, 28 Aug 2016 16:32:56 +0000 (10:32 -0600)] 
l10n: ca.po: update translation

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