git/git.git
2 years agolist-objects: check if filter is NULL before using
Jonathan Tan [Mon, 11 Jun 2018 21:51:26 +0000 (14:51 -0700)]
list-objects: check if filter is NULL before using

In partial_clone_get_default_filter_spec(), the
core_partial_clone_filter_default variable may be NULL; ensure that it
is not NULL before using it.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoupload-pack: disable object filtering when disabled by config
Jonathan Nieder [Wed, 28 Mar 2018 20:33:03 +0000 (13:33 -0700)]
upload-pack: disable object filtering when disabled by config

When upload-pack gained partial clone support (v2.17.0-rc0~132^2~12,
2017-12-08), it was guarded by the uploadpack.allowFilter config item
to allow server operators to control when they start supporting it.

That config item didn't go far enough, though: it controls whether the
'filter' capability is advertised, but if a (custom) client ignores
the capability advertisement and passes a filter specification anyway,
the server would handle that despite allowFilter being false.

This is particularly significant if a security bug is discovered in
this new experimental partial clone code.  Installations without
uploadpack.allowFilter ought not to be affected since they don't
intend to support partial clone, but they would be swept up into being
vulnerable.

Simplify and limit the attack surface by making uploadpack.allowFilter
disable the feature, not just the advertisement of it.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agounpack-trees: release oid_array after use in check_updates()
René Scharfe [Sun, 25 Mar 2018 16:31:48 +0000 (18:31 +0200)]
unpack-trees: release oid_array after use in check_updates()

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agot5616: test bulk prefetch after partial fetch
Jeff Hostetler [Fri, 8 Dec 2017 15:58:51 +0000 (15:58 +0000)]
t5616: test bulk prefetch after partial fetch

Add test to t5616 to bulk fetch missing objects following
a partial fetch.  A technique like this could be used in
a pre-command hook for example.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch: inherit filter-spec from partial clone
Jeff Hostetler [Fri, 8 Dec 2017 15:58:50 +0000 (15:58 +0000)]
fetch: inherit filter-spec from partial clone

Teach (partial) fetch to inherit the filter-spec used by
the partial clone.  Extend --no-filter to override this
inheritance.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agot5616: end-to-end tests for partial clone
Jeff Hostetler [Fri, 8 Dec 2017 15:58:49 +0000 (15:58 +0000)]
t5616: end-to-end tests for partial clone

Additional end-to-end tests for partial clone.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch-pack: restore save_commit_buffer after use
Jonathan Tan [Fri, 8 Dec 2017 15:58:48 +0000 (15:58 +0000)]
fetch-pack: restore save_commit_buffer after use

In fetch-pack, the global variable save_commit_buffer is set to 0, but
not restored to its original value after use.

In particular, if show_log() (in log-tree.c) is invoked after
fetch_pack() in the same process, show_log() will return before printing
out the commit message (because the invocation to
get_cached_commit_buffer() returns NULL, because the commit buffer was
not saved). I discovered this when attempting to run "git log -S" in a
partial clone, triggering the case where revision walking lazily loads
missing objects.

Therefore, restore save_commit_buffer to its original value after use.

An alternative to solve the problem I had is to replace
get_cached_commit_buffer() with get_commit_buffer(). That invocation was
introduced in commit a97934d ("use get_cached_commit_buffer where
appropriate", 2014-06-13) to replace "commit->buffer" introduced in
commit 3131b71 ("Add "--show-all" revision walker flag for debugging",
2008-02-13). In the latter commit, the commit author seems to be
deciding between not showing an unparsed commit at all and showing an
unparsed commit without the message (which is what the commit does), and
did not mention parsing the unparsed commit, so I prefer to preserve the
existing behavior.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agounpack-trees: batch fetching of missing blobs
Jonathan Tan [Fri, 8 Dec 2017 15:58:47 +0000 (15:58 +0000)]
unpack-trees: batch fetching of missing blobs

When running checkout, first prefetch all blobs that are to be updated
but are missing. This means that only one pack is downloaded during such
operations, instead of one per missing blob.

This operates only on the blob level - if a repository has a missing
tree, they are still fetched one at a time.

This does not use the delayed checkout mechanism introduced in commit
2841e8f ("convert: add "status=delayed" to filter process protocol",
2017-06-30) due to significant conceptual differences - in particular,
for partial clones, we already know what needs to be fetched based on
the contents of the local repo alone, whereas for status=delayed, it is
the filter process that tells us what needs to be checked in the end.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoclone: partial clone
Jonathan Tan [Fri, 8 Dec 2017 15:58:46 +0000 (15:58 +0000)]
clone: partial clone

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agopartial-clone: define partial clone settings in config
Jeff Hostetler [Fri, 8 Dec 2017 15:58:45 +0000 (15:58 +0000)]
partial-clone: define partial clone settings in config

Create get and set routines for "partial clone" config settings.
These will be used in a future commit by clone and fetch to
remember the promisor remote and the default filter-spec.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch: support filters
Jeff Hostetler [Fri, 8 Dec 2017 15:58:44 +0000 (15:58 +0000)]
fetch: support filters

Teach fetch to support filters. This is only allowed for the remote
configured in extensions.partialcloneremote.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch: refactor calculation of remote list
Jonathan Tan [Fri, 8 Dec 2017 15:58:43 +0000 (15:58 +0000)]
fetch: refactor calculation of remote list

Separate out the calculation of remotes to be fetched from and the
actual fetching. This will allow us to include an additional step before
the actual fetching in a subsequent commit.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch-pack: test support excluding large blobs
Jonathan Tan [Fri, 8 Dec 2017 15:58:42 +0000 (15:58 +0000)]
fetch-pack: test support excluding large blobs

Created tests to verify fetch-pack and upload-pack support
for excluding large blobs using --filter=blobs:limit=<n>
parameter.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch-pack: add --no-filter
Jeff Hostetler [Fri, 8 Dec 2017 15:58:41 +0000 (15:58 +0000)]
fetch-pack: add --no-filter

Fixup fetch-pack to accept --no-filter to be consistent with
rev-list and pack-objects.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofetch-pack, index-pack, transport: partial clone
Jeff Hostetler [Fri, 8 Dec 2017 15:58:40 +0000 (15:58 +0000)]
fetch-pack, index-pack, transport: partial clone

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoupload-pack: add object filtering for partial clone
Jeff Hostetler [Fri, 8 Dec 2017 15:58:39 +0000 (15:58 +0000)]
upload-pack: add object filtering for partial clone

Teach upload-pack to negotiate object filtering over the protocol and
to send filter parameters to pack-objects.  This is intended for partial
clone and fetch.

The idea to make upload-pack configurable using uploadpack.allowFilter
comes from Jonathan Tan's work in [1].

[1] https://public-inbox.org/git/f211093280b422c32cc1b7034130072f35c5ed51.1506714999.git.jonathantanmy@google.com/

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agogc: do not repack promisor packfiles
Jonathan Tan [Fri, 8 Dec 2017 15:27:16 +0000 (15:27 +0000)]
gc: do not repack promisor packfiles

Teach gc to stop traversal at promisor objects, and to leave promisor
packfiles alone. This has the effect of only repacking non-promisor
packfiles, and preserves the distinction between promisor packfiles and
non-promisor packfiles.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agorev-list: support termination at promisor objects
Jonathan Tan [Fri, 8 Dec 2017 15:27:15 +0000 (15:27 +0000)]
rev-list: support termination at promisor objects

Teach rev-list to support termination of an object traversal at any
object from a promisor remote (whether one that the local repo also has,
or one that the local repo knows about because it has another promisor
object that references it).

This will be used subsequently in gc and in the connectivity check used
by fetch.

For efficiency, if an object is referenced by a promisor object, and is
in the local repo only as a non-promisor object, object traversal will
not stop there. This is to avoid building the list of promisor object
references.

(In list-objects.c, the case where obj is NULL in process_blob() and
process_tree() do not need to be changed because those happen only when
there is a conflict between the expected type and the existing object.
If the object doesn't exist, an object will be synthesized, which is
fine.)

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agosha1_file: support lazily fetching missing objects
Jonathan Tan [Fri, 8 Dec 2017 15:27:14 +0000 (15:27 +0000)]
sha1_file: support lazily fetching missing objects

Teach sha1_file to fetch objects from the remote configured in
extensions.partialclone whenever an object is requested but missing.

The fetching of objects can be suppressed through a global variable.
This is used by fsck and index-pack.

However, by default, such fetching is not suppressed. This is meant as a
temporary measure to ensure that all Git commands work in such a
situation. Future patches will update some commands to either tolerate
missing objects (without fetching them) or be more efficient in fetching
them.

In order to determine the code changes in sha1_file.c necessary, I
investigated the following:
 (1) functions in sha1_file.c that take in a hash, without the user
     regarding how the object is stored (loose or packed)
 (2) functions in packfile.c (because I need to check callers that know
     about the loose/packed distinction and operate on both differently,
     and ensure that they can handle the concept of objects that are
     neither loose nor packed)

(1) is handled by the modification to sha1_object_info_extended().

For (2), I looked at for_each_packed_object and others.  For
for_each_packed_object, the callers either already work or are fixed in
this patch:
 - reachable - only to find recent objects
 - builtin/fsck - already knows about missing objects
 - builtin/cat-file - warning message added in this commit

Callers of the other functions do not need to be changed:
 - parse_pack_index
   - http - indirectly from http_get_info_packs
   - find_pack_entry_one
     - this searches a single pack that is provided as an argument; the
       caller already knows (through other means) that the sought object
       is in a specific pack
 - find_sha1_pack
   - fast-import - appears to be an optimization to not store a file if
     it is already in a pack
   - http-walker - to search through a struct alt_base
   - http-push - to search through remote packs
 - has_sha1_pack
   - builtin/fsck - already knows about promisor objects
   - builtin/count-objects - informational purposes only (check if loose
     object is also packed)
   - builtin/prune-packed - check if object to be pruned is packed (if
     not, don't prune it)
   - revision - used to exclude packed objects if requested by user
   - diff - just for optimization

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agointroduce fetch-object: fetch one promisor object
Jonathan Tan [Tue, 5 Dec 2017 16:58:49 +0000 (16:58 +0000)]
introduce fetch-object: fetch one promisor object

Introduce fetch-object, providing the ability to fetch one object from a
promisor remote.

This uses fetch-pack. To do this, the transport mechanism has been
updated with 2 flags, "from-promisor" to indicate that the resulting
pack comes from a promisor remote (and thus should be annotated as such
by index-pack), and "no-dependents" to indicate that only the objects
themselves need to be fetched (but fetching additional objects is
nevertheless safe).

Whenever "no-dependents" is used, fetch-pack will refrain from using any
object flags, because it is most likely invoked as part of a dynamic
object fetch by another Git command (which may itself use object flags).
An alternative to this is to leave fetch-pack alone, and instead update
the allocation of flags so that fetch-pack's flags never overlap with
any others, but this will end up shrinking the number of flags available
to nearly every other Git command (that is, every Git command that
accesses objects), so the approach in this commit was used instead.

This will be tested in a subsequent commit.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoindex-pack: refactor writing of .keep files
Jonathan Tan [Tue, 5 Dec 2017 16:58:48 +0000 (16:58 +0000)]
index-pack: refactor writing of .keep files

In a subsequent commit, index-pack will be taught to write ".promisor"
files which are similar to the ".keep" files it knows how to write.
Refactor the writing of ".keep" files, so that the implementation of
writing ".promisor" files becomes easier.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofsck: support promisor objects as CLI argument
Jonathan Tan [Tue, 5 Dec 2017 16:58:47 +0000 (16:58 +0000)]
fsck: support promisor objects as CLI argument

Teach fsck to not treat missing promisor objects provided on the CLI as
an error when extensions.partialclone is set.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofsck: support referenced promisor objects
Jonathan Tan [Tue, 5 Dec 2017 16:58:46 +0000 (16:58 +0000)]
fsck: support referenced promisor objects

Teach fsck to not treat missing promisor objects indirectly pointed to
by refs as an error when extensions.partialclone is set.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofsck: support refs pointing to promisor objects
Jonathan Tan [Tue, 5 Dec 2017 16:58:45 +0000 (16:58 +0000)]
fsck: support refs pointing to promisor objects

Teach fsck to not treat refs referring to missing promisor objects as an
error when extensions.partialclone is set.

For the purposes of warning about no default refs, such refs are still
treated as legitimate refs.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofsck: introduce partialclone extension
Jonathan Tan [Tue, 5 Dec 2017 16:58:44 +0000 (16:58 +0000)]
fsck: introduce partialclone extension

Currently, Git does not support repos with very large numbers of objects
or repos that wish to minimize manipulation of certain blobs (for
example, because they are very large) very well, even if the user
operates mostly on part of the repo, because Git is designed on the
assumption that every referenced object is available somewhere in the
repo storage. In such an arrangement, the full set of objects is usually
available in remote storage, ready to be lazily downloaded.

Teach fsck about the new state of affairs. In this commit, teach fsck
that missing promisor objects referenced from the reflog are not an
error case; in future commits, fsck will be taught about other cases.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoextension.partialclone: introduce partial clone extension
Jonathan Tan [Tue, 5 Dec 2017 16:58:43 +0000 (16:58 +0000)]
extension.partialclone: introduce partial clone extension

Introduce new repository extension option:
    `extensions.partialclone`

See the update to Documentation/technical/repository-version.txt
in this patch for more information.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agorev-list: support --no-filter argument
Jeff Hostetler [Tue, 5 Dec 2017 16:50:14 +0000 (16:50 +0000)]
rev-list: support --no-filter argument

Teach rev-list to support --no-filter to override a
previous --filter=<filter_spec> argument.  This is
to be consistent with commands that use OPT_PARSE
macros.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agolist-objects-filter-options: support --no-filter
Jeff Hostetler [Tue, 5 Dec 2017 16:50:13 +0000 (16:50 +0000)]
list-objects-filter-options: support --no-filter

Teach opt_parse_list_objects_filter() to take --no-filter
option and to free the contents of struct filter_options.
This command line argument will be automatically inherited
by commands using OPT_PARSE_LIST_OBJECTS_FILTER(); this
includes pack-objects.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agolist-objects-filter-options: fix 'keword' typo in comment
Christian Couder [Tue, 5 Dec 2017 16:50:12 +0000 (16:50 +0000)]
list-objects-filter-options: fix 'keword' typo in comment

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agopack-objects: add list-objects filtering
Jeff Hostetler [Tue, 21 Nov 2017 20:58:52 +0000 (20:58 +0000)]
pack-objects: add list-objects filtering

Teach pack-objects to use the filtering provided by the
traverse_commit_list_filtered() interface to omit unwanted
objects from the resulting packfile.

Filtering requires the use of the "--stdout" option.

Add t5317 test.

In the future, we will introduce a "partial clone" mechanism
wherein an object in a repo, obtained from a remote, may
reference a missing object that can be dynamically fetched from
that remote once needed.  This "partial clone" mechanism will
have a way, sometimes slow, of determining if a missing link
is one of the links expected to be produced by this mechanism.

This patch introduces handling of missing objects to help
debugging and development of the "partial clone" mechanism,
and once the mechanism is implemented, for a power user to
perform operations that are missing-object aware without
incurring the cost of checking if a missing link is expected.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Reviewed-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agorev-list: add list-objects filtering support
Jeff Hostetler [Tue, 21 Nov 2017 20:58:51 +0000 (20:58 +0000)]
rev-list: add list-objects filtering support

Teach rev-list to use the filtering provided by the
traverse_commit_list_filtered() interface to omit
unwanted objects from the result.

Object filtering is only allowed when one of the "--objects*"
options are used.

When the "--filter-print-omitted" option is used, the omitted
objects are printed at the end.  These are marked with a "~".
This option can be combined with "--quiet" to get a list of
just the omitted objects.

Add t6112 test.

In the future, we will introduce a "partial clone" mechanism
wherein an object in a repo, obtained from a remote, may
reference a missing object that can be dynamically fetched from
that remote once needed.  This "partial clone" mechanism will
have a way, sometimes slow, of determining if a missing link
is one of the links expected to be produced by this mechanism.

This patch introduces handling of missing objects to help
debugging and development of the "partial clone" mechanism,
and once the mechanism is implemented, for a power user to
perform operations that are missing-object aware without
incurring the cost of checking if a missing link is expected.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Reviewed-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agolist-objects: filter objects in traverse_commit_list
Jeff Hostetler [Tue, 21 Nov 2017 20:58:50 +0000 (20:58 +0000)]
list-objects: filter objects in traverse_commit_list

Create traverse_commit_list_filtered() and add filtering
interface to allow certain objects to be omitted from the
traversal.

Update traverse_commit_list() to be a wrapper for the above
with a null filter to minimize the number of callers that
needed to be changed.

Object filtering will be used in a future commit by rev-list
and pack-objects for partial clone and fetch to omit unwanted
objects from the result.

traverse_bitmap_commit_list() does not work with filtering.
If a packfile bitmap is present, it will not be used.  It
should be possible to extend such support in the future (at
least to simple filters that do not require object pathnames),
but that is beyond the scope of this patch series.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Reviewed-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agooidset: add iterator methods to oidset
Jeff Hostetler [Tue, 21 Nov 2017 20:58:49 +0000 (20:58 +0000)]
oidset: add iterator methods to oidset

Add the usual iterator methods to oidset.
Add oidset_remove().

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Reviewed-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agooidmap: add oidmap iterator methods
Jeff Hostetler [Tue, 21 Nov 2017 20:58:48 +0000 (20:58 +0000)]
oidmap: add oidmap iterator methods

Add the usual map iterator functions to oidmap.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Reviewed-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agodir: allow exclusions from blob in addition to file
Jeff Hostetler [Tue, 21 Nov 2017 20:58:47 +0000 (20:58 +0000)]
dir: allow exclusions from blob in addition to file

Refactor add_excludes() to separate the reading of the
exclude file into a buffer and the parsing of the buffer
into exclude_list items.

Add add_excludes_from_blob_to_list() to allow an exclude
file be specified with an OID without assuming a local
worktree or index exists.

Refactor read_skip_worktree_file_from_index() and add
do_read_blob() to eliminate duplication of preliminary
processing of blob contents.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Reviewed-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoGit 2.15 v2.15.0
Junio C Hamano [Mon, 30 Oct 2017 05:00:44 +0000 (14:00 +0900)]
Git 2.15

Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge tag 'l10n-2.15.0-rnd2.1' of git://github.com/git-l10n/git-po
Junio C Hamano [Mon, 30 Oct 2017 00:32:54 +0000 (09:32 +0900)]
Merge tag 'l10n-2.15.0-rnd2.1' of git://github.com/git-l10n/git-po

l10n for Git 2.15.0 round 2 with Catalan updates

* tag 'l10n-2.15.0-rnd2.1' of git://github.com/git-l10n/git-po:
  l10n: Update Catalan translation

3 years agol10n: Update Catalan translation
Jordi Mas [Wed, 25 Oct 2017 17:50:59 +0000 (19:50 +0200)]
l10n: Update Catalan translation

Signed-off-by: Jordi Mas <jmas@softcatala.org>
3 years agoHopefully final batch before 2.15
Junio C Hamano [Sat, 28 Oct 2017 01:20:30 +0000 (10:20 +0900)]
Hopefully final batch before 2.15

Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge branch 'sg/rev-list-doc-reorder-fix'
Junio C Hamano [Sat, 28 Oct 2017 01:18:42 +0000 (10:18 +0900)]
Merge branch 'sg/rev-list-doc-reorder-fix'

Doc flow fix.

* sg/rev-list-doc-reorder-fix:
  rev-list-options.txt: use correct directional reference

3 years agoMerge branch 'sb/rev-parse-show-superproject-root'
Junio C Hamano [Sat, 28 Oct 2017 01:18:40 +0000 (10:18 +0900)]
Merge branch 'sb/rev-parse-show-superproject-root'

Doc markup fix.

* sb/rev-parse-show-superproject-root:
  docs: fix formatting of rev-parse's --show-superproject-working-tree

3 years agoMerge branch 'ao/path-use-xmalloc'
Junio C Hamano [Sat, 28 Oct 2017 01:18:39 +0000 (10:18 +0900)]
Merge branch 'ao/path-use-xmalloc'

A possible oom error is now caught as a fatal error, instead of
continuing and dereferencing NULL.

* ao/path-use-xmalloc:
  path.c: use xmalloc() in add_to_trie()

3 years agoMerge branch 'np/config-path-doc'
Junio C Hamano [Sat, 28 Oct 2017 01:18:39 +0000 (10:18 +0900)]
Merge branch 'np/config-path-doc'

Doc update.

* np/config-path-doc:
  config doc: clarify "git config --path" example

3 years agodocs: fix formatting of rev-parse's --show-superproject-working-tree
Sebastian Schuberth [Thu, 26 Oct 2017 11:53:37 +0000 (11:53 +0000)]
docs: fix formatting of rev-parse's --show-superproject-working-tree

Signed-off-by: Sebastian Schuberth <sschuberth@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agorev-list-options.txt: use correct directional reference
SZEDER Gábor [Thu, 26 Oct 2017 15:26:37 +0000 (17:26 +0200)]
rev-list-options.txt: use correct directional reference

The descriptions of the options '--parents', '--children' and
'--graph' say "see 'History Simplification' below", although the
referred section is in fact above the description of these options.

Send readers in the right direction by saying "above" instead of
"below".

Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge branch 'mh/ref-locking-fix'
Junio C Hamano [Thu, 26 Oct 2017 03:29:23 +0000 (12:29 +0900)]
Merge branch 'mh/ref-locking-fix'

Transactions to update multiple references that involves a deletion
was quite broken in an error codepath and did not abort everything
correctly.

* mh/ref-locking-fix:
  files_transaction_prepare(): fix handling of ref lock failure
  t1404: add a bunch of tests of D/F conflicts

3 years agopath.c: use xmalloc() in add_to_trie()
Andrey Okoshkin [Tue, 24 Oct 2017 15:15:05 +0000 (18:15 +0300)]
path.c: use xmalloc() in add_to_trie()

Add usage of xmalloc() instead of malloc() in add_to_trie() as xmalloc wraps
and checks memory allocation result.

Signed-off-by: Andrey Okoshkin <a.okoshkin@samsung.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agofiles_transaction_prepare(): fix handling of ref lock failure
Michael Haggerty [Tue, 24 Oct 2017 15:16:25 +0000 (17:16 +0200)]
files_transaction_prepare(): fix handling of ref lock failure

Since dc39e09942 (files_ref_store: use a transaction to update packed
refs, 2017-09-08), failure to lock a reference has been handled
incorrectly by `files_transaction_prepare()`. If
`lock_ref_for_update()` fails in the lock-acquisition loop of that
function, it sets `ret` then breaks out of that loop. Prior to
dc39e09942, that was OK, because the only thing following the loop was
the cleanup code. But dc39e09942 added another blurb of code between
the loop and the cleanup. That blurb sometimes resets `ret` to zero,
making the cleanup code think that the locking was successful.

Specifically, whenever

* One or more reference deletions have been processed successfully in
  the lock-acquisition loop. (Processing the first such reference
  causes a packed-ref transaction to be initialized.)

* Then `lock_ref_for_update()` fails for a subsequent reference. Such
  a failure can happen for a number of reasons, such as the old SHA-1
  not being correct, lock contention, etc. This causes a `break` out
  of the lock-acquisition loop.

* The `packed-refs` lock is acquired successfully and
  `ref_transaction_prepare()` succeeds for the packed-ref transaction.
  This has the effect of resetting `ret` back to 0, and making the
  cleanup code think that lock acquisition was successful.

In that case, any reference updates that were processed prior to
breaking out of the loop would be carried out (loose and packed), but
the reference that couldn't be locked and any subsequent references
would silently be ignored.

This can easily cause data loss if, for example, the user was trying
to push a new name for an existing branch while deleting the old name.
After the push, the branch could be left unreachable, and could even
subsequently be garbage-collected.

This problem was noticed in the context of deleting one reference and
creating another in a single transaction, when the two references D/F
conflict with each other, like

    git update-ref --stdin <<EOF
    delete refs/foo
    create refs/foo/bar HEAD
    EOF

This triggers the above bug because the deletion is processed
successfully for `refs/foo`, then the D/F conflict causes
`lock_ref_for_update()` to fail when `refs/foo/bar` is processed. In
this case the transaction *should* fail, but instead it causes
`refs/foo` to be deleted without creating `refs/foo`. This could
easily result in data loss.

The fix is simple: instead of just breaking out of the loop, jump
directly to the cleanup code. This fixes some tests in t1404 that were
added in the previous commit.

Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agot1404: add a bunch of tests of D/F conflicts
Michael Haggerty [Tue, 24 Oct 2017 15:16:24 +0000 (17:16 +0200)]
t1404: add a bunch of tests of D/F conflicts

It is currently not allowed, in a single transaction, to add one
reference and delete another reference if the two reference names D/F
conflict with each other (e.g., like `refs/foo/bar` and `refs/foo`).
The reason is that the code would need to take locks

    $GIT_DIR/refs/foo.lock
    $GIT_DIR/refs/foo/bar.lock

But the latter lock couldn't coexist with the loose reference file

    $GIT_DIR/refs/foo

, because `$GIT_DIR/refs/foo` cannot be both a directory and a file at
the same time (hence the name "D/F conflict).

Add a bunch of tests that we cleanly reject such transactions.

In fact, many of the new tests currently fail. They will be fixed in
the next commit along with an explanation.

Reported-by: Jeff King <peff@peff.net>
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge tag 'l10n-2.15.0-rnd2' of git://github.com/git-l10n/git-po
Junio C Hamano [Tue, 24 Oct 2017 02:44:52 +0000 (11:44 +0900)]
Merge tag 'l10n-2.15.0-rnd2' of git://github.com/git-l10n/git-po

l10n for Git 2.15.0 round 2

* tag 'l10n-2.15.0-rnd2' of git://github.com/git-l10n/git-po: (22 commits)
  l10n: zh_CN: review for git v2.15.0 l10n round 2
  l10n: zh_CN: for git v2.15.0 l10n round 2
  l10n: de.po: fix typos
  l10n: de.po: translate 70 new messages
  l10n: ru.po: update Russian translation
  l10n: vi.po(3245t): Updated Vietnamese translation for v2.15.0 round 2
  l10n: sv.po: Update Swedish translation (3245t0f0u)
  l10n: fr.po: v2.15.0 round 2
  l10n: fr.po change translation of "First, rewinding"
  l10n: fr.po fix some mistakes
  l10n: Update Catalan translation
  l10n: ko.po: Update Korean translation
  l10n: es.po: v2.15.0 round 2
  l10n: git.pot: v2.15.0 round 2 (2 new, 2 removed)
  l10n: ru.po: update Russian translation
  l10n: bg.po: Updated Bulgarian translation (3245t)
  l10n: sv.po: Update Swedish translation (3245t0f0u)
  l10n: vi.po(3245t): Updated Vietnamese translation for v2.15.0
  l10n: es.po: Update translation v2.15.0 round 1
  l10n: git.pot: v2.15.0 round 1 (68 new, 36 removed)
  ...

3 years agoMerge branch 'jx/zh_CN-proposed' of github.com:jiangxin/git
Jiang Xin [Tue, 24 Oct 2017 02:11:48 +0000 (10:11 +0800)]
Merge branch 'jx/zh_CN-proposed' of github.com:jiangxin/git

* 'jx/zh_CN-proposed' of github.com:jiangxin/git:
  l10n: zh_CN: review for git v2.15.0 l10n round 2
  l10n: zh_CN: for git v2.15.0 l10n round 2

3 years agol10n: zh_CN: review for git v2.15.0 l10n round 2
Ray Chen [Mon, 23 Oct 2017 16:17:59 +0000 (00:17 +0800)]
l10n: zh_CN: review for git v2.15.0 l10n round 2

Signed-off-by: Ray Chen <oldsharp@gmail.com>
3 years agol10n: zh_CN: for git v2.15.0 l10n round 2
Jiang Xin [Sun, 8 Oct 2017 07:29:11 +0000 (15:29 +0800)]
l10n: zh_CN: for git v2.15.0 l10n round 2

Translate 69 messages (3245t0f0u) for git v2.15.0-rc2.

Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
Reviewed-by: 依云 <lilydjwg@gmail.com>
3 years agoMerge branch 'master' of https://github.com/ralfth/git-po-de
Jiang Xin [Tue, 24 Oct 2017 01:56:09 +0000 (09:56 +0800)]
Merge branch 'master' of https://github.com/ralfth/git-po-de

* 'master' of https://github.com/ralfth/git-po-de:
  l10n: de.po: fix typos
  l10n: de.po: translate 70 new messages

3 years agol10n: de.po: fix typos
Andre Hinrichs [Thu, 19 Oct 2017 06:25:25 +0000 (08:25 +0200)]
l10n: de.po: fix typos

Signed-off-by: Andre Hinrichs <andre.hinrichs@gmx.de>
Signed-off-by: Ralf Thielow <ralf.thielow@gmail.com>
3 years agol10n: de.po: translate 70 new messages
Ralf Thielow [Wed, 11 Oct 2017 10:48:48 +0000 (12:48 +0200)]
l10n: de.po: translate 70 new messages

Translate 70 new messages came from git.pot update in 25eab542b
(l10n: git.pot: v2.15.0 round 1 (68 new, 36 removed)) and 9c07fab78
(l10n: git.pot: v2.15.0 round 2 (2 new, 2 removed)).

Signed-off-by: Ralf Thielow <ralf.thielow@gmail.com>
3 years agoSync with 2.14.3
Junio C Hamano [Mon, 23 Oct 2017 05:54:30 +0000 (14:54 +0900)]
Sync with 2.14.3

3 years agoGit 2.14.3 v2.14.3
Junio C Hamano [Mon, 23 Oct 2017 05:44:17 +0000 (14:44 +0900)]
Git 2.14.3

Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge branch 'jk/info-alternates-fix' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:40:00 +0000 (14:40 +0900)]
Merge branch 'jk/info-alternates-fix' into maint

A regression fix for 2.11 that made the code to read the list of
alternate object stores overrun the end of the string.

* jk/info-alternates-fix:
  read_info_alternates: warn on non-trivial errors
  read_info_alternates: read contents into strbuf

3 years agoMerge branch 'jc/fetch-refspec-doc-update' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:39:08 +0000 (14:39 +0900)]
Merge branch 'jc/fetch-refspec-doc-update' into maint

"git fetch <there> <src>:<dst>" allows an object name on the <src>
side when the other side accepts such a request since Git v2.5, but
the documentation was left stale.

* jc/fetch-refspec-doc-update:
  fetch doc: src side of refspec could be full SHA-1

3 years agoMerge branch 'jk/write-in-full-fix' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:37:21 +0000 (14:37 +0900)]
Merge branch 'jk/write-in-full-fix' into maint

Many codepaths did not diagnose write failures correctly when disks
go full, due to their misuse of write_in_full() helper function,
which have been corrected.

* jk/write-in-full-fix:
  read_pack_header: handle signed/unsigned comparison in read result
  config: flip return value of store_write_*()
  notes-merge: use ssize_t for write_in_full() return value
  pkt-line: check write_in_full() errors against "< 0"
  convert less-trivial versions of "write_in_full() != len"
  avoid "write_in_full(fd, buf, len) != len" pattern
  get-tar-commit-id: check write_in_full() return against 0
  config: avoid "write_in_full(fd, buf, len) < len" pattern

3 years agoMerge branch 'rj/no-sign-compare' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:20:18 +0000 (14:20 +0900)]
Merge branch 'rj/no-sign-compare' into maint

Many codepaths have been updated to squelch -Wsign-compare
warnings.

* rj/no-sign-compare:
  ALLOC_GROW: avoid -Wsign-compare warnings
  cache.h: hex2chr() - avoid -Wsign-compare warnings
  commit-slab.h: avoid -Wsign-compare warnings
  git-compat-util.h: xsize_t() - avoid -Wsign-compare warnings

3 years agoMerge branch 'ma/ts-cleanups' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:19:02 +0000 (14:19 +0900)]
Merge branch 'ma/ts-cleanups' into maint

Assorted bugfixes and clean-ups.

* ma/ts-cleanups:
  ThreadSanitizer: add suppressions
  strbuf_setlen: don't write to strbuf_slopbuf
  pack-objects: take lock before accessing `remaining`
  convert: always initialize attr_action in convert_attrs

3 years agoMerge branch 'ls/travis-scriptify' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:17:53 +0000 (14:17 +0900)]
Merge branch 'ls/travis-scriptify' into maint

The scripts to drive TravisCI has been reorganized and then an
optimization to avoid spending cycles on a branch whose tip is
tagged has been implemented.

* ls/travis-scriptify:
  travis-ci: fix "skip_branch_tip_with_tag()" string comparison
  travis: dedent a few scripts that are indented overly deeply
  travis-ci: skip a branch build if equal tag is present
  travis-ci: move Travis CI code into dedicated scripts

3 years agoMerge branch 'er/fast-import-dump-refs-on-checkpoint' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:17:27 +0000 (14:17 +0900)]
Merge branch 'er/fast-import-dump-refs-on-checkpoint' into maint

The checkpoint command "git fast-import" did not flush updates to
refs and marks unless at least one object was created since the
last checkpoint, which has been corrected, as these things can
happen without any new object getting created.

* er/fast-import-dump-refs-on-checkpoint:
  fast-import: checkpoint: dump branches/tags/marks even if object_count==0

3 years agoMerge branch 'jt/fast-export-copy-modify-fix' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:14:51 +0000 (14:14 +0900)]
Merge branch 'jt/fast-export-copy-modify-fix' into maint

"git fast-export" with -M/-C option issued "copy" instruction on a
path that is simultaneously modified, which was incorrect.

* jt/fast-export-copy-modify-fix:
  fast-export: do not copy from modified file

3 years agoMerge branch 'nd/worktree-kill-parse-ref' into maint
Junio C Hamano [Mon, 23 Oct 2017 05:14:16 +0000 (14:14 +0900)]
Merge branch 'nd/worktree-kill-parse-ref' into maint

"git branch -M a b" while on a branch that is completely unrelated
to either branch a or branch b misbehaved when multiple worktree
was in use.  This has been fixed.

* nd/worktree-kill-parse-ref:
  branch: fix branch renaming not updating HEADs correctly

3 years agol10n: ru.po: update Russian translation
Dimitriy Ryazantcev [Sun, 22 Oct 2017 17:35:13 +0000 (20:35 +0300)]
l10n: ru.po: update Russian translation

Signed-off-by: Dimitriy Ryazantcev <dimitriy.ryazantcev@gmail.com>
3 years agoMerge branch 'master' of https://github.com/vnwildman/git
Jiang Xin [Sun, 22 Oct 2017 11:01:07 +0000 (19:01 +0800)]
Merge branch 'master' of https://github.com/vnwildman/git

* 'master' of https://github.com/vnwildman/git:
  l10n: vi.po(3245t): Updated Vietnamese translation for v2.15.0 round 2

3 years agoGit 2.15-rc2 v2.15.0-rc2
Junio C Hamano [Thu, 19 Oct 2017 05:49:17 +0000 (14:49 +0900)]
Git 2.15-rc2

Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge branch 'jc/branch-force-doc-readability-fix'
Junio C Hamano [Thu, 19 Oct 2017 05:45:45 +0000 (14:45 +0900)]
Merge branch 'jc/branch-force-doc-readability-fix'

Doc update.

* jc/branch-force-doc-readability-fix:
  branch doc: sprinkle a few commas for readability

3 years agoMerge branch 'dg/filter-branch-filter-order-doc'
Junio C Hamano [Thu, 19 Oct 2017 05:45:45 +0000 (14:45 +0900)]
Merge branch 'dg/filter-branch-filter-order-doc'

Update the documentation for "git filter-branch" so that the filter
options are listed in the same order as they are applied, as
described in an earlier part of the doc.

* dg/filter-branch-filter-order-doc:
  doc: list filter-branch subdirectory-filter first

3 years agoMerge branch 'jc/fetch-refspec-doc-update'
Junio C Hamano [Thu, 19 Oct 2017 05:45:44 +0000 (14:45 +0900)]
Merge branch 'jc/fetch-refspec-doc-update'

"git fetch <there> <src>:<dst>" allows an object name on the <src>
side when the other side accepts such a request since Git v2.5, but
the documentation was left stale.

* jc/fetch-refspec-doc-update:
  fetch doc: src side of refspec could be full SHA-1

3 years agoMerge branch 'wk/merge-options-gpg-sign-doc'
Junio C Hamano [Thu, 19 Oct 2017 05:45:43 +0000 (14:45 +0900)]
Merge branch 'wk/merge-options-gpg-sign-doc'

Doc updates.

* wk/merge-options-gpg-sign-doc:
  Documentation/merge-options.txt: describe -S/--gpg-sign for 'pull'

3 years agoconfig doc: clarify "git config --path" example
Nathan Payre [Wed, 18 Oct 2017 20:27:16 +0000 (22:27 +0200)]
config doc: clarify "git config --path" example

Change the word "bla" to "section.variable"; "bla" is a placeholder
for a variable name but it wasn't clear for everyone.

While we're here, also reformat this sample command line to use
monospace instead of italics, to better match the rest of the file.

Use a space instead of a dash in "git config", as is common in the
rest of Git's documentation.

Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: MOY Matthieu <matthieu.moy@univ-lyon1.fr>
Signed-off-by: Daniel Bensoussan <daniel.bensoussan--bohm@etu.univ-lyon1.fr>
Signed-off-by: Timothee Albertin <timothee.albertin@etu.univ-lyon1.fr>
Signed-off-by: Nathan Payre <nathan.payre@etu.univ-lyon1.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge branch 'l10n_fr_v2.15.0r2' of git://github.com/jnavila/git
Jiang Xin [Thu, 19 Oct 2017 00:17:23 +0000 (08:17 +0800)]
Merge branch 'l10n_fr_v2.15.0r2' of git://github.com/jnavila/git

* 'l10n_fr_v2.15.0r2' of git://github.com/jnavila/git:
  l10n: fr.po: v2.15.0 round 2
  l10n: fr.po change translation of "First, rewinding"
  l10n: fr.po fix some mistakes

3 years agoMerge branch 'master' of git://github.com/nafmo/git-l10n-sv
Jiang Xin [Thu, 19 Oct 2017 00:16:30 +0000 (08:16 +0800)]
Merge branch 'master' of git://github.com/nafmo/git-l10n-sv

* 'master' of git://github.com/nafmo/git-l10n-sv:
  l10n: sv.po: Update Swedish translation (3245t0f0u)

3 years agoMerge branch 'master' of https://github.com/Softcatala/git-po
Jiang Xin [Thu, 19 Oct 2017 00:14:55 +0000 (08:14 +0800)]
Merge branch 'master' of https://github.com/Softcatala/git-po

* 'master' of https://github.com/Softcatala/git-po:
  l10n: Update Catalan translation

3 years agoMerge branch 'translation' of https://github.com/ChrisADR/git-po
Jiang Xin [Thu, 19 Oct 2017 00:13:29 +0000 (08:13 +0800)]
Merge branch 'translation' of https://github.com/ChrisADR/git-po

* 'translation' of https://github.com/ChrisADR/git-po:
  l10n: es.po: v2.15.0 round 2

3 years agol10n: vi.po(3245t): Updated Vietnamese translation for v2.15.0 round 2
Tran Ngoc Quan [Thu, 19 Oct 2017 00:08:04 +0000 (07:08 +0700)]
l10n: vi.po(3245t): Updated Vietnamese translation for v2.15.0 round 2

Signed-off-by: Tran Ngoc Quan <vnwildman@gmail.com>
3 years agol10n: sv.po: Update Swedish translation (3245t0f0u)
Peter Krefting [Wed, 18 Oct 2017 18:35:32 +0000 (19:35 +0100)]
l10n: sv.po: Update Swedish translation (3245t0f0u)

Signed-off-by: Peter Krefting <peter@softwolves.pp.se>
3 years agol10n: fr.po: v2.15.0 round 2
Jean-Noel Avila [Sun, 8 Oct 2017 12:37:52 +0000 (14:37 +0200)]
l10n: fr.po: v2.15.0 round 2

Signed-off-by: Jean-Noel Avila <jn.avila@free.fr>
3 years agol10n: fr.po change translation of "First, rewinding"
Nicolas Cornu [Tue, 19 Sep 2017 07:37:34 +0000 (09:37 +0200)]
l10n: fr.po change translation of "First, rewinding"

Signed-off-by: Nicolas Cornu <nicolac76@yahoo.fr>
3 years agol10n: fr.po fix some mistakes
Jean-Noel Avila [Thu, 14 Sep 2017 12:05:41 +0000 (14:05 +0200)]
l10n: fr.po fix some mistakes

Reported-by: Christophe Jaillet <christophe.jaillet@wanadoo.fr>
Signed-off-by: Jean-Noel Avila <jean-noel.avila@scantech.fr>
3 years agoSync with maint
Junio C Hamano [Wed, 18 Oct 2017 05:26:53 +0000 (14:26 +0900)]
Sync with maint

* maint:
  Prepare for 2.14.3

3 years agoPrepare for 2.14.3
Junio C Hamano [Wed, 18 Oct 2017 05:24:09 +0000 (14:24 +0900)]
Prepare for 2.14.3

Signed-off-by: Junio C Hamano <gitster@pobox.com>
3 years agoMerge branch 'jk/ref-filter-colors-fix' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:20:43 +0000 (14:20 +0900)]
Merge branch 'jk/ref-filter-colors-fix' into maint

This is the "theoretically more correct" approach of simply
stepping back to the state before plumbing commands started paying
attention to "color.ui" configuration variable.

* jk/ref-filter-colors-fix:
  tag: respect color.ui config
  Revert "color: check color.ui in git_default_config()"
  Revert "t6006: drop "always" color config tests"
  Revert "color: make "always" the same as "auto" in config"
  color: make "always" the same as "auto" in config
  provide --color option for all ref-filter users
  t3205: use --color instead of color.branch=always
  t3203: drop "always" color test
  t6006: drop "always" color config tests
  t7502: use diff.noprefix for --verbose test
  t7508: use test_terminal for color output
  t3701: use test-terminal to collect color output
  t4015: prefer --color to -c color.diff=always
  test-terminal: set TERM=vt100

3 years agoMerge branch 'jc/doc-checkout' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'jc/doc-checkout' into maint

Doc update.

* jc/doc-checkout:
  checkout doc: clarify command line args for "checkout paths" mode

3 years agoMerge branch 'tb/complete-describe' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'tb/complete-describe' into maint

Docfix.

* tb/complete-describe:
  completion: add --broken and --dirty to describe

3 years agoMerge branch 'rs/rs-mailmap' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'rs/rs-mailmap' into maint

* rs/rs-mailmap:
  .mailmap: normalize name for René Scharfe

3 years agoMerge branch 'rs/fsck-null-return-from-lookup' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'rs/fsck-null-return-from-lookup' into maint

Improve behaviour of "git fsck" upon finding a missing object.

* rs/fsck-null-return-from-lookup:
  fsck: handle NULL return of lookup_blob() and lookup_tree()

3 years agoMerge branch 'jk/sha1-loose-object-info-fix' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'jk/sha1-loose-object-info-fix' into maint

Leakfix and futureproofing.

* jk/sha1-loose-object-info-fix:
  sha1_loose_object_info: handle errors from unpack_sha1_rest

3 years agoMerge branch 'sb/branch-avoid-repeated-strbuf-release' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'sb/branch-avoid-repeated-strbuf-release' into maint

* sb/branch-avoid-repeated-strbuf-release:
  branch: reset instead of release a strbuf

3 years agoMerge branch 'rs/qsort-s' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:14 +0000 (14:19 +0900)]
Merge branch 'rs/qsort-s' into maint

* rs/qsort-s:
  test-stringlist: avoid buffer underrun when sorting nothing

3 years agoMerge branch 'jn/strbuf-doc-re-reuse' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:13 +0000 (14:19 +0900)]
Merge branch 'jn/strbuf-doc-re-reuse' into maint

* jn/strbuf-doc-re-reuse:
  strbuf doc: reuse after strbuf_release is fine

3 years agoMerge branch 'rs/run-command-use-alloc-array' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:13 +0000 (14:19 +0900)]
Merge branch 'rs/run-command-use-alloc-array' into maint

Code clean-up.

* rs/run-command-use-alloc-array:
  run-command: use ALLOC_ARRAY

3 years agoMerge branch 'rs/tag-null-pointer-arith-fix' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:12 +0000 (14:19 +0900)]
Merge branch 'rs/tag-null-pointer-arith-fix' into maint

Code clean-up.

* rs/tag-null-pointer-arith-fix:
  tag: avoid NULL pointer arithmetic

3 years agoMerge branch 'rs/cocci-de-paren-call-params' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:12 +0000 (14:19 +0900)]
Merge branch 'rs/cocci-de-paren-call-params' into maint

Code clean-up.

* rs/cocci-de-paren-call-params:
  coccinelle: remove parentheses that become unnecessary

3 years agoMerge branch 'ad/doc-markup-fix' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:12 +0000 (14:19 +0900)]
Merge branch 'ad/doc-markup-fix' into maint

Docfix.

* ad/doc-markup-fix:
  doc: correct command formatting

3 years agoMerge branch 'mr/doc-negative-pathspec' into maint
Junio C Hamano [Wed, 18 Oct 2017 05:19:12 +0000 (14:19 +0900)]
Merge branch 'mr/doc-negative-pathspec' into maint

Doc updates.

* mr/doc-negative-pathspec:
  docs: improve discoverability of exclude pathspec