blob: 761f697f109e29f1c3d7b54b43e030e803a2b57a [file] [log] [blame]
Git 2.7 Release Notes
Updates since v2.6
UI, Workflows & Features
* "git remote" learned "get-url" subcommand to show the URL for a
given remote name used for fetching and pushing.
* There was no way to defeat a configured rebase.autostash variable
from the command line, as "git rebase --no-autostash" was missing.
* "git log --date=local" used to only show the normal (default)
format in the local timezone. The command learned to take 'local'
as an instruction to use the local timezone with other formats,
* The refs used during a "git bisect" session is now per-worktree so
that independent bisect sessions can be done in different worktrees
created with "git worktree add".
* Users who are too busy to type three extra keystrokes to ask for
"git stash show -p" can now set stash.showPatch configuration
varible to true to always see the actual patch, not just the list
of paths affected with feel for the extent of damage via diffstat.
* "quiltimport" allows to specify the series file by honoring the
$QUILT_SERIES environment and also --series command line option.
* The use of 'good/bad' in "git bisect" made it confusing to use when
hunting for a state change that is not a regression (e.g. bugfix).
The command learned 'old/new' and then allows the end user to
say e.g. "bisect start --term-old=fast --term-new=slow" to find a
performance regression.
* "git interpret-trailers" can now run outside of a Git repository.
* "git p4" learned to reencode the pathname it uses to communicate
with the p4 depot with a new option.
* Give progress meter to "git filter-branch".
* Allow a later "!/abc/def" to override an earlier "/abc" that
appears in the same .gitignore file to make it easier to express
"everything in /abc directory is ignored, except for ...".
* Teach "git p4" to send large blobs outside the repository by
talking to Git LFS.
* Prepare for Git on-disk repository representation to undergo
backward incompatible changes by introducing a new repository
format version "1", with an extension mechanism.
(merge 067fbd4 jk/repository-extension later to maint).
* "git worktree" learned a "list" subcommand.
* "git clone --dissociate" learned that it can be used even when
"--reference" was not used at the same time.
* "git blame" learnt to take "--first-parent" and "--reverse" at the
same time when it makes sense.
* "git checkout" did not follow the usual "--[no-]progress"
convention and implemented only "--quiet" that is essentially
a superset of "--no-progress". Extend the command to support the
usual "--[no-]progress".
* The semantics of tranfer.hideRefs configuration variable have been
extended to work better with the ref "namespace" feature that lets
you throw unrelated bunches of repositories in a single physical
repository and virtually serve them as separate ones.
* send-email config variables whose values are pathnames now go
through the ~username/ expansion.
* bash completion learnt to TAB-complete recipient addresses given
to send-email.
* The credential-cache daemon can be told to ignore SIGHUP to work
around issue when running Git from inside emacs.
Performance, Internal Implementation, Development Support etc.
* The infrastructure to rewrite "git submodule" in C is being built
incrementally. Let's polish these early parts well enough and make
them graduate to 'next' and 'master', so that the more involved
follow-up can start cooking on a solid ground.
* Some features from "git tag -l" and "git branch -l" have been made
available to "git for-each-ref" so that eventually the unified
implementation can be shared across all three. The version merged
to the 'master' branch earlier had a performance regression in "tag
--contains", which has since been corrected.
* Because "test_when_finished" in our test framework queues the
clean-up tasks to be done in a shell variable, it should not be
used inside a subshell. Add a mechanism to allow 'bash' to catch
such uses, and fix the ones that were found.
(merge 0968f12 jk/test-lint-forbid-when-finished-in-subshell later to maint).
* The debugging infrastructure for pkt-line based communication has
been improved to mark the side-band communication specifically.
(merge fd89433 jk/async-pkt-line later to maint).
* Update "git branch" that list existing branches, using the
ref-filter API that is shared with "git tag" and "git
* The test for various line-ending conversions has been enhanced.
* A few test scripts around "git p4" have been improved for
* Many allocations that is manually counted (correctly) that are
followed by strcpy/sprintf have been replaced with a less error
prone constructs such as xstrfmt.
* The internal stripspace() function has been moved to where it
logically belongs to, i.e. strbuf API, and the command line parser
of "git stripspace" has been updated to use the parse_options API.
(merge bed4452 tk/stripspace later to maint).
* "git am" used to spawn "git mailinfo" via run_command() API once
per each patch, but learned to make a direct call to mailinfo()
* The implementation of "git mailinfo" was refactored so that a
mailinfo() function can be directly called from inside a process.
* With a "debug" helper, debugging of a single "git" invocation in
our test scripts has become a lot easier.
* The "configure" script did not test for -lpthread correctly, which
upset some linkers.
* Cross completed task off of subtree project's todo list.
* Test cleanups for the subtree project.
* Clean up style in an ancient test t9300.
* Work around some test flakiness with p4d.
* Fsck did not correctly detect a NUL-truncated header in a tag.
* Use a safer behavior when we hit errors verifying remote certificates.
* Speed up filter-branch for cases where we only care about rewriting
commits, not tree data.
* The parse-options API has been updated to make "-h" command line
option work more consistently in all commands.
* "git svn rebase/mkdirs" got optimized by keeping track of empty
directories better.
* Fix some racy client/server tests by treating SIGPIPE the same as a
normal non-zero exit.
* The necessary infrastructure to build topics using the free Travis
CI has been added. Developers forking from this topic (and enabling
Travis) can do their own builds, and we can turn on auto-builds for
git/git (including build-status for pull requests that people
Also contains various documentation updates and code clean-ups.
Fixes since v2.6
Unless otherwise noted, all the fixes since v2.6 in the maintenance
track are contained in this release (see the maintenance releases'
notes for details).
* Very small number of options take a parameter that is optional
(which is not a great UI element as they can only appear at the end
of the command line). Add notice to documentation of each and
every one of them.
(merge 2b594bf mm/keyid-docs later to maint).
* "git blame --first-parent v1.0..v2.0" was not rejected but did not
limit the blame to commits on the first parent chain.
(merge 95a4fb0 jk/blame-first-parent later to maint).
* "git subtree" (in contrib/) now can take whitespaces in the
pathnames, not only in the in-tree pathname but the name of the
directory that the repository is in. (merge 5b6ab38
as/subtree-with-spaces later to maint).
* The ssh transport, just like any other transport over the network,
did not clear GIT_* environment variables, but it is possible to
use SendEnv and AcceptEnv to leak them to the remote invocation of
Git, which is not a good idea at all. Explicitly clear them just
like we do for the local transport.
(merge a48b409 jk/connect-clear-env later to maint).
* Correct "git p4 --detect-labels" so that it does not fail to create
a tag that points at a commit that is also being imported.
(merge b43702a ld/p4-import-labels later to maint).
* The Makefile always runs the library archiver with hardcoded "crs"
options, which was inconvenient for exotic platforms on which
people want to use programs with totally different set of command
line options.
(merge ac179b4 jw/make-arflags-customizable later to maint).
* Customization to change the behaviour with "make -w" and "make -s"
in our Makefile was broken when they were used together.
(merge ef49e05 jk/make-findstring-makeflags-fix later to maint).
* Allocation related functions and stdio are unsafe things to call
inside a signal handler, and indeed killing the pager can cause
glibc to deadlock waiting on allocation mutex as our signal handler
tries to free() some data structures in wait_for_pager(). Reduce
these unsafe calls.
(merge 507d780 ti/glibc-stdio-mutex-from-signal-handler later to maint).
* The way how --ref/--notes to specify the notes tree reference are
DWIMmed was not clearly documented.
(merge e14c92e jk/notes-dwim-doc later to maint).
* "git gc" used to barf when a symbolic ref has gone dangling
(e.g. the branch that used to be your upstream's default when you
cloned from it is now gone, and you did "fetch --prune").
(merge 14886b4 js/gc-with-stale-symref later to maint).
* "git clone --dissociate" runs a big "git repack" process at the
end, and it helps to close file descriptors that are open on the
packs and their idx files before doing so on filesystems that
cannot remove a file that is still open.
(merge 786b150 js/clone-dissociate later to maint).
* Description of the "log.follow" configuration variable in "git log"
documentation is now also copied to "git config" documentation.
(merge fd8d07e dt/log-follow-config later to maint).
* "git rebase -i" had a minor regression recently, which stopped
considering a line that begins with an indented '#' in its insn
sheet not a comment. Further, the code was still too picky on
Windows where CRLF left by the editor is turned into a trailing CR
on the line read via the "read" built-in command of bash. Both of
these issues are now fixed.
(merge 39743cf gr/rebase-i-drop-warn later to maint).
* After "git checkout --detach", "git status" reported a fairly
useless "HEAD detached at HEAD", instead of saying at which exact
(merge 0eb8548 mm/detach-at-HEAD-reflog later to maint).
* When "git send-email" wanted to talk over Net::SMTP::SSL,
Net::Cmd::datasend() did not like to be fed too many bytes at the
same time and failed to send messages. Send the payload one line
at a time to work around the problem.
(merge f60c483 sa/send-email-smtp-batch-data-limit later to maint).
* When "git am" was rewritten as a built-in, it stopped paying
attention to user.signingkey, which was fixed.
(merge 434c64d pt/am-builtin later to maint).
* It was not possible to use a repository-lookalike created by "git
worktree add" as a local source of "git clone".
(merge d78db84 nd/clone-linked-checkout later to maint).
* On a case insensitive filesystems, setting GIT_WORK_TREE variable
using a random cases that does not agree with what the filesystem
thinks confused Git that it wasn't inside the working tree.
(merge 63ec5e1 js/icase-wt-detection later to maint).
* Performance-measurement tests did not work without an installed Git.
(merge 31cd128 sb/perf-without-installed-git later to maint).
* A test script for the HTTP service had a timing dependent bug,
which was fixed.
(merge 362d8b6 sb/http-flaky-test-fix later to maint).
* There were some classes of errors that "git fsck" diagnosed to its
standard error that did not cause it to exit with non-zero status.
(merge 122f76f jc/fsck-dropped-errors later to maint).
* Work around "git p4" failing when the P4 depot records the contents
in UTF-16 without UTF-16 BOM.
(merge 1f5f390 ls/p4-translation-failure later to maint).
* When "git gc --auto" is backgrounded, its diagnosis message is
lost. Save it to a file in $GIT_DIR and show it next time the "gc
--auto" is run.
(merge 329e6e8 nd/gc-auto-background-fix later to maint).
* The submodule code has been taught to work better with separate
work trees created via "git worktree add".
(merge 11f9dd7 mk/submodule-gitdir-path later to maint).
* "git gc" is safe to run anytime only because it has the built-in
grace period to protect young objects. In order to run with no
grace period, the user must make sure that the repository is
(merge fae1a90 jc/doc-gc-prune-now later to maint).
* A recent "filter-branch --msg-filter" broke skipping of the commit
object header, which is fixed.
(merge a5a4b3f jk/filter-branch-use-of-sed-on-incomplete-line later to maint).
* The normalize_ceiling_entry() function does not muck with the end
of the path it accepts, and the real world callers do rely on that,
but a test insisted that the function drops a trailing slash.
(merge b2a7123 rd/test-path-utils later to maint).
* A test for interaction between untracked cache and sparse checkout
added in Git 2.5 days were flaky.
(merge 9b680fb dt/t7063-fix-flaky-test later to maint).
* A couple of commands still showed "[options]" in their usage string
to note where options should come on their command line, but we
spell that "[<options>]" in most places these days.
(merge d96a031 rt/placeholder-in-usage later to maint).
* The synopsis text and the usage string of subcommands that read
list of things from the standard input are often shown as if they
only take input from a file on a filesystem, which was misleading.
(merge 33e8fc8 jc/usage-stdin later to maint).
* "git am -3" had a small regression where it is aborted in its error
handling codepath when underlying merge-recursive failed in certain
ways, as it assumed that the internal call to merge-recursive will
never die, which is not the case (yet).
(merge c63d4b2 jc/am-3-fallback-regression-fix later to maint).
* The linkage order of libraries was wrong in places around libcurl.
(merge 7e91e8d rp/link-curl-before-ssl later to maint).
* The name-hash subsystem that is used to cope with case insensitive
filesystems keeps track of directories and their on-filesystem
cases for all the paths in the index by holding a pointer to a
randomly chosen cache entry that is inside the directory (for its
ce->ce_name component). This pointer was not updated even when the
cache entry was removed from the index, leading to use after free.
This was fixed by recording the path for each directory instead of
borrowing cache entries and restructuring the API somewhat.
(merge 41284eb dt/name-hash-dir-entry-fix later to maint).
* "git merge-file" tried to signal how many conflicts it found, which
obviously would not work well when there are too many of them.
(merge e34f802 jk/merge-file-exit-code later to maint).
* The error message from "git blame --contents --reverse" incorrectly
talked about "--contents --children".
(merge 9526197 mk/blame-error-message later to maint).
* "git imap-send" did not compile well with older version of cURL library.
(merge 71d9257 js/imap-send-curl-compilation-fix later to maint).
* Merging a branch that removes a path and another that changes the
mode bits on the same path should have conflicted at the path, but
it didn't and silently favoured the removal.
(merge 72fac66 jk/delete-modechange-conflict later to maint).
* "git --literal-pathspecs add -u/-A" without any command line
argument misbehaved ever since Git 2.0.
(merge 29abb33 jc/add-u-A-default-to-top later to maint).
* "git daemon" uses "run_command()" without "finish_command()", so it
needs to release resources itself, which it forgot to do.
(merge b1b49ff rs/daemon-plug-child-leak later to maint).
* "git status --branch --short" accessed beyond the constant string
"HEAD", which has been corrected.
(merge c72b49d rs/wt-status-detached-branch-fix later to maint).
* We peek objects from submodule's object store by linking it to the
list of alternate object databases, but the code to do so forgot to
correctly initialize the list.
(merge 9a6e4f0 jk/initialization-fix-to-add-submodule-odb later to maint).
* The code to prepare the working tree side of temporary directory
for the "dir-diff" feature forgot that symbolic links need not be
copied (or symlinked) to the temporary area, as the code already
special cases and overwrites them. Besides, it was wrong to try
computing the object name of the target of symbolic link, which may
not even exist or may be a directory.
(merge cfe2d4b da/difftool later to maint).
* A Range: request can be responded with a full response and when
asked properly libcurl knows how to strip the result down to the
requested range. However, we were hand-crafting a range request
and it did not kick in.
* Having a leftover .idx file without corresponding .pack file in
the repository hurts performance; "git gc" learned to prune them.
(merge 478f34d dk/gc-idx-wo-pack later to maint).
* Apple's common crypto implementation of SHA1_Update() does not take
more than 4GB at a time, and we now have a compile-time workaround
for it.
(merge 001fd7a ad/sha1-update-chunked later to maint).
* Produce correct "dirty" marker for shell prompts, even when we
are on an orphan or an unborn branch.
(merge c26f70c sg/bash-prompt-dirty-orphan later to maint).
* A build without NO_IPv6 used to use gethostbyname() when guessing
user's hostname, instead of getaddrinfo() that is used in other
codepaths in such a build.
(merge 00bce77 ep/ident-with-getaddrinfo later to maint).
* The exit code of git-fsck didnot reflect some types of errors found
in packed objects, which has been corrected.
(merge 8c24d83 dt/fsck-verify-pack-error later to maint).
* Code clean-up, minor fixes etc.
(merge 15ed07d jc/rerere later to maint).
(merge e7a7401 pt/pull-builtin later to maint).
(merge 29bc480 nd/ls-remote-does-not-have-u-option later to maint).
(merge be510e0 jk/asciidoctor-section-heading-markup-fix later to maint).
(merge 83e6bda tk/typofix-connect-unknown-proto-error later to maint).
(merge a43eb67 tk/doc-interpret-trailers-grammo later to maint).
(merge ba128e2 es/worktree-add-cleanup later to maint).
(merge 44cd91e cc/quote-comments later to maint).
(merge 147875f sb/submodule-config-parse later to maint).
(merge ae9f274 es/worktree-add later to maint).
(merge 3b19dba jc/em-dash-in-doc later to maint).
(merge f3f38c7 jc/everyday-markup later to maint).
(merge 77d5f71 xf/user-manual-markup later to maint).
(merge b2af482 xf/user-manual-ff later to maint).
(merge e510ab8 rs/pop-commit later to maint).
(merge fdcdb77 js/misc-fixes later to maint).
(merge c949b00 rs/show-branch-argv-array later to maint).
(merge 56a8aea nd/doc-check-ref-format-typo later to maint).