Documentation/RelNotes/2.37.0.txt

Git v2.37 Release Notes
=======================

UI, Workflows & Features

 * "vimdiff[123]" mergetool drivers have been reimplemented with a
   more generic layout mechanism.

 * "git -v" and "git -h" are now understood as "git --version" and
   "git --help".

 * The temporary files fed to external diff command are now generated
   inside a new temporary directory under the same basename.

 * "git log --since=X" will stop traversal upon seeing a commit that
   is older than X, but there may be commits behind it that is younger
   than X when the commit was created with a faulty clock.  A new
   option is added to keep digging without stopping, and instead
   filter out commits with timestamp older than X.

 * "git -c branch.autosetupmerge=simple branch $A $B" will set the $B
   as $A's upstream only when $A and $B shares the same name, and "git
   -c push.default=simple" on branch $A would push to update the
   branch $A at the remote $B came from.  Also more places use the
   sole remote, if exists, before defaulting to 'origin'.

 * A new doc has been added that lists tips for tools to work with
   Git's codebase.

 * "git remote -v" now shows the list-objects-filter used during
   fetching from the remote, if available.

 * With the new http.curloptResolve configuration, the CURLOPT_RESOLVE
   mechanism that allows cURL based applications to use pre-resolved
   IP addresses for the requests is exposed to the scripts.

 * "git add -i" was rewritten in C some time ago and has been in
   testing; the reimplementation is now exposed to general public by
   default.


Performance, Internal Implementation, Development Support etc.

 * The performance of the "untracked cache" feature has been improved
   when "--untracked-files=<mode>" and "status.showUntrackedFiles"
   are combined.

 * "git stash" works better with sparse index entries.

 * "git show :<path>" learned to work better with the sparse-index
   feature.

 * Introduce and apply coccinelle rule to discourage an explicit
   comparison between a pointer and NULL, and applies the clean-up to
   the maintenance track.


Fixes since v2.36
-----------------

 * "git submodule update" without pathspec should silently skip an
   uninitialized submodule, but it started to become noisy by mistake.
   (merge 4f1ccef87c gc/submodule-update-part2 later to maint).

 * "diff-tree --stdin" has been broken for about a year, but 2.36
   release broke it even worse by breaking running the command with
   <pathspec>, which in turn broke "gitk" and got noticed.  This has
   been corrected by aligning its behaviour to that of "log".
   (merge f8781bfda3 jc/diff-tree-stdin-fix later to maint).

 * Regression fix for 2.36 where "git name-rev" started to sometimes
   reference strings after they are freed.
   (merge 45a14f578e rs/name-rev-fix-free-after-use later to maint).

 * "git show <commit1> <commit2>... -- <pathspec>" lost the pathspec
   when showing the second and subsequent commits, which has been
   corrected.
   (merge 5cdb38458e jc/show-pathspec-fix later to maint).

 * "git fast-export -- <pathspec>" lost the pathspec when showing the
   second and subsequent commits, which has been corrected.
   (merge d1c25272f5 rs/fast-export-pathspec-fix later to maint).

 * "git format-patch <args> -- <pathspec>" lost the pathspec when
   showing the second and subsequent commits, which has been
   corrected.
   (merge 91f8f7e46f rs/format-patch-pathspec-fix later to maint).

 * "git clone --origin X" leaked piece of memory that held value read
   from the clone.defaultRemoteName configuration variable, which has
   been plugged.
   (merge 6dfadc8981 jc/clone-remote-name-leak-fix later to maint).

 * Get rid of a bogus and over-eager coccinelle rule.
   (merge 08bdd3a185 jc/cocci-xstrdup-or-null-fix later to maint).

 * The path taken by "git multi-pack-index" command from the end user
   was compared with path internally prepared by the tool withut first
   normalizing, which lead to duplicated paths not being noticed,
   which has been corrected.
   (merge 11f9e8de3d ds/midx-normalize-pathname-before-comparison later to maint).

 * Correct choices of C compilers used in various CI jobs.
   (merge 3506cae04f ab/cc-package-fixes later to maint).

 * Various cleanups to "git p4".
   (merge 4ff0108d9e jh/p4-various-fixups later to maint).

 * The progress meter of "git blame" was showing incorrect numbers
   when processing only parts of the file.
   (merge e5f5d7d42e ea/progress-partial-blame later to maint).

 * "git rebase --keep-base <upstream> <branch-to-rebase>" computed the
   commit to rebase onto incorrectly, which has been corrected.
   (merge 9e5ebe9668 ah/rebase-keep-base-fix later to maint).

 * Fix a leak of FILE * in an error codepath.
   (merge c0befa0c03 kt/commit-graph-plug-fp-leak-on-error later to maint).

 * Avoid problems from interaction between malloc_check and address
   sanitizer.
   (merge 067109a5e7 pw/test-malloc-with-sanitize-address later to maint).

 * The commit summary shown after making a commit is matched to what
   is given in "git status" not to use the break-rewrite heuristics.
   (merge 84792322ed rs/commit-summary-wo-break-rewrite later to maint).

 * Update a few end-user facing messages around eol conversion.
   (merge c970d30c2c ah/convert-warning-message later to maint).

 * Trace2 documentation updates.
   (merge a6c80c313c js/trace2-doc-fixes later to maint).

 * Build procedure fixup.
   (merge 1fbfd96f50 mg/detect-compiler-in-c-locale later to maint).

 * "git pull" without "--recurse-submodules=<arg>" made
   submodule.recurse take precedence over fetch.recurseSubmodules by
   mistake, which has been corrected.
   (merge 5819417365 gc/pull-recurse-submodules later to maint).

 * "git bisect" was too silent before it is ready to start computing
   the actual bisection, which has been corrected.
   (merge f11046e6de cd/bisect-messages-from-pre-flight-states later to maint).

 * macOS CI jobs have been occasionally flaky due to tentative version
   skew between perforce and the homebrew packager.  Instead of
   failing the whole CI job, just let it skip the p4 tests when this
   happens.
   (merge f15e00b463 cb/ci-make-p4-optional later to maint).

 * A bit of test framework fixes with a few fixes to issues found by
   valgrind.
   (merge 7c898554d7 ab/valgrind-fixes later to maint).

 * "git archive --add-file=<path>" picked up the raw permission bits
   from the path and propagated to zip output in some cases, without
   normalization, which has been corrected (tar output did not have
   this issue).
   (merge 6a61661967 jc/archive-add-file-normalize-mode later to maint).

 * "make coverage-report" without first running "make coverage" did
   not produce any meaningful result, which has been corrected.
   (merge 96ddfecc5b ep/coverage-report-wants-test-to-have-run later to maint).

 * The "--current" option of "git show-branch" should have been made
   incompatible with the "--reflog" mode, but this was not enforced,
   which has been corrected.
   (merge 41c64ae0e7 jc/show-branch-g-current later to maint).

 * "git fetch" unnecessarily failed when an unexpected optional
   section appeared in the output, which has been corrected.
   (merge 7709acf7be jt/fetch-peek-optional-section later to maint).

 * The way "git fetch" without "--update-head-ok" ensures that HEAD in
   no worktree points at any ref being updated was too wasteful, which
   has been optimized a bit.
   (merge f7400da800 os/fetch-check-not-current-branch later to maint).

 * "git fetch --recurse-submodules" from multiple remotes (either from
   a remote group, or "--all") used to make one extra "git fetch" in
   the submodules, which has been corrected.
   (merge 0353c68818 jc/avoid-redundant-submodule-fetch later to maint).

 * With a recent update to refuse access to repositories of other
   people by default, "sudo make install" and "sudo git describe"
   stopped working.  This series intends to loosen it while keeping
   the safety.
   (merge b9063afda1 cb/path-owner-check-with-sudo later to maint).

 * The tests that ensured merges stop when interfering local changes
   are present did not make sure that local changes are preserved; now
   they do.
   (merge 4b317450ce jc/t6424-failing-merge-preserve-local-changes later to maint).

 * Some real problems noticed by gcc 12 have been fixed, while false
   positives have been worked around.

 * Other code cleanup, docfix, build fix, etc.
   (merge e6b2582da3 cm/reftable-0-length-memset later to maint).
   (merge 0b75e5bf22 ab/misc-cleanup later to maint).
   (merge 52e1ab8a76 ea/rebase-code-simplify later to maint).
   (merge 756d15923b sg/safe-directory-tests-and-docs later to maint).
   (merge d097a23bfa ds/do-not-call-bug-on-bad-refs later to maint).
   (merge c36c27e75c rs/t7812-pcre2-ws-bug-test later to maint).
   (merge 1da312742d gf/unused-includes later to maint).
   (merge 465b30a92d pb/submodule-recurse-mode-enum later to maint).
   (merge 82b28c4ed8 km/t3501-use-test-helpers later to maint).
   (merge 72315e431b sa/t1011-use-helpers later to maint).
   (merge 95b3002201 cg/vscode-with-gdb later to maint).
   (merge fbe5f6b804 tk/p4-utf8-bom later to maint).
   (merge 17f273ffba tk/p4-with-explicity-sync later to maint).
   (merge 944db25c60 kf/p4-multiple-remotes later to maint).
   (merge b014cee8de jc/update-ozlabs-url later to maint).
   (merge 4ec5008062 pb/ggg-in-mfc-doc later to maint).
   (merge af845a604d tb/receive-pack-code-cleanup later to maint).
   (merge 2acf4cf001 js/ci-gcc-12-fixes later to maint).
   (merge 05e280c0a6 jc/http-clear-finished-pointer later to maint).