* ps/cat-file-remote-object-info:
cat-file: make remote-object-info allow-list adapt to the server
cat-file: add remote-object-info to batch-command
transport: add client support for object-info
serve: advertise object-info feature
protocol-caps: check object existence regardless of the attributes requested
fetch-pack: move fetch initialization
connect: make write_fetch_command_and_capabilities() more generic
fetch-pack: move write_fetch_command_and_capabilities() to connect.c
fetch-pack: use unsigned int for hash_algo variable
fetch-pack: drop the static advertise_sid variable
t1006: extract helper functions into new 'lib-cat-file.sh'
cat-file: declare loop counter inside for()
transport-helper: fix memory leak of helper on disconnect
Since the info command in cat-file --batch-command prints object
info for a given object, it is natural to add another command in
cat-file --batch-command to print object info for a given object
from a remote.
Add remote-object-info command to cat-file --batch-command.
While info takes object ids one at a time, this creates overhead when
making requests to a server. So remote-object-info instead can take
multiple object ids at once.
The cat-file --batch-command command is generally implemented in the
following manner:
- Receive and parse input from user
- Call respective function attached to command
- Get object info, print object info
In --buffer mode, this changes to:
- Receive and parse input from user
- Store respective function attached to command in a queue
- After flush, loop through commands in queue
- Call respective function attached to command
- Get object info, print object info
Notice how the getting and printing of object info is accomplished one
at a time. As described above, this creates a problem for making
requests to a server. Therefore, remote-object-info is implemented in
the following manner:
- Receive and parse input from user
If command is remote-object-info:
- Get object info from remote
- Loop through and print each object info
Else:
- Call respective function attached to command
- Parse input, get object info, print object info
And finally for --buffer mode remote-object-info:
- Receive and parse input from user
- Store respective function attached to command in a queue
- After flush, loop through commands in queue:
If command is remote-object-info:
- Get object info from remote
- Loop through and print each object info
Else:
- Call respective function attached to command
- Get object info, print object info
To summarize, remote-object-info gets object info from the remote and
then loops through the object info passed in, printing the info.
In order for remote-object-info to avoid remote communication
overhead in the non-buffer mode, the objects are passed in as such:
remote-object-info <remote> <oid> <oid> ... <oid>
rather than
remote-object-info <remote> <oid>
remote-object-info <remote> <oid>
...
remote-object-info <remote> <oid>
Placeholders in the format are validated against an allow-list of the
atoms the remote path supports: "objectname" and "objectsize".
Unsupported atoms expand to an empty string, honoring how for-each-ref
handles known but inapplicable atoms.
Without this, atoms like %(objecttype) would mark data->info.typep and
because the server only sends size, type_name() would later crash.
As extra safety, even outside of the remote path, initialize
expand_data's type to OBJ_BAD and handle type_name() returning NULL.
Helped-by: Jonathan Tan <jonathantanmy@google.com>
Helped-by: Christian Couder <chriscool@tuxfamily.org>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Mentored-by: Chandra Pratap <chandrapratap3519@gmail.com>
Signed-off-by: Calvin Wan <calvinwan@google.com>
Signed-off-by: Eric Ju <eric.peijian@gmail.com>
[pablo: added the atom allow-list validation]
Signed-off-by: Pablo Sabater <pabloosabaterr@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently, send_info() only checks for existence when the attribute
'size' is also requested. Requesting a bare OID, without attributes only
echoes back the OID.
Extract the existence check to be done regardless of the number of
attributes requested.
While at it, introduce a wrapper called get_object_info() similar to
odb_read_object_info() that returns OBJ_BAD on fail and adds
OBJECT_INFO_SKIP_FETCH_OBJECT and OBJECT_INFO_QUICK flags.
OBJECT_INFO_SKIP_FETCH_OBJECT is so a server with a partial clone
doesn't trigger fetching objects when it gets an object-info request
with an OID that is not available locally. A server should only report
what it has locally.
Tighten the condition used to determine whether an object is
recognized. get_object_info() returns OBJ_BAD for unknown objects,
but OBJ_NONE (0) can also mean "not found". Change the check from '< 0'
to '<= OBJ_NONE' to cover both as unrecognized.
With this patch, a bare OID has two possible responses:
1. Recognized OID: the server answers with "<OID>"
2. Unrecognized OID: the server answers with "<OID> SP"
Update the object-info section in 'gitprotocol-v2.adoc':
- Require full obj-oid explicitly.
- Fix parentheses.
- Define obj-size explicitly.
- Make obj-size optional in obj-info and document the behavior
for unrecognized object IDs.
- Describe the attr header as zero or more pkt-lines, one per attribute,
matching what the server implements. A request with no attributes gets
no header.
Signed-off-by: Pablo Sabater <pabloosabaterr@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The usage string and SYNOPSIS for 'git fast-export' have been
standardized to make them consistent with each other and with other
commands.
* cc/doc-fast-export-synopsis-fix:
fast-export: standardize usage string and SYNOPSIS
The '[includeIf "condition"]' conditional inclusion facility for
configuration files has been taught to use the location of the
worktree in its condition.
* cl/conditional-config-on-worktree-path:
config: add "worktree" and "worktree/i" includeIf conditions
config: refactor include_by_gitdir() into include_by_path()
The contributor guide has been updated to advise new contributors to
trim irrelevant quoted text when replying to review comments, matching
the existing advice given to reviewers.
* wy/doc-myfirstcontribution-trim-quotes:
MyFirstContribution: mention trimming quoted text in replies
The 'SubmittingPatches' document has been updated to explicitly
describe the expectation for contributors to retract or abandon their
patch series when they are no longer pursuing it.
* jc/submitting-patches-abandoning:
SubmittingPatches: document how to retract a topic
A description in the release notes for Git 2.55.0 has been
retroactively updated to clarify that Rust support is enabled by
default, but still optional, and will become mandatory in Git 3.0.
* jc/relnotes-2.55-rust-fix:
Rust: fix description in Release Notes to 2.55
The alignment of commit object name abbreviations in 'git blame'
output has been optimized to reserve a column for marks (caret,
question mark, or asterisk) only when such marks are actually shown.
* rs/blame-abbrev-marks:
blame: reserve mark column only if necessary
The experimental 'git history' command has been taught a new 'drop'
subcommand to remove a commit, with its descendants replayed onto its
parent.
* ps/history-drop:
builtin/history: implement "drop" subcommand
builtin/history: split handling of ref updates into two phases
replay: expose `replay_result_queue_update()`
reset: stop assuming that the caller passes in a clean index
reset: allow the caller to specify the current HEAD object
reset: introduce ability to skip updating HEAD
reset: introduce dry-run mode
reset: modernize flags passed to `reset_working_tree()`
reset: rename `reset_head()`
reset: drop `USE_THE_REPOSITORY_VARIABLE`
read-cache: split out function to drop unmerged entries to stage 0
The 'git refs' toolbox has been extended with new 'create', 'delete',
'update', and 'rename' subcommands to create, delete, update, and
rename references, respectively.
* ps/refs-writing-subcommands:
builtin/refs: add "rename" subcommand
builtin/refs: add "create" subcommand
builtin/refs: add "update" subcommand
builtin/refs: add "delete" subcommand
builtin/refs: drop `the_repository`
The output of `git fast-export -h` currently starts with:
usage: git fast-export [<rev-list-opts>]
while the SYNOPSIS section in this command's documentation shows:
'git fast-export' [<options>] | 'git fast-import'
Let's make both of these consistent with each other and with other Git
commands by describing the arguments with:
[<options>] [<revision-range>] [[--] <path>...]
This takes into account the following:
- `git fast-export` accepts both rev-list arguments and a number of
genuine options of its own (--[no-]progress, --[no-]signed-tags,
--[no-]signed-commits, etc).
- `git fast-export` was the only command using `[<rev-list-opts>]`
while many other commands describe their revision arguments as
`[<revision-range>] [[--] <path>...]`.
- In the DESCRIPTION section of the documentation, it's already
mentioned several times that the output should eventually be fed to
`git fast-import`.
This also enables us to remove fast-export from
"t/t0450/adoc-help-mismatches".
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When 'git push origin/main' or 'git branch origin main' is run, the
command is now recognized as a potential typo, and advice has been
added to offer a typo fix.
* hn/branch-push-slip-advice:
push: suggest <remote> <branch> for a slash slip
branch: suggest <remote>/<branch> on upstream slip
The includeIf mechanism already supports matching on the .git
directory path (gitdir) and the currently checked out branch
(onbranch). But in multi-worktree setups the .git directory of a
linked worktree points into the main repository's .git/worktrees/
area, which makes gitdir patterns cumbersome when one wants to
include config based on the working tree's checkout path instead.
Introduce two new condition keywords:
- worktree:<pattern> matches the realpath of the current worktree's
working directory (i.e. repo_get_work_tree()) against a glob
pattern. This is the path returned by git rev-parse
--show-toplevel.
- worktree/i:<pattern> is the case-insensitive variant.
The implementation reuses the include_by_path() helper introduced in
the previous commit, passing the worktree path in place of the
gitdir. The condition never matches in bare repositories (where
there is no worktree) or during early config reading (where no
repository is available).
Add documentation describing the new conditions, including a comparison
with extensions.worktreeConfig and a note that worktree matching currently
uses the realpath-resolved worktree location. Add tests covering bare
repositories, multiple worktrees, realpath-resolved symlinked worktree
paths, case-sensitive and case-insensitive matching, early config reading,
and non-repository scenarios.
Signed-off-by: Chen Linxuan <me@black-desk.cn>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
While this document outlines an idealized lifecycle where an author
develops a patch, refines it with reviewer feedback, and
successfully merges it into Git, reality is rarely so seamless.
Sometimes, a topic must be abandoned. Doing so explicitly is far
better than leaving it in limbo, especially since topics can always
be resurrected later.
Clearly state that we encourage contributors to retract any topic
that does not pan out.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Finish incomplete sentence to say that we
- build Git 2.55 by default with Rust,
- but you can opt out and build 2.55 without Rust,
- but Rust will become mandatory in Git 3.0 and later.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Documentation on community contribution guidelines has been updated to
encourage replying to review comments before rerolling, and to advise
a default limit of at most one reroll per day to give reviewers across
different time zones enough time to participate.
* wy/doc-clarify-review-replies:
doc: advise batching patch rerolls
doc: encourage review replies before rerolling
The "git repo info" command has been taught new keys to output both
absolute and relative paths for "gitdir" and "commondir", supported by
a new path-formatting helper extracted from "git rev-parse".
* jk/repo-info-path-keys:
repo: add path.gitdir with absolute and relative suffix formatting
repo: add path.commondir with absolute and relative suffix formatting
path: extract format_path() and use in rev-parse
"git log --follow" has been updated to better handle non-linear
history, in which the path being tracked gets renamed differently in
multiple history lines.
* mv/log-follow-mergy:
log: improve --follow following renames for non-linear history
The pack-objects command has been updated to support reachability
bitmaps and delta-islands concurrently with the `--path-walk` option,
allowing faster packaging by falling back to path-walk when bitmaps
cannot fully satisfy the request.
* tb/pack-path-walk-bitmap-delta-islands:
pack-objects: support `--delta-islands` with `--path-walk`
pack-objects: extract `record_tree_depth()` helper
pack-objects: support reachability bitmaps with `--path-walk`
t/perf: drop p5311's lookup-table permutation
The documentation in SubmittingPatches has been updated to clarify how
patch contributors should respond to design and viability critiques,
and how the resolution of such critiques should be recorded in the
final commit messages.
* jc/submittingpatches-design-critiques:
SubmittingPatches: address design critiques
The trailer sections in SubmittingPatches have been updated to
encourage use of standard trailers.
* kh/submittingpatches-trailers:
SubmittingPatches: note that trailer order matters
SubmittingPatches: be consistent with trailer markup
SubmittingPatches: document Based-on-patch-by trailer
SubmittingPatches: discourage common Linux trailers
SubmittingPatches: encourage trailer use for substantial help
The `fetch.followRemoteHEAD` configuration variable has been added to
provide a default for the per-remote `remote.<name>.followRemoteHEAD`
setting.
* mh/fetch-follow-remote-head-config:
fetch: fixup a misaligned comment
fetch: add configuration variable fetch.followRemoteHEAD
fetch: refactor do_fetch handling of followRemoteHEAD
fetch: return 0 on known git_fetch_config
fetch: rename function report_set_head
t5510: cleanup remote in followRemoteHEAD dangling ref test
doc: explain fetchRemoteHEADWarn advice
fetch: fixup set_head advice for warn-if-not-branch
Project-specific configuration for b4 has been introduced, and the
documentation has been updated to recommend using it as a
streamlined method for submitting patches.
* ps/doc-recommend-b4:
b4: introduce configuration for the Git project
MyFirstContribution: recommend the use of b4
MyFirstContribution: recommend shallow threading of cover letters
The handling of promisor-remote protocol capability has been updated
to allow the other side to add to the list of promisor remotes via the
'promisor.acceptFromServerURL' configuration variable.
* cc/promisor-auto-config-url-more:
doc: promisor: improve acceptFromServer entry
promisor-remote: auto-configure unknown remotes
promisor-remote: trust known remotes matching acceptFromServerUrl
promisor-remote: introduce promisor.acceptFromServerUrl
promisor-remote: add 'local_name' to 'struct promisor_info'
urlmatch: add url_normalize_pattern() helper
urlmatch: change 'allow_globs' arg to bool
t5710: simplify 'mkdir X' followed by 'git -C X init'
Add a "rename" subcommand to git-refs(1) with the syntax:
$ git refs rename <oldref> <newref>
It renames <oldref> together with its reflog to <newref>; even when used
on a local branch ref, the current value and the reflog of the ref are
the only things that are renamed. Document it and redirect casual users
to "git branch -m" if that is what they wanted to do.
Co-authored-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "update" subcommand cannot only update an existing reference, but it
can also create new branches and delete existing branches by specifying
the all-zeroes object ID as either old or new value. Despite that, we
already have the "delete" subcommand as a handy shortcut so that a user
can easily delete a branch. This relieves them of needing to understand
the more arcane uses of the "update" command, and of counting the number
of zeroes they need to pass.
But while we have a "delete" subcommand, we don't have an equivalent
that would allow the user to create a new branch, which creates a
certain asymmetry.
Add a new "create" subcommand to plug this gap.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a new "update" subcommand which mirrors `git update-ref <refname>
<oldoid> <newoid>`. This follows the same reasoning as the preceding
commit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Reference-related functionality in Git is currently spread across many
different commands: git-update-ref(1), git-for-each-ref(1),
git-show-ref(1), git-pack-refs(1) and git-symbolic-ref(1). This makes it
hard for users to discover what functionality we have available to work
with references.
We have thus started to consolidate this functionality into git-refs(1),
which is a toolbox of everything related to references. Until now, the
command doesn't handle functionality of git-update-ref(1).
Fix this gap by introducing a new "delete" subcommand, which is the
equivalent of `git update-ref -d`.
Note that we're intentionally not using a generic "write" subcommand
with a "-d" flag. This is rather harder to discover, and subcommands
that are implmented as flags tend to be hard to reason about in the code
as we'd have to handle mutually-exclusive flags that stem from the other
subcommand-like modes.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
git blame prepends commit hashes of boundary commits with "^", ignored
commits with "?" and unblamable commits with "*" and reserves one column
for them by extending the hash abbreviation, to avoid showing ambiguous
hashes.
This reserved column wastes precious screen space, which can be
especially irritating when using the option -b to blank out boundary
commit hashes and not ignoring any commits. Reserve it only as needed,
i.e. if any of those cases are actually shown.
Pointed-out-by: Laszlo Ersek <laszlo.ersek@posteo.net>
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A common operation when editing the commit history is to drop a specific
commit from the history entirely, but this operation is not currently
covered by git-history(1).
A couple of noteworthy bits:
- This is the first git-history(1) command that will ultimately result
in changes to both the index and the working tree. We thus have to
add logic to merge resulting changes into those.
- It is still not possible to replay merge commits, so this limitation
is inherited for the new "drop" command.
- For now we refuse to drop root commits. While we _can_ indeed drop
root commits in the general case, there are edge cases where the
resulting history would become completely empty. This is thus left
to a subsequent patch series.
Other than that, most of the logic is rather straight-forward as we can
continue to build on the preexisting logic in git-history(1) for most of
the part.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When pushing the 'main' branch to the remote 'origin', i.e.,
$ git push origin main
it is easy to mistakenly write
$ git push origin/main
That is parsed as the repository to push to, and since 'origin/main'
is neither a configured remote nor a path it dies with:
fatal: 'origin/main' does not appear to be a git repository
Often 'origin/main' does not exist as a repository, so the command
fails without doing any harm, but it gives no hint that a space was
meant instead of a slash and can leave the user puzzled.
When the argument is not an existing path or configured remote but
its part before the first slash names one, suggest the intended
'<remote> <branch>' form:
$ git push origin main
The suggestion is shown as advice so it can be silenced with
advice.pushRepoLooksLikeRef.
Signed-off-by: Harald Nordgren <haraldnordgren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Scripts need a stable way to locate the git directory without
parsing rev-parse output or relying on its flag-driven path format
selection. There is no way to retrieve this path from git repo info
today.
Introduce path.gitdir.absolute and path.gitdir.relative keys,
consistent with the path.commondir keys added in the previous patch.
Reuse the test_repo_info_path helper introduced there to validate
both variants.
Mentored-by: Justin Tobler <jltobler@gmail.com>
Mentored-by: Lucas Seiki Oshiro <lucasseikioshiro@gmail.com>
Signed-off-by: K Jayatheerth <jayatheerthkulkarni2005@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Scripts working with worktree setups need a reliable way to discover
the common directory, which diverges from the git directory when
multiple worktrees are in use. There is no way to retrieve this path
from git repo info today.
Introduce path.commondir.absolute and path.commondir.relative keys.
Exposing explicit format variants rather than a single key with a
default avoids ambiguity for scripts that require predictable output.
Mentored-by: Justin Tobler <jltobler@gmail.com>
Mentored-by: Lucas Seiki Oshiro <lucasseikioshiro@gmail.com>
Signed-off-by: K Jayatheerth <jayatheerthkulkarni2005@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Have a repo with a subtree merge, do a 'git log --follow prefix/test.c',
the output only contains history in the outer repo, not commits that
were merged via a subtree merge.
What happens is that 'git log --follow' stores the followed path only in
opt->diffopt.pathspec, so in case the commit history is non-linear, and
multiple parents have renames to the followed path, then the end result
isn't really defined: the first commit that happens to be visited in one
of the parents update opt->diffopt.pathspec, and from that point, only
that updated path is visited.
Fix the problem by introducing a commit -> path map
(follow_pathspec_slab) that stores what will be a path to follow when
visiting that parent. At the top of log_tree_commit(), if the slab has
an entry for this commit, we replace opt->diffopt.pathspec with a path
from this entry, so the correct path is followed, even if an unrelated
sub-tree changed the path to be followed to something else. After
log_tree_diff() runs, we record each parent's path in the slab. As a
result, the walk order doesn't matter, which was exactly the source of
problems previously.
This helps with subtree merges (rename happens inside the merge commit),
but also fixes the general case when the rename happens in the history
of parents, not in the merge commit itself.
Signed-off-by: Miklos Vajna <vmiklos@collabora.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since the inception of `--path-walk`, this option has had a documented
incompatibility with `--delta-islands`.
When discussing those original patches on the list, a message from
Stolee in [1] noted the following:
this could be remedied by [...] doing a separate walk to identify
islands using the normal method
In a related portion of the thread, Peff explains[2]:
The delta islands code already does its own tree walk to propagate
the bits down (it does rely on the base walk's show_commit() to
propagate through the commits).
Once each object has its island bitmaps, I think however you
choose to come up with delta candidates [...] you should be able
to use it. It's fundamentally just answering the question of "am
I allowed to delta between these two objects".
That is similar to what this patch does, and it turns out the cheaper
option is sufficient: perform the same island side effects from the
path-walk callback rather than doing a second walk.
Recall how delta-islands are computed during a normal repack:
- `show_commit()` calls `propagate_island_marks()` for each commit,
which merges the commit's island bitset onto its root tree object and
onto each of its parent commits.
- `show_object()` for a tree records the tree's depth derived from the
slash-separated pathname. Subsequent `resolve_tree_islands()` uses
that depth to walk trees in increasing-depth order, propagating each
tree's marks to its children.
- At delta-search time, `in_same_island()` enforces that a delta
target's island bitmap is a subset of its base's: every island that
reaches the target must also reach the base.
Path-walk's enumeration callback is `add_objects_by_path()`. It already
adds objects to `to_pack`, but until now did not perform the
island-related side effects. Two things are needed:
- For each commit batch, call `propagate_island_marks()` on commits,
exactly as `show_commit()` does.
We have to be careful about the order in which we call this function,
and we must see a commit before its parents in order to have
island marks to propagate.
The path-walk batch preserves that order. Path-walk appends commits
to its `OBJ_COMMIT` batch as they come back from the same
`get_revision()` loop the regular traversal uses, and
`add_objects_by_path()` iterates the batch in array order. So every
commit reaches `propagate_island_marks()` in the same sequence that
`show_commit()` would have seen it, and the descendant-first chain
that the algorithm relies on is intact.
Skip island propagation for excluded commits to match the regular
traversal, whose `show_commit()` callback is only invoked for
interesting commits. Boundary commits may still be present in
path-walk's callback so they can serve as thin-pack bases, but they
should not contribute island marks.
- For each tree batch, record the tree's depth from the path. Use the
`record_tree_depth()` helper from the previous commit so both
callbacks behave identically, including the max-depth-wins behavior
when a tree is reached via more than one path. The helper accepts
both the `show_object()` path shape ("foo", "foo/bar") and the
path-walk shape with a trailing slash ("foo/", "foo/bar/"), so depths
recorded from either traversal mode are directly comparable.
This is implicit in the implementation sketch from Peff above.
`resolve_tree_islands()` sorts trees by `oe->tree_depth` in
increasing-depth order before propagating marks down, so that a
parent tree's marks are finalized before its children inherit them.
Without recording the depth at path-walk time, every
path-walk-discovered tree would land at depth 0 in `to_pack`, the
sort would lose its ordering, and children could inherit marks from
parents whose own contributions had not yet been merged in.
With those two pieces in place, `resolve_tree_islands()` receives the
same island inputs from path-walk as it would from the regular
traversal, so the existing island checks can be reused unchanged.
Drop the documented incompatibility between `--path-walk` and
`--delta-islands`, and add t5320 coverage for path-walk island repacks
with and without bitmap writing, as well as the same-island case where a
delta remains allowed.
[1]: https://lore.kernel.org/git/9aa2471b-0850-4707-9733-d3b33609f5f2@gmail.com/
[2]: https://lore.kernel.org/git/20240911063203.GA1538586@coredump.intra.peff.net/
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When 'pack-objects' is invoked with '--path-walk', it prevents us from
using reachability bitmaps.
This behavior dates back to 70664d2865 (pack-objects: add --path-walk
option, 2025-05-16), which included a comment in the relevant portion of
the command-line arguments handling that read as follows:
/*
* We must disable the bitmaps because we are removing
* the --objects / --objects-edge[-aggressive] options.
*/
In fb2c309b7d3 (pack-objects: pass --objects with --path-walk,
2026-05-02), path-walk learned to pass '--objects' again, but still
kept bitmap traversal disabled. That leaves two useful cases
unsupported:
* A path-walk repack that writes bitmaps does not give the bitmap
selector any commits, because path-walk reveals commits through
`add_objects_by_path()` rather than through `show_commit()`, where
`index_commit_for_bitmap()` is normally called.
* An invocation like "git pack-objects --use-bitmap-index --path-walk"
never tries an existing bitmap, even when one is available and could
answer the request.
Fortunately for us, neither restriction is required.
* On the writing side: teach the path-walk object callback to call
`index_commit_for_bitmap()` for commits that it adds to the pack.
That gives the bitmap selector the commit candidates it would have
seen from the regular traversal.
* For bitmap reading, keep passing '--objects' to the internal rev_list
machinery, but stop clearing `use_bitmap_index`. If an existing
bitmap can answer the request, use it; otherwise fall back to
path-walk's own enumeration.
As a result, we can see significantly reduced pack generation times from
p5311 (with our `GIT_PERF_REPO` set to a recent clone of the fluentui
repository) before this commit:
Test HEAD^ HEAD
----------------------------------------------------------------------------------------
5311.40: server (1 days, --path-walk) 1.43(1.39+0.04) 0.01(0.01+0.00) -99.3%
5311.41: size (1 days, --path-walk) 139.6K 139.7K +0.0%
5311.42: client (1 days, --path-walk) 0.02(0.02+0.00) 0.02(0.02+0.00) +0.0%
5311.44: server (2 days, --path-walk) 1.43(1.39+0.04) 0.01(0.00+0.00) -99.3%
5311.45: size (2 days, --path-walk) 139.6K 139.7K +0.0%
5311.46: client (2 days, --path-walk) 0.02(0.02+0.00) 0.02(0.02+0.00) +0.0%
5311.48: server (4 days, --path-walk) 1.44(1.39+0.04) 0.01(0.01+0.00) -99.3%
5311.49: size (4 days, --path-walk) 238.1K 238.1K +0.0%
5311.50: client (4 days, --path-walk) 0.03(0.03+0.00) 0.03(0.03+0.00) +0.0%
5311.52: server (8 days, --path-walk) 1.43(1.39+0.03) 0.01(0.00+0.00) -99.3%
5311.53: size (8 days, --path-walk) 344.9K 344.9K +0.0%
5311.54: client (8 days, --path-walk) 0.07(0.07+0.00) 0.07(0.08+0.00) +0.0%
5311.56: server (16 days, --path-walk) 1.47(1.44+0.03) 0.10(0.08+0.01) -93.2%
5311.57: size (16 days, --path-walk) 844.0K 844.0K +0.0%
5311.58: client (16 days, --path-walk) 0.09(0.09+0.00) 0.09(0.09+0.00) +0.0%
5311.60: server (32 days, --path-walk) 1.52(1.50+0.05) 0.14(0.15+0.02) -90.8%
5311.61: size (32 days, --path-walk) 4.2M 4.2M +0.1%
5311.62: client (32 days, --path-walk) 0.34(0.48+0.02) 0.34(0.45+0.05) +0.0%
5311.64: server (64 days, --path-walk) 1.55(1.52+0.06) 0.15(0.15+0.04) -90.3%
5311.65: size (64 days, --path-walk) 6.4M 6.4M -0.0%
5311.66: client (64 days, --path-walk) 0.51(0.79+0.05) 0.51(0.80+0.06) +0.0%
5311.68: server (128 days, --path-walk) 1.59(1.57+0.06) 0.16(0.21+0.01) -89.9%
5311.69: size (128 days, --path-walk) 8.4M 8.4M -0.0%
5311.70: client (128 days, --path-walk) 0.72(1.44+0.08) 0.71(1.47+0.09) -1.4%
We get the same size of output pack, but this commit allows us to do so
in a significantly shorter amount of time. Intuitively, we're generating
the same pack (hence the unchanged 'test_size' output from run to run),
but varying how we get there. Before this commit, pack-objects prefers
'--path-walk' to '--use-bitmap-index', so we generate the output pack by
performing a normal '--path-walk' traversal. With this commit, we are
operating over a *repacked* state (that itself was done with a
'--path-walk' traversal), but are able to perform pack-reuse on that
repacked state via bitmaps.
When comparing the size of the repacked pack with/without '--path-walk'
on the previous commit versus this one, we see that (a) the repacked size
improves significantly with '--path-walk', and that (b) writing bitmaps
during repacking does not regress this improvement:
Test HEAD^ HEAD
----------------------------------------------------------------------------------------
5311.3: size of bitmapped pack 558.4M 558.5M +0.0%
5311.38: size of bitmapped pack (--path-walk) 164.4M 164.4M +0.0%
(Note that to observe an improvement here, we must repack with '-F' in
order to avoid reusing non-'--path-walk' deltas, which would otherwise
skew our results.)
There is one wrinkle when it comes to '--boundary', which we must not
pass into the bitmap walk in the presence of both '--path-walk' and
'--use-bitmap-index'. Path-walk needs boundary commits when it performs
its own traversal, in order to discover bases for thin packs, but the
bitmap traversal does not expect this. Work around this by setting
`revs->boundary` as late as possible within the '--path-walk' traversal,
after any bitmap attempt has either succeeded or declined to answer the
request.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Contributors often need guidance on how quickly to send later iterations
of a patch series. Add a rough default of no more than one new version
of the same series per day so feedback can be batched and reviewers have
time to comment regardless of their time zones.
Mention factors that can affect the timing, such as series size, review
depth, and substantial rework. Also point out that avoiding rapid
rerolls encourages authors to polish each version before sending it, so
reviewers can focus on substantial issues.
Helped-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Weijie Yuan <wy@wyuan.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Review feedback should not be answered only by sending a new patch
version. Encourage contributors to discuss their planned response in the
mailing-list thread before rerolling.
This makes the author's reasoning explicit before the next version is
prepared, instead of forcing reviewers to infer it from the rerolled
patches. It also encourages more direct social interaction between
contributors and helps foster a more collaborative review process.
Signed-off-by: Weijie Yuan <wy@wyuan.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Contributors sometimes fail to answer fundamental design or
viability comments from reviewers and submit subsequent rounds
without addressing them. When design decisions are resolved on the
mailing list, the final justification should be recorded in the
commit messages.
Instruct authors to be particularly mindful of critiques regarding
high-level design or viability, to defend their choices on the list,
and to accompany new iterations with clearer explanations in the cover
letter, responses, and revised commit messages. Also instruct them to
explicitly document the resolution of these concerns in the commit
message body to keep the historical record complete.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano