4 Besides specifying a range of commits that should be listed using the
5 special notations explained in the description, additional commit
6 limiting may be applied. Note that they are applied before commit
7 ordering and formatting options, such as '--reverse'.
9 Different options are ANDed: '--author=bar --grep=foo'
10 limits to commits which match both conditions.
12 Several occurences of the '--grep' option are ORed: '--grep=foo --grep=bar'
13 limits to commits matching any of these conditions.
14 (If '--all-match' is given, the conditions are ANDed.)
16 Several occurences of the '--author' and '--committer' options are ORed
17 (because there can be only one each per commit).
19 For all other options, only the last occurence of the same option is
20 taken into account: '-n 5 -3' is '-3' and '-n 3 -n 5' is '-n 5'.
25 --max-count=<number>::
27 Limit the number of commits to output.
31 Skip 'number' commits before starting to show the commit output.
36 Show commits more recent than a specific date.
41 Show commits older than a specific date.
44 --max-age=<timestamp>::
45 --min-age=<timestamp>::
47 Limit the commits output to specified time range.
51 --committer=<pattern>::
53 Limit the commits output to ones with author/committer
54 header lines that match the specified pattern (regular expression).
58 Limit the commits output to ones with log message that
59 matches the specified pattern (regular expression).
62 Limit the commits output to ones that match all given --grep,
63 instead of ones that match at least one.
66 --regexp-ignore-case::
68 Match the regexp limiting patterns without regard to letters case.
73 Consider the limiting patterns to be extended regular expressions
74 instead of the default basic regular expressions.
79 Consider the limiting patterns to be fixed strings (don't interpret
80 pattern as a regular expression).
84 Stop when a given path disappears from the tree.
88 Print only merge commits. This is exactly the same as `--min-parents=2`.
92 Do not print commits with more than one parent. This is
93 exactly the same as `--max-parents=1`.
95 --min-parents=<number>::
96 --max-parents=<number>::
100 Show only commits which have at least (or at most) that many
101 commits. In particular, `--max-parents=1` is the same as `--no-merges`,
102 `--min-parents=2` is the same as `--merges`. `--max-parents=0`
103 gives all root commits and `--min-parents=3` all octopus merges.
105 `--no-min-parents` and `--no-max-parents` reset these limits (to no limit)
106 again. Equivalent forms are `--min-parents=0` (any commit has 0 or more
107 parents) and `--max-parents=-1` (negative numbers denote no upper limit).
110 Follow only the first parent commit upon seeing a merge
111 commit. This option can give a better overview when
112 viewing the evolution of a particular topic branch,
113 because merges into a topic branch tend to be only about
114 adjusting to updated upstream from time to time, and
115 this option allows you to ignore the individual commits
116 brought in to your history by such a merge.
120 Reverses the meaning of the '{caret}' prefix (or lack thereof)
121 for all following revision specifiers, up to the next '--not'.
125 Pretend as if all the refs in `refs/` are listed on the
126 command line as '<commit>'.
128 --branches[=<pattern>]::
130 Pretend as if all the refs in `refs/heads` are listed
131 on the command line as '<commit>'. If '<pattern>' is given, limit
132 branches to ones matching given shell glob. If pattern lacks '?',
133 '{asterisk}', or '[', '/{asterisk}' at the end is implied.
137 Pretend as if all the refs in `refs/tags` are listed
138 on the command line as '<commit>'. If '<pattern>' is given, limit
139 tags to ones matching given shell glob. If pattern lacks '?', '{asterisk}',
140 or '[', '/{asterisk}' at the end is implied.
142 --remotes[=<pattern>]::
144 Pretend as if all the refs in `refs/remotes` are listed
145 on the command line as '<commit>'. If '<pattern>' is given, limit
146 remote-tracking branches to ones matching given shell glob.
147 If pattern lacks '?', '{asterisk}', or '[', '/{asterisk}' at the end is implied.
149 --glob=<glob-pattern>::
150 Pretend as if all the refs matching shell glob '<glob-pattern>'
151 are listed on the command line as '<commit>'. Leading 'refs/',
152 is automatically prepended if missing. If pattern lacks '?', '{asterisk}',
153 or '[', '/{asterisk}' at the end is implied.
157 Upon seeing an invalid object name in the input, pretend as if
158 the bad input was not given.
160 ifndef::git-rev-list[]
163 Pretend as if the bad bisection ref `refs/bisect/bad`
164 was listed and as if it was followed by `--not` and the good
165 bisection refs `refs/bisect/good-*` on the command
167 endif::git-rev-list[]
171 In addition to the '<commit>' listed on the command
172 line, read them from the standard input. If a '--' separator is
173 seen, stop reading commits and start reading paths to limit the
176 ifdef::git-rev-list[]
179 Don't print anything to standard output. This form
180 is primarily meant to allow the caller to
181 test the exit status to see if a range of objects is fully
182 connected (or not). It is faster than redirecting stdout
183 to /dev/null as the output does not have to be formatted.
184 endif::git-rev-list[]
188 Like `--cherry-pick` (see below) but mark equivalent commits
189 with `=` rather than omitting them, and inequivalent ones with `+`.
193 Omit any commit that introduces the same change as
194 another commit on the "other side" when the set of
195 commits are limited with symmetric difference.
197 For example, if you have two branches, `A` and `B`, a usual way
198 to list all commits on only one side of them is with
199 `--left-right` (see the example below in the description of
200 the `--left-right` option). It however shows the commits that were cherry-picked
201 from the other branch (for example, "3rd on b" may be cherry-picked
202 from branch A). With this option, such pairs of commits are
203 excluded from the output.
208 List only commits on the respective side of a symmetric range,
209 i.e. only those which would be marked `<` resp. `>` by
212 For example, `--cherry-pick --right-only A...B` omits those
213 commits from `B` which are in `A` or are patch-equivalent to a commit in
214 `A`. In other words, this lists the `+` commits from `git cherry A B`.
215 More precisely, `--cherry-pick --right-only --no-merges` gives the exact
220 A synonym for `--right-only --cherry-mark --no-merges`; useful to
221 limit the output to the commits on our side and mark those that
222 have been applied to the other side of a forked history with
223 `git log --cherry upstream...mybranch`, similar to
224 `git cherry upstream mybranch`.
229 Instead of walking the commit ancestry chain, walk
230 reflog entries from the most recent one to older ones.
231 When this option is used you cannot specify commits to
232 exclude (that is, '{caret}commit', 'commit1..commit2',
233 nor 'commit1\...commit2' notations cannot be used).
235 With '\--pretty' format other than oneline (for obvious reasons),
236 this causes the output to have two extra lines of information
237 taken from the reflog. By default, 'commit@\{Nth}' notation is
238 used in the output. When the starting commit is specified as
239 'commit@\{now}', output also uses 'commit@\{timestamp}' notation
240 instead. Under '\--pretty=oneline', the commit message is
241 prefixed with this information on the same line.
242 This option cannot be combined with '\--reverse'.
243 See also linkgit:git-reflog[1].
247 After a failed merge, show refs that touch files having a
248 conflict and don't exist on all heads to merge.
252 Output uninteresting commits at the boundary, which are usually
257 History Simplification
258 ~~~~~~~~~~~~~~~~~~~~~~
260 Sometimes you are only interested in parts of the history, for example the
261 commits modifying a particular <path>. But there are two parts of
262 'History Simplification', one part is selecting the commits and the other
263 is how to do it, as there are various strategies to simplify the history.
265 The following options select the commits to be shown:
269 Commits modifying the given <paths> are selected.
271 --simplify-by-decoration::
273 Commits that are referred by some branch or tag are selected.
275 Note that extra commits can be shown to give a meaningful history.
277 The following options affect the way the simplification is performed:
281 Simplifies the history to the simplest history explaining the
282 final state of the tree. Simplest because it prunes some side
283 branches if the end result is the same (i.e. merging branches
284 with the same content)
288 Same as the default mode, but does not prune some history.
292 Only the selected commits are shown, plus some to have a
297 All commits in the simplified history are shown.
301 Additional option to '--full-history' to remove some needless
302 merges from the resulting history, as there are no selected
303 commits contributing to this merge.
307 When given a range of commits to display (e.g. 'commit1..commit2'
308 or 'commit2 {caret}commit1'), only display commits that exist
309 directly on the ancestry chain between the 'commit1' and
310 'commit2', i.e. commits that are both descendants of 'commit1',
311 and ancestors of 'commit2'.
313 A more detailed explanation follows.
315 Suppose you specified `foo` as the <paths>. We shall call commits
316 that modify `foo` !TREESAME, and the rest TREESAME. (In a diff
317 filtered for `foo`, they look different and equal, respectively.)
319 In the following, we will always refer to the same example history to
320 illustrate the differences between simplification settings. We assume
321 that you are filtering for a file `foo` in this commit graph:
322 -----------------------------------------------------------------------
328 -----------------------------------------------------------------------
329 The horizontal line of history A---P is taken to be the first parent of
330 each merge. The commits are:
332 * `I` is the initial commit, in which `foo` exists with contents
333 "asdf", and a file `quux` exists with contents "quux". Initial
334 commits are compared to an empty tree, so `I` is !TREESAME.
336 * In `A`, `foo` contains just "foo".
338 * `B` contains the same change as `A`. Its merge `M` is trivial and
339 hence TREESAME to all parents.
341 * `C` does not change `foo`, but its merge `N` changes it to "foobar",
342 so it is not TREESAME to any parent.
344 * `D` sets `foo` to "baz". Its merge `O` combines the strings from
345 `N` and `D` to "foobarbaz"; i.e., it is not TREESAME to any parent.
347 * `E` changes `quux` to "xyzzy", and its merge `P` combines the
348 strings to "quux xyzzy". Despite appearing interesting, `P` is
349 TREESAME to all parents.
351 'rev-list' walks backwards through history, including or excluding
352 commits based on whether '\--full-history' and/or parent rewriting
353 (via '\--parents' or '\--children') are used. The following settings
358 Commits are included if they are not TREESAME to any parent
359 (though this can be changed, see '\--sparse' below). If the
360 commit was a merge, and it was TREESAME to one parent, follow
361 only that parent. (Even if there are several TREESAME
362 parents, follow only one of them.) Otherwise, follow all
367 -----------------------------------------------------------------------
371 -----------------------------------------------------------------------
373 Note how the rule to only follow the TREESAME parent, if one is
374 available, removed `B` from consideration entirely. `C` was
375 considered via `N`, but is TREESAME. Root commits are compared to an
376 empty tree, so `I` is !TREESAME.
378 Parent/child relations are only visible with --parents, but that does
379 not affect the commits selected in default mode, so we have shown the
382 --full-history without parent rewriting::
384 This mode differs from the default in one point: always follow
385 all parents of a merge, even if it is TREESAME to one of them.
386 Even if more than one side of the merge has commits that are
387 included, this does not imply that the merge itself is! In
390 -----------------------------------------------------------------------
392 -----------------------------------------------------------------------
394 `P` and `M` were excluded because they are TREESAME to a parent. `E`,
395 `C` and `B` were all walked, but only `B` was !TREESAME, so the others
398 Note that without parent rewriting, it is not really possible to talk
399 about the parent/child relationships between the commits, so we show
402 --full-history with parent rewriting::
404 Ordinary commits are only included if they are !TREESAME
405 (though this can be changed, see '\--sparse' below).
407 Merges are always included. However, their parent list is rewritten:
408 Along each parent, prune away commits that are not included
409 themselves. This results in
411 -----------------------------------------------------------------------
417 -----------------------------------------------------------------------
419 Compare to '\--full-history' without rewriting above. Note that `E`
420 was pruned away because it is TREESAME, but the parent list of P was
421 rewritten to contain `E`'s parent `I`. The same happened for `C` and
422 `N`. Note also that `P` was included despite being TREESAME.
424 In addition to the above settings, you can change whether TREESAME
429 Commits that are walked are included if they are not TREESAME
434 All commits that are walked are included.
436 Note that without '\--full-history', this still simplifies merges: if
437 one of the parents is TREESAME, we follow only that one, so the other
438 sides of the merge are never walked.
442 First, build a history graph in the same way that
443 '\--full-history' with parent rewriting does (see above).
445 Then simplify each commit `C` to its replacement `C'` in the final
446 history according to the following rules:
451 * Replace each parent `P` of `C'` with its simplification `P'`. In
452 the process, drop parents that are ancestors of other parents, and
455 * If after this parent rewriting, `C'` is a root or merge commit (has
456 zero or >1 parents), a boundary commit, or !TREESAME, it remains.
457 Otherwise, it is replaced with its only parent.
460 The effect of this is best shown by way of comparing to
461 '\--full-history' with parent rewriting. The example turns into:
463 -----------------------------------------------------------------------
469 -----------------------------------------------------------------------
471 Note the major differences in `N` and `P` over '--full-history':
474 * `N`'s parent list had `I` removed, because it is an ancestor of the
475 other parent `M`. Still, `N` remained because it is !TREESAME.
477 * `P`'s parent list similarly had `I` removed. `P` was then
478 removed completely, because it had one parent and is TREESAME.
481 Finally, there is a fifth simplification mode available:
485 Limit the displayed commits to those directly on the ancestry
486 chain between the "from" and "to" commits in the given commit
487 range. I.e. only display commits that are ancestor of the "to"
488 commit, and descendants of the "from" commit.
490 As an example use case, consider the following commit history:
492 -----------------------------------------------------------------------
495 B---C---G---H---I---J
497 A-------K---------------L--M
498 -----------------------------------------------------------------------
500 A regular 'D..M' computes the set of commits that are ancestors of `M`,
501 but excludes the ones that are ancestors of `D`. This is useful to see
502 what happened to the history leading to `M` since `D`, in the sense
503 that "what does `M` have that did not exist in `D`". The result in this
504 example would be all the commits, except `A` and `B` (and `D` itself,
507 When we want to find out what commits in `M` are contaminated with the
508 bug introduced by `D` and need fixing, however, we might want to view
509 only the subset of 'D..M' that are actually descendants of `D`, i.e.
510 excluding `C` and `K`. This is exactly what the '--ancestry-path'
511 option does. Applied to the 'D..M' range, it results in:
513 -----------------------------------------------------------------------
519 -----------------------------------------------------------------------
521 The '\--simplify-by-decoration' option allows you to view only the
522 big picture of the topology of the history, by omitting commits
523 that are not referenced by tags. Commits are marked as !TREESAME
524 (in other words, kept after history simplification rules described
525 above) if (1) they are referenced by tags, or (2) they change the
526 contents of the paths given on the command line. All other
527 commits are marked as TREESAME (subject to be simplified away).
529 ifdef::git-rev-list[]
535 Limit output to the one commit object which is roughly halfway between
536 included and excluded commits. Note that the bad bisection ref
537 `refs/bisect/bad` is added to the included commits (if it
538 exists) and the good bisection refs `refs/bisect/good-*` are
539 added to the excluded commits (if they exist). Thus, supposing there
540 are no refs in `refs/bisect/`, if
542 -----------------------------------------------------------------------
543 $ git rev-list --bisect foo ^bar ^baz
544 -----------------------------------------------------------------------
546 outputs 'midpoint', the output of the two commands
548 -----------------------------------------------------------------------
549 $ git rev-list foo ^midpoint
550 $ git rev-list midpoint ^bar ^baz
551 -----------------------------------------------------------------------
553 would be of roughly the same length. Finding the change which
554 introduces a regression is thus reduced to a binary search: repeatedly
555 generate and test new 'midpoint's until the commit chain is of length
560 This calculates the same as `--bisect`, except that refs in
561 `refs/bisect/` are not used, and except that this outputs
562 text ready to be eval'ed by the shell. These lines will assign the
563 name of the midpoint revision to the variable `bisect_rev`, and the
564 expected number of commits to be tested after `bisect_rev` is tested
565 to `bisect_nr`, the expected number of commits to be tested if
566 `bisect_rev` turns out to be good to `bisect_good`, the expected
567 number of commits to be tested if `bisect_rev` turns out to be bad to
568 `bisect_bad`, and the number of commits we are bisecting right now to
573 This outputs all the commit objects between the included and excluded
574 commits, ordered by their distance to the included and excluded
575 commits. Refs in `refs/bisect/` are not used. The farthest
576 from them is displayed first. (This is the only one displayed by
579 This is useful because it makes it easy to choose a good commit to
580 test when you want to avoid to test some of them for some reason (they
581 may not compile for example).
583 This option can be used along with `--bisect-vars`, in this case,
584 after all the sorted commit objects, there will be the same text as if
585 `--bisect-vars` had been used alone.
586 endif::git-rev-list[]
592 By default, the commits are shown in reverse chronological order.
595 Show no parents before all of its children are shown, but
596 otherwise show commits in the commit timestamp order.
599 Show no parents before all of its children are shown, and
600 avoid showing commits on multiple lines of history
603 For example, in a commit history like this:
605 ----------------------------------------------------------------
611 ----------------------------------------------------------------
613 where the numbers denote the order of commit timestamps, `git
614 rev-list` and friends with `--date-order` show the commits in the
615 timestamp order: 8 7 6 5 4 3 2 1.
617 With `--topo-order`, they would show 8 6 5 3 7 4 2 1 (or 8 7 4 2 6 5
618 3 1); some older commits are shown before newer ones in order to
619 avoid showing the commits from two parallel development track mixed
624 Output the commits in reverse order.
625 Cannot be combined with '\--walk-reflogs'.
630 These options are mostly targeted for packing of git repositories.
634 Print the object IDs of any object referenced by the listed
635 commits. '--objects foo ^bar' thus means "send me
636 all object IDs which I need to download if I have the commit
637 object 'bar', but not 'foo'".
641 Similar to '--objects', but also print the IDs of excluded
642 commits prefixed with a "-" character. This is used by
643 linkgit:git-pack-objects[1] to build "thin" pack, which records
644 objects in deltified form based on objects contained in these
645 excluded commits to reduce network traffic.
649 Only useful with '--objects'; print the object IDs that are not
652 --no-walk[=(sorted|unsorted)]::
654 Only show the given commits, but do not traverse their ancestors.
655 This has no effect if a range is specified. If the argument
656 "unsorted" is given, the commits are show in the order they were
657 given on the command line. Otherwise (if "sorted" or no argument
658 was given), the commits are show in reverse chronological order
663 Overrides a previous --no-walk.
668 ifdef::git-rev-list[]
669 Using these options, linkgit:git-rev-list[1] will act similar to the
670 more specialized family of commit log tools: linkgit:git-log[1],
671 linkgit:git-show[1], and linkgit:git-whatchanged[1]
672 endif::git-rev-list[]
674 include::pretty-options.txt[]
678 Synonym for `--date=relative`.
680 --date=(relative|local|default|iso|rfc|short|raw)::
682 Only takes effect for dates shown in human-readable format, such
683 as when using "--pretty". `log.date` config variable sets a default
684 value for log command's --date option.
686 `--date=relative` shows dates relative to the current time,
689 `--date=local` shows timestamps in user's local timezone.
691 `--date=iso` (or `--date=iso8601`) shows timestamps in ISO 8601 format.
693 `--date=rfc` (or `--date=rfc2822`) shows timestamps in RFC 2822
694 format, often found in E-mail messages.
696 `--date=short` shows only date but not time, in `YYYY-MM-DD` format.
698 `--date=raw` shows the date in the internal raw git format `%s %z` format.
700 `--date=default` shows timestamps in the original timezone
701 (either committer's or author's).
703 ifdef::git-rev-list[]
706 Print the contents of the commit in raw-format; each record is
707 separated with a NUL character.
708 endif::git-rev-list[]
712 Print also the parents of the commit (in the form "commit parent...").
713 Also enables parent rewriting, see 'History Simplification' below.
717 Print also the children of the commit (in the form "commit child...").
718 Also enables parent rewriting, see 'History Simplification' below.
720 ifdef::git-rev-list[]
722 Print the raw commit timestamp.
723 endif::git-rev-list[]
727 Mark which side of a symmetric diff a commit is reachable from.
728 Commits from the left side are prefixed with `<` and those from
729 the right with `>`. If combined with `--boundary`, those
730 commits are prefixed with `-`.
732 For example, if you have this topology:
734 -----------------------------------------------------------------------
739 o---x---a---a branch A
740 -----------------------------------------------------------------------
742 you would get an output like this:
744 -----------------------------------------------------------------------
745 $ git rev-list --left-right --boundary --pretty=oneline A...B
753 -----------------------------------------------------------------------
757 Draw a text-based graphical representation of the commit history
758 on the left hand side of the output. This may cause extra lines
759 to be printed in between commits, in order for the graph history
760 to be drawn properly.
762 This enables parent rewriting, see 'History Simplification' below.
764 This implies the '--topo-order' option by default, but the
765 '--date-order' option may also be specified.
767 ifdef::git-rev-list[]
769 Print a number stating how many commits would have been
770 listed, and suppress all other output. When used together
771 with '--left-right', instead print the counts for left and
772 right commits, separated by a tab. When used together with
773 '--cherry-mark', omit patch equivalent commits from these
774 counts and print the count for equivalent commits separated
776 endif::git-rev-list[]
779 ifndef::git-rev-list[]
783 Below are listed options that control the formatting of diff output.
784 Some of them are specific to linkgit:git-rev-list[1], however other diff
785 options may be given. See linkgit:git-diff-files[1] for more options.
789 With this option, diff output for a merge commit
790 shows the differences from each of the parents to the merge result
791 simultaneously instead of showing pairwise diff between a parent
792 and the result one at a time. Furthermore, it lists only files
793 which were modified from all parents.
797 This flag implies the '-c' option and further compresses the
798 patch output by omitting uninteresting hunks whose contents in
799 the parents have only two variants and the merge result picks
800 one of them without modification.
804 This flag makes the merge commits show the full diff like
805 regular commits; for each merge parent, a separate log entry
806 and diff is generated. An exception is that only diff against
807 the first parent is shown when '--first-parent' option is given;
808 in that case, the output represents the changes the merge
809 brought _into_ the then-current branch.
813 Show recursive diffs.
817 Show the tree objects in the diff output. This implies '-r'.
820 Suppress diff output.
821 endif::git-rev-list[]