"git log" learned a new "--diff-merges=<how>" option.
* so/log-diff-merge: (32 commits)
t4013: add tests for --diff-merges=first-parent
doc/git-show: include --diff-merges description
doc/rev-list-options: document --first-parent changes merges format
doc/diff-generate-patch: mention new --diff-merges option
doc/git-log: describe new --diff-merges options
diff-merges: add '--diff-merges=1' as synonym for 'first-parent'
diff-merges: add old mnemonic counterparts to --diff-merges
diff-merges: let new options enable diff without -p
diff-merges: do not imply -p for new options
diff-merges: implement new values for --diff-merges
diff-merges: make -m/-c/--cc explicitly mutually exclusive
diff-merges: refactor opt settings into separate functions
diff-merges: get rid of now empty diff_merges_init_revs()
diff-merges: group diff-merge flags next to each other inside 'rev_info'
diff-merges: split 'ignore_merges' field
diff-merges: fix -m to properly override -c/--cc
t4013: add tests for -m failing to override -c/--cc
t4013: support test_expect_failure through ':failure' magic
diff-merges: revise revs->diff flag handling
diff-merges: handle imply -p on -c/--cc logic for log.c
...
New options don't have any visible effect unless -p is either given or
implied, as unlike -c/-cc we don't imply -p with --diff-merges. To fix
this, this patch adds new functionality by letting new options enable
output of diffs for merge commits only.
Add 'merges_need_diff' field and set it whenever diff output for merges is
enabled by any of the new options.
Extend diff output logic accordingly, to output diffs for merges when
'merges_need_diff' is set even when no -p has been provided.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add 'combined_imply_patch' field and set it only for old --cc/-c
options, then imply -p if this flag is set instead of implying -p
whenever 'combined_merge' flag is set.
We don't want new --diff-merge options to imply -p, to make it
possible to enable output of diffs for merges independently from
non-merge commits. At the same time we want to preserve behavior of
old --c/-c/-m options and their interactions with --first-parent, to
stay backward-compatible.
This patch is first step in this direction: it separates old "--cc/-c
imply -p" logic from the rest of the options.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The relevant flags were somewhat scattered over definition of 'struct
rev_info'. Rearrange them to group them together.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
'ignore_merges' was 3-way field that served two distinct purposes that
we now assign to 2 new independent flags: 'separate_merges', and
'explicit_diff_merges'.
'separate_merges' tells that we need to output diff format containing
separate diff for every parent (as opposed to 'combine_merges').
'explicit_diff_merges' tells that at least one of diff-merges options
has been explicitly specified on the command line, so no defaults
should apply.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This new field allows us to separate format of diff for merges from
'first_parent_only' flag which primary purpose is limiting history
traversal.
This change further localizes diff format selection logic into the
diff-merges.c file.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Create separate diff-merges.c and diff-merges.h files, and move all
the code related to handling of diff merges there.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use these implementations from show_setup_revisions_tweak() and
log_setup_revisions_tweak() in builtin/log.c.
This completes moving of management of diff merges parameters to a
single place, where we can finally observe them simultaneously.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
For the past 15 years, we've used the hardcoded 64 as the length
limit of the filename of the output from the "git format-patch"
command. Since the value is shorter than the 80-column terminal, it
could grow without line wrapping a bit. At the same time, since the
value is longer than half of the 80-column terminal, we could fit
two or more of them in "ls" output on such a terminal if we allowed
to lower it.
Introduce a new command line option --filename-max-length=<n> and a
new configuration variable format.filenameMaxLength to override the
hardcoded default.
While we are at it, remove a check that the name of output directory
does not exceed PATH_MAX---this check is pointless in that by the
time control reaches the function, the caller would already have
done an equivalent of "mkdir -p", so if the system does not like an
overly long directory name, the control wouldn't have reached here,
and otherwise, we know that the system allowed the output directory
to exist. In the worst case, we will get an error when we try to
open the output file and handle the error correctly anyway.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add separate 'match_missing' field for diff-index to use and set it when we
encounter "-m" option. This field won't then be cleared when another meaning of
"-m" is reverted (e.g., by "--no-diff-merges"), nor it will be affected by
future option(s) that might drive 'ignore_merges' field.
Use this new field from diff-lib:do_oneway_diff() instead of reusing
'ignore_merges' field.
Signed-off-by: Sergey Organov <sorganov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git log --first-parent -p" showed patches only for single-parent
commits on the first-parent chain; the "--first-parent" option has
been made to imply "-m". Use "--no-diff-merges" to restore the
previous behaviour to omit patches for merge commits.
* jk/log-fp-implies-m:
doc/git-log: clarify handling of merge commit diffs
doc/git-log: move "-t" into diff-options list
doc/git-log: drop "-r" diff option
doc/git-log: move "Diff Formatting" from rev-list-options
log: enable "-m" automatically with "--first-parent"
revision: add "--no-diff-merges" option to counteract "-m"
log: drop "--cc implies -m" logic
Updates to the changed-paths bloom filter.
* ds/commit-graph-bloom-updates:
commit-graph: check all leading directories in changed path Bloom filters
revision: empty pathspecs should not use Bloom filters
revision.c: fix whitespace
commit-graph: check chunk sizes after writing
commit-graph: simplify chunk writes into loop
commit-graph: unify the signatures of all write_graph_chunk_*() functions
commit-graph: persist existence of changed-paths
bloom: fix logic in get_bloom_filter()
commit-graph: change test to die on parse, not load
commit-graph: place bloom_settings in context
The "-m" option sets revs->ignore_merges to "0", but there's no way to
undo it. This probably isn't something anybody overly cares about, since
"1" is already the default, but it will serve as an escape hatch when we
flip the default for ignore_merges to "0" in more situations.
We'll also add a few extra niceties:
- initialize the value to "-1" to indicate "not set", and then resolve
it to the normal 0/1 bool in setup_revisions(). This lets any tweak
functions, as well as setup_revisions() itself, avoid clobbering the
user's preference (which until now they couldn't actually express).
- since we now have --no-diff-merges, let's add the matching
--diff-merges, which is just a synonym for "-m". Then we don't even
need to document --no-diff-merges separately; it countermands the
long form of "-m" in the usual way.
The new test shows that this behaves just the same as the current
behavior without "-m".
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The file 'dir/subdir/file' can only be modified if its leading
directories 'dir' and 'dir/subdir' are modified as well.
So when checking modified path Bloom filters looking for commits
modifying a path with multiple path components, then check not only
the full path in the Bloom filters, but all its leading directories as
well. Take care to check these paths in "deepest first" order,
because it's the full path that is least likely to be modified, and
the Bloom filter queries can short circuit sooner.
This can significantly reduce the average false positive rate, by
about an order of magnitude or three(!), and can further speed up
pathspec-limited revision walks. The table below compares the average
false positive rate and runtime of
git rev-list HEAD -- "$path"
before and after this change for 5000+ randomly* selected paths from
each repository:
Average false Average Average
positive rate runtime runtime
before after before after difference
------------------------------------------------------------------
git 3.220% 0.7853% 0.0558s 0.0387s -30.6%
linux 2.453% 0.0296% 0.1046s 0.0766s -26.8%
tensorflow 2.536% 0.6977% 0.0594s 0.0420s -29.2%
*Path selection was done with the following pipeline:
git ls-tree -r --name-only HEAD | sort -R | head -n 5000
The improvements in runtime are much smaller than the improvements in
average false positive rate, as we are clearly reaching diminishing
returns here. However, all these timings depend on that accessing
tree objects is reasonably fast (warm caches). If we had a partial
clone and the tree objects had to be fetched from a promisor remote,
e.g.:
$ git clone --filter=tree:0 --bare file://.../webkit.git webkit.notrees.git
$ git -C webkit.git -c core.modifiedPathBloomFilters=1 \
commit-graph write --reachable
$ cp webkit.git/objects/info/commit-graph webkit.notrees.git/objects/info/
$ git -C webkit.notrees.git -c core.modifiedPathBloomFilters=1 \
rev-list HEAD -- "$path"
then checking all leading path component can reduce the runtime from
over an hour to a few seconds (and this is with the clone and the
promisor on the same machine).
This adjusts the tracing values in t4216-log-bloom.sh, which provides a
concrete way to notice the improvement.
Helped-by: Taylor Blau <me@ttaylorr.com>
Helped-by: René Scharfe <l.s.r@web.de>
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The bit fields in struct object have an unfortunate layout. Here's what
pahole reports on x86_64 GNU/Linux:
struct object {
unsigned int parsed:1; /* 0: 0 4 */
unsigned int type:3; /* 0: 1 4 */
/* XXX 28 bits hole, try to pack */
/* Force alignment to the next boundary: */
unsigned int :0;
unsigned int flags:29; /* 4: 0 4 */
/* XXX 3 bits hole, try to pack */
struct object_id oid; /* 8 32 */
/* size: 40, cachelines: 1, members: 4 */
/* sum members: 32 */
/* sum bitfield members: 33 bits, bit holes: 2, sum bit holes: 31 bits */
/* last cacheline: 40 bytes */
};
Notice the 1+3+29=33 bits in bit fields and 28+3=31 bits in holes.
There are holes inside the flags bit field as well -- while some object
flags are used for more than one purpose, 22, 23 and 24 are still free.
Use 23 and 24 instead of 27 and 28 for TOPO_WALK_EXPLORED and
TOPO_WALK_INDEGREE. This allows us to reduce FLAG_BITS by one so that
all bitfields combined fit into a single 32-bit slot:
struct object {
unsigned int parsed:1; /* 0: 0 4 */
unsigned int type:3; /* 0: 1 4 */
unsigned int flags:28; /* 0: 4 4 */
struct object_id oid; /* 4 32 */
/* size: 36, cachelines: 1, members: 4 */
/* last cacheline: 36 bytes */
};
With this tight packing the size of struct object is reduced by 10%.
Other architectures probably benefit as well.
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Introduce an extension to the commit-graph to make it efficient to
check for the paths that were modified at each commit using Bloom
filters.
* gs/commit-graph-path-filter:
bloom: ignore renames when computing changed paths
commit-graph: add GIT_TEST_COMMIT_GRAPH_CHANGED_PATHS test flag
t4216: add end to end tests for git log with Bloom filters
revision.c: add trace2 stats around Bloom filter usage
revision.c: use Bloom filters to speed up path based revision walks
commit-graph: add --changed-paths option to write subcommand
commit-graph: reuse existing Bloom filters during write
commit-graph: write Bloom filters to commit graph file
commit-graph: examine commits by generation number
commit-graph: examine changed-path objects in pack order
commit-graph: compute Bloom filters for changed paths
diff: halt tree-diff early after max_changes
bloom.c: core Bloom filter implementation for changed paths.
bloom.c: introduce core Bloom filter constructs
bloom.c: add the murmur3 hash implementation
commit-graph: define and use MAX_NUM_CHUNKS
"git log" learned "--show-pulls" that helps pathspec limited
history views; a merge commit that takes the whole change from a
side branch, which is normally omitted from the output, is shown
in addition to the commits that introduce real changes.
* ds/revision-show-pulls:
revision: --show-pulls adds helpful merges
The default file history simplification of "git log -- <path>" or
"git rev-list -- <path>" focuses on providing the smallest set of
commits that first contributed a change. The revision walk greatly
restricts the set of walked commits by visiting only the first
TREESAME parent of a merge commit, when one exists. This means
that portions of the commit-graph are not walked, which can be a
performance benefit, but can also "hide" commits that added changes
but were ignored by a merge resolution.
The --full-history option modifies this by walking all commits and
reporting a merge commit as "interesting" if it has _any_ parent
that is not TREESAME. This tends to be an over-representation of
important commits, especially in an environment where most merge
commits are created by pull request completion.
Suppose we have a commit A and we create a commit B on top that
changes our file. When we merge the pull request, we create a merge
commit M. If no one else changed the file in the first-parent
history between M and A, then M will not be TREESAME to its first
parent, but will be TREESAME to B. Thus, the simplified history
will be "B". However, M will appear in the --full-history mode.
However, suppose that a number of topics T1, T2, ..., Tn were
created based on commits C1, C2, ..., Cn between A and M as
follows:
A----C1----C2--- ... ---Cn----M------P1---P2--- ... ---Pn
\ \ \ \ / / / /
\ \__.. \ \/ ..__T1 / Tn
\ \__.. /\ ..__T2 /
\_____________________B \____________________/
If the commits T1, T2, ... Tn did not change the file, then all of
P1 through Pn will be TREESAME to their first parent, but not
TREESAME to their second. This means that all of those merge commits
appear in the --full-history view, with edges that immediately
collapse into the lower history without introducing interesting
single-parent commits.
The --simplify-merges option was introduced to remove these extra
merge commits. By noticing that the rewritten parents are reachable
from their first parents, those edges can be simplified away. Finally,
the commits now look like single-parent commits that are TREESAME to
their "only" parent. Thus, they are removed and this issue does not
cause issues anymore. However, this also ends up removing the commit
M from the history view! Even worse, the --simplify-merges option
requires walking the entire history before returning a single result.
Many Git users are using Git alongside a Git service that provides
code storage alongside a code review tool commonly called "Pull
Requests" or "Merge Requests" against a target branch. When these
requests are accepted and merged, they typically create a merge
commit whose first parent is the previous branch tip and the second
parent is the tip of the topic branch used for the request. This
presents a valuable order to the parents, but also makes that merge
commit slightly special. Users may want to see not only which
commits changed a file, but which pull requests merged those commits
into their branch. In the previous example, this would mean the
users want to see the merge commit "M" in addition to the single-
parent commit "C".
Users are even more likely to want these merge commits when they
use pull requests to merge into a feature branch before merging that
feature branch into their trunk.
In some sense, users are asking for the "first" merge commit to
bring in the change to their branch. As long as the parent order is
consistent, this can be handled with the following rule:
Include a merge commit if it is not TREESAME to its first
parent, but is TREESAME to a later parent.
These merges look like the merge commits that would result from
running "git pull <topic>" on a main branch. Thus, the option to
show these commits is called "--show-pulls". This has the added
benefit of showing the commits created by closing a pull request or
merge request on any of the Git hosting and code review platforms.
To test these options, extend the standard test example to include
a merge commit that is not TREESAME to its first parent. It is
surprising that that option was not already in the example, as it
is instructive.
In particular, this extension demonstrates a common issue with file
history simplification. When a user resolves a merge conflict using
"-Xours" or otherwise ignoring one side of the conflict, they create
a TREESAME edge that probably should not be TREESAME. This leads
users to become frustrated and complain that "my change disappeared!"
In my experience, showing them history with --full-history and
--simplify-merges quickly reveals the problematic merge. As mentioned,
this option is expensive to compute. The --show-pulls option
_might_ show the merge commit (usually titled "resolving conflicts")
more quickly. Of course, this depends on the user having the correct
parent order, which is backwards when using "git pull master" from a
topic branch.
There are some special considerations when combining the --show-pulls
option with --simplify-merges. This requires adding a new PULL_MERGE
object flag to store the information from the initial TREESAME
comparisons. This helps avoid dropping those commits in later filters.
This is covered by a test, including how the parents can be simplified.
Since "struct object" has already ruined its 32-bit alignment by using
33 bits across parsed, type, and flags member, let's not make it worse.
PULL_MERGE is used in revision.c with the same value (1u<<15) as
REACHABLE in commit-graph.c. The REACHABLE flag is only used when
writing a commit-graph file, and a revision walk using --show-pulls
does not happen in the same process. Care must be taken in the future
to ensure this remains the case.
Update Documentation/rev-list-options.txt with significant details
around this option. This requires updating the example in the
History Simplification section to demonstrate some of the problems
with TREESAME second parents.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When commit subjects or authors have non-ASCII characters, git
format-patch Q-encodes them so they can be safely sent over email.
However, if the patch transfer method is something other than email (web
review tools, sneakernet), this only serves to make the patch metadata
harder to read without first applying it (unless you can decode RFC 2047
in your head). git am as well as some email software supports
non-Q-encoded mail as described in RFC 6531.
Add --[no-]encode-email-headers and format.encodeEmailHeaders to let the
user control this behavior.
Signed-off-by: Emma Brooks <me@pluvano.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Revision walk will now use Bloom filters for commits to speed up
revision walks for a particular path (for computing history for
that path), if they are present in the commit-graph file.
We load the Bloom filters during the prepare_revision_walk step,
currently only when dealing with a single pathspec. Extending
it to work with multiple pathspecs can be explored and built on
top of this series in the future.
While comparing trees in rev_compare_trees(), if the Bloom filter
says that the file is not different between the two trees, we don't
need to compute the expensive diff. This is where we get our
performance gains. The other response of the Bloom filter is '`:maybe',
in which case we fall back to the full diff calculation to determine
if the path was changed in the commit.
We do not try to use Bloom filters when the '--walk-reflogs' option
is specified. The '--walk-reflogs' option does not walk the commit
ancestry chain like the rest of the options. Incorporating the
performance gains when walking reflog entries would add more
complexity, and can be explored in a later series.
Performance Gains:
We tested the performance of `git log -- <path>` on the git repo, the linux
and some internal large repos, with a variety of paths of varying depths.
On the git and linux repos:
- we observed a 2x to 5x speed up.
On a large internal repo with files seated 6-10 levels deep in the tree:
- we observed 10x to 20x speed ups, with some paths going up to 28 times
faster.
Helped-by: Derrick Stolee <dstolee@microsoft.com
Helped-by: SZEDER Gábor <szeder.dev@gmail.com>
Helped-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Garima Singh <garima.singh@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git format-patch" can take a set of configured format.notes values
to specify which notes refs to use in the log message part of the
output. The behaviour of this was not consistent with multiple
--notes command line options, which has been corrected.
* dl/format-patch-notes-config-fixup:
notes.h: fix typos in comment
notes: break set_display_notes() into smaller functions
config/format.txt: clarify behavior of multiple format.notes
format-patch: move git_config() before repo_init_revisions()
format-patch: use --notes behavior for format.notes
notes: extract logic into set_display_notes()
notes: create init_display_notes() helper
notes: rename to load_display_notes()
* hw/doc-in-header:
trace2: move doc to trace2.h
submodule-config: move doc to submodule-config.h
tree-walk: move doc to tree-walk.h
trace: move doc to trace.h
run-command: move doc to run-command.h
parse-options: add link to doc file in parse-options.h
credential: move doc to credential.h
argv-array: move doc to argv-array.h
cache: move doc to cache.h
sigchain: move doc to sigchain.h
pathspec: move doc to pathspec.h
revision: move doc to revision.h
attr: move doc to attr.h
refs: move doc to refs.h
remote: move doc to remote.h and refspec.h
sha1-array: move doc to sha1-array.h
merge: move doc to ll-merge.h
graph: move doc to graph.h and graph.c
dir: move doc to dir.h
diff: move doc to diff.h and diffcore.h
In 8164c961e1 (format-patch: use --notes behavior for format.notes,
2019-12-09), we introduced set_display_notes() which was a monolithic
function with three mutually exclusive branches. Break the function up
into three small and simple functions that each are only responsible for
one task.
This family of functions accepts an `int *show_notes` instead of
returning a value suitable for assignment to `show_notes`. This is for
two reasons. First of all, this guarantees that the external
`show_notes` variable changes in lockstep with the
`struct display_notes_opt`. Second, this prompts future developers to be
careful about doing something meaningful with this value. In fact, a
NULL check is intentionally omitted because causing a segfault here
would tell the future developer that they are meant to use the value for
something meaningful.
One alternative was making the family of functions accept a
`struct rev_info *` instead of the `struct display_notes_opt *`, since
the former contains the `show_notes` field as well. This does not work
because we have to call git_config() before repo_init_revisions().
However, if we had a `struct rev_info`, we'd need to initialize it before
it gets assigned values from git_config(). As a result, we break the
circular dependency by having standalone `int show_notes` and
`struct display_notes_opt notes_opt` variables which temporarily hold
values from git_config() until the values are copied over to `rev`.
To implement this change, we need to get a pointer to
`rev_info::show_notes`. Unfortunately, this is not possible with
bitfields and only direct-assignment is possible. Change
`rev_info::show_notes` to a non-bitfield int so that we can get its
address.
Signed-off-by: Denton Liu <liu.denton@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
get_revision_mark() used to return a `char *`, even though all of the
strings it was returning were string literals. Make get_revision_mark()
return a `const char *` so that callers won't be tempted to modify the
returned string.
Signed-off-by: Denton Liu <liu.denton@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the documentation from Documentation/technical/api-revision-walking.txt
to revision.h as it's easier for the developers to find the usage
information beside the code instead of looking for it in another doc file.
Also documentation/technical/api-revision-walking.txt is removed because the
information it has is now redundant and it'll be hard to keep it up to
date and synchronized with the documentation in the header file.
Signed-off-by: Heba Waly <heba.waly@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Output from "diff --cc" did not show the original paths when the
merge involved renames. A new option adds the paths in the
original trees to the output.
* en/combined-all-paths:
log,diff-tree: add --combined-all-paths option
The combined diff format for merges will only list one filename, even if
rename or copy detection is active. For example, with raw format one
might see:
::100644 100644 100644 fabadb8cc95eb04866510 MM describe.c
::100755 100755 100755 52b7a2d 6d1ac04 d2ac7d7 RM bar.sh
::100644 100644 100644 e07d6c5 9042e82 ee91881 RR phooey.c
This doesn't let us know what the original name of bar.sh was in the
first parent, and doesn't let us know what either of the original names
of phooey.c were in either of the parents. In contrast, for non-merge
commits, raw format does provide original filenames (and a rename score
to boot). In order to also provide original filenames for merge
commits, add a --combined-all-paths option (which must be used with
either -c or --cc, and is likely only useful with rename or copy
detection active) so that we can print tab-separated filenames when
renames are involved. This transforms the above output to:
::100644 100644 100644 fabadb8cc95eb04866510 MM desc.c desc.c desc.c
::100755 100755 100755 52b7a2d 6d1ac04 d2ac7d7 RM foo.sh bar.sh bar.sh
::100644 100644 100644 e07d6c5 9042e82 ee91881 RR fooey.c fuey.c phooey.c
Further, in patch format, this changes the from/to headers so that
instead of just having one "from" header, we get one for each parent.
For example, instead of having
--- a/phooey.c
+++ b/phooey.c
we would see
--- a/fooey.c
--- a/fuey.c
+++ b/phooey.c
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git pack-objects" learned another algorithm to compute the set of
objects to send, that trades the resulting packfile off to save
traversal cost to favor small pushes.
* ds/push-sparse-tree-walk:
pack-objects: create GIT_TEST_PACK_SPARSE
pack-objects: create pack.useSparse setting
revision: implement sparse algorithm
list-objects: consume sparse tree walk
revision: add mark_tree_uninteresting_sparse
In preparation for a new algorithm that walks fewer trees when
creating a pack from a set of revisions, create a method that
takes an oidset of tree oids and marks reachable objects as
UNINTERESTING.
The current implementation uses the existing
mark_tree_uninteresting to recursively walk the trees and blobs.
This will walk the same number of trees as the old mechanism. To
ensure that mark_tree_uninteresting walks the tree, we need to
remove the UNINTERESTING flag before calling the method. This
implementation will be replaced entirely in a later commit.
There is one new assumption in this approach: we are also given
the oids of the interesting trees. This implementation does not
use those trees at the moment, but we will use them in a later
rewrite of this method.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Put the allow_exclude_promisor_objects flag in setup_revision_opt. When
it was in rev_info, it was unclear when it was used, since rev_info is
passed to functions that don't use the flag. This resulted in
unnecessary setting of the flag in prune.c, so fix that as well.
Signed-off-by: Matthew DeVore <matvore@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The revision walker machinery learned to take advantage of the
commit generation numbers stored in the commit-graph file.
* ds/reachable-topo-order:
t6012: make rev-list tests more interesting
revision.c: generation-based topo-order algorithm
commit/revisions: bookkeeping before refactoring
revision.c: begin refactoring --topo-order logic
test-reach: add rev-list tests
test-reach: add run_three_modes method
prio-queue: add 'peek' operation
Operations on promisor objects make sense in the context of only a
small subset of the commands that internally use the revisions
machinery, but the "--exclude-promisor-objects" option were taken
and led to nonsense results by commands like "log", to which it
didn't make much sense. This has been corrected.
* md/exclude-promisor-objects-fix:
exclude-promisor-objects: declare when option is allowed
Documentation/git-log.txt: do not show --exclude-promisor-objects
The current --topo-order algorithm requires walking all
reachable commits up front, topo-sorting them, all before
outputting the first value. This patch introduces a new
algorithm which uses stored generation numbers to
incrementally walk in topo-order, outputting commits as
we go. This can dramatically reduce the computation time
to write a fixed number of commits, such as when limiting
with "-n <N>" or filling the first page of a pager.
When running a command like 'git rev-list --topo-order HEAD',
Git performed the following steps:
1. Run limit_list(), which parses all reachable commits,
adds them to a linked list, and distributes UNINTERESTING
flags. If all unprocessed commits are UNINTERESTING, then
it may terminate without walking all reachable commits.
This does not occur if we do not specify UNINTERESTING
commits.
2. Run sort_in_topological_order(), which is an implementation
of Kahn's algorithm. It first iterates through the entire
set of important commits and computes the in-degree of each
(plus one, as we use 'zero' as a special value here). Then,
we walk the commits in priority order, adding them to the
priority queue if and only if their in-degree is one. As
we remove commits from this priority queue, we decrement the
in-degree of their parents.
3. While we are peeling commits for output, get_revision_1()
uses pop_commit on the full list of commits computed by
sort_in_topological_order().
In the new algorithm, these three steps correspond to three
different commit walks. We run these walks simultaneously,
and advance each only as far as necessary to satisfy the
requirements of the 'higher order' walk. We know when we can
pause each walk by using generation numbers from the commit-
graph feature.
Recall that the generation number of a commit satisfies:
* If the commit has at least one parent, then the generation
number is one more than the maximum generation number among
its parents.
* If the commit has no parent, then the generation number is one.
There are two special generation numbers:
* GENERATION_NUMBER_INFINITY: this value is 0xffffffff and
indicates that the commit is not stored in the commit-graph and
the generation number was not previously calculated.
* GENERATION_NUMBER_ZERO: this value (0) is a special indicator
to say that the commit-graph was generated by a version of Git
that does not compute generation numbers (such as v2.18.0).
Since we use generation_numbers_enabled() before using the new
algorithm, we do not need to worry about GENERATION_NUMBER_ZERO.
However, the existence of GENERATION_NUMBER_INFINITY implies the
following weaker statement than the usual we expect from
generation numbers:
If A and B are commits with generation numbers gen(A) and
gen(B) and gen(A) < gen(B), then A cannot reach B.
Thus, we will walk in each of our stages until the "maximum
unexpanded generation number" is strictly lower than the
generation number of a commit we are about to use.
The walks are as follows:
1. EXPLORE: using the explore_queue priority queue (ordered by
maximizing the generation number), parse each reachable
commit until all commits in the queue have generation
number strictly lower than needed. During this walk, update
the UNINTERESTING flags as necessary.
2. INDEGREE: using the indegree_queue priority queue (ordered
by maximizing the generation number), add one to the in-
degree of each parent for each commit that is walked. Since
we walk in order of decreasing generation number, we know
that discovering an in-degree value of 0 means the value for
that commit was not initialized, so should be initialized to
two. (Recall that in-degree value "1" is what we use to say a
commit is ready for output.) As we iterate the parents of a
commit during this walk, ensure the EXPLORE walk has walked
beyond their generation numbers.
3. TOPO: using the topo_queue priority queue (ordered based on
the sort_order given, which could be commit-date, author-
date, or typical topo-order which treats the queue as a LIFO
stack), remove a commit from the queue and decrement the
in-degree of each parent. If a parent has an in-degree of
one, then we add it to the topo_queue. Before we decrement
the in-degree, however, ensure the INDEGREE walk has walked
beyond that generation number.
The implementations of these walks are in the following methods:
* explore_walk_step and explore_to_depth
* indegree_walk_step and compute_indegrees_to_depth
* next_topo_commit and expand_topo_walk
These methods have some patterns that may seem strange at first,
but they are probably carry-overs from their equivalents in
limit_list and sort_in_topological_order.
One thing that is missing from this implementation is a proper
way to stop walking when the entire queue is UNINTERESTING, so
this implementation is not enabled by comparisions, such as in
'git rev-list --topo-order A..B'. This can be updated in the
future.
In my local testing, I used the following Git commands on the
Linux repository in three modes: HEAD~1 with no commit-graph,
HEAD~1 with a commit-graph, and HEAD with a commit-graph. This
allows comparing the benefits we get from parsing commits from
the commit-graph and then again the benefits we get by
restricting the set of commits we walk.
Test: git rev-list --topo-order -100 HEAD
HEAD~1, no commit-graph: 6.80 s
HEAD~1, w/ commit-graph: 0.77 s
HEAD, w/ commit-graph: 0.02 s
Test: git rev-list --topo-order -100 HEAD -- tools
HEAD~1, no commit-graph: 9.63 s
HEAD~1, w/ commit-graph: 6.06 s
HEAD, w/ commit-graph: 0.06 s
This speedup is due to a few things. First, the new generation-
number-enabled algorithm walks commits on order of the number of
results output (subject to some branching structure expectations).
Since we limit to 100 results, we are running a query similar to
filling a single page of results. Second, when specifying a path,
we must parse the root tree object for each commit we walk. The
previous benefits from the commit-graph are entirely from reading
the commit-graph instead of parsing commits. Since we need to
parse trees for the same number of commits as before, we slow
down significantly from the non-path-based query.
For the test above, I specifically selected a path that is changed
frequently, including by merge commits. A less-frequently-changed
path (such as 'README') has similar end-to-end time since we need
to walk the same number of commits (before determining we do not
have 100 hits). However, get the benefit that the output is
presented to the user as it is discovered, much the same as a
normal 'git log' command (no '--topo-order'). This is an improved
user experience, even if the command has the same runtime.
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When running 'git rev-list --topo-order' and its kin, the topo_order
setting in struct rev_info implies the limited setting. This means
that the following things happen during prepare_revision_walk():
* revs->limited implies we run limit_list() to walk the entire
reachable set. There are some short-cuts here, such as if we
perform a range query like 'git rev-list COMPARE..HEAD' and we
can stop limit_list() when all queued commits are uninteresting.
* revs->topo_order implies we run sort_in_topological_order(). See
the implementation of that method in commit.c. It implies that
the full set of commits to order is in the given commit_list.
These two methods imply that a 'git rev-list --topo-order HEAD'
command must walk the entire reachable set of commits _twice_ before
returning a single result.
If we have a commit-graph file with generation numbers computed, then
there is a better way. This patch introduces some necessary logic
redirection when we are in this situation.
In v2.18.0, the commit-graph file contains zero-valued bytes in the
positions where the generation number is stored in v2.19.0 and later.
Thus, we use generation_numbers_enabled() to check if the commit-graph
is available and has non-zero generation numbers.
When setting revs->limited only because revs->topo_order is true,
only do so if generation numbers are not available. There is no
reason to use the new logic as it will behave similarly when all
generation numbers are INFINITY or ZERO.
In prepare_revision_walk(), if we have revs->topo_order but not
revs->limited, then we trigger the new logic. It breaks the logic
into three pieces, to fit with the existing framework:
1. init_topo_walk() fills a new struct topo_walk_info in the rev_info
struct. We use the presence of this struct as a signal to use the
new methods during our walk. In this patch, this method simply
calls limit_list() and sort_in_topological_order(). In the future,
this method will set up a new data structure to perform that logic
in-line.
2. next_topo_commit() provides get_revision_1() with the next topo-
ordered commit in the list. Currently, this simply pops the commit
from revs->commits.
3. expand_topo_walk() provides get_revision_1() with a way to signal
walking beyond the latest commit. Currently, this calls
add_parents_to_list() exactly like the old logic.
While this commit presents method redirection for performing the
exact same logic as before, it allows the next commit to focus only
on the new logic.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "rev-list --filter" feature learned to exclude all trees via
"tree:0" filter.
* md/filter-trees:
list-objects: support for skipping tree traversal
filter-trees: code clean-up of tests
list-objects-filter: implement filter tree:0
list-objects-filter-options: do not over-strbuf_init
list-objects-filter: use BUG rather than die
revision: mark non-user-given objects instead
rev-list: handle missing tree objects properly
list-objects: always parse trees gently
list-objects: refactor to process_tree_contents
list-objects: store common func args in struct
The --exclude-promisor-objects option causes some funny behavior in at
least two commands: log and blame. It causes a BUG crash:
$ git log --exclude-promisor-objects
BUG: revision.c:2143: exclude_promisor_objects can only be used
when fetch_if_missing is 0
Aborted
[134]
Fix this such that the option is treated like any other unknown option.
The commands that must support it are limited, so declare in those
commands that the flag is supported. In particular:
pack-objects
prune
rev-list
The commands were found by searching for logic which parses
--exclude-promisor-objects outside of revision.c. Extra logic outside of
revision.c is needed because fetch_if_missing must be turned on before
revision.c sees the option or it will BUG-crash. The above list is
supported by the fact that no other command is introspectively invoked
by another command passing --exclude-promisor-object.
Signed-off-by: Matthew DeVore <matvore@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Various codepaths in the core-ish part learn to work on an
arbitrary in-core index structure, not necessarily the default
instance "the_index".
* nd/the-index: (23 commits)
revision.c: reduce implicit dependency the_repository
revision.c: remove implicit dependency on the_index
ws.c: remove implicit dependency on the_index
tree-diff.c: remove implicit dependency on the_index
submodule.c: remove implicit dependency on the_index
line-range.c: remove implicit dependency on the_index
userdiff.c: remove implicit dependency on the_index
rerere.c: remove implicit dependency on the_index
sha1-file.c: remove implicit dependency on the_index
patch-ids.c: remove implicit dependency on the_index
merge.c: remove implicit dependency on the_index
merge-blobs.c: remove implicit dependency on the_index
ll-merge.c: remove implicit dependency on the_index
diff-lib.c: remove implicit dependency on the_index
read-cache.c: remove implicit dependency on the_index
diff.c: remove implicit dependency on the_index
grep.c: remove implicit dependency on the_index
diff.c: remove the_index dependency in textconv() functions
blame.c: rename "repo" argument to "r"
combine-diff.c: remove implicit dependency on the_index
...
Currently, list-objects.c incorrectly treats all root trees of commits
as USER_GIVEN. Also, it would be easier to mark objects that are
non-user-given instead of user-given, since the places in the code
where we access an object through a reference are more obvious than
the places where we access an object that was given by the user.
Resolve these two problems by introducing a flag NOT_USER_GIVEN that
marks blobs and trees that are non-user-given, replacing USER_GIVEN.
(Only blobs and trees are marked because this mark is only used when
filtering objects, and filtering of other types of objects is not
supported yet.)
This fixes a bug in that git rev-list behaved differently from git
pack-objects. pack-objects would *not* filter objects given explicitly
on the command line and rev-list would filter. This was because the two
commands used a different function to add objects to the rev_info
struct. This seems to have been an oversight, and pack-objects has the
correct behavior, so I added a test to make sure that rev-list now
behaves properly.
Signed-off-by: Matthew DeVore <matvore@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Previously, we assumed only blob objects could be missing. This patch
makes rev-list handle missing trees like missing blobs. The --missing=*
and --exclude-promisor-objects flags now work for trees as they already
do for blobs. This is demonstrated in t6112.
Signed-off-by: Matthew DeVore <matvore@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git format-patch" learned a new "--range-diff" option to explain
the difference between this version and the previous attempt in
the cover letter (or after the tree-dashes as a comment).
* es/format-patch-rangediff:
format-patch: allow --range-diff to apply to a lone-patch
format-patch: add --creation-factor tweak for --range-diff
format-patch: teach --range-diff to respect -v/--reroll-count
format-patch: extend --range-diff to accept revision range
format-patch: add --range-diff option to embed diff in cover letter
range-diff: relieve callers of low-level configuration burden
range-diff: publish default creation factor
range-diff: respect diff_option.file rather than assuming 'stdout'
"git format-patch" learned a new "--interdiff" option to explain
the difference between this version and the previous atttempt in
the cover letter (or after the tree-dashes as a comment).
* es/format-patch-interdiff:
format-patch: allow --interdiff to apply to a lone-patch
log-tree: show_log: make commentary block delimiting reusable
interdiff: teach show_interdiff() to indent interdiff
format-patch: teach --interdiff to respect -v/--reroll-count
format-patch: add --interdiff option to embed diff in cover letter
format-patch: allow additional generated content in make_cover_letter()
"git rev-list --stdin </dev/null" used to be an error; it now shows
no output without an error. "git rev-list --stdin --default HEAD"
still falls back to the given default when nothing is given on the
standard input.
* jk/rev-list-stdin-noop-is-ok:
rev-list: make empty --stdin not an error
When we originally did the series that contains 7ba826290a
(revision: add rev_input_given flag, 2017-08-02) the intent
was that "git rev-list --stdin </dev/null" would similarly
become a successful noop. However, an attempt at the time to
do that did not work[1]. The problem is that rev_input_given
serves two roles:
- it tells rev-list.c that it should not error out
- it tells revision.c that it should not have the "default"
ref kick (e.g., "HEAD" in "git log")
We want to trigger the former, but not the latter. This is
technically possible with a single flag, if we set the flag
only after revision.c's revs->def check. But this introduces
a rather subtle ordering dependency.
Instead, let's keep two flags: one to denote when we got
actual input (which triggers both roles) and one for when we
read stdin (which triggers only the first).
This does mean a caller interested in the first role has to
check both flags, but there's only one such caller. And any
future callers might want to make the distinction anyway
(e.g., if they care less about erroring out, and more about
whether revision.c soaked up our stdin).
In fact, we already keep such a flag internally in
revision.c for this purpose, so this is really just exposing
that to the caller (and the old function-local flag can go
away in favor of our new one).
[1] https://public-inbox.org/git/20170802223416.gwiezhbuxbdmbjzx@sigill.intra.peff.net/
Helped-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code hygiene improvement for the header files.
* en/incl-forward-decl:
Remove forward declaration of an enum
compat/precompose_utf8.h: use more common include guard style
urlmatch.h: fix include guard
Move definition of enum branch_track from cache.h to branch.h
alloc: make allocate_alloc_state and clear_alloc_state more consistent
Add missing includes and forward declarations
I looped over the toplevel header files, creating a temporary two-line C
program for each consisting of
#include "git-compat-util.h"
#include $HEADER
This patch is the result of manually fixing errors in compiling those
tiny programs.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The --range-diff option announces the embedded range-diff generically
as "Range-diff:", however, we can do better when --reroll-count is
specified by emitting "Range-diff against v{n}:" instead.
Signed-off-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When submitting a revised version of a patch series, it can be helpful
(to reviewers) to include a summary of changes since the previous
attempt in the form of a range-diff, however, doing so involves manually
copy/pasting the diff into the cover letter.
Add a --range-diff option to automate this process. The argument to
--range-diff specifies the tip of the previous attempt against which to
generate the range-diff. For example:
git format-patch --cover-letter --range-diff=v1 -3 v2
(At this stage, the previous attempt and the patch series being
formatted must share a common base, however, a subsequent enhancement
will make it possible to specify an explicit revision range for the
previous attempt.)
Signed-off-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
