Merge branch 'tb/blame-resurrect-convert-to-git' into maint
Some time ago, "git blame" (incorrectly) lost the convert_to_git()
call when synthesizing a fake "tip" commit that represents the
state in the working tree, which broke folks who record the history
with LF line ending to make their project portabile across
platforms while terminating lines in their working tree files with
CRLF for their platform.
* tb/blame-resurrect-convert-to-git:
blame: CRLF in the working tree and LF in the repo
* pt/xdg-config-path:
path.c: remove home_config_paths()
git-config: replace use of home_config_paths()
git-commit: replace use of home_config_paths()
credential-store.c: replace home_config_paths() with xdg_config_home()
dir.c: replace home_config_paths() with xdg_config_home()
attr.c: replace home_config_paths() with xdg_config_home()
path.c: implement xdg_config_home()
t0302: "unreadable" test needs POSIXPERM
t0302: test credential-store support for XDG_CONFIG_HOME
git-credential-store: support XDG_CONFIG_HOME
git-credential-store: support multiple credential files
"git rev-list --objects $old --not --all" to see if everything that
is reachable from $old is already connected to the existing refs
was very inefficient.
* jk/still-interesting:
limit_list: avoid quadratic behavior from still_interesting
"hash-object --literally" introduced in v2.2 was not prepared to
take a really long object type name.
* jc/hash-object:
write_sha1_file(): do not use a separate sha1[] array
t1007: add hash-object --literally tests
hash-object --literally: fix buffer overrun with extra-long object type
git-hash-object.txt: document --literally option
Merge branch 'jk/filter-branch-use-of-sed-on-incomplete-line' into maint
"filter-branch" corrupted commit log message that ends with an
incomplete line on platforms with some "sed" implementations that
munge such a line. Work it around by avoiding to use "sed".
* jk/filter-branch-use-of-sed-on-incomplete-line:
filter-branch: avoid passing commit message through sed
Merge branch 'jk/stash-require-clean-index' into maint
"git stash pop/apply" forgot to make sure that not just the working
tree is clean but also the index is clean. The latter is important
as a stash application can conflict and the index will be used for
conflict resolution.
* jk/stash-require-clean-index:
stash: require a clean index to apply
t3903: avoid applying onto dirty index
t3903: stop hard-coding commit sha1s
Merge branch 'jk/git-no-more-argv0-path-munging' into maint
We have prepended $GIT_EXEC_PATH and the path "git" is installed in
(typically "/usr/bin") to $PATH when invoking subprograms and hooks
for almost eternity, but the original use case the latter tried to
support was semi-bogus (i.e. install git to /opt/foo/git and run it
without having /opt/foo on $PATH), and more importantly it has
become less and less relevant as Git grew more mainstream (i.e. the
users would _want_ to have it on their $PATH). Stop prepending the
path in which "git" is installed to users' $PATH, as that would
interfere the command search order people depend on (e.g. they may
not like versions of programs that are unrelated to Git in /usr/bin
and want to override them by having different ones in /usr/local/bin
and have the latter directory earlier in their $PATH).
* jk/git-no-more-argv0-path-munging:
stop putting argv[0] dirname at front of PATH
Teach the codepaths that read .gitignore and .gitattributes files
that these files encoded in UTF-8 may have UTF-8 BOM marker at the
beginning; this makes it in line with what we do for configuration
files already.
* cn/bom-in-gitignore:
attr: skip UTF8 BOM at the beginning of the input file
config: use utf8_bom[] from utf.[ch] in git_parse_source()
utf8-bom: introduce skip_utf8_bom() helper
add_excludes_from_file: clarify the bom skipping logic
dir: allow a BOM at the beginning of exclude files
Access to objects in repositories that borrow from another one on a
slow NFS server unnecessarily got more expensive due to recent code
becoming more cautious in a naive way not to lose objects to pruning.
* jk/prune-mtime:
sha1_file: only freshen packs once per run
sha1_file: freshen pack objects before loose
reachable: only mark local objects as recent
Merge branch 'jk/init-core-worktree-at-root' into maint
We avoid setting core.worktree when the repository location is the
".git" directory directly at the top level of the working tree, but
the code misdetected the case in which the working tree is at the
root level of the filesystem (which arguably is a silly thing to
do, but still valid).
* jk/init-core-worktree-at-root:
init: don't set core.worktree when initializing /.git
Merge branch 'jc/diff-no-index-d-f' into maint-2.3
The usual "git diff" when seeing a file turning into a directory
showed a patchset to remove the file and create all files in the
directory, but "git diff --no-index" simply refused to work. Also,
when asked to compare a file and a directory, imitate POSIX "diff"
and compare the file with the file with the same name in the
directory, instead of refusing to run.
* jc/diff-no-index-d-f:
diff-no-index: align D/F handling with that of normal Git
diff-no-index: DWIM "diff D F" into "diff D/F F"
home_config_paths() combines distinct functionality already implemented
by expand_user_path() and xdg_config_home(), and it also hard-codes the
path ~/.gitconfig, which makes it unsuitable to use for other home
config file paths. Since its use will just add unnecessary complexity to
the code, remove it.
Signed-off-by: Paul Tan <pyokagan@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since home_config_paths() combines distinct functionality already
implemented by expand_user_path() and xdg_config_home(), and hides the
home config file path ~/.gitconfig. Make the code more explicit by
replacing the use of home_config_paths() with expand_user_path() and
xdg_config_home().
Signed-off-by: Paul Tan <pyokagan@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since home_config_paths() combines two distinct functionality already
implemented by expand_user_path() and xdg_config_home(), and hides the
home config file path ~/.gitconfig. Make the code more explicit by
replacing the use of home_config_paths() with expand_user_path() and
xdg_config_home().
Signed-off-by: Paul Tan <pyokagan@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
credential-store.c: replace home_config_paths() with xdg_config_home()
Since only the xdg credentials file path is required, and
home_config_paths() is unable to construct the path ~/.git-credentials,
simplify the code by replacing home_config_paths() with
xdg_config_home().
Signed-off-by: Paul Tan <pyokagan@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The XDG base dir spec[1] specifies that configuration files be stored in
a subdirectory in $XDG_CONFIG_HOME. To construct such a configuration
file path, home_config_paths() can be used. However, home_config_paths()
combines distinct functionality:
1. Retrieve the home git config file path ~/.gitconfig
2. Construct the XDG config path of the file specified by `file`.
This function was introduced in commit 21cf3227 ("read (but not write)
from $XDG_CONFIG_HOME/git/config file"). While the intention of the
function was to allow the home directory configuration file path and the
xdg directory configuration file path to be retrieved with one function
call, the hard-coding of the path ~/.gitconfig prevents it from being
used for other configuration files. Furthermore, retrieving a file path
relative to the user's home directory can be done with
expand_user_path(). Hence, it can be seen that home_config_paths()
introduces unnecessary complexity, especially if a user just wants to
retrieve the xdg config file path.
As such, implement a simpler function xdg_config_home() for constructing
the XDG base dir spec configuration file path. This function, together
with expand_user_path(), can replace all uses of home_config_paths().
When 01cec54e (daemon: deglobalize hostname information, 2015-03-07)
wrapped the global variables such as hostname inside a struct, it
forgot to convert one location that spelled "hostname" that needs to
be updated to "hi->hostname".
This was inside NO_IPV6 block, and was not caught by anybody.
write_sha1_file(): do not use a separate sha1[] array
In the beginning, write_sha1_file() did not have a way to tell the
caller the name of the object it wrote to the caller. This was
changed in d6d3f9d0 (This implements the new "recursive tree"
write-tree., 2005-04-09) by adding the "returnsha1" parameter to the
function so that the callers who are interested in the value can
optionally pass a pointer to receive it.
It turns out that all callers do want to know the name of the object
it just has written. Nobody passes a NULL to this parameter, hence
it is not necessary to use a separate sha1[] array to receive the
result from write_sha1_file_prepare(), and copy the result to the
returnsha1 supplied by the caller.
git-hash-object learned a --literally option in 5ba9a93
(hash-object: add --literally option, 2014-09-11). Check that
--literally allows object creation with a bogus type, with two
type strings whose length is reasonably short and very long.
Signed-off-by: Eric Sunshine <sunshine@sunshineco.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
hash-object --literally: fix buffer overrun with extra-long object type
"hash-object" learned in 5ba9a93 (hash-object: add --literally
option, 2014-09-11) to allow crafting a corrupt/broken object of
unknown type.
When the user-provided type is particularly long, however, it can
overflow the relatively small stack-based character array handed to
write_sha1_file_prepare() by hash_sha1_file() and write_sha1_file(),
leading to stack corruption (and crash). Introduce a custom helper
to allow arbitrarily long typenames just for "hash-object --literally".
[jc: Eric's original used a strbuf in the more common codepaths, and
I rewrote it to avoid penalizing the non-literally code. Bugs are mine]
Signed-off-by: Eric Sunshine <sunshine@sunshineco.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
completion: fix and update 'git log --decorate=' options
'git log --decorate=' understands the 'full', 'short' and 'no' options.
From these the completion script only offered 'short' and it offered
'long' instead of 'full'.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
blame: CRLF in the working tree and LF in the repo
A typical setup under Windows is to set core.eol to CRLF, and text
files are marked as "text" in .gitattributes, or core.autocrlf is
set to true.
After 4d4813a5 "git blame" no longer works as expected for such a
set-up. Every line is annotated as "Not Committed Yet", even though
the working directory is clean. This is because the commit removed
the conversion in blame.c for all files, with or without CRLF in the
repo.
Having files with CRLF in the repo and core.autocrlf=input is a
temporary situation, and the files, if committed as is, will be
normalized in the repo, which _will_ be a notable change. Blaming
them with "Not Committed Yet" is the right result. Revert commit 4d4813a5 which was a misguided attempt to "solve" a non-problem.
Add two test cases in t8003 to verify the correct CRLF conversion.
Suggested-By: Stepan Kasal <kasal@ucw.cz> Signed-off-by: Torsten Bögershausen <tboegi@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
filter-branch: avoid passing commit message through sed
On some systems (like OS X), if sed encounters input without
a trailing newline, it will silently add it. As a result,
"git filter-branch" on such systems may silently rewrite
commit messages that omit a trailing newline. Even though
this is not something we generate ourselves with "git
commit", it's better for filter-branch to preserve the
original data as closely as possible.
We're using sed here only to strip the header fields from
the commit object. We can accomplish the same thing with a
shell loop. Since shell "read" calls are slow (usually one
syscall per byte), we use "cat" once we've skipped past the
header. Depending on the size of your commit messages, this
is probably faster (you pay the cost to fork, but then read
the data in saner-sized chunks). This idea is shamelessly
stolen from Junio.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When the branch to be rebased is already up to date, we
"git checkout" the branch, print an "up to date" message,
and end the rebase early. However, our checkout may print
"Switched to branch 'foo'" or "Already on 'foo'", even if
the user has asked for "--quiet".
We should avoid printing these messages at all, "--quiet" or
no. Since the rebase is a noop, this checkout can be seen as
optimizing out these other two checkout operations (that
happen in a real rebase):
1. Moving to the detached HEAD to start the rebase; we
always feed "-q" to checkout there, and instead rely on
our own custom message (which respects --quiet).
2. Finishing a rebase, where we move to the final branch.
Here we actually use update-ref rather than
git-checkout, and produce no messages.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Merge branch 'tb/connect-ipv6-parse-fix' into maint
An earlier update to the parser that disects a URL broke an
address, followed by a colon, followed by an empty string (instead
of the port number), e.g. ssh://example.com:/path/to/repo.
* tb/connect-ipv6-parse-fix:
connect.c: ignore extra colon after hostname
The "git push --signed" protocol extension did not limit what the
"nonce" that is a server-chosen string can contain or how long it
can be, which was unnecessarily lax. Limit both the length and the
alphabet to a reasonably small space that can still have enough
entropy.
* jc/push-cert:
push --signed: tighten what the receiving end can ask to sign
When the "git" wrapper is invoked, we prepend the baked-in
exec-path to our PATH, so that any sub-processes we exec
will all find the git-foo commands that match the wrapper
version.
If you invoke git with an absolute path, like:
/usr/bin/git foo
we also prepend "/usr/bin" to the PATH. This was added long
ago by by 231af83 (Teach the "git" command to handle some
commands internally, 2006-02-26), with the intent that
things would just work if you did something like:
cd /opt
tar xzf premade-git-package.tar.gz
alias git=/opt/git/bin/git
as we would then find all of the related external commands
in /opt/git/bin. I.e., it made git runtime-relocatable,
since at the time of 231af83, we installed all of the git
commands into $(bindir). But these days, that is not enough.
Since f28ac70 (Move all dashed-form commands to libexecdir,
2007-11-28), we do not put commands into $(bindir), and you
actually need to convert "/usr/bin" into "/usr/libexec". And
not just for finding binaries; we want to find $(sharedir),
etc, the same way. The RUNTIME_PREFIX build knob does this
the right way, by assuming a sane hierarchy rooted at
"$prefix" when we run "$prefix/bin/git", and inferring
"$prefix/libexec/git-core", etc.
So this feature (prepending the argv[0] dirname to the PATH)
is broken for providing a runtime prefix, and has been for
many years. Does it do anything for other cases?
For the "git" wrapper itself, as well as any commands
shipped by "git", the answer is no. Those are already in
git's exec-path, which is consulted first. For third-party
commands which you've dropped into the same directory, it
does include them. So if you do
cd /opt
tar xzf git-built-specifically-for-opt-git.tar.gz
cp third-party/git-foo /opt/git/bin/git-foo
alias git=/opt/git/bin/git
it does mean that we will find the third-party "git-foo",
even if you do not put /opt/git/bin into your $PATH. But
the flipside of this is that we will bump the precedence of
_other_ third-party tools that happen to be in the same
directory as git. For example, consider this setup:
1. Git is installed by the system in /usr/bin. There are
other system utilities in /usr/bin. E.g., a system
"vi".
2. The user installs tools they prefer in /usr/local/bin.
E.g., vim with a "vi" symlink. They set their PATH to
/usr/local/bin:/usr/bin to prefer their custom tools.
3. Running /usr/bin/git puts "/usr/bin" at the front of
their PATH. When git invokes the editor on behalf of
the user, they get the system vi, not their normal vim.
There are other variants of this, including overriding
system ruby and python (which is quite common using tools
like "rvm" and "virtualenv", which use relocatable
hierarchies and $PATH settings to get a consistent
environment).
Given that the main motivation for git placing the argv[0]
dirname into the PATH has been broken for years, that the
remaining cases are obscure and unlikely (and easily fixed
by the user just setting up their $PATH sanely), and that
the behavior is hurting real, reasonably common use cases,
it's not worth continuing to do so.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
If you have staged contents in your index and run "stash
apply", we may hit a conflict and put new entries into the
index. Recovering to your original state is difficult at
that point, because tools like "git reset --keep" will blow
away anything staged. We can make this safer by refusing to
apply when there are staged changes.
It's possible we could provide better tooling here, as "git
stash apply" should be writing only conflicts to the index
(so we know that any stage-0 entries are potentially
precious). But it is the odd duck; most "mergy" commands
will update the index for cleanly merged entries, and it is
not worth updating our tooling to support this use case
which is unlikely to be of interest (besides which, we would
still need to block a dirty index for "stash apply --index",
since that case _would_ be ambiguous).
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
One of the tests in t3903 wants to make sure that applying a
stash that touches only "file" can still happen even if there
are working tree changes to "other-file". To do so, it adds
"other-file" to the index (since otherwise it is an
untracked file, voiding the purpose of the test).
But as we are about to refactor the dirty-index handling,
and as this test does not actually care about having a dirty
index (only a dirty working tree), let's bump the tracking
of "other-file" into the setup phase, so we can have _just_
a dirty working tree here.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When testing the diff output of "git stash list", we look
for the stash's subject of "WIP on master: $sha1", even
though it's not relevant to the diff output. This makes the
test brittle to refactoring, as any changes to earlier tests
may impact the commit sha1.
Since we don't care about the commit subject here, we can
simply ask stash not to print it.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
"diff-highlight" (in contrib/) used to show byte-by-byte
differences, which meant that multi-byte characters can be chopped
in the middle. It learned to pay attention to character boundaries
(assuming the UTF-8 payload).
* jk/colors:
diff-highlight: do not split multibyte characters
* jk/test-annoyances:
t5551: make EXPENSIVE test cheaper
t5541: move run_with_cmdline_limit to test-lib.sh
t: pass GIT_TRACE through Apache
t: redirect stderr GIT_TRACE to descriptor 4
t: translate SIGINT to an exit
An earlier update to the parser that disects an address broke an
address, followed by a colon, followed by an empty string (instead
of the port number).
* tb/connect-ipv6-parse-fix:
connect.c: ignore extra colon after hostname
* va/fix-git-p4-tests:
t9814: guarantee only one source exists in git-p4 copy tests
git-p4: fix copy detection test
t9814: fix broken shell syntax in git-p4 rename test
The "git push --signed" protocol extension did not limit what the
"nonce" that is a server-chosen string can contain or how long it
can be, which was unnecessarily lax. Limit both the length and the
alphabet to a reasonably small space that can still have enough
entropy.
* jc/push-cert:
push --signed: tighten what the receiving end can ask to sign
A broken or badly formatted commit might not record author or
committer lines or we may not find a valid name on them. The
function record_person() returned after calling get_commit_buffer()
without calling unuse_commit_buffer() on the memory it obtained in
such cases, potentially leaking it.
Helped-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since 33d4221 (write_sha1_file: freshen existing objects,
2014-10-15), we update the mtime of existing objects that we
would have written out (had they not existed). For the
common case in which many objects are packed, we may update
the mtime on a single packfile repeatedly. This can result
in a noticeable performance problem if calling utime() is
expensive (e.g., because your storage is on NFS).
We can fix this by keeping a per-pack flag that lets us
freshen only once per program invocation.
An alternative would be to keep the packed_git.mtime flag up
to date as we freshen, and freshen only once every N
seconds. In practice, it's not worth the complexity. We are
racing against prune expiration times here, which inherently
must be set to accomodate reasonable program running times
(because they really care about the time between an object
being written and it becoming referenced, and the latter is
typically the last step a program takes).
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When writing out an object file, we first check whether it
already exists and if so optimize out the write. Prior to 33d4221, we did this by calling has_sha1_file(), which will
check for packed objects followed by loose. Since that
commit, we check loose objects first.
For the common case of a repository whose objects are mostly
packed, this means we will make a lot of extra access()
system calls checking for loose objects. We should follow
the same packed-then-loose order that all of our other
lookups use.
Reported-by: Stefan Saasen <ssaasen@atlassian.com> Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When pruning and repacking a repository that has an
alternate object store configured, we may traverse a large
number of objects in the alternate. This serves no purpose,
and may be expensive to do. A longer explanation is below.
Commits d3038d2 and abcb865 taught prune and pack-objects
(respectively) to treat "recent" objects as tips for
reachability, so that we keep whole chunks of history. They
built on the object traversal in 660c889 (sha1_file: add
for_each iterators for loose and packed objects,
2014-10-15), which covers both local and alternate objects.
In both cases, covering alternate objects is unnecessary, as
both commands can only drop objects from the local
repository. In the case of prune, we traverse only the local
object directory. And in the case of repacking, while we may
or may not include local objects in our pack, we will never
reach into the alternate with "repack -d". The "-l" option
is only a question of whether we are migrating objects from
the alternate into our repository, or leaving them
untouched.
It is possible that we may drop an object that is depended
upon by another object in the alternate. For example,
imagine two repositories, A and B, with A pointing to B as
an alternate. Now imagine a commit that is in B which
references a tree that is only in A. Traversing from recent
objects in B might prevent A from dropping that tree. But
this case isn't worth covering. Repo B should take
responsibility for its own objects. It would never have had
the commit in the first place if it did not also have the
tree, and assuming it is using the same "keep recent chunks
of history" scheme, then it would itself keep the tree, as
well.
So checking the alternate objects is not worth doing, and
come with a significant performance impact. In both cases,
we skip any recent objects that have already been marked
SEEN (i.e., that we know are already reachable for prune, or
included in the pack for a repack). So there is a slight
waste of time in opening the alternate packs at all, only to
notice that we have already considered each object. But much
worse, the alternate repository may have a large number of
objects that are not reachable from the local repository at
all, and we end up adding them to the traversal.
We can fix this by considering only local unseen objects.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The old wording was somehow implying that <start> and <end> were not
regular expressions. Also, the common case is to use a plain function
name here so <funcname> makes sense (the fact that it is a regular
expression is documented in line-range-format.txt).
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* tag 'gitgui-0.20.0' of http://repo.or.cz/r/git-gui:
git-gui: set version 0.20
git-gui: sv.po: Update Swedish translation (547t0f0u)
git-gui i18n: Updated Bulgarian translation (547t,0f,0u)
git-gui: Makes chooser set 'gitdir' to the resolved path
git-gui: Fixes chooser not accepting gitfiles
git-gui: reinstate support for Tcl 8.4
git-gui: fix problem with gui.maxfilesdisplayed
git-gui: fix verbose loading when git path contains spaces.
git-gui/gitk: Do not depend on Cygwin's "kill" command on Windows
git-gui: add configurable tab size to the diff view
git-gui: Make git-gui lib dir configurable at runime
git-gui i18n: Updated Bulgarian translation (520t,0f,0u)
L10n: vi.po (543t): Init translation for Vietnamese
git-gui: align the new recursive checkbox with the radiobuttons.
git-gui: Add a 'recursive' checkbox in the clone menu.
limit_list: avoid quadratic behavior from still_interesting
When we are limiting a rev-list traversal due to
UNINTERESTING refs, we have to walk down the tips (both
interesting and uninteresting) to find where they intersect.
We keep a queue of commits to examine, pop commits off
the queue one by one, and potentially add their parents. The
size of the queue will naturally fluctuate based on the
"width" of the history graph; i.e., the number of
simultaneous lines of development. But for the most part it
will stay in the same ballpark as the initial number of tips
we fed, shrinking over time (as we hit common ancestors of
the tips). So roughly speaking, if we start with `N` tips,
we'll spend much of the time with a queue around `N` items.
For each UNINTERESTING commit we pop, we call
still_interesting to check whether marking its parents as
UNINTERESTING has made the whole queue uninteresting (in
which case we can quit early). Because the queue is stored
as a linked list, this is `O(N)`, where `N` is the number of
items in the queue. So processing a queue with `N` commits
marked UNINTERESTING (and one or more interesting commits)
will take `O(N^2)`.
If you feed a lot of positive tips, this isn't a problem.
They aren't UNINTERESTING, so they don't incur the
still_interesting check. It also isn't a problem if you
traverse from an interesting tip to some UNINTERESTING
bases. We order the queue by recency, so the interesting
commits stay at the front of the queue as we walk down them.
The linear check can exit early as soon as it sees one
interesting commit left in the queue.
But if you want to know whether an older commit is reachable
from a set of newer tips, we end up processing in the
opposite direction: from the UNINTERESTING ones down to the
interesting one. This may happen when we call:
git rev-list $commits --not --all
in check_everything_connected after a fetch. If we fetched
something much older than most of our refs, and if we have a
large number of refs, the traversal cost is dominated by the
quadratic behavior.
These commands simulate the connectivity check of such a
fetch, when you have `$n` distinct refs in the receiver:
# positive ref is 100,000 commits deep
git rev-list --all | head -100000 | tail -1 >input
# huge number of more recent negative refs
git rev-list --all | head -$n | sed s/^/^/ >>input
time git rev-list --stdin <input
Here are timings for various `n` on the linux.git
repository. The `n=1` case provides a baseline for just
walking the commits, which lets us see the still_interesting
overhead. The times marked with `+` subtract that baseline
to show just the extra time growth due to the large number
of refs. The `x` numbers show the slowdown of the adjusted
time versus the prior trial.
Each trial doubles `n`, so you can see the quadratic (`4x`)
behavior before this patch. Afterwards, we have a roughly
linear relationship.
The implementation is fairly straightforward. Whenever we do
the linear search, we cache the interesting commit we find,
and next time check it before doing another linear search.
If that commit is removed from the list or becomes
UNINTERESTING itself, then we fall back to the linear
search. This is very similar to the trick used by fce87ae
(Fix quadratic performance in rewrite_one., 2008-07-12).
I considered and rejected several possible alternatives:
1. Keep a count of UNINTERESTING commits in the queue.
This requires managing the count not only when removing
an item from the queue, but also when marking an item
as UNINTERESTING. That requires touching the other
functions which mark commits, and would require knowing
quickly which commits are in the queue (lookup in the
queue is linear, so we would need an auxiliary
structure or to also maintain an IN_QUEUE flag in each
commit object).
2. Keep a separate list of interesting commits. Drop items
from it when they are dropped from the queue, or if
they become UNINTERESTING. This again suffers from
extra complexity to maintain the list, not to mention
CPU and memory.
3. Use a better data structure for the queue. This is
something that could help the fix in fce87ae, because
we order the queue by recency, and it is about
inserting quickly in recency order. So a normal
priority queue would help there. But here, we cannot
disturb the order of the queue, which makes things
harder. We really do need an auxiliary index to track
the flag we care about, which is basically option (2)
above.
The "cache" trick is simple, and the numbers above show that
it works well in practice. This is because the length of
time it takes to find an interesting commit is proportional
to the length of time it will remain cached (i.e., if we
have to walk a long way to find it, it also means we have to
pop a lot of elements in the queue until we get rid of it
and have to find another interesting commit).
The worst case is still quadratic, though. We could have `N`
uninteresting commits at the front of the queue, followed by
`N` interesting commits, where commit `i` has parent `i+N`.
When we pop commit `i`, we will notice that the parent of
the next commit, `i+1+N` is still interesting and cache it.
But then handling commit `i+1`, we will mark its parent
`i+1+N` uninteresting, and immediately invalidate our cache.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When commit fe8e3b7 refactored type_from_string to allow
input that was not NUL-terminated, it switched to using
strncmp instead of strcmp. But this means we check only the
first "len" bytes of the strings, and ignore any remaining
bytes in the object_type_string. We should make sure that it
is also "len" bytes, or else we would accept "comm" as
"commit", and so forth.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
config: use utf8_bom[] from utf.[ch] in git_parse_source()
Because the function reads one character at the time, unfortunately
we cannot use the easier skip_utf8_bom() helper, but at least we do
not have to duplicate the constant string this way.