Suppose a fetch or push is requested between two shallow repositories
(with no history deepening or shortening). A pack that contains
necessary objects is transferred over together with .git/shallow of
the sender. The receiver has to determine whether it needs to update
.git/shallow if new refs needs new shallow comits.
The rule here is avoid updating .git/shallow by default. But we don't
want to waste the received pack. If the pack contains two refs, one
needs new shallow commits installed in .git/shallow and one does not,
we keep the latter and reject/warn about the former.
Even if .git/shallow update is allowed, we only add shallow commits
strictly necessary for the former ref (remember the sender can send
more shallow commits than necessary) and pay attention not to
accidentally cut the receiver history short (no history shortening is
asked for)
So the steps to figure out what ref need what new shallow commits are:
1. Split the sender shallow commit list into "ours" and "theirs" list
by has_sha1_file. Those that exist in current repo in "ours", the
remaining in "theirs".
2. Check the receiver .git/shallow, remove from "ours" the ones that
also exist in .git/shallow.
3. Fetch the new pack. Either install or unpack it.
4. Do has_sha1_file on "theirs" list again. Drop the ones that fail
has_sha1_file. Obviously the new pack does not need them.
5. If the pack is kept, remove from "ours" the ones that do not exist
in the new pack.
6. Walk the new refs to answer the question "what shallow commits,
both ours and theirs, are required in .git/shallow in order to add
this ref?". Shallow commits not associated to any refs are removed
from their respective list.
7. (*) Check reachability (from the current refs) of all remaining
commits in "ours". Those reachable are removed. We do not want to
cut any part of our (reachable) history. We only check up
commits. True reachability test is done by
check_everything_connected() at the end as usual.
8. Combine the final "ours" and "theirs" and add them all to
.git/shallow. Install new refs. The case where some hook rejects
some refs on a push is explained in more detail in the push
patches.
Of these steps, #6 and #7 are expensive. Both require walking through
some commits, or in the worst case all commits. And we rather avoid
them in at least common case, where the transferred pack does not
contain any shallow commits that the sender advertises. Let's look at
each scenario:
1) the sender has longer history than the receiver
All shallow commits from the sender will be put into "theirs" list
at step 1 because none of them exists in current repo. In the
common case, "theirs" becomes empty at step 4 and exit early.
2) the sender has shorter history than the receiver
All shallow commits from the sender are likely in "ours" list at
step 1. In the common case, if the new pack is kept, we could empty
"ours" and exit early at step 5.
If the pack is not kept, we hit the expensive step 6 then exit
after "ours" is emptied. There'll be only a handful of objects to
walk in fast-forward case. If it's forced update, we may need to
walk to the bottom.
3) the sender has same .git/shallow as the receiver
This is similar to case 2 except that "ours" should be emptied at
step 2 and exit early.
A fetch after "clone --depth=X" is case 1. A fetch after "clone" (from
a shallow repo) is case 3. Luckily they're cheap for the common case.
A push from "clone --depth=X" falls into case 2, which is expensive.
Some more work may be done at the sender/client side to avoid more
work on the server side: if the transferred pack does not contain any
shallow commits, send-pack should not send any shallow commits to the
receive-pack, effectively turning it into a normal push and avoid all
steps.
This patch implements all steps except #3, already handled by
fetch-pack and receive-pack, #6 and #7, which has their own patch due
to their size.
(*) in previous versions step 7 was put before step 3. I reorder it so
that the common case that keeps the pack does not need to walk
commits at all. In future if we implement faster commit
reachability check (maybe with the help of pack bitmaps or commit
cache), step 7 could become cheap and be moved up before 6 again.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
No callers pass a non-empty pointer as shallow_points at this
stage. As a result, all clients still refuse to talk to shallow
repository on the other end.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If either receive-pack or upload-pack is called on a shallow
repository, shallow commits (*) will be sent after the ref
advertisement (but before the packet flush), so that the receiver has
the full "shape" of the sender's commit graph. This will be needed for
the receiver to update its .git/shallow if necessary.
This breaks the protocol for all clients trying to push to a shallow
repo, or fetch from one. Which is basically the same end result as
today's "is_repository_shallow() && die()" in receive-pack and
upload-pack. New clients will be made aware of shallow upstream and
can make use of this information.
The sender must send all shallow commits that are sent in the
following pack. It may send more shallow commits than necessary.
upload-pack for example may choose to advertise no shallow commits if
it knows in advance that the pack it's going to send contains no
shallow commits. But upload-pack is the server, so we choose the
cheaper way, send full .git/shallow and let the client deal with it.
Smart HTTP is not affected by this patch. Shallow support on
smart-http comes later separately.
(*) A shallow commit is a commit that terminates the revision
walker. It is usually put in .git/shallow in order to keep the
revision walker from going out of bound because there is no
guarantee that objects behind this commit is available.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If we borrow objects from another repository, we should also pay
attention to their $GIT_DIR/shallow (and even info/grafts). But
current alternates code does not.
Reject alternate repos that are shallow because we do not do it
right. In future the alternate code may be updated to check
$GIT_DIR/shallow properly so that this restriction could be lifted.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
send-pack can send a pack with loose ends to the server. receive-pack
before 6d4bb38 (fetch: verify we have everything we need before
updating our ref - 2011-09-01) does not detect this and keeps the pack
anyway, which corrupts the repository, at least from fsck point of
view.
send-pack will learn to safely push from a shallow repository later.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The latter can do everything the former can and is used in many more
places.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 15a147e (rebase: use @{upstream} if no upstream specified,
2011-02-09) says:
Make it default to 'git rebase @{upstream}'. That is also what
'git pull [--rebase]' defaults to, so it only makes sense that
'git rebase' defaults to the same thing.
but that isn't actually the case. Since commit d44e712 (pull: support
rebased upstream + fetch + pull --rebase, 2009-07-19), pull has actually
chosen the most recent reflog entry which is an ancestor of the current
branch if it can find one.
Add a '--fork-point' argument to git-rebase that can be used to trigger
this behaviour. This option is turned on by default if no non-option
arguments are specified on the command line, otherwise we treat an
upstream specified on the command-line literally.
Signed-off-by: John Keeping <john@keeping.me.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* git://repo.or.cz/git-gui:
git-gui: correct spelling errors in comments
git-gui: add menu item to launch a bash shell on Windows.
git-gui: corrected setup of git worktree under cygwin.
git-gui: right half window is paned
git-gui: Add gui.displayuntracked option
git-gui: show the maxrecentrepo config option in the preferences dialog
git-gui: added gui.maxrecentrepo to extend the number of remembered repos
git-gui: Improve font rendering on retina macbooks
* git://ozlabs.org/~paulus/gitk:
gitk: Recognize -L option
gitk: Support showing the gathered inline diffs
gitk: Split out diff part in $commitinfo
gitk: Refactor per-line part of getblobdiffline and its support
gitk: Support -G option from the command line
gitk: Tag display improvements
Use only one return point from git_connect(), doing the
free();
return conn;
only at one place in the code.
There may be a little confusion what the variable "host" is for. At
some places it is only the host part, at other places it may include
the port number, so change host into hostandport here.
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Make the function is_local() in transport.c public, rename it into
url_is_local_not_ssh() and use it in both transport.c and connect.c
Use a protocol "local" for URLs for the local file system.
One note about using file:// under Windows:
The (absolute) path on Unix like system typically starts with "/".
When the host is empty, it can be omitted, so that a shell scriptlet
url=file://$pwd
will give a URL like "file:///home/user/repo".
Windows does not have the same concept of a root directory located in "/".
When parsing the URL allow "file://C:/user/repo"
(even if RFC1738 indicates that "file:///C:/user/repo" should be used).
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use get_host_and_port() even for ssh.
Remove the variable port git_connect(), and simplify parse_connect_url()
Use only one return point in git_connect(), doing the free() and return conn.
t5601 had 2 corner test cases which now pass.
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The documentation (in urls.txt) says that
"ssh://host:/~repo",
"host:/~repo" or
"host:~repo"
specify the repository "repo" in the home directory at "host".
This has not been working for "host:/~repo".
Before commit 356bec "Support [address] in URLs", the comparison
"url != hostname" could be used to determine if the URL had a scheme
or not: "ssh://host/host" != "host".
However, after 356bec "[::1]" was converted into "::1", yielding
url != hostname as well. To fix this regression, don't use
"if (url != hostname)", but look at the separator instead.
Rename the variable "c" into "separator" to make it easier to read.
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add test cases using git fetch-pack --diag-url:
- parse out host and path for URLs with a scheme (git:// file:// ssh://)
- parse host names embedded by [] correctly
- extract the port number, if present
- separate URLs like "file" (which are local)
from URLs like "host:repo" which should use ssh
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The main purpose is to trace the URL parser called by git_connect() in
connect.c
The main features of the parser can be listed as this:
- parse out host and path for URLs with a scheme (git:// file:// ssh://)
- parse host names embedded by [] correctly
- extract the port number, if present
- separate URLs like "file" (which are local)
from URLs like "host:repo" which should use ssh
Add the new parameter "--diag-url" to "git fetch-pack", which prints
the value for protocol, host and path to stderr and exits.
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
git_connect has grown large due to the many different protocols syntaxes
that are supported. Move the part of the function that parses the URL to
connect to into a separate function for readability.
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since day one, function git_connect() had a limit on the command line of
the command that is invoked to make a connection. 7a33bcbe converted the
code that constructs the command to strbuf. This would have been the
right time to remove the limit, but it did not happen. Remove it now.
git_connect() uses start_command() to invoke the command; consequently,
the limits of the system still apply, but are diagnosed only at execve()
time. But these limits are more lenient than the 1K that git_connect()
imposed.
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When rev-parse looks at whether an argument like "foo..bar" or
"foobar^@" is a difference or parent-shorthand, it internally
munges the arguments so that it can pass the individual rev
arguments to get_sha1(). However, we do not consistently un-munge
the result.
For cases where we do not match (e.g., "doesnotexist..HEAD"), we
would then want to try to treat the argument as a filename.
try_difference gets() this right, and always unmunges in this case.
However, try_parent_shorthand() never unmunges, leading to incorrect
error messages, or even incorrect results:
$ git rev-parse foobar^@
foobar
fatal: ambiguous argument 'foobar': unknown revision or path not in the working tree.
Use '--' to separate paths from revisions, like this:
'git <command> [<revision>...] -- [<file>...]'
$ >foobar
$ git rev-parse foobar^@
foobar
For cases where we do match, neither function unmunges. This does
not currently matter, since we are done with the argument. However,
a future patch will do further processing, and this prepares for
it. In addition, it's simply a confusing interface for some cases to
modify the const argument, and others not to.
Signed-off-by: Jeff King <peff@peff.net>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 66713ef (pull: allow pull to preserve merges when rebasing)
didn't include an update so 'git remote status' parses branch.<name>.rebase=preserve
correctly, let's do that.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Oftentimes people will make the same change in two branches, revert the change
in one branch, and then be surprised when a merge reinstitutes that change when
the branches are merged. Add an explanatory paragraph that explains that this
occurs and the reason why, so people are not surprised.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There's no mention of the 'origin' default, or the fact that the
upstream tracking branch remote is used.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
So that we check that UTF-8 and spaces work fine.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If the repository is moved, the absolute path of the shared repository
would fail.
Make sure it's always up-to-date.
Reported-by: Michael Davis <mjmdavis@gmail.com>
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Suggested-by: Roman Ovchinnikov <coolthecold@gmail.com>
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
error on pull: fatal: Invalid raw date "" in ident: remote-hg <>
Neither %s nor %z are officially supported by python, they may work on
some (most?) platforms, but not all.
removed strftime use of %s and %z, which are not officially supported by python, with standard formats
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since commit d96855f (merge-base: teach "--fork-point" mode, 2013-10-23)
we can replace a shell loop in git-pull with a single call to
git-merge-base. So let's do so.
Signed-off-by: John Keeping <john@keeping.me.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
After a86a8b9 (sb/parseopt-boolean-removal), the deprecated
OPT_BOOLEAN is not used anywhere except by OPT__* macros. Kill
OPT_BOOLEAN and make OPT__* use OPT_COUNTUP directly instead. This
should stop OPT_BOOLEAN from entering the tree again in new patches.
OPT__DRY_RUN() is converted to use OPT_BOOL though because it does not
make sense to increase the level of dryness. All OPT__DRY_RUN call
sites have been checked and they look safe for OPT_BOOL.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Rev-parse understands that a "--" may separate revisions and
filenames, and that anything after the "--" is taken as-is.
However, it does not understand that anything before the
token must be a revision (which is the usual rule
implemented by the setup_revisions parser).
Since rev-parse prefers revisions to files when parsing
before the "--", we end up with the correct result (if such
an argument is a revision, we parse it as one, and if it is
not, it is an error either way). However, we misdiagnose
the errors:
$ git rev-parse foobar -- >/dev/null
fatal: ambiguous argument 'foobar': unknown revision or path not in the working tree.
Use '--' to separate paths from revisions, like this:
'git <command> [<revision>...] -- [<file>...]'
$ >foobar
$ git rev-parse foobar -- >/dev/null
fatal: bad flag '--' used after filename
In both cases, we should know that the real error is that
"foobar" is meant to be a revision, but could not be
resolved.
Signed-off-by: Jeff King <peff@peff.net>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Additionally, precedence of negated patterns is exactly as outlined in
the DESCRIPTION section, we don't need to repeat this.
Signed-off-by: Karsten Blees <blees@dcon.de>
Reviewed-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When --prompt option is set, git-difftool displays a prompt for each
modified file to be viewed in an external diff program. At that
point, it could be useful to display a counter and the total number
of files in the diff queue.
Below is the current difftool prompt for the first of 5 modified files:
Viewing: 'diff.c'
Launch 'vimdiff' [Y/n]:
Consider the modified prompt:
Viewing (1/5): 'diff.c'
Launch 'vimdiff' [Y/n]:
The current GIT_EXTERNAL_DIFF mechanism does not tell the number of
paths in the diff queue nor the current counter. To make this
"counter/total" info available for GIT_EXTERNAL_DIFF programs
without breaking existing ones by doing the following:
- Keep track of the number of paths shown so far in diff_options;
- Export two new environment variables from run_external_diff() to
show the total number of paths (from diff_queue_struct) and the
current value of the counter (from diff_options); and
- Update git-difftool--helper to use these two environment variables.
Signed-off-by: Zoltan Klinger <zoltan.klinger@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Back in 233c3e6 (parse_pathspec: preserve prefix length via
PATHSPEC_PREFIX_ORIGIN - 2013-07-14), parse_pathspec() is taught to
save prefix length as a dynamic magic. This is needed when the
pathspec is passed to another process and and prefix lenght would be
lost.
Back then we support two cases. If the pathspec is normal, e.g. "abc",
we simply add the prefix to become ":(prefix:2)abc". If the pathspec
contains long magic, e.g. ":(foo,bar)abc" then we turn it to
":(foo,bar,prefix:2)abc". We do not support prefixing on short form,
because the only supported mnemonic '/' disappears after the the
preprocessing steps.
With the introduction of exclude magic with mnemonic '!', we need to
add support for the short form case so that ':!abc' becomes
':(exclude,prefix:2)abc'. Without this, it will break
cd Documentation
git add -p -- . ':!technical'
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Instead of simply ignoring the value passed to --depth option when
it is zero or negative, catch and report it as an error to let
people know that they were using the option incorrectly.
Original-patch-by: Andrés G. Aragoneses <knocte@gmail.com>
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
git-config used a static match array to hold the matches we want to
unset/replace when using --unset or --replace-all. Use a
variable-sized array instead.
This in particular fixes the symptoms git-svn had when storing large
numbers of svn-remote.*.added-placeholder entries in the config file.
While the tests are rather more paranoid than just --unset and
--replace-all, the other operations already worked. Indeed git-svn's
usage only breaks the first time *after* creating so many entries,
when it wants to unset and re-add them all.
Reported-by: Jess Hottenstein <jess.hottenstein@gmail.com>
Signed-off-by: Thomas Rast <tr@thomasrast.ch>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git diff -- ':(icase)makefile'" were rejected unnecessarily.
This needs to be merged to 'maint' later.
* nd/magic-pathspec:
diff: restrict pathspec limitations to diff b/f case only
Add a few formatting directives to "git for-each-ref --format=...",
to paint them in color, etc.
* rr/for-each-ref-decoration:
for-each-ref: avoid color leakage
for-each-ref: introduce %(color:...) for color
for-each-ref: introduce %(upstream:track[short])
for-each-ref: introduce %(HEAD) asterisk marker
t6300 (for-each-ref): don't hardcode SHA-1 hexes
t6300 (for-each-ref): clearly demarcate setup
Updates to remote-bzr and remote-hg in contrib.
* rh/remote-hg-bzr-updates:
remote-bzr, remote-hg: fix email address regular expression
test-hg.sh: help user correlate verbose output with email test
test-hg.sh: fix duplicate content strings in author tests
test-hg.sh: avoid obsolete 'test' syntax
test-hg.sh: eliminate 'local' bashism
test-bzr.sh, test-hg.sh: prepare for change to push.default=simple
test-bzr.sh, test-hg.sh: allow running from any dir
test-lib.sh: convert $TEST_DIRECTORY to an absolute path
Allow customizing the paths to Perl modules with the new
PERLLIB_EXTRA makefile variable.
* jn/perl-lib-extra:
Makefile: add PERLLIB_EXTRA variable that adds to default perl path
Makefile: rebuild perl scripts when perl paths change
Our current scheme for naming packfiles is to calculate the
sha1 hash of the sorted list of objects contained in the
packfile. This gives us a unique name, so we are reasonably
sure that two packs with the same name will contain the same
objects.
It does not, however, tell us that two such packs have the
exact same bytes. This makes things awkward if we repack the
same set of objects. Due to run-to-run variations, the bytes
may not be identical (e.g., changed zlib or git versions,
different source object reuse due to new packs in the
repository, or even different deltas due to races during a
multi-threaded delta search).
In theory, this could be helpful to a program that cares
that the packfile contains a certain set of objects, but
does not care about the particular representation. In
practice, no part of git makes use of that, and in many
cases it is potentially harmful. For example, if a dumb http
client fetches the .idx file, it must be sure to get the
exact .pack that matches it. Similarly, a partial transfer
of a .pack file cannot be safely resumed, as the actual
bytes may have changed. This could also affect a local
client which opened the .idx and .pack files, closes the
.pack file (due to memory or file descriptor limits), and
then re-opens a changed packfile.
In all of these cases, git can detect the problem, as we
have the sha1 of the bytes themselves in the pack trailer
(which we verify on transfer), and the .idx file references
the trailer from the matching packfile. But it would be
simpler and more efficient to actually get the correct
bytes, rather than noticing the problem and having to
restart the operation.
This patch simply uses the pack trailer sha1 as the pack
name. It should be similarly unique, but covers the exact
representation of the objects. Other parts of git should not
care, as the pack name is returned by pack-objects and is
essentially opaque.
One test needs to be updated, because it actually corrupts a
pack and expects that re-packing the corrupted bytes will
use the same name. It won't anymore, but we can easily just
use the name that pack-objects hands back.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
