remote.<remote>.vcs causes remote->foreign_vcs to be set on entry to
transport_get(). Unfortunately, the code assumed that any such entry
is stale from previous round.
Fix this by making VCS set by URL to be volatile w.r.t. transport_get()
instead.
Signed-off-by: Ilari Liusvaara <ilari.liusvaara@elisanet.fi> Acked-by: Daniel Barkalow <barkalow@iabervon.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jl/diff-submodule-ignore:
Teach diff --submodule that modified submodule directory is dirty
git diff: Don't test submodule dirtiness with --ignore-submodules
Make ce_uptodate() trustworthy again
work around an obnoxious bash "safety feature" on OpenBSD
Bash (4.0.24) on OpenBSD 4.6 refuses to run this snippet:
$ cat gomi.sh
#!/bin/sh
one="/var/tmp/1 1"
rm -f /var/tmp/1 "/var/tmp/1 1"
echo hello >$one
$ sh gomi.sh; ls /var/tmp/1*
/var/tmp/1 1
$ bash gomi.sh; ls /var/tmp/1*
gomi.sh: line 4: $one: ambiguous redirect
ls: /var/tmp/1*: No such file or directory
Every competent shell programmer knows that a <$word in redirection is not
subject to field splitting (POSIX.1 "2.7 Redirection" explicitly lists the
kind of expansion performed: "... the word that follows the redirection
operator shall be subjected to ...", and "Field Splitting" is not among
them).
Some clueless folks apparently decided that users need to be protected in
the name of "security", however.
Output from "git grep -e '> *\$' -- '*.sh'" indicates that rebase-i
suffers from this bogus "safety". Work it around by surrounding the
variable reference with a dq pair.
builtin-config: Fix crash when using "-f <relative path>" from non-root dir
When your current directory is not at the root of the working tree, and you
use the "-f" option with a relative path, the current code tries to read
from a wrong file, since argv[2] is now beyond the end of the rearranged
argument list.
This patch replaces the incorrect argv[2] with the variable holding the
given config file name.
The bug was introduced by d64ec16 (git config: reorganize to use parseopt).
[jc: added test]
Signed-off-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Allow using "["<host>"]":<port> and "["<host>"]" notations in git://
host addresses. This is needed to be able to connect to addresses
that contain ':' (e.g. numeric IPv6 addresses). Also send the host
header []-wrapped so it can actually be parsed by remote end.
Signed-off-by: Ilari Liusvaara <ilari.liusvaara@elisanet.fi> Signed-off-by: Junio C Hamano <gitster@pobox.com>
If host address could have ':' in it (e.g. numeric IPv6 address), then
host and port could not be uniquely parsed. Fix this by parsing the
"["<host>"]":<port> and "["<host>"]" notations. Currently the built-in
git:// client would send <host>:<port> or <host> for such thing, but
it doesn't matter as due to bugs, resolving address fails if <host>
contains ':'.
Signed-off-by: Ilari Liusvaara <ilari.liusvaara@elisanet.fi> Signed-off-by: Junio C Hamano <gitster@pobox.com>
grep: use REG_STARTEND (if available) to speed up regexec
BSD and glibc have an extension to regexec which takes a buffer + length pair
instead of a NUL-terminated string. Since we already have the length computed
this can save us a strlen call inside regexec.
Signed-off-by: Benjamin Kramer <benny.kra@googlemail.com> Acked-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
am: fix patch format detection for Thunderbird "Save As" emails
The patch detection wants to inspect all the headers of a rfc2822 message
and ensure that they look like header fields. The headers are always
separated from the message body with a blank line. When Thunderbird saves
the message the blank line separating the headers from the body includes a
CR. The patch detection is failing because a CRLF doesn't match /^$/. Fix
this by allowing a CR to exist on the separating line.
Signed-off-by: Stephen Boyd <bebarino@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
append_cr(), remove_cr(), q_to_nul() and q_to_cr() are defined in multiple
tests. Consolidate them into test-lib.sh so we can stop redefining them.
The use of remove_cr() in t0020 to test for a CR is replaced with a new
function has_cr() to accurately reflect what is intended (the output of
remove_cr() was being thrown away).
Signed-off-by: Stephen Boyd <bebarino@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t5541-http-push: make grep expression check for one line only
Don't feed a multiple-line pattern to grep and expect the them to match
with lines in order.
Simplify the grep expressions in the non-fast-forward tests to check
only for the first line of the non-fast-forward warning - having that
line should be enough assurance that the full warning is printed.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Teach diff --submodule that modified submodule directory is dirty
Since commit 8e08b4 git diff does append "-dirty" to the work tree side
if the working directory of a submodule contains new or modified files.
Lets do the same when the --submodule option is used.
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git diff: Don't test submodule dirtiness with --ignore-submodules
The diff family suppresses the output of submodule changes when
requested but checks them nonetheless. But since recently submodules
get examined for their dirtiness, which is rather expensive. There is
no need to do that when the --ignore-submodules option is used, as
the gathered information is never used anyway.
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
In Internet Explorer 8 (IE8) the 'blame_incremental' view, which uses
JavaScript to generate blame info using AJAX, sometimes hang at the
beginning (at 0%) of blaming, e.g. for larger files with long history
like git's own gitweb/gitweb.perl.
The error shown by JavaScript console is "Unspecified error" at char:2
of the following line in gitweb/gitweb.js:
if (xhr.readyState === 3 && xhr.status !== 200) {
Debugging it using IE8 JScript debuger shown that the error occurs
when trying to access xhr.status (xhr is XMLHttpRequest object).
Watch for xhr object shows 'Unspecified error.' as "value" of
xhr.status, and trying to access xhr.status from console throws error.
This bug is some intermittent bug, depending on XMLHttpRequest timing,
as it doesn't occur in all cases. It is probably caused by the fact
that handleResponse is called from timer (pollTimer), to work around
the fact that some browsers call onreadystatechange handler only once
for each state change, and not like required for 'blame_incremental'
as soon as new data is available from server. It looks like xhr
object is not properly initialized; still it is a bug to throw an
error when accessing xhr.status (and not use 'null' or 'undefined' as
value).
Work around this bug in IE8 by using try-catch block when accessing
xhr.status.
Signed-off-by: Jakub Narebski <jnareb@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jc/fix-tree-walk:
read-tree --debug-unpack
unpack-trees.c: look ahead in the index
unpack-trees.c: prepare for looking ahead in the index
Aggressive three-way merge: fix D/F case
traverse_trees(): handle D/F conflict case sanely
more D/F conflict tests
tests: move convenience regexp to match object names to test-lib.sh
* maint:
git-gui: work from the .git dir
git-gui: Fix applying a line when all following lines are deletions
git-gui: Correct file_states when unstaging partly staged entry
git-gui: Fix gitk for branch whose name matches local file
git-gui: Keep repo_config(gui.recentrepos) and .gitconfig in sync
git-gui: handle really long error messages in updateindex.
git-gui: Add hotkeys for "Unstage from commit" and "Revert changes"
git-gui: Makefile: consolidate .FORCE-* targets
Merge branch 'maint' of git://git.spearce.org/git-gui into maint
* 'maint' of git://git.spearce.org/git-gui:
git-gui: work from the .git dir
git-gui: Fix applying a line when all following lines are deletions
git-gui: Correct file_states when unstaging partly staged entry
git-gui: Fix gitk for branch whose name matches local file
git-gui: Keep repo_config(gui.recentrepos) and .gitconfig in sync
git-gui: handle really long error messages in updateindex.
git-gui: Add hotkeys for "Unstage from commit" and "Revert changes"
git-gui: Makefile: consolidate .FORCE-* targets
Commit 44626dc7 (MSVC: Windows-native implementation for subset
of threads API, 2010-01-15) introduces builtin replacement of
pthreadGC2.dll functionality, thus we can completely drop
dependency on this dll.
Signed-off-by: Michael Lukashov <michael.lukashov@gmail.com> Acked-by: Johannes Sixt <j6t@kdbg.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
bed575e (commit: support commit.status, --status, and --no-status,
2009-12-07) forgot to add the :: that sets off an item from the
paragraph that explains it, breaking the layout.
Signed-off-by: Thomas Rast <trast@student.ethz.ch> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Merge branch 'doc-style/for-next' of git://repo.or.cz/git/trast
* 'doc-style/for-next' of git://repo.or.cz/git/trast:
Documentation: merge: use MERGE_HEAD to refer to the remote branch
Documentation: simplify How Merge Works
Documentation: merge: add a section about fast-forward
Documentation: emphasize when git merge terminates early
Documentation: merge: add an overview
Documentation: merge: move merge strategy list to end
Documentation: suggest `reset --merge` in How Merge Works section
Documentation: merge: move configuration section to end
Documentation: emphasise 'git shortlog' in its synopsis
Documentation: show-files is now called git-ls-files
Documentation: tiny git config manual tweaks
Documentation: git gc packs refs by default now
Documentation: merge: use MERGE_HEAD to refer to the remote branch
commit 57bddb11 (Documentation/git-merge: reword references to
"remote" and "pull", 2010-01-07) fixed the manual to drop the
assumption that the other branch being merged is from a remote
repository. Unfortunately, in a few places, to do so it
introduced the antecedentless phrase "their versions". Worse, in
passages like the following, 'they' is playing two roles.
| highlighting changes from both the HEAD and their versions.
|
| * Look at the diffs on their own. 'git log --merge -p <path>'
Using HEAD and MERGE_HEAD nicely assigns terminology to "our" and
"their" sides. It also provides the reader with practice using
names that git will recognize on the command line.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Thomas Rast <trast@student.ethz.ch>
The user most likely does not care about the exact order of
operations because he cannot see it happening anyway. Instead,
try to explain what it means to merge two commits into a single
tree.
While at it:
- Change the heading to TRUE MERGE. The entire manual page is
about how merges work.
- Document MERGE_HEAD. It is a useful feature, since it makes
the parents of the intended merge commit easier to refer to.
- Do not assume commits named on the 'git merge' command line come
from another repository. For simplicity, the discussion of
conflicts still does assume that there is only one and it is a
branch head.
- Do not start list items with `code`. Otherwise, a toolchain bug
produces a line break in the generated nroff, resulting in odd
extra space.
Suggested-by: Thomas Rast <trast@student.ethz.ch> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Thomas Rast <trast@student.ethz.ch>
Documentation: emphasize when git merge terminates early
A merge-based operation in git can fail in two ways: one that
stops before touching anything, or one that goes ahead and
results in conflicts.
As the 'git merge' manual explains:
| A merge is always between the current `HEAD` and one or more
| commits (usually, branch head or tag), and the index file must
| match the tree of `HEAD` commit (i.e. the contents of the last commit)
| when it starts out.
Unfortunately, the placement of this sentence makes it easy to
skip over, and its formulation leaves the important point, that
any other attempted merge will be gracefully aborted, unspoken.
So give this point its own section and expand upon it.
Probably this could be simplified somewhat: after all, a change
registered in the index is just a special kind of local
uncommited change, so the second added paragraph is only a
special case of the first. It seemed more helpful to be explicit
here.
Inspired by <http://gitster.livejournal.com/25801.html>.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Thomas Rast <trast@student.ethz.ch>
Documentation: merge: move merge strategy list to end
So the section layout changes as follows:
NAME
SYNOPSIS
DESCRIPTION
OPTIONS
-MERGE STRATEGIES
HOW MERGE WORKS
HOW CONFLICTS ARE PRESENTED
HOW TO RESOLVE CONFLICTS
EXAMPLES
+MERGE STRATEGIES
CONFIGURATION
SEE ALSO
AUTHOR
DOCUMENTATION
GIT
NOTES
The first-time user will care more about conflicts than about
strategies other than 'recursive'.
One of the examples uses -s ours, but I do not think this hinders
readability.
Suggested-by: Thomas Rast <trast@student.ethz.ch> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Thomas Rast <trast@student.ethz.ch>
Documentation: suggest `reset --merge` in How Merge Works section
The 'merge' manual suggests 'reset' to cancel a merge at the end
of the Merge Strategies list. It is more logical to explain this
right before explaining how merge conflicts work, so the daunted
reader can have a way out when he or she needs it most.
While at it, make the advice more dependable and self-contained
by providing the --merge option.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Thomas Rast <trast@student.ethz.ch>
The rule has always been that a cache entry that is ce_uptodate(ce)
means that we already have checked the work tree entity and we know
there is no change in the work tree compared to the index, and nobody
should have to double check. Note that false ce_uptodate(ce) does not
mean it is known to be dirty---it only means we don't know if it is
clean.
There are a few codepaths (refresh-index and preload-index are among
them) that mark a cache entry as up-to-date based solely on the return
value from ie_match_stat(); this function uses lstat() to see if the
work tree entity has been touched, and for a submodule entry, if its
HEAD points at the same commit as the commit recorded in the index of
the superproject (a submodule that is not even cloned is considered
clean).
A submodule is no longer considered unmodified merely because its HEAD
matches the index of the superproject these days, in order to prevent
people from forgetting to commit in the submodule and updating the
superproject index with the new submodule commit, before commiting the
state in the superproject. However, the patch to do so didn't update
the codepath that marks cache entries up-to-date based on the updated
definition and instead worked it around by saying "we don't trust the
return value of ce_uptodate() for submodules."
This makes ce_uptodate() trustworthy again by not marking submodule
entries up-to-date.
The next step _could_ be to introduce a few "in-core" flag bits to
cache_entry structure to record "this entry is _known_ to be dirty",
call is_submodule_modified() from ie_match_stat(), and use these new
bits to avoid running this rather expensive check more than once, but
that can be a separate patch.
rebase -i: Export GIT_AUTHOR_* variables explicitly
There is no point doing self-assignments of these variables. Instead,
just export them to the environment, but do so in a sub-shell, because
VAR1=VAL1 VAR2=VAL2 ... command arg1 arg2...
does not mark the variables exported if command that is run
is a shell function, according to POSIX.1.
The callers of do_with_author do not rely on seeing the effect of any
shell variable assignments that may happen inside what was called through
this shell function (currently "output" is the only one), so running it in
the subshell doesn't have an adverse semantic effect.
git-gui: Fix applying a line when all following lines are deletions
If a diff looked like:
@@
context
-del1
-del2
and you wanted to stage the deletion 'del1', the generated patch
wouldn't apply because it was missing the line 'del2' converted to
context, but this line was counted in the @@-line
Signed-off-by: Jeff Epler <jepler@unpythonic.net> Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
git-gui: Correct file_states when unstaging partly staged entry
When unstaging a partly staged file or submodule, the file_states
list was not updated properly (unless unstaged linewise). Its
index_info part did not contain the former head_info as it should
have but kept its old value.
This seems not to have had any bad effects but diminishes the value
of the file_states list for future enhancements.
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de> Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
git-gui: Fix gitk for branch whose name matches local file
When trying to run gitk on a branch name whose name matches a local
file, it will toss an error saying that the name is ambiguous. Adding
a pair of dashes will make gitk parse the options to the left of
it as branch names. Since wish eats the first pair of dashes we
throw at it, we need to add a second one to ensure they get through.
Signed-off-by: Peter Krefting <peter@softwolves.pp.se> Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
git-gui: Keep repo_config(gui.recentrepos) and .gitconfig in sync
When the number of recent repo's gets to ten there can be a
situation where an item is removed from the .gitconfig file via
a call to git config --unset, but the internal representation of
that file (repo_config(gui.recentrepo)) is not updated. Then a
subsequent attempt to remove an item from the list fails because
git-gui attempts to call --unset on a value that has already
been removed. This leads to duplicates in the .gitconfig file,
which then also cause errors if the git-gui tries to --unset them
(rather than using --unset-all. --unset-all is not used because it
is not expected that duplicates should ever be allowed to exist.)
When loading the list of recent repositories (proc _get_recentrepos)
if a repo in the list is not considered a valid git reposoitory
then we should go ahead and remove it so it doesn't take up a slot
in the list (since we limit to 10 items). This will prevent a bunch
of invalid entries in the list (which are not shown) from making
valid entries dissapear off the list even when there are less than
ten valid entries.
See: http://code.google.com/p/msysgit/issues/detail?id=362 Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
git-gui: handle really long error messages in updateindex.
As reported to msysGit (bug #340) it is possible to get some very
long error messages when updating the index. The use of a label to
display this prevents scrolling the output. This patch replaces the
label with a scrollable text widget configured to look like a label.
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net> Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
* git://git.bogomips.org/git-svn:
git-svn: allow subset of branches/tags to be specified in glob spec
git-svn: allow UUID to be manually remapped via rewriteUUID
git-svn: update svn mergeinfo test suite
git-svn: document --username/commit-url for branch/tag
git-svn: add --username/commit-url options for branch/tag
git-svn: respect commiturl option for branch/tag
git-svn: fix mismatched src/dst errors for branch/tag
git-svn: handle merge-base failures
git-svn: ignore changeless commits when checking for a cherry-pick
git-svn: allow subset of branches/tags to be specified in glob spec
For very large projects it is useful to be able to clone a subset of the
upstream SVN repo's branches. Allow for this by letting the left-side of
the branches and tags glob specs contain a brace-delineated comma-separated
list of names. e.g.:
Imagine that only "git-svn clone" and "svnsync clone" are made available
to external users. Furthur, "git-svn clone" contains only trunk, and for
reasons unknown, "svnsync clone" is missing the revision properties that
normally provide the origin svn repo's UUID.
A git user who has cloned the "git-svn clone" repo now wishes to use
git-svn to pull in the missing branches from the "synsync clone" repo.
In order for git-svn to get the history correct for those branches,
it needs to know the origin svn repo's UUID. Hence rewriteUUID.
Signed-off-by: Jay Soffian <jaysoffian@gmail.com> Acked-by: Eric Wong <normalperson@yhbt.net>
When constructing a destination URL, use the property 'commiturl' if it
is specified in the configuration file; otherwise take 'url' as usual.
This accommodates the scenario where a user only wants to involve the
writable repository in operations performing a commit and defaults
everything else to a read-only URL.
[ew: shortened subject]
Signed-off-by: Igor Mironov <igor.a.mironov@gmail.com> Acked-by: Eric Wong <normalperson@yhbt.net>
git-svn: fix mismatched src/dst errors for branch/tag
This fixes the following issue:
$ git svn branch -t --username=svnuser \
--commit-url=https://myproj.domain.com/svn mytag
Copying http://myproj.domain.com/svn/trunk at r26 to
https://myproj.domain.com/svn/tags/mytag...
Trying to use an unsupported feature: Source and dest appear not to be
in the same repository (src: 'http://myproj.domain.com/svn/trunk';
dst: 'https://myproj.domain.com/svn/tags/mytag')
[ew: shortened subject]
Signed-off-by: Igor Mironov <igor.a.mironov@gmail.com> Acked-by: Eric Wong <normalperson@yhbt.net>
Change git-svn to warn and continue when merge-base fails while processing svn
merge tickets.
merge-base can fail when a partial branch is created and merged back to trunk
in svn, because it cannot find a common ancestor between the partial branch and
trunk.
Signed-off-by: Andrew Myrick <amyrick@apple.com> Acked-by: Sam Vilain <sam@vilain.net> Acked-by: Eric Wong <normalperson@yhbt.net>
git-svn: ignore changeless commits when checking for a cherry-pick
Update git-svn to ignore commits that do not change the tree when it is
deciding if an svn merge ticket represents a real branch merge or just a
cherry-pick.
Consider the following integration model in the svn repository:
F---G branch1
/ \
D tag1 \ E tag2
/ \ /
A---B C trunk
branch1 is merged to trunk in commit C.
With this patch, git-svn will correctly identify branch1 as a proper merge
parent, instead of incorrectly ignoring it as a cherry-pick.
Signed-off-by: Andrew Myrick <amyrick@apple.com> Acked-by: Sam Vilain <sam@vilain.net> Acked-by: Eric Wong <normalperson@yhbt.net>
msvc: Add a definition of NORETURN compatible with msvc compiler
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Acked-by: Sebastian Schuberth <sschuberth@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
msvc: Fix a compiler warning due to an incorrect pointer cast
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Acked-by: Sebastian Schuberth <sschuberth@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Having recently added support for building git-imap-send on
Windows, we now link against OpenSSL libraries, and the linker
issues the following warning:
In order to suppress the warning, we change the msvc linker
script to translate an '-lssl' parameter to the ssleay32.lib
library.
Note that the linker script was already including ssleay32.lib
(along with libeay32.lib) as part of the translation of the
'-lcrypto' library parameter. However, libeay32.dll does not
depend on ssleay32.dll and can be used stand-alone, so we remove
ssleay32.lib from the '-lcrypto' translation.
The dependence of ssleay32.dll on libeay32.dll is represented in
the Makefile by the NEEDS_CRYPTO_WITH_SSL build variable.
Also, add the corresponding change to the buildsystem generator.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Acked-by: Sebastian Schuberth <sschuberth@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jl/submodule-diff:
Performance optimization for detection of modified submodules
git status: Show uncommitted submodule changes too when enabled
Teach diff that modified submodule directory is dirty
Show submodules as modified when they contain a dirty work tree
Make difftool.prompt fall back to mergetool.prompt
The documentation states that "git-difftool falls back to git-mergetool
config variables when the difftool equivalents have not been defined".
Until now, this was not the case for "difftool.prompt".
Signed-off-by: Sebastian Schuberth <sschuberth@gmail.com> Acked-by: David Aguilar <davvid@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
engine.pl: Fix a recent breakage of the buildsystem generator
Commit ade2ca0c (Do not try to remove directories when removing
old links, 2009-10-27) added an expression to a 'test' using an
'-o' or connective. This resulted in the buildsystem generator
mistaking a conditional 'rm' for a linker command. In order to
fix the breakage, we filter out all 'test' commands before then
attempting to identify the commands of interest.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Acked-by: Sebastian Schuberth <sschuberth@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The function takes two paths, an early part of abs is supposed to match
base; otherwise abs is not a path under base and the function returns the
full path of abs. The caller can easily confuse the implementation by
giving duplicated and needless slashes in these path arguments.
Credit for test script, motivation and initial patch goes to Thomas Rast.
A follow-up fix (squashed) is by Hannes.
git-mv: fix moving more than one source to a single destination
The code used as if return value from basename(3) were stable, but
often the function is implemented to return a pointer to a static
storage internal to it.
Because basename(3) is also allowed to modify its input parameter in
place, casting constness away from the strings we obtained from the
caller and giving them to basename is a no-no.
Reported, and initial fix and test supplied by David Rydh.
Documentation: improve description of --glob=pattern and friends
Consolidate the descriptions of --branches, --tags and --remotes a
bit, to make it less repetitive. Improve the grammar a bit, and spell
out the meaning of the 'append /*' rule.
Signed-off-by: Thomas Rast <trast@student.ethz.ch> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Exal Sibeaz pointed out that some git files are way too big, and that
add_files_to_cache() brings in all the diff machinery to any git binary
that needs the basic git SHA1 object operations from read-cache.c. Which
is pretty much all of them.
It's doubly silly, since add_files_to_cache() is only used by builtin
programs (add, checkout and commit), so it's fairly easily fixed by just
moving the thing to builtin-add.c, and avoiding the dependency entirely.
I initially argued to Exal that it would probably be best to try to depend
on smart compilers and linkers, but after spending some time trying to
make -ffunction-sections work and giving up, I think Exal was right, and
the fix is to just do some trivial cleanups like this.
This trivial cleanup results in pretty stunning file size differences.
The diff machinery really is mostly used by just the builtin programs, and
you have things like these trivial before-and-after numbers:
(Of course, stripping all the binaries brings the 33MB down to 9MB, so the
whole debug information thing is still the bulk of it all, but that's a
separate issue entirely)
Now, I'm sure there are other things we should do, and changing our
compiler flags from -O2 to -Os would bring the text size down by an
additional almost 20%, but this thing Exal pointed out seems to be some
good low-hanging fruit.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
merge-recursive: do not return NULL only to cause segfault
merge-recursive calls write_tree_from_memory() to come up with a virtual
tree, with possible conflict markers inside the blob contents, while
merging multiple common ancestors down. It is a bug to call the function
with unmerged entries in the index, even if the merge to come up with the
common ancestor resulted in conflicts. Otherwise the result won't be
expressible as a tree object.
We _might_ want to suggest the user to set GIT_MERGE_VERBOSITY to 5 and
re-run the merge in the message. At least we will know which part of
process_renames() or process_entry() functions is not correctly handling
the unmerged paths, and it might help us diagnosing the issue.