Junio C Hamano [Wed, 13 May 2015 04:28:54 +0000 (21:28 -0700)]
Merge branch 'mh/write-refs-sooner-2.3' into mh/write-refs-sooner-2.4
* mh/write-refs-sooner-2.3:
ref_transaction_commit(): fix atomicity and avoid fd exhaustion
ref_transaction_commit(): remove the local flags variable
ref_transaction_commit(): inline call to write_ref_sha1()
rename_ref(): inline calls to write_ref_sha1() from this function
commit_ref_update(): new function, extracted from write_ref_sha1()
write_ref_to_lockfile(): new function, extracted from write_ref_sha1()
t7004: rename ULIMIT test prerequisite to ULIMIT_STACK_SIZE
update-ref: test handling large transactions properly
Michael Haggerty [Fri, 24 Apr 2015 11:35:49 +0000 (13:35 +0200)]
ref_transaction_commit(): fix atomicity and avoid fd exhaustion
The old code was roughly
for update in updates:
acquire locks and check old_sha
for update in updates:
if changing value:
write_ref_to_lockfile()
commit_ref_update()
for update in updates:
if deleting value:
unlink()
rewrite packed-refs file
for update in updates:
if reference still locked:
unlock_ref()
This has two problems.
Non-atomic updates
==================
The atomicity of the reference transaction depends on all pre-checks
being done in the first loop, before any changes have started being
committed in the second loop. The problem is that
write_ref_to_lockfile() (previously part of write_ref_sha1()), which
is called from the second loop, contains two more checks:
* It verifies that new_sha1 is a valid object
* If the reference being updated is a branch, it verifies that
new_sha1 points at a commit object (as opposed to a tag, tree, or
blob).
If either of these checks fails, the "transaction" is aborted during
the second loop. But this might happen after some reference updates
have already been permanently committed. In other words, the
all-or-nothing promise of "git update-ref --stdin" could be violated.
So these checks have to be moved to the first loop.
File descriptor exhaustion
==========================
The old code locked all of the references in the first loop, leaving
all of the lockfiles open until later loops. Since we might be
updating a lot of references, this could result in file descriptor
exhaustion.
The solution
============
After this patch, the code looks like
for update in updates:
acquire locks and check old_sha
if changing value:
write_ref_to_lockfile()
else:
close_ref()
for update in updates:
if changing value:
commit_ref_update()
for update in updates:
if deleting value:
unlink()
rewrite packed-refs file
for update in updates:
if reference still locked:
unlock_ref()
This fixes both problems:
1. The pre-checks in write_ref_to_lockfile() are now done in the first
loop, before any changes have been committed. If any of the checks
fails, the whole transaction can now be rolled back correctly.
2. All lockfiles are closed in the first loop immediately after they
are created (either by write_ref_to_lockfile() or by close_ref()).
This means that there is never more than one open lockfile at a
time, preventing file descriptor exhaustion.
To simplify the bookkeeping across loops, add a new REF_NEEDS_COMMIT
bit to update->flags, which keeps track of whether the corresponding
lockfile needs to be committed, as opposed to just unlocked. (Since
"struct ref_update" is internal to the refs module, this change is not
visible to external callers.)
This change fixes two tests in t1400.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Fri, 24 Apr 2015 11:35:48 +0000 (13:35 +0200)]
ref_transaction_commit(): remove the local flags variable
Instead, work directly with update->flags. This has the advantage that
the REF_DELETING bit, set in the first loop, can be read in the second
loop instead of having to be recomputed. Plus, it was potentially
confusing having both update->flags and flags, which sometimes had
different values.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sat, 9 May 2015 15:29:20 +0000 (17:29 +0200)]
ref_transaction_commit(): inline call to write_ref_sha1()
That was the last caller, so delete function write_ref_sha1().
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sat, 9 May 2015 15:20:39 +0000 (17:20 +0200)]
rename_ref(): inline calls to write_ref_sha1() from this function
Most of what it does is unneeded from these call sites.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sat, 9 May 2015 15:18:36 +0000 (17:18 +0200)]
commit_ref_update(): new function, extracted from write_ref_sha1()
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Fri, 24 Apr 2015 11:35:45 +0000 (13:35 +0200)]
write_ref_to_lockfile(): new function, extracted from write_ref_sha1()
This is the first step towards separating the checking and writing of
the new reference value to committing the change.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Stefan Beller [Tue, 14 Apr 2015 22:25:07 +0000 (15:25 -0700)]
t7004: rename ULIMIT test prerequisite to ULIMIT_STACK_SIZE
During creation of the patch series our discussion we could have a
more descriptive name for the prerequisite for the test so it stays
unique when other limits of ulimit are introduced.
Signed-off-by: Stefan Beller <redacted>
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Stefan Beller [Tue, 14 Apr 2015 22:25:06 +0000 (15:25 -0700)]
update-ref: test handling large transactions properly
Signed-off-by: Stefan Beller <redacted>
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Wed, 13 May 2015 04:26:09 +0000 (21:26 -0700)]
Merge branch 'mh/write-refs-sooner-2.2' into mh/write-refs-sooner-2.3
* mh/write-refs-sooner-2.2:
ref_transaction_commit(): fix atomicity and avoid fd exhaustion
ref_transaction_commit(): remove the local flags variable
ref_transaction_commit(): inline call to write_ref_sha1()
rename_ref(): inline calls to write_ref_sha1() from this function
commit_ref_update(): new function, extracted from write_ref_sha1()
write_ref_to_lockfile(): new function, extracted from write_ref_sha1()
t7004: rename ULIMIT test prerequisite to ULIMIT_STACK_SIZE
update-ref: test handling large transactions properly
Michael Haggerty [Sun, 10 May 2015 02:45:37 +0000 (04:45 +0200)]
ref_transaction_commit(): fix atomicity and avoid fd exhaustion
The old code was roughly
for update in updates:
acquire locks and check old_sha
for update in updates:
if changing value:
write_ref_to_lockfile()
commit_ref_update()
for update in updates:
if deleting value:
unlink()
rewrite packed-refs file
for update in updates:
if reference still locked:
unlock_ref()
This has two problems.
Non-atomic updates
==================
The atomicity of the reference transaction depends on all pre-checks
being done in the first loop, before any changes have started being
committed in the second loop. The problem is that
write_ref_to_lockfile() (previously part of write_ref_sha1()), which
is called from the second loop, contains two more checks:
* It verifies that new_sha1 is a valid object
* If the reference being updated is a branch, it verifies that
new_sha1 points at a commit object (as opposed to a tag, tree, or
blob).
If either of these checks fails, the "transaction" is aborted during
the second loop. But this might happen after some reference updates
have already been permanently committed. In other words, the
all-or-nothing promise of "git update-ref --stdin" could be violated.
So these checks have to be moved to the first loop.
File descriptor exhaustion
==========================
The old code locked all of the references in the first loop, leaving
all of the lockfiles open until later loops. Since we might be
updating a lot of references, this could result in file descriptor
exhaustion.
The solution
============
After this patch, the code looks like
for update in updates:
acquire locks and check old_sha
if changing value:
write_ref_to_lockfile()
else:
close_ref()
for update in updates:
if changing value:
commit_ref_update()
for update in updates:
if deleting value:
unlink()
rewrite packed-refs file
for update in updates:
if reference still locked:
unlock_ref()
This fixes both problems:
1. The pre-checks in write_ref_to_lockfile() are now done in the first
loop, before any changes have been committed. If any of the checks
fails, the whole transaction can now be rolled back correctly.
2. All lockfiles are closed in the first loop immediately after they
are created (either by write_ref_to_lockfile() or by close_ref()).
This means that there is never more than one open lockfile at a
time, preventing file descriptor exhaustion.
To simplify the bookkeeping across loops, add a new REF_NEEDS_COMMIT
bit to update->flags, which keeps track of whether the corresponding
lockfile needs to be committed, as opposed to just unlocked. (Since
"struct ref_update" is internal to the refs module, this change is not
visible to external callers.)
This change fixes two tests in t1400.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sun, 10 May 2015 02:45:36 +0000 (04:45 +0200)]
ref_transaction_commit(): remove the local flags variable
Instead, work directly with update->flags. This has the advantage that
the REF_DELETING bit, set in the first loop, can be read in the second
loop instead of having to be recomputed. Plus, it was potentially
confusing having both update->flags and flags, which sometimes had
different values.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sun, 10 May 2015 02:45:35 +0000 (04:45 +0200)]
ref_transaction_commit(): inline call to write_ref_sha1()
And remove the function write_ref_sha1(), as it is no longer used.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sun, 10 May 2015 02:45:34 +0000 (04:45 +0200)]
rename_ref(): inline calls to write_ref_sha1() from this function
Most of what it does is unneeded from these call sites.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sun, 10 May 2015 02:45:33 +0000 (04:45 +0200)]
commit_ref_update(): new function, extracted from write_ref_sha1()
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Sun, 10 May 2015 02:45:32 +0000 (04:45 +0200)]
write_ref_to_lockfile(): new function, extracted from write_ref_sha1()
This is the first step towards separating the checking and writing of
the new reference value to committing the change.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Stefan Beller [Sun, 10 May 2015 02:45:31 +0000 (04:45 +0200)]
t7004: rename ULIMIT test prerequisite to ULIMIT_STACK_SIZE
During creation of the patch series, our discussion revealed that
we could have a more descriptive name for the prerequisite for the
test so it stays unique when other limits of ulimit are introduced.
Let's rename the existing ulimit about setting the stack size to
a more explicit ULIMIT_STACK_SIZE.
Signed-off-by: Stefan Beller <redacted>
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Stefan Beller [Sun, 10 May 2015 02:45:30 +0000 (04:45 +0200)]
update-ref: test handling large transactions properly
Signed-off-by: Stefan Beller <redacted>
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Thu, 30 Apr 2015 18:25:06 +0000 (11:25 -0700)]
Git 2.4
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Tue, 28 Apr 2015 20:01:29 +0000 (13:01 -0700)]
Merge branch 'mh/multimail-renewal'
* mh/multimail-renewal:
Update git-multimail to version 1.0.2
Junio C Hamano [Tue, 28 Apr 2015 20:00:20 +0000 (13:00 -0700)]
Merge branch 'mg/show-notes-doc'
Documentation fix.
* mg/show-notes-doc:
rev-list-options.txt: complete sentence about notes matching
Junio C Hamano [Tue, 28 Apr 2015 20:00:19 +0000 (13:00 -0700)]
Merge branch 'nd/versioncmp-prereleases'
* nd/versioncmp-prereleases:
git tag: mention versionsort.prereleaseSuffix in manpage
Junio C Hamano [Tue, 28 Apr 2015 20:00:18 +0000 (13:00 -0700)]
Merge branch 'mg/status-v-v'
* mg/status-v-v:
status: document the -v/--verbose option
Michael Haggerty [Mon, 27 Apr 2015 11:17:25 +0000 (13:17 +0200)]
Update git-multimail to version 1.0.2
The only changes are to the README files, most notably the list of
maintainers and the project URL.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Mon, 27 Apr 2015 19:26:21 +0000 (12:26 -0700)]
Sync with 2.3.7
Junio C Hamano [Mon, 27 Apr 2015 19:25:36 +0000 (12:25 -0700)]
Git 2.3.7
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Mon, 27 Apr 2015 19:23:53 +0000 (12:23 -0700)]
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
Junio C Hamano [Mon, 27 Apr 2015 19:23:51 +0000 (12:23 -0700)]
Merge branch 'ma/bash-completion-leaking-x' into maint
The completion script (in contrib/) contaminated global namespace
and clobbered on a shell variable $x.
* ma/bash-completion-leaking-x:
completion: fix global bash variable leak on __gitcompappend
Junio C Hamano [Mon, 27 Apr 2015 19:23:47 +0000 (12:23 -0700)]
Merge branch 'jc/push-cert' into maint
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
Michael Haggerty [Thu, 23 Apr 2015 12:27:46 +0000 (14:27 +0200)]
status: document the -v/--verbose option
Document `git status -v`, including its new doubled `-vv` form.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Thu, 23 Apr 2015 12:27:50 +0000 (14:27 +0200)]
RelNotes: wordsmithing
Make many textual tweaks to the 2.4.0 release notes.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Thu, 23 Apr 2015 12:27:49 +0000 (14:27 +0200)]
RelNotes: refer to the rebase -i "todo list", not "insn sheet"
"Todo list" is the name that is used in the user-facing documentation.
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Thu, 23 Apr 2015 12:27:48 +0000 (14:27 +0200)]
RelNotes: correct name of versionsort.prereleaseSuffix
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Michael Haggerty [Thu, 23 Apr 2015 12:27:47 +0000 (14:27 +0200)]
git tag: mention versionsort.prereleaseSuffix in manpage
Signed-off-by: Michael Haggerty <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Wed, 22 Apr 2015 20:52:43 +0000 (13:52 -0700)]
Git 2.4.0-rc3
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Tue, 21 Apr 2015 19:58:50 +0000 (12:58 -0700)]
Sync with maint
Junio C Hamano [Tue, 21 Apr 2015 19:17:09 +0000 (12:17 -0700)]
Git 2.3.6
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Tue, 21 Apr 2015 19:12:25 +0000 (12:12 -0700)]
Merge branch 'jk/colors' into maint
"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
Junio C Hamano [Tue, 21 Apr 2015 19:12:24 +0000 (12:12 -0700)]
Merge branch 'jk/test-annoyances' into maint
Test fixes.
* 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
Junio C Hamano [Tue, 21 Apr 2015 19:12:23 +0000 (12:12 -0700)]
Merge branch 'pt/enter-repo-comment-fix' into maint
Documentation update.
* pt/enter-repo-comment-fix:
enter_repo(): fix docs to match code
Junio C Hamano [Tue, 21 Apr 2015 19:12:22 +0000 (12:12 -0700)]
Merge branch 'jz/gitweb-conf-doc-fix' into maint
Documentation update.
* jz/gitweb-conf-doc-fix:
gitweb.conf.txt: say "build-time", not "built-time"
Junio C Hamano [Tue, 21 Apr 2015 19:12:21 +0000 (12:12 -0700)]
Merge branch 'jk/cherry-pick-docfix' into maint
* jk/cherry-pick-docfix:
cherry-pick: fix docs describing handling of empty commits
Junio C Hamano [Tue, 21 Apr 2015 19:12:20 +0000 (12:12 -0700)]
Merge branch 'iu/fix-parse-options-h-comment' into maint
* iu/fix-parse-options-h-comment:
parse-options.h: OPTION_{BIT,SET_INT} do not store pointer to defval
Junio C Hamano [Tue, 21 Apr 2015 19:12:19 +0000 (12:12 -0700)]
Merge branch 'jg/cguide-we-cannot-count' into maint
* jg/cguide-we-cannot-count:
CodingGuidelines: update 'rough' rule count
Junio C Hamano [Tue, 21 Apr 2015 19:12:18 +0000 (12:12 -0700)]
Merge branch 'jk/pack-corruption-post-mortem' into maint
Documentation update.
* jk/pack-corruption-post-mortem:
howto: document more tools for recovery corruption
Junio C Hamano [Tue, 21 Apr 2015 19:12:17 +0000 (12:12 -0700)]
Merge branch 'jn/doc-fast-import-no-16-octopus-limit' into maint
Documentation update.
* jn/doc-fast-import-no-16-octopus-limit:
fast-import doc: remove suggested 16-parent limit
Junio C Hamano [Tue, 21 Apr 2015 18:09:19 +0000 (11:09 -0700)]
RelNotes: "merge --quiet" change has been reverted
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Mon, 20 Apr 2015 22:30:13 +0000 (15:30 -0700)]
Hopefully the last batch for 2.4
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Mon, 20 Apr 2015 22:28:34 +0000 (15:28 -0700)]
Merge branch 'ps/grep-help-all-callback-arg'
Code clean-up.
* ps/grep-help-all-callback-arg:
grep: correctly initialize help-all option
Junio C Hamano [Mon, 20 Apr 2015 22:28:33 +0000 (15:28 -0700)]
Merge branch 'tb/connect-ipv6-parse-fix'
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
Junio C Hamano [Mon, 20 Apr 2015 22:28:32 +0000 (15:28 -0700)]
Merge branch 'va/fix-git-p4-tests'
Test fixes for git-p4.
* 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
Junio C Hamano [Mon, 20 Apr 2015 22:28:31 +0000 (15:28 -0700)]
Merge branch 'jc/push-cert'
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
Junio C Hamano [Mon, 20 Apr 2015 22:28:29 +0000 (15:28 -0700)]
Merge branch 'ma/bash-completion-leaking-x'
The completion script (in contrib/) contaminated global namespace
and clobbered on a shell variable $x.
* ma/bash-completion-leaking-x:
completion: fix global bash variable leak on __gitcompappend
Junio C Hamano [Sun, 19 Apr 2015 01:35:48 +0000 (18:35 -0700)]
Merge tag 'gitgui-0.20.0' of repo.or.cz/r/git-gui
git-gui 0.20.0
* 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.
Pat Thoyts [Sat, 18 Apr 2015 11:15:32 +0000 (12:15 +0100)]
git-gui: set version 0.20
Signed-off-by: Pat Thoyts <redacted>
Peter Krefting [Fri, 27 Mar 2015 09:25:32 +0000 (10:25 +0100)]
git-gui: sv.po: Update Swedish translation (547t0f0u)
Signed-off-by: Peter Krefting <redacted>
Signed-off-by: Pat Thoyts <redacted>
Alexander Shopov [Tue, 7 Apr 2015 15:29:46 +0000 (18:29 +0300)]
git-gui i18n: Updated Bulgarian translation (547t,0f,0u)
Signed-off-by: Alexander Shopov <redacted>
Signed-off-by: Pat Thoyts <redacted>
Michael J Gruber [Fri, 17 Apr 2015 14:28:56 +0000 (16:28 +0200)]
rev-list-options.txt: complete sentence about notes matching
Signed-off-by: Michael J Gruber <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Thu, 16 Apr 2015 15:03:14 +0000 (08:03 -0700)]
Revert "merge: pass verbosity flag down to merge-recursive"
This reverts commit
2bf15a3330a26183adc8563dbeeacc11294b8a01, whose
intention was good, but the verbosity levels used in merge-recursive
turns out to be rather uneven. For example, a merge of two branches
with conflicting submodule updates used to report CONFLICT: output
with --quiet but no longer (which *is* desired), while the final
"Automatic merge failed; fix conflicts and then commit" message is
still shown even with --quiet (which *is* inconsistent).
Originally reported by Bryan Turner; it is too early to declare what
the concensus is, but it seems that we would need to level the
verbosity levels used in merge strategy backends before we can go
forward. In the meantime, we'd revert to the old behaviour until
that happens.
cf. $gmane/267245
Junio C Hamano [Tue, 14 Apr 2015 18:57:13 +0000 (11:57 -0700)]
Git 2.4.0-rc2
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Tue, 14 Apr 2015 18:49:13 +0000 (11:49 -0700)]
Merge branch 'jk/colors'
"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
Junio C Hamano [Tue, 14 Apr 2015 18:49:12 +0000 (11:49 -0700)]
Merge branch 'jk/merge-quiet'
"git merge --quiet" did not squelch messages from the underlying
merge-recursive strategy.
* jk/merge-quiet:
merge: pass verbosity flag down to merge-recursive
Junio C Hamano [Tue, 14 Apr 2015 18:49:11 +0000 (11:49 -0700)]
Merge branch 'jk/pack-corruption-post-mortem'
Documentation update.
* jk/pack-corruption-post-mortem:
howto: document more tools for recovery corruption
Junio C Hamano [Tue, 14 Apr 2015 18:49:10 +0000 (11:49 -0700)]
Merge branch 'jc/update-instead-into-void'
A push into an unborn branch, with "receive.denyCurrentBranch" set
to "updateInstead", did not check out the working tree as expected.
* jc/update-instead-into-void:
push-to-deploy: allow pushing into an unborn branch and updating it
Junio C Hamano [Tue, 14 Apr 2015 18:49:09 +0000 (11:49 -0700)]
Merge branch 'sb/plug-streaming-leak'
* sb/plug-streaming-leak:
streaming.c: fix a memleak
Junio C Hamano [Tue, 14 Apr 2015 18:49:08 +0000 (11:49 -0700)]
Merge branch 'jn/doc-fast-import-no-16-octopus-limit'
Documentation update.
* jn/doc-fast-import-no-16-octopus-limit:
fast-import doc: remove suggested 16-parent limit
Junio C Hamano [Tue, 14 Apr 2015 18:49:07 +0000 (11:49 -0700)]
Merge branch 'sb/plug-wt-shortstatus-tracking-leak'
* sb/plug-wt-shortstatus-tracking-leak:
wt-status.c: fix a memleak
Junio C Hamano [Tue, 14 Apr 2015 17:34:05 +0000 (10:34 -0700)]
Merge branch 'pt/enter-repo-comment-fix'
* pt/enter-repo-comment-fix:
enter_repo(): fix docs to match code
Junio C Hamano [Tue, 14 Apr 2015 17:34:01 +0000 (10:34 -0700)]
Merge branch 'jz/gitweb-conf-doc-fix'
* jz/gitweb-conf-doc-fix:
gitweb.conf.txt: say "build-time", not "built-time"
Junio C Hamano [Tue, 14 Apr 2015 17:33:54 +0000 (10:33 -0700)]
Merge branch 'jk/cherry-pick-docfix'
* jk/cherry-pick-docfix:
cherry-pick: fix docs describing handling of empty commits
Junio C Hamano [Tue, 14 Apr 2015 17:33:45 +0000 (10:33 -0700)]
Merge branch 'iu/fix-parse-options-h-comment'
* iu/fix-parse-options-h-comment:
parse-options.h: OPTION_{BIT,SET_INT} do not store pointer to defval
Junio C Hamano [Tue, 14 Apr 2015 16:55:29 +0000 (09:55 -0700)]
Merge branch 'jg/cguide-we-cannot-count'
* jg/cguide-we-cannot-count:
CodingGuidelines: update 'rough' rule count
Julian Gindi [Mon, 13 Apr 2015 12:54:14 +0000 (08:54 -0400)]
CodingGuidelines: update 'rough' rule count
Changed inaccurate count of "rough rules" from three to the more
generic 'a few'.
Signed-off-by: Julian Gindi <redacted>
Reviewed-by: Eric Sunshine <redacted>
Reviewed-by: Jonathan Nieder <redacted>
Signed-off-by: Junio C Hamano <redacted>
Patrick Steinhardt [Thu, 9 Apr 2015 19:59:06 +0000 (21:59 +0200)]
grep: correctly initialize help-all option
The "help-all" option is being initialized with a wrong value.
While being semantically wrong this can also cause a segmentation
fault in gcc on ARMv7 hardfloat platforms with a hardened
toolchain. Fix this by initializing with a NULL value.
Signed-off-by: Patrick Steinhardt <redacted>
Reviewed-by: René Scharfe <redacted>
Signed-off-by: Junio C Hamano <redacted>
Márcio Almada [Wed, 8 Apr 2015 05:45:58 +0000 (02:45 -0300)]
completion: fix global bash variable leak on __gitcompappend
Signed-off-by: Junio C Hamano <redacted>
Vitor Antunes [Sun, 5 Apr 2015 23:08:35 +0000 (00:08 +0100)]
t9814: guarantee only one source exists in git-p4 copy tests
By using a tree with multiple identical files and allowing copy detection to
choose any one of them, the check in the test is unnecessarily complex. We can
simplify by:
* Modify source file (file2) before copying the file.
* Check that only file2 is the source in the output of "p4 filelog".
* Remove all "case" statements and replace them with simple tests to check
that source is "file2".
Signed-off-by: Vitor Antunes <redacted>
Acked-by: Luke Diamand <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Thu, 9 Apr 2015 04:15:33 +0000 (21:15 -0700)]
Merge tag 'l10n-2.4.0-rnd2' of git://github.com/git-l10n/git-po
git l10n 2.4.0-rnd2
Torsten Bögershausen [Tue, 7 Apr 2015 20:03:25 +0000 (22:03 +0200)]
connect.c: ignore extra colon after hostname
Ignore an extra ':' at the end of the hostname in URL's like
"ssh://example.com:/path/to/repo"
The colon is meant to separate a port number from the hostname.
If the port is empty, the colon should be ignored, see RFC 3986.
It had been working for URLs with ssh:// scheme, but was unintentionally
broken in 86ceb3, "allow ssh://user@[2001:db8::1]/repo.git"
Reported-by: Reid Woodbury Jr. <redacted>
Signed-off-by: Torsten Bögershausen <redacted>
Signed-off-by: Junio C Hamano <redacted>
Jiang Xin [Thu, 9 Apr 2015 00:21:39 +0000 (08:21 +0800)]
Merge branch 'master' of git://github.com/alexhenrie/git-po
* 'master' of git://github.com/alexhenrie/git-po:
l10n: ca.po: update translation
Jiang Xin [Thu, 9 Apr 2015 00:00:10 +0000 (08:00 +0800)]
l10n: TEAMS: Change repository URL of zh_CN
Repository URL of zh_CN l10n for Git has been changed over 2 years,
update po/TEAMS for it.
Signed-off-by: Jiang Xin <redacted>
Alex Henrie [Wed, 8 Apr 2015 00:07:47 +0000 (18:07 -0600)]
l10n: ca.po: update translation
Signed-off-by: Alex Henrie <redacted>
Alexander Shopov [Tue, 7 Apr 2015 08:22:47 +0000 (11:22 +0300)]
l10n: Updated Bulgarian translation of git (2305t,0f,0u)
Signed-off-by: Alexander Shopov <redacted>
Peter Krefting [Tue, 7 Apr 2015 07:45:20 +0000 (08:45 +0100)]
l10n: sv.po: Update Swedish translation (2305t0f0u)
Signed-off-by: Peter Krefting <redacted>
Ralf Thielow [Sat, 4 Apr 2015 08:06:42 +0000 (10:06 +0200)]
l10n: de.po: translate one message
Translate one message came from git.pot update in
6eebb35
(l10n: git.pot: v2.4.0 round 2 (1 update)).
Signed-off-by: Ralf Thielow <redacted>
Kyle J. McKay [Fri, 3 Apr 2015 22:15:14 +0000 (15:15 -0700)]
diff-highlight: do not split multibyte characters
When the input is UTF-8 and Perl is operating on bytes instead of
characters, a diff that changes one multibyte character to another
that shares an initial byte sequence will result in a broken diff
display as the common byte sequence prefix will be separated from
the rest of the bytes in the multibyte character.
For example, if a single line contains only the unicode character
U+C9C4 (encoded as UTF-8 0xEC, 0xA7, 0x84) and that line is then
changed to the unicode character U+C9C0 (encoded as UTF-8 0xEC,
0xA7, 0x80), when operating on bytes diff-highlight will show only
the single byte change from 0x84 to 0x80 thus creating invalid UTF-8
and a broken diff display.
Fix this by putting Perl into character mode when splitting the line
and then back into byte mode after the split is finished.
The utf8::xxx functions require Perl 5.8 so we require that as well.
Also, since we are mucking with code in the split_line function, we
change a '*' quantifier to a '+' quantifier when matching the $COLOR
expression which has the side effect of speeding everything up while
eliminating useless '' elements in the returned array.
Reported-by: Yi EungJun <redacted>
Signed-off-by: Kyle J. McKay <redacted>
Acked-by: Jeff King <redacted>
Signed-off-by: Junio C Hamano <redacted>
Jean-Noel Avila [Sat, 4 Apr 2015 16:04:56 +0000 (18:04 +0200)]
l10n: fr.po v2.4.0 round 2
Signed-off-by: Jean-Noel Avila <redacted>
Dimitriy Ryazantcev [Fri, 3 Apr 2015 13:30:14 +0000 (16:30 +0300)]
l10n: ru: updated Russian translation
Signed-off-by: Dimitriy Ryazantcev <redacted>
Tran Ngoc Quan [Fri, 3 Apr 2015 01:54:35 +0000 (08:54 +0700)]
l10n: vi.po(2305t): Updated 1 new string
Signed-off-by: Tran Ngoc Quan <redacted>
Jiang Xin [Fri, 3 Apr 2015 00:39:57 +0000 (08:39 +0800)]
l10n: zh_CN: for git v2.4.0 l10n round 2
Translate 1 update message (2305t0f0u) for git v2.4.0 l10n round 2.
Signed-off-by: Jiang Xin <redacted>
Jiang Xin [Fri, 3 Apr 2015 00:16:53 +0000 (08:16 +0800)]
l10n: git.pot: v2.4.0 round 2 (1 update)
Generate po/git.pot from v2.4.0-rc1 for git v2.4.0 l10n round 2.
Signed-off-by: Jiang Xin <redacted>
Jiang Xin [Fri, 3 Apr 2015 00:13:24 +0000 (08:13 +0800)]
Merge branch 'master' of git://github.com/git-l10n/git-po
* 'master' of git://github.com/git-l10n/git-po:
l10n: de.po: translate 'symbolic link' as 'symbolische Verknüpfung'
l10n: de.po: translate 99 new messages
l10n: de.po: fix messages with abbreviated hashs
l10n: de.po: add space before ellipsis
l10n: vi.po: Updated Vietnamese translation
l10n: zh_CN: translations for git v2.4.0-rc0
l10n: fr.po v2.4.0-rc0 round 1
l10n: ca.po: update translation
l10n: ru: updated Russian translation
l10n: sv.po: Update Swedish translation (2305t0f0u)
l10n: git.pot: v2.4.0 round 1 (99 new, 92 removed)
l10n: ru: added Russian translation
l10n: de.po: fix negation for commit -a with paths
Jeff King [Thu, 2 Apr 2015 21:39:52 +0000 (17:39 -0400)]
merge: pass verbosity flag down to merge-recursive
This makes "git merge --quiet" really quiet when we call
into merge-recursive.
Note that we can't just pass our flag down as-is; the two
parts of the code use different scales. We center at "0" as
normal for git-merge (with "--quiet" giving a negative
value), but merge-recursive uses "2" as its center. This
patch passes a negative value to merge-recursive rather than
"1", though, as otherwise the user would have to use "-qqq"
to squelch all messages (but the downside is that the user
cannot distinguish between levels 0-2 if without resorting
to the GIT_MERGE_VERBOSITY variable).
We may want to review and renormalize the message severities
in merge-recursive, but that does not have to happen now.
This is at least in improvement in the sense that we are
respecting "--quiet" at all.
Signed-off-by: Jeff King <redacted>
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Thu, 2 Apr 2015 19:46:06 +0000 (12:46 -0700)]
Git 2.4.0-rc1
Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Thu, 2 Apr 2015 19:34:43 +0000 (12:34 -0700)]
Merge branch 'sb/atomic-push'
* sb/atomic-push:
send-pack: unify error messages for unsupported capabilities
Junio C Hamano [Thu, 2 Apr 2015 01:00:36 +0000 (18:00 -0700)]
push --signed: tighten what the receiving end can ask to sign
Instead of blindly trusting the receiving side to give us a sensible
nonce to sign, limit the length (max 256 bytes) and the alphabet
(alnum and a few selected punctuations, enough to encode in base64)
that can be used in nonce.
Signed-off-by: Junio C Hamano <redacted>
Ralf Thielow [Thu, 2 Apr 2015 17:28:48 +0000 (19:28 +0200)]
send-pack: unify error messages for unsupported capabilities
If --signed is not supported, the error message names the remote
"receiving end". If --atomic is not supported, the error message
names the remote "server". Unify the naming to "receiving end"
as we're in the context of "push".
Signed-off-by: Ralf Thielow <redacted>
Signed-off-by: Junio C Hamano <redacted>
Matthias Rüster [Sun, 29 Mar 2015 01:46:32 +0000 (03:46 +0200)]
l10n: de.po: translate 'symbolic link' as 'symbolische Verknüpfung'
The use of 'symbolische Verknüpfung' for 'symbolic link' is more common
than 'symbolischer Verweis'.
Signed-off-by: Matthias Rüster <redacted>
Signed-off-by: Ralf Thielow <redacted>
Ralf Thielow [Fri, 27 Mar 2015 15:58:26 +0000 (16:58 +0100)]
l10n: de.po: translate 99 new messages
Translate 99 messages came from git.pot update in
c2ea120
(l10n: git.pot: v2.4.0 round 1 (99 new, 92 removed)).
Signed-off-by: Ralf Thielow <redacted>
Ralf Thielow [Tue, 24 Mar 2015 18:37:47 +0000 (19:37 +0100)]
l10n: de.po: fix messages with abbreviated hashs
The three dots in messages where the hash is abbreviated
were misinterpreted and are fixed with this commit.
Noticed-by: Junio C Hamano <redacted>
Signed-off-by: Ralf Thielow <redacted>
Phillip Sz [Sat, 21 Mar 2015 12:52:37 +0000 (13:52 +0100)]
l10n: de.po: add space before ellipsis
Signed-off-by: Phillip Sz <redacted>
Signed-off-by: Ralf Thielow <redacted>