4534f49985 · matheusbsbs/matheus-bagliano

@@ -1,20 +1,18 @@

1 To: git@vger.kernel.org

2 Bcc: lwn@lwn.net

3 -Subject: What's cooking in git.git (Sep 2015, #02; Wed, 9)

4 -X-master-at: 7a2c4af7a82eebf2ed0e60a2b36c31e70c619264

5 -X-next-at: 8a6dc4c4eedd63a3d9286c8083e0c7f522faa35b

3 +Subject: What's cooking in git.git (Sep 2015, #03; Mon, 14)

4 +X-master-at: f4d9753a89bf04011c00e943d85211906e86a0f6

5 +X-next-at: 924a60862de14449284d454b8dd848b57a8e3f75

6

7 -What's cooking in git.git (Sep 2015, #02; Wed, 9)

7 +What's cooking in git.git (Sep 2015, #03; Mon, 14)

8 --------------------------------------------------

9

10 Here are the topics that have been cooking. Commits prefixed with

11 '-' are only in 'pu' (proposed updates) while commits prefixed with

12 '+' are in 'next'.

13

14 -The first release candidate 2.6-rc1 has been tagged. Let's make

15 -sure we catch any possible regressions from now on, without getting

16 -distracted by yet more shiny new features. On the 'maint' front,

17 -we'll have 2.5.2 tomorrow.

14 +The tip of 'master' is slightly past 2.6-rc2; hopefully l10n team

15 +can work on this without updates before the end of the cycle.

16

17 You can find the changes described here in the integration branches

18 of the repositories listed at

@@ -24,137 +22,154 @@ of the repositories listed at

22 --------------------------------------------------

23 [Graduated to "master"]

24

27 -* jc/builtin-am-signoff-regression-fix (2015-09-06) 1 commit

28 - (merged to 'next' on 2015-09-06 at 7e84b6f)

29 - + am: match --signoff to the original scripted version

25 +* ah/show-ref-usage-string (2015-08-31) 1 commit

26 + (merged to 'next' on 2015-09-08 at dc512c8)

27 + + show-ref: place angle brackets around variables in usage string

28

31 - Recent "git am" had regression when adding a Signed-off-by line

32 - with its "-s" option by an unintended tightening of how an existing

33 - trailer block is detected.

29 + Both "git show-ref -h" and "git show-ref --help" illustrated that the

30 + "--exclude-existing" option makes the command read list of refs

31 + from its standard input. Change only the "show-ref -h" output to

32 + have a pair of "<>" around the placeholder that designate an input

33 + file, i.e. "git show-ref --exclude-existing < <ref-list>".

34

35

36 -* jk/am-rerere-lock-fix (2015-09-01) 1 commit

37 - (merged to 'next' on 2015-09-02 at 686c5c0)

38 - + rerere: release lockfile in non-writing functions

36 +* gb/apply-comment-typofix (2015-09-09) 1 commit

37 + (merged to 'next' on 2015-09-09 at 29fb0ff)

38 + + apply: comment grammar fix

39

40 - Recent "git am" introduced a double-locking failure when used with

41 - the "--3way" option that invokes rerere machinery.

40

43 ---------------------------------------------------

44 -[New Topics]

41 +* jk/pack-protocol-doc (2015-09-03) 1 commit

42 + (merged to 'next' on 2015-09-08 at 98d861d)

43 + + pack-protocol: clarify LF-handling in PKT-LINE()

44

46 -* jk/connect-clear-env (2015-09-08) 2 commits

47 - (merged to 'next' on 2015-09-08 at 86b64f7)

48 - + git_connect: clarify conn->use_shell flag

49 - (merged to 'next' on 2015-09-06 at 9ddbed8)

50 - + git_connect: clear GIT_* environment for ssh

45 + Streamline documentation of the pkt-line protocol.

46

52 - The ssh transport, just like any other transport over the network,

53 - did not clear GIT_* environment variables, but it is possible to

54 - use SendEnv and AcceptEnv to leak them to the remote invocation of

55 - Git, which is not a good idea at all. Explicitly clear them just

56 - like we do for the local transport.

47

58 - Will merge to 'master'.

48 +* js/maint-am-skip-performance-regression (2015-09-09) 1 commit

49 + (merged to 'next' on 2015-09-14 at b669ea1)

50 + + am --skip/--abort: merge HEAD/ORIG_HEAD tree into index

51

52 + Later versions of scripted "git am" has a performance regression in

53 + "git am --skip" codepath, which no longer exists in the built-in

54 + version on the 'master' front. Fix the regression in the last

55 + scripted version that appear in 2.5.x maintenance track and older.

56

61 -* jk/pack-protocol-doc (2015-09-03) 1 commit

62 - (merged to 'next' on 2015-09-08 at 98d861d)

63 - + pack-protocol: clarify LF-handling in PKT-LINE()

57 + Will merge to 'maint' and older.

58

65 - Streamline documentation of the pkt-line protocol.

59

67 - Will merge to 'master'.

60 +* mp/t7060-diff-index-test (2015-09-02) 1 commit

61 + (merged to 'next' on 2015-09-08 at 80ff284)

62 + + t7060: actually test "git diff-index --cached -M"

63

64 + Fix an old test that was doing the same thing as another one.

65

70 -* ti/glibc-stdio-mutex-from-signal-handler (2015-09-04) 1 commit

71 - (merged to 'next' on 2015-09-08 at c8047ba)

72 - + pager: don't use unsafe functions in signal handlers

66

74 - Allocation related functions and stdio are unsafe things to call

75 - inside a signal handler, and indeed killing the pager can cause

76 - glibc to deadlock waiting on allocation mutex as our signal handler

77 - tries to free() some data structures in wait_for_pager(). Reduce

78 - these unsafe calls.

67 +* rt/help-strings-fix (2015-09-11) 2 commits

68 + (merged to 'next' on 2015-09-14 at 875908c)

69 + + tag, update-ref: improve description of option "create-reflog"

70 + + pull: don't mark values for option "rebase" for translation

71 +

72

80 - I seem to recall Peff had other ideas? Let's revisit this in the

81 - next cycle.

73 +* sg/help-group (2015-09-10) 1 commit

74 + (merged to 'next' on 2015-09-14 at cef17e2)

75 + + Makefile: use SHELL_PATH when running generate-cmdlist.sh

76

77 + Last minute regression fix.

78

84 -* as/subtree-with-spaces (2015-09-08) 2 commits

85 - (merged to 'next' on 2015-09-09 at f3d763d)

86 - + contrib/subtree: respect spaces in a repository path

87 - + t7900-subtree: test the "space in a subdirectory name" case

79 +--------------------------------------------------

80 +[New Topics]

81

89 - Update "git subtree" (in contrib/) so that it can take whitespaces

90 - in the pathnames, not only in the in-tree pathname but the name of

91 - the directory that the repository is in.

82 +* ld/p4-detached-head (2015-09-09) 2 commits

83 + - git-p4: work with a detached head

84 + - git-p4: add failing test for submit from detached head

85 +

86 + Will be rerolled.

87 + ($gmane/277574)

88 +

89 +

90 +* as/config-doc-markup-fix (2015-09-12) 1 commit

91 + (merged to 'next' on 2015-09-14 at faa4134)

92 + + Documentation/config: fix formatting for branch.*.rebase and pull.rebase

93

94 Will merge to 'master'.

95

96

96 -* gb/apply-comment-typofix (2015-09-09) 1 commit

97 - (merged to 'next' on 2015-09-09 at 29fb0ff)

98 - + apply: comment grammar fix

97 +* et/win32-poll-timeout (2015-09-14) 1 commit

98 + (merged to 'next' on 2015-09-14 at 308dff4)

99 + + poll: honor the timeout on Win32

100

101 Will merge to 'master'.

102

103

103 -* gb/filter-branch-progress (2015-09-08) 1 commit

104 - - filter-branch: add passed/remaining seconds on progress

104 +* jk/graph-format-padding (2015-09-14) 1 commit

105 + - pretty: pass graph width to pretty formatting for use in '%>|(N)'

106

106 - Give progress meter to "git filter-branch".

107 + Redefine the way '%>|(N)' padding and the "--graph" option

108 + interacts. It has been that the available columns to display the

109 + log message was measured from the edge of the area the graph ended,

110 + but with this it becomes the beginning of the entire output.

111

108 - Is a reroll coming?

112 + I have a suspicion that 50% of the users would appreciate this

113 + change, and the remainder see this break their expectation. If

114 + that is the case, we might need to introduce a similar but

115 + different alignment operator so that this new behaviour is

116 + available to those who want to use it, without negatively affecting

117 + existing uses.

118

119

111 -* jk/interpret-trailers-outside-a-repository (2015-09-08) 1 commit

112 - (merged to 'next' on 2015-09-09 at a75fce3)

113 - + interpret-trailers: allow running outside a repository

120 +* jk/make-findstring-makeflags-fix (2015-09-10) 1 commit

121 + - Makefile: fix MAKEFLAGS tests with multiple flags

122

115 - Allow "git interpret-trailers" to run outside of a Git repository.

123 + Customization to change the behaviour with "make -w" and "make -s"

124 + in our Makefile was broken when they were used together.

125

117 - Will merge to 'master'.

126 + Will merge to 'next'.

127

128

120 -* jk/test-lint-forbid-when-finished-in-subshell (2015-09-08) 5 commits

121 - (merged to 'next' on 2015-09-09 at 74ede61)

122 - + test-lib-functions: detect test_when_finished in subshell

123 - + t7800: don't use test_config in a subshell

124 - + test-lib-functions: support "test_config -C <dir> ..."

125 - + t5801: don't use test_when_finished in a subshell

126 - + t7610: don't use test_config in a subshell

129 +* jk/rebase-no-autostash (2015-09-10) 2 commits

130 + - Documentation/git-rebase: fix --no-autostash formatting

131 + - rebase: support --no-autostash

132

128 - Because "test_when_finished" in our test framework queues the

129 - clean-up tasks to be done in a shell variable, it should not be

130 - used inside a subshell. Add a mechanism to allow 'bash' to catch

131 - such uses, and fix the ones that were found.

133 + There was no way to defeat a configured rebase.autostash variable

134 + from the command line, as "git rebase --no-autostash" was missing.

135

133 - Will merge to 'master'.

136 + Will merge to 'next'.

137

138

136 -* ls/p4-lfs (2015-09-09) 5 commits

137 - - git-p4: add Git LFS backend for large file system

138 - - git-p4: add support for large file systems

139 - - git-p4: return an empty list if a list config has no values

140 - - git-p4: add gitConfigInt reader

141 - - git-p4: add optional type specifier to gitConfig reader

139 +* jw/make-arflags-customizable (2015-09-10) 1 commit

140 + - Makefile: allow $(ARFLAGS) specified from the command line

141

143 - Teach "git p4" to send large blobs outside the repository by

144 - talking to Git LFS.

142 + The Makefile always runs the library archiver with hardcoded "crs"

143 + options, which was inconvenient for exotic platforms on which

144 + people wants to use programs with totally different set of command

145 + line options.

146

147 + Will merge to 'next'.

148

147 -* nd/clone-linked-checkout (2015-09-07) 5 commits

148 - - clone: better error when --reference is a linked checkout

149 - - clone: allow --local from a linked checkout

150 - - enter_repo: allow .git files in strict mode

151 - - enter_repo: avoid duplicating logic, use is_git_directory() instead

152 - - path.c: delete an extra space

149

154 - It was not possible to use a repository-lookalike created by "git

155 - worktree add" as a local source of "git clone".

150 +* nd/ignore-then-not-ignore (2015-09-14) 2 commits

151 + - dir.c: don't exclude whole dir prematurely if neg pattern may match

152 + - dir.c: make last_exclude_matching_from_list() run til the end

153

157 - Is a reroll coming?

154 + Allow a later "!/abc/def" to override an earlier "/abc" that

155 + appears in the same .gitignore file to make it easier to express

156 + "everything in /abc directory is ignored, except for ...".

157 +

158 +

159 +* po/doc-branch-desc (2015-09-14) 1 commit

160 + (merged to 'next' on 2015-09-14 at 4934a96)

161 + + doc: show usage of branch description

162 +

163 + The branch descriptions that are set with "git branch --edit-description"

164 + option were used in many places but they weren't clearly documented.

165 +

166 + Will merge to 'master'.

167 +

168 +

169 +* sb/submodule-parallel-fetch (2015-09-14) 2 commits

170 + - fetch: fetch submodules in parallel

171 + - submodule.c: write "Fetching submodule <foo>" to stderr

172 + (this branch uses sb/submodule-helper.)

173

174 --------------------------------------------------

175 [Stalled]

@@ -166,17 +181,6 @@ of the repositories listed at

181 ($gmane/276104)

182

183

169 -* mk/submodule-gitdir-path (2015-08-05) 2 commits

170 - - path: implement common_dir handling in git_path_submodule()

171 - - submodule refactor: use git_path_submodule() in add_submodule_odb()

172 -

173 - The submodule code has been taught to work better with separate

174 - work trees created via "git worktree add".

175 -

176 - Waiting for a review.

177 - ($gmane/275340).

178 -

179 -

184 * nd/list-files (2015-02-09) 21 commits

185 - t3080: tests for git-list-files

186 - list-files: -M aka diff-cached

@@ -393,6 +397,113 @@ of the repositories listed at

397 --------------------------------------------------

398 [Cooking]

399

400 +* mk/submodule-gitdir-path (2015-09-14) 2 commits

401 + - path: implement common_dir handling in git_pathdup_submodule()

402 + - submodule refactor: use strbuf_git_path_submodule() in add_submodule_odb()

403 +

404 + The submodule code has been taught to work better with separate

405 + work trees created via "git worktree add".

406 +

407 +

408 +* jk/connect-clear-env (2015-09-08) 2 commits

409 + (merged to 'next' on 2015-09-08 at 86b64f7)

410 + + git_connect: clarify conn->use_shell flag

411 + (merged to 'next' on 2015-09-06 at 9ddbed8)

412 + + git_connect: clear GIT_* environment for ssh

413 +

414 + The ssh transport, just like any other transport over the network,

415 + did not clear GIT_* environment variables, but it is possible to

416 + use SendEnv and AcceptEnv to leak them to the remote invocation of

417 + Git, which is not a good idea at all. Explicitly clear them just

418 + like we do for the local transport.

419 +

420 + Will merge to 'master'.

421 +

422 +

423 +* ti/glibc-stdio-mutex-from-signal-handler (2015-09-04) 1 commit

424 + (merged to 'next' on 2015-09-08 at c8047ba)

425 + + pager: don't use unsafe functions in signal handlers

426 +

427 + Allocation related functions and stdio are unsafe things to call

428 + inside a signal handler, and indeed killing the pager can cause

429 + glibc to deadlock waiting on allocation mutex as our signal handler

430 + tries to free() some data structures in wait_for_pager(). Reduce

431 + these unsafe calls.

432 +

433 + I seem to recall Peff had other ideas? Let's revisit this in the

434 + next cycle.

435 +

436 +

437 +* as/subtree-with-spaces (2015-09-08) 2 commits

438 + (merged to 'next' on 2015-09-09 at f3d763d)

439 + + contrib/subtree: respect spaces in a repository path

440 + + t7900-subtree: test the "space in a subdirectory name" case

441 +

442 + Update "git subtree" (in contrib/) so that it can take whitespaces

443 + in the pathnames, not only in the in-tree pathname but the name of

444 + the directory that the repository is in.

445 +

446 + Will merge to 'master'.

447 +

448 +

449 +* gb/filter-branch-progress (2015-09-08) 1 commit

450 + - filter-branch: add passed/remaining seconds on progress

451 +

452 + Give progress meter to "git filter-branch".

453 +

454 + Is a reroll coming?

455 +

456 +

457 +* jk/interpret-trailers-outside-a-repository (2015-09-08) 1 commit

458 + (merged to 'next' on 2015-09-09 at a75fce3)

459 + + interpret-trailers: allow running outside a repository

460 +

461 + Allow "git interpret-trailers" to run outside of a Git repository.

462 +

463 + Will merge to 'master'.

464 +

465 +

466 +* jk/test-lint-forbid-when-finished-in-subshell (2015-09-08) 5 commits

467 + (merged to 'next' on 2015-09-09 at 74ede61)

468 + + test-lib-functions: detect test_when_finished in subshell

469 + + t7800: don't use test_config in a subshell

470 + + test-lib-functions: support "test_config -C <dir> ..."

471 + + t5801: don't use test_when_finished in a subshell

472 + + t7610: don't use test_config in a subshell

473 +

474 + Because "test_when_finished" in our test framework queues the

475 + clean-up tasks to be done in a shell variable, it should not be

476 + used inside a subshell. Add a mechanism to allow 'bash' to catch

477 + such uses, and fix the ones that were found.

478 +

479 + Will merge to 'master'.

480 +

481 +

482 +* ls/p4-lfs (2015-09-09) 5 commits

483 + - git-p4: add Git LFS backend for large file system

484 + - git-p4: add support for large file systems

485 + - git-p4: return an empty list if a list config has no values

486 + - git-p4: add gitConfigInt reader

487 + - git-p4: add optional type specifier to gitConfig reader

488 +

489 + Teach "git p4" to send large blobs outside the repository by

490 + talking to Git LFS.

491 +

492 +

493 +* nd/clone-linked-checkout (2015-09-14) 5 commits

494 + - clone: better error when --reference is a linked checkout

495 + - clone: allow --local from a linked checkout

496 + - enter_repo: allow .git files in strict mode

497 + - enter_repo: avoid duplicating logic, use is_git_directory() instead

498 + - path.c: delete an extra space

499 +

500 + It was not possible to use a repository-lookalike created by "git

501 + worktree add" as a local source of "git clone".

502 +

503 + Will be rerolled.

504 + ($gmane/277815)

505 +

506 +

507 * jh/quiltimport-explicit-series-file (2015-09-01) 1 commit

508 (merged to 'next' on 2015-09-02 at 1129ca4)

509 + git-quiltimport: add commandline option --series <file>

@@ -414,28 +525,6 @@ of the repositories listed at

525 Will merge to 'master'.

526

527

417 -* mp/t7060-diff-index-test (2015-09-02) 1 commit

418 - (merged to 'next' on 2015-09-08 at 80ff284)

419 - + t7060: actually test "git diff-index --cached -M"

420 -

421 - Fix an old test that was doing the same thing as another one.

422 -

423 - Will merge to 'master'.

424 -

425 -

426 -* ah/show-ref-usage-string (2015-08-31) 1 commit

427 - (merged to 'next' on 2015-09-08 at dc512c8)

428 - + show-ref: place angle brackets around variables in usage string

429 -

430 - Both "git show-ref -h" and "git show-ref --help" illustrated that the

431 - "--exclude-existing" option makes the command read list of refs

432 - from its standard input. Change only the "show-ref -h" output to

433 - have a pair of "<>" around the placeholder that designate an input

434 - file, i.e. "git show-ref --exclude-existing < <ref-list>".

435 -

436 - Will merge to 'master'.

437 -

438 -

528 * jk/date-local (2015-09-03) 11 commits

529 (merged to 'next' on 2015-09-08 at 66781d9)

530 + t6300: add tests for "-local" date formats

@@ -522,17 +611,18 @@ of the repositories listed at

611 Will merge to 'master'.

612

613

525 -* kn/for-each-tag (2015-08-25) 15 commits

526 - - SQUASH??? a tag longer than 15 bytes break alignment but keeps SP before the message

614 +* kn/for-each-tag (2015-09-11) 16 commits

615 - tag.c: implement '--merged' and '--no-merged' options

616 - tag.c: implement '--format' option

617 - tag.c: use 'ref-filter' APIs

618 - tag.c: use 'ref-filter' data structures

619 - ref-filter: add option to match literal pattern

620 - ref-filter: add support to sort by version

533 - - ref-filter: support printing N lines from tag annotation

621 + - ref-filter: add support for %(contents:lines=X)

622 - ref-filter: add option to filter out tags, branches and remotes

623 - ref-filter: implement an `align` atom

624 + - ref-filter: introduce match_atom_name()

625 + - ref-filter: introduce handler function for each atom

626 - utf8: add function to align a string into given strbuf

627 - ref-filter: introduce ref_formatting_state and ref_formatting_stack

628 - ref-filter: move `struct atom_value` to ref-filter.c

@@ -543,8 +633,7 @@ of the repositories listed at

633 The "ref-filter" code was taught about many parts of what "tag -l"

634 does and then "tag -l" is reimplemented in terms of "ref-filter".

635

546 - Almost there...

547 - A reroll exists at v16, but it seems there will be another one.

636 + Will merge to 'next'.

637

638

639 * nd/ita-cleanup (2015-09-06) 6 commits

@@ -581,7 +670,7 @@ of the repositories listed at

670 given remote name used for fetching and pushing.

671

672 Waiting for a reroll.

584 - ($gmane/275401)

673 + ($gmane/277313)

674

675

676 * sb/submodule-helper (2015-09-08) 4 commits

@@ -589,6 +678,7 @@ of the repositories listed at

678 - submodule: rewrite `module_name` shell function in C

679 - submodule: rewrite `module_list` shell function in C

680 - Merge 'hv/submodule-config' to 'sb/submodule-helper'

681 + (this branch is used by sb/submodule-parallel-fetch.)

682

683 The infrastructure to rewrite "git submodule" in C is being built

684 incrementally. Let's polish these early parts well enough and make

@@ -660,10 +750,10 @@ of the repositories listed at

750 Will merge to 'master'.

751

752

663 -* jc/rerere-multi (2015-07-30) 7 commits

664 - . t4200: rerere a merge with two identical conflicts

665 - . WIP

666 - . rerere: allow multiple variants to exist

753 +* jc/rerere-multi (2015-09-11) 7 commits

754 + - t4200: rerere a merge with two identical conflicts

755 + - WIP

756 + - rerere: allow multiple variants to exist

757 - rerere: delay the recording of preimage

758 - rerere: handle leftover rr-cache/$ID directory and postimage files

759 - rerere: scan $GIT_DIR/rr-cache/$ID when instantiating a rerere_id

What's cooking (2015/09 #03)