be70464a91 · matheusbsbs/matheus-bagliano

@@ -1,10 +1,10 @@

1 To: git@vger.kernel.org

2 Bcc: lwn@lwn.net

3 -Subject: What's cooking in git.git (Jun 2018, #06; Mon, 25)

4 -X-master-at: ed843436dd4924c10669820cc73daf50f0b4dabd

5 -X-next-at: fe4b3013947290a493c4417c35793593612e592c

3 +Subject: What's cooking in git.git (Jun 2018, #07; Thu, 28)

4 +X-master-at: e3331758f12da22f4103eec7efe1b5304a9be5e9

5 +X-next-at: 9cd7c0d54aa2ac12712fd5611825c618efdaf1e0

6

7 -What's cooking in git.git (Jun 2018, #06; Mon, 25)

7 +What's cooking in git.git (Jun 2018, #07; Thu, 28)

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

9

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

@@ -12,18 +12,10 @@ Here are the topics that have been cooking. Commits prefixed with

12 '+' are in 'next'. The ones marked with '.' do not appear in any of

13 the integration branches, but I am still holding onto them.

14

15 -Git 2.18 turned out to be an unusually large release in the recent

16 -history, with 900+ individual changes, compared to ~500 for a few

17 -releases before that. Thanks, everybody who participated.

18 -

19 -The tip of 'next' hasn't been rewound yet, but I've already reverted

20 -merges of a few topics from there to give them a fresh restart.

21 -

22 -I haven't annotated many topics still in 'pu' with short-term plans

23 -and my assessment on their done-ness. Comments and review summaries

24 -to help deciding are appreciated. I'd like to start accepting new

25 -topics only after getting the classification more-or-less right for

26 -the topics already in flight.

15 +The tip of 'next' has been rewound and it currently has only 4

16 +topics. Quite a many topics are cooking in 'pu' and need to be

17 +sifted into good bins (for 'next') and the remainder. Help is

18 +appreciated in that area ;-)

19

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

21 of the repositories listed at

@@ -33,363 +25,243 @@ of the repositories listed at

25 --------------------------------------------------

26 [Graduated to "master"]

27

36 -* ag/rebase-p (2018-06-01) 4 commits

37 - (merged to 'next' on 2018-06-13 at dd6f8a51d7)

38 - + rebase: remove -p code from git-rebase--interactive.sh

39 - + rebase: use the new git-rebase--preserve-merges.sh

40 - + rebase: strip unused code in git-rebase--preserve-merges.sh

41 - + rebase: introduce a dedicated backend for --preserve-merges

42 - (this branch is used by ag/rebase-i-append-todo-help and ag/rebase-i-rewrite-todo.)

28 +* ab/refspec-init-fix (2018-06-11) 3 commits

29 + (merged to 'next' on 2018-06-13 at 91d71d8435)

30 + + refspec: initalize `refspec_item` in `valid_fetch_refspec()`

31 + + refspec: add back a refspec_item_init() function

32 + + refspec: s/refspec_item_init/&_or_die/g

33

44 - Separate "rebase -p" codepath out of "rebase -i" implementation to

45 - slim down the latter and make it easier to manage.

34 + Make refspec parsing codepath more robust.

35

36

48 -* cc/tests-without-assuming-ref-files-backend (2018-06-04) 1 commit

49 - (merged to 'next' on 2018-06-13 at 7e2f74431c)

50 - + t9104: kosherly remove remote refs

51 -

52 - Instead of mucking with filesystem directly, use plumbing commands

53 - update-ref etc. to manipulate the refs in the tests.

54 -

55 -

56 -* ds/commit-graph-lockfile-fix (2018-05-22) 1 commit

57 - (merged to 'next' on 2018-05-24 at 3d12a02b0c)

58 - + commit-graph: fix UX issue when .lock file exists

59 - (this branch is used by ds/commit-graph-fsck; uses ds/generation-numbers.)

60 -

61 - Update to ds/generation-numbers topic.

62 -

63 -

64 -* ds/generation-numbers (2018-05-22) 11 commits

65 - (merged to 'next' on 2018-05-24 at 56fc38a1b6)

66 - + commit-graph.txt: update design document

67 - + merge: check config before loading commits

68 - + commit: use generation number in remove_redundant()

69 - + commit: add short-circuit to paint_down_to_common()

70 - + commit: use generation numbers for in_merge_bases()

71 - + ref-filter: use generation number for --contains

72 - + commit-graph: always load commit-graph information

73 - + commit: use generations in paint_down_to_common()

74 - + commit-graph: compute generation numbers

75 - + commit: add generation number to struct commit

76 - + ref-filter: fix outdated comment on in_commit_list

77 - (this branch is used by ds/commit-graph-fsck and ds/commit-graph-lockfile-fix.)

78 -

79 - A recently added "commit-graph" datafile has learned to store

80 - pre-computed generation numbers to speed up the decisions to stop

81 - history traversal.

82 -

83 -

84 -* en/merge-recursive-tests (2018-05-28) 5 commits

85 - (merged to 'next' on 2018-06-01 at 8490b560b4)

86 - + t6036: prefer test_when_finished to manual cleanup in following test

87 - + t6036, t6042: prefer test_cmp to sequences of test

88 - + t6036, t6042: prefer test_path_is_file, test_path_is_missing

89 - + t6036, t6042: use test_line_count instead of wc -l

90 - + t6036, t6042: use test_create_repo to keep tests independent

91 -

92 - Clean up tests in t6xxx series about 'merge' command.

93 -

94 -

95 -* en/rename-directory-detection-reboot (2018-06-18) 1 commit

96 - (merged to 'next' on 2018-06-18 at 95c454d3f4)

97 - + merge-recursive: use xstrdup() instead of fixed buffer

98 -

99 - Newly added codepath in merge-recursive had potential buffer

100 - overrun, which has been fixed.

37 +* as/safecrlf-quiet-fix (2018-06-11) 1 commit

38 + (merged to 'next' on 2018-06-13 at b163674843)

39 + + config.c: fix regression for core.safecrlf false

40

41 + Fix for 2.17-era regression around `core.safecrlf`.

42

103 -* jk/show-index (2018-05-29) 2 commits

104 - (merged to 'next' on 2018-06-01 at 4b3382d994)

105 - + show-index: update documentation for index v2

106 - + make show-index a builtin

107 -

108 - Modernize a less often used command.

109 -

110 -

111 -* ls/complete-remote-update-names (2018-06-01) 1 commit

112 - (merged to 'next' on 2018-06-13 at 86b4d23278)

113 - + completion: complete remote names too

114 -

115 - "git remote update" can take both a single remote nickname and a

116 - nickname for remote groups, and the completion script (in contrib/)

117 - has been taught about it.

118 -

119 -

120 -* nd/commit-util-to-slab (2018-05-21) 15 commits

121 - (merged to 'next' on 2018-05-24 at bb5643d75c)

122 - + commit.h: delete 'util' field in struct commit

123 - + merge: use commit-slab in merge remote desc instead of commit->util

124 - + log: use commit-slab in prepare_bases() instead of commit->util

125 - + show-branch: note about its object flags usage

126 - + show-branch: use commit-slab for commit-name instead of commit->util

127 - + name-rev: use commit-slab for rev-name instead of commit->util

128 - + bisect.c: use commit-slab for commit weight instead of commit->util

129 - + revision.c: use commit-slab for show_source

130 - + sequencer.c: use commit-slab to associate todo items to commits

131 - + sequencer.c: use commit-slab to mark seen commits

132 - + shallow.c: use commit-slab for commit depth instead of commit->util

133 - + describe: use commit-slab for commit names instead of commit->util

134 - + blame: use commit-slab for blame suspects instead of commit->util

135 - + commit-slab: support shared commit-slab

136 - + commit-slab.h: code split

137 -

138 - The in-core "commit" object had an all-purpose "void *util" field,

139 - which was tricky to use especially in library-ish part of the

140 - code. All of the existing uses of the field has been migrated to a

141 - more dedicated "commit-slab" mechanism and the field is eliminated.

142 -

143 -

144 -* nd/complete-config-vars (2018-05-29) 13 commits

145 - (merged to 'next' on 2018-06-13 at c2dd5546d0)

146 - + completion: complete general config vars in two steps

147 - + log-tree: allow to customize 'grafted' color

148 - + completion: support case-insensitive config vars

149 - + completion: keep other config var completion in camelCase

150 - + completion: drop the hard coded list of config vars

151 - + am: move advice.amWorkDir parsing back to advice.c

152 - + advice: keep config name in camelCase in advice_config[]

153 - + fsck: produce camelCase config key names

154 - + help: add --config to list all available config

155 - + fsck: factor out msg_id_info[] lazy initialization code

156 - + grep: keep all colors in an array

157 - + Add and use generic name->id mapping code for color slot parsing

158 - + Merge branch 'nd/command-list' into nd/complete-config-vars

159 -

160 - Continuing with the idea to programatically enumerate various

161 - pieces of data required for command line completion, teach the

162 - codebase to report the list of configuration variables

163 - subcommands care about to help complete them.

164 -

165 -

166 -* nd/diff-apply-ita (2018-05-29) 4 commits

167 - (merged to 'next' on 2018-05-30 at f98728de81)

168 - + apply: add --intent-to-add

169 - + t2203: add a test about "diff HEAD" case

170 - + diff: turn --ita-invisible-in-index on by default

171 - + diff: ignore --ita-[in]visible-in-index when diffing worktree-to-tree

172 -

173 - "git diff" compares the index and the working tree. For paths

174 - added with intent-to-add bit, the command shows the full contents

175 - of them as added, but the paths themselves were not marked as new

176 - files. They are now shown as new by default.

177 -

178 - "git apply" learned the "--intent-to-add" option so that an

179 - otherwise working-tree-only application of a patch will add new

180 - paths to the index marked with the "intent-to-add" bit.

181 -

182 -

183 -* nd/reject-empty-shallow-request (2018-06-04) 1 commit

184 - (merged to 'next' on 2018-06-13 at d6b6a1c3a7)

185 - + upload-pack: reject shallow requests that would return nothing

186 -

187 - "git fetch --shallow-since=<cutoff>" that specifies the cut-off

188 - point that is newer than the existing history used to end up

189 - grabbing the entire history. Such a request now errors out.

190 -

191 -

192 -* pc/submodule-helper-foreach (2018-05-11) 4 commits

193 - (merged to 'next' on 2018-05-22 at f22659ad46)

194 - + submodule: port submodule subcommand 'foreach' from shell to C

195 - + submodule foreach: document variable '$displaypath'

196 - + submodule foreach: document '$sm_path' instead of '$path'

197 - + submodule foreach: correct '$path' in nested submodules from a subdirectory

198 -

199 - The bulk of "git submodule foreach" has been rewritten in C.

200 -

201 -

202 -* sb/object-store-alloc (2018-05-16) 13 commits

203 - (merged to 'next' on 2018-06-13 at 2868c2db9d)

204 - + alloc: allow arbitrary repositories for alloc functions

205 - + object: allow create_object to handle arbitrary repositories

206 - + object: allow grow_object_hash to handle arbitrary repositories

207 - + alloc: add repository argument to alloc_commit_index

208 - + alloc: add repository argument to alloc_report

209 - + alloc: add repository argument to alloc_object_node

210 - + alloc: add repository argument to alloc_tag_node

211 - + alloc: add repository argument to alloc_commit_node

212 - + alloc: add repository argument to alloc_tree_node

213 - + alloc: add repository argument to alloc_blob_node

214 - + object: add repository argument to grow_object_hash

215 - + object: add repository argument to create_object

216 - + repository: introduce parsed objects field

217 - (this branch is used by sb/object-store-grafts.)

218 -

219 - The conversion to pass "the_repository" and then "a_repository"

220 - throughout the object access API continues.

43

44 +* jc/clean-after-sanity-tests (2018-06-15) 1 commit

45 + (merged to 'next' on 2018-06-22 at 2df77b8af9)

46 + + tests: clean after SANITY tests

47

223 -* sb/plug-misc-leaks (2018-06-04) 5 commits

224 - (merged to 'next' on 2018-06-13 at bf68cabe28)

225 - + SQUASH: tentatively cast const-ness away when calling free()

226 - + sequencer.c: plug mem leak in git_sequencer_config

227 - (merged to 'next' on 2018-06-04 at fbefac1c7a)

228 - + sequencer.c: plug leaks in do_pick_commit

229 - + submodule--helper: plug mem leak in print_default_remote

230 - + refs/packed-backend.c: close fd of empty file

48 + test cleanup.

49

232 - Misc leak plugging.

50

51 +* jh/partial-clone (2018-06-12) 1 commit

52 + (merged to 'next' on 2018-06-13 at 818f864b0c)

53 + + list-objects: check if filter is NULL before using

54

235 -* sg/update-ref-stdin-cleanup (2018-06-04) 1 commit

236 - (merged to 'next' on 2018-06-13 at 2b9924760d)

237 - + update-ref --stdin: use skip_prefix()

55 + The recent addition of "partial clone" experimental feature kicked

56 + in when it shouldn't, namely, when there is no partial-clone filter

57 + defined even if extensions.partialclone is set.

58

239 - Code cleanup.

59 +

60 +* jk/fetch-all-peeled-fix (2018-06-13) 2 commits

61 + (merged to 'next' on 2018-06-13 at 1333bb9d90)

62 + + fetch-pack: test explicitly that --all can fetch tag references pointing to non-commits

63 + + fetch-pack: don't try to fetch peel values with --all

64 +

65 + "git fetch-pack --all" used to unnecessarily fail upon seeing an

66 + annotated tag that points at an object other than a commit.

67 +

68 +

69 +* ms/send-pack-honor-config (2018-06-12) 1 commit

70 + (merged to 'next' on 2018-06-13 at e2cd933715)

71 + + builtin/send-pack: populate the default configs

72 +

73 + "git send-pack --signed" (hence "git push --signed" over the http

74 + transport) did not read user ident from the config mechanism to

75 + determine whom to sign the push certificate as, which has been

76 + corrected.

77 +

78 +

79 +* nd/completion-negation (2018-06-11) 3 commits

80 + (merged to 'next' on 2018-06-19 at a3be59b450)

81 + + completion: collapse extra --no-.. options

82 + + completion: suppress some -no- options

83 + + parse-options: option to let --git-completion-helper show negative form

84 +

85 + Continuing with the idea to programmatically enumerate various

86 + pieces of data required for command line completion, the codebase

87 + has been taught to enumerate options prefixed with "--no-" to

88 + negate them.

89 +

90 +

91 +* pw/add-p-recount (2018-06-11) 1 commit

92 + (merged to 'next' on 2018-06-19 at 1880ecc3f1)

93 + + add -p: fix counting empty context lines in edited patches

94 +

95 + When user edits the patch in "git add -p" and the user's editor is

96 + set to strip trailing whitespaces indiscriminately, an empty line

97 + that is unchanged in the patch would become completely empty

98 + (instead of a line with a sole SP on it). The code introduced in

99 + Git 2.17 timeframe failed to parse such a patch, but now it learned

100 + to notice the situation and cope with it.

101 +

102 +

103 +* sb/fix-fetching-moved-submodules (2018-06-14) 2 commits

104 + (merged to 'next' on 2018-06-22 at 16039dc62a)

105 + + t5526: test recursive submodules when fetching moved submodules

106 + + submodule: fix NULL correctness in renamed broken submodules

107 +

108 + The code to try seeing if a fetch is necessary in a submodule

109 + during a fetch with --recurse-submodules got confused when the path

110 + to the submodule was changed in the range of commits in the

111 + superproject, sometimes showing "(null)". This has been corrected.

112 +

113 +

114 +* sg/gpg-tests-fix (2018-06-11) 2 commits

115 + (merged to 'next' on 2018-06-13 at f3a05f1c41)

116 + + tests: make forging GPG signed commits and tags more robust

117 + + t7510-signed-commit: use 'test_must_fail'

118 +

119 + Some flaky tests have been fixed.

120 +

121 +

122 +* tz/cred-netrc-cleanup (2018-06-18) 4 commits

123 + (merged to 'next' on 2018-06-22 at a471dd714c)

124 + + git-credential-netrc: make "all" default target of Makefile

125 + + git-credential-netrc: fix exit status when tests fail

126 + + git-credential-netrc: use in-tree Git.pm for tests

127 + + git-credential-netrc: minor whitespace cleanup in test script

128 +

129 + Build and test procedure for netrc credential helper (in contrib/)

130 + has been updated.

131

132 --------------------------------------------------

133 [New Topics]

134

244 -* ld/p423 (2018-06-19) 6 commits

245 - - git-p4: python3: fix octal constants

246 - - git-p4: python3: use print() function

247 - - git-p4: python3: basestring workaround

248 - - git-p4: python3: remove backticks

249 - - git-p4: python3: replace dict.has_key(k) with "k in dict"

250 - - git-p4: python3: replace <> with !=

135 +* ao/config-from-gitmodules (2018-06-26) 6 commits

136 + - submodule-config: reuse config_from_gitmodules in repo_read_gitmodules

137 + - submodule-config: pass repository as argument to config_from_gitmodules

138 + - submodule-config: make 'config_from_gitmodules' private

139 + - submodule-config: add helper to get 'update-clone' config from .gitmodules

140 + - submodule-config: add helper function to get 'fetch' config from .gitmodules

141 + - config: move config_from_gitmodules to submodule-config.c

142

252 - Code preparation to make "git p4" closer to be usable with Python 3.

143 + Tighten the API to make it harder to misuse in-tree .gitmodules

144 + file, even though it shares the same syntax with configuration

145 + files, to read random configuration items from it.

146

147 Will merge to 'next'.

148

149

257 -* pw/rebase-i-keep-reword-after-conflict (2018-06-19) 1 commit

258 - - sequencer: do not squash 'reword' commits when we hit conflicts

150 +* as/sequencer-customizable-comment-char (2018-06-28) 1 commit

151 + - sequencer: use configured comment character

152

260 - Bugfix for "rebase -i" corner case regression.

153 + Honor core.commentchar when preparing the list of commits to replay

154 + in "rebase -i".

155 +

156 +

157 +* dj/runtime-prefix (2018-06-26) 1 commit

158 + - Makefile: tweak sed invocation

159 +

160 + POSIX portability fix in Makefile to fix a glitch introduced a few

161 + releases ago.

162

163 Will merge to 'next'.

164

165

265 -* xy/format-patch-prereq-patch-id-fix (2018-06-19) 1 commit

266 - - format-patch: clear UNINTERESTING flag before prepare_bases

166 +* ds/commit-graph (2018-06-28) 1 commit

167 + - commit-graph: fix documentation inconsistencies

168

268 - Recently added "--base" option to "git format-patch" command did

269 - not correctly generate prereq patch ids.

169 + Docfix.

170

171 Will merge to 'next'.

172

173

274 -* bw/config-refer-to-gitsubmodules-doc (2018-06-21) 1 commit

275 - - docs: link to gitsubmodules

174 +* en/rebase-i-microfixes (2018-06-27) 3 commits

175 + - git-rebase--merge: modernize "git-$cmd" to "git $cmd"

176 + - Fix use of strategy options with interactive rebases

177 + - t3418: add testcase showing problems with rebase -i and strategy options

178 +

179 + Will merge to 'next'.

180

277 - Docfix.

181 +

182 +* js/enhanced-version-info (2018-06-28) 1 commit

183 + - Makefile: fix the "built from commit" code

184 +

185 + Build fix.

186

187 Will merge to 'next'.

188

189

282 -* bw/ref-in-want (2018-06-21) 8 commits

283 - - fetch-pack: implement ref-in-want

284 - - fetch-pack: put shallow info in output parameter

285 - - fetch: refactor to make function args narrower

286 - - fetch: refactor fetch_refs into two functions

287 - - fetch: refactor the population of peer ref OIDs

288 - - upload-pack: test negotiation with changing repository

289 - - upload-pack: implement ref-in-want

290 - - test-pkt-line: add unpack-sideband subcommand

190 +* js/rebase-recreate-merge (2018-06-27) 1 commit

191 + - rebase: fix documentation formatting

192

193 + Docfix.

194

293 -* en/rebase-consistency (2018-06-21) 7 commits

294 - - git-rebase: make --allow-empty-message the default

295 - - git-rebase.txt: address confusion between --no-ff vs --force-rebase

296 - - git-rebase.txt: document behavioral inconsistencies between modes

297 - - git-rebase: error out when incompatible options passed

298 - - t3422: new testcases for checking when incompatible options passed

299 - - git-rebase.sh: update help messages a bit

300 - - git-rebase.txt: document incompatible options

195 + Will merge to 'next'.

196

197

303 -* jt/remove-pack-bitmap-global (2018-06-21) 2 commits

304 - - pack-bitmap: add free function

305 - - pack-bitmap: remove bitmap_git global variable

198 +* jt/connectivity-check-after-unshallow (2018-06-27) 1 commit

199 + - fetch: when deepening, check connectivity fully

200

201 + "git fetch" failed to correctly validate the set of objects it

202 + received when making a shallow history deeper, which has been

203 + corrected.

204

308 -* sb/submodule-move-head-error-msg (2018-06-25) 1 commit

309 - - submodule.c: report the submodule that an error occurs in

205

311 - Needs a reroll.

312 - cf. <20180622081713.5360-1-szeder.dev@gmail.com>

206 +* jt/fetch-nego-tip (2018-06-27) 1 commit

207 + - fetch-pack: support negotiation tip whitelist

208 + (this branch uses jt/fetch-pack-negotiator.)

209

210 + "git fetch" learned a new option "--negotiation-tip" to limit the

211 + set of commits it tells the other end as "have", to reduce wasted

212 + bandwidth and cycles, which would be helpful when the receiving

213 + repository has a lot of refs that have little to do with the

214 + history at the remote it is fetching from.

215

315 -* bw/protocol-v2 (2018-06-22) 1 commit

316 - - protocol-v2 doc: put HTTP headers after request

216

318 - Doc fix.

217 +* mb/filter-branch-optim (2018-06-26) 1 commit

218 + - filter-branch: skip commits present on --state-branch

219 +

220 + "git filter-branch" when used with the "--state-branch" option

221 + still attempted to rewrite the commits whose filtered result is

222 + known from the previous attempt (which is recorded on the state

223 + branch); the command has been corrected not to waste cycles doing

224 + so.

225

226 Will merge to 'next'.

227

228

323 -* jk/branch-l-1-repurpose (2018-06-22) 1 commit

324 - - branch: make "-l" a synonym for "--list"

325 - (this branch uses jk/branch-l-0-deprecation.)

229 +* ms/core-icase-doc (2018-06-28) 1 commit

230 + - Documentation: declare "core.ignoreCase" as internal variable

231

327 - Updated plan to repurpose the "-l" option to "git branch".

232 + Clarify that setting core.ignoreCase to deviate from reality would

233 + not turn a case-incapable filesystem into a case-capable one.

234

329 - Will hold in 'pu' until jk/branch-l-0-deprecation progresses sufficiently.

235 + Will merge to 'next'.

236

237

332 -* vs/typofixes (2018-06-22) 1 commit

333 - - Documentation: spelling and grammar fixes

238 +* rj/submodule-fsck-skip (2018-06-27) 1 commit

239 + - fsck: check skiplist for object in fsck_blob()

240

335 - Doc fix.

241 + "fsck.skipList" did not prevent a blob object listed there from

242 + being inspected for is contents (e.g. we recently started to

243 + inspect the contents of ".gitmodules" for certain malicious

244 + patterns), which has been corrected.

245

246 Will merge to 'next'.

247

248

340 -* cc/remote-odb (2018-06-25) 9 commits

341 - - Documentation/config: add odb.<name>.promisorRemote

342 - - t0410: test fetching from many promisor remotes

343 - - Use odb.origin.partialclonefilter instead of core.partialclonefilter

344 - - Use remote_odb_get_direct() and has_remote_odb()

345 - - remote-odb: add remote_odb_reinit()

346 - - remote-odb: implement remote_odb_get_many_direct()

347 - - remote-odb: implement remote_odb_get_direct()

348 - - Add initial remote odb support

349 - - fetch-object: make functions return an error code

249 +* tb/grep-only-matching (2018-06-27) 3 commits

250 + - SQUASH??? emerg compilation fix

251 + - grep.c: teach 'git grep --only-matching'

252 + - grep.c: extract show_line_header()

253 + (this branch uses tb/grep-column.)

254

255 + "git grep" learned the "--only-matching" option.

256

352 -* ds/multi-pack-index (2018-06-25) 24 commits

353 - - midx: clear midx on repack

354 - - packfile: skip loading index if in multi-pack-index

355 - - midx: prevent duplicate packfile loads

356 - - midx: use midx in approximate_object_count

357 - - midx: use existing midx when writing new one

358 - - midx: use midx in abbreviation calculations

359 - - midx: read objects from multi-pack-index

360 - - midx: prepare midxed_git struct

361 - - config: create core.multiPackIndex setting

362 - - midx: write object offsets

363 - - midx: write object id fanout chunk

364 - - midx: write object ids in a chunk

365 - - midx: sort and deduplicate objects from packfiles

366 - - midx: read pack names into array

367 - - multi-pack-index: write pack names in chunk

368 - - multi-pack-index: read packfile list

369 - - packfile: generalize pack directory list

370 - - multi-pack-index: expand test data

371 - - multi-pack-index: load into memory

372 - - midx: write header information to lockfile

373 - - multi-pack-index: add 'write' verb

374 - - multi-pack-index: add builtin

375 - - multi-pack-index: add format details

376 - - multi-pack-index: add design document

257

258 +* tz/exclude-doc-smallfixes (2018-06-27) 2 commits

259 + - dir.c: fix typos in core.excludesfile comment

260 + - gitignore.txt: clarify default core.excludesfile path

261

379 -* nd/use-the-index-compat-less (2018-06-25) 13 commits

380 - - wt-status.c: stop using index compat macros

381 - - sha1-name.c: stop using index compat macros

382 - - sequencer.c: stop using index compat macros

383 - - revision.c: stop using index compat macros

384 - - rerere.c: stop using index compat macros

385 - - merge.c: stop using index compat macros

386 - - merge-recursive.c: stop using index compat macros

387 - - entry.c: stop using index compat macros

388 - - diff.c: stop using index compat macros

389 - - diff-lib.c: stop using index compat macros

390 - - check-racy.c: stop using index compat macros

391 - - blame.c: stop using index compat macros

392 - - apply.c: stop using index compat macros

262 + Doc updates.

263 +

264 + Will merge to 'next'.

265

266 --------------------------------------------------

267 [Stalled]

@@ -518,6 +390,185 @@ of the repositories listed at

390 --------------------------------------------------

391 [Cooking]

392

393 +* ld/p423 (2018-06-19) 6 commits

394 + - git-p4: python3: fix octal constants

395 + - git-p4: python3: use print() function

396 + - git-p4: python3: basestring workaround

397 + - git-p4: python3: remove backticks

398 + - git-p4: python3: replace dict.has_key(k) with "k in dict"

399 + - git-p4: python3: replace <> with !=

400 +

401 + Code preparation to make "git p4" closer to be usable with Python 3.

402 +

403 + Will merge to 'next'.

404 +

405 +

406 +* pw/rebase-i-keep-reword-after-conflict (2018-06-19) 1 commit

407 + - sequencer: do not squash 'reword' commits when we hit conflicts

408 +

409 + Bugfix for "rebase -i" corner case regression.

410 +

411 + Will merge to 'next'.

412 +

413 +

414 +* xy/format-patch-prereq-patch-id-fix (2018-06-19) 1 commit

415 + - format-patch: clear UNINTERESTING flag before prepare_bases

416 +

417 + Recently added "--base" option to "git format-patch" command did

418 + not correctly generate prereq patch ids.

419 +

420 + Will merge to 'next'.

421 +

422 +

423 +* bw/config-refer-to-gitsubmodules-doc (2018-06-21) 1 commit

424 + - docs: link to gitsubmodules

425 +

426 + Docfix.

427 +

428 + Will merge to 'next'.

429 +

430 +

431 +* bw/ref-in-want (2018-06-28) 8 commits

432 + - fetch-pack: implement ref-in-want

433 + - fetch-pack: put shallow info in output parameter

434 + - fetch: refactor to make function args narrower

435 + - fetch: refactor fetch_refs into two functions

436 + - fetch: refactor the population of peer ref OIDs

437 + - upload-pack: test negotiation with changing repository

438 + - upload-pack: implement ref-in-want

439 + - test-pkt-line: add unpack-sideband subcommand

440 +

441 + Protocol v2 has been updated to allow slightly out-of-sync set of

442 + servers to work together to serve a single client, which would be

443 + useful with load-balanced servers that talk smart HTTP transport.

444 +

445 +

446 +* en/rebase-consistency (2018-06-27) 9 commits

447 + - git-rebase: make --allow-empty-message the default

448 + - t3401: add directory rename testcases for rebase and am

449 + - git-rebase.txt: document behavioral differences between modes

450 + - directory-rename-detection.txt: technical docs on abilities and limitations

451 + - git-rebase.txt: address confusion between --no-ff vs --force-rebase

452 + - git-rebase: error out when incompatible options passed

453 + - t3422: new testcases for checking when incompatible options passed

454 + - git-rebase.sh: update help messages a bit

455 + - git-rebase.txt: document incompatible options

456 +

457 + "git rebase" behaved slightly differently depending on which one of

458 + the three backends gets used; this has been documented and an

459 + effort to make them more uniform has begun.

460 +

461 +

462 +* jt/remove-pack-bitmap-global (2018-06-21) 2 commits

463 + - pack-bitmap: add free function

464 + - pack-bitmap: remove bitmap_git global variable

465 +

466 + The effort to move globals to per-repository in-core structure

467 + continues.

468 +

469 +

470 +* sb/submodule-move-head-error-msg (2018-06-25) 1 commit

471 + - submodule.c: report the submodule that an error occurs in

472 +

473 + Needs a reroll.

474 + cf. <20180622081713.5360-1-szeder.dev@gmail.com>

475 +

476 +

477 +* bw/protocol-v2 (2018-06-22) 1 commit

478 + - protocol-v2 doc: put HTTP headers after request

479 +

480 + Doc fix.

481 +

482 + Will merge to 'next'.

483 +

484 +

485 +* jk/branch-l-1-repurpose (2018-06-22) 1 commit

486 + - branch: make "-l" a synonym for "--list"

487 + (this branch uses jk/branch-l-0-deprecation.)

488 +

489 + Updated plan to repurpose the "-l" option to "git branch".

490 +

491 + Will hold in 'pu' until jk/branch-l-0-deprecation progresses sufficiently.

492 +

493 +

494 +* vs/typofixes (2018-06-22) 1 commit

495 + - Documentation: spelling and grammar fixes

496 +

497 + Doc fix.

498 +

499 + Will merge to 'next'.

500 +

501 +

502 +* cc/remote-odb (2018-06-25) 9 commits

503 + . Documentation/config: add odb.<name>.promisorRemote

504 + . t0410: test fetching from many promisor remotes

505 + . Use odb.origin.partialclonefilter instead of core.partialclonefilter

506 + . Use remote_odb_get_direct() and has_remote_odb()

507 + . remote-odb: add remote_odb_reinit()

508 + . remote-odb: implement remote_odb_get_many_direct()

509 + . remote-odb: implement remote_odb_get_direct()

510 + . Add initial remote odb support

511 + . fetch-object: make functions return an error code

512 +

513 + Needs a reroll.

514 +

515 +

516 +* ds/multi-pack-index (2018-06-25) 24 commits

517 + - midx: clear midx on repack

518 + - packfile: skip loading index if in multi-pack-index

519 + - midx: prevent duplicate packfile loads

520 + - midx: use midx in approximate_object_count

521 + - midx: use existing midx when writing new one

522 + - midx: use midx in abbreviation calculations

523 + - midx: read objects from multi-pack-index

524 + - midx: prepare midxed_git struct

525 + - config: create core.multiPackIndex setting

526 + - midx: write object offsets

527 + - midx: write object id fanout chunk

528 + - midx: write object ids in a chunk

529 + - midx: sort and deduplicate objects from packfiles

530 + - midx: read pack names into array

531 + - multi-pack-index: write pack names in chunk

532 + - multi-pack-index: read packfile list

533 + - packfile: generalize pack directory list

534 + - multi-pack-index: expand test data

535 + - multi-pack-index: load into memory

536 + - midx: write header information to lockfile

537 + - multi-pack-index: add 'write' verb

538 + - multi-pack-index: add builtin

539 + - multi-pack-index: add format details

540 + - multi-pack-index: add design document

541 +

542 + When there are too many packfiles in a repository (which is not

543 + recommended), looking up an object in these would require

544 + consulting many pack .idx files; a new mechanism to have a single

545 + file that consolidates all of these .idx files is introduced.

546 +

547 +

548 +* nd/use-the-index-compat-less (2018-06-25) 13 commits

549 + - wt-status.c: stop using index compat macros

550 + - sha1-name.c: stop using index compat macros

551 + - sequencer.c: stop using index compat macros

552 + - revision.c: stop using index compat macros

553 + - rerere.c: stop using index compat macros

554 + - merge.c: stop using index compat macros

555 + - merge-recursive.c: stop using index compat macros

556 + - entry.c: stop using index compat macros

557 + - diff.c: stop using index compat macros

558 + - diff-lib.c: stop using index compat macros

559 + - check-racy.c: stop using index compat macros

560 + - blame.c: stop using index compat macros

561 + - apply.c: stop using index compat macros

562 +

563 + Use of many convenience functions that operate on the_index

564 + "primary in-core index instance" have been rewritten to explicitly

565 + call the coutnerpart functions that take arbitrary index_state and

566 + pass the_index.

567 +

568 + I'd say that alone is a useless code churn, but people seem to be

569 + excited about the change, so it is queued here.

570 +

571 +

572 * tb/grep-column (2018-06-22) 7 commits

573 - contrib/git-jump/git-jump: jump to exact location

574 - grep.c: add configuration variables to show matched option

@@ -526,22 +577,12 @@ of the repositories listed at

577 - grep.[ch]: extend grep_opt to allow showing matched column

578 - grep.c: expose {,inverted} match column in match_line()

579 - Documentation/config.txt: camel-case lineNumber for consistency

580 + (this branch is used by tb/grep-only-matching.)

581

582 "git grep" learned the "--column" option that gives not just the

583 line number but the column number of the hit.

584

533 -

534 -* tz/cred-netrc-cleanup (2018-06-18) 4 commits

535 - (merged to 'next' on 2018-06-22 at a471dd714c)

536 - + git-credential-netrc: make "all" default target of Makefile

537 - + git-credential-netrc: fix exit status when tests fail

538 - + git-credential-netrc: use in-tree Git.pm for tests

539 - + git-credential-netrc: minor whitespace cleanup in test script

540 -

541 - Build and test procedure for netrc credential helper (in contrib/)

542 - has been updated.

543 -

544 - Will merge to 'master'.

585 + Will merge to 'next'.

586

587

588 * jt/fetch-pack-negotiator (2018-06-15) 7 commits

@@ -552,10 +593,13 @@ of the repositories listed at

593 - fetch-pack: directly end negotiation if ACK ready

594 - fetch-pack: clear marks before re-marking

595 - fetch-pack: split up everything_local()

596 + (this branch is used by jt/fetch-nego-tip.)

597

598 Code restructuring and a small fix to transport protocol v2 during

599 fetching.

600

601 + Is this ready for 'next'?

602 +

603

604 * ag/rebase-i-append-todo-help (2018-06-14) 2 commits

605 - rebase--interactive: rewrite append_todo_help() in C

@@ -564,6 +608,9 @@ of the repositories listed at

608

609 Stepwise rewriting of the machinery of "rebase -i" into C continues.

610

611 + Needs a reroll.

612 + cf. <nycvar.QRO.7.76.6.1806261125330.21419@tvgsbejvaqbjf.bet>

613 +

614

615 * ag/rebase-i-rewrite-todo (2018-06-15) 3 commits

616 - rebase--interactive: rewrite the edit-todo functionality in C

@@ -574,25 +621,14 @@ of the repositories listed at

621 Stepwise rewriting of the machinery of "rebase -i" into C continues.

622

623

577 -* sb/fix-fetching-moved-submodules (2018-06-14) 2 commits

578 - (merged to 'next' on 2018-06-22 at 16039dc62a)

579 - + t5526: test recursive submodules when fetching moved submodules

580 - + submodule: fix NULL correctness in renamed broken submodules

581 -

582 - The code to try seeing if a fetch is necessary in a submodule

583 - during a fetch with --recurse-submodules got confused when the path

584 - to the submodule was changed in the range of commits in the

585 - superproject, sometimes showing "(null)". This has been corrected.

586 -

587 - Will merge to 'master'.

588 -

589 -

624 * sb/submodule-core-worktree (2018-06-19) 3 commits

591 - (merged to 'next' on 2018-06-22 at 2f7f263b30)

625 + (merged to 'next' on 2018-06-28 at 96e1a8dbd1)

626 + submodule deinit: unset core.worktree

627 + submodule: ensure core.worktree is set after update

628 + submodule: unset core.worktree if no working tree is present

629

630 + Originally merged to 'next' on 2018-06-22

631 +

632 "git submodule" did not correctly adjust core.worktree setting that

633 indicates whether/where a submodule repository has its associated

634 working tree across various state transitions, which has been

@@ -602,7 +638,7 @@ of the repositories listed at

638

639

640 * ds/ewah-cleanup (2018-06-21) 10 commits

605 - (merged to 'next' on 2018-06-22 at ad0ab374a1)

641 + (merged to 'next' on 2018-06-28 at 9cd7c0d54a)

642 + ewah: delete unused 'rlwit_discharge_empty()'

643 + ewah: drop ewah_serialize_native function

644 + ewah: drop ewah_deserialize function

@@ -614,21 +650,14 @@ of the repositories listed at

650 + ewah/bitmap.c: delete unused 'bitmap_each_bit()'

651 + ewah/bitmap.c: delete unused 'bitmap_clear()'

652

653 + Originally merged to 'next' on 2018-06-22

654 +

655 Remove unused function definitions and declarations from ewah

656 bitmap subsystem.

657

658 Will merge to 'master'.

659

660

623 -* jc/clean-after-sanity-tests (2018-06-15) 1 commit

624 - (merged to 'next' on 2018-06-22 at 2df77b8af9)

625 - + tests: clean after SANITY tests

626 -

627 - test cleanup.

628 -

629 - Will merge to 'master'.

630 -

631 -

661 * is/parsing-line-range (2018-06-15) 2 commits

662 - log: prevent error if line range ends past end of file

663 - blame: prevent error if range ends past end of file

@@ -637,49 +666,8 @@ of the repositories listed at

666 take has been tweaked.

667

668

640 -* ab/refspec-init-fix (2018-06-11) 3 commits

641 - (merged to 'next' on 2018-06-13 at 91d71d8435)

642 - + refspec: initalize `refspec_item` in `valid_fetch_refspec()`

643 - + refspec: add back a refspec_item_init() function

644 - + refspec: s/refspec_item_init/&_or_die/g

645 -

646 - Make refspec parsing codepath more robust.

647 -

648 - Will merge to 'master'.

649 -

650 -

651 -* as/safecrlf-quiet-fix (2018-06-11) 1 commit

652 - (merged to 'next' on 2018-06-13 at b163674843)

653 - + config.c: fix regression for core.safecrlf false

654 -

655 - Fix for 2.17-era regression.

656 -

657 - Will merge to 'master'.

658 -

659 -

660 -* sg/gpg-tests-fix (2018-06-11) 2 commits

661 - (merged to 'next' on 2018-06-13 at f3a05f1c41)

662 - + tests: make forging GPG signed commits and tags more robust

663 - + t7510-signed-commit: use 'test_must_fail'

664 -

665 - Some flaky tests have been fixed.

666 -

667 - Will merge to 'master'.

668 -

669 -

670 -* jk/fetch-all-peeled-fix (2018-06-13) 2 commits

671 - (merged to 'next' on 2018-06-13 at 1333bb9d90)

672 - + fetch-pack: test explicitly that --all can fetch tag references pointing to non-commits

673 - + fetch-pack: don't try to fetch peel values with --all

674 -

675 - "git fetch-pack --all" used to unnecessarily fail upon seeing an

676 - annotated tag that points at an object other than a commit.

677 -

678 - Will merge to 'master'.

679 -

680 -

669 * en/merge-recursive-cleanup (2018-06-12) 6 commits

682 - (merged to 'next' on 2018-06-19 at d175cea416)

670 + (merged to 'next' on 2018-06-28 at 1a3646eb7d)

671 + merge-recursive: add pointer about unduly complex looking code

672 + merge-recursive: rename conflict_rename_*() family of functions

673 + merge-recursive: clarify the rename_dir/RENAME_DIR meaning

@@ -687,30 +675,9 @@ of the repositories listed at

675 + merge-recursive: fix numerous argument alignment issues

676 + merge-recursive: fix miscellaneous grammar error in comment

677

690 - Code cleanup.

691 -

692 - Will merge to 'master'.

693 -

694 -

695 -* jh/partial-clone (2018-06-12) 1 commit

696 - (merged to 'next' on 2018-06-13 at 818f864b0c)

697 - + list-objects: check if filter is NULL before using

698 -

699 - The recent addition of "partial clone" experimental feature kicked

700 - in when it shouldn't, namely, when there is no partial-clone filter

701 - defined even if extensions.partialclone is set.

702 -

703 - Will merge to 'master'.

704 -

678 + Originally merged to 'next' on 2018-06-19

679

706 -* ms/send-pack-honor-config (2018-06-12) 1 commit

707 - (merged to 'next' on 2018-06-13 at e2cd933715)

708 - + builtin/send-pack: populate the default configs

709 -

710 - "git send-pack --signed" (hence "git push --signed" over the http

711 - transport) did not read user ident from the config mechanism to

712 - determine whom to sign the push certificate as, which has been

713 - corrected.

680 + Code cleanup.

681

682 Will merge to 'master'.

683

@@ -726,24 +693,11 @@ of the repositories listed at

693 - checkout tests: index should be clean after dwim checkout

694

695

729 -* pw/add-p-recount (2018-06-11) 1 commit

730 - (merged to 'next' on 2018-06-19 at 1880ecc3f1)

731 - + add -p: fix counting empty context lines in edited patches

732 -

733 - When user edits the patch in "git add -p" and the user's editor is

734 - set to strip trailing whitespaces indiscriminately, an empty line

735 - that is unchanged in the patch would become completely empty

736 - (instead of a line with a sole SP on it). The code introduced in

737 - Git 2.17 timeframe failed to parse such a patch, but now it learned

738 - to notice the situation and cope with it.

739 -

740 - Will merge to 'master'.

741 -

742 -

743 -* ds/commit-graph-fsck (2018-05-29) 20 commits

696 +* ds/commit-graph-fsck (2018-06-27) 22 commits

697 - commit-graph: update design document

698 - gc: automatically write commit-graph files

699 - commit-graph: add '--reachable' option

700 + - commit-graph: use string-list API for input

701 - fsck: verify commit-graph

702 - commit-graph: verify contents match checksum

703 - commit-graph: test for corrupted octopus edge

@@ -761,9 +715,12 @@ of the repositories listed at

715 - commit-graph: parse commit from chosen graph

716 - commit-graph: fix GRAPH_MIN_SIZE

717 - commit-graph: UNLEAK before die()

718 + - t5318-commit-graph.sh: use core.commitGraph

719

765 - Expecting a reroll.

766 - cf. <ba3b8e06-b5e1-99a1-0fe4-ff97d6da8f15@gmail.com>

720 + "git fsck" learns to make sure the optional commit-graph file is in

721 + a sane state.

722 +

723 + Is this ready for 'next'?

724

725

726 * ma/wrapped-info (2018-05-28) 2 commits

@@ -773,44 +730,27 @@ of the repositories listed at

730 An attempt to help making multi-line messages fed to warning(),

731 error(), and friends more easily translatable.

732

776 - Waiting for the discussion to settle.

733 + Will discard and wait for a cleaned-up rewrite.

734 cf. <20180529213957.GF7964@sigill.intra.peff.net>

735

736

780 -* nd/completion-negation (2018-06-11) 3 commits

781 - (merged to 'next' on 2018-06-19 at a3be59b450)

782 - + completion: collapse extra --no-.. options

783 - + completion: suppress some -no- options

784 - + parse-options: option to let --git-completion-helper show negative form

785 -

786 - Continuing with the idea to programatically enumerate various

787 - pieces of data required for command line completion, the codebase

788 - has been taught to enumerate options prefixed with "--no-" to

789 - negate them.

790 -

791 - Will merge to 'master'.

792 -

793 -

794 -* jm/cache-entry-from-mem-pool (2018-06-21) 8 commits

737 +* jm/cache-entry-from-mem-pool (2018-06-28) 8 commits

738 - block alloc: add validations around cache_entry lifecyle

796 - - block alloc: allocate cache entries from mem_pool

739 + - block alloc: allocate cache entries from mem-pool

740 - mem-pool: fill out functionality

798 - - mem-pool: add lifecycle management functions

799 - - mem-pool: tweak math on mp_block allocation size

741 + - mem-pool: add life cycle management functions

742 - mem-pool: only search head block for available space

743 - block alloc: add lifecycle APIs for cache_entry structs

744 + - read-cache: make_cache_entry should take object_id struct

745 - read-cache: teach refresh_cache_entry() to take istate

746

747 For a large tree, the index needs to hold many cache entries

748 allocated on heap. These cache entries are now allocated out of a

749 dedicated memory pool to amortize malloc(3) overhead.

750

808 - Kicked back to 'pu' with a rerolled one.

809 - cf. <20180620201557.77155-1-jamill@microsoft.com>

810 -

751

752 * sb/object-store-grafts (2018-05-18) 19 commits

813 - (merged to 'next' on 2018-06-22 at 386a6cc66a)

753 + (merged to 'next' on 2018-06-28 at 02f70d6302)

754 + commit: allow lookup_commit_graft to handle arbitrary repositories

755 + commit: allow prepare_commit_graft to handle arbitrary repositories

756 + shallow: migrate shallow information into the object parser

@@ -831,6 +771,8 @@ of the repositories listed at

771 + object: move grafts to object parser

772 + object-store: move object access functions to object-store.h

773

774 + Originally merged to 'next' on 2018-06-22

775 +

776 The conversion to pass "the_repository" and then "a_repository"

777 throughout the object access API continues.

778

What's cooking (2018/06 #07)