06ad1f9e05 · 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 2020, #06; Mon, 29)

4 -X-master-at: a08a83db2bf27f015bec9a435f6d73e223c21c5e

5 -X-next-at: 050319c2ae82f2fac00eac9d80a1d91394d99aec

3 +Subject: What's cooking in git.git (Jul 2020, #01; Mon, 6)

4 +X-master-at: 4a0fcf9f760c9774be77f51e1e88a7499b53d2e2

5 +X-next-at: b3809f4674cea6e2f5db8a478a39b85851690c7d

6

7 -What's cooking in git.git (Jun 2020, #06; Mon, 29)

7 +What's cooking in git.git (Jul 2020, #01; Mon, 6)

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

9

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

@@ -12,11 +12,10 @@ only in 'seen' (formerly 'pu'---proposed updates) while commits prefixed

12 with '+' 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 -More topics graduate to the 'master' and to the 'next' branches. I

16 -expect that this week would be slower than usual due to US holidays

17 -but hopefully we'd be able to merge most of the impactful topics by

18 -the -rc0. Some large-ish topics may have to be left for the next

19 -cycle, as usual.

15 +We are about to tag the -rc0 "preview" and enter pre-release freeze

16 +period. The change to let users name the initial branch that gets

17 +prepared in a newly created repository has now been merged to be

18 +included in the upcoming release.

19

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

21 repositories listed at

@@ -24,139 +23,139 @@ repositories listed at

23 http://git-blame.blogspot.com/p/git-public-repositories.html

24

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

27 -[Graduated to "master"]

26 +[Graduated to 'master']

27

29 -* dl/diff-usage-comment-update (2020-06-23) 2 commits

30 - (merged to 'next' on 2020-06-23 at 48d685cd45)

31 - + builtin/diff: fix botched update of usage comment

32 - (merged to 'next' on 2020-06-19 at b53a12bb66)

33 - + builtin/diff: update usage comment

34 -

35 - An in-code comment in "git diff" has been updated.

36 -

37 -

38 -* rs/commit-reach-leakfix (2020-06-19) 1 commit

39 - (merged to 'next' on 2020-06-22 at fcc1e385f0)

40 - + commit-reach: plug minor memory leak after using is_descendant_of()

41 -

42 - Leakfix.

43 -

44 -

45 -* rs/pull-leakfix (2020-06-19) 1 commit

46 - (merged to 'next' on 2020-06-22 at d80a6b9ca3)

47 - + pull: plug minor memory leak after using is_descendant_of()

48 -

49 - Leakfix.

50 -

51 -

52 -* rs/retire-strbuf-write-fd (2020-06-19) 2 commits

53 - (merged to 'next' on 2020-06-22 at c175b54f2a)

54 - + strbuf: remove unreferenced strbuf_write_fd method.

55 - + bugreport.c: replace strbuf_write_fd with write_in_full

56 -

57 - A misdesigned strbuf_write_fd() function has been retired.

58 -

59 -

60 -* sk/diff-files-show-i-t-a-as-new (2020-06-22) 1 commit

61 - (merged to 'next' on 2020-06-22 at 825a823416)

62 - + diff-files: treat "i-t-a" files as "not-in-index"

63 - (this branch is used by js/diff-files-i-t-a-fix-for-difftool.)

64 -

65 - "git diff-files" has been taught to say paths that are marked as

66 - intent-to-add are new files, not modified from an empty blob.

67 -

68 -

69 -* xl/upgrade-repo-format (2020-06-05) 4 commits

70 - (merged to 'next' on 2020-06-19 at 02bf7a9d8c)

71 - + check_repository_format_gently(): refuse extensions for old repositories

72 - + sparse-checkout: upgrade repository to version 1 when enabling extension

73 - + fetch: allow adding a filter after initial clone

74 - + repository: add a helper function to perform repository format upgrade

28 +* ak/commit-graph-to-slab (2020-06-17) 4 commits

29 + (merged to 'next' on 2020-06-25 at 9a10ae18d3)

30 + + commit-graph: minimize commit_graph_data_slab access

31 + + commit: move members graph_pos, generation to a slab

32 + + commit-graph: introduce commit_graph_data_slab

33 + + object: drop parsed_object_pool->commit_count

34

76 - Allow runtime upgrade of the repository format version, which needs

77 - to be done carefully.

35 + A few fields in "struct commit" that do not have to always be

36 + present have been moved to commit slabs.

37

79 - There is a rather unpleasant backward compatibility worry with the

80 - last step of this series, but it is the right thing to do in the

81 - longer term.

38

83 ---------------------------------------------------

84 -[New Topics]

39 +* bc/http-push-flagsfix (2020-06-23) 1 commit

40 + (merged to 'next' on 2020-06-25 at 1dabb1f61b)

41 + + http-push: ensure unforced pushes fail when data would be lost

42

86 -* mt/hash-to-hex-thread-safety (2020-06-26) 2 commits

87 - - hex: make hash_to_hex_algop() and friends thread-safe

88 - - compat/win32/pthread: add pthread_once()

43 + The code to push changes over "dumb" HTTP had a bad interaction

44 + with the commit reachability code due to incorrect allocation of

45 + object flag bits, which has been corrected.

46

90 - hash_to_hex() used a set of rotating static buffers, which was not

91 - safe to use in a threaded environment. This has been made safer by

92 - using thread-local storage.

47

94 - Will merge to 'next'.

48 +* bc/sha-256-cvs-svn-updates (2020-06-22) 14 commits

49 + (merged to 'next' on 2020-06-25 at e0d1c87fcf)

50 + + git-cvsexportcommit: port to SHA-256

51 + + git-cvsimport: port to SHA-256

52 + + git-cvsserver: port to SHA-256

53 + + git-svn: set the OID length based on hash algorithm

54 + + perl: make SVN code hash independent

55 + + perl: make Git::IndexInfo work with SHA-256

56 + + perl: create and switch variables for hash constants

57 + + t/lib-git-svn: make hash size independent

58 + + t9101: make hash independent

59 + + t9104: make hash size independent

60 + + t9100: make test work with SHA-256

61 + + t9108: make test hash independent

62 + + t9168: make test hash independent

63 + + t9109: make test hash independent

64 + (this branch uses bc/sha-256-part-2.)

65

96 ---------------------------------------------------

97 -[Stalled]

66 + CVS/SVN interface have been prepared for SHA-256 transition

67

99 -* dr/push-remoteref-fix (2020-04-23) 1 commit

100 - - remote.c: fix handling of %(push:remoteref)

68

102 - The "%(push:remoteref)" placeholder in the "--format=" argument of

103 - "git format-patch" (and friends) only showed what got explicitly

104 - configured, not what ref at the receiving end would be updated when

105 - "git push" was used, as it ignored the default behaviour (e.g. update

106 - the same ref as the source).

69 +* bc/sha-256-part-2 (2020-06-19) 44 commits

70 + (merged to 'next' on 2020-06-22 at dc153f366d)

71 + + remote-testgit: adapt for object-format

72 + + bundle: detect hash algorithm when reading refs

73 + + t5300: pass --object-format to git index-pack

74 + + t5704: send object-format capability with SHA-256

75 + + t5703: use object-format serve option

76 + + t5702: offer an object-format capability in the test

77 + + t/helper: initialize the repository for test-sha1-array

78 + + remote-curl: avoid truncating refs with ls-remote

79 + + t1050: pass algorithm to index-pack when outside repo

80 + + builtin/index-pack: add option to specify hash algorithm

81 + + remote-curl: detect algorithm for dumb HTTP by size

82 + + builtin/ls-remote: initialize repository based on fetch

83 + + t5500: make hash independent

84 + + serve: advertise object-format capability for protocol v2

85 + + connect: parse v2 refs with correct hash algorithm

86 + + connect: pass full packet reader when parsing v2 refs

87 + + Documentation/technical: document object-format for protocol v2

88 + + t1302: expect repo format version 1 for SHA-256

89 + + builtin/show-index: provide options to determine hash algo

90 + + t5302: modernize test formatting

91 + + packfile: compute and use the index CRC offset

92 + + t3200: mark assertion with SHA1 prerequisite

93 + + setup: set the_repository's hash algo when checking format

94 + + fetch-pack: parse and advertise the object-format capability

95 + + t5562: pass object-format in synthesized test data

96 + + builtin/clone: initialize hash algorithm properly

97 + + remote-curl: implement object-format extensions

98 + + transport-helper: implement object-format extensions

99 + + docs: update remote helper docs for object-format extensions

100 + + builtin/receive-pack: detect when the server doesn't support our hash

101 + + connect: detect algorithm when fetching refs

102 + + fetch-pack: detect when the server doesn't support our hash

103 + + connect: make parse_feature_value extern

104 + + send-pack: detect when the server doesn't support our hash

105 + + connect: add function to detect supported v1 hash functions

106 + + transport: add a hash algorithm member

107 + + pkt-line: add a member for hash algorithm

108 + + connect: add function to fetch value of a v2 server capability

109 + + connect: add function to parse multiple v1 capability values

110 + + remote: advertise the object-format capability on the server side

111 + + wrapper: add function to compare strings with different NUL termination

112 + + connect: have ref processing code take struct packet_reader

113 + + Documentation: document v1 protocol object-format capability

114 + + t1050: match object ID paths in a hash-insensitive way

115 + (this branch is used by bc/sha-256-cvs-svn-updates.)

116

108 - Expecting a reroll.

109 - cf. <20200416152145.wp2zeibxmuyas6y6@feanor>

117 + SHA-256 migration work continues.

118

119

112 -* mr/bisect-in-c-2 (2020-04-23) 12 commits

113 - - bisect--helper: retire `--bisect-autostart` subcommand

114 - - bisect--helper: retire `--write-terms` subcommand

115 - - bisect--helper: retire `--check-expected-revs` subcommand

116 - - bisect--helper: reimplement `bisect_state` & `bisect_head` shell functions in C

117 - - bisect--helper: retire `--next-all` subcommand

118 - - bisect--helper: retire `--bisect-clean-state` subcommand

119 - - bisect--helper: finish porting `bisect_start()` to C

120 - - bisect--helper: reimplement `bisect_next` and `bisect_auto_next` shell functions in C

121 - - bisect--helper: reimplement `bisect_autostart` shell function in C

122 - - bisect--helper: introduce new `write_in_file()` function

123 - - bisect--helper: use '-res' in 'cmd_bisect__helper' return

124 - - bisect--helper: fix `cmd_*()` function switch default return

120 +* cb/is-descendant-of (2020-06-23) 1 commit

121 + (merged to 'next' on 2020-06-25 at ca0eaddf5e)

122 + + commit-reach: avoid is_descendant_of() shim

123

126 - Rewrite of the remainder of "git bisect" script in C continues.

124 + Code clean-up.

125

128 - Expecting a response to reviews.

129 - cf. <nycvar.QRO.7.76.6.2005230007260.56@tvgsbejvaqbjf.bet>

126

127 +* dl/test-must-fail-fixes-5 (2020-06-24) 4 commits

128 + (merged to 'next' on 2020-06-29 at efe53a7588)

129 + + lib-submodule-update: pass 'test_must_fail' as an argument

130 + + lib-submodule-update: prepend "git" to $command

131 + + lib-submodule-update: consolidate --recurse-submodules

132 + + lib-submodule-update: add space after function name

133

132 -* mk/use-size-t-in-zlib (2018-10-15) 1 commit

133 - - zlib.c: use size_t for size

134 + The effort to avoid using test_must_fail on non-git command continues.

135

135 - The wrapper to call into zlib followed our long tradition to use

136 - "unsigned long" for sizes of regions in memory, which have been

137 - updated to use "size_t".

136

139 ---------------------------------------------------

140 -[Cooking]

137 +* en/sparse-status (2020-06-22) 3 commits

138 + (merged to 'next' on 2020-06-25 at 404403a075)

139 + + git-prompt: include sparsity state as well

140 + + git-prompt: document how in-progress operations affect the prompt

141 + + wt-status: show sparse checkout status as well

142

142 -* bc/http-push-flagsfix (2020-06-23) 1 commit

143 - (merged to 'next' on 2020-06-25 at 1dabb1f61b)

144 - + http-push: ensure unforced pushes fail when data would be lost

143 + "git status" learned to report the status of sparse checkout.

144

146 - The code to push changes over "dumb" HTTP had a bad interaction

147 - with the commit reachability code due to incorrect allocation of

148 - object flag bits, which has been corrected.

145

150 - Will merge to 'master'.

146 +* es/get-worktrees-unsort (2020-06-22) 2 commits

147 + (merged to 'next' on 2020-06-25 at 732d93a36d)

148 + + worktree: drop get_worktrees() unused 'flags' argument

149 + + worktree: drop get_worktrees() special-purpose sorting option

150

151 + API cleanup for get_worktrees()

152

153 -* cb/is-descendant-of (2020-06-23) 1 commit

154 - (merged to 'next' on 2020-06-25 at ca0eaddf5e)

155 - + commit-reach: avoid is_descendant_of() shim

153

157 - Code clean-up.

154 +* es/worktree-code-cleanup (2020-06-24) 1 commit

155 + (merged to 'next' on 2020-06-29 at 91ffd85a31)

156 + + worktree: avoid dead-code in conditional

157

159 - Will merge to 'master'.

158 + Code cleanup.

159

160

161 * jk/fast-export-anonym-alt (2020-06-25) 11 commits

@@ -176,19 +175,43 @@ repositories listed at

175 "git fast-export --anonymize" learned to take customized mapping to

176 allow its users to tweak its output more usable for debugging.

177

179 - Will merge to 'master'.

178 +

179 +* jl/complete-git-prune (2020-06-22) 1 commit

180 + (merged to 'next' on 2020-06-25 at 64f30b8b41)

181 + + bash-completion: add git-prune into bash completion

182 +

183 + Add "git prune" to the completion (in contrib/), which could be

184 + typed by end-users from the command line.

185 +

186 +

187 +* js/default-branch-name (2020-06-30) 10 commits

188 + (merged to 'next' on 2020-07-01 at c38d766581)

189 + + contrib: subtree: adjust test to change in fmt-merge-msg

190 + (merged to 'next' on 2020-06-25 at 8f962f9219)

191 + + testsvn: respect `init.defaultBranch`

192 + + remote: use the configured default branch name when appropriate

193 + + clone: use configured default branch name when appropriate

194 + + init: allow setting the default for the initial branch name via the config

195 + + init: allow specifying the initial branch name for the new repository

196 + + docs: add missing diamond brackets

197 + + submodule: fall back to remote's HEAD for missing remote.<name>.branch

198 + + send-pack/transport-helper: avoid mentioning a particular branch

199 + + fmt-merge-msg: stop treating `master` specially

200 +

201 + The name of the primary branch in existing repositories, and the

202 + default name used for the first branch in newly created

203 + repositories, is made configurable, so that we can eventually wean

204 + ourselves off of the hardcoded 'master'.

205

206

182 -* js/diff-files-i-t-a-fix-for-difftool (2020-06-25) 2 commits

183 - (merged to 'next' on 2020-06-29 at a9fcdabb3b)

207 +* js/diff-files-i-t-a-fix-for-difftool (2020-07-01) 2 commits

208 + (merged to 'next' on 2020-07-01 at cf5d18635c)

209 + difftool -d: ensure that intent-to-add files are handled correctly

210 + diff-files --raw: show correct post-image of intent-to-add files

211

212 "git difftool" has trouble dealing with paths added to the index

213 with the intent-to-add bit.

214

190 - Will merge to 'master'.

191 -

215

216 * js/pu-to-seen (2020-06-25) 3 commits

217 (merged to 'next' on 2020-06-25 at 06c9e5a885)

@@ -199,8 +222,6 @@ repositories listed at

222 The documentation and some tests have been adjusted for the recent

223 renaming of "pu" branch to "seen".

224

202 - Will merge to 'master'.

203 -

225

226 * mk/pb-pretty-email-without-domain-part-fix (2020-06-23) 1 commit

227 (merged to 'next' on 2020-06-25 at 81a3aca46b)

@@ -208,18 +229,12 @@ repositories listed at

229

230 Docfix.

231

211 - Will merge to 'master'.

212 -

232

214 -* mf/submodule-summary-with-correct-repository (2020-06-24) 2 commits

215 - - submodule: use submodule repository when preparing summary

216 - - revision: use repository from rev_info when parsing commits

217 -

218 - "git diff/show" on a change that involves a submodule used to read

219 - the information on commits in the submodule from a wrong repository

220 - and gave a wrong information when the commit-graph is involved.

233 +* ps/ref-transaction-hook (2020-06-19) 1 commit

234 + (merged to 'next' on 2020-06-22 at 3a23dcdbdc)

235 + + refs: implement reference transaction hook

236

222 - Needs tests.

237 + A new hook.

238

239

240 * rs/pack-bits-in-object-better (2020-06-24) 1 commit

@@ -229,128 +244,212 @@ repositories listed at

244 By renumbering object flag bits, "struct object" managed to lose

245 bloated inter-field padding.

246

232 - Will merge to 'master'.

247 +--------------------------------------------------

248 +[New Topics]

249

250 +* bw/fail-cloning-into-non-empty (2020-07-06) 1 commit

251 + - git clone: don't clone into non-empty directory

252

235 -* es/worktree-code-cleanup (2020-06-24) 1 commit

236 - (merged to 'next' on 2020-06-29 at 91ffd85a31)

237 - + worktree: avoid dead-code in conditional

253 + "git clone --separate-git-dir=$elsewhere" used to stomp on the

254 + contents of the existing directory $elsewhere, which has been

255 + taught to fail when $elsewhere is not an empty directory.

256

239 - Code cleanup.

257 + Expecting a reroll.

258 + with improved commit log message.

259 + cf. <xmqqv9j08dxo.fsf@gitster.c.googlers.com>

260 +

261 +

262 +* cc/cat-file-usage-update (2020-07-01) 1 commit

263 + (merged to 'next' on 2020-07-06 at 270769a009)

264 + + cat-file: add missing [=<format>] to usage/synopsis

265 +

266 + Doc/usage update.

267

268 Will merge to 'master'.

269

270

244 -* ra/send-email-in-reply-to-from-command-line-wins (2020-06-29) 1 commit

245 - - send-email: restore --in-reply-to superseding behavior

271 +* cc/pretty-contents-size (2020-07-06) 2 commits

272 + - ref-filter: add support for %(contents:size)

273 + - Documentation: clarify %(contents:XXXX) doc

274

247 - "git send-email --in-reply-to=<msg>" did not use the In-Reply-To:

248 - header with the value given from the command line, and let it be

249 - overridden by the value on In-Reply-To: header in the messages

250 - being sent out (if exists).

275 + "git for-each-ref --format=<>" learned %(contents:size).

276

252 - Expecting a reroll/redesign.

253 - Does not behave sensibly for a multi-patch series.

277 + Expecting a reroll.

278 + with clarified semantics and tests with various object types

279 + cf. <xmqq5zb09v4e.fsf@gitster.c.googlers.com>

280

281

256 -* en/sparse-status (2020-06-22) 3 commits

257 - (merged to 'next' on 2020-06-25 at 404403a075)

258 - + git-prompt: include sparsity state as well

259 - + git-prompt: document how in-progress operations affect the prompt

260 - + wt-status: show sparse checkout status as well

282 +* ls/mergetool-meld-auto-merge (2020-07-06) 1 commit

283 + - Support auto-merge for meld to follow the vim-diff behavior

284

262 - "git status" learned to report the status of sparse checkout.

285 + The 'meld' backend of the "git mergetool" learned to give the

286 + underlying 'meld' the '--auto-merge' option, which would help

287 + reduce the amount of text that requires manual merging.

288

264 - Will merge to 'master'.

289 + Expecting a reroll.

290 + with final fixes and readability improvements.

291 + cf. <xmqqzh8c8eda.fsf@gitster.c.googlers.com>

292

293

267 -* ps/ref-transaction-hook (2020-06-19) 1 commit

268 - (merged to 'next' on 2020-06-22 at 3a23dcdbdc)

269 - + refs: implement reference transaction hook

294 +* pb/log-rev-list-doc (2020-07-06) 4 commits

295 + - rev-list-description.txt: fix Asciidoc syntax

296 + - git-log.txt: include rev-list-description.txt

297 + - git-rev-list.txt: move description to separate file

298 + - git-log.txt: add links to 'rev-list' and 'diff' docs

299

271 - A new hook.

300 + "git help log" has been enhanced by sharing more material from the

301 + documentation for the underlying "git rev-list" command.

302

273 - Will merge to 'master'.

303 + Expecting a reroll.

304 + with reordering the last one to make it preparatory clean-up.

305 + cf. <xmqqimf087pq.fsf@gitster.c.googlers.com>

306 + cf. <xmqqeepo87hz.fsf@gitster.c.googlers.com>

307

308

276 -* bc/sha-256-cvs-svn-updates (2020-06-22) 14 commits

277 - (merged to 'next' on 2020-06-25 at e0d1c87fcf)

278 - + git-cvsexportcommit: port to SHA-256

279 - + git-cvsimport: port to SHA-256

280 - + git-cvsserver: port to SHA-256

281 - + git-svn: set the OID length based on hash algorithm

282 - + perl: make SVN code hash independent

283 - + perl: make Git::IndexInfo work with SHA-256

284 - + perl: create and switch variables for hash constants

285 - + t/lib-git-svn: make hash size independent

286 - + t9101: make hash independent

287 - + t9104: make hash size independent

288 - + t9100: make test work with SHA-256

289 - + t9108: make test hash independent

290 - + t9168: make test hash independent

291 - + t9109: make test hash independent

292 - (this branch uses bc/sha-256-part-2; is tangled with hn/reftable.)

309 +* rs/line-log-until (2020-07-06) 1 commit

310 + (merged to 'next' on 2020-07-06 at 84d6bc67fa)

311 + + revision: disable min_age optimization with line-log

312

294 - CVS/SVN interface have been prepared for SHA-256 transition

313 + "git log -Lx,y:path --before=date" lost track of where the range

314 + should be because it didn't take the changes made by the youngest

315 + commits that are omitted from the output into account.

316

317 Will merge to 'master'.

318

319

299 -* ds/commit-graph-bloom-updates (2020-06-26) 10 commits

300 - - commit-graph: check all leading directories in changed path Bloom filters

301 - - revision: empty pathspecs should not use Bloom filters

302 - - revision.c: fix whitespace

303 - - commit-graph: check chunk sizes after writing

304 - - commit-graph: simplify chunk writes into loop

305 - - commit-graph: unify the signatures of all write_graph_chunk_*() functions

306 - - commit-graph: persist existence of changed-paths

307 - - bloom: fix logic in get_bloom_filter()

308 - - commit-graph: change test to die on parse, not load

309 - - commit-graph: place bloom_settings in context

310 - (this branch uses sg/commit-graph-cleanups.)

320 +* ta/wait-on-aliased-commands-upon-signal (2020-07-06) 2 commits

321 + - Wait for child on signal death for aliases to externals

322 + - Wait for child on signal death for aliases to builtins

323

312 - Updates to the changed-paths bloom filter.

324 + When an aliased command, whose output is piped to a pager by git,

325 + gets killed by a signal, the pager got into a funny state, which

326 + has been corrected (again).

327 +

328 + Expecting a reroll.

329 + with improved log messages.

330 + cf. <20200706211403.GB85133@coredump.intra.peff.net>

331

314 - Will merge to 'next'.

332

333 +* tb/upload-pack-filters (2020-07-06) 4 commits

334 + - upload-pack.c: introduce 'uploadpack.filter.tree.maxDepth'

335 + - upload-pack.c: pass 'struct list_objects_filter_options *'

336 + - upload-pack.c: allow banning certain object filter(s)

337 + - list_objects_filter_options: introduce 'list_object_filter_config_name'

338

317 -* es/get-worktrees-unsort (2020-06-22) 2 commits

318 - (merged to 'next' on 2020-06-25 at 732d93a36d)

319 - + worktree: drop get_worktrees() unused 'flags' argument

320 - + worktree: drop get_worktrees() special-purpose sorting option

339 + The component to respond to "git fetch" request is made more

340 + configurable to selectively allow or reject object filtering

341 + specification used for partial cloning.

342 +

343 + Waiting for reviews.

344 + cf. <cover.1593720075.git.me@ttaylorr.com>

345

322 - API cleanup for get_worktrees()

346 +

347 +* vs/completion-with-set-u (2020-07-01) 1 commit

348 + (merged to 'next' on 2020-07-06 at 72b1eca511)

349 + + completion: nounset mode fixes

350 +

351 + The command line completion support (in contrib/) used to be

352 + prepared to work with "set -u" but recent changes got a bit more

353 + sloppy. This has been corrected.

354

355 Will merge to 'master'.

356

357 +--------------------------------------------------

358 +[Stalled]

359

327 -* jl/complete-git-prune (2020-06-22) 1 commit

328 - (merged to 'next' on 2020-06-25 at 64f30b8b41)

329 - + bash-completion: add git-prune into bash completion

360 +* dr/push-remoteref-fix (2020-04-23) 1 commit

361 + - remote.c: fix handling of %(push:remoteref)

362

331 - Add "git prune" to the completion (in contrib/), which could be

332 - typed by end-users from the command line.

363 + The "%(push:remoteref)" placeholder in the "--format=" argument of

364 + "git format-patch" (and friends) only showed what got explicitly

365 + configured, not what ref at the receiving end would be updated when

366 + "git push" was used, as it ignored the default behaviour (e.g. update

367 + the same ref as the source).

368 +

369 + Expecting a reroll.

370 + cf. <20200416152145.wp2zeibxmuyas6y6@feanor>

371 +

372 +

373 +* mr/bisect-in-c-2 (2020-04-23) 12 commits

374 + - bisect--helper: retire `--bisect-autostart` subcommand

375 + - bisect--helper: retire `--write-terms` subcommand

376 + - bisect--helper: retire `--check-expected-revs` subcommand

377 + - bisect--helper: reimplement `bisect_state` & `bisect_head` shell functions in C

378 + - bisect--helper: retire `--next-all` subcommand

379 + - bisect--helper: retire `--bisect-clean-state` subcommand

380 + - bisect--helper: finish porting `bisect_start()` to C

381 + - bisect--helper: reimplement `bisect_next` and `bisect_auto_next` shell functions in C

382 + - bisect--helper: reimplement `bisect_autostart` shell function in C

383 + - bisect--helper: introduce new `write_in_file()` function

384 + - bisect--helper: use '-res' in 'cmd_bisect__helper' return

385 + - bisect--helper: fix `cmd_*()` function switch default return

386 +

387 + Rewrite of the remainder of "git bisect" script in C continues.

388 +

389 + Expecting a response to reviews.

390 + cf. <nycvar.QRO.7.76.6.2005230007260.56@tvgsbejvaqbjf.bet>

391 +

392 +

393 +* mk/use-size-t-in-zlib (2018-10-15) 1 commit

394 + - zlib.c: use size_t for size

395 +

396 + The wrapper to call into zlib followed our long tradition to use

397 + "unsigned long" for sizes of regions in memory, which have been

398 + updated to use "size_t".

399 +

400 +--------------------------------------------------

401 +[Cooking]

402 +

403 +* mt/hash-to-hex-thread-safety (2020-06-26) 2 commits

404 + - hex: make hash_to_hex_algop() and friends thread-safe

405 + - compat/win32/pthread: add pthread_once()

406 +

407 + hash_to_hex() used a set of rotating static buffers, which was not

408 + safe to use in a threaded environment. This has been made safer by

409 + using thread-local storage.

410 +

411 + Expecting a reroll.

412 + cf. <CAHd-oW6A2aBHg80R9kyifvF7thwzam5EYYoN9d2TaBcHJrcKWw@mail.gmail.com>

413 +

414 +

415 +* mf/submodule-summary-with-correct-repository (2020-06-24) 2 commits

416 + - submodule: use submodule repository when preparing summary

417 + - revision: use repository from rev_info when parsing commits

418 +

419 + "git diff/show" on a change that involves a submodule used to read

420 + the information on commits in the submodule from a wrong repository

421 + and gave a wrong information when the commit-graph is involved.

422 +

423 + Needs tests.

424 +

425 +

426 +* ra/send-email-in-reply-to-from-command-line-wins (2020-07-01) 1 commit

427 + (merged to 'next' on 2020-07-06 at 3892a1ab88)

428 + + send-email: restore --in-reply-to superseding behavior

429 +

430 + "git send-email --in-reply-to=<msg>" did not use the In-Reply-To:

431 + header with the value given from the command line, and let it be

432 + overridden by the value on In-Reply-To: header in the messages

433 + being sent out (if exists).

434

435 Will merge to 'master'.

436

437

337 -* js/default-branch-name (2020-06-24) 9 commits

338 - (merged to 'next' on 2020-06-25 at 8f962f9219)

339 - + testsvn: respect `init.defaultBranch`

340 - + remote: use the configured default branch name when appropriate

341 - + clone: use configured default branch name when appropriate

342 - + init: allow setting the default for the initial branch name via the config

343 - + init: allow specifying the initial branch name for the new repository

344 - + docs: add missing diamond brackets

345 - + submodule: fall back to remote's HEAD for missing remote.<name>.branch

346 - + send-pack/transport-helper: avoid mentioning a particular branch

347 - + fmt-merge-msg: stop treating `master` specially

348 - (this branch is used by hn/reftable.)

438 +* ds/commit-graph-bloom-updates (2020-07-01) 10 commits

439 + (merged to 'next' on 2020-07-06 at 177e6b362e)

440 + + commit-graph: check all leading directories in changed path Bloom filters

441 + + revision: empty pathspecs should not use Bloom filters

442 + + revision.c: fix whitespace

443 + + commit-graph: check chunk sizes after writing

444 + + commit-graph: simplify chunk writes into loop

445 + + commit-graph: unify the signatures of all write_graph_chunk_*() functions

446 + + commit-graph: persist existence of changed-paths

447 + + bloom: fix logic in get_bloom_filter()

448 + + commit-graph: change test to die on parse, not load

449 + + commit-graph: place bloom_settings in context

450 + (this branch uses sg/commit-graph-cleanups.)

451

350 - The name of the primary branch in existing repositories, and the

351 - default name used for the first branch in newly created

352 - repositories, is made configurable, so that we can eventually wean

353 - ourselves off of the hardcoded 'master'.

452 + Updates to the changed-paths bloom filter.

453

454 Will merge to 'master'.

455

@@ -370,42 +469,17 @@ repositories listed at

469 Waiting for a (hopefully final) review.

470

471

373 -* ak/commit-graph-to-slab (2020-06-17) 4 commits

374 - (merged to 'next' on 2020-06-25 at 9a10ae18d3)

375 - + commit-graph: minimize commit_graph_data_slab access

376 - + commit: move members graph_pos, generation to a slab

377 - + commit-graph: introduce commit_graph_data_slab

378 - + object: drop parsed_object_pool->commit_count

379 -

380 - A few fields in "struct commit" that do not have to always be

381 - present have been moved to commit slabs.

382 -

383 - Will merge to 'master'.

384 -

385 -

386 -* dl/test-must-fail-fixes-5 (2020-06-24) 4 commits

387 - (merged to 'next' on 2020-06-29 at efe53a7588)

388 - + lib-submodule-update: pass 'test_must_fail' as an argument

389 - + lib-submodule-update: prepend "git" to $command

390 - + lib-submodule-update: consolidate --recurse-submodules

391 - + lib-submodule-update: add space after function name

392 -

393 - The effort to avoid using test_must_fail on non-git command continues.

394 -

395 - Will merge to 'master'.

396 -

397 -

472 * sg/commit-graph-cleanups (2020-06-08) 10 commits

399 - - commit-graph: simplify write_commit_graph_file() #2

400 - - commit-graph: simplify write_commit_graph_file() #1

401 - - commit-graph: simplify parse_commit_graph() #2

402 - - commit-graph: simplify parse_commit_graph() #1

403 - - commit-graph: clean up #includes

404 - - diff.h: drop diff_tree_oid() & friends' return value

405 - - commit-slab: add a function to deep free entries on the slab

406 - - commit-graph-format.txt: all multi-byte numbers are in network byte order

407 - - commit-graph: fix parsing the Chunk Lookup table

408 - - tree-walk.c: don't match submodule entries for 'submod/anything'

473 + + commit-graph: simplify write_commit_graph_file() #2

474 + + commit-graph: simplify write_commit_graph_file() #1

475 + + commit-graph: simplify parse_commit_graph() #2

476 + + commit-graph: simplify parse_commit_graph() #1

477 + + commit-graph: clean up #includes

478 + + diff.h: drop diff_tree_oid() & friends' return value

479 + + commit-slab: add a function to deep free entries on the slab

480 + + commit-graph-format.txt: all multi-byte numbers are in network byte order

481 + + commit-graph: fix parsing the Chunk Lookup table

482 + + tree-walk.c: don't match submodule entries for 'submod/anything'

483 (this branch is used by ds/commit-graph-bloom-updates.)

484

485 The changed-path Bloom filter is improved using ideas from an

@@ -455,59 +529,6 @@ repositories listed at

529 cf. <CABPp-BGdEyEeajYZj_rdxp=MyEQdszuyjVTax=hhYj3fOtRQUQ@mail.gmail.com>

530

531

458 -* bc/sha-256-part-2 (2020-06-19) 44 commits

459 - (merged to 'next' on 2020-06-22 at dc153f366d)

460 - + remote-testgit: adapt for object-format

461 - + bundle: detect hash algorithm when reading refs

462 - + t5300: pass --object-format to git index-pack

463 - + t5704: send object-format capability with SHA-256

464 - + t5703: use object-format serve option

465 - + t5702: offer an object-format capability in the test

466 - + t/helper: initialize the repository for test-sha1-array

467 - + remote-curl: avoid truncating refs with ls-remote

468 - + t1050: pass algorithm to index-pack when outside repo

469 - + builtin/index-pack: add option to specify hash algorithm

470 - + remote-curl: detect algorithm for dumb HTTP by size

471 - + builtin/ls-remote: initialize repository based on fetch

472 - + t5500: make hash independent

473 - + serve: advertise object-format capability for protocol v2

474 - + connect: parse v2 refs with correct hash algorithm

475 - + connect: pass full packet reader when parsing v2 refs

476 - + Documentation/technical: document object-format for protocol v2

477 - + t1302: expect repo format version 1 for SHA-256

478 - + builtin/show-index: provide options to determine hash algo

479 - + t5302: modernize test formatting

480 - + packfile: compute and use the index CRC offset

481 - + t3200: mark assertion with SHA1 prerequisite

482 - + setup: set the_repository's hash algo when checking format

483 - + fetch-pack: parse and advertise the object-format capability

484 - + t5562: pass object-format in synthesized test data

485 - + builtin/clone: initialize hash algorithm properly

486 - + remote-curl: implement object-format extensions

487 - + transport-helper: implement object-format extensions

488 - + docs: update remote helper docs for object-format extensions

489 - + builtin/receive-pack: detect when the server doesn't support our hash

490 - + connect: detect algorithm when fetching refs

491 - + fetch-pack: detect when the server doesn't support our hash

492 - + connect: make parse_feature_value extern

493 - + send-pack: detect when the server doesn't support our hash

494 - + connect: add function to detect supported v1 hash functions

495 - + transport: add a hash algorithm member

496 - + pkt-line: add a member for hash algorithm

497 - + connect: add function to fetch value of a v2 server capability

498 - + connect: add function to parse multiple v1 capability values

499 - + remote: advertise the object-format capability on the server side

500 - + wrapper: add function to compare strings with different NUL termination

501 - + connect: have ref processing code take struct packet_reader

502 - + Documentation: document v1 protocol object-format capability

503 - + t1050: match object ID paths in a hash-insensitive way

504 - (this branch is used by bc/sha-256-cvs-svn-updates and hn/reftable.)

505 -

506 - SHA-256 migration work continues.

507 -

508 - Will merge to 'master'.

509 -

510 -

532 * jx/proc-receive-hook (2020-05-18) 11 commits

533 - doc: add documentation for the proc-receive hook

534 - transport: parse report options for tracking refs

@@ -527,36 +548,20 @@ repositories listed at

548 Needs review.

549

550

530 -* hn/reftable (2020-06-29) 23 commits

531 - - fixup! Reftable support for git-core

532 - - Add "test-tool dump-reftable" command.

533 - - Add reftable testing infrastructure

534 - - git-prompt: prepare for reftable refs backend

535 - - vcxproj: adjust for the reftable changes

536 - - Add GIT_DEBUG_REFS debugging mechanism

537 - - Hookup unittests for the reftable library.

538 - - Reftable support for git-core

539 - - Add standalone build infrastructure for reftable

540 - - Add reftable library

541 - - Add .gitattributes for the reftable/ directory

542 - - Iterate over the "refs/" namespace in for_each_[raw]ref

543 - - Move REF_LOG_ONLY to refs-internal.h

544 - - Treat REVERT_HEAD as a pseudo ref

545 - - Treat CHERRY_PICK_HEAD as a pseudo ref

551 +* hn/reftable (2020-07-06) 4 commits

552 - Treat BISECT_HEAD as a pseudo ref

547 - - Make refs_ref_exists public

548 - - Write pseudorefs through ref backends.

553 - checkout: add '\n' to reflog message

554 - t3432: use git-reflog to inspect the reflog for HEAD

555 - lib-t6000.sh: write tag using git-update-ref

552 - - Merge branch 'bc/sha-256-part-2' into HEAD

553 - - Merge branch 'js/default-branch-name' into HEAD

554 - (this branch uses bc/sha-256-part-2 and js/default-branch-name; is tangled with bc/sha-256-cvs-svn-updates.)

555 -

556 - A new refs backend "reftable" to replace the traditional

557 - combination of packed-refs files and one-file-per-ref loose refs

558 - has been implemented and integrated for improved performance and

559 - atomicity.

556 +

557 + Preliminary clean-up of the refs API in preparation for adding a

558 + new refs backend "reftable".

559 +

560 + Expecting a reroll.

561 + The second one should be dropped, and possibly replaced with a

562 + patch that moves message normalization from backends to the more

563 + generic layer.

564 + cf. <xmqqmu4ca31g.fsf@gitster.c.googlers.com>

565

566 --------------------------------------------------

567 [Discarded]

What's cooking (2020/07 #01)