9125ddae14 · 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, #01; Fri, 1)

4 -X-master-at: c2c7d17b030646b40e6764ba34a5ebf66aee77af

5 -X-next-at: fd7c087b0d85b48ffb998ee0349e5bb8945e292a

3 +Subject: What's cooking in git.git (Jun 2018, #02; Mon, 4)

4 +X-master-at: 3e5524907b43337e82a24afbc822078daf7a868f

5 +X-next-at: 61856ae69a2ceb241a90e47953e18f218e4d5f2f

6

7 -What's cooking in git.git (Jun 2018, #01; Fri, 1)

7 +What's cooking in git.git (Jun 2018, #02; Mon, 4)

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

9

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

@@ -18,358 +18,114 @@ of the repositories listed at

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

19

20 --------------------------------------------------

21 -[New Topics]

22 -

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

24 - - rebase: remove -p code from git-rebase--interactive.sh

25 - - rebase: use the new git-rebase--preserve-merges.sh

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

27 - - rebase: introduce a dedicated backend for --preserve-merges

28 -

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

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

21 +[Graduated to "master"]

22

23 +* bc/t3430-fixup (2018-06-04) 1 commit

24 + (merged to 'next' on 2018-06-04 at 892aab2dc8)

25 + + t3430: test clean-up

26

33 -* jk/index-pack-maint (2018-06-01) 2 commits

34 - - index-pack: handle --strict checks of non-repo packs

35 - - prepare_commit_graft: treat non-repository as a noop

27 + Test fix.

28

37 - "index-pack --strict" has been taught to make sure that it runs the

38 - final object integrity checks after making the freshly indexed

39 - packfile available to itself.

29

41 - Will merge to 'next' and then to 'master'.

30 +* bw/refspec-api (2018-06-01) 1 commit

31 + (merged to 'next' on 2018-06-04 at f244fe357b)

32 + + refspec-api: avoid uninitialized field in refspec item

33

34 + Hotfix.

35

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

45 - - completion: complete remote names too

36

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

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

49 - has been taught about it.

37 +* jt/submodule-pull-recurse-rebase (2018-05-25) 1 commit

38 + (merged to 'next' on 2018-06-01 at cd3e2a1008)

39 + + submodule: do not pass null OID to setup_revisions

40

51 - Will merge to and cook in 'next'.

41 + "git pull -recurse-submodules --rebase", when the submodule

42 + repository's history did not have anything common between ours and

43 + the upstream's, failed to execute. We need to fetch from them to

44 + continue even in such a case.

45

46

54 -* nd/remote-update-doc (2018-06-01) 1 commit

55 - - remote.txt: update documentation for 'update' command

47 +* nd/remote-update-doc (2018-06-04) 2 commits

48 + (merged to 'next' on 2018-06-04 at b2901b60ad)

49 + + remote: doc typofix

50 + (merged to 'next' on 2018-06-04 at 82cce447e5)

51 + + remote.txt: update documentation for 'update' command

52

53 "git remote update" can take both a single remote nickname and a

54 nickname for remote groups, but only one of them was documented.

55

60 - Will merge to 'next' and then to 'master'.

61 -

56

57 * rd/p4-doc-markup-env (2018-06-01) 1 commit

64 - - p4.txt: Use backquotes for variable names

58 + (merged to 'next' on 2018-06-04 at 80212295cf)

59 + + p4.txt: Use backquotes for variable names

60

61 Doc markup update.

62

68 - Will merge to 'next' and then to 'master'.

69 -

70 -

71 -* sb/plug-misc-leaks (2018-06-01) 2 commits

72 - - submodule--helper: plug mem leak in print_default_remote

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

74 -

75 - Misc leak plugging.

76 -

77 - Will merge to 'next' and then to 'master'.

78 -

63

64 * tg/doc-sec-list (2018-06-01) 2 commits

81 - - note git-security@googlegroups.com in more places

82 - - SubmittingPatches: replace numbered attributes with names

65 + (merged to 'next' on 2018-06-04 at e1f80ffe09)

66 + + note git-security@googlegroups.com in more places

67 + + SubmittingPatches: replace numbered attributes with names

68

69 Doc update.

70

86 - Will merge to 'next' and then to 'master'.

87 -

88 -* bw/refspec-api (2018-06-01) 1 commit

89 - - refspec-api: avoid uninitialized field in refspec item

90 -

91 - Hotfix.

92 -

93 - Will merge to 'next' and then to 'master'.

94 -

71 --------------------------------------------------

96 -[Graduated to "master"]

97 -

98 -* bw/refspec-api (2018-05-18) 36 commits

99 - (merged to 'next' on 2018-05-24 at 83d361419e)

100 - + submodule: convert push_unpushed_submodules to take a struct refspec

101 - + remote: convert check_push_refs to take a struct refspec

102 - + remote: convert match_push_refs to take a struct refspec

103 - + http-push: store refspecs in a struct refspec

104 - + transport: remove transport_verify_remote_names

105 - + send-pack: store refspecs in a struct refspec

106 - + transport: convert transport_push to take a struct refspec

107 - + push: convert to use struct refspec

108 - + push: check for errors earlier

109 - + remote: convert match_explicit_refs to take a struct refspec

110 - + remote: convert get_ref_match to take a struct refspec

111 - + remote: convert query_refspecs to take a struct refspec

112 - + remote: convert apply_refspecs to take a struct refspec

113 - + remote: convert get_stale_heads to take a struct refspec

114 - + fetch: convert prune_refs to take a struct refspec

115 - + fetch: convert get_ref_map to take a struct refspec

116 - + fetch: convert do_fetch to take a struct refspec

117 - + refspec: remove the deprecated functions

118 - + fetch: convert refmap to use struct refspec

119 - + fetch: convert fetch_one to use struct refspec

120 - + transport-helper: convert to use struct refspec

121 - + remote: remove add_prune_tags_to_fetch_refspec

122 - + remote: convert fetch refspecs to struct refspec

123 - + remote: convert push refspecs to struct refspec

124 - + fast-export: convert to use struct refspec

125 - + clone: convert cmd_clone to use refspec_item_init

126 - + remote: convert match_push_refs to use struct refspec

127 - + remote: convert check_push_refs to use struct refspec

128 - + transport: convert transport_push to use struct refspec

129 - + pull: convert get_tracking_branch to use refspec_item_init

130 - + submodule--helper: convert push_check to use struct refspec

131 - + refspec: convert valid_fetch_refspec to use parse_refspec

132 - + refspec: introduce struct refspec

133 - + refspec: factor out parsing a single refspec

134 - + refspec: rename struct refspec to struct refspec_item

135 - + refspec: move refspec parsing logic into its own file

136 - (this branch is used by bw/ref-prefix-for-configured-refspec.)

137 -

138 - An API update for handling of refspecs used by fetch & push

139 - codepath.

140 -

141 -

142 -* bc/hash-independent-tests (2018-05-21) 28 commits

143 - (merged to 'next' on 2018-05-24 at 40fb709f7c)

144 - + t5300: abstract away SHA-1-specific constants

145 - + t4208: abstract away SHA-1-specific constants

146 - + t4045: abstract away SHA-1-specific constants

147 - + t4042: abstract away SHA-1-specific constants

148 - + t4205: sort log output in a hash-independent way

149 - + t/lib-diff-alternative: abstract away SHA-1-specific constants

150 - + t4030: abstract away SHA-1-specific constants

151 - + t4029: abstract away SHA-1-specific constants

152 - + t4029: fix test indentation

153 - + t4022: abstract away SHA-1-specific constants

154 - + t4020: abstract away SHA-1-specific constants

155 - + t4014: abstract away SHA-1-specific constants

156 - + t4008: abstract away SHA-1-specific constants

157 - + t4007: abstract away SHA-1-specific constants

158 - + t3905: abstract away SHA-1-specific constants

159 - + t3702: abstract away SHA-1-specific constants

160 - + t3103: abstract away SHA-1-specific constants

161 - + t2203: abstract away SHA-1-specific constants

162 - + t: skip pack tests if not using SHA-1

163 - + t4044: skip test if not using SHA-1

164 - + t1512: skip test if not using SHA-1

165 - + t1007: annotate with SHA1 prerequisite

166 - + t0000: annotate with SHA1 prerequisite

167 - + t: switch $_x40 to $OID_REGEX

168 - + t/test-lib: introduce OID_REGEX

169 - + t: switch $_z40 to $ZERO_OID

170 - + t/test-lib: introduce ZERO_OID

171 - + t/test-lib: add an SHA1 prerequisite

172 -

173 - Many tests hardcode the raw object names, which would change once

174 - we migrate away from SHA-1. While some of them must test against

175 - exact object names, most of them do not have to use hardcoded

176 - constants in the test. The latter kind of tests have been updated

177 - to test the moral equivalent of the original without hardcoding the

178 - actual object names.

179 -

180 -

181 -* bw/ref-prefix-for-configured-refspec (2018-05-18) 2 commits

182 - (merged to 'next' on 2018-05-24 at a0d4af6180)

183 - + fetch: generate ref-prefixes when using a configured refspec

184 - + refspec: consolidate ref-prefix generation logic

185 - (this branch uses bw/refspec-api.)

186 -

187 - "git fetch $there $refspec" that talks over protocol v2 can take

188 - advantage of server-side ref filtering; the code has been extended

189 - so that this mechanism triggers also when fetching with configured

190 - refspec.

191 -

192 -

193 -* bw/remote-curl-compressed-responses (2018-05-23) 2 commits

194 - (merged to 'next' on 2018-05-24 at 98ee4308dc)

195 - + remote-curl: accept compressed responses with protocol v2

196 - + remote-curl: accept all encodings supported by curl

197 -

198 - Our HTTP client code used to advertise that we accept gzip encoding

199 - from the other side; instead, just let cURL library to advertise

200 - and negotiate the best one.

201 -

202 -

203 -* cc/tests-without-assuming-ref-files-backend (2018-05-30) 2 commits

204 - (merged to 'next' on 2018-05-30 at e9605bacf0)

205 - + t990X: use '.git/objects' as 'deep inside .git' path

206 - (merged to 'next' on 2018-05-24 at 19138b162f)

207 - + t: make many tests depend less on the refs being files

208 -

209 - Quite a many tests assumed that newly created refs are made as

210 - loose refs using the files backend, which have been updated to use

211 - proper plumbing like rev-parse and update-ref, to avoid breakage

212 - once we start using different ref backends.

213 -

214 -

215 -* en/rev-parse-invalid-range (2018-05-25) 1 commit

216 - (merged to 'next' on 2018-05-30 at aec27f8a94)

217 - + rev-parse: check lookup'ed commit references for NULL

218 -

219 - "git rev-parse Y..." etc. misbehaved when given endpoints were

220 - not committishes.

221 -

222 -

223 -* jk/config-blob-sans-repo (2018-05-21) 1 commit

224 - (merged to 'next' on 2018-05-24 at 3426d9508c)

225 - + config: die when --blob is used outside a repository

226 -

227 - Error codepath fix.

228 -

229 -

230 -* jk/snprintf-truncation (2018-05-21) 5 commits

231 - (merged to 'next' on 2018-05-24 at 4b78e15f66)

232 - + fmt_with_err: add a comment that truncation is OK

233 - + shorten_unambiguous_ref: use xsnprintf

234 - + fsmonitor: use internal argv_array of struct child_process

235 - + log_write_email_headers: use strbufs

236 - + http: use strbufs instead of fixed buffers

237 -

238 - Avoid unchecked snprintf() to make future code auditing easier.

239 -

240 -

241 -* js/empty-config-section-fix (2018-05-18) 1 commit

242 - (merged to 'next' on 2018-05-24 at 3d0875c7b4)

243 - + config: a user-provided invalid section is not a BUG

244 -

245 - Error codepath fix.

246 -

247 -

248 -* js/rebase-recreate-merge (2018-05-30) 2 commits

249 - (merged to 'next' on 2018-05-30 at e295dfb56e)

250 - + sequencer: ensure labels that are object IDs are rewritten

251 - + git-rebase--interactive: fix copy-paste mistake

252 -

253 - Hotfixes.

254 -

255 -

256 -* ld/p4-unshelve (2018-05-24) 1 commit

257 - (merged to 'next' on 2018-05-30 at 7367508399)

258 - + git-p4: add unshelve command

259 -

260 - "git p4" learned to "unshelve" shelved commit from P4.

261 -

262 -

263 -* ma/config-store-data-clear (2018-05-21) 3 commits

264 - (merged to 'next' on 2018-05-24 at cc6e4b9371)

265 - + config: let `config_store_data_clear()` handle `key`

266 - + config: let `config_store_data_clear()` handle `value_regex`

267 - + config: free resources of `struct config_store_data`

268 -

269 - Leak plugging.

270 -

271 -

272 -* ma/regex-no-regfree-after-comp-fail (2018-05-21) 1 commit

273 - (merged to 'next' on 2018-05-24 at 3e473da1b5)

274 - + regex: do not call `regfree()` if compilation fails

275 -

276 - We used to call regfree() after regcomp() failed in some codepaths,

277 - which have been corrected.

278 -

279 -

280 -* ma/unpack-trees-free-msgs (2018-05-22) 4 commits

281 - (merged to 'next' on 2018-05-24 at 84cbd459f7)

282 - + unpack_trees_options: free messages when done

283 - + argv-array: return the pushed string from argv_push*()

284 - + merge-recursive: provide pair of `unpack_trees_{start,finish}()`

285 - + merge: setup `opts` later in `checkout_fast_forward()`

286 -

287 - Leak plugging.

288 -

289 -

290 -* nd/command-list (2018-05-21) 17 commits

291 - (merged to 'next' on 2018-05-24 at 093e983b05)

292 - + completion: allow to customize the completable command list

293 - + completion: add and use --list-cmds=alias

294 - + completion: add and use --list-cmds=nohelpers

295 - + Move declaration for alias.c to alias.h

296 - + completion: reduce completable command list

297 - + completion: let git provide the completable command list

298 - + command-list.txt: documentation and guide line

299 - + help: use command-list.txt for the source of guides

300 - + help: add "-a --verbose" to list all commands with synopsis

301 - + git: support --list-cmds=list-<category>

302 - + completion: implement and use --list-cmds=main,others

303 - + git --list-cmds: collect command list in a string_list

304 - + git.c: convert --list-* to --list-cmds=*

305 - + Remove common-cmds.h

306 - + help: use command-list.h for common command list

307 - + generate-cmds.sh: export all commands to command-list.h

308 - + generate-cmds.sh: factor out synopsis extract code

309 - (this branch is used by nd/complete-config-vars.)

310 -

311 - The list of commands with their various attributes were spread

312 - across a few places in the build procedure, but it now is getting a

313 - bit more consolidated to allow more automation.

314 -

72 +[New Topics]

73

316 -* nd/use-opt-int-set-f (2018-05-24) 1 commit

317 - (merged to 'next' on 2018-05-30 at f8f5d612ed)

318 - + Use OPT_SET_INT_F() for cmdline option specification

74 +* ab/checkout-default-remote (2018-06-04) 8 commits

75 + - checkout & worktree: introduce checkout.defaultRemote

76 + - checkout: add advice for ambiguous "checkout <branch>"

77 + - builtin/checkout.c: use "ret" variable for return

78 + - checkout: pass the "num_matches" up to callers

79 + - checkout.c]: change "unique" member to "num_matches"

80 + - checkout.c: introduce an *_INIT macro

81 + - checkout.h: wrap the arguments to unique_tracking_name()

82 + - checkout tests: index should be clean after dwim checkout

83

320 - Code simplification.

84

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

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

87

323 -* pa/import-tars-long-names (2018-05-24) 1 commit

324 - (merged to 'next' on 2018-05-30 at 5e043b2fca)

325 - + import-tars: read overlong names from pax extended header

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

89 + point that is newer than the existing history used to end up

90 + grabbing the entire history. Such a request now errors out.

91

327 - The import-tars script (in contrib/) has been taught to handle

328 - tarballs with overly long paths that use PAX extended headers.

92

93 +* pw/add-p-recount (2018-06-04) 1 commit

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

95

331 -* rd/doc-options-placeholder (2018-05-25) 1 commit

332 - (merged to 'next' on 2018-05-30 at e3961b038f)

333 - + Use proper syntax for replaceables in command docs

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

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

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

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

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

101 + to notice the situation and cope with it.

102

335 - Docfix.

103 + Will merge to and cook in 'next'.

104

105

338 -* rd/init-typo (2018-05-30) 1 commit

339 - (merged to 'next' on 2018-05-30 at 8813a649eb)

340 - + init: fix grammar in "templates not found" msg

106 +* rd/comment-typofix-in-sha1-file (2018-06-04) 1 commit

107 + - sha1-file.c: correct $GITDIR to $GIT_DIR in a comment

108

342 - Message fix.

109 + In code comment typofix

110

111 + Will merge to 'next'.

112

345 -* rd/tag-doc-lightweight (2018-05-29) 1 commit

346 - (merged to 'next' on 2018-05-30 at 979ba9f720)

347 - + tag: clarify in the doc that a tag can refer to a non-commit object

113

349 - Docfix.

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

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

116

117 + Code cleanup.

118

352 -* sb/grep-die-on-unreadable-index (2018-05-16) 1 commit

353 - (merged to 'next' on 2018-05-24 at 6d78916503)

354 - + grep: handle corrupt index files early

119 + Will merge to 'next'.

120

356 - Error behaviour of "git grep" when it cannot read the index was

357 - inconsistent with other commands that uses the index, which has

358 - been corrected to error out early.

121

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

123 + - t9104: kosherly remove remote refs

124

361 -* sb/submodule-merge-in-merge-recursive (2018-05-18) 3 commits

362 - (merged to 'next' on 2018-05-24 at 6228d22182)

363 - + merge-recursive: give notice when submodule commit gets fast-forwarded

364 - + merge-recursive: i18n submodule merge output and respect verbosity

365 - + submodule.c: move submodule merging to merge-recursive.c

125 + Instead of mucking with filesystem directly, use plumbing commands

126 + update-ref etc. to manipulate the refs in the tests.

127

367 - By code restructuring of submodule merge in merge-recursive,

368 - informational messages from the codepath are now given using the

369 - same mechanism as other output, and honor the merge.verbosity

370 - configuration. The code also learned to give a few new messages

371 - when a submodule three-way merge resolves cleanly when one side

372 - records a descendant of the commit chosen by the other side.

128 + Will merge to 'next'.

129

130 --------------------------------------------------

131 [Stalled]

@@ -467,10 +223,9 @@ of the repositories listed at

223 cf. <0102015f5e5ee171-f30f4868-886f-47a1-a4e4-b4936afc545d-000000@eu-west-1.amazonses.com>

224

225

470 -* mk/http-backend-content-length (2017-11-27) 4 commits

471 - - SQUASH???

472 - - t5560-http-backend-noserver.sh: add CONTENT_LENGTH cases

226 +* mk/http-backend-content-length (2018-06-04) 3 commits

227 - SQUASH???

228 + - http-backend: respect CONTENT_LENGTH for receive-pack

229 - http-backend: respect CONTENT_LENGTH as specified by rfc3875

230

231 The http-backend (used for smart-http transport) used to slurp the

@@ -478,13 +233,6 @@ of the repositories listed at

233 that is supplied in the environment and instead expecting the Web

234 server to close the input stream. This has been fixed.

235

481 - Expecting a reroll.

482 - Suggested fixes to be used when rerolling is queued, but I'd

483 - prefer _not_ squashing them myself.

484 -

485 - Also, it may be too complex solution for the problem.

486 - cf. <20171204171308.GA13332@sigill.intra.peff.net>

487 -

236

237 * jk/drop-ancient-curl (2017-08-09) 5 commits

238 - http: #error on too-old curl

@@ -511,6 +259,53 @@ of the repositories listed at

259 --------------------------------------------------

260 [Cooking]

261

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

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

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

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

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

267 +

268 + Separate "rebase -p" codepath out of "rebase -i" implementation to

269 + slim down the latter and make it easier to manage.

270 +

271 + Will merge to 'next'.

272 +

273 +

274 +* jk/index-pack-maint (2018-06-01) 2 commits

275 + (merged to 'next' on 2018-06-04 at c553a485e8)

276 + + index-pack: handle --strict checks of non-repo packs

277 + + prepare_commit_graft: treat non-repository as a noop

278 +

279 + "index-pack --strict" has been taught to make sure that it runs the

280 + final object integrity checks after making the freshly indexed

281 + packfile available to itself.

282 +

283 + Will cook in 'next'.

284 +

285 +

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

287 + - completion: complete remote names too

288 +

289 + "git remote update" can take both a single remote nickname and a

290 + nickname for remote groups, and the completion script (in contrib/)

291 + has been taught about it.

292 +

293 + Will merge to and cook in 'next'.

294 +

295 +

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

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

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

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

300 + + sequencer.c: plug leaks in do_pick_commit

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

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

303 +

304 + Misc leak plugging.

305 +

306 + Will merge to 'next'.

307 +

308 +

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

310 - commit-graph: update design document

311 - gc: automatically write commit-graph files

@@ -561,18 +356,6 @@ of the repositories listed at

356 Will cook in 'next'.

357

358

564 -* jt/submodule-pull-recurse-rebase (2018-05-25) 1 commit

565 - (merged to 'next' on 2018-06-01 at cd3e2a1008)

566 - + submodule: do not pass null OID to setup_revisions

567 -

568 - "git pull -recurse-submodules --rebase", when the submodule

569 - repository's history did not have anything common between ours and

570 - the upstream's, failed to execute. We need to fetch from them to

571 - continue even in such a case.

572 -

573 - Will merge to 'master'.

574 -

575 -

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

360 - usage: prefix all lines in `vreportf()`, not just the first

361 - usage: extract `prefix_suffix_lines()` from `advise()`

@@ -580,7 +363,7 @@ of the repositories listed at

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

364 error(), and friends more easily translatable.

365

583 - Wait for discussion to settle.

366 + Waiting for the discussion to settle.

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

368

369

@@ -633,8 +416,6 @@ of the repositories listed at

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

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

418

636 - Is this ready for 'next'?

637 -

419

420 * rm/p4-submit-with-commit-option (2018-05-21) 1 commit

421 - git-p4: add options --commit and --disable-rebase

What's cooking (2018/06 #02)