What's cooking (2017/08 #05)

Junio C Hamano committed Aug 22, 2017 at 11:31 UTC dea9677694744e03fdb880aebb97b3605458af16
1 file changed +571 -431
whats-cooking.txt
+571 -431
@@ -1,10 +1,10 @@
1 To: git@vger.kernel.org
2 Bcc: lwn@lwn.net
3 -Subject: What's cooking in git.git (Aug 2017, #04; Fri, 18)
4 -X-master-at: b3622a4ee94e4916cd05e6d96e41eeb36b941182
5 -X-next-at: d436f9bbced11cd7ed9d109ea30eb62b4ef69635
3 +Subject: What's cooking in git.git (Aug 2017, #05; Tue, 22)
4 +X-master-at: f0294f474e16840f4efb3f4da3329ba46f53efa7
5 +X-next-at: 238414bde168866658281957364b80816e9be329
6
7 -What's cooking in git.git (Aug 2017, #04; Fri, 18)
7 +What's cooking in git.git (Aug 2017, #05; Tue, 22)
8 --------------------------------------------------
9
10 Here are the topics that have been cooking. Commits prefixed with
@@ -12,11 +12,13 @@ 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 -The top part of the draft release notes for the next cycle, which I
16 -tentatively called Git 2.15, declares that "git add ''" will still
17 -be supported up to this release but will become illegal after that.
18 -Given that a typical release cycle is 8-12 weeks, that will happen
19 -sometime early next year.
15 +The second batch of topics are in. This cycle is going a bit slower
16 +than the previous one (as of mid-week #3 of this cycle, we have
17 +about 200 patches on 'master' since v2.14, compared to about 300
18 +patches in the cycle towards v2.14 at a similar point in the cycle),
19 +but hopefully we can catch up eventually.
20 +
21 +I am planning to be offline most of the next week, by the way.
22
23 You can find the changes described here in the integration branches
24 of the repositories listed at
@@ -24,129 +26,410 @@ of the repositories listed at
26 http://git-blame.blogspot.com/p/git-public-repositories.html
27
28 --------------------------------------------------
27 -[New Topics]
29 +[Graduated to "master"]
30
29 -* cc/subprocess-handshake-missing-capabilities (2017-08-16) 1 commit
30 - - sub-process: print the cmd when a capability is unsupported
31 +* ab/ref-filter-no-contains (2017-08-07) 1 commit
32 + (merged to 'next' on 2017-08-18 at 7ec9d3d3a2)
33 + + tests: don't give unportable ">" to "test" built-in, use -gt
34
32 - When handshake with a subprocess filter notices that the process
33 - asked for an unknown capability, Git did not report what program
34 - the offending subprocess was running. This has been corrected.
35 + A test fix.
36
36 - Will merge to 'next'.
37
38 - We may want a follow-up fix to tighten the error checking, though.
38 +* bw/clone-recursive-quiet (2017-08-04) 1 commit
39 + (merged to 'next' on 2017-08-14 at fbd4473ce4)
40 + + clone: teach recursive clones to respect -q
41
42 + "git clone --recurse-submodules --quiet" did not pass the quiet
43 + option down to submodules.
44
41 -* tb/apply-with-crlf (2017-08-17) 3 commits
42 - - SQUASH???
43 - - apply: file commited with CRLF should roundtrip diff and apply
44 - - convert: add SAFE_CRLF_KEEP_CRLF
45 - (this branch is tangled with jc/apply-with-crlf.)
45
47 - "git apply" that is used as a better "patch -p1" failed to apply a
48 - taken from a file with CRLF line endings to a file with CRLF line
49 - endings. The root cause was because it misused convert_to_git()
50 - that tried to do "safe-crlf" processing by looking at the index
51 - entry at the same path, which is a nonsense---in that mode, "apply"
52 - is not working on the data in (or derived from) the index at all.
53 - This has been fixed.
46 +* bw/grep-recurse-submodules (2017-08-02) 10 commits
47 + (merged to 'next' on 2017-08-14 at dcfcfc94af)
48 + + grep: recurse in-process using 'struct repository'
49 + + submodule: merge repo_read_gitmodules and gitmodules_config
50 + + submodule: check for unmerged .gitmodules outside of config parsing
51 + + submodule: check for unstaged .gitmodules outside of config parsing
52 + + submodule: remove fetch.recursesubmodules from submodule-config parsing
53 + + submodule: remove submodule.fetchjobs from submodule-config parsing
54 + + config: add config_from_gitmodules
55 + + cache.h: add GITMODULES_FILE macro
56 + + repository: have the_repository use the_index
57 + + repo_read_index: don't discard the index
58 + (this branch is used by bw/submodule-config-cleanup.)
59
55 - Will merge to 'next' after squashing the fix in.
60 + "git grep --recurse-submodules" has been reworked to give a more
61 + consistent output across submodule boundary (and do its thing
62 + without having to fork a separate process).
63
64
58 -* rs/t1002-do-not-use-sum (2017-08-15) 1 commit
59 - - t1002: stop using sum(1)
65 +* bw/push-options-recursively-to-submodules (2017-07-20) 1 commit
66 + (merged to 'next' on 2017-08-14 at 421dc09fd0)
67 + + submodule--helper: teach push-check to handle HEAD
68
61 - Test simplification.
69 + "git push --recurse-submodules $there HEAD:$target" was not
70 + propagated down to the submodules, but now it is.
71
63 - Will merge to 'next'.
72
73 +* jc/perl-git-comment-typofix (2017-08-07) 1 commit
74 + (merged to 'next' on 2017-08-18 at b2ad043e6a)
75 + + perl/Git.pm: typofix in a comment
76
66 -* sb/sha1-file-cleanup (2017-08-15) 1 commit
67 - - sha1_file: make read_info_alternates static
77 + A comment fix.
78 +
79 +
80 +* jk/drop-sha1-entry-pos (2017-08-09) 1 commit
81 + (merged to 'next' on 2017-08-18 at 3a4d9bcf12)
82 + + sha1_file: drop experimental GIT_USE_LOOKUP search
83 + (this branch is used by jt/packmigrate.)
84
85 Code clean-up.
86
71 - Will merge to 'next'.
87
88 +* jk/hashcmp-memcmp (2017-08-09) 1 commit
89 + (merged to 'next' on 2017-08-18 at 27c4aa5520)
90 + + hashcmp: use memcmp instead of open-coded loop
91
74 -* as/grep-quiet-no-match-exit-code-fix (2017-08-17) 1 commit
75 - - git-grep: correct exit code with --quiet and -L
92 + Code clean-up.
93
77 - "git grep -L" and "git grep --quiet -L" reported different exit
78 - codes; this has been corrected.
94
80 - Will merge to 'next'.
95 +* ma/parse-maybe-bool (2017-08-07) 6 commits
96 + (merged to 'next' on 2017-08-18 at ba22bb836c)
97 + + parse_decoration_style: drop unused argument `var`
98 + + treewide: deprecate git_config_maybe_bool, use git_parse_maybe_bool
99 + + config: make git_{config,parse}_maybe_bool equivalent
100 + + config: introduce git_parse_maybe_bool_text
101 + + t5334: document that git push --signed=1 does not work
102 + + Doc/git-{push,send-pack}: correct --sign= to --signed=
103
104 + Code clean-up.
105
83 -* hv/t5526-andand-chain-fix (2017-08-17) 1 commit
84 - - t5526: fix some broken && chains
106
86 - Test fix.
107 +* mf/no-dashed-subcommands (2017-08-07) 1 commit
108 + (merged to 'next' on 2017-08-18 at 05365af2ff)
109 + + scripts: use "git foo" not "git-foo"
110
88 - Will merge to 'next'.
111 + Code clean-up.
112
113
91 -* jc/diff-sane-truncate-no-more (2017-08-17) 1 commit
92 - - diff: retire sane_truncate_fn
114 +* mh/packed-ref-store (2017-08-17) 32 commits
115 + (merged to 'next' on 2017-08-18 at 14c58936e1)
116 + + files-backend: cheapen refname_available check when locking refs
117 + (merged to 'next' on 2017-08-14 at 987b76d302)
118 + + packed_ref_store: handle a packed-refs file that is a symlink
119 + + read_packed_refs(): die if `packed-refs` contains bogus data
120 + + t3210: add some tests of bogus packed-refs file contents
121 + + repack_without_refs(): don't lock or unlock the packed refs
122 + + commit_packed_refs(): remove call to `packed_refs_unlock()`
123 + + clear_packed_ref_cache(): don't protest if the lock is held
124 + + packed_refs_unlock(), packed_refs_is_locked(): new functions
125 + + packed_refs_lock(): report errors via a `struct strbuf *err`
126 + + packed_refs_lock(): function renamed from lock_packed_refs()
127 + + commit_packed_refs(): use a staging file separate from the lockfile
128 + + commit_packed_refs(): report errors rather than dying
129 + + packed_ref_store: make class into a subclass of `ref_store`
130 + + packed-backend: new module for handling packed references
131 + + packed_read_raw_ref(): new function, replacing `resolve_packed_ref()`
132 + + packed_ref_store: support iteration
133 + + packed_peel_ref(): new function, extracted from `files_peel_ref()`
134 + + repack_without_refs(): take a `packed_ref_store *` parameter
135 + + get_packed_ref(): take a `packed_ref_store *` parameter
136 + + rollback_packed_refs(): take a `packed_ref_store *` parameter
137 + + commit_packed_refs(): take a `packed_ref_store *` parameter
138 + + lock_packed_refs(): take a `packed_ref_store *` parameter
139 + + add_packed_ref(): take a `packed_ref_store *` parameter
140 + + get_packed_refs(): take a `packed_ref_store *` parameter
141 + + get_packed_ref_cache(): take a `packed_ref_store *` parameter
142 + + validate_packed_ref_cache(): take a `packed_ref_store *` parameter
143 + + clear_packed_ref_cache(): take a `packed_ref_store *` parameter
144 + + packed_ref_store: move `packed_refs_lock` member here
145 + + packed_ref_store: move `packed_refs_path` here
146 + + packed_ref_store: new struct
147 + + add_packed_ref(): teach function to overwrite existing refs
148 + + t1408: add a test of stale packed refs covered by loose refs
149 +
150 + The "ref-store" code reorganization continues.
151 +
152 +
153 +* pw/am-signoff (2017-08-08) 1 commit
154 + (merged to 'next' on 2017-08-18 at 7678ed822c)
155 + + am: fix signoff when other trailers are present
156 +
157 + "git am -s" has been taught that some input may end with a trailer
158 + block that is not Signed-off-by: and it should refrain from adding
159 + an extra blank line before adding a new sign-off in such a case.
160 +
161 +
162 +* pw/sequence-rerere-autoupdate (2017-08-02) 6 commits
163 + (merged to 'next' on 2017-08-14 at 010cbb0cbe)
164 + + cherry-pick/revert: reject --rerere-autoupdate when continuing
165 + + cherry-pick/revert: remember --rerere-autoupdate
166 + + t3504: use test_commit
167 + + rebase -i: honor --rerere-autoupdate
168 + + rebase: honor --rerere-autoupdate
169 + + am: remember --rerere-autoupdate setting
170 +
171 + Commands like "git rebase" accepted the --rerere-autoupdate option
172 + from the command line, but did not always use it. This has been
173 + fixed.
174 +
175 +
176 +* rj/add-chmod-error-message (2017-08-09) 1 commit
177 + (merged to 'next' on 2017-08-18 at ba2afb696a)
178 + + builtin/add: add detail to a 'cannot chmod' error message
179 +
180 + Message fix.
181 +
182 +
183 +* rs/apply-lose-prefix-length (2017-08-09) 1 commit
184 + (merged to 'next' on 2017-08-18 at 6ce0dbb3f0)
185 + + apply: remove prefix_length member from apply_state
186
187 Code clean-up.
188
96 - Will merge to 'next'.
189
190 +* rs/find-pack-entry-bisection (2017-08-09) 1 commit
191 + (merged to 'next' on 2017-08-18 at b4130177a9)
192 + + sha1_file: avoid comparison if no packed hash matches the first byte
193 + (this branch is used by jt/packmigrate.)
194
99 -* ks/branch-set-upstream (2017-08-17) 3 commits
100 - - branch: quote branch/ref names to improve readability
101 - - builtin/branch: stop supporting the "--set-upstream" option
102 - - t3200: cleanup cruft of a test
195 + Code clean-up.
196
104 - "branch --set-upstream" that has been deprecated in Git 1.8 has
105 - finally been retired.
197
107 - Will merge to 'next'.
198 +* rs/fsck-obj-leakfix (2017-08-10) 1 commit
199 + (merged to 'next' on 2017-08-18 at 77e3d111ef)
200 + + fsck: free buffers on error in fsck_obj()
201
202 + Memory leak in an error codepath has been plugged.
203
110 -* mg/format-ref-doc-fix (2017-08-18) 2 commits
111 - - Documentation/git-for-each-ref: clarify peeling of tags for --format
112 - - Documentation: use proper wording for ref format strings
204
114 - Doc fix.
205 +* rs/in-obsd-basename-dirname-take-const (2017-08-07) 1 commit
206 + (merged to 'next' on 2017-08-18 at c739077e3b)
207 + + test-path-utils: handle const parameter of basename and dirname
208 +
209 + Portability fix.
210 +
211 +
212 +* rs/merge-microcleanup (2017-08-10) 1 commit
213 + (merged to 'next' on 2017-08-18 at 79ba7d301d)
214 + + merge: use skip_prefix()
215 +
216 + Code clean-up.
217 +
218 +
219 +* rs/obsd-getcwd-workaround (2017-08-07) 1 commit
220 + (merged to 'next' on 2017-08-18 at adb98cfe0c)
221 + + t0001: skip test with restrictive permissions if getpwd(3) respects them
222 +
223 + Test portability fix for BSDs.
224 +
225 +
226 +* rs/strbuf-getwholeline-fix (2017-08-10) 1 commit
227 + (merged to 'next' on 2017-08-18 at 5f879d8261)
228 + + strbuf: clear errno before calling getdelim(3)
229 +
230 + A helper function to read a single whole line into strbuf
231 + mistakenly triggered OOM error at EOF under certain conditions,
232 + which has been fixed.
233 +
234 +
235 +* rs/t3700-clean-leftover (2017-08-08) 1 commit
236 + (merged to 'next' on 2017-08-18 at 12232a8cd3)
237 + + t3700: fix broken test under !POSIXPERM
238 +
239 + A test fix.
240 +
241 +
242 +* rs/t4062-obsd (2017-08-09) 1 commit
243 + (merged to 'next' on 2017-08-18 at 02bff42d86)
244 + + t4062: use less than 256 repetitions in regex
245 +
246 + Test portability fix.
247 +
248 +
249 +* rs/unpack-entry-leakfix (2017-08-10) 1 commit
250 + (merged to 'next' on 2017-08-18 at 1e7d8f54b1)
251 + + sha1_file: release delta_stack on error in unpack_entry()
252 + (this branch is used by jt/packmigrate.)
253 +
254 + Memory leak in an error codepath has been plugged.
255 +
256 +
257 +* rs/win32-syslog-leakfix (2017-08-10) 1 commit
258 + (merged to 'next' on 2017-08-18 at 011eccb7bd)
259 + + win32: plug memory leak on realloc() failure in syslog()
260 +
261 + Memory leak in an error codepath has been plugged.
262 +
263 +
264 +* sb/retire-t1200 (2017-08-10) 1 commit
265 + (merged to 'next' on 2017-08-18 at d436f9bbce)
266 + + t1200: remove t1200-tutorial.sh
267 +
268 + A test script that outlived its usefulness has been removed.
269 +
270 +
271 +* ur/svn-local-zone (2017-08-08) 1 commit
272 + (merged to 'next' on 2017-08-18 at f222bb6db0)
273 + + git svn fetch: Create correct commit timestamp when using --localtime
274 +
275 + "git svn" used with "--localtime" option did not compute the tz
276 + offset for the timestamp in question and instead always used the
277 + current time, which has been corrected.
278 +
279 +--------------------------------------------------
280 +[New Topics]
281 +
282 +* jt/packmigrate (2017-08-19) 27 commits
283 + - pack: move for_each_packed_object()
284 + - pack: move has_pack_index()
285 + - pack: move has_sha1_pack()
286 + - pack: move find_pack_entry() and make it global
287 + - pack: move find_sha1_pack()
288 + - pack: move find_pack_entry_one(), is_pack_valid()
289 + - pack: move check_pack_index_ptr(), nth_packed_object_offset()
290 + - pack: move nth_packed_object_{sha1,oid}
291 + - pack: move clear_delta_base_cache(), packed_object_info(), unpack_entry()
292 + - pack: move unpack_object_header()
293 + - pack: move get_size_from_delta()
294 + - pack: move unpack_object_header_buffer()
295 + - pack: move {,re}prepare_packed_git and approximate_object_count
296 + - pack: move install_packed_git()
297 + - pack: move add_packed_git()
298 + - pack: move unuse_pack()
299 + - pack: move use_pack()
300 + - pack: move pack-closing functions
301 + - pack: move release_pack_memory()
302 + - pack: move open_pack_index(), parse_pack_index()
303 + - pack: move pack_report()
304 + - pack: move static state variables
305 + - pack: move pack name-related functions
306 + - Merge branch 'rs/unpack-entry-leakfix' into HEAD
307 + - Merge branch 'jt/sha1-file-cleanup' into HEAD
308 + - Merge branch 'jk/drop-sha1-entry-pos' into HEAD
309 + - Merge branch 'rs/find-pack-entry-bisection' into HEAD
310 + (this branch uses jt/sha1-file-cleanup.)
311 +
312 + Code movement to make it easier to hack later.
313 +
314 + Rebase this on the master after jt/sha1-file-cleanup graduates and
315 + merge it to 'next' (if reviewers found this a good change, that
316 + is).
317 +
318 +
319 +* jc/cutoff-config (2017-08-19) 2 commits
320 + - rerere: allow approxidate in gc.rerereResolved/gc.rerereUnresolved
321 + - rerere: represent time duration in timestamp_t internally
322 +
323 + "[gc] rerereResolved = 5.days" used to be invalid, as the variable
324 + is defined to take an integer counting the number of days. It now
325 + is allowed.
326 +
327 + Needs reroll.
328 + cf. <xmqqa82t7ov1.fsf@gitster.mtv.corp.google.com>
329 +
330 +
331 +* jc/simplify-progress (2017-08-19) 1 commit
332 + - progress: simplify "delayed" progress API
333 +
334 + The API to start showing progress meter after a short delay has
335 + been simplified.
336
337 Will merge to 'next'.
338
339
119 -* po/read-graft-line (2017-08-18) 4 commits
120 - - commit: rewrite read_graft_line
121 - - commit: allocate array using object_id size
122 - - commit: replace the raw buffer with strbuf in read_graft_line
123 - - sha1_file: fix definition of null_sha1
340 +* jt/stash-tests (2017-08-19) 3 commits
341 + (merged to 'next' on 2017-08-22 at 0b022697aa)
342 + + stash: add a test for stashing in a detached state
343 + + stash: add a test for when apply fails during stash branch
344 + + stash: add a test for stash create with no files
345
125 - Conversion from uchar[20] to struct object_id continues; this is to
126 - ensure that we do not assume sizeof(struct object_id) is the same
127 - as the length of SHA-1 hash (or length of longest hash we support).
346 + Test update to improve coverage for "git stash" operations.
347 +
348 + Will merge to 'master'.
349 +
350 +
351 +* rs/archive-excluded-directory (2017-08-19) 3 commits
352 + - archive: don't queue excluded directories
353 + - archive: factor out helper functions for handling attributes
354 + - t5001: add tests for export-ignore attributes and exclude pathspecs
355 +
356 + "git archive" did not work well with pathspecs and the
357 + export-ignore attribute.
358
359 Will merge to 'next'.
360
361 + We may want to resurrect the "we don't archive an empty directory"
362 + bonus patch, but I do not mind merging the above early to 'next'
363 + and leave it as a separate follow-up enhancement.
364 + cf. <20170820090629.tumvqwzkromcykjf@sigill.intra.peff.net>
365
132 -* sb/submodule-parallel-update (2017-08-17) 1 commit
133 - - submodule.sh: remove unused variable
366 +
367 +* rs/commit-h-single-parent-cleanup (2017-08-19) 1 commit
368 + (merged to 'next' on 2017-08-22 at 7db01c4035)
369 + + commit: remove unused inline function single_parent()
370 +
371 + Code clean-up.
372 +
373 + Will merge to 'master'.
374 +
375 +
376 +* bc/hash-algo (2017-08-20) 5 commits
377 + - hash-algo: switch empty tree and blob lookups to use hash abstraction
378 + - hash-algo: integrate hash algorithm support with repo setup
379 + - hash-algo: add structure representing hash algorithm
380 + - setup: expose enumerated repo info
381 + - Merge branch 'bc/vcs-svn-cleanup' into bc/hash-algo
382 + (this branch uses bc/vcs-svn-cleanup.)
383 +
384 +
385 +* bc/vcs-svn-cleanup (2017-08-20) 2 commits
386 + - vcs-svn: rename repo functions to "svn_repo"
387 + - vcs-svn: remove unused prototypes
388 + (this branch is used by bc/hash-algo.)
389
390 Code clean-up.
391
392 Will merge to 'next'.
393
394
140 -* jc/apply-with-crlf (2017-08-16) 6 commits
141 - . apply: clarify read_old_data() is about no-index case
142 - . apply: localize the CRLF business to read_old_data()
143 - . apply: only pay attention to CRLF in the preimage
144 - . apply: remove unused field apply_state.flags
145 - . apply: file commited with CRLF should roundtrip diff and apply
146 - - convert: add SAFE_CRLF_KEEP_CRLF
147 - (this branch is tangled with tb/apply-with-crlf.)
395 +* jk/doc-the-this (2017-08-20) 1 commit
396 + - doc: fix typo in sendemail.identity
397 +
398 + Doc clean-up.
399 +
400 + Will merge to 'next'.
401 +
402 +
403 +* kw/write-index-reduce-alloc (2017-08-21) 3 commits
404 + - read-cache: avoid allocating every ondisk entry when writing
405 + - read-cache: fix memory leak in do_write_index
406 + - perf: add test for writing the index
407 +
408 +
409 +* mg/killed-merge (2017-08-21) 3 commits
410 + - merge: save merge state earlier
411 + - merge: split write_merge_state in two
412 + - Documentation/git-merge: explain --continue
413 +
414 + Killing "git merge --edit" before the editor returns control left
415 + the repository in a state with MERGE_MSG but without MERGE_HEAD,
416 + which incorrectly tells the subsequent "git commit" that there was
417 + a squash merge in progress. This has been fixed.
418 +
419 + Looked mostly good with minor nits.
420 + cf. <36bea69b-010a-9542-ec87-38b00d2bb009@grubix.eu>
421 +
422 +
423 +* po/object-id (2017-08-20) 6 commits
424 + - sha1_file: convert index_stream to struct object_id
425 + - sha1_file: convert hash_sha1_file_literally to struct object_id
426 + - sha1_file: convert index_fd to struct object_id
427 + - sha1_file: convert index_path to struct object_id
428 + - read-cache: convert to struct object_id
429 + - builtin/hash-object: convert to struct object_id
430 +
431 + Will merge to 'next'.
432
149 - Will discard as it now is part of the tb/apply-with-crlf topic.
433
434 --------------------------------------------------
435 [Stalled]
@@ -226,292 +509,220 @@ of the repositories listed at
509 --------------------------------------------------
510 [Cooking]
511
229 -* ks/prepare-commit-msg-sample-fix (2017-08-14) 1 commit
230 - - hook: use correct logical variable
231 -
232 - An "oops" fix to a topic that is already in 'master'.
233 -
234 - Will merge to 'next'.
235 -
236 -
237 -* ab/ref-filter-no-contains (2017-08-07) 1 commit
238 - (merged to 'next' on 2017-08-18 at 7ec9d3d3a2)
239 - + tests: don't give unportable ">" to "test" built-in, use -gt
240 -
241 - A test fix.
242 -
243 - Will merge to 'master'.
244 -
245 -
246 -* ma/parse-maybe-bool (2017-08-07) 6 commits
247 - (merged to 'next' on 2017-08-18 at ba22bb836c)
248 - + parse_decoration_style: drop unused argument `var`
249 - + treewide: deprecate git_config_maybe_bool, use git_parse_maybe_bool
250 - + config: make git_{config,parse}_maybe_bool equivalent
251 - + config: introduce git_parse_maybe_bool_text
252 - + t5334: document that git push --signed=1 does not work
253 - + Doc/git-{push,send-pack}: correct --sign= to --signed=
254 -
255 - Code clean-up.
256 -
257 - Will merge to 'master'.
258 -
259 -
260 -* mf/no-dashed-subcommands (2017-08-07) 1 commit
261 - (merged to 'next' on 2017-08-18 at 05365af2ff)
262 - + scripts: use "git foo" not "git-foo"
263 -
264 - Code clean-up.
265 -
266 - Will merge to 'master'.
267 -
268 -
269 -* mk/diff-delta-uint-may-be-shorter-than-ulong (2017-08-10) 1 commit
270 - - diff-delta: fix encoding size that would not fit in "unsigned int"
271 -
272 - The machinery to create xdelta used in pack files received the
273 - sizes of the data in size_t, but lost the higher bits of them by
274 - storing them in "unsigned int" during the computation, which is
275 - fixed.
276 -
277 - Will merge to 'next'.
278 -
279 -
280 -* rs/obsd-getcwd-workaround (2017-08-07) 1 commit
281 - (merged to 'next' on 2017-08-18 at adb98cfe0c)
282 - + t0001: skip test with restrictive permissions if getpwd(3) respects them
283 -
284 - Test portability fix for BSDs.
285 -
286 - Will merge to 'master'.
287 -
288 -
289 -* rs/in-obsd-basename-dirname-take-const (2017-08-07) 1 commit
290 - (merged to 'next' on 2017-08-18 at c739077e3b)
291 - + test-path-utils: handle const parameter of basename and dirname
292 -
293 - Portability fix.
294 -
295 - Will merge to 'master'.
296 -
297 -
298 -* rs/t3700-clean-leftover (2017-08-08) 1 commit
299 - (merged to 'next' on 2017-08-18 at 12232a8cd3)
300 - + t3700: fix broken test under !POSIXPERM
301 -
302 - A test fix.
303 -
304 - Will merge to 'master'.
305 -
306 -
307 -* jc/perl-git-comment-typofix (2017-08-07) 1 commit
308 - (merged to 'next' on 2017-08-18 at b2ad043e6a)
309 - + perl/Git.pm: typofix in a comment
512 +* cc/subprocess-handshake-missing-capabilities (2017-08-16) 1 commit
513 + (merged to 'next' on 2017-08-19 at c512710fda)
514 + + sub-process: print the cmd when a capability is unsupported
515
311 - A comment fix.
516 + When handshake with a subprocess filter notices that the process
517 + asked for an unknown capability, Git did not report what program
518 + the offending subprocess was running. This has been corrected.
519
520 Will merge to 'master'.
521
315 -
316 -* jk/drop-ancient-curl (2017-08-09) 5 commits
317 - - http: #error on too-old curl
318 - - curl: remove ifdef'd code never used with curl >=7.19.4
319 - - http: drop support for curl < 7.19.4
320 - - http: drop support for curl < 7.16.0
321 - - http: drop support for curl < 7.11.1
322 -
323 - Some code in http.c that has bitrot is being removed.
324 -
325 - What is the status of the discussion around this area????
326 -
327 -
328 -* tc/curl-with-backports (2017-08-11) 2 commits
329 - - http: use a feature check to enable GSSAPI delegation control
330 - - http: fix handling of missing CURLPROTO_*
331 -
332 - Updates to the HTTP layer we made recently unconditionally used
333 - features of libCurl without checking the existence of them, causing
334 - compilation errors, which has been fixed. Also migrate the code to
335 - check feature macros, not version numbers, to cope better with
336 - libCurl that vendor ships with backported features.
337 -
338 - What is the doneness of this topic????
339 -
340 -
341 -* jk/drop-sha1-entry-pos (2017-08-09) 1 commit
342 - (merged to 'next' on 2017-08-18 at 3a4d9bcf12)
343 - + sha1_file: drop experimental GIT_USE_LOOKUP search
344 -
345 - Code clean-up.
346 -
347 - Will merge to 'master'.
522 + We may want a follow-up fix to tighten the error checking, though.
523
524
350 -* jk/hashcmp-memcmp (2017-08-09) 1 commit
351 - (merged to 'next' on 2017-08-18 at 27c4aa5520)
352 - + hashcmp: use memcmp instead of open-coded loop
525 +* tb/apply-with-crlf (2017-08-19) 2 commits
526 + (merged to 'next' on 2017-08-22 at f92bf67061)
527 + + apply: file commited with CRLF should roundtrip diff and apply
528 + + convert: add SAFE_CRLF_KEEP_CRLF
529 + (this branch is tangled with jc/apply-with-crlf.)
530
354 - Code clean-up.
531 + "git apply" that is used as a better "patch -p1" failed to apply a
532 + taken from a file with CRLF line endings to a file with CRLF line
533 + endings. The root cause was because it misused convert_to_git()
534 + that tried to do "safe-crlf" processing by looking at the index
535 + entry at the same path, which is a nonsense---in that mode, "apply"
536 + is not working on the data in (or derived from) the index at all.
537 + This has been fixed.
538
539 Will merge to 'master'.
540
541
359 -* jk/trailers-parse (2017-08-15) 8 commits
360 - - pretty: support normalization options for %(trailers)
361 - - t4205: refactor %(trailers) tests
362 - - pretty: move trailer formatting to trailer.c
363 - - interpret-trailers: add --parse convenience option
364 - - interpret-trailers: add an option to unfold values
365 - - interpret-trailers: add an option to show only existing trailers
366 - - interpret-trailers: add an option to show only the trailers
367 - - trailer: put process_trailers() options into a struct
368 -
369 - "git interpret-trailers" has been taught a "--parse" and a few
370 - other options to make it easier for scripts to grab existing
371 - trailer lines from a commit log message.
542 +* rs/t1002-do-not-use-sum (2017-08-15) 1 commit
543 + (merged to 'next' on 2017-08-19 at 15b2a5f21b)
544 + + t1002: stop using sum(1)
545
373 - Will merge to 'next'.
546 + Test simplification.
547
548 + Will merge to 'master'.
549
376 -* mk/use-size-t-in-zlib (2017-08-10) 1 commit
377 - - zlib.c: use size_t for size
550
379 - The wrapper to call into zlib followed our long tradition to use
380 - "unsigned long" for sizes of regions in memory, which have been
381 - updated to use "size_t".
551 +* sb/sha1-file-cleanup (2017-08-15) 1 commit
552 + (merged to 'next' on 2017-08-19 at 85898d1e7c)
553 + + sha1_file: make read_info_alternates static
554
383 - Needs review.
555 + Code clean-up.
556
385 - There were some concerns raised around s/off_t/size_t/ in this
386 - change.
557 + Will merge to 'master'.
558
559
389 -* pw/am-signoff (2017-08-08) 1 commit
390 - (merged to 'next' on 2017-08-18 at 7678ed822c)
391 - + am: fix signoff when other trailers are present
560 +* as/grep-quiet-no-match-exit-code-fix (2017-08-17) 1 commit
561 + (merged to 'next' on 2017-08-19 at 362f88fb41)
562 + + git-grep: correct exit code with --quiet and -L
563
393 - "git am -s" has been taught that some input may end with a trailer
394 - block that is not Signed-off-by: and it should refrain from adding
395 - an extra blank line before adding a new sign-off in such a case.
564 + "git grep -L" and "git grep --quiet -L" reported different exit
565 + codes; this has been corrected.
566
567 Will merge to 'master'.
568
569
400 -* rj/add-chmod-error-message (2017-08-09) 1 commit
401 - (merged to 'next' on 2017-08-18 at ba2afb696a)
402 - + builtin/add: add detail to a 'cannot chmod' error message
570 +* hv/t5526-andand-chain-fix (2017-08-17) 1 commit
571 + (merged to 'next' on 2017-08-19 at fa95053653)
572 + + t5526: fix some broken && chains
573
404 - Message fix.
574 + Test fix.
575
576 Will merge to 'master'.
577
578
409 -* rs/apply-lose-prefix-length (2017-08-09) 1 commit
410 - (merged to 'next' on 2017-08-18 at 6ce0dbb3f0)
411 - + apply: remove prefix_length member from apply_state
579 +* jc/diff-sane-truncate-no-more (2017-08-17) 1 commit
580 + (merged to 'next' on 2017-08-19 at 3ba3980eed)
581 + + diff: retire sane_truncate_fn
582
583 Code clean-up.
584
585 Will merge to 'master'.
586
587
418 -* rs/find-pack-entry-bisection (2017-08-09) 1 commit
419 - (merged to 'next' on 2017-08-18 at b4130177a9)
420 - + sha1_file: avoid comparison if no packed hash matches the first byte
588 +* ks/branch-set-upstream (2017-08-17) 3 commits
589 + - branch: quote branch/ref names to improve readability
590 + - builtin/branch: stop supporting the "--set-upstream" option
591 + - t3200: cleanup cruft of a test
592 +
593 + "branch --set-upstream" that has been deprecated in Git 1.8 has
594 + finally been retired.
595 +
596 + Will merge to 'next'.
597
422 - Code clean-up.
598 +
599 +* mg/format-ref-doc-fix (2017-08-18) 2 commits
600 + (merged to 'next' on 2017-08-19 at 6490525c54)
601 + + Documentation/git-for-each-ref: clarify peeling of tags for --format
602 + + Documentation: use proper wording for ref format strings
603 +
604 + Doc fix.
605
606 Will merge to 'master'.
607
608
427 -* rs/fsck-obj-leakfix (2017-08-10) 1 commit
428 - (merged to 'next' on 2017-08-18 at 77e3d111ef)
429 - + fsck: free buffers on error in fsck_obj()
609 +* po/read-graft-line (2017-08-18) 4 commits
610 + (merged to 'next' on 2017-08-22 at 1e3fe0d3a1)
611 + + commit: rewrite read_graft_line
612 + + commit: allocate array using object_id size
613 + + commit: replace the raw buffer with strbuf in read_graft_line
614 + + sha1_file: fix definition of null_sha1
615
431 - Memory leak in an error codepath has been plugged.
616 + Conversion from uchar[20] to struct object_id continues; this is to
617 + ensure that we do not assume sizeof(struct object_id) is the same
618 + as the length of SHA-1 hash (or length of longest hash we support).
619
620 Will merge to 'master'.
621
622
436 -* rs/merge-microcleanup (2017-08-10) 1 commit
437 - (merged to 'next' on 2017-08-18 at 79ba7d301d)
438 - + merge: use skip_prefix()
623 +* sb/submodule-parallel-update (2017-08-17) 1 commit
624 + (merged to 'next' on 2017-08-19 at c023a834f0)
625 + + submodule.sh: remove unused variable
626
627 Code clean-up.
628
629 Will merge to 'master'.
630
631
445 -* rs/strbuf-getwholeline-fix (2017-08-10) 1 commit
446 - (merged to 'next' on 2017-08-18 at 5f879d8261)
447 - + strbuf: clear errno before calling getdelim(3)
632 +* ks/prepare-commit-msg-sample-fix (2017-08-14) 1 commit
633 + (merged to 'next' on 2017-08-19 at e75c30de64)
634 + + hook: use correct logical variable
635
449 - A helper function to read a single whole line into strbuf
450 - mistakenly triggered OOM error at EOF under certain conditions,
451 - which has been fixed.
636 + An "oops" fix to a topic that is already in 'master'.
637
638 Will merge to 'master'.
639
640
456 -* rs/t4062-obsd (2017-08-09) 1 commit
457 - (merged to 'next' on 2017-08-18 at 02bff42d86)
458 - + t4062: use less than 256 repetitions in regex
641 +* mk/diff-delta-uint-may-be-shorter-than-ulong (2017-08-10) 1 commit
642 + . diff-delta: fix encoding size that would not fit in "unsigned int"
643
460 - Test portability fix.
644 + The machinery to create xdelta used in pack files received the
645 + sizes of the data in size_t, but lost the higher bits of them by
646 + storing them in "unsigned int" during the computation, which is
647 + fixed.
648
462 - Will merge to 'master'.
649 + Dropped, as it was rerolled for review as part of a larger series.
650 + cf. <1502914591-26215-1-git-send-email-martin@mail.zuhause>
651
652
465 -* rs/unpack-entry-leakfix (2017-08-10) 1 commit
466 - (merged to 'next' on 2017-08-18 at 1e7d8f54b1)
467 - + sha1_file: release delta_stack on error in unpack_entry()
653 +* jk/drop-ancient-curl (2017-08-09) 5 commits
654 + - http: #error on too-old curl
655 + - curl: remove ifdef'd code never used with curl >=7.19.4
656 + - http: drop support for curl < 7.19.4
657 + - http: drop support for curl < 7.16.0
658 + - http: drop support for curl < 7.11.1
659
469 - Memory leak in an error codepath has been plugged.
660 + Some code in http.c that has bitrot is being removed.
661
471 - Will merge to 'master'.
662 + Will be rerolled after tc/curl-with-backports stabilizes.
663
664
474 -* rs/win32-syslog-leakfix (2017-08-10) 1 commit
475 - (merged to 'next' on 2017-08-18 at 011eccb7bd)
476 - + win32: plug memory leak on realloc() failure in syslog()
665 +* tc/curl-with-backports (2017-08-11) 2 commits
666 + - http: use a feature check to enable GSSAPI delegation control
667 + - http: fix handling of missing CURLPROTO_*
668
478 - Memory leak in an error codepath has been plugged.
669 + Updates to the HTTP layer we made recently unconditionally used
670 + features of libCurl without checking the existence of them, causing
671 + compilation errors, which has been fixed. Also migrate the code to
672 + check feature macros, not version numbers, to cope better with
673 + libCurl that vendor ships with backported features.
674
480 - Will merge to 'master'.
675 + Will merge to 'next'.
676
677
483 -* sb/retire-t1200 (2017-08-10) 1 commit
484 - (merged to 'next' on 2017-08-18 at d436f9bbce)
485 - + t1200: remove t1200-tutorial.sh
678 +* jk/trailers-parse (2017-08-20) 9 commits
679 + (merged to 'next' on 2017-08-22 at 2d4d937141)
680 + + doc/interpret-trailers: fix "the this" typo
681 + (merged to 'next' on 2017-08-19 at faa63f8196)
682 + + pretty: support normalization options for %(trailers)
683 + + t4205: refactor %(trailers) tests
684 + + pretty: move trailer formatting to trailer.c
685 + + interpret-trailers: add --parse convenience option
686 + + interpret-trailers: add an option to unfold values
687 + + interpret-trailers: add an option to show only existing trailers
688 + + interpret-trailers: add an option to show only the trailers
689 + + trailer: put process_trailers() options into a struct
690
487 - A test script that outlived its usefulness has been removed.
691 + "git interpret-trailers" has been taught a "--parse" and a few
692 + other options to make it easier for scripts to grab existing
693 + trailer lines from a commit log message.
694
695 Will merge to 'master'.
696 + Possible leftover bits are mentioned in
697 + cf. <20170820094009.z23wclpku35txflg@sigill.intra.peff.net>
698
699
492 -* ur/svn-local-zone (2017-08-08) 1 commit
493 - (merged to 'next' on 2017-08-18 at f222bb6db0)
494 - + git svn fetch: Create correct commit timestamp when using --localtime
700 +* mk/use-size-t-in-zlib (2017-08-10) 1 commit
701 + . zlib.c: use size_t for size
702
496 - "git svn" used with "--localtime" option did not compute the tz
497 - offset for the timestamp in question and instead always used the
498 - current time, which has been corrected.
703 + The wrapper to call into zlib followed our long tradition to use
704 + "unsigned long" for sizes of regions in memory, which have been
705 + updated to use "size_t".
706
500 - Will merge to 'master'.
707 + Dropped, as it was rerolled for review as part of a larger series.
708 + cf. <1502914591-26215-1-git-send-email-martin@mail.zuhause>
709
710
711 * jt/sha1-file-cleanup (2017-08-11) 2 commits
504 - - sha1_file: remove read_packed_sha1()
505 - - sha1_file: set whence in storage-specific info fn
712 + (merged to 'next' on 2017-08-19 at 93f4c94578)
713 + + sha1_file: remove read_packed_sha1()
714 + + sha1_file: set whence in storage-specific info fn
715 + (this branch is used by jt/packmigrate.)
716
717 Preparatory code clean-up.
718
509 - Will merge to 'next'.
719 + Will merge to 'master'.
720
721
722 * kw/rebase-progress (2017-08-14) 2 commits
513 - - rebase: turn on progress option by default for format-patch
514 - - format-patch: have progress option while generating patches
723 + (merged to 'next' on 2017-08-19 at 1958f378dd)
724 + + rebase: turn on progress option by default for format-patch
725 + + format-patch: have progress option while generating patches
726
727 "git rebase", especially when it is run by mistake and ends up
728 trying to replay many changes, spent long time in silence. The
@@ -519,75 +730,82 @@ of the repositories listed at
730 long time preparing these many changes to replay (which would give
731 the user a chance to abort with ^C).
732
522 - Will merge to 'next'.
733 + Will merge to 'master'.
734
735
736 * mk/diff-delta-avoid-large-offset (2017-08-11) 1 commit
526 - - diff-delta: do not allow delta offset truncation
737 + . diff-delta: do not allow delta offset truncation
738
739 The delta format used in the packfile cannot reference data at
740 offset larger than what can be expressed in 4-byte, but the
741 generator for the data failed to make sure the offset does not
742 overflow. This has been corrected.
743
533 - Will merge to 'next'.
744 + Dropped, as it was rerolled for review as part of a larger series.
745 + cf. <1502914591-26215-1-git-send-email-martin@mail.zuhause>
746
747
748 * nm/stash-untracked (2017-08-11) 1 commit
537 - - stash: clean untracked files before reset
749 + (merged to 'next' on 2017-08-19 at 70990d7eb3)
750 + + stash: clean untracked files before reset
751
752 "git stash -u" used the contents of the committed version of the
753 ".gitignore" file to decide which paths are ignored, even when the
754 file has local changes. The command has been taught to instead use
755 the locally modified contents.
756
544 - Will merge to 'next'.
757 + Will merge to 'master'.
758
759
760 * rs/object-id (2017-08-14) 1 commit
548 - - tree-walk: convert fill_tree_descriptor() to object_id
761 + (merged to 'next' on 2017-08-19 at 0f7283cd19)
762 + + tree-walk: convert fill_tree_descriptor() to object_id
763
764 Conversion from uchar[20] to struct object_id continues.
765
552 - Will merge to 'next'.
766 + Will merge to 'master'.
767
768
769 * ah/doc-empty-string-is-false (2017-08-14) 1 commit
556 - - doc: clarify "config --bool" behaviour with empty string
770 + (merged to 'next' on 2017-08-19 at df47ffeffa)
771 + + doc: clarify "config --bool" behaviour with empty string
772
773 Doc update.
774
560 - Will merge to 'next'.
775 + Will merge to 'master'.
776
777
778 * jt/diff-color-move-fix (2017-08-16) 3 commits
564 - - diff: define block by number of alphanumeric chars
565 - - diff: respect MIN_BLOCK_LENGTH for last block
566 - - diff: avoid redundantly clearing a flag
779 + (merged to 'next' on 2017-08-19 at 6ae3831c8c)
780 + + diff: define block by number of alphanumeric chars
781 + + diff: respect MIN_BLOCK_LENGTH for last block
782 + + diff: avoid redundantly clearing a flag
783 (this branch uses sb/diff-color-move.)
784
785 A handful of bugfixes and an improvement to "diff --color-moved".
786
571 - Will merge to 'next'.
787 + Will merge to 'master'.
788
789
790 * kd/stash-with-bash-4.4 (2017-08-14) 1 commit
575 - - stash: prevent warning about null bytes in input
791 + (merged to 'next' on 2017-08-19 at 79b2c4b052)
792 + + stash: prevent warning about null bytes in input
793
794 bash 4.4 or newer gave a warning on NUL byte in command
795 substitution done in "git stash"; this has been squelched.
796
580 - Will merge to 'next'.
797 + Will merge to 'master'.
798
799
800 * kw/commit-keep-index-when-pre-commit-is-not-run (2017-08-16) 1 commit
584 - - commit: skip discarding the index if there is no pre-commit hook
801 + (merged to 'next' on 2017-08-19 at 2b5a25e5ae)
802 + + commit: skip discarding the index if there is no pre-commit hook
803
804 "git commit" used to discard the index and re-read from the filesystem
805 just in case the pre-commit hook has updated it in the middle; this
806 has been optimized out when we know we do not run the pre-commit hook.
807
590 - Will merge to 'next'.
808 + Will merge to 'master'.
809
810
811 * bw/git-clang-format (2017-08-14) 2 commits
@@ -598,51 +816,27 @@ of the repositories listed at
816 out coding style issues.
817
818
601 -* pw/sequence-rerere-autoupdate (2017-08-02) 6 commits
602 - (merged to 'next' on 2017-08-14 at 010cbb0cbe)
603 - + cherry-pick/revert: reject --rerere-autoupdate when continuing
604 - + cherry-pick/revert: remember --rerere-autoupdate
605 - + t3504: use test_commit
606 - + rebase -i: honor --rerere-autoupdate
607 - + rebase: honor --rerere-autoupdate
608 - + am: remember --rerere-autoupdate setting
609 -
610 - Commands like "git rebase" accepted the --rerere-autoupdate option
611 - from the command line, but did not always use it. This has been
612 - fixed.
613 -
614 - Will merge to 'master'.
615 -
616 -
617 -* bw/clone-recursive-quiet (2017-08-04) 1 commit
618 - (merged to 'next' on 2017-08-14 at fbd4473ce4)
619 - + clone: teach recursive clones to respect -q
620 -
621 - "git clone --recurse-submodules --quiet" did not pass the quiet
622 - option down to submodules.
623 -
624 - Will merge to 'master'.
625 -
626 -
819 * lg/merge-signoff (2017-07-25) 1 commit
628 - - merge: add a --signoff flag
820 + (merged to 'next' on 2017-08-19 at cb53d7b026)
821 + + merge: add a --signoff flag
822
823 "git merge" learned a "--signoff" option to add the Signed-off-by:
824 trailer with the committer's name.
825
633 - Will merge to 'next'.
826 + Will merge to 'master'.
827
828
829 * pb/trailers-from-command-line (2017-08-14) 4 commits
637 - - interpret-trailers: fix documentation typo
638 - - interpret-trailers: add options for actions
639 - - trailers: introduce struct new_trailer_item
640 - - trailers: export action enums and corresponding lookup functions
830 + (merged to 'next' on 2017-08-19 at b12342e49e)
831 + + interpret-trailers: fix documentation typo
832 + + interpret-trailers: add options for actions
833 + + trailers: introduce struct new_trailer_item
834 + + trailers: export action enums and corresponding lookup functions
835
836 "git interpret-trailers" learned to take the trailer specifications
837 from the command line that overrides the configured values.
838
645 - Will merge to 'next'.
839 + Will merge to 'master'.
840
841
842 * sb/submodule-recursive-checkout-detach-head (2017-07-28) 2 commits
@@ -689,20 +883,14 @@ of the repositories listed at
883 - t7411: check configuration parsing errors
884 - Merge branch 'bc/object-id' into bw/submodule-config-cleanup
885 - Merge branch 'bw/grep-recurse-submodules' into bw/submodule-config-cleanup
692 - (this branch uses bw/grep-recurse-submodules.)
886
887 Code clean-up to avoid mixing values read from the .gitmodules file
888 and values read from the .git/config file.
889
697 -
698 -* bw/push-options-recursively-to-submodules (2017-07-20) 1 commit
699 - (merged to 'next' on 2017-08-14 at 421dc09fd0)
700 - + submodule--helper: teach push-check to handle HEAD
701 -
702 - "git push --recurse-submodules $there HEAD:$target" was not
703 - propagated down to the submodules, but now it is.
704 -
705 - Will merge to 'master'.
890 + So, after the recent discussion around submodule.<name>.update (and
891 + its resolution "use submodule.<name>.active; "reset --hard" must
892 + ignore the .update thing"), this is now good to go, I presume?
893 + Please yell at me that I am clueless if that is not the case ;-)
894
895
896 * jk/check-ref-format-oor-fix (2017-07-14) 1 commit
@@ -713,27 +901,6 @@ of the repositories listed at
901 cf. <20170817102217.3yw7uxnkupdy3lh5@sigill.intra.peff.net>
902
903
716 -* bw/grep-recurse-submodules (2017-08-02) 10 commits
717 - (merged to 'next' on 2017-08-14 at dcfcfc94af)
718 - + grep: recurse in-process using 'struct repository'
719 - + submodule: merge repo_read_gitmodules and gitmodules_config
720 - + submodule: check for unmerged .gitmodules outside of config parsing
721 - + submodule: check for unstaged .gitmodules outside of config parsing
722 - + submodule: remove fetch.recursesubmodules from submodule-config parsing
723 - + submodule: remove submodule.fetchjobs from submodule-config parsing
724 - + config: add config_from_gitmodules
725 - + cache.h: add GITMODULES_FILE macro
726 - + repository: have the_repository use the_index
727 - + repo_read_index: don't discard the index
728 - (this branch is used by bw/submodule-config-cleanup.)
729 -
730 - "git grep --recurse-submodules" has been reworked to give a more
731 - consistent output across submodule boundary (and do its thing
732 - without having to fork a separate process).
733 -
734 - Will merge to 'master'.
735 -
736 -
904 * rs/sha1-file-micro-optim (2017-07-09) 2 commits
905 - SQUASH???
906 - sha1_file: add slash once in for_each_file_in_obj_subdir()
@@ -775,47 +942,8 @@ of the repositories listed at
942 sound too bad.
943
944 Will cook in 'next'.
778 -
779 -
780 -* mh/packed-ref-store (2017-08-17) 32 commits
781 - (merged to 'next' on 2017-08-18 at 14c58936e1)
782 - + files-backend: cheapen refname_available check when locking refs
783 - (merged to 'next' on 2017-08-14 at 987b76d302)
784 - + packed_ref_store: handle a packed-refs file that is a symlink
785 - + read_packed_refs(): die if `packed-refs` contains bogus data
786 - + t3210: add some tests of bogus packed-refs file contents
787 - + repack_without_refs(): don't lock or unlock the packed refs
788 - + commit_packed_refs(): remove call to `packed_refs_unlock()`
789 - + clear_packed_ref_cache(): don't protest if the lock is held
790 - + packed_refs_unlock(), packed_refs_is_locked(): new functions
791 - + packed_refs_lock(): report errors via a `struct strbuf *err`
792 - + packed_refs_lock(): function renamed from lock_packed_refs()
793 - + commit_packed_refs(): use a staging file separate from the lockfile
794 - + commit_packed_refs(): report errors rather than dying
795 - + packed_ref_store: make class into a subclass of `ref_store`
796 - + packed-backend: new module for handling packed references
797 - + packed_read_raw_ref(): new function, replacing `resolve_packed_ref()`
798 - + packed_ref_store: support iteration
799 - + packed_peel_ref(): new function, extracted from `files_peel_ref()`
800 - + repack_without_refs(): take a `packed_ref_store *` parameter
801 - + get_packed_ref(): take a `packed_ref_store *` parameter
802 - + rollback_packed_refs(): take a `packed_ref_store *` parameter
803 - + commit_packed_refs(): take a `packed_ref_store *` parameter
804 - + lock_packed_refs(): take a `packed_ref_store *` parameter
805 - + add_packed_ref(): take a `packed_ref_store *` parameter
806 - + get_packed_refs(): take a `packed_ref_store *` parameter
807 - + get_packed_ref_cache(): take a `packed_ref_store *` parameter
808 - + validate_packed_ref_cache(): take a `packed_ref_store *` parameter
809 - + clear_packed_ref_cache(): take a `packed_ref_store *` parameter
810 - + packed_ref_store: move `packed_refs_lock` member here
811 - + packed_ref_store: move `packed_refs_path` here
812 - + packed_ref_store: new struct
813 - + add_packed_ref(): teach function to overwrite existing refs
814 - + t1408: add a test of stale packed refs covered by loose refs
815 -
816 - The "ref-store" code reorganization continues.
817 -
818 - Will merge to 'master'.
945 + As the draft RelNotes says, let's merge this to 'master' after
946 + this release cycle.
947
948
949 * sd/branch-copy (2017-06-18) 3 commits
@@ -833,40 +961,40 @@ of the repositories listed at
961
962
963 * sb/diff-color-move (2017-06-30) 26 commits
836 - - diff: document the new --color-moved setting
837 - - diff.c: add dimming to moved line detection
838 - - diff.c: color moved lines differently, plain mode
839 - - diff.c: color moved lines differently
840 - - diff.c: buffer all output if asked to
841 - - diff.c: emit_diff_symbol learns about DIFF_SYMBOL_SUMMARY
842 - - diff.c: emit_diff_symbol learns about DIFF_SYMBOL_STAT_SEP
843 - - diff.c: convert word diffing to use emit_diff_symbol
844 - - diff.c: convert show_stats to use emit_diff_symbol
845 - - diff.c: convert emit_binary_diff_body to use emit_diff_symbol
846 - - submodule.c: migrate diff output to use emit_diff_symbol
847 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_REWRITE_DIFF
848 - - diff.c: emit_diff_symbol learns about DIFF_SYMBOL_BINARY_FILES
849 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_HEADER
850 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_FILEPAIR_{PLUS, MINUS}
851 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_INCOMPLETE
852 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_WORDS[_PORCELAIN]
853 - - diff.c: migrate emit_line_checked to use emit_diff_symbol
854 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_NO_LF_EOF
855 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_FRAGINFO
856 - - diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_MARKER
857 - - diff.c: introduce emit_diff_symbol
858 - - diff.c: factor out diff_flush_patch_all_file_pairs
859 - - diff.c: move line ending check into emit_hunk_header
860 - - diff.c: readability fix
861 - - Merge branch 'sb/hashmap-customize-comparison' into sb/diff-color-move
964 + (merged to 'next' on 2017-08-19 at a239d031bc)
965 + + diff: document the new --color-moved setting
966 + + diff.c: add dimming to moved line detection
967 + + diff.c: color moved lines differently, plain mode
968 + + diff.c: color moved lines differently
969 + + diff.c: buffer all output if asked to
970 + + diff.c: emit_diff_symbol learns about DIFF_SYMBOL_SUMMARY
971 + + diff.c: emit_diff_symbol learns about DIFF_SYMBOL_STAT_SEP
972 + + diff.c: convert word diffing to use emit_diff_symbol
973 + + diff.c: convert show_stats to use emit_diff_symbol
974 + + diff.c: convert emit_binary_diff_body to use emit_diff_symbol
975 + + submodule.c: migrate diff output to use emit_diff_symbol
976 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_REWRITE_DIFF
977 + + diff.c: emit_diff_symbol learns about DIFF_SYMBOL_BINARY_FILES
978 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_HEADER
979 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_FILEPAIR_{PLUS, MINUS}
980 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_INCOMPLETE
981 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_WORDS[_PORCELAIN]
982 + + diff.c: migrate emit_line_checked to use emit_diff_symbol
983 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_NO_LF_EOF
984 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_FRAGINFO
985 + + diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_MARKER
986 + + diff.c: introduce emit_diff_symbol
987 + + diff.c: factor out diff_flush_patch_all_file_pairs
988 + + diff.c: move line ending check into emit_hunk_header
989 + + diff.c: readability fix
990 + + Merge branch 'sb/hashmap-customize-comparison' into sb/diff-color-move
991 (this branch is used by jt/diff-color-move-fix.)
992
993 "git diff" has been taught to optionally paint new lines that are
994 the same as deleted lines elsewhere differently from genuinely new
995 lines.
996
868 - Will merge to 'next'.
869 - ... again.
997 + Will merge to 'master'.
998
999 --------------------------------------------------
1000 [Discarded]
@@ -890,3 +1018,15 @@ of the repositories listed at
1018 of other worktrees as the root for reachability traversal, making
1019 objects that are in use only in other worktrees to be subject to
1020 garbage collection.
1021 +
1022 +
1023 +* jc/apply-with-crlf (2017-08-16) 6 commits
1024 + . apply: clarify read_old_data() is about no-index case
1025 + . apply: localize the CRLF business to read_old_data()
1026 + . apply: only pay attention to CRLF in the preimage
1027 + . apply: remove unused field apply_state.flags
1028 + . apply: file commited with CRLF should roundtrip diff and apply
1029 + + convert: add SAFE_CRLF_KEEP_CRLF
1030 + (this branch is tangled with tb/apply-with-crlf.)
1031 +
1032 + This served as an input to updated tb/apply-with-crlf.