What's cooking (2014/04 #04)

Junio C Hamano committed Apr 15, 2014 at 15:06 UTC 5827fd485c816ff78c3c5c46752cf18acf49d2e4
1 file changed +153 -160
whats-cooking.txt
+153 -160
@@ -1,10 +1,10 @@
1 To: git@vger.kernel.org
2 Bcc: lwn@lwn.net
3 -Subject: What's cooking in git.git (Apr 2014, #03; Fri, 11)
3 +Subject: What's cooking in git.git (Apr 2014, #04; Tue, 15)
4 X-master-at: 68773ac915580e88016c34ce1269730382b501d4
5 -X-next-at: 5ae4ba539641e54f43a36c773306acca96600f24
5 +X-next-at: eaba9156adf2b84a54a025dcd2c3fecdea5dc8a0
6
7 -What's cooking in git.git (Apr 2014, #03; Fri, 11)
7 +What's cooking in git.git (Apr 2014, #04; Tue, 15)
8 --------------------------------------------------
9
10 Here are the topics that have been cooking. Commits prefixed with
@@ -12,7 +12,7 @@ Here are the topics that have been cooking. Commits prefixed with
12 '+' are in 'next'.
13
14 The number of topics cooking in 'next' has been shrinking, and the
15 -cycle is getting long. Hopefully we will have -rc0 late next week to
15 +cycle is getting long. Hopefully we will have -rc0 late this week to
16 close the 'master' branch and a few rounds of -rc iterations until
17 2.0 final. Other topics under discussion may continue to flow to
18 'next', as usual, during that time, to be pushed out early in the
@@ -26,164 +26,52 @@ of the repositories listed at
26 --------------------------------------------------
27 [New Topics]
28
29 -* nd/index-pack-one-fd-per-thread (2014-04-09) 1 commit
30 - - index-pack: work around thread-unsafe pread()
31 -
32 - Enable threaded index-pack on platforms without thread-unsafe
33 - pread() emulation.
34 -
35 - Will merge to 'next' and keep it there for the remainder of the cycle.
36 -
37 -
38 -* tb/unicode-6.3-zero-width (2014-04-09) 1 commit
39 - - utf8.c: partially update to version 6.3
40 -
41 - Teach our display-column-counting logic about decomposed umlauts
42 - and friends.
43 -
44 - Will merge to 'next'.
45 -
46 -
47 -* fc/complete-aliased-push (2014-04-09) 1 commit
48 - - completion: fix completing args of aliased "push", "fetch", etc.
49 -
50 - Will merge to 'next'.
29 +* fc/prompt-zsh-read-from-file (2014-04-14) 1 commit
30 + - prompt: fix missing file errors in zsh
31
52 -
53 -* fc/remote-helper-fixes (2014-04-09) 4 commits
54 - - remote-bzr: include authors field in pushed commits
55 - - remote-bzr: add support for older versions
56 - - remote-hg: always normalize paths
57 - - remote-helpers: allow all tests running from any dir
58 -
59 - Will merge to 'next'.
60 -
61 -
62 -* fc/publish-vs-upstream (2014-04-11) 8 commits
63 - - sha1_name: add support for @{publish} marks
64 - - sha1_name: simplify track finding
65 - - sha1_name: cleanup interpret_branch_name()
66 - - branch: display publish branch
67 - - push: add --set-publish option
68 - - branch: add --set-publish-to option
69 - - Add concept of 'publish' branch
70 - - t5516 (fetch-push): fix test restoration
71 -
72 - Add branch@{publish}; this round v3 hasn't yet seen much reviews
73 - yet.
74 -
75 - Seems to have some interactions to break tests when merged to 'pu'.
76 -
77 -
78 -* ym/fix-opportunistic-index-update-race (2014-04-10) 2 commits
79 - - read-cache.c: verify index file before we opportunistically update it
80 - - wrapper.c: add xpread() similar to xread()
81 -
82 - Duy raised a good point that we may need to do the same for the
83 - normal writeout codepath, not just the "opportunistic" update
84 - codepath. While that is true, nobody sane would be running two
85 - simultaneous operations that are clearly write-oriented competing
86 - with each other against the same index file. So in that sense that
87 - can be done as a less urgent follow-up for this topic.
88 -
89 -
90 -* km/avoid-bs-in-shell-glob (2014-04-11) 1 commit
91 - - test: fix t5560 on FreeBSD
92 -
93 - Portability fix.
32 Will merge to 'next' and to 'master'.
33
34
97 -* km/avoid-cp-a (2014-04-11) 1 commit
98 - - test: fix t7001 cp to use POSIX options
35 +* fc/transport-helper-sync-error-fix (2014-04-14) 5 commits
36 + - transport-helper: fix sync issue on crashes
37 + - transport-helper: trivial cleanup
38 + - transport-helper: propagate recvline() error pushing
39 + - remote-helpers: make recvline return an error
40 + - transport-helper: remove barely used xchgline()
41
100 - Portability fix.
101 - Will merge to 'next' and hopefully to 'master'.
42 + Make sure the marks are not written out when the transport helper
43 + did not finish happily, to avoid marks file that is out of sync
44 + with the reality.
45
46 + As I had to resolve some conflicts with the recent code, will wait
47 + until getting an OK from Felipe on the conflict resolution and then
48 + hopefully can be merged to 'next' and then later to 'master'.
49
104 -* rs/ref-closer-to-atomic (2014-04-11) 3 commits
105 - - refs.c: change ref_transaction_commit to run the commit loops once all work is finished
106 - - refs.c: split delete_ref_loose() into a separate flag-for-deletion and commit phase
107 - - refs.c: split writing and commiting a ref into two separate functions
108 - (this branch uses mh/ref-transaction.)
50
110 - Builds on top of Michael's ref-transaction series to shrink the
111 - race window during multiple ref updates.
112 -
113 - Appears to break some tests even standalone, and somewhat conflicts
114 - with Michael's mh/lockfile topic when merging to 'pu'.
115 -
116 ---------------------------------------------------
117 -[Graduated to "master"]
118 -
119 -* ib/rev-parse-parseopt-argh (2014-04-01) 1 commit
120 - (merged to 'next' on 2014-04-01 at 025578d)
121 - + rev-parse: fix typo in example on manpage
122 -
123 - Finishing touch to a new topic scheduled for 2.0
124 -
125 -
126 -* jc/rev-parse-argh-dashed-multi-words (2014-03-24) 3 commits
127 - (merged to 'next' on 2014-03-31 at 1c48649)
128 - + parse-options: make sure argh string does not have SP or _
129 - + update-index: teach --cacheinfo a new syntax "mode,sha1,path"
130 - + parse-options: multi-word argh should use dash to separate words
131 -
132 - Make sure that the help text given to describe the "<param>" part
133 - of the "git cmd --option=<param>" does not contain SP or _,
134 - e.g. "--gpg-sign=<key-id>" option for "git commit" is not spelled
135 - as "--gpg-sign=<key id>".
51 +* km/avoid-non-function-return-in-rebase (2014-04-14) 2 commits
52 + - Revert "rebase: fix run_specific_rebase's use of "return" on FreeBSD"
53 + - rebase: avoid non-function use of "return" on FreeBSD
54
55 + Work around /bin/sh that does not like "return" at the top-level
56 + of a file that is dot-sourced from inside a function definition.
57
138 -* jk/commit-dates-parsing-fix (2014-04-01) 2 commits
139 - (merged to 'next' on 2014-04-04 at c16eeb0)
140 - + t4212: loosen far-in-future test for AIX
141 - + date: recognize bogus FreeBSD gmtime output
142 -
143 - Finishing touches for portability.
144 -
145 -
146 -* jk/pack-bitmap (2014-04-04) 2 commits
147 - (merged to 'next' on 2014-04-04 at 0306834)
148 - + pack-objects: do not reuse packfiles without --delta-base-offset
149 - + add `ignore_missing_links` mode to revwalk
150 -
151 - Fixes the pack-bitmap already in 'master'.
152 -
153 -
154 -* jl/nor-or-nand-and (2014-03-31) 4 commits
155 - (merged to 'next' on 2014-04-04 at b5d1ac5)
156 - + code and test: fix misuses of "nor"
157 - + comments: fix misuses of "nor"
158 - + contrib: fix misuses of "nor"
159 - + Documentation: fix misuses of "nor"
160 -
161 - Eradicate mistaken use of "nor" (that is, essentially "nor" used
162 - not in "neither A nor B" ;-)) from in-code comments, command output
163 - strings, and documentations.
164 -
165 -
166 -* mh/update-ref-batch-create-fix (2014-04-02) 1 commit
167 - (merged to 'next' on 2014-04-04 at 97e3f12)
168 - + update-ref: fail create operation over stdin if ref already exists
58 + Will merge to 'next'.
59
170 - Requesting "update-ref --stdin" to create a ref that already exists
171 - should have errored out, but didn't.
60 + We may want to merge it to 'master' by -rc1 if no regressions are
61 + reported.
62
63
174 -* mr/msvc-link-with-invalidcontinue (2014-03-28) 1 commit
175 - (merged to 'next' on 2014-03-31 at 051a29e)
176 - + MSVC: link in invalidcontinue.obj for better POSIX compatibility
64 +* db/make-with-curl (2014-04-15) 2 commits
65 + - Makefile: allow static linking against libcurl
66 + - Makefile: use curl-config to determine curl flags
67
68 + Ask curl-config how to link with the curl library, instead of
69 + having only a limited configurability knobs in the Makefile.
70
179 -* mr/opt-set-ptr (2014-03-31) 3 commits
180 - (merged to 'next' on 2014-04-03 at a26385b)
181 - + parse-options: remove unused OPT_SET_PTR
182 - + parse-options: add cast to correct pointer type to OPT_SET_PTR
183 - + MSVC: fix t0040-parse-options crash
71 + Will merge to 'next'.
72
185 - OPT_SET_PTR() implementation was broken on IL32P64 platforms;
186 - it turns out that the macro is not used by any real user.
73 + We may want to merge it to 'master' by -rc1 if no regressions are
74 + reported.
75
76 --------------------------------------------------
77 [Stalled]
@@ -354,43 +242,136 @@ of the repositories listed at
242 --------------------------------------------------
243 [Cooking]
244
245 +* nd/index-pack-one-fd-per-thread (2014-04-09) 1 commit
246 + - index-pack: work around thread-unsafe pread()
247 +
248 + Enable threaded index-pack on platforms without thread-unsafe
249 + pread() emulation.
250 +
251 + The log message may need to be replaced ($gmane/246176).
252 +
253 +
254 +* tb/unicode-6.3-zero-width (2014-04-09) 1 commit
255 + (merged to 'next' on 2014-04-14 at 72ce72a)
256 + + utf8.c: partially update to version 6.3
257 +
258 + Teach our display-column-counting logic about decomposed umlauts
259 + and friends.
260 +
261 + Will merge to 'master'.
262 +
263 +
264 +* fc/complete-aliased-push (2014-04-09) 1 commit
265 + - completion: fix completing args of aliased "push", "fetch", etc.
266 +
267 + Will merge to 'next'.
268 +
269 +
270 +* fc/remote-helper-fixes (2014-04-14) 5 commits
271 + - remote-bzr: trivial test fix
272 + - remote-bzr: include authors field in pushed commits
273 + - remote-bzr: add support for older versions
274 + - remote-hg: always normalize paths
275 + - remote-helpers: allow all tests running from any dir
276 +
277 + Will merge to 'next'.
278 +
279 +
280 +* fc/publish-vs-upstream (2014-04-11) 8 commits
281 + - sha1_name: add support for @{publish} marks
282 + - sha1_name: simplify track finding
283 + - sha1_name: cleanup interpret_branch_name()
284 + - branch: display publish branch
285 + - push: add --set-publish option
286 + - branch: add --set-publish-to option
287 + - Add concept of 'publish' branch
288 + - t5516 (fetch-push): fix test restoration
289 +
290 + Add branch@{publish}; this round v3 hasn't yet seen much reviews
291 + yet.
292 +
293 + Seems to have some interactions to break t5516 when merged to 'pu'.
294 +
295 +
296 +* ym/fix-opportunistic-index-update-race (2014-04-10) 2 commits
297 + - read-cache.c: verify index file before we opportunistically update it
298 + - wrapper.c: add xpread() similar to xread()
299 +
300 + Duy raised a good point that we may need to do the same for the
301 + normal writeout codepath, not just the "opportunistic" update
302 + codepath. While that is true, nobody sane would be running two
303 + simultaneous operations that are clearly write-oriented competing
304 + with each other against the same index file. So in that sense that
305 + can be done as a less urgent follow-up for this topic.
306 +
307 + Will merge to 'next' and keep it there for the remainder of the cycle.
308 +
309 +
310 +* km/avoid-bs-in-shell-glob (2014-04-11) 1 commit
311 + (merged to 'next' on 2014-04-14 at a3d9a58)
312 + + test: fix t5560 on FreeBSD
313 +
314 + Portability fix.
315 + Will merge to 'master'.
316 +
317 +
318 +* km/avoid-cp-a (2014-04-11) 1 commit
319 + (merged to 'next' on 2014-04-14 at be661c4)
320 + + test: fix t7001 cp to use POSIX options
321 +
322 + Portability fix.
323 + Will merge to 'master'.
324 +
325 +
326 * jl/status-added-submodule-is-never-ignored (2014-04-07) 2 commits
327 - commit -m: commit staged submodules regardless of ignore config
328 - status/commit: show staged submodules regardless of ignore config
329
330 + There also are a few patches Ronald Weiss and Jens are working on
331 + polishing around this topic, and a patch from Jens each for gitk
332 + and git-gui.
333 +
334 + Waiting for the dust to settle until picking them up all.
335 +
336
337 * mh/multimail (2014-04-07) 1 commit
363 - - git-multimail: update to version 1.0.0
338 + (merged to 'next' on 2014-04-15 at eaba915)
339 + + git-multimail: update to version 1.0.0
340
341 + Will merge to 'master'.
342
366 -* mh/lockfile (2014-04-07) 25 commits
343 +
344 +* mh/lockfile (2014-04-15) 25 commits
345 - trim_last_path_elm(): replace last_path_elm()
346 - resolve_symlink(): take a strbuf parameter
347 - resolve_symlink(): use a strbuf for internal scratch space
370 - - Change lock_file::filename into a strbuf
348 + - change lock_file::filename into a strbuf
349 - commit_lock_file(): use a strbuf to manage temporary space
350 - try_merge_strategy(): use a statically-allocated lock_file object
351 - try_merge_strategy(): remove redundant lock_file allocation
352 + - struct lock_file: declare some fields volatile
353 - lockfile: avoid transitory invalid states
375 - - commit_lock_file(): make committing an unlocked lockfile a NOP
354 + - commit_lock_file(): die() if called for unlocked lockfile object
355 - commit_lock_file(): inline temporary variable
356 - remove_lock_file(): call rollback_lock_file()
357 - lock_file(): exit early if lockfile cannot be opened
358 - write_packed_entry_fn(): convert cb_data into a (const int *)
359 - prepare_index(): declare return value to be (const char *)
360 - delete_ref_loose(): don't muck around in the lock_file's filename
382 - - lockfile: define a constant LOCK_SUFFIX_LEN
361 + - cache.h: define constants LOCK_SUFFIX and LOCK_SUFFIX_LEN
362 - lockfile.c: document the various states of lock_file objects
384 - - struct lock_file: replace on_list field with flags field
363 - lock_file(): always add lock_file object to lock_file_list
364 - hold_lock_file_for_append(): release lock on errors
365 - lockfile: unlock file if lockfile permissions cannot be adjusted
366 - rollback_lock_file(): set fd to -1
367 - rollback_lock_file(): do not clear filename redundantly
390 - - unable_to_lock_die(): rename function from unable_to_lock_index_die()
368 - api-lockfile: expand the documentation
369 + - unable_to_lock_die(): rename function from unable_to_lock_index_die()
370 +
371 + Refactor and fix corner-case bugs in the lockfile API, all looked
372 + sensible.
373
393 - Refactor and fix corner-case bugs in the lockfile API.
374 + Still being commented on.
375
376
377 * mt/patch-id-stable (2014-03-31) 3 commits
@@ -408,7 +389,7 @@ of the repositories listed at
389 people complain. It could be that we do not have to worry about
390 the default flipping at all. We'll see.
391
411 - Will keep in 'next' for the remainder of this cycle.
392 + Will keep in 'next' for the remainder of the cycle.
393
394
395 * mh/ref-transaction (2014-04-07) 27 commits
@@ -444,7 +425,7 @@ of the repositories listed at
425 Update "update-ref --stdin [-z]" and then introduce a transactional
426 support for (multi-)reference updates.
427
447 - Is this ready to be merged to 'next' for wider exposure?
428 + Will merge to 'next' and keep it there for the remainder of the cycle.
429
430
431 * jc/apply-ignore-whitespace (2014-03-26) 1 commit
@@ -456,7 +437,7 @@ of the repositories listed at
437 inconsistent with the option of the same name "diff" and "git diff"
438 have.
439
459 - Will keep in 'next' for the remainder of this cycle.
440 + Will keep in 'next' for the remainder of the cycle.
441
442
443 * as/grep-fullname-config (2014-03-20) 1 commit
@@ -469,7 +450,7 @@ of the repositories listed at
450 This may cause regressions on scripted users that do not expect
451 this new behaviour.
452
472 - Will keep in 'next' for the remainder of this cycle.
453 + Will keep in 'next' for the remainder of the cycle.
454
455
456 * nd/multiple-work-trees (2014-03-25) 28 commits
@@ -543,7 +524,7 @@ of the repositories listed at
524 detections and such. And N=2 case nicely degenerates to the usual
525 2-way diff-tree, which is very nice.
526
546 - Will keep in 'next' for the remainder of this cycle.
527 + Will keep in 'next' for the remainder of the cycle.
528
529
530 * cc/interpret-trailers (2014-04-07) 12 commits
@@ -564,7 +545,7 @@ of the repositories listed at
545 messages.
546
547 I was planning to merge it to 'next' and keep it there for the
567 - remainder of this cycle, but it appears that there still will be
548 + remainder of the cycle, but it appears that there still will be
549 another round of reroll, at least for the documentation?
550
551 --------------------------------------------------
@@ -574,3 +555,15 @@ of the repositories listed at
555 . Enable index-pack threading in msysgit.
556
557 Queued a different attempt by Duy on nd/index-pack-one-fd-per-thread
558 +
559 +
560 +* rs/ref-closer-to-atomic (2014-04-14) 3 commits
561 + . refs.c: change ref_transaction_commit to run the commit loops once all work is finished
562 + . refs.c: split delete_ref_loose() into a separate flag-for-deletion and commit phase
563 + . refs.c: split writing and commiting a ref into two separate functions
564 + (this branch uses mh/ref-transaction.)
565 +
566 + Builds on top of Michael's ref-transaction series to shrink the
567 + race window during multiple ref updates.
568 +
569 + Deferred to be rerolled ($gmane/246289).