What's cooking (2020/08 #04)

Junio C Hamano committed Aug 18, 2020 at 15:48 UTC 9be24a30d9a64db5c83590d29fc0d4e1511cfdad
1 file changed +470 -416
whats-cooking.txt
+470 -416
@@ -1,10 +1,10 @@
1 To: git@vger.kernel.org
2 Bcc: lwn@lwn.net
3 -Subject: What's cooking in git.git (Aug 2020, #03; Thu, 13)
4 -X-master-at: 878e727637ec5815ccb3301eb994a54df95b21b8
5 -X-next-at: 1956fa8f8d7168157346f0cbc0f1f9dd1fca787a
3 +Subject: What's cooking in git.git (Aug 2020, #04; Tue, 18)
4 +X-master-at: 2befe97201e1f3175cce557866c5822793624b5a
5 +X-next-at: 76ed4a4daf6aec5583ad1c21b86f032a08bf353f
6
7 -What's cooking in git.git (Aug 2020, #03; Thu, 13)
7 +What's cooking in git.git (Aug 2020, #04; Tue, 18)
8 --------------------------------------------------
9
10 Here are the topics that have been cooking. Commits prefixed with '-' are
@@ -12,13 +12,6 @@ only in 'seen' (formerly 'pu'---proposed updates) while commits prefixed
12 with '+' are in 'next'. The ones marked with '.' do not appear in any of
13 the integration branches, but I am still holding onto them.
14
15 -We are at 7th batch. I'm trying to merge smaller batches of topics
16 -to the 'master' branch more often than the previous cycles. I will
17 -also experiment, for smaller topics, with shorter gestation period
18 -in 'next' before merging them to the 'master' branch. We'll see if
19 -this lowers stability of the tip of 'master' significantly; not that
20 -much, hopefully.
21 -
15 You can find the changes described here in the integration branches of the
16 repositories listed at
17
@@ -27,147 +20,78 @@ repositories listed at
20 --------------------------------------------------
21 [Graduated to 'master']
22
30 -* bc/sha-256-cvs-svn-updates (2020-08-05) 1 commit
31 - (merged to 'next' on 2020-08-07 at b1ce7e5ec5)
32 - + git-cvsexportcommit: support Perl before 5.10.1
33 -
34 - Portability fix.
35 -
36 -
37 -* bc/sha-256-part-3 (2020-07-30) 39 commits
38 - (merged to 'next' on 2020-08-04 at 57115e548f)
39 - + t: remove test_oid_init in tests
40 - + docs: add documentation for extensions.objectFormat
41 - + ci: run tests with SHA-256
42 - + t: make SHA1 prerequisite depend on default hash
43 - + t: allow testing different hash algorithms via environment
44 - + t: add test_oid option to select hash algorithm
45 - + repository: enable SHA-256 support by default
46 - + setup: add support for reading extensions.objectformat
47 - + bundle: add new version for use with SHA-256
48 - + builtin/verify-pack: implement an --object-format option
49 - + http-fetch: set up git directory before parsing pack hashes
50 - + t0410: mark test with SHA1 prerequisite
51 - + t5308: make test work with SHA-256
52 - + t9700: make hash size independent
53 - + t9500: ensure that algorithm info is preserved in config
54 - + t9350: make hash size independent
55 - + t9301: make hash size independent
56 - + t9300: use $ZERO_OID instead of hard-coded object ID
57 - + t9300: abstract away SHA-1-specific constants
58 - + t8011: make hash size independent
59 - + t8003: make hash size independent
60 - + t8002: make hash size independent
61 - + t7508: use $ZERO_OID instead of hard-coded constant
62 - + t7506: avoid checking for SHA-1-specific constants
63 - + t7405: make hash size independent
64 - + t7400: make hash size independent
65 - + t7102: abstract away SHA-1-specific constants
66 - + t7201: abstract away SHA-1-specific constants
67 - + t7063: make hash size independent
68 - + t7003: compute appropriate length constant
69 - + t6501: avoid hard-coded objects
70 - + t6500: specify test values for SHA-256
71 - + t6301: make hash size independent
72 - + t6101: make hash size independent
73 - + t6100: make hash size independent
74 - + t3404: prepare 'short SHA-1 collision' tests for SHA-256
75 - + t3305: make hash agnostic
76 - + t1001: use $ZERO_OID
77 - + t: make test-bloom initialize repository
78 -
79 - The final leg of SHA-256 transition.
80 -
81 -
82 -* es/worktree-doc-cleanups (2020-08-03) 5 commits
83 - (merged to 'next' on 2020-08-04 at cbb53ca464)
84 - + git-worktree.txt: link to man pages when citing other Git commands
85 - + git-worktree.txt: make start of new sentence more obvious
86 - + git-worktree.txt: fix minor grammatical issues
87 - + git-worktree.txt: consistently use term "working tree"
88 - + git-worktree.txt: employ fixed-width typeface consistently
89 -
90 - Doc cleanup around "worktree".
91 -
92 -
93 -* jt/has_object (2020-08-06) 4 commits
94 - (merged to 'next' on 2020-08-07 at ed08abb693)
95 - + fsck: do not lazy fetch known non-promisor object
96 - + pack-objects: no fetch when allow-{any,promisor}
97 - + apply: do not lazy fetch when applying binary
98 - + sha1-file: introduce no-lazy-fetch has_object()
99 -
100 - A new helper function has_object() has been introduced to make it
101 - easier to mark object existence checks that do and don't want to
102 - trigger lazy fetches, and a few such checks are converted using it.
103 -
104 -
105 -* ma/test-quote-cleanup (2020-08-06) 2 commits
106 - (merged to 'next' on 2020-08-10 at 63a95c2926)
107 - + t4104: modernize and simplify quoting
108 - + t: don't spuriously close and reopen quotes
109 -
110 - Test cleanup.
111 -
112 -
113 -* rp/blame-first-parent-doc (2020-08-06) 1 commit
114 - (merged to 'next' on 2020-08-10 at 3fdbebe1ea)
115 - + blame-options.txt: document --first-parent option
116 -
117 - The "git blame --first-parent" option was not documented, but now
118 - it is.
119 -
120 -
121 -* ss/cmake-build (2020-06-26) 8 commits
122 - (merged to 'next' on 2020-08-03 at a0d70165c1)
123 - + ci: modification of main.yml to use cmake for vs-build job
124 - + cmake: support for building git on windows with msvc and clang.
125 - + cmake: support for building git on windows with mingw
126 - + cmake: support for testing git when building out of the source tree
127 - + cmake: support for testing git with ctest
128 - + cmake: installation support for git
129 - + cmake: generate the shell/perl/python scripts and templates, translations
130 - + Introduce CMake support for configuring Git
23 +* al/bisect-first-parent (2020-08-07) 5 commits
24 + (merged to 'next' on 2020-08-11 at b232f7ca17)
25 + + bisect: combine args passed to find_bisection()
26 + + bisect: introduce first-parent flag
27 + + cmd_bisect__helper: defer parsing no-checkout flag
28 + + rev-list: allow bisect and first-parent flags
29 + + t6030: modernize "git bisect run" tests
30
132 - Originally merged to 'next' on 2020-08-01
31 + "git bisect" learns the "--first-parent" option to find the first
32 + breakage along the first-parent chain.
33
134 - CMake support to build with MSVC for Windows bypassing the Makefile.
34
35 +* dd/send-email-config (2020-07-23) 1 commit
36 + (merged to 'next' on 2020-08-11 at aaf9f3a301)
37 + + git-send-email: die if sendmail.* config is set
38
137 -* tb/upload-pack-filters (2020-08-05) 4 commits
138 - (merged to 'next' on 2020-08-05 at 918e7092fe)
139 - + t5616: use test_i18ngrep for upload-pack errors
140 - (merged to 'next' on 2020-08-04 at 3ae57cc90c)
141 - + upload-pack.c: introduce 'uploadpackfilter.tree.maxDepth'
142 - + upload-pack.c: allow banning certain object filter(s)
143 - + list_objects_filter_options: introduce 'list_object_filter_config_name'
39 + Stop when "sendmail.*" configuration variables are defined, which
40 + could be a mistaken attempt to define "sendemail.*" variables.
41
145 - The component to respond to "git fetch" request is made more
146 - configurable to selectively allow or reject object filtering
147 - specification used for partial cloning.
148 - cf. <20200804003722.GA2726931@coredump.intra.peff.net>
42
150 ---------------------------------------------------
151 -[New Topics]
43 +* es/test-cmp-typocatcher (2020-08-09) 1 commit
44 + (merged to 'next' on 2020-08-11 at ee5557b170)
45 + + test_cmp: diagnose incorrect arguments
46
153 -* en/merge-tests (2020-08-10) 11 commits
154 - (merged to 'next' on 2020-08-13 at eab952350b)
155 - + t6425: be more flexible with rename/delete conflict messages
156 - + t642[23]: be more flexible for add/add conflicts involving pair renames
157 - + t6422, t6426: be more flexible for add/add conflicts involving renames
158 - + t6423: add an explanation about why one of the tests does not pass
159 - + t6416, t6423: clarify some comments and fix some typos
160 - + t6422: fix multiple errors with the mod6 test expectations
161 - + t6423: fix test setup for a couple tests
162 - + t6416, t6422: fix incorrect untracked file count
163 - + t6422: fix bad check against missing file
164 - + t6418: tighten delete/normalize conflict testcase
165 - + Collect merge-related tests to t64xx
47 + Test framework update.
48
167 - Updates to "git merge" tests, in preparation for a new merge
168 - strategy backend.
49
170 - Will merge to 'master'.
50 +* hn/reftable-prep-part-2 (2020-07-27) 3 commits
51 + (merged to 'next' on 2020-08-11 at 43ac0bc60a)
52 + + Make HEAD a PSEUDOREF rather than PER_WORKTREE.
53 + + Modify pseudo refs through ref backend storage
54 + + t1400: use git rev-parse for testing PSEUDOREF existence
55 +
56 + Further preliminary change to refs API.
57 +
58 +
59 +* jc/noop-with-static-inline (2020-08-06) 1 commit
60 + (merged to 'next' on 2020-08-11 at 8a63ff648e)
61 + + compat-util: type-check parameters of no-op replacement functions
62 +
63 + A no-op replacement function implemented as a C preprocessor macro
64 + does not perform as good a job as one implemented as a "static
65 + inline" function in catching errors in parameters; replace the
66 + former with the latter in <git-compat-util.h> header.
67 +
68 +
69 +* jk/log-fp-implies-m (2020-07-29) 7 commits
70 + (merged to 'next' on 2020-08-03 at 39fefa6b82)
71 + + doc/git-log: clarify handling of merge commit diffs
72 + + doc/git-log: move "-t" into diff-options list
73 + + doc/git-log: drop "-r" diff option
74 + + doc/git-log: move "Diff Formatting" from rev-list-options
75 + + log: enable "-m" automatically with "--first-parent"
76 + + revision: add "--no-diff-merges" option to counteract "-m"
77 + + log: drop "--cc implies -m" logic
78 + (this branch is used by so/log-diff-merges-opt.)
79 +
80 + Originally merged to 'next' on 2020-08-01
81 +
82 + "git log --first-parent -p" showed patches only for single-parent
83 + commits on the first-parent chain; the "--first-parent" option has
84 + been made to imply "-m". Use "--no-diff-merges" to restore the
85 + previous behaviour to omit patches for merge commits.
86 +
87 + Will merge to 'master' together with 'so/log-diff-merges-opt'
88 +
89 +
90 +* jk/sideband-error-l10n (2020-08-07) 1 commit
91 + (merged to 'next' on 2020-08-11 at 27dbe05a65)
92 + + sideband: mark "remote error:" prefix for translation
93 +
94 + Mark error message for i18n.
95
96
97 * ma/stop-progress-null-fix (2020-08-10) 1 commit
@@ -176,25 +100,41 @@ repositories listed at
100
101 NULL dereference fix.
102
179 - Will merge to 'master'.
103
104 +* pd/mergetool-nvimdiff (2020-07-29) 2 commits
105 + (merged to 'next' on 2020-08-11 at 947ce686d0)
106 + + mergetools: add support for nvimdiff (neovim) family
107 + + mergetool--lib: improve support for vimdiff-style tool variants
108
182 -* ds/midx-repack-to-batch-size (2020-08-11) 1 commit
183 - - multi-pack-index: repack batches below --batch-size
109 + The existing backends for "git mergetool" based on variants of vim
110 + have been refactored and then support for "nvim" has been added.
111
185 - The "--batch-size" option of "git multi-pack-index repack" command
186 - is now used to specify that very small packfiles are collected into
187 - one until the total size roughly exceeds it.
112
189 - Will merge to 'next'.
113 +* ps/ref-transaction-hook (2020-08-11) 2 commits
114 + (merged to 'next' on 2020-08-11 at 311c045810)
115 + + t1416: avoid hard-coded sha1 ids
116 + (merged to 'next' on 2020-08-10 at d8ad7cc8f6)
117 + + refs: fix interleaving hook calls with reference-transaction hook
118
119 + The logic to find the ref transaction hook script attempted to
120 + cache the path to the found hook without realizing that it needed
121 + to keep a copied value, as the API it used returned a transitory
122 + buffer space. This has been corrected.
123 + to be followed by a removal of the caching feature, which does not
124 + seem to help even as a negative cache.
125
192 -* rs/upload-pack-sigchain-fix (2020-08-11) 1 commit
193 - - upload-pack: remove superfluous sigchain_pop() call
126
195 - Code clean-up.
127 +* rp/apply-cached-with-i-t-a (2020-08-09) 3 commits
128 + (merged to 'next' on 2020-08-11 at c28eaa6a0e)
129 + + t4140: test apply with i-t-a paths
130 + + apply: make i-t-a entries never match worktree
131 + + apply: allow "new file" patches on i-t-a entries
132
197 - Will merge to 'next'.
133 + Recent versions of "git diff-files" shows a diff between the index
134 + and the working tree for "intent-to-add" paths as a "new file"
135 + patch; "git apply --cached" should be able to take "git diff-files"
136 + and should act as an equivalent to "git add" for the path, but the
137 + command failed to do so for such a path.
138
139
140 * so/log-diff-merges-opt (2020-08-11) 3 commits
@@ -215,161 +155,200 @@ repositories listed at
155
156 Will merge to 'master' together with 'jk/lo-implies-m'.
157
158 +--------------------------------------------------
159 +[New Topics]
160
219 -* tb/bloom-improvements (2020-08-11) 14 commits
220 - - builtin/commit-graph.c: introduce '--max-new-filters=<n>'
221 - - commit-graph: rename 'split_commit_graph_opts'
222 - - commit-graph: add large-filters bitmap chunk
223 - - commit-graph.c: sort index into commits list
224 - - bloom/diff: properly short-circuit on max_changes
225 - - bloom: use provided 'struct bloom_filter_settings'
226 - - csum-file.h: introduce 'hashwrite_be64()'
227 - - bloom: split 'get_bloom_filter()' in two
228 - - commit-graph.c: store maximum changed paths
229 - - commit-graph: respect 'commitGraph.readChangedPaths'
230 - - t/helper/test-read-graph.c: prepare repo settings
231 - - commit-graph: pass a 'struct repository *' in more places
232 - - t4216: use an '&&'-chain
233 - - commit-graph: introduce 'get_bloom_filter_settings()'
161 +* bc/sha-256-doc-updates (2020-08-13) 2 commits
162 + (merged to 'next' on 2020-08-17 at 7154a0ce3c)
163 + + docs: fix step in transition plan
164 + + docs: document SHA-256 pack and indices
165
235 - Misc Bloom filter improvements.
166 + Further update of docs to adjust to the recent SHA-256 work.
167
237 - Expecting a reroll.
238 - It seems that the review is getting closer to result in another update.
239 - cf. <20200811220503.GC66656@syl.lan>
168 + Will merge to 'master'.
169
170
242 -* ak/sequencer-fix-find-uniq-abbrev (2020-08-13) 1 commit
243 - - rebase -i: fix possibly wrong onto hash in todo
171 +* ma/sha-256-docs (2020-08-17) 4 commits
172 + (merged to 'next' on 2020-08-17 at 6861142a20)
173 + + shallow.txt: document SHA-256 shallow format
174 + + protocol-capabilities.txt: clarify "allow-x-sha1-in-want" re SHA-256
175 + + index-format.txt: document SHA-256 index format
176 + + http-protocol.txt: document SHA-256 "want"/"have" format
177
245 - Ring buffer with size 4 used for bin-hex translation resulted in a
246 - wrong object name in the sequencer's todo output, which has been
247 - corrected.
178 + Further update of docs to adjust to the recent SHA-256 work.
179
249 - Will merge to 'next'.
180 + Will merge to 'master'.
181
182
252 -* en/dir-nonbare-embedded (2020-08-12) 2 commits
253 - - dir: avoid prematurely marking nonbare repositories as matches
254 - - t3000: fix some test description typos
183 +* dd/diff-customize-index-line-abbrev (2020-08-13) 2 commits
184 + - diff: index-line: respect --abbrev in object's name
185 + - revision: differentiate if --no-abbrev asked explicitly
186
256 - "ls-files -o" mishandled the top-level directory of another git
257 - working tree that hangs in the current git working tree.
187 + The output from the "diff" family of the commands had abbreviated
188 + object names of blobs involved in the patch, but its length was not
189 + affected by the --abbrev option. Now it is.
190
259 - Will merge to 'next'.
191 + The first patch is unneeded.
192
193
262 -* en/sequencer-merge-labels (2020-08-12) 1 commit
263 - - sequencer: avoid garbled merge machinery messages due to commit labels
194 +* ds/sha256-leftover-bits (2020-08-17) 3 commits
195 + (merged to 'next' on 2020-08-17 at a8e6f9f3f4)
196 + + multi-pack-index: use hash version byte
197 + + commit-graph: use the "hash version" byte
198 + + t/README: document GIT_TEST_DEFAULT_HASH
199
265 - The commit labels used to explain each side of conflicted hunks
266 - placed by the sequencer machinery have been made more readable by
267 - humans.
200 + midx and commit-graph files now use the byte defined in their file
201 + format specification for identifying the hash function used for
202 + object names.
203
269 - Will merge to 'next'.
204 + Will merge to 'master'.
205
206
272 -* jb/commit-graph-doc-fix (2020-08-13) 1 commit
273 - - docs: commit-graph: fix some whitespace in the diagram
207 +* hn/refs-fetch-head-is-special (2020-08-17) 4 commits
208 + - refs: read FETCH_HEAD generically
209 + - refs: move gitdir into base ref_store
210 + - refs: fix comment about submodule ref_stores
211 + - Split off reading loose ref data in separate function
212 + (this branch is used by hn/refs-pseudorefs.)
213
275 - Docfix.
214 + The FETCH_HEAD is now always read from the filesystem regardless of
215 + the ref backend in use, as its format is much richer than the
216 + normal refs, and written directly by "git fetch" as a plain file..
217
277 - Will merge to 'next'.
218
219 +* hv/ref-filter-misc (2020-08-17) 9 commits
220 + - ref-filter: add `sanitize` option for 'subject' atom
221 + - format-support: move `format_sanitized_subject()` from pretty
222 + - pretty: refactor `format_sanitized_subject()`
223 + - ref-filter: add `short` modifier to 'parent' atom
224 + - ref-filter: add `short` modifier to 'tree' atom
225 + - ref-filter: rename `objectname` related functions and fields
226 + - ref-filter: modify error messages in `grab_objectname()`
227 + - ref-filter: refactor `grab_objectname()`
228 + - ref-filter: support different email formats
229
280 -* jk/blame-coalesce-fix (2020-08-13) 3 commits
281 - - blame: only coalesce lines that are adjacent in result
282 - - t8003: factor setup out of coalesce test
283 - - t8003: check output of coalesced blame
230 + The "--format=" option to the "for-each-ref" command and friends
231 + learned a few more tricks, e.g. the ":short" suffix that applies to
232 + "objectname" now also can be used for "parent", "tree", etc.
233
285 - When given more than one target line ranges, "git blame -La,b
286 - -Lc,d" was over-eager to coalesce groups of original lines and
287 - showed incorrect results, which has been corrected.
234
289 - Will merge to 'next'.
235 +* jc/object-names-are-not-sha-1 (2020-08-14) 1 commit
236 + (merged to 'next' on 2020-08-17 at 400fe9e831)
237 + + messages: avoid SHA-1 in end-user facing messages
238
239 + A few end-user facing messages have been update to be
240 + hash-algorithm agnostic.
241
292 -* jk/slimmed-down (2020-08-13) 5 commits
293 - - drop vcs-svn experiment
294 - - make git-fast-import a builtin
295 - - make git-bugreport a builtin
296 - - make credential helpers builtins
297 - - Makefile: drop builtins from MSVC pdb list
242 + Will merge to 'master'.
243
299 - Trim an unused binary and turn a bunch of commands into built-in.
244
301 - In general this is a good thing to do, modulo 'bugreport'.
245 +* jh/mingw-unlink (2020-08-17) 1 commit
246 + (merged to 'next' on 2020-08-17 at fb53a3d1a1)
247 + + mingw: improve performance of mingw_unlink()
248
249 + "unlink" emulation on MinGW has been optimized.
250
304 -* jk/unleak-fixes (2020-08-13) 2 commits
305 - - ls-remote: simplify UNLEAK() usage
306 - - stop calling UNLEAK() before die()
251 + Will merge to 'master'.
252
308 - Fix some incorrect UNLEAK() annotations.
253 +
254 +* jk/leakfix (2020-08-17) 7 commits
255 + - submodule--helper: fix leak of core.worktree value
256 + - config: fix leak in git_config_get_expiry_in_days()
257 + - config: drop git_config_get_string_const()
258 + - config: fix leaks from git_config_get_string_const()
259 + - checkout: fix leak of non-existent branch names
260 + - submodule--helper: use strbuf_release() to free strbufs
261 + - clear_pattern_list(): clear embedded hashmaps
262 +
263 + Code clean-up.
264
265 Will merge to 'next'.
266
267
313 -* pb/set-url-docfix (2020-08-13) 1 commit
314 - - fetch, pull doc: correct description of '--set-upstream'
268 +* jk/refspecs-negative (2020-08-17) 3 commits
269 + - refspec: add support for negative refspecs
270 + - refspec: make sure stack refspec_item variables are zeroed
271 + - refspec: fix documentation referring to refspec_item
272
316 - Doc fix.
273 + "negative refspecs"
274
318 - Will merge to 'next'.
275
276 +* jt/fetch-pack-loosen-validation-with-packfile-uri (2020-08-14) 1 commit
277 + - fetch-pack: make packfile URIs work with transfer.fsckobjects
278
321 -* pb/userdiff-fortran-update (2020-08-13) 2 commits
322 - - userdiff: improve Fortran xfuncname regex
323 - - userdiff: add tests for Fortran xfuncname regex
279 + Bugfix for "git fetch" when the packfile URI capability is in use.
280
325 - The regexp to identify the function boundary for FORTRAN programs
326 - has been updated.
281 + Need to pick up a reroll.
282 +
283 +
284 +* mr/diff-hide-stat-wo-textual-change (2020-08-17) 1 commit
285 + - diff: teach --stat to ignore uninteresting modifications
286 +
287 + Expecting a reroll.
288 +
289 +
290 +* pw/add-p-allowed-options-fix (2020-08-17) 2 commits
291 + - add -p: fix checking of user input
292 + - add -p: use ALLOC_GROW_BY instead of ALLOW_GROW
293 +
294 + "git add -p" update.
295 +
296 + Doubts?
297 + cf. <pull.702.git.1597670589.gitgitgadget@gmail.com>
298 +
299 +
300 +* dl/subtree-docs (2020-08-18) 2 commits
301 + - contrib/subtree: document 'push' does not take '--squash'
302 + - contrib/subtree: fix "unsure" for --message in the document
303 +
304 + Doc updates for subtree (in contrib/)
305
306 Will merge to 'next'.
307
308
331 -* rs/fast-export-anon-simplify (2020-08-13) 1 commit
332 - - fast-export: factor out print_oid()
309 +* en/mem-pool (2020-08-18) 3 commits
310 + - mem-pool: use consistent pool variable name
311 + - mem-pool: use more standard initialization and finalization
312 + - mem-pool: add convenience functions for strdup and strndup
313
334 - Code simplification.
314 + API update.
315
316 Will merge to 'next'.
317
318
339 -* rs/preserve-merges-unused-code-removal (2020-08-12) 1 commit
340 - - rebase: remove unused function reschedule_last_action
319 +* jc/no-update-fetch-head (2020-08-18) 1 commit
320 + - fetch: optionally allow disabling FETCH_HEAD update
321 + (this branch is used by ds/maintenance-part-1, ds/maintenance-part-2 and jt/lazy-fetch.)
322
342 - Code clean-up.
323 + "git fetch" learned --no-write-fetch-head option to avoid writing
324 + the FETCH_HEAD file.
325
326 Will merge to 'next'.
327
328
347 -* ss/t7401-modernize (2020-08-12) 4 commits
348 - - t7401: add a NEEDSWORK
349 - - t7401: change indentation for enhanced readability
350 - - t7401: change test_i18ncmp syntax for clarity
351 - - t7401: modernize style
329 +* jt/lazy-fetch (2020-08-18) 7 commits
330 + - fetch-pack: remove no_dependents code
331 + - promisor-remote: lazy-fetch objects in subprocess
332 + - fetch-pack: do not lazy-fetch during ref iteration
333 + - fetch: only populate existing_refs if needed
334 + - fetch: avoid reading submodule config until needed
335 + - fetch: allow refspecs specified through stdin
336 + - negotiator/noop: add noop fetch negotiator
337 + (this branch uses jc/no-update-fetch-head; is tangled with ds/maintenance-part-1 and ds/maintenance-part-2.)
338
353 - Test clean-up.
339 + Updates to on-demand fetching code in lazily cloned repositories.
340
355 - Expecting a reroll.
356 - cf. <c70f7bb5d1289f5318da7b99e41cf5828a451174.camel@gmail.com>
357 - Almost there, it seems.
341
359 ---------------------------------------------------
360 -[Stalled]
342
362 -* pw/rebase-i-more-options (2020-07-16) 5 commits
363 - - rebase: add --reset-author-date
364 - - rebase -i: support --ignore-date
365 - - sequencer: rename amend_author to author_to_free
366 - - rebase -i: support --committer-date-is-author-date
367 - - rebase -i: add --ignore-whitespace flag
343 +* mt/checkout-entry-dead-code-removal (2020-08-18) 1 commit
344 + - checkout_entry(): remove unreachable error() call
345
369 - "git rebase -i" learns a bit more options.
346 + Code clean-up.
347
371 - Waiting for a (hopefully final) review.
348 + Will merge to 'next'.
349
350 +--------------------------------------------------
351 +[Stalled]
352
353 * mt/grep-sparse-checkout (2020-06-12) 6 commits
354 - config: add setting to ignore sparsity patterns in some cmds
@@ -386,14 +365,6 @@ repositories listed at
365 cf. <CABPp-BGdEyEeajYZj_rdxp=MyEQdszuyjVTax=hhYj3fOtRQUQ@mail.gmail.com>
366
367
389 -* rs/more-buffered-io (2020-08-02) 3 commits
390 - - upload-pack: use buffered I/O to talk to rev-list
391 - - midx: use buffered I/O to talk to pack-objects
392 - - connected: use buffered I/O to talk to rev-list
393 -
394 - Expecting a reroll.
395 -
396 -
368 * ls/mergetool-meld-auto-merge (2020-07-12) 2 commits
369 - SQUASH???
370 - Support auto-merge for meld to follow the vim-diff behavior
@@ -405,26 +376,6 @@ repositories listed at
376 Expecting a reroll.
377
378
408 -* jx/proc-receive-hook (2020-05-18) 11 commits
409 - . doc: add documentation for the proc-receive hook
410 - . transport: parse report options for tracking refs
411 - . t5411: test updates of remote-tracking branches
412 - . receive-pack: new config receive.procReceiveRefs
413 - . refs.c: refactor to reuse ref_is_hidden()
414 - . receive-pack: feed report options to post-receive
415 - . doc: add document for capability report-status-v2
416 - . New capability "report-status-v2" for git-push
417 - . receive-pack: add new proc-receive hook
418 - . t5411: add basic test cases for proc-receive hook
419 - . transport: not report a non-head push as a branch
420 -
421 - "git receive-pack" that accepts requests by "git push" learned to
422 - outsource most of the ref updates to the new "proc-receive" hook.
423 -
424 - Ejected out of 'seen'; somehow its tests seem to break with clang
425 - cf. https://travis-ci.org/github/git/git/builds/713443572
426 -
427 -
379 * mf/submodule-summary-with-correct-repository (2020-06-24) 2 commits
380 - submodule: use submodule repository when preparing summary
381 - revision: use repository from rev_info when parsing commits
@@ -482,207 +433,299 @@ repositories listed at
433 --------------------------------------------------
434 [Cooking]
435
485 -* ds/maintenance-part-2 (2020-08-06) 9 commits
486 - - maintenance: add incremental-repack auto condition
487 - - maintenance: auto-size incremental-repack batch
488 - - maintenance: add incremental-repack task
489 - - midx: use start_delayed_progress()
490 - - midx: enable core.multiPackIndex by default
491 - - maintenance: create auto condition for loose-objects
492 - - maintenance: add loose-objects task
493 - - maintenance: add prefetch task
494 - - fetch: optionally allow disabling FETCH_HEAD update
495 - (this branch uses ds/maintenance.)
436 +* jx/proc-receive-hook (2020-08-17) 10 commits
437 + - doc: add documentation for the proc-receive hook
438 + - transport: parse report options for tracking refs
439 + - t5411: test updates of remote-tracking branches
440 + - receive-pack: new config receive.procReceiveRefs
441 + - receive-pack: feed report options to post-receive
442 + - doc: add document for capability report-status-v2
443 + - New capability "report-status-v2" for git-push
444 + - receive-pack: add new proc-receive hook
445 + - t5411: add basic test cases for proc-receive hook
446 + - transport: not report a non-head push as a branch
447
497 - "git maintenance", an extended big brother of "git gc", continues
498 - to evolve.
448 + "git receive-pack" that accepts requests by "git push" learned to
449 + outsource most of the ref updates to the new "proc-receive" hook.
450
500 - Under review.
501 - A "review club" reviews have been posted for an early few steps.
451 + Needs review.
452
453
504 -* ma/doc-sha-256-is-experimental (2020-08-06) 1 commit
505 - - Documentation: mark `--object-format=sha256` as experimental
454 +* pw/rebase-i-more-options (2020-08-17) 5 commits
455 + - rebase: add --reset-author-date
456 + - rebase -i: support --ignore-date
457 + - rebase -i: support --committer-date-is-author-date
458 + - am: stop exporting GIT_COMMITTER_DATE
459 + - rebase -i: add --ignore-whitespace flag
460
507 - The recent addition of SHA-256 support is marked as experimental in
508 - the documentation.
461 + "git rebase -i" learns a bit more options.
462 +
463 + Waiting for a (hopefully final) review.
464 +
465 +
466 +* rs/more-buffered-io (2020-08-17) 3 commits
467 + - upload-pack: use buffered I/O to talk to rev-list
468 + - midx: use buffered I/O to talk to pack-objects
469 + - connected: use buffered I/O to talk to rev-list
470 +
471 + Use more buffered I/O where we used to call many small write(2)s.
472
473 Will merge to 'next'.
511 - cf. <xmqqmu2y1e0t.fsf@gitster.c.googlers.com>
474
475
514 -* rp/apply-cached-with-i-t-a (2020-08-09) 3 commits
515 - (merged to 'next' on 2020-08-11 at c28eaa6a0e)
516 - + t4140: test apply with i-t-a paths
517 - + apply: make i-t-a entries never match worktree
518 - + apply: allow "new file" patches on i-t-a entries
476 +* en/merge-tests (2020-08-10) 11 commits
477 + (merged to 'next' on 2020-08-13 at eab952350b)
478 + + t6425: be more flexible with rename/delete conflict messages
479 + + t642[23]: be more flexible for add/add conflicts involving pair renames
480 + + t6422, t6426: be more flexible for add/add conflicts involving renames
481 + + t6423: add an explanation about why one of the tests does not pass
482 + + t6416, t6423: clarify some comments and fix some typos
483 + + t6422: fix multiple errors with the mod6 test expectations
484 + + t6423: fix test setup for a couple tests
485 + + t6416, t6422: fix incorrect untracked file count
486 + + t6422: fix bad check against missing file
487 + + t6418: tighten delete/normalize conflict testcase
488 + + Collect merge-related tests to t64xx
489
520 - Recent versions of "git diff-files" shows a diff between the index
521 - and the working tree for "intent-to-add" paths as a "new file"
522 - patch; "git apply --cached" should be able to take "git diff-files"
523 - and should act as an equivalent to "git add" for the path, but the
524 - command failed to do so for such a path.
490 + Updates to "git merge" tests, in preparation for a new merge
491 + strategy backend.
492
493 Will merge to 'master'.
494
495
529 -* jc/noop-with-static-inline (2020-08-06) 1 commit
530 - (merged to 'next' on 2020-08-11 at 8a63ff648e)
531 - + compat-util: type-check parameters of no-op replacement functions
496 +* ds/midx-repack-to-batch-size (2020-08-11) 1 commit
497 + (merged to 'next' on 2020-08-17 at eee94634aa)
498 + + multi-pack-index: repack batches below --batch-size
499
533 - A no-op replacement function implemented as a C preprocessor macro
534 - does not perform as good a job as one implemented as a "static
535 - inline" function in catching errors in parameters; replace the
536 - former with the latter in <git-compat-util.h> header.
500 + The "--batch-size" option of "git multi-pack-index repack" command
501 + is now used to specify that very small packfiles are collected into
502 + one until the total size roughly exceeds it.
503
504 Will merge to 'master'.
505
506
541 -* ps/ref-transaction-hook (2020-08-11) 2 commits
542 - (merged to 'next' on 2020-08-11 at 311c045810)
543 - + t1416: avoid hard-coded sha1 ids
544 - (merged to 'next' on 2020-08-10 at d8ad7cc8f6)
545 - + refs: fix interleaving hook calls with reference-transaction hook
507 +* rs/upload-pack-sigchain-fix (2020-08-11) 1 commit
508 + (merged to 'next' on 2020-08-17 at 997473f309)
509 + + upload-pack: remove superfluous sigchain_pop() call
510
547 - The logic to find the ref transaction hook script attempted to
548 - cache the path to the found hook without realizing that it needed
549 - to keep a copied value, as the API it used returned a transitory
550 - buffer space. This has been corrected.
511 + Code clean-up.
512
513 Will merge to 'master'.
553 - to be followed by a removal of the caching feature, which does not
554 - seem to help even as a negative cache.
514
515
557 -* ss/submodule-summary-in-c (2020-08-12) 4 commits
558 - - submodule: port submodule subcommand 'summary' from shell to C
559 - - t7421: introduce a test script for verifying 'summary' output
560 - - submodule: rename helper functions to avoid ambiguity
561 - - submodule: remove extra line feeds between callback struct and macro
516 +* tb/bloom-improvements (2020-08-11) 14 commits
517 + - builtin/commit-graph.c: introduce '--max-new-filters=<n>'
518 + - commit-graph: rename 'split_commit_graph_opts'
519 + - commit-graph: add large-filters bitmap chunk
520 + - commit-graph.c: sort index into commits list
521 + - bloom/diff: properly short-circuit on max_changes
522 + - bloom: use provided 'struct bloom_filter_settings'
523 + - csum-file.h: introduce 'hashwrite_be64()'
524 + - bloom: split 'get_bloom_filter()' in two
525 + - commit-graph.c: store maximum changed paths
526 + - commit-graph: respect 'commitGraph.readChangedPaths'
527 + - t/helper/test-read-graph.c: prepare repo settings
528 + - commit-graph: pass a 'struct repository *' in more places
529 + - t4216: use an '&&'-chain
530 + - commit-graph: introduce 'get_bloom_filter_settings()'
531
563 - Yet another subcommand of "git submodule" is getting rewritten in C.
532 + Misc Bloom filter improvements.
533
565 - Will merge to 'next'.
534 + Expecting a reroll.
535 + It seems that the review is getting closer to result in another update.
536 + cf. <20200811220503.GC66656@syl.lan>
537
538
568 -* am/ci-wsfix (2020-08-10) 1 commit
569 - - ci: fix inconsistent indentation
539 +* ak/sequencer-fix-find-uniq-abbrev (2020-08-13) 1 commit
540 + (merged to 'next' on 2020-08-17 at 6ce0f1d111)
541 + + rebase -i: fix possibly wrong onto hash in todo
542
571 - Aesthetic fix to a CI configuration file.
543 + Ring buffer with size 4 used for bin-hex translation resulted in a
544 + wrong object name in the sequencer's todo output, which has been
545 + corrected.
546
573 - Waiting for a reroll.
574 - It needs a better log message.
547 + Will merge to 'master'.
548
549
577 -* es/init-no-separate-git-dir-in-bare (2020-08-10) 1 commit
578 - - init: disallow --separate-git-dir with bare repository
550 +* en/dir-nonbare-embedded (2020-08-12) 2 commits
551 + (merged to 'next' on 2020-08-17 at ab180b7fcb)
552 + + dir: avoid prematurely marking nonbare repositories as matches
553 + + t3000: fix some test description typos
554
580 - The purpose of "git init --separate-git-dir" is to initialize a
581 - new project with the repository separate from the working tree,
582 - or, in the case of an existing project, to move the repository
583 - (the .git/ directory) out of the working tree. It does not make
584 - sense to use --separate-git-dir with a bare repository for which
585 - there is no working tree, so disallow its use with bare
586 - repositories.
555 + "ls-files -o" mishandled the top-level directory of another git
556 + working tree that hangs in the current git working tree.
557
588 - Will merge to 'next'.
558 + Will merge to 'master'.
559
560
591 -* es/test-cmp-typocatcher (2020-08-09) 1 commit
592 - (merged to 'next' on 2020-08-11 at ee5557b170)
593 - + test_cmp: diagnose incorrect arguments
561 +* en/sequencer-merge-labels (2020-08-14) 1 commit
562 + (merged to 'next' on 2020-08-17 at 2010e329ed)
563 + + sequencer: avoid garbled merge machinery messages due to commit labels
564
595 - Test framework update.
565 + The commit labels used to explain each side of conflicted hunks
566 + placed by the sequencer machinery have been made more readable by
567 + humans.
568
569 Will merge to 'master'.
570
571
600 -* jk/sideband-error-l10n (2020-08-07) 1 commit
601 - (merged to 'next' on 2020-08-11 at 27dbe05a65)
602 - + sideband: mark "remote error:" prefix for translation
572 +* jb/commit-graph-doc-fix (2020-08-13) 1 commit
573 + (merged to 'next' on 2020-08-17 at d83ee98955)
574 + + docs: commit-graph: fix some whitespace in the diagram
575
604 - Mark error message for i18n.
576 + Docfix.
577
578 Will merge to 'master'.
579
580
609 -* rp/ita-diff-modefix (2020-08-09) 1 commit
610 - - diff-lib: use worktree mode in diffs from i-t-a entries
581 +* jk/blame-coalesce-fix (2020-08-13) 3 commits
582 + (merged to 'next' on 2020-08-17 at 122c701ec7)
583 + + blame: only coalesce lines that are adjacent in result
584 + + t8003: factor setup out of coalesce test
585 + + t8003: check output of coalesced blame
586
612 - "git diff [<tree-ish>] $path" for a $path that is marked with i-t-a
613 - bit was not showing the mode bits from the working tree.
587 + When given more than one target line ranges, "git blame -La,b
588 + -Lc,d" was over-eager to coalesce groups of original lines and
589 + showed incorrect results, which has been corrected.
590
615 - Will merge to 'next'.
591 + Will merge to 'master'.
592
593
618 -* jk/log-fp-implies-m (2020-07-29) 7 commits
619 - (merged to 'next' on 2020-08-03 at 39fefa6b82)
620 - + doc/git-log: clarify handling of merge commit diffs
621 - + doc/git-log: move "-t" into diff-options list
622 - + doc/git-log: drop "-r" diff option
623 - + doc/git-log: move "Diff Formatting" from rev-list-options
624 - + log: enable "-m" automatically with "--first-parent"
625 - + revision: add "--no-diff-merges" option to counteract "-m"
626 - + log: drop "--cc implies -m" logic
627 - (this branch is used by so/log-diff-merges-opt.)
594 +* jk/slimmed-down (2020-08-13) 5 commits
595 + - drop vcs-svn experiment
596 + - make git-fast-import a builtin
597 + - make git-bugreport a builtin
598 + - make credential helpers builtins
599 + - Makefile: drop builtins from MSVC pdb list
600
629 - Originally merged to 'next' on 2020-08-01
601 + Trim an unused binary and turn a bunch of commands into built-in.
602
631 - "git log --first-parent -p" showed patches only for single-parent
632 - commits on the first-parent chain; the "--first-parent" option has
633 - been made to imply "-m". Use "--no-diff-merges" to restore the
634 - previous behaviour to omit patches for merge commits.
603 + In general this is a good thing to do, modulo 'bugreport'.
604
636 - Will merge to 'master' together with 'so/log-diff-merges-opt'
605
606 +* jk/unleak-fixes (2020-08-13) 2 commits
607 + (merged to 'next' on 2020-08-17 at f9bd296b25)
608 + + ls-remote: simplify UNLEAK() usage
609 + + stop calling UNLEAK() before die()
610
639 -* pd/mergetool-nvimdiff (2020-07-29) 2 commits
640 - (merged to 'next' on 2020-08-11 at 947ce686d0)
641 - + mergetools: add support for nvimdiff (neovim) family
642 - + mergetool--lib: improve support for vimdiff-style tool variants
611 + Fix some incorrect UNLEAK() annotations.
612
644 - The existing backends for "git mergetool" based on variants of vim
645 - have been refactored and then support for "nvim" has been added.
613 + Will merge to 'master'.
614 +
615 +
616 +* pb/set-url-docfix (2020-08-13) 1 commit
617 + (merged to 'next' on 2020-08-17 at 19b56c6ebb)
618 + + fetch, pull doc: correct description of '--set-upstream'
619 +
620 + Doc fix.
621
622 Will merge to 'master'.
623
624
650 -* al/bisect-first-parent (2020-08-07) 5 commits
651 - (merged to 'next' on 2020-08-11 at b232f7ca17)
652 - + bisect: combine args passed to find_bisection()
653 - + bisect: introduce first-parent flag
654 - + cmd_bisect__helper: defer parsing no-checkout flag
655 - + rev-list: allow bisect and first-parent flags
656 - + t6030: modernize "git bisect run" tests
625 +* pb/userdiff-fortran-update (2020-08-13) 2 commits
626 + (merged to 'next' on 2020-08-17 at 534af3f35a)
627 + + userdiff: improve Fortran xfuncname regex
628 + + userdiff: add tests for Fortran xfuncname regex
629
658 - "git bisect" learns the "--first-parent" option to find the first
659 - breakage along the first-parent chain.
630 + The regexp to identify the function boundary for FORTRAN programs
631 + has been updated.
632
633 Will merge to 'master'.
634
635
664 -* dd/send-email-config (2020-07-23) 1 commit
665 - (merged to 'next' on 2020-08-11 at aaf9f3a301)
666 - + git-send-email: die if sendmail.* config is set
636 +* rs/preserve-merges-unused-code-removal (2020-08-12) 1 commit
637 + (merged to 'next' on 2020-08-17 at b96393455c)
638 + + rebase: remove unused function reschedule_last_action
639
668 - Stop when "sendmail.*" configuration variables are defined, which
669 - could be a mistaken attempt to define "sendemail.*" variables.
640 + Code clean-up.
641
642 Will merge to 'master'.
643
644
674 -* hn/reftable-prep-part-2 (2020-07-27) 3 commits
675 - (merged to 'next' on 2020-08-11 at 43ac0bc60a)
676 - + Make HEAD a PSEUDOREF rather than PER_WORKTREE.
677 - + Modify pseudo refs through ref backend storage
678 - + t1400: use git rev-parse for testing PSEUDOREF existence
645 +* ss/t7401-modernize (2020-08-12) 4 commits
646 + - t7401: add a NEEDSWORK
647 + - t7401: change indentation for enhanced readability
648 + - t7401: change test_i18ncmp syntax for clarity
649 + - t7401: modernize style
650 +
651 + Test clean-up.
652 +
653 + Expecting a reroll.
654 + cf. <c70f7bb5d1289f5318da7b99e41cf5828a451174.camel@gmail.com>
655 + Almost there, it seems.
656 +
657 +
658 +* ds/maintenance-part-2 (2020-08-18) 8 commits
659 + - maintenance: add incremental-repack auto condition
660 + - maintenance: auto-size incremental-repack batch
661 + - maintenance: add incremental-repack task
662 + - midx: use start_delayed_progress()
663 + - midx: enable core.multiPackIndex by default
664 + - maintenance: create auto condition for loose-objects
665 + - maintenance: add loose-objects task
666 + - maintenance: add prefetch task
667 + (this branch uses ds/maintenance-part-1 and jc/no-update-fetch-head; is tangled with jt/lazy-fetch.)
668 +
669 + "git maintenance", an extended big brother of "git gc", continues
670 + to evolve.
671 +
672 +
673 +* ma/doc-sha-256-is-experimental (2020-08-17) 1 commit
674 + - Documentation: mark `--object-format=sha256` as experimental
675 +
676 + The recent addition of SHA-256 support is marked as experimental in
677 + the documentation.
678 +
679 + Will merge to 'next'.
680
680 - Further preliminary change to refs API.
681 +
682 +* ss/submodule-summary-in-c (2020-08-12) 4 commits
683 + (merged to 'next' on 2020-08-17 at 9bc352cb70)
684 + + submodule: port submodule subcommand 'summary' from shell to C
685 + + t7421: introduce a test script for verifying 'summary' output
686 + + submodule: rename helper functions to avoid ambiguity
687 + + submodule: remove extra line feeds between callback struct and macro
688 +
689 + Yet another subcommand of "git submodule" is getting rewritten in C.
690
691 Will merge to 'master'.
692
693
685 -* ds/maintenance (2020-08-06) 11 commits
694 +* am/ci-wsfix (2020-08-10) 1 commit
695 + - ci: fix inconsistent indentation
696 +
697 + Aesthetic fix to a CI configuration file.
698 +
699 + Waiting for a reroll.
700 + It needs a better log message.
701 +
702 +
703 +* es/init-no-separate-git-dir-in-bare (2020-08-10) 1 commit
704 + (merged to 'next' on 2020-08-17 at 80498c8659)
705 + + init: disallow --separate-git-dir with bare repository
706 +
707 + The purpose of "git init --separate-git-dir" is to initialize a
708 + new project with the repository separate from the working tree,
709 + or, in the case of an existing project, to move the repository
710 + (the .git/ directory) out of the working tree. It does not make
711 + sense to use --separate-git-dir with a bare repository for which
712 + there is no working tree, so disallow its use with bare
713 + repositories.
714 +
715 + Will merge to 'master'.
716 +
717 +
718 +* rp/ita-diff-modefix (2020-08-09) 1 commit
719 + (merged to 'next' on 2020-08-17 at c03a9cfb6e)
720 + + diff-lib: use worktree mode in diffs from i-t-a entries
721 +
722 + "git diff [<tree-ish>] $path" for a $path that is marked with i-t-a
723 + bit was not showing the mode bits from the working tree.
724 +
725 + Will merge to 'master'.
726 +
727 +
728 +* ds/maintenance-part-1 (2020-08-18) 11 commits
729 - maintenance: add trace2 regions for task execution
730 - maintenance: add auto condition for commit-graph task
731 - maintenance: use pointers to check --auto
@@ -694,25 +737,13 @@ repositories listed at
737 - maintenance: replace run_auto_gc()
738 - maintenance: add --quiet option
739 - maintenance: create basic maintenance runner
697 - (this branch is used by ds/maintenance-part-2.)
740 + (this branch is used by ds/maintenance-part-2; uses jc/no-update-fetch-head; is tangled with jt/lazy-fetch.)
741
742 A "git gc"'s big brother has been introduced to take care of more
743 repository maintenance tasks, not limited to the object database
744 cleaning.
745
746
704 -* mt/hash-to-hex-thread-safety (2020-06-26) 2 commits
705 - - hex: make hash_to_hex_algop() and friends thread-safe
706 - - compat/win32/pthread: add pthread_once()
707 -
708 - hash_to_hex() used a set of rotating static buffers, which was not
709 - safe to use in a threaded environment. This has been made safer by
710 - using thread-local storage.
711 -
712 - Expecting a reroll.
713 - cf. <CAHd-oW6A2aBHg80R9kyifvF7thwzam5EYYoN9d2TaBcHJrcKWw@mail.gmail.com>
714 -
715 -
747 * es/config-hooks (2020-07-30) 6 commits
748 - hook: add 'run' subcommand
749 - parse-options: parse into argv_array
@@ -728,3 +759,26 @@ repositories listed at
759 on top of it. Are there unresolved issues, or does the topic need
760 a round of detailed review?
761 cf. <xmqqmu3i9kvg.fsf@gitster.c.googlers.com>
762 +
763 +--------------------------------------------------
764 +[Discarded]
765 +
766 +* rs/fast-export-anon-simplify (2020-08-13) 1 commit
767 + . fast-export: factor out print_oid()
768 +
769 + Code simplification.
770 +
771 + Retracted.
772 + cf. <6e2d4472-8293-4f10-0ba6-82ae83f7a465@web.de>
773 +
774 +
775 +* mt/hash-to-hex-thread-safety (2020-06-26) 2 commits
776 + . hex: make hash_to_hex_algop() and friends thread-safe
777 + . compat/win32/pthread: add pthread_once()
778 +
779 + hash_to_hex() used a set of rotating static buffers, which was not
780 + safe to use in a threaded environment. This has been made safer by
781 + using thread-local storage.
782 +
783 + Retracted.
784 + cf. <CAHd-oW7Wd8oSaMhPFeRcEeKTJ-k_hC7b6e28efhXT5LFu1E_Uw@mail.gmail.com>