1Git 2.15 Release Notes 2====================== 3 4Backward compatibility notes and other notable changes. 5 6 * Use of an empty string as a pathspec element that is used for 7 'everything matches' is still warned and Git asks users to use a 8 more explicit '.' for that instead. The hope is that existing 9 users will not mind this change, and eventually the warning can be 10 turned into a hard error, upgrading the deprecation into removal of 11 this (mis)feature. That is now scheduled to happen in the upcoming 12 release. 13 14 * Git now avoids blindly falling back to ".git" when the setup 15 sequence said we are _not_ in Git repository. A corner case that 16 happens to work right now may be broken by a call to die("BUG"). 17 We've tried hard to locate such cases and fixed them, but there 18 might still be cases that need to be addressed--bug reports are 19 greatly appreciated. 20 21 * "branch --set-upstream" that has been deprecated in Git 1.8 has 22 finally been retired. 23 24 25Updates since v2.14 26------------------- 27 28UI, Workflows & Features 29 30 * An example that is now obsolete has been removed from a sample hook, 31 and an old example in it that added a sign-off manually has been 32 improved to use the interpret-trailers command. 33 34 * The advice message given when "git rebase" stops for conflicting 35 changes has been improved. 36 37 * The "rerere-train" script (in contrib/) learned the "--overwrite" 38 option to allow overwriting existing recorded resolutions. 39 40 * "git contacts" (in contrib/) now lists the address on the 41 "Reported-by:" trailer to its output, in addition to those on 42 S-o-b: and other trailers, to make it easier to notify (and thank) 43 the original bug reporter. 44 45 * "git rebase", especially when it is run by mistake and ends up 46 trying to replay many changes, spent long time in silence. The 47 command has been taught to show progress report when it spends 48 long time preparing these many changes to replay (which would give 49 the user a chance to abort with ^C). 50 51 * "git merge" learned a "--signoff" option to add the Signed-off-by: 52 trailer with the committer's name. 53 54 * "git diff" learned to optionally paint new lines that are the same 55 as deleted lines elsewhere differently from genuinely new lines. 56 57 * "git interpret-trailers" learned to take the trailer specifications 58 from the command line that overrides the configured values. 59 60 * "git interpret-trailers" has been taught a "--parse" and a few 61 other options to make it easier for scripts to grab existing 62 trailer lines from a commit log message. 63 64 * "gitweb" shows a link to visit the 'raw' contents of blbos in the 65 history overview page. 66 67 * "[gc] rerereResolved = 5.days" used to be invalid, as the variable 68 is defined to take an integer counting the number of days. It now 69 is allowed. 70 71 * The code to acquire a lock on a reference (e.g. while accepting a 72 push from a client) used to immediately fail when the reference is 73 already locked---now it waits for a very short while and retries, 74 which can make it succeed if the lock holder was holding it during 75 a read-only operation. 76 77 * "branch --set-upstream" that has been deprecated in Git 1.8 has 78 finally been retired. 79 80 * The codepath to call external process filter for smudge/clean 81 operation learned to show the progress meter. 82 83 * "git rev-parse" learned "--is-shallow-repository", that is to be 84 used in a way similar to existing "--is-bare-repository" and 85 friends. 86 87 * "git describe --match <pattern>" has been taught to play well with 88 the "--all" option. 89 90 91Performance, Internal Implementation, Development Support etc. 92 93 * Conversion from uchar[20] to struct object_id continues. 94 95 * Start using selected c99 constructs in small, stable and 96 essentialpart of the system to catch people who care about 97 older compilers that do not grok them. 98 99 * The filter-process interface learned to allow a process with long 100 latency give a "delayed" response. 101 102 * Many uses of comparision callback function the hashmap API uses 103 cast the callback function type when registering it to 104 hashmap_init(), which defeats the compile time type checking when 105 the callback interface changes (e.g. gaining more parameters). 106 The callback implementations have been updated to take "void *" 107 pointers and cast them to the type they expect instead. 108 109 * Because recent Git for Windows do come with a real msgfmt, the 110 build procedure for git-gui has been updated to use it instead of a 111 hand-rolled substitute. 112 113 * "git grep --recurse-submodules" has been reworked to give a more 114 consistent output across submodule boundary (and do its thing 115 without having to fork a separate process). 116 117 * A helper function to read a single whole line into strbuf 118 mistakenly triggered OOM error at EOF under certain conditions, 119 which has been fixed. 120 (merge 642956cf45 rs/strbuf-getwholeline-fix later to maint). 121 122 * The "ref-store" code reorganization continues. 123 124 * "git commit" used to discard the index and re-read from the filesystem 125 just in case the pre-commit hook has updated it in the middle; this 126 has been optimized out when we know we do not run the pre-commit hook. 127 (merge 680ee550d7 kw/commit-keep-index-when-pre-commit-is-not-run later to maint). 128 129 * Updates to the HTTP layer we made recently unconditionally used 130 features of libCurl without checking the existence of them, causing 131 compilation errors, which has been fixed. Also migrate the code to 132 check feature macros, not version numbers, to cope better with 133 libCurl that vendor ships with backported features. 134 135 * The API to start showing progress meter after a short delay has 136 been simplified. 137 (merge 8aade107dd jc/simplify-progress later to maint). 138 139 * Code clean-up to avoid mixing values read from the .gitmodules file 140 and values read from the .git/config file. 141 142 * We used to spend more than necessary cycles allocating and freeing 143 piece of memory while writing each index entry out. This has been 144 optimized. 145 146 * Platforms that ship with a separate sha1 with collision detection 147 library can link to it instead of using the copy we ship as part of 148 our source tree. 149 150 * Code around "notes" have been cleaned up. 151 (merge 3964281524 mh/notes-cleanup later to maint). 152 153 * The long-standing rule that an in-core lockfile instance, once it 154 is used, must not be freed, has been lifted and the lockfile and 155 tempfile APIs have been updated to reduce the chance of programming 156 errors. 157 158 * Our hashmap implementation in hashmap.[ch] is not thread-safe when 159 adding a new item needs to expand the hashtable by rehashing; add 160 an API to disable the automatic rehashing to work it around. 161 162 * Many of our programs consider that it is OK to release dynamic 163 storage that is used throughout the life of the program by simply 164 exiting, but this makes it harder to leak detection tools to avoid 165 reporting false positives. Plug many existing leaks and introduce 166 a mechanism for developers to mark that the region of memory 167 pointed by a pointer is not lost/leaking to help these tools. 168 169 * As "git commit" to conclude a conflicted "git merge" honors the 170 commit-msg hook, "git merge" that records a merge commit that 171 cleanly auto-merges should, but it didn't. 172 173 * The codepath for "git merge-recursive" has been cleaned up. 174 175 * Many leaks of strbuf have been fixed. 176 177 * "git imap-send" has our own implementation of the protocol and also 178 can use more recent libCurl with the imap protocol support. Update 179 the latter so that it can use the credential subsystem, and then 180 make it the default option to use, so that we can eventually 181 deprecate and remove the former. 182 183 * "make style" runs git-clang-format to help developers by pointing 184 out coding style issues. 185 186 * A test to demonstrate "git mv" failing to adjust nested submodules 187 has been added. 188 (merge c514167df2 hv/mv-nested-submodules-test later to maint). 189 190 * On Cygwin, "ulimit -s" does not report failure but it does not work 191 at all, which causes an unexpected success of some tests that 192 expect failures under a limited stack situation. This has been 193 fixed. 194 195 * Many codepaths have been updated to squelch -Wimplicit-fallthrough 196 warnings from Gcc 7 (which is a good code hygiene). 197 198 * Add a helper for DLL loading in anticipation for its need in a 199 future topic RSN. 200 201 * "git status --ignored", when noticing that a directory without any 202 tracked path is ignored, still enumerated all the ignored paths in 203 the directory, which is unnecessary. The codepath has been 204 optimized to avoid this overhead. 205 206Also contains various documentation updates and code clean-ups. 207 208 209Fixes since v2.14 210----------------- 211 212 * "%C(color name)" in the pretty print format always produced ANSI 213 color escape codes, which was an early design mistake. They now 214 honor the configuration (e.g. "color.ui = never") and also tty-ness 215 of the output medium. 216 217 * The http.{sslkey,sslCert} configuration variables are to be 218 interpreted as a pathname that honors "~[username]/" prefix, but 219 weren't, which has been fixed. 220 221 * Numerous bugs in walking of reflogs via "log -g" and friends have 222 been fixed. 223 224 * "git commit" when seeing an totally empty message said "you did not 225 edit the message", which is clearly wrong. The message has been 226 corrected. 227 228 * When a directory is not readable, "gitweb" fails to build the 229 project list. Work this around by skipping such a directory. 230 231 * Some versions of GnuPG fails to kill gpg-agent it auto-spawned 232 and such a left-over agent can interfere with a test. Work it 233 around by attempting to kill one before starting a new test. 234 235 * A recently added test for the "credential-cache" helper revealed 236 that EOF detection done around the time the connection to the cache 237 daemon is torn down were flaky. This was fixed by reacting to 238 ECONNRESET and behaving as if we got an EOF. 239 240 * "git log --tag=no-such-tag" showed log starting from HEAD, which 241 has been fixed---it now shows nothing. 242 243 * The "tag.pager" configuration variable was useless for those who 244 actually create tag objects, as it interfered with the use of an 245 editor. A new mechanism has been introduced for commands to enable 246 pager depending on what operation is being carried out to fix this, 247 and then "git tag -l" is made to run pager by default. 248 249 * "git push --recurse-submodules $there HEAD:$target" was not 250 propagated down to the submodules, but now it is. 251 252 * Commands like "git rebase" accepted the --rerere-autoupdate option 253 from the command line, but did not always use it. This has been 254 fixed. 255 256 * "git clone --recurse-submodules --quiet" did not pass the quiet 257 option down to submodules. 258 259 * Test portability fix for OBSD. 260 261 * Portability fix for OBSD. 262 263 * "git am -s" has been taught that some input may end with a trailer 264 block that is not Signed-off-by: and it should refrain from adding 265 an extra blank line before adding a new sign-off in such a case. 266 267 * "git svn" used with "--localtime" option did not compute the tz 268 offset for the timestamp in question and instead always used the 269 current time, which has been corrected. 270 271 * Memory leak in an error codepath has been plugged. 272 273 * "git stash -u" used the contents of the committed version of the 274 ".gitignore" file to decide which paths are ignored, even when the 275 file has local changes. The command has been taught to instead use 276 the locally modified contents. 277 278 * bash 4.4 or newer gave a warning on NUL byte in command 279 substitution done in "git stash"; this has been squelched. 280 281 * "git grep -L" and "git grep --quiet -L" reported different exit 282 codes; this has been corrected. 283 284 * When handshake with a subprocess filter notices that the process 285 asked for an unknown capability, Git did not report what program 286 the offending subprocess was running. This has been corrected. 287 288 * "git apply" that is used as a better "patch -p1" failed to apply a 289 taken from a file with CRLF line endings to a file with CRLF line 290 endings. The root cause was because it misused convert_to_git() 291 that tried to do "safe-crlf" processing by looking at the index 292 entry at the same path, which is a nonsense---in that mode, "apply" 293 is not working on the data in (or derived from) the index at all. 294 This has been fixed. 295 296 * Killing "git merge --edit" before the editor returns control left 297 the repository in a state with MERGE_MSG but without MERGE_HEAD, 298 which incorrectly tells the subsequent "git commit" that there was 299 a squash merge in progress. This has been fixed. 300 301 * "git archive" did not work well with pathspecs and the 302 export-ignore attribute. 303 304 * In addition to "cc: <a@dd.re.ss> # cruft", "cc: a@dd.re.ss # cruft" 305 was taught to "git send-email" as a valid way to tell it that it 306 needs to also send a carbon copy to <a@dd.re.ss> in the trailer 307 section. 308 (merge cc90750677 mm/send-email-cc-cruft later to maint). 309 310 * "git branch -M a b" while on a branch that is completely unrelated 311 to either branch a or branch b misbehaved when multiple worktree 312 was in use. This has been fixed. 313 (merge 31824d180d nd/worktree-kill-parse-ref later to maint). 314 315 * "git gc" and friends when multiple worktrees are used off of a 316 single repository did not consider the index and per-worktree refs 317 of other worktrees as the root for reachability traversal, making 318 objects that are in use only in other worktrees to be subject to 319 garbage collection. 320 321 * A regression to "gitk --bisect" by a recent update has been fixed. 322 (merge 1d0538e486 mh/packed-ref-store-prep later to maint). 323 324 * "git -c submodule.recurse=yes pull" did not work as if the 325 "--recurse-submodules" option was given from the command line. 326 This has been corrected. 327 328 * Unlike "git commit-tree < file", "git commit-tree -F file" did not 329 pass the contents of the file verbatim and instead completed an 330 incomplete line at the end, if exists. The latter has been updated 331 to match the behaviour of the former. 332 (merge c818e74332 rk/commit-tree-make-F-verbatim later to maint). 333 334 * Many codepaths did not diagnose write failures correctly when disks 335 go full, due to their misuse of write_in_full() helper function, 336 which have been corrected. 337 (merge f48ecd38cb jk/write-in-full-fix later to maint). 338 339 * "git help co" now says "co is aliased to ...", not "git co is". 340 (merge b3a8076e0d ks/help-alias-label later to maint). 341 342 * "git archive", especially when used with pathspec, stored an empty 343 directory in its output, even though Git itself never does so. 344 This has been fixed. 345 (merge 4318094047 rs/archive-excluded-directory later to maint). 346 347 * API error-proofing which happens to also squelch warnings from GCC. 348 (merge c788c54cde tg/refs-allowed-flags later to maint). 349 350 * The explanation of the cut-line in the commit log editor has been 351 slightly tweaked. 352 (merge 8c4b1a3593 ks/commit-do-not-touch-cut-line later to maint). 353 354 * "git gc" tries to avoid running two instances at the same time by 355 reading and writing pid/host from and to a lock file; it used to 356 use an incorrect fscanf() format when reading, which has been 357 corrected. 358 (merge afe2fab72c aw/gc-lockfile-fscanf-fix later to maint). 359 360 * The scripts to drive TravisCI has been reorganized and then an 361 optimization to avoid spending cycles on a branch whose tip is 362 tagged has been implemented. 363 (merge 8376eb4a8f ls/travis-scriptify later to maint). 364 365 * The test linter has been taught that we do not like "echo -e". 366 (merge 1a6d46895d tb/test-lint-echo-e later to maint). 367 368 * Code cmp.std.c nitpick. 369 (merge ac7da78ede mh/for-each-string-list-item-empty-fix later to maint). 370 371 * A regression fix for 2.11 that made the code to read the list of 372 alternate object stores overrun the end of the string. 373 (merge f0f7bebef7 jk/info-alternates-fix later to maint). 374 375 * "git describe --match" learned to take multiple patterns in v2.13 376 series, but the feature ignored the patterns after the first one 377 and did not work at all. This has been fixed. 378 (merge da769d2986 jk/describe-omit-some-refs later to maint). 379 380 * "git filter-branch" cannot reproduce a history with a tag without 381 the tagger field, which only ancient versions of Git allowed to be 382 created. This has been corrected. 383 (merge b2c1ca6b4b ic/fix-filter-branch-to-handle-tag-without-tagger later to maint). 384 385 * "git cat-file --textconv" started segfaulting recently, which 386 has been corrected. 387 (merge cc0ea7c9e5 jk/diff-blob later to maint). 388 389 * The built-in pattern to detect the "function header" for HTML did 390 not match <H1>..<H6> elements without any attributes, which has 391 been fixed. 392 (merge 9c03caca2c ik/userdiff-html-h-element-fix later to maint). 393 394 * "git mailinfo" was loose in decoding quoted printable and produced 395 garbage when the two letters after the equal sign are not 396 hexadecimal. This has been fixed. 397 (merge c8cf423eab rs/mailinfo-qp-decode-fix later to maint). 398 399 * The machinery to create xdelta used in pack files received the 400 sizes of the data in size_t, but lost the higher bits of them by 401 storing them in "unsigned int" during the computation, which is 402 fixed. 403 404 * The delta format used in the packfile cannot reference data at 405 offset larger than what can be expressed in 4-byte, but the 406 generator for the data failed to make sure the offset does not 407 overflow. This has been corrected. 408 409 * The documentation for '-X<option>' for merges was misleadingly 410 written to suggest that "-s theirs" exists, which is not the case. 411 (merge c25d98b2a7 jc/merge-x-theirs-docfix later to maint). 412 413 * "git fast-export" with -M/-C option issued "copy" instruction on a 414 path that is simultaneously modified, which was incorrect. 415 (merge b3e8ca89cf jt/fast-export-copy-modify-fix later to maint). 416 417 * Many codepaths have been updated to squelch -Wsign-compare 418 warnings. 419 (merge 071bcaab64 rj/no-sign-compare later to maint). 420 421 * Memory leaks in various codepaths have been plugged. 422 (merge 4d01a7fa65 ma/leakplugs later to maint). 423 424 * Other minor doc, test and build updates and code cleanups. 425 (merge f094b89a4d ma/parse-maybe-bool later to maint). 426 (merge 39b00fa4d4 jk/drop-sha1-entry-pos later to maint). 427 (merge 6cdf8a7929 ma/ts-cleanups later to maint). 428 (merge 7560f547e6 ma/up-to-date later to maint). 429 (merge 0db3dc75f3 rs/apply-epoch later to maint). 430 (merge 74f1bd912b dw/diff-highlight-makefile-fix later to maint). 431 (merge f991761eb8 jk/config-lockfile-leak-fix later to maint). 432 (merge 150efef1e7 ma/pkt-line-leakfix later to maint). 433 (merge 5554451de6 mg/timestamp-t-fix later to maint). 434 (merge 276d0e35c0 ma/split-symref-update-fix later to maint). 435 (merge 3bc4b8f7c7 bb/doc-eol-dirty later to maint). 436 (merge c1bb33c99c jk/system-path-cleanup later to maint). 437 (merge ab46e6fc72 cc/subprocess-handshake-missing-capabilities later to maint). 438 (merge f7a32dd97f kd/doc-for-each-ref later to maint). 439 (merge be94568bc7 ez/doc-duplicated-words-fix later to maint). 440 (merge 01e4be6c3d ks/test-readme-phrasofix later to maint). 441 (merge 217bb56d4f hn/typofix later to maint). 442 (merge c08fd6388c jk/doc-read-tree-table-asciidoctor-fix later to maint). 443 (merge c3342b362e ks/doc-use-camelcase-for-config-name later to maint).