Junio C Hamano | 3c3d3f6 | 2015-10-16 14:40:04 -0700 | [diff] [blame] | 1 | Git v2.6.2 Release Notes |
| 2 | ======================== |
| 3 | |
| 4 | Fixes since v2.6.1 |
| 5 | ------------------ |
| 6 | |
| 7 | * There were some classes of errors that "git fsck" diagnosed to its |
| 8 | standard error that did not cause it to exit with non-zero status. |
| 9 | |
| 10 | * A test script for the HTTP service had a timing dependent bug, |
| 11 | which was fixed. |
| 12 | |
| 13 | * Performance-measurement tests did not work without an installed Git. |
| 14 | |
| 15 | * On a case insensitive filesystems, setting GIT_WORK_TREE variable |
| 16 | using a random cases that does not agree with what the filesystem |
| 17 | thinks confused Git that it wasn't inside the working tree. |
| 18 | |
| 19 | * When "git am" was rewritten as a built-in, it stopped paying |
| 20 | attention to user.signingkey, which was fixed. |
| 21 | |
| 22 | * After "git checkout --detach", "git status" reported a fairly |
| 23 | useless "HEAD detached at HEAD", instead of saying at which exact |
| 24 | commit. |
| 25 | |
| 26 | * "git rebase -i" had a minor regression recently, which stopped |
| 27 | considering a line that begins with an indented '#' in its insn |
| 28 | sheet not a comment, which is now fixed. |
| 29 | |
| 30 | * Description of the "log.follow" configuration variable in "git log" |
| 31 | documentation is now also copied to "git config" documentation. |
| 32 | |
| 33 | * Allocation related functions and stdio are unsafe things to call |
| 34 | inside a signal handler, and indeed killing the pager can cause |
| 35 | glibc to deadlock waiting on allocation mutex as our signal handler |
| 36 | tries to free() some data structures in wait_for_pager(). Reduce |
| 37 | these unsafe calls. |
| 38 | |
| 39 | * The way how --ref/--notes to specify the notes tree reference are |
| 40 | DWIMmed was not clearly documented. |
| 41 | |
| 42 | * Customization to change the behaviour with "make -w" and "make -s" |
| 43 | in our Makefile was broken when they were used together. |
| 44 | |
| 45 | * The Makefile always runs the library archiver with hardcoded "crs" |
| 46 | options, which was inconvenient for exotic platforms on which |
| 47 | people want to use programs with totally different set of command |
| 48 | line options. |
| 49 | |
| 50 | * The ssh transport, just like any other transport over the network, |
| 51 | did not clear GIT_* environment variables, but it is possible to |
| 52 | use SendEnv and AcceptEnv to leak them to the remote invocation of |
| 53 | Git, which is not a good idea at all. Explicitly clear them just |
| 54 | like we do for the local transport. |
| 55 | |
| 56 | * "git blame --first-parent v1.0..v2.0" was not rejected but did not |
| 57 | limit the blame to commits on the first parent chain. |
| 58 | |
| 59 | * Very small number of options take a parameter that is optional |
| 60 | (which is not a great UI element as they can only appear at the end |
| 61 | of the command line). Add notice to documentation of each and |
| 62 | every one of them. |
| 63 | |
| 64 | Also contains typofixes, documentation updates and trivial code |
| 65 | clean-ups. |