src


Log

Author Commit Date CI Message
Vicent Marti 967d3f2e 2014-03-05T21:06:59 Merge pull request #2163 from ethomson/nobackend_odb_write ODB writing fails gracefully when unsupported
Edward Thomson 7bd2f401 2014-03-05T11:35:47 ODB writing fails gracefully when unsupported If no ODB backends support writing, we should fail gracefully.
Brian Lambert 06a8f5c3 2014-03-05T00:00:41 Fixed missing error check on call to git_remote_download in git_remote_fetch. Moved error check to statement following git_remote_disconnect so that the disconnect happens regardless of the result of the download call.
Vicent Marti 66d9e046 2014-03-04T21:01:39 Merge pull request #2157 from libgit2/cmn/write-object-mem pack-objects: free memory safely
Carlos Martín Nieto a14aa1e7 2014-03-04T20:09:17 pack-objects: free memory safely A few fixes have accumulated in this area which have made the freeing of data a bit muddy. Make sure to free the data only when needed and once. When we are going to write a delta to the packfile, we need to free the data, otherwise leave it. The current version of the code mixes up the checks for po->data and po->delta_data.
Jacques Germishuys 4636ca93 2014-03-04T12:22:27 Remove ignored files from the working directory if they were stashed
Vicent Marti 0511b15c 2014-03-03T15:05:26 Merge pull request #2141 from ravselj/development BUGFIX - Fetching twice from the same remote causes a segfault
Vicent Marti ebb3c506 2014-03-03T12:40:25 features: Rename `_HAS_` to `_FEATURE_`
Vicent Marti c9f5298b 2014-03-03T12:09:17 caps: Rename to features to avoid confusion
Miha b5212858 2014-03-03T11:40:22 Merge remote-tracking branch 'remotes/upstream/development' into development
Ben Straub 494be429 2014-03-02T09:00:00 Merge pull request #2144 from linquize/branch-f-current Do not allow git_branch_create() to force update branch
Russell Belfer 6789b7a7 2014-02-27T14:13:22 Add buffer to buffer diff and patch APIs This adds `git_diff_buffers` and `git_patch_from_buffers`. This also includes a bunch of internal refactoring to increase the shared code between these functions and the blob-to-blob and blob-to-buffer APIs, as well as some higher level assert helpers in the tests to also remove redundancy.
Linquize 59b1dbcd 2014-02-27T23:56:25 Do not allow git_branch_create() to force update branch
Ben Straub 1574d388 2014-02-26T16:58:20 Merge pull request #2137 from jru/blame-first-parent Blame first-parent history
Juan Rubén 0276f0f5 2014-02-26T19:22:19 Reset num_parents to 1 only for merge commits Also, correct test case to account for the boundary flag
Miha 3536c168 2014-02-25T14:57:47 - need_pack was not set to 0 when local fetch was already present causing negotiate_fetch access violation
Miha 300f4412 2014-02-25T11:56:11 - BUGFIX #2133 (@fourplusone) in smart_protocol.c - added MSVC cmake definitions to disable warnings - general.c is rewritten so it is ansi-c compatible and compiles ok on microsoft windows - some MSVC reported warning fixes
Vicent Marti cb81c3a7 2014-02-25T10:46:41 Merge pull request #2138 from ethomson/sysdir Move system directory cache out of utils
Edward Thomson 4f46a98b 2014-02-24T23:32:25 Remove now-duplicated stdarg.h include
Edward Thomson ab0af298 2014-02-24T21:56:08 Include stdarg.h for the va_copy test
Edward Thomson 83634d38 2014-02-24T17:43:10 Move system directory cache out of utils
Ben Straub c8893d1f 2014-02-24T13:39:04 Use a portable cast
Ben Straub b1f2c2e2 2014-02-22T10:18:42 Prevent icc warning
Ben Straub 8ac7da79 2014-02-22T10:15:11 Avoid casting warning
Juan Rubén c7c83394 2014-02-21T00:22:07 Add option to limit blame to first parent
Russell Belfer 72556cc6 2014-02-20T14:27:10 Address PR comments * Make GIT_INLINE an internal definition so it cannot be used in public headers * Fix language in CONTRIBUTING * Make index caps API use signed instead of unsigned values
Russell Belfer 978a4ed5 2014-02-20T11:00:31 Make git_oid_equal a non-inline API
Ben Straub 864535cf 2014-02-18T14:07:42 Readability
Ben Straub 9bda5fb8 2014-02-18T14:05:30 Improve error propagation in shallow call
brian m. carlson 0197d410 2014-02-15T23:09:01 Check for EWOULDBLOCK as well as EAGAIN on write. On some systems, notably HP PA-RISC systems running Linux or HP-UX, EWOULDBLOCK and EAGAIN are not the same value. POSIX (and these OSes) allow EWOULDBLOCK to occur on write(2) (and send(2), etc.), so check explicitly for this case as well as EAGAIN by defining and using a macro GIT_ISBLOCKED that considers both. The macro is necessary because MSYS does not provide EWOULDBLOCK and compilation fails if an attempt is made to use it unconditionally. On most systems, where the two values are the same, the compiler will simply optimize this check out and it will have no effect.
Edward Thomson b033f3a3 2014-02-11T16:52:08 Never convert CRLF->LF Core git performs no conversion on systems that use LF, emulate that.
Russell Belfer 5d195cf7 2014-02-11T15:56:04 Merge pull request #2110 from libgit2/ed/crlf_input Handle `core.autocrlf=input` when checking out
Carlos Martín Nieto 15284a2c 2014-02-10T14:52:28 refs: move current_id before the reflog parameters Keep the reflog parameters as the last two, as they're the optional parameters.
Carlos Martín Nieto 77ad6754 2014-02-10T14:38:01 refs: conditional wording fixups This addresses arrbee's concerns about wording in the conditional reference udpate functions.
Edward Thomson 66b2626c 2014-02-09T13:43:56 core.autocrlf=input w/ text=auto attr to workdir
Russell Belfer 80c29fe9 2014-01-17T10:45:11 Add git_commit_amend API This adds an API to amend an existing commit, basically a shorthand for creating a new commit filling in missing parameters from the values of an existing commit. As part of this, I also added a new "sys" API to create a commit using a callback to get the parents. This allowed me to rewrite all the other commit creation APIs so that temporary allocations are no longer needed.
Russell Belfer 2d929194 2014-02-07T16:14:17 Merge pull request #2099 from libgit2/bs/more-reflog-stuff More reflogness
Russell Belfer 57c47af1 2014-02-07T16:05:19 Merge pull request #2042 from libgit2/cmn/conditional-ref refs: conditional ref updates
Russell Belfer 3158e2fe 2014-02-07T15:24:39 Fix some Windows warnings This fixes a number of warnings with the Windows 64-bit build including a test failure in test_repo_message__message where an invalid pointer to a git_buf was being used.
Vicent Marti c4ee3b54 2014-02-07T18:32:06 Merge pull request #2100 from libgit2/rb/update-pqueue Replace priority queue code with implementation from hashsig
Ben Straub db55bb73 2014-02-06T11:18:10 Correct default reflog message for git_remote_fetch
Edward Thomson 55d257e7 2014-02-05T15:03:00 Remove unused utf8 -> utf16 conversion code
Ben Straub 0de2c4e3 2014-02-05T13:15:57 Merge remote-tracking branch 'libgit2/development' into bs/more-reflog-stuff
Russell Belfer df70de07 2014-02-05T10:25:50 Merge pull request #2094 from libgit2/cmn/push-non-commit Add flexibility to the revwalk API
Carlos Martín Nieto 24f3024f 2014-02-05T14:31:01 Split p_strlen into its own header We need this from util.h and posix.h, but the latter includes common.h which includes util.h, which means p_strlen is not defined by the time we get to git__strndup(). Split the definition on p_strlen() off into its own header so we can use it in util.h.
Carlos Martín Nieto 1e6f0ac4 2014-02-05T13:49:51 utils: don't reimplement strnlen The standard library provides a very nice strnlen function, which knows to use SSE, let's not reimplement it ourselves.
Carlos Martín Nieto a6563619 2014-02-05T13:01:54 commit: faster parsing The current code issues a lot of strncmp() calls in order to check for the end of the header, simply in order to copy it and start going through it again. These are a lot of calls for something we can check as we go along. Knowing the amount of parents beforehand to reduce allocations in extreme cases does not make up for them. Instead start parsing immediately and check for the double-newline after each header field, leaving the raw_header allocation for the end, which lets us go through the header once and reduces the amount of strncmp() calls significantly. In unscientific testing, this has reduced a shortlog-like usage (walking though the whole history of a branch and extracting data from the commits) of git.git from ~830ms to ~700ms and makes the time we spend in strncmp() negligible.
Carlos Martín Nieto c74077d1 2014-02-02T12:08:18 revparse: do look at all refs when matching text Now that we no longer fail to push non-commits on a glob, let's search on all refs when we rev-parse syntax asks us to match text.
Carlos Martín Nieto af817202 2014-02-01T15:29:16 revwalk: remove usage of foreach
Carlos Martín Nieto d465e4e9 2014-02-01T15:19:13 revwalk: ignore wrong object type in glob pushes Pushing a whole namespace can cause us to attempt to push non-committish objects. Catch this situation and special-case it for ignoring this.
Carlos Martín Nieto f61272e0 2014-02-01T12:51:36 revwalk: accept committish objects Let the user push committish objects and peel them to figure out which commit to push to our queue. This is for convenience and for allowing uses of git_revwalk_push_glob(w, "tags") with annotated tags.
Carlos Martín Nieto 5367ec4b 2014-02-05T12:02:52 refs: add an unconditional delete Add it under the git_reference_remove() name, letting the user pass the repo and name, analogous to unconditional setting/creation.
Carlos Martín Nieto b7ae71ec 2014-02-05T11:47:33 refs: catch cases where the ref type has changed If the type of the on-disk reference has changed, the old value comparison should fail.
Carlos Martín Nieto f44fd59e 2014-02-05T11:21:14 refs: check the ref's old value when deleting Recognize when the reference has changed since we loaded it.
Carlos Martín Nieto 7ee8c7e6 2014-02-05T11:07:34 refs: placeholder conditional delete We don't actually pass the old value yet.
Carlos Martín Nieto f8621dde 2014-02-05T10:42:42 refs: factor out old value comparison We will reuse this later for deletion.
Carlos Martín Nieto 878fb66f 2014-02-05T10:19:17 refs: bring conditional symbolic updates to the frontend Bring the race detection goodness to symbolic references as well.
Carlos Martín Nieto 91123661 2014-02-04T22:04:00 refdb: add conditional symbolic updates Add a parameter to the backend to allow checking for the old symbolic target.
Carlos Martín Nieto 1202c7ea 2014-02-04T21:35:44 refs: fix leak on successful update Free the old ref even on success.
Carlos Martín Nieto fc4728e3 2014-01-29T14:07:18 refs: return GIT_EMODIFIED if the ref target moved In case we loose the race to update the reference, return GIT_EMODIFIED to let the user distinguish it from other types of errors.
Carlos Martín Nieto 5d96fe88 2014-01-14T15:33:29 refs: changes from feedback Change the name to _matching() intead of _if(), and force _set_target() to be a conditional update. If the user doesn't care about the old value, they should use git_reference_create().
Carlos Martín Nieto 9b148098 2013-12-18T19:58:16 refs: conditional ref updates Allow updating references if the old value matches the given one.
Ben Straub 010cec3a 2014-02-04T20:50:40 Add reflog params to git_repository_detach_head
Ben Straub c3ab1e5a 2014-02-04T20:38:13 Add reflog parameters to remote apis Also added a test for git_remote_fetch.
Ben Straub 491cecfe 2014-02-04T20:13:50 Add reflog parameters to git_push_update_tips
Russell Belfer 1bbacc9f 2014-02-04T16:46:43 Avoid extra copying in pqueue operations This tweaks the pqueue_up and pqueue_down routines so that they will not do full element swaps but instead carry over the state of the previous loop iteration and only assign elements for which we know the final position. This will avoid a little bit of data assignment which should improve performance in theory. Also got rid of some vector helpers that I'm no longer using.
Ben Straub 0adb0606 2014-02-04T15:32:57 Fix reflog message when creating commits
Russell Belfer 43709ca8 2014-02-04T10:33:30 Fix typo setting sorted flag when reloading index This fixes a typo I made for setting the sorted flag on the index after a reload. That typo didn't actually cause any test failures so I'm also adding a test that explicitly checks that the index is correctly sorted after a reload when ignoring case and when not.
Russell Belfer 882c7742 2014-02-04T10:01:37 Convert pqueue to just be a git_vector This updates the git_pqueue to simply be a set of specialized init/insert/pop functions on a git_vector. To preserve the pqueue feature of having a fixed size heap, I converted the "sorted" field in git_vectors to a more general "flags" field so that pqueue could mix in it's own flag. This had a bunch of ramifications because a number of places were directly looking at the vector "sorted" field - I added a couple new git_vector helpers (is_sorted, set_sorted) so the specific representation of this information could be abstracted.
Russell Belfer 4075e060 2014-02-03T21:02:08 Replace pqueue with code from hashsig heap I accidentally wrote a separate priority queue implementation when I was working on file rename detection as part of the file hash signature calculation code. To simplify licensing terms, I just adapted that to a general purpose priority queue and replace the old priority queue implementation that was borrowed from elsewhere. This also removes parts of the COPYING document that no longer apply to libgit2.
Edward Thomson dbfd83bc 2014-02-03T19:56:13 Remove unused pointer assignment
Edward Thomson c0b10c25 2014-02-01T12:05:00 Merge wd validation tests against index not HEAD Validating the workdir should not compare HEAD to working directory - this is both inefficient (as it ignores the cache) and incorrect. If we had legitimately allowed changes in the index (identical to the merge result) then comparing HEAD to workdir would reject these changes as different. Further, this will identify files that were filtered strangely as modified, while testing with the cache would prevent this. Also, it's stupid slow.
Ben Straub 86746b4b 2014-02-03T15:06:47 Add reset tests for reflog
Ben Straub 586be3b8 2014-02-03T15:05:55 Add reflog parameters to git_reset
Vicent Marti 3b6a5bac 2014-02-03T10:36:04 Merge pull request #2095 from libgit2/update-head-reflog Correct "new" id for reattached-HEAD reflog entry
Arthur Schreiber 50ad7cc2 2014-02-02T18:20:06 Add `git_reference_is_note`.
Ben Straub 7f058b86 2014-02-01T19:29:48 Check for errors when dereferencing symbolic refs
Ben Straub ee8e6afd 2014-02-01T11:46:44 Reflog: correct "new" id for reattaching HEAD
Ben Straub a1710a28 2014-01-29T10:35:46 Enhance testing of signature parameters
Ben Straub 59bb1126 2014-01-28T11:45:30 Provide good default reflog messages in branch api
Ben Straub ccf6ce5c 2014-01-28T11:30:36 Ensure renaming a reference updates the reflog
Ben Straub 540c1809 2014-01-28T10:44:33 Add reflog parameters to git_branch_move
Ben Straub 48110f67 2014-01-28T10:31:54 Deleting a branch deletes its reflog
Ben Straub 1cc974ab 2014-01-27T14:40:31 Augment clone API with reflog parameters
Ben Straub b31ebfbc 2014-01-27T14:12:29 Add reflog params to git_branch_create
Ben Straub 6357388e 2014-01-27T13:24:10 Enhance clarity
Ben Straub a2311f92 2014-01-27T13:12:31 Ensure updating HEAD updates reflog
Ben Straub 94f263f5 2014-01-25T08:04:49 Add reflog params to set-head calls
Vicent Marti 8646b0a0 2014-01-30T15:10:39 Merge pull request #2085 from libgit2/rb/index-tree-blob-collision Index tree-bob collision
Russell Belfer 25babd02 2014-01-30T11:38:59 Fix checkout NONE to not remove file If you are checking out NONE, then don't remove.
Russell Belfer b794cbcd 2014-01-30T11:38:25 Rename conflict to collision to prevent confusion
Russell Belfer 71ae7601 2014-01-30T11:30:16 Force explicit remove of files instead of defer The checkout code used to defer removal of "blocking" files in checkouts until the blocked item was actually being written (since we have already checked that the removing the block is acceptable according to the update rules). Unfortunately, this resulted in an intermediate index state where both the blocking and new items were in the index which is no longer allowed. Now we just remove the blocking item in the first pass so it never needs to coexist. In cases where there are typechanges, this could result in a bit more churn of removing and recreating intermediate directories, but I'm going to assume that is an unusual case and the churn will not be too costly.
Russell Belfer 19459b1e 2014-01-30T10:23:35 Defer zstream NUL termination to end And don't terminate if there isn't space for it (since it's binary data, it's not worth a reallocation).
Russell Belfer 8606f33b 2014-01-30T09:59:15 Expand zstream tests and fix off-by-one error
Russell Belfer d9b04d78 2014-01-29T15:02:35 Reorganize zstream API and fix wrap problems There were some confusing issues mixing up the number of bytes written to the zstream output buffer with the number of bytes consumed from the zstream input. This reorganizes the zstream API and makes it easier to deflate an arbitrarily large input while still using a fixed size output.
Russell Belfer e9d5e5f3 2014-01-28T16:25:42 Some fixes for Windows x64 warnings
Russell Belfer 3cf11eef 2014-01-28T11:47:33 Misc cleanups
Russell Belfer c0644c3f 2014-01-28T11:45:06 Make submodule fetchRecurse match other options This removes the fetchRecurse compiler warnings and makes the behavior match the other submodule options (i.e. the in-memory setting can be reset to the on-disk value).
Russell Belfer 5572d2b8 2014-01-28T11:44:17 Some missing oid to id renames
Edward Thomson 0972c592 2014-01-29T13:14:00 Two-phase index merging When three-way merging indexes, we previously changed each path as we read them, which would lead to us adding an index entry for 'foo', then removing an index entry for 'foo/file'. With the new index requirements, this is not allowed. Removing entries in the merged index, then adding them, resolves this. In the previous example, we now remove 'foo/file' before adding 'foo'.