|
de3cf801
|
2014-07-02T15:28:24
|
|
Merge pull request #2456 from libgit2/cmn/ssh-send-everything
ssh: libssh2_channel_write() behaves like send()
|
|
0963716b
|
2014-07-02T12:49:51
|
|
ssh: libssh2_channel_write() behaves like send()
When the stream writing function was written, it assume that
libssh2_channel_write() would always write all of the data to the
wire. This is only true for the first 32k of data, which it tries to
fit into one ssh packet.
Since it can perform short writes, call it in a loop like we do for
send(), advancing the buffer offset.
|
|
5fa8cda9
|
2014-06-30T12:05:25
|
|
Round up pool alloc sizes for alignment
To make sure that items returned from pool allocations are aligned
on nice boundaries, this rounds up all pool allocation sizes to a
multiple of 8. This adds a small amount of overhead to each item.
The rounding up could be made optional with an extra parameter to
the pool initialization that turned on rounding only for pools
where item alignment actually matters, but I think for the extra
code and complexity that would be involved, that it makes sense
just to burn a little bit of extra memory and enable this all the
time.
|
|
|
dcdb8500
|
2014-06-30T17:35:42
|
|
Merge pull request #2440 from phkelley/transports
Improvements to git_transport extensibility
|
|
bc8a0886
|
2014-06-27T11:51:35
|
|
Fix assert when receiving uncommon sideband packet
|
|
|
1697cd6f
|
2014-06-25T13:20:27
|
|
Improvements to git_transport extensibility
git_remote_set_transport now takes a transport factory rather than a transport
git_clone_options now allows the caller to specify a remote creation callback
|
|
|
f36d57b9
|
2014-06-26T07:48:09
|
|
Fixes #2443 Zero size arrays are an extension
|
|
|
c19b1c04
|
2014-06-25T21:35:58
|
|
pack: clean up error returns
Set a message when we fail to lock.
Also make the put function void, since it's called from free, which
cannot report errors. The only errors we can experience here are
internal state corruption, so we assert that we are trying to put a
pack which we have previously got.
|
|
|
966fb207
|
2014-06-25T21:25:44
|
|
tree: free in error conditions
As reported by coverity, we would leak some memory in error conditions.
|
|
|
5e0f47c3
|
2014-06-25T21:20:39
|
|
pack: free the new pack struct if we fail to insert
If we fail to insert the packfile in the map, make sure to free it.
This makes the free function only attempt to remove its mwindows from
the global list if we have opened the packfile to avoid accessing the
list unlocked.
|
|
d412165f
|
2014-06-18T16:54:32
|
|
Update text=auto / core.autocrlf=false behavior
Git for Windows 1.9.4 changed the behavior when the text=auto
attribute is specified and core.autocrlf=false. Previous observed
behavior would *not* filter files when going into the working
directory, the new behavior *does* filter. Update our behavior to match.
|
|
|
3ddd0d92
|
2014-06-24T17:55:15
|
|
Merge remote-tracking branch 'upstream/cmn/mixed-eol-passthrough'
|
|
|
5e37874d
|
2014-06-24T17:51:45
|
|
Merge remote-tracking branch 'upstream/cmn/treebuilder-perf'
|
|
|
58152669
|
2014-06-24T17:35:14
|
|
Merge branch 'cmn/global-mwf'
|
|
|
5a76ad35
|
2014-06-19T11:45:46
|
|
crlf: pass-through mixed EOL buffers from LF->CRLF
When checking out files, we're performing conversion into the user's
native line endings, but we only want to do it for files which have
consistent line endings. Refuse to perform the conversion for mixed-EOL
files.
The CRLF->LF filter is left as-is, as that conversion is considered to be
normalization by git and should force a conversion of the line endings.
|
|
|
b3b66c57
|
2014-06-18T17:13:12
|
|
Share packs across repository instances
Opening the same repository multiple times will currently open the same
file multiple times, as well as map the same region of the file multiple
times. This is not necessary, as the packfile data is immutable.
Instead of opening and closing packfiles directly, introduce an
indirection and allocate packfiles globally. This does mean locking on
each packfile open, but we already use this lock for the global mwindow
list so it doesn't introduce a new contention point.
|
|
|
461da57a
|
2014-06-23T17:32:30
|
|
Merge remote-tracking branch 'upstream/cmn/filebuf-atomic-unlock'
|
|
|
e93206e0
|
2014-06-14T12:58:03
|
|
Merge pull request #2421 from libgit2/cmn/init-ssl-once
netops: init OpenSSL once under lock
|
|
|
9c3e4e97
|
2014-06-13T02:35:33
|
|
http: fix typo in credentials logic
We want to check whether the credentials callback is NULL, not whether
the payload is.
|
|
|
081e76ba
|
2014-06-12T16:20:52
|
|
ssl: init everything all the time
Bring together all of the OpenSSL initialization to
git_threads_init() so it's together and doesn't need locks.
Moving it here also gives us libssh2 thread safety (when built against
openssl).
|
|
|
8f897b6f
|
2014-06-12T14:50:08
|
|
ssl: init also without threads
|
|
|
cf15ac8a
|
2014-06-12T03:20:34
|
|
ssl: cargo-cult thread safety
OpenSSL's tests init everything in the main thread, so let's do that.
|
|
|
5fa04943
|
2014-06-11T23:19:48
|
|
ssl: use locking
When using in a multithreaded context, OpenSSL needs to lock, and leaves
it up to application to provide said locks.
We were not doing this, and it's just luck that's kept us from crashing
up to now.
|
|
|
1d3364ac
|
2014-06-11T20:52:15
|
|
netops: init OpenSSL once under lock
The OpenSSL init functions are not reentrant, which means that running
multiple fetches in parallel can cause us to crash.
Use a mutex to init OpenSSL, and since we're adding this extra checks,
init it only once.
|
|
|
f9a97667
|
2014-06-11T00:06:44
|
|
revwalk: more sensible array handling
Instead of using a sentinel empty value to detect the last commit, let's
check for when we get a NULL from popping the stack, which lets us know
when we're done.
The current code causes us to read uninitialized data, although only on
RHEL/CentOS 6 in release mode. This is a readability win overall.
|
|
|
fcc60066
|
2014-06-09T22:59:32
|
|
treentry: no need for manual size book-keeping
We can simply ask the hasmap.
|
|
|
978fbb4c
|
2014-06-09T22:45:23
|
|
treebuilder: don't keep removed entries around
If the user wants to keep a copy for themselves, they should make a
copy. It adds unnecessary complexity to make sure the returned entries
are valid until the builder is cleared.
|
|
|
4d3f1f97
|
2014-06-09T04:38:22
|
|
treebuilder: use a map instead of vector to store the entries
Finding a filename in a vector means we need to resort it every time we
want to read from it, which includes every time we want to write to it
as well, as we want to find duplicate keys.
A hash-map fits what we want to do much more accurately, as we do not
care about sorting, but just the particular filename.
We still keep removed entries around, as the interface let you assume
they were going to be around until the treebuilder is cleared or freed,
but in this case that involves an append to a vector in the filter case,
which can now fail.
The only time we care about sorting is when we write out the tree, so
let's make that the only time we do any sorting.
|
|
|
2c11d2ee
|
2014-06-09T23:23:53
|
|
treebuilder: insert sorted
By inserting in the right position, we can keep the vector sorted,
making entry insertion almost twice as fast.
|
|
|
ce5e6617
|
2014-06-08T16:44:32
|
|
Merge pull request #2407 from libgit2/cmn/remote-rename-more
More remote rename fixes
|
|
1b4e29b7
|
2014-06-07T13:56:39
|
|
React to review feedback
|
|
|
fb591767
|
2014-06-07T12:51:48
|
|
Win32: Fix object::cache::threadmania test on x64
|
|
|
231f350d
|
2014-06-06T22:55:34
|
|
remote: don't free the remote on delete
This was a bad idea. Don't free except in the free function.
|
|
|
d1544564
|
2014-06-06T22:38:26
|
|
remote: handle symrefs when renaming
A symref inside the namespace gets renamed, we should make it point to
the target's new name.
This is for the origin/HEAD -> origin/master type of situations.
|
|
|
72bca13e
|
2014-06-06T16:33:54
|
|
remote: return problem refspecs instead of using a callback
There is no reason why we need to use a callback here. A string array
fits better with the usage, as this is not an event and we don't need
anything from the user.
|
|
|
61dcfe14
|
2014-06-06T15:57:37
|
|
remote: make sure the name stays valid on rename
We must make sure that the name pointer remains valid, so make sure to
allocate the new one before freeing the old one and swap them so the
user never sees an invalid pointer.
|
|
|
5a49ff9f
|
2014-06-06T15:54:42
|
|
remote: remove rename code for anonymous remotes
We don't allow renames of anonymous remotes, so there's no need to
handle them.
A remote is always associated with a repository, so there's no need to
check for that.
|
|
|
a52ab4b8
|
2014-06-06T01:09:49
|
|
remote: tighten up reference renaming
Tighten up which references we consider for renaming so we don't try to
rename unrelated ones and end up with unexplained references.
If there is a reference on the target namespace, git overwrites it, so
let's do the same.
|
|
|
e58281aa
|
2014-04-04T14:40:38
|
|
filebuf: make unlocking atomic
When renaming a lock file to its final location, we need to make sure
that it is replaced atomically.
We currently have a workaround for Windows by removing the target file.
This means that the target file, which may be a ref or a packfile, may
cease to exist for a short wile, which shold be avoided.
Implement the workaround only in Windows, by making sure that the file
we want to replace is writable.
|
|
|
90befde4
|
2014-06-03T22:10:34
|
|
Merge pull request #2399 from libgit2/cmn/path-to-path
clone: re-use the local transport's path resolution
|
|
|
dfcba09e
|
2014-06-03T13:05:20
|
|
Merge pull request #2395 from libgit2/cmn/ref-iter-concurrent
Concurrent ref iterator access
|
|
|
18d7896c
|
2014-06-03T21:47:53
|
|
clone: re-use the local transport's path resolution
Whe already worked out the kinks with the function used in the local
transport. Expose it and make use of it in the local clone method
instead of trying to work it out again.
|
|
|
bccb36eb
|
2014-06-03T17:25:59
|
|
Merge pull request #2389 from arthurschreiber/arthur/set-error-when-no-remote-found
Remote: Set an error when a remote cannot be found.
|
|
|
2d945f82
|
2014-06-02T17:44:51
|
|
refs: copy the packed refs on iteration
This lets us work without worrying about what's happening but work on a
snapshot.
|
|
|
8a9419aa
|
2014-06-01T02:16:07
|
|
remote: build up the list of refs to remove
When removing the remote-tracking branches, build up the list and remove
in two steps, working around an issue with the iterator. Removing while
we're iterating over the refs can cause us to miss references.
|
|
|
bc81220d
|
2014-05-31T10:19:55
|
|
minor cleanups
|
|
|
947a58c1
|
2014-05-30T13:19:49
|
|
Clean up the handling of large binary diffs
|
|
d723dbed
|
2014-05-30T19:26:49
|
|
Remote: Set an error when a remote cannot be found.
Inside `git_remote_load`, the calls to `get_optional_config` use
`giterr_clear` to unset any errors that are set due to missing config
keys. If neither a fetch nor a push url config was found for a remote,
we should set an error again.
|
|
|
49837fd4
|
2014-05-30T11:30:53
|
|
Ignore core.safecrlf=warn until we have a warn infrastructure
|
|
|
824f755f
|
2014-05-20T17:31:53
|
|
Refs: Introduce `git_refname_t`.
|
|
|
68f9d6b2
|
2014-05-15T22:44:50
|
|
Refs: Fix some issue when core.precomposeunicode = true.
This fixes two issues I found when core.precomposeunicode is enabled:
* When creating a reference with a NFD string, the returned
git_reference would return this NFD string as the reference’s
name. But when looking up the reference later, the name would
then be returned as NFC string.
* Renaming a reference would not honor the core.precomposeunicode and
apply no normalization to the new reference name.
|
|
|
5f0527ae
|
2014-05-30T13:06:34
|
|
config: initialize the error
The error would be uninitialized if we take a snapshot of a config with
no backends.
|
|
|
31c55152
|
2014-05-29T14:50:57
|
|
Merge pull request #2011 from libgit2/cmn/clone-local
Local clone
|
|
|
2614819c
|
2014-05-28T11:28:57
|
|
clone: allow for linking in local clone
If requested, git_clone_local_into() will try to link the object files
instead of copying them.
This only works on non-Windows (since it doesn't have this) when both
are on the same filesystem (which are unix semantics).
|
|
|
94f742ba
|
2014-05-28T10:18:05
|
|
fileops: allow linking files when copying directory structures
When passed the LINK_FILES flag, the recursive copy will hardlink files
instead of copying them.
|
|
|
c1dbfcbb
|
2014-05-28T10:07:23
|
|
clone: add flag not to link
|
|
|
121b2673
|
2013-12-23T11:12:31
|
|
clone: add flags to override whether to perform a local clone
|
|
|
a0b5f785
|
2013-12-22T15:39:54
|
|
clone: store the realpath when given a relative one
A call like git_clone("./foo", "./foo1") writes origin's url as './foo',
which makes it unusable, as they're relative to different things.
Go with git's behaviour and store the realpath as the url.
|
|
|
4386d80b
|
2013-12-21T17:18:21
|
|
clone: perform a "local clone" when given a local path
When git is given such a path, it will perform a "local clone",
bypassing the git-aware protocol and simply copying over all objects
that exist in the source.
Copy this behaviour when given a local path.
|
|
|
5d91bea2
|
2014-05-28T13:55:57
|
|
Merge pull request #2380 from libgit2/cmn/index-add-modes
index: check for valid filemodes on add
|
|
|
07c0eacd
|
2014-05-28T13:50:58
|
|
Merge pull request #2359 from e45lee/chmod-fix
Fixed permissions on template directories.
|
|
|
eff531e1
|
2014-05-27T20:58:20
|
|
Modify GIT_MERGE_CONFIG -> GIT_MERGE_PREFERENCE
|
|
|
de3f851e
|
2014-05-21T09:44:05
|
|
Staticify `merge_config`
|
|
|
22ab8881
|
2014-05-20T22:07:15
|
|
Use a config snapshot
|
|
|
a3622ba6
|
2014-05-16T13:54:40
|
|
Move GIT_MERGE_CONFIG_* to its own enum
|
|
|
d362093f
|
2014-05-08T15:41:36
|
|
Introduce GIT_MERGE_CONFIG_* for merge.ff settings
git_merge_analysis will now return GIT_MERGE_CONFIG_NO_FASTFORWARD
when merge.ff=false and GIT_MERGE_CONFIG_FASTFORWARD_ONLY when
merge.ff=true
|
|
517341c5
|
2014-05-23T22:41:35
|
|
Address style concerns in setting mkdir/copy flags.
|
|
|
530594c0
|
2014-05-23T05:53:41
|
|
odb: clear backend errors on successful read
We go through the different backends in order, so it's not an error if
at least one of the backends has the data we want.
|
|
|
052a2ffd
|
2014-05-22T16:01:02
|
|
index: check for valid filemodes on add
|
|
|
9331f98a
|
2014-05-22T12:52:31
|
|
smart: initialize the error variable
|
|
|
4c4408c3
|
2014-05-22T12:28:39
|
|
Plug leaks and fix a C99-ism
We have too many places where we repeat free code, so when adding the
new free to the generic code, it didn't take for the local transport.
While there, fix a C99-ism that sneaked through.
|
|
|
a5982644
|
2014-05-22T02:28:42
|
|
Merge pull request #2376 from libgit2/cmn/remote-symref
Add support for the symref extension
|
|
|
7ee233a9
|
2014-05-22T02:27:33
|
|
Merge pull request #2375 from libgit2/rb/safecrlf-on-lf-platform
Make core.safecrlf not generate an error on LF-ending platforms
|
|
|
2bc36fac
|
2014-05-21T13:07:45
|
|
Merge pull request #2372 from libgit2/cmn/fetch-terminate
smart: send a flush when we disconnect
|
|
|
2a597116
|
2014-05-21T11:54:10
|
|
clone: get rid of head_info
Since we no longer need to push data to callbacks, there's no need for
this truct.
|
|
|
cdb8a608
|
2014-05-21T11:51:33
|
|
clone: make use of the remote's default branch guessing
Let's use the remote's default branch guessing instead of reinventing
one ourselves with callbacks.
|
|
|
d22db24f
|
2014-05-21T09:32:35
|
|
remote: add api to guess the remote's default branch
If the remote supports the symref protocol extension, then we return
that, otherwise we guess with git's rules.
|
|
|
04865aa0
|
2014-05-21T10:01:44
|
|
local transport: expose the symref data
When using the local transport, we always have the symbolic information
available, so fill it.
|
|
|
306475eb
|
2014-05-20T09:55:26
|
|
remote: expose the remote's symref mappings
Add a symref_target field to git_remote_head to expose the symref
mappings to the user.
|
|
|
8156835d
|
2014-05-20T09:29:39
|
|
smart: store reported symrefs
The protocol has a capability which allows the server to tell us which
refs are symrefs, so we can e.g. know which is the default branch.
This capability is different from the ones we already support, as it's
not setting a flag to true, but requires us to store a list of
refspec-formatted mappings.
This commit does not yet expose the information in the reference
listing.
|
|
60cdf495
|
2014-05-19T09:13:45
|
|
Minor fix for cmn/clone-into-mirror.
A recently added check might skip initialization of old_fetchhead and go
directly to cleanup. So, destruct in the opposite order of construction.
|
|
|
ac11219b
|
2014-05-19T16:54:19
|
|
smart: send a flush when we disconnect
The git server wants to hear a flush from us when we disconnect,
particularly when we want to perform a fetch but are up to date.
|
|
|
c094197b
|
2014-05-19T15:05:39
|
|
Just don't CRLF filter if there are no CRs
|
|
|
16798d08
|
2014-05-19T14:57:09
|
|
Make core.safecrlf work on LF-ending platforms
If you enabled core.safecrlf on an LF-ending platform, we would
error even for files with all LFs. We should only be warning on
irreversible mappings, I think.
|
|
|
588a43af
|
2014-05-19T15:43:13
|
|
Merge pull request #2371 from martinwoodward/attrib_fnmatch
Restore attributions for fnmatch
|
|
213a269a
|
2014-05-19T14:39:45
|
|
Restore attributions for fnmatch
|
|
|
124a45ea
|
2014-05-19T15:00:50
|
|
Merge pull request #2354 from libgit2/cmn/clone-into-mirror
Allow mirror-clone via `git_clone_into()`
|
|
|
32332fcc
|
2014-05-19T14:15:40
|
|
clone: don't error out if the branch already exists
We set up the current branch after we fetch from the remote. This means
that the user's refspec may have already created this reference. It is
therefore not an error if we cannot create the branch because it already
exists.
This allows for the user to replicate git-clone's --mirror option.
|
|
|
3c607685
|
2014-05-19T13:36:00
|
|
clone: duplicate the remote
Instead of changing the user-provided remote, duplicate it so we can add
the extra refspec without having to worry about unsetting it before
returning.
|
|
|
52b597b6
|
2014-05-19T12:21:37
|
|
Merge pull request #2364 from libgit2/cmn/comment-char
message: don't assume the comment char
|
|
|
138af337
|
2014-05-19T12:20:31
|
|
Merge pull request #2303 from jacquesg/mingw-lseek
WIP: Windows fixes
|
|
|
9c4feef9
|
2014-05-17T12:44:21
|
|
Fix warning on uninitialized variable.
|
|
|
49e369b2
|
2014-05-18T10:06:49
|
|
message: don't assume the comment char
The comment char is configurable and we need to provide a way for the
user to specify which comment char they chose for their message.
|
|
|
4c9ffdff
|
2014-05-17T12:45:34
|
|
Fix printf format string from previous commit
|
|
|
c6320bec
|
2014-05-17T12:19:32
|
|
print_binary_hunk: Treat types with respect
|
|
|
bf6a5b61
|
2014-05-17T11:38:03
|
|
Merge remote-tracking branch 'upstream/cmn/indexer-mmap' into development
|
|
|
191ff936
|
2014-05-17T02:37:13
|
|
Merge pull request #2362 from libgit2/rb/update-4k-to-8k
Test and fix Git diff binary detection compatibility
|
|
|
0731a5b4
|
2014-05-14T19:12:48
|
|
indexer: mmap fixes for Windows
Windows has its own ftruncate() called _chsize_s().
p_mkstemp() is changed to use p_open() so we can make sure we open for
writing; the addition of exclusive create is a good thing to do
regardless, as we want a temporary path for ourselves.
Lastly, MSVC doesn't quite know how to add two numbers if one of them is a
void pointer, so let's alias it to unsigned char.C
|
|
|
f7310540
|
2014-05-13T02:41:48
|
|
indexer: use mmap for writing
Some OSs cannot keep their ideas about file content straight when mixing
standard IO with file mapping. As we use mmap for reading from the
packfile, let's make writing to the pack file use mmap.
|
|
|
d0f00de4
|
2014-05-16T11:08:19
|
|
Increase binary detection len to 8k
|