midx and commit-graph files now use the byte defined in their file
format specification for identifying the hash function used for
object names.
* ds/sha256-leftover-bits:
multi-pack-index: use hash version byte
commit-graph: use the "hash version" byte
t/README: document GIT_TEST_DEFAULT_HASH
Further update of docs to adjust to the recent SHA-256 work.
* ma/sha-256-docs:
shallow.txt: document SHA-256 shallow format
protocol-capabilities.txt: clarify "allow-x-sha1-in-want" re SHA-256
index-format.txt: document SHA-256 index format
http-protocol.txt: document SHA-256 "want"/"have" format
A few end-user facing messages have been updated to be
hash-algorithm agnostic.
* jc/object-names-are-not-sha-1:
messages: avoid SHA-1 in end-user facing messages
Further update of docs to adjust to the recent SHA-256 work.
* bc/sha-256-doc-updates:
docs: fix step in transition plan
docs: document SHA-256 pack and indices
The regexp to identify the function boundary for FORTRAN programs
has been updated.
* pb/userdiff-fortran-update:
userdiff: improve Fortran xfuncname regex
userdiff: add tests for Fortran xfuncname regex
When given more than one target line ranges, "git blame -La,b
-Lc,d" was over-eager to coalesce groups of original lines and
showed incorrect results, which has been corrected.
* jk/blame-coalesce-fix:
blame: only coalesce lines that are adjacent in result
t8003: factor setup out of coalesce test
t8003: check output of coalesced blame
Ring buffer with size 4 used for bin-hex translation resulted in a
wrong object name in the sequencer's todo output, which has been
corrected.
* ak/sequencer-fix-find-uniq-abbrev:
rebase -i: fix possibly wrong onto hash in todo
The commit labels used to explain each side of conflicted hunks
placed by the sequencer machinery have been made more readable by
humans.
* en/sequencer-merge-labels:
sequencer: avoid garbled merge machinery messages due to commit labels
"git diff [<tree-ish>] $path" for a $path that is marked with i-t-a
bit was not showing the mode bits from the working tree.
* rp/ita-diff-modefix:
diff-lib: use worktree mode in diffs from i-t-a entries
Updates to "git merge" tests, in preparation for a new merge
strategy backend.
* en/merge-tests:
t6425: be more flexible with rename/delete conflict messages
t642[23]: be more flexible for add/add conflicts involving pair renames
t6422, t6426: be more flexible for add/add conflicts involving renames
t6423: add an explanation about why one of the tests does not pass
t6416, t6423: clarify some comments and fix some typos
t6422: fix multiple errors with the mod6 test expectations
t6423: fix test setup for a couple tests
t6416, t6422: fix incorrect untracked file count
t6422: fix bad check against missing file
t6418: tighten delete/normalize conflict testcase
Collect merge-related tests to t64xx
The FETCH_HEAD and MERGE_HEAD refs must be stored in a file, regardless of the
type of ref backend. This is because they can hold more than just a single ref.
To accomodate them for alternate ref backends, read them from a file generically
in refs_read_raw_ref()
Signed-off-by: Han-Wen Nienhuys <hanwen@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This prepares for handling FETCH_HEAD (which is not a regular ref)
separately from the ref backend.
Signed-off-by: Han-Wen Nienhuys <hanwen@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The dir structure seemed to have a number of leaks and problems around
it. First I noticed that parent_hashmap and recursive_hashmap were
being leaked (though Peff noticed and submitted fixes before me). Then
I noticed in the previous commit that clear_directory() was only taking
responsibility for a subset of fields within dir_struct, despite the
fact that entries[] and ignored[] we allocated internally to dir.c.
That, of course, resulted in many callers either leaking or haphazardly
trying to free these arrays and their contents.
Digging further, I found that despite the pretty clear documentation
near the top of dir.h that folks were supposed to call clear_directory()
when the user no longer needed the dir_struct, there were four callers
that didn't bother doing that at all. However, two of them clearly
thought about leaks since they had an UNLEAK(dir) directive, which to me
suggests that the method to free the data was too unclear. I suspect
the non-obviousness of the API and its holes led folks to avoid it,
which then snowballed into further problems with the entries[],
ignored[], parent_hashmap, and recursive_hashmap problems.
Rename clear_directory() to dir_clear() to be more in line with other
data structures in git, and introduce a dir_init() to handle the
suggested memsetting of dir_struct to all zeroes. I hope that a name
like "dir_clear()" is more clear, and that the presence of dir_init()
will provide a hint to those looking at the code that they need to look
for either a dir_clear() or a dir_free() and lead them to find
dir_clear().
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The calling convention for the dir API is supposed to end with a call to
clear_directory() to free up no longer needed memory. However,
clear_directory() didn't free dir->entries or dir->ignored. I believe
this was an oversight, but a number of callers noticed memory leaks and
started free'ing these. Unfortunately, they did so somewhat haphazardly
(sometimes freeing the entries in the arrays, and sometimes only
free'ing the arrays themselves). This suggests the callers weren't
trying to make sure any possible memory used might be free'd, but just
the memory they noticed their usecase definitely had allocated.
Fix this mess by moving all the duplicated free'ing logic into
clear_directory(). End by resetting dir to a pristine state so it could
be reused if desired.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Whitespace is ignored when calculating patch IDs. This is done by
removing all whitespace from diff lines before hashing them, including
a newline at the end of a file. If that newline is missing, however,
diff reports that fact in a separate line containing "\ No newline at
end of file\n", and this marker is hashed like a context line.
This goes against our goal of making patch IDs independent of
whitespace. Use the same heuristic that 2485eab55c (git-patch-id: do
not trip over "no newline" markers, 2011-02-17) added to git patch-id
instead and skip diff lines that start with a backslash and a space
and are longer than twelve characters.
Reported-by: Tilman Vogel <tilman.vogel@web.de>
Initial-test-by: Tilman Vogel <tilman.vogel@web.de>
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This if statement never evaluates to true since we already check
state->force a few lines above, and immediately return when it is
false.
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If you run fetch but record the result in remote-tracking branches,
and either if you do nothing with the fetched refs (e.g. you are
merely mirroring) or if you always work from the remote-tracking
refs (e.g. you fetch and then merge origin/branchname separately),
you can get away with having no FETCH_HEAD at all.
Teach "git fetch" a command line option "--[no-]write-fetch-head".
The default is to write FETCH_HEAD, and the option is primarily
meant to be used with the "--no-" prefix to override this default,
because there is no matching fetch.writeFetchHEAD configuration
variable to flip the default to off (in which case, the positive
form may become necessary to defeat it).
Note that under "--dry-run" mode, FETCH_HEAD is never written;
otherwise you'd see list of objects in the file that you do not
actually have. Passing `--write-fetch-head` does not force `git
fetch` to write the file.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
About half the function declarations in mem-pool.h used 'struct mem_pool
*pool', while the other half used 'struct mem_pool *mem_pool'. Make the
code a bit more consistent by just using 'pool' in preference to
'mem_pool' everywhere.
No behavioral changes included; this is just a mechanical rename (though
a line or two was rewrapped as well).
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A typical memory type, such as strbuf, hashmap, or string_list can be
stored on the stack or embedded within another structure. mem_pool
cannot be, because of how mem_pool_init() and mem_pool_discard() are
written. mem_pool_init() does essentially the following (simplified
for purposes of explanation here):
void mem_pool_init(struct mem_pool **pool...)
{
*pool = xcalloc(1, sizeof(*pool));
It seems weird to require that mem_pools can only be accessed through a
pointer. It also seems slightly dangerous: unlike strbuf_release() or
strbuf_reset() or string_list_clear(), all of which put the data
structure into a state where it can be re-used after the call,
mem_pool_discard(pool) will leave pool pointing at free'd memory.
read-cache (and split-index) are the only current users of mem_pools,
and they haven't fallen into a use-after-free mistake here, but it seems
likely to be problematic for future users especially since several of
the current callers of mem_pool_init() will only call it when the
mem_pool* is not already allocated (i.e. is NULL).
This type of mechanism also prevents finding synchronization
points where one can free existing memory and then resume more
operations. It would be natural at such points to run something like
mem_pool_discard(pool...);
and, if necessary,
mem_pool_init(&pool...);
and then carry on continuing to use the pool. However, this fails badly
if several objects had a copy of the value of pool from before these
commands; in such a case, those objects won't get the updated value of
pool that mem_pool_init() overwrites pool with and they'll all instead
be reading and writing from free'd memory.
Modify mem_pool_init()/mem_pool_discard() to behave more like
strbuf_init()/strbuf_release()
or
string_list_init()/string_list_clear()
In particular: (1) make mem_pool_init() just take a mem_pool* and have
it only worry about allocating struct mp_blocks, not the struct mem_pool
itself, (2) make mem_pool_discard() free the memory that the pool was
responsible for, but leave it in a state where it can be used to
allocate more memory afterward (without the need to call mem_pool_init()
again).
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
fast-import had a special mem_pool_strdup() convenience function that I
want to be able to use from the new merge algorithm I am writing. Move
it from fast-import to mem-pool, and also add a mem_pool_strndup()
while at it that I also want to use.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
git subtree push does not support --squash, as previously illustrated in
6ccc71a9 (contrib/subtree: there's no push --squash, 2015-05-07)
Signed-off-by: Danny Lin <danny0838@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Revise the documentation and remove previous "unsure" after making sure
that --message supports only 'add', 'merge', 'pull', and 'split --rejoin'.
Signed-off-by: Danny Lin <danny0838@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Earlier, to countermand the implicit "-m" option when the
"--first-parent" option is used with "git log", we added the
"--[no-]diff-merges" option in the jk/log-fp-implies-m topic. To
leave the door open to allow the "--diff-merges" option to take
values that instructs how patches for merge commits should be
computed (e.g. "cc"? "-p against first parent?"), redefine
"--diff-merges" to take non-optional value, and implement "off"
that means the same thing as "--no-diff-merges".
* so/log-diff-merges-opt:
t/t4013: add test for --diff-merges=off
doc/git-log: describe --diff-merges=off
revision: change "--diff-merges" option to require parameter
"git log --first-parent -p" showed patches only for single-parent
commits on the first-parent chain; the "--first-parent" option has
been made to imply "-m". Use "--no-diff-merges" to restore the
previous behaviour to omit patches for merge commits.
* jk/log-fp-implies-m:
doc/git-log: clarify handling of merge commit diffs
doc/git-log: move "-t" into diff-options list
doc/git-log: drop "-r" diff option
doc/git-log: move "Diff Formatting" from rev-list-options
log: enable "-m" automatically with "--first-parent"
revision: add "--no-diff-merges" option to counteract "-m"
log: drop "--cc implies -m" logic
Recent versions of "git diff-files" shows a diff between the index
and the working tree for "intent-to-add" paths as a "new file"
patch; "git apply --cached" should be able to take "git diff-files"
and should act as an equivalent to "git add" for the path, but the
command failed to do so for such a path.
* rp/apply-cached-with-i-t-a:
t4140: test apply with i-t-a paths
apply: make i-t-a entries never match worktree
apply: allow "new file" patches on i-t-a entries
"git bisect" learns the "--first-parent" option to find the first
breakage along the first-parent chain.
* al/bisect-first-parent:
bisect: combine args passed to find_bisection()
bisect: introduce first-parent flag
cmd_bisect__helper: defer parsing no-checkout flag
rev-list: allow bisect and first-parent flags
t6030: modernize "git bisect run" tests
A no-op replacement function implemented as a C preprocessor macro
does not perform as good a job as one implemented as a "static
inline" function in catching errors in parameters; replace the
former with the latter in <git-compat-util.h> header.
* jc/noop-with-static-inline:
compat-util: type-check parameters of no-op replacement functions
The existing backends for "git mergetool" based on variants of vim
have been refactored and then support for "nvim" has been added.
* pd/mergetool-nvimdiff:
mergetools: add support for nvimdiff (neovim) family
mergetool--lib: improve support for vimdiff-style tool variants
Further preliminary change to refs API.
* hn/reftable-prep-part-2:
Make HEAD a PSEUDOREF rather than PER_WORKTREE.
Modify pseudo refs through ref backend storage
t1400: use git rev-parse for testing PSEUDOREF existence
Stop when "sendmail.*" configuration variables are defined, which
could be a mistaken attempt to define "sendemail.*" variables.
* dd/send-email-config:
git-send-email: die if sendmail.* config is set
The logic to find the ref transaction hook script attempted to
cache the path to the found hook without realizing that it needed
to keep a copied value, as the API it used returned a transitory
buffer space. This has been corrected.
* ps/ref-transaction-hook:
t1416: avoid hard-coded sha1 ids
refs: fix interleaving hook calls with reference-transaction hook
Similar to the commit-graph format, the multi-pack-index format has a
byte in the header intended to track the hash version used to write the
file. This allows one to interpret the hash length without having the
context of the repository config specifying the hash length. This was
not modified as part of the SHA-256 work because the hash length was
automatically up-shifted due to that config.
Since we have this byte available, we can make the file formats more
obviously incompatible instead of relying on other context from the
repository.
Add a new oid_version() method in midx.c similar to the one in
commit-graph.c. This is specifically made separate from that
implementation to avoid artificially linking the formats.
The test impact requires a few more things than the corresponding change
in the commit-graph format. Specifically, 'test-tool read-midx' was not
writing anything about this header value to output. Since the value
available in 'struct multi_pack_index' is hash_len instead of a version
value, we output "20" or "32" instead of "1" or "2".
Since we want a user to not have their Git commands fail if their
multi-pack-index has the incorrect hash version compared to the
repository's hash version, we relax the die() to an error() in
load_multi_pack_index(). This has some effect on 'git multi-pack-index
verify' as we need to check that a failed parse of a file that exists is
actually a verify error. For that test that checks the hash version
matches, we change the corrupted byte from "2" to "3" to ensure the test
fails for both hash algorithms.
Helped-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Reviewed-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The commit-graph format reserved a byte among the header of the file to
store a "hash version". During the SHA-256 work, this was not modified
because file formats are not necessarily intended to work across hash
versions. If a repository has SHA-256 as its hash algorithm, it
automatically up-shifts the lengths of object names in all necessary
formats.
However, since we have this byte available for adjusting the version, we
can make the file formats more obviously incompatible instead of relying
on other context from the repository.
Update the oid_version() method in commit-graph.c to add a new value, 2,
for sha-256. This automatically writes the new value in a SHA-256
repository _and_ verifies the value is correct. This is a breaking
change relative to the current 'master' branch since 092b677 (Merge
branch 'bc/sha-256-cvs-svn-updates', 2020-08-13) but it is not breaking
relative to any released version of Git.
The test impact is relatively minor: the output of 'test-tool
read-graph' lists the header information, so those instances of '1' need
to be replaced with a variable determined by GIT_TEST_DEFAULT_HASH. A
more careful test is added that specifically creates a repository of
each type then swaps the commit-graph files. The important value here is
that the "git log" command succeeds while writing a message to stderr.
Helped-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Reviewed-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Helped-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Reviewed-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the ensure_core_worktree() function, we load the core.worktree value
of the submodule repository using repo_config_get_string(). This
function copies the string, but we never free it, leaking the memory.
We can instead use the "tmp" version of that function to avoid the
allocation at all. We don't have to worry about lifetime issues, since
we never even look at the value (we just want to know if it's set).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We use git_config_get_string() to retrieve the expiry value in a newly
allocated string. But after parsing it, we never free it, leaking the
memory.
We could fix this with a free() obviously, but there's an even better
solution: we can use the non-allocating "tmp" variant of the function;
we only need it to be valid for the lifetime of our parse function.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As evidenced by the leak fixes in the previous commit, the "const" in
git_config_get_string_const() clearly misleads people into thinking that
it does not allocate a copy of the string. We can fix this by renaming
it, but it's easier still to just drop it. Of the four remaining
callers:
- The one in git_config_parse_expiry() still needs to allocate, since
that's what its callers expect. We can just use the non-const
version and cast our pointer. Slightly ugly, but the damage is
contained in one spot.
- The two in apply are writing to global "const char *" variables, and
need to continue allocating. We often mark these as const because we
assign default string literals to them. But in this case we don't do
that, so we can just declare them as real "char *" pointers and use
the non-const version.
- The call in checkout doesn't actually need a copy; it can just use
the non-allocating "tmp" version of the function.
The function is also mentioned in the MyFirstContribution document. We
can swap that call out for the non-allocating "tmp" variant, which fits
well in the example given.
We'll drop the "configset" and "repo" variants, as well (which are
unused).
Note that this frees up the "const" name, so we could rename the "tmp"
variant back to that. But let's give some time for topics in flight to
adapt to the new code before doing so (if we do it too soon, the
function semantics will change but the compiler won't alert us).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Update mingw_unlink() to first try to delete the file with existing
permissions before trying to force it.
Windows throws an error when trying to delete a read-only file. The
mingw_unlink() compatibility wrapper always tries to _wchmod(666) the
file before calling _wunlink() to avoid that error. However, since
most files in the worktree are already writable, this is usually
wasted effort.
Update mingw_unlink() to just call DeleteFileW() directly and if that
succeeds return. If that fails, fall back into the existing code path
to update the permissions and use _wunlink() to get the existing
error code mapping.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
After eff45daab8 ("repository: enable SHA-256 support by default",
2020-07-29), vanilla builds of Git enable the user to run, e.g.,
git init --object-format=sha256
and hack away. This can be a good way to gain experience with the
SHA-256 world, e.g., to find bugs that
GIT_TEST_DEFAULT_HASH=sha256 make test
doesn't spot.
But it really is a separate world: Such SHA-256 repos will live entirely
separate from the (by now fairly large) set of SHA-1 repos. Interacting
across the border is possible in principle, e.g., through "diff + apply"
(or "format-patch + am"), but even that has its limitations: Applying a
SHA-256 diff in a SHA-1 repo works in the simple case, but if you need
to resort to `-3`, you're out of luck.
Similarly, "push + pull" should work, but you really will be operating
mostly offset from the rest of the world. That might be ok by the time
you initialize your repository, and it might be ok for several months
after that, but there might come a day when you're starting to regret
your use of `git init --object-format=sha256` and have dug yourself into
a fairly deep hole.
There are currently topics in flight to document our data formats and
protocols regarding SHA-256 and in some cases (midx and commit-graph),
we're considering adjusting how the file formats indicate which object
format to use.
Wherever `--object-format` is mentioned in our documentation, let's make
it clear that using it with "sha256" is experimental. If we later need
to explain why we can't handle data we generated back in 2020, we can
always point to this paragraph we're adding here.
By "include::"-ing a small blurb, we should be able to be consistent
throughout the documentation and can eventually gradually tone down the
severity of this text. One day, we might even use it to start phasing
out `--object-format=sha1`, but let's not get ahead of ourselves...
There's also `extensions.objectFormat`, but it's only mentioned three
times. Twice where we're adding this new disclaimer and in the third
spot we already have a "do not edit" warning. From there, interested
readers should eventually find this new one that we're adding here.
Because `GIT_DEFAULT_HASH` provides another entry point to this
functionality, document the experimental nature of it too.
Signed-off-by: Martin Ågren <martin.agren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A couple of functions that used struct refspec_item did not zero out the
structure memory. This can result in unexpected behavior, especially if
additional parameters are ever added to refspec_item in the future. Use
memset to ensure that unset structure members are zero.
It may make sense to convert most of these uses of struct refspec_item
to use either struct initializers or refspec_item_init_or_die. However,
other similar code uses memset. Converting all of these uses has been
left as a future exercise.
Signed-off-by: Jacob Keller <jacob.keller@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano