In 5bf96e0c39 (reftable/generic: move seeking of records into the
iterator, 2024-05-13) we have refactored the reftable codebase such that
iterators can be initialized once and then re-seeked multiple times.
This feature is used by 1869525066 (refs/reftable: wire up support for
exclude patterns, 2024-09-16) in order to skip records based on exclude
patterns provided by the caller.
The logic to re-seek the merged iterator is insufficient though because
we don't drain the priority queue on a re-seek. This means that the
queue may contain stale entries and thus reading the next record in the
queue will return the wrong entry. While this is an obvious bug, it is
harmless in the context of above exclude patterns:
- If the queue contained stale entries that match the pattern then the
caller would already know to filter out such refs. This is because
our codebase is prepared to handle backends that don't have a way to
efficiently implement exclude patterns.
- If the queue contained stale entries that don't match the pattern
we'd eventually filter out any duplicates. This is because the
reftable code discards items with the same ref name and sorts any
remaining entries properly.
So things happen to work in this context regardless of the bug, and
there is no other use case yet where we re-seek iterators. We're about
to introduce a caching mechanism though where iterators are reused by
the reftable backend, and that will expose the bug.
Fix the issue by draining the priority queue when seeking and add a
testcase that surfaces the issue.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Reftable stacks are reloaded in two cases:
- When calling `reftable_stack_reload()`, if the stat-cache tells us
that the stack has been modified.
- When committing a reftable addition.
While callers can figure out the second case, they do not have a
mechanism to figure out whether `reftable_stack_reload()` led to an
actual reload of the on-disk data. All they can do is thus to assume
that data is always being reloaded in that case.
Improve the situation by introducing a new `on_reload()` callback to the
reftable options. If provided, the function will be invoked every time
the stack has indeed been reloaded. This allows callers to invalidate
data that depends on the current stack data.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Refactor the callback function that expires reflog entries in the
reftable backend to use `reftable_backend_read_ref()` instead of
accessing the reftable stack directly. This ensures that the function
will benefit from the new caching layer that we're about to introduce.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Refactor the callback function that reads symbolic references in the
reftable backend to use `reftable_backend_read_ref()` instead of
accessing the reftable stack directly. This ensures that the function
will benefit from the new caching layer that we're about to introduce.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Refactor `read_ref_without_reload()` to accept `struct reftable_backend`
as parameter instead of `struct reftable_stack`. Rename the function to
`reftable_backend_read_ref()` to clarify its scope and move it close to
other functions operating on `struct reftable_backend`.
This change allows us to implement an additional caching layer when
reading refs where we can reuse reftable iterators.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The function `read_ref_without_reload()` accepts a ref store as input
only so that we can figure out the hash function used by it. This is
duplicate information though because the reftable stack knows about its
hash function, too.
Drop the superfluous parameter to simplify the calling convention a bit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add an accessor function that allows callers to access the hash ID of a
reftable stack. This function will be used in a subsequent commit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When accessing a stack we almost always have to reload the stack before
reading data from it. This is mostly because Git does not have a
notification mechanism for when underlying data has been changed, and
thus we are forced to opportunistically reload the stack every single
time to account for any changes that may have happened concurrently.
Handle the reload internally in `backend_for()`. For one this forces
callsites to think about whether or not they need to reload the stack.
But second this makes the logic to access stacks more self-contained by
letting the `struct reftable_backend` manage themselves.
Update callsites where we don't reload the stack to document why we
don't. In some cases it's unclear whether it is the right thing to do in
the first place, but fixing that is outside of the scope of this patch
series.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The reftable ref store needs to keep track of multiple stacks, one for
the main worktree and an arbitrary number of stacks for worktrees. This
is done by storing pointers to `struct reftable_stack`, which we then
access directly.
Wrap the stack in a new `struct reftable_backend`. This will allow us to
attach more data to each respective stack in subsequent commits.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 9b1cb5070f (builtin: add a repository parameter for builtin
functions, 2024-09-13) the repository was passed down to all builtin
commands. This allowed the repository to be passed down to lower layers
without depending on the global `the_repository` variable.
Continue this work by also passing down the repository parameter from
the command to sub-commands. This will help pass down the repository to
other subsystems and cleanup usage of global variables like
'the_repository' and 'the_hash_algo'.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If a user specified e.g.
M 100644 :1 ../some-file
then fast-import previously would happily create a git history where
there is a tree in the top-level directory named "..", and with a file
inside that directory named "some-file". The top-level ".." directory
causes problems. While git checkout will die with errors and fsck will
report hasDotdot problems, the user is going to have problems trying to
remove the problematic file. Simply avoid creating this bad history in
the first place.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Coverity has started to warn about a potential double-free in
`find_bisection()`. This warning is triggered because we may modify the
list head of the passed-in `commit_list` in case it is an UNINTERESTING
commit, but still call `free_commit_list()` on the original variable
that points to the now-freed head in case where `do_find_bisection()`
returns a `NULL` pointer.
As far as I can see, this double free cannot happen in practice, as
`do_find_bisection()` only returns a `NULL` pointer when it was passed a
`NULL` input. So in order to trigger the double free we would have to
call `find_bisection()` with a commit list that only consists of
UNINTERESTING commits, but I have not been able to construct a case
where that happens.
Drop the `else` branch entirely as it seems to be a no-op anyway.
Another option might be to instead call `free_commit_list()` on `list`,
which is the modified version of `commit_list` and thus wouldn't cause a
double free. But as mentioned, I couldn't come up with any case where a
passed-in non-NULL list becomes empty, so this shouldn't be necessary.
And if it ever does become necessary we'd notice anyway via the leak
sanitizer.
Interestingly enough we did not have a single test exercising this
branch: all tests pass just fine even when replacing it with a call to
`BUG()`. Add a test that exercises it.
Reported-by: Jeff King <peff@peff.net>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ps/leakfixes-part-10: (27 commits)
t: remove TEST_PASSES_SANITIZE_LEAK annotations
test-lib: unconditionally enable leak checking
t: remove unneeded !SANITIZE_LEAK prerequisites
t: mark some tests as leak free
t5601: work around leak sanitizer issue
git-compat-util: drop now-unused `UNLEAK()` macro
global: drop `UNLEAK()` annotation
t/helper: fix leaking commit graph in "read-graph" subcommand
builtin/branch: fix leaking sorting options
builtin/init-db: fix leaking directory paths
builtin/help: fix leaks in `check_git_cmd()`
help: fix leaking return value from `help_unknown_cmd()`
help: fix leaking `struct cmdnames`
help: refactor to not use globals for reading config
builtin/sparse-checkout: fix leaking sanitized patterns
split-index: fix memory leak in `move_cache_to_base_index()`
git: refactor builtin handling to use a `struct strvec`
git: refactor alias handling to use a `struct strvec`
strvec: introduce new `strvec_splice()` function
line-log: fix leak when rewriting commit parents
...
The rebase todo editor has commands like `fixup -c` which affects
the commit messages of the rebased commits.[1] For example:
pick hash1 <msg>
fixup hash2 <msg>
fixup -c hash3 <msg>
This says that hash2 and hash3 should be squashed into hash1 and
that hash3’s commit message should be used for the resulting commit.
So the user is presented with an editor where the two first commit
messages are commented out and the third is not. However this does
not work if `core.commentChar`/`core.commentString` is in use since
the comment char is hardcoded (#) in this `sequencer.c` function.
As a result the first commit message will not be commented out.
† 1: See 9e3cebd97c (rebase -i: add fixup [-C | -c] command,
2021-01-29)
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Co-authored-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Reported-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Kristoffer Haugsbakk <code@khaugsbakk.name>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
`git revert --reference <commit>` leaves behind a comment in the
first line:[1]
# *** SAY WHY WE ARE REVERTING ON THE TITLE LINE ***
Meaning that the commit will just consist of the next line if the user
exits the editor directly:
This reverts commit <--format=reference commit>
But the comment char here is hardcoded (#). Which means that the
comment line will inadvertently be included in the commit message if
`core.commentChar`/`core.commentString` is in use.
† 1: See 43966ab315 (revert: optionally refer to commit in the
"reference" format, 2022-05-26)
Signed-off-by: Kristoffer Haugsbakk <code@khaugsbakk.name>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
`git rebase --update-ref` does not insert commands for dependent/sub-
branches which are checked out.[1] Instead it leaves a comment about
that fact. The comment char is hardcoded (#). In turn the comment
line gets interpreted as an invalid command when `core.commentChar`/
`core.commentString` is in use.
† 1: See 900b50c242 (rebase: add --update-refs option, 2022-07-19)
Signed-off-by: Kristoffer Haugsbakk <code@khaugsbakk.name>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Both `struct block_writer` and `struct reftable_writer` have a `buf`
member that is being reused to optimize the number of allocations.
Rename the variable to `scratch` to clarify its intend and provide a
comment explaining why it exists.
Suggested-by: Christian Couder <christian.couder@gmail.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `initial_transaction` flag is tracked as a signed integer, but we
typically pass around flags via unsigned integers. Adapt the type
accordingly.
Suggested-by: Christian Couder <christian.couder@gmail.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We have recently added a new test to t7900 that exercises whether
git-maintenance(1) fails as expected when the "schedule.lock" file
exists. The test depends on whether or not the host has the required
executables present to schedule maintenance tasks in the first place,
like systemd or launchctl -- if not, the test fails with an unrelated
error before even checking for the lock file. This fails for example in
our CI systems, where macOS images do not have launchctl available.
Fix this issue by creating a stub systemctl(1) binary and using the
systemd scheduler.
Reported-by: Jeff King <peff@peff.net>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Even though `git help cli` recommends users to prefer using
"--option=value" over "--option value", there can be reasons why
giving them separately is a good idea. One reason is that shells do
not perform tilde expansion for `--option=~/path/name` but they
expand `--options ~/path/name` just fine.
This is not a problem for many options whose option parsing is
properly written using OPT_FILENAME(), because the value given to
OPT_FILENAME() is tilde-expanded internally by us, but some commands
take a pathname as a mere string, which needs this trick to have the
shell help us.
I think the reason we originally decided to recommend the stuck form
was because an option that takes an optional value requires you to
use it in the stuck form, and it is one less thing for users to
worry about if they get into the habit to always use the stuck form.
But we should be discouraging ourselves from adding an option with
an optional value in the first place, and we might want to weaken
the current recommendation.
In any case, let's describe this one case where it is necessary to
use the separate form, with an example.
Reviewed-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code clean-up.
* jk/output-prefix-cleanup:
diff: store graph prefix buf in git_graph struct
diff: return line_prefix directly when possible
diff: return const char from output_prefix callback
diff: drop line_prefix_length field
line-log: use diff_line_prefix() instead of custom helper
Doc update to clarify how periodical maintenance are scheduled,
spread across time to avoid thundering hurds.
* sk/doc-maintenance-schedule:
doc: add a note about staggering of maintenance
* 'master' of https://github.com/j6t/gitk:
Makefile(s): avoid recipe prefix in conditional statements
doc: switch links to https
doc: update links to current pages
Since the introduction of 'initialize_merge_tool' in de8dafbada
(mergetool: break setup_tool out into separate initialization function,
2021-02-09), any errors from this function are ignored in
git-difftool--helper.sh::launch_merge_tool, which is not the case for
its call in git-mergetool.sh::merge_file.
Despite the in-code comment, initialize_merge_tool (via its call to
setup_tool) does different checks than run_merge_tool, so it makes sense
to abort early if it encounters errors. Add exit calls if
initialize_merge_tool fails.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In setup_tool, we check if the given tool is a known variant of a tool,
and quietly return with an error if not. This leads to the following
invocation quietly failing:
git mergetool --tool=vimdiff4
Add an error message before returning in this case.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In git-mergetool--lib.sh::setup_tool, we check if the given tool is a
known builtin tool, a known variant, or a user-defined tool by calling
setup_user_tool, and we return with the exit code from setup_user_tool
if it was called. setup_user_tool checks if {diff,merge}tool.$tool.cmd
is set and quietly returns with an error if not.
This leads to the following invocation quietly failing:
git mergetool --tool=unknown
which is not very user-friendly. Adjust setup_tool to output an error
message before returning if setup_user_tool returned with an error.
Note that we do not check the result of the second call to
setup_user_tool in setup_tool, as this call is only meant to allow users
to redefine 'cmd' for a builtin tool; it is not an error if they have
not done so.
Note that this behaviour of quietly failing is a regression dating back
to de8dafbada (mergetool: break setup_tool out into separate
initialization function, 2021-02-09), as before this commit an unknown
mergetool would be diagnosed in get_merge_tool_path when called from
run_merge_tool.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In git-mergetool--lib.sh::get_merge_tool_path, we check if the chosen
tool is valid via valid_tool and exit with an error message if not. This
error message mentions "Unknown merge tool", even if the command the
user tried was 'git difftool --tool=unknown'. Use the global 'TOOL_MODE'
variable for a more correct error message.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When adding a remote to bare repository with "git remote add --mirror",
running fetch will fail to update HEAD to the remote's HEAD, since it
does not know how to handle bare repositories. On the other hand HEAD
already has content, since "git init --bare" has already set HEAD to
whatever is the default branch set for the user. Unless this - by chance
- is the same as the remote's HEAD, HEAD will be pointing to a bad
symref. Teach set_head to handle bare repositories, by overwriting HEAD
so it mirrors the remote's HEAD.
Note, that in this case overriding the local HEAD reference is
necessary, since HEAD will exist before fetch can be run, but this
should not be an issue, since the whole purpose of --mirror is to be an
exact mirror of the remote, so following any changes to HEAD makes
sense.
Also note, that although "git remote set-head" also fails when trying to
update the remote's locally tracked HEAD in a mirrored bare repository,
the usage of the command does not make much sense after this patch:
fetch will update the remote HEAD correctly, and setting it manually to
something else is antithetical to the concept of mirroring.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When cloning a repository remote/HEAD is created, but when the user
creates a repository with git init, and later adds a remote, remote/HEAD
is only created if the user explicitly runs a variant of "remote
set-head". Attempt to set remote/HEAD during fetch, if the user does not
have it already set. Silently ignore any errors.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Allow the caller to specify that it only wants to update the symref if
it does not already exist. Silently ignore the error from the
transaction API if the symref already exists.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently there is only one special error for transaction, for when
there is a naming conflict, all other errors are dumped under a generic
error. Add a new special error case for when the caller requests the
reference to be updated only when it does not yet exist and the
reference actually does exist.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently, set-head --auto will print a message saying "remote/HEAD set
to branch", which implies something was changed.
Change the output of --auto, so the output actually reflects what was
done: a) set a previously unset HEAD, b) change HEAD because remote
changed or c) no updates. As edge cases, if HEAD is changed from
a previous symbolic reference that was not a remote branch, explicitly
call attention to this fact, and also notify the user if the previous
reference was not a symbolic reference.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Make two different readability refactors:
Rename strbufs "buf" and "buf2" to something more explanatory.
Instead of calling get_main_ref_store(the_repository) multiple times,
call it once and store the result in a new refs variable. Although this
change probably offers some performance benefits, the main purpose is to
shorten the line lengths of function calls using this variable.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When updating a symref with update_symref it's currently not possible to
know for sure what was the previous value that was overwritten. Extend
refs_update_symref under a new function name, to record the value after
the ref has been locked if the caller of refs_update_symref_extended
requests it via a new variable in the function call. Make the return
value of the function notify the caller, if the previous value was
actually not a symbolic reference. Keep the original refs_update_symref
function with the same signature, but now as a wrapper around
refs_update_symref_extended.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When the symbolic reference we want to read with refs_read_symbolic_ref
is actually not a symbolic reference, the files and the reftable
backends return different values (1 and -1 respectively). Standardize
the returned values so that 0 is success, -1 is a generic error and -2
is that the reference was actually non-symbolic.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The test coverage was missing a test for the failure branch of remote
set-head auto's output. Add the missing text and while we are at it,
correct a small grammatical mistake in the error's output ("setup" is
the noun, "set up" is the verb).
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Consider the bare repository called "mirror" in the test. Running `git
remote add --mirror -f origin ../one` will not change HEAD, consequently
if init.defaultBranch is not the same as what HEAD in the remote
("one"), HEAD in "mirror" will be pointing to a non-existent reference.
Hence if "mirror" is used as a remote by yet another repository,
ls-remote will not show HEAD. On the other hand, if init.defaultBranch
happens to match HEAD in "one", then ls-remote will show HEAD.
Since the "ci/run-build-and-tests.sh" script globally exports
GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME=main for some (but not all) jobs,
there may be a drift in some tests between how the test repositories are
set up in the CI and during local testing, if the test itself uses
"master" as default instead of "main". In particular, this happens in
t5505-remote.sh. This issue does not manifest currently, as the test
does not do any remote HEAD manipulation where this would come up, but
should such things be added, a locally passing test would break the CI
and vice-versa.
Set GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME=main in t5505-remote to be
consistent with the CI.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
