"git fetch" can clobber a symref that is dangling when the
remote-tracking HEAD is set to auto update, which has been
corrected.
* jk/no-clobber-dangling-symref-with-fetch:
refs: do not clobber dangling symrefs
t5510: prefer "git -C" to subshell for followRemoteHEAD tests
t5510: stop changing top-level working directory
t5510: make confusing config cleanup more explicit
Code clean-ups.
* ps/reftable-libgit2-cleanup:
refs/reftable: always reload stacks when creating lock
reftable: don't second-guess errors from flock interface
reftable/stack: handle outdated stacks when compacting
reftable/stack: allow passing flags to `reftable_stack_add()`
reftable/stack: fix compiler warning due to missing braces
reftable/stack: reorder code to avoid forward declarations
reftable/writer: drop Git-specific `QSORT()` macro
reftable/writer: fix type used for number of records
"git remote rename origin upstream" failed to move origin/HEAD to
upstream/HEAD when origin/HEAD is unborn and performed other
renames extremely inefficiently, which has been corrected.
* ps/remote-rename-fix:
builtin/remote: only iterate through refs that are to be renamed
builtin/remote: rework how remote refs get renamed
builtin/remote: determine whether refs need renaming early on
builtin/remote: fix sign comparison warnings
refs: simplify logic when migrating reflog entries
refs: pass refname when invoking reflog entry callback
"git refs migrate" to migrate the reflog entries from a refs
backend to another had a handful of bugs squashed.
* ps/reflog-migrate-fixes:
refs: fix invalid old object IDs when migrating reflogs
refs: stop unsetting REF_HAVE_OLD for log-only updates
refs/files: detect race when generating reflog entry for HEAD
refs: fix identity for migrated reflogs
ident: fix type of string length parameter
builtin/reflog: implement subcommand to write new entries
refs: export `ref_transaction_update_reflog()`
builtin/reflog: improve grouping of subcommands
Documentation/git-reflog: convert to use synopsis type
When given an expected "before" state, the ref-writing code will avoid
overwriting any ref that does not match that expected state. We use the
null oid as a sentinel value for "nothing should exist", and likewise
that is the sentinel value we get when trying to read a ref that does
not exist.
But there's one corner case where this is ambiguous: dangling symrefs.
Trying to read them will yield the null oid, but there is potentially
something of value there: the dangling symref itself.
For a normal recursive write, this is OK. Imagine we have a symref
"FOO_HEAD" that points to a ref "refs/heads/bar" that does not exist,
and we try to write to it with a create operation like:
oid=$(git rev-parse HEAD) ;# or whatever
git symbolic-ref FOO_HEAD refs/heads/bar
echo "create FOO_HEAD $oid" | git update-ref --stdin
The attempt to resolve FOO_HEAD will actually resolve "bar", yielding
the null oid. That matches our expectation, and the write proceeds. This
is correct, because we are not writing FOO_HEAD at all, but writing its
destination "bar", which in fact does not exist.
But what if the operation asked not to dereference symrefs? Like this:
echo "create FOO_HEAD $oid" | git update-ref --no-deref --stdin
Resolving FOO_HEAD would still result in a null oid, and the write will
proceed. But it will overwrite FOO_HEAD itself, removing the fact that
it ever pointed to "bar".
This case is a little esoteric; we are clobbering a symref with a
no-deref write of a regular ref value. But the same problem occurs when
writing symrefs. For example:
echo "symref-create FOO_HEAD refs/heads/other" |
git update-ref --no-deref --stdin
The "create" operation asked us to create FOO_HEAD only if it did not
exist. But we silently overwrite the existing value.
You can trigger this without using update-ref via the fetch
followRemoteHEAD code. In "create" mode, it should not overwrite an
existing value. But if you manually create a symref pointing to a value
that does not yet exist (either via symbolic-ref or with "remote add
-m"), create mode will happily overwrite it.
Instead, we should detect this case and refuse to write. The correct
specification to overwrite FOO_HEAD in this case is to provide an
expected target ref value, like:
echo "symref-update FOO_HEAD refs/heads/other ref refs/heads/bar" |
git update-ref --no-deref --stdin
Note that the non-symref "update" directive does not allow you to do
this (you can only specify an oid). This is a weakness in the update-ref
interface, and you'd have to overwrite unconditionally, like:
echo "update FOO_HEAD $oid" | git update-ref --no-deref --stdin
Likewise other symref operations like symref-delete do not accept the
"ref" keyword. You should be able to do:
echo "symref-delete FOO_HEAD ref refs/heads/bar"
but cannot (and can only delete unconditionally). This patch doesn't
address those gaps. We may want to do so in a future patch for
completeness, but it's not clear if anybody actually wants to perform
those operations. The symref update case (specifically, via
followRemoteHEAD) is what I ran into in the wild.
The code for the fix is relatively straight-forward given the discussion
above. But note that we have to implement it independently for the files
and reftable backends. The "old oid" checks happen as part of the
locking process, which is implemented separately for each system. We may
want to factor this out somehow, but it's beyond the scope of this
patch. (Another curiosity is that the messages in the reftable code are
marked for translation, but the ones in the files backend are not. I
followed local convention in each case, but we may want to harmonize
this at some point).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When creating a new addition via either `reftable_stack_new_addition()`
or its convenince wrapper `reftable_stack_add()` we:
1. Create the "tables.list.lock" file.
2. Verify that the current version of the "tables.list" file is
up-to-date.
3. Write the new table records if so.
By default, the second step would cause us to bail out if we see that
there has been a concurrent write to the stack that made our in-memory
copy of the stack out-of-date. This is a safety mechanism to not write
records to the stack based on outdated information.
The downside though is that concurrent writes may now cause us to bail
out, which is not a good user experience. In addition, this isn't even
necessary for us, as Git knows to perform all checks for the old state
of references under the lock. (Well, in all except one case: when we
expire the reflog we first create the log iterator before we create the
lock, but this ordering is fixed as part of this commit.)
Consequently, most writers pass the `REFTABLE_STACK_NEW_ADDITION_RELOAD`
flag. The effect of this flag is that we reload the stack after having
acquired the lock in case the stack is out-of-date. This plugs the race
with concurrent writers, but we continue performing the verifications of
the expected old state to catch actual conflicts in the references we
are about to write.
Adapt the remaining callsites that don't yet pass this flag to do so.
While at it, drop a needless manual reload.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `reftable_stack_add()` function is a simple wrapper to lock the
stack, add records to it via a callback and then commit the
result. One problem with it though is that it doesn't accept any flags
for creating the addition. This makes it impossible to automatically
reload the stack in case it was modified before we managed to lock the
stack.
Add a `flags` field to plug this gap and pass it through accordingly.
For now this new flag won't be used by us, but it will be used by
libgit2.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
With `refs_for_each_reflog_ent()` callers can iterate through all the
reflog entries for a given reference. The callback that is being invoked
for each such entry does not receive the name of the reference that we
are currently iterating through. This isn't really a limiting factor, as
callers can simply pass the name via the callback data.
But this layout sometimes does make for a bit of an awkward calling
pattern. One example: when iterating through all reflogs, and for each
reflog we iterate through all refnames, we have to do some extra book
keeping to track which reference name we are currently yielding reflog
entries for.
Change the signature of the callback function so that the reference name
of the reflog gets passed through to it. Adapt callers accordingly and
start using the new parameter in trivial cases. The next commit will
refactor the reference migration logic to make use of this parameter so
that we can simplify its logic a bit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ps/reflog-migrate-fixes:
refs: fix invalid old object IDs when migrating reflogs
refs: stop unsetting REF_HAVE_OLD for log-only updates
refs/files: detect race when generating reflog entry for HEAD
refs: fix identity for migrated reflogs
ident: fix type of string length parameter
builtin/reflog: implement subcommand to write new entries
refs: export `ref_transaction_update_reflog()`
builtin/reflog: improve grouping of subcommands
Documentation/git-reflog: convert to use synopsis type
When migrating reflog entries between different storage formats we end
up with invalid old object IDs for the migrated entries: instead of
writing the old object ID of the to-be-migrated entry, we end up with
the all-zeroes object ID.
The root cause of this issue is that we don't know to use the old object
ID provided by the caller. Instead, we manually resolve the old object
ID by resolving the current value of its matching reference. But as that
reference does not yet exist in the target ref storage we always end up
resolving it to all-zeroes.
This issue got unnoticed as there is no user-facing command that would
even show the old object ID. While `git log -g` knows to show the new
object ID, we don't have any formatting directive to show the old object
ID.
Fix the bug by introducing a new flag `REF_LOG_USE_PROVIDED_OIDS`. If
set, backends are instructed to use the old and new object IDs provided
by the caller, without doing any manual resolving. Set this flag in
`ref_transaction_update_reflog()`.
Amend our tests in t1460-refs-migrate to use our test tool to read
reflog entries. This test tool prints out both old and new object ID of
each reflog entry, which fixes the test gap. Furthermore it also prints
the full identity used to write the reflog, which provides test coverage
for the previous commit in this patch series that fixed the identity for
migrated reflogs.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `REF_HAVE_OLD` flag indicates whether a given ref update has its old
object ID set. If so, the value of that field is used to verify whether
the current state of the reference matches this expected state. It is
thus an important part of mitigating races with a concurrent process
that updates the same set of references.
When writing reflogs though we explicitly unset that flag. This is a
sensible thing to do: the old state of reflog entry updates may not
necessarily match the current on-disk state of its accompanying ref, but
it's only intended to signal what old object ID we want to write into
the new reflog entry. For example when migrating refs we end up writing
many reflog entries for a single reference, and most likely those reflog
entries will have many different old object IDs.
But unsetting this flag also removes a useful signal, namely that the
caller _did_ provide an old object ID for a given reflog entry. This
signal will become useful in a subsequent commit, where we add a new
flag that tells the transaction to use the provided old and new object
IDs to write a reflog entry. The `REF_HAVE_OLD` flag is then used as a
signal to verify that the caller really did provide an old object ID.
Stop unsetting the flag so that we can use it as this described signal
in a subsequent commit. Skip checking the old object ID for log-only
updates so that we don't expect it to match the current on-disk state.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When updating a reference that is being pointed to HEAD we don't only
write a reflog message for that particular reference, but also generate
one for HEAD. This logic is handled by `split_head_update()`, where we:
1. Verify that the condition actually triggered. This is done by
reading HEAD at the start of the transaction so that we can then
check whether a given reference update refers to its target.
2. Queue a new log-only update for HEAD in case it did.
But the logic is unfortunately not free of races, as we do not lock the
HEAD reference after we have read its target. This can lead to the
following two scenarios:
- HEAD gets concurrently updated to point to one of the references we
have already processed. This causes us not writing a reflog message
even though we should have done so.
- HEAD gets concurrently updated to no longer point to a reference
anymore that we have already processed. This causes us to write a
reflog message even though we should _not_ have done so.
Improve the situation by introducing a new `REF_LOG_VIA_SPLIT` flag that
is specific to the "files" backend. If set, we will double check that
the HEAD reference still points to the reference that we are creating
the reflog entry for after we have locked HEAD. Furthermore, instead of
manually resolving the old object ID of that entry, we now use the same
old state as for the parent update.
If we detect such a racy update we abort the transaction. This is a bit
heavy-handed: the user didn't even ask us to write a reflog update for
"HEAD", so it might be surprising if we abort the transaction. That
being said:
- Normal users wouldn't typically hit this case as we only hit the
relevant code when committing to a branch that is being pointed to
by "HEAD" directly. Commands like git-commit(1) typically commit to
"HEAD" itself though.
- Scripted users that use git-update-ref(1) and related plumbing
commands are unlikely to hit this case either, as they would have to
update the pointed-to-branch at the same as "HEAD" is being updated,
which is an exceedingly rare event.
The alternative would be to instead drop the log-only update completely,
but that would require more logic that is hard to verify without adding
infrastructure specific for such a test. So we rather do the pragmatic
thing and don't worry too much about an edge case that is very unlikely
to happen.
Unfortunately, this change only helps with the second race. We cannot
reliably plug the first race without locking the HEAD reference at the
start of the transaction. Locking HEAD unconditionally would effectively
serialize all writes though, and that doesn't seem like an option. Also,
double checking its value at the end of the transaction is not an option
either, as its target may have flip-flopped during the transaction.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The config API had a set of convenience wrapper functions that
implicitly use the_repository instance; they have been removed and
inlined at the calling sites.
* ps/config-wo-the-repository: (21 commits)
config: fix sign comparison warnings
config: move Git config parsing into "environment.c"
config: remove unused `the_repository` wrappers
config: drop `git_config_set_multivar()` wrapper
config: drop `git_config_get_multivar_gently()` wrapper
config: drop `git_config_set_multivar_in_file_gently()` wrapper
config: drop `git_config_set_in_file_gently()` wrapper
config: drop `git_config_set()` wrapper
config: drop `git_config_set_gently()` wrapper
config: drop `git_config_set_in_file()` wrapper
config: drop `git_config_get_bool()` wrapper
config: drop `git_config_get_ulong()` wrapper
config: drop `git_config_get_int()` wrapper
config: drop `git_config_get_string()` wrapper
config: drop `git_config_get_string()` wrapper
config: drop `git_config_get_string_multi()` wrapper
config: drop `git_config_get_value()` wrapper
config: drop `git_config_get_value()` wrapper
config: drop `git_config_get()` wrapper
config: drop `git_config_clear()` wrapper
...
Code clean-up.
* kn/for-each-ref-skip-updates:
ref-filter: use REF_ITERATOR_SEEK_SET_PREFIX instead of '1'
t6302: add test combining '--start-after' with '--exclude'
for-each-ref: reword the documentation for '--start-after'
for-each-ref: fix documentation argument ordering
ref-cache: use 'size_t' instead of int for length
"git for-each-ref" learns "--start-after" option to help
applications that want to page its output.
* kn/for-each-ref-skip:
ref-cache: set prefix_state when seeking
for-each-ref: introduce a '--start-after' option
ref-filter: remove unnecessary else clause
refs: selectively set prefix in the seek functions
ref-cache: remove unused function 'find_ref_entry()'
refs: expose `ref_iterator` via 'refs.h'
The commit 090eb5336c (refs: selectively set prefix in the seek
functions, 2025-07-15) modified the ref-cache iterator to support
seeking to a specified marker without setting the prefix.
The commit adds and uses an integer 'len' to capture the length of the
seek marker to compare with the entries of a given directory. Since the
type of the variable is 'int', this is met with a typecast of converting
a `strlen` to 'int' so it can be assigned to the 'len' variable.
This is whole operation is a bit wrong:
1. Since the 'len' variable is eventually used in a 'strncmp', it should
have been of type 'size_t'.
2. This also truncates the value provided from 'strlen' to an int, which
could cause a large refname to produce a negative number.
Let's do the correct thing here and simply use 'size_t' for `len`.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 090eb5336c (refs: selectively set prefix in the seek functions,
2025-07-15) we separated the seeking functionality of reference
iterators from the functionality to set prefix to an iterator. This
allows users of ref iterators to seek to a particular reference to
provide pagination support.
The files-backend, uses the ref-cache iterator to iterate over loose
refs. The iterator tracks directories and entries already processed via
a stack of levels. Each level corresponds to a directory under the files
backend. New levels are added to the stack, and when all entries from a
level is yielded, the corresponding level is popped from the stack.
To accommodate seeking, we need to populate and traverse the levels to
stop the requested seek marker at the appropriate level and its entry
index. Each level also contains a 'prefix_state' which is used for
prefix matching, this allows the iterator to skip levels/entries which
don't match a prefix. The default value of 'prefix_state' is
PREFIX_CONTAINS_DIR, which yields all entries within a level. When
purely seeking without prefix matching, we want to yield all entries.
The commit however, skips setting the value explicitly. This causes the
MemorySanitizer to issue a 'use-of-uninitialized-value' error when
running 't/t6302-for-each-ref-filter'.
Set the value explicitly to avoid to fix the issue.
Reported-by: Kyle Lippincott <spectral@google.com>
Helped-by: Kyle Lippincott <spectral@google.com>
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 036876a106 (config: hide functions using `the_repository` by
default, 2024-08-13) we have moved around a bunch of functions in the
config subsystem that depend on `the_repository`. Those function have
been converted into mere wrappers around their equivalent function that
takes in a repository as parameter, and the intent was that we'll
eventually remove those wrappers to make the dependency on the global
repository variable explicit at the callsite.
Follow through with that intent and remove `git_config_get_int()`. All
callsites are adjusted so that they use
`repo_config_get_int(the_repository, ...)` instead. While some callsites
might already have a repository available, this mechanical conversion is
the exact same as the current situation and thus cannot cause any
regression. Those sites should eventually be cleaned up in a later patch
series.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 036876a106 (config: hide functions using `the_repository` by
default, 2024-08-13) we have moved around a bunch of functions in the
config subsystem that depend on `the_repository`. Those function have
been converted into mere wrappers around their equivalent function that
takes in a repository as parameter, and the intent was that we'll
eventually remove those wrappers to make the dependency on the global
repository variable explicit at the callsite.
Follow through with that intent and remove `git_config()`. All callsites
are adjusted so that they use `repo_config(the_repository, ...)`
instead. While some callsites might already have a repository available,
this mechanical conversion is the exact same as the current situation
and thus cannot cause any regression. Those sites should eventually be
cleaned up in a later patch series.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When a ref creation at refs/heads/foo/bar fails, the files backend
now removes refs/heads/foo/ if the directory is otherwise not used.
* ps/refs-files-remove-empty-parent:
refs/files: remove empty parent dirs when ref creation fails
The ref iterator exposes a `ref_iterator_seek()` function. The name
suggests that this would seek the iterator to a specific reference in
some ways similar to how `fseek()` works for the filesystem.
However, the function actually sets the prefix for refs iteration. So
further iteration would only yield references which match the particular
prefix. This is a bit confusing.
Let's add a 'flags' field to the function, which when set with the
'REF_ITERATOR_SEEK_SET_PREFIX' flag, will set the prefix for the
iteration in-line with the existing behavior. Otherwise, the reference
backends will simply seek to the specified reference and clears any
previously set prefix. This allows users to start iteration from a
specific reference.
In the packed and reftable backend, since references are available in a
sorted list, the changes are simply setting the prefix if needed. The
changes on the files-backend are a little more involved, since the files
backend uses the 'ref-cache' mechanism. We move out the existing logic
within `cache_ref_iterator_seek()` to `cache_ref_iterator_set_prefix()`
which is called when the 'REF_ITERATOR_SEEK_SET_PREFIX' flag is set. We
then parse the provided seek string and set the required levels and
their indexes to ensure that seeking is possible.
Helped-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The 'find_ref_entry' function is no longer used, so remove it.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `ref_iterator` is an internal structure to the 'refs/'
sub-directory, which allows iteration over refs. All reference iteration
is built on top of these iterators.
External clients of the 'refs' subsystem use the various
'refs_for_each...()' functions to iterate over refs. However since these
are wrapper functions, each combination of functionality requires a new
wrapper function. This is not feasible as the functions pile up with the
increase in requirements. Expose the internal reference iterator, so
advanced users can mix and match options as needed.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git push" and "git fetch" are taught to update refs in batches to
gain performance.
* kn/fetch-push-bulk-ref-update:
receive-pack: handle reference deletions separately
refs/files: skip updates with errors in batched updates
receive-pack: use batched reference updates
send-pack: fix memory leak around duplicate refs
fetch: use batched reference updates
refs: add function to translate errors to strings
When creating a new reference in the "files" backend we first create the
directory hierarchy for that reference, then create the lockfile for
that reference, and finally rename the lockfile into place. When the
transaction gets aborted we prune the lockfile, but we don't clean up
the directory hierarchy that we may have created for the lockfile.
In some egde cases this can lead to lots of empty directories being
cluttered in the ".git/refs" directory that really serve no purpose at
all. We know to prune such empty directories when packing refs, but that
only patches over the issue.
Improve this by removing empty parents when cleaning up still-locked
references in `files_transaction_cleanup()`. This function is also
called when preparing or committing the transaction, so this change also
helps when not explicitly aborting the transaction.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The commit 23fc8e4f61 (refs: implement batch reference update support,
2025-04-08) introduced support for batched reference updates. This
allows users to batch updates together, while allowing some of the
updates to fail.
Under the hood, batched updates use the reference transaction mechanism.
Each update which fails is marked as such. Any failed updates must be
skipped over in the rest of the code, as they wouldn't apply any more.
In two of the loops within 'files_transaction_finish()' of the files
backend, the failed updates aren't skipped over. This can cause a
SEGFAULT otherwise. Add the missing skips and a test to validate the
same.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git verify-refs" (and hence "git fsck --reference") started
erroring out in a repository in which secondary worktrees were
prepared with Git 2.43 or lower.
* sj/ref-contents-check-fix:
fsck: ignore missing "refs" directory for linked worktrees
"git refs verify" doesn't work if there are worktrees created on Git
v2.43.0 or older versions. These versions don't automatically create the
"refs" directory, causing the error:
error: cannot open directory .git/worktrees/<worktree name>/refs:
No such file or directory
Since 8f4c00de95 (builtin/worktree: create refdb via ref backend,
2024-01-08), we automatically create the "refs" directory for new
worktrees. And in 7c78d819e6 (ref: support multiple worktrees check for
refs, 2024-11-20), we assume that all linked worktrees have this
directory and would wrongly report an error to the user, thus
introducing compatibility issue.
Check for ENOENT errno before reporting directory access errors for
linked worktrees to maintain backward compatibility.
Reported-by: Kristoffer Haugsbakk <code@khaugsbakk.name>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
During fsck, we use "strbuf_read" to read the content of "packed-refs"
without using mmap mechanism. This is a bad practice which would consume
more memory than using mmap mechanism. Besides, as all code paths in
"packed-backend.c" use this way, we should make "fsck" align with the
current codebase.
As we have introduced the helper function "allocate_snapshot_buffer", we
can simply use this function to use mmap mechanism.
Suggested-by: Jeff King <peff@peff.net>
Suggested-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"load_contents" would choose which way to load the content of the
"packed-refs". However, we cannot directly use this function when
checking the consistency due to we don't want to open the file. And we
also need to reuse the logic to avoid causing repetition.
Let's create a new helper function "allocate_snapshot_buffer" to extract
the snapshot allocation logic in "load_contents" and update the
"load_contents" to align with the behavior.
Suggested-by: Jeff King <peff@peff.net>
Suggested-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We assume the "packed-refs" won't be empty and instead has at least one
line in it (even when there are no refs packed, there is the file header
line). Because there is no terminating LF in the empty file, we will
report "packedRefEntryNotTerminated(ERROR)" to the user.
However, the runtime code paths would accept an empty "packed-refs"
file, for example, "create_snapshot" would simply return the "snapshot"
without checking the content of "packed-refs". So, we should skip
checking the content of "packed-refs" when it is empty during fsck.
After 694b7a1999 (repack_without_ref(): write peeled refs in the
rewritten file, 2013-04-22), we would always write a header into the
"packed-refs" file. So, versions of Git that are not too ancient never
write such an empty "packed-refs" file.
As an empty file often indicates a sign of a filesystem-level issue, the
way we want to resolve this inconsistency is not make everybody totally
silent but notice and report the anomaly.
Let's create a "FSCK_INFO" message id "EMPTY_PACKED_REFS_FILE" to report
to the users that "packed-refs" is empty.
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code clean-up.
* ps/object-file-cleanup:
object-store: merge "object-store-ll.h" and "object-store.h"
object-store: remove global array of cached objects
object: split out functions relating to object store subsystem
object-file: drop `index_blob_stream()`
object-file: split up concerns of `HASH_*` flags
object-file: split out functions relating to object store subsystem
object-file: move `xmmap()` into "wrapper.c"
object-file: move `git_open_cloexec()` to "compat/open.c"
object-file: move `safe_create_leading_directories()` into "path.c"
object-file: move `mkdir_in_gitdir()` into "path.c"
Incorrect sorting of refs with bytes with high-bit set on platforms
with signed char led to a BUG, which has been corrected.
* ps/refname-avail-check-optim:
refs/packed: fix BUG when seeking refs with UTF-8 characters
Updating multiple references have only been possible in all-or-none
fashion with transactions, but it can be more efficient to batch
multiple updates even when some of them are allowed to fail in a
best-effort manner. A new "best effort batches of updates" mode
has been introduced.
* kn/non-transactional-batch-updates:
update-ref: add --batch-updates flag for stdin mode
refs: support rejection in batch updates during F/D checks
refs: implement batch reference update support
refs: introduce enum-based transaction error types
refs/reftable: extract code from the transaction preparation
refs/files: remove duplicate duplicates check
refs: move duplicate refname update check to generic layer
refs/files: remove redundant check in split_symref_update()
The object layer has been updated to take an explicit repository
instance as a parameter in more code paths.
* ps/object-wo-the-repository:
hash: stop depending on `the_repository` in `null_oid()`
hash: fix "-Wsign-compare" warnings
object-file: split out logic regarding hash algorithms
delta-islands: stop depending on `the_repository`
object-file-convert: stop depending on `the_repository`
pack-bitmap-write: stop depending on `the_repository`
pack-revindex: stop depending on `the_repository`
pack-check: stop depending on `the_repository`
environment: move access to "core.bigFileThreshold" into repo settings
pack-write: stop depending on `the_repository` and `the_hash_algo`
object: stop depending on `the_repository`
csum-file: stop depending on `the_repository`
The `safe_create_leading_directories()` function and its relatives are
located in "object-file.c", which is not a good fit as they provide
generic functionality not related to objects at all. Move them into
"path.c", which already hosts `safe_create_dir()` and its relative
`safe_create_dir_in_gitdir()`.
"path.c" is free of `the_repository`, but the moved functions depend on
`the_repository` to read the "core.sharedRepository" config. Adapt the
function signature to accept a repository as argument to fix the issue
and adjust callers accordingly.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It was reported that using git-pull(1) in a repository whose remote
contains branches with emojis leads to the following bug:
$ git pull
remote: Enumerating objects: 161255, done.
remote: Counting objects: 100% (55884/55884), done.
remote: Compressing objects: 100% (5518/5518), done.
remote: Total 161255 (delta 54253), reused 50509 (delta 50364),
pack-reused 105371 (from 4)
Receiving objects: 100% (161255/161255), 309.90 MiB | 16.87 MiB/s, done.
Resolving deltas: 100% (118048/118048), completed with 13416 local objects.
From github.com:github/github
97ab7ae3f3745..8fb2f9fa180ed master -> origin/master
[...snip many screenfuls of updates to origin remotes...]
BUG: refs/packed-backend.c:984: packed-refs backend yielded reference
preceding its prefix
error: fetch died of signal 6
This issue bisects to 22600c0452 (refs/iterator: implement seeking for
packed-ref iterators, 2025-03-12) where we have implemented seeking for
the packed-ref iterator. As part of that change we introduced a check
that verifies that the iterator only returns refnames bigger than the
prefix. In theory, this check should always hold: when a prefix is set
we know that we would've seeked that prefix first, so we should never
see a reference sorting before that prefix.
But in practice the check itself is misbehaving when handling unicode
characters. The particular issue triggered with a branch that got the
"shaved ice" unicode character in its name, which is composed of the
bytes "0xEE 0x90 0xBF". The bug triggers when we compare the refname
"refs/heads/<shaved-ice>" to something like "refs/heads/z", and it
specifically hits when comparing the first byte, "0xEE".
The root cause is that the most-significant bit of 0xEE is set. The
`refname` and `prefix` pointers that we use to compare bytes with one
another are both pointers to signed characters. As such, when we
dereference the 0xEE byte the result is a _negative_ value, and this
value will of course compare smaller than "z".
We can see that this issue is avoided in `cmp_packed_refname()`, where
we explicitly cast each byte to its unsigned form. Fix the bug by doing
the same in `packed_ref_iterator_advance()`.
Reported-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ps/object-wo-the-repository:
hash: stop depending on `the_repository` in `null_oid()`
hash: fix "-Wsign-compare" warnings
object-file: split out logic regarding hash algorithms
delta-islands: stop depending on `the_repository`
object-file-convert: stop depending on `the_repository`
pack-bitmap-write: stop depending on `the_repository`
pack-revindex: stop depending on `the_repository`
pack-check: stop depending on `the_repository`
environment: move access to "core.bigFileThreshold" into repo settings
pack-write: stop depending on `the_repository` and `the_hash_algo`
object: stop depending on `the_repository`
csum-file: stop depending on `the_repository`
The `refs_verify_refnames_available()` is used to batch check refnames
for F/D conflicts. While this is the more performant alternative than
its individual version, it does not provide rejection capabilities on a
single update level. For batched updates, this would mean a rejection of
the entire transaction whenever one reference has a F/D conflict.
Modify the function to call `ref_transaction_maybe_set_rejected()` to
check if a single update can be rejected. Since this function is only
internally used within 'refs/' and we want to pass in a `struct
ref_transaction *` as a variable. We also move and mark
`refs_verify_refnames_available()` to 'refs-internal.h' to be an
internal function.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Git supports making reference updates with or without transactions.
Updates with transactions are generally better optimized. But
transactions are all or nothing. This means, if a user wants to batch
updates to take advantage of the optimizations without the hard
requirement that all updates must succeed, there is no way currently to
do so. Particularly with the reftable backend where batching multiple
reference updates is more efficient than performing them sequentially.
Introduce batched update support with a new flag,
'REF_TRANSACTION_ALLOW_FAILURE'. Batched updates while different from
transactions, use the transaction infrastructure under the hood. When
enabled, this flag allows individual reference updates that would
typically cause the entire transaction to fail due to non-system-related
errors to be marked as rejected while permitting other updates to
proceed. System errors referred by 'REF_TRANSACTION_ERROR_GENERIC'
continue to result in the entire transaction failing. This approach
enhances flexibility while preserving transactional integrity where
necessary.
The implementation introduces several key components:
- Add 'rejection_err' field to struct `ref_update` to track failed
updates with failure reason.
- Add a new struct `ref_transaction_rejections` and a field within
`ref_transaction` to this struct to allow quick iteration over
rejected updates.
- Modify reference backends (files, packed, reftable) to handle
partial transactions by using `ref_transaction_set_rejected()`
instead of failing the entire transaction when
`REF_TRANSACTION_ALLOW_FAILURE` is set.
- Add `ref_transaction_for_each_rejected_update()` to let callers
examine which updates were rejected and why.
This foundational change enables batched update support throughout the
reference subsystem. A following commit will expose this capability to
users by adding a `--batch-updates` flag to 'git-update-ref(1)',
providing both a user-facing feature and a testable implementation.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Replace preprocessor-defined transaction errors with a strongly-typed
enum `ref_transaction_error`. This change:
- Improves type safety and function signature clarity.
- Makes error handling more explicit and discoverable.
- Maintains existing error cases, while adding new error cases for
common scenarios.
This refactoring paves the way for more comprehensive error handling
which we will utilize in the upcoming commits to add batch reference
update support.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Extract the core logic for preparing individual reference updates from
`reftable_be_transaction_prepare()` into `prepare_single_update()`. This
dedicated function now handles all validation and preparation steps for
each reference update in the transaction, including object ID
verification, HEAD reference handling, and symref processing.
The refactoring consolidates all reference update validation into a
single logical block, which improves code maintainability and
readability. More importantly, this restructuring lays the groundwork
for implementing batched reference update support in the reftable
backend, which will be introduced in a followup commit.
No functional changes are included in this commit - it is purely a code
reorganization to support future enhancements.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Within the files reference backend's transaction's 'finish' phase, a
verification step is currently performed wherein the refnames list is
sorted and examined for multiple updates targeting the same refname.
It has been observed that this verification is redundant, as an
identical check is already executed during the transaction's 'prepare'
stage. Since the refnames list remains unmodified following the
'prepare' stage, this secondary verification can be safely eliminated.
The duplicate check has been removed accordingly, and the
`ref_update_reject_duplicates()` function has been marked as static, as
its usage is now confined to 'refs.c'.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the tracking of refnames in `affected_refnames` from individual
backends into the generic layer in 'refs.c'. This centralizes the
duplicate refname detection that was previously handled separately by
each backend.
Make some changes to accommodate this move:
- Add a `string_list` field `refnames` to `ref_transaction` to contain
all the references in a transaction. This field is updated whenever
a new update is added via `ref_transaction_add_update`, so manual
additions in reference backends are dropped.
- Modify the backends to use this field internally as needed. The
backends need to check if an update for refname already exists when
splitting symrefs or adding an update for 'HEAD'.
- In the reftable backend, within `reftable_be_transaction_prepare()`,
move the `string_list_has_string()` check above
`ref_transaction_add_update()`. Since `ref_transaction_add_update()`
automatically adds the refname to `transaction->refnames`,
performing the check after will always return true, so we perform
the check before adding the update.
This helps reduce duplication of functionality between the backends and
makes it easier to make changes in a more centralized manner.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In `split_symref_update()`, there were two checks for duplicate
refnames:
- At the start, `string_list_has_string()` ensures the refname is not
already in `affected_refnames`, preventing duplicates from being
added.
- After adding the refname, another check verifies whether the newly
inserted item has a `util` value.
The second check is unnecessary because the first one guarantees that
`string_list_insert()` will never encounter a preexisting entry.
The `item->util` field is assigned to validate that a rename doesn't
already exist in the list. The validation is done after the first check.
As this check is removed, clean up the validation and the assignment of
this field in `split_head_update()` and `files_transaction_prepare()`.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The code paths to check whether a refname X is available (by seeing
if another ref X/Y exists, etc.) have been optimized.
* ps/refname-avail-check-optim:
refs: reuse iterators when determining refname availability
refs/iterator: implement seeking for files iterators
refs/iterator: implement seeking for packed-ref iterators
refs/iterator: implement seeking for ref-cache iterators
refs/iterator: implement seeking for reftable iterators
refs/iterator: implement seeking for merged iterators
refs/iterator: provide infrastructure to re-seek iterators
refs/iterator: separate lifecycle from iteration
refs: stop re-verifying common prefixes for availability
refs/files: batch refname availability checks for initial transactions
refs/files: batch refname availability checks for normal transactions
refs/reftable: batch refname availability checks
refs: introduce function to batch refname availability checks
builtin/update-ref: skip ambiguity checks when parsing object IDs
object-name: allow skipping ambiguity checks in `get_oid()` family
object-name: introduce `repo_get_oid_with_flags()`
"git fsck" becomes more careful when checking the refs.
* sj/ref-consistency-checks-more:
builtin/fsck: add `git refs verify` child process
packed-backend: check whether the "packed-refs" is sorted
packed-backend: add "packed-refs" entry consistency check
packed-backend: check whether the refname contains NUL characters
packed-backend: add "packed-refs" header consistency check
packed-backend: check if header starts with "# pack-refs with: "
packed-backend: check whether the "packed-refs" is regular file
builtin/refs: get worktrees without reading head information
t0602: use subshell to ensure working directory unchanged
Implement seeking for "files" iterators. As we simply use a ref-cache
iterator under the hood the implementation is straight-forward. Note
that we do not implement seeking on reflog iterators, same as with the
"reftable" backend.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Implement seeking of `packed-ref` iterators. The implementation is again
straight forward, except that we cannot continue to use the prefix
iterator as we would otherwise not be able to reseek the iterator
anymore in case one first asks for an empty and then for a non-empty
prefix. Instead, we open-code the logic to in `advance()`.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Implement seeking of ref-cache iterators. This is done by splitting most
of the logic to seek iterators out of `cache_ref_iterator_begin()` and
putting it into `cache_ref_iterator_seek()` so that we can reuse the
logic.
Note that we cannot use the optimization anymore where we return an
empty ref iterator when there aren't any references, as otherwise it
wouldn't be possible to reseek the iterator to a different prefix that
may exist. This shouldn't be much of a performance concern though as we
now start to bail out early in case `advance()` sees that there are no
more directories to be searched.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Implement seeking of reftable iterators. As the low-level reftable
iterators already support seeking this change is straight-forward. Two
notes though:
- We do not support seeking on reflog iterators. It is unclear what
seeking would even look like in this context, as you typically would
want to seek to a specific entry in the reflog for a specific ref.
There is currently no use case for this, but if one arises in the
future, we can still implement seeking at that later point.
- We start to check whether `reftable_stack_init_ref_iterator()` is
successful.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Implement seeking on merged iterators. The implementation is rather
straight forward, with the only exception that we must not deallocate
the underlying iterators once they have been exhausted.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Reftable iterators need to be scrapped after they have either been
exhausted or aren't useful to the caller anymore, and it is explicitly
not possible to reuse them for iterations. But enabling for reuse of
iterators may allow us to tune them by reusing internal state of an
iterator. The reftable iterators for example can already be reused
internally, but we're not able to expose this to any users outside of
the reftable backend.
Introduce a new `.seek` function in the ref iterator vtable that allows
callers to seek an iterator multiple times. It is expected to be
functionally the same as calling `refs_ref_iterator_begin()` with a
different (or the same) prefix.
Note that it is not possible to adjust parameters other than the seeked
prefix for now, so exclude patterns, trimmed prefixes and flags will
remain unchanged. We do not have a usecase for changing these parameters
right now, but if we ever find one we can adapt accordingly.
Implement the callback for trivial cases. The other iterators will be
implemented in subsequent commits.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The ref and reflog iterators have their lifecycle attached to iteration:
once the iterator reaches its end, it is automatically released and the
caller doesn't have to care about that anymore. When the iterator should
be released before it has been exhausted, callers must explicitly abort
the iterator via `ref_iterator_abort()`.
This lifecycle is somewhat unusual in the Git codebase and creates two
problems:
- Callsites need to be very careful about when exactly they call
`ref_iterator_abort()`, as calling the function is only valid when
the iterator itself still is. This leads to somewhat awkward calling
patterns in some situations.
- It is impossible to reuse iterators and re-seek them to a different
prefix. This feature isn't supported by any iterator implementation
except for the reftable iterators anyway, but if it was implemented
it would allow us to optimize cases where we need to search for
specific references repeatedly by reusing internal state.
Detangle the lifecycle from iteration so that we don't deallocate the
iterator anymore once it is exhausted. Instead, callers are now expected
to always call a newly introduce `ref_iterator_free()` function that
deallocates the iterator and its internal state.
Note that the `dir_iterator` is somewhat special because it does not
implement the `ref_iterator` interface, but is only used to implement
other iterators. Consequently, we have to provide `dir_iterator_free()`
instead of `dir_iterator_release()` as the allocated structure itself is
managed by the `dir_iterator` interfaces, as well, and not freed by
`ref_iterator_free()` like in all the other cases.
While at it, drop the return value of `ref_iterator_abort()`, which
wasn't really required by any of the iterator implementations anyway.
Furthermore, stop calling `base_ref_iterator_free()` in any of the
backends, but instead call it in `ref_iterator_free()`.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "files" backend explicitly carves out special logic for its initial
transaction so that it can avoid writing out every single reference as
a loose reference. While the assumption is that there shouldn't be any
preexisting references, we still have to verify that none of the newly
written references will conflict with any other new reference in the
same transaction.
Refactor the initial transaction to use batched refname availability
checks. This does not yet have an effect on performance as we still call
`refs_verify_refname_available()` in a loop. But this will change in
subsequent commits and then impact performance when cloning a repository
with many references or when migrating references to the "files" format.
This will improve performance when cloning a repository with many
references or when migrating references from any format to the "files"
format once the availability checks have learned to optimize checks for
many references in a subsequent commit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Same as the "reftable" backend that we have adapted in the preceding
commit to use batched refname availability checks we can also do so for
the "files" backend. Things are a bit more intricate here though, as we
call `refs_verify_refname_available()` in a set of different contexts:
1. `lock_raw_ref()` when it hits either EEXISTS or EISDIR when creating
a new reference, mostly to create a nice, user-readable error
message. This is nothing we have to care about too much, as we only
hit this code path at most once when we hit a conflict.
2. `lock_raw_ref()` when it _could_ create the lockfile to check
whether it is conflicting with any packed refs. In the general case,
this code path will be hit once for every (successful) reference
update.
3. `lock_ref_oid_basic()`, but it is only executed when copying or
renaming references or when expiring reflogs. It will thus not be
called in contexts where we have many references queued up.
4. `refs_refname_ref_available()`, but again only when copying or
renaming references. It is thus not interesting due to the same
reason as the previous case.
5. `files_transaction_finish_initial()`, which is only executed when
creating a new repository or migrating references.
So out of these, only (2) and (5) are viable candidates to use the
batched checks.
Adapt `lock_raw_ref()` accordingly by queueing up reference names that
need to be checked for availability and then checking them after we have
processed all updates. This check is done before we (optionally) lock
the `packed-refs` file, which is somewhat flawed because it means that
the `packed-refs` could still change after the availability check and
thus create an undetected conflict. But unconditionally locking the file
would change semantics that users are likely to rely on, so we keep the
current locking sequence intact, even if it's suboptmial.
The refactoring of `files_transaction_finish_initial()` will be done in
the next commit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Refactor the "reftable" backend to batch the availability check for
refnames. This does not yet have an effect on performance as
`refs_verify_refnames_available()` effectively still performs the
availability check for each refname individually. But this will be
optimized in subsequent commits, where we learn to optimize some parts
of the logic when checking multiple refnames for availability.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `null_oid()` function returns the object ID that only consists of
zeroes. Naturally, this ID also depends on the hash algorithm used, as
the number of zeroes is different between SHA1 and SHA256. Consequently,
the function returns the hash-algorithm-specific null object ID.
This is currently done by depending on `the_hash_algo`, which implicitly
makes us depend on `the_repository`. Refactor the function to instead
pass in the hash algorithm for which we want to retrieve the null object
ID. Adapt callsites accordingly by passing in `the_repository`, thus
bubbling up the dependency on that global variable by one layer.
There are a couple of trivial exceptions for subsystems that already got
rid of `the_repository`. These subsystems instead use the repository
that is available via the calling context:
- "builtin/grep.c"
- "grep.c"
- "refs/debug.c"
There are also two non-trivial exceptions:
- "diff-no-index.c": Here we know that we may not have a repository
initialized at all, so we cannot rely on `the_repository`. Instead,
we adapt `diff_no_index()` to get a `struct git_hash_algo` as
parameter. The only caller is located in "builtin/diff.c", where we
know to call `repo_set_hash_algo()` in case we're running outside of
a Git repository. Consequently, it is fine to continue passing
`the_repository->hash_algo` even in this case.
- "builtin/ls-files.c": There is an in-flight patch series that drops
`USE_THE_REPOSITORY_VARIABLE` in this file, which causes a semantic
conflict because we use `null_oid()` in `show_submodule()`. The
value is passed to `repo_submodule_init()`, which may use the object
ID to resolve a tree-ish in the superproject from which we want to
read the submodule config. As such, the object ID should refer to an
object in the superproject, and consequently we need to use its hash
algorithm.
This means that we could in theory just not bother about this edge
case at all and just use `the_repository` in "diff-no-index.c". But
doing so would feel misdesigned.
Remove the `USE_THE_REPOSITORY_VARIABLE` preprocessor define in
"hash.c".
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The path.[ch] API takes an explicit repository parameter passed
throughout the callchain, instead of relying on the_repository
singleton instance.
* ps/path-sans-the-repository:
path: adjust last remaining users of `the_repository`
environment: move access to "core.sharedRepository" into repo settings
environment: move access to "core.hooksPath" into repo settings
repo-settings: introduce function to clear struct
path: drop `git_path()` in favor of `repo_git_path()`
rerere: let `rerere_path()` write paths into a caller-provided buffer
path: drop `git_common_path()` in favor of `repo_common_path()`
worktree: return allocated string from `get_worktree_git_dir()`
path: drop `git_path_buf()` in favor of `repo_git_path_replace()`
path: drop `git_pathdup()` in favor of `repo_git_path()`
path: drop unused `strbuf_git_path()` function
path: refactor `repo_submodule_path()` family of functions
submodule: refactor `submodule_to_gitdir()` to accept a repo
path: refactor `repo_worktree_path()` family of functions
path: refactor `repo_git_path()` family of functions
path: refactor `repo_common_path()` family of functions
With the preceding refactorings we now only have a couple of implicit
users of `the_repository` left in the "path" subsystem, all of which
depend on global state via `calc_shared_perm()`. Make the dependency on
`the_repository` explicit by passing the repo as a parameter instead and
adjust callers accordingly.
Note that this change bubbles up into a couple of subsystems that were
previously declared as free from `the_repository`. Instead of marking
all of them as `the_repository`-dependent again, we instead use the
repository that is available in the calling context. There are three
exceptions though with "copy.c", "pack-write.c" and "tempfile.c".
Adjusting these would require us to adapt callsites all over the place,
so this is left for a future iteration.
Mark "path.c" as free from `the_repository`.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When there is a "sorted" trait in the header of the "packed-refs" file,
it means that each entry is sorted increasingly by comparing the
refname. We should add checks to verify whether the "packed-refs" is
sorted in this case.
Update the "packed_fsck_ref_header" to know whether there is a "sorted"
trail in the header. It may seem that we could record all refnames
during the parsing process and then compare later. However, this is not
a good design due to the following reasons:
1. Because we need to store the state across the whole checking
lifetime, we would consume a lot of memory if there are many entries
in the "packed-refs" file.
2. We cannot reuse the existing compare function "cmp_packed_ref_records"
which cause repetition.
Because "cmp_packed_ref_records" needs an extra parameter "struct
snaphost", extract the common part into a new function
"cmp_packed_ref_records" to reuse this function to compare.
Then, create a new function "packed_fsck_ref_sorted" to parse the file
again and user the new fsck message "packedRefUnsorted(ERROR)" to report
to the user if the file is not sorted.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"packed-backend.c::next_record" will parse the ref entry to check the
consistency. This function has already checked the following things:
1. Parse the main line of the ref entry to inspect whether the oid is
not correct. Then, check whether the next character is oid. Then
check the refname.
2. If the next line starts with '^', it would continue to parse the
peeled oid and check whether the last character is '\n'.
As we decide to implement the ref consistency check for "packed-refs",
let's port these two checks and update the test to exercise the code.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"packed-backend.c::next_record" will use "check_refname_format" to check
the consistency of the refname. If it is not OK, the program will die.
However, it is reported in [1], we cannot catch some corruption. But we
already have the code path and we must miss out something.
We use the following code to get the refname:
strbuf_add(&iter->refname_buf, p, eol - p);
iter->base.refname = iter->refname_buf.buf
In the above code, `p` is the start pointer of the refname and `eol` is
the next newline pointer. We calculate the length of the refname by
subtracting the two pointers. Then we add the memory range between `p`
and `eol` to get the refname.
However, if there are some NUL characters in the memory range between `p`
and `eol`, we will see the refname as a valid ref name as long as the
memory range between `p` and first occurred NUL character is valid.
In order to catch above corruption, create a new function
"refname_contains_nul" by searching the first NUL character. If it is
not at the end of the string, there must be some NUL characters in the
refname.
Use this function in "next_record" function to die the program if
"refname_contains_nul" returns true.
[1] https://lore.kernel.org/git/6cfee0e4-3285-4f18-91ff-d097da9de737@rd10.de/
Reported-by: R. Diez <rdiez-temp3@rd10.de>
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In "packed-backend.c::create_snapshot", if there is a header (the line
which starts with '#'), we will check whether the line starts with "#
pack-refs with: ". However, we need to consider other situations and
discuss whether we need to add checks.
1. If the header does not exist, we should not report an error to the
user. This is because in older Git version, we never write header in
the "packed-refs" file. Also, we do allow no header in "packed-refs"
in runtime.
2. If the header content does not start with "# packed-ref with: ", we
should report an error just like what "create_snapshot" does. So,
create a new fsck message "badPackedRefHeader(ERROR)" for this.
3. If the header content is not the same as the constant string
"PACKED_REFS_HEADER". This is expected because we make it extensible
intentionally and runtime "create_snapshot" won't complain about
unknown traits. In order to align with the runtime behavior. There is
no need to report.
As we have analyzed, we only need to check the case 2 in the above. In
order to do this, use "open_nofollow" function to get the file
descriptor and then read the "packed-refs" file via "strbuf_read". Like
what "create_snapshot" and other functions do, we could split the line
by finding the next newline in the buffer. When we cannot find a
newline, we could report an error.
So, create a function "packed_fsck_ref_next_line" to find the next
newline and if there is no such newline, use
"packedRefEntryNotTerminated(ERROR)" to report an error to the user.
Then, parse the first line to apply the checks. Update the test to
exercise the code.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We always write a space after "# pack-refs with:" but we don't align
with this rule in the "create_snapshot" method where we would check
whether header starts with "# pack-refs with:". It might seem that we
should undoubtedly tighten this rule, however, we don't have any
technical documentation about this and there is a possibility that we
would break the compatibility for other third-party libraries.
By investigating influential third-party libraries, we could conclude
how these libraries handle the header of "packed-refs" file:
1. libgit2 is fine and always writes the space. It also expects the
whitespace to exist.
2. JGit does not expect th header to have a trailing space, but expects
the "peeled" capability to have a leading space, which is mostly
equivalent because that capability is typically the first one we
write. It always writes the space.
3. gitoxide expects the space t exist and writes it.
4. go-git doesn't create the header by default.
As many third-party libraries expect a single space after "# pack-refs
with:", if we forget to write the space after the colon,
"create_snapshot" won't catch this. And we would break other
re-implementations. So, we'd better tighten the rule by checking whether
the header starts with "# pack-refs with: ".
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Although "git-fsck(1)" and "packed-backend.c" will check some
consistency and correctness of "packed-refs" file, they never check the
filetype of the "packed-refs". Let's verify that the "packed-refs" has
the expected filetype, confirming it is created by "git pack-refs"
command.
We could use "open_nofollow" wrapper to open the raw "packed-refs" file.
If the returned "fd" value is less than 0, we could check whether the
"errno" is "ELOOP" to report an error to the user. And then we use
"fstat" to check whether the "packed-refs" file is a regular file.
Reuse "FSCK_MSG_BAD_REF_FILETYPE" fsck message id to report the error to
the user if "packed-refs" is not a regular file.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code clean-up.
* kn/reflog-migration-fix-followup:
reftable: prevent 'update_index' changes after adding records
refs: use 'uint64_t' for 'ref_update.index'
refs: mark `ref_transaction_update_reflog()` as static
reflog entries for symbolic ref updates were broken, which has been
corrected.
* kn/reflog-symref-fix:
refs: fix creation of reflog entries for symrefs
When migrating reflogs between reference backends, maintaining the
original order of the reflog entries is crucial. To achieve this, an
`index` field is stored within the `ref_update` struct that encodes the
relative order of reflog entries. This field is used by the reftable
backend as update index for the respective reflog entries to maintain
that ordering.
These update indices must be respected when writing table headers, which
encode the minimum and maximum update index of contained records in the
header and footer. This logic was added in commit bc67b4ab5f (reftable:
write correct max_update_index to header, 2025-01-15), which started to
use `reftable_writer_set_limits()` to propagate the mininum and maximum
update index of all records contained in a ref transaction.
However, we only set the maximum update index for the first transaction
argument, even though there can be multiple such arguments. This is the
case when we write to multiple stacks in a single transaction, e.g. when
updating references in two different worktrees at once. Consequently,
the update index for all but the first argument remain uninitialized,
which may cause undefined behaviour.
Fix this by moving the assignment of the maximum update index in
`reftable_be_transaction_finish()` inside the loop, which ensures that
all elements of the array are correctly initialized.
Furthermore, initialize the `max_index` field to 0 when queueing a new
transaction argument. This is not strictly necessary, as all elements of
`write_transaction_table_arg.max_index` are now assigned correctly.
However, this initialization is added for consistency and to safeguard
against potential future changes that might inadvertently introduce
uninitialized memory access.
Reported-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The commit 297c09eabb (refs: allow multiple reflog entries for the
same refname, 2024-12-16) added logic to exit early in
`lock_ref_for_update()` after obtaining the required lock. This was
added as a performance optimization on a false assumption that no
further processing was required for reflog-only updates.
However the assumption was wrong. For a symref's reflog entry, the
update needs to be populated with the old_oid value, but the early
exit skipped this necessary step.
This caused a bug in Git 2.48 in the files backend where target
references of symrefs being updated would create a corrupted reflog
entry for the symref since the old_oid is not populated.
Everything the early exit skipped in the code path is necessary for
both regular and symbolic ref, so eliminate the mistaken
optimization, and also add a test to ensure that such an issue
doesn't arise in the future.
Reported-by: Nika Layzell <nika@thelayzells.com>
Co-authored-by: Jeff King <peff@peff.net>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The function `reftable_writer_set_limits()` allows updating the
'min_update_index' and 'max_update_index' of a reftable writer. These
values are written to both the writer's header and footer.
Since the header is written during the first block write, any subsequent
changes to the update index would create a mismatch between the header
and footer values. The footer would contain the newer values while the
header retained the original ones.
To protect against this bug, prevent callers from updating these values
after any record is written. To do this, modify the function to return
an error whenever the limits are modified after any record adds. Check
for record adds within `reftable_writer_set_limits()` by checking the
`last_key` and `next` variable. The former is updated after each record
added, but is reset at certain points. The latter is set after writing
the first block.
Modify all callers of the function to anticipate a return type and
handle it accordingly. Add a unit test to also ensure the function
returns the error as expected.
Helped-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The 'ref_update.index' variable is used to store an index for a given
reference update. This index is used to order the updates in a
predetermined order, while the default ordering is alphabetical as per
the refname.
For large repositories with millions of references, it should be safer
to use 'uint64_t'. Let's do that. This also is applied for all other
code sections where we store 'index' and pass it around.
Reported-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 297c09eabb (refs: allow multiple reflog entries for the same refname,
2024-12-16), the reftable backend learned to handle multiple reflog
entries within the same transaction. This was done modifying the
`update_index` for reflogs with multiple indices. During writing the
logs, the `max_update_index` of the writer was modified to ensure the
limits were raised to the modified `update_index`s.
However, since ref entries are written before the modification to the
`max_update_index`, if there are multiple blocks to be written, the
reftable backend writes the header with the old `max_update_index`. When
all logs are finally written, the footer will be written with the new
`min_update_index`. This causes a mismatch between the header and the
footer and causes the reftable file to be corrupted. The existing tests
only spawn a single block and since headers are lazily written with the
first block, the tests didn't capture this bug.
To fix the issue, the appropriate `max_update_index` limit must be set
even before the first block is written. Add a `max_index` field to the
transaction which holds the `max_index` within all its updates, then
propagate this value to the reftable backend, wherein this is used to
the set the `max_update_index` correctly.
Add a test which creates a few thousand reference updates with multiple
reflog entries, which should trigger the bug.
Reported-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git refs migrate" learned to also migrate the reflog data across
backends.
* kn/reflog-migration:
refs: mark invalid refname message for translation
refs: add support for migrating reflogs
refs: allow multiple reflog entries for the same refname
refs: introduce the `ref_transaction_update_reflog` function
refs: add `committer_info` to `ref_transaction_add_update()`
refs: extract out refname verification in transactions
refs/files: add count field to ref_lock
refs: add `index` field to `struct ref_udpate`
refs: include committer info in `ref_update` struct
Start working to make the codebase buildable with -Wsign-compare.
* ps/build-sign-compare:
t/helper: don't depend on implicit wraparound
scalar: address -Wsign-compare warnings
builtin/patch-id: fix type of `get_one_patchid()`
builtin/blame: fix type of `length` variable when emitting object ID
gpg-interface: address -Wsign-comparison warnings
daemon: fix type of `max_connections`
daemon: fix loops that have mismatching integer types
global: trivial conversions to fix `-Wsign-compare` warnings
pkt-line: fix -Wsign-compare warning on 32 bit platform
csum-file: fix -Wsign-compare warning on 32-bit platform
diff.h: fix index used to loop through unsigned integer
config.mak.dev: drop `-Wno-sign-compare`
global: mark code units that generate warnings with `-Wsign-compare`
compat/win32: fix -Wsign-compare warning in "wWinMain()"
compat/regex: explicitly ignore "-Wsign-compare" warnings
git-compat-util: introduce macros to disable "-Wsign-compare" warnings
When "git fetch $remote" notices that refs/remotes/$remote/HEAD is
missing and discovers what branch the other side points with its
HEAD, refs/remotes/$remote/HEAD is updated to point to it.
* bf/set-head-symref:
fetch set_head: handle mirrored bare repositories
fetch: set remote/HEAD if it does not exist
refs: add create_only option to refs_update_symref_extended
refs: add TRANSACTION_CREATE_EXISTS error
remote set-head: better output for --auto
remote set-head: refactor for readability
refs: atomically record overwritten ref in update_symref
refs: standardize output of refs_read_symbolic_ref
t/t5505-remote: test failure of set-head
t/t5505-remote: set default branch to main
The reference transaction only allows a single update for a given
reference to avoid conflicts. This, however, isn't an issue for reflogs.
There are no conflicts to be resolved in reflogs and when migrating
reflogs between backends we'd have multiple reflog entries for the same
refname.
So allow multiple reflog updates within a single transaction. Also the
reflog creation logic isn't exposed to the end user. While this might
change in the future, currently, this reduces the scope of issues to
think about.
In the reftable backend, the writer sorts all updates based on the
update_index before writing to the block. When there are multiple
reflogs for a given refname, it is essential that the order of the
reflogs is maintained. So add the `index` value to the `update_index`.
The `index` field is only set when multiple reflog entries for a given
refname are added and as such in most scenarios the old behavior
remains.
This is required to add reflog migration support to `git refs migrate`.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Introduce a new function `ref_transaction_update_reflog`, for clients to
add a reflog update to a transaction. While the existing function
`ref_transaction_update` also allows clients to add a reflog entry, this
function does a few things more, It:
- Enforces that only a reflog entry is added and does not update the
ref itself.
- Allows the users to also provide the committer information. This
means clients can add reflog entries with custom committer
information.
The `transaction_refname_valid()` function also modifies the error
message selectively based on the type of the update. This change also
affects reflog updates which go through `ref_transaction_update()`.
A follow up commit will utilize this function to add reflog support to
`git refs migrate`.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `ref_transaction_add_update()` creates the `ref_update` struct. To
facilitate addition of reflogs in the next commit, the function needs to
accommodate setting the `committer_info` field in the struct. So modify
the function to also take `committer_info` as an argument and set it
accordingly.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When refs are updated in the files-backend, a lock is obtained for the
corresponding file path. This is the case even for reflogs, i.e. a lock
is obtained on the reference path instead of the reflog path. This
works, since generally, reflogs are updated alongside the ref.
The upcoming patches will add support for reflog updates in ref
transaction. This means, in a particular transaction we want to have ref
updates and reflog updates. For a given ref in a given transaction there
can be at most one update. But we can theoretically have multiple reflog
updates for a given ref in a given transaction. A great example of this
would be when migrating reflogs from one backend to another. There we
would batch all the reflog updates for a given reference in a single
transaction.
The current flow does not support this, because currently refs & reflogs
are treated as a single entity and capture the lock together. To
separate this, add a count field to ref_lock. With this, multiple
updates can hold onto a single ref_lock and the lock will only be
released when all of them release the lock.
This patch only adds the `count` field to `ref_lock` and adds the logic
to increment and decrement the lock. In a follow up commit, we'll
separate the reflog update logic from ref updates and utilize this
functionality.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The reftable backend, sorts its updates by refname before applying them,
this ensures that the references are stored sorted. When migrating
reflogs from one backend to another, the order of the reflogs must be
maintained. Add a new `index` field to the `ref_update` struct to
facilitate this.
This field is used in the reftable backend's sort comparison function
`transaction_update_cmp`, to ensure that indexed fields maintain their
order.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The reference backends obtain the committer information from
`git_committer_info(0)` when adding a reflog. The upcoming patches
introduce support for migrating reflogs between the reference backends.
This requires an interface to creating reflogs, including custom
committer information.
Add a new field `committer_info` to the `ref_update` struct, which is
then used by the reference backends. If there is no `committer_info`
provided, the reference backends default to using
`git_committer_info(0)`. The field itself cannot be set to
`git_committer_info(0)` since the values are dynamic and must be
obtained right when the reflog is being committed.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Optimize reading random references out of the reftable backend by
allowing reuse of iterator objects.
* ps/reftable-iterator-reuse:
refs/reftable: reuse iterators when reading refs
reftable/merged: drain priority queue on reseek
reftable/stack: add mechanism to notify callers on reload
refs/reftable: refactor reflog expiry to use reftable backend
refs/reftable: refactor reading symbolic refs to use reftable backend
refs/reftable: read references via `struct reftable_backend`
refs/reftable: figure out hash via `reftable_stack`
reftable/stack: add accessor for the hash ID
refs/reftable: handle reloading stacks in the reftable backend
refs/reftable: encapsulate reftable stack
Isolates the reftable subsystem from the rest of Git's codebase by
using fewer pieces of Git's infrastructure.
* ps/reftable-detach:
reftable/system: provide thin wrapper for lockfile subsystem
reftable/stack: drop only use of `get_locked_file_path()`
reftable/system: provide thin wrapper for tempfile subsystem
reftable/stack: stop using `fsync_component()` directly
reftable/system: stop depending on "hash.h"
reftable: explicitly handle hash format IDs
reftable/system: move "dir.h" to its only user
We have a bunch of loops which iterate up to an unsigned boundary using
a signed index, which generates warnigs because we compare a signed and
unsigned value in the loop condition. Address these sites for trivial
cases and enable `-Wsign-compare` warnings for these code units.
This patch only adapts those code units where we can drop the
`DISABLE_SIGN_COMPARE_WARNINGS` macro in the same step.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Mark code units that generate warnings with `-Wsign-compare`. This
allows for a structured approach to get rid of all such warnings over
time in a way that can be easily measured.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A double-free that may not trigger in practice by luck has been
corrected in the reference resolution code.
* sj/refs-symref-referent-fix:
ref-cache: fix invalid free operation in `free_ref_entry`
"git fsck" learned to issue warnings on "curiously formatted" ref
contents that have always been taken valid but something Git
wouldn't have written itself (e.g., missing terminating end-of-line
after the full object name).
* sj/ref-contents-check:
ref: add symlink ref content check for files backend
ref: check whether the target of the symref is a ref
ref: add basic symref content check for files backend
ref: add more strict checks for regular refs
ref: port git-fsck(1) regular refs check for files backend
ref: support multiple worktrees check for refs
ref: initialize ref name outside of check functions
ref: check the full refname instead of basename
ref: initialize "fsck_ref_report" with zero
In cfd971520e (refs: keep track of unresolved reference value in
iterators, 2024-08-09), we added a new field "referent" into the "struct
ref" structure. In order to free the "referent", we unconditionally
freed the "referent" by simply adding a "free" statement.
However, this is a bad usage. Because when ref entry is either directory
or loose ref, we will always execute the following statement:
free(entry->u.value.referent);
This does not make sense. We should never access the "entry->u.value"
field when "entry" is a directory. However, the change obviously doesn't
break the tests. Let's analysis why.
The anonymous union in the "ref_entry" has two members: one is "struct
ref_value", another is "struct ref_dir". On a 64-bit machine, the size
of "struct ref_dir" is 32 bytes, which is smaller than the 48-byte size
of "struct ref_value". And the offset of "referent" field in "struct
ref_value" is 40 bytes. So, whenever we create a new "ref_entry" for a
directory, we will leave the offset from 40 bytes to 48 bytes untouched,
which means the value for this memory is zero (NULL). It's OK to free a
NULL pointer, but this is merely a coincidence of memory layout.
To fix this issue, we now ensure that "free(entry->u.value.referent)" is
only called when "entry->flag" indicates that it represents a loose
reference and not a directory to avoid the invalid memory operation.
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When reading references the reftable backend has to:
1. Create a new ref iterator.
2. Seek the iterator to the record we're searching for.
3. Read the record.
We cannot really avoid the last two steps, but re-creating the iterator
every single time we want to read a reference is kind of expensive and a
waste of resources. We couldn't help it in the past though because it
was not possible to reuse iterators. But starting with 5bf96e0c39
(reftable/generic: move seeking of records into the iterator,
2024-05-13) we have split up the iterator lifecycle such that creating
the iterator and seeking are two different concerns.
Refactor the code such that we cache iterators in the reftable backend.
This cache is invalidated whenever the respective stack is reloaded such
that we know to recreate the iterator in that case. This leads to a
sizeable speedup when creating many refs, which requires a lot of random
reference reads:
Benchmark 1: update-ref: create many refs (refcount = 100000, revision = master)
Time (mean ± σ): 1.793 s ± 0.010 s [User: 0.954 s, System: 0.835 s]
Range (min … max): 1.781 s … 1.811 s 10 runs
Benchmark 2: update-ref: create many refs (refcount = 100000, revision = HEAD)
Time (mean ± σ): 1.680 s ± 0.013 s [User: 0.846 s, System: 0.831 s]
Range (min … max): 1.664 s … 1.702 s 10 runs
Summary
update-ref: create many refs (refcount = 100000, revision = HEAD) ran
1.07 ± 0.01 times faster than update-ref: create many refs (refcount = 100000, revision = master)
While 7% is not a huge win, you have to consider that the benchmark is
_writing_ data, so _reading_ references is only one part of what we do.
Flame graphs show that we spend around 40% of our time reading refs, so
the speedup when reading refs is approximately ~2.5x that. I could not
find better benchmarks where we perform a lot of random ref reads.
You can also see a sizeable impact on memory usage when creating 100k
references. Before this change:
HEAP SUMMARY:
in use at exit: 19,112,538 bytes in 200,170 blocks
total heap usage: 8,400,426 allocs, 8,200,256 frees, 454,367,048 bytes allocated
After this change:
HEAP SUMMARY:
in use at exit: 674,416 bytes in 169 blocks
total heap usage: 7,929,872 allocs, 7,929,703 frees, 281,509,985 bytes allocated
As an additional factor, this refactoring opens up the possibility for
more performance optimizations in how we re-seek iterators. Any change
that allows us to optimize re-seeking by e.g. reusing data structures
would thus also directly speed up random reads.
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>
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>
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>
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>
Besides the textual symref, we also allow symbolic links as the symref.
So, we should also provide the consistency check as what we have done
for textual symref. And also we consider deprecating writing the
symbolic links. We first need to access whether symbolic links still
be used. So, add a new fsck message "symlinkRef(INFO)" to tell the
user be aware of this information.
We have already introduced "files_fsck_symref_target". We should reuse
this function to handle the symrefs which use legacy symbolic links. We
should not check the trailing garbage for symbolic refs. Add a new
parameter "symbolic_link" to disable some checks which should only be
executed for textual symrefs.
And we need to also generate the "referent" parameter for reusing
"files_fsck_symref_target" by the following steps:
1. Use "strbuf_add_real_path" to resolve the symlink and get the
absolute path "ref_content" which the symlink ref points to.
2. Generate the absolute path "abs_gitdir" of "gitdir" and combine
"ref_content" and "abs_gitdir" to extract the relative path
"relative_referent_path".
3. If "ref_content" is outside of "gitdir", we just set "referent" with
"ref_content". Instead, we set "referent" with
"relative_referent_path".
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Ideally, we want to the users use "git symbolic-ref" to create symrefs
instead of writing raw contents into the filesystem. However, "git
symbolic-ref" is strict with the refname but not strict with the
referent. For example, we can make the "referent" located at the
"$(gitdir)/logs/aaa" and manually write the content into this where we
can still successfully parse this symref by using "git rev-parse".
$ git init repo && cd repo && git commit --allow-empty -mx
$ git symbolic-ref refs/heads/test logs/aaa
$ echo $(git rev-parse HEAD) > .git/logs/aaa
$ git rev-parse test
We may need to add some restrictions for "referent" parameter when using
"git symbolic-ref" to create symrefs because ideally all the
nonpseudo-refs should be located under the "refs" directory and we may
tighten this in the future.
In order to tell the user we may tighten the above situation, create
a new fsck message "symrefTargetIsNotARef" to notify the user that this
may become an error in the future.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We have code that checks regular ref contents, but we do not yet check
the contents of symbolic refs. By using "parse_loose_ref_content" for
symbolic refs, we will get the information of the "referent".
We do not need to check the "referent" by opening the file. This is
because if "referent" exists in the file system, we will eventually
check its correctness by inspecting every file in the "refs" directory.
If the "referent" does not exist in the filesystem, this is OK as it is
seen as the dangling symref.
So we just need to check the "referent" string content. A regular ref
could be accepted as a textual symref if it begins with "ref:", followed
by zero or more whitespaces, followed by the full refname, followed only
by whitespace characters. However, we always write a single SP after
"ref:" and a single LF after the refname. It may seem that we should
report a fsck error message when the "referent" does not apply above
rules and we should not be so aggressive because third-party
reimplementations of Git may have taken advantage of the looser syntax.
Put it more specific, we accept the following contents:
1. "ref: refs/heads/master "
2. "ref: refs/heads/master \n \n"
3. "ref: refs/heads/master\n\n"
When introducing the regular ref content checks, we created two fsck
infos "refMissingNewline" and "trailingRefContent" which exactly
represents above situations. So we will reuse these two fsck messages to
write checks to info the user about these situations.
But we do not allow any other trailing garbage. The followings are bad
symref contents which will be reported as fsck error by "git-fsck(1)".
1. "ref: refs/heads/master garbage\n"
2. "ref: refs/heads/master \n\n\n garbage "
And we introduce a new "badReferentName(ERROR)" fsck message to report
above errors by using "is_root_ref" and "check_refname_format" to check
the "referent". Since both "is_root_ref" and "check_refname_format"
don't work with whitespaces, we use the trimmed version of "referent"
with these functions.
In order to add checks, we will do the following things:
1. Record the untrimmed length "orig_len" and untrimmed last byte
"orig_last_byte".
2. Use "strbuf_rtrim" to trim the whitespaces or newlines to make sure
"is_root_ref" and "check_refname_format" won't be failed by them.
3. Use "orig_len" and "orig_last_byte" to check whether the "referent"
misses '\n' at the end or it has trailing whitespaces or newlines.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We have already used "parse_loose_ref_contents" function to check
whether the ref content is valid in files backend. However, by
using "parse_loose_ref_contents", we allow the ref's content to end with
garbage or without a newline.
Even though we never create such loose refs ourselves, we have accepted
such loose refs. So, it is entirely possible that some third-party tools
may rely on such loose refs being valid. We should not report an error
fsck message at current. We should notify the users about such
"curiously formatted" loose refs so that adequate care is taken before
we decide to tighten the rules in the future.
And it's not suitable either to report a warn fsck message to the user.
We don't yet want the "--strict" flag that controls this bit to end up
generating errors for such weirdly-formatted reference contents, as we
first want to assess whether this retroactive tightening will cause
issues for any tools out there. It may cause compatibility issues which
may break the repository. So, we add the following two fsck infos to
represent the situation where the ref content ends without newline or
has trailing garbages:
1. refMissingNewline(INFO): A loose ref that does not end with
newline(LF).
2. trailingRefContent(INFO): A loose ref has trailing content.
It might appear that we can't provide the user with any warnings by
using FSCK_INFO. However, in "fsck.c::fsck_vreport", we will convert
FSCK_INFO to FSCK_WARN and we can still warn the user about these
situations when using "git refs verify" without introducing
compatibility issues.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git-fsck(1)" implicitly checks the ref content by passing the
callback "fsck_handle_ref" to the "refs.c::refs_for_each_rawref".
Then, it will check whether the ref content (eventually "oid")
is valid. If not, it will report the following error to the user.
error: refs/heads/main: invalid sha1 pointer 0000...
And it will also report above errors when there are dangling symrefs
in the repository wrongly. This does not align with the behavior of
the "git symbolic-ref" command which allows users to create dangling
symrefs.
As we have already introduced the "git refs verify" command, we'd better
check the ref content explicitly in the "git refs verify" command thus
later we could remove these checks in "git-fsck(1)" and launch a
subprocess to call "git refs verify" in "git-fsck(1)" to make the
"git-fsck(1)" more clean.
Following what "git-fsck(1)" does, add a similar check to "git refs
verify". Then add a new fsck error message "badRefContent(ERROR)" to
represent that a ref has an invalid content.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We have already set up the infrastructure to check the consistency for
refs, but we do not support multiple worktrees. However, "git-fsck(1)"
will check the refs of worktrees. As we decide to get feature parity
with "git-fsck(1)", we need to set up support for multiple worktrees.
Because each worktree has its own specific refs, instead of just showing
the users "refs/worktree/foo", we need to display the full name such as
"worktrees/<id>/refs/worktree/foo". So we should know the id of the
worktree to get the full name. Add a new parameter "struct worktree *"
for "refs-internal.h::fsck_fn". Then change the related functions to
follow this new interface.
The "packed-refs" only exists in the main worktree, so we should only
check "packed-refs" in the main worktree. Use "is_main_worktree" method
to skip checking "packed-refs" in "packed_fsck" function.
Then, enhance the "files-backend.c::files_fsck_refs_dir" function to add
"worktree/<id>/" prefix when we are not in the main worktree.
Last, add a new test to check the refname when there are multiple
worktrees to exercise the code.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We passes "refs_check_dir" to the "files_fsck_refs_name" function which
allows it to create the checked ref name later. However, when we
introduce a new check function, we have to allocate redundant memory and
re-calculate the ref name. It's bad for us to allocate redundant memory
and duplicate logic. Instead, we should allocate and calculate it only
once and pass the ref name to the check functions.
In order not to do repeat calculation, rename "refs_check_dir" to
"refname". And in "files_fsck_refs_dir", create a new strbuf "refname",
thus whenever we handle a new ref, calculate the name and call the check
functions one by one.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In "files-backend.c::files_fsck_refs_name", we validate the refname
format by using "check_refname_format" to check the basename of the
iterator with "REFNAME_ALLOW_ONELEVEL" flag.
However, this is a bad implementation. Although we doesn't allow a
single "@" in ".git" directory, we do allow "refs/heads/@". So, we will
report an error wrongly when there is a "refs/heads/@" ref by using one
level refname "@".
Because we just check one level refname, we either cannot check the
other parts of the full refname. And we will ignore the following
errors:
"refs/heads/ new-feature/test"
"refs/heads/~new-feature/test"
In order to fix the above problem, enhance "files_fsck_refs_name" to use
the full name for "check_refname_format". Then, replace the tests which
are related to "@" and add tests to exercise the above situations using
for loop to avoid repetition.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In "fsck.c::fsck_refs_error_function", we need to tell whether "oid" and
"referent" is NULL. So, we need to always initialize these parameters to
NULL instead of letting them point to anywhere when creating a new
"fsck_ref_report" structure.
The original code explicitly initializes the "path" member in the
"struct fsck_ref_report" to NULL (which implicitly 0-initializes other
members in the struct). It is more customary to use "{ 0 }" to express
that we are 0-initializing everything. In order to align with the
codebase, initialize "fsck_ref_report" with zero.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Reference transactions use `refs_verify_refname_available()` to check
for colliding references. This check consists of two parts:
- Checks for whether multiple ref updates in the same transaction
conflict with each other.
- Checks for whether existing refs conflict with any refs part of the
transaction.
While we generally cannot avoid the first check, the second check is
superfluous in cases where the transaction is an initial one in an
otherwise empty ref store. The check results in multiple ref reads as
well as the creation of a ref iterator for every ref we're checking,
which adds up quite fast when performing the check for many refs.
Introduce a new flag that allows us to skip this check and wire it up in
such that the backends pass it when running an initial transaction. This
leads to significant speedups when migrating ref storage backends. From
"files" to "reftable":
Benchmark 1: migrate files:reftable (refcount = 100000, revision = HEAD~)
Time (mean ± σ): 472.4 ms ± 6.7 ms [User: 175.9 ms, System: 285.2 ms]
Range (min … max): 463.5 ms … 483.2 ms 10 runs
Benchmark 2: migrate files:reftable (refcount = 100000, revision = HEAD)
Time (mean ± σ): 86.1 ms ± 1.9 ms [User: 67.9 ms, System: 16.0 ms]
Range (min … max): 82.9 ms … 90.9 ms 29 runs
Summary
migrate files:reftable (refcount = 100000, revision = HEAD) ran
5.48 ± 0.15 times faster than migrate files:reftable (refcount = 100000, revision = HEAD~)
And from "reftable" to "files":
Benchmark 1: migrate reftable:files (refcount = 100000, revision = HEAD~)
Time (mean ± σ): 452.7 ms ± 3.4 ms [User: 209.9 ms, System: 235.4 ms]
Range (min … max): 445.9 ms … 457.5 ms 10 runs
Benchmark 2: migrate reftable:files (refcount = 100000, revision = HEAD)
Time (mean ± σ): 95.2 ms ± 2.2 ms [User: 73.6 ms, System: 20.6 ms]
Range (min … max): 91.7 ms … 100.8 ms 28 runs
Summary
migrate reftable:files (refcount = 100000, revision = HEAD) ran
4.76 ± 0.11 times faster than migrate reftable:files (refcount = 100000, revision = HEAD~)
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "files" backend has implemented special logic when committing
the first transactions in an otherwise empty ref store: instead of
writing all refs as separate loose files, it instead knows to write them
all into a "packed-refs" file directly. This is significantly more
efficient than having to write each of the refs as separate "loose" ref.
The only user of this optimization is git-clone(1), which only uses this
mechanism to write regular refs. Consequently, the implementation does
not know how to handle both symbolic and root refs. While fine in the
context of git-clone(1), this keeps us from using the mechanism in more
cases.
Adapt the logic to also support symbolic and root refs by using a second
transaction that we use for all of the refs that need to be written as
loose refs.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There are two different ways to commit a transaction:
- `ref_transaction_commit()` can be used to commit a regular
transaction and is what almost every caller wants.
- `initial_ref_transaction_commit()` can be used when it is known that
the ref store that the transaction is committed for is empty and
when there are no concurrent processes. This is used when cloning a
new repository.
Implementing this via two separate functions has a couple of downsides.
First, every reference backend needs to implement a separate callback
even in the case where they don't special-case the initial transaction.
Second, backends are basically forced to reimplement the whole logic for
how to commit the transaction like the "files" backend does, even though
backends may wish to only tweak certain behaviour of a "normal" commit.
Third, it is awkward that callers must never prepare the transaction as
this is somewhat different than how a transaction typically works.
Refactor the code such that we instead mark initial transactions via a
separate flag when starting the transaction. This addresses all of the
mentioned painpoints, where the most important part is that it will
allow backends to have way more leeway in how exactly they want to
handle the initial transaction.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the logic to commit initial transactions such that we can start to
call it in `files_transaction_finish()` in a subsequent commit without
requiring a separate function declaration.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Allow passing flags when setting up a transaction such that the
behaviour of the transaction itself can be altered. This functionality
will be used in a subsequent patch.
Adapt callers accordingly.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ps/reftable-detach:
reftable/system: provide thin wrapper for lockfile subsystem
reftable/stack: drop only use of `get_locked_file_path()`
reftable/system: provide thin wrapper for tempfile subsystem
reftable/stack: stop using `fsync_component()` directly
reftable/system: stop depending on "hash.h"
reftable: explicitly handle hash format IDs
reftable/system: move "dir.h" to its only user
We're executing `fsync_component()` directly in the reftable library so
that we can fsync data to disk depending on "core.fsync". But as we're
in the process of converting the reftable library to become standalone
we cannot use that function in the library anymore.
Refactor the code such that users of the library can inject a custom
fsync function via the write options. This allows us to get rid of the
dependency on "write-or-die.h".
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We include "hash.h" in "reftable/system.h" such that we can use hash
format IDs as well as the raw size of SHA1 and SHA256. As we are in the
process of converting the reftable library to become standalone we of
course cannot rely on those constants anymore.
Introduce a new `enum reftable_hash` to replace internal uses of the
hash format IDs and new constants that replace internal uses of the hash
size. Adapt the reftable backend to set up the correct hash function.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Used regex to find these typos:
(?<!struct )(?<=\s)([a-z]{1,}) \1(?=\s)
Signed-off-by: Sven Strickroth <email@cs-ware.de>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
The reftable library is now prepared to expect that the memory
allocation function given to it may fail to allocate and to deal
with such an error.
* ps/reftable-alloc-failures: (26 commits)
reftable/basics: fix segfault when growing `names` array fails
reftable/basics: ban standard allocator functions
reftable: introduce `REFTABLE_FREE_AND_NULL()`
reftable: fix calls to free(3P)
reftable: handle trivial allocation failures
reftable/tree: handle allocation failures
reftable/pq: handle allocation failures when adding entries
reftable/block: handle allocation failures
reftable/blocksource: handle allocation failures
reftable/iter: handle allocation failures when creating indexed table iter
reftable/stack: handle allocation failures in auto compaction
reftable/stack: handle allocation failures in `stack_compact_range()`
reftable/stack: handle allocation failures in `reftable_new_stack()`
reftable/stack: handle allocation failures on reload
reftable/reader: handle allocation failures in `reader_init_iter()`
reftable/reader: handle allocation failures for unindexed reader
reftable/merged: handle allocation failures in `merged_table_init_iter()`
reftable/writer: handle allocation failures in `reftable_new_writer()`
reftable/writer: handle allocation failures in `writer_index_hash()`
reftable/record: handle allocation failures when decoding records
...
Handle allocation failures in `merged_table_init_iter()`. While at it,
merge `merged_iter_init()` into the function. It only has a single
caller and merging them makes it easier to handle allocation failures
consistently.
This change also requires us to adapt `reftable_stack_init_*_iterator()`
to bubble up the new error codes of `merged_table_iter_init()`. Adapt
callsites accordingly.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Give timeout to the locking code to write to reftable.
* ps/reftable-concurrent-writes:
refs/reftable: reload locked stack when preparing transaction
reftable/stack: allow locking of outdated stacks
refs/reftable: introduce "reftable.lockTimeout"
The reftable backend learned to more efficiently handle exclude
patterns while enumerating the refs.
* ps/reftable-exclude:
refs/reftable: wire up support for exclude patterns
reftable/reader: make table iterator reseekable
t/unit-tests: introduce reftable library
Makefile: stop listing test library objects twice
builtin/receive-pack: fix exclude patterns when announcing refs
refs: properly apply exclude patterns to namespaced refs
When starting a reftable transaction we lock all stacks we are about to
modify. While it may happen that the stack is out-of-date at this point
in time we don't really care: transactional updates encode the expected
state of a certain reference, so all that we really want to verify is
that the _current_ value matches that expected state.
Pass `REFTABLE_STACK_NEW_ADDITION_RELOAD` when locking the stack such
that an out-of-date stack will be reloaded after having been locked.
This change is safe because all verifications of the expected state
happen after this step anyway.
Add a testcase that verifies that many writers are now able to write to
the stack concurrently without failures and with a deterministic end
result.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In `reftable_stack_new_addition()` we first lock the stack and then
check whether it is still up-to-date. If it is not we return an error to
the caller indicating that the stack is outdated.
This is overly restrictive in our ref transaction interface though: we
lock the stack right before we start to verify the transaction, so we do
not really care whether it is outdated or not. What we really want is
that the stack is up-to-date after it has been locked so that we can
verify queued updates against its current state while we know that it is
locked for concurrent modification.
Introduce a new flag `REFTABLE_STACK_NEW_ADDITION_RELOAD` that alters
the behaviour of `reftable_stack_init_addition()` in this case: when we
notice that it is out-of-date we reload it instead of returning an error
to the caller.
This logic will be wired up in the reftable backend in the next commit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When multiple concurrent processes try to update references in a
repository they may try to lock the same lockfiles. This can happen even
when the updates are non-conflicting and can both be applied, so it
doesn't always make sense to abort the transaction immediately. Both the
"loose" and "packed" backends thus have a grace period that they wait
for the lock to be released that can be controlled via the config values
"core.filesRefLockTimeout" and "core.packedRefsTimeout", respectively.
The reftable backend doesn't have such a setting yet and instead fails
immediately when it sees such a lock. But the exact same concepts apply
here as they do apply to the other backends.
Introduce a new "reftable.lockTimeout" config that controls how long we
may wait for a "tables.list" lock to be released. The default value of
this config is 100ms, which is the same default as we have it for the
"loose" backend.
Note that even though we also lock individual tables, this config really
only applies to the "tables.list" file. This is because individual
tables are only ever locked when we already hold the "tables.list" lock
during compaction. When we observe such a lock we in fact do not want to
compact the table at all because it is already in the process of being
compacted by a concurrent process. So applying the same timeout here
would not make any sense and only delay progress.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code clean-up.
* ps/environ-wo-the-repository: (21 commits)
environment: stop storing "core.notesRef" globally
environment: stop storing "core.warnAmbiguousRefs" globally
environment: stop storing "core.preferSymlinkRefs" globally
environment: stop storing "core.logAllRefUpdates" globally
refs: stop modifying global `log_all_ref_updates` variable
branch: stop modifying `log_all_ref_updates` variable
repo-settings: track defaults close to `struct repo_settings`
repo-settings: split out declarations into a standalone header
environment: guard state depending on a repository
environment: reorder header to split out `the_repository`-free section
environment: move `set_git_dir()` and related into setup layer
environment: make `get_git_namespace()` self-contained
environment: move object database functions into object layer
config: make dependency on repo in `read_early_config()` explicit
config: document `read_early_config()` and `read_very_early_config()`
environment: make `get_git_work_tree()` accept a repository
environment: make `get_graft_file()` accept a repository
environment: make `get_index_file()` accept a repository
environment: make `get_object_directory()` accept a repository
environment: make `get_git_common_dir()` accept a repository
...
Exclude patterns can be used by reference backends to skip over blocks
of references that are uninteresting to the caller. Reference backends
do not have to wire up support for them, and all callers are expected to
behave as if the backend didn't support them. In fact, the only backend
that supports exclude patterns right now is the "packed" backend.
Exclude patterns can be quite an important performance optimization in
repositories that have loads of references. The patterns are set up in
case "transfer.hideRefs" and friends are configured during a fetch, so
handling these patterns becomes important once there are lots of hidden
refs in a served repository.
Now that we have properly re-seekable reftable iterators we can also
wire up support for these patterns in the "reftable" backend. Doing so
is conceptually simple: once we hit a reference whose prefix matches the
current exclude pattern we re-seek the iterator to the first reference
that doesn't match the pattern anymore. This schema only works for
trivial patterns that do not have any globbing characters in them, but
this restriction also applies do the "packed" backend.
This makes t1419 work with the "reftable" backend with some slight
modifications. Of course it also speeds up listing of references with
hidden refs. The following benchmark prints one reference with 1 million
hidden references:
Benchmark 1: HEAD~
Time (mean ± σ): 93.3 ms ± 2.1 ms [User: 90.3 ms, System: 2.5 ms]
Range (min … max): 89.8 ms … 97.2 ms 33 runs
Benchmark 2: HEAD
Time (mean ± σ): 4.2 ms ± 0.6 ms [User: 2.2 ms, System: 1.8 ms]
Range (min … max): 3.1 ms … 8.1 ms 765 runs
Summary
HEAD ran
22.15 ± 3.19 times faster than HEAD~
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git pack-refs --auto" for the files backend was too aggressive,
which has been a bit tamed.
* ps/pack-refs-auto-heuristics:
refs/files: use heuristic to decide whether to repack with `--auto`
t0601: merge tests for auto-packing of refs
wrapper: introduce `log2u()`
Same as the preceding commit, storing the "core.preferSymlinkRefs" value
globally is misdesigned as this setting may be set per repository.
There is only a single user of this value anyway, namely the "files"
backend. So let's just remove the global variable and read the value of
this setting when initializing the backend.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The value of "core.logAllRefUpdates" is being stored in the global
variable `log_all_ref_updates`. This design is somewhat aged nowadays,
where it is entirely possible to access multiple repositories in the
same process which all have different values for this setting. So using
a single global variable to track it is plain wrong.
Remove the global variable. Instead, we now provide a new function part
of the repo-settings subsystem that parses the value for a specific
repository. While that may require us to read the value multiple times,
we work around this by reading it once when the ref backends are set up
and caching the value there.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In refs-related code we modify the global `log_all_ref_updates`
variable, which is done because `should_autocreate_reflog()` does not
accept passing an `enum log_refs_config` but instead accesses the global
variable. Adapt its interface such that the value is provided by the
caller, which allows us to compute the proper value locally without
having to modify global state.
This change requires us to move the enum to "repo-settings.h", or
otherwise we get compilation errors due to include cycles. We're about
to fully move this setting into the repo-settings subsystem anyway, so
this is fine.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In "environment.h" we have quite a lot of functions and variables that
either explicitly or implicitly depend on `the_repository`.
The implicit set of stateful declarations includes for example variables
which get populated when parsing a repository's Git configuration. This
set of variables is broken by design, as their state often depends on
the last repository config that has been parsed. So they may or may not
represent the state of `the_repository`.
Fixing that is quite a big undertaking, and later patches in this series
will demonstrate a solution for a first small set of those variables. So
for now, let's guard these with `USE_THE_REPOSITORY_VARIABLE` so that
callers are aware of the implicit dependency.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `--auto` flag for git-pack-refs(1) allows the ref backend to decide
whether or not a repack is in order. This switch has been introduced
mostly with the "reftable" backend in mind, which already knows to
auto-compact its tables during normal operations. When the flag is set,
then it will use the same auto-compaction mechanism and thus end up
doing nothing in most cases.
The "files" backend does not have any such heuristic yet and instead
packs any loose references unconditionally. So we rewrite the complete
"packed-refs" file even if there's only a single loose reference to be
packed.
Even worse, starting with 9f6714ab3e (builtin/gc: pack refs when using
`git maintenance run --auto`, 2024-03-25), `git pack-refs --auto` is
unconditionally executed via our auto maintenance, so we end up repacking
references every single time auto maintenance kicks in. And while that
commit already mentioned that the "files" backend unconditionally packs
refs now, the author obviously didn't quite think about the consequences
thereof. So while the idea was sound, we really should have added a
heuristic to the "files" backend before implementing it.
Introduce a heuristic that decides whether or not it is worth to pack
loose references. The important factors to decide here are the number of
loose references in comparison to the overall size of the "packed-refs"
file. The bigger the "packed-refs" file, the longer it takes to rewrite
it and thus we scale up the limit of allowed loose references before we
repack.
As is the nature of heuristics, this mechansim isn't obviously
"correct", but should rather be seen as a tradeoff between how much
resources we spend packing refs and how inefficient the ref store
becomes. For all I can say, we have successfully been using the exact
same heuristic in Gitaly for several years by now.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is needed to build things with -Werror=unused-parameter on a
platform without symbolic link support.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Mark unused parameters as UNUSED to squelch -Wunused warnings.
* jk/mark-unused-parameters:
t-hashmap: stop calling setup() for t_intern() test
scalar: mark unused parameters in dummy function
daemon: mark unused parameters in non-posix fallbacks
setup: mark unused parameter in config callback
test-mergesort: mark unused parameters in trivial callback
t-hashmap: mark unused parameters in callback function
reftable: mark unused parameters in virtual functions
reftable: drop obsolete test function declarations
reftable: ignore unused argc/argv in test functions
unit-tests: ignore unused argc/argv
t/helper: mark more unused argv/argc arguments
oss-fuzz: mark unused argv/argc argument
refs: mark unused parameters in do_for_each_reflog_helper()
refs: mark unused parameters in ref_store fsck callbacks
update-ref: mark more unused parameters in parser callbacks
imap-send: mark unused parameter in ssl_socket_connect() fallback
Drop unused parameters from functions.
* jk/drop-unused-parameters:
diff-lib: drop unused index argument from get_stat_data()
ref-filter: drop unused parameters from email_atom_option_parser()
pack-bitmap: drop unused parameters from select_pseudo_merges()
pack-bitmap: load writer config from repository parameter
refs: drop some unused parameters from create_symref_lock()
Use of API functions that implicitly depend on the_repository
object in the config subsystem has been rewritten to pass a
repository object through the callchain.
* ps/config-wo-the-repository:
config: hide functions using `the_repository` by default
global: prepare for hiding away repo-less config functions
config: don't depend on `the_repository` with branch conditions
config: don't have setters depend on `the_repository`
config: pass repo to functions that rename or copy sections
config: pass repo to `git_die_config()`
config: pass repo to `git_config_get_expiry_in_days()`
config: pass repo to `git_config_get_expiry()`
config: pass repo to `git_config_get_max_percent_split_change()`
config: pass repo to `git_config_get_split_index()`
config: pass repo to `git_config_get_index_threads()`
config: expose `repo_config_clear()`
config: introduce missing setters that take repo as parameter
path: hide functions using `the_repository` by default
path: stop relying on `the_repository` in `worktree_git_path()`
path: stop relying on `the_repository` when reporting garbage
hooks: remove implicit dependency on `the_repository`
editor: do not rely on `the_repository` for interactive edits
path: expose `do_git_common_path()` as `repo_common_pathv()`
path: expose `do_git_path()` as `repo_git_pathv()`
The reftable code uses a lot of virtual function pointers, but many of
the concrete implementations do not need all of the parameters.
For the most part these are obviously fine to just mark as UNUSED (e.g.,
the empty_iterator functions unsurprisingly do not do anything). Here
are a few cases where I dug a little deeper (but still ended up just
marking them UNUSED):
- the iterator exclude_patterns is best-effort and optional (though it
would be nice to support in the long run as an optimization)
- ignoring the ref_store in many transaction functions is unexpected,
but works because the ref_transaction itself carries enough
information to do what we need.
- ignoring "err" for in some cases (e.g., transaction abort) is OK
because we do not return any errors. It is a little odd for
reftable_be_create_reflog(), though, since we do return errors
there. We should perhaps be creating string error messages at this
layer, but I've punted on that for now.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit ab6f79d8df (refs: set up ref consistency check infrastructure,
2024-08-08) added virtual functions to the ref store for doing fsck
checks. But the packed and reftable backends do not yet do anything.
Let's annotate them to silence -Wunused-parameter.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This function was factored out in 57d0b1e2ea (files-backend: extract out
`create_symref_lock()`, 2024-05-07), but we never look at the ref_store
or refname parameters. We just need the path, which is already contained
in the lockfile struct.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git fsck" infrastructure has been taught to also check the sanity
of the ref database, in addition to the object database.
* sj/ref-fsck:
fsck: add ref name check for files backend
files-backend: add unified interface for refs scanning
builtin/refs: add verify subcommand
refs: set up ref consistency check infrastructure
fsck: add refs report function
fsck: add a unified interface for reporting fsck messages
fsck: make "fsck_error" callback generic
fsck: rename objects-related fsck error functions
fsck: rename "skiplist" to "skip_oids"
The refs API has been taught to give symref target information to
the users of ref iterators, allowing for-each-ref and friends to
avoid an extra ref_resolve_* API call per a symbolic ref.
* jc/refs-symref-referent:
ref-filter: populate symref from iterator
refs: add referent to each_ref_fn
refs: keep track of unresolved reference value in iterators
We're about to hide config functions that implicitly depend on
`the_repository` behind the `USE_THE_REPOSITORY_VARIABLE` macro. This
will uncover a bunch of dependents that transitively relied on the
global variable, but didn't define the macro yet.
Adapt them such that we define the macro to prepare for this change.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a parameter to each_ref_fn so that callers to the ref APIs
that use this function as a callback can have acess to the
unresolved value of a symbolic ref.
Signed-off-by: John Cai <johncai86@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since ref iterators do not hold onto the direct value of a reference
without resolving it, the only way to get ahold of a direct value of a
symbolic ref is to make a separate call to refs_read_symbolic_ref.
To make accessing the direct value of a symbolic ref more efficient,
let's save the direct value of the ref in the iterators for both the
files backend and the reftable backend.
Signed-off-by: John Cai <johncai86@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The git-fsck(1) only implicitly checks the reference, it does not fully
check refs with bad format name such as standalone "@".
However, a file ending with ".lock" should not be marked as having a bad
ref name. It is expected that concurrent writers may have such lock files.
We currently ignore this situation. But for bare ".lock" file, we will
report it as error.
In order to provide such checks, add a new fsck message id "badRefName"
with default ERROR type. Use existing "check_refname_format" to explicit
check the ref name. And add a new unit test to verify the functionality.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
For refs and reflogs, we need to scan its corresponding directories to
check every regular file or symbolic link which shares the same pattern.
Introduce a unified interface for scanning directories for
files-backend.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "struct ref_store" is the base class which contains the "be" pointer
which provides backend-specific functions whose interfaces are defined
in the "ref_storage_be". We could reuse this polymorphism to define only
one interface. For every backend, we need to provide its own function
pointer.
The interfaces defined in the `ref_storage_be` are carefully structured
in semantic. It's organized as the five parts:
1. The name and the initialization interfaces.
2. The ref transaction interfaces.
3. The ref internal interfaces (pack, rename and copy).
4. The ref filesystem interfaces.
5. The reflog related interfaces.
To keep consistent with the git-fsck(1), add a new interface named
"fsck_refs_fn" to the end of "ref_storage_be". This semantic cannot be
grouped into any above five categories. Explicitly add blank line to
make it different from others.
Last, implement placeholder functions for each ref backends.
Mentored-by: Patrick Steinhardt <ps@pks.im>
Mentored-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: shejialuo <shejialuo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano