"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>
Junio C Hamano