At least since v1.7, the elements of the `refs_to_prune` linked list
have been leaked. Fix the leak by teaching `prune_refs()` to free the
list elements as it processes them.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Now that the packed reference store supports transactions, we can use
a transaction to write the packed versions of references that we want
to pack. This decreases the coupling between `files_ref_store` and
`packed_ref_store`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Implement `packed_delete_refs()` using a reference transaction. This
means that `files_delete_refs()` can use `refs_delete_refs()` instead
of `repack_without_refs()` to delete any packed references, decreasing
the coupling between the classes.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Implement the methods needed to support reference transactions for
the packed-refs backend. The new methods are not yet used.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
`packed_ref_store` is going to want to store some transaction-wide
data, so make a place for it.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The old code incremented the packed ref cache reference count when
acquiring the packed-refs lock, and decremented the count when
releasing the lock. This is unnecessary because:
* Another process cannot change the packed-refs file because it is
locked.
* When we ourselves change the packed-refs file, we do so by first
modifying the packed ref-cache, and then writing the data from the
ref-cache to disk. So the packed ref-cache remains fresh because any
changes that we plan to make to the file are made in the cache first
anyway.
So there is no reason for the cache to become stale.
Moreover, the extra reference count causes a problem if we
intentionally clear the packed refs cache, as we sometimes need to do
if we change the cache in anticipation of writing a change to disk,
but then the write to disk fails. In that case, `packed_refs_unlock()`
would have no easy way to find the cache whose reference count it
needs to decrement.
This whole issue will soon become moot due to upcoming changes that
avoid changing the in-memory cache as part of updating the packed-refs
on disk, but this change makes that transition easier.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since the tempfile code recently relaxed the rule that
tempfile structs (and thus locks) need to hang around
forever, we no longer have to leak our lock_file structs.
In fact, we don't even need to heap-allocate them anymore,
since their lifetime can just match that of the surrounding
ref_lock (and if we forget to delete a lock, the effect is
the same as before: it will eventually go away at program
exit).
Note that there is a check in unlock_ref() to only rollback
a lock file if it has been allocated. We don't need that
check anymore; we zero the ref_lock (and thus the
lock_file), so at worst we pass a NULL pointer to
delete_tempfile(), which considers that a noop.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The previous commit taught the tempfile code to give up
ownership over tempfiles that have been renamed or deleted.
That makes it possible to use a stack variable like this:
struct tempfile t;
create_tempfile(&t, ...);
...
if (!err)
rename_tempfile(&t, ...);
else
delete_tempfile(&t);
But doing it this way has a high potential for creating
memory errors. The tempfile we pass to create_tempfile()
ends up on a global linked list, and it's not safe for it to
go out of scope until we've called one of those two
deactivation functions.
Imagine that we add an early return from the function that
forgets to call delete_tempfile(). With a static or heap
tempfile variable, the worst case is that the tempfile hangs
around until the program exits (and some functions like
setup_shallow_temporary rely on this intentionally, creating
a tempfile and then leaving it for later cleanup).
But with a stack variable as above, this is a serious memory
error: the variable goes out of scope and may be filled with
garbage by the time the tempfile code looks at it. Let's
see if we can make it harder to get this wrong.
Since many callers need to allocate arbitrary numbers of
tempfiles, we can't rely on static storage as a general
solution. So we need to turn to the heap. We could just ask
all callers to pass us a heap variable, but that puts the
burden on them to call free() at the right time.
Instead, let's have the tempfile code handle the heap
allocation _and_ the deallocation (when the tempfile is
deactivated and removed from the list).
This changes the return value of all of the creation
functions. For the cleanup functions (delete and rename),
we'll add one extra bit of safety: instead of taking a
tempfile pointer, we'll take a pointer-to-pointer and set it
to NULL after freeing the object. This makes it safe to
double-call functions like delete_tempfile(), as the second
call treats the NULL input as a noop. Several callsites
follow this pattern.
The resulting patch does have a fair bit of noise, as each
caller needs to be converted to handle:
1. Storing a pointer instead of the struct itself.
2. Passing the pointer instead of taking the struct
address.
3. Handling a "struct tempfile *" return instead of a file
descriptor.
We could play games to make this less noisy. For example, by
defining the tempfile like this:
struct tempfile {
struct heap_allocated_part_of_tempfile {
int fd;
...etc
} *actual_data;
}
Callers would continue to have a "struct tempfile", and it
would be "active" only when the inner pointer was non-NULL.
But that just makes things more awkward in the long run.
There aren't that many callers, so we can simply bite
the bullet and adjust all of them. And the compiler makes it
easy for us to find them all.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since the lockfile code is based on the tempfile code, it
has some of the same problems, including that close_lock_file()
erases the tempfile's filename buf, making it hard for the
caller to write a good error message.
In practice this comes up less for lockfiles than for
straight tempfiles, since we usually just report the
refname. But there is at least one buggy case in
write_ref_to_lockfile(). Besides, given the coupling between
the lockfile and tempfile modules, it's less confusing if
their close() functions have the same semantics.
Just as the previous commit did for close_tempfile(), let's
teach close_lock_file() and its wrapper close_ref() not to
rollback on error. And just as before, we'll give them new
"gently" names to catch any new callers that are added.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The code to acquire a lock on a reference (e.g. while accepting a
push from a client) used to immediately fail when the reference is
already locked---now it waits for a very short while and retries,
which can make it succeed if the lock holder was holding it during
a read-only operation.
* mh/ref-lock-entry:
refs: retry acquiring reference locks for 100ms
refs/bisect is unfortunately per-worktree, so we need to look in
per-worktree logs/refs/bisect in addition to per-repo logs/refs. The
current iterator only goes through per-repo logs/refs.
Use merge iterator to walk two ref stores at the same time and pick
per-worktree refs from the right iterator.
PS. Note the unsorted order of for_each_reflog in the test. This is
supposed to be OK, for now. If we enforce order on for_each_reflog()
then some more work will be required.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The philosophy of reference locking has been, "if another process is
changing a reference, then whatever I'm trying to do to it will
probably fail anyway because my old-SHA-1 value is probably no longer
current". But this argument falls down if the other process has locked
the reference to do something that doesn't actually change the value
of the reference, such as `pack-refs` or `reflog expire`. There
actually *is* a decent chance that a planned reference update will
still be able to go through after the other process has released the
lock.
So when trying to lock an individual reference (e.g., when creating
"refs/heads/master.lock"), if it is already locked, then retry the
lock acquisition for approximately 100 ms before giving up. This
should eliminate some unnecessary lock conflicts without wasting a lot
of time.
Add a configuration setting, `core.filesRefLockTimeout`, to allow this
setting to be tweaked.
Note: the function `get_files_ref_lock_timeout_ms()` cannot be private
to the files backend because it is also used by `write_pseudoref()`
and `delete_pseudoref()`, which are defined in `refs.c` so that they
can be used by other reference backends.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "ref-store" code reorganization continues.
* mh/packed-ref-store: (32 commits)
files-backend: cheapen refname_available check when locking refs
packed_ref_store: handle a packed-refs file that is a symlink
read_packed_refs(): die if `packed-refs` contains bogus data
t3210: add some tests of bogus packed-refs file contents
repack_without_refs(): don't lock or unlock the packed refs
commit_packed_refs(): remove call to `packed_refs_unlock()`
clear_packed_ref_cache(): don't protest if the lock is held
packed_refs_unlock(), packed_refs_is_locked(): new functions
packed_refs_lock(): report errors via a `struct strbuf *err`
packed_refs_lock(): function renamed from lock_packed_refs()
commit_packed_refs(): use a staging file separate from the lockfile
commit_packed_refs(): report errors rather than dying
packed_ref_store: make class into a subclass of `ref_store`
packed-backend: new module for handling packed references
packed_read_raw_ref(): new function, replacing `resolve_packed_ref()`
packed_ref_store: support iteration
packed_peel_ref(): new function, extracted from `files_peel_ref()`
repack_without_refs(): take a `packed_ref_store *` parameter
get_packed_ref(): take a `packed_ref_store *` parameter
rollback_packed_refs(): take a `packed_ref_store *` parameter
...
When locking references in preparation for updating them, we need to
check that none of the newly added references D/F conflict with
existing references (e.g., we don't allow `refs/foo` to be added if
`refs/foo/bar` already exists, or vice versa).
Prior to 524a9fdb51 (refs_verify_refname_available(): use function in
more places, 2017-04-16), conflicts with existing loose references
were checked by looking directly in the filesystem, and then conflicts
with existing packed references were checked by running
`verify_refname_available_dir()` against the packed-refs cache.
But that commit changed the final check to call
`refs_verify_refname_available()` against the *whole* files ref-store,
including both loose and packed references, with the following
comment:
> This means that those callsites now check for conflicts with all
> references rather than just packed refs, but the performance cost
> shouldn't be significant (and will be regained later).
That comment turned out to be too sanguine. User s@kazlauskas.me
reported that fetches involving a very large number of references in
neighboring directories were slowed down by that change.
The problem is that when fetching, each reference is updated
individually, within its own reference transaction. This is done
because some reference updates might succeed even though others fail.
But every time a reference update transaction is finished,
`clear_loose_ref_cache()` is called. So when it is time to update the
next reference, part of the loose ref cache has to be repopulated for
the `refs_verify_refname_available()` call. If the references are all
in neighboring directories, then the cost of repopulating the
reference cache increases with the number of references, resulting in
O(N²) effort.
The comment above also claims that the performance cost "will be
regained later". The idea was that once the packed-refs were finished
being split out into a separate ref-store, we could limit the
`refs_verify_refname_available()` call to the packed references again.
That is what we do now.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
One of the tricks that `contrib/workdir/git-new-workdir` plays is to
making `packed-refs` in the new workdir a symlink to the `packed-refs`
file in the original repository. Before
42dfa7ecef ("commit_packed_refs(): use a staging file separate from
the lockfile", 2017-06-23), a lockfile was used as the staging file,
and because the `LOCK_NO_DEREF` was not used, the pointed-to file was
locked and modified.
But after that commit, the staging file was created using a tempfile,
with the end result that rewriting the `packed-refs` file in the
workdir overwrote the symlink rather than the original `packed-refs`
file.
Change `commit_packed_refs()` to use `get_locked_file_path()` to find
the path of the file that it should overwrite. Since that path was
properly resolved when the lockfile was created, this restores the
pre-42dfa7ecef behavior.
Also add a test case to document this use case and prevent a
regression like this from recurring.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The old code ignored any lines that it didn't understand, including
unterminated lines. This is dangerous. Instead, `die()` if the
`packed-refs` file contains any unterminated lines or lines that we
don't know how to handle.
This fixes the tests added in the last commit.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change `repack_without_refs()` to expect the packed-refs lock to be
held already, and not to release the lock before returning. Change the
callers to deal with lock management.
This change makes it possible for callers to hold the packed-refs lock
for a longer span of time, a possibility that will eventually make it
possible to fix some longstanding races.
The only semantic change here is that `repack_without_refs()` used to
forget to release the lock in the `if (!removed)` exit path. That
omission is now fixed.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Bugfix for a topic that is (only) in 'master'.
* mh/packed-ref-store-prep:
for_each_bisect_ref(): don't trim refnames
lock_packed_refs(): fix cache validity check
A common pattern to free a piece of memory and assign NULL to the
pointer that used to point at it has been replaced with a new
FREE_AND_NULL() macro.
* ab/free-and-null:
*.[ch] refactoring: make use of the FREE_AND_NULL() macro
coccinelle: make use of the "expression" FREE_AND_NULL() rule
coccinelle: add a rule to make "expression" code use FREE_AND_NULL()
coccinelle: make use of the "type" FREE_AND_NULL() rule
coccinelle: add a rule to make "type" code use FREE_AND_NULL()
git-compat-util: add a FREE_AND_NULL() wrapper around free(ptr); ptr = NULL
Fix configuration codepath to pay proper attention to commondir
that is used in multi-worktree situation, and isolate config API
into its own header file.
* bw/config-h:
config: don't implicitly use gitdir or commondir
config: respect commondir
setup: teach discover_git_directory to respect the commondir
config: don't include config.h by default
config: remove git_config_iter
config: create config.h
Instead, change the callers of `commit_packed_refs()` to call
`packed_refs_unlock()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The existing callers already check that the lock isn't held just
before calling `clear_packed_ref_cache()`, and in the near future we
want to be able to call this function when the lock is held.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add two new public functions, `packed_refs_unlock()` and
`packed_refs_is_locked()`, with which callers can manage and query the
`packed-refs` lock externally.
Call `packed_refs_unlock()` from `commit_packed_refs()` and
`rollback_packed_refs()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
That way the callers don't have to come up with error messages
themselves.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Rename `lock_packed_refs()` to `packed_refs_lock()` for consistency
with how other methods are named. Also, it's about to get some
companions.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We will want to be able to hold the lockfile for `packed-refs` even
after we have activated the new values. So use a separate tempfile,
`packed-refs.new`, as a place to stage the new contents of the
`packed-refs` file. For now this is all done within
`commit_packed_refs()`, but that will change shortly.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Report errors via a `struct strbuf *err` rather than by calling
`die()`. To enable this goal, change `write_packed_entry()` to report
errors via a return value and `errno` rather than dying.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add the infrastructure to make `packed_ref_store` implement
`ref_store`, at least formally (few of the methods are actually
implemented yet). Change the functions in its interface to take
`ref_store *` arguments. Change `files_ref_store` to store a pointer
to `ref_store *` and to call functions via the virtual `ref_store`
interface where possible. This also means that a few
`packed_ref_store` functions can become static.
This is a work in progress. Some more `ref_store` methods will soon be
implemented (e.g., those having to do with reference transactions).
But some of them will never be implemented (e.g., those having to do
with symrefs or reflogs).
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Now that the interface between `files_ref_store` and
`packed_ref_store` is relatively narrow, move the latter into a new
module, "refs/packed-backend.h" and "refs/packed-backend.c". It still
doesn't quite implement the `ref_store` interface, but it will soon.
This commit moves code around and adjusts its visibility, but doesn't
change anything.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a new function, `packed_read_raw_ref()`, which is nearly a
`read_raw_ref_fn`. Use it in place of `resolve_packed_ref()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add the infrastructure to iterate over a `packed_ref_store`. It's a
lot of boilerplate, but it's all part of a campaign to make
`packed_ref_store` implement `ref_store`. In the future, this iterator
will work much differently.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This will later become a method of `packed_ref_store`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It only cares about the packed-refs part of the reference store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the `packed_refs_lock` member from `files_ref_store` to
`packed_ref_store`, and rename it to `lock` since it's now more
obvious what it is locking.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move `packed_refs_path` from `files_ref_store` to `packed_ref_store`,
and rename it to `path` since its meaning is clear from its new
context.
Inline `files_packed_refs_path()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Start extracting the packed-refs-related data structures into a new
class, `packed_ref_store`. It doesn't yet implement `ref_store`, but
it will.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Teach `add_packed_ref()` to overwrite an existing entry if one already
exists for the specified `refname`. This means that we can call it
from `files_pack_refs()`, thereby reducing the amount that the latter
function needs to know about the internals of packed-reference
handling.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add the ability to --copy a branch and its reflog and configuration,
this uses the same underlying machinery as the --move (-m) option
except the reflog and configuration is copied instead of being moved.
This is useful for e.g. copying a topic branch to a new version,
e.g. work to work-2 after submitting the work topic to the list, while
preserving all the tracking info and other configuration that goes
with the branch, and unlike --move keeping the other already-submitted
branch around for reference.
Like --move, when the source branch is the currently checked out
branch the HEAD is moved to the destination branch. In the case of
--move we don't really have a choice (other than remaining on a
detached HEAD) and in order to keep the functionality consistent, we
are doing it in similar way for --copy too.
The most common usage of this feature is expected to be moving to a
new topic branch which is a copy of the current one, in that case
moving to the target branch is what the user wants, and doesn't
unexpectedly behave differently than --move would.
One outstanding caveat of this implementation is that:
git checkout maint &&
git checkout master &&
git branch -c topic &&
git checkout -
Will check out 'maint' instead of 'master'. This is because the @{-N}
feature (or its -1 shorthand "-") relies on HEAD reflogs created by
the checkout command, so in this case we'll checkout maint instead of
master, as the user might expect. What to do about that is left to a
future change.
Helped-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Sahil Dua <sahildua2305@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Replace occurrences of `free(ptr); ptr = NULL` which weren't caught by
the coccinelle rule. These fall into two categories:
- free/NULL assignments one after the other which coccinelle all put
on one line, which is functionally equivalent code, but very ugly.
- manually spotted occurrences where the NULL assignment isn't right
after the free() call.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Apply the result of the just-added coccinelle rule. This manually
excludes a few occurrences, mostly things that resulted in many
FREE_AND_NULL() on one line, that'll be manually fixed in a subsequent
change.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stop including config.h by default in cache.h. Instead only include
config.h in those files which require use of the config system.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 28ed9830b1 (get_packed_ref_cache(): assume "packed-refs" won't
change while locked, 2017-05-22) assumes that the "packed-refs" file
cannot change while we hold the lock. That assumption is
justified *if* the lock has been held the whole time since the
"packed-refs" file was last read.
But in `lock_packed_refs()`, we ourselves lock the "packed-refs" file
and then call `get_packed_ref_cache()` to ensure that the cache agrees
with the file. The intent is to guard against the possibility that
another process changed the "packed-refs" file the moment before we
locked it.
This check was defeated because `get_packed_ref_cache()` saw that the
file was locked, and therefore didn't do the `stat_validity_check()`
that we want.
The mistake was compounded with a misleading comment in
`lock_packed_refs()` claiming that it was doing the right thing. That
comment came from an earlier draft of the mh/packed-ref-store-prep
patch series when the commits were in a different order.
So instead:
* Extract a function `validate_packed_ref_cache()` that does the
validity check independent of whether the lock is held.
* Change `get_packed_ref_cache()` to call the new function, but only
if the lock *isn't* held.
* Change `lock_packed_refs()` to call the new function in any case
before calling `get_packed_ref_cache()`.
* Fix the comment in `lock_packed_refs()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When iterating over references, reference priming is used to make sure
that loose references are read into the ref-cache before packed
references, to avoid races. It used to be that the prefix passed to
reference iterators almost always ended in `/`, for example
`refs/heads/`. In that case, the priming code would read all loose
references under `find_containing_dir("refs/heads/")`, which is
"refs/heads/". That's just what we want.
But now that `ref-filter` knows how to pass refname prefixes to
`for_each_fullref_in()`, the prefix might come from user input; for
example,
git for-each-ref refs/heads
Since the argument doesn't include a trailing slash, the reference
iteration code would prime all of the loose references under
`find_containing_dir("refs/heads")`, which is "refs/". Thus we would
unnecessarily read tags, remote-tracking references, etc., when the
user is only interested in branches.
It is a bit awkward to get around this problem. We can't just append a
slash to the argument, because we don't know ab initio whether an
argument like `refs/tags/release` corresponds to a single tag or to a
directory containing tags.
Moreover, until now a `prefix_ref_iterator` was used to make the final
decision about which references fall within the prefix (the
`cache_ref_iterator` only did a rough cut). This is also inefficient,
because the `prefix_ref_iterator` can't know, for example, that while
you are in a subdirectory that is completely within the prefix, you
don't have to do the prefix check.
So:
* Move the responsibility for doing the prefix check directly to
`cache_ref_iterator`. This means that `cache_ref_iterator_begin()`
never has to wrap its return value in a `prefix_ref_iterator`.
* Teach `cache_ref_iterator_begin()` (and `prime_ref_dir()`) to be
stricter about what they iterate over and what directories they
prime.
* Teach `cache_ref_iterator` to keep track of whether the current
`cache_ref_iterator_level` is fully within the prefix. If so, skip
the prefix checks entirely.
The main benefit of these optimizations is for loose references, since
packed references are always read all at once.
Note that after this change, `prefix_ref_iterator` is only ever used
for its trimming feature and not for its "prefix" feature. But I'm not
ripping out the latter yet, because it might be useful for another
patch series that I'm working on.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Only one caller was using it, so move the check to that caller.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Instead of handling `GIT_REF_PARANOIA` in
`files_ref_iterator_begin()`, handle it in
`refs_ref_iterator_begin()`, where it will cover all reference stores.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The old code ignored any errors encountered when trying to fopen the
"packed-refs" file, treating all such failures as if the file didn't
exist. But it could be that there is some other error opening the
file (e.g., permissions problems), and we don't want to silently
ignore such problems. So report any failures that are not due to
ENOENT.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Teach `read_packed_refs()` to also
* Allocate and initialize the new `packed_ref_cache`
* Open and close the `packed-refs` file
* Update the `validity` field of the new object
This decreases the coupling between `packed_refs_cache` and
`files_ref_store` by a little bit.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If we've got the "packed-refs" file locked, then it can't change;
there's no need to keep calling `stat_validity_check()` on it.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Extract a function for deciding whether a reference should be packed.
It is a self-contained bit of logic, so splitting it out improves
readability.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the future, compound reference stores will sometimes need to modify
references in two different reference stores at the same time, meaning
that a single logical reference transaction might have to be
implemented as two internal sub-transactions. They won't want to call
`ref_transaction_commit()` for the two sub-transactions one after the
other, because that wouldn't be atomic (the first commit could succeed
and the second one fail). Instead, they will want to prepare both
sub-transactions (i.e., obtain any necessary locks and do any
pre-checks), and only if both prepare steps succeed, then commit both
sub-transactions.
Start preparing for that day by adding a new, optional
`ref_transaction_prepare()` step to the reference transaction
sequence, which obtains the locks and does any prechecks, reporting
any errors that occur. Also add a `ref_transaction_abort()` function
that can be used to abort a sub-transaction even if it has already
been prepared.
That is on the side of the public-facing API. On the side of the
`ref_store` VTABLE, get rid of `transaction_commit` and instead add
methods `transaction_prepare`, `transaction_finish`, and
`transaction_abort`. A `ref_transaction_commit()` now basically calls
methods `transaction_prepare` then `transaction_finish`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the check that `transaction->state` is valid from
`files_transaction_commit()` to `ref_transaction_commit()`, where
other future reference backends can benefit from it as well.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Extract the cleanup functionality from `files_transaction_commit()`
into a new function. It will soon have another caller.
Use the common cleanup code even on early exit if the transaction is
empty, to reduce code duplication.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Instead of using a global `lock_file` instance for the main
"packed-refs" file and using a pointer in `files_ref_store` to keep
track of whether it is locked, embed the `lock_file` instance directly
in the `files_ref_store` struct and use the new
`is_lock_file_locked()` function to keep track of whether it is
locked. This keeps related data together and makes the main reference
store less of a special case.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the `lock` member from `packed_ref_cache` to `files_ref_store`,
since at most one cache can have a locked "packed-refs" file
associated with it. Rename it to `packed_refs_lock` to make its
purpose clearer in its new home. More changes are coming here shortly.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Just because the files backend can't retain reflogs for deleted
references is no reason that they shouldn't be supported by the
virtual method interface. Also, `delete_ref()` and `refs_delete_ref()`
have already gained `msg` parameters. Now let's add them to
`delete_refs()` and `refs_delete_refs()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Eliminate any chance of integer overflow on platforms where the two
types have different sizes.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `trim` parameter can be set independently of `prefix`. So if some
caller were to set `trim` to be greater than `strlen(prefix)`, we
could end up pointing the `refname` field of the iterator past the NUL
of the actual reference name string.
That can't happen currently, because `trim` is always set either to
zero or to `strlen(prefix)`. But even the latter could lead to
confusion, if a refname is exactly equal to the prefix, because then
we would set the outgoing `refname` to the empty string.
And we're about to decouple the `prefix` and `trim` arguments even
more, so let's be cautious here. Report a bug if ever asked to trim a
reference whose name is not longer than `trim`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The former is by far more common in our codebase.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The iterator returned by this function only includes references whose
names start with the whole prefix, not all of those in
`find_containing_dir(prefix)` as the old docstring claimed. This
docstring was probably copy-pasted from old ref-cache code, which had
the old specification. But now, `cache_ref_iterator_begin()`
(from which the files reference iterator gets its values)
automatically wraps its output using `prefix_ref_iterator_begin()`
when necessary, so it has the stricter behavior.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* bc/object-id: (53 commits)
object: convert parse_object* to take struct object_id
tree: convert parse_tree_indirect to struct object_id
sequencer: convert do_recursive_merge to struct object_id
diff-lib: convert do_diff_cache to struct object_id
builtin/ls-tree: convert to struct object_id
merge: convert checkout_fast_forward to struct object_id
sequencer: convert fast_forward_to to struct object_id
builtin/ls-files: convert overlay_tree_on_cache to object_id
builtin/read-tree: convert to struct object_id
sha1_name: convert internals of peel_onion to object_id
upload-pack: convert remaining parse_object callers to object_id
revision: convert remaining parse_object callers to object_id
revision: rename add_pending_sha1 to add_pending_oid
http-push: convert process_ls_object and descendants to object_id
refs/files-backend: convert many internals to struct object_id
refs: convert struct ref_update to use struct object_id
ref-filter: convert some static functions to struct object_id
Convert struct ref_array_item to struct object_id
Convert the verify_pack callback to struct object_id
Convert lookup_tag to struct object_id
...
Some platforms have ulong that is smaller than time_t, and our
historical use of ulong for timestamp would mean they cannot
represent some timestamp that the platform allows. Invent a
separate and dedicated timestamp_t (so that we can distingiuish
timestamps and a vanilla ulongs, which along is already a good
move), and then declare uintmax_t is the type to be used as the
timestamp_t.
* js/larger-timestamps:
archive-tar: fix a sparse 'constant too large' warning
use uintmax_t for timestamps
date.c: abort if the system time cannot handle one of our timestamps
timestamp_t: a new data type for timestamps
PRItime: introduce a new "printf format" for timestamps
parse_timestamp(): specify explicitly where we parse timestamps
t0006 & t5000: skip "far in the future" test when time_t is too limited
t0006 & t5000: prepare for 64-bit timestamps
ref-filter: avoid using `unsigned long` for catch-all data type
"git gc" did not interact well with "git worktree"-managed
per-worktree refs.
* nd/worktree-kill-parse-ref:
refs: kill set_worktree_head_symref()
worktree.c: kill parse_ref() in favor of refs_resolve_ref_unsafe()
refs: introduce get_worktree_ref_store()
refs: add REFS_STORE_ALL_CAPS
refs.c: make submodule ref store hashmap generic
environment.c: fix potential segfault by get_git_common_dir()
Convert many of the internals of the files backend to use struct
object_id. Avoid converting public APIs (except one change to
refs/ref-cache.c) to limit the scope of the changes.
Convert one use of get_sha1_hex to parse_oid_hex, and rely on the fact
that a strbuf will be NUL-terminated and that parse_oid_hex will fail on
truncated input to avoid the need to check the length.
This is a requirement to convert parse_object later on.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert struct ref_array_item to use struct object_id by changing the
definition and applying the following semantic patch, plus the standard
object_id transforms:
@@
struct ref_update E1;
@@
- E1.new_sha1
+ E1.new_oid.hash
@@
struct ref_update *E1;
@@
- E1->new_sha1
+ E1->new_oid.hash
@@
struct ref_update E1;
@@
- E1.old_sha1
+ E1.old_oid.hash
@@
struct ref_update *E1;
@@
- E1->old_sha1
+ E1->old_oid.hash
This transformation allows us to convert write_ref_to_lockfile, which is
required to convert parse_object.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Adjust the callback functions to take struct object_id * instead of
unsigned char *, and modify related static functions accordingly.
Introduce a temporary object_id instance into files_reflog_expire and
copy the SHA-1 value passed in. This is necessary because the sha1
parameter can come indirectly from get_sha1. Without the temporary, it
would require much more refactoring to be able to convert this function.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Git's source code assumes that unsigned long is at least as precise as
time_t. Which is incorrect, and causes a lot of problems, in particular
where unsigned long is only 32-bit (notably on Windows, even in 64-bit
versions).
So let's just use a more appropriate data type instead. In preparation
for this, we introduce the new `timestamp_t` data type.
By necessity, this is a very, very large patch, as it has to replace all
timestamps' data type in one go.
As we will use a data type that is not necessarily identical to `time_t`,
we need to be very careful to use `time_t` whenever we interact with the
system functions, and `timestamp_t` everywhere else.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The internals of the refs API around the cached refs has been
streamlined.
* mh/separate-ref-cache:
do_for_each_entry_in_dir(): delete function
files_pack_refs(): use reference iteration
commit_packed_refs(): use reference iteration
cache_ref_iterator_begin(): make function smarter
get_loose_ref_cache(): new function
get_loose_ref_dir(): function renamed from get_loose_refs()
do_for_each_entry_in_dir(): eliminate `offset` argument
refs: handle "refs/bisect/" in `loose_fill_ref_dir()`
ref-cache: use a callback function to fill the cache
refs: record the ref_store in ref_cache, not ref_dir
ref-cache: introduce a new type, ref_cache
refs: split `ref_cache` code into separate files
ref-cache: rename `remove_entry()` to `remove_entry_from_dir()`
ref-cache: rename `find_ref()` to `find_ref_entry()`
ref-cache: rename `add_ref()` to `add_ref_entry()`
refs_verify_refname_available(): use function in more places
refs_verify_refname_available(): implement once for all backends
refs_ref_iterator_begin(): new function
refs_read_raw_ref(): new function
get_ref_dir(): don't call read_loose_refs() for "refs/bisect"
70999e9cec (branch -m: update all per-worktree HEADs - 2016-03-27)
added this function in order to update HEADs of all relevant
worktrees, when a branch is renamed.
It, as a public ref api, kind of breaks abstraction when it uses
internal functions of files backend. With the introduction of
refs_create_symref(), we can move back pretty close to the code before
70999e9cec, where create_symref() was used for updating HEAD.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Various small fixes.
* rs/misc-cppcheck-fixes:
server-info: avoid calling fclose(3) twice in update_info_file()
files_for_each_reflog_ent_reverse(): close stream and free strbuf on error
am: close stream on error, but not stdin
Currently, Git's source code treats all timestamps as if they were
unsigned longs. Therefore, it is okay to write "%lu" when printing them.
There is a substantial problem with that, though: at least on Windows,
time_t is *larger* than unsigned long, and hence we will want to switch
away from the ill-specified `unsigned long` data type.
So let's introduce the pseudo format "PRItime" (currently simply being
defined to "lu") to make it easier to change the data type used for
timestamps.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently, Git's source code represents all timestamps as `unsigned
long`. In preparation for using a more appropriate data type, let's
introduce a symbol `parse_timestamp` (currently being defined to
`strtoul`) where appropriate, so that we can later easily switch to,
say, use `strtoull()` instead.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Exit the loop orderly through the cleanup code, instead of dashing out
with logfp still open and sb leaking.
Found with Cppcheck.
Signed-off-by: Rene Scharfe <l.s.r@web.de>
Reviewed-by: Jeff King <peff@peff.net>
Reviewed-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use reference iteration rather than `do_for_each_entry_in_dir()` in
the definition of `files_pack_refs()`. This makes the code shorter and
easier to follow, because the logic can be inline rather than spread
between the main function and a callback function, and it removes the
need to use `pack_refs_cb_data` to preserve intermediate state.
This removes the last callers of `entry_resolves_to_object()` and
`get_loose_ref_dir()`, so delete those functions.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use reference iteration rather than do_for_each_entry_in_dir() in the
definition of commit_packed_refs().
Note that an internal consistency check that was previously done in
`write_packed_entry_fn()` is not there anymore. This is actually an
improvement:
The old error message was emitted when there is an entry in the
packed-ref cache that is not `REF_KNOWS_PEELED`, and when we attempted
to peel the reference, the result was `PEEL_INVALID`,
`PEEL_IS_SYMREF`, or `PEEL_BROKEN`. Since a packed ref cannot be a
symref, `PEEL_IS_SYMREF` and `PEEL_BROKEN` can be ruled out. So we're
left with `PEEL_INVALID`.
An entry without `REF_KNOWS_PEELED` can get into the packed-refs cache
in the following two ways:
* The reference was read from a `packed-refs` file that didn't have
the `fully-peeled` attribute. In that case, we *don't want* to emit
an error, because the broken value is presumably a stale value of
the reference that is now masked by a loose version of the same
reference (which we just don't happen to be packing this time). This
is a perfectly legitimate situation and doesn't indicate that the
repository is corrupt. The old code incorrectly emits an error
message in this case. (It was probably never reported as a bug
because this scenario is rare.)
* The reference was a loose reference that was just added to the
packed ref cache by `files_packed_refs()` via
`pack_if_possible_fn()` in preparation for being packed. The latter
function refuses to pack a reference for which
`entry_resolves_to_object()` returns false, and otherwise calls
`peel_entry()` itself and checks the return value. So an entry added
this way should always have `REF_KNOWS_PEELED` and shouldn't trigger
the error message in either the old code or the new.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change `cache_ref_iterator_begin()` to take two new arguments:
* `prefix` -- to iterate only over references with the specified
prefix.
* `prime_dir` -- to "prime" (i.e., pre-load) the cache before starting
the iteration.
The new functionality makes it possible for
`files_ref_iterator_begin()` to be made more ignorant of the internals
of `ref_cache`, and `find_containing_dir()` and `prime_ref_dir()` to
be made private.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Extract a new function, `get_loose_ref_cache()`, from
get_loose_ref_dir(). The function returns the `ref_cache` for the
loose refs of a `files_ref_store`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The new name is more analogous to `get_packed_ref_dir()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
That "refs/bisect/" has to be handled specially when filling the
ref_cache for loose references is a peculiarity of the files backend,
and the ref-cache code shouldn't need to know about it. So move this
code to the callback function, `loose_fill_ref_dir()`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It is a leveling violation for `ref_cache` to know about
`files_ref_store` or that it should call `read_loose_refs()` to lazily
fill cache directories. So instead, have its constructor take as an
argument a callback function that it should use for lazy-filling, and
change `files_ref_store` to supply a pointer to function
`read_loose_refs` (renamed to `loose_fill_ref_dir`) when creating the
ref cache for its loose refs.
This means that we can generify the type of the back-pointer in
`struct ref_cache` from `files_ref_store` to `ref_store`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Instead of keeping a pointer to the `ref_store` in every `ref_dir`
entry, store it once in `struct ref_cache`, and change `struct
ref_dir` to include a pointer to its containing `ref_cache` instead.
This makes it easier to add to the information that is accessible from
a `ref_dir` without increasing the size of every `ref_dir` instance.
Note that previously, every `ref_dir` pointed at the containing
`files_ref_store` regardless of whether it was a part of the loose or
packed reference cache. Now we have to be sure to initialize the
instances to point at the correct containing `ref_cache`. So change
`create_dir_entry()` to take a `ref_cache` parameter, and change its
callers to pass the correct `ref_cache` depending on the purpose of
the new `dir_entry`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
For now, it just wraps a `ref_entry *` that points at the root of the
tree. Soon it will hold more information.
Add two new functions, `create_ref_cache()` and `free_ref_cache()`.
Make `free_ref_entry()` private.
Change files-backend to use this type to hold its caches.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `ref_cache` code is currently too tightly coupled to
`files-backend`, making the code harder to understand and making it
awkward for new code to use `ref_cache` (as we indeed have planned).
Start loosening that coupling by splitting `ref_cache` into a separate
module.
This commit moves code, adds declarations, and changes the visibility
of some functions, but doesn't change any code.
The modules are still too tightly coupled, but the situation will be
improved in subsequent commits.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This function's visibility is about to be increased, so give it a more
distinctive name.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This function's visibility is about to be increased, so give it a more
distinctive name.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This function's visibility is about to be increased, so give it a more
distinctive name.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change `lock_raw_ref()` and `lock_ref_sha1_basic()` to use
`refs_verify_refname_available()` instead of
`verify_refname_available_dir()`. This means that those callsites now
check for conflicts with all references rather than just packed refs,
but the performance cost shouldn't be significant (and will be
regained later).
These were the last callers of `verify_refname_available_dir()`, so
also delete that (very complicated) function.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
