files-backend: don't rewrite the `packed-refs` file unnecessarily
Even when we are deleting references, we needn't overwrite the
`packed-refs` file if the references that we are deleting only exist
as loose references. Implement this optimization as follows:
* Add a function `is_packed_transaction_needed()`, which checks
whether a given packed-refs transaction actually needs to be carried
out (i.e., it returns false if the transaction obviously wouldn't
have any effect). This function must be called while holding the
`packed-refs` lock to avoid races.
* Change `files_transaction_prepare()` to check whether the
packed-refs transaction is actually needed. If not, squelch it, but
continue holding the `packed-refs` lock until the end of the
transaction to avoid races.
This fixes a mild regression caused by dc39e09942 (files_ref_store:
use a transaction to update packed refs, 2017-09-08). Before that
commit, unnecessary rewrites of `packed-refs` were suppressed by
`repack_without_refs()`. But the transaction-based writing introduced
by that commit didn't perform that optimization.
Note that the pre-dc39e09942 code still had to *read* the whole
`packed-refs` file to determine that the rewrite could be skipped, so
the performance for the cases that the write could be elided was
`O(N)` in the number of packed references both before and after dc39e09942. But after that commit the constant factor increased.
This commit reimplements the optimization of eliding unnecessary
`packed-refs` rewrites. That, plus the fact that since cfa2e29c34 (packed_ref_store: get rid of the `ref_cache` entirely,
2017-03-17) we don't necessarily have to read the whole `packed-refs`
file at all, means that deletes of one or a few loose references can
now be done with `O(n lg N)` effort, where `n` is the number of loose
references being deleted and `N` is the total number of packed
references.
This commit fixes two tests in t1409.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t1409: check that `packed-refs` is not rewritten unnecessarily
There is no need to rewrite the `packed-refs` file except for the case
that we are deleting a reference that has a packed version. Verify
that `packed-refs` is not rewritten when it shouldn't be.
In fact, two of these tests fail:
* A new (empty) `packed-refs` file is created when deleting any loose
reference and no `packed-refs` file previously existed.
* The `packed-refs` file is rewritten unnecessarily when deleting a
loose reference that has no packed counterpart.
Both problems will be fixed in the next commit.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
files_transaction_finish(): delete reflogs before references
If the deletion steps unexpectedly fail, it is less bad to leave a
reference without its reflog than it is to leave a reflog without its
reference, since the latter is an invalid repository state.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Now the outside world interacts with the packed ref store only via the
generic refs API plus a few lock-related functions. This allows us to
delete some functions that are no longer used, thereby completing the
encapsulation of the packed ref store.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
files_ref_store: use a transaction to update packed refs
When processing a `files_ref_store` transaction, it is sometimes
necessary to delete some references from the "packed-refs" file. Do
that using a reference transaction conducted against the
`packed_ref_store`.
This change further decouples `files_ref_store` from
`packed_ref_store`. It also fixes multiple problems, including the two
revealed by test cases added in the previous commit.
First, the old code didn't obtain the `packed-refs` lock until
`files_transaction_finish()`. This means that a failure to acquire the
`packed-refs` lock (e.g., due to contention with another process)
wasn't detected until it was too late (problems like this are supposed
to be detected in the "prepare" phase). The new code acquires the
`packed-refs` lock in `files_transaction_prepare()`, the same stage of
the processing when the loose reference locks are being acquired,
removing another reason why the "prepare" phase might succeed and the
"finish" phase might nevertheless fail.
Second, the old code deleted the loose version of a reference before
deleting any packed version of the same reference. This left a moment
when another process might think that the packed version of the
reference is current, which is incorrect. (Even worse, the packed
version of the reference can be arbitrarily old, and might even point
at an object that has since been garbage-collected.)
Third, if a reference deletion fails to acquire the `packed-refs` lock
altogether, then the old code might leave the repository in the
incorrect state (possibly corrupt) described in the previous
paragraph.
Now we activate the new "packed-refs" file (sans any references that
are being deleted) *before* deleting the corresponding loose
references. But we hold the "packed-refs" lock until after the loose
references have been finalized, thus preventing a simultaneous
"pack-refs" process from packing the loose version of the reference in
the time gap, which would otherwise defeat our attempt to delete it.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t1404: demonstrate two problems with reference transactions
Currently, a loose reference is deleted even before locking the
`packed-refs` file, let alone deleting any packed version of the
reference. This leads to two problems, demonstrated by two new tests:
* While a reference is being deleted, other processes might see the
old, packed value of the reference for a moment before the packed
version is deleted. Normally this would be hard to observe, but we
can prolong the window by locking the `packed-refs` file externally
before running `update-ref`, then unlocking it before `update-ref`'s
attempt to acquire the lock times out.
* If the `packed-refs` file is locked so long that `update-ref` fails
to lock it, then the reference can be left permanently in the
incorrect state described in the previous point.
In a moment, both problems will be fixed.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
files_initial_transaction_commit(): use a transaction for packed refs
Use a `packed_ref_store` transaction in the implementation of
`files_initial_transaction_commit()` rather than using internal
features of the packed ref store. This further decouples
`files_ref_store` from `packed_ref_store`.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
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>
files_pack_refs(): use a reference transaction to write packed refs
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>
packed-backend: don't adjust the reference count on lock/unlock
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>
Killing "git merge --edit" before the editor returns control left
the repository in a state with MERGE_MSG but without MERGE_HEAD,
which incorrectly tells the subsequent "git commit" that there was
a squash merge in progress. This has been fixed.
* mg/killed-merge:
merge: save merge state earlier
merge: split write_merge_state in two
merge: clarify call chain
Documentation/git-merge: explain --continue
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
"[gc] rerereResolved = 5.days" used to be invalid, as the variable
is defined to take an integer counting the number of days. It now
is allowed.
* jc/cutoff-config:
rerere: allow approxidate in gc.rerereResolved/gc.rerereUnresolved
rerere: represent time duration in timestamp_t internally
t4200: parameterize "rerere gc" custom expiry test
t4200: gather "rerere gc" together
t4200: make "rerere gc" test more robust
t4200: give us a clean slate after "rerere gc" tests
We used to spend more than necessary cycles allocating and freeing
piece of memory while writing each index entry out. This has been
optimized.
* kw/write-index-reduce-alloc:
read-cache: avoid allocating every ondisk entry when writing
read-cache: fix memory leak in do_write_index
perf: add test for writing the index
"git apply" that is used as a better "patch -p1" failed to apply a
taken from a file with CRLF line endings to a file with CRLF line
endings. The root cause was because it misused convert_to_git()
that tried to do "safe-crlf" processing by looking at the index
entry at the same path, which is a nonsense---in that mode, "apply"
is not working on the data in (or derived from) the index at all.
This has been fixed.
* tb/apply-with-crlf:
apply: file commited with CRLF should roundtrip diff and apply
convert: add SAFE_CRLF_KEEP_CRLF
Test update to improve coverage for "git stash" operations.
* jt/stash-tests:
stash: add a test for stashing in a detached state
stash: add a test for when apply fails during stash branch
stash: add a test for stash create with no files
"git interpret-trailers" has been taught a "--parse" and a few
other options to make it easier for scripts to grab existing
trailer lines from a commit log message.
* jk/trailers-parse:
doc/interpret-trailers: fix "the this" typo
pretty: support normalization options for %(trailers)
t4205: refactor %(trailers) tests
pretty: move trailer formatting to trailer.c
interpret-trailers: add --parse convenience option
interpret-trailers: add an option to unfold values
interpret-trailers: add an option to show only existing trailers
interpret-trailers: add an option to show only the trailers
trailer: put process_trailers() options into a struct
A handful of bugfixes and an improvement to "diff --color-moved".
* jt/diff-color-move-fix:
diff: define block by number of alphanumeric chars
diff: respect MIN_BLOCK_LENGTH for last block
diff: avoid redundantly clearing a flag
"git diff" has been taught to optionally paint new lines that are
the same as deleted lines elsewhere differently from genuinely new
lines.
* sb/diff-color-move: (25 commits)
diff: document the new --color-moved setting
diff.c: add dimming to moved line detection
diff.c: color moved lines differently, plain mode
diff.c: color moved lines differently
diff.c: buffer all output if asked to
diff.c: emit_diff_symbol learns about DIFF_SYMBOL_SUMMARY
diff.c: emit_diff_symbol learns about DIFF_SYMBOL_STAT_SEP
diff.c: convert word diffing to use emit_diff_symbol
diff.c: convert show_stats to use emit_diff_symbol
diff.c: convert emit_binary_diff_body to use emit_diff_symbol
submodule.c: migrate diff output to use emit_diff_symbol
diff.c: emit_diff_symbol learns DIFF_SYMBOL_REWRITE_DIFF
diff.c: emit_diff_symbol learns about DIFF_SYMBOL_BINARY_FILES
diff.c: emit_diff_symbol learns DIFF_SYMBOL_HEADER
diff.c: emit_diff_symbol learns DIFF_SYMBOL_FILEPAIR_{PLUS, MINUS}
diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_INCOMPLETE
diff.c: emit_diff_symbol learns DIFF_SYMBOL_WORDS[_PORCELAIN]
diff.c: migrate emit_line_checked to use emit_diff_symbol
diff.c: emit_diff_symbol learns DIFF_SYMBOL_NO_LF_EOF
diff.c: emit_diff_symbol learns DIFF_SYMBOL_CONTEXT_FRAGINFO
...
Updates to the HTTP layer we made recently unconditionally used
features of libCurl without checking the existence of them, causing
compilation errors, which has been fixed. Also migrate the code to
check feature macros, not version numbers, to cope better with
libCurl that vendor ships with backported features.
* tc/curl-with-backports:
http: use a feature check to enable GSSAPI delegation control
http: fix handling of missing CURLPROTO_*
When handshake with a subprocess filter notices that the process
asked for an unknown capability, Git did not report what program
the offending subprocess was running. This has been corrected.
* cc/subprocess-handshake-missing-capabilities:
sub-process: print the cmd when a capability is unsupported
Both sha1_file.c and packfile.c now need read_object(), so a copy of
read_object() was created in packfile.c.
This patch makes both mark_bad_packed_object() and has_packed_and_bad()
global. Unlike most of the other patches in this series, these 2
functions need to remain global.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently, sha1_file.c and cache.h contain many functions, both related
to and unrelated to packfiles. This makes both files very large and
causes an unclear separation of concerns.
Create a new file, packfile.c, to hold all packfile-related functions
currently in sha1_file.c. It has a corresponding header packfile.h.
In this commit, the pack name-related functions are moved. Subsequent
commits will move the other functions.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jt/subprocess-handshake:
sub-process: refactor handshake to common function
Documentation: migrate sub-process docs to header
convert: add "status=delayed" to filter process protocol
convert: refactor capabilities negotiation
convert: move multiple file filter error handling to separate function
convert: put the flags field before the flag itself for consistent style
t0021: write "OUT <size>" only on success
t0021: make debug log file name configurable
t0021: keep filter log files on comparison
Merge branch 'dl/credential-cache-socket-in-xdg-cache' into maint
A recently added test for the "credential-cache" helper revealed
that EOF detection done around the time the connection to the cache
daemon is torn down were flaky. This was fixed by reacting to
ECONNRESET and behaving as if we got an EOF.
* dl/credential-cache-socket-in-xdg-cache:
credential-cache: interpret an ECONNRESET as an EOF
When a directory is not readable, "gitweb" fails to build the
project list. Work this around by skipping such a directory.
It might end up hiding a problem under the rug and a better
solution might be to loudly complain to the administrator pointing
out the problematic directory, but this will at least make it
"work".
Numerous bugs in walking of reflogs via "log -g" and friends have
been fixed.
* jk/reflog-walk:
reflog-walk: apply --since/--until to reflog dates
reflog-walk: stop using fake parents
rev-list: check reflog_info before showing usage
get_revision_1(): replace do-while with an early return
log: do not free parents when walking reflog
log: clarify comment about reflog cycles
revision: disallow reflog walking with revs->limited
t1414: document some reflog-walk oddities
Merge branch 'jc/http-sslkey-and-ssl-cert-are-paths' into maint
The http.{sslkey,sslCert} configuration variables are to be
interpreted as a pathname that honors "~[username]/" prefix, but
weren't, which has been fixed.
* jc/http-sslkey-and-ssl-cert-are-paths:
http.c: http.sslcert and http.sslkey are both pathnames
"%C(color name)" in the pretty print format always produced ANSI
color escape codes, which was an early design mistake. They now
honor the configuration (e.g. "color.ui = never") and also tty-ness
of the output medium.
* jk/ref-filter-colors:
ref-filter: consult want_color() before emitting colors
pretty: respect color settings for %C placeholders
rev-list: pass diffopt->use_colors through to pretty-print
for-each-ref: load config earlier
color: check color.ui in git_default_config()
ref-filter: pass ref_format struct to atom parsers
ref-filter: factor out the parsing of sorting atoms
ref-filter: make parse_ref_filter_atom a private function
ref-filter: provide a function for parsing sort options
ref-filter: move need_color_reset_at_eol into ref_format
ref-filter: abstract ref format into its own struct
ref-filter: simplify automatic color reset
t: use test_decode_color rather than literal ANSI codes
docs/for-each-ref: update pointer to color syntax
check return value of verify_ref_format()
Merge branch 'js/git-gui-msgfmt-on-windows' into maint
Because recent Git for Windows do come with a real msgfmt, the
build procedure for git-gui has been updated to use it instead of a
hand-rolled substitute.
* js/git-gui-msgfmt-on-windows:
git-gui (MinGW): make use of MSys2's msgfmt
git gui: allow for a long recentrepo list
git gui: de-dup selected repo from recentrepo history
git gui: cope with duplicates in _get_recentrepo
git-gui: remove duplicate entries from .gitconfig's gui.recentrepo
"git commit" used to discard the index and re-read from the filesystem
just in case the pre-commit hook has updated it in the middle; this
has been optimized out when we know we do not run the pre-commit hook.
* kw/commit-keep-index-when-pre-commit-is-not-run:
commit: skip discarding the index if there is no pre-commit hook
"git rebase", especially when it is run by mistake and ends up
trying to replay many changes, spent long time in silence. The
command has been taught to show progress report when it spends
long time preparing these many changes to replay (which would give
the user a chance to abort with ^C).
* kw/rebase-progress:
rebase: turn on progress option by default for format-patch
format-patch: have progress option while generating patches
"git stash -u" used the contents of the committed version of the
".gitignore" file to decide which paths are ignored, even when the
file has local changes. The command has been taught to instead use
the locally modified contents.
* nm/stash-untracked:
stash: clean untracked files before reset
vcs-svn: move remaining repo_tree functions to fast_export.h
These used to be for manipulating the in-memory repo_tree structure,
but nowadays they are convenience wrappers to handle a few git-vs-svn
mismatches:
1. Git does not track empty directories but Subversion does. When
looking up a path in git that Subversion thinks exists and finding
nothing, we can safely assume that the path represents a
directory. This is needed when a later Subversion revision
modifies that directory.
2. Subversion allows deleting a file by copying. In Git fast-import
we have to handle that more explicitly as a deletion.
These are details of the tool's interaction with git fast-import.
Move them to fast_export.c, where other such details are handled.
This way the function names do not start with a repo_ prefix that
would clash with the repository object introduced in
v2.14.0-rc0~38^2~16 (repository: introduce the repository object,
2017-06-22) or an svn_ prefix that would clash with libsvn (in case
someone wants to link this code with libsvn some day).
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since v1.7.10-rc0~118^2~4^2~4^2~3 (vcs-svn: pass paths through to
fast-import, 2010-12-13) this is an alias for fast_export_delete.
Remove the unnecessary layer of indirection.
No functional change intended.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>