In c8aed5e8da (repository: stop setting SHA1 as the default object
hash), we got rid of the default hash algorithm for the_repository.
Due to this change, it is now the responsibility of the callers to set
their own default when this is not present.
As stated in the docs, show-index should use SHA1 as the default hash
algorithm when run outside a repository. Make sure this promise is
met by falling back to SHA1 when the_hash_algo is not present (i.e.
when the command is run outside a repository). Also add a test that
verifies this behavior.
Signed-off-by: Abhijeet Sonar <abhijeet.nkt@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When called with '--left-right' and '--use-bitmap-index', 'rev-list'
will produce output without any left/right markers, which has been
corrected.
* jk/left-right-bitmap:
rev-list: skip bitmap traversal for --left-right
When trying to describe a commit, we'll traverse from the commit,
collecting candidate tags that point to its ancestors. But once we've
seen all of the tags in the repo, there's no point in traversing
further. There's nothing left to find!
For a default "git describe", this isn't usually a big problem. In a
large repo you'll probably have multiple tags, so we'll eventually find
10 candidates (the default for max_candidates) and stop there. And in a
small repo, it's quick to traverse to the root.
But you can imagine a large repo with few tags. Or, as we saw in a real
world case, explicitly limiting the set of matches like this (on
linux.git):
git describe --match=v6.12-rc4 HEAD
which goes all the way to the root before realizing that no, there are
no other tags under consideration besides the one we fed via --match.
If we add in "--candidates=1" there, it's much faster (at least as of
the previous commit).
But we should be able to speed this up without the user asking for it.
After expanding all matching tags, we know the total number of names. We
could just stop the traversal there, but as hinted at above we already
have a mechanism for doing that: the max_candidate limit. So we can just
reduce that limit to match the number of possible candidates.
Our p6100 test shows this off:
Test HEAD^ HEAD
---------------------------------------------------------------------------------------
6100.2: describe HEAD 0.71(0.65+0.06) 0.72(0.68+0.04) +1.4%
6100.3: describe HEAD with one max candidate 0.01(0.00+0.00) 0.01(0.00+0.00) +0.0%
6100.4: describe HEAD with one tag 0.72(0.66+0.05) 0.01(0.00+0.00) -98.6%
Now we are fast automatically, just as if --candidates=1 were supplied
by the user.
Reported-by: Josh Poimboeuf <jpoimboe@kernel.org>
Helped-by: Rasmus Villemoes <ravi@prevas.dk>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
By default, describe considers only 10 candidate matches, and stops
traversing when we have enough. This makes things much faster in a large
repository, where collecting all candidates requires walking all the way
down to the root (or at least to the oldest tag). This goes all the way
back to 8713ab3079 (Improve git-describe performance by reducing
revision listing., 2007-01-13).
However, we don't stop immediately when we have enough candidates. We
keep traversing and only bail when we find one more candidate that we're
ignoring. Usually this is not too expensive, if the tags are sprinkled
evenly throughout history. But if you are unlucky, you might hit the max
candidate quickly, and then have a huge swath of history before finding
the next one.
Our p6100 test has exactly this unlucky case: with a max of "1", we find
a recent tag quickly and then have to go all the way to the root to find
the old tag that will be discarded.
A more interesting real-world case is:
git describe --candidates=1 --match=v6.12-rc4 HEAD
in the linux.git repo. There we restrict the set of tags to a single
one, so there is no older candidate to find at all! But despite
--candidates=1, we keep traversing to the root only to find nothing.
So why do we keep traversing after hitting thet max? There are two
reasons I can see:
1. In theory the extra information that there was another candidate
could be useful, and we record it in the gave_up_on variable. But
we only show this information with --debug.
2. After finding the candidate, there's more processing we do in our
loop. The most important of this is propagating the "within" flags
to our parent commits, and putting them in the commit_list we'll
use for finish_depth_computation().
That function continues the traversal until we've counted all
commits reachable from the starting point but not reachable from
our best candidate tag (so essentially counting "$tag..$start", but
avoiding re-walking over the bits we've seen). If we break
immediately without putting those commits into the list, our depth
computation will be wrong (in the worst case we'll count all the
way down to the root, not realizing those commits are included in
our tag).
But we don't need to find a new candidate for (2). As soon as we finish
the loop iteration where we hit max_candidates, we can then quit on the
next iteration. This should produce the same output as the original code
(which could, after all, find a candidate on the very next commit
anyway) but ends the traversal with less pointless digging.
We still have to set "gave_up_on"; we've popped it off the list and it
has to go back. An alternative would be to re-order the loop so that it
never gets popped, but it's perhaps still useful to show in the --debug
output, so we need to know it anyway. We do have to adjust the --debug
output since it's now just a commit where we stopped traversing, and not
the max+1th candidate.
p6100 shows the speedup using linux.git:
Test HEAD^ HEAD
---------------------------------------------------------------------------------------
6100.2: describe HEAD 0.70(0.63+0.06) 0.71(0.66+0.04) +1.4%
6100.3: describe HEAD with one max candidate 0.70(0.64+0.05) 0.01(0.00+0.00) -98.6%
6100.4: describe HEAD with one tag 0.70(0.67+0.03) 0.70(0.63+0.06) +0.0%
Reported-by: Josh Poimboeuf <jpoimboe@kernel.org>
Helped-by: Rasmus Villemoes <ravi@prevas.dk>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The documentation for the --shallow-exclude option to clone/fetch/etc.
claims that the option takes a revision, but it does not. As per
upload-pack.c's process_deepen_not(), it passes the option to
expand_ref() and dies if it does not find exactly one ref matching the
name passed. Further, this has always been the case ever since these
options were introduced by the commits merged in a460ea4a3c (Merge
branch 'nd/shallow-deepen', 2016-10-10). Fix the documentation to
match the implementation.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The `obuf` member of `struct merge_options` is used to buffer output in
some cases. In order to not discard its allocated memory we only release
its contents in `merge_finalize()` when we're not currently recursing
into a subtree.
This results in some situations where we seemingly do not release the
buffer reliably. We thus have calls to `strbuf_release()` for this
buffer scattered across the codebase. But we're missing one callsite in
git-merge(1), which causes a memory leak.
We should ideally refactor this interface so that callers don't have to
know about any such internals. But for now, paper over the issue by
adding one more `strbuf_release()` call.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We do not free the key ID when signing a tag fails. Do so by using
the common exit path.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The cleanup string set by the config is leaking when it is being
overridden by an option. Fix this by tracking these via two separate
variables such that we can free the old value.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
While we free the worktree change data, we never free its contents. Fix
this.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When executing with `--max-count=0` we'll return early from git-grep(1)
without performing any cleanup, which causes memory leaks. Plug these.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The list of server options populated via `OPT_STRING_LIST()` is never
cleared, causing a memory leak. Plug it.
This leak is exposed by t5702, but plugging it alone does not make the
whole test suite pass.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The dumb-http code regressed when the result of re-indexing a pack
yielded an *.idx file that differs in content from the *.idx file it
downloaded from the remote. This has been corrected by no longer
relying on the *.idx file we got from the remote.
* jk/dumb-http-finalize:
packfile: use oidread() instead of hashcpy() to fill object_id
packfile: use object_id in find_pack_entry_one()
packfile: convert find_sha1_pack() to use object_id
http-walker: use object_id instead of bare hash
packfile: warn people away from parse_packed_git()
packfile: drop sha1_pack_index_name()
packfile: drop sha1_pack_name()
packfile: drop has_pack_index()
dumb-http: store downloaded pack idx as tempfile
t5550: count fetches in "previously-fetched .idx" test
midx: avoid duplicate packed_git entries
Teach 'git notes add' and 'git notes append' a new '-e' flag,
instructing them to open the note in $GIT_EDITOR before saving.
* sa/notes-edit:
notes: teach the -e option to edit messages in editor
Treat ECONNABORTED the same as ECONNRESET in 'git credential-cache' to
work around a possible Cygwin regression. This resolves a race condition
caused by changes in Cygwin's handling of socket closures, allowing the
client to exit cleanly when encountering ECONNABORTED.
* rj/cygwin-exit:
credential-cache: treat ECONNABORTED like ECONNRESET
Various platform compatibility fixes split out of the larger effort
to use Meson as the primary build tool.
* ps/platform-compat-fixes:
t6006: fix prereq handling with `test_format ()`
http: fix build error on FreeBSD
builtin/credential-cache: fix missing parameter for stub function
t7300: work around platform-specific behaviour with long paths on MinGW
t5500, t5601: skip tests which exercise paths with '[::1]' on Cygwin
t3404: work around platform-specific behaviour on macOS 10.15
t1401: make invocation of tar(1) work with Win32-provided one
t/lib-gpg: fix setup of GNUPGHOME in MinGW
t/lib-gitweb: test against the build version of gitweb
t/test-lib: wire up NO_ICONV prerequisite
t/test-lib: fix quoting of TEST_RESULTS_SAN_FILE
Running:
git rev-list --left-right --use-bitmap-index one...two
will produce output without any left-right markers, since the bitmap
traversal returns only a single set of reachable commits. Instead we
should refuse to use bitmaps here and produce the correct output using a
traditional traversal.
This is probably not the only remaining option that misbehaves with
bitmaps, but it's particularly egregious in that it feels like it
_could_ work. Doing two separate traversals for the left/right sides and
then taking the symmetric set differences should yield the correct
answer, but our traversal code doesn't know how to do that.
It's not clear if naively doing two separate traversals would always be
a performance win. A traditional traversal only needs to walk down to
the merge base, but bitmaps always fill out the full reachability set.
So depending on your bitmap coverage, we could end up walking old bits
of history twice to fill out the same uninteresting bits on both sides.
We'd also of course end up with a very large --boundary set, if the user
asked for that.
So this might or might not be something worth implementing later. But
for now, let's make sure we don't produce the wrong answer if somebody
tries it.
The test covers this, but also the same thing with "--count" (which is
what I originally tried in a real-world case). Ironically the
try_bitmap_count() code already realizes that "--left-right" won't work
there. But that just causes us to fall back to the regular bitmap
traversal code, which itself doesn't handle counting (we produce a list
of objects rather than a count).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
The main function we use to search a pack index for an object is
find_pack_entry_one(). That function still takes a bare pointer to the
hash, despite the fact that its underlying bsearch_pack() function needs
an object_id struct. And so we end up making an extra copy of the hash
into the struct just to do a lookup.
As it turns out, all callers but one already have such an object_id. So
we can just take a pointer to that struct and use it directly. This
avoids the extra copy and provides a more type-safe interface.
The one exception is get_delta_base() in packfile.c, when we are chasing
a REF_DELTA from inside the pack (and thus we have a pointer directly to
the mmap'd pack memory, not a struct). We can just bump the hashcpy()
from inside find_pack_entry_one() to this one caller that needs it.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
The find_sha1_pack() function has a few problems:
- it's badly named, since it works with any object hash
- it takes the hash as a bare pointer rather than an object_id struct
We can fix both of these easily, as all callers actually have a real
object_id anyway.
I also found the existence of this function somewhat confusing, as it is
about looking in an arbitrary set of linked packed_git structs. It's
good for things like dumb-http which are looking in downloaded remote
packs, and not our local packs. But despite the name, it is not a good
way to find the pack which contains a local object (it skips the use of
the midx, the pack mru list, and so on).
So let's also add an explanatory comment above the declaration that may
point people in the right direction.
I suspect the calls in fast-import.c, which use the packed_git list from
the repository struct, could actually just be using find_pack_entry().
But since we'd need to keep it anyway for dumb-http, I didn't dig
further there. If we eventually drop dumb-http support, then it might be
worth examining them to see if we can get rid of the function entirely.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Like sha1_pack_name() that we dropped in the previous commit, this
function uses an error-prone static strbuf and the somewhat misleading
name "sha1".
The only caller left is in pack-redundant.c. While this command is
marked for potential removal in our BreakingChanges document, we still
have it for now. But it's simple enough to convert it to use its own
strbuf with the underlying odb_pack_name() function, letting us drop the
otherwise obsolete function.
Note that odb_pack_name() does its own strbuf_reset(), so it's safe to
use directly within a loop like this.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Teaches 'shortlog' to explicitly use SHA-1 when operating outside of
a repository.
* wm/shortlog-hash:
builtin/shortlog: explicitly set hash algo when there is no repo
Commands that can also work outside Git have learned to take the
repository instance "repo" when we know we are in a repository, and
NULL when we are not, in a parameter. The uses of the_repository
variable in a few of them have been removed using the new calling
convention.
* jc/a-commands-without-the-repo:
archive: remove the_repository global variable
annotate: remove usage of the_repository global
git: pass in repo to builtin based on setup_git_directory_gently
A "git fetch" from the superproject going down to a submodule used
a wrong remote when the default remote names are set differently
between them.
* db/submodule-fetch-with-remote-name-fix:
submodule: correct remote name with fetch
An extra worktree attached to a repository points at each other to
allow finding the repository from the worktree and vice versa
possible. Turn this linkage to relative paths.
* cw/worktree-relative:
worktree: add test for path handling in linked worktrees
worktree: link worktrees with relative paths
worktree: refactor infer_backlink() to use *strbuf
worktree: repair copied repository and linked worktrees
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>
Notes can be added to a commit using:
- "-m" to provide a message on the command line.
- -C to copy a note from a blob object.
- -F to read the note from a file.
When these options are used, Git does not open an editor,
it simply takes the content provided via these options and
attaches it to the commit as a note.
Improve flexibility to fine-tune the note before finalizing it
by allowing the messages to be prefilled in the editor and edited
after the messages have been provided through -[mF].
Signed-off-by: Abraham Samuel Adekunle <abrahamadekunle50@gmail.com>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
On Cygwin, t0301 fails because "git credential-cache exit" returns a
non-zero exit code. What's supposed to happen here is:
1. The client (the "credential-cache" invocation above) connects to a
previously-spawned credential-cache--daemon.
2. The client sends an "exit" command to the daemon.
3. The daemon unlinks the socket and then exits, closing the
descriptor back to the client.
4. The client sees EOF on the descriptor and exits successfully.
That works on most platforms, and even _used_ to work on Cygwin. But
that changed in Cygwin's ef95c03522 (Cygwin: select: Fix FD_CLOSE
handling, 2021-04-06). After that commit, the client sees a read error
with errno set to ECONNABORTED, and it reports the error and dies.
It's not entirely clear if this is a Cygwin bug. It seems that calling
fclose() on the filehandles pointing to the sockets is sufficient to
avoid this error return, even though exiting should in general look the
same from the client's perspective.
However, we can't just call fclose() here. It's important in step 3
above to unlink the socket before closing the descriptor to avoid the
race mentioned by 7d5e9c9849 (credential-cache--daemon: clarify "exit"
action semantics, 2016-03-18). The client will exit as soon as it sees
the descriptor close, and the daemon may or may not have actually
unlinked the socket by then. That makes test code like this:
git credential exit &&
test_path_is_missing .git-credential-cache
racy.
So we probably _could_ fix this by calling:
delete_tempfile(&socket_file);
fclose(in);
fclose(out);
before we exit(). Or by replacing the exit() with a return up the stack,
in which case the fclose() happens as we unwind. But in that case we'd
still need to call delete_tempfile() here to avoid the race.
But simpler still is that we can notice that we already special-case
ECONNRESET on the client side, courtesy of 1f180e5eb9 (credential-cache:
interpret an ECONNRESET as an EOF, 2017-07-27). We can just do the same
thing here (I suspect that prior to the Cygwin commit that introduced
this problem, we were really just seeing ECONNRESET instead of
ECONNABORTED, so the "new" problem is just the switch of the errno
values).
There's loads more debugging in this thread:
https://lore.kernel.org/git/9dc3e85f-a532-6cff-de11-1dfb2e4bc6b6@ramsayjones.plus.com/
but I've tried to summarize the useful bits in this commit message.
[jk: commit message]
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Ramsay Jones <ramsay@ramsayjones.plus.com>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
"git maintenance start" crashed due to an uninitialized variable
reference, which has been corrected.
* ps/maintenance-start-crash-fix:
builtin/gc: fix crash when running `git maintenance start`
Whilst git-shortlog(1) does not explicitly need any repository
information when run without reference to one, it still parses some of
its arguments with parse_revision_opt() which assumes that the hash
algorithm is set. However, in c8aed5e8da (repository: stop setting SHA1
as the default object hash, 2024-05-07) we stopped setting up a default
hash algorithm and instead require commands to set it up explicitly.
This was done for most other commands like in ab274909d4 (builtin/diff:
explicitly set hash algo when there is no repo, 2024-05-07) but was
missed for builtin/shortlog, making git-shortlog(1) segfault outside of
a repository when given arguments like --author that trigger a call to
parse_revision_opt().
Fix this for now by explicitly setting the hash algorithm to SHA1. Also
add a regression test for the segfault.
Thanks-to: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Wolfgang Müller <wolf@oriole.systems>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
When not compiling the credential cache we may use a stub function for
`cmd_credential_cache()`. With commit 9b1cb5070f (builtin: add a
repository parameter for builtin functions, 2024-09-13), we have added a
new parameter to all of those top-level `cmd_*()` functions, and did
indeed adapt the non-stubbed-out `cmd_credential_cache()`. But we didn't
adapt the stubbed-out variant, so the code does not compile.
Fix this by adding the missing parameter.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
On macOS, fsmonitor can fall into a race condition that results in
a client waiting forever to be notified for an event that have
already happened. This problem has been corrected.
* jk/fsmonitor-event-listener-race-fix:
fsmonitor: initialize fs event listener before accepting clients
simple-ipc: split async server initialization and running
A new configuration variable remote.<name>.serverOption makes the
transport layer act as if the --serverOption=<value> option is
given from the command line.
* xx/remote-server-option-config:
ls-remote: leakfix for not clearing server_options
fetch: respect --server-option when fetching multiple remotes
transport.c:🤝 make use of server options from remote
remote: introduce remote.<name>.serverOption configuration
transport: introduce parse_transport_option() method
Deprecate the "trailer_info" struct name and replace it with
"trailer_block". This is more readable, for two reasons:
1. "trailer_info" on the surface sounds like it's about a single
trailer when in reality it is a collection of one or more trailers,
and
2. the "*_block" suffix is more informative than "*_info", because it
describes a block (or region) of contiguous text which has trailers
in it, which has been parsed into the trailer_block structure.
Rename the
size_t trailer_block_start, trailer_block_end;
members of trailer_info to just "start" and "end". Rename the "info"
pointer to "trailer_block" because it is more descriptive. Update
comments accordingly.
Signed-off-by: Linus Arver <linus@ucla.edu>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
As part of the effort to get rid of global state due to the global
the_repository variable, replace the_repository with the repository
argument that gets passed down through the builtin function.
The repo might be NULL, but we should be safe in write_archive() because
it detects if we are outside of a repository and calls
setup_git_directory() which will error.
Signed-off-by: John Cai <johncai86@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As part of the effort to get rid of global state due to the_repository
variable, remove the the_repository with the repository argument that
gets passed down through the builtin function.
Signed-off-by: John Cai <johncai86@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The current code in run_builtin() passes in a repository to the builtin
based on whether cmd_struct's option flag has RUN_SETUP.
This is incorrect, however, since some builtins that only have
RUN_SETUP_GENTLY can potentially take a repository.
setup_git_directory_gently() tells us whether or not a command is being
run inside of a repository.
Use the output of setup_git_directory_gently() to help determine whether
or not there is a repository to pass to the builtin. If not, then we
just pass NULL.
As part of this patch, we need to modify add to check for a NULL repo
before calling repo_git_config(), since add -h can be run outside of a
repository.
Signed-off-by: John Cai <johncai86@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We can only check out commits or branches, not refs in general. And the
problem here is if another worktree is using the branch that we want to
check out.
Let’s be more direct and just talk about branches instead of refs.
Also replace “be held” with “in use”. Further, “in use” is not
restricted to a branch being checked out (e.g. the branch could be busy
on a rebase), hence generalize to “or otherwise in use” in the option
description.
Signed-off-by: Kristoffer Haugsbakk <code@khaugsbakk.name>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It was reported on the mailing list that running `git maintenance start`
immediately segfaults starting with b6c3f8e12c (builtin/maintenance: fix
leak in `get_schedule_cmd()`, 2024-09-26). And indeed, this segfault is
trivial to reproduce up to a point where one is scratching their head
why we didn't catch this regression in our test suite.
The root cause of this error is `get_schedule_cmd()`, which does not
populate the `out` parameter in all cases anymore starting with the
mentioned commit. Callers do assume it to always be populated though and
will e.g. call `strvec_split()` on the returned value, which will of
course segfault when the variable is uninitialized.
So why didn't we catch this trivial regression? The reason is that our
tests always set up the "GIT_TEST_MAINT_SCHEDULER" environment variable
via "t/test-lib.sh", which allows us to override the scheduler command
with a custom one so that we don't accidentally modify the developer's
system. But the faulty code where we don't set the `out` parameter will
only get hit in case that environment variable is _not_ set, which is
never the case when executing our tests.
Fix the regression by again unconditionally allocating the value in the
`out` parameter, if provided. Add a test that unsets the environment
variable to catch future regressions in this area.
Reported-by: Shubham Kanodia <shubham.kanodia10@gmail.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The code fetches the submodules remote based on the superproject remote name
instead of the submodule remote name[1].
Instead of grabbing the default remote of the superproject repository, ask
the default remote of the submodule we are going to run 'git fetch' in.
1. https://lore.kernel.org/git/ZJR5SPDj4Wt_gmRO@pweza/
Signed-off-by: Daniel Black <daniel@mariadb.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The synopsis for `git config unset` mentions two positional arguments:
`<name>` and `<value>`. While the first argument is correct, the second
is not. Users are expected to provide the value via `--value=<value>`.
Remove the positional argument. The `--value=<value>` option is already
documented correctly, so this is all we need to do to fix the
documentation.
Signed-off-by: Josh Heinrichs <joshiheinrichs@gmail.com>
Reviewed-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There's a racy hang in fsmonitor on macOS that we sometimes see in CI.
When we serve a client, what's supposed to happen is:
1. The client thread calls with_lock__wait_for_cookie() in which we
create a cookie file and then wait for a pthread_cond event
2. The filesystem event listener sees the cookie file creation, does
some internal book-keeping, and then triggers the pthread_cond.
But there's a problem: we start the listener that accepts client threads
before we start the fs event thread. So it's possible for us to accept a
client which creates the cookie file and starts waiting before the fs
event thread is initialized, and we miss those filesystem events
entirely. That leaves the client thread hanging forever.
In CI, the symptom is that t9210 (which is testing scalar, which always
enables fsmonitor under the hood) may hang forever in "scalar clone". It
is waiting on "git fetch" which is waiting on the fsmonitor daemon.
The race happens more frequently under load, but you can trigger it
predictably with a sleep like this, which delays the start of the fs
event thread:
--- a/compat/fsmonitor/fsm-listen-darwin.c
+++ b/compat/fsmonitor/fsm-listen-darwin.c
@@ -510,6 +510,7 @@ void fsm_listen__loop(struct fsmonitor_daemon_state *state)
FSEventStreamSetDispatchQueue(data->stream, data->dq);
data->stream_scheduled = 1;
+ sleep(1);
if (!FSEventStreamStart(data->stream)) {
error(_("Failed to start the FSEventStream"));
goto force_error_stop_without_loop;
One solution might be to reverse the order of initialization: start the
fs event thread before we start the thread listening for clients. But
the fsmonitor code explicitly does it in the opposite direction. The fs
event thread wants to refer to the ipc_server_data struct, so we need it
to be initialized first.
A further complication is that we need a signal from the fs event thread
that it is actually ready and listening. And those details happen within
backend-specific fsmonitor code, whereas the initialization is in the
shared code.
So instead, let's use the ipc_server init/start split added in the
previous commit. The generic fsmonitor code will init the ipc_server but
_not_ start it, leaving that to the backend specific code, which now
needs to call ipc_server_start_async() at the right time.
For macOS, that is right after we start the FSEventStream that you can
see in the diff above.
It's not clear to me if Windows suffers from the same problem (and we
simply don't trigger it in CI), or if it is immune. Regardless, the
obvious place to start accepting clients there is right after we've
established the ReadDirectoryChanges watch.
This makes the hangs go away in our macOS CI environment, even when
compiled with the sleep() above.
Helped-by: Koji Nakamaru <koji.nakamaru@gree.net>
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Koji Nakamaru <koji.nakamaru@gree.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
To start an async ipc server, you call ipc_server_run_async(). That
initializes the ipc_server_data object, and starts all of the threads
running, which may immediately start serving clients.
This can create some awkward timing problems, though. In the fsmonitor
daemon (the sole user of the simple-ipc system), we want to create the
ipc server early in the process, which means we may start serving
clients before the rest of the daemon is fully initialized.
To solve this, let's break run_async() into two parts: an initialization
which allocates all data and spawns the threads (without letting them
run), and a start function which actually lets them begin work. Since we
have two simple-ipc implementations, we have to handle this twice:
- in ipc-unix-socket.c, we have a central listener thread which hands
connections off to worker threads using a work_available mutex. We
can hold that mutex after init, and release it when we're ready to
start.
We do need an extra "started" flag so that we know whether the main
thread is holding the mutex or not (e.g., if we prematurely stop the
server, we want to make sure all of the worker threads are released
to hear about the shutdown).
- in ipc-win32.c, we don't have a central mutex. So we'll introduce a
new startup_barrier mutex, which we'll similarly hold until we're
ready to let the threads proceed.
We again need a "started" flag here to make sure that we release the
barrier mutex when shutting down, so that the sub-threads can
proceed to the finish.
I've renamed the run_async() function to init_async() to make sure we
catch all callers, since they'll now need to call the matching
start_async().
We could leave run_async() as a wrapper that does both, but there's not
much point. There are only two callers, one of which is fsmonitor, which
will want to actually do work between the two calls. And the other is
just a test-tool wrapper.
For now I've added the start_async() calls in fsmonitor where they would
otherwise have happened, so there should be no behavior change with this
patch.
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Koji Nakamaru <koji.nakamaru@gree.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Git currently stores absolute paths to both the main repository and
linked worktrees. However, this causes problems when moving repositories
or working in containerized environments where absolute paths differ
between systems. The worktree links break, and users are required to
manually execute `worktree repair` to repair them, leading to workflow
disruptions. Additionally, mapping repositories inside of containerized
environments renders the repository unusable inside the containers, and
this is not repairable as repairing the worktrees inside the containers
will result in them being broken outside the containers.
To address this, this patch makes Git always write relative paths when
linking worktrees. Relative paths increase the resilience of the
worktree links across various systems and environments, particularly
when the worktrees are self-contained inside the main repository (such
as when using a bare repository with worktrees). This improves
portability, workflow efficiency, and reduces overall breakages.
Although Git now writes relative paths, existing repositories with
absolute paths are still supported. There are no breaking changes
to workflows based on absolute paths, ensuring backward compatibility.
At a low level, the changes involve modifying functions in `worktree.c`
and `builtin/worktree.c` to use `relative_path()` when writing the
worktree’s `.git` file and the main repository’s `gitdir` reference.
Instead of hardcoding absolute paths, Git now computes the relative path
between the worktree and the repository, ensuring that these links are
portable. Locations where these respective file are read have also been
updated to properly handle both absolute and relative paths. Generally,
relative paths are always resolved into absolute paths before any
operations or comparisons are performed.
Additionally, `repair_worktrees_after_gitdir_move()` has been introduced
to address the case where both the `<worktree>/.git` and
`<repo>/worktrees/<id>/gitdir` links are broken after the gitdir is
moved (such as during a re-initialization). This function repairs both
sides of the worktree link using the old gitdir path to reestablish the
correct paths after a move.
The `worktree.path` struct member has also been updated to always store
the absolute path of a worktree. This ensures that worktree consumers
never have to worry about trying to resolve the absolute path themselves.
Signed-off-by: Caleb White <cdwhite3@pm.me>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Ensure `server_options` is properly cleared using `string_list_clear()`
in `builtin/ls-remote.c:cmd_ls_remote`.
Although we cannot yet enable `TEST_PASSES_SANITIZE_LEAK=true` for
`t/t5702-protocol-v2.sh` due to other existing leaks, this fix ensures
that "git-ls-remote" related server options tests pass the sanitize leak
check:
...
ok 12 - server-options are sent when using ls-remote
ok 13 - server-options from configuration are used by ls-remote
...
Signed-off-by: Xing Xin <xingxin.xx@bytedance.com>
Reviewed-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Fix an issue where server options specified via the command line
(`--server-option` or `-o`) were not sent when fetching from multiple
remotes using Git protocol v2.
To reproduce the issue with a repository containing multiple remotes:
GIT_TRACE_PACKET=1 git -c protocol.version=2 fetch --server-option=demo --all
Observe that no server options are sent to any remote.
The root cause was identified in `builtin/fetch.c:fetch_multiple`, which
is invoked when fetching from more than one remote. This function forks
a `git-fetch` subprocess for each remote but did not include the
specified server options in the subprocess arguments.
This commit ensures that command-line specified server options are
properly passed to each subprocess. Relevant tests have been added.
Signed-off-by: Xing Xin <xingxin.xx@bytedance.com>
Reviewed-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add the `parse_transport_option()` method to parse the `push.pushOption`
configuration. This method will also be used in the next commit to
handle the new `remote.<name>.serverOption` configuration for setting
server options in Git protocol v2.
Signed-off-by: Xing Xin <xingxin.xx@bytedance.com>
Reviewed-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
