"git push --atomic" that goes over the transport-helper (namely,
the smart http transport) failed to prevent refs to be pushed when
it can locally tell that one of the ref update will fail without
having to consult the other end, which has been corrected.
* es/local-atomic-push-failure-with-http:
transport-helper: avoid var decl in for () loop control
transport-helper: enforce atomic in push_refs_with_push
Teach transport-helper how to notice if skipping a ref during push would
violate atomicity on the client side. We notice that a ref would be
rejected, and choose not to send it, but don't notice that if the client
has asked for --atomic we are violating atomicity if all the other
pushes we are sending would succeed. Asking the server end to uphold
atomicity wouldn't work here as the server doesn't have any idea that we
tried to update a ref that's broken.
The added test-case is a succinct way to reproduce this issue that fails
today. The same steps work fine when we aren't using a transport-helper
to get to the upstream, i.e. when we've added a local repository as a
remote:
git remote add ~/upstream upstream
Signed-off-by: Emily Shaffer <emilyshaffer@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The ownership rule for the file descriptor to fast-import remote
backend was mixed up, leading to unrelated file descriptor getting
closed, which has been fixed.
* mh/import-transport-fd-fix:
Use xmmap_gently instead of xmmap in use_pack
dup() the input fd for fast-import used for remote helpers
When a remote helper exposes the "import" capability, stdout of the
helper is sent to stdin of a new fast-import process. This is done by
setting the corresponding child_process's in field to the value of the
out field of the helper child_process.
The child_process API is defined to close the file descriptors it's
given when calling start_command. This means when start_command is
called for the fast-import process, its input fd (the output fd of the
helper), is closed.
But when the transport helper is later destroyed, in disconnect_helper,
its input and output are closed, which means close() is called with
an invalid fd (since it was already closed as per above). Or worse, with
a valid fd owned by something else (since fd numbers can be reused).
Signed-off-by: Mike Hommey <mh@glandium.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add trace2 child classification for transport processes.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Update the protocol message specification to allow only the limited
use of scaled quantities. This is ensure potential compatibility
issues will not go out of hand.
* js/filter-options-should-use-plain-int:
filter-options: expand scaled numbers
tree:<depth>: skip some trees even when collecting omits
list-objects-filter: teach tree:# how to handle >0
Portability updates for the HPE NonStop platform.
* rb/hpe:
compat/regex/regcomp.c: define intptr_t and uintptr_t on NonStop
git-compat-util.h: add FLOSS headers for HPE NonStop
config.mak.uname: support for modern HPE NonStop config.
transport-helper: drop read/write errno checks
transport-helper: use xread instead of read
When communicating with a remote server or a subprocess, use
expanded numbers rather than numbers with scaling suffix in the
object filter spec (e.g. "limit:blob=1k" becomes
"limit:blob=1024").
Update the protocol docs to note that clients should always perform this
expansion, to allow for more compatibility between server
implementations.
Signed-off-by: Josh Steadmon <steadmon@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since we use xread() and xwrite() here, EINTR, EAGAIN, and
EWOULDBLOCK retries are already handled for us, and we will
never see these errno values ourselves. We can drop these
conditions entirely, making the code easier to follow.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This fix was needed on HPE NonStop NSE and NSX where SSIZE_MAX is less than
BUFFERSIZE resulting in EINVAL. The call to read in transport-helper.c
was the only place outside of wrapper.c where it is used instead of xread.
Signed-off-by: Randall S. Becker <rsbecker@nexbridge.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The built-in bundle transport and the transport helper interface do not
work when transport_fetch_refs() is called immediately after transport
creation. This will be needed in a subsequent patch, so fix this.
Evidence: fetch_refs_from_bundle() relies on data->header being
initialized in get_refs_from_bundle(), and fetch() in transport-helper.c
relies on either data->fetch or data->import being set by get_helper(),
but neither transport_helper_init() nor fetch() calls get_helper().
Up until the introduction of the partial clone feature, this has not
been a problem, because transport_fetch_refs() is always called after
transport_get_remote_refs(). With the introduction of the partial clone
feature, which involves calling transport_fetch_refs() (to fetch objects
by their OIDs) without transport_get_remote_refs(), this is still not a
problem, but only coincidentally - we do not support partially cloning a
bundle, and as for cloning using a transport-helper-using protocol, it
so happens that before transport_fetch_refs() is called, fetch_refs() in
fetch-object.c calls transport_set_option(), which means that the
aforementioned get_helper() is invoked through set_helper_option() in
transport-helper.c.
This could be fixed by fixing the transports themselves, but it doesn't
seem like a good idea to me to open up previously untested code paths;
also, there may be transport helpers in the wild that assume that "list"
is always called before "fetch". Instead, fix this by having
transport_fetch_refs() call transport_get_remote_refs() to ensure that
the latter is always called at least once, unless the transport
explicitly states that it supports fetching without listing refs.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git fetch" sometimes failed to update the remote-tracking refs,
which has been corrected.
* jt/connectivity-check-after-unshallow:
fetch-pack: unify ref in and out param
Many more strings are prepared for l10n.
* nd/i18n: (23 commits)
transport-helper.c: mark more strings for translation
transport.c: mark more strings for translation
sha1-file.c: mark more strings for translation
sequencer.c: mark more strings for translation
replace-object.c: mark more strings for translation
refspec.c: mark more strings for translation
refs.c: mark more strings for translation
pkt-line.c: mark more strings for translation
object.c: mark more strings for translation
exec-cmd.c: mark more strings for translation
environment.c: mark more strings for translation
dir.c: mark more strings for translation
convert.c: mark more strings for translation
connect.c: mark more strings for translation
config.c: mark more strings for translation
commit-graph.c: mark more strings for translation
builtin/replace.c: mark more strings for translation
builtin/pack-objects.c: mark more strings for translation
builtin/grep.c: mark strings for translation
builtin/config.c: mark more strings for translation
...
"git fetch" learned a new option "--negotiation-tip" to limit the
set of commits it tells the other end as "have", to reduce wasted
bandwidth and cycles, which would be helpful when the receiving
repository has a lot of refs that have little to do with the
history at the remote it is fetching from.
* jt/fetch-nego-tip:
fetch-pack: support negotiation tip whitelist
When a user fetches:
- at least one up-to-date ref and at least one non-up-to-date ref,
- using HTTP with protocol v0 (or something else that uses the fetch
command of a remote helper)
some refs might not be updated after the fetch.
This bug was introduced in commit 989b8c4452 ("fetch-pack: put shallow
info in output parameter", 2018-06-28) which allowed transports to
report the refs that they have fetched in a new out-parameter
"fetched_refs". If they do so, transport_fetch_refs() makes this
information available to its caller.
Users of "fetched_refs" rely on the following 3 properties:
(1) it is the complete list of refs that was passed to
transport_fetch_refs(),
(2) it has shallow information (REF_STATUS_REJECT_SHALLOW set if
relevant), and
(3) it has updated OIDs if ref-in-want was used (introduced after
989b8c4452).
In an effort to satisfy (1), whenever transport_fetch_refs()
filters the refs sent to the transport, it re-adds the filtered refs to
whatever the transport supplies before returning it to the user.
However, the implementation in 989b8c4452 unconditionally re-adds the
filtered refs without checking if the transport refrained from reporting
anything in "fetched_refs" (which it is allowed to do), resulting in an
incomplete list, no longer satisfying (1).
An earlier effort to resolve this [1] solved the issue by readding the
filtered refs only if the transport did not refrain from reporting in
"fetched_refs", but after further discussion, it seems that the better
solution is to revert the API change that introduced "fetched_refs".
This API change was first suggested as part of a ref-in-want
implementation that allowed for ref patterns and, thus, there could be
drastic differences between the input refs and the refs actually fetched
[2]; we eventually decided to only allow exact ref names, but this API
change remained even though its necessity was decreased.
Therefore, revert this API change by reverting commit 989b8c4452, and
make receive_wanted_refs() update the OIDs in the sought array (like how
update_shallow() updates shallow information in the sought array)
instead. A test is also included to show that the user-visible bug
discussed at the beginning of this commit message no longer exists.
[1] https://public-inbox.org/git/20180801171806.GA122458@google.com/
[2] https://public-inbox.org/git/86a128c5fb710a41791e7183207c4d64889f9307.1485381677.git.jonathantanmy@google.com/
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Many messages will be marked for translation in the following
commits. This commit updates some of them to be more consistent and
reduce diff noise in those commits. Changes are
- keep the first letter of die(), error() and warning() in lowercase
- no full stop in die(), error() or warning() if it's single sentence
messages
- indentation
- some messages are turned to BUG(), or prefixed with "BUG:" and will
not be marked for i18n
- some messages are improved to give more information
- some messages are broken down by sentence to be i18n friendly
(on the same token, combine multiple warning() into one big string)
- the trailing \n is converted to printf_ln if possible, or deleted
if not redundant
- errno_errno() is used instead of explicit strerror()
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
During negotiation, fetch-pack eventually reports as "have" lines all
commits reachable from all refs. Allow the user to restrict the commits
sent in this way by providing a whitelist of tips; only the tips
themselves and their ancestors will be sent.
Both globs and single objects are supported.
This feature is only supported for protocols that support connect or
stateless-connect (such as HTTP with protocol v2).
This will speed up negotiation when the repository has multiple
relatively independent branches (for example, when a repository
interacts with multiple repositories, such as with linux-next [1] and
torvalds/linux [2]), and the user knows which local branch is likely to
have commits in common with the upstream branch they are fetching.
[1] https://kernel.googlesource.com/pub/scm/linux/kernel/git/next/linux-next/
[2] https://kernel.googlesource.com/pub/scm/linux/kernel/git/torvalds/linux/
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Expand the transport fetch method signature, by adding an output
parameter, to allow transports to return information about the refs they
have fetched. Then communicate shallow status information through this
mechanism instead of by modifying the input list of refs.
This does require clients to sometimes generate the ref map twice: once
from the list of refs provided by the remote (as is currently done) and
potentially once from the new list of refs that the fetch mechanism
provides.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert 'apply_refspecs()' to take a 'struct refspec' as a parameter instead
of a list of 'struct refspec_item'.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert the refspecs in transport-helper.c to be stored in a
'struct refspec'.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In preparation for introducing an abstraction around a collection of
refspecs (much like how a 'struct pathspec' is a collection of 'struct
pathspec_item's) rename the existing 'struct refspec' to 'struct
refspec_item'.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In preparation for performing a refactor on refspec related code, move
the refspec parsing logic into its own file.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The beginning of the next-gen transfer protocol.
* bw/protocol-v2: (35 commits)
remote-curl: don't request v2 when pushing
remote-curl: implement stateless-connect command
http: eliminate "# service" line when using protocol v2
http: don't always add Git-Protocol header
http: allow providing extra headers for http requests
remote-curl: store the protocol version the server responded with
remote-curl: create copy of the service name
pkt-line: add packet_buf_write_len function
transport-helper: introduce stateless-connect
transport-helper: refactor process_connect_service
transport-helper: remove name parameter
connect: don't request v2 when pushing
connect: refactor git_connect to only get the protocol version once
fetch-pack: support shallow requests
fetch-pack: perform a fetch using v2
upload-pack: introduce fetch server command
push: pass ref prefixes when pushing
fetch: pass ref prefixes when fetching
ls-remote: pass ref prefixes when requesting a remote's refs
transport: convert transport_get_remote_refs to take a list of ref prefixes
...
Introduce the transport-helper capability 'stateless-connect'. This
capability indicates that the transport-helper can be requested to run
the 'stateless-connect' command which should attempt to make a
stateless connection with a remote end. Once established, the
connection can be used by the git client to communicate with
the remote end natively in a stateless-rpc manner as supported by
protocol v2. This means that the client must send everything the server
needs in a single request as the client must not assume any
state-storing on the part of the server or transport.
If a stateless connection cannot be established then the remote-helper
will respond in the same manner as the 'connect' command indicating that
the client should fallback to using the dumb remote-helper commands.
A future patch will implement the 'stateless-connect' capability in our
http remote-helper (remote-curl) so that protocol v2 can be used using
the http transport.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A future patch will need to take advantage of the logic which runs and
processes the response of the connect command on a remote helper so
factor out this logic from 'process_connect_service()' and place it into
a helper function 'run_connect()'.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 266f1fdfa (transport-helper: be quiet on read errors from
helpers, 2013-06-21) removed a call to 'die()' which printed the name of
the remote helper passed in to the 'recvline_fh()' function using the
'name' parameter. Once the call to 'die()' was removed the parameter
was no longer necessary but wasn't removed. Clean up 'recvline_fh()'
parameter list by removing the 'name' parameter.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert the 'struct transport' virtual function 'get_refs_list()' to
optionally take an argv_array of ref prefixes. When communicating with
a server using protocol v2 these ref prefixes can be sent when
requesting a listing of their refs allowing the server to filter the
refs it sends based on the sent prefixes. This list will be ignored
when not using protocol v2.
Signed-off-by: Brandon Williams <bmwill@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The machinery to clone & fetch, which in turn involves packing and
unpacking objects, have been told how to omit certain objects using
the filtering mechanism introduced by the jh/object-filtering
topic, and also mark the resulting pack as a promisor pack to
tolerate missing objects, taking advantage of the mechanism
introduced by the jh/fsck-promisors topic.
* jh/partial-clone:
t5616: test bulk prefetch after partial fetch
fetch: inherit filter-spec from partial clone
t5616: end-to-end tests for partial clone
fetch-pack: restore save_commit_buffer after use
unpack-trees: batch fetching of missing blobs
clone: partial clone
partial-clone: define partial clone settings in config
fetch: support filters
fetch: refactor calculation of remote list
fetch-pack: test support excluding large blobs
fetch-pack: add --no-filter
fetch-pack, index-pack, transport: partial clone
upload-pack: add object filtering for partial clone
Move the definition of the transport-specific functions provided by
transports, whether declared in transport.c or transport-helper.c, into
an internal header. This means that transport-using code (as opposed to
transport-declaring code) can no longer access these functions (without
importing the internal header themselves), making it clear that they
should use the transport_*() functions instead, and also allowing the
interface between the transport mechanism and an individual transport to
independently evolve.
This is superficially a reversal of commit 824d5776c3 ("Refactor
struct transport_ops inlined into struct transport", 2007-09-19).
However, the scope of the involved variables was neither affected nor
discussed in that commit, and I think that the advantages in making
those functions more private outweigh the advantages described in that
commit's commit message. A minor additional point is that the code has
gotten more complicated since then, in that the function-pointer
variables are potentially mutated twice (once initially and once if
transport_take_over() is invoked), increasing the value of corralling
them into their own struct.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Transfer ownership of detached strbufs to string_lists of the
duplicating variety by calling string_list_append_nodup() instead of
string_list_append() to avoid duplicating and then leaking the buffer.
While at it make sure to release the string_list when done;
push_refs_with_export() already does that.
Reported-by: Jeff King <peff@peff.net>
Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert resolve_ref_unsafe to take a pointer to struct object_id by
converting one remaining caller to use struct object_id, removing the
temporary NULL pointer check in expand_ref, converting the declaration
and definition, and applying the following semantic patch:
@@
expression E1, E2, E3, E4;
@@
- resolve_ref_unsafe(E1, E2, E3.hash, E4)
+ resolve_ref_unsafe(E1, E2, &E3, E4)
@@
expression E1, E2, E3, E4;
@@
- resolve_ref_unsafe(E1, E2, E3->hash, E4)
+ resolve_ref_unsafe(E1, E2, E3, E4)
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
All but two of the call sites already have parameters using the hash
parameter of struct object_id, so convert them to take a pointer to the
struct directly. Also convert refs_read_refs_full, the underlying
implementation.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert update_ref, refs_update_ref, and write_pseudoref to use struct
object_id. Update the existing callers as well. Remove update_ref_oid,
as it is no longer needed.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Many codepaths did not diagnose write failures correctly when disks
go full, due to their misuse of write_in_full() helper function,
which have been corrected.
* jk/write-in-full-fix:
read_pack_header: handle signed/unsigned comparison in read result
config: flip return value of store_write_*()
notes-merge: use ssize_t for write_in_full() return value
pkt-line: check write_in_full() errors against "< 0"
convert less-trivial versions of "write_in_full() != len"
avoid "write_in_full(fd, buf, len) != len" pattern
get-tar-commit-id: check write_in_full() return against 0
config: avoid "write_in_full(fd, buf, len) < len" pattern
Many leaks of strbuf have been fixed.
* rs/strbuf-leakfix: (34 commits)
wt-status: release strbuf after use in wt_longstatus_print_tracking()
wt-status: release strbuf after use in read_rebase_todolist()
vcs-svn: release strbuf after use in end_revision()
utf8: release strbuf on error return in strbuf_utf8_replace()
userdiff: release strbuf after use in userdiff_get_textconv()
transport-helper: release strbuf after use in process_connect_service()
sequencer: release strbuf after use in save_head()
shortlog: release strbuf after use in insert_one_record()
sha1_file: release strbuf on error return in index_path()
send-pack: release strbuf on error return in send_pack()
remote: release strbuf after use in set_url()
remote: release strbuf after use in migrate_file()
remote: release strbuf after use in read_remote_branches()
refs: release strbuf on error return in write_pseudoref()
notes: release strbuf after use in notes_copy_from_stdin()
merge: release strbuf after use in write_merge_heads()
merge: release strbuf after use in save_state()
mailinfo: release strbuf on error return in handle_boundary()
mailinfo: release strbuf after use in handle_from()
help: release strbuf on error return in exec_woman_emacs()
...
The return value of write_in_full() is either "-1", or the
requested number of bytes[1]. If we make a partial write
before seeing an error, we still return -1, not a partial
value. This goes back to f6aa66cb95 (write_in_full: really
write in full or return error on disk full., 2007-01-11).
So checking anything except "was the return value negative"
is pointless. And there are a couple of reasons not to do
so:
1. It can do a funny signed/unsigned comparison. If your
"len" is signed (e.g., a size_t) then the compiler will
promote the "-1" to its unsigned variant.
This works out for "!= len" (unless you really were
trying to write the maximum size_t bytes), but is a
bug if you check "< len" (an example of which was fixed
recently in config.c).
We should avoid promoting the mental model that you
need to check the length at all, so that new sites are
not tempted to copy us.
2. Checking for a negative value is shorter to type,
especially when the length is an expression.
3. Linus says so. In d34cf19b89 (Clean up write_in_full()
users, 2007-01-11), right after the write_in_full()
semantics were changed, he wrote:
I really wish every "write_in_full()" user would just
check against "<0" now, but this fixes the nasty and
stupid ones.
Appeals to authority aside, this makes it clear that
writing it this way does not have an intentional
benefit. It's a historical curiosity that we never
bothered to clean up (and which was undoubtedly
cargo-culted into new sites).
So let's convert these obviously-correct cases (this
includes write_str_in_full(), which is just a wrapper for
write_in_full()).
[1] A careful reader may notice there is one way that
write_in_full() can return a different value. If we ask
write() to write N bytes and get a return value that is
_larger_ than N, we could return a larger total. But
besides the fact that this would imply a totally broken
version of write(), it would already invoke undefined
behavior. Our internal remaining counter is an unsigned
size_t, which means that subtracting too many byte will
wrap it around to a very large number. So we'll instantly
begin reading off the end of the buffer, trying to write
gigabytes (or petabytes) of data.
Signed-off-by: Jeff King <peff@peff.net>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a file .tsan-suppressions and list two functions in it: want_color()
and transfer_debug(). Both of these use the pattern
static int foo = -1;
if (foo < 0)
foo = bar();
where bar always returns the same non-negative value. This can cause
ThreadSanitizer to diagnose a race when foo is written from two threads.
That is indeed a race, although it arguably doesn't matter in practice
since it's always the same value that is written.
Add NEEDSWORK-comments to the functions so that this problem is not
forever swept way under the carpet.
The suppressions-file is used by setting the environment variable
TSAN_OPTIONS to, e.g., "suppressions=$(pwd)/.tsan-suppressions". Observe
that relative paths such as ".tsan-suppressions" might not work.
Signed-off-by: Martin Ågren <martin.agren@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>
We can use xsnprintf to do our truncation check with less
code. The error message isn't as specific, but the point is
that this isn't supposed to trigger in the first place
(because our buffer is big enough to handle any int).
Signed-off-by: Jeff King <peff@peff.net>
"git ls-remote" and "git archive --remote" are designed to work
without being in a directory under Git's control. However, recent
updates revealed that we randomly look into a directory called
.git/ without actually doing necessary set-up when working in a
repository. Stop doing so.
* jn/remote-helpers-with-git-dir:
remote helpers: avoid blind fall-back to ".git" when setting GIT_DIR
remote: avoid reading $GIT_DIR config in non-repo
Junio C Hamano
Emily Shaffer
Mike Hommey
Jeff Hostetler
Josh Steadmon
Jeff King
Randall S. Becker
Nguyễn Thái Ngọc Duy
Jonathan Tan
Brandon Williams
René Scharfe
brian m. carlson
Martin Ågren
Ævar Arnfjörð Bjarmason