On platforms where `size_t` does not have the same width as `unsigned
long`, passing a pointer to the former when a pointer to the latter is
expected can lead to problems.
Windows and 32-bit Linux are among the affected platforms.
In this instance, we want to store the size of the blob that was read in
that variable. However, `read_blob_data_from_index()` passes that
pointer to `read_object_file()` which expects an `unsigned long *`.
Which means that on affected platforms, the variable is not fully
populated and part of its value is left uninitialized. (On Big-Endian
platforms, this problem would be even worse.)
The consequence is that depending on the uninitialized memory's
contents, we may erroneously reject perfectly fine attributes.
Let's address this by passing a pointer to a variable of the expected
data type.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The existing text read as if "git checkout -b/-B name" were
equivalent to "git branch [-f] name", which clearly was not
what we wanted to say.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git branch -f name start" forces to recreate the named branch, but
the forcing does not defeat the "do not touch a branch that is
checked out elsewhere" safety valve.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The CURLOPT_PROTOCOLS (and matching CURLOPT_REDIR_PROTOCOLS) flag was
deprecated in curl 7.85.0, and using it generate compiler warnings as of
curl 7.87.0. The path forward is to use CURLOPT_PROTOCOLS_STR, but we
can't just do so unilaterally, as it was only introduced less than a
year ago in 7.85.0.
Until that version becomes ubiquitous, we have to either disable the
deprecation warning or conditionally use the "STR" variant on newer
versions of libcurl. This patch switches to the new variant, which is
nice for two reasons:
- we don't have to worry that silencing curl's deprecation warnings
might cause us to miss other more useful ones
- we'd eventually want to move to the new variant anyway, so this gets
us set up (albeit with some extra ugly boilerplate for the
conditional)
There are a lot of ways to split up the two cases. One way would be to
abstract the storage type (strbuf versus a long), how to append
(strbuf_addstr vs bitwise OR), how to initialize, which CURLOPT to use,
and so on. But the resulting code looks pretty magical:
GIT_CURL_PROTOCOL_TYPE allowed = GIT_CURL_PROTOCOL_TYPE_INIT;
if (...http is allowed...)
GIT_CURL_PROTOCOL_APPEND(&allowed, "http", CURLOPT_HTTP);
and you end up with more "#define GIT_CURL_PROTOCOL_TYPE" macros than
actual code.
On the other end of the spectrum, we could just implement two separate
functions, one that handles a string list and one that handles bits. But
then we end up repeating our list of protocols (http, https, ftp, ftp).
This patch takes the middle ground. The run-time code is always there to
handle both types, and we just choose which one to feed to curl.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The IOCTLFUNCTION option has been deprecated, and generates a compiler
warning in recent versions of curl. We can switch to using SEEKFUNCTION
instead. It was added in 2008 via curl 7.18.0; our INSTALL file already
indicates we require at least curl 7.19.4.
But there's one catch: curl says we should use CURL_SEEKFUNC_{OK,FAIL},
and those didn't arrive until 7.19.5. One workaround would be to use a
bare 0/1 here (or define our own macros). But let's just bump the
minimum required version to 7.19.5. That version is only a minor version
bump from our existing requirement, and is only a 2 month time bump for
versions that are almost 13 years old. So it's not likely that anybody
cares about the distinction.
Switching means we have to rewrite the ioctl functions into seek
functions. In some ways they are simpler (seeking is the only
operation), but in some ways more complex (the ioctl allowed only a full
rewind, but now we can seek to arbitrary offsets).
Curl will only ever use SEEK_SET (per their documentation), so I didn't
bother implementing anything else, since it would naturally be
completely untested. This seems unlikely to change, but I added an
assertion just in case.
Likewise, I doubt curl will ever try to seek outside of the buffer sizes
we've told it, but I erred on the defensive side here, rather than do an
out-of-bounds read.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The two options do exactly the same thing, but the latter has been
deprecated and in recent versions of curl may produce a compiler
warning. Since the UPLOAD form is available everywhere (it was
introduced in the year 2000 by curl 7.1), we can just switch to it.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
On platforms where `size_t` does not have the same width as `unsigned
long`, passing a pointer to the former when a pointer to the latter is
expected can lead to problems.
Windows and 32-bit Linux are among the affected platforms.
In this instance, we want to store the size of the blob that was read in
that variable. However, `read_blob_data_from_index()` passes that
pointer to `read_object_file()` which expects an `unsigned long *`.
Which means that on affected platforms, the variable is not fully
populated and part of its value is left uninitialized. (On Big-Endian
platforms, this problem would be even worse.)
The consequence is that depending on the uninitialized memory's
contents, we may erroneously reject perfectly fine attributes.
Let's address this by passing a pointer to a variable of the expected
data type.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stop initializing "name" because it is set again before use.
Let quote_c_style() write directly to "sb" instead of taking a detour
through "quoted". This avoids an allocation and a string copy. The
result is the same because the function only appends.
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
expand_show_tree() borrows the base strbuf given to us by read_tree() to
build the full path of the current entry when handling %(path). Only
its indirect caller, show_tree_fmt(), removes the added entry name.
That works fine as long as %(path) is only included once in the format
string, but accumulates duplicates if it's repeated:
$ git ls-tree --format='%(path) %(path) %(path)' HEAD M*
Makefile MakefileMakefile MakefileMakefileMakefile
Reset the length after each use to get the same expansion every time;
here's the behavior with this patch:
$ ./git ls-tree --format='%(path) %(path) %(path)' HEAD M*
Makefile Makefile Makefile
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
'ORIG_HEAD' is written at the start of the rebase, but is not guaranteed
to still point to the original branch tip at the end of the rebase.
Indeed, using other commands that write 'ORIG_HEAD' during the rebase,
like splitting a commit using 'git reset HEAD^', will lead to 'ORIG_HEAD'
being overwritten. This causes confusion for some users [1].
Add a note about that in the 'Description' section, and mention the more
robust alternative of using the branch's reflog.
[1] https://lore.kernel.org/git/28ebf03b-e8bb-3769-556b-c9db17e43dbb@gmail.com/T/#m827179c5adcfb504d67f76d03c8e6942b55e5ed0
Reported-by: Erik Cervin Edin <erik@cervined.in>
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Acked-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When mentioning 'ORIG_HEAD', be explicit about which command write that
pseudo-ref, namely 'git am', 'git merge', 'git rebase' and 'git reset'.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Acked-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The fact that 'git merge' writes 'ORIG_HEAD' before performing the merge
is missing from the documentation of the command.
Mention it in the 'Description' section.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Acked-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The fact that 'git reset' writes 'ORIG_HEAD' before changing HEAD is
mentioned in an example, but is missing from the 'Description' section.
Mention it in the discussion of the "'git reset' [<mode>] [<commit>]"
form of the command.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Acked-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 67ac1e1d57 (cherry-pick/revert: add support for
-X/--strategy-option, 2010-12-10) added an example to the documentation
of 'git cherry-pick'. This example mentions how to abort a failed
cherry-pick and retry with an additional merge strategy option.
The command used in the example to abort the cherry-pick is 'git reset
--merge ORIG_HEAD', but cherry-pick does not write 'ORIG_HEAD' before
starting its operation. So this command would checkout a commit
unrelated to what was at HEAD when the user invoked cherry-pick.
Use 'git cherry-pick --abort' instead.
Signed-off-by: Philippe Blain <levraiphilippeblain@gmail.com>
Acked-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When 1819ad327b (grep: fix multibyte regex handling under macOS,
2022-08-26) started to use the native regex library instead of Git's
own (compat/regex/), it lost support for alternation in basic
regular expressions.
Bring it back by enabling the flag REG_ENHANCED on macOS when
compiling basic regular expressions.
Reported-by: Marco Nenciarini <marco.nenciarini@enterprisedb.com>
Suggested-by: Jeff King <peff@peff.net>
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The prepare_temp_file() function takes a diff_filespec as well as a
filename. But it is almost certainly an error to pass in a name that
isn't the filespec's "path" parameter, since that is the only thing that
reliably tells us how to find the content (and indeed, this was the
source of a recently-fixed bug).
So let's drop the redundant "name" parameter and just use one->path
throughout the function. This simplifies the interface a little bit, and
makes it impossible for calling code to get it wrong.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since the previous commit, setting up the tempfile for an external diff
uses df->path from the diff_filespec, rather than the logical name. This
means add_external_diff_name() does not need to take a "name" parameter
at all, and we can drop it. And that in turn lets us simplify the
conditional for handling renames (when the "other" name is non-NULL).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we're going to run an external diff, we have to make the contents
of the pre- and post-images available either by dumping them to a
tempfile, or by pointing at a valid file in the worktree. The logic of
this is all handled by prepare_temp_file(), and we just pass in the
filename and the diff_filespec.
But there's a gotcha here. The "filename" we have is a logical filename
and not necessarily a path on disk or in the repository. This matters in
at least one case: when using "--relative", we may have a name like
"foo", even though the file content is found at "subdir/foo". As a
result, we look for the wrong path, fail to find "foo", and claim that
the file has been deleted (passing "/dev/null" to the external diff,
rather than the correct worktree path).
We can fix this by passing the pathname from the diff_filespec, which
should always be a full repository path (and that's what we want even if
reusing a worktree file, since we're always operating from the top-level
of the working tree).
The breakage seems to go all the way back to cd676a5136 (diff
--relative: output paths as relative to the current subdirectory,
2008-02-12). As far as I can tell, before then "name" would always have
been the same as the filespec's "path".
There are two related cases I looked at that aren't buggy:
1. the only other caller of prepare_temp_file() is run_textconv(). But
it always passes the filespec's path field, so it's OK.
2. I wondered if file renames/copies might cause similar confusion.
But they don't, because run_external_diff() receives two names in
that case: "name" and "other", which correspond to the two sides of
the diff. And we did correctly pass "other" when handling the
post-image side. Barring the use of "--relative", that would always
match "two->path", the path of the second filespec (and the rename
destination).
So the only bug is just the interaction with external diff drivers and
--relative.
Reported-by: Carl Baldwin <carl@ecbaldwin.net>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The sparse checkout documentation states that the cone mode pattern set
is limited to patterns that either recursively include directories or
patterns that match all files in a directory. In the sparse checkout
file, the former manifest in the form:
/A/B/C/
while the latter become a pair of patterns either in the form:
/A/B/
!/A/B/*/
or in the special case of matching the toplevel files:
/*
!/*/
The 'add_pattern_to_hashsets()' function contains checks which serve to
disable cone-mode when non-cone patterns are encountered. However, these
do not catch when the pattern list attempts to match a single file or
directory, e.g. a pattern in the form:
/A/B/C
This causes sparse-checkout to exhibit unexpected behaviour when such a
pattern is in the sparse-checkout file and cone mode is enabled.
Concretely, with the pattern like the above, sparse-checkout, in
non-cone mode, will only include the directory or file located at
'/A/B/C'. However, with cone mode enabled, sparse-checkout will instead
just manifest the toplevel files but not any file located at '/A/B/C'.
Relatedly, issues occur when supplying the same kind of filter when
partial cloning with '--filter=sparse:oid=<oid>'. 'upload-pack' will
correctly just include the objects that match the non-cone pattern
matching. Which means that checking out the newly cloned repo with the
same filter, but with cone mode enabled, fails due to missing objects.
To fix these issues, add a cone mode pattern check that asserts that
every pattern is either a directory match or the pattern '/*'. Add a
test to verify the new pattern check and modify another to reflect that
non-directory patterns are caught earlier.
Signed-off-by: William Sprent <williams@unity3d.com>
Acked-by: Victoria Dye <vdye@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Improve the usage we emit on e.g. "git bundle create" to note why
we're showing the usage, it's because the "<file>" argument is
missing.
We know that'll be the case for all parse_options_cmd_bundle() users,
as they're passing the "char **bundle_file" parameter, which as the
context shows we're expected to populate.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As noted in 891cb09db6 (bundle: don't segfault on "git bundle
<subcmd>", 2022-12-20) the "newargc" in this function is redundant to
using our own "argc". Let's refactor the function to avoid needlessly
introducing another variable.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
At GitLab we have recently received a report where a repository was left
with a corrupted `packed-refs` file after the node hard-crashed even
though `core.fsync=reference` was set. This is something that in theory
should not happen if we correctly did the atomic-rename dance to:
1. Write the data into a temporary file.
2. Synchronize the temporary file to disk.
3. Rename the temporary file into place.
So if we crash in the middle of writing the `packed-refs` file we should
only ever see either the old or the new state of the file.
And while we do the dance when writing the `packed-refs` file, there is
indeed one gotcha: we use a `FILE *` stream to write the temporary file,
but don't flush it before synchronizing it to disk. As a consequence any
data that is still buffered will not get synchronized and a crash of the
machine may cause corruption.
Fix this bug by flushing the file stream before we fsync.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since aef7d75e58 (builtin/bundle.c: let parse-options parse
subcommands, 2022-08-19) we've been segfaulting if no argument was
provided.
The fix is easy, as all of the "git bundle" subcommands require a
non-option argument we can check that we have arguments left after
calling parse-options().
This makes use of code added in 73c3253d75 (bundle: framework for
options before bundle file, 2019-11-10), before this change that code
has always been unreachable. In 73c3253d75 we'd never reach it as we
already checked "argc < 2" in cmd_bundle() itself.
Then when aef7d75e58 (whose segfault we're fixing here) migrated this
code to the subcommand API it removed that "argc < 2" check, but we
were still checking the wrong "argc" in parse_options_cmd_bundle(), we
need to check the "newargc". The "argc" will always be >= 1, as it
will necessarily contain at least the subcommand name
itself (e.g. "create").
As an aside, this could be safely squashed into this, but let's not do
that for this minimal segfault fix, as it's an unrelated refactoring:
--- a/builtin/bundle.c
+++ b/builtin/bundle.c
@@ -55,13 +55,12 @@ static int parse_options_cmd_bundle(int argc,
const char * const usagestr[],
const struct option options[],
char **bundle_file) {
- int newargc;
- newargc = parse_options(argc, argv, NULL, options, usagestr,
+ argc = parse_options(argc, argv, NULL, options, usagestr,
PARSE_OPT_STOP_AT_NON_OPTION);
- if (!newargc)
+ if (!argc)
usage_with_options(usagestr, options);
*bundle_file = prefix_filename(prefix, argv[0]);
- return newargc;
+ return argc;
}
static int cmd_bundle_create(int argc, const char **argv, const char *prefix) {
Reported-by: Hubert Jasudowicz <hubertj@stmcyber.pl>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Tested-by: Hubert Jasudowicz <hubertj@stmcyber.pl>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Get rid of deprecation warnings in the CI runs. Also gets the latest
security patches.
Signed-off-by: Chris. Webster <chris@webstech.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A message in the step log will refer to the Summary output.
The job summary output is using markdown to improve readability. The
git commands and commits with errors are now in ordered lists.
Commits and files in error are links to the user's repository.
Signed-off-by: Chris. Webster <chris@webstech.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Make the errors more visible by adding them to the job summary and
display the git commands that will usually fix the problem.
Signed-off-by: Chris. Webster <chris@webstech.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It has been a frequent matter of contention that the win+VS jobs not
only take a long time to run, but are also more easily broken than the
other jobs (because they do not use the same `Makefile`-based builds as
all other jobs), and to make matters worse, these breakages are also
much harder to diagnose and fix than other jobs', especially for
contributors who are happy to stay away from Windows.
The purpose of these win+VS jobs is to maintain the CMake-based build
of Git, with the target audience being Visual Studio users on Windows
who are typically quite unfamiliar with `make` and POSIX shell
scripting, but the benefit of whose expertise we want for the Git
project nevertheless.
The CMake support was introduced for that specific purpose, and already
early on concerns were raised that it would put an undue burden on
contributors to ensure that these jobs pass in CI, when they do not have
access to Windows machines (nor want to have that).
This developer's initial hope was that it would be enough to fix win+VS
failures and provide the changes to be squashed into contributors'
patches, and that it would be worth the benefit of attracting
Windows-based developers' contributions.
Neither of these hopes have panned out.
To lower the frustration, and incidentally benefit from using way less
build minutes, let's just not run the win+VS jobs by default, which
appears to be the consensus of the mail thread leading up to
https://lore.kernel.org/git/xmqqk0311blt.fsf@gitster.g/
Since the Git for Windows project still needs to at least try to attract
more of said Windows-based developers, let's keep the jobs, but disable
them everywhere except in Git for Windows' fork. This will help because
Git for Windows' branch thicket is "continuously rebased" via automation
to the `shears/maint`, `shears/main`, `shears/next` and `shears/seen`
branches at https://github.com/git-for-windows/git. That way, the Git
for Windows project will still be notified early on about potential
breakages, but the Git project won't be burdened with fixing them
anymore, which seems to be the best compromise we can get on this issue.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When the -L argument to "git log" is passed the zero-width regular
expression "$" (as in "-L :$:line-range.c"), this results in an
infinite loop in find_funcname_matching_regexp().
Modify find_funcname_matching_regexp to correctly match the entire line
instead of the zero-width match at eol and update the loop condition to
prevent an infinite loop in the event of other undiscovered corner cases.
The primary change is that we pre-decrement the beginning-of-line marker
('bol') before comparing it to '\n'. In the case of '$', where we match the
'\n' at the end of the line and start the loop with bol == eol, this
ensures that bol will find the beginning of the line on which the match
occurred.
Originally reported in <https://stackoverflow.com/q/74690545/147356>.
Signed-off-by: Lars Kellogg-Stedman <lars@oddbit.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
