| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
We currently disable deprecation synchronization warnings in case we're
building with Clang. We check for Clang by doing a string comparison on
the compiler identification, but this seems to have been broken by an
update in macOS' image as the compiler ID has changed to "AppleClang".
Let's just unconditionally disable this warning on Unix platforms. We
never add the deprecated attribute anyway, so the warning doesn't help
us at all.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The `CMAKE_MINIUM_REQUIRE()` function not only sets up the minimum
required CMake version of a project, but it will also at the same time
set the CMake policy version. In effect this means that all policies
that have been introduced before the minimum CMake version will be
enabled automatically.
When updating our minimum required version ebabb88f2 (cmake: update
minimum CMake version to v3.5.1, 2019-10-10), we didn't remove any of
the policies we've been manually enabling. The newest CMake policy we've
been enabling is CMP0054, which was introduced back in CMake v3.1. As a
result, we can now just remove all manual calls to `CMAKE_POLICY()`.
|
|
|
|
|
|
|
|
|
| |
We currently have an option that adds options for profiling to both our
CFLAGS and LDFLAGS. Having such flags behind various build options is
not really sensible at all, since users should instead set up those
flags via environment variables supported by CMake itself.
Let's remove this option.
|
|
|
|
|
|
|
|
|
|
|
|
| |
We currently have support for generating tags via ctags as part of our
build system. We aren't really in the place of supporting any tooling
that exists apart from the actual build environment, as doing so adds
additional complexity and maintenance burden to our build instructions.
This is in fact nicely demonstrated by this particular option, as it
hasn't been working anymore since commit e5c9723d0 (cmake: move library
build instructions into subdirectory, 2017-06-30).
As a result, this commit removes support for building CTags
|
|
|
|
|
|
|
|
| |
Our custom CMake module currently live in "cmake/Modules". As the
"cmake/" directory doesn't contain anything except the "Modules"
directory, it doesn't really make sense to have the additional
intermediate directory. So let's instead move the modules one level up
into the "cmake/" top level directory.
|
|
|
|
|
|
| |
This feature requires Visual Studio 2015 (MSVC_VERSION = 1900) or later. As the
minimum required CMake version is currently less than 3.7, GREATER_EQUAL is not
available to us and we must invert the result of the LESS operator.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We currently hand-code logic to configure where to install our artifacts
via the `LIB_INSTALL_DIR`, `INCLUDE_INSTALL_DIR` and `BIN_INSTALL_DIR`
variables. This is reinventing the wheel, as CMake already provide a way
to do that via `CMAKE_INSTALL_<DIR>` paths, e.g. `CMAKE_INSTALL_LIB`.
This requires users of libgit2 to know about the discrepancy and will
require special hacks for any build systems that handle these variables
in an automated way. One such example is Gentoo Linux, which sets up
these paths in both the cmake and cmake-utils eclass.
So let's stop doing that: the GNUInstallDirs module handles it in a
better way for us, especially so as the actual values are dependent on
CMAKE_INSTALL_PREFIX. This commit removes our own set of variables and
instead refers users to use the standard ones.
As a second benefit, this commit also fixes our pkgconfig generation to
use the GNUInstallDirs module. We had a bug there where we ignored the
CMAKE_INSTALL_PREFIX when configuring the libdir and includedir keys, so
if libdir was set to "lib64", then libdir would be an invalid path. With
GNUInstallDirs, we can now use `CMAKE_INSTALL_FULL_LIBDIR`, which
handles the prefix for us.
|
|
|
|
|
|
| |
This commit also switches our SOVERSION to be "$MAJOR.$MINOR" instead of
"$MINOR", only. This is in preparation of v1.0, where the previous
scheme would've stopped working in an obvious way.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
OpenSSL doesn't initialize bytes on purpose in order to generate
additional entropy. Valgrind isn't too happy about that though, causing
it to generate warninings about various issues regarding use of
uninitialized bytes.
We traditionally had some infrastructure to silence these errors in our
OpenSSL stream implementation, where we invoke the Valgrind macro
`VALGRIND_MAKE_MEMDEFINED` in various callbacks that we provide to
OpenSSL. Naturally, we only include these instructions if a preprocessor
define "VALGRIND" is set, and that in turn is only set if passing
"-DVALGRIND" to CMake. We do that in our usual Azure pipelines, but we
in fact forgot to do this in our nightly build. As a result, we get a
slew of warnings for these nightly builds, but not for our normal
builds.
To fix this, we could just add "-DVALGRIND" to our nightly builds. But
starting with commit d827b11b6 (tests: execute leak checker via CTest
directly, 2019-06-28), we do have a secondary variable that directs
whether we want to use memory sanitizers for our builds. As such, every
user wishing to use Valgrind for our tests needs to pass both options
"VALGRIND" and "USE_LEAK_CHECKER", which is cumbersome and error prone,
as can be seen by our own builds.
Instead, let's consolidate this into a single option, removing the old
"-DVALGRIND" one. Instead, let's just add the preprocessor directive if
USE_LEAK_CHECKER equals "valgrind" and remove "-DVALGRIND" from our own
pipelines.
|
|
|
|
|
|
|
|
|
| |
Tracing is meant to be extremely low-impact when not enabled. We
currently ship no tracing calls in libgit2, but if / when we do, the
tracing infrastructure is created to skip tracing as quickly as
possible. It should compile to a simple test when tracing is off.
Thus, there's on reason to not enable it by default.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Back in commit cf9f34521 (cmake: bump minimum version to 2.8.11,
2017-09-06), we have bumped the minimum CMake version to require at
least v2.8.11. The main hold-backs back then were distributions like
RHEL/CentOS as well as Ubuntu Trusty, which caused us to not target a
more modern version. Nowadays, Ubuntu Trusty has been EOL'd and CentOS 6
has CMake v3.6.1 available via the EPEL6 repository, and thus it seems
fair to upgrade to a more recent version.
Going through repology [1], one can see that all supported mainstream
distributions do in fact have CMake 3 available. Going through the list,
the minimum version that is supported by all mainstream distros is in
fact v3.5.1:
- CentOS 6 via EPEL6: 3.6.1
- Debian Oldstable: 3.7.2
- Fedora 26: 3.8.2
- OpenMandriva 3.x: 3.5.1
- Slackware 14.2: 3.5.2
- Ubuntu 16.04: 3.5.1
Consequentally, let's upgrade CMake to the minimum version of 3.5.1 and
remove all the version CMake checks that aren't required anymore.
[1]: https://repology.org/project/cmake/versions
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Right now, we have an awful hack in our test CI setup that extracts the
test command from CTest's output and then prepends the leak checker.
This is dependent on non-machine-parseable output from CMake and also
breaks on various ocassions, like for example when we have spaces in the
current path or when the path contains backslashes. Both conditions may
easily be triggered on Win32 systems, and in fact they do break our
Azure Pipelines builds.
Remove the awful hack in favour of a new CMake build option
"USE_LEAK_CHECKER". If specifying e.g. "-DUSE_LEAK_CHECKER=valgrind",
then we will set up all tests to be run under valgrind. Like this, we
can again simply execute ctest without needing to rely on evil sourcery.
|
|
|
|
|
|
|
|
|
|
|
| |
Add the `-Wno-documentation-deprecated-sync` switch when compiling with
clang, since our documentation adds `deprecated` markers, but we do not
add the deprecation attribute in the code itself. (ie, the code is out
of sync with the docs).
In fact, we do not _want_ to mark these items as deprecated in the code,
at least not yet, as we are not quite ready to bother our end-users with
this since they're not going away.
|
|
|
|
|
|
|
| |
MinGW uses gcc, which expects POSIX formatting for printf, but uses the
Windows C library, which uses its own format specifiers. Therefore, it
gets confused about format specifiers. Disable warnings for format
specifiers.
|
|
|
|
|
|
|
|
|
| |
If building libgit2 with `-DUSE_HTTPS=NO`, then CMake will
generate an error complaining that there's no usable HTTPS
backend for NTLM. In fact, it doesn't make sense to support NTLM
when we don't support HTTPS. So let's should just have
NTLM default to OFF when HTTPS is disabled to make life easier
and to fix our OSSFuzz builds failing.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The interactions between `USE_HTTPS` and `SHA1_BACKEND` have been
streamlined. Previously we would have accepted not quite working
configurations (like, `-DUSE_HTTPS=OFF -DSHA1_BACKEND=OpenSSL`) and, as
the OpenSSL detection only ran with `USE_HTTPS`, the link would fail.
The detection was moved to a new `USE_SHA1`, modeled after `USE_HTTPS`,
which takes the values "CollisionDetection/Backend/Generic", to better
match how the "hashing backend" is selected, the default (ON) being
"CollisionDetection".
Note that, as `SHA1_BACKEND` is still used internally, you might need to
check what customization you're using it for.
|
|
|
|
|
|
|
|
|
|
|
| |
Our bundled http-parser includes bugfixes, therefore we should prefer
our http-parser until such time as we can identify that the system
http-parser has these bugfixes (using a version check).
Since these bugs are - at present - minor, retain the ability for users
to force that they want to use the system http-parser anyway. This does
change the cmake specification so that people _must_ opt-in to the new
behavior knowingly.
|
|
|
|
| |
Include https://github.com/ethomson/ntlmclient as a dependency.
|
|
|
|
| |
This avoids any misunderstanding with the REGEX keyword in cmake.
|
|
|
|
|
|
|
| |
Use PCRE2 and its POSIX compatibility layer if requested by the user.
Although PCRE2 is adequate for our needs, the PCRE2 POSIX layer as
installed on Debian and Ubuntu systems is broken, so we do not opt-in to
it by default to avoid breaking users on those platforms.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Users can now select which regex implementation they want to use: one of
the system `regcomp_l`, the system PCRE, the builtin PCRE or the
system's `regcomp`.
By default the system `regcomp_l` will be used if it exists, otherwise
the system PCRE will be used. If neither of those exist, then the
builtin PCRE implementation will be used.
The system's `regcomp` is not used by default due to problems with
locales.
|
|
|
|
|
|
|
| |
Add a CMake option to enable hard deprecation; the resultant library
will _not_ include any deprecated functions. This may be useful for
internal CI builds that create libraries that are not shared with
end-users to ensure that we do not use deprecated bits internally.
|
|
|
|
|
|
|
|
|
|
| |
Although the error functions were deprecated, we did not properly mark
them as deprecated. We need to include the `deprecated.h` file in order
to ensure that the functions get their export attributes.
Similarly, do not define `GIT_DEPRECATE_HARD` within the library, or
those functions will also not get their export attributes. Define that
only on the tests and examples.
|
|
|
|
|
|
|
| |
A number of source files have their implementation #ifdef'd out (because
they target another platform). MSVC warns on empty compilation units
(with warning LNK4221). Ignore warning 4221 when creating the object
library.
|
|
|
|
|
| |
Ensure that we do not use any deprecated functions in the library
source, test code or examples.
|
|
|
|
|
|
|
| |
To explicitly break end-users who were specifying STDCALL, explicitly
fail the cmake process to ensure that they know that they need to change
their bindings. Otherwise, we would quietly ignore their option and the
resulting cdecl library would produced undefined behavior.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The recommendation from engineers within Microsoft is that libraries
should have a calling convention specified in the public API, and that
calling convention should be cdecl unless there are strong reasons to
use a different calling convention.
We previously offered end-users the choice between cdecl and stdcall
calling conventions. We did this for presumed wider compatibility: most
Windows applications will use cdecl, but C# and PInvoke default to
stdcall for WINAPI compatibility. (On Windows, the standard library
functions are are stdcall so PInvoke also defaults to stdcall.)
However, C# and PInvoke can easily call cdecl APIs by specifying an
annotation.
Thus, we will explicitly declare ourselves cdecl and remove the option
to build as stdcall.
|
|
|
|
|
|
|
|
|
|
|
|
| |
After taking into consideration the following, I think this should be
removed :
- OpenSSL isn't the default on Apple platforms
- you have to jump through hoops to get CMake to use OpenSSL on macOS
(headers aren't in `/usr/include`, so you have to provide `-DOPENSSL_*`
overrides)
- users are likely (as getting anywhere near the installed 0.9.8 version
is insanity IMHO) to package a "modern" version, which wouldn't be
marked as deprecated
|
|\
| |
| | |
Unused function warnings
|
| |
| |
| |
| |
| |
| |
| |
| | |
Together with the warnings for unused warnings, we always had warnings
for unused constant variables disabled since commit 823c0e9cc (Fix
broken logic for attr cache invalidation, 2014-04-17). As we have now
fixed all occurrences of such variables, we can safely enable those
warnings again.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Ever since commit 823c0e9cc (Fix broken logic for attr cache invalidation,
2014-04-17), we have completely disabled warnings for unused functions. The only
comment that was added back then is about "annoying extra warnings" from Clang,
but in fact we shouldn't just ignore warnings about functions which aren't used
at all. Instead, the right thing would be to either only conditionally compile
functions that aren't used in all configurations or, alternatively, to remove
functions that aren't required at all.
As remaining instances of unused functions have been removed in the last two
commits, re-enable the warning.
|
| |
| |
| |
| |
| |
| |
| |
| | |
The `-Wdocumentation-deprecated-sync` option will warn when there is a
doxygen `\deprecated` tag but there is no corresponding deprecation
attribute on the function. We want to encourage users to not use
particular APIs by marking them deprecated in the documentation without
necessarily raising a compiler warning by marking an item as deprecated.
|
|/
|
|
|
| |
We previously used cURL to support HTTP proxies. Now that we've added
this support natively, we can remove the curl dependency.
|
|
|
|
|
| |
We use libssh2. We do not use libssh. Make sure to disambiguate them
correctly.
|
|
|
|
|
|
|
|
| |
While GCC enables int-conversion warnings by default, it will currently
only warn about such errors even in case where "-DENABLE_WERROR=ON" has
been passed to CMake. Explicitly enable int-conversion warnings by using
our `ENABLE_WARNINGS` macro, which will automatically use
"-Werror=int-conversions" in case it has been requested by the user.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Quoting from CMP0054's documentation:
Only interpret if() arguments as variables or keywords when
unquoted.
CMake 3.1 and above no longer implicitly dereference variables or
interpret keywords in an if() command argument when it is a Quoted
Argument or a Bracket Argument.
The OLD behavior for this policy is to dereference variables and
interpret keywords even if they are quoted or bracketed. The NEW
behavior is to not dereference variables or interpret keywords that
have been quoted or bracketed.
The previous behaviour could be quite unexpected. Quoted arguments might
be expanded in case where the value of the argument corresponds to a
variable. E.g. `IF("MONKEY" STREQUAL "MONKEY")` would have been expanded
to `IF("1" STREQUAL "1")` iff `SET(MONKEY 1)` was set. This behaviour
was weird, and recent CMake versions have started to complain about this
if they see ambiguous situations. Thus we want to disable it in favor of
the new behaviour.
|
|
|
|
|
|
| |
Our CMake coding style dictates that there should be no space between
`IF` and its opening `(`. Adjust our policy statements to honor this
style.
|
|\
| |
| | |
CI: refactoring
|
| |
| |
| |
| |
| | |
This reverts commit a2d73f5643814cddf90d5bf489332e14ada89ab8.
Using clar to propagate the XML settings was a mistake.
|
|/
|
|
|
| |
We do not currently have any warnings in this regard, but it's good practice to
have them on in case we introduce something.
|
|
|
|
|
|
| |
Introduce a CLAR_XML option, to run the `ctest` commands with the new
`-r` flag to clar. Permitted values are `OFF`, `ON` and a directory to
write the XML test results to.
|
|
|
|
|
|
|
|
| |
This performs a compile-check by using CMake support, to differentiate the GNU
version from the BSD version of qsort_r.
Module taken from 4f252abea5f1d17c60f6ff115c9c44cc0b6f1df6, which I've checked
against CMake 2.8.11.
|
|
|
|
|
|
|
|
|
|
|
| |
Both the USE_SANITIZER and USE_COVERAGE options are convenience options
that turn on a set of CFLAGS. Despite our own set of CFLAGS required to
build libgit2, we have no real business to mess with them, though, as
they can easily be passed in by the user via specifying the CFLAGS
environment variable. The reasoning behind not providing them is that as
soon as we start adding those for some usecases, users might ask for
other sets of CFLAGS catering to their specific need in another usecase.
Thus, we do not want to support them here.
|
| |
|
|
|
|
|
|
|
| |
We do want to notify users compiling our source code early on if they
try to use C flags which aren't supported. Add a new macro `AddCFlag`,
which results in a fatal error in case the flag is not supported, and
use it for our fuzzing flags.
|
|
|
|
|
|
|
|
|
|
|
| |
Our layout uses names like "examples" or "tests" which is why the "fuzz"
directory doesn't really fit in here. Rename the directory to be called
"fuzzers" instead. Furthermore, we rename the fuzzer "fuzz_packfile_raw"
to "packfile_raw_fuzzer", which is also in line with the already
existing fuzzer at google/oss-fuzz.
While at it, rename the "packfile_raw" fuzzer to instead just be called
"packfile" fuzzer.
|
|
|
|
|
|
| |
This change adds support for building a fuzz target for exercising the
packfile parser, as well as documentation. It also runs the fuzzers in
Travis to avoid regressions.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
While the aim of libgit2 was to conform to C90 code, we never instructed
the compiler to enforce C90 compliance. Thus, quite a few violations
were able to get into our code base, which have been removed with the
previous commits. As we are now able to build libgit2 with C90 enforced,
we can set the C_STANDARD property for our own build targets.
Note that we explicitly avoid setting the C standard for our third-party
dependencies. At least the zlib target does not build with C90 enforced,
and we do not want to fix them by deviating from upstream. Thus we
simply enforce no standard for them.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Libraries found by CMake modules are usually handled with their full
path. This makes linking against those libraries a lot more robust when
it comes to libraries in non-standard locations, as otherwise we might
mix up libraries from different locations when link directories are
given.
One excemption are libraries found by PKG_CHECK_MODULES. Instead of
returning libraries with their complete path, it will return the
variable names as well as a set of link directories. In case where
multiple sets of the same library are installed in different locations,
this can lead the compiler to link against the wrong libraries in the
end, when link directories of other dependencies are added.
To fix this shortcoming, we need to manually resolve library paths
returned by CMake against their respective library directories. This is
an easy task to do with `FIND_LIBRARY`.
|