This was committed previously as 4975ae172 and reverted, because the flags were
applied even if the checks didn't pass. This is the same commit, fixed up to
actually disable the functionality when necessary.
Enabled automatically if boost >= 1.49.
See: https://svn.boost.org/trac/boost/ticket/2309
Also, check for a default visibility attribute, so that we can mark future
api functions correctly.
Enabled automatically if boost >= 1.49.
See: https://svn.boost.org/trac/boost/ticket/2309
Also, check for a default visibility attribute, so that we can mark future
api functions correctly.
When the libpath doesn't line up with the value from config.sub, we don't find
the correct path to boost's libs. This adds a hack to try another path before
giving up.
Should close#3219.
While we're at it, reduce the use of LIBS as well. This makes dependencies
explicit.
Fixes building with (the not-yet-merged) libsecp256k1 as well.
Github-Pull: #4689
Rebased-By: Wladimir J. van der laan <laanwj@gmail.com>
Rebased-From: 909b347 c0e5dda
When QT is linked statically, macdeploy can't infer its paths. While plugins
and frameworks don't need to be packaged, translations still do (for now).
If clock_gettime is implemented outside of libc (librt in this case), configure
would fail when testing boost. Since clock_gettime is not present on all OSs,
boost only uses it when it can. Check for it in librt and add it to LIBS if
found, but don't fail if it's not (since boost won't be expecting it in this
case).
Also, reverse the link order as necessary for static libs.
Note that it's possible that there are other similar cases for boost, which may
be handled the same way.
This is the first part of a huge effort to rework the handling of dependencies.
To start, this change allows all supported platforms to build against a static
Qt. 5.2.1 and 5.3 have been successfully tested against osx64, win32, win64,
linux32, and linux64.
It also makes a small change to the windows config, to allow linking against
qt builds with or without built-in libjpeg/libpng/libpcre/libz.
The actual build processes to take advantage of these changes (for gitian and
pull-tester) are coming soon. Until then, this should be a no-op.
Note: This is added to our existing automake targets rather than as a
libtool-style lib. The switch to libtool-style targets can come later if it
proves to not add any complications.
bitcoin-config.h moved, but the old file is likely to still exist when
reconfiguring or switching branches. This would've caused files to not rebuild
correctly, and other strange problems.
Make the path explicit so that the old one cannot be found.
Core libs use config/bitcoin-config.h.
Libs (like crypto) which don't want access to bitcoin's headers continue
to use -Iconfig and #include bitcoin-config.h.
The rcc tool is quirky and only honors files in the same directory as the qrc.
When doing an out-of-tree build (as 'make distcheck' does), the generated
translation files end up in a different path, so rcc can't find them.
Split them up so that rcc is run twice: once for static source files and once
for generated files.
Now that the build is non-recursive, adding to AM_CPPFLAGS means adding to
_all_ cppflags.
Logical groups of includes have been added instead, and are used individually
by various targets.
- Some file generation was still noisy, silence it.
- AM_V_GEN is used rather than @ so that 'make V=1' works as intended
- Cut down on file copies and moves when using sed, use pipes instead
- Avoid the use of top_ and abs_ dirs where possible
Build logic moves from individual Makefile.am's to include files, which
the main src/Makefile.am includes. This avoids having to manage a gigantic
single Makefile.
TODO: Move the rules from the old Makefile.include to where they actually
belong and nuke the old file.
Qt5 Removed the qt_mac_set_dock_menu function and left no replacement. It was
later re-added and deprecated for backwards-compatibility.
Qt5.2 adds the non-deprecated QMenu::setAsDockMenu(). Use that when possible.
Because this class replaces some usages of CBigNum, tests have been added to
verify that they function the same way. The only difference in their usage is
the handling of out-of-range numbers.
While operands are constrained to [-0x7FFFFFFF,0x7FFFFFFF], the results may
overflow. The overflowing result is technically unbounded, but in practice
it can be no bigger than the result of an operation on two operands. This
implementation limits them to the size of an int64.
CBigNum was unaware of this constraint, so it allowed for unbounded results,
which were then checked before use. CScriptNum asserts if an arithmetic
operation will overflow an int64_t, since scripts are not able to reach those
numbers anyway. Additionally, CScriptNum will throw an exception when
constructed from a vector containing more than 4 bytes This mimics the previous
CastToBigNum behavior.
glibc/libstdc++ have added new symbols in later releases. When running a new
binary against an older glibc, the run-time linker is unable to resolve the
new symbols and the binary refuses to run.
This can be fixed by adding our own versions of those functions, so that the
build-time linker does not emit undefined symbols for them.
This enables our binary releases to work on older Linux distros, while not
incurring the downsides of a fully static binary.
For qt5.2 on osx, the qcocoa plugin is mandatory. However, it fails to load
when qt.conf specifies the "plugin" path instead of the expected "Plugin". This
is in line with the documentation:
https://qt-project.org/doc/qt-5.0/qtdoc/qt-conf.html
I'm not sure how the plugins were loading before, unless the case-sensitivity
for OSX is new.
Since we're now properly linking against static plugins for qt binaries,
we need to ensure that they're Imported properly.
Without these Imports, the linker drops some of the unused linked libs, causing
undefined symbols in QtCore.
At least Debian/Ubuntu use 'qtchooser' for switching between qt4/qt5 binaries.
It is a wrapper for all qt tools, and calls the named tool of the default
version unless overridden by the -qt= option or QT_SELECT environment variable.
QT_SELECT is set by configure once the qt version has been chosen.
Take for example, moc.
$ which moc
/usr/bin/moc
$ ls -go /usr/bin/moc
lrwxrwxrwx 1 9 Jul 3 21:33 /usr/bin/moc -> qtchooser
$ qtchooser -print-env
QT_SELECT="default"
QTTOOLDIR="/usr/lib/x86_64-linux-gnu/qt4/bin"
QTLIBDIR="/usr/lib/x86_64-linux-gnu"
$ QT_SELECT=qt5 qtchooser -print-env
QT_SELECT="qt5"
QTTOOLDIR="/usr/lib/x86_64-linux-gnu/qt5/bin"
QTLIBDIR="/usr/lib/x86_64-linux-gnu"
$ moc -v
Qt Meta Object Compiler version 63 (Qt 4.8.4)
$ QT_SELECT=qt5 moc -v
Qt Meta Object Compiler version 67 (Qt 5.0.1)
This should be harmless elsewhere.
If BDB_CPPFLAGS returns only "-I", the next argument sent to the preprocessor
is treated as a path. There are 2 fixes here:
1. Check in CPPFLAGS, as a user might have manually passed a path to check.
2. Ensure the value is not empty before setting BDB_CPPFLAGS to "-I value"
This change moves test data into the binaries rather than reading them from
the disk at runtime.
Advantages:
- Tests become distributable
- Cross-compile friendly. Build on one machine and execute in an arbitrary
location on another.
- Easier testing for backports. Users can verify that tests pass without having
to track down corresponding test data.
- More trustworthy test results and easier quality assurance as tests make
fewer assumptions about their environment.
- Tests could theoretically run at client/daemon startup and exit on failure.
Disadvantages:
- Required 'hexdump' build-dependency. This is a standard bsd tool that should
be usable everywhere. It is likely already installed on all build-machines.
- Tests can no longer be fudged after build by altering test-data.
libleveldb.a and libmemenv.a should be able to build in parallel, but in
practice calling the leveldb makefile ends up rewriting build_config.mk. If
one target tries to build while the other is halfway through writing the
.mk, the make ends up in an undefined state.
Fix that by making one depend on the other. This also reorders the variables
to be passed by param rather than via the environment, and combines the targets
into a single rule to avoid needless duplication.
Cory Fields