de4238f92f4c067f099663f68d9772105de81d75 build: consolidate reduced export checks (fanquake)
012bdec1b7df01906566a6526e56f27d57d1653b build: add building libconsensus to end-of-configure output (fanquake)
8f360e349e365870b40a6873917c81de714ae41a build: remove ax_gcc_func_attribute macro (fanquake)
f054a089ecfbdc4732e6f705a10e93189074f41c build: remove AX_GCC_FUNC_ATTRIBUTE test for dllimport (fanquake)
7cd0a696643a824ab6f6911278f116f01c5af662 build: test for __declspec(dllexport) in configure (fanquake)
1624e17b5430dfe808bb3b1b79dfa53bf45aa053 build: remove duplicate visibility attribute detection (fanquake)
Pull request description:
Darwin targets do not have a `protected` visibility function attribute, see [LLVM explanation](8e9a505139/clang/lib/Basic/Targets/OSTargets.h (L131)). This means that the `AX_GCC_FUNC_ATTRIBUTE` check for `visibility` fails:
```bash
configure:24513: checking for __attribute__((visibility))
configure:24537: g++ -std=c++11 -o conftest -g -O2 -DHAVE_BUILD_INFO -D__STDC_FORMAT_MACROS -DMAC_OSX -DOBJC_OLD_DISPATCH_PROTOTYPES=0 -Wl,-headerpad_max_install_names conftest.cpp >&5
conftest.cpp:35:56: warning: target does not support 'protected' visibility; using 'default' [-Wunsupported-visibility]
int foo_pro( void ) __attribute__((visibility("protected")));
^
1 warning generated.
configure:24537: $? = 0
configure:24550: result: no
```
This leads to `EXPORT_SYMBOL` being [defined to nothing](f4de89edfa/src/script/bitcoinconsensus.h (L29)), as `HAVE_FUNC_ATTRIBUTE_VISIBILITY` is not defined, and when building with reduced exports, you end up with a libbitcoinconsensus.dylib that doesn't export any `_bitcoinconsensus_*` symbols.
```bash
➜ git:(master) nm -C src/.libs/libbitcoinconsensus.dylib | rg _bitcoinconsensus_
➜ git:(master)
```
We do have a [second check](f4de89edfa/configure.ac (L882)) for the `visibility` attribute, which works for Darwin as it's only testing for default visibility, however the result of this check isn't used at all. It was added in #4725, along with the `--enable-reduce-exports` option, however when libbitcoinconsensus was added in #5235, it used the results of the added `AX_GCC_FUNC_ATTRIBUTE` calls.
This PR removes our usage of the AX_GCC_FUNC_ATTRIBUTE macro entirely, in favour of our own checks in configure. This meant adding a check for `dllexport`, which I've tested as working with both [GCC](https://gcc.gnu.org/onlinedocs/gcc/Microsoft-Windows-Function-Attributes.html) and [Clang](https://releases.llvm.org/10.0.0/tools/clang/docs/AttributeReference.html#dllexport) when building for Windows. I haven't added an equivalent check for `dllimport`, as we weren't actually using the result of that check, we're just testing that `MSC_VER` was defined before using.
With these changes building a libbitcoinconsensus with reduced exports, when targeting Darwin, works as expected:
```bash
./autogen.sh
./configure --disable-tests --disable-bench --with-utils=no --with-daemon=no --with-gui=no --disable-wallet --with-libs=yes --enable-reduce-exports
make -j8
...
nm -C src/.libs/libbitcoinconsensus.dylib | rg _bitcoinconsensus_
000000000000a340 T _bitcoinconsensus_verify_script
00000000000097e0 T _bitcoinconsensus_verify_script_with_amount
000000000000a3c0 T _bitcoinconsensus_version
```
```python
>>> import ctypes
>>> consensus = ctypes.CDLL("src/.libs/libbitcoinconsensus.dylib")
>>> print(consensus.bitcoinconsensus_version())
1
>>> exit()
```
TODO: Modify a CI job to compile with --enable-reduce-exports and check for symbols in shared lib?
ACKs for top commit:
laanwj:
Code review ACK de4238f92f4c067f099663f68d9772105de81d75
Tree-SHA512: d148f3c55d14dac6e9e5b718cc65bb557bcf6f663218d24bc9044b86281bd5dd3d931ebea79c336a58e8ed50d683218c0a9e75494f2267b91097665043e252ae
060a2a64d40d75fecb60b7d2b9946a67e46aa6fc ci: remove boost thread installation (fanquake)
06e1d7d81d5a56d136c6fc88f09a2b0654a164f9 build: don't build or use Boost Thread (fanquake)
7097add83c8596f81be9edd66971ffd2486357eb refactor: replace Boost shared_mutex with std shared_mutex in sigcache (fanquake)
8e55981ef834490c438436719f95cbaf888c4914 refactor: replace Boost shared_mutex with std shared_mutex in cuckoocache tests (fanquake)
Pull request description:
This replaces `boost::shared_mutex` and `boost::unique_lock` with [`std::shared_mutex`](https://en.cppreference.com/w/cpp/thread/shared_mutex) & [`std::unique_lock`](https://en.cppreference.com/w/cpp/thread/unique_lock).
Even though [some concerns were raised](https://github.com/bitcoin/bitcoin/issues/16684#issuecomment-726214696) in #16684 with regard to `std::shared_mutex` being unsafe to use across some glibc versions, I still think this change is an improvement. As I mentioned in #21022, I also think trying to restrict standard library feature usage based on bugs in glibc is not only hard to do, but it's not currently clear exactly how we do that in practice (does it also extend to patching out use in our dependencies, should we be implementing more runtime checks for features we are using, when do we consider an affected glibc "old enough" not to worry about? etc). If you take a look through the [glibc bug tracker](https://sourceware.org/bugzilla/describecomponents.cgi?product=glibc) you'll no doubt find plenty of (active) bug reports for standard library code we already using. Obviously not to say we shouldn't try and avoid buggy code where possible.
Two other points:
[Cory mentioned in #21022](https://github.com/bitcoin/bitcoin/pull/21022#issuecomment-769274179):
> It also seems reasonable to me to worry that boost hits the same underlying glibc bug, and we've just not happened to trigger the right conditions yet.
Moving away from Boost to the standard library also removes the potential for differences related to Boosts configuration. Boost has multiple versions of `shared_mutex`, and what you end up using, and what it's backed by depends on:
* The version of Boost.
* The platform you're building for.
* Which version of `BOOST_THREAD_VERSION` is defined: (2,3,4 or 5) default=2. (see [here](https://www.boost.org/doc/libs/1_70_0/doc/html/thread/build.html#thread.build.configuration) for some of the differences).
* Is `BOOST_THREAD_V2_SHARED_MUTEX` defined? (not by default). If so, you might get the ["less performant, but more robust"](https://github.com/boostorg/thread/issues/230#issuecomment-475937761) version of `shared_mutex`.
A lot of these factors are eliminated by our use of depends, but users will have varying configurations. It's also not inconceivable to think that a distro, or some package manager might start defining something like `BOOST_THREAD_VERSION=3`. Boost tried to change the default from 2 to 3 at one point.
With this change, we no longer use Boost Thread, so this PR also removes it from depends, the build system, CI etc.
Previous similar PRs were #19183 & #20922. The authors are included in the commits here.
Also related to #21022 - pthread sanity checking.
ACKs for top commit:
laanwj:
Code review ACK 060a2a64d40d75fecb60b7d2b9946a67e46aa6fc
vasild:
ACK 060a2a64d40d75fecb60b7d2b9946a67e46aa6fc
Tree-SHA512: 572d14d8c9de20bc434511f20d3f431836393ff915b2fe9de5a47a02dca76805ad5c3fc4cceecb4cd43f3ba939a0508178c4e60e62abdbaaa6b3e8db20b75b03
d796091b04f3b02d2280aaa761c2b94950199da8 build: Bump AX_PTHREAD macro to the latest version (Hennadii Stepanov)
Pull request description:
This PR silents autoconf >2.69 (this [one](https://formulae.brew.sh/formula/autoconf), for instance) warnings about the obsolete `$as_echo`:
```
% ./autogen.sh
...
configure.ac:847: warning: $as_echo is obsolete; use AS_ECHO(["message"]) instead
lib/m4sugar/m4sh.m4:692: _AS_IF_ELSE is expanded from...
lib/m4sugar/m4sh.m4:699: AS_IF is expanded from...
./lib/autoconf/general.m4:2249: AC_CACHE_VAL is expanded from...
./lib/autoconf/general.m4:2270: AC_CACHE_CHECK is expanded from...
build-aux/m4/ax_pthread.m4:89: AX_PTHREAD is expanded from...
configure.ac:847: the top level
...
```
No other behavior changes.
ACKs for top commit:
fanquake:
ACK d796091b04f3b02d2280aaa761c2b94950199da8 - matches upstream at serial 31.
Tree-SHA512: aa9b60698f453427221444a5a63420d833c4c5dd23f8b0c74e5bd4639daec9c6cff0907a5281c00103ccb030e394998cf05653be750d4a3bf0f37ca41ff6fbe1
747cd17404832604c50d03d58e11ba816bb229f7 build: no-longer fail default configure if BDB isn't available (fanquake)
Pull request description:
Inline with moving to descriptor (sqlite) wallets by default for 0.23,
this adapts the build system so that a default `./configure` invocation
no-longer fails if BDB isn't present. Currently, if configure is run
with no options, and no BDB is present, we'll fail with:
```bash
checking for Berkeley DB C++ headers... no
configure: error: libdb_cxx headers missing, Bitcoin Core requires this library for BDB wallet support (--without-bdb to disable BDB wallet support)
```
If descriptor wallets are to be the default, this behaviour no longer
makes sense, as a builder should be able to configure and build, to use
a wallet, without BDB installed, and without passing additional
arguments, i.e `--without-bdb` or `--with-incompatible-bdb`, to
configure.
With this change, running configure will no-longer fail, but will
instead print:
```bash
checking for Berkeley DB C++ headers... no
configure: WARNING: libdb_cxx headers missing
configure: WARNING: Bitcoin Core requires this library for BDB (legacy) wallet support
configure: WARNING: Passing --without-bdb will suppress this warning
checking for sqlite3 >= 3.7.17... yes
checking whether to build wallet with support for sqlite... yes
```
ACKs for top commit:
hebasto:
ACK 747cd17404832604c50d03d58e11ba816bb229f7, tested on Linux Mint 20.2 (x86_64) with the (un)installed system packages `libdb-dev` and `libdb++-dev`.
Tree-SHA512: ae316d71ad0803c9d4b02a5fedcade08242650d987cc047840493ba4a881e71ff48b099075bb7c325307d44744fcdeccb57f7fa8db4135c81a5835841f562afa
4783115fd4cccb46a7f8c592b34fa7c094c29410 net: add ifaddrs.h include (fanquake)
879215e665a9f348c8d3fa92701c34065bc86a69 build: check if -lsocket is required with *ifaddrs (fanquake)
87deac66aa747481e6f34fc80599e1e490de3ea0 rand: only try and use freeifaddrs if available (fanquake)
Pull request description:
Fixes#21485 by linking against `-lsocket` when it's required for using `*ifaddrs` functions.
ACKs for top commit:
laanwj:
Code review ACK 4783115fd4cccb46a7f8c592b34fa7c094c29410
hebasto:
ACK 4783115fd4cccb46a7f8c592b34fa7c094c29410, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: 4542e036e9b029de970eff8a9230fe45d9204bb22313d075f474295d49bdaf1f1cbb36c0c6e2fa8dbbcdba518d8d3a68a6116ce304b82414315f333baf9af0e4
e251726affe97da745362c82567c2377ceb07d21 build, qt: Fix typo in QtInputSupport check (Hennadii Stepanov)
Pull request description:
This typo slipped in fecb3723b63d1441f1f786352e9ef330c1ac57a5 (#21565).
I apologize for that as a reviewer of #21565.
ACKs for top commit:
fanquake:
ACK e251726affe97da745362c82567c2377ceb07d21
Tree-SHA512: c14d603ed5a9b7e4b6503bfb93e66b800d5a107f21a8375c6f98465ef937e3e55b9c4fc160407c7e3ee402b495c5820ebb35ec6cb92956c1ee4092415635b915
a8bd5ea01720e5639cabdc9897cf9cf7c02c47c6 build, qt: Fix libraries linking order for Linux hosts (Hennadii Stepanov)
Pull request description:
`-lxcb-shm` should follow `-lxcb` when linking libraries for Linux hosts.
Fixes#22105.
ACKs for top commit:
prayank23:
ACK a8bd5ea017
fanquake:
ACK a8bd5ea01720e5639cabdc9897cf9cf7c02c47c6
Tree-SHA512: 6ea2a1ba5440dd01e64428e5c3aa3e7a6e1380b723071ad38a0516b23b56dc252cb2a6544e014b13a0b2f89b3fe45c04a82e498669b5cc74a5f9b70d9c270926
6aab7649d30b19d136a27f1287fd2c8b00fb460c doc: Fix whitespace errs in .md files, bitcoin.conf, Info.plist.in, and find_bdb48.m4 (Jon Layton)
Pull request description:
Although there is an existing `test/lint/lint-whitespace.sh` linter, it only prevents new errors from being introduced. This commit removes all existing whitespace errors from Core markdown files (skips `src/crypto/ctaes/`, `leveldb/`, and `doc/release-notes/`), `bitcoin.conf`, and `Info.plist.in`.
Further formatting could be done on the markdown documents, but seeing as there several coexisting styles that break a few `markdownlint` rules, a first step would be to define and add a linter to Travis. For now, the small fix is made.
ACKs for top commit:
fanquake:
ACK 6aab7649d30b19d136a27f1287fd2c8b00fb460c - Thanks for following up. Hopefully we now never have to deal with whitespace again.
Tree-SHA512: 810cc31ae4364b2dedf85783e67315d7b4e11589e4b32c599606e1b1ba8de0663bcae9ddb1bd8c9762a3636a2d65bdcd64ec22d2e90943f374a0c9574b77ca23
ffdd7de690d8dbe813de0108eb2a7997a8027773 build, qt: Fix regression introduced in #21363 (Hennadii Stepanov)
Pull request description:
After #21363 the `_BITCOIN_QT_CHECK_STATIC_PLUGIN` never fails due to the ill-formed code.
Sorry for breaking it.
ACKs for top commit:
fanquake:
ACK ffdd7de690d8dbe813de0108eb2a7997a8027773
Tree-SHA512: 0af3334af5332f3b349dba219565ee336090541c2c4fae53d7dc4dd2aa05291b3c7c13bfc506219f642fb854f3c887f0f9621b42d0df68d5d794e0563b2cf5d4
366913e307d2dc13bc00d6bf7b6b2426c359ac30 build: AX_BOOST_THREAD serial 33 (Igor Cota)
cf0681133ae7301ead7091eaee55c945da5cdfcc build: disable D-Bus on Android by default (Igor Cota)
Pull request description:
I've been trying to build for Android on different OSes/Gitian with varying success. Build system is quite the beast and sometimes it doesn't get it right. To make sure it does these three little tweaks make the Android build more robust:
- disable D-Bus (Android doesn't support it and has its own way to trigger notifications)
- don't flag `-lpthread` when linking Boost, [Bionic has built-in support](https://stackoverflow.com/questions/30801752/android-ndk-and-pthread)
- ~~add `-static-libstdc++` to linker flags. This avoids having to bundle `libc++_shared` with CLI apps, still necessary with `bitcoin-qt` though (thanks Sjors)~~
I think these are small and fairly straightforward so I put them all into this one PR.
ACKs for top commit:
fanquake:
ACK 366913e307d2dc13bc00d6bf7b6b2426c359ac30
Tree-SHA512: 31465fd228a5877c20aa2a05f98242d4eeb328b9b35bd1a7a3dcfb1ef51379d84053a81ade5a65436ffc1bc8ccd21f11ed0539eb10e827d182c0c04394629af0
9a0969585fce03b45be7004bba865bc15909904c build: Add /opt/homebrew to path to look for boost libraries (Fu Yong Quah)
Pull request description:
Following the instruction in https://github.com/bitcoin/bitcoin/blob/master/doc/build-osx.md for building on the M1 Macs don't work out of the box, because homebrew now defaults to /opt/homebrew instead of /usr/local. This PR fixes that.
ACKs for top commit:
jonasschnelli:
utACK 9a0969585fce03b45be7004bba865bc15909904c
promag:
Tested ACK 9a0969585fce03b45be7004bba865bc15909904c.
Tree-SHA512: 472568b97fbd8623481fe6fd43b0509fa32fe7f1c1d8090321a6a6a5bdc7343d4ad4122c10dcc7c9c93068db8a3f009a73befaf1ba11e4af54a66afd2c2dbe14
da0842dcd44f8c9c9b167917fac0949b4978c3b0 build: Update ax_boost_mase.m4 to the latest serial (Hennadii Stepanov)
Pull request description:
Picked from the upstream 90814f1895Fix#17010.
This PR is [alternative](https://github.com/bitcoin/bitcoin/issues/17010#issuecomment-610651736) to #18501.
ACKs for top commit:
laanwj:
ACK da0842dcd44f8c9c9b167917fac0949b4978c3b0
Tree-SHA512: 5e43e12c524e4ea6b967c9be02c81a75948eac6cf55b819e3339222a2e3414731581d40af3524ad865abae7c5247c190448ebf2aa5e0d9a338edb501cc23ba38
bb99c4e684bbd3053ecf7a789049b11b29260189 build: update boost macros to latest upstream (fanquake)
Pull request description:
Fixes: #16803
I opened an [upstream PR](https://github.com/autoconf-archive/autoconf-archive/pull/197) to improve the Boost error reporting, so pull the latest macros.
ACKs for top commit:
laanwj:
Code review ACK bb99c4e684bbd3053ecf7a789049b11b29260189
jonatack:
Sanity check ACK bb99c4e684bbd3053ecf7a789049b11b29260189, light code read, built and ran tests on Debian 4.19.37-5+deb10u2 (2019-08-08) x86_64 GNU/Linux. Only tested the happy path.
Tree-SHA512: 34704ed623ac0085215fd874a23fde8f6e39a69fa20d78472b0c4d2306dc101c0571fa26c4c8821600746b94daaaf05faf6d15546899d588081c26357d29ec46
f447a0a7079619f0d650084df192781cca9fd826 Remove program options from build system (Chun Kuan Lee)
11588c639e8912f1b28e981c1a2a0e4306dbd093 Replace boost program_options (Chun Kuan Lee)
Pull request description:
Concept from #12744, but without parsing negated options.
Tree-SHA512: 7f418744bb8934e313d77a5f162633746ef5d043de802b9c9cd9f7c1842e7e566eb5f171cd9e2cc13317281b2449c6fbd553fa4f09b837e6af2f5d2b2aabdca2
4af4672525f698c7b491023fcd36a17a4e982070 build, qt: Add Qt version checking (Hennadii Stepanov)
30e336f785e1b662cade3cfacea91a9785ea1023 build: Drop unused bitcoin_cv_qt58 (Hennadii Stepanov)
Pull request description:
Now `configure` script checks that Qt version is not less then minimum required (currently [5.5.1](https://github.com/bitcoin/bitcoin/pull/15393)).
This PR is an alternative to #15706 (see https://github.com/bitcoin/bitcoin/pull/15706#issuecomment-629076962).
Closes#15688.
The first commit removes dead code (see https://github.com/bitcoin/bitcoin/pull/18297#issuecomment-603252662).
ACKs for top commit:
fanquake:
ACK 4af4672525f698c7b491023fcd36a17a4e982070 - this looks ok. I've tested this with Qt 5.15.0 and Qt 5.7.1 system libs, as well as 5.9.8 from depends.
Tree-SHA512: 8e3b82fa3a98926814923331038185633fabad962c271f31bd158e1ab293dcde52ab1dbf997745540a9ed27e16835cf5b5f3701d405876d877fa561eb03cc619
8a26848c460160e1279f26bc413f693a34e33b9d build: Fix m4 escaping (Hennadii Stepanov)
9123ec15db104397998f5084afc69403d2f9e4b8 build: Remove extra tokens warning (Hennadii Stepanov)
fded4f48c33742d7c790335c8de59c15b80d94e6 build: Remove duplicated QT_STATICPLUGIN define (Hennadii Stepanov)
05a93d5d96101b45d87571af5b772c7a1e82fd27 build: Fix indentation in bitcoin_qt.m4 (Hennadii Stepanov)
ddbb41931019ed4226af3df37874c7eb7cf570f1 build: Use pkg-config in BITCOIN_QT_CONFIGURE for all hosts (Hennadii Stepanov)
492971de35bab26346545f68365872211f458b00 build: Fix mingw pkgconfig file and dependency naming (Hennadii Stepanov)
Pull request description:
This PR makes `bitcoin_qt.m4` to use `pkg-config` for all hosts and removes non-pkg-config paths from it. This is a step towards the idea which was clear [stated](https://github.com/bitcoin/bitcoin/pull/8314#issue-76644643) by Cory Fields:
> I believe the consensus is to treat Windows like the others and require pkg-config across the board. We can drop all of the non-pkg-config paths, and simply AC_REQUIRE(PKG_PROG_PKG_CONFIG)
There are two unsolved problems with this PR. If depends is built with `DEBUG=1` the `configure` script fails to pickup Qt:
- for macOS host (similar to, but not the same as #16391)
- for Windows host (regression)
The fix is ~on its way~ submitted in #18298 (as a followup).
Also this PR picks some small improvements from #17820.
ACKs for top commit:
theuni:
Code review ACK 8a26848c460160e1279f26bc413f693a34e33b9d
dongcarl:
Code Review ACK 8a26848c460160e1279f26bc413f693a34e33b9d
laanwj:
Code review ACK 8a26848c460160e1279f26bc413f693a34e33b9d
Tree-SHA512: 3b25990934b939121983df7707997b31d61063b1207d909f539d69494c7cb85212f353092956d09ecffebb9fef28b869914dd1216a596d102fcb9744bb5487f7
2620e24b83d16bf0f2bfe360dee1e98b4be59ca5 [depends] boost: update to 1.70 (Sjors Provoost)
Pull request description:
Version [1.70](https://www.boost.org/users/history/version_1_70_0.html) is most recent.
Versions needed for:
* 1.66: #12557: fixes the single arm64 configuration ([06ee5b5](06ee5b54ef))
ACKs for commit 2620e2:
Tree-SHA512: 6e0174f1d92c2c24314c0689d4809e048914f8f42d17aa73799f5ee232169e0dd0ed71f5f973903c44c08309f2837c629c493f15e5c31ec6c7bd1daae5f3b25f
Ugrade Qt depends to Qt5.9.4
Depends can now be built with Qt5.9.4 , which is Qt's new long term
support version.
Fix depends Qt5.9.4 mac build
Apply patch from QTBUG-67286
Upgrade Qt depends to 5.9.6
8ae413235 Remove redundant checks for MSG_* from configure.ac (Vasil Dimov)
71129e026 Do not check for main() in libminiupnpc (Vasil Dimov)
8c632f73c ax_boost_{chrono,unit_test_framework}.m4: take changes from upstream (Vasil Dimov)
Pull request description:
Tree-SHA512: a99ef98c0b94f892eadeda24b3d55c25bedf225b98c6e4178cf6c2d886b44d43e9f75414d0b37db9ac261cec2350666e5e64fab9c104249dd34ff485c51663cb
