8f7d1b39ef Fix QPainter non-determinism on macOS (Andrew Chow)
Pull request description:
Aplies a patch to Qt that fixes the non-determinism by modifying Qt. The source of the non-determinism is how LLVM 8 optimizes qt_intersect_spans when compiling. The particular optimization that seems to be causing the problems is that a temp variable is being added for spans->y. For some reason, when it does this, it chooses different instructions to use when making that variable. We bypass this problem by patching qt_intersect_spans to always make and use this local variable.
Potential alternative to #20436 and #20440
ACKs for top commit:
hebasto:
re-ACK 8f7d1b39ef ~for merging into the 0.21 branch, but [not into the master](https://github.com/bitcoin/bitcoin/pull/20454) branch.~
fanquake:
ACK 8f7d1b39ef
Tree-SHA512: b0d00a77643554021736524fb64611462ef2ec849a220543c12d99edb0f52f2e8128d2cc61fa82176b7e13b294574774a92d6b649badf8b7630c6d6a7e70ce10
Aplies a patch to Qt that fixes the non-determinism by modifying Qt. The
source of the non-determinism is how LLVM 8 optimizes qt_intersect_spans
when compiling. The particular optimization that seems to be causing the
problems is that a temp variable is being added for spans->y. For some
reason, when it does this, it chooses different instructions to use when
making that variable. We bypass this problem by patching
qt_intersect_spans to always make and use this local variable.
This has been around since the original import of Qt
(38be0d13830efd2d98281c645c3a60afe05ffece), however there
are now only two instatnces of it left in the qt codebase,
and from what I can gather, it's unused.
plugin_no_soname was removed from Qt some time ago, see upstream commit
1d034244c261520d5e739534dc264c2500e02b5f. It was replaced with
plugin_with_soname, however that is currently only used (as of 5.15.x)
in the Android Clang mkspec.
This should mostly be a no-op, however it would seem to make more sense
that we pass through the XCODE_VERSION we now have in depends, rather
than leaving the version set to 4.3.
This is an alternative to #19751 that fixes the build without requiring
splitting out libpng. This patch can be dropped once we are building qt
5.12.0 or later.
libevent uses getaddrinfo when available, and falls back to gethostbyname
Windows has both, but gethostbyname only supports IPv4
libevent fails to detect Windows's getaddrinfo due to not including the right headers
This patches libevent's configure script to check it correctly
This change adds the correct suffix to debug mode .pc filenames for
MinGW and also to the Qt libraries listed in the `Requires` field.
The filename adjustment fixes the accidental overwriting of release
mode .pc files with the debug mode variant which required the wrong
variant of the libraries when `debug_and_release` is active.
Note that macOS also supports the `debug_and_release' configuration
but may use the regular library names together with DYLD_IMAGE_SUFFIX.
Creation of *_debug.pc files is turned off as they're identical to their
non-debug counterparts.
More info:
- QTBUG-4155
- Qt commit a0d8fb4ac3cb7bafdb39f340055eacee4f957513
02ac445b2f bump libevent to 2.1.11 in depends (stefanwouldgo)
Pull request description:
this doesn't need patches on Android anymore like 2.1.8 did.
ACKs for top commit:
laanwj:
ACK 02ac445b2f
Tree-SHA512: 1fbfe342ee15fa4c5cb417979bd6c443f7c7aa40a489accf8ccd7c919e5b08e859b3da6edeee3de484f6f156b35dd4e97c7e2c7971b59fc31029865585ccb296
f9af3ced1c Android: add all arch support (Block Mechanic)
d419ca7e32 depends: export dynamic JNI symbols from static qtforandroid.a (Igor Cota)
ed30684d03 Qt: patch androidjnimain.cpp to make sure JNI is initialised when statically compiled (Igor Cota)
e4c319e8a1 builds: remove superfluous config_opts_aarch64_android (Igor Cota)
24ffef0c27 Patch libevent when building for Android (fix arc4random_addrandom) (Igor Cota)
f1e40b3e71 Update bitcoin_qt.m4 (BlockMechanic)
b4057d8261 Define TARGET_OS when host is android (Igor Cota)
80b475f159 Fix Android zlib cross compilation issue (https://stackoverflow.com/questions/21396988/zlib-build-not-configuring-properly-with-cross-compiler-ignores-ar) (Igor Cota)
45f8219015 Add full Android build example command and instructions on getting SDK/NDK (Igor Cota)
b68f2a68c2 Add config opts and patch for aarch64_android build of Qt (Igor Cota)
9c4cb0166e Add ranlib to android.mk hosts file (fix OSX Android NDK build) (Igor Cota)
c2a749c9c1 Add example Android host-platform-triplet and options (Igor Cota)
0b0cff3c61 Add support for building Android dependencies (Igor Cota)
Pull request description:
This allows one to build the dependencies with the Android SDK and goes towards fixing #11844. It has been tested to work with:
`make HOST=aarch64-linux-android ANDROID_API_LEVEL=28 ANDROID_TOOLCHAIN_BIN=/home/user/Android/Sdk/ndk-bundle/toolchains/llvm/prebuilt/linux-x86_64/bin NO_QT=1 NO_WALLET=1`
ACKs for top commit:
Sjors:
ACK f9af3ce. I'm OK with merging and then improving later.
Tree-SHA512: cb805115ebe5c9e33db2bf3eab8628808fe3f50052053d8877d8b8e4406d6fea1ed9e5c4dff85d777fb99c81be6ffb9d95a0e6d32344e728e5e0da6c653e2ce7
463a1d5244 Refresh ZeroMQ 4.3.1 patch (Nathan Marley)
Pull request description:
Currently in Alpine Linux (latest, 3.10) in the depends system, one of the ZeroMQ patches won't apply cleanly because the context around the patch has changed and Alpine's `patch` implementation can't handle the diff.
Some patch implementations can't handle fuzz / too much divergence from the original code.
This PR just tweaks the context code around the patch so that less-sophisticated patch implementations (such as on Alpine Linux) can apply the patch without errors.
This partially fixes#16925
ACKs for top commit:
fanquake:
ACK 463a1d5244 - Tested building zeromq in depends inside an [Alpine container](https://github.com/fanquake/core-review/blob/master/docker/alpine.dockerfile) as well as on macOS.
Tree-SHA512: d6e3cb60835cdd090b9b864ca9cb33961687606bc9184fbbeb7a54ec23db4057b9317b65c5c276fb8c5492cb3cfcc4a7f3369f049551f4eb0915db971f2290ce
Some patch implementations can't handle fuzz / too much divergence from the
original code.
This just tweaks the context code around the patch so that less-sophisticated
patch implementations (such as on Alpine Linux) can apply the patch without
errors.
We're no longer building QT with libX11/XLib, however, libX11/XLib
headers are still required for parts of QT. In this commit we add a
minimal configure.ac for libX11/XLib that is headers-only.
This change allows us to remove all of libX11/XLib's dependencies.
Qt's configure grabs the path to xkb's data root during configure, but the
build changes in 5.8 apparently broke the handling for cross builds. As a
result, the string embedded in the binary depends on whether or not some files
are present in the builder's filesystem.
The "-xkb-config-root" configure setting is intended to allow manual overriding
but it is also broken. See: https://bugreports.qt.io/browse/QTBUG-60005
This has since been fixed upstream, so just hard-code the path for now. We can
drop this patch when we bump to a fixed Qt.
Also, fix the "-qt-xkbcommon-x11" config param which was renamed. This does not
appear to affect build results, presumably because auto-detection is working,
but it does not hurt to be explicit.
The patch Bitcoin Core has been maintaining for mac_alias was pulled by the mac_alias maintainer in commit 4f31cb084c1c6a8626128b0b00842020b6db9037. Delete the patch and remove the patch from the depends system.
Note that this PR won't be complete until a new version of mac_alias containing the path has been released, and the depends system is updated to reflect the new version.
fanquake