c23f6f84ef Add depends qt fix for ARM macs (Jonas Schnelli)
Pull request description:
With this, depends builds fine on macOS 11 on an Apple Silicon Mac (ARM64).
ACKs for top commit:
laanwj:
Code review ACK c23f6f84ef
Tree-SHA512: a8354cec99969cff9e7dab150c335050ddb4b3c93a9f12a4db5e8046f02b11ce692ac17c2b96cbbe7f380c1aa110b15b8d6d48d51bc9c560282c702e99fd8a8d
c82d15b6d1 depends: Do not force Precompiled Headers (PCH) for building Qt on Linux (Hennadii Stepanov)
Pull request description:
On CentOS 8 (Cirrus CI job) the forced `-pch` option breaks Qt build.
Removing `-pch` option does not affect build time for other systems:
- master (e2ff5e7b35):
```
$ time make -j 9 -C depends/ qt
...
Caching qt...
make: Leaving directory '/home/hebasto/guix/GitHub/bitcoin/depends'
real 4m22,359s
user 18m3,719s
sys 1m24,769s
```
- this PR:
```
$ time make -j 9 -C depends/ qt
...
Caching qt...
make: Leaving directory '/home/hebasto/guix/GitHub/bitcoin/depends'
real 4m14,862s
user 18m3,355s
sys 1m24,506s
```
Qt docs: https://doc.qt.io/qt-5/qmake-precompiledheaders.htmlFixes#20423
ACKs for top commit:
MarcoFalke:
review ACK c82d15b6d1
Tree-SHA512: 0f2a3712e90de881d00f8e56c363edde33dd4f5c117df5744ab4e51d0a8146331de7236bc8329d68ddd91535cd853e68ee80ef4cceb6a909786abfd8881b01e8
2f5dfe4a7f depends: build qt in c++17 mode (fanquake)
104e859c97 builds: don't pass -silent to qt when building in debug mode (fanquake)
e2c500636c depends: build zeromq with -std=c++17 (fanquake)
2374f2fbef depends: build Boost with -std=c++17 (fanquake)
2dde55702d depends: build bdb with -std=c++17 (fanquake)
Pull request description:
In packages where we are passing `-std=c++11` switch to `-std=c++17`, or, `-std=c++1z` in the case of Qt.
This PR also contains a [commit](104e859c97) that improves debug output when building Qt for debugging (`DEBUG=1`).
Now we'll get output like this:
```bash
g++ -c -pipe -ffunction-sections -O2 -fPIC -std=c++11 -fno-exceptions <lots more> ../../corelib/kernel/qcoreapplication.cpp
```
rather than just:
```bash
compiling ../../corelib/kernel/qcoreapplication.cpp
```
Note that when you look at the DEBUG output for these changes when building Qt, you'll see objects being compiled with a mix of C++11 and C++17. The breakdown is roughly:
1. `qmake` built with `-std=c++11`:
```bash
Creating qmake...
make[1]: Entering directory '<trim>/qt/5.9.8-4110fa99945/qtbase/qmake'
g++ -c -o project.o -std=c++11 -ffunction-sections -O2 -g <trim> <trim>/qt/5.9.8-4110fa99945/qtbase/qmake/project.cpp
# when qmake, Qt also builds some of it's corelib, such as corelib/global/qmalloc.cpp
g++ -c -o qmalloc.o -std=c++11 -ffunction-sections -O2 -g <trim> <trim>/qt/5.9.8-4110fa99945/qtbase/src/corelib/global/qmalloc.cpp
```
2. `qmake` is run, and passed our build options, including `-c++std`:
```bash
make[1]: Entering directory '<trim>/qt/5.9.8-4110fa99945/qtbase'
<trim>qt/5.9.8-4110fa99945/qtbase/bin/qmake -o Makefile qtbase.pro -- -bindir <trim>/native/bin -c++std c++1z -confirm-license <trim>
```
3. After some cleaning and configuring, we actually start to build Qt, as well as it's tools and internal libs:
```bash
Building qt...
make[1]: Entering directory '<trim>/qt/5.9.8-4110fa99945/qtbase/src'
# build libpng, zlib etc
gcc -c -m64 -pipe -pipe -O1 <trim> -o .obj/png.o png.c
# build libQt5Bootstrap, using C++11, which again compiles qmalloc.cpp
make[2]: Entering directory '<trim>/qt/5.9.8-4110fa99945/qtbase/src/tools/bootstrap'
g++ -c -pipe -ffunction-sections -O2 -fPIC -std=c++11 <trim> -o .obj/qmalloc.o ../../corelib/global/qmalloc.cpp
# build a bunch of tools like moc, rcc, uic, qfloat16-tables, qdbuscpp2xml, using C++11
g++ -c -pipe -O2 -std=c++11 -fno-exceptions -Wall -W <trim> -o .obj/rcc.o rcc.cpp
# from here, Qt is compiled with -std=c++1z, including qmalloc.cpp, for the third and final time:
g++ -c -include .pch/Qt5Core <trim> -g -Og -fPIC -std=c++1z -fvisibility=hidden <trim> -o .obj/qmalloc.o global/qmalloc.cpp
```
4. Finally, build tools like `lrelease`, `lupdate`, etc, but back to using -std=c++11
```bash
make[1]: Entering directory '<trim>/qt/5.9.8-4110fa99945/qttools/src/linguist/lrelease'
g++ -c -pipe -O2 -std=c++11 -fno-exceptions -Wall -W <trim> -o .obj/translator.o ../shared/translator.cpp
```
If you dump the debug info from the built Qt libs, they should also tell you that they were compiled with `C++17`:
```bash
objdump -g bitcoin/depends/x86_64-pc-linux-gnu/lib/libQt5Core.a
GNU C++17 9.3.0 -m64 -mtune=generic -march=x86-64 -g -O1 -Og -std=c++17 -fPIC -fvisibility=hidden -fvisibility-inlines-hidden -fasynchronous-unwind-tables -fstack-protector-strong -fstack-clash-protection -fcf-protection
```
ACKs for top commit:
laanwj:
Code review ACK 2f5dfe4a7f
practicalswift:
cr ACK 2f5dfe4a7f: patch looks correct
fjahr:
Code review ACK 2f5dfe4a7f
hebasto:
ACK 2f5dfe4a7f, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: fc5e9d7c7518c68349c8228fb1aead829850373efc960c9b8c079096a83d1dad19c62a9730fce5802322bf07e320960fd47851420d429eda0a87c307f4e8b03a
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
This means we'll get build output like this when building with DEBUG=1:
g++ -c -pipe -ffunction-sections -O2 -fPIC -std=c++11 -fno-exceptions <lots more> ../../corelib/kernel/qcoreapplication.cpp
rather than just:
compiling ../../corelib/kernel/qcoreapplication.cpp
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.
a52ecc936a build: set minimum supported macOS to 10.14 (fanquake)
Pull request description:
This is a requirement for C++17 support. See my comments [here](https://github.com/bitcoin/bitcoin/issues/16684#issuecomment-643722538):
> You cannot use std::get with std::variant on macOS < 10.14, because Apples libc++ doesn't support the std::bad_variant_access exception. [Relevant comment](https://github.com/bitcoin/bitcoin/pull/19183#discussion_r439794318) in #19183.
> While we could work around this in our own code, using std::get_if, this would still be a problem for 3rd-party dependencies.
> I've been testing Qt 5.15LTS (we'll have to enable C++17 in qt, and may upgrade to a newer version at the same time), and you can't enable -std c++17, while targeting a macOS deployment version < 10.14, configuring will fail. They are making use of std::get with std::variant throughout their cocoa code.
We would have to had to have bumped to at least 10.13 in any case, as Qt 5.15 (#19716) [requires 10.13+](https://doc.qt.io/qt-5/supported-platforms.html).
ACKs for top commit:
hebasto:
ACK a52ecc936a, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: f669b2fc777aeea1e9afdbbc7bd9afe3997418211db6ba53c934cae0e62a9b999603da539518c229f34961d275c9e2f315c7b022cf5fb97bd201a69c85d470cc
e1f2553e11 build: remove global_init_link_order from mac qt qmake.conf (fanquake)
498fa16bea build: document preprocessing steps in qt package (fanquake)
bd5d9336d9 build: don't copy Info.plist.* into mkspec for macOS qt build (fanquake)
bfd7e33b4b build: remove plugin_no_soname from mac qt qmake.conf (fanquake)
fdde4c7ce6 build: pass XCODE_VERSION through to qt macOS cross compile conf (fanquake)
49473ef211 build: convert "echo" usage into a patch in qt package (fanquake)
Pull request description:
Follow up on removing `sed` usage in #19761. Also nice to revisit & cleanup before 5.15.x.
ACKs for top commit:
laanwj:
Code review ACK e1f2553e11
Tree-SHA512: 4e6489d877aaa300f69e091d7117136da49611bd80afd45adfbd7ddeb5b3c9c76fb0f87a3249cbe63ba93129df56281fd4a9389daadc852211325c5ca9ac6567
7087440894 depends: native_ds_store 1.3.0 (fanquake)
Pull request description:
`ds_store` [now takes advantage](36fb607940) of Pythons ability to decode binary [plists](https://docs.python.org/3/library/plistlib.html) (since 3.4), so we can drop its biplist dependency.
The call to `biplist.Data()` in `custom_dsstore.py` doesn't seem to do anything, and from what I can tell can just be removed. i.e:
```diff
diff --git a/contrib/macdeploy/custom_dsstore.py b/contrib/macdeploy/custom_dsstore.py
index dc1c1882d..e475bc6c3 100755
--- a/contrib/macdeploy/custom_dsstore.py
+++ b/contrib/macdeploy/custom_dsstore.py
@@ -47,6 +47,7 @@ alias.volume.disk_image_alias.target.filename = package_name_ns + '.temp.dmg'
alias.volume.disk_image_alias.target.carbon_path = 'Macintosh HD:Users:\x00bitcoinuser:\x00Documents:\x00bitcoin:\x00bitcoin:\x00' + package_name_ns + '.temp.dmg'
alias.volume.disk_image_alias.target.posix_path = 'Users/bitcoinuser/Documents/bitcoin/bitcoin/' + package_name_ns + '.temp.dmg'
alias.target.carbon_path = package_name_ns + ':.background:\x00background.tiff'
+assert(biplist.Data(alias.to_bytes()) == alias.to_bytes())
icvp['backgroundImageAlias'] = biplist.Data(alias.to_bytes())
ds['.']['icvp'] = icvp
```
ACKs for top commit:
laanwj:
ACK 7087440894
Tree-SHA512: 8ba3cf561937efe4a3daae8b0cb4de3bf9e425b3a9244161b09d94ee2b1bd4c3e21315fa70e495b19a052aabdc1731b3b6f346b63272d72d2762ced83237d02f
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.
We generate our own Info.plist as part of make deploy, and as far as I
can tell, it doesn't seem to have an effect wether these are present
during qt's build.
I also can't find a single mention of the .app plist in the qt code,
whereas there are multiple instances of .lib.
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.
native_ds_store now takes advantage of Pythons ability to decode binary
plists (since 3.4), so we can drop its biplist dependency.
The call to biplist.Data() in custom_dsstore doesn't seem to do anything,
and from what I can tell can just be removed.
Previously, when running ./configure:
1. With CONFIG_SITE pointed to our depends config.site.in, and
2. PYTHONPATH was not set either in the environment or by the user
The configure would output something like:
PYTHONPATH='depends/x86_64-pc-linux-gnu/share/../native/lib/python3/dist-packages:'
When we really mean:
PYTHONPATH='depends/x86_64-pc-linux-gnu/share/../native/lib/python3/dist-packages'
...without the colon
This change makes sure that:
1. There's no trailing colon, and
2. We use the $PATH_SEPARATOR variable instead of a colon
Files like config.site.in are not referenced by any other script in our
tree, so we need to mark it manually with a "shellcheck shell="
directive and make sure that shellcheck is run on them.
Previously, if ./configure was invoked with:
```
$ env CONFIG_SITE=depends/x86_64-pc-linux-gnu/share/config.site ./configure
```
Where $CONFIG_SITE was a relative path, ./configure would fail with the
following misleading output:
```
checking for boostlib >= 1.58.0 (105800)... yes
checking whether the Boost::System library is available... yes
configure: error: Could not find a version of the Boost::System library!
```
Fully resolving depends_prefix in config.site.in fixes this. To make
sure that there are no other side effects I ran a diff on the
config.status generated by:
1. The scripts prior to this change with CONFIG_SITE set to a full path:
env CONFIG_SITE=$PWD/depends/x86_64-pc-linux-gnu/share/config.site ./configure
2. The scripts after this change with CONFIG_SITE set to a relative path:
env CONFIG_SITE=depends/x86_64-pc-linux-gnu/share/config.site ./configure
And it looks good!
Diff: https://paste.sr.ht/~dongcarl/95b469fbc555c128046e85723d87a9082a754f6b
d0a829e963 build: fix mutex detection when building bdb on macOS (fanquake)
Pull request description:
Starting with the Apple Clang shipped with Xcode 12, [Apple has enabled -Werror=implicit-function-declaration by default](https://developer.apple.com/documentation/xcode-release-notes/xcode-12-release-notes):
> Clang now reports an error when you use a function without an explicit declaration when building C or Objective-C code for macOS (-Werror=implicit-function-declaration flag is on). This additional error detection unifies Clang’s behavior for iOS/tvOS and macOS 64-bit targets for this diagnostic. (49917738)
This causes bdbs mutex detection to fail when building on macOS (not cross-compiling):
```bash
checking for mutexes... UNIX/fcntl
configure: WARNING: NO SHARED LATCH IMPLEMENTATION FOUND FOR THIS PLATFORM.
configure: error: Unable to find a mutex implementation
```
as previously emitted warnings are being turned into errors. i.e:
```bash
configure:18704: checking for mutexes
configure:18815: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/clang -mmacosx-version-min=10.12 --sysroot /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk -o conftest -pipe -O2 -I/Users/michael/github/fanquake-bitcoin/depends/x86_64-apple-darwin19.6.0/include -L/Users/michael/github/fanquake-bitcoin/depends/x86_64-apple-darwin19.6.0/lib conftest.c -lpthread >&5
conftest.c:46:1: warning: type specifier missing, defaults to 'int' [-Wimplicit-int]
main() {
^
conftest.c:51:2: error: implicitly declaring library function 'exit' with type 'void (int) __attribute__((noreturn))' [-Werror,-Wimplicit-function-declaration]
exit (
^
conftest.c:51:2: note: include the header <stdlib.h> or explicitly provide a declaration for 'exit'
1 warning and 1 error generated.
```
Append `-Wno-error=implicit-function-declaration` to `cflags` so that `-Wimplicit-function-declaration` [returns to being a warning](https://clang.llvm.org/docs/UsersManual.html#cmdoption-wno-error), and the configure checks succeed.
Fixes#19411.
ACKs for top commit:
laanwj:
Code review ACK d0a829e963
Tree-SHA512: 7813005b1fc0b370f843b6c0672acab32c999416e92c3f02b75d866e9c7aa41fe5822704fc74de6b65f0d7d94f2cdd05cc7c3ee83295ff1ecbc71d8492b9a2bf
Starting with the Clang shipped with Xcode 12, Apple has enabled
-Werror=implicit-function-declaration by default. This causes bdbs mutex
detection to fail when building on macOS (not cross-compiling):
checking for mutexes... UNIX/fcntl
configure: WARNING: NO SHARED LATCH IMPLEMENTATION FOUND FOR THIS PLATFORM.
configure: error: Unable to find a mutex implementation
as previously emitted warnings are being turned into errors. i.e:
error: implicitly declaring library function 'exit' with type 'void (int) __attribute__((noreturn))' [-Werror,-Wimplicit-function-declaration]
Append -Wno-error=implicit-function-declaration to cflags so that
-Wimplicit-function-declaration returns to being a warning, and the
configure checks will succeed.
Fixes#19411.
7a89f2e6c5 build: Fix target name (Hennadii Stepanov)
Pull request description:
It seems like a typo :)
This PR:
- fixes errors when building a package in depends for `HOST=x86_64-apple-darwin16` (fix#19799)
- is a correct alternative to d25e0e308f from #19764
ACKs for top commit:
icota:
tACK 7a89f2e6c5
dongcarl:
Code Review ACK 7a89f2e6c5
theuni:
ACK 7a89f2e6c5.
Tree-SHA512: a0bcbc6805d3450e201476ef1e22e0eb53903db1586c5515314c19afd337bded887e56de0fbe62feaf359b2de15dbccd49a44f1a8b566b4c64f5ae3d94a2ab6d
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.
b893688357 depends: Specify LDFLAGS to cmake as well (Carl Dong)
b3f541f618 depends: Prepend CPPFLAGS to C{,XX}FLAGS for CMake (Carl Dong)
8e121e5509 depends: Cleanup CMake invocation (Carl Dong)
8c7cd0c6d9 depends: More robust cmake invocation (Carl Dong)
3ecf0eca63 depends: Use $($(package)_cmake) instead of cmake (Carl Dong)
Pull request description:
- Use `$($(package)_cmake)` instead of invoking `cmake` directly
- Use well-known env vars instead of overriding CMake variables
ACKs for top commit:
ryanofsky:
Code review ACK b893688357. Only changes since last review are new commits adding whitespace, cppflags and ldflags to cmake invocation
Tree-SHA512: cfcd8cc9dcd0b336cf48b82fca9fe4bbc7930ed397cb7a68a07066680eb4c1906a6a9b5bd2589b4b4999e8f16232fa30ee9b376b60f4456d0fff931fbf9cc19a
