9381052194 doc: Bash is needed in gen_id and is not installed on FreeBSD by default (Hennadii Stepanov)
Pull request description:
On FreeBSD 14.0, in the `depends` directory:
- without `bash`:
```
$ gmake print-bdb_build_id_long
env: bash: No such file or directory
env: bash: No such file or directory
bdb_build_id_long=bdb-4.8.30-4b0c6f8e95251b9c6731844fc34111c04b75fd9f15c671d6e34f2a4d014ec1be-release
$ gmake print-final_build_id
env: bash: No such file or directory
env: bash: No such file or directory
final_build_id=722b2d3e264
```
- with `bash`:
```
$ gmake print-bdb_build_id_long
bdb_build_id_long=bdb-4.8.30-4b0c6f8e95251b9c6731844fc34111c04b75fd9f15c671d6e34f2a4d014ec1be-release 1ed47cefe468014c79dedb275cf921f44ab28d91dd56bf94712409b81326d765
$ gmake print-final_build_id
final_build_id=7b4f9aaa683
```
ACKs for top commit:
vasild:
ACK 9381052194
kristapsk:
ACK 9381052194
alfonsoromanz:
ACK 9381052194
Tree-SHA512: da3f3469ac416518180194f09fb054fb352a2793848fb9a7982439de08244ff6149a7f449ad21fcdf0e9bd79b6949a91751f9cc35833953d2b6a35cea5c6ae21
1a9aa8d4ee build: better scope usage of -Wl,-headerpad_max_install_names (fanquake)
3bee51427a build: don't use install_name_tool for macOS deploy when cross-compiling (fanquake)
78b6b5c485 build: don't pass strip to macOS deploy if cross-compiling (fanquake)
Pull request description:
Neither of these tools are actually used when we are cross-compiling for macOS. They are used when we have to adjust non-static libs during a deploy after building on a macOS machine. Simplies #29739 (will be rebased on top).
Guix (aarch64):
```bash
8f29bce75d7f574306a0e38d793e0e4e145b547a4b9e5a755a54976121d8ac41 guix-build-5afd3c302051/output/arm64-apple-darwin/SHA256SUMS.part
9ba01fe46be715adcbe80f39dc7dbe449f32ca9d9b660da698f933aef3e6d80b guix-build-5afd3c302051/output/arm64-apple-darwin/bitcoin-5afd3c302051-arm64-apple-darwin-unsigned.tar.gz
37719437e951449341d0e10dcc4afe93e955d59de5312ce6351e1fa01b4927ac guix-build-5afd3c302051/output/arm64-apple-darwin/bitcoin-5afd3c302051-arm64-apple-darwin-unsigned.zip
06a79fc871dcd4290f5f7e7e9de19a5a535203d20279f4555d1c319d07abe2d0 guix-build-5afd3c302051/output/arm64-apple-darwin/bitcoin-5afd3c302051-arm64-apple-darwin.tar.gz
98d2b8b37197dcad36a04eb2f3ff2130b859220a17b83a4186a78dcf0af4eafd guix-build-5afd3c302051/output/dist-archive/bitcoin-5afd3c302051.tar.gz
df63ff44ef41565ff13ce6dde5485173a18d5866ebc316df86f9ebd91fda18f5 guix-build-5afd3c302051/output/x86_64-apple-darwin/SHA256SUMS.part
28362ce9e80d5e78db198efa5f89434fbe76ca91df5fde7455da4d50ceb8523a guix-build-5afd3c302051/output/x86_64-apple-darwin/bitcoin-5afd3c302051-x86_64-apple-darwin-unsigned.tar.gz
534745b679eb9e8e408dd251a6bf0829e62e12f7a41772b8a57a044ded14208c guix-build-5afd3c302051/output/x86_64-apple-darwin/bitcoin-5afd3c302051-x86_64-apple-darwin-unsigned.zip
f53d0c9a1bb83d548c7d274c7d39653a3989fb1b4efec49e73dd1cac7c92074c guix-build-5afd3c302051/output/x86_64-apple-darwin/bitcoin-5afd3c302051-x86_64-apple-darwin.tar.gz
```
ACKs for top commit:
TheCharlatan:
ACK 1a9aa8d4ee
Tree-SHA512: 0aa77ea4d6dc45c226806bb1758b6aa7e8ca17f91045bab4fc6891af7b9de476211cd5692c11cb9d5bcf59744fd86a2534812a77fe304ae10c3518e08fc412be
3c1ae3ee33 depends: switch libnatpmp to CMake (Cory Fields)
72ba7b5d26 depends: libnatpmp f2433bec24ca3d3f22a8a7840728a3ac177f94ba (fanquake)
Pull request description:
This picks up one of the changes from https://github.com/bitcoin/bitcoin/pull/29232, which is a switch to building libnatpmp with CMake. It includes an update to the most recent version of libnatpmp (f2433bec24), which includes (https://github.com/miniupnp/libnatpmp/pull/43).
From an initial look I couldn't find any significant difference between the Autotools and CMake produced libs.
ACKs for top commit:
m3dwards:
ACK 3c1ae3ee33
hebasto:
ACK 3c1ae3ee33.
TheCharlatan:
ACK 3c1ae3ee33
Tree-SHA512: 1dd9d9933a5fceb9f8c4e1d68cd5cb4456a10a6dd27a6f6316f14493f9d2efad981ef8be9570c09ca82d45163aebd7f4cb2b2449989ec6084268ddba9a564c83
007ea322a6 depends: switch to building libqrencode with CMake (fanquake)
884330c0a5 guix: make cmake-minimal a global requirement (fanquake)
Pull request description:
Switch to building libqrencode with CMake. Note that upstream (https://github.com/fukuchi/libqrencode) hasn't seen any activity for ~4 years, so the odds of getting anything upstream seems low, but I've made two minor changes to the source here, which I will PR in any case.
From an initial look I couldn't find any significant difference between the Autotools and CMake produced libs. As part of this change we move cmake-minimal in Guix into the global package set.
ACKs for top commit:
TheCharlatan:
ACK 007ea322a6
Tree-SHA512: c784f790ddea958082c8ae96d3744bdf99331a8799765f9d44f00861b8e2cfcab1a88a3d64af5b10e51a8d5938d55eb6a3d271790b565e50492a39d00dc0e30f
84fbf9b284 depends: remove -g from sqlite debug flags (fanquake)
eef51afc6a depends: add -g to DEBUG=1 flags (fanquake)
Pull request description:
Add `-g` to the base DEBUG=1 flags in depends.
Avoids the need to specify it per-package.
More alignment with `--enable-debug` behaviour in configure.
We also want to align the optimization flags, currently -O1 vs -O0, however that can be it's own PR.
ACKs for top commit:
theuni:
ACK 84fbf9b284
Tree-SHA512: 1ee98ba0c13e4b80bb87632658b4f53ce49c73e0e7712990c30da60deca4a349a744232f2d78f243dee9a07f5b9b70f9c2c4ae34082c34ae51b37b054fac61fd
430f319f73 depends: qt 5.15.13 (fanquake)
Pull request description:
I need to try and patch around Qts buildsystem to keep #21778 moving along (the issue being that even when you tell Qt to build using Clang on Linux, it still calls out to GCC, breaking our ability to have a macOS release build env that doesn't have a GCC toolchain installed, and thus no `ld` binary).
Before trying to patch Qt any further, update to the latest LTS release, and update the current patch set.
ACKs for top commit:
TheCharlatan:
Re-ACK 430f319f73
hebasto:
re-ACK 430f319f73.
Tree-SHA512: 11122c04d63f4ec79e5d7945e1686eaf308cb1f98c3ff8bc52e265d7353cd4f1a532f4d78d55b43183245aeebf93cc1092087d9fe83e42eefa8a69ad9017d2da
601edd8ee8 ci: use codespell 2.2.6 (fanquake)
52fa0d285f doc: fix some typos (crazeteam)
b5ed13a240 doc: Fix typos (RoboSchmied)
Pull request description:
Combines the recent PRs to fix typos so they can be merged.
ACKs for top commit:
brunoerg:
crACK 601edd8ee8
tdb3:
crACK 601edd8ee8
kristapsk:
cr utACK 601edd8ee8
Tree-SHA512: d054b1dad1336d6b9291cc5d5252d4debf6424a993d4edd6a97d7c15055a7fc48a333d30967f72e7dc9c6c1d9a9038ca8bb5e219c529f4c2365ea48404a508d0
I need to try and patch around Qts buildsystem to keep #21778 moving
along (the issue being that even when you tell Qt to build using
Clang on Linux, it still calls out to GCC, breaking our ability to have
a macOS release build env that doesn't have a GCC toolchain installed,
and thus no ld binary).
Before trying to patch Qt any further, update to the latest LTS
release, and update the current patch set.
The issue is that compilation is done with `x86_64-w64-mingw32-g++-posix`,
but then linking is done with `x86_64-w64-mingw32-g++`.
I'm guessing this has been broken since #24131
(01d1845a80), but have not checked.
Fixes#29734.
Unblocks #29527 (now DEBUG=1 builds can be tested).
Rather than potentially having to set this per-package, set it globally,
as this should always be what we want. Without doing this, changes in
later commits will have to add this per-package.
Similar to https://github.com/bitcoin/bitcoin/pull/29488, which is the
Autotools equivalent.
Rather than setting this per package, set it globally, as this is always
what we want. Without doing this, later commit will have to add the same
doc + change to more packages.
e037c4fe09 depends: always configure with --with-pic (fanquake)
Pull request description:
We currently do this sporadically. Not only amongst packages, but across OS's, i.e sometimes it's done for BSDs/Android, and sometimes not.
Configure with `--with-pic` globally instead. I think this generally makes more sense, and should not have any downsides.
See related discussion in https://github.com/bitcoin/bitcoin/pull/28846#discussion_r1399123100.
ACKs for top commit:
hebasto:
ACK e037c4fe09.
Tree-SHA512: efc743ff92f9f99f3ac16514e98363ad395c6f956cd4be7e785b5c573685baf7fcd68c51d6a705ee8761fc676eb045b7e61676595be0eb0f70f34e99174cddc0
We currently do this sporadically. Not only amongst packages, but across
OS's, i.e sometimes it's done for BSDs/Android, and sometimes not.
Configure with `--with-pic` globally instead. I think this generally
makes more sense, and should not have any downsides.
See related discussion in
https://github.com/bitcoin/bitcoin/pull/28846#discussion_r1399123100.
Compiling C++ code with `-D_XOPEN_SOURCE=600` causes problems on
OpenBSD. If that define is set, the C++ standard header detection
routine in BDB's configure script fails. This results in
`HAVE_CXX_STDHEADERS` not being defined, which then it turn leads to
the inclusion of `<iostream.h>` (rather than `<iostream>`), which
doesn't exist.
According to a mailing list post discussing a similar problem [1],
"OpenBSD provides the POSIX APIs by default", so we don't need this
define anyway and can remove it. This fixes the BDB build problem as
described in issue #28963.
Tested on OpenBSD 7.4 with clang 13.0.0.
[1] https://www.mail-archive.com/tech@openbsd.org/msg63386.html
5b9d5bf866 depends: remove (darwin) libtool now that it's no longer used (Cory Fields)
3ef6563495 depends: use ar rather than libtool for miniupnpc/libnatpmp (Cory Fields)
Pull request description:
An alternative to https://github.com/bitcoin/bitcoin/pull/29232
Rather than switching to the CMake builds which [proved problematic](https://github.com/bitcoin/bitcoin/pull/29232#issuecomment-1898513919), do the quick and dirty thing of just patching out libtool. Doesn't seem to introduce any new issues.
This should buy us time to upstream the necessary CMake fixes.
ACKs for top commit:
TheCharlatan:
ACK 5b9d5bf866
fanquake:
ACK 5b9d5bf866
Tree-SHA512: c75c4bcc9332d8c1fc3395e2b5fc7265849186afc7005700f662ab291e6ea1f111025fad733d0b0b39d35029d1b757d3f1937d63aad3c0c3b88d0f8ac902ee18
The `--enable-debug` configure option for the SQLite package does two
things. It adds three preprocessor definitions and overrides CFLAGS with
"-g -O0". The latter breaks the user's ability to provide sanitizer and
LTO flags.
6ec2813cd8 depends: add NM output to gen_id (fanquake)
Pull request description:
`NM` is part of the current toolset, and can be set by the user. Include it in `gen_id`.
ACKs for top commit:
TheCharlatan:
Re-ACK 6ec2813cd8
Tree-SHA512: 2ada61e03783f9eb441f285ef5da50557ad729cb52ce2d2c4b2c38103dab29920a26262d4545fd2ac7fbf1cedc4902cd2359833544fbc0debf829c12a63e9769
2d1b1c7dae build: remove --enable-lto (fanquake)
Pull request description:
This has outlived its usefulness, doesn't gel well with newer compilers & `-flto` related options, i.e thin vs full, or `=auto`, and having `-flto` as the only option means that sometimes this just needs to be worked around, i.e in oss-fuzz:
https://github.com/google/oss-fuzz/blob/master/projects/bitcoin-core/build.sh.
While it was convenient when `-flto` was newer, support for `-flto` is now in all compilers we use, and there's also no-longer any real need for us to treat `-flto` different to any other optimization option.
Remove it, to remove build complexity, and so there's no need to port a similar option to CMake.
Note that the LTO option remains in depends, because we still a way to build packages that have LTO specific patches/options.
ACKs for top commit:
TheCharlatan:
ACK 2d1b1c7dae
hebasto:
ACK 2d1b1c7dae.
Tree-SHA512: 91812de7da35346f51850714a188fcffbac478bc8b348bf756c2555fcbde86ba622ac2fb77d294dea0378c741d3656f06121ef3a795aeed63fd170fc31bfa5af
ff3f51b402 depends: Include `config.guess` and `config.sub` into `meta_depends` (Hennadii Stepanov)
Pull request description:
ACKs for top commit:
theuni:
ACK ff3f51b402.
Tree-SHA512: e8575473d3fca2293181131c76bd6d43017fe753d2e670c53227a646b64b069dc542a0fc50a77b43e74bc6a0c0159ffa2fb1c3ff3aef9625684e0f78c16ad960
The `ALLOW_HOST_PACKAGES` variable was introduced in bitcoin#10508 "to
speed up build and avoid timeout".
It is no longer the case for our CI infrastructure, which uses self-
hosted persistent workers and depends caching.
In the current circumstances, it does not seem worth porting this
feature to the upcoming CMake-based build system.
This has outlived its usefulness, doesn't gel well with
newer compilers & `-flto` related options, i.e thin vs full, or `=auto`,
and having `-flto` as the only option means that sometimes this just
needs to be worked around, i.e in oss-fuzz:
https://github.com/google/oss-fuzz/blob/master/projects/bitcoin-core/build.sh.
While it was convenient when `-flto` was newer, support for `-flto` is now
in all compilers we use, and there's also no-longer any real need
for us to treat `-flto` different to any other optimization option.
Remove it, to remove build complexity, and so there's no need
to port a similar option to CMake.
Note that the LTO option remains in depends, because we still a way to
build packages that have LTO specific patches/options.
If we decide to merge this, I'll follow up downstream in oss-fuzz first,
to make sure we don't break the build.
