Fixes #29248
The std::result_of type was removed in c++20, but was being referenced in some old, unused code in the library. The issue was fixed in:
This update also includes other recent libmultiprocess changes to improve C++20 support and fix build issues:
<!--e57a25ab6845829454e8d69fc972939a-->
The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.
<!--006a51241073e994b41acfe9ec718e94-->
For detailed information about the code coverage, see the test coverage report.
<!--021abf342d371248e50ceaed478a90ca-->
See the guideline for information on the review process.
| Type | Reviewers |
|---|---|
| ACK | fanquake |
| Concept ACK | edilmedeiros |
| Stale ACK | jarolrod |
If your review is incorrectly listed, please react with 👎 to this comment and the bot will ignore it on the next update.
<!--174a7506f384e20aa4161008e828411d-->
Reviewers, this pull request conflicts with the following ones:
If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.
ACK 5dfd24581a7c5497601966a5371d8c33eabcee8a - thanks. Checked this fixes the macOS build. Also tested the build on aarch64.
There is another macos issue I'm debugging here: https://github.com/chaincodelabs/libmultiprocess/issues/92 and I don't think it shows up in depends builds because the depends build skips tests. But it might trigger errors in bitcoin builds if more code from #10102 is merged.
There is no need to hold off on this PR, but if it is not merged, I will probably update it again with whatever the fix is for https://github.com/chaincodelabs/libmultiprocess/issues/92
I probably need to figure out a better way of doing CI to catch these issues.
ACK 5dfd24581a7c5497601966a5371d8c33eabcee8a
This fixes the build issue.
Fixes #29248
The std::result_of type was removed in c++20, but was being referenced in some
old, unused code in the library. The issue was fixed in:
https://github.com/chaincodelabs/libmultiprocess/pull/91 util: Drop Bind, BindTuple, ComposeFn, GetFn, and ThrowFn helpers
This update also includes other recent libmultiprocess changes to improve C++20
support and fix build issues:
https://github.com/chaincodelabs/libmultiprocess/pull/89 pkgconfig: Drop -std=c++17 compile flag
https://github.com/chaincodelabs/libmultiprocess/pull/90 pkgconfig: Use @CMAKE_INSTALL_LIBDIR@ variable
https://github.com/chaincodelabs/libmultiprocess/pull/93 Fix support for vector<bool> serialization with libc++
Updated 5dfd24581a7c5497601966a5371d8c33eabcee8a -> b8105b3ed7c97cd6545dba99d0e13c33f183e450 (pr/c14.1 -> pr/c14.2, compare) just adding a new macos build fix: https://github.com/chaincodelabs/libmultiprocess/pull/93.
I'm pretty sure the new fix is not needed for any current depends builds because depends builds skip tests, but since code in #10102 does need the fix, it could potentially help review of that code if the fix is in place now. Sorry for the churn.
Tested ACK.
I was watching the issue in chaincodelabs/libmultiprocess#92 and confirmed the fix in chaincodelabs/libmultiprocess#93.
I built and executed bitcoin-core tests with this PR following the official setup with Brew.
I could not compile this with dependencies coming from Macports, but the issue seems to be not related to the fix here as it does not compile either with --disable-multiprocess (I understand Macports is not supported anyhow, see #15175).
ACK b8105b3ed7c97cd6545dba99d0e13c33f183e450.