Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Frankly, I'm not sure what to do about mpir, which is a fork of libgmp, used to build ext/gmp on Windows. The latest release (3.0.0) was published more than seven years ago, and the project seems to be abandoned. There is https://github.com/BrianGladman/mpir, a fork that focuses on Windows, which had some activity, but according to wbhart/mpir#272, neither the original work, nor Brian Gladman's fork are still maintained.
If no further maintenance will happen, we may consider to switch to libgmp, although there is no support for native Windows builds (only autoconf), which is likely one of the reasons why we had chosen mpir; the other is that mpir provides optimized implementations (in Assembler, using SIMD instructions), although I don't think that any of these have been used for our official builds. Those that are offered out of the box by BrianGladman/mpir use AVX and maybe AVX2 instructions, and that would not fit with our PHP builds which are constrained to SSE2. If we switch to libgmp, we may even consider using mini-gmp which is a minimal variant consisting of a few C files only, but claims to be slow since it's lacking optimizations. It might not be that bad, considering that we didn't deploy the optimizations of mpir.
Anyhow, to make some progress, this PR add a builder for BrianGladman/mpir (which has at least one relevant bug fix). A test build showed no issues running the ext/gmp test suite on
master
/x64. Since that repo never released tags, and the version numbers have never been bumped, I have no idea what to do about that. Of course, we could modify the version numbers, or just call it mpir-3.0.0-1 or something.