Xi Ruoyao

The addition of `pushy_https` is an action to fix some security vulnerabilities, details at http://blogs.perl.org/users/neilb/2021/11/addressing-cpan-vulnerabilities-related-to-checksums.html. I guess the content of security note should depend on "how much level a normal...

This thing is just like "cascaded" `dynamic_cast`s in C++. We should always refactor it to a function call via a `interface` if `n` is really large.

> > It's difficult for the nix maintainers to validate changes for platforms which are not supported by `rustc`, including ensuring they don't immediately regress in support as part of...

Please remove this line: https://github.com/openssl/openssl/pull/21510/files#diff-cfb77b1f843910cbcc75a53d1e387ab16a6a630beff34c95978e22ca3a16317aR687 and the related `$disable` probing above it. (I added the detection at the wrong place and with your change it can be removed.)

How does the other architectures handle the issue? With Perl scripts it might be easy (?) to rewrite the vector instructions with some Perl functions emitting `.word `...

上游不可能同意把符号版本改成 2.34 (我记得最早的一批 patch 因为这事遭到了批判)，所以只能动 openEuler，建议在他们那边开 issue。

Maybe related to #683.

目前都是将 -gnu 视为 -gnuf64 的别名，之前讨论的时候也没人反对继续这么做：https://github.com/loongson/LoongArch-Documentation/issues/29#issuecomment-972645700

> 其它架构常常把 `-gnu` 作为最常用的 fabi 的别名，也就是可以省略 fabi suffix，所以我倾向明确把 `-gnu` 作为 `-gnuf64` 的别名。 > > 需要注意这里讨论的是 GNU tuple 的别名问题。Multiarch 架构标识符的话只能用 `loongarch64-linux-gnuf64`，不允许用 `loongarch64-linux-gnu`。 同意，然而 Debian (multiarch 这东西的发明者) 居然在用 `loongarch64-linux-gnu`：https://wiki.debian.org/Ports/loong64 谁联系他们改一下罢……

FWIW it can be worked around by removing `src/api.c` but it will then require `valac` to regenerate this file.