The rest of the 6.15 merge window

Linus Torvalds released 6.15-rc1 and closed the 6.15 merge window on April 6. By that time, 12,633 non-merge changesets had found their way into his repository; that is substantially more than were merged during the entire 6.14 development cycle. Just under 6,000 of those changesets were merged after the first-half merge-window summary was written.
Linus Torvalds 于 4 月 6 日发布了 6.15-rc1，并关闭了 6.15 的合并窗口。截至当时，共有 12,633 个非合并变更集进入了他的代码仓库；这个数量明显超过了整个 6.14 开发周期中的合并总数。其中将近 6,000 个变更集是在第一阶段合并窗口总结完成之后合入的。

Some of the most interesting changes from the second half of the 6.15 merge window are:
以下是 6.15 合并窗口后半阶段中一些最有趣的改动：

Architecture-specific
体系结构相关

The RISC-V architecture has gained support for the BFloat16 floating-point extension, the Zaamo and Zalrsc extensions, and the ZBKB extension.
RISC-V 架构新增了对 BFloat16 浮点扩展、Zaamo 与 Zalrsc 扩展，以及 ZBKB 扩展的支持。

Core kernel
内核核心

The function and function-graph tracers have gained the ability to record the arguments to called functions; those arguments can then be examined in the trace output.
函数和函数图追踪器现在能够记录被调用函数的参数；这些参数随后可以在追踪输出中查看。

The io_uring subsystem now supports zero-copy reception of network data, with eventual plans for allowing reception directly into device memory. See this commit for documentation.
io_uring 子系统现在支持网络数据的零拷贝接收，未来计划支持直接接收到设备内存中。相关文档可见于该提交。

It is also now possible to read epoll events via an io_uring operation. From the merge message:
现在也可以通过 io_uring 操作读取 epoll 事件。以下是合并消息中的内容：

While this may seem counter-intuitive (and/or productive), the reasoning here is that quite a few existing epoll event loops can easily do a partial conversion to a completion based model, but are still stuck with one (or few) event types that remain readiness based.
尽管这看起来可能不太直观（或不够高效），但这里的逻辑是，许多现有的 epoll 事件循环可以轻松地部分转换为基于完成（completion-based）的模型，但仍受限于一个或少数几个仍基于就绪状态（readiness-based）的事件类型。

For that case, they then need to add the io_uring fd to the epoll context, and continue to rely on epoll_wait(2) for waiting on events. This misses out on the finer grained waiting that io_uring can do, to reduce context switch