blocks|key|1222294|text|TL;DR:使用本质|type|unstyled|depth|inlineStyleRanges|offset|length|style|BOLD|entityRanges|data|1222295|检查asm输出以确保它不是哑的并不是个坏主意，但是使用intrinsics可以让编译器进行常量传播，并对顺序内核进行调度/软件管道。|1222296|如果您阅读从2009年起你联系的那个帖子上的注释线程，您会发现来自NEON的坏代码是一个gcc错误修复于2011年。|1222297|编译器现在非常擅长处理内部的问题，并且不断改进。尤其是Clang可以做很多事情，比如使用不同的洗牌指令，而不是用本质写的东西。|1222298|至少它们适用于x86；用于ARM的编译器有时仍在与本质进行斗争，特别是当您试图访问16字节向量的两个8字节部分时，就像您通常希望在32位ARM代码中进行横向操作时那样。参见https://stackoverflow.com/questions/46910799/arm-neon-intrinsics-convert-d-64-bit-register-to-low-half-of-q-128-bit-regis+/+https://stackoverflow.com/questions/49511244/neon-intrinsic-for-sum-of-two-subparts-of-a-q-register+-杰克·李报道说，早在2018年，一些clang版本就把它搞得一团糟，但GCC6.x并没有那么糟糕。|1222299|这在AArch64中可能没有那么大的问题。|1222300|asm-级别差异：|1222301|我根本不是这方面的专家，但主要的霓虹灯变化之一是，Aarch64有32个128+B霓虹灯寄存器(v0+-+v31)，而不是每个q寄存器混叠成两个d半。|CODE|1222302|还请参阅一些关于元素大小语法的官方ARM文件，其中可以使用.16B来指示由16个字节元素组成的向量。(相对于意味着每个元素是8位。)|1222303|entityMap|0|LINK|mutability|MUTABLE|url|http://hilbert-space.de/?p=22|1|https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43118|2|https://stackoverflow.com/questions/46910799/arm-neon-intrinsics-convert-d-64-bit-register-to-low-half-of-q-128-bit-regis|3|https://stackoverflow.com/questions/49511244/neon-intrinsic-for-sum-of-two-subparts-of-a-q-register|4|http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.den0024a/ch07s03.html|5|https://stackoverflow.com/questions/38035351/rgba-to-abgr-inline-arm-neon-asm-for-ios-xcode/38040651?noredirect=1#comment63616468_38040651^0|0|A|0|0|5|F|0|18|D|1|0|0|H|A|2E|3D|2|5U|2R|3|0|0|0|1C|8|1R|1|20|1|0|T|4|F|7|4|1I|A|5|0^^$0|@$1|2|3|4|5|6|7|1E|8|@$9|1F|A|1G|B|C]]|D|@]|E|$]]|$1|F|3|G|5|6|7|1H|8|@]|D|@]|E|$]]|$1|H|3|I|5|6|7|1I|8|@]|D|@$9|1J|A|1K|1|1L]|$9|1M|A|1N|1|1O]]|E|$]]|$1|J|3|K|5|6|7|1P|8|@]|D|@]|E|$]]|$1|L|3|M|5|6|7|1Q|8|@$9|1R|A|1S|B|C]]|D|@$9|1T|A|1U|1|1V]|$9|1W|A|1X|1|1Y]]|E|$]]|$1|N|3|O|5|6|7|1Z|8|@]|D|@]|E|$]]|$1|P|3|Q|5|6|7|20|8|@]|D|@]|E|$]]|$1|R|3|S|5|6|7|21|8|@$9|22|A|23|B|T]|$9|24|A|25|B|T]|$9|26|A|27|B|T]]|D|@]|E|$]]|$1|U|3|V|5|6|7|28|8|@$9|29|A|2A|B|T]]|D|@$9|2B|A|2C|1|2D]|$9|2E|A|2F|1|2G]]|E|$]]|$1|W|3|-4|5|6|7|2H|8|@]|D|@]|E|$]]]|X|$Y|$5|Z|10|11|E|$12|13]]|14|$5|Z|10|11|E|$12|15]]|16|$5|Z|10|11|E|$12|17]]|18|$5|Z|10|11|E|$12|19]]|1A|$5|Z|10|11|E|$12|1B]]|1C|$5|Z|10|11|E|$12|1D]]]]

TL;DR: use intrinsics
It's not a bad idea to check the asm output to make sure it's not dumb, but using intrinsics lets compilers do constant-propagation, and schedule / software-pipeline for in-order cores.
If you read the comment thread on <a href="http://hilbert-space.de/?p=22" rel="nofollow noreferrer">that post from 2009 you linked</a>, you'd see that the bad code from NEON intrinsics was a <a href="https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43118" rel="nofollow noreferrer">gcc bug fixed in 2011</a>.
Compilers are quite good at handling intrinsics these days, and continually improving. Clang especially can do quite a lot, like use different shuffle instructions than what you wrote with intrinsics.
At least they are for x86; compilers for ARM still sometimes struggle with intrinsics, especially when trying to access the two 8-byte halves of a 16-byte vector like you often want to in 32-bit ARM code for horizontal operations. See <a href="https://stackoverflow.com/questions/46910799/arm-neon-intrinsics-convert-d-64-bit-register-to-low-half-of-q-128-bit-regis">ARM NEON intrinsics convert D (64-bit) register to low half of Q (128-bit) register, leaving upper half undefined</a> / <a href="https://stackoverflow.com/questions/49511244/neon-intrinsic-for-sum-of-two-subparts-of-a-q-register">NEON intrinsic for sum of two subparts of a Q register</a> - Jake Lee reports that as recently as 2018, some clang versions made a total mess out of it, but GCC6.x was not as bad.
This might not be as much of a problem with AArch64.
<hr />
<h3>asm-level differences:</h3>
I'm not at all an expert on this, but one of the major NEON changes is that Aarch64 has thirty-two 128b NEON registers (<code>v0 - v31</code>), instead of each <code>q</code> register aliasing onto two <code>d</code> halves.
See also some <a href="http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.den0024a/ch07s03.html" rel="nofollow noreferrer">official ARM documentation</a> about syntax for element-size, where you can use <code>.16B</code> to indicate a vector of 16 byte elements. (As opposed to <a href="https://stackoverflow.com/questions/38035351/rgba-to-abgr-inline-arm-neon-asm-for-ios-xcode/38040651?noredirect=1#comment63616468_38040651">the old syntax where <code>.8</code> meant each element was 8 bits</a>.)

I am trying to use MSFA (googles music synth) on 64-bit iOS devices, and it has four NEON assembly source files for DSP operations that are apparently written for 32-bit architectures. I was initially told that it would be better to rewrite this as NEON intrinsics so that it would be architecture-agnostic. However, after reading some articles (such as <a href="http://hilbert-space.de/?p=22" rel="nofollow">http://hilbert-space.de/?p=22</a>), it appears that it still is ideal to have this as pure hand written assembly.

My question is, is it trivial to convert this to 64-bit? If so, how would I get started doing this?

The .s files are:

<a href="https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_fir.s" rel="nofollow">https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_fir.s</a>

<a href="https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_fm_kernel.s" rel="nofollow">https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_fm_kernel.s</a>

<a href="https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_iir.s" rel="nofollow">https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_iir.s</a>

<a href="https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_ladder.s" rel="nofollow">https://github.com/google/music-synthesizer-for-android/blob/master/cpp/src/neon_ladder.s</a>

How do I convert 32-bit NEON assembly to 64-bit?

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

教程

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云智能顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云AI代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

功能1上新10个字符

功能2描述100个字符功能2描述100个字符功能2描述100个字符功能2描述100个字符功能2描述100个字符功能2描述100个字符功能2描述100个字符功能2描述100个字符功能2描述100个字符。

功能2上新100个字符功能2上新100个字符功能2上新100个字符功能2上新100个字符功能2上新100个字符功能2上新100个字符功能2上新100个字符功能2上新100个字符功能2上新100个字符。

功能5描述100个字符功能5描述100个字符功能5描述100个字符功能5描述100个字符功能5描述100个字符功能5描述100个字符

功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符功能5上新100个字符

功能4上新

文章&问答评论现已支持表情

全新交互，全新视觉，新增快捷键、悬浮工具栏、高亮块等功能并同时优化现有功能，全面提升创作效率和体验

社区富文本编辑器全新改版！诚邀体验～ 

精选全网热门MCP server，让你的AI更好用 🚀

💥开发者 MCP广场重磅上线！

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

聚焦“写作效率、视觉美观与运行性能”三方面进行全面升级，为您提供更高效、稳定的创作环境

社区富文本&Markdown编辑器全新改版上线，欢迎大家体验!

诚挚邀请您参与本次调研，分享您的真实使用感受与建议。您的反馈至关重要，感谢您的支持与参与！

社区新版编辑器体验调研

我试图在64位iOS设备上使用MSFA (googles音乐合成)，它有四个用于DSP操作的霓虹灯组装源文件，这些文件显然是为32位架构编写的。我最初被告知，最好把它改写成霓虹灯的本质，这样它就会成为建筑不可知论者。但是，在阅读了一些文章(如)之后，将其作为纯手工编写的程序集似乎仍然是理想的。我的问题是，把这个转换成64位是否微不足道？如果是的话，我怎么开始这么做呢？.s文件是：

问如何将32位霓虹灯组件转换为64位？
EN

回答 1

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问如何将32位霓虹灯组件转换为64位？EN