Path: csiph.com!eternal-september.org!feeder.eternal-september.org!nntp.eternal-september.org!.POSTED!not-for-mail From: Tim Rentsch Newsgroups: comp.arch Subject: Re: endian history Date: Tue, 22 Sep 2026 16:55:50 -0700 Organization: A noiseless patient Spider Lines: 39 Message-ID: <861paku9h5.fsf@linuxsc.com> References: <10uks4f$1dqo$1@gal.iecc.com> <10vnjp6$1df4$1@gal.iecc.com> <10vns1f$3a252$1@dont-email.me> <10vpblp$3mkd1$1@dont-email.me> <10vphh0$3o9jl$1@dont-email.me> <117gj6t$1b95v$2@dont-email.me> <6a9c1c93.39172234@news.eternal-september.org> <117j3es$24qgq$6@dont-email.me> <117n6ho$35j69$1@paganini.bofh.team> <6a9f7664.1891062@news.eternal-september.org> <2026Sep8.125659@mips.complang.tuwien.ac.at> <117pnpg$f179$2@dont-email.me> <11898n9$1nc06$2@dont-email.me> <118d6ld$1tkm9$1@paganini.bofh.team> <118dl5f$37cfv$1@dont-email.me> <118e8kg$3f05h$1@dont-email.me> <868q4zuhz4.fsf@linuxsc.com> <118jitv$1cm6b$1@dont-email.me> <86ik42tcf4.fsf@linuxsc.com> <1790100995-5857@newsgrouper.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Injection-Date: Tue, 22 Sep 2026 23:55:51 +0000 (UTC) Injection-Info: dont-email.me; logging-data="1377169"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX19zDK/mwpORIN87aI/nMIkGYy0QYDDtxbk="; posting-host="03472381143ebeb5b8a249e157345c91" User-Agent: Gnus/5.11 (Gnus v5.11) Emacs/22.4 (gnu/linux) Cancel-Lock: sha1:GAgMQpax03SwBBdBROLvDkBj+wY= sha1:U55AVDl2nXzsWAsg0sep+Lgu3ew= sha256:fYu1yCdxNuhhaHqHdlR4TKUmeSp+T+ulQF/9paURx58= sha1:ZZart4Ba8S+ecda9O8JvljwPnhQ= sha256:TJQxS41yRcwx8A/3Db1QjE5LVMmuo2OPGMYRScvsNgA= Xref: csiph.com comp.arch:118137 MitchAlsup writes: > Tim Rentsch posted: > >> Terje Mathisen writes: > > ----------------------- > >>> In order to be really fast, the compiler has to recognize that both >>> the div and mod should use reciprocal mul (standard), and if possible, >>> only do it once. >>> ... >>> Checking Godbolt: >>> >>> Turns out it is actually generating two IMULs, but it should work for >>> both positive and negative inputs... >> >> I suspect the two multiplys are a consequence of the usual way for >> producing a quotient and remainder: >> >> quotient = n / BASE; >> remainder = n - quotient*BASE > > What is wrong with the standard way:: > > quotient = n / BASE; // 1 divide > remainder = n % BASE; // no additional instruction > > ?? I didn't mean to suggest a way to write the code. I was trying to explain how the compiler might have transformed the code, so as to need two multiplies rather than one. As it turns out, since I posted the earlier message, I have looked into this matter a bit more, and this proposed explanation isn't right. I think I now understand why there were two multiplies, but I haven't investigated that further because of possible differences between C and Rust.