From -564861454353879676 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: f78e5,7522e5259f7267d6 X-Google-Attributes: gidf78e5,public From: jpotter@falcon.lhup.edu (John Potter) Subject: Re: x = y = z; Undefined? Date: 2000/01/11 Message-ID: <3879ca08.3142429@news.csrlink.net>#1/1 X-Deja-AN: 570713747 Approved: Fergus Henderson , moderator of comp.std.c++ References: <3876254a.30903561@news.csrlink.net> X-Authentication-Warning: mulga.cs.mu.OZ.AU: fjh set sender to devnull@stump.algebra.com using -f X-Complaints-To: abuse@demon.net X-Mail2News-Path: news.demon.net!mulga.cs.mu.oz.au X-Trace: mail2news.demon.co.uk 947512967 mail2news:5192 mail2news mail2news.demon.co.uk Organization: Newscene Public Access Usenet News Service (http://www.newscene.com/) X-Robomod: STUMP, ichudov@algebra.com (Igor Chudov) Nntp-Posting-Date: Mon, 10 Jan 2000 06:34:50 CST Newsgroups: comp.std.c++ On Mon, 10 Jan 2000 08:28:54 CST, Anthony DeRobertis wrote: : In article <3876254a.30903561@news.csrlink.net>, : jpotter@falcon.lhup.edu (John Potter) wrote: : : : >could someone give an example of translation which would produce : >something other than expected? : > : > (i1 = i2) += i3; : : Well, on my PowerPC this would generate (unoptimized): : -- sequence point -- : or i1,i2,i2 : add i1,i1,i3 : -- sequence point -- : : Which could legally be re-arranged: : add i1,i1,i3 : or i1, i2, i2 I don't think so. i3 += i1 = i2; or i1,i2,i2 add i3,i3,i1 Re-arrangement violates the semantics of the language in both cases. The implementation is allowed to assume that no stored value is modified more than once. I guess that you could stretch it to the point of being allowed to detect an expression which invokes undefined behavior and generate purposely bad code; however, the compiler writers are in the business of selling compilers not harassing users. In both cases, the second instruction uses the value of the first instruction. : Which could then be optimized: : or i1, i2, i2 : : Which means: : i1 = i2; : : As there are no sequencing points. This could concievably be done for : scheduling purposes. I don't seem to be able to get my compile to do it I'm not surprised. : (it does even weirder stuff...) Given i2 == 2 and i3 == 3, does it do anything which produces other than i1 == 5? Assuming, of course, that there is a reason to do anything. My compiler nicely optimized void f () { int i1, i2(2), i3(3); (i1 = i2) += i3; } to void f () { } Changing to int f () and adding return i1, it optimized to return 5. int f (int i2, int i3) { int i1; (i1 = i2) += i3; return i1; } optimized to add r3,r3,r4 which is i1 = i2 + i3. Streatching my imagination, thanks to James Kanze load r1,i2 copy r2,r1 add r2,i3 sto i1,r2 sto i1,r1 is a valid translation. With multiprocessors, the last two instructions could be executed in parallel opening the door for many results depending on the hardware. The ++ ++ i example also has strange valid translations. I'm convinced, finally. John --- [ comp.std.c++ is moderated. To submit articles, try just posting with ] [ your news-reader. If that fails, use mailto:std-c++@ncar.ucar.edu ] [ --- Please see the FAQ before posting. --- ] [ FAQ: http://reality.sgi.com/austern_mti/std-c++/faq.html ]