Date: Thu, 18 Dec 1997 00:57:20 +0100 (MET) From: Oliver Fromme <olli@dorifer.heim3.tu-clausthal.de> To: freebsd-sparc@FreeBSD.ORG Subject: Re: Data types Message-ID: <199712172357.AAA20211@dorifer.heim3.tu-clausthal.de>
next in thread | raw e-mail | index | archive | help
Robert S. Sciuk <rob@ControlQ.com> wrote in list.freebsd-sparc: > I'm not sure that this message went out to the list, but relevant info > appears at the following sites: > > http://www.UNIX-systems.org/version2/whatsnew/datasize.html > > and ... > > http://www.UNIX-systems.org/version2/whatsnew/lp64_wp.html > > I hope that this proves useful. Indeed, very interesting documents. Thanks! Especially, the chapter "Performance Characteristics" in the second document illuminates the question whether ints should be 32 or 64 bits on a hardware platform which is natively 64 bits (they mention Alpha, MIPS, Sparc V9 [AKA Ultra] and PowerPC). They come to the conclusion that ints should be 32 bits, and longs and pointers should be 64 bits, with which I tend to agree. (Note: ILP64 means ints, longs, and pointers are all 64 bits, LP64 means that ints are 32 bits.) <QUOTE> Some software vendors have experimented with an ILP64 model, which can be approximated on LP64 systems by changing all int declarations to long. In these cases, the conclusion reached after these experiments was not to use ILP64, since the application did not benefit from the additional range of int values and did not wish to pay the performance penalty of extra memory use. [...] All this, as well as natural use of the native C datatypes to support all the widths needed in a 64-bit system make a compelling argument for the inherent advantage of LP64. </QUOTE> This is just a small excerpt; I recommend to read the whole document. Regards Oliver -- Oliver Fromme, Leibnizstr. 18-61, 38678 Clausthal, Germany (Info: finger userinfo:olli@dorifer.heim3.tu-clausthal.de)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199712172357.AAA20211>