Date: Mon, 28 Feb 2005 21:53:58 +0100 From: Francois Tigeot <ftigeot@wolfpond.org> To: freebsd-amd64@freebsd.org Subject: Re: ia32 ports... Message-ID: <20050228205358.GA73467@aoi.wolfpond.org> In-Reply-To: <20050228195158.GC59327@dragon.nuxi.com> References: <16931.17425.441677.617748@canoe.dclg.ca> <20050228181056.GC69361@aoi.wolfpond.org> <20050228195158.GC59327@dragon.nuxi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Feb 28, 2005 at 11:51:58AM -0800, David O'Brien wrote: > On Mon, Feb 28, 2005 at 07:10:56PM +0100, Francois Tigeot wrote: > > On Mon, Feb 28, 2005 at 11:17:21AM -0500, David Gilbert wrote: > > > Having wrestled with a few ports (including java) on amd64, what I > > > don't understand is why we don't have the capability to compile ia32 > > > versions of ports that won't compile in amd64 mode... > > > > It would be even better to have an option to compile amd64 binaries with > > 32-bit pointers. > > Unclean code would be able to run an still benefit from the amd64 > > architectural improvements (more registers, better abi, etc...) > > The extra registers (and thus ABI) is not supported in 32-bit mode for > compatability reasons. I know that. What I meant was not to compile to i386 but to amd64 with 32-bit pointers. The binary would technically still be a 64-bit one. This page explains the mechanism: http://alpha.caspur.it/C++6.3A_DOC/uguhdr.htm#xtaso_sec It was possible to compile "32-bit limited" applications on the Alpha, which was a 64-bit only architecture. I don't know if it is possible on amd64. I'm not sure I'm very clear either. -- Francois Tigeot
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050228205358.GA73467>