Date: Thu, 08 May 1997 16:31:58 -0700 From: Jason Thorpe <thorpej@nas.nasa.gov> To: Warner Losh <imp@village.org> Cc: Ben Black <black@zen.cypher.net>, Brandon Gillespie <brandon@cold.org>, hackers@freebsd.org Subject: Re: Alpha questions.. Message-ID: <199705082331.QAA20330@lestat.nas.nasa.gov>
next in thread | raw e-mail | index | archive | help
On Thu, 08 May 1997 17:19:20 -0600 Warner Losh <imp@village.org> wrote: > I was unaware that the FSF folks were so dead set against it. The > concetp looks like a very fast way to dispatch the system calls based > on the type of OS that you have (or more generally, whhc ABI the > program conforms to). However, there are ways of doing this with ELF > which aren't exactly fast, but aren't exactly slow either (assuming > that the binaries are branded or otherwise marked). ...and it's not like it really _matters_... the dispatch happens exactly once... and, really, how expensive is looking for PT_NOTE compared to doing the VM stuff necessary to map the executable? OLF has at least two problems: (1) doesn't scale (2) unnecessarily breaks compatibility Jason R. Thorpe thorpej@nas.nasa.gov NASA Ames Research Center Home: 408.866.1912 NAS: M/S 258-6 Work: 415.604.0935 Moffett Field, CA 94035 Pager: 415.428.6939
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199705082331.QAA20330>