Date: Fri, 10 Mar 2000 22:43:45 +0100 From: Martin Cracauer <cracauer@cons.org> To: Alfred Perlstein <bright@wintelcom.net> Cc: Doug Rabson <dfr@nlsystems.com>, Martin Cracauer <cracauer@FreeBSD.ORG>, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG Subject: Re: cvs commit: src/sys/i386/include npx.h Message-ID: <20000310224345.A20522@cons.org> In-Reply-To: <20000310133936.B14279@fw.wintelcom.net>; from bright@wintelcom.net on Fri, Mar 10, 2000 at 01:39:36PM -0800 References: <200003101756.JAA90710@freefall.freebsd.org> <Pine.BSF.4.21.0003102057080.79394-100000@salmon.nlsystems.com> <20000310133936.B14279@fw.wintelcom.net>
next in thread | previous in thread | raw e-mail | index | archive | help
In <20000310133936.B14279@fw.wintelcom.net>, Alfred Perlstein wrote: > * Doug Rabson <dfr@nlsystems.com> [000310 13:32] wrote: > > On Fri, 10 Mar 2000, Martin Cracauer wrote: > > > > > cracauer 2000/03/10 09:56:33 PST > > > > > > Modified files: > > > sys/i386/include npx.h > > > Log: > > > Change the default FPU control word so that exceptions for new > > > processes are now masked until set by fpsetmask(3). > > > > Excellent! > > Is there any reason for doing this other than so we can emulate > linux's bog^H^H^H nifty ability to divide by zero? > > What's the point of this except to mask obvious programming error? Please, before flooding -committers, everyone with a strong opinion in this regard should check the archives of -current, -hacker, -bugs and whatnot for message by Bruce, me, Jordan and other people. Or (no pun intended, really) could have taken part in the discussion. Overall, it is not clear what the right thing is (we can't claim to be IEEE 754 compliant for other reasons), so the convincing thing (for me) was that every application that really needs a specific behaviour *must* set the mask by itself. We can't really do anything for those applications that don't announce their expectations and we now go the way of least resistance by users (my intepretation). The only real drawback now is that our experienced user can't really choose anymore. In the past, you could change your own machines to masked exceptions, but now ports will stop inserting fpsetmask(3) calls soon and hence the other way for the cautios user will not work for long (unless the mass is high enough to catch those ports). But since you don't gain more from blindly inserted fpsetmask calls than from a default-to-masked the real paranoid people will maybe see the situation as improved, not weakend. Martin -- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Martin Cracauer <cracauer@cons.org> http://www.cons.org/cracauer/ Tel.: (private) +4940 5221829 Fax.: (private) +4940 5228536 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000310224345.A20522>