From owner-freebsd-hackers Wed Feb 15 11:29:20 1995 Return-Path: hackers-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id LAA00260 for hackers-outgoing; Wed, 15 Feb 1995 11:29:20 -0800 Received: from uivlsi.csl.uiuc.edu (uivlsi.csl.uiuc.edu [128.174.57.133]) by freefall.cdrom.com (8.6.9/8.6.6) with SMTP id LAA00253 for ; Wed, 15 Feb 1995 11:29:19 -0800 Received: by uivlsi.csl.uiuc.edu id AA13080 (5.67b/IDA-1.3.4 for hackers@freebsd.org); Wed, 15 Feb 1995 13:29:16 -0600 Date: Wed, 15 Feb 1995 13:29:16 -0600 From: Terry Lee Message-Id: <199502151929.AA13080@uivlsi.csl.uiuc.edu> To: hackers@FreeBSD.org Subject: /tmp or /var/tmp? Sender: hackers-owner@FreeBSD.org Precedence: bulk Shouldn't gcc be using /tmp instead of /var/tmp? Making /tmp a MFS is a common practice. /var/tmp shouldn't be made MFS because vi.recover is there. By the way, how can I make FreeBSD not use the FPU? Disabling npx0 makes all programs abort with error 8. I'm trying to figure out how FPU intensive maplay is. Would compiling it with -msoft-float do it or would libm still use the FPU? My 486DX40 can't even keep up with one channel in real time. If decoding MPEG layer 2 is FPU intensive, then this could be a real blow for 486 chips. Terry Lee terry@uivlsi.csl.uiuc.edu