From owner-freebsd-current Tue Apr 9 01:04:29 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id BAA20058 for current-outgoing; Tue, 9 Apr 1996 01:04:29 -0700 (PDT) Received: from mail.think.com (Mail1.Think.COM [131.239.33.245]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id BAA20050 for ; Tue, 9 Apr 1996 01:04:25 -0700 (PDT) Received: from Early-Bird-1.Think.COM by mail.think.com; Tue, 9 Apr 96 04:04:23 -0400 Received: from compound.think.com ([206.10.99.151]) by Early-Bird.Think.COM; Tue, 9 Apr 96 04:04:19 EDT Received: (from alk@localhost) by compound.think.com (8.7.5/8.6.112) id DAA01446; Tue, 9 Apr 1996 03:04:35 -0500 (CDT) Date: Tue, 9 Apr 1996 03:04:35 -0500 (CDT) Message-Id: <199604090804.DAA01446@compound.think.com> From: Tony Kimball To: current@freebsd.org Subject: msun does not use i387/*.S using make -DHAVE_FPU Sender: owner-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk When I made world using pgcc at ctm 1665, I noticed that the resulting awk executable dumped core during floor() in builtins.c if it was linked dynamically. This led me on a little adventure, in which I discovered (among many other strange and wonderful things) that -DHAVE_FPU does not suffice to compile msun from the i387/*.S sources, as the Makefile intimates that it is. I'm guessing this is a .mk problem. As for the core, it was a pgcc bug, but that is another story altogether. //alk