Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!spool.mu.edu!agate!ucbvax!hplabs!unix.sri.com!headwall.Stanford.EDU!leland.Stanford.EDU!yergeau From: yergeau@leland.Stanford.EDU (Dan Yergeau) Newsgroups: comp.os.386bsd.apps Subject: Re: FPU math library? Message-ID: <2svmiq$qn2@nntp2.Stanford.EDU> Date: 6 Jun 94 17:32:10 GMT References: <2somn9$6n7@news.ysu.edu> Organization: DSG, Stanford University, CA 94305, USA Lines: 127 NNTP-Posting-Host: elaine24.stanford.edu In article <2somn9$6n7@news.ysu.edu>, ap713@yfn.ysu.edu (Christopher L. Mikkelson) writes: |> |> Is there a good math library (preferable compatible with the standard) |> for FreeBSD. I had tried with some success to use libfpu (mathlib2.1-386bsd) |> that I found laying around somewhere, but programs running it had the |> nasty habit of dumping core practically everywhere. If you want to stick with mathlib2.1, apply these patches. The pow() bug clobbered ebx, esi, and edi. *** acosf.s~ Fri May 29 14:34:47 1992 --- acosf.s Mon May 31 14:54:25 1993 *************** *** 42,48 **** ret .Lzero: ! flds .Lhalfpi .Ldone: leave --- 42,48 ---- ret .Lzero: ! fldl .Lhalfpi .Ldone: leave *** atan2f.s~ Fri May 29 14:34:44 1992 --- atan2f.s Mon May 31 14:55:04 1993 *************** *** 55,61 **** fdivp fld1 fpatan ! flds .Lmpi fsubrp leave --- 55,61 ---- fdivp fld1 fpatan ! fldl .Lmpi fsubrp leave *************** *** 65,71 **** fdivp fld1 fpatan ! flds .Lpi fsubrp leave --- 65,71 ---- fdivp fld1 fpatan ! fldl .Lpi fsubrp leave *************** *** 80,86 **** jz .Lzero jc .Lneg ! flds .Lhalfpi leave ret --- 80,86 ---- jz .Lzero jc .Lneg ! fldl .Lhalfpi leave ret *************** *** 92,98 **** ret .Lneg: ! flds .Lmhalfpi leave ret --- 92,98 ---- ret .Lneg: ! fldl .Lmhalfpi leave ret *** pow.s~ Fri May 29 14:34:51 1992 --- pow.s Mon May 31 14:59:28 1993 *************** *** 155,161 **** fchs /*/NO_APP*/ .L1: ! leal -28(%ebp),%esp popl %ebx popl %esi popl %edi --- 155,161 ---- fchs /*/NO_APP*/ .L1: ! leal -40(%ebp),%esp popl %ebx popl %esi popl %edi -- Dan Yergeau You are in a twisty little passage yergeau@gloworm.Stanford.EDU of standards, all conflicting. #include <std.disclaimer>