Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 07 Nov 2013 13:30:21 -0700
From:      Ian Lepore <ian@FreeBSD.org>
To:        Steve Kargl <sgk@troutmask.apl.washington.edu>
Cc:        svn-src-head@FreeBSD.org, svn-src-all@FreeBSD.org, src-committers@FreeBSD.org, Steve Kargl <kargl@FreeBSD.org>
Subject:   Re: svn commit: r257770 - head/lib/msun/src
Message-ID:  <1383856221.31172.226.camel@revolution.hippie.lan>
In-Reply-To: <20131107201946.GA58797@troutmask.apl.washington.edu>
References:  <201311062344.rA6Niqkf001246@svn.freebsd.org> <1383854281.31172.214.camel@revolution.hippie.lan> <20131107201946.GA58797@troutmask.apl.washington.edu>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 2013-11-07 at 12:19 -0800, Steve Kargl wrote:
> On Thu, Nov 07, 2013 at 12:58:01PM -0700, Ian Lepore wrote:
> > On Wed, 2013-11-06 at 23:44 +0000, Steve Kargl wrote:
> > > Author: kargl
> > > Date: Wed Nov  6 23:44:52 2013
> > > New Revision: 257770
> > > URL: http://svnweb.freebsd.org/changeset/base/257770
> > > 
> > > Log:
> > >   * Use "math.h" instead of <math.h>.
> > >   * Use bit twiddling.  This requires inclusion of math_private.h
> > >     and inclusion of float.h in s_roundl.c.  Raise invalid exception.
> > >   * Use literal integer constants where possible.  Let the compiler
> > >     do the appropriate conversion.
> > >   * In s_roundf.c, use an F suffix on float constants instead of
> > >     promoting float to double and then converting the result back
> > >     to float. In s_roundl.c, use an L suffix.
> > >   * In s_roundl.c, use the ENTERI and RETURNI macros.  This requires
> > >     the inclusion of fpmath.h and on __i386__ class hardware ieeefp.h.
> > >   
> > >   Reviewed by:	bde
> > > 
> > > Modified:
> > >   head/lib/msun/src/s_round.c
> > >   head/lib/msun/src/s_roundf.c
> > >   head/lib/msun/src/s_roundl.c
> > > [snip]
> > 
> > FYI, this seems to be causing build failures on non-x86 platforms.  For
> > example...
> > 
> > ===> lib/msun (obj,depend,all,install)
> > cc  -fpic -DPIC  -O -pipe  -I/local/build/staging/freebsd/bb/src/lib/msun/src -I/local/build/staging/freebsd/bb/src/lib/msun/../libc/include  -I/local/build/staging/freebsd/bb/src/lib/msun/../libc/arm -std=gnu99 -Qunused-arguments -Wsystem-headers -Werror -Wno-pointer-sign -Wno-empty-body -Wno-string-plus-int -Wno-tautological-compare -Wno-unused-value -Wno-parentheses-equality -Wno-unused-function -Wno-conversion -Wno-switch -Wno-switch-enum -Wno-knr-promoted-parameter -Wno-parentheses -c /local/build/staging/freebsd/bb/src/lib/msun/src/s_roundl.c -o s_roundl.So
> > /local/build/staging/freebsd/bb/src/lib/msun/src/s_roundl.c:45:2: error: no member named 'xbits' in 'union IEEEl2bits'; did you mean 'bits'?
> >         GET_LDBL_EXPSIGN(hx, x);
> >         ^
> > /local/build/staging/freebsd/bb/src/lib/msun/src/math_private.h:224:14: note: expanded from macro 'GET_LDBL_EXPSIGN'
> >   (i) = ge_u.xbits.expsign;                                     \
> >              ^
> > /local/build/staging/freebsd/bb/src/lib/msun/../libc/arm/_fpmath.h:54:4: note: 'bits' declared here
> >         } bits;
> 
> I zero about floating-point on arm.  Does long double have
> 64, 80, or 113-bit significands?
> 

I don't know either (for some reason I'm thinking long double is synonym
for double on arm, but I can't remember why I think that).

This compile failure isn't just arm, btw, it's also powerpc and mips.
sparc and x86 seem to be okay.

-- Ian





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1383856221.31172.226.camel>