bug-gnulib
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 'round' modules takes 3


From: Bruno Haible
Subject: Re: 'round' modules takes 3
Date: Sun, 21 Oct 2007 19:12:06 +0200
User-agent: KMail/1.5.4

Hi Ben,

Thanks for committing this!

Ben Pfaff wrote:
> I always have a hard time telling what people expect,
> indentation-wise.  I now work with a group of anti-tab bigots and
> another group of pro-tab bigots.  To please both groups, perhaps
> I should write my code so that no line is ever indented past
> column 6 ;-)

It depends on the code to which you contribute. The persons can be
pro-tab bigots and anti-tab bigots, depending on the code they are
working.

> > In m4/roundf.m4 you need an AC_REQUIRE([gl_FUNC_FLOORF]), so that the
> > determination of FLOORF_LIBM comes before it is used in the assignment to
> > ROUNDF_LIBM, not afterwards. Likewise for m4/roundl.m4.
> 
> I am not sure about this.  gl_FUNC_FLOORF checks for floorf and
> provides a substitute if it is not available.  But for roundf I
> was planning to use the system floorf if it was available and, if
> not, use the roundf implementation that does not need floorf.
> Thus, gl_FUNC_FLOORF does more than what roundf needs.

OK, I can provide a macro that just tests whether floorf() needs libm.

Compiling the roundf and roundl modules on HP-UX 11, with CC="cc -Ae -O",
I see these failures:


roundf on HP-UX 11:

        cc -Ae -O -DHAVE_CONFIG_H -I. -I..      -g -c printf-frexp.c
cpp: "./math.h", line 250: error 4038: Bad syntax for #if condition.
*** Error exit code 1
HAVE_DECL_ROUNDF is defined to empty; it should be defined to 0 or 1.

Later:
        cc -Ae -O  -g   -o test-roundf1 test-roundf1.o ../gllib/libgnu.a  
/usr/ccs/bin/ld: Unsatisfied symbols:
   rpl_roundf (first referenced in test-roundf1.o) (code)
   _Isnanf (first referenced in test-roundf1.o) (code)
*** Error exit code 1

There is no roundf.o. I.e. apparently an AC_LIBOBJ([roundf]) is missing or
not executed.

_Isnanf is defined in libm, not in libc. The test already depends on the
isnanf-nolibm module; I think it should include "isnanf.h" and use isnanf().


roundl on HP-UX 11:

        cc -Ae -O  -g   -o test-roundl test-roundl.o ../gllib/libgnu.a  
/usr/ccs/bin/ld: Unsatisfied symbols:
   rpl_roundl (first referenced in test-roundl.o) (code)

There is no roundl.o. I.e. apparently an AC_LIBOBJ([roundl]) is missing or
not executed.


You can get access to HP-UX machines via testdrive.hp.com.

Bruno





reply via email to

[Prev in Thread] Current Thread [Next in Thread]