KAMINSKI@ikp.tu-darmstadt.de writes:

> While i'm not expierienced with Make filesyntax, in
> src/lib/libm/Makefile there is one thing that seems odd to me:
>
> .elif (${MACHINE_ARCH} == "vax")
> #.PATH: ${.CURDIR}/arch/vax
> #NOIEEE_ARCH= n_infnan.S n_argred.S n_sqrt.S
> #ARCH_SRCS = n_atan2.S n_cabs.S n_cbrt.S n_support.S n_sincos.S n_tan.S
> # XXX - ripped out due to lack of the insn polyd in the Mariah chip,
> # and emulation code isn't written yet.
> WARNS?=2
> .endif
>
> Above the n_atan2.S is ripped out, but...
>
> .if (${MACHINE_ARCH} == "vax") # XXX until POLYD is written.
> .PATH:  ${.CURDIR}/arch/vax
> SRCS:=${SRCS} n_sqrt.S n_argred.S n_infnan.S n_atan2.S n_cabs.S n_cbrt.S \
>         n_support.S
>
> Here it is included. Maybe this is the source of the problems.

The Makefile also includes some magic that causes it to use the .S
source even if you explicitly specify that you want the .c from the
noieee directory -- and the .c doesn't compile properly anyway,
because of multiple warnings generated by the compiler.  I haven't
gotten around to working this out yet, but I'll give it a try.

-tih
-- 
Don't ascribe to stupidity what can be adequately explained by ignorance.