> > > math/cln                   6       adam@NetBSD.org
> >
> > Please try patch-ac.
>
> (was attached)
>
> Well, that exposed another problem, apparently a missing newline
> at the end of an umul:
>
>         umul %g2,%g3,%g1sub %g5,%g1,%g1
>
> and there were a few occurrances of that.  New patch-ac candidate
> diff attached below.
>
> There's also a bunch of warnings related to redefinition of
> cl_word_alignment:
>
> In file included from ../include/cln/string.h:6,
>                  from ../include/cln/symbol.h:6,
>                  from base/symbol/cl_symbol.cc:9:
> ../include/cln/object.h:29:1: warning: "cl_word_alignment" redefined
> ../include/cln/object.h:26:1: warning: this is the location of the pr=
evious definition
>
> The reason is that both __sparc__ and __sparc64__ is defined.
>
> It's the __sparc64__ define which takes effect, so this appears
> to be benign.  New suggested patch-ad attached below.
>
> With that diff in place it continues the build for a long time (I got=

> tired of waiting for it to complete before sending this message).

Well, it failed in the end after all:

c++ -O2 exam.o exam_I.o exam_RA.o exam_SF.o exam_FF.o exam_DF.o exam_LF=
.o exam_I
_gcd.o exam_I_sqrtp.o -Wl,-R/usr/pkg/lib -o .libs/exam  ../src/.libs/li=
bcln.so -
L/usr/pkgsrc/math/cln/work/.buildlink/lib -lm -Wl,--rpath -Wl,/usr/pkg/=
lib
../src/.libs/libcln.so: undefined reference to `gf2_mul32'
../src/.libs/libcln.so: undefined reference to `divu_3232_3232_'
gmake[1]: *** [exam] Error 1
gmake[1]: Leaving directory `/usr/pkgsrc/math/cln/work/cln-1.1.9/tests'=

gmake: *** [all] Error 2
*** Error code 2

Not sure how what the deal is with that.

- H=E5vard