Re: OpenBSD frexpl failure

2007-12-05 Thread Eric Blake
According to Bruno Haible on 12/5/2007 2:59 PM: > Eric Blake wrote: >> it appears that is botched for OpenBSD 4.0, since >> the above makes it clear that long double occupies 80 bits with a 15-bit >> exponent field, ... >> It looks like float.in.h needs to be updated to cater for another platform

Re: gnulib-tool: making parts of .gitignore and .cvsignore files optional

2007-12-05 Thread Bruno Haible
Jim Meyering wrote: > Would you be interested in a gnulib-tool patch to add an option so that > a project that version-controls the sources does not get the primary > sources listed in the generated .gitignore and .cvsignore files? The discussion in the thread "Updating a package's gnulib files in

Re: mingw frexp failure

2007-12-05 Thread Bruno Haible
Eric Blake wrote: > When compiling natively: > > checking whether frexp() can be used without linking with libm... yes > checking whether frexp works... no > > The failure is with +inf, where frexp mistakenly returns NaN instead of > inf. > > But when cross-compiling, mingw was omitted from the lis

Re: OpenBSD frexpl failure

2007-12-05 Thread Bruno Haible
Eric Blake wrote: > it appears that is botched for OpenBSD 4.0, since > the above makes it clear that long double occupies 80 bits with a 15-bit > exponent field, ... > It looks like float.in.h needs to be updated to cater for another platform > with a broken . Can you try it out (change float_h.

base64 build problem on DragonFly 1.8.0

2007-12-05 Thread Martin Lambers
Jeremy C. Reed received the following bug report about building mpop-1.0.12 on DragonFly 1.8.0. This version of mpop uses current gnulib files from 2007-11-27. > Here is the error when attempting to build base64.o: > > In file included from /usr/include/wchar.h:73, > from ./std

no license on some test modules

2007-12-05 Thread Jim Meyering
Hi Bruno, I'm importing some gnulib modules into a lgpl'd project (libvirt) along with their tests. I noticed that several test modules don't specify a license, which makes gnulib-tool --lgpl fail. alloca-opt-tests is one example. I can see a couple ways of fixing it, assuming you are game: -

gnulib-tool: making parts of .gitignore and .cvsignore files optional

2007-12-05 Thread Jim Meyering
Hi Bruno, Would you be interested in a gnulib-tool patch to add an option so that a project that version-controls the sources does not get the primary sources listed in the generated .gitignore and .cvsignore files? Of course, including names of derived files (like .h from .h.in) is still useful.