Gary V. Vaughan wrote: > Hi, > >> On Dec 9, 2014, at 6:23 AM, KO Myung-Hun <[email protected]> wrote: >> >> Ping ? >> >> KO Myung-Hun wrote: >>> UNIXROOT on OS/2 is used to specify a drive where unix FHS is >>> installed. If M4 is not executed on a drive where it exists without >>> UNIXROOT support, modules cannot be located. Because '/' is a root of >>> a current drive. >>> >>> * m4/path.c (UNIXROOT): Define it on OS/2 or empty string on others. >>> (m4__include_init): Prepend UNIXROOT to PKGLIBDIR. >>> --- >>> m4/path.c | 8 +++++++- >>> 1 file changed, 7 insertions(+), 1 deletion(-) >>> >>> diff --git a/m4/path.c b/m4/path.c >>> index bc8b376..10cc098 100644 >>> --- a/m4/path.c >>> +++ b/m4/path.c >>> @@ -378,6 +378,12 @@ m4_load_filename (m4 *context, const m4_call_info >>> *caller, >>> } >>> >>> >>> +#if OS2 >>> +# define UNIXROOT "/@unixroot" >>> +#else >>> +# define UNIXROOT "" >>> +#endif >>> + >>> void >>> m4__include_init (m4 *context) >>> { >>> @@ -393,7 +399,7 @@ m4__include_init (m4 *context) >>> search_path_add (info, "", false); >>> >>> /* Non-core modules installation directory. */ >>> - search_path_add (info, PKGLIBDIR, false); >>> + search_path_add (info, UNIXROOT PKGLIBDIR, false); >>> } >>> >>> #ifdef DEBUG_INCL > > This seems like papering over the cracks to me. > > Wouldn't it be better to fix this for every project by patching the gnulib > configmake module to set the appropriate prefix for all the configured > directories? >
You're right. > If we fix just PKGLIBDIR, for just M4, you'll have a zillion tiny patches > to discover, write and shepherp through their individual projects otherwise. > I agree. I'll try another ways. Thanks. -- KO Myung-Hun Using Mozilla SeaMonkey 2.7.2 Under OS/2 Warp 4 for Korean with FixPak #15 In VirtualBox v4.1.32 on Intel Core i7-3615QM 2.30GHz with 8GB RAM Korean OS/2 User Community : http://www.ecomstation.co.kr _______________________________________________ M4-patches mailing list [email protected] https://lists.gnu.org/mailman/listinfo/m4-patches
