On Sat, Jan 24, 2009 at 12:53, Emanuele Giaquinta <[email protected]> wrote: > On Thu, Jan 22, 2009 at 04:02:38PM -0500, Russell Harmon wrote: > >> +if test x$support_mlock = xyes; then >> + dnl# glibc versions < 2.5.20 don't work with mlockall >> + AC_CACHE_CHECK([for glibc >= 2.5], rxvt_mlock, >> + [AC_RUN_IFELSE([ >> + AC_LANG_PROGRAM([ >> +#include <stdlib.h> >> + ],[ >> +#if ! ( ( defined(__GLIBC__) && defined(__GLIBC_MINOR__) ) && ( >> __GLIBC__ > 2 || ( __GLIBC__ == 2 && __GLIBC_MINOR__ >= 5 ) ) ) >> + return 1; >> +#endif >> + ])], >> + [],[support_mlock=no] >> + )]) >> +fi > > Doesn't this check prevent mlockall to be enabled on non glibc systems? I don't believe so... on a non-glibc system, that statement should resolve to:
#if ! ( ( false && false ) && ( undef > 2 || (undef == 2 && undef >= 5 ) ) ) which becomes #if ! ( ( false && false ) ) which becomes #if ! false which becomes #if true which compiles it in. _______________________________________________ rxvt-unicode mailing list [email protected] http://lists.schmorp.de/cgi-bin/mailman/listinfo/rxvt-unicode
