> On Sat, Jul 09, 2011 at 09:23:30PM +0200, Peter Mazinger wrote: > > > That would mean that CONFIG_LIBRESOLV_STUB has influence to whether > > > the resolver symbols end up in the libc? Or would you build and link > > > them unconditionally? > > > > libresolv.* is built only if the above option is enabled, but does > > not provide any usable functions, allowing apps that hardcode > > -lresolv to find something > > The only libresolv.* should be libresolv.a. If you also make a .so, it > will get linked, loaded, and waste 4k or more in every process that > does so, all for no usefulness whatsoever.
good idea, there is really no need for the lib*.so stubs (nsl and resolv) Peter > The approach I took in musl is to make completely empty (8 byte) .a > files for all the legacy library names (libm, libpthread, etc.) just > to satisfy programs that use -lm etc. and only a single .so file, > libc.so. > > Rich > _______________________________________________ > uClibc mailing list > [email protected] > http://lists.busybox.net/mailman/listinfo/uclibc -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de _______________________________________________ uClibc mailing list [email protected] http://lists.busybox.net/mailman/listinfo/uclibc
