I am wondering why we don't default on --enable-plugin in gcc 4.6 (and perhaps 4.5.1) for those hosts that are known to have working testsuite results of plugin.exp? The additional overhead for building the plugin support is close to nil and the user has to explicitly invoke the loading of a compiler plugin (so that it shouldn't destabilize in the default usage of the compiler). Jack
- --enable-plugin as default? Jack Howarth
- Re: --enable-plugin as default? Richard Guenther
- Re: --enable-plugin as default? Duncan Sands
- Re: --enable-plugin as default? Basile Starynkevitch
- Re: --enable-plugin as default? Diego Novillo
- Re: --enable-plugin as default? Jack Howarth