On Thu, Dec 01, 2011 at 07:42:49PM +0100, Manuel Bouyer wrote: > On Thu, Dec 01, 2011 at 05:50:14PM +0000, David Laight wrote: > > I'd rather the .c file wasn't polluted with #ifdefs. > > Probably better to #define fss_compat_ioctl(...) EINVAL somewhere. > > > > We ought to sort out a way of making the compat code loadable > > - in which case this would need to be a real function call > > (possibly an indirect one). > > I'm not sure bloating the kernel and infrastrucure with functions call, > functions pointer and others to be able to compile out 4 lines of code > is worth it.
Use a weak alias, fss_compat_ioctl() -> einval(). Let the compat code supply a strong symbol for the real implementation. IIRC, our module loader cannot (yet) load a strong alias over a weak alias and restore the weak alias when the module containing the strong alias is unloaded. That would have been nice to have when bpf was modularized. Dave -- David Young dyo...@pobox.com Urbana, IL (217) 721-9981