Hi Peter, * Peter Ekberg wrote on Thu, Sep 22, 2005 at 12:43:46PM CEST: > * Ralf Wildenhues wrote on Thursday, September 22, 2005 10:06 CEST: > > * Peter Ekberg wrote on Thu, Sep 22, 2005 at 10:00:32AM CEST: > > > * Ralf Wildenhues wrote on Wednesday, September 21, 2005 22:37 CEST: > > > > * Peter Ekberg wrote on Mon, Sep 19, 2005 at 09:05:04PM CEST: > > > > > > > Thinking about it further, include_expsyms is perhaps not > > > > > buggy with skipped exports, as then the needed symbols perhaps > > > > > get exported anyway with the export_symbols_cmds functionality. > > > > > A check wouldn't hurt though... > > > > > But looking closer, include_expsyms is a disaster area. Since it > > > isn't used, my vote is to zap it completely. > > > > > > The problem is that it potentially clobbers user provided files, > > > so that building repeatedly will eat up all your disk space > > > eventually (but that isn't the worst problem with clobbering > > > user provided files of course). > > > > Why not make a copy of the user-provided file and append to that? > > C'mon, that isn't rocket science. Unless there are other reasons > > against include_expsyms, that is. > > Well, it seemed to complicate matters for little reason, that's all.
That's just because you try to avoid any extra copy, which requires quite some logic, as seen in your proposed patch. > Ok, here's a patch. I renamed the orig_export_symbols variable > as it's now not always the original export_symbols value... I tried hard to find an error in the patch, but failed. Please install, I think it's ok. Cheers, Ralf > * libltdl/config/ltmain.m4sh (func_mode_link): Don't modify > user provided symbols file when $include_expsyms is active. > Rename orig_export_symbols variable to filt_export_symbols.
