Alex Bennee wrote: > Christian Biere said: > > Alex Bennee wrote: > >> So I did this: > >> - G_LIST_FOREACH(list_nodes, (GFunc) update_row, &now); > >> + g_list_foreach(list_nodes, (GFunc) update_row, &now);
> > Yes, this should work but I want to keep using a macro instead and > > fixed the macro definition therefore. Could you please confirm that the > > attached patch fixes the problem as well? > I'll give it a spin when I get home and let you know. I haven't yet tried with GCC 3.4 but maybe the patch doesn't fix the problem. If it doesn't fix it, try removing "inline" from the prototype of "update_row". This could be related to a change in the way inline is handled by GCC. Maybe it doesn't generate any code at all for it because it's nowhere directly used so that the address of the function is undefined. OTOH, the patch should take of this as well because the function is now called "by name" instead of by address. At first I thought dereferencing "func" was the problem. > Sorry, I mean it missed it when compiling from scratch as 147 files > generates a lot of output and the final link had scrolled the warnings of > the top of my screen. > I wonder if there is an easy way of summerising the number of warnings in > the build at the end? Whenever I compile something I use "make >& make.log". -- Christian
pgpAtICvrhAKc.pgp
Description: PGP signature
