Hi,
> > +# needed to make gcc accept wide unicode chars without warning
> > +curses.o-cflags := -std=gnu99
> > +
> > ifeq ($(CONFIG_SDLABI),1.2)
> > sdl.mo-objs := sdl.o sdl_zoom.o
> > endif
> > --
> > 1.8.3.1
>
> I'm not sure about this.
>
> Do we really want gnu99 and not c99? Should we just enable
> std=c99 for all source files, given that we already assume
> C99 anyway?
-std=c99 doesn't fly, throws errors on (gcc-specific) asm.
So we have to use -std=gnu99.
Tried to do that tree-wide by patching configure but got build errors
then (somewhere in iscsi code).
> It would also be helpful if the commit message quoted the
> compiler warning that you get otherwise, so it's easier
> to see why we're doing this.
With the wide char curses patches by samuel applied I get errors like
this one:
/home/kraxel/projects/qemu/ui/curses.c:627:18: error: universal
character names are only valid in C++ and C99 [-Werror]
case L'\u23bd':
^
I'm open to better ideas to handle this.
Failing that I can cut+paste this reply into the commit message to
clarify things.
cheers,
Gerd
