> On Mon, Oct 06, 2014 at 09:00:51PM +0200, Philipp Klaus Krause wrote: >> Currently, in sdcc, char is signed char by default. >> >> I would like to change this to unsigned char. >> >> The current --funsigned-char would be replaced by a --fsigned char >> switch to get the non-default behaviour. >> >> char being unsigned char by default has advantages: >> >> People compare chars. For many architectures, unsigned comparisons are >> more efficient than unsigned comparisons. >> >> People cast chars to ints (many standard functions int arguments for >> characters). Casts from unsigned char to int are muh more efficient than >> signed char to int (the latter needs sign extension, the former >> doesn't). >> >> The C standard states that char should be either signed char or unsigned >> char. > > The only concern I have is backward-compability. I mean if someone's > program relies the default sign char, will this change breaks his code?
It might. And that is why we're asking here first. We already thought of this concern and we want to hear from those people that rely on this current default. There may even be often used libraries that depend on it. But if we hear nothing of that kind we will continue and do what we think best. There will always be the fall-back scenario of using the to be introduced --fsigned-char option. Maarten ------------------------------------------------------------------------------ Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk _______________________________________________ Sdcc-user mailing list Sdcc-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sdcc-user
