On Sun, 30 Dec 2012, Jonas Walld?n @ Pike developers forum wrote:
Too bad with the name clash, but a Pike-prefixed name would also be quite annoying due to its length.
Yes I agree. I will add an undef before the libvpx includes, which should fix that.
Do you mean that my latest changes introduced the gcc overflow issues or only that you noticed them as of now? I'm also concerned about the overflow detection method using costly integer division on every Pike variable adjustment (machine-code excepted) so if we can optimize that it would be great. Here's a link that I found that could be useful: https://www.securecoding.cert.org/confluence/display/seccode/INT32-C.+Ensure+that+operations+on+signed+integers+do+not+result+in+overflow?showComments=false
As far as I can see its the latest change, which turned the macro into fuctions. I think its overall quite mysterious under which circumstances gcc (or clang) optimize that check away. In particular because compilers change and something that works now might break tomorrow. That said, I think it makes sense to use something which does _not_ involve any undefined behavior. Thanks for the link. I will have a look at that. I was thinking about using __int128 for those compilers that support it and "manual" multiplication otherwise (see http://www.fefe.de/intof.html). arne
