https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118940
--- Comment #8 from Miao Wang <shankerwangmiao at gmail dot com> --- (In reply to Andrew Pinski from comment #5) > Note I suspect using C with some (generic) builtins might be faster than > what the inline-asm could provide these days than the inline-asm that was > used. Plus I doubt the speed of big-int would ever be the bottle neck speed > of ipxe because the network will always be the bottle neck. So I wonder why > they optimized this code in the first place 13 years ago. I assume the reason why the author chose to use the inline-asm is that the author want to use fewer instructions so that the generated boot image would be smaller.
