> > > 64 bit math must be explicitly requested ... see do_div and co. > > __div64_32() from src/lib/div64.c > > Which is what do_div() calls. AFAIK, do_div() is what any code in
I did some small research into this. There are two definitions for do_div() macro, in asm-generic and asm-i386. The asm-<arch> definition provides the optimized method (i386 in assembly) and generic version implements it in software, using repeated subtraction method using div64_32(). I am confused at one point. Even though lib/div64.c : __div64_32() is exported, I could not find it in the nm | grep __div64_32. When tried inserting a module with explicit call to __div64_32(), it refused to load stating unresolved symbol. Any idea? Om -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
