Re: [PATCH 2/4] util: Add l_strlcpy

> > how does a direct for/while loop implementation compare speed wise to > calling strlen & memcpy? It's about twice as fast to use strlen/memcpy. The direct implementation I used for comparison:         const char *orig_src = src;         if (len) {                 while (*src != '\0' && --len) {                         *dst++ = *src++;                 }                 *dst = '\0';         }         return (src - orig_src) + strlen(src); I modified the unit test to run a bunch of iterations with some more string sizes and to take performance samples during the strlcpy test. gperftools showed that the direct code spent 57.39% of execution time inside l_strlcpy and the strlen/memcpy technique spent 16.06% of execution time inside l_strlcpy. Turning off the profiling stuff and just using 'time', direct took 53.973 seconds and strlen/memcpy took 25.176 seconds of user time.