On: Improvements to static analysis in GCC 14

[teo_zero]

I'd rather put the final nul at d+l-1 than at d+l, so that l can be the size of d, not "one more than the size of d":

  strncpyz(buf,src,sizeof buf);