Jakub Jelinek <ja...@redhat.com> > > Those are the generic definitions, all targets that care about performance > obviously should replace them with assembly code.
No, that's exactly my point, it is not true that it is always best to write assembly code. For example there is absolutely no benefit in writing an optimized mempcpy. At best it is as fast as memcpy, and therefore expanding mempcpy into memcpy (p, q, n) + n would have the same performance. In actual use memcpy will then be slightly faster due to lower I-cache pressure. Wilco
