On 06/08/15 01:21, Chris Edwards wrote:
So Jessie's standard libc-2.19 is about half as fast as 2.17 on the same hardware and otherwise same OS. I don't know if this accounts entirely for the poor performance I'm seeing, but it probably doesn't help. Might this be worth taking up with the GNU libc people? My guess is it might be due to slow SSE2 execution on this old AMD64 CPU, whereas the glibc code is probably optimised for much newer hardware.
Just to clarify further, the libc-2.17 library is specifically "GNU C Library (Ubuntu EGLIBC 2.17-93ubuntu4) stable release version 2.17", and Jessie's native libc-2.19 is "GNU C Library (Debian GLIBC 2.19-18) stable release version 2.19". I hadn't realised you could invoke a .so as an executable to find out such info!
And, of course, I don't mean that the entire libc is slower, just the memcpy it chooses on my particular hardware.
I'm not sure what memcpy implementation was actually being used internally for the libc-2.17 test, as it lacked the debugging symbols.
Chris