From: Ian Lance Taylor <i...@google.com> Date: Tue, 27 Jan 2009 13:16:31 -0800
> A co-worker of mine at Google did some experiments along those lines > using gold. He was not able to demonstrate any performance > improvements, unfortunately (x86_64 target, proprietary test cases). > He was hacking linker scripts. This sort of optimization helps more on cpus with f.e. software managed TLBs. The higher the TLB miss cost, the more the benefit. This is why Sun's tools have supported this optimization for a long time, since 64-bit sparc has been predominantly using software managed TLB refill. I think it's the "-xlinkopt" options that do this with the SunPRO compiler tools.
