Kernel Compile Time Decreased with FastWalk on 16-way NUMA


On a kernel compile test on a 16-way NUMA the FastWalk patch:
 
Decreased compile time by 2.5%

Decreased contention on the dcache_lock by 50%