Re: [patch] mm, vmacache: hash addresses based on pmd

2018-07-11 Thread David Rientjes
On Wed, 11 Jul 2018, Andrew Morton wrote: > > > Did you consider LRU-sorting the array instead? > > > > > > > It adds 40 bytes to struct task_struct, > > What does? LRU sort? It's a 4-entry array, just do it in place, like > bh_lru_install(). Confused. > I was imagining an optimized sort ra

Re: [patch] mm, vmacache: hash addresses based on pmd

2018-07-11 Thread Andrew Morton
On Mon, 9 Jul 2018 18:37:37 -0700 (PDT) David Rientjes wrote: > > Did you consider LRU-sorting the array instead? > > > > It adds 40 bytes to struct task_struct, What does? LRU sort? It's a 4-entry array, just do it in place, like bh_lru_install(). Confused. > but I'm not sure the least >

Re: [patch] mm, vmacache: hash addresses based on pmd

2018-07-09 Thread David Rientjes
On Mon, 9 Jul 2018, Andrew Morton wrote: > > When perf profiling a wide variety of different workloads, it was found > > that vmacache_find() had higher than expected cost: up to 0.08% of cpu > > utilization in some cases. This was found to rival other core VM > > functions such as alloc_pages_vm

Re: [patch] mm, vmacache: hash addresses based on pmd

2018-07-09 Thread Andrew Morton
On Mon, 9 Jul 2018 17:50:03 -0700 (PDT) David Rientjes wrote: > When perf profiling a wide variety of different workloads, it was found > that vmacache_find() had higher than expected cost: up to 0.08% of cpu > utilization in some cases. This was found to rival other core VM > functions such as