Re: [RFC PATCH] mm: Promote slow memory in advance to improve performance

[Baolin Wang]

On 2021/11/23 3:34, Yang Shi wrote:
> On Mon, Nov 22, 2021 at 2:22 AM Baolin Wang
> <baolin.wang@xxxxxxxxxxxxxxxxx> wrote:
> > Some workloads access a set of data entities will follow the data locality,
> > also known as locality of reference, which means the probability of accessing
> > some data soon after some nearby data has been accessed.
> >
> > On some systems with different memory types, which will rely on the numa
> > balancing to promote slow hot memory to fast memory to improve performance.
> > So we can promote several sequential pages on slow memory at one time
> > according to the data locality for some workloads to improve the performance.
>
> Fault around for NUMA fault definitely could reduce the overhead for
> NUMA balancing by having fewer faults. I think this could be extended
> to regular NUMA balancing too. But I'm not sure whether false
> positives are worth concerning or not.

OK. Like Huang Ying said, maybe we can add some algorithm to adjust the 
window of proactive numa faults dynamically. Thanks for your input.

> I recall Mel proposed fault around too (not in patch, but shared some
> ideas). Added Mel in this thread.