2002-02-27 21:31:17

by Niels Christiansen

[permalink] [raw]
Subject: Re: [Lse-tech] lockmeter results comparing 2.4.17, 2.5.3, and 2.5.5

Andrew Morton wrote:
> I have a concern about the lockmeter results. Lockmeter appears
> to be measuring lock frequency and hold times and contention. But
> is it measuring the cost of the cacheline transfers?

No.

> I expect that with delayed allocation and radix-tree pagecache, one
> of the major remaining bottlenecks will be ownership of the superblock
> semaphore's cacheline. Is this measurable?
When you ask if this is measurable, exactly what do you mean? The cost
of cacheline transfers? Expressed in which unit of measure? Bottlenecks?

If the data you are after is available, a tool can surely be made
to capture and present it...

-nc-