Messages in this thread | | | Date | Sat, 8 Jan 2005 20:07:34 -0800 (PST) | From | Linus Torvalds <> | Subject | Re: Make pipe data structure be a circular list of pages, rather than |
| |
On Sat, 8 Jan 2005, Linus Torvalds wrote: > > Short and sweet: the latency changes are in the noise for SMP, but can be > seen on UP. I'll look at it a bit more: since I had to add the coalescing > code anyway, I might also decide to re-use a buffer page rather than free > it immediately, since that may help latency for small writes.
Side note: the SMP numbers seem to fluctuate wildly making any benchmarking a bit suspect. They probably depend very much on just how things get scheduled, and on inter-CPU cache behaviour.
In contrast, the UP numbers are a lot more reliable, so I'll try to make sure that everything looks good on UP - which definitely means that I'll work to make sure that latency looks good too.
Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |