Messages in this thread | | | Date | Fri, 31 Jan 2014 13:28:06 -0500 | From | Waiman Long <> | Subject | Re: [PATCH v3 1/2] qspinlock: Introducing a 4-byte queue spinlock implementation |
| |
On 01/30/2014 02:35 PM, Peter Zijlstra wrote: > On Tue, Jan 28, 2014 at 01:19:10PM -0500, Waiman Long wrote: >> This patch introduces a new queue spinlock implementation that can >> serve as an alternative to the default ticket spinlock. Compared with >> the ticket spinlock, this queue spinlock should be almost as fair as >> the ticket spinlock. It has about the same speed in single-thread and >> it can be much faster in high contention situations. Only in light to >> moderate contention where the average queue depth is around 1-2 will >> this queue spinlock be potentially a bit slower due to the higher >> slowpath overhead. > But light to moderate contention should be the most common case. If your > spinlock is very heavily contended, it should be broken up. > > I would really like more performance numbers for this on single and dual > socket machines and 'normal' workloads (ha! as if those exist), ie. > kbuild is all very important to us :-)
Yes, I am planning to get more performance data on low end machines. The v3 patch was sent out to spur discussion on this topic while I am trying to get more data.
-Longman
| |