Messages in this thread | | | Subject | Re: [RFC][PATCH 00/11] perf pmu interface -v2 | From | Peter Zijlstra <> | Date | Wed, 30 Jun 2010 20:11:28 +0200 |
| |
On Wed, 2010-06-30 at 10:19 -0700, Corey Ashford wrote: > However, there still are concerns relative to writing these control > registers from an interrupt handler because of the latency that will be > incurred, however slow we choose to do the event rotation. So at least > for the Wire-Speed processor, we may need a worker thread of some sort > to hand off the work to.
Right, once we have per-pmu contexts we can look at having means to over-ride the means of rotation, such that the pmu can optionally drive it itself instead of through the common tick.
| |