Messages in this thread | | | Subject | Re: 2.6.14-rt1 - xruns in a certain circumstance | From | Lee Revell <> | Date | Mon, 31 Oct 2005 11:18:47 -0500 |
| |
On Mon, 2005-10-31 at 15:22 +0100, Ingo Molnar wrote: > this could be some sort of hardware latency, as Lee suspects. > Videocards are known to be pretty agressively holding the system bus, > for the last few percentiles of Quake performance ... Also, mainboard > chipsets are sometimes not that good at enforcing fairness between DMA > agents - possibly starving the CPU itself for lengthly amounts of > time. We have seen such incidents before, and latency tracing ought to > be able to show this with reasonable certainty.
Ingo,
IIRC when I had this problem with the via X driver the latency traces actually didn't show anything useful. The disk controller induced DMA starvation problems did show up in the tracer.
Lee
- 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/
| |