On 07/06/2018 10:41 AM, Federico Sbalchiero wrote: > Il 05/07/2018 11:14, Philippe Gerum ha scritto: >> On 07/04/2018 07:06 PM, Federico Sbalchiero wrote: >>> Hi, >>> first I want to say thanks to everyone involved in Xenomai for their >>> job. >>> >>> I'm testing Xenomai 3.0.7 and ipipe-arm/4.14 on Freescale/NXP i.MX6q >>> sabresd board using Yocto. System boots fine and is stable, but latency >>> under load (xeno-test) is higher than in my reference system (Xenomai >>> 2.6.5 on Freescale kernel 3.10.17 + ipipe 3.10.18). >>> This is after disabling power management, frequency scaling, CMA, >>> graphics, tracing, debug. >>> >>> I have found that a simple non-realtime user space process writing a >>> buffer in memory (memwrite) is able to trigger such high latencies. >>> Latency worsen a lot running a copy of the process on each core. >>> There is a correlation between buffer size and cache size suggesting >>> an L2 cache issue, like the L2 write allocate discussed in the mailing >>> list, but I can confirm L2 WA is disabled (see log). >>> >>> I'm looking for comments or suggestions. >>> >> A basic dd if=/dev/zero of=/dev/null loop in the background is enough to >> raise the latency actually. Could you try the Xenomai 3 + 3.18 combo on >> your hw and let us know whether you see the same regression? >> >> TIA, >> > > kernel 3.18.20-ipipe + xenomai 3.0.7 > > latency under load (four memwrite instances) > RTT| 00:00:01 (periodic user-mode task, 1000 us period, priority 99) > RTH|----lat min|----lat avg|----lat max|-overrun|---msw|---lat > best|--lat worst > RTD| 24.985| 41.374| 76.351| 0| 0| 24.985| 76.351 > RTD| 26.889| 41.203| 68.070| 0| 0| 24.985| 76.351 > RTD| 22.828| 41.376| 67.681| 0| 0| 22.828| 76.351 > RTD| 20.969| 41.043| 74.143| 0| 0| 20.969| 76.351 > RTD| 27.027| 41.441| 68.037| 0| 0| 20.969| 76.351 > RTD| 24.413| 41.585| 81.062| 0| 0| 20.969| 81.062 > RTD| 27.234| 41.168| 76.516| 0| 0| 20.969| 81.062 > RTD| 23.779| 41.141| 70.466| 0| 0| 20.969| 81.062 > RTD| 24.824| 41.273| 75.322| 0| 0| 20.969| 81.062 > RTD| 25.627| 41.195| 71.157| 0| 0| 20.969| 81.062 > RTD| 28.874| 41.089| 66.579| 0| 0| 20.969| 81.062 > RTD| 26.672| 41.638| 75.995| 0| 0| 20.969| 81.062 > RTD| 25.139| 41.040| 69.543| 0| 0| 20.969| 81.062 > RTD| 26.215| 41.099| 66.336| 0| 0| 20.969| 81.062 > RTD| 24.192| 41.117| 76.828| 0| 0| 20.969| 81.062 > RTD| 27.310| 41.942| 79.888| 0| 0| 20.969| 81.062 > RTD| 24.348| 40.955| 66.484| 0| 0| 20.969| 81.062 > RTD| 26.679| 41.260| 80.242| 0| 0| 20.969| 81.062 > RTD| 26.820| 41.251| 74.986| 0| 0| 20.969| 81.062 > RTD| 27.635| 41.301| 73.961| 0| 0| 20.969| 81.062 > RTD| 26.877| 41.305| 72.789| 0| 0| 20.969| 81.062 > >
Ok, if all goes well, we should soon be able to see the worst-case latency drop to ~65 us under high mm stress on i.MX6q over 4.14, the fewer the cores, the better the results with the i.MX6 series. -- Philippe. _______________________________________________ Xenomai mailing list [email protected] https://xenomai.org/mailman/listinfo/xenomai
