Robert Schwebel wrote: > On Wed, Jul 18, 2007 at 12:08:48PM +0200, M. Koehrer wrote: >> a while ago I posted already a mail concerning this question. >> http://www.mail-archive.com/rtnet-users@lists.sourceforge.net/msg01687.html >> The question was if rtnet will support Ingo Molnar's kernel_preempt >> patch. As there is already a native rtdm the missing pieces are to >> extend the native rtdm (or to create another variant of rtdm for the >> kernel preempt_patch). Also, the build system of rtnet has to be >> adapted to support this. Are there any news on this issue? > > We are currently doing some research in order to find out what is > possible with the vanilla Linux network stack and determinism. Although > it is a little bit too early for solid results, the first tests we did > today indicate that it is possible to get 200 ... 300 us round trip time > between two boxes, both running 2.6.22.1-rt9 (although no real load on > the boxes yet, so no worst case). The histograms show that most round > trip times are around 100 us. The code is standard userspace POSIX code, > using raw and packet sockets. > > Before we'll provide more measurement results we have to code proper > load scenarios.
For this, I would suggest to start with analysing the TX and RX code paths in normal Linux or -rt. User-triggered latency tracing or LTTng might help. Then the next step would be a lock-dependency analysis. I would be curious to hear about your findings. BTW, do you already use some of the emergency pool reservation patches for this? I read interesting ideas recently on netdev, though no code for those is written yet. In any case, memory pressure should be part of your load scenarios. Jan
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________ RTnet-users mailing list RTnet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rtnet-users
