Has anyone looked at xenomai? They show to have support for the beagleboard and I think all OMAP processors. I'm looking at it right now and am thinking about making an rtapi wrapper for it and getting emc to run with it.
One thing with the OMAP-L138 that the hawkboard uses is that it has a universal parallel port with a cycle time of 13 or 26 ns. I'm not sure what all the specs there mean but it looks much faster than 240ns. I don't know how it handles gpio, but it may be faster with that as well. Moses El mié, 22-09-2010 a las 12:24 -0500, Jon Elson escribió: > Adam Hunt wrote: > > Just out of curiosity, what's the state of EMC's RTLinux support? Has > > it totally rotted out? > > > Apparently, no one knows! Most likely there has been some small > incompatibility > that might have developed over the last few years, as it hasn't been > tested by > compiling and running it since maybe 2007 or so. I know when John Kasunich > ran the 8-CPU build farm he was testing at least the ability to compile > against > the RTLinux kernel, but that has been a while. I'm real hazy on that > 2007 date, > too, that is a very rough guess. > > Anyway, as far as I know, Torsten Koschorrek, the ARM maintainer for > RTAI, is still > working on finishing the RTAI port for the Cortex-A8 branch of ARM > CPUs. His > project is not funded, so he is picking at it a little at a time. Last > I heard, he has about > one month full-time of work to complete the port. So, hopefully, this > will eventually > get done. I am somewhat anxious to see what kind of latency numbers he > can provide. > > One big problem with the Beagle Board and other OMAP3530 implementations > is that the > bare GPIO hardware is multiplexed, and although the CPU is quite fast, > now at 720 MHz, > the GPIO pins are only updated every 240 ns. Without that multiplexing, > GPIO traffic > could have been maybe 10 X faster. I don't know if other Cortex-A8 or > other ARM > CPUs have this limitation. There are other options like I2S that can go > to 32 mbit/second, > which is hardly faster than the bare GPIO on a byte-wide port. > > I was just working on a project for a portable EMC system, and we had to > go with an Atom-based > PC motherboard as the Beagle just isn't ready yet. > > Jon > > ------------------------------------------------------------------------------ > Start uncovering the many advantages of virtual appliances > and start using them to simplify application deployment and > accelerate your shift to cloud computing. > http://p.sf.net/sfu/novell-sfdev2dev > _______________________________________________ > Emc-developers mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-developers ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
